Te Linde's Operative Gynecology 9th

Te Linde's Operative Gynecology Editor: Rock, John A.; Jones, Howard W. Publisher: Lippincott Williams & Wilkins Edition: 9th Edition ISBN: 0-7817-2859-2 [email protected]

Section I: General Topics Affecting Gynecologic Surgery Practice Chapter 1: A Brief History of Operative Gynecology Chapter 2:The Ethics of Pelvic Surgery Chapter 3: Psychological Aspects of Pelvic Surgery Chapter 4: Professional Liability and Risk Management for the Gynecologic Surgeon Section II: Principles Of Anatomy And Perioperative Considerations Chapter 5: Surgical Anatomy of the Female Pelvis Chapter 6: Preoperative Care Chapter 7: Postanesthesia and Postoperative Care Chapter 8: Water, Electrolyte, and Acid–Base Metabolism Chapter 9: Postoperative Infections: Prevention and Management Chapter 10: Shock in the Gynecologic Patient Chapter 11: Wound Healing, Suture Material, and Surgical Instrumentation Section III: Principles Of Gynecologic Surgical Techniques And Management Of Endoscopy Chapter 12: Incisions for Gynecologic Surgery Chapter 13: Principles of Electrosurgery as Applied to Gynecology Chapter 14: Ultrasonic Surgery Chapter 15: Application of Laser in Gynecology Chapter 16: Diagnostic and Operative Laparoscopy Chapter 17: Operative Hysteroscopy Chapter 18: Control of Pelvic Hemorrhage Section IV: Surgery for Fertility and Benign Gynecologic Conditions Chapter 19: Evolving Aspects of Reparative Surgery Chapter 20: Normal and Abnormal Uterine Bleeding Chapter 21: Management of Abortion Chapter 22: Ectopic Pregnancy Chapter 23: Tubal Sterilization Chapter 24: Reconstructive Tubal Surgery Chapter 25: Endometriosis Chapter 26: Surgery for Benign Disease of the Ovary Chapter 27: Persistent or Chronic Pelvic Pain Chapter 28: Pelvic Inflammatory Disease Chapter 29: Surgery for Anomalies of the Müllerian Ducts Chapter 30: Leiomyomata Uteri and Myomectomy Chapter 31: Hysterectomy Chapter 32 - Gynecologic Surgery for Obstetric Patients A. Obstetrics Problems B. Ovarian Tumers Complicating Pregnancy C. Fetal Surgery Chapter 33: Surgical Conditions of the Vulva Chapter 34: Surgical Conditions of the Vagina and Urethra Section V: Surgery for Correction of Defects in Pelvic Support and Pelvic Fistula Chapter 35 - Surgical Correction of Defects in Pelvic Support A. Pelvic Organ Prolapse B. Correction of Anterior Compartment Defects C. Posterior Compartment Defects D. Vaginal Hysterectomy with Repair of Enterocele, Cystocele, and Rectocele E. Vaginal Vault Prolapse F. The Nonsurgical Management of Pelvic Organ Prolapse: The Use of Vaginal Pessaries Chapter 36: Urinary Stress Incontinence Chapter 37: Operative Injuries to the Ureter Chapter 38: Vesicovaginal and Urethrovaginal Fistulas Chapter 39: Anal Incontinence and Rectovaginal Fistulas

Section VI: Related Surgery Chapter 40: Breast Diseases: Benign and Malignant Chapter 41: The Vermiform Appendix in Relation to Gynecology Chapter 42: Intestinal Tract in Gynecologic Surgery Chapter 43: Nongynecologic Conditions Encountered by the Gynecologic Surgeon Section VII: Gynecologic Oncology Chapter 44: Malignancies of the Vulva Chapter 45: Cervical Cancer Precursors and Their Management Chapter 46: Cancer of the Cervix Chapter 47: Malignant Tumors of the Uterine Corpus Chapter 48: Ovarian Cancer: Etiology, Screening, and Surgery Chapter 49: Pelvic Exenteration Chapter 50: Pelvic Reconstruction After Gynecologic Cancer Surgery Chapter 51: Training the Gynecologic Surgeon

Section I: General Topics Affecting Gynecologic Surgery Practice

Chapter 1

A Brief History of Operative Gynecology Gert H. Brieger

Gynecology, spelled gynaecology, is defined by the Oxford English Dictionary as “That department of medical science which treats of the functions and diseases peculiar to women.” The word was first used as such in the middle of the 19th century. In 1867, gynecology represented the physiology and pathology of the nonpregnant state. Although most histories of gynecology trace its roots back to antiquity, the field of medicine we call by that name today really has had a fairly recent origin. The successful removal of an ovarian tumor by Ephraim McDowell in 1809 was as rare an event as it was a spectacular one. In the preceding centuries, the history of gynecologic surgery was closely tied to the history of general surgery, and the obstacles that had to be overcome were the same. Infection, hemorrhage and shock, and pain were all effective barriers to any but emergency surgical procedures in the days before anesthesia. “The history of gynecology,” Howard Kelly wrote in 1912, “seems to me more full of dramatic interest than the evolution of any other medical or surgical specialty.” Himself an accomplished historian of medicine, among his many other skills, Kelly noted that, “It was, notably, anesthesia which robbed surgery of its horrors, asepsis which robbed it of its dangers, and cellular pathology which came as a godsend to enable the operator to discriminate between malignant and non-malignant growths.” Here, in a nutshell, we have the landmarks of much of the history of gynecology of the last 150 years. There are many ways to approach the history of a medical and surgical specialty such as gynecology. The usual practice in textbooks that make an attempt to include some history is to tell the story in terms of who discovered what and who did which operation first. These facts are of interest but hardly constitute the history of the field. Besides the surgical operations of gynecology, the techniques devised, and the instruments to carry them out, there is much to be learned from the changing picture of diseases and their diagnoses; from the professionalization of the field, including the societies, journals, and textbooks that have been created; and from the education required to master the science and practice of operative gynecology. It is in these terms, rather than in tracing simply the great ideas and their creators, that this historical introduction proceeds. Any major medical textbook can itself serve as a convenient window through which we can see history unfold. Robert Hahn has vividly described the changing world view of obstetrics by examining the succeeding editions of Williams' Obstetrics since its first edition in 1903. Likewise, the 50 years that have elapsed since the first edition of Richard Wesley Te Linde's Operative Gynecology provide an equal opportunity to describe the major developments in the companion field of gynecology.

BARRIERS TO SURGICAL PROGRESS BEGINNINGS OF GYNECOLOGIC SURGERY IN 19TH-CENTURY AMERICA WOMEN AS PATIENTS IN THE 19TH CENTURY THE RELATION BETWEEN SURGERY AND GYNECOLOGY GYNECOLOGY IN THE 20TH CENTURY AND DR. TE LINDE'S BOOK OPERATIVE GYNECOLOGY, FIRST EDITION, 1946

BARRIERS TO SURGICAL PROGRESS Part of "Chapter 1 - A Brief History of Operative Gynecology" In ancient times, the lack of real anatomic knowledge was a barrier to the development of surgery. It is sometimes said that because the ancient Egyptians had effective techniques for the evisceration of bodies for mummification, they must have had a good knowledge of the body. However, removal of the internal organs during the embalming process was performed by technicians who did not concern themselves with the structure of the bodies they were preparing. Anatomy was pursued in Alexandria during the Hellenistic period, but it had few, if any, practical applications until a later time. By the end of the 13th century, anatomic dissection again became more common, but often it was limited to one or two public dissections a year or the study of animals. Surgeons were responsible for the few autopsies that were performed to determine the cause of death. This was especially important if a crime was suspected or drowning had to be established. Soranus, the Roman physician and writer who practiced in the reign of the Emperors Trajan (98–117) and Hadrian (117–138), is perhaps best known for his text entitled, Gynecology. This book is somewhat mistitled because it is mostly devoted to what we would call obstetrics. Soranus wrote about prenatal and postnatal problems, as well as those associated with delivery itself. This ancient text has been translated and has an excellent introduction by Owsei Temkin. Recently, it has been reissued in a paperback edition. Although Soranus' Gynecology still makes interesting reading, it hardly qualifies as an early text on the subject of operative gynecology. However, like other physicians of his time, Soranus clearly noted that the best midwife was one who was trained in all branches of therapy, “… for some cases must be treated by diet, others by surgery, while still others must be cured by drugs.” Although there were instances of human anatomy in earlier times, we generally begin the story with the work of Andreas Vesalius and the publication of his De humani corporis fabrica in 1543. Before this time, anatomic knowledge was not tied to the teaching and practice of medicine. The tradition of the surgeon-anatomists, of whom Vesalius was a stellar example, culminated in the late 18th century with the work of the English surgical teacher John Hunter (1728–1793) and his older brother William (1718–1783). It was William's classic book about the gravid uterus with its detailed engravings that shed new light on the structures of the female pelvis. In the 19th century, for all types of surgery, the problems of pain, hemorrhage, and infection had to be solved before operations could be undertaken safely. The problems of surgical dressings and postoperative infections were generally a matter of trial and error. The Scottish surgeon and gynecologist Sir James Simpson (1811–1870) urged his surgical colleagues to perform their operations on the kitchen tables of their patients to avoid the dangers of hospital infections, or “hospitalism” as it came to be called. In the 1840s, the Hungarian obstetrician Ignaz Semmelweis (1818–1865) showed clearly that puerperal fever could be prevented by disinfecting the hands of doctors before they examined their patients during the course of delivery. Despite good statistical evidence, his method of washing hands in chlorinated lime solution was not widely adopted. In fact, it met with outright resistance from most physicians. In this country, the Harvard anatomist and writer Oliver Wendell Holmes (1809–1894) met similar disbelief and resistance when he suggested in 1842 that it was the physicians themselves who were carrying the dreaded puerperal infections to their patients. In the middle 1860s, Joseph Lister (1827–1912), while working in Glasgow, began experiments using carbolic acid, a phenol derivative, to clean the instruments, sutures, and dressings he was using in his operations. He based his work on an understanding of the germ theory of disease, which was then just in its infancy as a major theory of disease causation. Lister believed it was important to prevent the germs present in the air or on instruments and sutures from entering the wound, which would prevent the formation of the heretofore much desired laudable pus. Lister, too, met much opposition to his method of antisepsis. Partly because of the frequent changes in the system he was developing, which made it difficult for others to follow him, and because of the inadequate understanding of the germ theory by most surgeons, it took nearly two decades for antiseptic surgery to become routine. In Lister's case, as was also true for Holmes and Semmelweis, some of the resistance undoubtedly stemmed from the fact that doctors never like being told that what they are doing is actually causing harm to their patients.

Lister encountered a great deal of opposition, particularly in his own country. Lawson Tait (1845–1899), an active and polemical gynecologist who settled in Birmingham, was staunchly opposed to Lister's system of antisepsis. Tait paid much attention to general cleanliness when he was operating, and he actually achieved quite good results. However, his older colleague, Spencer Wells (1818–1897) of London, was a devoted follower of the antiseptic system in his many ovarian operations, perhaps because he had a clear grasp of the role of microbes. In 1864, the year before Lister began using carbolic acid in Glasgow and 3 years before he published his first results, Wells published a paper in the British Medical Journal entitled “Some Causes of Excessive Mortality After Surgical Operations.” Wells clearly described the recent work on germs by Louis Pasteur (1822–1895) in France. There is no definite proof that Lister was aware of the paper, but it is hard to imagine that he did not know what was appearing in the national medical journal. Thus, gynecologists probably had a much greater hand in the development of safe surgery in the last century than is usually acknowledged.

BEGINNINGS OF GYNECOLOGIC SURGERY IN 19TH-CENTURY AMERICA Part of "Chapter 1 - A Brief History of Operative Gynecology" Opening the abdominal cavity to remove extrauterine pregnancies was successfully accomplished several times in the later 18th century but did not become routine until the advent of anesthesia and antisepsis/asepsis. Ephraim McDowell (1771–1830) (Fig. 1.1) made surgical history with his successful removal of a large ovarian cyst in his patient Jane Todd Crawford, who in 1809 rode 60 miles to her doctor's house in Danville, Kentucky, to undergo an untried operation without any assurance of cure and without the benefit of anesthesia. Although McDowell is often referred to as a backwoods physician, he was in fact a well-trained surgeon. His Edinburgh training probably gave him confidence in his diagnosis and courage to attempt a surgical cure rather than have his patient face certain death from her relentlessly growing tumor. During his study tour in Scotland, he probably heard that in the previous century the popular surgical teacher John Hunter had suggested such an operation, believing that “women could bear spaying just as well as did animals.” FIGURE 1.1. Ephraim McDowell (1771–1830). One of the earliest abdominal surgeons.

The drama of McDowell's case is best described in the words of the surgeon himself: In December, 1809, I was called to see a Mrs. Crawford, who had for several months thought herself pregnant. She was affected with pains similar to labor pains, from which she could find no relief. So strong was the presumption of her being in the last stage of pregnancy, that two physicians, who were consulted on her case, requested my aid in delivering her. The abdomen was considerably enlarged, and had the appearance of pregnancy, though the inclination of the tumor was to one side, admitting of an easy removal to the other. Upon examination, per vaginum, I found nothing in the uterus; which induced the conclusion that it must be an enlarged ovarium. Having never seen so large a substance extracted, nor heard of an attempt, or success attending any operation, such as this required, I gave to the unhappy woman information of her dangerous situation. She appeared willing to undergo an experiment, which I promised to perform if she would come to Danville…. With the assistance of my nephew and colleague, James McDowell, M.D., I commenced the operation, which was concluded as follows: Having placed her on a table of the ordinary height, on her back, and removed all her dressing which might in any way impede the operation, I made an incision about three inches from the musculus rectus abdominis, on the left side, continuing the same nine inches in length, parallel with the fibers of the above named muscle, extending into the cavity of the abdomen, the parietes of which were a good deal contused, which we ascribed to the resting of the tumor on the horn of the saddle during her journey. The tumor then appeared in full view, but was so large that we could not take it away entire. We put a strong ligature around the fallopian tube near to the uterus; we then cut open the tumor, which was the ovarium and fibrinous part of the fallopian tube very much enlarged. We took out fifteen pounds of a dirty, gelatinous looking substance. After which we cut through the fallopian tube, and extracted the sack, which weighed seven pounds and one half. As soon as the external opening was made, the intestines rushed out upon the table; and so completely was the abdomen filled by the tumor, that they could not be replaced during the operation, which was terminated in about twenty-five minutes. We then turned her upon her left side, so as to permit

the blood to escape; after which, we closed the external opening with the interrupted suture, leaving out, at the lower end of the incision, the ligature which surrounded the fallopian tube. Between every two stitches we put a strip of adhesive plaster, which, by keeping the parts in contact, hastened the healing of the incision. We then applied the usual dressing, put her to bed, and prescribed a strict observance of the antiphlogistic regimen. In five days I visited her, and much to my astonishment found her engaged in making up her bed. I gave her particular caution for the future; and in twenty five days, she returned home as she came, in good health, which she continues to enjoy. McDowell's patient long outlived her surgeon. He did not publish his feat until 1816, by which time he had performed several more oophorectomies. McDowell is sometimes cited as a pioneer of early ambulation, unwitting as it was in his case. If his sturdy patient had not recovered so well, her failure would surely have been blamed on rising too early from her bed after such extensive surgery. McDowell also did not mention the intense drama of this Christmas Day operation. When the townsfolk of Danville heard about his plan, they were incensed. They gathered in a tense group outside his house, with a rope slung over a tree, ready to lynch the surgeon if his “experiment” proved a failure. McDowell certainly had the nature of a true pioneer. T.G. Thomas, in his 1876 centennial review of obstetrics and gynecology, reported that Alexander Dunlap of Springfield, Ohio claimed he did his first ovarian operation in 1843. Dunlap said he sent the report of this case to a medical journal, which sent it back to him saying that they “could not publish the case of such an unjustifiable operation.” By 1876, Thomas wrote, “It is to estimate the amount of good this operation has bestowed upon humanity. Practised today in every civilized country in the world, yielding the statistics of seventy to seventy-five per cent of recoveries, and daily being improved in its various steps, it may well be regarded as one of the greatest surgical triumphs of the century.” In the middle decades of the 19th century, another American surgeon working in the South helped to popularize gynecologic surgery by another set of pioneering feats. James Marion Sims (1813–1883) told the dramatic tale of his development of a successful technique to repair vesicovaginal fistulas in his widely read autobiography The Story of My Life, which was published the year after his death (Fig. 1.2). He described his repeated attempts to achieve a permanent closure of these fistulas in a few of his young slave-women patients. Sims began his experiments in 1845 and continued them for 4 years. In these preanesthesia and preantiseptic days, Sims produced remarkable results. He had had no experience in pelvic surgery, and in fact claimed that he disliked it. It was his custom to turn away patients with pelvic disorders, referring them to other doctors in his Alabama neighborhood. Many of his planter friends owned slaves, some of whom suffered from vesicovaginal fistulas as a result of traumatic births. These wounds were considered incurable and made the young women unacceptable for household work. After several entreaties to help one of his planter friends who had such a slave, Sims began with a small group of women, operating on some of them repeatedly over the course of 4 years.

FIGURE 1.2. James Marion Sims (1813–1883).

View Figure

Sims' many failures only increased his determination to succeed. The colleagues who at first assisted him at the operations abandoned him, and his friends, he claimed, begged him to give up what was considered to be a hopeless effort. He trained other young slave patients to assist him, and on his 29th operation on one of the patients, he finally succeeded. In reviewing his work in 1852, Sims did cite several successful cases by other American surgeons between 1839 and 1849. He claimed originality for: 1st. for the discovery of a method by which the vagina can be thoroughly explored, and the operation easily performed [the Sims, or lateral, position]. 2nd. For the introduction of a new suture apparatus, which lies imbedded in the tissues for an indefinite period without danger of cutting its way out, as do silk ligatures. And 3rd. For the invention of a self-retaining catheter, which can be worn with the greatest comfort by the patient during the whole process of treatment. The new “suture apparatus” used silver wire. This provided the breakthrough needed for the successful repair of vesico-vaginal fistulae. Sims used silver in many of his other operations. In a tenth anniversary lecture at the New York Academy of Medicine in 1858, Sims somewhat immodestly told his august audience that the use of silver suture was one of the great achievements of 19th-century surgery. On the 21st of June in 1849, Sims proclaimed, “After nearly four years of fruitless labor, silver wire was fortunately substituted, for silk as a suture, and lo! a new era dawns upon surgery.” Sims soon left the South, and after a stay in Europe during the Civil War, he settled in New York, where a newly founded Woman's Hospital allowed him to develop gynecologic surgery with great success. With the advent of anesthesia and the use of antiseptic techniques, such surgery became increasingly routine. The repair of vesicovaginal fistulas and the removal of ovaries for a wide variety of indications were the beginning of the field of operative gynecology as it is known today. The story is, of course, not purely an American one. The English, French, and German contributions were important and can be found in any general history of medicine or of obstetrics and gynecology. In 1876 Sims became President of the American Medical Association, and in the same year he and others founded the American Gynecological Association. Even with the advent of effective and relatively safe anesthesia after 1846, it was several decades before surgeons were ready to increase the number of their operations. At mid-century and during the Civil War in the 1860s, surgery was generally confined to amputations after accidents; hernia repair when the intestine became incarcerated in the hernia sac, thus threatening life; an occasional ligation of a major vessel for aneurysm; and cystotomy for bladder stones. Therefore, Sims, operating in the 1840s, was truly a pioneer.

Also pioneers in the field of gynecologic surgery by mid-century were the Atlee brothers of Lancaster, Pennsylvania. They rediscovered oophorectomy, which was also being done in England by the 1860s, and were among the early leaders who performed myomectomy for fibroid tumors of the uterus. Of semantic interest is the changing terminology for ovarian surgery. Ovariotomy, often used imprecisely to refer to removal of the ovary, actually was first used in that way in the 1850s by James Simpson and other British gynecologists. Ovariotomy means to cut into the ovary for removal of a cyst or tumor. In the 1870s, gynecologists such as Edmund Peaslee of New York, in his book on ovarian tumors, stated that oophorectomy was a more precise and distinctive term for removal of the ovary. John Light Atlee (1799–1885) actively practiced medicine for 65 years, during which time he performed over 2,000 operations and attended 3,200 births. John Atlee performed 78 ovarian operations between 1843 and 1883, with 64 recoveries and only 14 deaths. Thus, he validated McDowell's work of the early part of the 19th century. Atlee's younger brother, Washington Lemuel Atlee (1808–1878) (Fig. 1.3), also was involved in some of the ovarian cases, but deserves separate credit for being one of the first to successfully treat the problem of uterine leiomyomata. FIGURE 1.3. Washington Lemuel Atlee (1808–1878).

View Figure

The Atlee brothers were relatively conservative gynecologic surgeons. It was their careful approach coupled with their obvious successes that gave other surgeons increasing confidence to operate. Thus, they played an important role in the early stages of operative gynecology as it developed into the specialty it would become in the next generation. Ovariotomy, the most controversial of gynecologic procedures, was also the key to making it a surgical specialty. Indeed, some gynecologists claimed that operating for ovarian cysts and tumors laid the groundwork for all abdominal surgery in the last decades of the 19th century. By the 1880s, the specialty of gynecology, or the science of women, as some historians have called it, was well on its way to being established as one of the subdivisions of medical labor. Ornella Moscucci, in her perceptive history of gynecology in Britain, quotes the eminent surgeon from Birmingham, Lawson Tait, in his aptly titled book of 1889, Diseases of Women and Abdominal Surgery: The great function of woman's life has for years made her the subject of specialists, male and female, the obstetricians. The subsidiary relations of her special organs and the special requirements of her physique, based upon these, have necessitated the establishment of another class of specialist, the gynecologist.

THE RELATION BETWEEN SURGERY AND GYNECOLOGY Part of "Chapter 1 - A Brief History of Operative Gynecology" The complex relation between general surgery and gynecology played a continuing role in the professional definition of gynecology as a 20th-century specialty. Moreover, several important contributions to surgery, such as chloroform anesthesia, rubber gloves, and early ambulation, were influenced by gynecologists as well as surgeons. The latter two items are discussed subsequently. By 1905, the Chicago gynecologist Franklin H. Martin (1857–1935) was convinced that the three closely allied fields—surgery, gynecology, and obstetrics—were making sufficient progress to warrant a new journal. There was a shared feeling, Martin wrote in the opening editorial of Surgery, Gynecology, and Obstetrics, “… that the field of the three allied specialties represented by its title is not over-cultivated, and that there is already a place for a creditable magazine representing in one publication these three divisions of surgery.” Another of the founding editors of the journal, the gynecologist J. Clarence Webster, wrote a provocative editorial in the first issue on “The Future of Gynecology.” Webster firmly laid to rest an idea that had gained some acceptance by 1905—that gynecology was doomed to extinction, to be gradually merged with the practice of the general surgeon. Webster assured his readers that contrary to what some had claimed, much advance had occurred in the preceding decades, and, moreover, “… it is very evident that almost all the important advances have resulted from the work of men who have given their entire energies to the specialty. At the present day the leading authorities everywhere are those who still limit their attention to this sphere of work.” In a programmatic statement to the American Gynecological Society in 1920, Robert L. Dickinson (1861–1950) contended that gynecologists promote surgery. “But if we be just surgeons, by surgeons we may be displaced.” In this presidential address to the Society, Dickinson claimed that gynecologic procedures constituted one fourth of all surgery, but this hardly accounted for the extent of the field, “… since operation is needed by less than one-tenth of the patients that come to the doctor for ailments peculiar to women (childbearing not included).” It was true, of course, that for much of the preceding century, gynecology was a medical rather than a surgical discipline, often taught in medical schools as part of the course on diseases of women and children. In the early decades of the 20th century, the professional battles between the general surgeons (who increasingly dominated the field of abdominal surgery) and the gynecologists (who wished to lay claim to the same territory) waxed and waned. Dr. Howard Longyear of Detroit noted in 1917 that general surgeons tended to scorn the area of the pelvis, while this area was being increasingly perfected by gynecologists. These surgeons wanted to move upward in the body from surgery of the female genitalia and the pelvis to the abdomen. Longyear also noted that the Sims operation for vesicovaginal fistula did more to establish operative gynecology as a specialty than did any other single procedure or development. The complex relations between surgery and gynecology also can be traced by following the name changes in the American Medical Association specialty section. In 1903, at its founding, it was called Section on Obstetrics and Gynecology. From 1912 until 1936, it was called Section on Obstetrics, Gynecology, and Abdominal Surgery. Then the name was changed once again and dropped the abdominal surgery component. One area of joint progress forged by surgeons and gynecologists was the introduction of the use of rubber gloves, which helped to expand the work of all surgeons. The idea of using some form of protective covering for the surgeon's hands occasionally appeared in the medical literature in the early decades of the 19th century, but it was not until the end of the century that some of the associates of Dr. William S. Halsted (1852–1922) at the Johns Hopkins Hospital in Baltimore began to use gloves routinely. About two decades after their introduction, Dr. Halsted recalled the story: In the winter of 1889 and 1890—I cannot recall the month—the nurse in charge of my operating room complained that the solutions of mercuric chloride produced a dermatitis of her arms and hands. As she was an unusually efficient woman, I gave the matter my consideration and one day in New York requested the Goodyear Rubber Company to make as an experiment two pair of thin rubber gloves with gauntlets. On trial these proved to be so satisfactory that additional gloves were ordered. In the autumn, on my return to town, the assistant who passed the instruments and threaded the needles was also provided with

rubber gloves to wear at the operations. At first the operator wore them only when exploratory incisions into joints were made. After a time the assistants became so accustomed to working in gloves that they also wore them as operators and would remark that they seemed to be less expert with the bare hands than with the gloved hands. I think it was Dr. Bloodgood, my house surgeon, who first made this comment and that he was the first to wear them invariably, when operating…. Dr. Hunter Robb in 1894, in his book on aseptic technic recommended that the operator wear rubber gloves. Dr. Robb was, at that time, resident gynecologist of the Johns Hopkins Hospital and had frequent opportunities to observe the technic of the surgical clinic. Gynecologists were also closely involved in the form of postoperative care we have now come to take for granted—early ambulation after surgery. With the change from 2 or 3 weeks of enforced bed rest after surgery to active ambulation within a few hours of the operation, we have improved recovery, shortened hospital stays, and reduced costs as well as postoperative complications. But like all new techniques or practices, early rising after surgery did not win rapid acceptance. Ephraim McDowell's patient in 1809 not only was ambulant early, but also engaged in physical tasks such as making her own bed. Her surgeon clearly was not pleased with her activity, which was not in keeping with customary and usual practices of the day. What we call early ambulation was not found again in the medical literature until the very last year of the 19th century, when Emil Ries, a professor of gynecology in Chicago, published a landmark paper, which soon disappeared from view. It was rediscovered four decades later. Ries noted in his 1899 paper that he wanted to change treatment radically by freeing patients from “… many irksome and disagreeable features of convalescence following vaginal and abdominal surgery.” Ries found that his patients could be fed and allowed out of bed much sooner than was the usual custom. “Very soon I found,” he wrote, “that the period for which it was advisable to confine such cases to bed could be counted by hours instead of days, so that of late I have allowed my patients to get up within twenty-four to forty-eight hours and to leave the hospital four to six days after their vaginal celiotomy.” These patients, Ries also noted, did not have the listlessness or muscular weakness that was usually seen after 2 or 3 weeks in bed. In the preoperative preparation of his patients, Ries also went against the usual custom of completely emptying the bowel. Most textbooks, he said, claimed that early action of the bowels helped to prevent peritonitis. However, in most patients with an empty intestinal tract, regular movements did not resume until after they were eating a regular diet. Ries maintained that cause and effect were confused because it was not movement of the bowels that prevented peritonitis, but freedom from inflammation that allowed the bowels to move. At the meeting of the Southern Surgical and Gynecological Society in Baltimore in 1906, H.J. Boldt described 384 cases of early ambulation that he had accumulated since 1890. All recovered well. Ironically, Boldt reported, the most serious objection raised by his colleagues was that early ambulation increased the risk of thrombosis. This was clearly wrong, he said, from both a theoretic and an empirical point of view, because his patients had a better circulation from exercising. Early ambulation was discussed repeatedly in the succeeding decade, but it received far from universal acceptance. Even Howard Kelly, the country's leading teacher of gynecology, noted in 1911 that great progress was made as a result of Boldt's and Ries' work, but that it was far from standard practice. Early ambulation really became a routine practice with the exigencies of World War II (which resulted in a shortage of hospital personnel) and with the work of Daniel J. Leithauser, a general surgeon from Detroit who rediscovered Boldt and Ries. Although doctors may not have prescribed early ambulation, as in the case of McDowell, patients probably were up and about far more often than we realize. Dr. Bert Dunphy of San Francisco told me that when he had a hernia repair while he was a house officer at the Peter Bent Brigham Hospital in 1938, his surgeon prescribed strict bed rest after the operation. Dunphy was up on the first day and thereafter and felt perfectly well, if a bit guilty.

GYNECOLOGY IN THE 20TH CENTURY AND DR. TE LINDE'S BOOK Part of "Chapter 1 - A Brief History of Operative Gynecology" In the 1890s, when Thomas Cullen (1868–1953) was a medical student in Toronto, he recalled that “… there were anteversions, anteflexions, retroversions, and retroflexions and that some of the displacements might be relieved by appropriate pessaries.” Abdominal gynecologic operations, Cullen continued, “… were limited almost entirely to the removal of large ovarian cysts. An occasional myomatous uterus was removed, but the fatality in this class of cases was so high that the operation was rarely attempted.” Cullen also said that he did hear of cancers of the uterus in his student days, but only cauterization or curettage was performed. Entire removal of the uterus was not yet being done. By the turn of the 20th century, the leadership of gynecology in this country had clearly moved to the new Johns Hopkins Hospital, where Howard A. Kelly (1858–1943) (Fig. 1.6) began to train a series of young men who put gynecology on a strong academic footing in the next two generations. Kelly received both his bachelor's and medical degrees from the University of Pennsylvania. After his medical graduation in 1882, Kelly spent some time in Germany learning the latest surgical and pathologic techniques. Back in Philadelphia at Kensington Hospital, he soon acquired a reputation as a brilliant operator. When his fellow Philadelphian William Osler became Chief of Medicine at the opening of the Johns Hopkins Hospital in 1889, he urged the trustees to hire Kelly as Chief of Obstetrics and Gynecology. FIGURE 1.6. Howard A. Kelly (1858–1943). (From: Davis AW. Dr. Kelly of Hopkins. Baltimore: The Johns Hopkins Press, 1959, with permission.)

View Figure

At age 31, the youthful-appearing Kelly, who many patients thought was still a student or resident, initiated a residency program in gynecology with a strong link to the pathology department. Even more than half a century later, the leading texts in the field—Eastman's (Williams') Obstetrics, Te Linde's Operative Gynecology, and Novak's Gynecologic and Obstetric Pathology—were written by professors in Baltimore who had received their training at Hopkins with Kelly and his assistants. Kelly soon found that his interests and skills were in gynecologic surgery; therefore, he turned the obstetric service over to J. Whitridge Williams (1866–1931), who became a leader in that field and the author of the most widely used textbook of the time. Kelly had a great interest in the female urinary system, realizing that the symptomatology of urinary tract disease is often intertwined with that of the reproductive organs. He invented the air cystoscope and devised ureteral catheters. He was the first to plicate the vesical sphincter for stress incontinence of urine. Physicians from all over the world came to Baltimore to watch him operate (Fig. 1.7 and Fig. 1.8).

FIGURE 1.7. Dr. Howard A. Kelly's operating room at the Johns Hopkins Hospital. To the left is the door to the corridor. In the center is the door to the ether room. The rubber pad was used for drainage during irrigation of the abdomen. A similar pad was developed by Dr. Kelly for drainage of blood and amniotic fluid during and after a vaginal delivery.

View Figure

FIGURE 1.8. Howard Kelly operates. Grouped about the operating table, left to right, are: Emma Beckwith, head nurse; Jay Durkee (seated); Thomas S. Cullen; Max Br?del (center); Elisabeth Hurdon; J.E. Stokes; and John G. Clark. (From: Davis AW. Dr. Kelly of Hopkins. Baltimore: The Johns Hopkins Press, 1959, with permission.)

View Figure

Kelly's legendary operative skill was well described by Cullen, who later became one of Kelly's outstanding residents and successors to the chair at Hopkins. Kelly and Hunter Robb, his earlier resident, went to the Toronto General Hospital not long after Kelly became Chief at Hopkins. Cullen was an intern in Toronto at the time and handled the instruments during an operation that Kelly and Robb had agreed to perform. Cullen's description speaks for itself: I turned around to thread a needle and when I turned back found to my amazement that the operator had the abdomen open. Operators in the General often took ten minutes to get that far. After cutting through the skin, fat and fascia they were apt to get lost in the muscles. Kelly and Robb working together used dissecting forceps as I had never seen them used. One man pulling each way, the cleavage between the muscles was seen at once and the opening in the abdomen could be completed without difficulty. I watched, fascinated, while Kelly went ahead and finished that operation and did the second, working with clock-like precision and at a speed I had not imagined possible. By the time he had finished, the course of my professional life was decided. Up to that afternoon I had intended to be a physician. From that afternoon I knew I had to be a surgeon. Chance often determines the course of one's life, so it was fortunate for Cullen that he had to wait 6 months for his residency with Kelly to start. He used this time to begin the study of pathology with William H. Welch at Hopkins, and it was the close alliance of gynecology and pathology, begun by Kelly and continued by Cullen, that shaped the careers of many future gynecologists at Hopkins and elsewhere and determined the course of the field itself. In 1898, Kelly published a two-volume textbook called Operative Gynecology, certainly the direct ancestor of the volume you have in your hands. Kelly wrote in the preface, “My aim in writing this book has been to place in the hands of the many friends who have from time to time visited me and followed my work, a convenient summary of the various gynecological operations I have found best in my own practice.”

Although gynecology at the end of the last century was still a very young science, in Kelly's words, change was at hand: “Although I have spent several years in the preparation of my book, so rapid have been the changes in the gynecological field that I have found it necessary to rewrite some of the chapters two and even three times.” A little more than a dozen years later, in the preface to his text entitled Medical Gynecology, Kelly reiterated the pace of the changes: “What a transformation two generations have witnessed in the field of gynecology! From modest beginnings, as a sort of minor specialty coupled with diseases of children and often professed by general practitioners with no special training, it has grown to the dignity of a major surgical specialty, so extensive that many gynecologists of today [1912] claim the entire field of abdominal surgery as their proper domain by right of discovery and conquest.” This was also a time when radical or complete removal of tumors and repair of hernias became increasingly common. Kelly and his residents were pioneers in radical hysterectomy when Hugh Young of Hopkins introduced radical prostatectomy. What imparted even greater value to all of Kelly's texts were the illustrations of Max Br?del, a German medical illustrator brought to Hopkins by Kelly (Fig. 1.9). Br?del's contributions to operative gynecology, including Te Linde's text 50 years after Kelly's, were portrayals of operative techniques, pelvic anatomy, and pathologic conditions. He set a standard for medical illustration never attained before or surpassed since. FIGURE 1.9. Max Br?del. (From: Robinson J. Tom Cullen of Baltimore. New York: Oxford University Press, 1949, with permission.)

View Figure

As long ago as 1900, in his classic text, Cancer of the Uterus, Thomas Cullen, student of and successor to Kelly (Fig. 1.10), wrote that, “The number of cases of cancer of the genital tract coming too late for operation is so appalling that the surgeon is ever seeking to devise ways and means by which the dread malady may be more generally detected at the earliest possible moment—at a time when complete removal of the malignant tissue is still possible…. But since it is the general practitioner who, as a rule, is the first consulted, upon him largely falls the responsibility of arriving at a timely diagnosis.”

FIGURE 1.10. Thomas S. Cullen (1868–1953). (From: Robinson J. Tom Cullen of Baltimore. New York: Oxford University Press, 1949, with permission.)

View Figure

One of the greatest advances in gynecology in this century has been the improvement in the early detection and cure rate of cancer of the uterine cervix that has resulted from the development of cytology and the recognition of carcinoma in situ. In 1943, George N. Papanicolaou (1883–1962) and Herbert Traut (1894–1963) published their seminal monograph entitled Diagnosis of Uterine Cancer by the Vaginal Smear. Papanicolaou had worked on this technique since the 1920s, but, like many other innovations in medicine, it took years to find widespread acceptance. Further publications by Papanicolaou and others, notably Ruth Graham, demonstrated beyond a doubt that cytologic studies could almost infallibly detect cervical cancer. Cancer in situ was recognized early in the century by Cullen and in 1912 by J. Schottlander and F. Kermauner, but its relation to invasive cancer was not well understood. This relation was more clearly described in 1944 by G.A. Galvin and Te Linde in several reports. Since then, the relation has been amply confirmed, and early cervical cancer has become a detectable and curable disease. Since its inception by Hans Hinselmann in Germany in the 1920s, colposcopy has given a new dimension to the assessment of cervical carcinoma, making blind, random cervical biopsies unnecessary and providing more accuracy in finding and treating localized lesions. By the early 1970s, the editor of a new journal, Gynecologic Oncology, pointed out that “… the scientific importance of gynecologic oncology may be gained from the observation that the tumors that we study and treat are prototypes for cancer in other areas of the body, for the histogenesis of the two principal uterine cancers is probably understood better than that of any other tumor in the body.” Another diagnostic and therapeutic procedure that has been a milestone in the history of pelvic surgery is the use of the laparoscope. The idea of viewing the intraabdominal organs was explained in 1911 by Bertram Bernheim, a surgical resident at Hopkins. He described two cases in which a proctoscope was passed through a small abdominal incision. He observed the organs in the abdomen by using a reflected light. Others soon tried the technique, and by the late 1920s, H. Kalk, a German surgeon, was avidly promoting peritoneoscopy. The method gained credibility through his many examinations and publications. Diagnostic possibilities were further increased in the late 1940s with the introduction of fiberoptics. Culdoscopy has been replaced by laparoscopy now. In recent years, many operative and diagnostic procedures formerly requiring a major pelvic operation have been performed successfully through the laparoscope. Among these are tubal sterilization, lysis of pelvic adhesions, evaluation of chronic pelvic pain and infertility, evaluation of treated pelvic malignancies, and, more recently, use of the carbon dioxide laser for vaporization of endometriosis implants throughout the pelvis. The immediate post–World War II years were a period of truly astounding medical developments and saw

the explosive growth of medical research funding and new hospital construction. After 1945, penicillin became available for civilian use, and this was soon followed by other antibiotics. Hormone replacement became increasingly possible, and in 1946, the year that Richard Te Linde published the first edition of this textbook, Congress passed the Hill-Burton Act, making federal funds available to localities for the construction of new hospitals. These and other developments of the time greatly changed and expanded the work of medicine.

OPERATIVE GYNECOLOGY, FIRST EDITION, 1946 Part of "Chapter 1 - A Brief History of Operative Gynecology" Richard Wesley Te Linde was born in Wisconsin in 1894, and except for the years he attended a small liberal arts college in Holland, Michigan, he spent all his formative years in Wisconsin. When he was ready to go to medical school, he went to Madison, but in 1916 the University of Wisconsin had only a 2-year school. Te Linde completed the 2 preclinical years and then transferred to Johns Hopkins for the final 2 years. He graduated with the class of 1920 and spent the rest of his professional career associated with Hopkins, where he became chief of the gynecology division of the Department of Surgery and then chair of the separate Department of Gynecology in 1939. He held that post until his retirement in 1960, when the newly reunified Department of Obstetrics and Gynecology was reestablished. Just as his teacher Howard Kelly had felt the need to compile a textbook of operative gynecology half a century earlier, Te Linde believed that the many-sided specialty that gynecology had become by World War II required a new text. With Kelly's earlier text as a model, Te Linde wished to incorporate the vast changes that had occurred in the period separating the two books. During this 50-year span, there were changes in our knowledge of the hormones, new surgical techniques, and the ability to visualize the abdominal and pelvic organs. Te Linde chose the same simple title for his own text. Although Kelly's had been published by Appleton in New York, Te Linde chose Lippincott in Philadelphia. Gynecology, Te Linde wrote in the preface in 1946, was no longer to be considered simply a branch of general surgery. The gynecologist, he stressed, must still be a good surgeon but must also master the pathology of gynecologic disorders and the newly burgeoning field of endocrinology. New books were appearing in all these fields except gynecologic surgery, and it was this void that Te Linde wished to fill. Te Linde wrote his text with the “… primary purpose of describing the technic of the usual and some of the rarer operative procedures. It also includes indications for and against operations as well as pre- and postoperative care of patients.” Gynecologic pathology, Te Linde stressed in the Hopkins tradition, is the bedrock of good gynecologic surgery. “Without an understanding of it, surgery becomes merely a mechanical job, and errors in surgical judgment are inevitable.” In the organization of his text and in the subsequent editions over the succeeding half century to the present edition, one can readily see important landmarks in the history of operative gynecology. Some of these were discussed in the preceding. The 751-page first edition of 1946, all of it written by Te Linde, had a first printing of 5,000 copies, which quickly sold out. A second printing was equally successful. The reviews have always been laudatory. Of the sixth edition of 1985, edited by Richard Mattingly and John Thompson, the Journal of the American Medical Association reviewer ended by saying, “I cannot imagine any gynecologist who performs surgery doing without it, first as a primer and then as a reminder.” By 1962, when the third edition appeared and Te Linde had retired from the chairmanship of his department, he decided that, like all the other major medical textbooks of the time, his book needed a group of authors to bring out new revisions. In the preface to that edition, he states that his book has never been simply a manual of surgical technique—that surgical philosophy is equally important. “What does it profit a woman if the operation is technically perfect and the procedure unnecessary or even harmful?” One reason unnecessary procedures still prevailed, Te Linde noted, was the lack of knowledge of gynecologic pathology, still the “bedrock upon which good surgery is done.” Therefore, Te Linde justified including a considerable amount of pathology in his text. Pathology is what has differentiated gynecologic surgery from general surgery since Howard Kelly's years at the turn of the century. Surgical texts, and by implication their surgical readers, have generally not devoted nearly as much attention to pathology as have gynecologists, some of whose leaders have actually been very well versed in pathology. The fact that Te Linde could produce three editions, each larger than the first, is a testament to his broad knowledge of his field, his ability as a writer, and his stamina for hard work. He died in Baltimore in 1989 at the age of 95. In the decades since the third edition of 1962, the world of medicine and the society in which it is practiced have seen much change. By the mid-1960s, when significant advances in the treatment of infections, malignancies, and hormonal disorders had become evident, these successes had an impact on gynecology just as they did in other areas of medicine. The reduction in mastoid infections, for instance, has changed the practice of the otolaryngologist considerably. In gynecology, the reduction in major pelvic

inflammatory disease forced gynecologists to focus more of their attention on other disorders. Also affecting gynecologic surgery by the middle of this century were significant improvements in obstetric practices, which sharply reduced injuries to the bladder and rectum. Hysterectomies and suspensory operations were not performed for vague complaints of illness as often as they had been. We have also lived through social revolutions that have changed the way our society carries on its business and dispenses its social prerogatives. Especially prominent in the 1960s, a civil rights movement, greater concern for our environment, a resurgence of consumer rights, and a revitalized women's movement profoundly affected our social institutions, including medicine. Within medicine, no specialty has been more touched by these trends than obstetrics and gynecology. The new feminism viewed abortion, childbirth, contraception, and gynecologic surgery as a means of social control of female patients by doctors, most of whom were male. The feminist movement challenged not only the domination of doctors but also the supposed benevolence of their knowledge and practices. As the world has changed, so have our expectations. In the decades after the first edition of Te Linde's book appeared, when wonder drugs were touted as curing previously untreatable illnesses, the public began to expect much from its doctors, and we were not shy in claiming that ever-greater investments in medical research would lead to more cures. It is hardly surprising, then, that in these last few decades, as we began to spend increasing amounts of our gross national product for health, those who paid the bill became increasingly interested in seeing just what their money was actually buying. Like most other social institutions, medicine lost much of the autonomy it had for so long taken for granted. Although as a profession we did not always get all we wanted, we were for decades amazingly adept at preventing those things we did not want. Now that, too, has changed, as has the practice of medicine. The division of labor in all areas of medicine grew as the 20th century progressed. In the last decades of the century, what used to be called general practice became the specialty of family practice. In the Anglo-American world of the late 20th century, both obstetrics and gynecology were caught in the middle of the battles between specialists and generalists. Likewise, they became involved in the tensions among primary, secondary, and tertiary medical care. Similar strife occurred in earlier centuries among those vying for a place and for status among physicians caring for women in childbirth and in disease. If one looks at the Table of Contents of this edition and compares it to a simpler period of half a century or a century ago, one will see what great breadth the field of operative gynecology continues to enjoy.

Chapter 2 The Ethics of Pelvic Surgery Kenneth J. Ryan

THEORY AND PRACTICE IN MEDICAL ETHICS ETHICAL ISSUES IN SURGICAL TRAINING ETHICAL ISSUES IN RESEARCH ON HUMAN SUBJECTS INFORMED CONSENT AND THE OPERATIVE PERMIT SURGICAL COMPETENCE AS A MORAL COMMITMENT MORAL AND LEGAL ISSUES OF ABORTION

TERMINAL ILLNESS, ADVANCE DIRECTIVES, AND EUTHANASIA

THEORY AND PRACTICE IN MEDICAL ETHICS Part of "Chapter 2 - The Ethics of Pelvic Surgery" Ethics is concerned with the right and wrong, the good and bad of human behavior as it materially affects other human beings and some would say even all other sentient beings. In medicine, ethics has been embodied in the venerable Hippocratic oath and aphorisms that have provided a guide over the centuries on how physicians should deal with their patients. The major goals of medicine have been to provide relief from pain and suffering, to try to treat disease, and at least minimize harm in the process. Over the years, medical ethics has also been influenced by religious traditions of charity and the Good Samaritan and by the evolution of medicine into a profession with formal licensure and societal obligations. For all its good works, medicine also has had morally problematic traditions of being paternalistic and autocratic. These traditions were severely criticized during the social upheavals of the 1960s concerned with civil rights, women's issues, and the celebration of individual autonomy. It was this social unrest in the 1960s and 1970s, as well as the revolutionary changes in medical technology such as organ transplantation, artificial organs, mechanical life support systems, genetic engineering, and assisted reproductive technologies that gave birth to a broader field of bioethics and generated the public discussions and need for formal teaching of ethics in medical schools and residency training. Physicians have learned a new vocabulary as well as the theoretical basics of applied and normative ethics, have participated in public debates, and have communicated with moral philosophers, lawyers, and social scientists, who have been attracted to the problems in this field and often participate on hospital ethics committees. Hospital ethics committees are now reviewed by the Joint Commission for the Accreditation of Hospitals when evaluating a hospital's protection of patients' rights in matters such as terminal care and informed consent. In a 1962 edition of this text, Dr. Te Linde wrote a short piece on the moral issues of induced abortion at a time when it seemed to everyone that abortion was the only moral problem the gynecologist faced. This narrow view of ethics was common in most medical texts and medical schools. The teaching of specific courses in medical ethics began in earnest only in the past 20 years or so. This chapter recognizes the more explicit role that ethics now plays in the practice of the discipline and acknowledges the profound changes in ethical concerns brought about by advances in medical technology and surgery. During the course of the events described in the preceding, the concept of medical ethics changed from the paternalistic Hippocratic ideal, which emphasized physician concern for the patient's well-being, to a principle- and rights-based ethic in which physicians' obligations were more clearly defined. In 1973, the American Hospital Association issued “A Patient's Bill of Rights,” which emphasized institutional responsibility and the new dimension in the traditional physician–patient relationship when care is provided in the hospital. The disclosure by the physician to the patient would now include not only information on choices in diagnosis and treatment, but also how residents and students participate in patient care and procedures, whether any research is involved, and any financial relationship to laboratories or diagnostic facilities that might pose conflicts of interest for the doctor. In 1978, the Belmont Report was issued by the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. This report responded to the query by the US Congress concerning the ethical principles that should underlie the conduct of research and medical practice when funded by the US government. The report emphasized the following principles. Respect for persons. Treating research subjects and patients as autonomous individuals and obtaining informed consent for all research and medical treatments proposed. This principle was derived with Emanuel Kant's practical imperative in mind: to treat others as an end and never simply as a means. Beneficence. Doing a risk-benefit analysis and favoring procedures and treatments that protect or benefit subjects and patients and avoid unnecessary harm. This principle was derived with the classic Hippocratic admonition: to help and at least not harm. Justice. Being fair in the process of selecting subjects or patients and fair in the risks borne and the benefits received by them. This final principle was derived with John Rawls' elaborate “Theory of Justice” in mind (Table 2.1).

TABLE 2.1. Ethical Principles for Medical Research and Practice

Show respect for patients and research subjects by obtaining voluntary informed consent.

Demonstrate beneficence toward patients and research subjects by seeking their well-being and avoiding harm.

Provide justice to patients and research subjects by treating them fairly in the distribution of benefits and burdens.

These principles are guidelines for considering ethical issues in research and practice, but they provide less assistance when there is conflict in applying them to specific cases. In general, precedence in the law and most ethical teaching is that the patient's autonomy and right to respect require that a physician must not do anything even to benefit the patient without the patient's express permission, unless in an emergency when such permission cannot be obtained readily. Individuals of diminished autonomy, such as small children and those with mental handicaps, are protected by a requirement that physicians seek informed consent from the next of kin or a designated proxy. There are many other approaches to ethical analysis. Some have advocated a case-based method, working from analogy or the pragmatic approach, which is to consider in the broadest terms what difference there would be in taking one or another course of action. There is much to medical ethics not covered by this discussion that still requires wisdom and judgment by the practicing physician for resolution and, when needed, the assistance of colleagues and the hospital ethics committee.

ETHICAL ISSUES IN SURGICAL TRAINING Part of "Chapter 2 - The Ethics of Pelvic Surgery" Most surgeons trained before World War II attended medical schools and participated in residency programs affiliated with large, inner-city hospitals, where patients without adequate resources or health insurance received care. It was accepted that these patients would receive treatment and even surgery by students, interns, and residents, preferably under the watchful eyes of skilled volunteers or paid clinical faculty and only when the trainee had reached the necessary level of competence. The trade-off was free care for helping to train the doctors of the future. The care was not as luxurious as in the private pavilions or hospitals, but the quality of care was usually as good and sometimes better than private care because of the openness of the process and the enthusiasm and dedication of the young doctors-in-training and their supervisors. When health care became an entitlement under Medicare and Medicaid government funding and health insurance became more prevalent, many patients sought the services of private physicians. The so-called free care or resident services were often unable to recruit sufficient patients to provide adequate training for new surgeons. This was particularly true in pelvic surgery; given the choice, these patients sought refuge from the clinic, where privacy and dignity were hard to maintain, and fled in large numbers to private doctors' offices. Surgical teaching thereafter often involved the private patient, with the resident ultimately doing complete procedures under the watchful eye and assistance of the patient's private doctor. The problem with this arrangement was that patients were poorly informed or uninformed about the role played by physicians-in-training, and this often became apparent only when complications arose or the medical records were reviewed in malpractice cases. There was even an expose of this on national television when an investigative reporter informed a deceived patient on camera that it was the resident and not the professor who had actually done his complex surgery. We recall that the patient said he did not believe it. Patients are typically better informed now, and it is ethically necessary to inform the patient about teaching or training in each case so that objections can be dealt with or other arrangements made. Most surgeons involved in teaching programs tell their patients that modern surgery involves a team effort and that residents may be involved in assisting or operating with them, but that, as the private surgeons of record, they not only will be present but also will be in charge and responsible for all that takes place. Teaching should not take place in the operating room without such a disclosure and the patient's informed consent. The resident's role, status, and experience should be clear to the patient. We are long past introducing medical students to patients as doctors rather than revealing their true status. Another troubling aspect of modern surgical care is the disappearance of preoperative admission time so that surgery is scheduled on the day of arrival. This further diminishes the opportunity for contact preoperatively of the resident-in-training with the patient and compromises care and education even further.

ETHICAL ISSUES IN RESEARCH ON HUMAN SUBJECTS Part of "Chapter 2 - The Ethics of Pelvic Surgery" Although the Nuremberg medical war crimes trials of Nazi doctors and the ensuing Nuremberg Code in 1949 represent a landmark in bioethics on human subject research, they made little impact on research practices in the United States. Biomedical research expanded exponentially after World War II without much in the way of regulations or oversight. The Nuremberg Code established the need for the “voluntary consent of the human subject.” It covers such other matters as the need to justify a study in terms of expected beneficial results and risks, avoiding harm and injury to subjects, freedom for the subject to withdraw, and obligation of the investigator to stop a study if continuation would likely cause injury, disability, or death. This had to do with Nazi Germany, however, and did not seem applicable to democratic countries like the United States. Much of the US medical community lost sight of the fact that the German doctors involved were distinguished professors and academicians and that they were breaking their own laws regarding the protection of vulnerable populations. In the war crimes trials, these doctors used familiar excuses to justify their research, such as that the end (more knowledge) justified the means they employed. In 1966, the US Public Health Service introduced the requirement that all human subject research funded by the government must be peer reviewed by a local institutional review board. The objective was to protect the rights of the individuals involved, and review the quality of the informed consent and the risks and benefits of the study. This regulation was the result of a growing awareness that these kinds of safeguards often were not followed even in the best academic institutions in the United States. In 1966, Dr. Henry K. Beecher, professor of anesthesia at Harvard University, published an article in the New England Journal of Medicine describing ethically troubling studies published in good journals from prestigious institutions. One example involved surgeons who performed thymectomies to check the survival of skin homografts in children from 3½ months to 18 years of age as an add-on to cardiac surgery. Another publication described transplantation of a malignant melanoma from a terminally ill daughter to her consenting mother to look for tumor antibodies. The mother died of diffuse melanoma a little more than 1 year later. The most notorious nonsurgical studies that reached the attention of the public included the following. In the Jewish Chronic Disease Hospital, cancer cells were injected into patients in 1963 without their knowledge or consent to see if they would reject them. This study originated at the Memorial Sloan-Kettering Cancer Institute. The Tuskegee study, which ran from 1932 to 1972, was organized by the US Public Health Service to follow the natural history of syphilis in a cohort of 400 rural black men without providing any treatment and without any real informed consent. In the Willowbrook study, new residents on admission to an institution for the retarded were infected with hepatitis virus to study the course of the disease and search for a vaccine. When news of such studies reached the public, the US Congress in 1974 formed the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research (see the preceding) to craft suitable federal regulations on the ethical conduct of research and develop the Belmont Report to establish the ethical standards that should prevail. Although the regulations apply to research funded by the government, they should cover all research involving human subjects, regardless of the source of funding. Surgeons continually try to innovate and improve their skills and the procedures used, but they must be mindful of the possible need of antecedent work in basic science and the animal laboratory and the need for informed consent of the patient and peer review when significant deviation from standard practice is contemplated. In most instances, new surgical procedures should be developed in appropriately controlled research trials with all the safeguards needed for the protection of human subjects.

INFORMED CONSENT AND THE OPERATIVE PERMIT Part of "Chapter 2 - The Ethics of Pelvic Surgery" It is difficult now to operate on a patient without first having a signed consent form in the medical record. Strange as this may seem to the surgeon today, this has not always been the case. The practice of requiring a signed operative permit or consent gained momentum in the 1970s and has followed the change in attitudes about patient rights and medical ethics. It has also followed successful court cases in which the patient—suffering a complication of surgery not caused by negligence—could prove that the physician failed to inform him or her adequately of the gravity or frequency of a risk or failed to mention the risk altogether. Physicians have sometimes withheld complete disclosure of risks to avoid frightening the patient or ensure that the patient would not refuse surgery. The courts have held that the physician has a duty to disclose all the information a reasonable person would require to make an informed decision. According to the courts, the patient has a right to the information even if it results in a decision to forgo a life-saving procedure. Before 1960, there seldom was a requirement for a written document; the transfer of information and patient consent, if they occurred, were private transactions between patients and their doctors. The demand for informed consent started in research (see the preceding) and then spread to the practice of medicine, fostered by more openness in medicine, a greater interest in patient autonomy, and court decisions requiring disclosure. Today, informed consent is required for all operative procedures and has at least five elements (Table 2.2).

TABLE 2.2. Elements and Content of Informed Consent

Disclosure: Advise patients of their diagnoses and suggest surgical therapy, alternatives, and risks and benefits of all options, including no therapy, in sufficient detail to satisfy the needs of a reasonable person to make an informed choice.

Comprehension: Use language and descriptive materials appropriate to the patient's level of comprehension, and determine that the patient understands by asking for a summary in the patient's own words.

Voluntariness: Be certain the patient is free of coercion or constraints on his or her ability to choose freely.

Competence: Determine that the patient has no evidence of limitation in ability to understand and act independently on the information disclosed.

Validation: Obtain the written consent.

In practice, this involves informing the patient of the diagnosis, including the degree of certainty of the diagnosis; the recommended surgery; and possible alternatives with their expected outcomes, risks, and benefits, including those of no therapy. Determining the correct content to disclose to the patient still requires some judgment because it is impossible to disclose every contingency. The shared experience of colleagues and an institutional record of previous operative complications are helpful, as are preprinted consent forms that have been prepared for various procedures based on a collective experience. The consent should be obtained by the operative surgeon well in advance of surgery in a comfortable setting with adequate time for the patient to ask questions and make an informed decision. It is important to remember that risks have two dimensions—incidence and severity—and that sometimes a rare occurrence needs mentioning if it is of a sufficiently severe nature. Although an important obligation of the physician is to benefit the patient, there are few instances in which withholding important information can be justified on this basis.

SURGICAL COMPETENCE AS A MORAL COMMITMENT Part of "Chapter 2 - The Ethics of Pelvic Surgery" Competence of the surgeon is a moral commitment to the patient, especially before undertaking a new procedure. It is essential that adequate preparation in the basic and clinical sciences and training in surgical techniques be accomplished before any new surgical procedure is introduced into clinical medicine. An illustrative example of an overenthusiastic rush into a procedure was the sudden popularity of cardiac transplantation in the 1960s. The preparatory laboratory work in cardiac transplantation started in 1905, but the first successful replacement of the heart in a dog took place in 1960. Immune suppression, which is crucial to the procedure, was introduced in 1958, and long-time survival of grafts occurred by 1965. A chimpanzee heart transplanted into a human in 1964 functioned for only a brief time. Christian Barnard reported the first successful human heart transplantation in December 1967 in South Africa. By the end of 1968, 101 human heart transplantations had been performed by 64 surgical groups in 22 countries. Most patients improved briefly and then died of rejection of the transplant or infection. In 2 years, the procedure was largely discredited, and it took more than 10 additional years to reestablish wide acceptance of the operation. This experience illustrates the need to limit difficult and complex procedures to specialized centers that have the resources and adequately trained surgeons to perform them. This is equally true for complex pelvic surgery. The move to the subspecialty boards in gynecologic oncology and reproductive biology and the creation of programs in advanced gynecologic surgery have helped to emphasize the need for specialized training in surgical techniques for pelvic surgeons. The rapid expansion of the use of laparoscopic surgery, which started in a disorganized manner with many accompanying complications, is another example of a rush into routine practice. The use of this procedure is now better organized, with the establishment of guidelines for training, and preceptorships, before a surgeon is qualified. It is likely that standards will be developed for the number and type of operative cases that must be performed each year by a pelvic surgeon in order to maintain competence. This will further ensure the qualifications needed to maintain operative privileges.

MORAL AND LEGAL ISSUES OF ABORTION Part of "Chapter 2 - The Ethics of Pelvic Surgery" The moral controversy in the United States has been bitter and divisive since the Supreme Court's Roe v. Wade decision of 1973, which legalized abortion throughout the country. In keeping with the Puritan and other religious traditions during our first 200 years of development as a nation, abortion had always been considered by some to be morally wrong, but it was covered legally under the common law and generally dealt with leniently until the mid-19th century. During this period, abortion before quickening typically was not considered criminal. The American Medical Association was founded in 1847 and became the framework for an organized effort to change public policy on abortion. A physician-led crusade against abortion started in the 1850s and continued well into the 20th century with the objective of making abortion a crime. The expressed objectives were not only to establish a moral standard, but also to protect women from being harmed by the procedure and to drive “abortionists,” who were usually nonphysicians, out of business. This set the standard for state laws until the Supreme Court action in 1973, which established a national standard. Abortion in the pre–Roe v. Wade era usually was performed illegally by poorly trained individuals and often was performed under life- or health-threatening conditions. By the 1950s, fatal illegal abortion had become the number one cause of maternal death in New York City. Some abortions were carried out safely by physicians in hospitals to protect the mother's life from medical or pregnancy complications, but a new ploy was introduced. Psychiatric consultants recommended abortions under the justification of the threat of suicide by the pregnant woman, and this became a popular way to get around the laws and perform an abortion safely in a hospital. However, this latter mechanism for obtaining a safe abortion was demeaning to women and was available mainly to private patients and the wealthy. Because many countries other than the United States had more liberal abortion laws, the wealthy could also simply travel abroad to seek abortions. After World War II, attitudes toward abortion softened in the United States. In 1959, the American Law Institute, reflecting the popular mood, introduced model legislation to allow abortion for so-called hard reasons: if the woman's life or health is endangered, if the infant will have a severe birth defect, or if the pregnancy was the result of rape or incest. This was not enacted until some states, such as New York and California, liberalized their laws in the early 1970s. In the 1960s, there was a widespread and severe rubella epidemic, during which many women of all walks of life sought and received abortions to avoid having children with severe congenital anomalies. These abortions were performed using the deception of a psychiatric indication under threat of suicide (noted in the preceding). In many instances, however, the demand for abortion by pregnant women with rubella was so great and widespread that the laws were simply ignored. Also, amniocentesis and prenatal diagnosis were just beginning to be used, and this new technology further changed public attitudes. The Roe v. Wade decision was more liberal than the American Law Institute's model in that it made abortion a privacy right and, at least in the first trimester of pregnancy, allowed for abortion without restrictions. In the second trimester, restrictions on where and how abortions were performed could be set by law if necessary to protect the woman's health. In the third trimester, the state can have a legitimate interest in protecting fetal life and can restrict abortion, but not if it adversely affects the pregnant woman's life or health. Some have claimed that the animus and intensity of the antiabortion movement is based on the too-liberal nature of the law. Since the time of the Supreme Court decision, much has been written about both the validity of the logic of the Court's majority legal opinion and the moral arguments for and against abortion itself. The moral arguments are deceptively simple in the extremes. One argument is based on the moral status of the fetus, and the other argument is based on the rights of the pregnant woman versus those of the fetus when their interests conflict. In the first argument, abortion is morally wrong because the fetus is an innocent human life; it is wrong to harm the fetus from conception on, or deprive it of life, just as it is wrong to harm any human living creature. The counter argument is that abortion is not morally wrong because the developing life in the uterus—a zygote (fertilized egg), preembryo, embryo, or the early fetus—although technically living and human, is still developing from a microscopic single cell into a complex organism. At no stage before viability does it constitute the status of a freely living human person, and at these early stages of development it does not require the absolute respect for life that is afforded the late fetus or a born child.

In the second argument, abortion is morally right because a woman should not be forced to use her body to bear a child against her will. The universal moral justification would be that no one should be forced to use his or her body for the benefit of someone else. The counter argument is that abortion is not morally right because the woman has an obligation to the fetus (even one resulting from rape) and should not harm or destroy it, regardless of her preferences or needs about the use of her body. One can embellish or modify these two arguments, but they constitute the core of the rationales generally offered for and against abortion. One can even see the possibilities of accommodation between the two positions based on the gestational age of the pregnancy (early), the specification of certain justifiable circumstances (e.g., rape and incest), or the use of the American Law Institute's model code mentioned in the preceding. In fact, there has been little willingness to give any ground on either side of the argument. We believe both sides have merit depending on one's beliefs, religious or otherwise, about the moral status of the fetus and about a woman's rights. There is no single morally correct answer for everyone in our diverse society, but even this position of seeing merit in the opposing arguments is attacked by extremists on both sides. The major challenge is not to convince one or the other side that they are wrong, but rather to try to convince the two entrenched opposing sides that they have to learn to live together in a democratic society without the intense rhetoric and deadly violence that has occurred. It is important for the gynecologist to determine how he or she feels about these issues and to deal fairly and honestly with patients who ask for advice about abortion or request abortion services. It is generally accepted that physicians should be free to follow their own beliefs and not be forced to engage in practices that are in their own minds morally wrong, but it is difficult as a practicing gynecologist to stay completely neutral on this issue. In the year 2000, the Food and Drug Administration approved the use of the oral abortion drug RU 486 (mifepristone), which has done little to change the dynamic of the controversy.

TERMINAL ILLNESS, ADVANCE DIRECTIVES, AND EUTHANASIA Part of "Chapter 2 - The Ethics of Pelvic Surgery" The moral agenda on issues regarding death and dying has been set largely by the courts as doctors, patients, patients' families, and hospital administrators have sought to have the intractable controversies that arise among them resolved before a judge. Problems with managing terminal illness and persistent coma typically stem from the introduction of technology, such as the respirator, that can keep the human body functioning biologically and either delay or prolong dying without the prospect of a return to even a diminished level of normalcy. The problem with treatment of terminal cancer is often the willingness of the surgeon or patient to try extreme courses of chemotherapy and surgery beyond reasonable bounds of hope for even limited prolongation of a meaningful life. When patients, their families, and their health care teams all agree on one of these courses of action, there is usually no overt problem, except the question of the prudent use of scarce resources. When there is agreement among all parties, there is seldom a need to go to court, and courts discourage bringing these cases before them. When there is disagreement and the patient, the patient's family, or both want therapy and life support stopped and the doctors want it continued, or when the family and patient insist on so-called futile or nonindicated therapy or life support and the doctors believe otherwise, the stage is set for conflict that often ends up in court. The courts have provided guidance in each of these types of cases. The cases followed one another over the years and influenced public attitudes. Although not every case and its outcome are comparable, the general consensus is that the patient's autonomous choices are binding and that the ultimate goal of medicine is not to preserve life at any cost, but to relieve pain and suffering in keeping with the patient's wishes. Karen Ann Quinlan was a 21-year-old woman who had been in a prolonged coma after an overdose of alcohol and drugs at a birthday party. She had been put on a respirator when she developed pneumonia, and it was expected she would die if the respirator were withdrawn. When the likelihood of recovery disappeared, her family requested removal of the respirator. The hospital went to court because of the uncertainty about liability. Although the lower courts denied the parents' request, the New Jersey Supreme Court overturned these decisions and allowed removal of the respirator. Ironically, Karen lived another 10 years in the persistent vegetative state before dying. It was in this decision that the court suggested the use of a hospital ethics committee to help in prognosis. This case, as well as the subsequent need for ethics committees for treatment decisions in the newborn nursery, prompted the development of hospital ethics committees nationwide. In the case of Brophy v. New England Sinai Hospital, the patient was a 49-year-old firefighter from Boston who was in a persistent vegetative state from which he would not recover because of a massive brain hemorrhage. He had previously indicated to his family that he would not want to be kept alive under such circumstances, and his family requested the withdrawal of his feeding tube. There was much public discussion and debate in the press, including objections by religious leaders, but the Massachusetts Supreme Court in 1986 granted the family's wish, and the patient was allowed to die peacefully at home. It took essentially 10 years between these two cases to go from withdrawing a respirator to withdrawing food and fluid, and the moral arguments changed. Allowing someone to die under such circumstances became not only permissible, but also desirable, when this is what they would have wanted. Maintaining life without cognition was no longer an obligation. In the case of Nancy Cruzan of Missouri, which reached the US Supreme Court and was decided in 1990, the issue was similar to the Brophy case. Nancy Cruzan was in a prolonged coma, and her parents wanted to withdraw food and fluid. There was no record of her prior stated wishes, and Missouri has a state law forbidding withdrawal without such evidence. The US Supreme Court found the Missouri statute constitutional; therefore, the parents did not have immediate relief. In the course of handing down this decision, the US Supreme Court established the right of competent patients to refuse any kind of treatment, even life-sustaining treatment, and that no distinction was necessary between artificial feeding and other forms of therapy. Finally, evidence was found of Nancy Cruzan's prior wishes that she would not want to be kept alive in a permanent coma, and she was allowed to die after treatment was withdrawn. In a diametrically opposite type of case, Helen Wanglie was an 86-year-old woman in a persistent vegetative state. Her family wanted everything done to maintain her, whereas the doctors objected because there was no hope for recovery and the cost was enormous. The case went to court. In this case, the family maintained that this is what the patient would want because she felt life should be preserved at

all costs. The court found for the family and treatment continued. This is nonetheless consistent with most of the other court decisions that make the patient's prior wishes a determining factor in the determination whether or not to continue therapy. Many of these cases did not deal directly with the issue of withholding or withdrawing therapy from patients with a terminal illness, but they provided the moral and legal environment in which such requests can more readily be honored by the physician. There had been sufficient consensus on this issue that in 1990, the Patient Self Determination Act was passed by Congress, requiring hospitals to promote the use of advance directives by their patients to qualify for federal funds for Medicare and Medicaid. Practically every state in the country has enacted laws that allow health care proxies, durable power of attorney, living wills, or advance directives for withholding or withdrawing care and protecting the rights of the terminally ill. Pelvic surgeons should familiarize themselves with applicable laws and discuss these issues with their patients and their families, preferably before such decisions are actually needed for implementation. The advanced directive provides physicians with essential guidance on the need for do-not-resuscitate (DNR) orders and the type of care and treatment to be provided for patients with terminal or fatal diseases when discharge from the hospital or recovery is not expected. A problem arises when a patient with DNR orders must go to surgery and either the anesthetist or surgeon objects to having restrictions on resuscitation during a surgical procedure. Often the surgery is for a problem incidental to the long-term illness. Institutional policy should be developed for such cases; the tendency has been to allow patients to keep their DNR orders and still be allowed surgery. Although there is widespread acceptance of foregoing therapy in terminal illness or a permanent coma, also known as passive euthanasia, there is still dissension about whether active euthanasia or physician-assisted suicide should be allowed or encouraged. The argument has been advanced that there is no real moral difference between passive and active euthanasia because the objective is to relieve pain or suffering by death in both cases. This argument misses the emotional significance for physicians and patients of the conceptual difference between allowing death and killing. It also misses the concerns about possible abuse, “brutalization” of the physician, and an erosion of trust between the public and the medical profession. Physician-assisted suicide is a halfway measure that at least ensures patient participation. There is much public interest in active euthanasia and physician-assisted suicide, and one or the other issue has been on several state ballots for popular vote. So far, these have all been defeated with close margins, except for the passage of a law in Oregon permitting physician-assisted suicide under restricted conditions. In the first year of the Oregon law in 1998, 15 patients availed themselves of assisted suicide, most with terminal cancer. Unless controls are stringent, there are concerns that patients with undetected depression might request assisted suicide or that patients might seek this solely for economic reasons. Holland has had many years of experience dealing with a permissive policy on euthanasia, and based on interpretations of this experience, arguments have been advanced to support or condemn the practice, depending on the observer's bias. In the year 2000, the Dutch Parliament finally legalized Euthanasia and Assisted Suicide. Actually the first legalization of euthanasia occurred in the Northern Territory of Australia in 1996. As a profession, we do a poor job in providing adequate medication for pain relief, especially with cancer. The Agency for Health Care Policy and Research of the Department of Health and Human Services issued new practice guidelines in 1994 to improve performance in this area. It is possible that more attention to pain relief and wider availability of hospice care may relieve some of the pressure for euthanasia. Although many physicians and their medical societies are against changing the laws about euthanasia at this time, the public may be moving in this direction faster than the profession. The shift in attitude and more lenient enforcement of laws on use of morphinelike drugs allow physicians more latitude. Physicians have reported a greater willingness to use doses of drugs for pain relief that may in the end hasten the patient's death.

Chapter 3

Psychological Aspects of Pelvic Surgery Betty Ruth Speir

PSYCHOLOGICAL ASPECTS OF PELVIC SURGERY ANKH COMMUNICATION PSYCHOLOGICAL PREPARATION FOR SURGERY COMMON EMOTIONAL RESPONSES TO SURGERY PSYCHODYNAMICS SPECIFIC TO DIAGNOSIS AND SURGERY PSYCHOSEXUAL REHABILITATION SPECIAL CASES DEALING WITH DEATH FUTURE TRENDS CONCLUSION

PSYCHOLOGICAL ASPECTS OF PELVIC SURGERY Part of "Chapter 3 - Psychological Aspects of Pelvic Surgery" The technological revolution has presented today's surgeon with a vast array of sophisticated equipment. In capable hands, miraculous surgical feats can be performed on a woman's body, but at what cost to her psychological self? The removal or reconstruction of diseased or dysfunctional anatomy sets off a chain reaction of parallel events in a woman's psyche. For a surgeon, the gynecologic operation is an ordinary event of usually simple dimension. For the patient, each procedure is a unique experience. Her sense of well-being and health are threatened. She will definitely lose control over her body for some indefinite period of time. She may perceive the planned procedure as temporarily or even permanently affecting her sexual identity. As even complicated procedures become routine, the surgeon risks losing perspective about the impact of surgery on the life of the individual woman. The patient who experiences ablative genital (or breast) surgery is strongly influenced by her emotions. These vary in degree but are usually cumulative. As the patient passes through the presurgical, surgical, and postsurgical experiences, she is stressed and may be inundated beyond her capacity to compensate. If help is not available to facilitate emotional healing and rehabilitation, permanent psychological damage might result. The majority of women do heal and take up their lives, raise their children, work at their jobs, and relate well to their husbands or lovers. For them, the healing interval is relatively quick, and the stress is modest. Do not underestimate what can be learned from this segment of psychologically healthy women patients. In the 1940s, Abraham Maslow studied people with exceptional mental health to develop his hierarchy of needs theory. He discovered that these people's potential had never been weakened by negative thinking, destructive outlook, or destructive self-image. “The rest of us,” he declared, “fixate at a lower level because someone or thing has implanted notions of limitations.” Once you become cognizant of the success statistics in your own patient population, you may discover why most women recover after gynecologic surgery to zestfully re-embrace life, whereas others begin to slowly turn away from its possibilities. It will never be enough for a surgeon to be a trained mechanic, able only to diagnose and repair. He or she must also be prepared to predict, recognize, and begin treatment of the psychological consequences of gynecologic disorders. This chapter is designed to help surgeons and other physicians better understand the female perspective and to use that knowledge to facilitate multidimensional healing.

ANKH Part of "Chapter 3 - Psychological Aspects of Pelvic Surgery" The ansate cross, Egyptian emblem of generation, symbol for life and soul and eternity, is a mark universally designated to depict femaleness (Fig. 3.1). By using this hieroglyph, modern scientists continue an ancient tradition of respect and reverence for women. Within the inner sanctum of every woman's mind lie symbols that are powerful and personal. These feminine identity markers are guarded and guided by her instincts, and no surgeon can successfully maneuver in this labyrinth of psychological and psychosexual emotion and cause no harm unless the female patient acts as a guide.

FIGURE 3.1. The ansate cross, or ankh amulet, is an ancient Egyptian symbol of life and sexuality.

View Figure

Egyptian writings from 2,000 years ago in the Kahan Papyrus depict the uterus as having an important and powerful effect on mental life, and current research tends to agree with the ancient scribes. The uterus has great symbolic value for many modern women, too. Beginning with the rite of passage called menstruation, a strong invisible bond is formed between a woman and her body. Her biological clock has been set. For the next 40 years or so, she will be reminded each month that she is a woman, and menstruation is regarded by many as palpable proof of their femininity. Others believe that menstruation is part of a natural cleansing cycle that purges the body of poisons that accumulate during the month. They know from experience that the premenstrual symptoms of edema, bloating, headaches, and emotional tension will be washed away in the tide of their monthly flow. For some, the rhythm of the menstrual cycle is used as a way to time and order their lives. Like the phases of the moon, this cycle bestows a sense of routine, regularity, and predictability that has emotional significance. With the onset of menstruation, a woman is forever changed. Surgical removal of any of the reproductive organs will change her again, but how? The fact that many medical conditions are affected by a woman's menstrual cycle is well documented. Medical suppression of ovulation is commonly used as a way to evaluate and treat such chronic conditions as migraine headache, epilepsy, asthma, rheumatoid arthritis, irritable bowel syndrome, and diabetes. Current research is trying to determine if removal of malignant breast tumors during specific times of the menstrual cycle might affect long-term survival rates. Other scientists are investigating the circulating levels of sex hormones to determine whether or not the fluctuations produce a direct effect on the immune system and consequently, the disease processes. Hormones secreted during a woman's menstrual cycle certainly affect her mood. Could the decline of these hormones during the natural aging process contribute

to dementia? The future is alive with fascinating possibilities of eventual elucidation, but today's woman is backlit by the present. Her concerns about her body and the way it will work after surgery are immediate and quite often, heart-wrenching. A woman about to undergo a hysterectomy might wonder if her lover will be able to detect the absence of her uterus? After the surgery, will she be thought of as less of a woman? Will her partner abandon her for someone who is still complete, either in the sense of being able to offer the possibility of a child or complete in the sense of having experienced no unnatural surgical transformations? Will orgasm be as pleasurable for her after the surgery? For many women, the uterus has symbolic significance as a sexual organ. The uterus contracts during orgasm. Some women perceive this as most pleasurable. If the patient believes the uterus is essential to sexual response, then, in fact, it often becomes so, and women with this mindset may become sexually dysfunctional when it is removed. Once the procedure is accomplished, will she still look and sound like a woman, or will she become noticeably more masculine? For others, the uterus is closely tied to feelings of attractiveness and sexual desirability. To a few women, removal of the uterus or ovaries or both constitutes a desexing, a permanent destruction of female identity and function. Sadly, certain members of the medical community as well as the feminist community perpetuate this notion and increase the attendant fear when they refer to women who must have their ovaries removed as castrates. Some women become distressed when they learn they must deal with the certainty of absolute sterility. For those who choose motherhood, the personality, uterus, ovaries, breasts, and vagina work in harmony to attract a necessary mate and get down to the business of creating new life. These organs become vital coconspirators in the sexual and reproductive aspects of a woman's life. Impending loss of these physical structures because of disease or dysfunction sometimes creates deep angst that must be resolved before surgery is attempted. Many women who have all the children they want and who do not wish to get pregnant again are still sometimes disturbed by the finality of the decision. The gynecologist should assure these maternal women that, although they will no longer be able to conceive a child, the powerful urge to create will never leave them. In time, they will learn to direct this primal energy into other areas of their lives and be immensely satisfied with the results. Today's modern woman, for the most part, is an avid information seeker. She surfs the net, buys the latest books, reads magazine articles, and conducts in-depth interviews with peers who have experienced similar gynecologic problems. A certain proportion of the harvested material is useful to her and perhaps even illuminating for the physician, but unfortunately, some of the sources are inherently flawed, prejudicial, illogical, or without scientific basis in medical reality. It is the physician's responsibility to separate the grains of truth from the chaff. All the information she has gathered represents her attempt to prepare herself psychologically for the ordeal ahead, and she must never be condemned or made to feel small for trying to protect herself. For the gynecologist, it is not necessary to change a patient's basic attitude or feelings. It is, however, vital to acknowledge them. The right information, reassurance, and support usually quickly modify many negative factors and lead to a healthier attitude and understanding of the surgical process. It is crucial that the patient be allowed to vent her anxiety and speak of her fears.

COMMUNICATION Part of "Chapter 3 - Psychological Aspects of Pelvic Surgery" Patients in General Diplomats the world over understand the significance of effective communication. Indeed, the possibility of world peace depends on their ability to interact efficiently. A female patient's psychological bearing before, during, and after the gynecologic procedure can depend on the communication techniques employed by her physician. Once you establish rapport, begin to speak of the real and present danger to her health and then, elaborate on your plans to protect her. Whenever possible, provide pleasant surroundings where you and your patient can comfortably hold a private conversation. Push all other thoughts out of your mind. These few minutes belong exclusively to her, and the quality of the time you spend actively communicating will pay healthy dividends for you both. Whether explaining the simplicity of a needle-directed breast biopsy or the intricacies of hysterectomy, you must be able to highlight the technical details of the procedure and its postsurgical realities in a nonthreatening but utterly truthful manner. Begin the presentation using simple but thorough explanations and examples of the procedures. Your patient will let you know how much information she can or wishes to process by the questions she asks. Besides technical competence, the patient needs two assurances from her surgeon during the presurgical consultation. One, is to know that she has your complete attention, that her problem is being taken seriously, and two, that you are qualified to competently help her cope with all of the ramifications of this new experience. Dealing with the patient's feelings usually is not difficult if the surgeon accepts the viewpoint that many of the patient's emotions are part of the gynecologic situation. To the physician, the presurgical tension is predictable, familiar, transient, and simply comes with the territory, but to each new patient, it represents a dramatic and life-altering experience. To interpret the patient's true feelings, simply listen to her. Listening well is both an art and a skill, and the surgeon who cares about the patient's physical and mental health as well as her quick recovery works diligently to hone a sharp edge on this valuable tool. Listen and you will hear the woman give a name to her most profound doubts and fears. Repeat her words so that you are absolutely certain you understand what she said, and she is absolutely assured that you are listening to her. Begin the communication process by finding out what the patient perceives will be done to her body and why the procedure is necessary. Find out what she believes the consequences of the surgery will be. How does she think the surgery will impact her life? As she discusses the implications of her decision, her knowledge, fears, and biases emerge. At this point, you should be able to supplement the patient's perspective with appropriate explanations about anatomy, physiology, and pathology. After this, it is time to describe in detail the usual preoperative, operative, and postoperative routines. Address the patient's questions and fears in as many ways as it takes for her to become confident that she understands what is about to happen to her. Carefully explain what you are going to do to help her. If she wants to know, or needs to know, describe the common physical sensations, bandages, incisions, catheters, tubing, and medications that are associated with her particular procedure. Define the patient's role in her own convalescence and recovery. Give her a general timetable for how long she will feel discomfort, have to use pain medication, when she will be ambulatory, and finally, when she will be discharged from the hospital or outpatient center. Relate the most common complications that might occur as the result of the pelvic surgery. Injuries that might affect the quality of her life, even temporarily, should never come as a postsurgical surprise, nor should they be glossed over during the signing of the consent forms. Iatrogenic injury remains the most common cause of lower urinary tract trauma. An understanding of the prevention, recognition, and treatment of urologic complications is important for every surgeon performing major pelvic surgery. Injury to a woman's genitourinary system may take 10 to 20 years to develop full-blown symptoms, about the same time line as from multiple childbirths, but it remains a possibility. Physicians, however, disagree on whether or not to tell the patient that urinary incontinence after

hysterectomy because of damage to the pelvic nerves or pelvic supportive structures could be a long-term adverse effect because only 4% of the hysterectomies performed are for relief from symptoms of incontinence. Complications such as the formation of adhesions may occur in as many as 55% to 100% of patients after gynecologic surgery. These adhesions can become a critical issue from a standpoint of reproductive potential, and their presence is also strongly associated with pelvic pain, abnormal bowel function, and small bowel obstruction. The mentally competent patient has a moral, legal, and ethical right to make an intelligent informed decision and can only do it if she is privy to all of the available facts concerning her situation. It is essential that you involve her in the decision-making processes, because the more committed she is to the proposed treatment, the more involved she will become in her own preparation and rehabilitation. The art of touching, the therapeutic laying on of the hands, is important. Being lonely, frightened, and sick is a reason for the patient to be touched by her physician, especially if she seems particularly overwhelmed by her situation. To hold the patient's hand while talking to her or touch her shoulder is therapeutic. Even a comforting hug is appropriate if it fits the circumstances and the patient reaches out to you. Before fetal monitoring, quality of labor was evaluated by sitting at the patient's bedside with the physician's hand on her abdomen to feel uterine contractions. Often, the presence of the physician and a warm hand on her abdomen made the patient relax, rest, and become calmer during active labor. Cancer patients, especially, on hearing bad news, need immediate human-to-human contact to stay grounded enough to face that terrible moment. The professional boundaries of roles, time, place and space, gifts and services, chaperoned examinations, physical contact, money and formal language, were never intended to be an impermeable membrane separating a doctor's ability to administer human kindness from the patient's need to receive it. A healer's touch can often comfort a distressed patient when words are inadequate. A patient's family is a vital part of her support system, too, and can be a potent ally, or not, to the health care team. If your patient requests that family members be present during her consultation, allow it, but speak directly to her whenever possible. Once the initial presurgical discussions are complete, your patient may need some time to digest all the new information she has received. After assimilation, expect her to contact you for clarification or to ask more questions. You should be available to her. Too often, office personnel believe their role is to shield the physician from patients rather than to facilitate meaningful contact. Multicultural Patients Multicultural patients abound in the United States and their numbers are increasing dramatically. The populations are sometimes situated in dense clusters in specific regions and physicians in these areas are often multilingual or have assistants who are able to act as interpreters. No matter where your practice is located, chances are that at some point during your career as a physician, an individual from another culture will need your help. Mull makes the following suggestions for communicating effectively with a person whose language you do not speak and whose culture is foreign to you: Make sure that your office staff are courteous and respectful. Show the genuine concern you feel. Be friendly and helpful to build rapport and develop a repertoire of knowledge. Familiarize yourself with the general principles of their traditional medicine. Learn a few key phrases from their language, and use these for initial greeting and during examinations. Include, at least in discussion, any family members your patient considers influential. Always ask what they have done to treat themselves. Have they consulted an influential family member? A healer? Have they used home or herbal remedies? Physicians should be aware of common themes that exist in cross-cultural medicine, including the following. Fear of blood loss Fear of cold Tradition of male dominance Conservatism in sexual matters relative to teenage girls Poorly developed concept of preventive medicine Intolerance of side effects from medication

Expectation of expeditious wellness Reluctance to discuss emotions with people who are not family members Diaz-Gilbert cautions health care professionals to check their prejudices at the door and not to assume that a non–English speaking person is uneducated. The following are some guidelines for effective communication with multicultural patients. Allot extra time for the multicultural patient. Address every patient initially in English. Ask if the patient carries a bilingual dictionary. Gesture or write down simple words or phrases. Use visual cues, such as insurance forms, calendars, medication bottles, or anatomic sketches. If necessary, draw a picture to convey what you mean. If the patient needs to perform a specific task, such as disrobing, carefully pantomime each step. Remain aware that direct eye contact, certain hand or finger gestures, and physical touch are offensive, disrespectful, or can be construed as sexually suggestive in certain cultures. When a cultural language barrier exists, reliance on body language becomes crucially important. Work at learning to read it. These patients experience the same emotional responses to surgery and need the same education and reassurances. Having to use the services of a foreign doctor in an alien land makes the surgical process even more frightening for them.

PSYCHOLOGICAL PREPARATION FOR SURGERY Part of "Chapter 3 - Psychological Aspects of Pelvic Surgery" Massler and Devansan said there is an emotional response to any physical assault on the body. The magnitude of the response is expected to be proportional to the degree of emotional investment one has in the part of the body under siege. Among women, anatomic entities most vulnerable to this emotional reaction are the face, hair, breasts, genitalia, and abdominal wall. After Caesarian birth, hysterectomy is the most frequently performed major surgical procedure done on reproductive-aged women. Over 600,000 women a year have a hysterectomy in the United States. The majority recover quickly from the procedure and enjoy the new freedom that comes to an individual when a chronic health problem has been solved. Indeed, studies show that hysterectomy is not associated with severe depression in older women and bilateral oophorectomy is associated with only slight increases in depression scores. Your patients, however, are not always older, wiser women and they do not always recover quickly. If you know the signs to look for, you will be able to help that patient who is having a harder time adjusting to pelvic surgery, especially if it was a radical, life-threatening, or unexpected event. In his book, Matters of Life and Death, Daniel Bruns writes, “Even the strongest person can be shaken by the horrors of some medical cures. Beyond this, life anxiety is even more common in persons with pre-existing emotional difficulties or characterological disorders. These persons may go through life like eggshells, intact and functioning, but with psychological fragility. When faced with an extreme life stressor, such a person may simply shatter.” How will you, as the physician, recognize the vulnerable? Roeske researched 13 factors related to poor prognosis for excellent mental health after hysterectomy. These factors begin to define for you the patient who might react negatively to genital surgical stress. Gender identity Previous adverse reactions to stress Previous depressive episodes Family history of mental illness History of multiple physical complaints, especially lower back pain Numerous hospitalizations or surgeries Age less than 35 years at time of hysterectomy Desire for a child or more children Fear of loss of libido Significant other's negative attitude toward procedure Marital dissatisfaction or instability Cultural or religious disapproval Lack of vocation or hobbies Barnes and Tinkham's research also indicates that patients tend to react to current stress in much the same way as they reacted to past crises and personal losses. Well-established patterns of behavior repeat themselves. By taking a patient's history, the surgeon can be forewarned about which patients are likely to have the most difficult time handling the emotional aspects of gynecologic surgery. Equipped with this information, the surgeon can prepare to offer extra support in the form of reassurance, educational information, and, if indicated or requested, the names of psychotherapists who are trained to deal with women's health issues.

COMMON EMOTIONAL RESPONSES TO SURGERY Part of "Chapter 3 - Psychological Aspects of Pelvic Surgery" Insecurity Feelings of insecurity and vulnerability are often a realistic appraisal of the patient's situation. Giving up control of one's body, even temporarily, is uncomfortable for all of us, but it is terrifying for people who feel generally insecure. One of the most common defense mechanisms against feelings of insecurity is to institute rigid controls over all aspects of life. These patients had no control over getting sick. They will have little control, if any, during the surgical procedures. Postsurgical setting will be a hospital room where the staff tells the patient when to awake, take medicine, eat, bathe, walk, have visitors, and get blood drawn. The health care workers will probe such personal matters as urination, defecation, and passage of flatus. Anxiety, anger, and feelings of being assaulted combine with insecurity to produce an unhappy, fearful, and sometimes raging patient. These feelings are greatly diminished when the patient believes the surgery will improve the quality of her life. There will eventually be relief of pain, removal of cancer, an end to heavy bleeding, restoration of fertility, or some other positive result. She will be better than she was before the surgery. The transition to becoming this healthier person is made easier when she trusts and believes in her physician. Anxiety or Fear Anxiety or fear associated with surgery is essentially universal. Most common is fear of the unknown or of what the patient imagines she will be forced to endure during hospitalization. Factual information about the surgical and recovery process and competent care by a compassionate hospital staff help to diminish this fear. Surgeons are responsible for the behavior of the hospital staff toward their particular patients. When a patient complains of ill treatment by any of the health care providers, the physician should personally deal with the situation, because this will decrease the probability of a future recurrence of the offensive or thoughtless behavior. Patients also fear the loss of economic competence. A woman who has worked hard every day for many years is afraid she will be partially or totally disabled for some variable length of time. The loss of this familiar role, even temporarily, jeopardizes her sense of usefulness. Whether she is a major or minor financial player, the contribution she makes to her family's economic stability is important to her. Many women create an identity for themselves based on their job description. Any threat to the survival of the career also menaces a woman's self-esteem. Fear of anesthesia is often a thinly disguised fear of dying as well as a fear of loss of control. It may be appropriate to confront the fear of dying directly so that the patient has the opportunity to express why she is afraid. Is the fear general or specific? Did a close relative suffer from the same affliction and die during or shortly after surgery? Does the patient have a strong intuitive feeling that something will go wrong? If so, ask her what you can do to modify the surgical situation. Determine whether the scheduled date of surgery or the particular hospital is significant. The fact that you consider her feelings an important issue and a normal part of the gynecologic disease process may be enough to calm her fears. Regression and Dependency In most people who are ill or who undergo surgery, regression to a more dependent state is fairly common. Dealing with a woman who is no longer self-sufficient or emotionally stable is difficult for the patient's family and friends. These members of her supporting cast are accustomed to her presurgical roles as wage earner, wife, mother, friend, cook, advisor, shopper, housekeeper, taxi driver, entertainer, and more. When she becomes ill and can no longer function to make their lives easier, family members and friends often become frustrated and angry. They may apply overt or subtle pressure to try to force the woman to exert herself and fulfill her usual roles. A change in roles is difficult, but often the illness teaches family and friends why this particular woman is valuable to them. Those who temporarily assume her normal duties or help her cope with the surgical experience will have the opportunity to learn that taking care of another human being somehow makes us stronger than we ever were before. When the disease and prospect of surgery are new, all these factors, as well as a feeling of non-health, contribute to an emotional fragility that yields extremely labile emotions, including feelings of sadness,

despondency, tearfulness, and irritability. The usual defense mechanisms are often temporarily weakened or destroyed. The woman is vulnerable to attack on all personal and professional fronts. She needs time with people she cares about, and she needs time alone to sort out her thoughts. Grief Grief is a normal, natural reaction to illness or loss of any kind, and is essential to emotional healing. Recognizing the various stages of grief allows the surgeon to help the female patient understand what is happening to her. Denial is the first and most primitive emotional response to loss and can take many forms. The patient may demonstrate denial by not going to the doctor when she finds a lump in her breast or when she notices abnormal bleeding. She may pretend the symptoms do not exist or are a temporary nuisance. Not remembering instructions the physician gave her could be a manifestation of denial. She may forget important facts or deny the seriousness of the problem. Denial allows people to function for a little while in a make-believe world. With this primitive mechanism, they survive emotional stresses they might otherwise be unable to handle. Bargaining with a higher power is the second stage of grief. Patients feel they have carte blanche to bargain when they are experiencing a loss. “Make this bad thing that has happened to me go away, and I swear, I will become a better person.” Guilt can surface before or after a loss. Most guilty feelings are completely inappropriate, in that the guilty act rarely is directly related to the cause of the loss. When sick, many people feel that they are being punished for not being perfect. Explain to your guilt-ridden patients that their feelings are normal for the circumstances. Although guilt can sometimes deliver devastating, incapacitating blows, the good news is that it is usually transitory. Depression comes in varying degrees to most people experiencing grief and is characterized by feelings of helplessness, hopelessness, and worthlessness. Other symptoms include middle-of-the-night insomnia, nightmares, loss of appetite or excessive eating, lethargy, difficulty making decisions, psychosomatic symptoms, and fatigue unexplained by activity. Ask the patient if she has any of these symptoms. Postsurgical depression is common. Depressed patients usually admit that the sad feelings are routine and occur daily. When prolonged, they indicate the patient has been unable to work through the grief process. Something emotional has yet to be resolved. Rage turned inward often manifests as depression. When the patient is able to identify what she is angry about, to ventilate the rage, the depression usually begins to lift. When she takes charge of her life again and makes decisions, even small ones, she begins to feel better, and feelings of helplessness, hopelessness, and worthlessness abate. However, when the depressed patient becomes suicidal, stringent intervention must occur. The suicidal patient presents serious challenges and so is discussed in specific detail later in this chapter. The stage at which the patient ventilates her anger can be difficult for those providing care, but it should be accepted as healthy. The patient may go to extremes, writing letters to the newspaper or speaking of suing her physician. She may complain bitterly about the nursing staff and her hospital bill. Such actions are a form of protest at the stress that has been dealt to her body and mind. In most instances, this behavior means that the depression is lifting and the patient is beginning to move toward the resolution of her grief. The depressed patient should be encouraged to ventilate by talking, establish an enjoyable form of physical exercise, and begin to take charge of her own life. Resolution and integration eventually occur. The stressful experience of loss finally becomes an accepted part of her life. The memory causes sadness and regret, but no longer the devastating immobilization found in the earlier stages of grief. Integration does not mean the experience is forgotten, only that it has less trauma associated with it. After integration, certain stimuli can provoke flashback grief. The painful emotional tapes begin to play again, but the patient learns that the bad time is a rerun and will not last long.

The stages of grief do not always occur in order. The patient may feel fragments of several of them at the same time. If a female patient's behavior seems bizarre, excessive, or out of the realm of what would usually be anticipated during her particular surgical experience, look for the role grief might be playing in her life.

PSYCHODYNAMICS SPECIFIC TO DIAGNOSIS AND SURGERY Part of "Chapter 3 - Psychological Aspects of Pelvic Surgery" Patient–Physician Bonding Neither scalpel nor laser can divide the psyche and soma. As technical skills have improved and multiplied, comprehensive care of the gynecologic patient has declined. Increased emphasis on scientific and procedural care usually means less time in the consultation room and more time in the examining or procedure rooms. This is unfortunate for all concerned, because it takes time to explore the concepts, fears, and psychological well-being of the individual patient both before and after surgery. It takes time to at least scan the books and magazines women read, to search the same Internet sites, to listen to the voices of their advocates, and essentially, evaluate, critique, and learn from their sources of information. It is important to make the effort. Otherwise, you run the risk of appearing out of touch, or worse still, arrogant and condescending. Women in general have a sixth sense about these attitudes, and women about to undergo surgery or those recovering from surgical trauma to their bodies are hypersensitive to all manner of psychological stimuli. The medicolegal climate has also potentiated perioperative anxiety. When you inform your patient, as she prepares for surgery, that she can bleed to death, be subjected to blood transfusions, have an adverse reaction to anesthesia, or sustain bowel or urinary tract injuries, you augment her innate fear of surgery. With strong emphasis from the women's rights groups on the patient to seek and maintain a controlling role in her life, over her destiny, and over the surgeon, it is more necessary than ever before that the pelvic surgeon take the time to explore the patient's psyche in the preoperative and postoperative periods to avoid undesirable psychological sequelae. It is a good thing to have personable, competent health care workers in your office who assist in preparing your patient for surgery. Anatomic charts have become works of art and they, as well as video movie tapes and other teaching aids, are useful in explaining the technical details; however, despite all the educational tools and patient assistance, the most important person remains you, the doctor. Unless you sit and answer questions on a one-to-one basis, you are neglecting your responsibility to her. Much of the time, the questions will deal with information the patient has gleaned from literature, popular talk shows, and the Internet. Some of the opinions she reads or hears will frighten her or make her suspicious, and a few patients may initially come to your office thinking of you as a potential enemy. A staunch feminist may express the belief that you are just another insistently prosurgical doctor out to highjack her womb and add it to your trophy collection. Popular literature today often stresses sexism, ageism, and greed on the part of doctors. The Silent Passage and Our Bodies, Ourselves were among the first widely read books on which patients depended for their gynecologic information. In these, they read that “for well over a century in the United States, women's uteri and ovaries have been subject to routine medical abuse,” and “one should not be railroaded into hysterectomy nor onto hormones.” Hysterectomy is described as “devastating” surgery, and for some women, it certainly can be. These books found a wide audience and led to the publication of other books, which took an even more radical approach, all in the name of protecting women from castrating medical experts who might use their position of authority to hurt, not help, them. The Ultimate Rape: What Every Woman Should Know About Hysterectomies and Ovarian Removal was inspired after the author underwent a hysterectomy. She suffered extreme physical and emotional trauma following the surgery, but when she complained to her physicians, they advised her to go see a psychiatrist, because all her symptoms were in her head. The book's title is evidence of the rage she felt at their pronouncement. Now her voice is joined by others who believe every woman has the right to be thoroughly informed about procedures and consequences before consenting to gynecologic modifications. And certainly, a woman should. In No More Hysterectomies, touted as the first living textbook on the web, the reader learns how the male-dominated medical profession and the insurance industry have sanctioned millions of unwarranted hysterectomies. One testimonial to the ideas presented in the book describes the current medical environment as a “woman's hormonal holocaust.”

The enlightening news is that interest generated by these sources and their legions of followers has had a positive and direct effect on women's health research. Global studies are numerous and are concentrating on traditional as well as alternative methods of treatment for menopause, hysterectomy, hormone replacement therapy, cancer, endometriosis, fibroids, and dementia. For the first time in the history of medical science, ethnic research is being conducted on a large scale to determine how women in various cultures and with variances in their physiology react to menopause, gynecologic surgery, hormone therapy, and sexual function. Future generations of women will reap the benefits from this research, but the overwhelming aura that prevails in today's gynecologic patient is one of confusion. After reading just a sampling of the lay literature, some women feel that surgical removal of their female organs and commencement of hormonal therapy constitutes an unnatural, chemotherapy-like assault on their physical bodies. It is the task of the physician to admit into evidence the medical facts necessary to correct any gross misconceptions that could affect patient care. Sometimes, it may seem as if the patient, armed with advice about natural remedies for her severe pelvic pain, heavy bleeding, or hot flashes, wants to drag you with her back into the Dark Ages. Be patient and also prepared, if necessary, to explain the medically sound benefits of life lived outside the cave. Be compassionate. No matter how routine the job becomes, compassion is a vital requisite to becoming an exceptional communicator and healer. Empathy often follows experience, and those times when you are able to make a noticeably positive difference in your patient's life are inspirational. To try the one new thing that might help many patients in the future, it is necessary to earn the trust of a single patient in the present. The days are gone when a doctor was considered omnipotent, when he, and rarely, she, received a hock of ham for the birth of a child, or had to tell a woman that she would have to live with the eventual hump on her back because it was a natural process of aging. Patients know about osteoporosis and heart disease and reproductive technology and brain neurotransmitters. The media have turned every living room into a medical school. Some patients present videos and clippings detailing current research and experimental treatments relative to their specific diagnoses. They know a little, and they want to know more. Many patients want to participate in their health care and absolutely should be encouraged to do so. Unlike the doctors of old, who, for the most part, had to contend with an uneducated populace, the modern physician must form a partnership with the modern patient. Mutual responsibility, respect, and trust eventually strengthen this bond. The cornerstone of the initial work is truth. Use good judgment about when to tell all the facts, particularly those that point to a devastating diagnosis, but never lie. In 1961, 90% of physicians surveyed in a single large urban hospital stated that they withheld the diagnosis of cancer from their patients. By 1977, the position had been totally reversed, with 97% reporting that they did reveal the true diagnosis of cancer. Doctors, however, are not the only members of the team with ethical considerations. Patients also have the responsibility to tell the physician the truth relative to their symptoms, medications, allergies, past medical histories, and to relate any significant traumas or family history that could bear on their current situations. Question your patient specifically about stressful life events. Did she respond to these in a positive or negative way? Of all inquiries, this is the most important indicator of how the patient will respond to any current stress. Once the physician knows the answer, psychological preparation for diagnosis or surgery can begin in earnest. Researchers in the United Kingdom have compiled data from multiple trial studies confirming that psychological preparation for surgery is effective. The general hypothesis was that communication and counseling are important determinants of numerous factors, including the following. Accuracy of the diagnosis Effectiveness of disease management Disease or problem prevention Patient satisfaction Adherence to treatment Psychological well-being Patient understanding of procedures

Professional satisfaction and levels of stress Information about each of these parameters was compiled, and considerable evidence existed to support all the hypotheses. In review, Davis and Johnston reported that psychological preparation is effective in reducing negative effect, pain, medication, length of hospital stay, and in improving behavioral recovery and physiologic functioning. Surgical Whispers The surgeon should make it a point to be with the patient while anesthesia is administered. Knowing that you are there with her from the beginning will help her feel safe. At the end of each surgical procedure, whisper into your patient's ear, “You're going to be well very soon.” You may be surprised to find she needs less pain medication and recovers quicker than patients without the benefit of this positive prophecy, because the mind itself is a powerful force. Youngs and colleagues believe the surgeon or a familiar associate also should be present immediately after surgery to reassure the patient, orient her to her surroundings, and make certain she has adequate pain relief. Even if the patient appears unresponsive during the immediate postoperative phase, a familiar voice and reassuring word can be immensely beneficial. Immediate Postoperative Care Hospital stays are much shorter than in the past, and outpatient hysterectomy is performed in some areas. This has a positive psychological effect on the patient. She knows that she is getting well when she no longer requires needles and is able to ambulate and urinate without assistance. It has long been known that early ambulation significantly reduces morbidity as well as the incidence of phlebitis and pneumonia. As long as intravenous therapy and urethral catheterization are maintained, the patient remains immobile and consequently at higher risk for venous stasis, ileus, and pulmonary complications. The incidence of pulmonary embolus in the postoperative hysterectomy patient has decreased in the past decades with shortened stay and early ambulation. Crisis Intervention It is important to recognize a patient who is overwhelmed by stress. This event can occur during any phase between diagnosis and recuperation. A crisis has been described as an obstacle to important life goals that becomes insurmountable when the individual employs customary methods of problem solving. Kaplan highlighted the following four phases of crisis. Arousal occurs and attempts are made at problem solving. Increased tension leads to distress and disorganization because arousal hinders rather than promotes coping behavior. Insomnia and fatigue frequently result. Internal and external emergency resources are mobilized. Novel methods of coping are tried. A state of progressive deterioration, exhaustion, and decompensation ensues as the problems drag on and on without resolution. Dennerstein and van Hall report that the types of problems dealt with in crisis therapy include loss, change in status or role, interpersonal problems, and problems of choice between two or more alternatives. As an advocate, encourage your patient to communicate her feelings, to understand the problem enough to identify and define it, and then, help her rehearse alternative ways of coping. One of the best concepts to apply during any stressful situation is to give the assurance that this particular moment, no matter how painful, is temporary. The surgical procedure and all the stages of mending that follow will have a beginning, middle, and an end.

PSYCHOSEXUAL REHABILITATION Part of "Chapter 3 - Psychological Aspects of Pelvic Surgery" The goal of psychosexual rehabilitation after gynecologic or breast surgery is to restore sexual function, sexual identity, body image, and self-esteem. Most of the work must be done by the patient herself, but she may need assistance from her doctor and other health care providers because the experience is new to her. She has no gauge to measure what is normal for her particular situation and what is aberrant. You do. She may expect too much too soon from herself or she may head off in the opposite direction and begin to assume the role of invalid, but in most cases, she will be caught somewhere in-between these two extremes. Once she begins to exhibit her normal patterns of relating to others, you will know she has officially begun the process of genuine healing. You will be able to tell when she enters the healing phase because she will become less dependent on you, the nurses, and even her family members. As her strength increases, she will want to resume her usual activities. The inevitable, normal, uncomfortable grief process will commence. Encourage the patient to talk about her feelings rather than repress them and begin to brood, because worry and rumination are forms of repetitive thought that are concomitant and predictive of negative mood. Dreary thoughts fuel a depressed mood and turn it into something ugly and dangerous that has the potential to burn the thinker beyond recognition. The patient has the power within to effect change in herself. Family members and friends should be cautioned, at this point, to allow verbal ventilation. It's a form of healthy discontent that frequently provides the impetus to hurry and lose the sick image and begin to see herself well and strong again. Cosmetics, dress, and grooming are important parts of the rehabilitation process. When a postoperative patient combs her hair, puts on lipstick, and demands her own nightgown instead of hospital garb, she has begun to heal. When a patient feels that the surgery was disfiguring, she needs to compensate by learning new ways to dress or groom. She needs to feel whole and complete and responsible again as quickly as possible. The surgical patient begins to see herself as a sexual person when her sexual identity is validated by her sexual partner, friends, family, and even admiring strangers she passes on the street. The woman who has had a mastectomy or other body-altering surgery needs to know her partner still finds her attractive and desirable. Without this affirmation, she may have a great deal of trouble seeing herself as a sexual being. Some sexual partners cannot accept an incomplete person. This is another potential problem. Some surgical procedures result in loss of vulva, clitoris, or vagina. Radical pelvic surgery can leave a woman with a colostomy or urinary diversion. A severely altered body image concurrent with loss of health and vigor poses a serious threat to a woman's self-esteem. The woman who has lost her sexual identity feels damaged beyond repair. Some complain of continuing pelvic pain without obvious structural cause. Interest in sex vanishes, and the patient may actually leave her sexual partner or force the partner to abandon her. As she terminates her sexual identity, she feels old before her time and begins to draw in the edges of her life. These women need intense psychosexual therapy if they are ever to heal emotionally. Table 3.1 outlines the major factors that occur with psychosexual dysfunction.

TABLE 3.1. Major Factors in Psychosexual Dysfunction

Symptomatic

Interpersonal (discord with significant other)

Organic (disease, malnutrition, malfunction of body organs)

Psychiatric (anxiety, depression, schizophrenia)

Alcohol or drug abuse

Iatrogenic (suggestions, medication, surgery)

Learned

Family (childhood negative sexual associations, experiences)

Religion (imposed prohibitions internalized)

Early unpleasant sexual experiences

Gynecologic disorders (damaged genitalia, loss of breasts, uterus)

Intrapsychic conflict

Failure to develop psychosexually

Restrictive childrearing

Restrictive childrearing

Religious influences

Most of the time, the mate or lover of the woman is good to her. There is genuine concern for her health, hope for a quick recovery, and the willingness to assume many aspects of her role until she is well. Often there is a deepening of affection between the couple as gifts of love and concern are given and received. That special someone is in the waiting room during the surgical ordeal and by the patient's bedside when she awakes. There are flowers and gifts and promises made and kept. There is an abundance of reciprocal love. Adjustment to new roles is relatively smooth, causing new bonds to form and old ones to strengthen. In other cases, the woman's partner becomes a bigger problem than her physical disability. It is possible her significant other constructed a fragile emotional bond with body parts rather than with the actual woman. If she had or has cancer, the partner may irrationally feel that the cancer is contagious. If she is receiving radiation treatment, he may feel that if he resumes sexual relations with her, he, too, might absorb radiation from her body and be burned. The couple may be accustomed to frequent sex and any change in the woman's availability stresses the relationship. The fear of causing pain also has an inhibiting effect. Emotional isolation and loss of nurturing occur in both partners when the woman experiences physical disability. As surgeon to the postoperative patient, you are her first line of psychosexual defense and yours will not be an easy job. Depending on the study cited, sexual dysfunction exists in 23% to 43% of women and 31% of the men surveyed in the general population. One third of the women lacked sexual interest, one fourth were unable to experience orgasm in the menopause, one fifth reported lubrication difficulties, and another one fifth said they did not find sex pleasurable. These figures come from members of the population willing to discuss sexual dysfunction. Many women and their physicians, who sometimes fear they are not qualified to help, are reluctant to speak of personal problems such as libido, arousal, coital pain, or past traumatic sexual events. Much of this reluctance can be overcome if the gynecologist knows what questions to ask when taking a sexual history, preferably during an initial or annual examination prior to any body-altering surgery. Sexual Cycle Primer Davis suggests that the physician ask the following open-ended questions to obtain a sexual history: Are you sexually active? Are you or your partner having any sexual difficulties at this time? Has there been any change in your sexual activity? Have you ever experienced any unwanted or harmful sexual activity? Another good question is: What sort of sexual problems do you have? Even if the patient is initially reluctant to discuss such personal issues, she will have learned that you are willing to discuss them should the need arise. Davis also believes that a physician's confidence in dealing with sexual issues increase when the cycles of sexual response (desire, arousal, plateau, orgasm, and resolution) are learned and factors that affect them (psychological, environmental, and physiologic) understood. Davis, in a sexual and sexual dysfunction tutorial describes the following stages. Desire is the motivation and inclination to be sexual. It is dependent on internal (fantasies) and external sexual cues and also on adequate neuroendocrine functioning. Arousal is characterized by erotic feelings and vaginal lubrication as blood flow increases to the vagina. In addition to feelings of sexual tension, the sexually excited woman may experience tachycardia, rapid breathing, elevated blood pressure, breast engorgement, muscle tension, nipple erection, and other

physical signs of arousal such as a flush. This is the stage where the vagina lengthens, distends, and dilates, and the uterus elevates partially out of the pelvis. During the plateau phase, sexual tension, erotic feelings, and vasocongestion reach maximum intensity. The labia become more swollen and turn dark red, the lower third of the vagina swells and thickens to form the orgasmic platform. The clitoris becomes more swollen and elevated, and the uterus elevates fully out of the pelvis. Eventually, women reach the threshold point of orgasmic inevitability. Orgasm is a myotonic response mediated by the sympathetic nervous system and is experienced as a sudden release of the tension built up during previous phases. Women, unlike men, experience no refractory period but can experience multiple orgasms during a single cycle. They can also experience orgasms before, during, and after intercourse provided they receive enough clitoral stimulation. The last phase is called the resolution phase. Women experience a feeling of relaxation and well-being. The body returns to a resting state. Complete uterine descent, detumescence of the clitoris and orgasmic platform, and decongestion of the vagina and labia take about 5 to 10 minutes. Sexual adjustment is often significantly impaired in women after pelvic exenteration and gracilis myocutaneous vaginal reconstruction. Eighty-four percent of the patients in one of the few studies that exist resumed sexual activity within the first year after surgery. A modified version of the Sexual Adjustment Questionnaire was used and the responses outlined the most common problems patients face after the surgery: self-consciousness about a urostomy or colostomy, being seen in the nude by their partner, vaginal dryness, and vaginal discharge. It is hoped that future modifications in surgical technique, more realistic patient counseling, and aggressive postoperative support will minimize these problems in the future. Less serious matters can cause self-esteem and body image problems, too, if their aftermath includes or leads to bowel incontinence, urinary incontinence, vaginal vault prolapse, and scarring. Bowel incontinence is rarely discussed even with a woman's physician because it is so embarrassing. Whether from obstetric injuries, injury to the anal muscles, infections, or diminished muscle strength from aging, once the cause and severity are determined, treatments can begin that include dietary changes, constipating medications, muscle strengthening exercises, biofeedback techniques, and sometimes, surgical repair of the muscle. Some or all of these remedies help the woman control the discharge of embarrassing gas or stool. It is most important to discuss possible remedies because many women feel there is nothing that can be done for them but the frightening colostomy, when in actual fact, colostomy is a procedure that is rarely required. As many as 50% of all women experience occasional urinary incontinence. In an attempt to lessen the blow to a woman's ego and make the event more socially acceptable, manufacturers hire movie stars to make commercials about the effectiveness of diapers for grown women. Diapers do treat the symptoms and allow for more freedom of movement, but not in an intimate setting. For many years, gynecologists have instructed patients about Kegel exercises to tighten the muscles of the pelvic floor, but this may not be enough to stop the embarrassing leakage of urine. The patient needs to know that there are tests that can determine the exact cause of the problem, and treatment using bladder retaining therapy, medications, and surgery. Urinary incontinence may be more socially acceptable today, but it is never normal, no matter what the woman's age. Both bowel and urinary incontinence can be caused by vaginal vault prolapse and this condition must be ruled out because it drastically affects sexual functioning. The presence of a mass can cause painful intercourse, difficulty accepting penetration, and a great deal of psychological anxiety when the tissue can be seen in the vaginal opening. This condition, if left untreated, only worsens with time, but techniques that correct female organ–supporting defects in the pelvis can restore sexual functioning and with it, a woman's sense of vitality and feminine allure. Patients who talk about their sex lives frequently describe four pleasures associated with sexuality. These universal elements are touching, genital caressing, orgasm, and gratifying a partner. When a patient is recovering from surgery or has experienced surgical loss of coital function, genital caressing as a receiver or giver can be satisfying. Once a woman learns early in life how to be orgasmic, she can often learn to be so again despite major genital loss, including her clitoris. When the ability to experience orgasm by one favorite means is destroyed by disease, the patient can be encouraged to experiment with alternative

methods that do not conflict with her value system. Women who will never experience vaginal intercourse again can discover they are able, with education and imagination, to fulfill their feminine role as givers of pleasure if they choose to do so. When a patient's psychosexual rehabilitation after surgery seems to be impaired and she fails to make steady progress toward resumption of her usual role, with appropriate self-esteem, energy, identity, and ability to handle stress, she should be offered help. Help should be offered as soon as she mentions the problem. Early intervention is often easy and brief. The surgeon should be the first person to help the patient, with counseling and, if necessary, suitable medications. Hormonal Therapy A 16-year study that involved 60,000 postmenopausal female nurses found that those who took hormone replacement therapy for 10 years reduced their risk of dying from all causes by 37%, with the most dramatic reduction being death from cardiac disease. After 10 years, the reduced risk for all causes was 16% because of the increased risk of dying from breast cancer. That risk rose to 43% but the women who contracted breast cancer during the first 10 years had a lower death rate from the disease than women who had never taken hormones, probably because of early detection. Chances of early detection of breast cancer are probably better for hormone users because they receive regular check-ups. Before starting therapy, patients need to be aware of their risk factors for cardiovascular disease, osteoporosis, and breast cancer in order to make informed choices. The screening process that provides such information may include a thorough history and physical and an accurate measurement of body weight and height, blood pressure, cholesterol level, and, for some women, bone density. In an extensive review of current literature on the subject, dubbed “the New Science of Estrogen,” Hammond provides an overview of the risks and benefits of hormone replacement therapy and also includes information on therapeutic alternatives. Current theories indicate that estrogen has extraordinarily complex biological effects that translate into a variety of actions in diverse tissues. There is growing scientific evidence that estrogen exerts its beneficial actions on tissues of the skeletal, urogenital, digestive, cardiovascular, ocular, and nervous systems. However, many women are afraid to use it because the media repeatedly tell them that estrogen greatly increases their risk for breast cancer. Statistics show overwhelmingly that cardiovascular disease (CVD)—not cancer—is the leading cause of mortality for postmenopausal women. In fact, one in two women will eventually die of heart disease or stroke, whereas only one in 25 women die of breast cancer. Although the incidence of heart disease, including coronary artery disease and stroke, is low in premenopausal women, heart disease is the most frequent cause of death in women over the age of 50. Since 1984, the death rate from CVD in men has decreased, whereas the death rate for women has increased. Numerous epidemiologic studies support the long-term benefit of estrogen in preventing CVD. Observational studies, such as the Postmenopausal Estrogen/Progestin Intervention Study (PEPI) sponsored by the National Institutes of Health, revealed that hormone replacement therapy (HRT) can increase high-density lipoprotein cholesterol and decrease low-density lipoprotein cholesterol. The Nurses' Health Study demonstrated a reduction in the risk of CVD of up to 50% among current HRT users. Women who use estrogen have significantly less coronary artery stenosis than women who do not use it. Moreover, patients with the most advanced coronary artery disease experience the most benefit from estrogen replacement therapy (ERT), but only 35% of women surveyed were aware of the connection between heart disease and menopause. ERT/HRT is also first-line therapy for osteoporosis for most women, and treatment should begin as soon as possible after the menopause. Discontinuation of therapy is followed by bone loss, which could result in a subsequent increase in the occurrence of fractures. Preliminary data suggest that even the elderly respond to estrogen replacement. However, there are therapeutic alternatives and lifestyle modifications (diet and routine exercise) that perimenopausal women must be counseled about to create a comprehensive preventive program. Such an effort can have a significant impact on long-term morbidity and mortality associated with osteoporosis. Women have phenomenal memories because one of their jobs is to find every needle that gets lost in the proverbial haystacks of their homes. When they become less adept at remembering where they and other people put their things, they fear the worst—that they are losing their minds, and this fear is not illogical.

Women comprise 72% of the population over the age of 85 years, and roughly half of this group has Alzheimer disease (AD). Not only do women constitute a greater proportion of this older population, but AD is expressed earlier in women than men. This may be related to the estrogen loss that occurs with menopause. Hammond cites a study that found women who took estrogen for more than 1 year experienced a dramatic delay in AD onset. But even the group of women who averaged only 4 months of estrogen therapy and most likely took the medication to control symptoms such as hot flushes experienced a delay in AD onset. It has been speculated that a brief exposure to estrogen influenced AD expression 20 to 30 years later by preventing an irreversible loss of neurons associated with the occurrence of hot flushes. Research is ongoing but one study found that estrogen replacement therapy in postmenopausal women is associated with a 50% reduction in the risk of developing AD because it slows the decline of visual memory. Colon cancer occurs more often in women than men and is a leading cause of cancer incidence and cancer deaths in women. Even though mortality rates for colon cancer have decreased 25% among women in the last 20 years, it remains the third leading cause of cancer deaths in this group. The concept that postmenopausal ERT may decrease the risk of colorectal cancer has received considerable attention, even though the hormone has no indication for this use. Some 20 epidemiologic studies have been published that examined this relationship. The majority of these suggest an inverse, protective effect for estrogen, particularly with current use. Although the precise mechanism by which estrogen reduces colon cancer risk is unknown, it has been hypothesized that it affects bile acid metabolism or promotes tumor suppressor activity. The inclusion of estrogen as a measure to prevent colon cancer should be part of the discussions between menopausal women and their physicians. Counseling should include the American Cancer Society recommendations for annual digital rectal examination and fecal occult blood testing as well as a flexible sigmoidoscopy every 5 years or colonoscopy every 10 years. Age-related macular degeneration (AMD) may be reduced by estrogen administration. This disease is the leading cause of legal blindness in the United States, accounting for as many as 60% of all new cases. There is no medical treatment, and surgical management in the form of photocoagulation is effective in only a small percentage of patients with the wet type of the disease. In the Rotterdam study, women who experienced menopause at an earlier age had a 90% increased risk of exhibiting signs of late AMD compared with those who experienced menopause at a later age. These data suggest that HRT reduces the risk of developing AMD. Counseling women about replacement therapy must be combined with discussions about the importance of lifestyle changes, including the following. Normalization of weight Dietary intervention Smoking cessation Regular exercise Control of hypertension Control of diabetes Control of alcohol consumption Control of lipid elevations Routinely, ERT/HRT counseling should go beyond simple symptom control to include both short- and long-term benefits, contraindications, common patient concerns, and misconceptions. The contraindications to estrogen replacement, which have been established by the Food and Drug Administration (FDA) include: known or suspected pregnancy or breast cancer, estrogen-dependent neoplasia, undiagnosed abnormal genital bleeding, and active thromboembolic disorders. However, ongoing research suggests that some of these contraindications may not be absolute. In the meantime, all of the relative contraindications must be carefully discussed and weighed against the risk of not prescribing ERT/HRT. This is also a good time to discuss the common concerns and misconceptions that women have about estrogen, even if patients do not raise them. For example, many women are concerned that estrogen may bring on the return of monthly bleeding, restore fertility, or produce weight gain. For perimenopausal women who choose HRT, there are two common regimens: continuous estrogen and progestin or cyclic (continuous estrogen plus progestin for only 10 to 14 days per month). In ERT,

unopposed estrogen is administered without interruption to women without a uterus. For the perimenopausal woman, counseling must go beyond symptom control to include prevention of heart disease and osteoporosis, control of body weight, restrictions on alcohol and tobacco consumption, encouragement of regular exercise, maintenance of mental health (including sexuality), and cancer screening. When bilateral oophorectomy is anticipated in a premenopausal patient, hormonal replacement therapy should be discussed before surgery because one of the greatest fears of younger women is surgery-induced menopause. Patients should be told that estrogen therapy can be started immediately after surgery and that hot flashes and other menopausal symptoms can be avoided. The natural conjugated estrogens do not cause hypercoagulability and are safe during the immediate convalescent period. The long-term benefits of estrogen replacement therapy in preventing osteoporosis, cardiovascular disease, and colon cancer are well established. The health of the vagina and lower urinary tract is maintained. The vagina lubricates more easily with sexual arousal, and intercourse is more comfortable with an estrogenic vaginal mucosa. Many women report an increased interest in and enjoyment of sex. For women who do experience a loss of libido, even while taking estrogenic hormones, the new androgen therapies look promising as a way to improve sexual function and psychological well-being. Testosterone delivered via transdermal patches or gel bypasses the liver and has no negative effect on cholesterol. The skin serves as a constant reservoir; therefore, blood levels show fewer fluctuations. However, there are physicians who believe hormonal balances induced by prescription medications should only be offered for relief of extreme menopausal symptoms and only for a short while. The author of Dr. Susan Love's Hormone Book and Dr. Susan Love's Breast Book, is one such physician. She is a staunch supporter of eating soybean products and using herbal remedies such as black cohosh to maintain estrogen levels, the use of acupuncture and paced-respiration for hot flashes, exercise, and using vitamin and calcium supplements. But even she, admitted in an interview, “If my symptoms [for menopause] worsen, I may feel that I want to take some kind of drug. I certainly would be open to that.”

SPECIAL CASES Part of "Chapter 3 - Psychological Aspects of Pelvic Surgery" Every patient is special, of course, but some women in tumultuous life situations require a great deal of compassion and sensitivity from a surgeon. These women can present the greatest challenges to the surgeon's ability to handle patients in emotional turmoil. The Teenager Bluestein and Starling report that one million teenagers a year conceive. Of these million pregnancies, 400,000 end in abortion and 100,000 in miscarriage. Childbearing teenagers face a 60% excess in maternal mortality compared with adults and are most likely to suffer toxemia, anemia, hemorrhage, cervical trauma, cephalopelvic disproportion, excessive weight gain, and premature labor. These complications are due more to social and behavioral correlates than inherent adolescent aspects. Correlates include inadequate prenatal care, poor nutrition, substance abuse, and emotional distress. One to two teenagers per 1,000 who have first-trimester abortions experience fever, hemorrhage, and emergency abdominal surgical procedures. These numbers are lower than for older women, but the rate for cervical injury, which could affect future childbearing, is 5.5 per 1,000 teenagers, notably higher than the 1.7 to 3.1 per 1,000 for adults. Bluestein and Starling recommend the following communication techniques for the teenage patient: Guarantee confidentiality to build trust. Conduct the initial interview alone. Teenage girls may want to relate private information to the physician but not in the presence of a parent. Be patient. Gear communication to the patient's emotional and intellectual development. Be nonjudgmental. Gently explore the teenager's family and social environments. Will family members and friends support or harm your patient? Discuss the teenager's long-term plans. Be aware of ethnic differences in health-related matters. Female patients who belong to local or national gangs come into the health care system with a unique set of problems and values. It is imperative that the surgeon not add to the patient's stress more than is absolutely necessary. Sometimes, the simple act of being treated like a human being worthy of consideration is a novel and humbling experience. The Senior Woman It is easy to imagine a grandmother baking cookies or rocking a new baby in the family. Our culture makes it a little more difficult to imagine this same woman flushed and happy from an energetic romp in the sack with her favorite beau. Sex after 60 is a reality for many women and men. These people enjoyed sex when they were young, they perfected it as the years flowed by, and the thought of doing without expression of the natural urges, even temporarily, is discomforting. When an older woman faces gynecologic surgery, do not assume she is asexual. She may simply view her sex life as her private business and be loath to discuss it. Do not antagonize or humiliate the senior patient by assuming her sexuality is a thing of the past. If the surgical procedure will affect her sexually, explain the consequences. Assure her that the surgery will leave her as whole and as functional as possible. Give this woman the opportunity to express her anxieties and to ask questions. If her mate or lover is present, include this person in the discussion. In the senior woman who has a large cystocele, rectocele, or uterine or vault prolapse, every effort should be made to avoid colpocleisis. Sacrospinous colpopexies and vaginal reconstructive surgery are indicated. The fact that a woman is older and sexually inactive at the time of surgery does not mean that she will never be active again. Much preparation and explanation must be given to evaluate and prepare her for the closing off of her vagina or its subsequent reconstruction should that become necessary. Butler and colleagues advise that, when taking a sexual history, ask if the symptoms started after a period of sexual activity. Excretory urogram and cystoscopy sometimes can be avoided with a diagnosis of postcoital cystitis and subsequent antibiotic treatment. A history of traumatic intercourse in the presence of an atrophic vagina can lead to estrogen therapy and lubricants rather than dilation and curettage. Butler and colleagues, also, caution physicians not to assume an older woman is sexually inactive. They advise checking for sexually transmitted diseases in older women as you would in younger ones. When surgery is indicated, perform it as if the senior woman was still young and had many years of

sexuality ahead of her, because she probably does. If the surgery occurs in the genital area, do not shave her pubic hair. The procedure is archaic and microbiologically unnecessary if competent sterilization procedures are followed. As men age, they sometimes lose hair on the head. As women age, the pubic hair sometimes thins and once shaved, may never grow back. Many older women have had to curb their sexual appetites to compensate for physiologic changes that have occurred in a mate. Removal of excessive vaginal mucosa during pelvic surgery compromises the vagina, which is already losing elasticity. This inhibits penetration during coitus and can cause painful intercourse. Be understanding and helpful if solutions are possible. If the dysfunction resulting from the surgery will be permanent, encourage the couple to experiment with various ways to please each other or seek the advice of a sex therapist that is knowledgeable about creative sexual play. Butler instructs physicians to educate themselves about the effects of medications on sexuality. More than 200 medications have sexual dysfunction as a side effect. Many are effective in lower doses that do not harm the libido or the patient's physical ability. The positive benefits of a healthy sex life are multiple. Emotional intimacy and the ability to connect physically with another human being brings great joy and satisfaction. This need to connect intensifies rather than diminishes with age. The Sexually Assaulted Patient Reported rape and sexual assault in the United States, at the rate of 34.4 victims per 100,000 people showed no evidence in the late 1990s of being on the decline. It is estimated that only about 16% of the rapes that occur are ever reported. One in three sexually assaulted victims is under the age of 12. Convicted rape and sexual assault offenders testify that two thirds of their victims are younger than 18. Two thirds of the victims older than 18 knew their attacker prior to the rape. A sad fact is that almost one fifth of the women who are raped before the age of 18 are raped again after the age of 18. Many require medical attention, up to 22% suffer genital trauma, up to 40% incur sexually transmitted disease, and 1% to 5% become pregnant as a result of the rape. Rape survivors are 13 times more likely than the general population to attempt suicide. Most hospital emergency rooms have strict protocols to follow when treating sexually assaulted patients. The proper collection of evidence and initial treatment of injuries is a priority. The surgeon who repairs the gynecologic damage done to these female patients should be aware of the general characteristics of psychological trauma associated with rape. The symptoms of rape trauma syndrome compiled by Blair and Warner can be found in Table 3.2. TABLE 3.2. Symptoms of Rape Trauma Syndrome

Recurrent, painful recollections or dreams of the event

Suddenly acting or feeling as if the event were recurring

Demonstrations of fear, anger, or anxiety

Crying, restlessness, or tenseness

Controlled feelings masked by a false demeanor of calmness, composure, or subdued attitude

Matter-of-fact answering of questions

Inappropriate smiles or laughter

Inability to remember parts of the event because re-exposure to stimuli present during the traumatic moment reinvokes the associated pain

Decreased interest in important activities

Lack of future plans

Limited range of affect

Detachment toward others

Sleep disorders

Difficulty concentrating

Hypervigilance

Irritability

Irritability

Angry outbursts

Many of these symptoms are the same as seen in posttraumatic stress disorder. These authors also outlined the following necessary skills that any caregivers attempting to help a sexually assaulted victim should possess the following. Understand rape and sexual assault Assess how the patient perceives the act Identify and reinforce patient's ability to cope Assist significant others Coordinate care and help if victim needs assistance Mobilize community resources The female gynecologic patient, from the youngest child to the oldest adult, has survived the attack on her physical body. Most gynecologists are familiar with the legacy of these attacks (up to 75% of patients seen with chronic pelvic pain were physically or emotionally abused). The assaulted patient now needs reassurance from her surgeon and other health care workers that she will also survive the damage done to her psyche. The Cancer Patient Bloch cautions physicians that the manner in which the diagnosis of cancer is disclosed to the patient can determine whether the patient lives or dies. It is imperative to instill hope and the desire to fight the disease. A diagnosis of cancer by telephone can be devastating and is viewed by many physicians and patients as a form of cruelty. The cancer patient shares the same fears common to all surgical patients. She may be psychologically attached to the body part that must be removed. She may fear anesthesia, disfigurement, the unknown, the hospital experience and staff, debility, loss of economic competence, and sexual function. In addition to the normal realm of anxiety associated with surgery, a diagnosis of cancer brings special stresses to a woman's life. Schain describes universal concerns experienced by people diagnosed with cancer no matter where in the body the cancer is located. These people fear death, postoperative adjuvant treatment, and recurrence. The cancer patient is concerned about dying or being injured during the operation. Preoperative anxiety can manifest as anorexia, insomnia, tachycardia, fear, and panic. Acute depression is not uncommon and can lead to suicidal tendencies. The patient is afraid of becoming unable to take care of her family members. Will she be sick for long? Who will raise her children if she dies? Who will care for her parents? Will her loved ones be supportive, or will they begin to back away from her? If the woman is alone, without mate or family, she fears becoming unable to care for herself. The treating surgeon can ease some of the cancer patient's fear by free and open communication. Educate the patient about exactly what to expect, and offer reassurance if possible. Explain the procedure she is about to undergo and the positive benefits that you both hope will result. Surgery and cancer in the urogenital region can be consciously or unconsciously interpreted as mutilation. The possibility of some degree of postoperative sexual dysfunction can create a fear of abandonment by the sexual partner. Assure both the patient and her partner that they will continue to be able to bring joy, pleasure, and comfort to each other.

The psychological effect of cancer on a woman is largely determined by whether the malignancy is primary, recurrent, or terminal. The surgeon is usually the patient's first big gun aimed at the cancer and the patient's greatest hope for a complete cure. The surgeon most likely is the first physician to learn the stage of growth of the cancer, and the family will be waiting to hear the results of the surgical experience. How advanced is the cancer? What is the prognosis? What happens next? If the news is good, the patient recovers and has routine checkups to monitor her health for the rest of her life. If the news is not good, the results should be revealed with as much sensitivity and hope as possible. No patient should ever be given an absolute death sentence. Ethically and morally, the surgeon must wrestle with personal internal conflicts when the surgery for the cancer patient is palliative. Should procedures be performed that extend death and not life? If she is able, the patient herself must make the choice. If the cancer patient elects to have the procedure, the surgeon's job is primarily over once she has recovered from her operation. If the cancer has spread, radiologists, chemotherapists, and others who specialize in cancer treatment will assume the lead roles in trying to get the patient well or at least enhance the quality of the remainder of her life. Meaningful psychosocial support should be based on established concepts of crisis intervention because cancer is a major life crisis. Counseling should be designed to support adaptive behaviors and feelings that reduce the psychological stress, restore and consolidate the patient's self-image, and normalize sexual functioning as quickly as possible. It is the surgeon's job to foster courage, not hope, in the terminal patient.

The Suicidal Patient Approximately 30,000 people commit suicide each year in the United States. Cooper, Rosa, and Daniel studied over 6,000 individuals and found that hopelessness ranked the highest of all symptoms associated with suicidal ideation in psychiatric patients and adolescents. A surgeon probably will not know if a patient is contemplating suicide unless the patient admits to thinking about ending her life or unless a family member voices concern. Subtle warning signs include a chronic state of depression, lethargy, an inability to relate to others, weight loss or gain, lack of interest in life in general or the surgical procedure in question, a change in personal appearance, abnormal sleep patterns, or any strange behavior uncharacteristic of the particular patient. Many of these symptoms are normal for the presurgical and postsurgical patient, so it is easy to see how suicidal tendencies could hide within the maze. To determine a patient's suicide potential, ask: Have you been troubled with thoughts of hurting yourself? Barbee advises physicians to ask directly: Have you ever thought of taking your own life? Do you feel like taking your life now? If the patient answers yes, the physician should find out if the patient has a simple, straightforward suicide plan that is likely to succeed. A patient with low to moderate suicide potential is noticeably depressed but can identify some support system. There may be suicidal thoughts but no specific plan. A patient with a high suicide potential feels profoundly helpless. Little or no support system seems to exist. Thoughts of suicide are frequent, and a plan exists that is likely to succeed. The surgeon may be the first person to realize the patient is in emotional trouble and could be the only support system available if the patient has significantly withdrawn from others. Crisis therapy for the suicidal patient begins when the patient requests help or when someone recognizes her potential for self-destruction. By showing interest in her feelings and concern for her welfare, the surgeon initiates crisis intervention. Medcast News Networks reported an interesting study that was presented at the 152nd American Psychiatric Association. Antti Tanskanen and colleagues followed approximately 40,000 Finnish men and women for 15 years. Observing their health habits and causes of death, they found that those individuals with high cholesterol (309.4 mg/dL) were more than twice as likely to commit violent suicide compared with those whose blood lipids were in the normal range (193.3 mg/dL). This is a simple thing to check and

one that might make a life-or-death difference for a patient.

DEALING WITH DEATH Part of "Chapter 3 - Psychological Aspects of Pelvic Surgery" Death is a taboo subject for most physicians. Many doctors go to great lengths to avoid discussing the eventuality, and most are uncomfortable dealing with the emotional consequences of the death of a patient. To many physicians, death is viewed as a personal failure. Regardless of any internal attitudes, the surgeon is responsible for communicating the facts of the death to the family and for being the first person to help them cope with the loss of their loved one. When a patient dies, explain to the family, to the best of your ability, exactly what happened. Assure them that everything medically possible was done. Ufema suggests the following guidelines to make the death experience as bearable as possible for all concerned. Ask about donations for transplantation. Provide an area where the family members can say goodbye. If it is true, tell the family members how the patient affected you personally. Transfer your protective feelings for the patient to the family members. Begin the grief process by saying, “Jane's body is ready for the morgue now. Would you like to say a final goodbye?” If family members want a lock of hair, allow them to take it. Give family members the patient's belongings. Provide an escort for any family member who is alone. Help the hospital staff cope with the loss. Remember that it is permissible for the physician to show emotion with the family if the emotion is genuine. Someone is gone and will never see tomorrow.

FUTURE TRENDS Part of "Chapter 3 - Psychological Aspects of Pelvic Surgery" The new millennium was heralded with great joy and celebration. Its predominant theme was one of hope, and nowhere is this more evident than in the field of medicine, specifically, for women's health issues. The hand that rocks the cradle now rules important aspects of the research world. During the next 20 years, 45 million women will enter menopause and live one third to one half of their lives postmenopausally. Research into improving the quality of their lives and subsequently, all those they touch, has never been more important or timely. As a multitude of women move through the life stages of maiden, mother, and crone, the health care they receive during one stage will be reflected in all those that follow. Boomer women, born between 1946 and 1964, will continue their legacy of reform by participating in medical clinical studies concerning endocrinology, gynecology, neurology, oncology, genetics, and finally, geriatrics. The daughters and sons they presented to the world will advance reproductive technology in all its genetic and social ramifications. Virtual surgery is a recent gift of the technological world. Medical students will soon be able to perfect their surgical techniques using cyberscalpels, simulated life-like patient bodies, three-dimensional models, and long distance mentors. The space industry has also yielded a smart surgical probe that will be used for breast biopsies. The small disposable needle with multiple sensors will be able to distinguish normal tissue from tumor tissue and greatly reduce the 18,000 breast biopsies per week now performed on women with suspicious lesions. Originally conceived as a robotic tool to aid astronaut/physicians during long-duration space flights, it has found a medical application on Earth in neurosurgery. Voice-activated language programs will make it possible to effectively communicate with multicultural patients and international colleagues. Advances in documentation techniques will make keeping accurate records less of a burden. More ethics classes will emerge to better prepare the medical student for the real-world challenges such as end-of-life issues, fiduciary responsibilities, confidentiality, informed consent, and religious or philosophical conflicts. As surgeons, it will be your duty and privilege to witness, record, and practice the new knowledge and incorporate it with what is already known so that you can make a positive difference in the lives of women now living as well as those yet to be born.

CONCLUSION Part of "Chapter 3 - Psychological Aspects of Pelvic Surgery" Technology will never change the fact that women are emotional creatures by nature, even when they have the choice to menstruate or not. Few experiences give women more satisfaction than sexuality and conception. Surgery on or near her reproductive tract is potentially fraught with emotional sequelae. The gynecologic surgeon's responsibility transcends a stunning performance in the surgical arena. Making appropriate decisions about the surgical procedure and considerately managing the patient's return to physical and psychological health will be the hallmarks of an accomplished surgeon in this millennium and the next.

Chapter 4

Professional Liability and Risk Management for the Gynecologic Surgeon Charles J. Ward As we enter the new millennium the gynecologic surgeon, both in the United States and Europe, continues to be faced with increasing pressures from the threat of malpractice litigation. This chapter outlines practical steps that should reduce the exposure to malpractice claims. It also discusses the litigation process and its emotional impact on the physician. This chapter is neither to be considered a substitute for legal counsel nor a source of legal advice. A bibliography is furnished for those who wish to explore this subject further. The history of liability suits extends back to the 14th century. The first recorded malpractice suit in English law occurred in 1374. This case involved an action brought before the King's Bench against a surgeon, J. Mort. The plaintiff sustained an injury to the hand, and treatment of this injury left the hand maimed. The defendant surgeon was found not liable because of a legal technicality in the writ of complaint. But the rule was clearly laid down that if negligence were proven, the law would provide a remedy. The court further held that, “if the surgeon does well as he can and employs all his diligence to the cure, it is not right that he should be held culpable.” The civil liability of surgeons arises out of the rule laid down in 1534 by the English jurist, Sir Anthony Fitzherbert, which stated, “it is the duty of every artificer to exercise his art right and truly as he ought.” The first recorded malpractice suit in the United States occurred in Connecticut in 1794 and was also against a surgeon. In this case, Cross v. Guthrey, the patient's husband sued the surgeon, Dr. Guthrey, after the patient died from postoperative mastectomy complications. The suit alleged the doctor was guilty of negligence in operating on the plaintiff's wife … in the most unskillful, ignorant, and cruel manner, contrary to all the well-known rules and principles of practice in such cases, that the patient survived by but three hours, and that the defendant had wholly broken and violated his undertaking and promise to the plaintiff to perform said operation skillfully and with safety to his wife. The jury found the surgeon liable and awarded damages of £40. The 1970s and 1980s found the United States engulfed in a malpractice crisis. The 1990s have continued this trend, which had now spread to Europe. The millennium continues to bring additional pressures to bear on the gynecologic surgeon. These pressures arise from areas such as the transformation of the practice from a fee-for-service (FFS) reimbursement to various managed care programs, third-party payer interventions, declining incomes yet increasing overhead costs, longer work hours, and sizable malpractice premiums. Because the obstetrician/gynecologist performs six of the 10 most common surgical procedures, it is not surprising to find this specialty to be the most frequently sued. The 1999 survey of the American College of Obstetricians and Gynecologists membership found that 76.5% of all fellows have been sued at least one time. The obstetricians/gynecologists surveyed had an average of 2.53 claims filed against them during their career. Even resident physicians in training reported that 27.8% had at least one professional liability claim filed against them during their training. Gynecologic care involved 41.1% of all the claims filed. Of the gynecologic claims, patient injury accounted for the most frequent primary allegation and represented 25.8% of these claims. Failure to diagnose was the second most common allegation and accounted for 25.1% of gynecologic claims. Of these, failure to diagnose breast cancer amounted to 63.3%, failure to diagnose cervical cancer 18.1%, and failure to diagnose ovarian cancer accounted for 5.9% of the gynecologic claims. On average, 4.2 years elapsed from the onset of the claim to its closing. Fifty-four percent of all the claims were dropped or settled without any payment on behalf of the physician. Obstetricians/gynecologists won 65.5% of those claims that went to arbitration or a jury verdict.

Contrary to popular impressions that “bad doctors” are the cause of most malpractice actions, studies have shown that physicians who possessed higher qualifications, were board certified, and had more experience were more likely to be sued. This is in part because of the fact that such physicians manage higher-risk patients; therefore, they are more likely to encounter adverse outcomes that can be a source of litigation. Medical malpractice occurs when the treatment rendered is constituted to be below the degree of care exercised by physicians generally under the same or similar set of circumstances. Today most communities are held to a national standard of care that is defined as a duty to exercise the degree of care and skill expected of a reasonably competent practitioner in the same specialty acting under similar circumstances. Maloccurrence, or a poor outcome, is not malpractice unless the outcome arises from a direct effect of a breach of the standard of care. Substandard care per se does not always imply malpractice. There must exist a direct cause-and-effect relationship between the breach of the standard of care and the outcome to constitute malpractice. RISK MANAGEMENT INFORMED CONSENT LITIGATION PROCESS

INFORMED CONSENT Part of "Chapter 4 - Professional Liability and Risk Management for the Gynecologic Surgeon" Informed consent is a legal doctrine that requires physicians to obtain consent for treatment, whether it is diagnostic or therapeutic, medical or surgical, invasive, or noninvasive. Without informed consent the physician can be held liable for violating the patient's rights regardless of whether the treatment was appropriate and rendered within the standard of care. Although failure to provide informed consent is seldom the primary charge in a malpractice claim, it has been found to be a secondary issue in almost one third of malpractice claims. Informed consent is an ongoing process that includes the exchange of information and the development of choices. Informed consent is a process of ongoing-shared information. It provides for the development of choices and requires active participation from both the patient and the physician. It involves respect for the patient's bodily integrity and right to self-determination; thus, it respects the patient's autonomy. Informed consent should never be confused with the mere signing of a consent sheet. Without informed consent, a physician may be held liable of violating the patient's rights even if the treatment was appropriate and administered within the standards of care. Failure to obtain informed consent can result in the physician being accused of “battery” under common law. Since the 1947 Nuremberg Code, valid consent has consistently been described as having four characteristics: voluntary, competent, informed, and understanding. The American College of Obstetricians and Gynecologists, in its statement of May 1992 on “Ethical Dimensions of Informed Consent” explores the concept of voluntary or free consent and understanding or comprehension. The concept of freedom implies both an ability to choose as well as an ability to refuse treatment. Freedom also implies a lack of coercion, manipulation, or infringement on the patient's decision-making process. Recognition of different values, preferences, and alternatives is important in the process of free consent. Courts have recognized and states have adopted three different degrees of disclosure in informed consent. The first, the professional or reasonable physician standard, was prevalent prior to the 1970s. This degree of disclosure was based on the type and amount of information that a reasonable physician would tell the patient about the risks and benefits of a particular treatment. This paternalistic approach began to give way in the 1970s to the second type of disclosure, The materiality or reasonable patient viewpoint standard. This gave patients more input into the decision-making process. Under this standard, the physician had to disclose what a “reasonable person” would want to know under similar circumstances concerning the risks and benefits of a particular treatment. This concept is based on the patient's need rather than the professional perception of what the patient should know. The third, but not widely held, disclosure is the subjective patient viewpoint. Under this disclosure, physicians must disclose varying amounts of information based on the individual's personal needs and peculiar requirements. This form renders a standard extremely difficult for physicians to understand and apply. The need for informed consent can be suspended in an emergency situation, but specific criteria must be met for a situation to be declared an emergency. The patient must be unconscious or incapacitated and suffering from a life-threatening or serious health-threatening condition requiring immediate medical treatment. It is important in these situations for the physician to document the following: a description of the patient's condition at the time of the emergency, the reason the emergency existed, and an explanation for the need for immediate attention. Patients have the right to refuse treatment after receiving informed consent. Under these circumstances, for legal protection, the physician should document the reason the patient gave for refusal of the proposed treatment, the reason that the physician felt that the proposed treatment was indicated, and the possible jeopardy to the future health and well-being of the patient that might occur from the refusal of the treatment. It would be prudent to have the patient sign a statement acknowledging the refusal for treatment and listing the potential adverse consequences that might occur. The office setting provides the best environment for providing informed consent for several reasons. In this setting, the time for adequate consideration and adequate discussion is provided. Family members

who have a definite influence on the patient's decision can attend or be encouraged to attend. If the physician lacks the ability or time to provide the necessary adequate informed consent, the physician can delegate this responsibility to a nurse practitioner or physician assistant who has been previously trained to provide informed consent and is able to answer the patient's questions. Pamphlets, audiovisual, and even interactive visuals can greatly assist in this process. Because people rarely retain more than 30% of verbal communications, the average patient leaves the physician's office having forgotten or not understanding most of what was shown and explained. Thus, the more the patient receives in writing or by audiovisual instruction, the better is her understanding. Once the process of informed consent is complete, the physician can document the record to reflect when it took place, what was disclosed, that the patient had time to ask questions and have her questions answered to her satisfaction, and that the patient then requested the proposed treatment. Well-informed patients are more likely to overlook less than desirable outcomes and accept that medicine is still an art and not an exact science. Less-well informed patients are more likely to have unrealistic expectations and sue when these expectations are not met. Therefore, the best protection to a malpractice claim may come from taking full advantage of the legal doctrine of informed consent. The degree of disclosure in informed consent varies from state to state. Some states require elaborate consent, others much less. When a complication develops that was disclosed in the informed consent; however, there is no cause for legal action so long as the complication did not arise from a negligent act. There is almost universal agreement that informed consent should encompass the following six areas: Diagnosis Nature and purpose of the procedure Risks of the procedure Likelihood of success Reasonable alternatives Prognosis if the treatment is refused Figure 4.1 is a general surgical informed consent meeting these requirements. There are several points to note regarding this particular form. First, throughout the form, the words “I request” or “request for surgery” are included. These words are used to place greater emphasis on the patient's responsibility for her choice in selecting the proposed procedure. Next, a list of the “general risks of surgery” is provided. This list covers those risks that can arise from any surgical procedure as well as anesthesia. In describing the risks and complications of the procedures, the words “may include but are not limited to such complications as” have been added and are used to indicate that the physician did not attempt to list every potential complication but rather those that a reasonable patient would expect to know before making a decision. In the section on alternative forms of treatment, the words “such as” indicate that only those alternatives considered to be appropriate need to be listed. Finally, space is provided to list any additional materials that the patient was given or might have reviewed in making her decision. If a claim arises, the physician can then explain to a jury the situation surrounding the alleged act of negligence and use the same material in his or her defense to show that a reasonable person was properly informed.

FIGURE 4.1. Informed consent and request for surgery.

View Figure

Figures 4.2, Figures 4.3, Figures 4.4, Figures 4.5 and Figures 4.6 are informed consents covering the five most common gynecologic operative procedures. These forms describe each procedure in words that a patient can comprehend easily. A reading specialist has reviewed each of these forms, as well as the general surgery consent form described in the preceding. The language used has been placed at the level of an individual with a sixth-grade education, which is generally that used by daily newspapers. Each of these forms covers the risks and complications as well as alternatives that apply to the specific procedure along with the “general risks of surgery.”

FIGURE 4.2. Informed consent and request for dilation and curettage, hysteroscopy, or cervical biopsy.

View Figure

FIGURE 4.3. Informed consent and request for hysterectomy.

View Figure

FIGURE 4.4. Informed consent and request for sterilization.

View Figure

FIGURE 4.5. Informed consent and request for laparoscopy.

View Figure

FIGURE 4.6. Informed consent and request for repair of relaxation of pelvic organs and/or correction of urinary incontinence.

View Figure

Figure 4.2, the informed consent and request for dilatation and curettage, cervical biopsy, and hysteroscopy, lists the most common complication seen, namely perforation of the uterus. It also notes that the procedure can have an adverse impact on subsequent pregnancies. Figure 4.3, the informed consent and request for hysterectomy, covers complications seen, including injuries to the genitourinary tract, the most frequently seen complication. It also discusses the risk associated with blood transfusion. Figure 4.4, the informed consent and request for sterilization, clearly states that the procedure is not designed to be reversible. Many patients have a sterilization procedure performed in their twenties and in this age of divorce and remarriage, later request that the sterilization be reversed. When the procedure cannot be reversed, patients may sue with a claim of “lost opportunities.” This statement in the informed

consent is designed to cover that particular area. Ectopic pregnancy, one of the known and accepted risks of any sterilization procedure, is clearly spelled out. Figure 4.5, the informed consent and request for diagnostic and therapeutic laparoscopy, reviews the more commonly seen complications of injury to the gastrointestinal, or genitourinary tracts and vascular injury. The form also notes the risk of exploratory laparotomy. This provides an opportunity for the physician to discuss this potential extension of a laparoscopy procedure so that the physician is not hesitant to proceed with laparotomy when the findings require this procedure. Figure 4.6, the informed consent for repair of relaxation of pelvic organs and/or the correction of urinary incontinence problems, notes that recurrence of urinary incontinence is a potential complication. This is designed so that false expectations from the surgical procedure are not created. These procedure-specific informed consent forms are detailed but have been found to be accepted by patients. They also have made the defense of surgical complications much easier, especially when there is no negligence. The legal requirements of informed consent vary from state to state and jurisdiction to jurisdiction. The surgeon should be aware of the degree of disclosure required as well as any specific laws on informed consent that exist in his or her state. County or state medical societies are usually able to assist in providing this necessary information.

LITIGATION PROCESS Part of "Chapter 4 - Professional Liability and Risk Management for the Gynecologic Surgeon The Claim Several occurrences can forewarn a physician that a legal action may be impending. These include a complication or poor outcome, a disgruntled patient or dissatisfied family member, a request for medical records, or direct contact by the plaintiff's attorney. Under no circumstances should the physician ever discuss a potential case with the plaintiff's attorney unless he or she has been so advised by their attorney or insurance carrier claim investigator. Some plaintiff attorneys have been known to call a physician and ask for the physician's opinion about an alleged incident. The attorney may tape this conversation or later in a courtroom allege that on the date of such a contact the physician said something different from what the physician is now alleging occurred. This situation could possibly lead to the jury questioning a physician's credibility. Thus, it is best for the physician never to respond to a plaintiff's attorney except through his or her own defense team. A lawsuit actually begins when a plaintiff's attorney files a formal Complaint or Declaration. This is a legal document that lists the allegations to support a claim of medical malpractice. Some states require that an affidavit from an expert witness supporting the contention of malpractice be filed with the Complaint or Declaration. Once the filing occurs, the court serves the defendant physician with a summons. The summons is attached to the complaint and may contain questions to be answered by the physician. Responses to the complaint must be filed within a specified period of time, usually within 20 to 30 days. The defendant physician must respond within that period of time or he or she can be found guilty by default. Thus, it is critical that the physician immediately contact his or her insurance carrier and/or attorney on receiving the summons. The physician should send the original summons by certified mail or hand deliver these documents to the insurance carrier. Keep a photocopy of the summons in a separate file. Most insurance carriers require that the physician notify them of any incidence that might result in a claim. Failure to do so could negate the physician's policy. It is prudent to know what the requirements of the insurance carrier are in this regard. Once a suit has been filed, do not communicate with the patient, her family, or her attorney. When records are requested, be sure that a proper authorization from the patient has been received and then send a complete copy of the chart. Never change or alter a record once the suit has been filed, because there are many methods by which a plaintiff's attorney can prove that records have been tampered with or altered. Prepare a thorough chronological account of all events surrounding the incident. Include any oral communications that the physician had with the patient or her family. Gather all records, radiographs, laboratory tests, and consultant's reports and maintain these documents along with the complaint in a separate file. The physician's defense attorney will know all aspects of the law but may not be knowledgeable about the specific medical condition. Thus, it is imperative that the physician take an active role in educating the attorney in all medical aspects of the case, discussing not only the positive aspects of the therapy, but also any questionable areas. The defense attorney should thoroughly understand all medical aspects of the case and the physician's reasoning behind the particular treatment chosen. The physician should research literature for articles by authors who completely support the treatment rendered. The physician should help select expert witnesses who can support his or her views on the treatment that was rendered in the case. During the discovery process, both parties submit questions called interrogatories that are to be answered under oath within a specific period of time. The defendant physician should help prepare answers to these interrogatories. If a physician has only peripheral involvement in a case and receives a summons requesting records, it is prudent that he or she notify the insurance carrier as well as his or her attorney before any response is given. There are times when poorly worded answers result in the physician who has only a peripheral involvement being named in the actual suit. The physician should not take unnecessary chances, but rather exercise his or her legal right and have appropriate counsel before responding to any questions. The defendant physician should not discuss the suit with any other physicians unless so advised by his or

her attorney or the insurance claims representative. Idle conversation in a surgical dressing room may be overheard by a physician who could become the plaintiff's expert. The only people with whom it is safe to discuss a case are the defense attorney and the insurance claims representative. The Deposition The deposition is a standard legal process that takes place as part of the discovery process. A deposition is taken under oath in front of a court reporter and is generally admissible during the trial phase of the lawsuit. The deposition is serious in that what a defendant physician states is cast in stone and if not properly articulated can end up as a millstone about the physician's neck. Thus, adequate preparation cannot be overemphasized. The physician should be familiar with all of the records—both office and hospital—including office and hospital protocols, rules, and regulations. The physician should insist on considerable preparation and education from the attorney about this important legal process before the deposition. The deposition has many purposes, including to: (a) discover facts; (b) lock in testimony; (c) narrow and clarify issues; (d) discover additional witnesses; and (e) provide an opportunity for settlement if appropriate. The deposition is primarily taken so that each side can learn the opinions, theories, and strategies of the opposing experts and those involved in the case. The opposing attorneys can thus prepare their cases for trial with the knowledge gained. The demeanor of the physician is important. He or she should be professional, honest, and confident. Boredom, frustration, and hostility should not be shown. Anger can disrupt a physician's ability to concentrate, and under those circumstances, the physician may divulge things that should not be discussed. The physician should neither attempt to teach nor lecture during a deposition, remembering that the plaintiff already has an expert who stated that the physician committed malpractice. Therefore, any attempt to educate the lawyer is a frivolous and potentially dangerous act. During the deposition phase, the defendant physician is not acting as his or her own expert and thus should confine answers to short factual statements with no elaboration unless requested by further questioning. Answers should be specific and simple. Do not bring into a deposition any personal notes or records that include the physician's predeposition preparation. Such records are discoverable by the plaintiff's attorney if they are in the room. However, the physician should have a complete set of records to refer to. Never guess about an answer. If the physician does not understand the question, then simply state, “I do not understand the question,” or “I do not know,” or “I cannot recall.” Avoid complex questions. Make the attorney break the questions down into simple ones by stating, “I do not understand the question. Would you please rephrase it?” Avoid repetitive questions by noting, “I have already answered that.” When a question is asked, listen attentively, think, and organize your thoughts before responding. Then respond slowly, clearly, and as concisely as possible. Given a hypothetical question, the physician should be sure the facts are consistent with his or her case and reply only as it applies to the case. If the hypothetical situation is not applicable to the physician's case, the physician should qualify the response and so state that the given facts are not applicable to the case. Again, remember that you are defending yourself and not acting as your own expert witness. If the defense attorney objects to a question or its phrasing, do not answer the question until instructed to do so by the defense attorney. The physician should listen carefully to the objection raised by the defense attorney so that he or she can appropriately word the response. The physician should know the various alternative methods of therapy that might have been employed. The physician should have reasons why he or she selected the chosen therapy and be prepared to explain them in the courtroom. The deposition, however, is not the time to divulge any literature research that took place. Avoid the word authoritative. Legally, it implies that every single statement in a book or journal is absolute fact. Because medicine is an art rather than an exact science, experts may have different opinions on the same subject. The physician can admit that a book is scholarly or written by intelligent, respected people, but should never concede that every statement in the text is absolutely correct, which would make such a text authoritative by legal definition. Do not engage in bantering or make comments that could be overheard by the plaintiff's attorney during a recess or off-the-record moment. If the plaintiff's attorney overhears things stated during the recess, he or she has the right to come back on record and bring up such topics.

Do not accept the plaintiff's attorney's statement regarding prior testimony as being factual unless you are sure that the statements accurately reflect the stated facts. If you are not sure, state, “I do not know,” or “I do not have sufficient experience in that matter to be able to give an honest answer.” When a deposition is over, the physician should politely excuse himself or herself, leave the room and wait for further instructions from the defense attorney. Prepare for a deposition as if it were a trial. Although the setting may be more casual and appear friendlier, your answers are given under oath and carry the same weight as if they were given before a jury in a courtroom. After the deposition is completed, the physician is given an opportunity to read and correct his or her testimony. The physician has the right to correct the deposition, even change the meaning of an entire sentence, as long as this is accomplished during the specified time, usually 30 days. Because this is the only time the physician has the right to change his or her testimony, it is extremely important that the physician read and correct the deposition before it is used in legal proceedings. The Trial The trial is an adversarial experience during which allegations are made and facts are presented. In a courtroom, the plaintiff's attorney states his or her perception of the facts that are beneficial to the client. This perception may not always correlate with the defense's understanding of the facts. The jury, however, is instructed to judge a case by the facts, as they understand them and not by what an attorney alleges. Often sharp, even bitter controversy develops over the interpretation of significant facts. Legal maneuvering can distort or even suppress the presentation of some facts. This is an emotionally draining experience. Preparation for the trial is critical and should cover even demeanor and dress. Dress should be plain, conservative, and neat. Flashy or casual clothes project the wrong image. The physician wants to present himself or herself to the jury as a kind, considerate, compassionate person who projects warmth and sincerity, as well as a voice of authority. The defendant physician should go to the court before the trial to become familiar with the surroundings. He or she may even want to sit in on a trial to get a feel for this new and foreign environment. The physician should discuss in great detail any questions that he or she has regarding the trial, its proceedings, or the defense. Listen to your attorney's advice! The more familiar a physician is with the court and its proceedings, the more confident and less frightened he or she will be as the trial proceeds. There are several things a physician should do when called to the stand to testify. The physician should make and keep eye contact with the jury throughout the testimony. Physicians generally tend to focus their attention on the plaintiff's attorney as he or she asks questions; however, this is not helpful to the physician. Physicians are trained to read people's facial expressions. By maintaining eye contact with the jury, the physician can closely observe whether the jurors comprehend the testimony. If the physician feels that the jury does not understand an area of the testimony, he or she can elaborate further on a simpler level of explanation. Physicians must remember that the jury, not the plaintiff's attorney, reaches the verdict; therefore, it is critically important that the physician employ all of his or her talents to be sure that the jury understands the testimony. Be polite and humble. Important words in the courtroom are Sir and Madam. When a physician uses these words, they generate an air of humility that is pleasing to a jury. When a question is asked, be sure to take time to think before a response is given. Respond in nontechnical language that a 12-year-old child could understand. Listen carefully, and then speak clearly and audibly. Jurors like a physician who really tries to explain what he or she did step by step. Unlike during the deposition, in the courtroom the physician should be a teacher, using a blackboard or other audiovisual aids to explain complex issues. If the plaintiff's attorney tries to be overbearing and abrasive, react by being humble yet dignified. Repeat significant facts often so that the jury will remember them. There are also several things a defendant physician should not do on the stand. Do not be arrogant, pompous, or overbearing. Never try to outsmart or manipulate the plaintiff's attorney; this is the way attorneys make their living. Never instigate a fight or argue with an attorney because it will cause you to lose dignity and respect in front of the jury. Never belittle the patient's injury or pain, but show compassion and understanding. Never guess at an answer. If you do not know, then state, “I do not know,” or “I do not recall.” The physician should not worry if he or she misspeaks. The defense attorney, with further appropriate questions, generally is able to correct the mistake. Never get angry or show anger because anger breaks your concentration and often leads you to say things that should not be said. Answer the plaintiff's questions as simply and briefly as possible. Use a yes-or-no answer whenever

possible. The defense attorney, on direct examination, can have the physician elaborate in detail those issues that are important to the defense. It is important that the physician stay in the courtroom during the entire trial. The physician can help the defense attorney during the cross-examination of expert witnesses and reveal where the experts are inconsistent with known and accepted facts. The physician's very presence and demeanor can greatly influence the jury's perception. This can play a major role in the final outcome. The trial begins with opening statements. The plaintiff's attorney proceeds first because he or she has the burden to prove the existence of malpractice and causation of injury. Then the trial testimony phase begins. Again, the plaintiff's attorney begins with testimony and expert witnesses to try to establish that the defendant is negligent and the negligence caused the plaintiff's injury. During this initial presentation, the plaintiff's attorney has the right to (and frequently does) call the defendant physician to the stand. At this stage of the trial, the defendant's attorney cannot question or cross-examine the physician to clarify any response made by the physician, so it is extremely important that the physician take the time to understand the plaintiff's attorney's questions and then formulate and give a complete response so that the jury can understand the physician's position on the matter in question. Do not leave any questionable areas unexplained, because at this stage, the plaintiff's attorney is trying to establish points that are beneficial to their case. If the plaintiff's attorney asks a series of short questions that only require a yes-or-no answer and do not seem to give the physician time to explain his or her position completely, it is very important that the physician break such a sequence and take as much time as needed to clarify the answer. The physician can simply say to the plaintiff's attorney, “Sir or Madam, I did not finish my response to the last question.” During the testimony phase, the plaintiff's attorney usually employs one of four basic forms of attack on the defendant's reputation. The plaintiff may challenge the physician's competence; accuse the physician of being careless; allege that the physician is not compassionate but rather an indifferent, uncaring, wealthy person; and finally, attack the physician's credibility, especially if records were altered or if the physician contradicted testimony given under oath during the deposition. This is why it is so important for a physician to know exactly what he or she said in the deposition. The physician should outline the deposition, noting significant facts and any figures he or she might have used. Again, as in preparation for the deposition, the physician should have a thorough knowledge of the office and hospital records, including any rules, protocols, or policies that have any bearing on the case. Ignorance of any of these rules, protocols, policies, or records may provide all the evidence a plaintiff's attorney needs to raise a question of the physician's competence. If at the end of the presentation, the plaintiff has not produced sufficient evidence to establish the cause-and-effect relationship, the defense may request and be granted a directed verdict, and the suit is over. Next, through testimony, expert witnesses, and medical documents, the defense tries to prove that the allegation of negligence is false and that the injuries were not a direct result of the defendant's negligence. After the defense finishes its presentation, the plaintiff has the chance to rebut any new evidence brought out during this defense presentation. Following this, each side presents closing arguments. During this phase each side summarizes the major facts supporting their position. The plaintiff's attorney begins with his or her initial closing arguments. After the defense gives its closing arguments, the plaintiff's attorney responds with their final closing arguments. In other words, the plaintiff's attorney has the last word before the jury. Once this is completed, the judge instructs the jurors on the applicable laws for the case. The jury is then given the case and retires to the jury room to begin deliberations. The formal decision or finding made by the jury or the judge is called the verdict. Emotional Impact The practice of medicine in general, and of obstetrics and gynecology in particular, is filled with daily stresses uncommon to other professions. Physicians must deal with pain and suffering, problem patients, intense interpersonal relationships, and often-unrealistic expectations. A relatively new, but unfortunately rather frequent, stress is the threat of malpractice litigation. The stress of a malpractice suit has been equated to the stress and emotional reaction one undergoes during the loss of a loved one. Unlike other professionals, a physician is unable to accept that a malpractice suit is part of the risk of doing business owing to the very nature of the doctor–patient relationship. Part of the intense anger the physician experiences at the onset of a suit arises because of the shock that

he or she experiences when a bond built on trust has been broken. The physician feels betrayed. The accusation of being incompetent or at least practicing one's job badly challenges the very core of one's professional integrity and stirs strong feelings of anger. This anger may be so intense that it spills out in all directions, frequently in an inappropriate manner that may strain many personal and professional relationships.In a study of physicians who had been sued, conducted by psychiatrist Dr. Sarah Charles, 96% of physicians reported not only emotional reactions, but also physical problems related to the suit. Eight percent had the onset of physical ailments, of which one fourth were life-threatening, such as coronaries, strokes, ulcers, and hypertension. Forty percent showed symptoms suggestive of major depressive disorders. Commonly experienced reactions were anger, mood changes, depression, tension, frustration, insomnia, fatigue, and alcohol and drug abuse. Fifteen percent of physicians lost confidence in themselves. Nineteen percent lost confidence in certain clinical situations. Fear is another common emotion experienced by sued physicians. They fear the loss of their patient's respect and confidence. They fear the loss of their own self-confidence and, in this day of excess awards, they fear the loss of financial security. The physician's reputation, ego, and self-worth are challenged. Professional training demands that the physician hold his or her reactions in check. This training becomes counterproductive at the time of a malpractice suit. The physician's marriage also is put under considerable stress. The spouse carries a great part of the emotional burden of litigation. There are many sleepless nights and interrupted plans as depositions and trials are scheduled, postponed, and rescheduled by a seemingly uncaring legal system. Mood changes and periods of irritability have to be handled by the spouse. These stressful situations can be the downfall of a weak or borderline marriage. Children are often disturbed by the knowledge that a parent is legally accused of malpractice. They may indeed feel that their parent has deceived or shamed them. Younger children especially may misunderstand the nature of the accusation and believe the allegation as fact, especially if broadcast on the radio, television, or in the newspaper. Physicians must take action to protect their mental health during this major period of stress, realizing that we live in a litigious environment and that the stress surrounding malpractice suits not only affects physicians, but also adversely impacts all professions and occupations. Physicians need to discuss their feelings with their spouse, loved ones, mentors, or counselor. Physicians and their spouses should become active in support groups. These groups have received the support of the American College of Obstetricians and Gynecologists and American Academy of Pediatrics as outlined in their December 1998 issue, “Coping with Malpractice Litigations Stress.” Such forums are not intended to provide psychiatric therapy, but rather are safe and informal groups where participants can share their common stresses and experiences in a sympathetic environment. The physician should not hesitate to seek counseling if the stress becomes so great that daily function becomes impaired. An open discussion with your children that shares their feelings and concerns as well as yours, often relieves their anxiety. This type of communication often strengthens the family relationship and ultimately results in a positive effect from the terrifying ordeal of a malpractice suit.

Section II: Principles Of Anatomy And Perioperative Considerations

Chapter 5

Surgical Anatomy of the Female Pelvis John O. L. De Lancey VULVA AND ERECTILE STRUCTURES THE PELVIC FLOOR THE PELVIC VISCERA PELVIC CONNECTIVE TISSUE AND CLEAVAGE PLANES RETROPERITONEAL SPACES AND LATERAL PELVIC WALL

THE ABDOMINAL WALL

VULVA AND ERECTILE STRUCTURES Part of "Chapter 5 - Surgical Anatomy of the Female Pelvis" The bony pelvic outlet is bordered by the ischiopubic rami anteriorly and the coccyx and sacrotuberous ligaments posteriorly. It can be divided into anterior and posterior triangles, which share a common base along a line between the ischial tuberosities. The tissues filling the anterior triangle have a layered structure similar to that of the abdominal wall (Table 5.1). There is a skin and adipose layer (vulva) overlying a fascial layer (perineal membrane) that lies superficial to a muscular layer (levator ani muscles). TABLE 5.1. Layers of the Anterior Triangle of the Perineum

Skin

Subcutaneous tissue

Camper's fascia

Colles fascia

Superficial space

Clitoris and its crura

Ischiocavernous muscle

Vestibular bulb

Bulbocavernous muscle

Greater vestibular gland

Superficial transverse perineal muscle

Deep space-perineal membrane

Compressor urethrae

Compressor urethrae

Urethrovaginal sphincter

Subcutaneous Tissues of the Vulva The structures of the vulva lie on the pubic bones and extend caudally under its arch (Fig. 5.1). They consist of the mons, labia, clitoris, vestibule, and associated erectile structures and their muscles. The mons consists of hair-bearing skin over a cushion of adipose tissue that lies on the pubic bones. Extending posteriorly from the mons, the labia majora are composed of similar hair-bearing skin and adipose tissue, which contain the termination of the round ligaments of the uterus and the obliterated processus vaginalis (canal of Nuck). The round ligament can give rise to leiomyomas in this region, and the obliterated processus vaginalis can be a dilated embryonic remnant in the adult. FIGURE 5.1. External genitalia.

View Figure

The labia minora, vestibule, and glans clitoris can be seen between the two labia majora. The labia minora are hairless skin folds, each of which splits anteriorly to run over, and under, the glans of the clitoris. The more anterior folds unite to form the hood-shaped prepuce of the clitoris, whereas the posterior folds insert into the underside of the glans as the frenulum. Unlike the skin of the labia majora, the cutaneous structures of the labia minora and vestibule do not lie on an adipose layer but on a connective-tissue stratum that is loosely organized and permits mobility of the skin during intercourse. This loose attachment of the skin to underlying tissues allows the skin to be easily dissected off the underlying fascia during skinning vulvectomy in the area of the labia minora and vestibule. In the posterior lateral aspect of the vestibule, the duct of the major vestibular gland can be seen 3 to 4 mm outside the hymenal ring. The minor vestibular gland openings are found along a line extending anteriorly from this point, parallel to the hymenal ring and extending toward the urethral orifice. The urethra bulges slightly around the surrounding vestibular skin anterior to the vagina and posterior to the clitoris. Its orifice is flanked on either side by two small labia. Skene ducts open into the inner aspect of these labia and can be seen as small, punctate openings when the urethral labia are separated. Within the skin of the vulva are specialized glands that can become enlarged and thereby require surgical

removal. The holocrine sebaceous glands in the labia majora are associated with hair shafts, and in the labia minora they are freestanding. They lie close to the surface, which explains their easy recognition with minimal enlargement. In addition, lateral to the introitus and anus, there are numerous apocrine sweat glands, along with the normal eccrine sweat glands. The former structures undergo change with the menstrual cycle, having increased secretory activity in the premenstrual period. They can become chronically infected, as in hidradenitis suppurativa, or neoplastically enlarged, as in hidradenomas, both of which may require surgical therapy. The eccrine sweat glands present in the vulvar skin rarely present abnormalities, but on occasion they can form palpable masses as syringomas. The subcutaneous tissue of the labia majora is similar in composition to that of the abdominal wall. It consists of lobules of fat interlaced with connective tissue septa. Although there are no well-defined layers in the subcutaneous tissue, regional variations in the relative quantity of fat and fibrous tissue exist. The superficial region of this tissue, where fat predominates, has been called Camper's fascia, because it is on the abdomen. In this region there is a continuation of fat from the anterior abdominal wall, called the digital process of fat. In the deeper layers of the vulva there is less fat, and the interlacing fibrous connective tissue septa are much more evident than those in Camper's fascia. This more fibrous layer is called Colles fascia and is similar to Scarpa's fascia on the abdomen. Its interlacing fibrous septa of the subcutaneous tissue attach laterally to the ischiopubic rami and fuse posteriorly with the posterior edge of the perineal membrane (i.e., urogenital diaphragm). Anteriorly, however, there is no connection to the pubic rami, and this permits communication between the area deep to this layer and the abdominal wall. These fibrous attachments to the ischiopubic rami and the posterior aspect of the perineal membrane limit the spread of hematomas or infection in this compartment posterolaterally but allow spread into the abdomen. This clinical observation has led to the consideration of Colles fascia as a separate entity from the superficial Camper fascia, which lacks these connections. Superficial Compartment The space between the subcutaneous tissues and perineal membrane, which contains the clitoris, crura, vestibular bulbs, and ischiocavernous and bulbocavernous muscles, is called the superficial compartment of the perineum (Fig. 5.2). The deep compartment is the region just above the perineal membrane; it is discussed later. FIGURE 5.2. Superficial compartment and perineal membrane.

View Figure

The erectile bodies and their associated muscles within the superficial compartment lie on the caudal surface of the perineal membrane. The clitoris is composed of a midline shaft (body) capped with the

glans. This shaft lies on, and is suspended from, the pubic bones by a subcutaneous suspensory ligament. The paired crura of the clitoris bend downward from the shaft and are firmly attached to the pubic bones, continuing dorsally to lie on the inferior aspects of the pubic rami. The ischiocavernous muscles originate at the ischial tuberosities and the free surfaces of the crura, to insert on the upper crura and body of the clitoris. A few muscle fibers, called the superficial transverse perineal muscles, originate in common with the ischiocavernous muscle from the ischial tuberosity and lie medial to the perineal body. The paired vestibular bulbs lie immediately under the vestibular skin and are composed of erectile tissue. They are covered by the bulbocavernous muscles, which originate in the perineal body and lie over their lateral surfaces. These muscles, along with the ischiocavernous muscles, insert into the body of the clitoris and act to pull it downward. The Bartholin greater vestibular gland is found at the tail end of the bulb of the vestibule and is connected to the vestibular mucosa by a duct lined with squamous epithelium. The gland lies on the perineal membrane and beneath the bulbocavernous muscle. The intimate relation between the enormously vascular erectile tissue of the vestibular bulb and the Bartholin gland is responsible for the hemorrhage associated with removal of this latter structure. The perineal membrane and perineal body are important to the support of the pelvic organs. They are discussed in the section on the pelvic floor. Pudendal Nerve and Vessels The pudendal nerve is the sensory and motor nerve of the perineum. Its course and distribution in the perineum parallel the pudendal artery and veins that connect with the internal iliac vessels (Fig. 5.3). The course and division of the nerve are described with the understanding that the vascular channels parallel them. FIGURE 5.3. Pudendal nerve and vessels, with the position of the ischiorectal fossa (asterisk) and the perineal body (cross) indicated. (From: Anson BJ. An atlas of human anatomy. Philadelphia: WB Saunders, 1950, with permission.)

View Figure

The pudendal nerve arises from the sacral plexus (52–54), and the vessels originate from the anterior division of the internal iliac artery. They leave the pelvis through the greater sciatic foramen by hooking around the ischial spine and sacrospinous ligament to enter the pudendal (Alcock) canal through the lesser sciatic foramen. The nerve and vessels have three branches: the clitoral, perineal, and inferior hemorrhoidal. The clitoral branch lies on the perineal membrane along its path to supply the clitoris. The perineal branch (the largest of the three branches) enters the subcutaneous tissues of the vulva behind the perineal membrane. Here it supplies the bulbocavernous, ischiocavernous, and transverse perineal muscles. It also supplies the skin of the inner portions of the labia majora, labia minora, and vestibule. The inferior hemorrhoidal branch goes to the external anal sphincter and perianal skin. Lymphatic Drainage The pattern of the vulvar lymphatic vessels and drainage into the superficial inguinal group of lymph nodes

has been established by both injection studies and clinical observation. It is important to the treatment of vulvar malignancies; an overview of this system is provided here. This area is described and illustrated in more detail in Chapter 33. Tissues external to the hymenal ring are supplied by an anastomotic series of vessels in the superficial tissues that coalesce to a few trunks lateral to the clitoris and proceed laterally to the superficial inguinal nodes (Fig. 5.4). The vessels draining the labia majora also run in an anterior direction, lateral to those of the labia minora and vestibule. These lymphatic channels lie medial to the labiocrural fold, establishing it as the lateral border of surgical resection. FIGURE 5.4. Lymphatic drainage of the vulva and femoral triangle. Superficial inguinal nodes are shown in the right thigh, and deep inguinal nodes are shown in the left thigh. Fascia lata has been removed on the left.

View Figure

Injection studies of the urethral lymphatics have shown that lymphatic drainage of this region terminates in either the right or left inguinal nodes. The clitoris has been said to have some direct drainage to deep pelvic lymph nodes, bypassing the usual superficial nodes, but the clinical significance of this appears to be minimal. The inguinal lymph nodes are divided into two groups—the superficial and the deep nodes. There are 12 to 20 superficial nodes, and they lie in a T-shaped distribution parallel to and 1 cm below the inguinal ligament, with the stem extending down along the saphenous vein. The nodes are often divided into four quadrants with the center of the division at the saphenous opening. The vulvar drainage goes primarily to the medial nodes of the upper quadrant. These nodes lie deep in the adipose layer of the subcutaneous tissues, in the membranous layer, just superficial to the fascia lata. The large saphenous vein joins the femoral vein through the saphenous opening. Within 2 cm of the inguinal ligament, several superficial blood vessels branch from the saphenous vein and femoral artery. They include the superficial epigastric vessels that supply the subcutaneous tissues of the lower abdomen; the superficial circumflex iliac vessels that course laterally to the region of the iliac crest; and the superficial external pudendal vessels that supply the mons, labia majora, and clitoral hood. Lymphatics from the superficial nodes enter the fossa ovalis and drain into one to three deep inguinal nodes, which lie in the femoral canal of the femoral triangle. They pass through the fossa ovalis (saphenous opening) in the fascia lata that lies approximately 3 cm below the inguinal ligament, lateral to the pubic tubercle, along with the saphenous vein on its way to the femoral vein. The membranous layer of the subcutaneous tissues spans this opening as a trabeculate layer called a fascia cribrosa, pierced by lymphatics. The deep nodes are found under this fascia in the femoral triangle. The femoral triangle is the subfascial space of the upper one third of the thigh. It is bounded by the inguinal ligament, sartorius muscle, and adductor longus muscle. Its floor is formed by the pectineal, adductor longus, and iliopsoas muscles. The femoral artery bisects it vertically between the anterosuperior iliac spine and pubic tubercle. The femoral vein lies medial to the artery; the femoral nerve is lateral to it. As these vessels pass under the inguinal ligament, they carry with them an extension of the transversalis fascia, which is the extraperitoneal connective tissue deep to the rectus abdominis muscle called the femoral sheath. These sheaths extend about 2 to 3 cm below the inguinal ligament before fusing with the vascular adventitia. Besides the two parts of the femoral sheath that accompany these vessels, a third portion—the femoral canal—can be found in the space medial to the vein. The abdominal opening of this is the femoral ring. The femoral canal contains the deep inguinal lymph nodes. Lymph channels from these nodes pierce the membrane filling the femoral ring to communicate with the external iliac nodes.

Also within this region, the femoral vessels give rise to the deep external pudendal vessels. The external pudendal vessels run deep to the femoral vein over the pectineal muscle to pierce the fascia lata. Here they become subcutaneous and form anastomoses with branches of the internal pudendal vessels as well as the deep femoral and lateral circumflex femoral arteries.

THE PELVIC FLOOR Part of "Chapter 5 - Surgical Anatomy of the Female Pelvis" When humans assumed the upright posture, the opening in the bony pelvis came to lie at the bottom of the abdominopelvic cavity. This required the evolution of a supportive system to prevent the pelvic organs from being pushed downward through this opening. In the female, this system must withstand these downward forces but allow for the passage of the large and cranially dominant human fetus. The supportive system that has evolved to meet these needs consists of a fibromuscular floor that forms a shelf spanning the pelvic outlet and that contains a cleft for the birth canal and excretory drainage. A series of visceral ligaments and fasciae tethers the organs and maintains their position over the closed portions of the floor. The floor consists of the levator ani muscles and perineal membrane. The openings in these structures for parturition and elimination have required the development of ancillary fibrous elements that are concentrated over open areas in the muscular floor to support the viscera in these weak areas. This section discusses the structures of the pelvic floor; the fibrous supportive system is described in the section on the pelvic viscera and cleavage planes and fascia. Perineal Membrane (Urogenital Diaphragm) The perineal membrane forms the inferior portion of the anterior pelvic floor. It is a triangular sheet of dense, fibromuscular tissue that spans the anterior half of the pelvic outlet (see Fig. 5.2). It was previously called the urogenital diaphragm, and this change in name reflects the appreciation that it is not a two-layered structure with muscle in between, as was previously thought. It lies just caudal to the skeletal muscle of the striated urogenital sphincter (formerly the deep transverse perineal muscle). Because of the presence of the vagina, the perineal membrane cannot form a continuous sheet to close off the anterior pelvis in the female, as it does in the male. It does provide support for the posterior vaginal wall by attaching the perineal body and vagina and perineal body to the ischiopubic rami, thereby limiting their downward descent. This layer of the floor arises from the inner aspect of the inferior ischiopubic rami above the ischiocavernous muscles and the crura of the clitoris. The medial attachments of the perineal membrane are to the urethra, walls of the vagina, and perineal body. Just cephalad to the perineal membrane lie two arch-shaped muscles that begin posteriorly to arch over the urethra (Fig. 5.5). These are the compressor urethrae and the urethrovaginal sphincter. They are a part of the striated urogenital sphincter muscle in the female and are continuous with the sphincter urethrae muscle. They act to compress the distal urethra. Posteriorly, intermingled within the membrane are skeletal muscle fibers of the transverse vaginal muscle and some smooth muscle fibers. The dorsal and deep nerve and vessels of the clitoris are also found within this membrane and are described later. FIGURE 5.5. Structures visible after removal of the perineal membrane and superficial perineal muscles. (From: DeLancey, copyright 1995, with permission.)

View Figure

The primary function of the perineal membrane is related to its attachment to the vagina and perineal body. By attaching these structures to the bony pelvic outlet, the perineal membrane supports the pelvic floor against the effects of increases in intraabdominal pressure, and against the effects of gravity. The amount of downward descent that is permitted by this mechanism can be assessed by placing a finger in the rectum, hooking it forward, and pulling the perineal body downward. If the perineal membrane has been torn during parturition, then an abnormal amount of descent is detectable, and the pelvic floor sags and the introitus gapes.

Perineal Body Within the area bounded by the lower vagina, perineal skin, and anus is a mass of connective tissue called the perineal body (see Fig. 5.3). The term central tendon of the perineum has also been applied to this structure and is descriptive, suggesting its role as a central point into which many muscles insert. The perineal body is attached to the inferior pubic rami and ischial tuberosities through the perineal membrane and superficial transverse perineal muscles. Anterolaterally, it receives the insertion of the bulbocavernous muscles. On its lateral margins, the upper portions of the perineal body are connected with some fibers of the pelvic diaphragm. Posteriorly, the perineal body is indirectly attached to the coccyx by the external anal sphincter that is embedded in the perineal body, and it is attached at its other end to the coccyx. These connections anchor the perineal body and its surrounding structures to the bony pelvis and help to keep it in place. Posterior Triangle: Ischiorectal Fossa In the posterior triangle of the pelvis, the ischiorectal fossa lies between the pelvic walls and the levator ani muscles (see Fig. 5.3). It has an anterior recess that lies above the perineal membrane. It is bounded medially by the levator ani muscles and anterolaterally by the obturator internus muscle. The main portion of the fossa is lateral to the levator ani and external anal sphincter, and it has a posterior portion that extends above the gluteus maximus. Traversing this region is the pudendal neurovascular trunk. Anal Sphincters The external sphincter lies in the posterior triangle of the perineum (see Fig. 5.6). It is a single mass of muscle, which has traditionally been divided into superficial and deep portions. The subcutaneous portion lies attached to the perianal skin and forms an encircling ring around the anal canal. It is responsible for the characteristic radially oriented folds in the perianal skin. The superficial part attaches to the coccyx posteriorly and sends a few fibers into the perineal body anteriorly and forms the bulk of the anal sphincter seen separated in third-degree midline obstetric tears. The fibers of the deep part generally encircle the rectum and blend indistinguishably with the puborectalis, which forms a loop under the dorsal surface of the anorectum and which is attached anteriorly to the pubic bone (see Fig. 5.6). FIGURE 5.6. Semidiagrammatic dissection of the anorectal region in the female with the external sphincter cut in the anterior midsagittal plane and reflected posteriorly (mucosa removed). The origin of the anterior muscle-bundle is clarified and the remaining anterolateral portions of the external sphincters are interdigitated into the transverse perinei. (From: Oh C, Kark AE. Anatomy of the external anal sphincter. Br J Surg 1972;59:717–772, with permission.)

View Figure

The internal anal sphincter is a thickening in the circular smooth muscle of the anal wall. It lies just inside the external anal sphincter and is separated from it by a visible intersphincteric groove. It extends downward inside the external anal sphincter to within a few millimeters of the external sphincter's caudal extent. The internal sphincter can be identified just beneath the anal submucosa in repair of a chronic fourth-degree laceration of the perineum as a rubber white layer that is often erroneously been referred to as fascia. The longitudinal smooth muscle layer of the bowel, along with some fibers of the levator ani, separates the external and internal sphincters as they descend in the intersphincteric groove. Levator Ani and Pelvic Wall Unfortunately, the extreme abdominal pressures generated during embalming greatly distort the levator

ani muscles by forcing them downward. Most anatomy atlases therefore fail to give a true picture of the horizontal nature of this strong supportive shelf of muscle. Examination of the normal standing patient is the best way to appreciate the nature of this closure mechanism, because the lithotomy position causes some relaxation of the musculature. During routine pelvic examination of the nullipara, the effectiveness of this closure can be appreciated, because it is often difficult to insert a speculum if the muscles are contracted and not relaxed. The opening between the bones and muscles of the pelvic wall is spanned by the muscles of the pelvic diaphragm: the pubococcygeal, iliococcygeal, puborectal, and coccygeal muscles (Fig. 5.7). The most medial of these muscles is the puborectal–pubococcygeal complex. The pubococcygeal portion of these muscles has an insertion into the anococcygeal raphe and the superior surface of the coccyx, whereas the puborectal portion represents those inferior fibers that pass behind and insert into the rectum. Both portions arise from the inner surface of the pubic bones and pass the urethra without attaching to it. Some fibers attach to the lateral vaginal wall and external anal sphincter and form a sling around the rectum before returning to a similar course on the other side. The pubococcygeal portion passes posteriorly from its origin ventral to the iliococcygeal muscle, where its fibers insert between the internal and external anal sphincter muscles in the intersphincteric groove, and form a sling behind the rectum. A few fibers also run on the cephalic surface of the iliococcygeal muscle to reach the inner surface of the sacrum and the coccyx. FIGURE 5.7. Anatomy of the pelvic floor. The asterisk indicates the puborectalis portion of the pubococcygeal muscle. (From: Anson BJ. An atlas of human anatomy. Philadelphia: WB Saunders, 1950, with permission.)

View Figure

The iliococcygeal muscle arises from a fibrous band overlying the obturator internus called the arcus tendineus levatoris ani. From these broad origins, the fibers of the iliococcygeal muscle pass behind the rectum and insert into the midline anococcygeal raphe and the coccyx. The coccygeal muscle arises from the ischial spine and sacrospinous ligament to insert into the borders of the coccyx and the lowest segment of the sacrum. These muscles are covered on their superior and inferior surfaces by superior and inferior fasciae. When the levator ani muscles and their fasciae are considered together, they are called the pelvic diaphragm, not to be confused with the urogenital diaphragm (perineal membrane). The muscle fibers of the pelvic diaphragm form a broad U-shaped layer of muscle with the open end of the U directed anteriorly. The open area within the U through which the urethra, vagina, and rectum pass, is called the urogenital hiatus. The normal tone of the muscles of the pelvic diaphragm keep the base of the U pressed against the backs of the pubic bones, keeping the vagina and rectum closed. The region of the levator ani between the anus and coccyx formed by the anococcygeal raphe (see previous discussion) is clinically called the levator plate. It forms a supportive shelf on which the rectum, upper vagina, and uterus can rest. The relatively horizontal position of this shelf is determined by the anterior traction on the fibrous levator plane by the pubococcygeal and puborectal muscles and is important to vaginal and uterine support. The iliococcygeal and coccygeal muscles receive their innervation from an anterior branch of the ventral

ramus of the third and fourth sacral nerves, whereas the medial portions of the puborectal and pubococcygeal muscles are probably supplied by the pudendal nerve.

THE PELVIC VISCERA Part of "Chapter 5 - Surgical Anatomy of the Female Pelvis" This section on the pelvic viscera discusses the structure of the individual pelvic organs and considers specific aspects of their interrelations (Fig. 5.8). Those aspects of blood supply, innervation, and lymphatic drainage that are idiosyncratic to the specific pelvic viscera are covered here. However, the section on the retroperitoneum, where the overall description of these systems is given, provides the general consideration of these latter three topics. FIGURE 5.8. The pelvic viscera.

View Figure

Genital Structures Vagina The vagina is a pliable hollow viscus whose shape is determined by the structures that surround it and by its attachments to the pelvic wall. These attachments are to the lateral margins of the vagina, so that its lumen is a transverse slit, with the anterior and posterior walls in contact with one another. The lower portion of the vagina is constricted as it passes through the urogenital hiatus in the levator ani. The upper part is much more capacious. The vagina is bent at an angle of 120 degrees by the anterior traction of the levator ani muscles at the junction of the lower one third and upper two thirds of the vagina (Fig. 5.9). The cervix lies within the anterior vaginal wall, making it shorter than the posterior wall by about 3 cm. The former is about 7 to 9 cm in length, although there is great variability in this dimension.

FIGURE 5.9. Bead chain cystourethrogram with barium in the vagina showing normal vaginal axis in a patient in the standing position.

View Figure

When the lumen of the vagina is inspected through the introitus, many landmarks can be seen. The anterior and posterior walls have a midline ridge, called the anterior and posterior columns, respectively. These are caused by the impression of the urethra and bladder and the rectum on the vaginal lumen. The caudal portion of the anterior column is distinct and is called the urethral carina. The recesses in front of and behind the cervix are commonly called the anterior and posterior fornices of the vagina, and the creases along the side of the vagina, where the anterior and posterior walls meet, are called the lateral vaginal sulci. The vagina's relations to other parts of the body can be understood by dividing it into thirds. In the lower third, the vagina is fused anteriorly with the urethra, posteriorly with the perineal body, and laterally to each levator ani by the “fibers of Luschka.” In the middle third are the vesical neck and trigone anteriorly, the rectum posteriorly, and the levators laterally. In the upper third, the anterior vagina is adjacent to the bladder and ureters (which allow these latter structures to be palpated on pelvic examination), posterior to the cul-de-sac, and lateral to the cardinal ligaments of the vagina. The vaginal wall contains the same layers as all hollow viscera (i.e., mucosa, submucosa, muscularis, and adventitia). Except for the area covered by the cul-de-sac, it has no serosal covering. The mucosa is of the nonkeratinized stratified squamous type and lies on a dense, dermislike submucosa. The similarity of these layers to dermis and epidermis has resulted in their being called the “vaginal skin.” The vaginal muscularis is adherent to the submucosa, and the pattern of the muscularis is a bihelical arrangement. Outside the muscularis there is an adventitia that has varying degrees of development in different areas of the vagina. This layer is a portion of the connective tissue in the pelvis called the endopelvic fascia and has been given a separate name because of its unusual development. When it is dissected in the operating room, the muscularis is usually adherent to it, and this combination of specialized adventitia and muscularis is the surgeon's “fascia,” which might better be called the fibromuscular layer of the vagina, as Nichols suggested in Vaginal Surgery. Uterus The uterus is a fibromuscular organ whose shape, weight, and dimensions vary considerably, depending on both estrogenic stimulation and previous parturition. It has two portions, an upper muscular corpus and a lower fibrous cervix. In a woman of reproductive age, the corpus is considerably larger than the cervix, but before menarche, and after the menopause, their sizes are similar. Within the corpus, there is a triangularly shaped endometrial cavity surrounded by a thick muscular wall. That portion of the corpus that extends above the top of the endometrial cavity (i.e., above the insertions of the fallopian tubes) is called the fundus.

The muscle fibers that make up most of the uterine corpus are not arranged in a simple layered manner, as is true in the gastrointestinal tract, but are arranged in a more complex pattern. This pattern reflects the origin of the uterus from paired paramesonephric primordia, with the fibers from each half crisscrossing diagonally with those of the opposite side. The uterus is lined by a unique mucosa, the endometrium. It has both a columnar epithelium that forms glands and a specialized stroma. The superficial portion of this layer undergoes cyclic change with the menstrual cycle. Spasm of hormonally sensitive spiral arterioles that lie within the endometrium causes shedding of this layer after each cycle, but a deeper basal layer of the endometrium remains to regenerate a new lining. Separate arteries supply the basal endometrium, explaining its preservation at the time of menses. The cervix is divided into two portions: the portio vaginalis, which is that part protruding into the vagina; and the portio supravaginalis, which lies above the vagina and below the corpus. The substance of the cervical wall is made up of dense fibrous connective tissue with only a small (about 10%) amount of smooth muscle. What smooth muscle there is lies on the periphery of the cervix, connecting the myometrium with the muscle of the vaginal wall. This smooth muscle and accompanying fibrous tissue are easily dissected off the fibrous cervix and form the layer reflected during intrafascial hysterectomy. It is circularly arranged around the fibrous cervix and is the tissue into which the cardinal and uterosacral ligaments and pubocervical fascia insert. The portio vaginalis is covered by nonkeratinizing squamous epithelium. Its canal is lined by a columnar mucus-secreting epithelium that is thrown into a series of V-shaped folds that appear like the leaves of a palm and are therefore called plicae palmatae. These form compound clefts in the endocervical canal, not tubular racemose glands, as formerly thought. The upper border of the cervical canal is marked by the internal os, where the narrow cervical canal widens out into the endometrial cavity. The lower border of the canal, the external os, contains the transition from squamous epithelium of the portio vaginalis to the columnar epithelium of the endocervical canal. This occurs at a variable level relative to the os and changes with hormonal variations that occur during a woman's life. It is in this active area of cellular transition that the cervix is most susceptible to malignant transformation. There is little adventitia in the uterus, with the peritoneal serosa being directly attached to most of the corpus. The anterior portion of the uterine cervix is covered by the bladder; therefore, it has no serosa. Similarly, as discussed in the following, the broad ligament envelops the lateral aspects of the cervix and corpus; therefore, it has no serosal covering there. The posterior cervix does have a serosal covering. Adnexal Structures and Broad Ligament The fallopian tubes are paired tubular structures 7 to 12 cm in length (Fig. 5.10). Each has four recognizable portions. At the uterus, the tube passes through the cornu as an interstitial portion. On emerging from the corpus, a narrow isthmic portion begins with a narrow lumen and thick muscular wall. Proceeding toward the abdominal end, next is the ampulla, which has an expanding lumen and more convoluted mucosa. The fimbriated end of the tube has many frondlike projections to provide a wide surface for ovum pickup. The distal end of the fallopian tube is attached to the ovary by the fimbria ovarica, which is a smooth muscle band responsible for bringing the fimbria and ovary close to one another at the time of ovulation. The outer layer of the tube's muscularis is composed of longitudinal fibers; the inner layer has a circular orientation.

FIGURE 5.10. Uterine adnexa and collateral circulation of uterine and ovarian arteries. The uterine artery crosses over the ureter in the cardinal ligament and gives off cervical and vaginal branches before ascending adjacent to the wall of the uterus and anastomosing with the medial end of the ovarian artery. Note the small branch of the uterine or ovarian artery that nourishes the round ligament (Sampson artery).

View Figure

The lateral pole of the ovary is attached to the pelvic wall by the infundibulopelvic ligament and the ovarian artery and vein contained therein. Medially, it is connected to the uterus through the uteroovarian ligament. During reproductive life, it measures about 2.5 to 5 cm long, 1.5 to 3 cm thick, and 0.7 to 1.5 cm wide, varying with its state of activity or suppression, as with oral contraceptive medications. Its surface is mostly free but has an attachment to the broad ligament through the mesovarium, as discussed in the following. The ovary has a cuboidal to columnar covering and consists of a cortex and medulla. The medullary portion is primarily fibromuscular, with many blood vessels and much connective tissue. The cortex is composed of a more specialized stroma, punctuated with follicles, corpora lutea, and corpora albicantia. The round ligaments are extensions of the uterine musculature and represent the homolog of the gubernaculum testis. They begin as broad bands that arise on each lateral aspect of the anterior corpus. They assume a more rounded shape before they enter the retroperitoneal tissue, where they pass lateral to the deep inferior epigastric vessels and enter each internal inguinal ring. After traversing the inguinal canal, they exit the external ring and enter the subcutaneous tissue of the labia majora. They have little to do with uterine support. The ovaries and tubes constitute the uterine adnexa. They are covered by a specialized series of peritoneal folds called the broad ligament. During embryonic development, the paired müllerian ducts and ovaries arise from the lateral abdominopelvic walls. As they migrate toward the midline, a mesentery of peritoneum is pulled out from the pelvic wall from the cervix on up. This leaves the midline uterus connected on either side to the pelvic wall by a double layer of peritoneum. Within the upper layers of these two folds, called the broad ligaments, lie the fallopian tubes, round ligaments, and ovaries (Fig. 5.11). The cardinal and uterosacral ligaments are at the lower margin of the broad ligament. These structures are visceral ligaments; therefore, they are composed of varying amounts of smooth muscle, vessels, connective tissue, and other structures. They are not the pure ligaments associated with joints in the skeleton.

FIGURE 5.11. Composition of the broad ligament.

View Figure

The ovary, tube, and round ligament each have their own separate mesentery, called the mesovarium, mesosalpinx, and mesoteres, respectively. These are arranged in a constant pattern, with the round ligament placed ventrally, where it exits the pelvis through the inguinal ligament, and the ovary placed dorsally. The tube is in the middle and is the most cephalic of the three structures. At the lateral end of the fallopian tube and ovary, the broad ligament ends where the infundibulopelvic ligament blends with the pelvic wall. The cardinal ligaments lie at the base of the broad ligament and are described under the section on supportive tissues and cleavage planes. Blood Supply and Lymphatics of the Genital Tract The blood supply to the genital organs comes from the ovarian arteries and uterine and vaginal branches of the internal iliac arteries. A continuous arterial arcade connects these vessels on the lateral border of the adnexa, uterus, and vagina (see Fig. 5.10). The blood supply of the upper adnexal structures comes from the ovarian arteries that arise from the anterior surface of the aorta just below the level of the renal arteries. The accompanying plexus of veins drains into the vena cava on the right and the renal vein on the left. The arteries and veins follow a long, retroperitoneal course before reaching the cephalic end of the ovary. They pass along the mesenteric surface of the ovary to connect with the upper end of the marginal artery of the uterus. Because the ovarian artery runs along the hilum of the ovary, it not only supplies the gonad but also sends many small vessels through the mesosalpinx to supply the fallopian tube, including a prominent fimbrial branch at the lateral end of the tube. The uterine artery originates from the internal iliac artery. It usually arises independently from this source but can have a common origin with either the internal pudendal or vaginal artery. It joins the uterus near the junction of the corpus and cervix, but this position varies considerably, both with the individual and the amount of upward or downward traction placed on the uterus. Accompanying each uterine artery are several large uterine veins that drain the corpus and cervix. On arriving at the lateral border of the uterus (after passing over the ureter and giving off a small branch to this structure), the uterine artery flows into the side of the marginal artery that runs along the side of the uterus. Through this connection it sends blood both upward toward the corpus and downward to the cervix. Because the marginal artery continues along the lateral aspect of the cervix, it eventually crosses over the cervicovaginal junction and lies on the side of the vagina. The vagina receives its blood supply from a downward extension of the uterine artery along the lateral sulci of the vagina and from a vaginal branch of the internal iliac artery. These form an anastomotic arcade along the lateral aspect of the vagina at the 3- and 9-o'clock positions. Branches from these vessels also merge along the anterior and posterior vaginal walls. The distal vagina also receives a supply from the pudendal vessels, and the posterior wall has a contribution from the middle and inferior hemorrhoidal vessels. Lymphatic drainage of the upper two thirds of the vagina and uterus is primarily to the obturator and internal and external iliac nodes, and the distal-most vagina drains with the vulvar lymphatics to the inguinal nodes. In addition, some lymphatic channels from the uterine corpus extend along the round ligament to the superficial inguinal nodes, and some nodes extend posteriorly along the uterosacral

ligaments to the lateral sacral nodes. These routes of drainage are discussed more fully in the discussion of the retroperitoneal space. The lymphatic drainage of the ovary follows the ovarian vessels to the region of the lower abdominal aorta, where they drain into the lumbar chain of nodes (paraaortic nodes). The uterus receives its nerve supply from the uterovaginal plexus (Frankenh?user ganglion) that lies in the connective tissue of the cardinal ligament. Details of the organization of the pelvic innervation are contained in the section on retroperitoneal structures. Lower Urinary Tract Ureter The ureter is a tubular viscus about 25 cm long, divided into abdominal and pelvic portions of equal length. Its small lumen is surrounded by an inner longitudinal and outer circular muscle layer. In the abdomen, it lies in the extraperitoneal connective tissue on the posterior abdominal wall, crossed anteriorly by the left and right colic vessels. Its course and blood supply are described in the section on the retroperitoneum. Bladder The bladder can be divided into two portions; the dome and base (Fig. 5.12). The musculature of the spherical bladder does not lie in simple layers, as do the muscular walls of tubular viscera such as the gut and ureter. It is best described as a meshwork of intertwining muscle bundles. The musculature of the dome is relatively thin when the bladder is distended. The base of the bladder, which is thicker and varies less with distention of the dome, consists of the urinary trigone and a thickening of the detrusor, called the detrusor loop. This is a U-shaped band of musculature, open posteriorly, that forms the bladder base anterior to the intramural portion of the ureter. The trigone is made of smooth muscle that arises from the ureters that occupy two of its three corners. It continues as the muscle of the vesical neck and urethra. There it rests on the upper vagina. The shape of the bladder depends on its state of filling. When empty, it is a somewhat flattened disk, slightly concave upward. As it fills, the dome rises off the base, eventually assuming a more spherical shape. FIGURE 5.12. Lateral view of the pelvic organs showing the urethra and bladder. Inset 3: Two portions of the striated urogenital sphincter muscle, namely, the urethrovaginal sphincter and the urethral sphincter. The compressor urethra is not seen. (From: The Br?del Collection, Department of Art as Applied to Medicine, Johns Hopkins Medical Institution, Baltimore, MD, with permission.)

View Figure

The distinction between the base and dome has functional importance, because they have differing innervations. The bladder base has alpha-adrenergic receptors that contract when stimulated and thereby favor continence. The dome is responsive to beta or cholinergic stimulation, with contraction that causes bladder emptying. Anteriorly, the bladder lies against the lower abdominal wall. It lies against the pubic bones laterally and inferiorly and abuts the obturator internus and levator ani. Posteriorly, it rests against the vagina and cervix. These relations are discussed further in consideration of the pelvic planes and spaces. The blood supply of the bladder comes from the superior vesical artery, which comes off the obliterated umbilical artery and inferior vesical artery, which is either an independent branch of the internal pudendal artery or arises from the vaginal artery.

Urethra The urethral lumen begins at the internal urinary meatus and has a series of regional differences in its structure. It passes through the bladder base in an intramural portion for a little less than a centimeter. This region of the bladder, where the urethral lumen traverses the bladder base, is called the vesical neck. The urethra itself begins outside the bladder wall. In its distal two thirds it is fused with the vagina (see Fig. 5.12), with which it shares a common embryologic derivation. From the vesical neck to the perineal membrane, which starts at the junction of the middle and distal thirds of the bladder, the urethra has several layers. An outer, circularly oriented skeletal muscle layer (urogenital sphincter) mingles with some circularly oriented smooth muscle fibers. Inside this layer is a longitudinal layer of smooth muscle that surrounds a remarkably vascular submucosa and nonkeratinized squamous epithelium that responds to estrogenic stimulation. Within the submucosa is a group of tubular glands that lie on the vaginal surface of the urethra. These paraurethral (or Skene's) glands empty into the lumen at several points on the dorsal surface of the urethra, but two prominent openings on the inner aspects of the external urethral orifice can be seen when the orifice is opened. Chronic infection of these glands can lead to urethral diverticula, and obstruction of their terminal duct can result in cyst formation. Their location on the dorsal surface of the urethra reflects the distribution of the structures from which they arise. At the level of the perineal membrane, the distal portion of the urogenital sphincter begins. Here the skeletal muscle of the urethra leaves the urethral wall to form the urethrovaginal sphincter (see Fig. 5.5) and compressor urethrae (formerly called the deep transverse perineal muscle). Distal to this portion, the urethral wall is fibrous and forms a nozzle for aiming the urinary stream. The mechanical support of the vesical neck and urethra, which are so important to urinary continence, is discussed in the section of this chapter devoted to the supportive tissues of the urogenital system. The urethra receives its blood supply both from an inferior extension of the vesical vessels and from the pudendal vessels. Sigmoid Colon and Rectum The sigmoid colon begins its S-shaped curve at the pelvic brim. It has the characteristic structure of the colon, with three tenia coli lying over a circular smooth muscle layer. Unlike much of the colon, which is retroperitoneal, the sigmoid has a definite mesentery in its midportion. The length of the mesentery and the pattern of the sigmoid's curvature vary considerably. It receives its blood supply from the lowermost portion of the inferior mesenteric artery, the branches called the sigmoid arteries. As it enters the pelvis, the colon straightens its course and becomes the rectum. This portion extends from the pelvic brim until it loses its final anterior peritoneal investment below the cul-de-sac. It has two bands of smooth muscle (anterior and posterior). Its lumen has three transverse rectal folds that contain the mucosa, submucosa, and circular layers of the bowel wall. The most prominent fold, the middle one, lies anteriorly on the right about 8 cm above the anus, and it must be negotiated during high rectal examination or sigmoidoscopy. As the rectum passes posterior to the vagina, it expands into the rectal ampulla. This portion of the bowel begins under the cul-de-sac peritoneum and fills the posterior pelvis from the side. At the distal end of the rectum, the anorectal junction is bent at an angle of 90 degrees where it is pulled ventrally by the puborectalis fibers' attachment to the pubes and posteriorly by the external anal sphincter's dorsal attachment to the coccyx. Below this level, the gut is called the anus. It has many distinguishing features. There is a thickening of the circular involuntary muscle called the internal sphincter. The canal has a series of anal valves to assist in closure, and at their lower border the mucosa of the colon gives way to a transitional layer of non–hair-bearing squamous epithelium before becoming the hair-bearing perineal skin. The relations of the rectum and anus can be inferred from their course. They lie against the sacrum and levator plate posteriorly and against the vagina anteriorly. Inferiorly, each half of the levator ani abuts its lateral wall and sends fibers to mingle with the longitudinal involuntary fibers between the internal and external sphincters. Its distal terminus is surrounded by the external anal sphincter. The anorectum receives its blood supply from a number of sources (Fig. 5.13). From above, the superior rectal (hemorrhoidal) branch of the inferior mesenteric artery lies within the layers of the sigmoid mesocolon. As it reaches the beginning of the rectum, it divides into two branches and ends in the wall of the gut. A direct branch from the internal iliac artery arises from the pelvic wall on either side and supplies the rectum and ampulla above the pelvic floor. The anus and external sphincter receive their blood supply from the inferior rectal (hemorrhoidal) branch of the internal pudendal artery, which reaches the terminus of the gastrointestinal tract through the ischiorectal fossa.

FIGURE 5.13. Rectosigmoid colon and anal canal, showing collateral arterial circulation from superior hemorrhoidal (inferior mesenteric), middle hemorrhoidal (hypogastric or internal iliac), and inferior hemorrhoidal (internal pudendal) arteries.

View Figure

PELVIC CONNECTIVE TISSUE AND CLEAVAGE PLANES Part of "Chapter 5 - Surgical Anatomy of the Female Pelvis" The pelvic viscera are connected to the lateral pelvic wall by their adventitial layers and thickenings of the connective tissue that lie over the pelvic wall muscles (Fig. 5.14). These attachments, as well as the attachments of one organ to another, separate the different surgical cleavage planes from one another. These condensations of the adventitial layers of the pelvic organs have assumed supportive roles, connecting the viscera to the pelvic walls, in addition to their role in transmitting the organs' neurovascular supply from the pelvic wall. They are somewhat like a mesentery that connects the bowel, for example, to the body wall. It has a supportive function as well as a role in carrying vessels and nerves to the organ. An understanding of their disposition is important to both vaginal and abdominal surgery. FIGURE 5.14. Cross section of the pelvis showing cleavage planes.

View Figure

The tissue that connects the organs to the pelvic wall has been given the special designation of endopelvic fascia. It is not a layer similar to the layer encountered during abdominal incisions (rectus abdominis “fascia”). It is composed of blood vessels and nerves, interspersed with a supportive meshwork of irregular connective tissue containing collagen and elastin. These structures connect the muscularis of the visceral organs to pelvic wall muscles. In some areas there is considerable smooth muscle within this tissue, as is true in the area of the uterosacral ligaments. Although surgical texts often speak of this fascia as a specific structure separate from the viscera, this is not strictly true. These layers can be separated from the viscera, just as the superficial layers of the bowel wall can be artificially separated from the deeper layers, but they are not themselves separate structures. Pelvic Connective Tissue The term ligament is most familiar when it describes a dense connective tissue band that links two bones, but it also describes ridges in the peritoneum or thickenings of the endopelvic fascia. The ligaments of the genital tract are diverse. Although they share a common designation (i.e., ligament), they are composed of many types of tissue and have many different functions. Uterine Ligaments The broad ligaments are peritoneal folds that extend laterally from the uterus and cover the adnexal structures. They have no supportive function and were discussed in the section on the pelvic viscera. Within the broad ligament, beginning just caudal to the uterine arteries, there is a thickening in the endopelvic fascia that attaches the cervix and upper vagina to the pelvic side walls (Fig. 5.15), consisting of the cardinal and uterosacral ligaments (parametrium). The term uterosacral ligaments refers to that portion of this tissue that forms the medial margin of the parametrium and that borders the cul-de-sac of Douglas. The term cardinal ligament is used to refer to that portion that attaches the lateral margins of the

cervix and vagina to the pelvic walls. The cardinal and uterosacral ligaments, therefore, are simply two parts of a single body of suspensory tissue. The term parametrium refers to all of the tissue that attaches to the uterus (both cardinal and uterosacral ligaments), and the term paracolpium refers to the portion that attaches to the vagina (cardinal ligament of the vagina). The uterosacral ligament portion of the parametrium is composed predominantly of smooth muscle, the autonomic nerves of the pelvic organs, and some intermixed connective tissue and blood vessels, whereas the cardinal ligament portion consists primarily of perivascular connective tissue and the pelvic vessels. Although they are often described as extending laterally from the cervix to the pelvic wall, in the standing position they are almost vertical as one would expect for a suspensory tissue. Near the cervix, they are discrete, but they fan out in the retroperitoneal layer to have a broad, if somewhat ill-defined, area of attachment over the second, third, and fourth segments of the sacrum. These ligaments hold the cervix posteriorly in the pelvis over the levator plate of the pelvic diaphragm. FIGURE 5.15. A: Suspensory ligaments of the female genital tract seen with the bladder removed. B: Close-up of the lower portion of the middle vagina (level II) shows how the lateral attachments of the vagina result in an anterior layer under the bladder (pubocervical fascia) and a posterior layer in front of the rectum (rectovaginal fascia). The cephalic surfaces of the transected distal urethra and vagina (level III) are shown. (From: DeLancey JOL. Anatomic aspects of vaginal eversion after hysterectomy. Am J Obstet Gynecol 1992;166:1717, with permission.)

View Figure

The cardinal ligaments lie at the lower edge of the broad ligaments, between their peritoneal leaves, beginning just caudal to the uterine arteries. They attach to the cervix below the isthmus and fan out to attach to the pelvic walls over the piriformis muscle in the area of the greater sciatic foramen. Although when placed under tension they feel like ligamentous bands, they are composed simply of perivascular connective tissue and nerves that surround the uterine artery and veins. Nevertheless, these structures have considerable strength, and the lack of a separate “ligamentous band” in this area does not detract from their supportive role. They provide support not only to the cervix and uterus but also to the upper portion of the vagina (paracolpium) to keep these structures positioned posteriorly over the levator plate of the pelvic diaphragm and away from the urogenital hiatus.

Vaginal Fasciae and Attachments The attachments of the vagina to the pelvic walls are important in maintaining the pelvic organs in their normal positions. Failure of these attachments, along with damage to the levator ani muscles, result in the clinical conditions of uterine prolapse, cystocele, rectocele, and enterocele. The cervix and upper one third of the vagina are suspended within the pelvis by the downward extension of the cardinal ligaments (see Fig. 5.15). Anterior to the vagina in this area is the vesicovaginal space; posterior to it is the cul-de-sac and rectovaginal space. In its middle third, the vagina is attached laterally to the arcus tendineus fasciae pelvis. The arcus tendineus fasciae pelvis is a fibrous band that extends from its ventral attachment at the pubic bone to its dorsal attachment to the ischial spine. These lateral attachments suspend the anterior vaginal wall across the pelvis and prevent its downward descent with increases in abdominal pressure. The structural layer formed by the vaginal wall and its lateral attachments to the arcus tendineus is clinically referred to as the pubocervical fascia.

Support of the posterior vaginal wall prevents the rectum from bulging forward in the clinical condition known as rectocele. This support varies in different levels of the vagina. In the distal 2 or 3 cm of the posterior vaginal wall, attachments of the perineal body to the ischiopubic rami hold the perineal body in place and prevent protrusion of the distal rectum (Fig. 5.16). In the midvagina above this, the vaginal is attached laterally to the fascia covering the inside of the levator ani muscles (Fig. 5.17). This connection prevents the middle of the posterior vaginal wall from moving forward and downward during increases in abdominal pressure. The combination of these attachments results in a structural layer that has been referred to by the term “fascia of Denonvilliers.” Detailed histologic studies of this area, however, have failed to reveal a separate layer in the midline between the muscularis of the vagina and the rectovaginal space except in the distal vagina where the dense connective tissue of the perineal body separates these structures. FIGURE 5.16. The peripheral attachments of the perineal membrane to the ischiopubic rami and direction of tension on fibers uniting through the perineal body. (From: DeLancey JOL. Structural anatomy of the posterior compartment as it relates to rectocele. Am J Obstet Gynecol 1999;180:815–823, with permission.)

View Figure

View Figure

Urethral Supports

FIGURE 5.17. Lateral view of the pelvic organs after removal of the left ischial bone and ischial tuberosity. The bladder, vagina, and cervix have been cut in the sagittal plane to reveal their lumens. The rectum has been left intact. A strip of the posterior/lateral vaginal wall with its attached endopelvic fascia are shown indicating their position relative to the levator ani muscle and this fascia's course and attachment. The two portions of the levator ani muscle (puborectalis and iliococcygeus) are visible. The ischial spine and the intact sacrospinous ligament are above the level of the removed ischial tuberosity. The left half of the perineal membrane (urogenital diaphragm) is shown just caudal to the puborectalis portion of the levator ani muscle after its detachment from the inferior pubic ramus that has been removed. (From: DeLancey JOL. Structural anatomy of the posterior compartment as it relates to rectocele. Am J Obstet Gynecol 1999;180:815–823, with permission.)

The support of the proximal urethra is important in the maintenance of urinary continence during times of increased abdominal pressure. The distal portion of the urethra is inseparable from the vagina, because of their common embryologic derivation. These tissues are fixed firmly in position by connections of the periurethral tissues and vagina to the pubic bones through the perineal membrane (Fig. 5.18). A hammocklike layer composed of the endopelvic fascia and anterior vaginal wall provides the support of the proximal urethra. This layer is stabilized by its lateral attachments both to the arcus tendineus fasciae pelvis and the medial margin of the levator ani muscles. The arcus tendineus fasciae pelvis is a fibrous band stretched from a ventral attachment at the lower portion of the pubic bones about 1 cm above the lower margin of the pubic bones and 1 cm from the midline to the ischial spine. The muscular attachment of the endopelvic fascia allows contraction and relaxation of the levator ani muscles to elevate the urethra and to let it descend. FIGURE 5.18. Lateral view of the urethral supportive mechanism transected just lateral to the midline. The lateral wall of the vagina and a portion of the endopelvic fascia have been removed so that one can see the deeper structures. (Redrawn from: DeLancey JOL. Structural support of the urethra as it relates to stress urinary incontinence: the hammock hypothesis. Am J Obstet Gynecol 1994;170:1713, with permission.)

View Figure

During increases in abdominal pressure, the downward force caused by increased abdominal pressure on the ventral surface of the urethra compresses the urethra closed against the hammocklike supportive layer, thereby closing the urethral lumen against the increases in intravesical pressure. The stability of the fascial layer determines the effectiveness of this closure mechanism. If the layer is unyielding, it forms a firm backstop against which the urethra can be compressed closed; however, if it is unstable, the effectiveness of this closure is compromised. Therefore, the integrity of the attachment to the arcus tendineus and the levator ani is critical to the stress continence mechanism. The muscular attachment is responsible for the voluntary control of vesical neck position visible during vaginal examination or fluoroscopy when the pelvic muscles are contracted and relaxed. Relaxation of these muscles with descent of the vesical neck is associated with the initiation of urination and contraction with arrest of the urinary stream. The limit of downward vesical neck motion is determined by the connective tissue elasticity in the attachments to the arcus tendineus fasciae pelvis. Cul-de-sacs, Cleavage Planes, and Spaces Each of the pelvic viscera can expand somewhat independently of its neighboring organs. The ability to do this comes from their relatively loose attachment to one another, which permits the bladder, for example, to expand without equally elongating the adjacent cervix. This allows the viscera to be easily separated from one another along these lines of cleavage. These surgical cleavage planes are called spaces, although they are not empty but rather are filled with fatty or areolar connective tissue. The pelvic spaces are separated from one another by the connections of the viscera to one another and to the pelvic walls. Anterior and Posterior Cul-de-sacs Properly termed the vesicouterine and rectouterine pouches, the anterior and posterior cul-de-sac separate the uterus from the bladder and rectum. The anterior cul-de-sac is a recess between the dome of the bladder and the anterior surface of the uterus

(Fig. 5.19). The peritoneum is loosely applied in the region of the anterior cul-de-sac unlike its dense attachment to the upper portions of the uterine corpus. This allows the bladder to expand without stretching its overlying peritoneum. This loose peritoneum forms the vesicouterine fold that can easily be lifted and incised to create a bladder flap during abdominal hysterectomy or Caesarean section. It is the point at which the vesicocervical space is normally accessed during abdominal surgery. FIGURE 5.19. Sagittal section from a 28-year-old cadaver showing the anterior cul-de-sac (aCDS) and the posterior cul-de-sac (pCDS). Note how the posterior cul-de-sac peritoneum lies on the vaginal wall, whereas the anterior cul-de-sac lies several centimeters from the depth of the peritoneum in this area. (Peritoneum digitally enhanced in photograph to aid visibility.) (From: DeLancey, copyright 2001, with permission.)

View Figure

The posterior cul-de-sac is bordered ventrally by the vagina anteriorly, the rectosigmoid posteriorly and the uterosacral ligaments laterally. Its peritoneum extends for approximately 4 cm along the posterior vaginal wall below the posterior vaginal fornix where the vaginal wall attaches to the cervix. This allows direct entry into the peritoneum from the vagina when performing a vaginal hysterectomy, culdocentesis, or colpotomy. The anatomy here contrasts with the anterior cul-de-sac. Anteriorly, the peritoneum lies several centimeters above the vagina where posteriorly the peritoneum covers the vagina. Keeping this anatomic difference in mind facilitates entering both the anterior and the posterior cul-de-sacs during vaginal hysterectomy. Prevesical Space The prevesical space of Retzius (see Fig. 5.14) is separated from the undersurface of the rectus abdominis muscles by the transversalis fascia and can be entered by perforating this layer. Ventrolaterally, it is bounded by the bony pelvis and the muscles of the pelvic wall; cranially, it is bounded by the abdominal wall. The proximal urethra and bladder lie in a dorsal position. The dorsolateral limit to this space is the attachment of the bladder to the cardinal ligament and the attachment of the pubocervical fascia to the arcus tendineus fasciae pelvis. These separate this space from the vesicovaginocervical space. This lateral attachment is to the arcus tendineus fasciae pelvis, which lies on the inner surface of the obturator internus and pubococcygeal and puborectal muscles. Important structures lying within this space include the dorsal clitoral vessels under the symphysis at its lower border and the obturator nerve and vessels as they enter the obturator canal. A branch to the obturator canal often comes off the external iliac artery and lies on the pubic bone; therefore, dissection in this area should be performed with care (Fig. 5.20). Lateral to the bladder and vesical neck is a dense plexus of vessels that lie at the border of the lower urinary tract. They are deep to the pubovesical muscle, and although they bleed when sutures are placed here, this venous ooze usually stops when the sutures are tied. Also within this tissue, lateral to the bladder and urethra, lie the nerves of the lower urinary tract. The upper border of the pubic bones that form the anterior surface of this region has a ridgelike fold of periosteum called the iliopectineal line. This is sometimes used to anchor sutures during urethral suspension operations.

FIGURE 5.20. Structures of the pelvic wall. (From: Anson BJ. An atlas of human anatomy. Philadelphia: WB Saunders, 1950, with permission.)

View Figure

Vesicovaginal and Vesicocervical Space The space between the lower urinary tract and the genital tract is separated into the vesicovaginal and vesicocervical spaces (see Fig. 5.14). The lower extent of the space is the junction of the proximal one third and distal two thirds of the urethra, where it fuses with the vagina, and it extends to lie under the peritoneum at the vesicocervical peritoneal reflection. It extends laterally to the pelvic side walls, separating the vesical and genital aspects of the cardinal ligaments Rectovaginal Space On the dorsal surface of the vagina lies the rectovaginal space (see Fig. 5.14). It begins at the apex of the perineal body, about 2 to 3 cm above the hymenal ring. It extends upward to the cul-de-sac and laterally around the sides of the rectum to the attachment of the rectovaginal septum to the parietal endopelvic fascia. It contains loose areolar tissue and is easily opened with finger dissection. At the level of the cervix, some fibers of the cardinal-uterosacral ligament complex extend downward behind the vagina, connecting it to the lateral walls of the rectum and then to the sacrum. These are called the rectal pillars. They separate the midline rectovaginal space in this region from the lateral pararectal spaces. These pararectal spaces allow access to the sacrospinous ligaments (mentioned afterward). They also form the lateral boundaries of the retrorectal space between the rectum and sacrum. Region of the Sacrospinous Ligament The area around the sacrospinous ligament is another region that has become more important to the gynecologist operating for problems of vaginal support. The sacrospinous ligament lies on the dorsal aspect of the coccygeal muscle (see Fig. 5.20). The rectal pillar separates it from the rectovaginal space. PAs its name implies, the sacrospinous ligament courses from the lateral aspect of the sacrum to the ischial spine. In its medial portion it fuses with the sacrotuberous ligament and is a distinct structure only laterally. It can be reached from the rectovaginal space by perforation of the rectal pillar to enter the pararectal space or by dissection directly under the enterocele peritoneum. This area is covered in more detail in Chapter 35. Many structures are near the sacrospinous ligament, and their location must be remembered during surgery in this region. The sacral plexus lies immediately next to the ligament on its cephalic border and comes to lie on its lateral surface as the nerve passes through the greater sciatic foramen. Just before its exit, the plexus gives off the pudendal nerve, which, with its accompanying vessels, passes lateral to the sacrospinous ligament at its attachment to the ischial spine. The nerve to the levator ani muscles lies on the inner surface of the coccygeal muscle in its midportion. In developing this space, the tissues that are reflected medially and cranially to gain access contain the pelvic venous plexus of the internal iliac vein, as well as the middle rectal vessels. If they are mobilized too vigorously, they can cause considerable hemorrhage.

RETROPERITONEAL SPACES AND LATERAL PELVIC WALL Part of "Chapter 5 - Surgical Anatomy of the Female Pelvis" The retroperitoneal space of the posterior abdomen, presacral space, and pelvic retroperitoneum contain the major neural, vascular, and lymphatic supply to the pelvic viscera. These areas are explored during operations to identify the ureter, interrupt the pelvic nerve supply, arrest serious pelvic hemorrhage, and remove potentially malignant lymph nodes. Because this area is free of the adhesions from serious pelvic infection or endometriosis, it can be used as a plane of dissection when the peritoneal cavity has become obliterated. The structures found in these spaces are discussed in a regional context, because that is the way they are usually approached in the operating room. Retroperitoneal Structures of the Lower Abdomen The aorta lies on the lumbar spine slightly to the left of the vena cava, which it overlies. The portion of this vessel below the renal vessels is encountered during retroperitoneal dissection to identify the paraaortic lymph nodes (Fig. 5.21). The renal blood vessels arise at the second lumbar vertebra. The ovarian vessels also arise from the anterior surface of the aorta in this region. In general, the branches of the vena cava follow those of the aorta, except for the vessels of the intestine, which flow into the portal vein, and the left ovarian vein, which empties into the renal vein on that side. FIGURE 5.21. Structures of the retroperitoneum. Note the anomalous origin of the left ovarian artery from the left renal artery rather than from the aorta. (From: The Br?del Collection, Department of Art as Applied to Medicine, Johns Hopkins Medical Institution, Baltimore, with permission.)

View Figure

Below the level of the renal vessels and just below the third portion of the duodenum, the inferior mesenteric artery arises from the anterior aorta. It gives off ascending branches of the left colic artery and continues caudally to supply the sigmoid through the three or four sigmoid arteries that lie in the sigmoid mesentery. These vessels follow the bowel as it is pulled from side to side, so that their position can vary, depending on retraction. Inferiorly, a continuation of the inferior mesenteric artery forms the superior rectal artery. This vessel crosses over the external iliac vessels to lie on the dorsum of the lower sigmoid. It supplies the rectum, as described in the section concerning that viscus. The aorta and vena cava have segmental branches that arise at each lumbar level and are called the lumbar arteries and veins. They are situated somewhat posteriorly to the aorta and vena cava and are not visible from the front. When the vessels are mobilized, as is done in excising the lymphatic tissue in this area, they come into view. At the level of the fourth lumbar vertebra (just below the umbilicus), the aorta bifurcates into the left and right common iliac arteries. After about 5 cm, the common iliac arteries (and the medially placed veins) give off the internal iliac vessels from their medial side and continue toward the inguinal ligament as the external iliac arteries. These internal iliac vessels lie within the pelvic retroperitoneal region and are discussed afterward.

The aorta and vena cava in this region are surrounded by lymph nodes on all sides. Surgeons usually refer to this lumbar chain of nodes as the paraaortic nodes, reflecting their position. They receive the drainage from the common iliac nodes and are the final drainage of the pelvic viscera. In addition, they collect the lymphatic drainage from the ovaries that follows the ovarian vessels and does not pass through the iliac nodes. The nodes of the lumbar chain extend from the right side of the vena cava to the left of the aorta and can be found both anterior and posterior to the vessels. The ureters are attached loosely to the posterior abdominal wall in this region, and when the overlying colon is mobilized, they remain on the body wall. They are crossed anteriorly by the ovarian vessels, which contribute a branch to supply the ureter. Additional blood supply to the abdominal portion comes from the renal vessels at the kidney and the common iliac artery. This region can be exposed either by a midline peritoneal incision to the left of the small bowel mesentery or, retroperitoneally, by reflection of the colon. During embryonic development, the colon and its mesentery fuse with the abdominal wall. A cleavage plane exists here that allows the colon and its vessels to be elevated to expose the structures of the posterior abdominal wall. Because the ureter and ovarian vessels originally arise in this area, they are not elevated with the colon. Presacral Space The presacral space begins below the bifurcation of the aorta and is bounded laterally by the internal iliac arteries (Fig. 5.22 and Fig. 5.23). Lying directly on the sacrum are the middle sacral artery and vein, which originate from the dorsal aspect of the aorta and vena cava (and not from the point of bifurcation, as sometimes shown). Caudal and lateral to this are the lateral sacral vessels. The venous plexus of these vessels can be extensive, and bleeding from it can be considerable. FIGURE 5.22. Presacral nerve plexus, showing passage of sympathetic trunk over bifurcation of aorta. Observe the division of the trunk into left and right presacral nerves. (Redrawn from: Curtis AH, Anson BJ, Ashley FL, et al. The anatomy of the pelvic autonomic nerves in relation to gynecology. Surg Gynecol Obstet 1942;75:743, with permission.)

View Figure

FIGURE 5.23. Nerves of the female pelvis. (From: Anson BJ. An atlas of human anatomy. Philadelphia: WB Saunders, 1950, with permission.)

View Figure

Within this area lies the most familiar part of the pelvic autonomic nervous system, the presacral nerve (superior hypogastric plexus). The autonomic nerves of the pelvic viscera can be divided into a sympathetic (thoracolumbar) and parasympathetic (craniosacral) system. The former is also called the adrenergic system, and the latter is called the cholinergic system, according to their neurotransmitters. Alpha-adrenergic stimulation causes increased urethral and vesical neck tone, and cholinergic stimulation increases contractility of the detrusor muscle. Similarly, adrenergic stimulation in the colon and rectum favors storage, and cholinergic stimulation favors evacuation. ?-adrenergic agonists, which are used for tocolysis, suggest that these influence contractility of the uterus. As is true in the male, damage to the autonomic nerves during pelvic lymphadenectomy can have a significant influence on orgasmic function in the female. How these autonomic nerves reach the organs that they innervate has surgical importance. The terminology of this area is somewhat confusing, because many authors use idiosyncratic terms. However, the structure is simple: It consists of a single ganglionic midline plexus overlying the lower aorta (superior hypogastric plexus) that splits into two trunks without ganglia (hypogastric nerves), each of which connects with a plexus of nerves and ganglia lateral to the pelvic viscera (inferior hypogastric plexus). The superior hypogastric plexus lies in the retroperitoneal connective tissue on the ventral surface of the lower aorta and receives input from the sympathetic chain ganglia through the thoracic and lumbar splanchnic nerves. It also contains important afferent pain fibers from the pelvic viscera, which makes its transection effective in primary dysmenorrhea. It passes over the bifurcation of the aorta and extends over the proximal sacrum before splitting into two hypogastric nerves that descend into the pelvis in the region of the internal iliac vessels. The hypogastric nerves end in the inferior hypogastric plexus. The hypogastric plexi are broad expansions of the hypogastric nerves. Their sympathetic fibers come from the downward extensions of the superior hypogastric plexus and pelvic splanchnic nerves from the continuation of the sympathetic chain into the pelvis. Parasympathetic fibers come from sacral segments 2 through 4 by way of the pelvic splanchnic nerves (nervi erigentes) to join these ganglia. They lie in the pelvic connective tissue of the lateral pelvic wall, lateral to the uterus and vagina. The inferior hypogastric plexus (sometimes called the pelvic plexus) is divided into three portions: the vesical plexus anteriorly, uterovaginal plexus (Frankenh?user ganglion), and the middle rectal plexus. The uterovaginal plexus contains fibers that derive from two sources. It receives sympathetic and sensory fibers from the tenth thoracic through the first lumbar spinal cord segments. The second input comes from the second, third, and fourth sacral segments and consists primarily of parasympathetic nerves that reach the inferior hypogastric plexus through the pelvic splanchnic nerves. The uterovaginal plexus lies on the dorsal (medial) surface of the uterine vessels, lateral to the sacrouterine ligaments' insertion into the uterus. It has continuations cranially along the uterus and caudally along the vagina. This latter extension contains the fibers that innervate the vestibular bulbs and clitoris. These nerves lie in the tissue just lateral to the area where the uterine artery, cardinal ligament, and uterosacral ligament pedicles are made during

a hysterectomy for benign disease, and within the tissue removed during a radical hysterectomy. The location of the sensory fibers from the uterine corpus in the superior hypogastric nerve (the presacral nerve) allows the surgeon to alleviate visceral pain from the corpus by transecting this structure. It does not provide sensory innervation to the adnexal structures or to the peritoneum and is therefore not useful for alleviating pain in those sites. Another important way in which the autonomic nervous system is involved is through damage to the inferior hypogastric plexus during radical hysterectomy. The extension of the surgical field lateral to the viscera interrupts the connection of the bladder and sometimes the rectum to their central attachments. The ovary and uterine tube receive their neural supply from the plexus of nerves that accompany the ovarian vessels and that originate in the renal plexus. These fibers originate from the tenth thoracic segment, and the parasympathetic fibers come from extensions of the vagus. As the lumbar and sacral nerves exit from the intervertebral and sacral foramina, they form the lumbar and sacral plexuses. The lumbar nerves and plexus lie deep within the psoas muscle on either side of the spine. The sacral plexus lies on the piriformis muscle, and its major branch, the sciatic nerve, leaves the pelvis through the lower part of the greater sciatic foramen. The sacral plexus supplies nerves to the muscles of the hip, pelvic diaphragm, and perineum, as well as to the lower leg (through the sciatic nerve). The femoral nerve from the lumbar plexus is primarily involved in supplying the muscles of the thigh. Pelvic Retroperitoneal Space Division of the internal and external iliac vessels occurs in the area of the sacroiliac joint. Just before passing under the inguinal ligament to become the femoral vessels, the external iliac vessels contribute the deep inferior epigastric and deep circumflex iliac arteries. There are no other major branches of the external iliac artery in this region. Internal Iliac Vessels Unlike the external iliac artery, which is constant and relatively simple in its morphology, the branching pattern of the internal iliac arteries and veins is extremely variable (Fig. 5.24 and Fig. 5.25). A description of a common variant is included here. The internal iliac artery supplies the viscera of the pelvis and many muscles of the pelvic wall and gluteal region. It usually divides into an anterior and posterior division about 3 to 4 cm after leaving the common iliac artery (Table 5.2). The vessels of the posterior division (the iliolumbar, lateral sacral, and superior gluteal) leave the internal iliac artery from its lateral surface to provide some of the blood supply to the pelvic wall and gluteal muscles. Trauma to these hidden vessels should be avoided during internal iliac artery ligation as the suture is passed around behind vessels. TABLE 5.2. Collateral Circulation After Internal Iliac Artery Ligation

Internal Iliac Systemic

Iliolumbar

Lateral sacral

Middle hemorrhoidal

Lumbar

Middle sacral

Superior hemorrhoidal

FIGURE 5.24. Arteries and veins of the pelvis. (From: Anson BJ. An atlas of human anatomy. Philadelphia: WB Saunders, 1950, with permission.)

View Figure

FIGURE 5.25. Collateral circulation of the pelvis.

View Figure

The anterior division has both parietal and visceral branches. The obturator, internal pudendal, and inferior gluteal vessels primarily supply muscles, whereas the uterine, superior vesical, vaginal (inferior vesical), and middle rectal vessels supply the pelvic organs. The internal iliac veins begin lateral and posterior to the arteries. These veins form a large and complex plexus within the pelvis, rather than having single branches, as do the arteries. They tend to be deeper in this area than the arteries, and their pattern is highly variable. Ligation of the internal iliac artery has proved helpful in the management of postpartum hemorrhage. Burchell's arteriographic studies showed that physiologically active anastomoses between the systemic and pelvic arterial supplies were immediately patent after ligation of the internal iliac artery (see Fig. 5.25). These anastomoses, shown in Table 5.2, connected the arteries of the internal iliac system with systemic blood vessels either directly from the aorta, as is true for the lumbar and middle sacral artery, or indirectly through the inferior mesenteric artery, as with the superior hemorrhoidal vessels. These in vivo pathways were quite different from the anastomoses that had previously been hypothesized on purely anatomic grounds. Pelvic Ureter The course of the ureter within the pelvis is important to gynecologic surgeons and is fully considered in Chapter 37. A few of the important anatomic landmarks are considered here (see Fig. 5.24). After passing over the bifurcation of the internal and external iliac arteries, just medial to the ovarian vessels, the ureter descends within the pelvis. Here it lies in a special connective tissue sheath that is attached to the peritoneum of the lateral pelvic wall and medial leaf of the broad ligament. This explains why the ureter still adheres to the peritoneum and does not remain laterally with the vessels when the peritoneal space is entered. The ureter crosses under the uterine artery (“water flows under the bridge”) in its course through the cardinal ligament. There is a loose areolar plane around it to allow for its peristalsis here. At this point it lies along the anterolateral surface of the cervix, usually about 1 cm from it. From there it comes to lie on the anterior vaginal wall and then proceeds for a distance of about 1.5 cm through the wall of the bladder. During its pelvic course, the ureter receives blood from the vessels that it passes, specifically the common iliac, internal iliac, uterine, and vesical arteries. Within the wall of the ureter, these vessels are connected to one another by a convoluted vessel that can be seen running longitudinally along its outer surface. Lymphatics The lymph nodes and lymphatic vessels that drain the pelvic viscera vary in their number and distribution, but they can be organized into coherent groups. Because of the extensive interconnection of the lymph nodes, spread of lymph flow, and thus malignancy, is somewhat unpredictable. Therefore, some important

generalizations about the distribution and drainage of these tissues are still helpful. Distribution of the pelvic lymph nodes is discussed further in Chapter 46 on invasive carcinoma of the cervix. Figure 46.22, Figure 46.23, Figure 46.24 and Figure 46.25 show this anatomy. The nodes of the pelvis can be divided into the external iliac, internal iliac, common iliac, medial sacral, and pararectal nodes. The medial sacral nodes are few and follow the middle sacral artery. The pararectal nodes drain the part of the rectosigmoid above the peritoneal reflection that is supplied by the superior hemorrhoidal artery. The medial and pararectal nodes are seldom involved in gynecologic disease. The internal and external iliac nodes lie next to their respective blood vessels, and both end in the common iliac chain of nodes, which then drain into the nodes along the aorta. The external iliac nodes receive the drainage from the leg through the inguinal nodes. Nodes in the external iliac group can be found lateral to the artery, between the artery and vein, and on the medial aspect of the vein. These groups are called the anterosuperior, intermediate, and posteromedial groups, respectively. They can be separated from the underlying muscular fascia and periosteum of the pelvic wall along with the vessels, thereby defining their lateral extent. Some nodes at the distal end of this chain lie in direct relation to the deep inferior epigastric vessels and are named according to these adjacent vessels. Similarly, nodes that lie at the point where the obturator nerve and vessels enter the obturator canal are called obturator nodes. The internal iliac nodes drain the pelvic viscera and receive some drainage from the gluteal region along the posterior division of the internal iliac vessels as well. These nodes lie within the adipose tissue that is interspersed among the many branches of the vessels. The largest and most numerous nodes lie on the lateral pelvic wall, but many smaller nodes lie next to the viscera themselves. These nodes are named for the organ by which they are found (e.g., parauterine). Not only is it difficult in the operating room to make some of the fine distinctions mentioned in this anatomic discussion, but also there is little clinical importance in doing so. Surgeons generally refer to those nodes that are adjacent to the external iliac artery as the external iliac group of nodes and to those next to the internal iliac artery as the internal iliac nodes. This leaves those nodes that lie between the external iliac vein and internal artery, which are called interiliac nodes. The direction of lymph flow from the uterus tends to follow its attachments, draining along the cardinal, uterosacral, and even round ligaments. This latter connection can lead to metastasis from the uterus to the superficial inguinal nodes, whereas the former connections are to the internal iliac nodes, with free communication to the external iliac nodes and sometimes to the lateral sacral nodes. The anastomotic connection of the uterine and ovarian vessels makes lymphatic connections between these two drainage systems likely, and metastasis in this direction possible. The vagina and lower urinary tract have a divided drainage. Superiorly (upper two thirds of the vagina and the bladder), drainage occurs along with the uterine lymphatics to the internal iliac nodes, whereas the lower one third of the vagina and distal urethra drain to the inguinal nodes. However, this demarcation is far from precise. The common iliac nodes can be found from the medial to the lateral border of the vessels of the same name. They continue above the pelvic vessels and occur around the aorta and the vena cava. These nodes can lie anterior, lateral, or posterior to the vessels.

THE ABDOMINAL WALL Part of "Chapter 5 - Surgical Anatomy of the Female Pelvis" Knowledge of the layered structure of the abdominal wall allows the surgeon to enter the abdominal cavity with maximum efficiency and safety. A general summary of these layers is provided in Table 5.3. The abdomen's superior border is the lower edge of the rib cage (ribs 7 through 12). Inferiorly, it ends at the iliac crests, inguinal ligaments, and pubic bones. It ends posterolaterally at the lumbar spine and its adjacent muscles. TABLE 5.3. Table of Abdominal Wall Layers

Skin

Subcutaneous layer

Camper's fascia

Scarpa's fascia

Musculoaponeurotic layer

Rectus sheath-formed by conjoined aponeuroses of the external oblique muscle

Internal oblique muscle: fused in lower abdomen

Transverse abdominal muscle

Transversalis fascia

Peritoneum

Skin and Subcutaneous Tissue The fibers in the dermal layer of the abdominal skin are oriented in a predominantly transverse direction

following a gently curving concave upward line. This predominance of transversely oriented fibers results in more tension on the skin of a vertical incision and in a wider scar. Between the skin and musculoaponeurotic layer of the abdominal wall lie the subcutaneous tissues. It is made of globules of fat held in place and supported by a series of branching fibrous septa. In the more superficial portion of the subcutaneous layer, called Camper's fascia, the fat predominates, and the fibrous tissue is less apparent. Closer to the rectus sheath, the fibrous tissue predominates relative to the fat in the region known as Scarpa's fascia. Camper's and Scarpa's fasciae are not discrete or well-defined layers but represent regions of the subcutaneum. Scarpa's fascia is best developed laterally and is not seen as a well-defined layer during vertical incisions. Musculoaponeurotic Layer Deep to the subcutaneous tissue is a layer of muscle and fibrous tissue that holds the abdominal viscera in place and controls movement of the lower torso (Fig. 5.26 and Fig. 5.27). Within this area are two groups of muscles: vertical muscles in the anterior abdominal wall and oblique flank muscles. The rectus abdominis muscle is found on either side of the midline, and the pyramidalis muscle is located just above the pubes. Lateral to these are the flank muscles: the external oblique, internal oblique, and transverse abdominal. The broad, sheetlike tendons of these muscles form aponeuroses that unite with their corresponding member of the other side, forming a dense white covering of the rectus abdominis muscle properly called the rectus sheath (rectus “fascia”). FIGURE 5.26. External oblique, internal oblique, and pyramidal muscles. (From: Kelly HA. Gynecology. New York: Appleton, 1928, with permission.)

View Figure

FIGURE 5.27. Transverse abdominal and rectoabdominal muscles. (From: Kelly HA. Gynecology. New York: Appleton, 1928, with permission.)

View Figure

P.96 Rectus Abdominis and Pyramidal Muscles Each paired rectus abdominis muscle originates from the sternum and cartilages of ribs 5 through 7 and inserts into the anterior surface of the pubic bone. Each muscle has three tendinous inscriptions. These are fibrous interruptions within the muscle that firmly attach it to the rectus abdominis sheath. In general, they are confined to the region above the umbilicus, but they can be found below it. When this happens, the rectus sheath is attached to the rectus muscle there, and these two structures become difficult to separate during a Pfannenstiel incision. The pyramidal muscles arise from the pubic bones and insert into the linea alba in an area several centimeters above the symphysis. Their development varies considerably among individuals. Their strong attachment to the midline makes separation of their attachment here difficult by blunt dissection.

Flank Muscles Lateral to the rectus abdominis muscles lie the broad, flat muscles of the flank. The aponeurotic insertions of these muscles join to form the conjoined tendon, or rectus sheath, which covers the rectus abdominis. Because of its importance, it is discussed separately subsequently. The most superficial of these muscles is the external oblique. Its fibers run obliquely anteriorly and inferiorly from their origin on the lower eight ribs and iliac crest. Unlike the external oblique muscle's fibers, which run obliquely downward, the fibers of the internal oblique muscle fan out from their origin in the anterior two thirds of the iliac crest, the lateral part of the inguinal ligament, and the thoracolumbar fascia in the lower posterior flank. In most areas, they are perpendicular to the fibers of the external oblique muscle, but in the lower abdomen, their fibers arch somewhat more caudally and run in a direction similar to those of the external oblique muscle. As the name transversus abdominis implies, the fibers of the deepest of the three layers have a primarily transverse orientation. They arise from the lower six costal cartilages, the thoracolumbar fascia, the anterior three fourths of the iliac crest, and the lateral inguinal ligament. The caudal portion of the transverse abdominal muscle is fused with the internal oblique muscle. This explains why, during transverse incisions of the lower abdomen, only two layers are discernible at the lateral portion of the incision. Although the fibers of the flank muscles are not strictly parallel to one another, their primarily transverse orientation and the transverse pull of their attached muscular fibers place vertical suture lines in the rectus sheath under more tension than transverse ones. For this reason, vertical incisions are more prone to dehiscence.

Rectus Sheath (Conjoined Tendon) The line of demarcation between the muscular and aponeurotic portions of the external oblique muscle in the lower abdomen occurs along a vertical line through the anterosuperior iliac spine (Fig. 5.28). The internal oblique and transverse abdominal muscles extend farther toward the midline, coming closest at their inferior margin, at the pubic tubercle. Because of this, fibers of the internal oblique muscle are found underneath the aponeurotic portion of the external oblique muscle during a transverse incision. In addition, it is between the internal oblique and transverse abdominal muscles that the nerves and blood vessels of the flank are found and their injury avoided.

View Figure

FIGURE 5.28. Cross section of lower abdominal wall. A: The anterior fascial sheath of the rectus muscle from external oblique (1) and split aponeurosis of internal oblique (2) muscles. The posterior sheath is formed by aponeurosis of the transverse abdominal muscle (3) and split aponeurosis of the internal oblique muscle. B: Lower portion of the abdominal wall below arcuate line (linea semicircularis) with absence of a posterior fascial sheath of the rectus muscle and all of the fascial aponeuroses (1,2,3) forming the anterior rectus muscle sheath.

In forming the rectus sheath, the conjoined aponeuroses of the flank are separable lateral to the rectus muscles but fuse near the midline. As they reach the midline, these layers lose their separate directions and fuse. Many specialized aspects of the rectus sheath are important to the surgeon. In its lower one fourth, the sheath lies entirely anterior to the rectus muscle. Above that point, it splits to lie both ventral and dorsal to it. The transition between these two arrangements occurs midway between the umbilicus and the pubes and is called the arcuate line. Cranial to this line, the midline ridge of the rectus sheath, the linea alba, unites these two layers. Sharp dissection is usually required to separate these layers during a Pfannenstiel incision. A vertical peritoneal incision cuts the posterior sheath. The lateral border of the rectus muscle is marked by the semilunar line of the rectus sheath. Above the arcuate line, this is the level at which the anterior and posterior layers of the sheath split. Below it the transversalis fascia fuses with the sheath. The semilunar line is not always where the three layers of flank muscles join. During a transverse lower abdominal incision, the external and internal oblique aponeuroses are often separable near the midline. The inguinal canal lies at the lower edge of the musculofascial layer of the abdominal wall. Through the inguinal canal, in the female, the round ligament extends to its termination in the labium majus. In addition, the ilioinguinal nerve and the genital branch of the genitofemoral nerve pass through the canal. Transversalis Fascia, Peritoneum, and Bladder Reflection Inside the muscular layers, and outside the peritoneum, lies the transversalis fascia, a layer of fibrous tissue that lines the abdominopelvic cavity. It is visible during abdominal incisions as the layer just underneath the rectus abdominis muscles suprapubically. It is separated from the peritoneum by a variable layer of adipose tissue. It is frequently incised or bluntly dissected off the bladder to take the tissues in this region “down by layers.” The peritoneum is a single layer of serosa. It is thrown into five vertical folds by underlying ligaments or vessels that converge toward the umbilicus. The single median umbilical fold is caused by the presence of the urachus (median umbilical ligament). Lateral to this are paired medial umbilical folds that are raised by the obliterated umbilical arteries that connected the internal iliac vessels to the umbilical cord in fetal life, and the corresponding lateral umbilical folds caused by the inferior epigastric arteries and veins. The reflection of the bladder onto the abdominal wall is triangular in shape, with its apex blending into the medial umbilical ligament. Because the apex is highest in the midline, incision in the peritoneum lateral to the midline is less likely to result in bladder injury. Neurovascular Supply of the Abdominal Wall

Vessels of the Abdominal Wall Knowing the location and course of the abdominal wall blood vessels helps the surgeon anticipate their location during abdominal incisions and during the insertion of laparoscopic trocars (Fig. 5.29). The blood vessels that supply the abdominal wall can be separated into those that supply the skin and subcutaneous tissues and those that supply the musculofascial layer. Although there is only one set of epigastric vessels in the subcutaneous tissues (superficial epigastric), there are both superior and inferior epigastric vessels in the musculofascial layer, so care must be taken in using these terms to avoid confusion. FIGURE 5.29. Normal variation in epigastric vessels. A, B, and C designate safe spots for laparoscopic trocar insertion. Dotted lines indicate lateral border of rectus muscle. (From: Hurd WW, Bude RO, DeLancey JOL, et al. The location of abdominal wall blood vessels in relationship to abdominal landmarks apparent at laparoscopy. Am J Obstet Gynecol 1994;171:642, with permission.)

View Figure

The superficial epigastric vessels run a diagonal course in the subcutaneum from the femoral vessels toward the umbilicus, beginning as a single artery that branches extensively as it nears the umbilicus. Its position can be anticipated midway between the skin and musculofascial layer, in a line between the palpable femoral pulse and the umbilicus. The external pudendal artery runs a diagonal course from the femoral artery medially to supply the region of the mons pubis. It has many midline branches, and bleeding in its territory of distribution is heavier than that from the abdominal subcutaneous tissues. The superficial circumflex iliac vessels proceed laterally from the femoral vessels toward the flank. The blood supply to the lower abdominal wall's musculofascial layer parallels the subcutaneous vessels. The branches of the external iliac, the inferior epigastric, and the deep circumflex iliac arteries parallel their superficial counterparts (see Fig. 5.29). The circumflex iliac artery lies between the internal oblique and transverse abdominal muscle. The inferior epigastric artery and its two veins originate lateral to the rectus muscle. They run diagonally toward the umbilicus and intersect the muscle's lateral border midway between the pubis and umbilicus. Below the point at which the vessels pass under the rectus, they are found lateral to the muscle deep to the transversalis fascia. After crossing the lateral border of the muscle, they lie on the muscle's dorsal surface, between it and the posterior rectus sheath. As the vessels enter the rectus sheath, they branch extensively, so that they no longer represent a single trunk. The angle between the vessel and the border of the rectus muscle forms the apex of the Hesselbach triangle (inguinal triangle), whose base is the inguinal ligament. Lateral laparoscopic trocars are placed in a region of the lower abdomen where injury to the inferior epigastric and superficial epigastric vessels can occur easily. The inferior epigastric arteries and the superficial epigastric arteries run similar courses toward the umbilicus. Knowing the average location of these blood vessels helps in choosing insertion sites that will minimize their injury and the potential hemorrhage and hematomas that this injury can cause. Just above the pubic symphysis, the vessels lie approximately 5.5 cm from the midline, whereas at the level of the umbilicus, they are 4.5 cm from the midline (Fig. 5.30). Therefore, placement either lateral or medial to the line connecting these points minimizes potential vascular injury. In addition, the location of the inferior epigastric vessel can often be seen (Fig. 5.31) by following the round ligament to its point of entry into the inguinal ring, recognizing that the vessel lies just lateral to this point.

FIGURE 5.30. Nerve supply to the abdomen. Right: Deep innervation of T6BT12 to the transverse abdominal, internal oblique, and rectal muscles. Left: Superficial distribution, including cutaneous nerves, after penetration and innervation of the external oblique muscle and fascia. Innervation of the groin and thigh also is shown.

View Figure

FIGURE 5.31. Sagittal view of female pelvis, showing inguinal and femoral anatomy.

View Figure

Nerves of the Abdominal Wall The innervation of the abdominal wall (see Fig. 5.30) comes from the abdominal extension of intercostal nerves 7 through 11, subcostal nerves (T12), iliohypogastric nerves (T12 and L1), and ilioinguinal (L1) nerves. Dermatome T10 lies at the umbilicus. After giving off a lateral cutaneous branch, each intercostal nerve pierces the lateral border of the rectus sheath. There it provides a lateral branch that ends in the rectus muscle. The anterior branch then passes through the muscle and perforates the rectus sheath to supply the subcutaneous tissues and skin as the anterior cutaneous branches. Incisions along the lateral border of the rectus lead to denervation of the muscle, which can render it atrophic and weaken the abdominal wall. Elevation of the rectus sheath off the muscle during the Pfannenstiel incision stretches the perforating nerve, which is sometimes ligated to provide hemostasis from the accompanying artery. This may leave an area of cutaneous anesthesia. The iliohypogastric and ilioinguinal nerves pass medial to the anterosuperior iliac spine in the abdominal

wall. The former supplies the skin of the suprapubic area. The latter supplies the lower abdominal wall, and by sending a branch through the inguinal canal, it supplies the upper portions of the labia majora and medial portions of the thigh. These nerves can be entrapped in the lateral closure of a transverse incision and may lead to chronic pain syndromes. The genitofemoral (L1 and L2) and femorocutaneous (L2 and L3) nerves can be injured during gynecologic surgery. The genitofemoral nerve lies on the psoas muscle (see Fig. 5.31), where pressure from a retractor can damage it and lead to anesthesia in the medial thigh and lateral labia. The femoral cutaneous nerve can be compressed either by a retractor blade lateral to the psoas or by too much flexion of the hip in the lithotomy position, causing anesthesia over the anterior thigh.

Chapter 6

Preoperative Care Sanford M. Markham John A. Rock The preoperative care and management of women prior to gynecologic surgery has proven to be a critical factor in achieving anticipated and successful outcomes of both emergent and scheduled gynecologic surgical procedures. Although the importance of a thorough history and physical examination remain key elements in the preoperative evaluation of all gynecologic patients, the use of “routine” preoperative laboratory evaluation and imaging procedures have undergone considerable change since the last edition of this textbook. This is because of the impact of recent patient care outcome reviews that have focused on indicated preoperative testing compared to “routine” testing. Cost effectiveness and evidence-based medicine are terms that are now commonly used by both the medical professionals and by managed care organizations (MCOs) when ordering and approving preoperative laboratory assessment and imaging tests. This approach makes essential the understanding of what constitutes necessary and essential preoperative testing based on current evidence from both prospective and retrospective studies. A thorough review of medical sources does not yet provide sufficient controlled trial data to make evidence-based decisions on most preoperative laboratory and imaging testing for gynecologic surgery. Sources such as the Cochrane Library do not currently contain controlled trial studies that relate to cost effectiveness of any gynecologic preoperative testing or procedures. However, significant other case series data do exist that support the need for minimal preoperative testing in the uncomplicated gynecologic patient as well as specific preoperative testing and imaging for the complicated case. This chapter is designed to provide to gynecologic surgeons an in-depth understanding of the essential features of preoperative care from the preoperative examination in the office, or emergency room, to the time of surgery. Included are suggestions relating to appropriate preoperative testing and evaluation based on the experience of gynecologic surgeons and anesthesiologists. Also included are accumulated data that demonstrate the benefits of preoperative evaluation to patient care. Foremost, it is essential to keep in mind that each woman must be considered individually, based on her medical findings and needs, and that no suggestions can be completely adapted to all women preparing for gynecologic surgery. IMPORTANCE OF PREOPERATIVE CARE HISTORY AND PHYSICAL EXAMINATION LABORATORY ASSESSMENT PREOPERATIVE EVALUATION: OVERVIEW PREOPERATIVE EVALUATION, UNCOMPLICATED GYNECOLOGIC PATHOLOGY, UNCOMPLICATED MEDICAL/SURGICAL STATUS PREOPERATIVE EVALUATION, COMPLICATED GYNECOLOGIC PATHOLOGY, UNCOMPLICATED MEDICAL/SURGICAL STATUS PREOPERATIVE EVALUATION, UNCOMPLICATED OR COMPLICATED GYNECOLOGIC PATHOLOGY, COMPLICATED MEDICAL/SURGICAL STATUS PREOPERATIVE MANAGEMENT AND PREPARATION PREOPERATIVE PROCEDURES IN THE OPERATING SUITE UNIVERSAL PRECAUTIONS FOR THE PREVENTION OF SEROPOSITIVITY FOR ACQUIRED IMMUNODEFICIENCY SYNDROME SUMMARY

IMPORTANCE OF PREOPERATIVE CARE Part of "Chapter 6 - Preoperative Care" Successful surgical outcomes of operative gynecologic procedures occur as the result of several factors in addition to good surgical skills and techniques. These factors include: An appropriate preoperative evaluation (the ability to accurately assess and diagnose gynecologic pathology, defects, and injury) An appropriate patient selection (the ability to determine when surgical intervention is a necessary course of action) An appropriate discussion with the patient regarding the benefits and risks of the surgery (the ability to communicate to the patient both short- and long-term complications in a manner that can be understood) An ability to work with MCO organizations in terms of obtaining preoperative approval and complying with individual health care plan guidelines. Once a gynecologic pathology or defect has been detected and surgical intervention is thought to be the appropriate course of action, surgical planning should be instituted. In nonemergent cases this planning should include a specific time for preoperative evaluation. The purpose of the preoperative evaluation is to accomplish the following tasks as described by Fischer: (a) decrease surgical morbidity; (b) minimize expensive delays and cancellations on the day of surgery; (c) evaluate and optimize patient health status; (d) facilitate the planning of anesthesia and perioperative care; (e) reduce patient anxiety through education; and (f) obtain informed consent. Although these objectives were presented in the context of accomplishing the tasks in a setting of a preoperative evaluation clinic, they are equally applicable to surgical planning in smaller communities where preoperative evaluation clinics are not yet available, provided that the major components of the operating team, specifically the gynecologist, anesthesiologist, and consultants are available. The importance of effective preoperative evaluations should not be underestimated. Many studies have repeatedly shown that preoperative patient conditions are significant predictors of postoperative morbidity. It is essential that all women undergoing preoperative assessment have a complete history and thorough physical examination as a key element in their work-up. This examination is important to determine factors that could affect surgical outcome. Where medical status questions arise that can not be answered by the gynecologist, then laboratory testing and imaging procedures, as well as consultations, become important to promote optimal outcomes. In the past, preoperative testing developed around the use of a history and physical examination. Batteries of “routine” individual and multiphasic laboratory tests and imaging procedures were used to detect subclinical or presymptomatic medical problems that might affect the outcomes of the surgical procedure. Additionally, it was felt that less than desirable outcomes of gynecologic surgical procedures could be minimized through the use of a wide battery of tests to prove or disprove normalcy prior to surgery. This reduction in less-than-desirable outcomes might also provide some legal protection. Unfortunately, the use of multiple “routine” tests has resulted in considerable additional costs for surgery and has created a problem as to what should be done when preoperative test results are found to be unexpectedly abnormal. It has been reported that data from the last two decades indicate 60% to 70% of laboratory tests ordered preoperatively are not required based on a review of the history and/or physical examination. Other studies suggest that only 1% or less of routinely ordered preoperative tests revealed abnormalities that might have influenced perioperative management. Furthermore, between 30% and 60% of all unexpected abnormalities detected by preoperative laboratory tests were not actually noted or investigated before surgery. This fact alone suggests that not only does the ordering of multiple “routine” tests not provide legal protection but quite possibly it also sets up an opportunity for increased legal liability. Finally, the cost of accomplishing these unnecessary preoperative “routine” tests adds many millions of dollars to health-care costs each year without any proven benefit to patient care. Using this information and applying it to the preoperative assessment of gynecologic surgery patients in the future, a preoperative evaluation should strive to answer the following three questions as outlined by Roizen. Is the patient in optimal health? Can, or should, the patient's physical or mental condition be improved before surgery? Does the patient have health problems or use any medications that could unexpectedly influence perioperative events? Therefore, all preoperative assessment and care should be directed toward answering these questions, using only those preoperative testing modalities that are expected to give information that leads to answers, as opposed to the random “routine” batteries of tests used in the past.

It is most important to dedicate a portion of the preoperative care time to a discussion with the patient of options for management of her gynecologic problem, including both short- and long-term potential complications. All patients must be given sufficient medical information to allow them to make an educated decision about whether to proceed with the planned surgery. Examples of gynecologic issues that require informed decisions include the recently reported significant increase in urinary incontinence following hysterectomy (60%), or the higher than previously reported failure rate of tubal sterilization (CREST [Collaborative Review of Sterilization] study: 10-year accumulative failure rate of 18.5 per 1,000), or the common recurrence of abnormal uterine bleeding leading to a subsequent hysterectomy in women with a history of abnormal uterine bleeding controlled by oral contraceptive pills who elect to undergo a tubal sterilization (relative risk 1.8). Not only is the discussion time useful in fostering a good physician–patient relationship, but it becomes extremely important if outcomes of surgery are less than expected, particularly if the discussion was documented in the patient's record.

HISTORY AND PHYSICAL EXAMINATION Part of "Chapter 6 - Preoperative Care" History Preoperative care of a patient always begins by carefully taking a complete history and doing a thorough preoperative examination. For the surgeon, the process of history and physical examination are fundamental to good surgical results. The physical and mental preparation of the patient are key to patient satisfaction after surgery. It is essential that the operating gynecologist personally take the history. This personal contact with the patient is of value to both the patient and surgeon. During this time the gynecologist can gain an individual perspective of not only the gynecologic problems but also the patient's general status. The patient can, in turn, ask direct questions relating to the surgical procedure as well as express her concerns. This interchange gives the surgeon a unique impression of the problem and allows the surgeon to better assess the pathology or defect that requires surgery. This places the gynecologist in a much better position to arrive at an accurate diagnosis, determine the essential preoperative testing that will be needed, and select the most appropriate surgery to manage the problem. Good history taking and a thorough physical examination require time and patience, neither of which is easily available given the expanded demands of gynecologic practice. However, the reward for the gynecologist who takes the time to listen to the patient and accomplish a complete physical examination is the avoidance of unnecessary surgery. Unnecessary operations, particularly for patients already troubled by some difficult problem of life, may prove to be unsuccessful in relieving the patient's symptoms and also concentrate attention unnecessarily on the pelvic organs. If the condition is not urgent, do not make a firm decision regarding recommendation for major pelvic surgery on the first consultation. Emergent and urgent conditions by necessity require a different approach of rapid assessment and action. In most gynecologic cases, however, the patient is counseled on the need for surgery only after all physical and psychological aspects of her case are thoroughly evaluated. The surgeon who does not ask questions and offer explanations, and who does not carefully consider choices, usually has less than desirable outcomes. A patient's history should be concise, but accuracy should not be sacrificed for the sake of brevity. It has become a common practice to use a preprinted form when obtaining a history. Unfortunately, no medical form is applicable to every case. Maximum efficiency can be achieved by using a standard patient-completed history form (Fig. 6.1) along with a physician-completed history and physical examination form (Fig. 6.2). The physician can use the patient-completed form to efficiently assimilate the information and direct the history and physical examination process. The physician's history and physical examination form serve as a structured reminder of the essentials of the history-taking process as well as summarize the patient's medical history to facilitate the decision to accomplish surgery. This form additionally serves as a record for Current Procedural Terminology (CPT) coding documentation, as well as for later clinical research. Experience has shown that important omissions are much less frequent when information is compiled on a standard form. Care should be taken, however, to not allow such a form to restrict the accurate recording of the present illness. Forms always can be expanded to document an event in the patient's history that might have an important bearing on the present illness. View PDF FIGURE 6.1. Patient history (Hx) form. (From: University of Iowa Hospitals and Clinics, with permission.)

FIGURE 6.2. Physician's history (Hx) and physical examination (PE) form. (From: University of Iowa Hospitals and Clinics, with permission.)

View Figure

A few points should be stressed concerning proper gynecologic history taking. The menstrual history must be accurate and detailed. The clue to a correct diagnosis of the gynecologic condition for which surgery is being considered often appears in the pattern of menstrual irregularity, whether the menstrual disturbance results from an organic lesion or a dysfunctional cause. In fact, differentiation between dysfunctional and early organic disease is one of the most common and difficult clinical distinctions that must be made prior to surgery. Accurate dates of the last and previous menstrual periods are of major importance. When there is a discrepancy between menstrual dates and pelvic findings in a patient of reproductive age, pregnancy should be suspected and a pregnancy test accomplished. A discrepancy between menstrual dates and pelvic findings in a patient with suspected pregnancy requires quantitative human chorionic gonadotropin (hCG) testing to identify appropriate progression of the pregnancy along with a pelvic ultrasound when the pelvic examination and hCG test results are at odds. Good gynecologic history taking can also provide valuable clues relating to findings in the physical examination. In a woman younger than 50 years of age with a history of maternal diethylstilbestrol (DES) exposure, attention should be paid to the potential presence of a variety of anatomical defects such as a “T-shaped” uterus, vaginal adenosis, or cervical defects (“cock's comb” cervix). A documented history of Type II herpes simplex viral infection of the lower genital tract has long-term implications for future recurrences and symptomatology. Of equal importance to postmenopausal women is obtaining an accurate date of the menopause. Some women are now cycling well into their fifties. Variations in vaginal bleeding in these patients may be of far different importance than vaginal bleeding in a woman 10 to 15 years postmenopausal. The patient's reproductive history is also of great importance, particularly a history of previous pregnancies and complications of pregnancy such as dystocia, Cesarean section, postpartum infection, abortion, urinary tract infections, excessive infant size, vaginal lacerations, deep vein thrombophlebitis, and pulmonary embolization. A well-taken marital history may reveal dyspareunia and/or unsatisfactory sexual relations, which may explain some symptoms that resemble organic pelvic disease. Because the symptoms of urinary tract disease so closely resemble those of reproductive tract disease, a complete urologic history is important along with laboratory investigation of the urinary tract before a final diagnosis is determined and a decision made regarding surgery. Too often the symptoms of urinary frequency, urgency, and dysuria are diagnosed as a mechanical support defect of the urinary bladder and treated by surgical repair with plication of the bladder neck, although the real problem of chronic urinary tract infection or neurogenic dysfunction of the bladder remains undiagnosed. Put in another way, a disorder of the lower urinary tract can produce symptoms suggestive of reproductive tract disease and vice versa. For this reason basic urologic training is advocated for every gynecologist. The gynecologist who is adept with a urologic workup, including cystoscopy and cystometrics, can evaluate a case better than one who must depend entirely on consultation and urology reports. Symptomatology of gastrointestinal tract disease can also mimic disease of the reproductive tract. Thus, a

proper gastrointestinal tract history along with laboratory investigation make a significant contribution toward arriving at a correct diagnosis and treatment plan prior to surgical intervention. Constipation, irritable bowel syndrome, colitis, Crohn's disease, and diverticulitis can cause abdominal and pelvic pain that is not dissimilar to the pain of endometriosis, pelvic adhesions, or ovarian neoplasms. Therefore, basic gastrointestinal training is also advocated for all gynecologists. The gynecologist experienced in accomplishing sigmoidoscopy is better prepared to assess gynecologic cases being considered for surgery than the gynecologist who lacks such experience. Because abnormal uterine bleeding can result from a variety of endocrine and metabolic disorders, a history of hypothyroidism, hyperprolactemia, insulin metabolism errors, as well as other endocrine and metabolic defects in the patient and in her family are of considerable importance in assessing and treating menorrhagia and metrorrhagia. A positive history followed by appropriate endocrine testing can give physicians significant insight into the etiology of the problem and allow them to institute proper management without always requiring surgery. Finally, musculoskeletal and neurologic defects of the low back, pelvis, and hips can result in pain similar to that found in gynecologic pathology and disease. Therefore, an orthopedic and neurologic history is an important addition to a thorough gynecologic history. General Health Examination Experience has shown that the gynecologist/obstetrician often is the only physician whom a patient consults, particularly if the symptoms of her problem seem to involve the reproductive tract. As such, preoperative care requires a complete physical examination and not just a focused examination on the lower abdomen and pelvis. This complete physical examination should include blood pressure assessment, weight and height measurement, temperature recording, thyroid and neck examination, auscultation of the heart and lungs, examination of the breasts, neurologic and orthopedic assessment, and examination of the abdomen and pelvis. During the physical examination particular attention should be given to evidence of abnormal sexual development; abnormal growth of hair on the face, chest, abdomen, extremities, back, and pubic regions; and to sexual ambiguity of the external and internal female genitalia. In addition, it is the responsibility of the gynecologist to carry out a critical evaluation of cardiac and pulmonary function prior to the accomplishment of any surgery. The need for additional medical and/or anesthesia consultation must be determined before surgical scheduling to complete the preoperative care assessment. In any patient who presents with gynecologic disease and has symptoms remotely suggestive of urinary tract infection and/or disease, a meticulously collected, midstream clean-catch, or catheterized urine specimen should be examined and cultures obtained. Past data have suggested that complications of a single catheterization in terms of urinary tract infection or significant bacteriuria is minimal, whereas the complication of performing a gynecologic surgery in the presence of an undetected preexisting urinary tract infection offers a somewhat higher risk. Doing a transurethral catheterization in gynecologic patients with urinary tract symptoms is not considered to be a hazard to the normal bladder and can provide valuable information for the total assessment of gynecologic symptoms. Gynecologic Examination The gynecologic examination includes a thorough inspection and palpation of the breasts, abdomen, pelvis, and rectum. Ample time should be dedicated to this portion of the preoperative evaluation because defects detected during this examination affects surgical planning. This examination should be completed by the gynecologist performing the surgery rather than by other physicians or staff. In some institutions a gynecologic team approach is used in the preoperative care process, and in such cases the operating gynecologist may not always be the preoperative evaluation gynecologist. In this situation the evaluating and the operating surgeon together must review the completed preoperative evaluation and plan, along with the patient's concerns and wishes. The operating surgeon must then make time before the surgery to meet the patient, review the plan of management with her, and respond to all of her questions. Breast Examination The breasts are inspected for symmetry, size, condition of the nipples, the presence of gross lesions, and the presence of discharge. Normal breast tissue, which feels rather shotty to the fingertips, is often erroneously suspected to be tumorous by the patient and is sometimes even misjudged by the physician who is unfamiliar with the proper method of breast palpation. The breasts are examined in both the upright and supine positions for symmetry, contour, and a palpable mass. The supine shoulder of the breast being examined should be raised slightly to bring the lateral aspect of the breast tissue level with the remaining

portion of the breast. The arm is then raised above the head to flatten the breast against the thoracic cage. This action permits easy examination of the full thickness of the breast tissue. Examination using the flat ventral surface of the fingers and palm almost always allows identification of an existent significant lesion. Any suspicious lesion is evaluated by mammography, ultrasonography, aspiration, and/or biopsy to confirm or discount the existence of a significant breast pathology. The nipples and adjacent areolar tissue are gently compressed to detect the presence of discharge or secretion (galactorrhea). Cytologic examination of breast secretions has been reported in the past to be useful in the diagnosis of very early breast carcinoma before the clinical detection of a gross lesion. Current imaging techniques, along with fine needle biopsy offer diagnostic accuracy before abnormal breast secretion is usually experienced, however. Nonetheless, the observation of bilateral secretion showing only the presence of fat cells on an unstained microscopic examination is reassuring and can be accomplished in an office setting. Minimal galactorrhea is not uncommon, particularly in parous women and those in early pregnancy. Other causes of galactorrhea must be considered and include prolactin-secreting tumors of the pituitary, dopamine-agonist medications, birth control pills, and primary hypothyroidism. In these cases galactorrhea is usually found to be bilateral. Unilateral secretion should be evaluated by placing a drop of the secretion on a slide and sending the slide to cytology for examination and diagnosis. The importance of a thorough breast examination is not only to detect a previously undiagnosed breast pathology but also to detect other medical problems (galactorrhea) that could affect the outcome of a planned gynecologic surgery. Abdominal Examination Examination of the abdomen requires both visual inspection and palpation. Percussion and auscultation also may be useful. Bulging of the flanks suggests free abdominal fluid, but thin-walled ovarian cysts and irregularly shaped uterine leiomyomas can give a similar clinical picture. Although large ovarian cysts and leiomyomas most often cause protrusion of the anterior abdominal wall, there are a number of confusing exceptions. Palpation for a fluid wave through the lateral quadrants of the abdomen is useful. Percussion for areas of flatness or tympany and for shifting dullness can help determine whether distention is due to intraperitoneal fluid or to intestinal gas. Auscultation is especially useful to differentiate between a large tumor, a distended bowel, or an advanced pregnancy as the cause of abdominal enlargement. When physical findings are conflicting or inconclusive, imaging procedures such as abdominal-pelvic ultrasound or abdominal computed tomography scan or magnetic resonance imaging are quite helpful in completing the assessment of an abnormal abdominal examination. Areas of tenderness and acute pain should be noted along with the consistency of the pain and whether the pain is experienced with palpation and/or rebound. Location may give some insight into the abdominal organ or tissues involved. Palpation tenderness is more commonly related to pathology of a specific organ, whereas rebound tenderness would suggest peritoneal involvement. Pelvis and Rectum Examination An accurate evaluation of the female reproductive tract is essential to establish the underlying cause of gynecologic symptoms. Although a detailed description of a pelvic examination is not provided in this chapter, it is important to stress a few of the steps necessary for proper evaluation of the female pelvis. Before an adequate pelvic examination can be performed, the bladder must be emptied by voiding. A clean-catch specimen is obtained for complete urinalysis and for culture and antibiotic sensitivity studies, if indicated. On the other hand, complaints of urinary incontinence requires examination with a full bladder in the lithotomy and in the erect positions to demonstrate stress incontinence of the urethral sphincter. Inspection of the vulva for gross lesions includes examination of the Bartholin and Skene glands for evidence of cyst formation and purulent exudate as sources of gynecologic infection. Particular attention is given to the mons pubis and labia majora and minora for subtle changes in skin pigmentation, for vesicle formation, and for small, raised lesions that may be evidence of viral or bacterial infection or of early neoplasia. The outlet is closely inspected for relaxation of the anterior and posterior vaginal walls and for uterine descensus. The vaginal mucosa is observed for any visible lesions, evidence of infection, and for estrogen effect. The patient is asked to bear down and cough without the use of a tenaculum to demonstrate the degree of relaxation of the anterior and posterior vaginal walls and the extent of uterine descensus. The urethra is compressed along its entire length to assess the possibility of a suburethral diverticulum, which often is manifested by a purulent discharge from the urethral meatus or a tender suburethral mass. The cervix is evaluated for abnormal gross pathology, particularly ulceration, neoplastic growths, inflammation, and abnormal discharge. A Papanicolaou smear is obtained by a combined sampling of cells taken from the portio of the cervix by means of a flat stick, and from the endocervical canal by means

of a small circular brush. This type of combined cytologic smear is extremely valuable in detecting cervical and endocervical lesions and is always a part of a complete gynecologic examination. Pelvic surgery always should be preceded by a recent cytologic study of the cervix. Patients with abnormal Papanicolaou smears showing repeated mild dysplasia or moderate to severe squamous cell dysplasia should be evaluated by colposcopy and suspicious lesions should be biopsied. Abnormal smears showing glandular cell dysplasia require both an endocervical and endometrial evaluation such as an endocervical curettage and endometrial biopsy before proceeding to pelvic surgery. A negative Papanicolaou smear does not, however, exclude the possibility of a cervical, endocervical, or endometrial neoplasm. False-negative cervical smears have been reported. Because 80% to 90% of all preclinical malignancies of the cervix demonstrate no significant gross lesion, it is impossible to be certain of the condition of the cervix without a Papanicolaou smear or a colposcopy and a colposcopically directed cervical biopsy if lesions are identified. The use of 3% acetic acid or Lugol's (strong iodine solution) staining of the cervix may be beneficial in identifying lesions for biopsy. The uterus is examined bimanually by the abdominal-vaginal route for position, size, mobility, irregularity, and tenderness to motion. Both adnexal regions are evaluated vaginally and by rectovaginal examination. The rectal examination should never be omitted from the routine pelvic examination. Rectal examinations provide information that cannot be obtained through the vaginal examination alone. The rectal examination provides insight into the competence of the anal sphincter as well the presence of lesions of the anal canal and lower rectum. The rectal and vaginal examinations together are an effective method for detecting pelvic pathology and are especially useful for evaluating the broad and uterosacral ligaments, cul-de-sac of Douglas, uterus, and adnexa. The index finger is inserted into the vagina while the middle finger is inserted into the rectum to a higher level than is possible with the vaginal index finger (see Fig. 6.3). This method offers the most effective opportunity of evaluating the ovaries, posterior cul-de-sac, and posterior aspect of the broad ligament. When pelvic findings are doubtful or inconclusive, imaging techniques may be helpful in determining the preoperative diagnosis. When imaging techniques are inconclusive or unavailable, however, a more adequate examination may be performed under general anesthesia before a final decision for or against gynecologic surgery is made. Indeed, a complete pelvic examination always should precede any gynecologic surgery, whether it be major or minor. The pelvic organ findings of this examination should be described carefully in the operative note for future reference. Suspected pelvic pathology frequently can be ruled in or out after a thorough preoperative pelvic examination, and needless laparotomy or laparoscopy can be avoided. The most common area of clinical confusion occurs in determining the presence of an ovarian cyst, which can be confused with bowel, bladder, or uterine leiomyomas. If a normal ovary is palpated and a cyst not identified by imaging techniques such as pelvic ultrasound, an unnecessary operative procedure frequently can be spared. FIGURE 6.3. Rectovaginal-abdominal examination.

View Figure

LABORATORY ASSESSMENT Part of "Chapter 6 - Preoperative Care" Prior to the era of evidence-based medicine most patients undergoing gynecologic surgery were assessed by a thorough history and physical examination followed by a battery of laboratory tests, the purpose of which was to detect a medical disease or defect that could adversely affect surgical outcome. These tests were accomplished on almost all patients, irrespective of age or concurrent medical or surgical pathologies, and might include: (a) blood count with hemoglobin and hematocrit; (b) urinalysis; (c) coagulation studies; (d) chemistry panel; (e) chest x-ray; and (f) electrocardiogram. This relatively universal use of “routine” preoperative testing has not proven to be beneficial in terms of good patient care and, additionally, it has added significantly to the cost of medical care. In one study an average of 72.5% of preoperative tests ordered by surgeons were considered unnecessary based on a review of the patient's history and physical examination. This same study suggested a medical care savings of between 4 and 10 billion dollars per year in the United States by eliminating these unnecessary tests with no adverse effect on outcome. Over the last 10 years data has surfaced allowing gynecologists to better assess needed preoperative tests from those that in the past were labeled as “routine” tests. Because of information such as this and because of pressures from MCOs, the volume of preoperative testing has been reduced. With this reduction in the ordering of unnecessary tests, however, comes a reduction in the ordering of necessary tests. This finding emphasizes the need to accurately differentiate between indicated preoperative testing and the past practice of routine blanketed testing of gynecologic patients preparing for surgery. A large portion of the remainder of this chapter focuses on the best case series data to help guide the gynecologist when ordering appropriate preoperative tests based on the individual patient's history and physical examination. Key to the understanding of appropriate preoperative testing is an understanding of: (a) the risk category of the patient; and (b) the degree of complexity of the planned surgery. The risk category of any preoperative gynecology patient may be divided into one of six different classes as established by the American Society of Anesthesiologists (Table 6.1). The degree of complexity of the planned surgery may be divided into one of three separate types (Table 6.2). Therefore, preoperative testing should be planned around the assessed risk of the patient based on her age, history and physical examination, and the degree of complexity based on the invasiveness of the proposed surgery. Therefore, current data as well as experience suggest that, rather than a routine set of tests, all preoperative testing should be inextricably related to the type and complexity of the proposed gynecologic surgery and to the presence of the confounding medical or surgical condition. TABLE 6.1. Classification of Physical Status, Established by the American Society of Anesthesiologists

Class

Description

P1

A normal healthy patient

P2

A patient with mild systemic disease

P3

A patient with severe systemic disease

P4

A patient with severe systemic disease that is a constant threat to life

P5

A moribund patient who is not expected to survive without the operation

P6

A declared brain-dead patient whose organs are being removed for donor purposes

From: ASA Manual for Anesthesia Department Organization and Management, American Society of Anesthesiologists, Park Ridge, Ill., 1995, with permission.

TABLE 6.2. Types of Surgical Procedures for Which Anesthesia May Be Administered

Type

General Definition

Special Examples

Type A

Minimally invasive procedures that have little potential to disrupt normal physiology and are associated with only rare periprocedural morbidity related to anesthesia. These procedures rarely require blood administration, invasive monitoring, and/or postoperative management in critical care setting.

Cataract extraction, diagnostic arthroscopy, postpartum tubal ligation

Type B

Moderately invasive procedures that have a modest or intermediate potential to disrupt normal physiology. These procedures may require blood administration, invasive monitoring, or postoperative management in a critical care setting.

Carotid endarterectomy, transurethral resection of the prostate, and laparoscopic cholecystectomy

Type C

Highly invasive procedures that typically produce significant disruption of normal physiology. These procedures commonly require blood administration, or postoperative management in a critical care setting.

Total hip replacement, open aortic aneurysm, aortic valve replacement, and posterior fossa craniotomy for aneurysm

From: Roizen MF, Foss JF, Fischer SP. Preoperative evaluation. In: Miller RD, ed. Anesthesia, 5th ed. Philadelphia: Churchill Livingstone, 2000:843, with permission.

Asymptomatic Patients In an extensive review of currently available evidence on the value of routine preoperative testing in healthy or asymptomatic patients, Monro (1997) found that there were no controlled trials assessing the value of basic tests previously thought to be essential in presurgical evaluation and care. These tests included chest x-ray, electrocardiogram, blood counts and hemoglobin, coagulation studies, blood chemistries, and urinalysis. The authors further noted that all currently available evidence on preoperative testing of healthy or asymptomatic patients came only from case series studies. After reviewing all of the available case series data they concluded that the power of preoperative tests to predict adverse postoperative outcomes in asymptomatic patients is either weak or nonexistent. Conclusions such as this have resulted in: (a) a marked reduction in the recommendation of routine testing; and (b) the suggestion that the amount of routine testing of preoperative healthy asymptomatic patients be related to the patient's age. Currently, in this category of patients, it has been recommended that a hemoglobin or hematocrit be accomplished on all patients over the age of 6 months; an electrocardiogram (ECG) on all patients over the age of 40; and a blood urea nitrogen (BUN) test and a glucose test on all patients over the age of 65. Additionally, a pregnancy test should be accomplished on

all reproductive-age women who are at risk of early pregnancy (sexually active, no contraception, or questionably effective contraception) (Table 6.3). TABLE 6.3. Recommended Preoperative Testing for Healthy Asymptomatic Gynecologic Patients

Above Age 6 Months

Above Age 40a

Above Age 65a

Hct or Hbg

Hct or Hbg

Hct or Hbg

ECG

ECG

BUN/glucose

All Women in Reproductive Age, Sexually Active, Questionable Contraception

Pregnancy test

a

Recommendation of MF Roizen, Preoperative evaluation. In: Miller RD, ed. Anesthesia, 5th ed. Philadelphia: Churchill Livingstone, 2000:854, with permission.

Symptomatic Patients and Patients with Medical or Surgical Pathologies or Defects Women considered for gynecologic surgery who are symptomatic and/or have other medical or surgical pathologies or defects must be considered in a different light from asymptomatic, healthy women during their preoperative evaluation and testing. Preoperative testing should be accomplished to determine the current status of medical or surgical pathologies and provide data on the potential effect that the compounding problem will have on the outcome of the proposed surgery. This would allow, if possible, for necessary medical correction or improvement of the problem prior to surgery in an attempt to minimize

adverse outcomes of surgical intervention. Despite the lack of an evidence-based, preoperative evaluation plan to guide preoperative testing, experience supports adopting a diagnosis-based preoperative testing protocol when planning any gynecologic surgery. A number of diagnosed-based or clinical condition-based protocols have been suggested. Because the diagnosed-based preoperative testing evaluation recommended by Fischer considers both clinical outcomes as well as cost-effectiveness, it is most appropriate in the preoperative evaluation of gynecologic patients who are other than healthy and asymptomatic (Table 6.4). This approach links necessary preoperative laboratory and imaging testing with concurrent medical disease, including cardiovascular, pulmonary, and endocrine pathologies, as well as with malignancies, and the use of many common drug therapies.

TABLE 6.4. Diagnosed-based Preoperative Testing

Preoperative Diagnosis

ECG

CXR

Hct/Hb

CBC

Lytes

Renal

Cardiac disease

MI history

X

±

Stable angina

X

±

CHF

X

±

HTN

X

±

Chronic atrial fib

X

Periph vascular disease

X

Valvular heart disease

X

Pulmonary disease

±

±

Xa

X

Glucose

Emphysema

X

±

X

±

Asthma

Chronic bronchitis

Diabetes

X

X

±

X

X

X

X

Hepatic disease

Infectious hepatitis

Alcohol/drug induced

Tumor infiltration

Renal disease

X

Hematological disorders

X

Coagulopathies

X

CNS disorders

Stroke

X

X

X

X

Seizures

X

X

X

X

Tumor

X

X

Vascular/aneurysms

X

Malignancy

X

X

Malignancy

X

Hyperthyroidism

X

X

X

Hypothyroidism

X

X

X

Cushing's syndrome

X

X

X

Addison's disease

X

X

X

Hyperparathyroidism

X

Hypoparathyroidism

X

Morbid obesity

X

Malabsorption/poor nutrition

X

X

X

±

X

X

X

X

X

Select drug therapies

Digoxin (Digitalis)

Anticoagulants

X

±

X

Dilantin

Phenobarbital

Diuretics

X

Steroids

X

Chemotherapy

X

Aspirin/NSAID

X

X

Theophylline

X = Obtained.

± = Consider.

a

Patients on diuretics.

b

Patients on digoxin.

c

Patients on theophyilline.

From: Fischer SP. Cost-effective preoperative evaluation and testing. Chest 1999;115:98S, with permission.

PREOPERATIVE EVALUATION: OVERVIEW Part of "Chapter 6 - Preoperative Care" The outcome of pelvic surgery depends largely on four preoperative factors. The skill and judgment of the gynecologic surgeon The surgical correctability of the gynecologic abnormality or disease process The severity, stability, and reversibility of concurrent medical/surgical pathologies The availability of experienced support professionals for consultation as indicated, specifically from anesthesia, medicine, and surgery Preoperative evaluation and care of the women undergoing gynecologic surgery must include not only a thorough physical assessment but also correction or stabilization of any concurrent medical or surgical pathologies that could adversely affect the surgical outcome. A skilled and appropriate gynecologic surgical procedure can have an undesirable outcome because of an unstable condition or incompletely prepared patient. For this reason, when the preoperative history, physical examination, or laboratory testing reveals instability of the cardiovascular, pulmonary, renal, or hematologic systems, a consultation should be considered. The intent is to achieve maximal stabilization before surgery. If the patient can not be stabilized, then delaying or postponing the surgery must be considered based on the risk and benefit of the surgical procedure for the patient. A preoperative diminished cardiac, pulmonary or renal reserve, a blood coagulopathy, or dehydration and/or hypovolemia can play a critical role in the outcome of the surgery. Particular attention must be given to the senior gynecologic patient. Not only does this group of women represent the fastest growing segment of gynecologic patients, but they also are more frequently prone to have other medical issues that can affect the outcome of gynecologic surgery. The US Department of Health and Human Services has recently noted that the US population over the age of 65 was growing faster than the population as a whole. They stated that currently 13 out of every 100 Americans are 65 years of age or older, and in 2030 this number will increase to 20 out of 100. They additionally noted that there were significantly more women than men in this older population, with women comprising 59% of the group aged 65 years and older and 71% of those aged 85 years and older. Considerable data as well as experience have shown that this group of senior gynecologic patients has significantly greater health care problems and therefore will require a greater effort in the preoperative evaluation in terms of laboratory evaluation and consultation. With this increase in concurrent medical and surgical problems, and the resulting increase in operative risks of gynecologic surgery in the senior population, it is important to note that chronologic age, by itself, is not always an accurate indicator of organ function. Atheromatous changes of the cardiovascular system are uncommon in women until well past menopause. This biologic phenomenon is just one of the many factors that promote female (compared with male) longevity. The average female life expectancy in 1997 was just under 80 years. As a consequence of increased longevity, a high percentage of women are in the postmenopausal period of their life when gynecologic disease becomes more prevalent and surgery becomes more necessary. In spite of the increased concurrent medical and surgical problems and greater operative risks, excellent surgical skills, meticulous medical control of concurrent disease, and anesthesia carefully attuned to the physiologic requirements of the senior gynecologic patient have kept the risk of surgery at a level not much greater than that for the premenopausal patient. Guidelines for the preoperative evaluation of the gynecologic patient may be divided into three separate areas. Uncomplicated gynecologic pathology with an uncomplicated medical/surgical status Complicated gynecologic pathology with uncomplicated medical/surgical status Uncomplicated or complicated gynecologic pathology with a complicated medical/surgical status The following preoperative evaluation recommendations are made based on the most current data from case series studies, evidence-based medical data, and consensus opinions of gynecologic preoperative care.

PREOPERATIVE EVALUATION, UNCOMPLICATED GYNECOLOGIC PATHOLOGY, UNCOMPLICATED MEDICAL/SURGICAL STATUS Part of "Chapter 6 - Preoperative Care" A gynecologic patient who has an uncomplicated gynecologic pathology and no concurrent medical or surgical conditions that would affect surgical outcome should undergo the following preoperative evaluation. Thorough history Complete physical examination Hematocrit or hemoglobin if over 6 months of age Electrocardiogram if over 40 years of age Blood urea nitrogen and glucose if over 65 years of age Pregnancy test if in the reproductive age, sexually active and not on contraception or if using questionably effective contraception Sexually transmitted diseases testing (chlamydia, gonococcus, syphilis, hepatitis, and human immunodeficiency virus [HIV]) with suspected or documented exposure Blood type and screen if the potential exists for more than minimal surgical blood loss

PREOPERATIVE EVALUATION, COMPLICATED GYNECOLOGIC PATHOLOGY, UNCOMPLICATED MEDICAL/SURGICAL STATUS Part of "Chapter 6 - Preoperative Care" A gynecologic patient who has a complicated gynecologic pathology and no concurrent medical or surgical conditions that would additionally affect surgical outcome should undergo the following preoperative evaluation (complicated gynecologic pathology includes past abdominal/pelvic surgery with evidence of or anticipation of pelvic adhesive disease, tumors, or cysts of a size making surgery more difficult or complicated; suspected or proven cancerous lesions; concurrent infection of the reproductive tract; active or chronic bleeding from the reproductive tract resulting in a demonstrated or highly probable hematologic instability). Thorough history Complete physical examination Laboratory testing listed under uncomplicated gynecologic pathology White blood cell (WBC) count with suspected or evidence of pelvic infection Prothrombin time (PT) and partial thromboplastin time (PTT) with hemorrhage or anemia Platelet count with hemorrhage or anemia or with a recent history of radiation or chemotherapy Liver function tests and renal tests with any suspected hepatic or renal pathology Type and cross match in an anticipated bloody surgery, otherwise only a type and screen Anesthesia consultation

PREOPERATIVE EVALUATION, UNCOMPLICATED OR COMPLICATED GYNECOLOGIC PATHOLOGY, COMPLICATED MEDICAL/SURGICAL STATUS Part of "Chapter 6 - Preoperative Care" A gynecologic patient who has either an uncomplicated or complicated gynecologic pathology and a concurrent medical or surgical problem offers the greatest risk for gynecologic surgery (concurrent medical or surgical problem includes cardiac, pulmonary, vascular, renal, intestinal, endocrine, neurologic, orthopedic pathologies as well as use of medications for these defects). These patients should undergo the following preoperative evaluation. Thorough history Complete physical examination Laboratory testing listed under uncomplicated gynecologic pathology Laboratory testing listed under complicated gynecologic pathology WBC with suspected infection outside of the pelvis or with immunosuppressive therapy, anemias, white blood cell pathologies, steroid therapy, or collagen diseases PT/PTT with any suspected or known coagulation defect, history of thrombosis or embolization, anticoagulation therapy, liver or intestinal disease Platelet count with any suspected or known platelet pathology, leukemia, or splenic defect Liver function tests, renal function tests, electrolytes, and/or blood sugars in women with renal, liver, or intestinal disease, on diuretics, with an unexplained fever, with endocrine disease, including diabetes, hypoglycemia, parathyroid disease, adrenal and pituitary disease, or with a recent history of radiation or chemotherapy Electrocardiogram with suspected or known cardiac pathology Anesthesia consultation Medicine, cardiology, pulmonary medicine, endocrinology, urology, and surgery consultation as indicated

PREOPERATIVE MANAGEMENT AND PREPARATION Part of "Chapter 6 - Preoperative Care" Preoperative care includes not only preoperative evaluation and laboratory testing, but also any medical or gynecologic management in the months preceding the surgical procedure to help achieve maximal physical status. Achieving this goal is rewarded by a less complicated surgical procedure with better outcomes. Examples of this medical or gynecologic management includes a number of options. Ovarian suppression through the use of a gonadotropin-releasing hormone agonist in the 3 months before surgery has proven to be beneficial in hysteroscopic resection of uterine submucous leiomyomas larger than 2 cm in size and in myomectomies when the uterine volume is equivalent to or larger than a 12-week pregnant uterus size. A similar suppression is also useful in decreasing the thickness of the endometrial lining in an endometrial ablation, although an endometrial suction curettage can achieve a somewhat similar result. Perioperative antibiotic treatment of postmenopausal women undergoing reparative surgery for genital prolapse can be beneficial in reducing recurrent cystitis but has not been shown to significantly alter the outcome of surgery. Use of preoperative vaginal estrogen cream starting 4 to 6 weeks before surgery may help in controlling uropathogens as well as thickening the vaginal mucosa, which results in an easier dissection of the vagina and in reducing postoperative morbidity. Routine preoperative endometrial sampling before hysterectomy has not been found to be cost effective unless there is suspicion of endometrial pathology such as manifested by abnormal perimenopausal or menopausal bleeding or the presence of abnormal glandular cells on Papanicolaou smear. In the latter case endocervical sampling additionally is necessary. Another area of preoperative planning requiring an experienced decision by the gynecologist and possibly the anesthesiologist is the regulation of medications taken by the patient prior to surgery. When and how to modify insulin in diabetic women, surgery on women taking anticoagulants, and continuation or discontinuation of birth control pills are examples of issues that must be addressed in the preoperative period and conveyed to the patient in a manner that she understands. Because of the uniqueness of each patient it is not possible to provide a table of preoperative medication management that can be applied to every patient. However, some general suggestions serve as guidelines in the preoperative management of the more common concurrent diseases. Preoperative insulin control in patients with diabetes is thought to be essential to achieve good surgical outcome. Older animal data has shown a relationship between hyperglycemia and wound healing with reduced tensile strength and wound failure. Experience has suggested a similar relationship in the gynecologic patient. It is therefore important to co-manage each diabetic woman undergoing gynecologic surgery with her primary care physician or internist to achieve this optimal control, and once achieved to continue the insulin regime right up to the time of surgery. Every attempt should be made to schedule the diabetic patient's surgery as a first morning case to minimize the time period between the patient's last oral intake and the onset of the surgical procedure. A protocol for patient preoperative management of insulin regulation may be found in Table 6.5. Patients who are scheduled for gynecologic surgery who are on coumadin (warfarin), heparin or low-molecular-weight heparin (LMWH) represent another preoperative management issue. A number of options are available for converting the patient from warfarin to heparin. Experience has suggested that stopping the warfarin 4 to 5 days prior to the planned gynecologic surgery and at the same time converting to low dose heparin 5,000 U subcutaneously every 12 hours provides satisfactory anticoagulation protection. The low-dose heparin can be continued postoperatively until the patient is back on oral feeding, at which time the warfarin can be reinitiated. It is further suggested that these women use elastic stockings or intermittent pneumatic compression (IPC) at the beginning of the surgery and continue with this mechanical support through to the point of full postoperative ambulation. TABLE 6.5. Recommendation for Preoperative Insulin Management Classic “Nontight Control” Regimen of Roizen

Day before surgery: Patient should be given nothing by mouth after midnight; a 13-ounce glass of clear orange juice should be at the bedside or in the car for emergency use. At 6 a.m. on the day of surgery, institute intravenous fluids using plastic cannulae and a solution containing 5% dextrose, infused at a rate of 125 mL/h/70 kg body weight. After institution of intravenous infusion, give one-half the usual morning insulin dose (and usual type of insulin) subcutaneously. Continue 5% dextrose solutions through operative period, giving at least 125 mL/h/70 kg body weight. In recovery room, monitor blood glucose concentration and treat on a sliding scale.

Roizen MF. Anesthetic implications of concurrent diseases. In: Miller RD, ed. Anesthesia, 5th ed. Philadelphia: Churchill Livingstone, 2000:903, with permission.

Another preoperative medication issue involves the continuation or discontinuation of oral contraceptive pills prior to gynecologic surgery. Studies in the 1970s suggested a relationship between use of preoperative oral contraceptive pills and intraoperative or postoperative venous thrombosis. For this reason it has been the practice to discontinue oral contraceptive pills 2 to 4 weeks before surgery and convert to mechanical contraception. This practice, however, is unsupported by any current prospective controlled studies and places the patient at risk for unwanted pregnancy as well as menstrual irregularities. Therefore, routine discontinuation of oral contraception prior to gynecologic surgery is not recommended, but instead mechanical venous support such as elastic stockings or intermittent pneumatic compression be used at the time of surgery. The immediate preoperative preparation of the patient includes the preoperative examination by the gynecologic surgeon and the anesthesia assessment in the anesthesia preoperative clinic or by an individual anesthesiologist. In most cases the assessment for anesthesia risks is accomplished before the day of surgery. If significant anesthesia risk is present, such as cardiovascular or pulmonary pathology, this allows time for relevant information to be obtained, additional testing accomplished, other consultation carried out, and treatment instituted in an attempt to have the patient in optimal condition on the day of surgery. In some cases involving low-risk ambulatory procedures, however, the anesthesia assessment is accomplished on the day of surgery. Experience suggests that an open dialog between the gynecologic surgeon and the anesthesiologist regarding the planned surgery is an essential element in achieving a successful outcome with the lowest patient risk. Most gynecologic surgical patients are admitted on the day of surgery. Because of this, preoperative guidelines and instructions for patient activity and actions at home are important and need to be made clear to every patient. The goal is to have the patient rested and in the optimal physical condition with an empty stomach and reduced contents in the lower gastrointestinal tract at the time of surgery. There is no evidence to support the idea that marked reduction of activity on the day before surgery is beneficial; however, it would be reasonable to recommend planning activities so that the patient is not overstressed. Food intake on the day before surgery need not be restricted except for the evening meal before the morning of surgery, which should be light and easily digestible. An overloaded intestinal tract during surgery is particularly hazardous not only because it poses an anesthetic risk but also because it increases postoperative nausea and gas formation. The patient should be instructed to not eat or drink after midnight on the evening before surgery unless the surgery is scheduled for the late afternoon. Some exceptions to this rule might occur with the taking of indicated medications with water. Such an exception should be discussed between the gynecologist and the anesthesiologist in the preoperative evaluation sessions. Women who are scheduled for a late afternoon surgery may have a light breakfast of a liquid

diet if taken no fewer than 6 hours preoperatively. Women undergoing major abdominal surgery in which bowel entry or injury is anticipated (or is a high probability) should undergo a complete bowel prep. This bowel preparation should consist of the single use of a commercially available cleansing preparation such as GoLYTELY or NuLYTELY (Braintree Laboratories, Inc., Braintree, MA). In all other major abdominal cases the lower colon should be cleansed by a preoperative enema the evening before surgery. If the colon is not completely emptied, then a repeat enema may need to be given prior to performing the operation, allowing adequate time for evacuation. The patient needs to be given careful instructions for the use of enemas at home and the possible need for a repeat enema in the hospital prior to surgery. This issue is often overlooked in preoperative care and results in a more difficult surgical procedure because of space limitation and a less comfortable patient in the postoperative period. An adequate night's rest before surgery is also important for the patient. In some cases the use of a mild sedative is advisable. Preoperative prophylactic broad-spectrum antibiotics or surgical antimicrobial prophylaxis has frequently been used in gynecologic surgery on the basis of the potential for vaginal flora contamination of the operative field and because of the close proximity of the rectum and intestinal tract. Data do not support the routine use of preoperative broad-spectrum antibiotics in uncomplicated, noninfected gynecologic surgery except in the case of vaginal hysterectomy or possibly abdominal hysterectomy. In these procedures the occurrence of postoperative cuff cellulitis and pelvic abscess has been significantly reduced with use of a preoperative antibiotic. First-, second-, or third-generation cephalosporins (e.g., cefazolin, cefotetan, or cefotaxime, 2 g i.m./i.v.) are effective as a prophylactic coverage, as are many of the semisynthetic penicillin family (e.g., ampicillin 1g i.m./i.v.) or the semisynthetic broad spectrum B-lactamase penicillin combinations (e.g., piperacillin tazobactam 3.75 g i.v. or ticarcillin clavulanate 3.1 g i.v.). When used, the prophylactic antibiotic should be given as a single dose approximately 1 to 2 hours prior to beginning surgery and may be repeated if the operation lasts longer than 3 hours or if there is significant blood loss (in excess of 1,500 mL). If the preoperative examination and testing identifies vaginal infections such as bacterial vaginosis, treatment prior to surgery with metronidazole intravaginal gel (0.75%) or clindamycin vaginal cream (2%) should be accomplished. In like manner, any sexually transmitted disease discovered during preoperative examination and testing should be fully treated prior to surgery. Infections occurring after surgery in the female reproductive tract arise from the introduction of normal vaginal flora into the surgical field. Surgery on the reproductive tract accomplished through a bacteriologically contaminated field (e.g., the vagina), seeds bacteria into the pedicles and surgical margins of pelvic tissues and provides an excellent nidus for infection in devitalized tissue beds. Therefore, pelvic surgery provides an ideal condition for aerobic (principally polymicrobial rather than monomicrobial) infections. Tissue destruction and sutures lowers the tissue oxidation-reduction (redox) potential. Lower tissue oxygen levels enhance the growth of facultative anaerobes that normally inhabit the vagina. As tissue hypoxia progresses, primary anaerobic bacteria survive and proliferate. Therefore, the usual postoperative infection in the vaginal vault, although initially polymicrobial, usually can be prevented with the use of the preoperative prophylactic antibiotic when the vaginal apex has been opened during a vaginal or abdominal hysterectomy. Although prophylactic antibiotics are effective in reducing the incidence of postoperative infectious morbidity, they should never be used as a substitute for the time-honored principles of adequate hemostasis and gentle handling of tissue. Despite Wangensteen's disparaging statement that, “antibiotics will turn a third-class surgeon into a second-class surgeon, but will never turn a second-class surgeon into a first-class surgeon,” current data suggest that even in the hands of a highly skilled surgeon prophylactic preoperative antibiotics offer improved outcomes in gynecologic pelvic surgery such as vaginal and abdominal hysterectomy.

PREOPERATIVE PROCEDURES IN THE OPERATING SUITE Part of "Chapter 6 - Preoperative Care" Just prior to surgery the patient is brought to the operating theater and transferred either directly to the operating table in the operating room or to an operating table in the anesthesia room adjoining the surgical suite. Preoperative preparation includes any trimming of pubic hair, preparation of abdominal and vaginal skin, and placement of a catheter, if indicated. Preoperative shaving of pubic and abdominal hair prior to gynecologic surgery is generally not recommended; in fact, preoperative shaving is associated with a significantly higher surgical site infection (SSI) rate, particularly if completed the night before the operation. These guidelines note studies showing SSI rates of 5.6% in patients who had hair removed by shaving compared to 0.6% in those who had no hair removed. Furthermore, shaving the night before surgery resulted in a significantly higher rate of SSI than shaving just before the operative procedure (7.1% versus 3.1%). Experience has shown that in some gynecologic procedures removal of pubic and abdominal hair is useful, and in these situations hair clipping is recommended immediately before the surgery. After hair trimming and catheterization, the patient is usually sufficiently anesthetized for a careful bimanual pelvic examination, at which time the surgeon can obtain very valuable information not easily obtainable when the patient is awake. Detection of reduced mobility, identification of cysts or masses not previously known, and determination of the position of pelvic organs not appreciated in past examinations may persuade the gynecologist to alter the planned surgical approach or incision type. After the pelvic examination, the perineum and vagina are cleansed, followed by the abdominal preparation. The pelvic cleansing should be accomplished before all pelvic or abdominal surgery. There is always the possibility that findings at the time of operation may make a total abdominal hysterectomy advisable, even when the preoperative plan did not include such an extensive procedure. It is extremely disconcerting to find that a total abdominal hysterectomy must be performed at the time of a laparotomy if the vagina is not properly prepared. For this reason it is strongly recommended that preoperative vaginal preparation be accomplished as a routine procedure. To clean the perineum and vagina, the vulva and perineum are first scrubbed by a nurse or surgical assistant with a sponge soaked in surgical soap or an iodophor (povidone-iodine) solution using a gloved hand. Prior to this preparation Kelly pads should be placed under the buttocks and low back to catch wash and preparation solutions, and the perineum and vulva washed and wiped free of gross contamination (mucus, blood, bowel content). These pads should be removed prior to draping the patient to prevent contaminated fluids and concentrated iodine wash solutions from pooling and remaining under the buttocks during the operative procedure. After the perineal and vulvar preparation, the vagina is next scrubbed with a soapy sponge held in the gloved hand. After the vaginal scrub the nurse or operative assistant inserts his or her fingers in the vagina and spreads the fingers to enlarge the vaginal outlet. At the same time the perineum is depressed to allow the soapy water or povidone-iodine solution to run out of the vagina. This solution is then flushed away with sterile water poured into the vagina. After this the remaining cleanup is accomplished with a sterile sponge on a sponge forceps, which is used several times to clean the vagina with an appropriate antiseptic solution. After the vaginal, vulvar, and perineal cleanup, the abdomen is prepared with a 5-minute scrub using a povidone-iodine or similar solution. Particular attention should be paid to cleansing the umbilicus with a Q-tip/cotton swab applicator. The surgically prepared area should extend superiorly from the inferior rib cage to the midthigh, inferiorly. The lateral margins of the skin preparation should extend to the anterior iliac crest and the anterior axillary line. The actual skin washing and preparation should be accomplished using concentric circles moving toward the periphery. To additionally protect the incision from contamination many surgeons use a clear plastic adhesive (3M Steri-Drape Ioban 2 or similar product) placed over the skin at the site of the incision. The Hospital Infection Control Practices Advisory Committee of the Centers for Disease Control and Prevention has published Guidelines for Prevention of Surgical Site Infection (1999), which are important in gynecologic surgery and should be noted (Table 6.6).

TABLE 6.6. Recommendations for Preoperative Preparation of the Patient to Prevent Surgical Site Infections

Whenever possible, identify and treat all infections remote to the surgical site before elective operation and postpone elective operations on patients with remote site infections until the infection has resolved. Do not remove hair preoperatively unless hair at or around the incision site will interfere with the operation. If hair is removed, remove immediately before the operation, preferably with electric clippers. Adequately control serum blood glucose levels in all diabetic patients and particularly avoid hyperglycemia perioperatively. Encourage tobacco cessation. At a minimum, instruct patients to abstain for at least 30 days before elective operation from smoking cigarettes, cigars, pipes, or other form of tobacco consumption. Do not withhold necessary blood products from surgical patients as a means to prevent surgical site infections. Require patients to shower or bathe with an antiseptic agent on at least the night before the operative day. Thoroughly wash and clean at and around the incision site to remove gross contamination before performing antiseptic skin preparation. Use appropriate antiseptic agent for skin preparation. Apply preoperative antiseptic skin preparation in concentric circles moving toward the periphery. The prepared area must be large enough to extend the incision or create new incisions or drain sites, if necessary. Keep preoperative hospital stay as short as possible while allowing for adequate preoperative preparation of the patient. No recommendation to taper or discontinue systemic steroid use (when medically permissible) before elective operation. (unresolved issue) No recommendation to enhance nutritional support for surgical patients solely as a means to prevent surgical site infection. (unresolved issue) No recommendation to preoperatively apply mupirocin to nares to prevent surgical site infection. (unresolved issue) No recommendation to provide measures that enhance would space oxygenation to prevent surgical site infections. (unresolved issue)

Hospital Infection Control Practices Committee. Centers for Disease Control and Prevention (CDC). Am J Infect Control 1999;27:266–267, with permission.

UNIVERSAL PRECAUTIONS FOR THE PREVENTION OF SEROPOSITIVITY FOR ACQUIRED IMMUNODEFICIENCY SYNDROME Part of "Chapter 6 - Preoperative Care" Because surgeons are at risk for acquiring seropositivity for the acquired immunodeficiency syndrome through contamination from body fluids and blood products, certain precautions should be taken at the time of surgery. In 1988, the Centers for Disease Control and Prevention published a document recommending that blood and other body fluid precautions be consistently used for all patients, regardless of their blood-borne infectious status. This extension of blood and other body fluid precautions to all patients is referred to as Universal Blood and Fluid Precautions, or simply Universal Precautions. Under universal precautions, blood and body fluids of all patients are considered potentially infectious for HIV, hepatitis B virus (HBV), and other blood-borne pathogens. Universal precautions are intended to prevent parenteral, mucous membrane, and nonintact skin exposure of the surgeon to blood-borne pathogens. Immunization with HBV vaccine is also recommended as an important adjunct to the universal precautions for surgeons who are exposed to the risks. Protective barriers reduce the risk of exposure of the surgeon's skin or mucous membrane to potentially infectious materials. In the operating room, protective barriers include gowns, gloves, masks, and protective eyewear. Masks and protective eyewear or face shields reduce the incidence of contamination of mucous membranes of the mouth, nose, and eyes. Gloves reduce the incidence of contamination of the hands, but they cannot prevent penetrating injuries such as puncture by a needle or other sharp instrument. Special care must be taken to prevent any injury by needles, scalpels, or other sharp instruments or devices. Should a contamination occur in spite of these precautions, immediately and thoroughly wash the hands and other skin surfaces that have been contaminated and institute the medical institution's policy on rapid antigen testing to determine if the surgeon should consider prophylaxis for HIV.

SUMMARY Part of "Chapter 6 - Preoperative Care" Gynecologic surgery encompasses a wide variety of surgical procedures from minimally invasive operations to radical and extensive dissections of pelvic pathology. No matter what the extent of the procedure all gynecologic surgery should be considered to have three separate parts: (a) the preoperative care and management phase; (b) the surgical procedure itself; and (c) postoperative care and the management phase. Each part is inextricably related to the others and each part is vital in the overall outcome of the gynecologic pathology being managed. Preoperative care and management sets the stage for a successful surgical procedure. Best outcomes are linked to a patient being in optimal physical and medical status prior to surgery. The preoperative process begins with an accurate diagnosis and an appropriate decision to operate. The preoperative evaluation of each gynecologic patient serves to identify other pathologies that could interfere with optimal surgical outcome and to implement corrections prior to surgery. As a result, both the gynecologic surgeon and the patient arrive in the operating room with little doubt as to the pathology, the overall medical status, and the plan for surgical correction.

Chapter 7

Postanesthesia and Postoperative Care Ira R. Horowitz Jack B. Basil The most critical period of a patient's postoperative course occurs within the first 72 hours of surgery. Precise monitoring of the cardiovascular, renal, and respiratory systems provides the most valuable information about the patient's postoperative condition. Postoperative morbidity can be decreased with an appropriate preoperative evaluation of the surgical candidate. Emphasis should be placed on identifying those patients at risk for developing venous thrombosis and administering appropriate prophylaxis. Improved nutritional status in the preoperative and postoperative periods also has been shown to improve wound healing and decrease the postoperative recovery time. VASCULAR COMPLICATIONS PULMONARY COMPLICATIONS CARE OF THE URINARY BLADDER GASTROINTESTINAL TRACT MANAGEMENT TOTAL PARENTERAL NUTRITION ENTERAL NUTRITION ROUTINE ORDERS ESTROGEN REPLACEMENT THERAPY

ACKNOWLEDGMENT

VASCULAR COMPLICATIONS Part of "Chapter 7 - Postanesthesia and Postoperative Care" Venous Thrombosis About half a million hospitalizations are associated with deep venous thrombosis (DVT) or pulmonary embolism (PE). A fatal outcome occurs in about 50,000 of these patients. The incidence of venous thrombosis in the gynecology patient is 15%, with a range of 5% to 45%, depending on the procedure performed and associated risk factors. PE is responsible for 40% of the postoperative deaths in the gynecologic patient. The sudden occurrence of respiratory distress in a postoperative patient, followed by hypotension, chest pain, cardiac arrhythmias, and death, is a complication that converts a successful operative procedure into a postoperative mortality. In only 70% of patients who die of PE is it considered in the differential diagnosis. In recent years, diagnostic studies have provided more accurate information on the frequency of this vascular complication and have identified those patients with DVT who are at risk of shedding emboli. Preoperative and postoperative prophylaxis with heparin, dextran, antiembolic stockings, and intermittent pneumatic compression devices for moderate- and high-risk patients has reduced the incidence of pulmonary emboli. Clarke-Pearson and coworkers, using univariate and regression analysis, designed a prognostic model to evaluate the risks of postoperative DVT for an individual patient. The preoperative prognostic factors identified in a prospective study of 411 gynecology patients were type of surgery, age, leg edema, nonwhite patients, severity of venous varicosities, prior radiation therapy, and prior history of DVT. Essentially all of the major factors contributing to postoperative venous thrombosis were described by Virchow more than 125 years ago. These factors include an increase in blood coagulability, venous stasis, and trauma to the vessel wall. Tissue injury activates blood coagulation by the extrinsic and intrinsic pathways (Fig. 7.1) by exposing the blood to increased levels of tissue thromboplastin (extrinsic) or to subendothelial collagen in the vessel wall, which activates factor XII (intrinsic). Venous damage is particularly prevalent in patients undergoing radical surgery in which skeletonization of the pelvic vasculature is performed. When postoperative infection and pelvic cellulitis occur, there is an acceleration of the clotting mechanism caused by the release of tissue thromboplastin (extrinsic) and the activation of factor VII (intrinsic) by collagen. The activation of both extrinsic and intrinsic pathways results in the conversion of factor X to an active form that, in turn, interacts with factor V, calcium, and phospholipid from platelet factor III to convert prothrombin to thrombin (Fig. 7.2). Thrombin is the rate-regulating proteolytic enzyme that controls the conversion of fibrinogen to fibrin, the basic component of a venous thrombus. Other coagulation factors that are known to be increased after surgery are, principally, factors XI, IX, and VIII, which increase as a result of activation of the intrinsic pathway, and factor XII, which is activated by tissue collagen. There also is an increase in circulating platelets, platelet adhesiveness, and platelet aggregation within 72 to 96 hours of surgery. In addition, there is an increase in fibrinogen and in circulating fibrinolysin inhibitors. Normally, the fibrinolytic system, which is mostly plasmin formed from its inactive precursor, plasminogen, balances the clotting mechanism by digesting fibrin and fibrinogen and inactivating factors V and VIII. An acceleration of the clotting mechanism leads to a thrombus in either the venous or the arterial system. An excess of fibrinolytic activity causes failure of blood clotting and can produce serious hemorrhage. A proper balance of both is required for normal circulation.

FIGURE 7.1. Formation of venous thrombus following various surgical procedures with the activation of clotting factors and aggregation of platelets.

View Figure

FIGURE 7.2. Schematic representation of the cascade clotting mechanism, illustrating the role of extrinsic and intrinsic factors. Increases in tissue thromboplastin-like substance and collagen-activated factor XII initiate the formation of fibrin through the extrinsic and intrinsic pathways, principally by the activation of factor X.

View Figure

Venous stasis is known to be the cornerstone of postoperative thrombus formation. Venous stasis in the lower extremities and pelvis results in platelet aggregation and the adhesion of platelets to the vein wall, with the release of a thromboplastin-like substance that forms a platelet-fibrin-red cell network with resultant thrombus formation. These physiologic changes in venous hemodynamics occur in the preoperative, operative, and postoperative periods. Many investigators have found it advisable to perform the preoperative studies on high-risk patients on an ambulatory basis or, if the patient has been hospitalized for a significant period of time, to discharge her home for several weeks to increase her physical activity before elective surgery. Doran and others have shown that venous return from the lower extremities is decreased to one half its normal rate during the operative procedure. The decrease results from the loss of muscle tone that is caused by muscle relaxation from anesthetic agents. 125I-labeled fibrinogen scanning studies have demonstrated that venous clot is initiated during the operative procedure in 50% of the patients who develop thrombosis in the postoperative period. Blood flow from the lower extremity has been shown to undergo further reduction to about 75% of the normal drainage flow

immediately after surgery. This reduced flow rate persists for 10 to 14 days because of the loss of pump action of the leg muscles. The major site of clot formation is in the soleal venous sinuses of the calf, a portion of the venous arcade that joins the posterior tibial and peroneal veins that drain the soleal muscle. Thrombi from these sinuses often occur behind the valves that are located at the point at which the sinuses drain into the collecting veins. Thrombi often form in these large sinuses or valve cusps in bedridden patients. Another factor contributing to venous stasis is prolonged surgery with tight packing of the intestines in the upper abdomen and obstruction of the underlying vena cava. The type and length of operation are directly related to the incidence of postoperative venous thrombosis, as outlined in Table 7.1. TABLE 7.1. Levels of Thromboembolism Risk in Surgical Patients Without Prophylaxis

Level of Risk Examples

Calf DVT, %

Proximal DVT, %

Clinical PE, %

Fatal PE, %

Successful Prevention Strategies

Low risk

2

0.4

0.2

0.002

No specific measures Aggressive mobilization

10–20

2–4

1–2

0.1–0.4

LDUH q12h, LMWH, ES, or IPC

20–40

4–8

2–4

0.4–1.0

LDUH q8h,

Minor surgery in patients 40 yr or with additional risk factors

Highest risk

Major surgery in patients >40 yr plus prior VTE, cancer, or molecular hypercoagulable state; hip or knee arthroplasty, hip fracture surgery; major trauma; spinal cord injury

Modified from Gallus et al. and International Consensus Statement.

Source: Geerts WH, Heit JA, Clagett GP, et al. Prevention of venous thromboembolism. [Sixth ACCP Concensus Conference on Antithrombotic Therapy] Chest 2001;119(Suppl 1):132S–175S.

Diagnosis of Venous Thrombosis The traditional clinical methods used to diagnose venous thrombosis of the lower extremity are of limited value. Such methods may be in error in 50% of the cases and provide both false-positive and false-negative information. This diagnostic problem results from the silent and insidious nature of the venous thrombus formation process, which takes place principally in the deep soleal veins in the lower

extremity. Because the clinical diagnosis of this vascular complication is so inaccurate, more objective methods of diagnosis were required. In recent years, venography, 125I-labeled fibrinogen scanning, Doppler ultrasound, and impedance plethysmography (IPG) have been used along with a variety of other imaging techniques. Venography is regarded as the most definitive method for the diagnosis of venous thrombosis and has been used as the reference source against which other techniques are measured (Table 7.2).

TABLE 7.2. Diagnosis of Deep Venous Thrombosis (DVT)

Method

Sensitivity and Specificity

Clinical history and physical examination

Indication and Comments

Classic symptoms often absent in proven DVT 60 Non-major surgery

Obesity

Moderate

75–90 kg or >20% above ideal weight

Morbid

?115 kg or >30% above ideal weight with reduced fibrinolysin and immobility

Immobility

Preoperative

Prolonged hospitalization; venous stasis

Intraoperative

Prolonged operative time; loss of pump action of calf muscles; compression of vena cava

Postoperative

Prolonged bed confinement; venous stasis

Trauma

Damage of wall of pelvic veins

Radical pelvic surgery

Malignancy

Release of tissue thromboplastin a

Activation of factor X; reduced fibrinolysin

Radiation

Prior radiation therapy

Radiation

Prior radiation therapy

Medical diseases

Diabetes mellitus

Cardiac disease; heart failure

Severe varicose veins

Previous venous thrombosis with or without embolizationa

Chronic pulmonary disease Molecular hypercoagulable state

a Highest

risk.

Age In autopsy studies by Sevitt and Gallagher, the incidence of DVT was greatest in patients older than 60 years of age. Many additional studies have demonstrated a linear increase in fatal pulmonary emboli as patient age increases. The major correlation in such patients relates to degenerative changes in the vascular tree. The incidence of PE after major surgery increases sharply in women over 40 years of age. Obesity The risk of thromboembolism is decidedly increased in the obese patient. Breneman recognized obesity as one of the most significant factors in thromboembolic disease in the operative patient. His study of patients with thromboembolic disease demonstrated that they were 21.6% above their ideal weight. The major way that obesity influences thrombus formation is by venous stasis, which is aggravated by postoperative immobility. Immobility Prolonged inactivity in the preoperative patient causes an impairment in the venous circulation of the lower extremities. Preoperative immobilization, such as that required for prolonged diagnostic evaluation, produces a decrease in muscle tone of the lower extremity with diminished venous flow. These hemodynamic changes result in sludging of the platelets and red cells and set the stage for venous thrombosis during the operative period. Flane and associates and Kemble have clearly demonstrated this fact using 125I-labeled fibrinogen scanning before and immediately after surgery. They observed that in 50% of patients who ultimately developed postoperative thrombosis, the onset of the clot formation was during the operative procedure. Similarly, the patient who has undergone prolonged anesthesia with generalized muscle relaxation has an increased incidence of venous stagnation of the lower extremities and a higher incidence of thromboembolism postoperatively. It is important to reemphasize the fact that the nidus for venous thrombosis is frequently initiated either before or, more commonly, during the operative procedure. Consequently, prophylactic treatment for the high-risk patient should be initiated before surgery and should be continued until the

patient is fully ambulatory. Postoperative immobilization provides an added physiologic insult to preexisting venous stasis. As many as 66% of patients who develop postoperative venous thrombosis have evidence of thrombosis within 48 hours of surgery, as detected by 125I-labeled fibrinogen scanning. Such anatomic positions as sitting with the legs crossed, sitting with the legs dangling from the bed, and the exaggerated Fowler position produce impairment in venous return from the lower extremities. The high-risk patient (Table 7.3), in particular, should be ambulated vigorously. When confinement to bed is necessary, the legs and trunk should be elevated approximately 15 degrees above the horizontal. Sharnoff has demonstrated that immobilization in bed is perhaps more hazardous than the physiologic impact of advancing age. Other Factors Factors that lead to an increased risk of thrombosis and PE include varicose veins, previous thromboembolism, severe diabetes, cardiac failure, and chronic pulmonary disease. All of these produce impairment of circulation with resultant stasis and an increased frequency of venous thrombosis. Malignancy also is a recognized factor in venous thrombosis, although the exact mechanism is not fully understood. Many tumors undergoing tissue breakdown are known to elaborate a thromboplastin-like substance that may predispose to increased thrombosis. Hypercoagulable states, both congenital and acquired, increase the risk of thromboembolism in the gynecologic patient (Table 7.4).

TABLE 7.4. Risk of Thromboembolism

Deficiency/dysfunction

Antithrombin

Protein C

Protein S

Heparin cofactor II

Factor V Leiden

Prothrombin variant 20210A

Antiphospholipid antibodies

Lupus anticoagulant

Anticardiolipin

Anticardiolipin

Hyperhonocystireria

Dysfibrinogenemia

Decreased levels of plasminogen

Decreased levels of plasminogen activators

Heparin-induced thrombocytopenia

It is evident that a clinical profile can be developed for the high-risk patient. The paradigmatic high-risk patient is a woman who is older than 40 years of age, is morbidly obese (greater than twice her ideal weight), is diabetic, has varicose veins or heart disease (or both), has been in the hospital for a prolonged period for medical evaluation or treatment, and has a malignant tumor. Regardless of the nature of the pelvic disease, the greater the number of risk factors, the longer the operation, and the more difficult the surgical procedure, the more frequent is the occurrence of venous thrombosis and pulmonary embolus. Patients who embody one or more of these surgical risks should have intensive monitoring and prophylactic treatment for venous thrombosis. Treatment of Venous Thrombosis Prophylaxis The most effective treatment of venous thrombosis is prevention. In view of the evidence that between 5% and 45% of patients undergoing major gynecologic surgery develop venous thrombosis of the lower extremity, a prophylactic method of preventing this complication should be considered before surgery. In about 20% of cases, venous thrombi in the calf veins extend to the popliteal or femoral vessels, and in an estimated 40% of these, the patient develops pulmonary emboli. These facts are sufficient to warrant the prophylactic treatment of the high-risk patient in an effort to avoid this life-threatening sequence of events (Table 7.5). TABLE 7.5. Agents Used in Venous Thromboembolism

Agent

Mechanism of Action

Comments

Heparin

Combines with AT-III and neutralizes activated factors: IIa (thrombin activity)

Prevention and treatment of venous thromboembolism

Xa (responsible for thrombin generation) XIIa, XIa, IXa

Risk of heparin-induced thrombocytopenia Requires monitoring (aPTT) when used for treatment

LMWH Ardeparin Dalteparin Enoxaparin

Combines with AT-III and prevents thrombin generation through its anti-factor Xa effect

Prevention and treatment of venous thromboembolism Risk of heparin-induced thrombocytopenia No anti-IIa activity (if molecular weight 100 g in adults) usually in the presence of a decreased capacity to excrete mannitol (decreased GFR). Also to be excluded are patients with spurious hyponatremia that results from the abnormal accumulation of plasma lipids or proteins. In such circumstances, the concentration of sodium in plasma water is normal; however, the concentration of sodium expressed per liter of whole plasma is reduced because an abnormally large volume of whole plasma is occupied by the lipids or proteins, which do not contain plasma water and electrolytes. Thus, when aliquots of hyperlipemic or hyperproteinemic plasma are analyzed, a lower amount of sodium is determined to be present in a given volume of whole plasma. Plasma osmolality, however, is normal because lipids and proteins do not contribute importantly to plasma osmolality (see section on osmotic forces). Patients with spurious hyponatremia can be readily identified by the presence of markedly elevated total serum protein levels (e.g., multiple myeloma) or grossly lipemic serum. The distinction can be readily made if lipemic serum is subjected to centrifugation and the lipoprotein layer is removed before evaluation, if flame photometry is being used for measurement of serum Na+. Spurious hyponatremia is no longer a consideration in most laboratories, because serum Na+ concentration is determined by ion-specific electrodes and increased levels are not affected by lipemic serum. Symptoms of hyponatremia include increased tendon reflexes, lethargy, mental confusion, and muscle twitching, which are followed by convulsions, coma, and possibly death if levels fall beneath 115 mEq/L.

Hyponatremia and Volume Depletion Associated with Renal Sodium Wasting The normal renal response to volume depletion and hyponatremia is the virtual elimination of sodium from the urine (Fig. 8.4; see section on Sodium Balance). Thus, the presence of an excessive amount of urinary sodium under these conditions indicates that renal sodium loss is the cause or a major contributing factor to the state of sodium depletion. A spot urine sodium concentration greater than 40 mEq/L, a %E/F Na

over 1%, or a urinary sodium excretion rate greater than intake indicates such renal sodium wasting. The conditions discussed in the following sections are associated with hyponatremia, IVV depletion, and renal sodium wasting.

View Figure

FIGURE 8.4. Approach to the assessment of a hyponatremic patient. This approach considers only patients with true hyponatremia (i.e., in nonazotemic patients, serum osmolality is reduced in proportion to the decrease in serum sodium). Thus, patients are excluded who have lowered concentrations of serum sodium because of hyperlipidemia, hyperproteinemia, or the abnormal accumulations of solutes in the extracellular fluid, such as glucose or mannitol. ADH, antidiuretic hormone; %E/F Na , fractional excretion of sodium; GI, gastrointestinal; IVV, intravascular volume.

CHRONIC RENAL DISEASE All types of renal disease can be associated with renal salt wasting. In adults with such a disorder, the serum creatinine level is virtually always above 2 mg and usually much higher before a significant salt leak develops. These azotemic patients usually require 85 to 170 mEq of sodium daily (5 to 10 g of sodium chloride) to maintain salt balance at a normal effective IVV. Thus, if sodium intake is decreased in azotemic patients by anorexia or vomiting, or if additional sodium losses occur (e.g., diarrhea or diuretic therapy), the inability of the diseased kidneys to conserve sodium and water normally may rapidly lead to the development of significant sodium and water deficits. Water intake usually continues; therefore, sodium balance is more adversely affected than water balance. As a consequence, the patient becomes volume contracted with hyponatremia. With the onset of congestive heart failure or the nephrotic syndrome, the salt leak of chronic renal failure usually disappears, and salt intake must be restricted.

DIURETIC THERAPY The diuretics include thiazide agents or loop diuretics such as furosemide, bumetanide, and ethacrynic acid. Diuretics induce a renal salt-wasting state, and if the urinary output of sodium exceeds intake, sodium depletion ensues. Rarely, diuretics cause hyponatremia without evidence of volume depletion if severe potassium depletion has resulted from their use (see Fig. 8.4).

ADRENAL INSUFFICIENCY (ADDISON'S DISEASE) Destruction of the adrenal gland or sudden withdrawal of chronic, daily glucocorticoid therapy results in inadequate adrenal function. The lack of mineralocorticoid causes wasting of renal salt but retention of renal potassium, and leads to sodium depletion. The lack of glucocorticoid results in a decreased capacity to excrete a water load and leads to hyponatremia but not to volume depletion or hyperkalemia.

Hyponatremia and Volume Depletion Associated with Renal Sodium Conservation A spot urine sodium concentration of less than 20 mEq/L or a %E/FNa below 1% in a hyponatremic, volume-contracted patient is evidence of normal renal sodium conservation and indicates that the cause of the sodium depletion is nonrenal in origin or that it occurred during previous diuretic therapy. The fact that

the serum sodium concentration is lower than normal indicates that water balance is less negative than sodium balance. The conditions discussed in the following sections can result in volume depletion and hyponatremia as a result of extrarenal losses of sodium.

GASTROINTESTINAL LOSSES If losses of fluid from the upper gastrointestinal tract (e.g., vomiting, gastric aspiration) cause the hyponatremia, and if the gastric juice is normally acid, metabolic alkalosis is present. If diarrheal losses cause the hyponatremia, metabolic acidosis may be present. In patients with gastric achlorhydria, upper gastrointestinal losses also can lead to metabolic acidosis.

LOSSES OF SODIUM FROM THE SKIN Sweat contains about 50 mEq/L of sodium and is a hypotonic fluid. If sweat losses are not replaced, then hypernatremia can develop. In most situations, the water losses from the skin are replaced more adequately than the sodium losses. Thus, most patients with significant sodium losses due to sweating become hyponatremic. Skin losses of fluid and electrolytes also can occur after burns or other skin injuries. These are isotonic losses of sodium and lead to hyponatremia if the water losses are more adequately replaced than the sodium losses.

LOSSES OF SODIUM FROM PRIOR DIURETIC THERAPY The natriuretic action of most diuretics lasts less than 24 hours. Hyponatremia is made worse if water intake is excessive.

Hyponatremia and Normal Volume Status Associated with Water Diuresis In a patient with normal renal function who has become hyponatremic as a result of the administration or ingestion of excessive amounts of water, intravascular and ECF volume are normal to slightly expanded, and high rates of urine flow in association with maximally, or nearly maximally, dilute urine can be expected (see section on water balance). In a patient with preexisting renal functional impairment, water loading also increases urine flow rate and dilution of the urine; however, maximally dilute urine cannot be formed. Hyponatremia secondary to water loading may occur in compulsive water drinkers, who usually are severely neurotic or psychotic, or after excessive i.v. administration of hypotonic fluids. Many of these patients also have high levels of ADH for various reasons (e.g., drugs, psychosis). Without this elevation of ADH, presuming normal renal function, consumption of 20 L of water a day would be necessary for development of frank hyponatremia.

Hyponatremia and Normal to Slightly Elevated Volume Status Associated with Water Conservation As discussed, it is appropriate to observe a brisk water diuresis in a patient with normal renal function who is hyponatremic and has evidence of normal or slightly elevated IVV without edema. When high flow rates of hypotonic urine are not observed, the patient is exhibiting an inappropriate antidiuresis. This may result from the inappropriate release of ADH, although other mechanisms also can be involved (e.g., decreased renal blood flow, certain drugs). Another characteristic of such patients is that administered sodium is promptly excreted in the urine, perhaps because of the effect of atrial natriuretic factors. On the other hand, when sodium intake is curtailed, renal sodium conservation is observed. These patients also exhibit normal adrenal and renal function, and are not edematous. The syndrome of inappropriate antidiuresis has been associated with various clinical states, including malignant tumors (e.g., in the lung or pancreas), central nervous system (CNS) disorders (e.g., head trauma, meningitis), infections (e.g., tuberculosis, bacterial pneumonias), the postoperative state, hypopituitarism, and myxedema, as well as with many

drugs (Table 8.6). Infusion of oxytocin to induce uterine contraction also can cause hyponatremia because of the antidiuretic effects of oxytocin. TABLE 8.6. Antidiuretic Drugs

Sulfonylureas (chlorpropamide, tolbutamide) Cytotoxic agents (vincristine, cyclophosphamide) Nicotine Morphine Barbiturates Carbamazepine Psychotropics (tricyclics) Clofibrate Isoproterenol Nonsteroidals Salicylates Acetaminophen Vasopressin Oxytocin

Within the category of hyponatremia associated with normal IVV are three special categories. The feature that sets these apart is that patients may exhibit evidence of water conservation when water is withdrawn or an appropriate or nearly appropriate water diuresis when water is administered. That is, it appears that osmoregulation has been reset to “defend” a lowered plasma osmolality. The first special category includes patients who have an unusual response to diuretic therapy, characterized by hyponatremia, severe potassium depletion, and metabolic alkalosis. Despite the hyponatremia and normal IVV, exchangeable sodium is nearly normal, suggesting intracellular movement of sodium. Potassium replacement must be accomplished before specific treatment of hyponatremia. The second category involves patients with an unusual manifestation of a chronic illness, such as pulmonary tuberculosis, that resets the osmostat. The third category includes patients with sodium depletion resulting from any cause in whom the decrease in effective IVV is minimized by excessive water intake and retention. This effect of excessive water intake can occur in any of the causes of sodium depletion.

Hyponatremia Associated with Increased Effective Intravascular Volume or Increased Extracellular Fluid Volume (Edema or Ascites) CONGESTIVE HEART FAILURE When hyponatremia develops spontaneously in the course of chronic congestive heart failure (i.e., is not the result of excessive water administration or diuretic therapy), it usually is indicative of severe cardiac insufficiency and has a poor prognosis. The cause of the hyponatremia in such patients has been ascribed to a decreased capacity to increase renal free water clearance perhaps because of: (a) increased fractional reabsorption of glomerular filtrate proximal to the renal diluting sites of the distal nephron; and

(b) an elevated ADH level.

CIRRHOSIS OF THE LIVER Patients with cirrhosis and ascites have a decreased capacity to excrete a water load, possibly because of the same mechanisms at work in patients with congestive heart failure.

EXCESSIVE ADMINISTRATION OF HYPOTONIC FLUIDS This usually is an iatrogenic situation and must be especially guarded against in postoperative patients whose ADH levels are elevated because of stress, pain, hypovolemia, or drugs, as well as in elderly patients who are unable to maximally dilute their urine.

Hypernatremia All patients with hypernatremia are volume contracted except those in whom the disorder develops as a result of excessive administration of hypertonic saline or sodium bicarbonate and the rare patients with essential hypernatremia (Fig. 8.5). The following discussion considers only the first group of patients; the latter section on treatment discusses all forms of hypernatremia. Patients with hypernatremia usually have CNS deficits, and they may also have confusion and neuroseizures. Autopsy findings often reveal hemorrhages or thromboses of brain tissue. FIGURE 8.5. Approach to the assessment of a hypernatremic patient. This approach does not consider patients with hypernatremia secondary to excessive administration of hypertonic saline.

View Figure

Hypernatremia Associated with Formation of Concentrated Urine The normal renal response to decreased intake of water or increased extrarenal losses of water is the formation of maximally concentrated urine (see section on water balance). In most clinical situations in which hypernatremia is the result of water depletion, the expected renal response is a U osm:Posm ratio greater than 1.5 and a specific gravity above 1.015. Thus, the finding of hypernatremia with evidence of renal conservation of water indicates that the hypernatremia is caused by excessive nonrenal losses of water or solute diuresis.

EXCESSIVE NONRENAL WATER LOSS Hypernatremia typically develops in patients with accelerated rates of nonrenal water loss owing to a hot environment, fever, or hyperventilation and in whom water losses are not replaced because the patient cannot perceive or communicate thirst. Despite the hypernatremia, sodium deficits usually are present because initially, as water deficits develop, renal sodium excretion increases to maintain normal plasma osmolality and serum sodium concentration. When more than about 15% of ECF volume is lost, renal conservation of sodium occurs; if the water losses continue, hypernatremia develops. The presence of volume deficits is indicated by the signs of IVV depletion, as previously described. Urine flow rate usually

is less than 35 mL/h.

SOLUTE DIURESIS The amount of water that must accompany the excretion of a given amount of solute in the urine is determined by the osmolality of the renal medullary interstitial fluid (with which the collecting duct fluid must equilibrate) and the plasma level of ADH activity (which determines the permeability of the collecting duct to water and, therefore, the rate at which water moves from the collecting duct to medullary interstitial fluid to achieve osmotic equilibrium). Hypernatremia results if water intake does not keep pace with renal water losses, because although renal sodium excretion also is increased in solute diuresis, renal sodium reabsorption is affected proportionately less than water reabsorption. Large amounts of mannitol infused intravenously or high-protein mixtures fed by nasogastric tube (each gram of protein yields 8 mOsm as urea, phosphate, and potassium) can cause a solute diuresis sufficient to cause hypernatremia if water intake is inadequate. In solute diuresis, urine volume usually is greater than 35 mL/h.

Hypernatremia Associated with Formation of Dilute Urine The finding of hypernatremia in combination with isotonic or hypotonic urine indicates that, at least in part, the hypernatremia results from failure of normal renal conservation of water. Failure to concentrate the urine under these conditions may result from the lack of ADH (hypothalamic–pituitary diabetes insipidus) or impaired renal tubular function that interferes with the development of a hypertonic medullary interstitium (renal tubular damage). Central diabetes insipidus or nephrogenic diabetes insipidus should be suspected immediately in a patient with hypernatremia when the urine is very dilute (a Uosm:Posm ratio less than 0.5, or specific gravity 70°C

Coagulation (Latin: coagulatio = clotting). Collagens are converted to glucose.

>100°C

Phase transition from liquid to vapor of the intracellular and extracellular water.

Tissue rapidly dries out (dessication) (Latin: ex sico = dehydration). Glucose has an adhesive effect after dehydration.

>200°C

Carbonization (Latin: carbo = coal).

Medical pathologic burns of the 4th

Medical pathologic burns of the 4th degree.

* According to Bender and Schramm, 1968.

The Nd:YAG laser beam can be transmitted equally well with any distending medium, whereas monopolar electrosurgical devices operate most effectively in an electrolyte-free medium. Uterine perforation by either a laser fiber or an electrode is much more serious than perforation by scissors or another mechanical device, because the thermal energy can inflict great damage to surrounding structures (e.g., bowel or bladder). The injury may not attain its maximum damage until 2 or 3 days after surgery. Therefore, either laparoscopy or laparotomy is indicated in such cases to determine the extent of injury. The surgeon must be familiar with the physics governing the actions of lasers or electrosurgical tools and with the tissue actions exerted by these energized devices. A knowledgeable surgeon would not use a ball device to cut or a loop electrode to coagulate tissue. Proper selection of wattage depends on disease pathology and location. High power applied for a long period of time is risky, is inappropriate, and will inevitably lead to unwanted tissue injury. Regardless of whether a laser, resectoscope, or handheld electrode is used, depth of tissue action is extremely important; transmural injury is possible at high-power densities or with prolonged exposure. One must keep in mind that the thickness of the distended uterine wall (0.5 to 1 cm) is considerably less than that of the nondistended uterine wall (1.5 to 2 cm) (Fig. 17.21A,B).

View Figure

FIGURE 17.21. A: A uterus that is removed at hysterectomy shows the thick walls of the myometrium. These walls average about 1.5 cm in thickness, with the exception of the cornu, where the myometrium is thinner. B: The uterus is distended with the liquid medium. The walls of the uterus are now much thinner, averaging 0.5 to 1 cm in thickness.

THE PROCEDURES OF HYSTEROSCOPIC SURGERY Part of "Chapter 17 - Operative Hysteroscopy"

Septate Uterus Modern hysteroscopy has rendered the correction of septate uterus relatively simple and straightforward by the transcervical route, as documented by DeCherney and associates (1986) and by March and Israel (1987). Uterine septa are a treatable factor contributing to pregnancy wastage, usually secondary to premature labor and late second trimester abortion. The diagnosis of a uterine septum is usually made at hysterosalpingography or during a diagnostic hysteroscopy. Unfortunately, neither of the studies mentioned above differentiates between septate and bicornuate uteri. A diagnostic laparoscopy is most helpful for an accurate differential diagnosis. A laparoscopic view of a septate uterus will reveal a wide but otherwise normal fundus, whereas the bicornuate uterus typically appears heart-shaped. A bicornuate uterus should be treated by the Jones or Strassman procedure. A septate uterus should be treated hysteroscopically. The standard technique, reported by March et al. (1978), is to cut the septum with scissors under direct hysteroscopic view. Cararach and colleagues compared hysteroscopic incision of septate uteri during a 5-year period (81 women) using a scissors or resectoscope approach and found only marginal benefit in favor of the former. Choe and Baggish used the Nd:YAG laser fiber to transect septa in 14 women. Of 13 patients who conceived, 10 delivered a liveborn, term infant (87%), compared with a preoperative term pregnancy rate of 11%. Clearly, the Nd:YAG laser, resectoscope, and needle electrode are more appropriate for the broad and usually vascular septum. Uterine rupture during pregnancy and more specifically in labor has been reported after hysteroscopic metroplasty with and without uterine perforation. It would be prudent to inform the patient who will undergo metroplasty of the subsequent risk so that she is knowledgeable and can inform her obstetrician should she become pregnant.

Hysteroscopic Technique The uterine septum is viewed from the level of the internal cervical os. The endoscope is moved into each chamber of the divided uterine cavity, and the locations of the tubal ostia are marked. The hysteroscope is again withdrawn to a level just above the internal os. The appropriate operating instrument is inserted through the sheath, and the septum is cut from below and upward, as described by March (Fig. 17.22). As the fundus is approached, the operator depends on a signal from the assistant to indicate when the quality of the hysteroscope light demonstrates transmission through the intact uterine wall. A dialog between the hysteroscopist and laparoscopist prevents perforation. A new technique permits the operator to scan the uterus ultrasonographically to determine whether the myometrium has been entered and to monitor the amount of space existing between the operating device and the serosal surface of the uterus. It is unnecessary to excise the septum completely. Transection eliminates the septum and unites the uterus into a single cavity. We take the opportunity here to stress an important technical point. The surgeon should be aware of the common tendency of the cutting instrument to drift posteriorly and should clip the septum squarely in the middle. When the drift goes unnoticed, the operating instrument invariably cuts into the myometrium and causes pulsatile bleeding. Similarly, correcting the septum too perfectly at the level of the fundus results in deep penetration into the myometrium and subsequent hemorrhage (Fig. 17.23). If a multichannel hysteroscope is used, a 3-mm ball electrode can be used to coagulate the bleeding vessel. The double-needle bipolar electrode is a safe alternative method for electrocoagulation.

FIGURE 17.22. During this technique for a hysteroscopic section of a uterine septum, the operator cuts the septum from below and moves upward (A–D) until a single cavity is seen (E). (From Baggish MS, Barbot J, Valle RF. Diagnostic and operative hysteroscopy. Chicago: Mosby–Year Book, 1989, with permission.)

View Figure

FIGURE 17.23. The scissors or laser fiber should be maintained in the center of the septum to avoid bleeding. (From Baggish MS, Barbot J, Valle RF. Diagnostic and operative hysteroscopy. Chicago: Mosby–Year Book, 1989, with permission.)

View Figure

If bleeding does ensue, a Foley catheter with a 10-mL balloon is inserted into the endometrial cavity at the terminus of the operation and inflated to 5 to 6 mL. The pressure exerted by the bag on the uterine walls is sufficient to control the bleeding promptly. The bag is deflated 6 to 12 hours postoperatively and is removed if no further bleeding ensues. Overinflation of a catheter bag can lead to uterine rupture; therefore, one should add increments of water, 1 to 2 mL at a time, to the catheter bag until bleeding ceases. Patients are usually advised to take 2.5 mg of estrogen daily by mouth for 30 days after surgery. Antibiotics are not routinely administered.

Uterine Synechiae Adhesions form between the anterior and posterior walls of the uterus as a result of trauma or infection in a milieu of estrogen deprivation. Classically, this problem follows an abortion or postpartum hemorrhage for which a vigorous curettage was performed to control the bleeding, as reported by March and associates. Friedler and associates report the incidence of adhesions after one abortion to be 16.3%. This figure rises to 32% after three or more abortions. The severity of adhesions also typically rises as the number of abortions increases. Microscopic sections obtained from the curettage invariably reveal fragments of myometrium interspersed with inflamed decidua and glands. Historically, the patient does not resume menstruation; however, a minority of patients continues to menstruate normally. Because the patient is subsequently infertile or amenorrheic, a hysterogram is performed. The radiograph reveals filling defects that vary from minimal to severe (i.e., virtually obliterating the endometrial cavity). Past treatment of uterine synechiae consisted of blind curettage; the results were predictably poor. With the advent of operative panoramic hysteroscopy, treatment has progressed to identification of adhesions and sharp incision of the adhesions with scissors. Adhesiolysis surgery is probably the most difficult of the hysteroscopic operations. Because numerous vascular channels are opened up, the risk of intravascular absorption of the medium is high. Rock et al.

report a technique of laparoscopically injecting the uterus with leukomethylene blue dye to help identify the junction at which the anterior and posterior walls are adhered.

Hysteroscopic Technique A thorough diagnostic hysteroscopy is performed to assess the degree of adhesion formation and deformity of the cavity. Small openings in the curtain of adhesions in which there are flow patterns of tiny blood fragments and tissue debris are helpful and should be sought out, as are any normal anatomic landmarks. Photographs, videotapes, and detailed drawings are helpful reminders in planning the strategy for cutting these adhesions. Simultaneous laparoscopy is a prudent measure to prevent perforation of the uterus. Flexible or semirigid scissors and the Nd:YAG laser are the operating instruments of choice, although some operators use the monopolar needle electrode at 40 to 50 W of cutting power, blend 1 or 2. The laser is initially set to deliver 30 to 50 W of power. The medium is instilled into the cavity by way of an operating sheath. Continuous maintenance of distention is one key to success. Filmy and central adhesions should be cut first, always following the fluid flow. Marginal and dense adhesions should be tackled last, always cutting from below and moving upward. A second key to success is to maintain the hysteroscope in midchannel relative to the uterine walls. The cavity can usually be restored to reasonably normal architecture. Bleeding is not uncommon during this operation, particularly when cutting marginal adhesions, because the border between adhesion and myometrium is blurred. Hyskon provides an advantage here. The patient should be placed on conjugated estrogens, 2.5 mg daily, during postoperative recovery. Placement of an IUD within the cavity to keep the walls from adhering is clearly not based on scientific fact but has been used for so many years that it is a standard postoperative procedure.

Cannulation of the Fallopian Tube Novy et al. described a technique for passing a special catheter into the tubal ostium and through the obstructed interstitial portion of the tube. This procedure was successful in 92% of cases. Dumesic and Dhillon reported a tubal cannulation procedure in which they used a flexible guiding insert to facilitate passage of the cornual cannulation catheter. These techniques are useful for treating interstitial obstruction secondary to cellular debris and tubal spasm. The obvious advantage of this cannulation technique is its usefulness in treating cases that might otherwise require tubocornual anastomosis. Pregnancy rates range from 25% to 54% in 6 months.

Hysteroscopic Technique A 5.5F Teflon cannula with a metal obturator (Cook OB/GYN) is introduced through the operating channel of the hysteroscopic sheath. The obturator is removed. A 3F catheter with a guide cannula wire is introduced into the 5.5F cannula by way of a Y-adapter on the end of the cannula, engaged into the tubal ostium, and gently advanced into the tube. When the cornual portion of the tube is negotiated or when resistance is encountered, the guide wire is withdrawn and leukomethylene blue or indigo carmine dye is injected through the 3F catheter. Simultaneous laparoscopy allows one to see the dye exit the fimbriated end of the tube and to confirm patency. Alternatively, one can place a radiologic plate beneath the patient and inject radiopaque dye.

Uterine Polyps Functional and nonfunctional polyps produce intermenstrual bleeding, as reported by Barbot. Functional polyps tend to be smaller than nonfunctional polyps. If a hysterogram is performed, then a focal filling defect will be seen. Diagnosis is directly and readily made by hysteroscopy. Polyps protrude into the endometrial cavity. A functioning polyp has a lining identical to the surrounding endometrium. A nonfunctioning polyp presents as a white protuberance covered with branching surface vessels; thick-walled vessels are usually seen within the depths of the polyp. Polyps are relatively easy to diagnose

and to treat.

Hysteroscopic Technique A multichannel operating hysteroscope is inserted into the uterine cavity, and a retractable electric snare loop is inserted through the 3-mm channel of the operating sheath. The polyp is encircled by the loop such that the loop encompasses the polyp base as it is tightened. The polyp is cut off at the base with 30 to 40 W of power for cutting current. The snare is then removed, and an alligator jaw forceps is inserted. The polyp is grabbed by the forceps. The hysteroscope is withdrawn, removing with it the freed polyp, which is sent to the pathology lab for histologic evaluation. The site of removal is inspected again, and the procedure is terminated. If any bleeding is observed, a 3-mm ball electrode is applied to the site for coagulation (40 to 50 W).

Myomata Uteri Submucous myomas characteristically appear as white spherical masses covered with a network of fragile thin-walled vessels when viewed by hysteroscopy. Myomas typically are sessile or pedunculated. A hysterogram shows a filling defect that is not dissimilar to that produced by a polyp. Unfortunately, blind D&C is a grossly inaccurate method of diagnosing this disorder. Although subserous and intramural myomas rarely produce alarming symptoms, even when they attain relatively large size, smaller lesions in the submucous location invariably cause considerable bleeding. Additionally, submucous myomas are commonly associated with chronic endometritis, which interferes with implantation of the fertilized ovum and becomes a factor contributing to subfertility. In the past, a diagnosis of submucous myoma was usually followed by a recommendation for hysterectomy. Today, hysteroscopic surgery offers a therapeutic alternative to that radical approach. Various regimens of drug therapy (e.g., danazol [Danocrine] and gonadotropin-releasing hormone analogs such as leuprolide acetate [Lupron] or goserelin acetate [Zoladex]) have been recommended as supplementary preoperative medical therapy. The general plan is to treat symptomatic patients for 2 to 3 months preoperatively in order to reduce the size and vascularity of the lesion during surgery. All patients should be given detailed information concerning the need for typing and holding blood and the possibility of hysterectomy if intractable bleeding occurs. Valle (1990) reported data on 59 cases of abnormal bleeding, dysmenorrhea, and infertility that were diagnosed as submucous myomas. Hysteroscopy eliminated or markedly decreased bleeding in 52 of these cases. Baggish and Sze treated 71 patients with symptomatic myomas and four patients with incidental submucous myomas. The treatment methods used with the multichannel hysteroscope were Nd:YAG laser (n = 41), monopolar loop (n = 6), monopolar needle (n = 6), bipolar needles (n = 10), and electrosurgery or scissors and laser (n = 12). As with Valle's series, results were excellent; 65 of 75 (87%) returned to normal menses postoperatively. Barbot and Parent (personal communication, 1994) performed resectoscopic myomectomies in 825 women, of whom 83% were relieved of abnormal bleeding and suffered no recurrence.

Hysteroscopic Technique Several variations of hysteroscopic procedures are now available to manage submucous myomas. Current resectoscopic techniques differ little from those described by Neuwirth (1978) and by DeCherney and Polan (1983). However, the resectoscopic instrumentation has vastly improved compared with those earlier instruments. Self-flushing sheaths, straight and offset cutting loops, and diminished-diameter, low-profile scopes are among these recent improvements. In addition, electrosurgical generators have been modernized and are safer devices than instruments from 1970s and 1980s (Fig. 17.24A,B). Under video control, the resectoscopic technique consists of progressive shaving of the myoma and harvesting the pieces of tissue for subsequent histologic evaluation. For fundal myomas, the straight electrode is the most effective device, whereas the angulated electrode is preferred for lesions located on the anterior or posterior walls (Fig. 17.25A–C). (The electrode should be activated only while returning toward the hysteroscope, never while advancing outward away from the lens.)

FIGURE 17.24. A: A modern, state-of-the-art, computerized, constant-voltage electrosurgical generator. This apparatus is divided into cut, coagulation, monopolar, and bipolar functions. It incorporates numerous and sophisticated safety features. B: Close-up of the electrosurgical generator panel. The high cut feature is for bipolar cutting (vaporization).

View Figure

View Figure

FIGURE 17.25. The shaving technique for the elimination of a submucous myoma is shown using an angulated loop electrode via the resectoscope. (From Baggish MS, Barbot J, Valle RF. Diagnostic and operative hysteroscopy, second edition. St. Louis: Mosby–Year Book, 1999, with permission.)

The four Nd:YAG laser techniques described by Baggish and associates (1999) use power levels of 30 to 60 W. The first uses a hysteroscopic needle inserted through the operating channel, through which about 5 to 10 mL of 1:100 vasopressin solution (1 mL vasopressin in 99 mL sterile water) is injected into the myoma. The conical, sculpted, 1-mm laser fiber is brought into contact with the myoma to cut across its base. Scissors can be combined with the laser to free the myoma from its base. The myoma is extracted intact by way of the cervical canal. This technique is useful for myomas up to 3 cm in diameter. The second Nd:YAG technique uses a 1-mm ball of sculpted fiber that is drawn over the myoma multiple times for ablation of the myoma until it is level and flat in relation to the surrounding endometrium. This technique is used for 1- to 2-cm myomas. The third technique is similar to that used with the resectoscope. Layer upon layer of the myoma is sliced off until the base is reached. The fourth technique is used for a large (2- to 5-cm) lesion. The laser is used to devascularize the myoma by making multiple punctures into its substance. The large myoma can then be quartered with the laser and extracted piece by piece (Fig. 17.26A–C).

View Figure

FIGURE 17.26. A: Several Nd:YAG laser techniques are for the treatment of submucous myoma. Here the laser fiber cuts the small sessile myoma across its base. B: A large myoma is quartered by the laser fiber and then extracted in pieces. C: Layered cleaving of the myoma can be accomplished with a resectoscope or a laser. (From Baggish MS, Barbot J, Valle RF. Diagnostic and operative hysteroscopy, second edition. St. Louis: Mosby–Year Book, 1999, with permission.)

Other electrosurgical techniques are now performed with the large isolated-channel, flushing hysteroscope. The 3-mm needle, shaving loop, and bipolar electrodes may be used to perform all of the optional operations described above for the resectoscope and laser. The fine-needle electrode can be substituted for the laser fiber to excise pedunculated or section sessile myomas. The 3-mm retractable cutting loop can perform shaving procedures in a fashion similar to that of the resectoscope loop. The bipolar needles can be plunged many times into the substance of a submucous myoma of any size to coagulate the interior of the myoma (myolysis). If postoperative bleeding occurs, a 10-mL Foley balloon is placed in the cavity and blown up to 5 mL for 6 to 12 hours. If the cavity is large, a 30-mL balloon inflated with 10 to 15 mL of water can be used. We prefer to do a simultaneous laparoscopy when large myomas (2 to 5 cm) are resected and extracted. Regardless of myoma size, a simultaneous laparoscopy should be performed whenever concern for perforation exists. The central fundal myoma is associated with the greatest risk of uterine perforation. Reports caution that uterine rupture can occur during pregnancy after hysteroscopic myomectomy. This is particularly the case when the operator attempts to resect the intramural portion of the submucous myoma.

Endometrial Ablation More than 600,000 hysterectomies are performed annually in the United States, although the advent and growth of integrated health care will target this largely elective operation for reduction because of its substantial cost. The numbers will be further reduced because 40% of hysterectomies are unnecessary and 20% show no pathology; hence, cheaper alternatives are continuously sought. According to a recent study of hysterectomy by the New York State Department of Health, 30,065 hysterectomies were performed in that state during 1986, 10% of which were performed for the principal diagnosis of disorders of menstruation. Endometrial ablation or resection is the hysteroscopic alternative to hysterectomy as treatment for abnormal uterine bleeding. Two earlier reports by Droegemueller et al. describe blind procedures such as cryocoagulation that were used in an attempt to control dysfunctional uterine bleeding by creating physical destruction of the endometrium without sacrificing the uterus. Unfortunately, either the techniques themselves were associated with significant side effects or the endometrium promptly regenerated. Since the first practical method of hysteroscopic ablation was described in 1981, several thousand procedures have been performed by a variety of techniques, including the Nd:YAG laser, the resectoscopic roller ball or loop, and, most recently, the long hysteroscopic ball electrodes. Garry and associates (1995) reported 600 endometrial laser ablations performed on 524 women. No major operative morbidity was reported. The success rate (mean age, 43 years) was 83.4%. Baggish and Sze have performed 568 ablations; 401 of these were performed with the Nd:YAG laser, 167 by electrosurgery. Excellent results were obtained in 89% of the women treated, and amenorrhea was achieved in 58%. Again, no major operative complications were observed. Magos and coworkers reported 250 cases of endometrial resection with a 92% improvement in abnormal bleeding. However, data obtained from the Royal College of Obstetricians and Gynaecologist's Mistletoe (Minimally Invasive Surgical Technique Laser, Endothermal or Endoresection) Study in 1997 revealed a 6.4% rate of significant complications associated with endometrial resection alone and a rate of rate of 11.4/1,000 for emergency hysterectomy.

This compares to complication rates of 2.7% and 2.1% for laser and rollerball, respectively. The latter two techniques had emergency hysterectomy rates of 1.3/1,000 (i.e., 11 times less than endometrial resection). Two large controlled, randomized studies compared hysterectomy with hysteroscopic ablation-resection. Dwyer et al. prospectively compared 100 cases of endometrial resection and 100 cases of abdominal hysterectomy for menorrhagia. Postoperative morbidity, length of hospital stay, and time to return to work, normal daily activities, and sexual intercourse were significantly lower for the endometrial resection group. Dysmenorrheic premenstrual symptoms were significantly higher in the endometrial resection group. Pinion et al. randomized 204 patients to abdominal or vaginal hysterectomy (n = 99), endometrial laser ablation (n = 53), or endometrial resection (n = 52). Women treated by ablation or resection had less morbidity and a shorter recovery time. After 12 months, 89% of the hysterectomy group and 78% of the hysteroscopy group were very satisfied with the effect of surgery; 95% in the first group and 90% in the second reported acceptable improvement in symptoms. Equal numbers in each group stated that they would recommend the same operation to others. Several published reports confirm the cost effectiveness and efficacy of endometrial ablation for the control of abnormal uterine bleeding compared with hysterectomy.

Hysteroscopic Technique All patients who might be candidates for endometrial ablation should be managed first by hormonal treatment in an attempt to control the abnormal uterine bleeding. If this strategy fails, and if the woman does not desire to bear children, then she is a candidate for endometrial ablation. A preoperative diagnostic hysteroscopy, endometrial sampling, or both should be performed to exclude endometrial carcinoma or atypical hyperplasia, and all pertinent hematologic studies and consultations should be performed. All patients are pretreated to atrophy the endometrium. The drugs available to accomplish this effect include Danocrine, Lupron, Zoladex, Megace, and Depo-Provera. It is our experience that the best endometrial suppression is seen after 6 weeks of drug therapy. A simultaneous laparoscopy is not performed during endometrial ablation unless a perforation or other transmural injury is suspected. Depending on the technique selected, either 5% mannitol or 0.9% saline is used as the distending medium. The operating hysteroscope or resectoscope is inserted into the uterine cavity. With the hysteroscope, a 9F aspirating cannula (Cook OB/GYN) is inserted, and blood and debris are evacuated until the cavity is clear. We prefer to treat the fundus by dragging the laser fiber or the ball electrode from side to side (cornu to cornu) (Fig. 17.27). The anterior and lateral walls are ablated next, before the posterior wall. Ablation should not be extended below the internal os into the cervix. Power settings for the electrosurgical generator range from 50 to 150 W, depending on the size of the ball, barrel, or loop electrode (Fig. 17.28 and Fig. 17.29). Laser power is set at 40 to 60 W. The goal of the ablation operation is to destroy the visible endometrium, including the cornual endometrium, to a depth of 1 to 2 mm. The conduction heat will actually spread deeper, usually to 3 to 5 mm, depending on how long the device remained on the tissue. This penetration translates into extensive superficial myometrial destruction and coagulation of the radial branches of the uterine artery (Fig. 17.30). When the endometrium sloughs, regeneration is prevented because basal and spiral arterioles do not survive the 100°C heat exposure. Over a period of 6 to 8 weeks, the uterine walls scar and shrink. Subsequent sampling or hysteroscopy is possible after endometrial ablation. The mean duration of the operation is about 30 minutes. Patients usually are sent home on the day of surgery. The operation is usually completed with little or no blood loss (Fig. 17.31).

FIGURE 17.27. Endometrial ablation can be performed by use of a ball electrode in direct contact with the endometrium. The energized ball is pulled down to create a 1- to 2-mm furrow of endometrium. The conduction injury can extend down another 1 to 2 mm.

View Figure

FIGURE 17.28. The resectoscope with ball electrode attached is inserted into the uterine cavity. Initially, the cornu and fundus are carefully ablated taking care to keep dwell time low in order to reduce the risk of deep heat-conduction injury. Next, the anterior wall is ablated, followed by the posterior wall. (From Baggish MS, Barbot J, Valle RF. Diagnostic and operative hysteroscopy, second edition. St. Louis: Mosby–Year Book, 1999, with permission.)

View Figure

FIGURE 17.29. Endometrial resection is performed in a manner similar to ablation; however, instead of a ball electrode, a cutting loop is substituted. This is clearly a riskier procedure compared with ablation by either laser or ball electrode, particularly relative to deep myometrial resection and the accompanying risks of hemorrhage and/or perforation. (From Baggish MS, Barbot J, Valle RF. Diagnostic and operative hysteroscopy, second edition. St. Louis: Mosby–Year Book, 1999, with permission.) View Figure

FIGURE 17.30. The uterus was removed 5 days after a Nd:YAG laser ablation. Note the extensive laser injury involving about half the thickness of the myometrium. Laser penetration depends not only on power but also on the length of time the laser beam remains in contact with the tissue.

View Figure

FIGURE 17.31. Photomicrograph of a section of endometrium that was ablated by a Nd:YAG laser. The troughs represent places in which the laser fiber has cut into the tissue. The tissue between the troughs, as well as at the lower extremity, is necrotic.

View Figure

Nonhysteroscopic Minimally Invasive Techniques for Endometrial Ablation Although these new techniques are not hysteroscopic in the strictest sense, they are nonetheless closely related and should be included in this chapter. This basis for these methods encompass the following logic: Removal of the skill factor as a variable for endometrial ablation. Elimination of requirement for a distending medium. Reduction in the time required to perform the operation. Equivalency of efficacy compared with hysteroscopic ablation. Performance in an office setting.

The first practical technique described by Phipps et al. (1990) used a microwave technique of heating the endometrium by a probe inserted into the uterine cavity, exposing the endometrium to temperatures of 60° to 65°C for 15 minutes. Thijssen reported a large multicenter study using the technique. The report described a number of serious complications, including fistula formation and third-degree burn injuries. Several techniques using balloons containing hot water, balloons covered with monopolar electrodes, computerized continuously circulating in situ hot saline, and cryosurgical and photochemical techniques have been reported (Fig. 17.32, Fig. 17.33 and Fig. 17.34). The only device which has been practically used in the current marketplace has been the Thermachoice balloon technique (Ethicon Endo-Surgery, Cincinnati, Ohio) which was developed by Neuwirth. A cannula fitted with a terminal balloon is placed in the uterine cavity (Fig. 17.35). Sterile water distends the balloon and is heated in situ to 80° to 90°C. Thermistors mounted on the balloon give a continuous readout of temperature within the uterine cavity (Fig. 17.36). NaDH diaphorase staining showed destruction to a depth of 3.3 to 5 mm. The published data show efficacy to be lower than that for hysteroscopic ablation, and cost to be no less than that for hysteroscopic ablation. The dream of an office-based procedure has not yet been realized. FIGURE 17.32. Innerdyne (Tyco) “Enable” cannula. The terminal mushroom fitting and the acorn seal the cervix and isolate the corpal cavity from the cervical canal. The apparatus contains an in situ heater and thermistors.

View Figure

FIGURE 17.33. The “Enable” computerized continuously circulating hot saline cannula is engaged into a uterus and will ablate the endometrium at 80° to 90°C for 15 minutes.

View Figure

FIGURE 17.34. Hemiresection of a uterus removed at hysterectomy after “Enable” 15-minute ablation. Notice the brown, “cooked” appearance of the endometrium. See color version of figure.

View Figure

FIGURE 17.35. The Gynecare balloon cannula and control equipment. The cannula is inserted into the uterus. The balloon is inflated with water, which in turn is heated to 80° to 90°C.

View Figure

FIGURE 17.36. Schematic view of the inflated Gynecare hot-water balloon. See color version of figure.

View Figure

Miscellaneous Procedures

IUD Removal Although the number of IUD insertions has diminished over recent years, the gynecologist is occasionally called on to search for and remove a device with an indicator string that is not seen in the cervix. In such circumstances, the operating hysteroscope is a vital tool with which to locate the device and remove it under direct vision, according to Valle et al. The hysteroscope is inserted, and the device is viewed. If a string is seen, an alligator-jaw forceps is inserted, and the string is grasped. The hysteroscope is withdrawn, pulling the device through the uterine cavity and the cervix to the exterior. If the IUD is embedded, then a rigid grasping forceps is required. The IUD is located, and the large jaws of the rigid instrument grab the extruded portion of the IUD itself. Strong pressure is exerted on the jaws as the sheath of the hysteroscope is slowly withdrawn from the uterus, into the cervix, and out of the vagina.

Biopsy of Intrauterine Lesions When a tumor is suspected, the operative hysteroscope is inserted into the cavity, a 9F biopsy forceps is directed to the tumor site, and multiple biopsy specimens are obtained in a fashion analogous to that used with colposcopic biopsies. A 9F plastic cannula is inserted by way of the operating channel, and strong suction is applied to the mouth of the cannula by means of a 30-mL syringe (Fig. 17.37). The cannula is removed, and the contents are flushed out with saline into a bottle of fixative. Similarly, a 9F curette can be inserted under direct vision. Alternatively, a diagnostic hysteroscope can be inserted into the uterus. The site of pathology is noted. The endoscope is withdrawn, a Novak curette is inserted into the cavity, and biopsy specimens are taken at the previously located site. Finally, the hysteroscope is pulled back to the level of the internal cervical os, a small Novak curette is inserted alongside the hysteroscope, and a directed biopsy specimen is obtained. FIGURE 17.37. Direct sampling of an intrauterine lesion can be accomplished with a plastic cannula attached to a 30-mL syringe.

View Figure

Hemangiomas and Arteriovenous Malformations Hemangiomas and arteriovenous malformations can be diagnosed by their characteristic hysteroscopic appearance and by a history of massive unresponsive bleeding. Historically, women with these conditions are young and of low parity. Hysteroscopic examination shows the subsurface of the endometrium to be covered with irregular bluish purple vessels that form an abnormal tangle of distended channels that differ markedly from the normal fine-capillary net pattern. The abnormal channels are not unlike the vessels that cover the surface of submucous myomas. The Nd:YAG fiber is inserted through the operating channel, and the fiber is held several millimeters above the vascular abnormality at a power up to 50 to 60 W. The laser is discharged without touching the vessels or the surface of the endometrium. The laser energy causes the vessels to collapse and coagulate and the surface to blanch white. The endometrium

neighboring the abnormality is also treated and coagulated. The fiber is then withdrawn, the field is aspirated clear, and the hysteroscope is withdrawn. Similar treatment is repeated two or three times at 1-month intervals or until all evidence of the abnormality is obliterated.

Complications Unfortunately, accurate data concerning complications are hard to obtain, although one simple fact is clear: As greater numbers of gynecologists have begun to perform operative hysteroscopy, the rate of complications has increased. Voluntary surveys are worthless. Only state-mandated reports, such as those that are required for laparoscopic cholecystectomy in New York, have rendered any useful information. The exception to this is an excellent report by Smith et al., which details complications encountered by 42 gynecologists performing 257 endoscopic procedures (operative laparoscopy and hysteroscopy) in 218 patients (mean, 5.4 cases per surgeon) over a period of 15 months at the Swedish Hospital Medical Center in Seattle. Of 43 endometrial ablations, perforation was observed in three, fluid imbalance in two, and technical failure to complete in two. Myomectomy was performed in 30 and ablation plus myomectomy in 13. Perforation occurred in two, fluid imbalance in four, and fistula, sepsis, or both in three. Lysis of septa or synechiae was performed in 14 women, with perforation or hemorrhage in six patients. The overall complication rate ranged from 12% to 43%. Similarly, the MISTLETOE data cited in the Endometrial Ablation section are useful accurate data relative to endometrial ablation/resection complications.

Intraoperative and Postoperative Bleeding The most common complications inherent to hysteroscopic surgical procedures are intraoperative and postoperative bleeding. Generally speaking, intraoperative bleeding can be managed by aspirating the blood and by increasing the pressure of the distending medium so that it exceeds arterial pressure and compresses the walls of the uterus sufficiently to stop bleeding. Then the bleeding vessel can be coagulated with a 3-mm ball electrode with the use of forced coagulation at 30 to 40 W of power or by multiple jabs with bipolar needles at 20 to 30 W of power with the generator set for automatic bipolar. If the counterpressure of the medium is relaxed (at the termination of the procedure) and bleeding continues, then control is best obtained by inserting an intrauterine balloon initially inflated to 2 to 5 mL. If this pressure does not promptly stop the bleeding, then a larger balloon can be distended to 10 mL until the bleeding has stopped. More distention may be required for larger uteri. Care must be taken because overinflation of an intrauterine balloon can itself rupture the uterus. The balloon remains in place for 6 to 8 hours, is partially deflated for 6 hours, and, finally, is totally deflated before removal. When the bleeding is pulsatile, the source is arterial rather then venous. If this type of bleeding is not immediately controlled by balloon compression, then hysterectomy will usually be required. Delayed postoperative bleeding is most commonly associated with endometrial slough (after ablation), chronic endometritis, or spontaneous extrusion and expulsion of the intramyometrial portion of a previously resected submucous myoma (Fig. 17.38 and Fig. 17.39). Bleeding-clotting studies should be obtained in cases of late postoperative bleeding, particularly if these studies were not performed preoperatively in women with a diagnosis of abnormal uterine bleeding (preoperative endometrial ablation or myomectomy).

FIGURE 17.38. A new method for ablating a myoma via a ridged electrode or vaportrode, which develops very high power densities. (From Baggish MS, Barbot J, Valle RF. Diagnostic and operative hysteroscopy, second edition. St. Louis: Mosby–Year Book, 1999, with permission.)

View Figure

FIGURE 17.39. The part of the myoma protruding into the uterine cavity (submucous portion) is destroyed by either resection or vaporization. That portion remaining within the myometrium may be sufficiently devascularized so as to subsequently extrude itself into the cavity and be expelled via the cervix. (From Baggish MS, Barbot J, Valle RF. Diagnostic and operative hysteroscopy, second edition. St. Louis: Mosby–Year Book, 1999, with permission.)

View Figure

Uterine Perforation Uterine perforation can occur during any operative hysteroscopy procedure but is most common during septum resection, myomectomy operations, and adhesion takedown. The best insurance against this complication is simultaneous laparoscopy. Among novice operators, perforation can occur even during insertion of the hysteroscope. With appropriate care, this sort of perfora-tion should not happen, because the cervix and internal os should be negotiated under direct vision, and the cavity should likewise be entered under direct vision. Examination under anesthesia is also simple and lets the operator know the direction of the uterine axis. As we noted above, the most dangerous perforations are those associated with lasers and electrosurgical devices. The risk of this type of injury can be reduced by not activating the energy device during a thrusting or forward movement. The foot pedal is activated only during the return phase of the laser fiber or electrosurgical electrode. If a perforation does happen with an energy device, then laparotomy is

required to ensure that no injury has been inflicted on the intestine, bladder, or ureter (Fig. 17.40). FIGURE 17.40. The operator should never apply power to an energy device while advancing the electrode. The power can safely be applied as the electrode returns toward the sheath. (From Baggish MS, Barbot J, Valle RF. Diagnostic and operative hysteroscopy, second edition. St. Louis: Mosby–Year Book, 1999, with permission.)

View Figure

A risk of perforation is associated with septum transection in its final phase at the level of the uterine fundus because the operator may have some difficulty determining where the septum ends and the myometrium begins. This risk is constant regardless of the cutting instrument used. The operator rapidly becomes aware that uterine perforation has occurred because distention becomes difficult to maintain and the flow of the distending medium exits at the perforation site. An alert assistant viewing by laparoscope should warn the hysteroscopist of impending perforation the moment any increasing intensity of light transmission through the thinning uterine wall is observed. If perforation is unnoticed and if simultaneous laparoscopy is not performed, a serious complication is even possible with a nonenergy instrument, but this is far less common than those occurring with lasers or electrodes. Nevertheless, if a perforation is suspected, the patient should be carefully observed in the hospital. Injuries to the iliac vessels can occur as the result of uterine perforation. An unexplained falling blood pressure, together with medium leakage, should alert the surgeon to this possibility. Perforation of the uterus during hysteroscopy can place a woman at an increased risk for uterine rupture during a future pregnancy (Fig. 17.41). FIGURE 17.41. Perforation should be immediately suspected when the endometrial cavity depressurizes and collapses around the hysteroscope, creating a compromised view of the cavity.

View Figure

Poor Visibility in the Operative Field

Inability to see the operative field is a common problem. The usual cause of this problem is deep insertion of the hysteroscope so that the telescope lies directly in contact with the endometrium. The surgeon will see nothing but a red blur. The natural tendency is to push the hysteroscope deeper in. This strategic mistake invariably leads to perforation. Another cause of visibility problems is blood within the uterine cavity secondary to dilatation. The fastest way to deal with a bloody cavity is rapid flushing with the hysteroscopic medium combined with aspiration using a cannula placed into the cavity via the operating channel. Overdilatation of the cervix is an equally common mistake that results in excessive leakage of distending medium and an inability to maintain distention, with the resultant inability to perform the operative hysteroscopy. Blood and debris can cloud the field to such a degree that accurate operative endoscopy is impossible. Overdilatation is a less common occurrence when Hyskon is selected as the distending medium. If the operator cannot clearly see the field, it is better to discontinue the procedure than to press on and risk a catastrophic error. It is easy to become disoriented in the uterine cavity if normal anatomic landmarks cannot be recognized.

Gas Embolus CO2 embolism may occur during diagnostic or operative hysteroscopy. This will happen if an inappropriate method (e.g., laparoscopic insufflator) is used to infuse CO2 into the uterus. However, this complication may also occur when a proper CO2hysteroscopic insufflator is used. The diagnosis is made by the presence of a cog-wheel murmur accompanied by a rapid fall in expired CO2 . Brundin and Thomasson observed 70 women during CO2 hysteroscopy and reported the presence of the mill (cog)-wheel murmur in seven (10%). When the hysteroscopy was stopped, the murmur disappeared. Corson et al. infused CO2 gas directly into the circulation of ewes. At 90 cc per minute, the PCO2 dropped; however, at lower flow rates only transient drops in PCO2 were observed. The investigators concluded that CO2 clearance is very efficient owing to its high solubility in blood. Air embolism is exceedingly dangerous and has been reported to occur during hysteroscopy. Corson et al. and Perry and Baughman have reported this type of complication. These investigators recommended avoiding the Trendelenburg position, purging air from all tubing and sheaths, and careful dilatating to avoid opening venous channels. Baggish and Daniell reported air embolism that resulted in death secondary to the use of gas-cooled coaxial Nd:YAG laser fibers.

Infection The endometrium seems to be peculiarly resistant to infection, and infection is an unlikely complication associated with or coming after hysteroscopy. Hysteroscopy should be avoided in the presence of gross cervical infection, uterine infection, or salpingitis. Infection is otherwise uncommon after even extensive intrauterine surgery (e.g., adhesiolysis or myomectomy). Prophylactic antibiotics should be administered only when indications such as a history of rheumatic carditis, congenital heart defect, or prolapsed mitral valve exist, or in cases of suspected chronic endometritis (submucous myoma or embedded IUD). Baggish et al. (1999) observed only 13 infections out of 5,000 cases that could be casually related to the hysteroscopic operation. Salat-Baroux et al. reported seven mild infections out of 4,000 hysteroscopic examinations. On the other hand, McCausland et al. reported three cases of tuboovarian abscess after operative hysteroscopy.

Operator Technique The most serious complications happen because of operator error. Most often, these are the result of inexperience and are avoidable. Difficult cases beyond the capabilities of the primary care gynecologist should be referred to an expert hysteroscopist. Skill in one area of endoscopy (e.g., operative laparoscopy) does not confer similar expertise in operative hysteroscopy. Indeed, the opposite may be more true.

During the postoperative period, operative complications should be the initial exclusion diagnosis for any patient who is not recovering according to the usual pattern. Worsening postoperative pain, fever, nausea, distention, and free intraperitoneal air are the signals of bowel injury. Diminished urinary output, fever, and distention suggest bladder or ureteral trauma. Falling blood pressure and rapid thready pulse, with or without distention, should raise concerns of a vascular problem and third-space hemorrhage. Most negligence cases adjudicated in favor of the plaintiff have involved delayed initiation of appropriate treatment for an operative complication. Cases involving injury recognized at the time of surgery and correctly managed in a timely fashion do not usually become medicolegal problems.

Cancer After Ablation Valle and Baggish reported on eight patients who had an endometrial ablation and who were found to have carcinoma of the endometrium between 5 months and 5 years (one patient diagnosed at endometrial resection) later. The authors identified risk factors for carcinoma in all patients, including obesity, hypertension, and diabetes. Additionally, five out of eight showed hyperplasia on preoperative biopsy. The investigators considered that women with abnormal uterine bleeding who fall into the high-risk category might be better served by hysterectomy rather than by endometrial ablation. The risk of leiomyosarcoma is less than 1%. Nevertheless, any myoma or part of a myoma that is excised should be sent to the pathology laboratory for evaluation. This, of course, includes resectoscopic fragments.

Pregnancy After Hysteroscopic Ablation Fortunately, this is not a common complication after hysteroscopic endometrial ablation. Rogerson et al. reported on four cases of pregnancy after ablation of the endometrium, with pregnancy loss of 75%. Two of the cases were associated with adherent placentas, including the one viable pregnancy that ended in the delivery of a preterm infant.

Chapter 18

Control of Pelvic Hemorrhage Howard W. Jones III William A. Rock Jr. The successful performance of any surgical procedure involves control of pain, control of bleeding, and control of infection. The subject of this chapter is control of bleeding. Recognition and correction of an abnormal hemostatic mechanism, and the prevention and control of bleeding are fundamental to the success of any operation. Preoperative, intraoperative, and postoperative hemorrhage are potential complications in every patient undergoing gynecologic surgery. Preoperative hemorrhage is encountered in a variety of circumstances, such as in patients with intraperitoneal bleeding from a ruptured tubal pregnancy or in patients taking heparin who have a massive intraperitoneal hemorrhage with ovulation. Intraoperative and postoperative hemorrhage can result from vascular injury and failure to control bleeding during surgery, and postoperative bleeding is often a carryover from bleeding, owing to reflex vasoconstriction or hypotension, that was not apparent when the abdomen was closed. In all settings (preoperative, intraoperative, and postoperative), the bleeding can be caused or aggravated by a systemic bleeding diathesis that may or may not be related to the patient's other reason for hemorrhage. Many benign gynecologic conditions are associated with an increase in menstrual blood loss (menorrhagia), an increase in the duration of menstrual flow (metrorrhagia), an increase in the frequency of menstrual periods (polymenorrhea), or combinations of all three. Repeated small menstrual hemorrhages, such as those that occur with menorrhagia, will reduce the iron stores in the body over time. The daily dietary intake of iron usually is sufficient to replace the iron lost with normal menstruation, but it is inadequate to replace the increased loss of iron associated with heavy menstruation. In gynecologic patients with a history of heavy or prolonged menstrual blood loss, it is a good idea to check the hematocrit or hemoglobin before setting a date for elective surgery. Preoperative iron supplementation is indicated in these women because a good hemoglobin level and adequate iron stores are the first step in managing perioperative hemorrhage. Transfusion before elective gynecologic surgery is rarely if ever indicated in women with chronic blood loss anemia. Menstrual blood loss may be controlled with hormonal therapy while surgery is delayed and iron supplementation given to enable the patient to replete her own hemoglobin stores. The preoperative use of epoetin alfa (recombinate erythropoietin) for correction of preoperative anemia has been used successfully in orthopedics. Its application in gynecologic surgery remains unclear. It is probably most applicable in gynecologic patients with chronic renal failure, nonmyeloid (hematopoietic) leukemia, or human immunodeficiency virus (HIV). Occasionally, other forms of anemia will be encountered that require a more extensive evaluation and treatment before elective surgery. On the other hand, some women who present with acute blood loss from a ruptured ectopic pregnancy or malignancy may require urgent transfusion even as preparations are being made for surgical intervention. FUNDAMENTAL CONCEPTS OF NORMAL COAGULATION CONGENITAL CAUSES OF INAPPROPRIATE CLOTTING IMPAIRED FIBRINOLYSIS ACQUIRED CAUSES OF INAPPROPRIATE CLOTTING PREOPERATIVE COAGULATION SCREENING COMPONENT THERAPY FOR REPLACEMENT BEFORE SURGERY COMPONENT THERAPY FOR REPLACEMENT DURING SURGERY COMPONENT THERAPY FOR POSTOPERATIVE REPLACEMENT RISKS OF BLOOD TRANSFUSION AUTOLOGOUS BLOOD TRANSFUSION BASIC SURGICAL PRINCIPLES TO AVOID EXCESSIVE BLEEDING IN PELVIC SURGERY INTRAOPERATIVE MEASURES TO CONTROL PELVIC HEMORRHAGE SPECIAL SITUATIONS AND TECHNIQUES POSTOPERATIVE BLEEDING UNUSUAL METHODS OF CONTROLLING PELVIC HEMORRHAGE

FUNDAMENTAL CONCEPTS OF NORMAL COAGULATION Part of "Chapter 18 - Control of Pelvic Hemorrhage" Every surgeon should understand at least the basic mechanisms of normal hemostasis that can be relied on when surgical injury to tissue is inflicted. Bleeding during gynecologic surgery usually results from cutting or lacerating small or large vessels, but occasionally it may result from or be complicated by some preexisting or intraoperative defect in the clotting mechanism. The surgeon should be able to recognize when normal hemostasis is interdicted, so that available remedies to protect against or remedy excessive bleeding can be found. Hemostasis is a complex, intricate, integrated, complementary, and countervailing system that maintains a delicate balance between normal coagulation and hypocoagulation or hypercoagulation. Unusual clinical situations can arise that require hematologic consultation for resolution. A specialist in coagulation disorders can provide invaluable assistance in the diagnosis and treatment of many rare disorders of coagulation. The following is a discussion of the principles and concepts of normal hemostasis, abnormal hemostasis (congenital and acquired), and management techniques. Effective hemostasis is the result of all aspects of the coagulation system functioning together to stop bleeding. Coagulation is the working interrelation of five aspects of a complex biochemical and vascular system that causes the formation and dissolution of the fibrin platelet plug. These five components are (a) vasculature; (b) platelets; (c) plasma clotting proteins; (d) fibrinolysis and clot inhibition; and (e) the hypercoagulable response. How these five components interrelate in the normal setting must be understood before one can appreciate how the five relate to bleeding or abnormal clotting in disease states.

Vasculature The vasculature presents an endothelial-lined flexible conduit through which red cells, white cells, platelets, and all of the plasma proteins flow. At the interface between the flowing blood and vessel wall are several inhibitory biochemical systems that prevent the generation of the platelet–thrombin clot. The antiplatelet substance prostacyclin, produced in the vessel wall, inhibits platelet adhesion to the vessel wall. The surface antithrombin III–heparan sulfate complex inhibits deposition of thrombin and fibrin. A tear in the vessel wall removes the endothelial cell layer, exposing the basement membrane, smooth muscle, collagen, and supporting adventitia. These substances are biochemical activators of platelets and have their own thromboplastic activity, which initiates fibrin generation and deposition. Therefore, the disruption in the vessel wall removes the protective covering of the endothelial cells, exposing platelet clumping and clot-initiating substances that produce a platelet–fibrin mass that will plug the tear in the vessel wall. A disease or medication that interferes with or intensifies this process can cause bleeding or inappropriate clotting. The vessel wall is diagrammed in Fig. 18.1. FIGURE 18.1. Vessel cross section.

View Figure

Congenital diseases associated with inadequate connective tissue and vascular dysfunction associated

with bleeding are rare. The more frequently seen conditions are hereditary hemorrhagic telangiectasia, Ehlers-Danlos' syndrome, and Marfan's syndrome, which are characterized by defects in the quality of collagen. Defective collagen is responsible for poor clot formation and platelet activation at the injured site. No disease is known to be associated with excessive inappropriate clotting related to the vasculature as a structure. The congenital diseases closest to that definition are a predisposition to atherosclerosis owing to abnormalities in lipid metabolism, such as hypercholesterolemia, homocystinemia, and diabetes mellitus. Acquired diseases of the vessel associated with bleeding include deficiencies in vitamin C, Cushing's syndrome, acute and chronic inflammatory diseases such as infectious vasculitis and immune vasculitis, pyrogenic purpura, embolic purpura, and anaphylactoid reactions from drugs. Myeloproliferative disorders, such as multiple myeloma and Waldenstr?m's macroglobulinemia, produce abnormal proteins that interfere with vascular function and therefore permit bleeding. Routine laboratory assessment of vascular function is extremely primitive. The capillary fragility test, the only routinely available test used to assess vascular function, has limited value. It is sensitive to only the severest vascular structure abnormalities. More in-depth studies include vascular biopsies and skin window testing procedures, which are research procedures. There are no routinely available methods for assessing increased vascular activity in the area of inappropriate clotting.

Platelet Function Platelets are disk-shaped fragments of the large multinucleated megakaryocytes released from the bone marrow on a daily basis (normal count is 150 × 103/µL to 400 × 103/µL) (Fig. 18.2). Their life span is 8 to 10 days. These microscopic fragments have a well-defined substructure that can be directly correlated with platelet function. FIGURE 18.2. Platelet cross section.

View Figure

The surface activation of the receptor sites on the platelet causes it to change first to a sphere and finally to a spider-like structure, with pseudopods in all directions. This release reaction is the summation of biochemical and structural changes in the platelet, which are characterized as follows. The surface receptor sets up a biochemical chain reaction, resulting in the generation of thromboxane A2. This causes contraction of the protein thrombosthenin, which causes the ejection of the platelet contents. Of great importance are the dense granules with nonmetabolic adenosine diphosphate (ADP). ADP is a potent platelet-aggregating agent that, in a domino-like sequence, stimulates more platelets, generating a large platelet plug. The congenital diseases associated with poor platelet function are divided into four types of dysfunction: (a) adhesion to collagen, (b) adhesion to subendothelium, (c) release reaction defects, and (d) ADP aggregation defects. With the exception of von Willebrand's disease, a defect in the adhesion to subendothelium, all the congenital defects are rare and not essential to this discussion. von Willebrand's disease (Table 18.1) is a classically autosomal, dominantly inherited disorder resulting from absence, decreased production, or abnormal function of a large multimeric protein synthesized by megakaryocytes

and vascular endothelium. This protein is responsible for the proper binding of platelets to the collagen surface exposed in vascular trauma. Its absence results in the failure of platelets to bind normally to disruptions in the vasculature, preventing formation of the platelet plug necessary for normal hemostasis. The condition remains undetected in most patients until some form of vascular trauma occurs or surgery is performed. In addition, such patients are particularly sensitive to aspirin or other antiplatelet medications, and bleed excessively in surgery while taking this kind of medication. von Willebrand's disease is the most common congenital platelet disorder and is the disease most likely to go undetected until surgery. This disorder is particularly dangerous because, in its milder forms, a history of bleeding in surgery is negative and the preoperative coagulation screen is normal. Acquired defects in platelet function are much more common and can be classified into two groups: (a) those that are the result or consequence of a disease, such as renal failure, myeloproliferative disorders (polycythemia vera, chronic myelogenous leukemia), and increased fibrin split products in consumptive coagulopathies; and (b) those that are iatrogenic, such as defects caused by medications (aspirin, nonsteroidal antiinflammatory drugs, antibiotics, antihistamines, tricyclic antidepressants, dextran) and cardiopulmonary bypass surgery. TABLE 18.1. More Commonly Seen Rare Congenital Clotting Disorders

Name

Incidence (per million)

Treatment

Factor VIII (classic hemophilia A, sex-linked)

60–80

FVIII concentrate

Factor IX (classic hemophilia B, sex-linked)

15–20

FIX concentrate

von Willebrand's disease (dominant; autosomal)

5–10

Cryoprecipitate (DDAVP), factor VIII concentrate with von Willebrand factor

DDAVP, Deamino-D-arginine vasopressin.

The remainder of the known congenital clotting factors are very rare and occur with such low frequency that their discussion, diagnosis, and management can be found elsewhere. (See Harker LA, Hemostasis manual, second edition, Philadelphia: FA Davis, 1974; Corriveau DM, Fritsma GA, Hemostasis and thrombosis, Philadelphia: JB Lippincott Co, 1988; Triplett DA, ed. Laboratory evaluation of coagulation. Chicago: ASCP Press, 1982.)

The remainder of the known congenital clotting factors are very rare and occur with such low frequency that their discussion, diagnosis, and management can be found elsewhere. (See Harker LA, Hemostasis manual, second edition, Philadelphia: FA Davis, 1974; Corriveau DM, Fritsma GA, Hemostasis and thrombosis, Philadelphia: JB Lippincott Co, 1988; Triplett DA, ed. Laboratory evaluation of coagulation. Chicago: ASCP Press, 1982.)

Congenitally increased platelet function has not been described. Acquired disorders associated with increased platelet function, however, are common. The stress of routine surgery or trauma (fractured hip, femur, or pelvis) can create a hypercoagulable state with thrombocytosis and increased platelet activity. The laboratory assessment of platelet function has been expanded from the research laboratory and is more readily available to the surgeon. The routine analysis of platelet function should begin with a platelet count and PFA-100. In special cases, platelet adhesion and platelet aggregation are useful in identifying the inadequate or overstimulated platelet. In addition, biochemical markers for increased platelet use or turnover can be demonstrated with platelet factor IV and ?-thromboglobulin assays. Recent studies by Gewirtz et al. confirm previous studies that the bleeding time is not a good prediction of surgical bleeding.

Plasma-Clotting Proteins Plasma-clotting proteins are a group of serine proteases and cofactors that interact in a synergistic system to generate fibrin. The activation of the clotting system can be initiated in two ways: either by contact activation with factor XII or through thromboplastin activation of factor VII. The clotting cascade is diagrammed in Fig. 18.3. As we will see later in the discussion of fibrinolysis and antithrombin systems, anticoagulation forces are initiated at the inception of clotting. The tear in the vessel wall, described earlier, begins the orderly activation of the plasma-clotting system. The fibrin contribution to the platelet—fibrin plug is initiated with the activation of factor XII by collagen and of factor VII by tissue juice (thromboplastin). Any congenital or acquired disorder of the clotting factors can lead to inadequate or no generation of fibrin. Each clotting factor has a different role and significance in the overall generation of fibrin. This also is true with abnormal increases in some clotting factors that are associated with inappropriate clotting. FIGURE 18.3. Coagulation system. Dashed boxes indicate destruction of factors. HMKa, high-molecular-weight kininogen; PF3, platelet factor 3; TPA, tissue plasminogen activator; TFPI, tissue factor pathway inhibitor.

View Figure

The congenital-factor deficiencies associated with bleeding are either relatively common or rare. The relatively common group includes hemophilia A (factor VIII deficiency) and hemophilia B (factor IX deficiency). Both are seen in the male and rarely in the female disorders with sex-linked inheritance patterns. The rare group includes all the remaining factors that have an autosomal recessive inheritance pattern or a dominant pattern with variable penetrance. The acquired factor deficiencies are common. Multiple deficiency is usually owing to iatrogenic vitamin K deficiency with loss of factors II, VII, IX, and X. This deficiency often is the result of multiple-antibiotic therapy, which kills the vitamin K–producing bacterial flora in the intestine, and the nothing-by-mouth status of many critically ill patients, which results in the loss of food sources of vitamin K. Other common acquired multifactor deficiencies are seen in acute and chronic liver disease, as in viral hepatitis and alcoholic cirrhosis; consumptive coagulopathies, as in sepsis and placenta abruptio; washout coagulopathies, as in multiple-transfusion patients after severe blood loss (such as from ruptured abdominal aneurysms); and major trauma, as from automobile accidents or gunshot wounds.

The laboratory assessment of the plasma clotting factors has traditionally begun with the prothrombin time (PT; factors V, VII, and X, prothrombin, and fibrinogen) and the activated partial thromboplastin time (APTT; factors VIII, IX, XI, and XII). Specific factor assays also can identify the exact deficiencies. One must remember that a factor deficiency as low as 30% can generate a normal PT and APTT. This relation is important in investigating minimal prolongations of the PT or APTT that appear insignificant but could be hiding a moderately severe deficiency. The tissue factor pathway inhibitor modulates activated factors X and VIII but is not apparently significant in disease. The sensitivity of the PT and APTT reagents is essential to the appreciation of the proper use of these tests as preoperative screening tests, or in monitoring warfarin and heparin anticoagulant therapy. Recent publications from Europe and the United States stress the importance of and need for a standardized prothrombin reagent system in the United States. The lack of sensitivity of the rabbit brain thromboplastin used in the United States has led to the over-coumarinization of some patients. The original value of 2.0 to 2.5 times the control was based on the more-sensitive human thromboplastin. Current recommendations have lower ratios (Table 18.2). These ratios are applicable only in stable, coumarinized patients. Studies of different APTT reagents have revealed a similar variability of sensitivity to heparin. TABLE 18.2. Therapeutic Ranges for the International Normalized Ratios

Condition

Therapeutic Ranges

Prophylaxis for venous thrombo-embolism in high-risk surgery and in hip surgery

2.0–3.0

Treatment of venous thrombosis and pulmonary embolism

2.0–3.0

Prevention of systemic embolism Tissue heart valves Acute myocardial infarction Valvular heart disease Atrial fibrillation Bileaflet mechanical value in aortic position

Treatment for mechanical prosthetic heart valves (high risk)

Prevention of recurrent systemic embolism

Prevention of recurrent myocardial infarction

2.5–3.5

From: Hirsh J, Dalen JE, Anderson DR, et al. Oral anticoagulants: mechanism of action, clinical effectiveness, and optimal therapeutic range. Chest 2002;119:8S.

Fibrinolysis The activation of the fibrinolytic system begins with the activation of the plasma substrate plasminogen. This substrate is converted by naturally occurring activators such as urokinase, kallikrein, and clot-activated proteases to the active enzyme plasmin. Plasmin is the active enzyme that, if free or clot-bound, lyses fibrin clots and destroys fibrinogen. This enzyme is modulated by ?2-antiplasmin and antitrypsin, which destroy the active enzyme plasmin. This enzymatic conversion of fibrinolysis normally is initiated by clot formation or by a direct activator such as urokinase or tissue plasminogen activator (tPA). tPA released from the endothelium activates tissue plasminogen and is neutralized by PAI-1 inhibitor. Sometimes direct activation is seen in liver disease and during extracorporeal bypass. This activation also can be secondary to disease, as in a consumptive coagulopathy, such as bacterial sepsis, or a large abdominal aneurysm.

Hypercoagulable State With physiologic stress, such as emotional stress and surgical stress, there is a response of fright or flight. This response to stress is evident in the coagulation system. The plasma-clotting proteins, such as fibrinogen and factor VIII, increase, and the platelet count and stickiness can increase as well. This normal response is important in ensuring hemostasis at the time of increased need. When this process is exaggerated, uncontrolled, or unmodulated, inappropriate clotting can occur, which produces venous and arterial clots and all their sequelae. In gynecologic surgery, the normal physical hypercoagulable state, as well as the inappropriate state, must be understood to appreciate the diagnosis, intervention, and management of postoperative vascular occlusive complications. Virchow, in 1845, was the first to conceptualize the triad of blood flow, vessel wall, and content of blood itself as a basis for inappropriate clotting. An understanding of the relation of the three parts is essential to explain what has occurred in the problem patient.

CONGENITAL CAUSES OF INAPPROPRIATE CLOTTING Part of "Chapter 18 - Control of Pelvic Hemorrhage" The congenital etiology of inappropriate arterial and venous clotting has long been ill defined. Only recently has it been more completely elucidated (Table 18.3 and Table 18.4). Procoagulants, when increased on a congenital basis, have been associated with a propensity to generate clots. These procoagulants include fibrinogen and factor VIII; however, they are not present frequently enough to warrant testing every suspect case. Naturally occurring inhibitors of clotting are defined as those factors that actively destroy clotting factors or substrates as they are formed. The more common of these rare deficiencies are antithrombin III, protein C, protein S, factor V Leiden (R506Q), factor II mutation (G21201A), and methylenetetrahydrofolate reductase mutation (C677T).

TABLE 18.3. Risk Factors for Arterial Thrombosis

INHERITED Elevated cholesterol, triglycerides, lipoprotein (a), decreased high-density lipoprotein Diabetes FVII polymorphism Hyperhomocysteinemia Methylenetetrahydrofolate reductase mutation C677T PLA2 glycoprotein IIb/IIIa Gender, male > female Acquired Antiphospholipid antibodies Lupus anticoagulant Hypertension Diet with increased fat Infection: chlamydia, cytomegalovirus Heparin-induced thrombocytopenia Social class, body mass index MIXED HEREDITARY/ACQUIRED Factor VIII Fibrinogen FVII Homocysteine C-reactive protein Von Willebrand factor

From: Triplett DA. Thrombophilia: laboratory evaluation. ASCC Clinical Laboratory News 2002;28:12.

TABLE 18.4. Risk Factors for Venous Thrombosis

INHERITED

INHERITED Common Factor V Leiden (R506Q) Factor II Mutation (G21201A) Factor VIII Rare Antithrombin III deficiency Protein C deficiency Protein S deficiency PAI-1 polymorphism Dysfibrinogenemia Factor XII deficiency Prekalikrein (Fletcher factor) deficiency Plasminogen deficiency Tissue plasminogen activator deficiency ACQUIRED Surgery and trauma Prolonged immobilization Older age Cancer Myeloproliferative disorders Previous venous thrombosis Pregnancy/puerperium Contraceptives/hormone replacement APC resistance not due to FV Leiden Antiphospholipid antibodies Mild-to-moderate hyperhomocysteinemia Obesity

From: Seligsohn U, Lubetsky A. Genetic susceptibility to venous thrombosis. N Engl J Med 2001;344:1222.

IMPAIRED FIBRINOLYSIS Part of "Chapter 18 - Control of Pelvic Hemorrhage" A congenital decrease in the plasma substrate plasminogen results in inadequate fibrinolysis of thrombi. This deficiency can be qualitative and quantitative, with similar effects. A congenital decrease in tPA that normally is released from the vascular endothelium is associated with impaired fibrinolysis. An abnormal increase in plasminogen activator inhibitor also will reduce the level of tPA, resulting in inappropriate clotting. The decrease or absence of Fletcher factor (prekallikrein) and factor XII also can result in impaired fibrinolysis because of a decrease in activation of circulating plasminogen at the time of clot activation.

ACQUIRED CAUSES OF INAPPROPRIATE CLOTTING Part of "Chapter 18 - Control of Pelvic Hemorrhage" The number of acquired causes of inappropriate clotting is much greater than the number of congenital causes and is expanding every day because the same chemistry found in the congenital mechanism can be identified as a deficiency in an ongoing disease process. Factor VIII has been shown to determine the rate of thrombin production and is a cause of thrombogenesis and coronary artery disease. Increases in dietary fat also increase factor VIII:C levels, resulting in an increased aggregability of platelets, which causes platelet thrombi to increase. Also noteworthy is the fact that smoking as a cause of coronary artery disease may be mediated through a rise in fibrinogen. Natural physiologic states also can increase the levels of plasma-clotting factors. Instead of a single factor being the cause of inappropriate clotting, in these cases, it is likely that the complementary activity of all factors working synergistically produces inappropriate clotting. In pregnancy, factor VIIIc and fibrinogen are increased. A common reaction to trauma such as a leg fracture or surgery is an increase in factor VIIIc and fibrinogen levels and in the platelet count. Disease states associated with inappropriate clotting include both acute and chronic forms. The acute forms are seen in diseases such as thrombotic thrombocytopenic purpura (owing to the acquired or congenitally absent von Willebrand factor cleaving protease) and nephrotic syndrome, with loss of antithrombin III in the urine along with other plasma proteins. The chronic forms are seen in diseases such as diabetes mellitus (endothelial hyperplasia of smaller arterioles, reduced prostaglandin I2 production, and hypersensitivity of platelets), heavy cigarette smoking, and diets high in fat and cholesterol. Neoplastic diseases such as carcinoma of the lung, colon, and prostate are associated with severe thromboembolic complications. Myeloproliferative diseases, including polycythemia vera, chronic myelogenous leukemia, and essential thrombocythemia, are associated with inappropriate clotting. The lupuslike inhibitor and the anticardiolipin antibodies seen in lupus patients and in patients with infectious diseases and other autoimmune diseases are associated with inappropriate clotting and spontaneous abortions. Iatrogenic causes of inappropriate clotting are common findings in the hospital setting and generate great concern. Such causes include the postsurgical state, medication, vascular prosthetic devices, and immobilization for any reason. As a physiologic acute-phase response to surgical stress, an exaggerated outpouring of clotting factors and platelets in combination with a decrease in physiologic inhibitors can result in clot formation. This often occurs in deep leg veins, particularly in association with venous stasis. Prosthetic devices such as grafts, shunts, and artificial heart valves can provide a clottable surface that will form a nidus for initial thrombosis quickly followed by further clot formation, resulting in obstruction or embolization. The vascular component of acquired thrombotic disease has only recently been described in detail. It appears that decreased blood flow through a vein can decrease the contact between thrombin and thrombomodulin, diminishing the contact with protein C and predisposing the vein to thrombosis. However, the arterial side with high blood flow rates has a rich capillary bed with greater contact with protein C, lysing clots more efficiently. Local thrombus formation can be generated by direct mechanical disruption of the vascular endothelium, traumatic damage to the vessel wall, infectious or chemical damage to the vessel wall, and vasculitis.

PREOPERATIVE COAGULATION SCREENING Part of "Chapter 18 - Control of Pelvic Hemorrhage" For the preoperative evaluation, gynecologic patients must of necessity be divided into two categories: those having routine or elective surgery and those having emergency surgery.

Elective Surgery The elective gynecologic surgical patient must be evaluated in two ways: general medical history and specific nature of the surgery. The medical history taken at bedside, with review of the medical chart when available, is an excellent place to begin. Table 18.5 highlights the most important positive and negative findings to be identified. TABLE 18.5. Pertinent Medical History to Screen for Coagulation Problems

History of spontaneous bruising or bleeding History of unusual bruising or excessive bleeding after surgery Family history of bruising or bleeding after surgery Medication associated with bruising or bleeding Current medication within past week Previous coagulation testing Current coagulation testing

Preoperative coagulation screening is of limited value without complete knowledge of the patient's past and current history. It does not replace a good history and physical examination. One should not expect this screening to reveal the estimated blood loss in a routine surgical procedure. It is essential, however, for resolving and eliminating risk factors that can affect postoperative bleeding (Table 18.6 and Table 18.7). TABLE 18.6. Tests to Indicate Coagulation Status

Test

Reference Range*

Level of Alarm

Significance

Hematocrit (%)

37–47

25

Tissue anoxia

White cell count (µL)

4 × 103—12 × 103

3 × 103 –25 × 103

Susceptibility to infection, leukemia

Platelet count (µL)

140 × 103–400 × 103

100 × 103–700 ×

Bleeding, myeloproliferative disorder

103

Fibrinogen (mg/dL)

150–400

100

Bleeding, liver disease, intravascular consumption

Prothrombin time (s)

10–13

14

Bleeding factor deficiency

Activated partial thromboplastin time (s)

28–38

40

Bleeding factor deficiency, inhibitor

Clot retraction

Complete clot in 60 min: retraction complete in 120 min

No clot Clot lysis

Low platelets or fibrinogen Fibrinolysis

PFA-100

Collagen–epinephrine

Prolonged closure time

Screen for medication effect Bleeding (will not predict surgical bleeding)

*Reference ranges may vary in each laboratory, reflecting method, instrumentation, and reagents.

TABLE 18.7. Coagulation Profiles

Brief Coagulation Profile

Complete Coagulation Profile

CBC (includes WBC differential)

CBC (includes WBC differential)

Platelet count

Platelet count

Prothrombin time

Prothrombin time

Partial activated thromboplastin time

Partial activated thromboplastin time Fibrinogen Bleeding time

CBC, complete blood count; WBC, white blood cell.

Risk factors such as unknown history, or known history in an emergency surgical procedure; positive personal or family history of bleeding or bleeding with or without surgery; and known history of taking medications that can affect coagulation, such as antiplatelet P.420 medication, acquired vitamin K deficiency (nothing-by-mouth status with long use of antibiotics), and fibrinolytic therapy (decreased fibrinogen), are assessed by preoperative screening. Preoperative coagulation screen is not usually indicated unless the medical history and physical examination reveal suspicious or explained findings that suggest a risk of surgical bleeding. Items such as a history of unexplained surgical bleeding, family history of bleeding, bleeding after medication, or evidence of bruising or bleeding on examination to mention only a few. The decision to transfuse blood and blood components must be made with all the current knowledge of the patient's status. The surgeon must actively seek the patient's past history, hematology and coagulation test results, and chemistry results as appropriate (Fig. 18.4). The surgeon must be aware of the patient's hematologic and coagulation status throughout the case. Then, and only then, does the proper selection of

blood components solve problems. The surgeon's surgical dictation and progress notes also should reflect the observations, test results, and course of action taken. FIGURE 18.4. Evaluation of candidate for elective surgery.

View Figure

The risk of bloodborne infections and adverse reactions is always present, but the documented need for blood as a life-saving substance will validate the decision. When blood is transfused when indicated but is not justified in writing, this life-saving substance becomes a liability to all who use it. The routine preoperative orders for blood require knowledge of the specific needs of the patient and the surgeon's usual transfusion requirements for a specific surgical procedure. For the routine gynecologic procedure, such as simple hysterectomy in an otherwise healthy woman, a type and antibody screen are appropriate. With the type and antibody screen, the patient's blood is screened for unexpected antibodies. No specific blood units are set aside, but blood is available from the general inventory in an emergency. If an unexpected antibody is identified, the blood bank should notify the ordering physician and set aside 2 U of antigen-negative crossmatched compatible blood for use in an emergency situation. In an emergency, the blood bank can release blood immediately (with a type and match to follow) with a 99.99% safety factor when the previous screen for unexpected antibodies was negative. Additionally, when the surgeon can wait 10 to 15 minutes, an immediate spin crossmatch can be performed to further verify ABO compatibility between donor and recipient. The value of the type and antibody screen is in monetary savings for the patient, and there is no undue or unnecessary risk to the patient. In more complex procedures, such as pelvic exenteration for cancer, where there usually is significant blood loss, a type and crossmatch for the average number of units used is appropriate. With extremely difficult procedures or other complicating diseases, additional blood, fresh-frozen plasma, and platelets may be required during the procedure and should be requested preoperatively. Ideal or time-proven guidelines are difficult to establish for every operative case. Each surgical experience will benefit the surgeon, and over time he or she will establish usual transfusion requirements for both type and antibody screen, as well as type and crossmatch. The hospital quality assurance program, in planning with the transfusion service or blood bank and transfusion committee, should establish guidelines to assist the surgeon in identifying the usual blood transfusion needs. The use of either Guidelines for Transfusion Therapy (Boral) or Maximal Surgical Blood Order Schedule (Judd) is helpful in developing hospital guidelines.

Emergency Surgery As the emergency procedure is begun, decisions regarding blood replacement must be made. A direct approach to blood replacement therapy and the complications of such therapy depends on a clear understanding of the following concepts. As bank blood replacement with just packed red cells corrects the blood loss problem, it may create an acquired bleeding disorder. Thrombocytopenic hemophilia. Platelets and fresh-frozen plasma may be indicated. The patient's bleeding potential is dynamic and will change rapidly and frequently with the loss of blood and replacement therapy. Direct monitoring before, during, and after surgery offers the best chance to diagnose and manage the bleeding. Direct monitoring also allows formulation of plans and adjustment of the replacement therapy program.

COMPONENT THERAPY FOR REPLACEMENT BEFORE SURGERY Part of "Chapter 18 - Control of Pelvic Hemorrhage" With surgery planned, the preoperative data can be evaluated. Assuming the patient does not have hemophilia, von Willebrand's disease, severe liver disease, or liver failure, a prolonged PT and APTT may suggest a less common acquired or congenital bleeding disorder. (The blood sample must be properly drawn and mixed well and must not be taken from an A-line containing heparin or from an infusion site.) Assistance from a clinical pathologist or hematologist should be requested if an intrinsic bleeding disorder is suspected. von Willebrand's disease is the most common hereditary bleeding disorder transmitted predominately as an autosomal dominant defect. Effecting men and women equally, this disease often undiagnosed at the time of surgery can be a cause of increased morbidity and mortality. von Willebrand's disease presents commonly with a history of mucosal bleeding, such as epistaxis, or with a history of easy bruising or menorrhagia. A familial history of this kind of bleeding in many female members of a family is also a clue to the possible presence of von Willebrand's disease. This missing von Willebrand factor is manufactured in the vessel endothelial cell Weibel-Palade bodies and megakaryocytes. This factor has a major role in hemostasis, including (a) binding platelets to subendothelial collagen, and (b) joining with and stabilizing circulating factor VIII. Without von Willebrand factor, platelets adhere poorly to damaged endothelium, and clotting factor VIII can be significantly reduced. Treatment will evolve around the type, severity of the disease, and the nature of the surgical procedure. The more common type I and some of the rarer variants can be managed with DDAVP (deamino-D-arginine vasopressin; a vasopressor analog) that increases the release of the stores of von Willebrand factor in epithelial Weible-Palade bodies. This is contraindicated in most cases of type IIB. Cryoprecipitate and the manufactured Humate P (fractionated FVIII) are sources of von Willebrand factor that are infused before, during, and after a surgical procedure. The amount and frequency of replacement will depend on the type of surgery, blood loss, and risk of bleeding in a critical tissue or structure. Diagnosis and management of this disorder is often complex and difficult. Consultation with a clinical pathologists or hematologist should be considered. If the patient is bleeding before surgery, packed red blood cells should be given. If bleeding is severe, fresh-frozen plasma, cryoprecipitate, or platelets should be given as indicated. Whole blood (8 days old) is deficient in coagulation factor V and factor VIII in the plasma portion, as well as in platelets. Although the levels of these factors can produce normal PT and APTT readings, they are insufficient for a patient undergoing surgery and blood loss. (Mild hemophilia also can produce a normal APTT reading.) When packed red blood cells and fresh-frozen plasma are not available, whole blood can be used, although it may create a greater coagulation deficit when given in large amounts, making intraoperative monitoring even more critical. The goals of emergency preoperative screening are as follows: To determine whether a coagulation defect exists before surgery is begun and possibly to identify the cause To establish a baseline for assessing the changes owing to massive blood replacement and the success of specific component therapy To establish immediate component therapy needs The minimum preferred values to be achieved before surgery are listed in Table 18.8. TABLE 18.8. Minimum Preferred Values Before Surgery

Hematocrit

>25%

Platelet count

>150 × 10 3/µL

Fibrinogen

>150 mg/dL

Prothrombin time

21

Curettage (D&E)

98.6

93.7

85.0

Intrauterine saline instillation

0.0

0.3

0.7

Intrauterine prostaglandin instillation

0.2

2.9

2.9

Medical (nonsurgical)

0.2

0.6

0.9

Other

1.0

2.5

10.5

TOTAL

100.0

100.0

100.0

From Koonin LM, Strauss LT, Chrisman CE, et al. Abortion surveillance: United States, 1997. In: CDC surveillance summaries, December 8, 2000. MMWR 2000;49 (no. SS-11):1, with permission.

Based on data from the 1970s, D&E is clearly safer than alternative methods through 16 weeks' gestation. At later gestational ages, the distinction blurs between D&E and labor induction methods in terms of morbidity and mortality. Hence, the choice of abortion method at this later stage usually hinges on nonmedical considerations: cost, convenience, comfort, and compassion. Errors in estimating gestational age, especially underestimation, can have serious consequences during a D&E procedure. Hence, confirmation of gestational age by ultrasonography is important before D&E. D&E differs from suction curettage in two principal ways: D&E requires wider cervical dilation, and physicians need forceps to evacuate more advanced pregnancies. To achieve adequate dilation, many physicians insert osmotic dilators several hours to several days before D&E (Fig 21.13). For example, five laminaria placed overnight result in 1.5- to 2-cm dilation with minimal or no discomfort for most women.

Use of a single Lamicel for about 4 hours produces as much dilation as do several laminaria at 14 to 16 weeks' gestation. FIGURE 21.13. Laminaria in place after overnight preparation of the cervix. (From: Grimes DA, Hulka JF. Midtrimester dilatation and evacuation abortion. South Med J 1980;73:448, with permission.)

View Figure

Patients must understand that once osmotic dilators have been inserted, the abortion needs completion. Rarely, a patient changes her mind about abortion after placement of an osmotic dilator. Although some women have continued their pregnancies uneventfully after removal of the devices, others have developed severe chorioamnionitis and aborted. Dilating the cervix to a large diameter over several minutes can damage the cervix. Indeed, the first large study of this question revealed a higher incidence of low-birthweight infants in subsequent desired pregnancies. Hence, in the absence of evidence to the contrary, D&E procedures beyond about 14 weeks' gestation should use osmotic dilators. In the 13- to 16-week interval, vacuum alone is adequate; thereafter, forceps extraction predominates. Although some physicians have advocated administering feticidal agents such as digoxin or potassium chloride under ultrasound guidance, a randomized, controlled trial by Jackson et al. found no benefit of this intrusive practice. A cannula 14 mm in diameter can evacuate pregnancies through about 16 menstrual weeks' gestation. For later pregnancies, the cannula primarily drains amniotic fluid at the beginning of the evacuation and draws tissue into the lower uterus for forceps extraction. Specially designed forceps for D&E are far superior to standard sponge forceps. As with suction curettage, extraction should occur from the lower uterus to minimize the risks of perforation. Some physicians use a flexible 8-mm cannula to confirm complete evacuation. The physician must confirm completion by identifying all major fetal parts (extremities, spine, and calvarium). The calvarium is the component most frequently missed during the initial evacuation. Gentle exploration of the fundal and cornual areas with a large curette or forceps usually enables location and removal of the calvarium. Intraoperative ultrasonography can be helpful. If ultrasonography is not available, the physician may be able to remove the speculum and insert one digit into the uterine cavity to locate the missing part. If the abortion cannot be completed with ease, the physician should interrupt the procedure. A safe and simple remedy is to discontinue the operation and administer intravenous oxytocin to the woman for 2 to 3 hours in the recovery room. After the patient returns to the operating room, the physician usually finds the retained tissue at the internal os, from which it can be removed in a few seconds. No D&E abortion needs to finish in a single session; time often helps. Once membranes rupture, the uterus contracts to expel its contents. In D&E abortions, unlike medical abortion, the skill of the physician is critical. D&E abortion is an eclectic area of gynecology. Physicians should not dabble in these activities. This is not to say that D&E is too difficult for most physicians to learn. On the contrary, residents can quickly learn to do D&E procedures skillfully. For example, in a number of institutions, residents learn to perform D&E abortions with local anesthesia and ultrasound guidance up to 22 weeks' gestation. A before–after study from San Francisco General Hospital found use of ultrasound guidance for residents learning D&E was associated with a dramatic lowering of the perforation rate and a reduction in operating time. Facility with suction curettage is a prerequisite to learning to perform D&E abortions. The physician should study operative technique. He or she should then observe and assist skilled physicians and then perform D&E procedures only under direct supervision. The gestational age range can advance as skill grows. In

summary, D&E is not a trivial undertaking. However, like vaginal hysterectomy, it can be learned.

Labor Induction Although D&E has supplanted many labor-induction abortions, the need for such abortions continues, particularly at later gestational ages. In contrast to D&E, the proportion of abortions performed by labor induction increases with gestational age (Table 21.1). Abortifacients include two broad groups: hypertonic solutions (e.g., saline or urea) and uterotonic agents (e.g., oxytocin or misoprostol). The mechanism of action of hypertonic solutions is unclear, but these agents usually result in fetal death from osmotic insult; labor then usually ensues. Uterotonics act directly on the myometrium to stimulate contractions. Common doses of hypertonic solutions include 200 mL of 20% saline or 80 g of urea. Prostaglandins available in the United States include misoprostol, prostaglandin E2 (PGE2 ) vaginal suppositories, and 15-methyl PGF 2? for intramuscular injection. Instillation of the abortifacient requires amniocentesis. Some physicians prefer a “blind” insertion in the midline, several centimeters inferior to the level of the fundus. Others mark on the woman's abdomen the location of a pocket of amniotic fluid identified at ultrasonography. Still others use real-time ultrasonography during the needle insertion. Preparation for the amniocentesis is similar to that for diagnostic amniocentesis. The bladder should be empty. Aseptic technique should apply. The physician infiltrates the full thickness of the abdominal wall with several milliliters of local anesthetic. After needle insertion, a free flow of clear amniotic fluid should occur. Nitrazine paper or urine dipsticks for protein can differentiate amniotic fluid from urine if necessary. Blood-tinged fluid that clears does not present a problem; grossly bloody fluid that does not clear is a contraindication to injection of hypertonic solutions. If the uterus is small, lifting the uterus anteriorly with two fingers in the vagina can facilitate amniocentesis. Draining amniotic fluid is unnecessary before the injection of abortifacient: withdrawing fluid does not shorten the induction-to-abortion time. On the other hand, withdrawing several hundred milliliters of fluid (and the corresponding decrease in the size of the uterus) may displace the needle tip. Hence, remove only enough fluid to confirm free flow. Gravity-drip infusions should administer hypertonic solutions. If the physician loses correct needle placement during the injection, the flow stops. On the other hand, with syringe injection, the physician can inject hypertonic solutions into the myometrium in this situation. Careful observation of the patient is necessary during instillation. The patient should not receive a sedative or narcotic. She should be alert and watchful for any symptom (e.g., burning abdominal pain, severe thirst, headache, or nausea) that might indicate faulty administration or rapid systemic uptake of the abortifacient. Any such symptom dictates immediate cessation of the infusion and close observation of the patient. If the symptom persists, the instillation should stop, and the physician should choose an alternative method or agent for the abortion. Administration of prostaglandins has the advantage of simplicity. Amniocentesis is unnecessary, and inadvertent intravascular administration of hypertonic solutions cannot occur. Disadvantages of prostaglandins alone for abortion include a high frequency of nausea, vomiting, and diarrhea. Routine prophylactic use of antiemetics and antidiarrheal drugs can reduce but not eliminate these noxious side effects. Fever occurs in about one-third to one-half of patients given prostaglandin E2. Another serious drawback of prostaglandins alone for abortion is that these drugs are not inherently feticidal. Vaginal PGE2 suppositories appear to be more effective, but more noxious than intramuscular 15-methyl PGF 2?. With the 20-mg vaginal suppositories given every 3 hours, the mean abortion time is about 13 hours, and 90% of abortions occur within 24 hours. With intramuscular 15-methyl PGF 2?, mean abortion times are longer, and about 80% of abortions occur within 24 hours. Hypertonic solutions are useful in combination with uterotonic agents such as vaginal or parenteral prostaglandins. The discovery of the uterotonic effects of misoprostol led to trials of this drug for midtrimester abortion. Regimens with vaginal administration of misoprostol ranging from 100 to 200 µg vaginally every 12 hours have achieved success rates of about 90% within 48 hours. A randomized, controlled trial found the efficacy of 200 µg every 12 hours to be comparable to that achieved with vaginal PGE2 20 mg suppositories given every 3 hours. On the other hand, the misoprostol regimen was much less expensive and less noxious. Another trial found that 15-methyl PGF2? 2.5 mg injected in the amniotic fluid resulted in a higher success rate at 24 hours than did misoprostol 200 µg administered vaginally twice, 12 hours

apart. However, misoprostol use does not require amniocentesis. Interest in high-dose oxytocin infusion for abortion has renewed. Small comparative studies have suggested that this method may be a safe alternative to vaginal PGE2 suppositories given every 4 hours. The frequency of fever, vomiting, and diarrhea was significantly lower with oxytocin. One regimen includes oxytocin, 50 U in 500 mL of dextrose and normal saline administered intravenously over 3 hours; maintenance fluid (dextrose in normal saline) then follows for 1 hour. In stepwise fashion, the concentration of oxytocin increases by 50 U every 4 hours to a maximum of 300 U/500 mL. The investigators administered oxytocin in isotonic fluid and interrupted the oxytocin infusion every 4 hours for diuresis. With prolonged high-dose oxytocin infusion in hypotonic solutions, water intoxication and death can result. Much of the morbidity (and mortality) associated with labor-induction abortion is preventable. Women in labor with abortions need the same meticulous, attentive obstetric care as do women in labor with childbirth. Induction-to-abortion times of 13 to 24 hours appear to have the lowest complication rates; thus, abortion within this interval should be the goal. Serious complications increase significantly with increasing induction-to-abortion times. If labor is ineffective, the physician should stimulate labor. If the membranes rupture, then labor must conclude within a reasonable period. Similarly, active management of a retained placenta after abortion prevents morbidity. A passive approach to desultory labor in the presence of ruptured membranes exposes the woman to unnecessary risk of complications.

Ancillary Measures Several adjuncts can expedite instillation abortions. Administering intravenous oxytocin shortens induction-to-abortion times with saline instillation. Similarly, prostaglandins can augment labor; these agents are the pharmacologic treatments of choice for slow or failed instillation abortions. Direct cervical dilation is also useful. Osmotic dilators shorten induction-to-abortion times and protect against cervicovaginal fistulae, although this protection is not absolute. Sequential packing of dilators can be useful if uterotonic agents do not achieve adequate dilation. Alternatively, if progress stalls, a metreurynter can accomplish abortion. The physician inserts a sterile Foley catheter with a 30- to 75-mL balloon into the uterus, inflates the balloon with a sterile solution (not air), and ties the catheter to 0.5-kg orthopedic traction at the foot of the bed. A liter bag of intravenous fluids hung over the foot of the bed and tied to the catheter by a string also suffices. This method has the disadvantage of placing a foreign body in the uterus. In many cases, the preferred means of concluding a slow induction abortion is D&E. Twenty-four hours is a reasonable limit for labor-induction abortions. Frequently, the cervix is open several centimeters, and D&E proceeds quickly.

Hysterotomy and Hysterectomy Neither hysterotomy nor hysterectomy should be a primary method of abortion. The morbidity, mortality, expense, and pain associated with these operations are greater than with alternative methods. Hysterotomy for abortion is an archaic operation that should be used only when usual surgical and medical approaches fail. Fig. 21.14 depicts a hysterotomy to evacuate a 14-week pregnancy in the left horn of a bicornuate uterus; both labor induction and attempts to enter to pregnant horn through the cervix with ultrasonography and laparoscopy guidance had failed in this primigravida. Hysterectomy is appropriate in rare cases in which preexisting pathology, such as large leiomyomas (Fig. 21.15) or carcinoma in situ of the cervix, justify hysterectomy.

FIGURE 21.14. Hysterotomy incision in left uterine cavity, 14 weeks' gestation, after failed labor-induction abortion and failed attempted dilation and evacuation. Patient had a single cervix and two cavities. Non-pregnant cavity seen at right.

View Figure

FIGURE 21.15. Total abdominal hysterectomy for abortion at 15 weeks' gestation in a patient with multiple symptomatic leiomyomas with resultant anemia. Gestational sac and fetus visible in situ. Bottom: 15-cm ruler.

View Figure

Complications

Morbidity Legal abortion in the United States is safe. Fewer than one woman in 100 develops a serious complication, and fewer than one in 100,000 dies as a result of the operation. Gestational age is one of two important determinants of the likelihood of morbidity. In Table 21.2, which lists P.500 serious complication rates for abortion by gestational age, the term serious complication rates refers to the percentage of women who had fever of 38°C or higher for 3 or more days, hemorrhage requiring transfusion, or unintended surgery. These data, derived from a 1970s multicenter study including 84,000 abortions, relate to those women without concurrent sterilization or preexisting medical conditions and for whom follow-up information was available. Abortions performed at the 7 to 10 weeks' gestation interval had the lowest incidence of serious complications. Thereafter, complications increased progressively with advancing gestational age. The finding that serious complications are more frequent at or before 6 weeks'

gestation than at later gestational ages is consistent with two previous large studies in the United States. TABLE 21.2. Serious Complication Rates for Legal Abortions by Gestational Age: United States, 1975–1978*

Gestational Age (wk)

Rate†

?6

0.4

7–8

0.2

9–10

0.1

11–12

0.3

13–14

0.6

15–16

1.3

17–20

1.9

* For women with follow-up and without concurrent sterilization or pre-existing conditions. Serious complications include temperature of 38°C or higher for 3 days or more, hemorrhage requiring blood transfusion, and any complication requiring unintended surgery (excluding curettage).

† Per 100 abortions.

The method of abortion is the second principal determinant of the likelihood of complications. Table 21.3, derived from the same study as Table 21.2, demonstrates that suction curettage is the safest available abortion method. The risk of serious complications with D&E at 13 weeks' gestation or later is higher than that with suction curettage and lower than that with labor-induction abortion.

TABLE 21.3. Serious Complication Rates for Legal Abortions by Method: United States, 1975–1978*

Method

Rate†

Suction curettage

0.2

Dilation and evacuation

0.7

Saline instillation

2.1

Prostaglandin instillation

2.5

Urea—prostaglandin instillation

1.3

* For women with follow-up and without concurrent sterilization or pre-existing conditions. Serious complications include temperature of 38°C or higher for 3 days or more, hemorrhage requiring blood transfusion, and any complication requiring unintended surgery (excluding curettage).

† Per 100 abortions.

Abortion complications have three temporal categories: immediate, delayed, and late complications. Immediate complications are those that develop during or within 3 hours of the operation. Delayed complications occur more than 3 hours and up to 28 days after the procedure. Late complications develop thereafter.

Immediate Complications HEMORRHAGE Reported rates of hemorrhage vary widely, reflecting both diverse definitions (100 to 1000 mL blood loss)

and imprecision in estimating volumes of blood loss. Rates of hemorrhage range from 0.05 to 4.9 per 100 abortions in large case-series reports. The best index of clinically important hemorrhage is probably the rate of blood transfusion. The rate of transfusion associated with suction curettage in a large multicenter study was 0.06 per 100 abortions. For abortions performed later in pregnancy, investigators have reported rates of 0.26 for D&E, 0.32 for urea-prostaglandin, and 1.72 for saline. Vasopressin administered with paracervical anesthesia decreases blood loss with D&E abortion after 14 weeks' gestation. As little as 4 U (0.2 mL) mixed in with the anesthetic lowers the blood loss significantly; overall, vasopressin lowers four-fold the risk of a hemorrhage of 500 mL or more. When hemorrhage occurs after suction curettage, administration of uterotonic agents and manual compression usually resolve the problem. In addition to oxytocic agents, the physician can inject vasopressin in the paracervical tissue to slow bleeding. Should bleeding persist, assessment of the endometrial cavity by hysteroscopy and internal compression of the cavity by a large Foley catheter balloon or a vasopressin-soaked pack can be helpful. With the increasing cohort of women who have had Caesarean deliveries, the risk of encountering placenta accreta during abortion is increasing as well. In a series of more than 16,000 D&E abortions, the incidence of placenta accreta leading to hysterectomy was four per 10,000 cases.

CERVICAL INJURY Cervical injury encompasses a broad spectrum of trauma. The most common type is a superficial laceration caused by the tenaculum tearing off during dilation. At the other extreme are the cervicovaginal fistula and the longitudinal laceration ascending to the level of the uterine vessels. Rates of cervical injury range from 0.01 to 1.6 per 100 suction curettage abortions. In older studies, the incidence of cervical injury requiring sutures was about one per 100 suction curettage abortions. Several risk factors for cervical injury during suction curettage have emerged. Among factors within the control of the physician, use of laminaria and performance of the abortion by an attending physician (rather than a resident) lower the risk significantly, whereas use of general anesthesia raises the risk significantly. Among factors beyond the control of the physician, a history of a prior abortion lowers the risk, and age of 17 years or under increases the risk. Use of laminaria and performance of the abortion under local anesthesia by an attending physician together yield a 27-fold protective effect. Cervical preparation with misoprostol may confer similar benefits as laminaria, although more extensive experience will be needed to confirm this.

ACUTE HEMATOMETRA Also termed the postabortal syndrome or the redo syndrome, acute hematometra is an important complication of suction curettage; its cause is unknown. The incidence of this syndrome ranges from 0.1 to one per 100 suction curettage abortions, according to the available literature. Women with this condition develop severe cramping, usually within 2 hours of the abortion. Vaginal bleeding is less than expected. The woman may be weak and sweaty, and her uterus is large and markedly tender. Treatment consists of prompt repeat curettage, usually without anesthesia or dilation. Evacuation of both liquid and clotted blood leads to rapid resolution of the symptoms. The physician can aspirate the blood with a suction cannula, a Karman cannula and syringe, or even a catheter attached to wall suction. Administration of an oxytocic after the repeat evacuation is standard. Whether routine prophylactic use of an oxytocic would reduce the incidence of acute hematometra is unknown.

ANESTHESIA COMPLICATIONS Pain experienced during abortion relates not only to the choice of anesthesia but also to the characteristics of the patient. Young women (age, 13 to 17 years) and those with depression before the abortion report more pain than do other women. Local anesthesia is safer than general anesthesia for both first- and second-trimester abortions. In an Italian study, use of general anesthesia had a relative risk for all complications combined of 1.8 (95% confidence interval, 1.4 to 2.5). The largest effect occurred with hemorrhage. Similarly, use of general anesthesia for D&E abortion in the United States increases the risk of serious complications. Overall, the

attributable risk related to general anesthesia is low, and many women are willing to assume incremental risks in order to have no discomfort during the operation.

PERFORATION Perforation is a potentially serious, but infrequent complication of abortion. According to most reports, the incidence of perforation is about 0.2 per 100 suction curettage abortions. Several risk factors for perforation exist. Performance of a curettage abortion by a resident rather than by an attending physician increases the risk more than five-fold; on the other hand, cervical dilation by laminaria decreases the risk about five-fold. The risk of perforation increases significantly with advancing gestational age. Multiparous women have three times the risk of nulliparous women. The two principal dangers of perforation are hemorrhage and damage to the abdominal contents. Lateral perforations in the cervico–isthmic region are particularly hazardous because of the proximity of the uterine vessels. Perforations of the fundus are more likely to be innocuous. Indeed, most perforations are not suspected or detected. In a series of patients undergoing combined abortion and sterilization by laparoscopy, the investigators found a six-fold higher rate of uterine perforation than they had suspected clinically (20 versus three per 1,000 abortions). Not all perforations require treatment. Many suspected or documented perforations require only observation. Perforation with a dilator or sound is unlikely to damage abdominal contents. On the other hand, a suction cannula or forceps in the abdominal cavity can be devastating. If the physician suspects a perforation, the procedure should stop immediately. If unmanageable hemorrhage, expanding hematoma, or injury to abdominal contents occurs, prompt laparotomy is necessary. Laparoscopy can be useful in documenting perforation and assessing damage; if necessary, the physician can complete the abortion under laparoscopic visualization. Any woman with severe pain within hours after the abortion should be evaluated for possible perforation with bowel injury.

Delayed Complications RETAINED TISSUE Although retained tissue after abortion can pass without incident, retained tissue can lead to hemorrhage, infection, or both. This complication occurs infrequently, however. Its incidence after suction curettage abortion is less than one per 100 abortions. This complication usually manifests itself within several days of the abortion. Cramping and bleeding can be accompanied by fever. When women develop pain, bleeding, and low-grade fever after abortion, retained tissue may be present. Prompt outpatient suction curettage usually resolves the symptoms, but close follow-up is advisable.

INFECTION Postabortal infection can result from retained tissue. The likelihood of febrile morbidity after abortion depends on the method used. The incidence of fever of 38°C or higher for one or more days is usually less than one per 100 abortions by suction curettage. Corresponding figures for D&E are 1.5 per 100 abortions; for urea-prostaglandin, 6.3; and for hypertonic saline, 5.0. The organisms responsible for postabortal infection are similar to those responsible for other gynecologic infections. A number of risk factors for infection exist. Women are at increased risk if they have untreated endocervical gonorrhea or chlamydial infection. Late abortions also increase the risk. Likewise, use of labor-induction abortion instead of D&E and use of local rather than general anesthesia for suction curettage increase the risk. Administration of broad-spectrum antibiotics and, if needed, uterine curettage are the cornerstones of therapy.

Late Complications

RH SENSITIZATION Legal abortion is a potentially important cause of Rh sensitization for women at risk. The likelihood of sensitization increases with advancing gestational age (and, hence, larger volumes of fetal erythrocytes). One study has quantified the risk of Rh sensitization from first-trimester suction curettage without RhIG prophylaxis. A total of 3.1% of secundigravidas whose first pregnancy terminated by suction curettage without RhIG prophylaxis had antibodies in their second pregnancy. Subtracting 0.5% (the percentage of women estimated to have become sensitized primarily during the second pregnancy), the investigators estimated the risk of sensitization from suction curettage to be 2.6%. Thus, on a nationwide basis, the clinical impact of failure to administer RhIG to candidates after abortion may be substantial. Candidates should receive 50 µg of RhIG after abortions performed at 12 weeks' gestation or earlier or 300 µg after abortions performed later in pregnancy.

ADVERSE PREGNANCY OUTCOMES Investigators have linked induced abortion with a broad array of adverse reproductive outcomes, ranging from infertility to ectopic pregnancy. Most published reports, however, suffer from serious methodologic shortcomings that limit their usefulness. To examine the potential association between first-trimester induced abortion and subsequent reproductive performance, epidemiologists have performed an exhaustive review and analysis of the world literature. This includes more than 150 epidemiologic studies published in 11 languages. The findings of this analysis are largely reassuring. No increase in the risk of secondary infertility and ectopic pregnancy appears, even in studies with substantial power to detect differences in rates. Midtrimester spontaneous abortion is no more common among women who have had one previous abortion than among women pregnant for the first time. Similarly, the risk of premature delivery does not increase for women having undergone induced abortion. On the other hand, low birthweight is more frequent in first births after abortion by sharp curettage performed under general anesthesia compared with first-pregnancy births. This does not occur after other methods of abortion, such as suction curettage. The questions of the effect of repeat induced abortion and second-trimester abortion remain unresolved, but repeat sharp curettage may carry increased risks. First-born infants of women who had one induced abortion have risks of morbidity and mortality similar to those of other first-born children. Additional studies have corroborated the absence of adverse effects of induced abortion on subsequent reproduction. Outcomes studied included infertility, ectopic pregnancy, spontaneous abortion, and adverse obstetric outcomes. One unresolved issue is placenta previa. Sophisticated studies have found either no or a marginally significant increase in the risk (relative risk, 1.3; 95% confidence interval, 1.0 to 1.6), which was comparable to that with spontaneous abortion. Induced abortion does not threaten a woman's emotional health. In contrast, the dominant emotional reaction to induced abortion is a sense of relief. In several studies, abortion appeared to improve the emotional well-being of women by resolving an intense personal crisis. Specifically, claims of a postabortion trauma syndrome lack scientific merit. The putative association between induced abortion and breast cancer remains controversial. Although a number of case-control studies have found an association, this appears because of recall bias among controls. Women who are well (controls) are less likely to report prior induced abortions than are women with breast cancer (cases). This type of information bias has been documented in studies from Sweden. Two large cohort studies, which are less likely to be biased than are case-control studies, have shown either no effect or a protective effect of induced abortion on later breast cancer. No firm evidence links abortion to other cancers.

Mortality Since 1972, when the Centers for Disease Control and Prevention first began nationwide surveillance of abortion deaths, the safety of abortion has improved dramatically. As shown in Figure 21.16, the case-fatality rate fell

from 4.1 deaths per 100,000 abortions in 1972 to 0.4 in 1987. FIGURE 21.16. Case-fatality rates from legal abortion by year, United States, 1972 to 1987. (From: Lawson HW, Frye A, Atrash HK, et al. Abortion mortality, United States, 1972 through 1987. Am J Obstet Gynecol 1994;171:1365, with permission.)

View Figure

The causes of death from legal abortion have changed as well. From 1972 to 1977, infection and hemorrhage were the leading causes of death. Thereafter, complications of anesthesia (usually general anesthesia) became more important, emerging in 1983 as the leading cause of death from abortion. Most of these deaths were owing to hypoventilation or loss of airway resulting in hypoxia. The message from these deaths is clear: people administering general anesthesia for abortion must be skilled in airway management and observant for signs of hypoxia. The risk of death from legal abortion increases with gestational age: the earlier the abortion, the safer the abortion. As shown in Table 21.4, which includes deaths from 1972 to 1987, the death-to-case rate for abortions at 8 weeks' gestation or earlier was 0.4 per 100,000 abortions. The risk was nearly eight times higher in the 13- to 15-week interval. For abortions at 21 weeks and later, the risk of death was more than 30 times that for abortions at 8 weeks' gestation or earlier. TABLE 21.4. Legal Abortion Case-Fatality Rates and Relative Risks by Selected Categories: United States, 1972–1987

Risk Category

Rate*

Relative Risk and 95% Confidence Interval

?19

1.0

Referent

20–24

1.3

1.3 (0.9–1.8)

25–29

1.2

1.2 (0.8–1.8)

30–34

1.6

1.5 (1.0–2.4)

35–39

2.5

2.3 (1.4–3.8)

?40

3.1

3.0 (1.5–6.0)

White

0.9

Referent

Black and other

2.3

2.4 (1.9–3.2)

0

0.9

Referent

1

1.1

1.3 (0.9–1.9)

2

1.1

1.3 (0.8–1.9)

?3

2.5

2.8 (2.0–4.0)

AGE GROUP (y)

RACE

PARITY†

?3

2.5

2.8 (2.0–4.0)

?8

0.4

Referent

9–10

0.8

2.1 (1.3–3.3)

11–12

1.4

3.7 (2.3–5.8)

13–15

2.9

7.7 (5.0–11.7)

16–20

9.3

24.5 (18.6–32.3)

?20

12.0

31.5 (22.2–44.5)

Curettage

0.5

Referent

Dilation and evacuation

3.7

6.8 (4.9–9.5)

Labor induction

7.1

13.0 (9.7–17.5)

Hysterotomy or hysterectomy

51.6

95.0 (69.6–129.5)

Other

1.9

3.6 (1.5–8.3)

GESTATIONAL AGE (WK)‡

PROCEDURE§

* Number of legal abortion deaths per 100,000 abortions.

† Denominators for calculating rates by parity use previous live-birth data from abortion surveillance. Deaths with unknown parity are excluded.

‡ Deaths with unknown gestational age are excluded.

‡ Deaths with unknown gestational age are excluded.

§ Figures for 1972 and 1973 are excluded because data necessary to calculate rates by procedure were not collected. Deaths with unknown procedure are excluded.

(Lawson HW, Frye A, Atrash HK, et al. Abortion mortality, United States, 1972 through 1987. Am J Obstet Gynecol 1994;171:1365)

Curettage abortion (which includes suction curettage and D&E) is the safest method overall (Table 21.4). The risk of death for curettage abortions was 0.5 per 100,000 abortions, whereas that for D&E was nearly seven times higher. Labor-induction abortions had a rate 13 times higher than that of curettage. Hysterotomy and hysterectomy had a rate 95 times higher, although this rate reflects only 10 deaths. Women of minority race, those of advanced maternal age, and multiparas also had higher risks of death from abortion. At 13 to 15 weeks' gestation, D&E carries about half the risk of death as does labor induction (Table 21.5). At later gestational ages, the risks of death associated with both methods are similar. Because of the large number of D&E abortions performed at 13 to 15 weeks' gestation, the overall safety of D&E at 13 weeks' gestation or later is superior to that of labor induction. TABLE 21.5. Case-Fatality Ratesast; for Dilation and Evacuation and Labor Induction at 13 Weeks' Gestation or Later, by Method and Gestational Age: United States, 1972–1987

Gestational Age (wk)

Method

13–15

16–20

?21

All

Dilation and evacuation

2.0

6.5

11.9

3.7

Labor induction

3.8

7.9

10.3

7.1

* Per 100,000 abortions.

(Lawson HW, Frye A, Atrash HK, et al. Abortion mortality, United States, 1972 through 1987. Am J Obstet Gynecol 1994;171:1365)

(Lawson HW, Frye A, Atrash HK, et al. Abortion mortality, United States, 1972 through 1987. Am J Obstet Gynecol 1994;171:1365)

CONCLUSION Part of "Chapter 21 - Management of Abortion" Abortion is the most frequent outcome of human conception; thus, management of abortion and its complications is an important responsibility for physicians. Chromosomal anomalies are the single most important cause of spontaneous abortion. For women with threatened abortion, use of ultrasonography and ?-hCG monitoring can help predict the outcome of the pregnancy. Evacuation of the uterus in cases of fetal death is primarily for psychological rather than medical indications; either curettage or labor induction may be appropriate. Premature cervical dilation without labor is a poorly understood cause of spontaneous abortion. There is no uniform case definition and no valid diagnostic test. Randomized clinical trials reveal modest benefit of cervical cerclage on obstetric outcomes. Small numbers of illegal abortions continue to occur in the United States. Legally induced abortion, however, is one of the most frequently performed—and one of the safest—operations in contemporary practice. Fewer than one per 100 of those having an abortion suffer a major complication, and fewer than one per 100,000 die from causes associated with the procedure. The marketing of mifepristone in the United States for early abortion has broadened the options available to women, and use of misoprostol for cervical preparation before curettage represent important advances in gynecology.

Chapter 22

Ectopic Pregnancy Mark A. Damario John A. Rock Ectopic pregnancy was first recognized in 1693 by Busiere, when he was examining the body of a prisoner executed in Paris. Gifford of England made a more complete report in 1731 that described the condition of a fertilized ovum implanted outside the uterine cavity. Ectopic pregnancy has since become recognized as one of the more serious complications of pregnancy. One of the leading causes of maternal morbidity and mortality in the United States, it still accounted for 9% of all maternal deaths from 1990 to 1992, according to the Centers for Disease Control. Even despite significant advances in diagnosis and treatment, ectopic pregnancy remains the leading cause of maternal death in the first trimester. Today, early diagnosis of ectopic pregnancy is possible with highly sensitive and rapid ?-human chorionic gonadotropin (?-hCG) assays and the aid of advanced vaginal ultrasonographic equipment. The benefit of early diagnosis is that expectant medical therapy or conservative surgery becomes possible. Conservative management in the case of a small ectopic pregnancy that is present without rupture is usually successful when preservation of the oviduct to maintain or enhance fertility is important. Physicians should maintain a high index of suspicion for ectopic pregnancy and should be cognizant of the importance of early diagnosis and early intervention. This chapter summarizes the contemporary methods for diagnosis and treatment of ectopic pregnancy. EPIDEMIOLOGY OF ECTOPIC PREGNANCY PATHOLOGY ETIOLOGY OF ECTOPIC PREGNANCY SITES OF ECTOPIC PREGNANCY EFFECTS OF ECTOPIC PREGNANCY ON FUTURE REPRODUCTION TUBAL ECTOPIC PREGNANCY DIAGNOSIS TREATMENT FOR ECTOPIC PREGNANCY INTERSTITIAL PREGNANCY OVARIAN ECTOPIC PREGNANCY ABDOMINAL ECTOPIC PREGNANCY CERVICAL ECTOPIC PREGNANCY HETEROTOPIC PREGNANCY RH IMMUNOGLOBULIN USE AFTER ECTOPIC PREGNANCY

EPIDEMIOLOGY OF ECTOPIC PREGNANCY Part of "Chapter 22 - Ectopic Pregnancy" Although the total number of pregnancies has declined over the past three decades, the rate of ectopic pregnancy has continued to increase in most western nations. In the United States, the incidence of ectopic pregnancy has increased from 4.5 per 1,000 pregnancies in 1970 to 19.7 per 1,000 pregnancies in 1992. In Norway, an increase from 12.5 to 18.0 per 1,000 pregnancies was reported from 1976 to 1993. One contributing factor for the rising ratio of extrauterine to intrauterine pregnancies is felt to be the rising incidence of sexually transmitted diseases as well as the efficacy of modern antibiotic treatments for pelvic inflammatory disease (PID). A second factor may be the increased ability to detect the disease. Although the risk of death from ectopic pregnancy continues to decline among all races and ages in the United States, women of black and other minority races remain at significantly increased risk of death from ectopic pregnancy compared with white women. Although the overall incidence of ectopic pregnancy in the United States during 1970 to 1989 increased approximately five-fold, the risk of death from ectopic pregnancy declined by 90%. This decline in mortality from ectopic pregnancies may be related both to the increased awareness of the condition as well as improved diagnostic and therapeutic methods.

PATHOLOGY Part of "Chapter 22 - Ectopic Pregnancy" A tubal gestation traditionally has been defined as one that implants and grows within the tubal lumen. Budowick and associates have suggested that tubal implantation actually occurs in the lumen but is soon followed by penetration into the lamina propria and muscularis to become extraluminal. Pauerstein, Hodgson and Kramen demonstrated that trophoblastic infiltration can be predominantly intraluminal or predominantly extraluminal, or, occasionally, mixed. It is impossible to ascertain in the operating room the predominant pattern of growth of a given tubal pregnancy. In any event, fimbrial expression usually is an unacceptable method for removal of ectopic pregnancy. Not only is the method traumatic, but it frequently does not remove all of the trophoblastic tissue. The resultant persistent ectopic pregnancy may therefore require additional therapy.

ETIOLOGY OF ECTOPIC PREGNANCY Part of "Chapter 22 - Ectopic Pregnancy"

Tubal Damage Secondary to Inflammation Both the increased incidence of sexually transmitted disease resulting in salpingitis and the efficacy of antibiotic therapy in preventing total tubal occlusion after an episode of salpingitis are related to the increasing incidence of ectopic pregnancy. Levin and associates have demonstrated that the risk of ectopic pregnancy is increased in women with a primary history of PID. Westrom compared women with PID confirmed by laparoscopy with healthy women, matched by age and parity, and found a six-fold greater incidence of ectopic pregnancy in women with PID, an alarming rate of 1 ectopic pregnancy out of every 24 gestations. Similar statistics have been reported by other authors. Many of the patients in these studies had received antibiotic treatment for salpingitis. Before antibiotics became available for the treatment of PID, salpingitis was usually so acute that the inflamed tube became totally occluded, and permanent sterility was the result. Women who attempted to conceive after a pelvic infection were successful less than 40% of the time. Today, the rate of pregnancy exceeds 60% for patients adequately treated with antibiotics. After initial appropriate treatment of an infection with antibiotics, agglutination of the cilia can still occur and synechial bands can form within the tubal lumen to cause partial tubal obstruction. Westrom has demonstrated by laparoscopy that bilateral tubal occlusion occurs in approximately 12.8% of patients after treatment for the first tubal infection, in 35% after two infections, and in 75% after three or more infections. In addition, Westrom found that approximately 4% of all pregnancies subsequent to salpingitis were ectopic. Fallopian tubes containing a gestation are frequently normal on macroscopic visualization and gross histologic examination. Vasquez, Winston and Brosens using scanning electron microscopy and light microscopy studies of tubal biopsies from five groups of women, discovered marked differences in their ciliated surfaces. The proportion of ciliated cells was significantly lower in biopsy specimens taken from 25 women with tubal pregnancies as compared to biopsy specimens from seven women with intrauterine pregnancies at the same stage of gestation. Marked deciliation was likewise seen in eight women who had undergone biopsies during tubal reconstructive surgery. In another study, Gerard and colleagues found that seven of ten fallopian tube samples from patients with ectopic pregnancy were PCR-positive for C. trachomatis DNA. Therefore, the increased occurrence of sexually transmitted diseases contributing to subclinical tubal epithelial damage may be an important contributor to ectopic pregnancy. Comprehensive programs to prevent sexually transmitted diseases undertaken in Sweden and Wisconsin have been found to not only decrease the incidence of C. trachomatis infections and other sexually transmitted diseases but also the rate of ectopic pregnancies.

Contraceptive Devices The use of IUDs has been associated with an increased incidence of ectopic pregnancy. In a summary of published reports on ectopic pregnancy, Tatum and Schmidt observed that 4% of the pregnancies that occurred with an IUD in place were ectopic. In a recent meta-analysis, Mol and associates reported a range of odds ratios from 4.2 to 45.0 from heterogenous studies of IUD use and ectopic pregnancy. Subtle tubal epithelial damage or actual PID episodes are likely responsible for the observed association between IUDs and ectopic pregnancy.

Oral Contraceptives The overall risk of an ectopic pregnancy is lowered in women using oral contraceptives. When oral contraceptives fail, however, the risk of an ectopic pregnancy is slightly increased. This increase is presumed secondary to the inhibitory progestin effect on tubal motility. This hypothesis is supported by several studies implicating progestin-only oral contraceptives in the etiology of ectopic pregnancies.

Prior Tubal Surgery An operative procedure on the oviduct, whether a sterilization procedure or tubal reconstructive surgery, can cause an ectopic pregnancy. The incidence of ectopic pregnancies occurring after neosalpingostomy for distal tubal obstruction ranges from 2% to 18% (Table 22.1). The rate of ectopic pregnancy after a microsurgical reversal of a sterilization procedure is only about 4%, presumably because the tubes have not been damaged by prior infection. TABLE 22.1. Summary: Ectopic Pregnancy After Tubal Surgery

Procedure

Technique

Total Pregnancy(%)

Pregnancy Range(%)

Ectopic(%)

Ectopic Range(%)

Salpingoscopy

Macrosurgery

42

35–65

3.4

1–20

Microsurgery

52

31–69

1.8

0–16

Macrosurgery

42

36–50

14

10–18

Microsurgery

59

26–68

6

4–11

Macrosurgery

27

20–38

4.2

2–10

Microsurgery

26

17–44

7.7

0–18

Macrosurgery

44

25–83

9.2

0–15

Microsurgery

62

35–78

2.3

1–6.2

Salpingectomy

42

38–49

12

8–17

Salpingostomy

57

39–73

11

0–20

Fimbrioplasty

Neosalpingostomy

Tubal anastomosis

Removal of ectopic pregnancy

Salpingostomy

57

39–73

11

0–20

(Lavy G, Diamond MP, DeCherney AH. Ectopic pregnancy: relationship to tubal reconstructive surgery. Fertil Steril 1987;47:543–556.)

The United States Collaborative Review of Sterilization Working Group followed a total of 10,685 women undergoing tubal sterilization in a multicenter, prospective cohort study. The overall cumulative probability of pregnancy in the study cohort 10 years after sterilization was 18.5 per 1000 procedures (failure rate of 1.85%). The 10-year cumulative probability of ectopic pregnancy for all methods of tubal sterilization was 7.3 per 1000 procedures. From these data, one can therefore estimate that in the setting of a positive pregnancy following tubal sterilization, there is an approximately 40% risk that the pregnancy will be ectopic. The type of sterilization procedure and age of the patient at the time of sterilization appear to be relevant factors. Women sterilized by bipolar tubal coagulation before the age of 30 years had a probability of ectopic pregnancy that was 27 times as high as that of women of similar age who underwent postpartum partial salpingectomy (31.9 versus 1.2 ectopic pregnancies per 1000 procedures). In addition, ectopic pregnancy was often seen many years after the sterilization procedure. The annual rates of ectopic pregnancy in the fourth through tenth years after sterilization were no lower than that seen in the first three years. The pathophysiology of ectopic pregnancy after elective tubal sterilization is not clear. It is possible that a tuboperitoneal fistula in a previously coagulated segment of fallopian tube may allow spermatozoa to escape and reach the oocyte. Such fistulas have been demonstrated radiographically by Shah and colleagues in 11% of 150 women after laparoscopic electrocoagulation. Improper surgical technique (such as incomplete coagulation or misplacement of a mechanical device) may also influence the sterilization failure rate and incidence of ectopic pregnancy, although their likelihood is presumably low.

Assisted Reproductive Technologies Ectopic pregnancies are known to occur with increased frequency after in vitro fertilization (IVF) and related techniques. The Society for Assisted Reproductive Technology (SART) reported that 2.1% of pregnancies established after IVF in the United States during 1997 were ectopic. Several theories have been proposed regarding the occurrence of ectopic implantation after transcervical intrauterine embryo transfer. Potential factors include the possibility of direct injection of embryos into the fallopian tube, uterine contractions provoked by the transfer catheter that propel the embryos retrograde, position or depth of the transfer catheter in the uterine cavity, and the volume of transfer medium. Verhulst and colleagues reported that tubal damage was a major risk factor. These researchers found that the ectopic pregnancy rate after IVF was significantly greater in patients with tubal disease (3.65% of pregnancies) than in those without tubal disease (1.19% of pregnancies). Strandell, Thorburn and Hamberger found that a history of a previous ectopic pregnancy and a history of a previous myomectomy also appear to be risk factors for ectopic pregnancies following IVF. Tummon and coworkers reported a 2% risk of heterotopic pregnancy in women undergoing IVF who had distorted tubal anatomy. This is about 100 to 200 times the reported incidence of combined intrauterine and extrauterine pregnancies occurring spontaneously. These authors also found that the risk of heterotopic pregnancy appeared to increase proportionately with the number of embryos transferred.

Developmental Anomalies Intramural polyps and tubal diverticula can block or alter tubal transport of fertilized ova. Congenital absence of segments of the fallopian tube with peritoneal fistulas can also predispose to tubal pregnancy. Women exposed to diethylstilbestrol (DES) in utero are at higher risk of ectopic pregnancy. These women may have absent or minimal fimbriae and fallopian tubes that are shorter and thinner than normal.

Other Causal Factors Several studies have demonstrated that cigarette smoking seems to be an independent, dose-related risk factor for ectopic pregnancy. Other lifestyle factors such as multiple sexual partners and early age at first intercourse are associated with an increased risk. Vaginal douching has also been associated with a slightly increased risk of ectopic pregnancy, probably by increasing the overall risk of pelvic infections and resultant tubal damage. A summary of risk factors related to ectopic pregnancy is summarized in Table 22.2. TABLE 22.2. Risk Factors for Ectopic Pregnancy

Chronic pelvic inflammatory disease

Prior tubal surgery

Surgical sterilization

Use of an intrauterine device

Previous ectopic pregnancy

DES exposure

Progestin-only contraceptives

Assisted reproductive technologies

Infertility

Developmental tubal anomalies

Multiple sexual partners

Early age at first intercourse

Cigarette smoking

Vaginal douching

SITES OF ECTOPIC PREGNANCY Part of "Chapter 22 - Ectopic Pregnancy" About 95% of extrauterine implantations occur in the oviduct. About 55% of these tubal implantations occur in the ampulla, the most common site: implantation in the isthmic portion accounts for 20% to 25%, implantation in the infundibulum and fimbria account for 17%, and implantation in the interstitial segment (cornua) accounts for 2% to 4%. Ectopic implantations occur less often in the ovary, the cervix, and the peritoneal cavity (Fig. 22.1). FIGURE 22.1. Sites and incidence of ectopic pregnancy.

View Figure

Walters, Eddy, and Pauerstein reported that 16% of tubal pregnancies result from a contralateral ovulation. Transmigration of the ovum in the peritoneal cavity can occur because the oviducts and ovaries may be situated close together in the cul-de-sac. Alternatively, this phenomenon could also result from transmigration of the embryo through the endometrial cavity into the opposite oviduct.

EFFECTS OF ECTOPIC PREGNANCY ON FUTURE REPRODUCTION Part of "Chapter 22 - Ectopic Pregnancy" Tubal pregnancy is associated with a poor prognosis for subsequent reproduction. In most cases, an extrauterine pregnancy represents an impairment of the fertilized ovum's ability to migrate through the deep rugae of the oviduct as a result of altered tubal function. The morphologic abnormality is usually bilateral, irreversible and can produce repeated ectopic pregnancies or permanent sterility. In a 1975 study, Shoen and Nowak concluded that about 70% of patients who have an ectopic first pregnancy are unable to produce a living child. As many as 30% of the patients who have an ectopic first pregnancy will have a repeat ectopic pregnancy, which compares with the total repeat ectopic rate of 10% to 15% for the overall population of reproductive-age women. More than half of the subsequent extrauterine pregnancies will occur within a 2-year period, and 80% will occur within 4 years of the initial ectopic pregnancy. In reviewing the experience of the Kaiser Foundation hospitals, Hallatt reported a 9.2% overall incidence of repeat ectopic pregnancies among 1330 women who had extrauterine pregnancies. The potential reproductive capacity for a patient who has had an ectopic pregnancy therefore depends on her reproductive history. If an ectopic pregnancy was the result of her first reproductive effort, then the prognosis for future pregnancies is much worse than if the complication occurred after one or more successful pregnancies. Mueller and associates have estimated that 92% of infertility in women who have had a tubal pregnancy results from tubal damage due to the tubal pregnancy itself or other factors that had predisposed to its occurrence. A history of infertility itself is a risk factor for ectopic pregnancy. A two-fold increase in the risk of tubal pregnancy exists among infertile women with no evident abnormality during infertility evaluation.

TUBAL ECTOPIC PREGNANCY Part of "Chapter 22 - Ectopic Pregnancy" The morbidity and mortality associated with extrauterine pregnancy are directly related to the length of time required for diagnosis. In a Centers for Disease Control survey, two thirds of all patients who were later proven to have an ectopic pregnancy were previously seen by a physician, and either the diagnosis was deferred or the condition was incorrectly assessed. The mortality rate from an ectopic pregnancy is higher in rural areas, where patients are less likely to receive early medical care. For a successful outcome, an ectopic pregnancy must be diagnosed early. In some clinics where the condition is treated frequently, more than 50% of cases are diagnosed and treated before tubal rupture occurs. In some cases, however, the symptoms that bring a patient to seek medical care are caused by an already leaking or ruptured ectopic pregnancy. As many as 15% of all tubal pregnancies rupture before the first missed menstrual period, particularly if a patient's usual menstrual pattern is very irregular. Diagnostic accuracy is often improved in repeat ectopic pregnancies. The vast majority of patients with repeat ectopic pregnancies will be diagnosed and treated before tubal rupture. A difference with a repeat ectopic pregnancy is that the patient herself often raises the question of an extrauterine pregnancy. Being suspicious, the patient may seek medical care earlier and provides a more specific medical history than does a patient experiencing her first ectopic pregnancy. The result is often an earlier diagnosis and an improved chance for a successful outcome. Some form of vaginal bleeding occurs around the expected time of menses in more than 50% of women with an ectopic pregnancy, so that many patients and their physicians are unaware that a pregnancy has occurred. The vaginal bleeding may be followed by a period of amenorrhea. Clinical symptoms of an ectopic pregnancy usually appear 6 to 10 weeks after the last normal menstrual period.

DIAGNOSIS Part of "Chapter 22 - Ectopic Pregnancy"

Classic Symptoms: Pain, Bleeding, and Adnexal Mass The classic presentation of pain and uterine bleeding with the finding of an adnexal mass has been the clinical hallmark of an extrauterine pregnancy, but even classic presentations can be misleading. Schwartz and DiPietro observed that of the patients who presented with the classic signs and symptoms, only 14% had an ectopic pregnancy. The severity of the symptoms and signs depends on the stage of the condition, but in the early stages of an ectopic gestation, symptoms are less predictive than in the more advanced stages of the disease. A discrete, unilateral mass separate from the adjacent ovary has been detected in less than one third of all proven ectopic pregnancies. Locating a mass depends on many factors, including the diagnostic skill of the examiner, the degree of pelvic peritonitis present, the presence or absence of tubal rupture, and the degree of stoicism and cooperation of the patient. Even when all factors are optimal, an adnexal mass can be felt in only half of the cases.

Diagnostic Studies Three major advances have made early diagnosis of extrauterine pregnancy possible: (1) the development of highly sensitive and rapid ?-hCG assays; (2) the ability to use ultrasound to evaluate the uterus and the adnexa (vaginal sonography further increases the accuracy of diagnosis); and (3) the application of laparoscopy as a diagnostic tool. Culdocentesis or suction curettage, or both, can be useful under certain circumstances (e.g., to help establish the presence of a hemoperitoneum or a nonviable intrauterine pregnancy, respectively). Other newer diagnostic methods, such as serum progesterone assays or color Doppler flow analyses, can also provide useful information.

?-hCG Assays The principal endocrine marker of pregnancy is human chorionic gonadotropin (hCG), which is synthesized by the trophoblast. Human chorionic gonadotropin is a glycoprotein consisting of two subunits: ? and ?. The ? subunit has significant homology with other glycoprotein hormones, such as follicle-stimulating hormone, luteinizing hormone and thyroid-stimulating hormone. The ?-subunit, on the other hand, is specific to hCG, and antibodies against the ?-subunit form the basis for current radioimmunoassay and monoclonal antibody laboratory assays. A radioimmunoassay for ?-hCG can detect levels of hCG as low as 5 to 10 IU/L of serum with less than a 0.5% incidence of false-negative results. ?-hCG can be detected in maternal serum as early as 7 to 8 days after ovulation, or approximately the day after blastocyst implantation. One-step qualitative urinary pregnancy tests, which positively identify a threshold level of hCG of 20 to 50 IU/L, have also become available and have been successfully used in an emergency department setting as the initial step in triaging women of reproductive age who present with abdominal pain or abnormal vaginal bleeding. The quantification of serum ?-hCG levels is useful in determining the viability of pregnancy. To optimally use ?-hCG data in treating a patient with a problematic pregnancy, one should first have a thorough understanding of the particular assay used. The World Health Organization has established reference standards for ?-hCG assays. The Third International Standard (Third IS) is the most commonly used reference standard used by the available commercial kits of today. This standard is roughly equivalent to the First International Reference Preparation (First IRP), but is quite a bit different from the Second International Standard (Second IS). The Second IS contains about 20% intact hCG and was initially developed for use in hCG bioassays. One international unit of ?-hCG based on the First IRP is equal to approximately 0.58 IU of ?-hCG using the Second IS. Fortunately, the Second IS has been exhausted and is no longer used but can still be found in some studies and publications. The First IRP and Third IS are highly purified preparations that were developed to overcome the deficiencies seen in the use of a heterogeneous standard. Serum hCG concentrations increase in an exponential fashion in early pregnancy. During the period of

gestation in which the hCG concentration is less than 10,000 IU/L (First IRP), or about 25 to 30 days postovulation, the time required for doubling of hCG levels remains constant, with a mean of 1.9 days. Kadar, Caldwell and Romero reported that 87% of women with ectopic pregnancies and 15% of women with normal intrauterine pregnancies could expect to have hCG doubling times of more than 2.7 days when the hCG concentration measured less than 6000 IU/L. The lower limits of the increase in serum hCG for viable intrauterine pregnancies have also been established in the authors' laboratory by use of the First IRP. Serum quantitative ?-hCG levels appear to increase between 30% and 50% (mean, 39.1%) for an interval of 24 hours in a normal intrauterine pregnancy. A more meaningful interval determination is at 48 hours with a minimum 66% (range, 65% to 100%) increase over preceding hCG values expected. Interval ?-hCG determinations interpreted within the context of several values can therefore be of prognostic significance in the differentiation between normal intrauterine versus extrauterine pregnancies. A normal rise in hCG production, however, does not always differentiate an ectopic from a viable intrauterine pregnancy. Shepherd and associates reported that, in their experience, a normal rise in hCG production did not reliably differentiate an ectopic from a viable intrauterine pregnancy in the symptomatic patient. Early ectopic pregnancies can initially secrete appropriate amounts of hCG because of a well-vascularized placental bed.

Serum Progesterone Assay Serum progesterone levels reflect the production of progesterone by the corpus luteum in early pregnancy. During the first 8 to 10 weeks of gestation, serum progesterone concentrations change little; as pregnancy fails, the levels decrease. Matthews, Coulson and Weld reported progesterone levels in 29 patients with ectopic pregnancy using a direct radioimmunoassay that offers results within 4 hours. Patients with normal intrauterine pregnancies had serum progesterone levels greater than 20 ng/mL, and all patients with ectopic pregnancies had progesterone levels less than 15 ng/mL. Yeko and associates proposed that all ectopic pregnancies could be potentially diagnosed at the first emergency visit with a single serum progesterone determination using a discriminatory value of 15 ng/mL. Other authors, however, have demonstrated some overlap in the serum progesterone concentrations in ectopic and normal intrauterine pregnancies. One large study by Gelder, Boots, and Younger reported that 98% of patients with a normal intrauterine pregnancy had progesterone levels greater than 10 ng/mL, and that 98% of patients with ectopic pregnancies not associated with ovulation induction had progesterone levels less than 20 ng/mL. Unfortunately, 31% of viable intrauterine pregnancies, 23% of abnormal intrauterine pregnancies, and 51% of ectopic pregnancies in this series had progesterone levels that fell between 10 and 20 ng/mL, which greatly limited the clinical usefulness of the test. Hahlin, Sjoblom and Lindblom reported that a serum progesterone value of less than 9.4 ng/mL combined with an abnormal hCG increase had a positive predictive value of 1.0 for pathologic pregnancy. In 1992, Stovall and colleagues reported that in a group of more than 1,000 first-trimester pregnant patients, the lowest serum progesterone level associated with a viable pregnancy was 5.1 ng/mL. Therefore, these investigators established the lower cutoff limit of serum progesterone levels of 5 ng/mL; patients below this threshold had a nonviable pregnancy with 100% certainty and therefore underwent curettage. Patients with serum progesterone levels greater than 25 ng/mL had a 97% likelihood of having a viable intrauterine pregnancy in this study.

Transvaginal Ultrasonography Pelvic ultrasound has revolutionized the diagnostic process of ectopic pregnancy. Transvaginal ultrasonography, in particular, may identify masses in the adnexa as small as 10 mm in diameter and can provide more detail about the character of the mass than clinical exam (Fig. 22.2). At the same time, transvaginal ultrasonography can evaluate the contents of the endometrial cavity and can document the presence of a viable intrauterine pregnancy with great accuracy. In addition, transvaginal ultrasonography allows for the simultaneous assessment for the presence of free peritoneal fluid.

FIGURE 22.2. Tubal ectopic pregnancy documented by endovaginal sonography.

View Figure

Transvaginal ultrasonography is usually considered superior to transabdominal ultrasonography in the diagnosis of ectopic pregnancy. Although the latter provides a broader perspective of the abdominal cavity and pelvis, transvaginal ultrasonography generally provides better resolution of the internal female genitalia. A 5-MHz transvaginal transducer allows for a deeper penetration of the pelvis than transducers of higher frequency, whereas a 7.5 MHz transvaginal transducer provides for better near-resolution at the cost of shallower penetration. On rare occasions, an ectopic pregnancy may be located beyond the reach of the transvaginal transducer's scanning field. On these particular occasions, incorporation of transabdominal ultrasonography may be an important adjunctive step. Jain, Hamper and Sanders compared endovaginal and transabdominal ultrasound results in 90 patients with a positive serum pregnancy test (Table 22.3). The specific diagnosis of ectopic pregnancy was impossible utilizing only transabdominal ultrasound before 7 gestational weeks. Normal intrauterine pregnancies could be detected earlier with endovaginal ultrasound because the yolk sac, fetal pole, and fetal heart motion could be seen sooner. Fetal heart motion was detected as early as 34 days after the last menstrual period in patients with identifiable fetal poles at the time the crown to rump length was 0.3 cm. TABLE 22.3. Pregnancy Earliest Seen with Ultrasonography

Early Intrauterine Pregnancy

Endovaginal

Transabdominal

Gestational sac size

0.5 cm

0.5 cm

Gestational sac age

4.3 wk

4.3 wk

Gestational sac size

0.6–0.7 cm

1.0 cm

Gestational sac age

4.4 wk

5.0 wk

Gestational sac size

0.7 cm

1.0 cm

Gestational sac age

4.6 wk (34 d)

5.0 wk (35 d)

Gestational sac size

0.7 cm

1.7 cm

Gestational sac age

4.6 wk

6.0 wk

0.3 cm

0.6 cm

GESTATIONAL SAC SEEN

DOUBLE DECIDUAL OUTLINE

YOLK SAC SEEN

FETAL POLE SEEN

FETAL HEART MOTION SEEN

Crown-rump length

Gestational sac age

4.6 wk (34 d)

6.5 wk (47 d)

(Jain K, Hamper VM, Sanders RC. Comparison of transvaginal and transabdominal sonography in the detection of early pregnancy and its complications. Am J Radiol 1988;151:1139–1143.)

Although diagnosis by transvaginal ultrasound can be quite useful, it may at times be confusing. One problem is that a “pseudo” gestational sac due to a decidual cast can be mistaken for an amniotic sac. A useful differentiating feature is the “double-line” image, caused by the faint hypoechoic decidual lining of the uterus and the hyperechogenic rim of the trophoblast surrounding the gestational sac. The “double-line” image can be seen as early as 5 weeks after the last menstrual period. Even in the presence of the “double-line” image, however, it is important to further follow the course of pregnancy and subsequently confirm a viable intrauterine pregnancy with the ascertainment of ultrasonographically-imaged intrauterine cardiac motion. Although not always seen, Frates and Laing reported that the presence of a non-cystic extraovarian adnexal mass, extrauterine cardiac motion, or a “tubal ring” by transvaginal ultrasonography is highly specific for ectopic pregnancy (98.9%), with a high positive predictive value (96.3%). These authors described that the direct imaging of the ectopic pregnancy utilizing any of these differentiating features is possible in 84% of cases. Many authors have reported on correlations between threshold levels of hCG above which an intrauterine gestational sac is expected by ultrasonography in a normal pregnancy (discriminatory zone). Early on, Kadar, DeVore and Romero described a threshold level of hCG above which an intrauterine gestational sac was expected by abdominal sonography in a normal pregnancy (discriminatory zone). This threshold hCG level was initially characterized as a titer of 6500 IU/L or higher using the First IRP. Presently, transvaginal ultrasonography reliably detects intrauterine gestations as early as 1 week after missed menses (?-hCG ? 1500 IU/L; 5-6 weeks gestation). Barnhart and associates reported that with a ?-hCG concentration of 1500 IU/L or higher, an empty uterus on transvaginal ultrasonography indentified an ectopic pregnancy with 100% accuracy. Even using a discriminatory serum ?-hCG concentration of 1000 IU/L, Cacciatore, Stenman and Ylostalo indentified an intrauterine gestation in all intrauterine pregnancies and in none of the ectopic pregnancies. Furthermore, these investigators reported that the detection of an adnexal mass in combination with an empty uterus had a sensitivity of 97%, specificity of 99%, positive predictive value of 98% and negative predictive value of 98% provided serum ?-hCG concentrations exceeded 1000 IU/L. The coupling of hCG titers with transvaginal ultrasonographic findings has therefore greatly facilitated the early diagnosis of ectopic gestation. It must be stressed, however, that considering the variations in ?-hCG assays, ultrasound equipment and sonographer experience, each institution must determine their own discriminatory thresholds for the sonographic detection of an intrauterine pregnancy. The advent of color-flow Doppler technology may even further improve the accuracy of noninvasive diagnostic methods. Kurjak, Zalud and Schulman reported that ectopic pregnancies are characterized by the identification of peritrophoblastic flow associated with an adnexal mass by color Doppler techniques. Kirchler and coworkers showed that color Doppler qualitative blood flow analyses of the tubal arteries can help localize the side of a tubal ectopic pregnancy. These investigators reported a between-side difference in tubal blood flow of 20% to 45%, with increased blood flow seen on the side of the ectopic pregnancy. Emerson and colleagues demonstrated that color flow Doppler capability can help differentiate between viable intrauterine pregnancy, completed abortion, incomplete abortion, and ectopic pregnancy by analyzing uterine color Doppler appearance, intrauterine venous flow, the presence or absence of intrauterine peritrophoblastic flow, corpus luteal flow, and the presence or absence of peritrophoblastic flow in the adnexa. Pellerito and associates reported that color flow imaging increased the sensitivity of detecting an ectopic pregnancy. Of 65 patients with surgically confirmed ectopic pregnancies, 36 (sensitivity, 54%) cases were detected by endovaginal sonography alone, whereas 62 (sensitivity, 95%) cases were detected by a combination of endovaginal sonography and color flow imaging.

Dilation and Curettage At one time, the histologic changes in the endometrium that accompany an ectopic gestation were routinely confirmed by dilatation and curettage (D&C). Today, more accurate diagnostic methods, such as the radioimmunoassay for ?-hCG, transvaginal ultrasonography and laparoscopy exist. In this setting, a D&C may therefore not always be necessary. If, however, the plateau level of ?-hCG is low, a D&C can be helpful to establish the presence of degenerating villi, and if a patient is bleeding excessively, a D&C may also be required. In either case, assessment of the removed material and findings of decidua without chorionic villi suggests the diagnosis of ectopic pregnancy. Such findings do not provide absolute proof, however, because they also occur with spontaneous abortion. A frozen section may be obtained immediately after curettage, providing an opportunity to confirm the diagnosis within minutes while the patient is still in the operating room under anesthesia. If no chorionic villi is present, further assessment and treatment by laparoscopy may be undertaken. In a recent report of 87 consecutive frozen section samples taken from uterine curettings, Spandorfer and colleagues found that 93.1% of these specimens were identified correctly after further analyses of the tissue by permanent section. The atypical epithelial changes of the gestational endometrium in a case of tubal pregnancy were first described by Polak and Wolfe in 1924 and were further expanded upon by Arias-Stella in 1954 (Fig. 22.3). These comprise a highly controversial set of histologic criteria which depends for its accuracy on the precise definition of the particular cell type involved in the morphologic change, together with ill-defined physiologic events that reportedly produce the changes. Arias-Stella and others were convinced that these histologic changes are a progressive phenomenon resulting from the exaggerated proliferative and secretory endometrial responses to the elevated hormonal levels of pregnancy. Fienberg and Lloyd disagreed, maintaining that these endometrial changes are regressive, involutional and are the result of declining hormonal levels. Whichever hypothesis is ultimately proven, similar endometrial changes may be seen with a normal pregnancy, spontaneous abortion or ectopic pregnancy. Histologic endometrial criteria, therefore, seems to have limited value in the specific diagnosis of extrauterine pregnancies. FIGURE 22.3. Arias-Stella reaction in endometrial cells associated with ectopic pregnancy, showing nuclear enlargement, irregularity, and hyperchromasia with cytoplasmic vacuolation.

View Figure

Culdocentesis Culdocentesis is a diagnostic tool for identifying the presence of intraperitoneal bleeding. This simple procedure of inserting an 18-gauge spinal needle attached to a 50 mL aspirating syringe into the

cul-de-sac between the uterosacral ligaments (Fig. 22.4) provides immediate clinical information when unclotted blood is aspirated from the cul-de-sac. The procedure cannot be used for a definitive diagnosis, of course, because a tubal pregnancy may not have ruptured or leaked into the peritoneal cavity. In addition, a culdocentesis does not provide information concerning whether the blood is from an ectopic pregnancy or from some other cause of intraabdominal bleeding. The rupture of a corpus luteal hemorrhaghic cyst, for instance, may cause a similar bleeding pattern. FIGURE 22.4. Culdocentesis. An 18-gauge spinal needle is inserted through the posterior fornix and enters the cul-de-sac between the uterosacral ligaments.

View Figure

The availability of sensitive transvaginal ultrasonographic technology presently limits the specific usefulness of the culdocentesis procedure. Free intraperitoneal blood has a characteristic ultrasonographic appearance and can be seen in nearly all cases in which a significant intraperitoneal hemorrhage has occurred. In the absence of the immediate availability of transvaginal ultrasonography or in an emergency setting, however, culdocentesis can still be of significant value.

Laparoscopy Laparoscopy remains the “gold standard” in the detection of ectopic pregnancy, although non-invasive diagnostic methods continue to improve. In addition to permitting the diagnosis of an ectopic pregnancy, it enables surgical treatment. Laparoscopy also provides an opportunity to visualize the entire pelvis and other peritoneal organs. In particular, the condition of the unaffected fallopian tube can be assessed, as well as the presence of pelvic adhesions and endometriosis. This information may be particularly valuable for those patients interested in future fertility. The disadvantage of laparoscopy is that it is an invasive procedure which carries some risk of complications. Utilizing standard methods, it requires general anesthesia and an operating room setting, thereby contributing to increased medical costs. Recent investigators, however, have been exploring the potential of “microlaparoscopy,” in which improved optics and smaller diameter laparoscopes and trocars allow for a definitive diagnosis and possible treatment in the non-operating room setting. Several authors have reported the encouraging use of “microlaparoscopy” in the office setting, utilizing primarily local rather than general anesthesia. The specific utility of “microlaparoscopy”, however, for the primary evaluation and treatment of ectopic pregnancy remains to be established. Laparoscopy may be useful when an ectopic pregnancy is suspected but no signs of an ultrasonographically visualized extrauterine gestational sac is evident. This includes situations in which there is an inability to visualize an intrauterine gestational sac and serial ?-hCG determinations are rising inappropriately. This also includes situations in which a D & C fails to indentify products of conception. One must be careful, however, in settings in which the ?-hCG determinations are very low or the gestational age is limited. In these settings, the ectopically implanted gestational mass may be so small

that it is not able be seen at laparoscopy. The clinician and patient might therefore be falsely reassured by negative laparoscopic findings. All patients without a definitive diagnosis established at laparoscopy should be followed closely.

Other Potential Diagnostic Aids Gleicher, Parilli and Pratt described the use of hysterosalpingography and selective salpingography in differentiating early (biochemical) intrauterine from failing intratubal gestations. A characteristic tubal opacification pattern was seen in the cases of early tubal pregnancy. Confino and coworkers reported that selective salpingography was useful in diagnosing early tubal pregnancies in some patients with equivocal clinical, laboratory, and sonographic findings. In addition, these investigators injected a single dose of methotrexate through the selective salpingography catheter after cannulation of the tubal ostia and identification of a characteristic ampullary radiolucency in seven patients. Each had subsequent complete resolution of the pregnancy without complication. Risquez and colleagues reported the successful visualization of two ectopic pregnancies by transcervical tubal cannulation and falloposcopy. The falloposcope is a microendoscopic instrument 0.5 mm in external diameter that is introduced by a 1 mm coaxial catheter. Although limited by the presence of blood in the tubal lumen, direct visualization of the ectopic pregnancies was accomplished in both cases and confirmed by concurrent laparoscopy. Other investigative teams have explored the potential of other imaging methods, such as magnetic resonance imaging (MRI), in the diagnostic work-up for ectopic pregnancy. MRI might be useful if the sonographic image is inconclusive, although it is likely to be rarely needed, particularly if laparoscopy is generally considered in uncertain cases.

Summary of Diagnostic Methods for Detecting Tubal Ectopic Pregnancy When a patient is seen with a clinical history suggestive of ectopic pregnancy, a careful examination is performed (Fig. 22.5). Quantitative serum ?-hCG and rapid serum progesterone levels (if available) are obtained. If the ?-hCG titre is positive, a transvaginal ultrasound is performed. If an intrauterine sac is visualized with fetal heart activity, then the diagnosis of intrauterine pregnancy is established. If, however, there are no intrauterine sacs or there is a questionable intrauterine sac without fetal heart activity, then the asymptomatic patient may be expectantly treated awaiting further testing. If the serum progesterone level is, with certainty, below the threshold level for viability, then a uterine curettage can be performed. The subsequent failure to find chorionic villi on curettage is very suggestive, but not diagnostic, of an ectopic pregnancy. If the serum progesterone level is more than 25 ng/mL in the absence of ovulation induction, then there is a strong likelihood that a viable pregnancy is present. If the quantitative serum ?-hCG level is above the discriminatory zone for a particular institution and no intrauterine gestational sacs are apparent using transvaginal ultrasonography, then an ectopic pregnancy is likely. The level of hCG at which a intrauterine gestational sac should be visible varies, however, depending on the ?-hCG assay and the method of pelvic ultrasonography. Most contemporary investigators report that with transvaginal ultrasonography, an intrauterine gestational sac should be identified at a ?-hCG level of 1500 IU/L (Third IS) with high sensitivity and specificity.

FIGURE 22.5. Evaluation of the stable patient with suspected ectopic pregnancy. Hormonal parameters can vary depending on the assay technique and reference standard used. The discriminatory threshold for sonographic detection of an intrauterine gestational sac must be established by each institution.

View Figure

Commonly, results from initial testing are equivocal. The ectopic pregnancy can produce a low level of ?-hCG from the aborting or degenerating trophoblast. The differential diagnosis should include spontaneous abortion or blighted ovum. When the diagnosis is uncertain and the patient is in an unstable condition or significant intraperitoneal fluid is seen, then evaluation should proceed immediately by laparoscopy and, if necessary, by laparotomy. If the patient is in stable condition, tests for serial ?-hCG levels should be taken at 48-hour intervals and the assays should be correlated with the patient's previous values. If the hCG level increases more than 66% within a 48-hour period, then the patient probably has a normal intrauterine pregnancy, and nonsurgical care is indicated. If the increase in the serial hCG level is less than 66% of the original value, then an ectopic pregnancy should be suspected. Ultrasound can often then corroborate the diagnosis, with the demonstration of a gestational sac in the adnexa or fluid in the cul-de-sac. Initially, a normal increase in the ?-hCG level may be observed. Over time, however, the level may slowly fall, never reaching the discriminatory zone. If a possible intrauterine pregnancy is thought to be nonviable, then a D&C can be performed. If there is any question of viability, however, laparoscopy is preferred to rule out ectopic pregnancy first. Once fetal heart motion is observed within the uterine cavity, the possibility of a tubal ectopic pregnancy is virtually excluded. Certain patients should continue to be observed closely, however, particularly after a superovulation regimen, which is associated with an increased risk of a simultaneous intrauterine and ectopic pregnancy. Nevertheless, the overall risk of the two existing simultaneously even after a superovulation regimen is quite small. The use of vaginal ultrasonography with improved resolution and the addition of color Doppler flow analysis will invariably further define a lower discriminatory zone in the future. Both modalities appear to complement each other in attaining improved sensitivity and specificity in the diagnosis of ectopic pregnancy. Other investigative diagnostic techniques, such as selective salpingography and falloposcopy, should be considered strictly experimental.

TREATMENT FOR ECTOPIC PREGNANCY Part of "Chapter 22 - Ectopic Pregnancy"

Expectant Therapy Prior to the advent of effective therapy for ectopic pregnancy, it was noted that the condition was not uniformly fatal and that some patients had spontaneous resolution of the ectopic gestation, either through spontaneous regression or tubal abortion. The natural history of ectopic pregnancy therefore suggests that a number of tubal pregnancies can resolve without treatment. In 1988, Fernandez and associates observed a spontaneous resolution of ectopic pregnancy in 64% of carefully selected patients. The mean time for resolution was 20±13 days. Spontaneous resolution occurred more frequently when the initial hCG concentration was less than 1000 IU/L. The authors observed that a ?-hCG threshold of 1000 mIU/mL and a hemoperitoneum of less than 50 mL with a hematosalpinx of less than 2 cm appeared to be most compatible with successful expectant management. Subsequent larger studies have demonstrated similar results with expectant therapy. Korhonen, Stenman and Ylostalo have published the largest series to date. Criteria for patient selection included decreasing ?-hCG levels, an absent intrauterine pregnancy by transvaginal ultrasonography, and an adnexal mass of less than 4 cm without an embryonic heartbeat. Seventy-seven (65%) of 118 patients had spontaneous resolution of the ectopic pregnancy. The remaining patients had laparoscopy for either increasing abdominal pain, increasing cul-de-sac fluid volume or plateauing or increasing ?-hCG levels. One patient required a salpingectomy for a ruptured ectopic pregnancy. Similar to other reports, these authors also noted an increased expectant management success rate with lower initial ?-hCG concentrations.

Medical Treatment

Systemic Medical Therapy Methotrexate (MTX) is a folic acid antagonist that can be administered to eradicate trophoblastic tissue in an ectopic pregnancy. MTX is the chemotherapeutic agent of choice in the treatment of gestational trophoblastic disease. Long-term follow-up of women who have taken MTX for gestational trophoblastic disease has failed to demonstrate an increase in congenital malformations, spontaneous abortions, or second tumors after chemotherapy. Tanaka and colleagues reported the first use of systemic MTX for an ectopic pregnancy in 1982. This group successfully treated an interstitial pregnancy with a course of systemic MTX. Shortly thereafter, Farabow and coworkers described the use of systemic MTX in the treatment of a cervical pregnancy. Ory and associates reported the use of high-dose, short-course MTX therapy (plus citrovorum factor) to resolve small unruptured ectopic pregnancies without the need for conservative surgery. Six patients with ampullary pregnancies were treated, and resolution of the ectopic pregnancy was achieved in five patients. Surgical intervention was required in the sixth patient. Two of the five patients who experienced resolution, however, had protracted courses and required blood transfusions. Sauer and associates also reported the use of systemic MTX and citrovorum factor (CF) in 21 patients with ectopic pregnancy. Inclusion criteria included ?-hCG levels that were reaching a plateau or slightly rising, laparoscopic confirmation of the ectopic pregnancy, and a tubal diameter of less than 3 cm with the tubal serosa intact and no evidence of bleeding. Treatment consisted of 1.0 mg/kg MTX administered intramuscularly on postoperative days 1, 3, 5, and 7, along with 0.1 mg/kg CF administered intramuscularly on postoperative days 2, 4, 6, and 8. Twenty of 21 pregnancies resolved without the need for laparotomy. Two patients required blood transfusions, including one patient who required laparotomy and salpingectomy for a hemoperitoneum. In both of these cases, fetal heart activity in the adnexa was identified initially on ultrasound examination. This led the authors to suggest that MTX+CF can be safely used in selected cases of unruptured ectopic pregnancies that have not formed fetal elements that can be visualized by ultrasound. In 1991, Stovall and colleagues reviewed the results of several series of tubal ectopic pregnancies treated

with MTX + CF. Of 100 cases, 50 were diagnosed by laparoscopy, and 50 were diagnosed by a nonlaparoscopic algorithm. Complete resolution was achieved in 96 patients over a range of 14 to 92 days. In four patients, laparotomy was necessary because of tubal rupture; in one case, rupture occurred as late as 23 days after MTX administration. In five patients, cardiac activity was observed on ultrasound, and treatment was successful in four of them. Three patients experienced minor side effects. In 49 of 58 (84%) women who underwent subsequent hysterosalpingograms, tubal patency was demonstrated on the ipsilateral side. Of 56 patients desiring to conceive, 37 subsequently became pregnant; 33 of these were intrauterine pregnancies and 4 were repeat ectopic pregnancies. Stovall and Ling later studied the efficacy and safety of a simplified regimen of single-dose systemic methotrexate. All patients were diagnosed with a nonlaparoscopic algorithm by the use of serial hCG titers, serum progesterone, transvaginal ultrasonography, and curettage. Patients were treated with a single dose of 50 mg/m2 MTX intramuscularly if they were hemodynamically stable and the unruptured ectopic pregnancy did not exceed 3.5 cm in diameter. In the initial report, one hundred twenty patients were treated, including 14 (11.7%) with visualized cardiac activity. One hundred thirteen (94.2%) patients had compete resolution with treatment, with a mean time to resolution of 35.5 days. Four (3.3%) of the successfully treated patients required a second course of MTX on day 7. Seven (5.8%) patients required surgical management of the ectopic pregnancy, including two of the patients with cardiac activity. No major chemotherapy-related side effects were seen. Posttreatment hysterosalpingograms demonstrated tubal patency on the ipsilateral side in 51 of 62 (82.3%) patients. Of those attempting pregnancy, 79.6% subsequently became pregnant; 87.2% of these were intrauterine and 12.8% were ectopic. Lipscomb and colleagues further reported on the expanded Memphis cohort of patients treated with “single-dose” methotrexate. They utilized similar inclusion criteria, with the exception of further allowing pregnancies up to 4.0 cm in diameter provided that ectopic cardiac activity was not present. In this series, 287 of 315 (90.1%) patients were successfully treated with methotrexate. Forty-four patients with positive ectopic cardiac activity were treated with an 87.5% success rate. Of note, however, is that approximately 20% of these patients required more than one cycle of treatment. The authors' protocol reported that following the methotrexate dosing on day 1, serum chorionic gonadotropin was measured on days 1, 4 and 7. In many patients, they noted that ?-hCG levels frequently continued to rise until day 4. If the chorionic gonadotropin levels then declined less than 15% between days 4 and 7, the MTX protocol was repeated. If the levels declined 15 percent or more between days 4 and 7, serum ?-hCG was measured weekly until the level was less than 15 IU/L. If the chorionic gonadotropin level declined less than 15% during any subsequent week of follow-up, the MTX protocol was also then repeated. Other potentially difficult issues with “single-dose” methotrexate therapy include the management of “resolution pain” (which may occur in 20% of patients) and the prolonged time to resolution sometimes seen. A further review of the variables related to the success of single-dose methotrexate in the treatment of singleton ectopic pregnancy has been compiled. In this review, logistic regression analysis demonstrated that the serum chorionic gonadotropin level before treatment was the only factor that contributed significantly to the failure rate (Table 22.4). Interestingly, the size and volume of the mass, the volume of hematoma, and the presence or absence of free peritoneal blood in the pelvis were not associated with a significant risk of treatment failure. TABLE 22.4. Success Rates of Methotrexate Treatment With Ectopic Pregnancies as a Function of Their Initial Serum Chorionic Gonadotropin Concentrations

Serum Chorionic Gonadotropin Concentration (IU/L)

Success Rate (95% confidence interval)

40 points.

b

If the fimbriated end of the fallopian tube is completely enclosed, change the point assignment to 16.

This revised AFS classification has been widely used by investigators to categorize disease states. Nevertheless, direct comparison of treatment outcome is compromised by inconsistencies in the application of the staging criteria and by the great variations in medical and surgical therapeutic options being applied in the management of endometriosis. Evaluation of the extent of disease by laparoscopy may be limited by a lack of recognition of atypical implants, particularly if the patient is hypoestrogenic as a result of recent discontinuation of medical therapy for endometriosis. Furthermore, the divisions between stages of endometriosis remained arbitrary, the point score for ovarian involvement was weighted too heavily, and the classification scheme did not address disease involving the fallopian tubes, intestines, or urinary tract. Also, there were no parameters to indicate the present activity and state of evolution of the disease. Active endometrial tissue is usually found in the white lesions. The Endometriosis Classification Subcommittee of the American Society for Reproductive Medicine released new recommendations in 1996 for the documentation of the extent and location of disease. One concern over the reproducibility of the scoring system was directed at the variability in assessing ovarian endometriosis and cul-de-sac obliteration. The subcommittee indicated that an endometriotic cyst should be confirmed by histology or by the presence of the following features: (1) cyst diameter less than 12 cm; (2) adhesion to pelvic side wall and/or broad ligament; (3) endometriosis on surface of ovary; (4) tarry, thick, chocolate-colored fluid content. Cul-de-sac obliteration should be considered partial if some normal peritoneum is visible below the uterosacral ligaments, but adhesions or endometriosis have obliterated part of the cul-de-sac. Complete obliteration exists when no peritoneum is visible below the uterosacral ligaments. Because information is accumulating to suggest that the morphologic appearance of the endometriotic implants may correlate with biologic activity and consequently fertility, the newly revised classification scheme requests the categorization of lesions as red, white, and black. The percentage of surface involvement of each implant type is to be documented. The revised AFS (ASRM) classification system is oriented toward the infertile population. Muzii and colleagues, using a pain questionnaire administered to women before surgery, found a significant correlation between the severity of dysmenorrhea and total revised AFS score, partial score for deep disease, and partial score for adhesions. However, they found no correlation between the pain score for dysmenorrhea and the partial score for superficial disease, number of typical and atypical implants, or the total number of implants. Limited knowledge of the specific pathophysiologic alterations by which endometriosis can cause these symptoms has so far prevented any precise categorization of disease based on response to conventional therapies for these symptoms.

THERAPIES Part of "Chapter 25 - Endometriosis" Although women with endometriosis can present with a range of symptoms, therapy is usually initiated for the correction of pain, infertility, or a persistent pelvic mass. Pain and infertility can coexist in a patient; nevertheless, many women with endometriosis-associated infertility have relatively little or no discomfort. Treatment options vary depending on the clinical history and findings at the time of surgery.

Expectant Management Treatment of mild and moderate endometriosis with hormonal preparations may not offer any advantage over expectant management in promoting conception. In studies by Seibel and colleagues, Hull and associates, and Telimaa, patients assigned to expectant management conceived earlier than the medically treated group, and the cumulative pregnancy rate was not higher for women receiving progestogens or danazol. This lack of enhancement of fecundity may be related to the lower number of estrogen, progesterone, and androgen receptors in endometriotic lesions as compared with normal endometrium. Nevertheless, patients with minimal or mild disease who have pelvic pain or dysmenorrhea do benefit from hormonal therapy. The age of the patient and the duration of her infertility are important factors to consider in determining the appropriate therapy for the symptomatic individual. Laparoscopic laser ablation of milder stages of endometriosis appears to lessen the interval to conception, although the cumulative pregnancy rate may not be greater than that of women managed expectantly. Surgical therapy for more advanced disease results in a higher pregnancy rate than does expectant management or hormonal treatment, partly because of correction of mechanical factors that may be inhibiting ovulation or tubal function. There is no direct evidence to support the contention that surgical treatment of minimal or mild endometriosis in the asymptomatic patient will hinder future disease progression and sequelae. The potential benefits of cytoreductive therapy must be weighed against the risk of adhesion formation through surgical devitalization of peritoneal surfaces.

Medical Treatment Mild pain symptoms associated with endometriosis may be effectively treated with nonsteroidal antiinflammatory agents and oral contraceptives. Additional endocrinologic therapies include progestogens, danazol, and GnRH agonists. These agents have similar degrees of efficacy in the relief of pain symptoms; side effects vary depending on their mechanism of action.

Progestogens High-dose combination estrogen/progestogen regimens were introduced in the late 1950s for the symptomatic relief of moderate to severe pain associated with endometriosis. The rationale for this therapy was based on the observation that pregnancy provided subjective and objective improvement in many patients with extensive pelvic endometriosis. As is often seen in pregnancy, high doses of estrogens and progestogens are thought to transform endometrial tissue into decidua that ultimately undergoes necrosis and involution. Oral contraceptives with strongly progestational properties have been prescribed in the past. Typical regimens included an initial dose of 2 tablets daily; the dose was increased by 1 to 2 tablets at biweekly intervals until the patient was amenorrheic or was receiving the equivalent of 20 mg of norethynodrel. This dose was continued for 6 to 9 months. Although this combination therapy did relieve pelvic pain and dysmenorrhea in 50% to 80% of patients, significant side effects were encountered, including weight gain, mastalgia, nausea, headaches, and irregular bleeding. As a result, the discontinuation rate was high. Because of these side effects and the potential risks of the high-dose administration of estrogen in some patients, progestogen-only regimens have gained favor over the continuous high-dose oral contraceptive schedule for creating a pseudopregnancy state. Progestogens inhibit the pituitary release of LH and thereby suppress ovarian steroidogenesis and promote secretory changes in the glandular epithelium and decidualization of the endometrial stroma. Progestogens oppose the growth-promoting effects of estrogens on the endometrial tissue by altering the clearance of the nuclear estrogen receptor and inducing 17?-hydroxysteroid dehydrogenase, which converts estradiol to the weaker estrone. Moreover, by eliminating cyclic bleeding and suppressing uterine contractility, progestogens prevent reflux menstruation, a potential stimulus for continued endometriosis development. Luciano and colleagues administered medroxyprogesterone acetate, 50 mg daily for 4 months, to symptomatic women with moderate to severe endometriosis. Improvement of pain, pelvic nodularity, and tenderness on examination occurred in 80% of patients. Twenty percent of women experienced

breakthrough bleeding, and an additional 10% reported persistent cyclic bleeding. Minor weight gain, edema, and increased irritability were other described side effects, which were generally well tolerated. A lower daily dose of 30 mg may provide equivalent relief of symptoms. Compared with the cost of danazol and GnRH agonists, which are the other commonly prescribed agents, the low cost of the medroxyprogesterone acetate is a notable advantage. Parenteral depot medroxyprogesterone acetate has also been used to produce long periods of amenorrhea and elicit direct progestational changes of the endometrial tissue. A regimen of 150 mg intramuscularly every 3 months for 1 year has been used to manage endometriosis patients with moderate to severe pelvic pain. Twenty-nine of 40 subjects (72.5%) were satisfied with their pain relief after 1 year of therapy. A similar response rate can be obtained with megestrol acetate. Doses of 40 mg per day for up to 24 months resulted in significant relief of dysmenorrhea, noncyclic pelvic pain, and dyspareunia in 86% of subjects. The rate of recurrence of symptomatic endometriosis after progestogen therapy appears to be related to the length of follow-up. Riva and colleagues reported an 18% rate after an average of 11 months, whereas Moghissi and Boyce described a 42% recurrence rate during a 2-year interval after discontinuation of medication. Relief of pelvic pain, particularly cramping associated with menstruation, can be achieved with cyclic administration of low-dose oral contraceptive pills. This line of therapy should be considered for the woman with mild symptoms who is not attempting to conceive. Low dose (20 to 35 µg ethinyl estradiol) combination oral contraceptives may be given daily for 6 to 9 months without break to relieve pain or more severe dysmenorrhea. The dose may be increased to 2 or more tablets per day for several days to alleviate episodes of breakthrough bleeding.

Danazol Danazol is a synthetic (2,3-isoxazole) derivative of 17?-ethinyl testosterone that was introduced into clinical practice by Greenblatt and colleagues in 1971 after good performance in uncontrolled trials. The drug gained rapid acceptance because of its effectiveness in relieving pain associated with endometriosis and in enhancing fertility. All of the progestational and the weak androgenic effects of the drug result from retention of the methyl group in the 19 position of the steroid nucleus, whereas the oral activity of danazol is ascribed to the ethinyl group at position 17. The pharmacologic action of danazol is complex. By directly inhibiting GnRH secretion, the midcycle LH surge is ablated, although basal gonadotropin concentrations are maintained. The drug interacts with endometrial androgen and progesterone receptors, suppresses the activity of multiple enzymes necessary for ovarian and adrenal steroidogenesis, and displaces androgens from sex hormone–binding globulin, thereby augmenting androgen action on endometrial receptors. The decline in sex hormone–binding globulin induced by danazol lowers estradiol binding, increases estradiol clearance, and promotes a decline in the circulating level of this hormone. Hence, the derivative has direct androgenic and antiprogestational action on endometrial implants and creates a hypoestrogenic, hypoprogestational environment antagonistic to endometriosis. Moreover, by producing amenorrhea, danazol prevents peritoneal seeding of refluxed endometrial tissue. In addition, danazol is capable of suppressing elevated autoantibodies in several autoimmune diseases and has been shown to decrease immunoglobulin and autoantibody levels in women with endometriosis. In contrast to GnRH agonists, danazol use maintains a normal estrogenic state and increases bone mineral density over baseline. The adverse effects of danazol reflect its anabolic, androgenic, and antiestrogenic properties and may be dose-related (Fig. 25.3). Weight gain, muscle cramps, decreased breast size, and vasomotor symptoms are noted in 50% or more of patients maintained on doses of 400 to 800 mg per day. In Buttram's 1985 series, 41% of patients treated with the standard dose of 800 mg per day gained more than 10 pounds during the course of therapy. The threefold increase in free testosterone can cause acne, oily skin, and deepening of the voice in a small percentage of recipients. High-density lipoprotein (HDL) cholesterol declines by 50% or more in response to the altered steroid concentrations; an 80% decrease in the HDL2 subfraction has been reported. Most series have described a concomitant increase in low-density lipoprotein (LDL) cholesterol; the alteration in the ratio of HDL to LDL cholesterol may be an unacceptable risk to some patients. Because danazol is metabolized by the liver, modest elevations in serum glutamic oxaloacetic transaminase and serum glutamate pyruvate transaminase may arise. Reported idiopathic drug reactions include gastrointestinal disturbances, weakness, dizziness, skin rashes, headaches, and muscle cramps. Bothersome side effects occur in as many as 85% of patients, and at least 10% of women receiving danazol discontinue pharmacologic treatment because the adverse effects are

intolerable. Combining danazol therapy with aerobic exercise appears to reduce the incidence of many of these androgenic side effects. Preliminary data from trials using danazol rings suggest that this route of administration may result in symptomatic improvement of pain while avoiding the androgenic side effects noted with oral administration. FIGURE 25.3. Summary of most frequent adverse effects associated with therapy in patients receiving danazol or nafarelin. *Women with hot flashes before treatment were excluded (one had danazol, five had nafarelin). + p < .01 for between-treatment difference. ++ p < .05 for between-treatment difference. (Redrawn from Burry KA. Nafarelin in the management of endometriosis: quality of life assessment. Am J Obstet Gynecol 1992;166:735) View Figure

Because of the potential androgenic action of this hormone on the developing fetus, the patient must not be pregnant when initiating therapy. Barrier contraception has been recommended for the entire course of treatment to eliminate the possibility of conception, although high doses of danazol usually cause anovulation. The amenorrhea induced by danazol has been found to benefit patients with dysmenorrhea, dyspareunia, and cyclic pelvic pain associated with endometriosis. Young and Blackmore reviewed the effects of different dosages of danazol with respect to relief of symptoms in 452 patients. At a dose of 800 mg, 95% of patients noted relief of dysmenorrhea, and 89% reported relief of pelvic pain. At a dose of 400 mg, posttherapeutic relief was reduced by 10%. Moore and associates reported that pain associated with minimal and moderate pelvic endometriosis appeared to respond well to doses of danazol of 400 mg or less per day, whereas severe endometriosis was best treated with doses greater than 400 mg per day. A 6-year prospective study that evaluated the effectiveness of danazol at two doses (400 mg and 800 mg) in carefully classified patients concluded that there was no difference in side effects between the two doses and that gross resolutions of disease at second-look laparoscopy were similar. However, ovarian endometriosis greater than 1 cm did not respond as well to either dose of danazol as did peritoneal or ovarian disease less than 1 cm. Recurrence of symptoms within 4 to 12 months of discontinuation of danazol therapy approaches 50% in most studies. Puleo and Hammond found that pain recurred in 38% of patients after a mean of 6.9 months; active disease was found within 1 year in 51% of women. Lower daily doses of medication or courses of treatment less than 4 months in duration may result in a shorter symptom-free interval. Danazol has been extensively prescribed as therapy for endometriosis-associated infertility, although there are no well-controlled studies to support this indication. Pregnancy rates after the use of danazol as the sole therapy have ranged from 30.9% to 52.6% in mild endometriosis, 23.1% to 50% in moderate disease, and 0% to 100% in severe stages. Monthly fecundity rates range from 1.6% to 6.8%. The findings of many early studies have been questioned because of their lack of randomization, failure to include an expectant management control group, and failure to account for other infertility factors. Recent data concerning medical therapy of minimal, mild, and moderate stages of disease refute the notion that danazol may enhance conception. When all other infertility factors are excluded, the estimated monthly fecundity rate with expectant management of mild endometriosis is 8.7%. Furthermore, conception is delayed while the patient is receiving danazol.

Gonadotropin-Releasing Hormone Agonists GnRH agonists are available for use in the treatment of estrogen-dependent diseases such as endometriosis. Some of the more frequently studied analogs include leuprolide, nafarelin, buserelin, and goserelin. Alteration of the amino acid at position 6 and ethylamide replacement of the C-terminal amino acid of the native decapeptide hormone results in a GnRH agonist with increased resistance to lysosomal degradation. Pituitary receptor binding is enhanced, resulting in a decline in the number of receptors available for further occupancy. Continued administration of the GnRH agonist leads to a desensitization

of the pituitary gonadotrope receptor and a reversible downregulation of the pituitary-ovarian axis. Ovarian estrogen secretion may reach castrate levels. The initial response to GnRH agonist administration is a markedly increased secretion of pituitary stores of follicle-stimulating hormone (FSH) and LH. If therapy is begun in the follicular phase of the menstrual cycle, the developing follicle may respond to the flare in circulating gonadotropin levels with a rapid increase in estradiol production. Estradiol levels may remain elevated for 3 weeks before declining. GnRH agonist administration in the luteal phase leads to a more rapid decline in estrogen secretion, although FSH and LH levels remain elevated for 1 and 4 weeks, respectively. GnRH agonist treatment results in improvement or resolution of pain symptoms in all stages of disease. Lemay and colleagues reported resolution of pain in 70% and improvement in discomfort in 15% of 24 subjects after 2 to 4 months of treatment with the agonist buserelin. Dyspareunia improved in 9% and disappeared in 91% of patients studied. The depot formulation of leuprolide acetate has also been shown to significantly reduce dysmenorrhea, pelvic pain, and pelvic tenderness in patients with endometriosis. Henzl and associates, in a double-blind, multicenter study, treated 213 patients with either danazol or nafarelin. After 6 months of treatment, more than 80% of patients in all groups experienced a significant reduction in visible implants. A 43% reduction in AFS score was noted for each treatment group; there was no difference in response among patients receiving the 0.4- and 0.8-mg daily dose of nafarelin. Most patients continued to demonstrate some visible implants at the time of follow-up laparoscopy, and, as with danazol, there was some diminution in size of endometriomas but no effect on preexisting adhesions. The optimal interval of GnRH analog administration has been widely debated. Six months of medication has been traditionally prescribed, although a significant reduction in implant volume occurs as early as 2 weeks after initiation of treatment in the rat model. A maximal effect was measured after 4 weeks of therapy in this animal study, suggesting that short courses of drug may be as efficacious as 6 months of continuous therapy. The regrowth of lesions after estrogen therapy has been reported years after the menopause; hence, hypoestrogenism results in inactivation rather than resolution of the disease. Response to therapy may be dependent on route of administration. Donnez and colleagues reported that buserelin administration by a long-acting subcutaneous implant led to a greater reduction in endometriosis score, mitotic index, and endometrial cyst diameter than when given in an intranasal form. This may have been due to a greater consistency in hormonal release by the injected preparation. As occurs with danazol and progestogen regimens, symptoms recur at variable periods after discontinuation of GnRH analog therapy. Subjective return of pain occurred in 57% of patients within 6 months of discontinuing leuprolide, although 37% with moderate or severe pelvic pain at baseline were still improved at 1 year. Franssen and colleagues noted a lasting and significant amelioration of dysmenorrhea and dyspareunia 6 months after completion of treatment; however, scores for chronic pelvic pain had nearly reached their pretreatment level once this time had elapsed. Patients with a higher disease stage at the onset are more likely to experience recurrence and to experience it earlier than patients with minimal disease. One treatment option for such patients may be a second 3-month course of GnRH analog. Henzl reported a significant decrease in mean pain scores and essentially no change in compact bone density in most patients when nafarelin was readministered for 3 months after a treatment-free interval of 6 months or more. The effect of GnRH analog on endometriosis-associated infertility is difficult to assess because of a lack of an expectant management control group in most clinical studies. The preliminary pregnancy rates, which range from 0% to 60%, are derived from trials that do categorize response based on stage of disease. Most of the side effects associated with GnRH analog therapy are related to hypoestrogenism. Hot flashes are common and can lead to sleep disturbances and chronic fatigue in extreme cases (Fig. 25.3). Vaginal dryness, superficial dyspareunia, headaches, and depression have been reported. In general, these adverse effects are better tolerated than those experienced with danazol use. In addition, there are no undesirable changes in HDL, LDL, or total cholesterol throughout the prolonged period of hypoestrogenism induced by GnRH analog, unlike the changes accompanying danazol intake. A decline in trabecular bone mineral content and an increase in urinary calcium excretion to menopausal levels occur during the course of GnRH analog therapy in about two thirds of patients. Quantitated computed tomographic studies consistently show significant loss of trabecular bone of the vertebrae and hip with GnRH analog exposure. Restoration of normal estrogen production after cessation of therapy appears to at least partially reverse these bone changes. In a study of the GnRH agonist goserelin, an 8.2% decline in density of the lumbar spine was measured after completion of 6 months of treatment; this improved to a mean loss of 5.4% at 6 months postcessation. Others found no significant change from baseline after a 6-month course of GnRH analog when bone density was assessed 6 months after treatment.

Concomitant administration of a progestogen during the course of GnRH analog therapy has been examined to ameliorate vasomotor symptoms and retard both urinary calcium excretion and radiologic evidence of loss of bone mineral density. Cedars and co-workers reported a diminution in the side effects mentioned earlier when medroxyprogesterone acetate was administered at a dose of 20 to 30 mg per day during the 6-month course of agonist therapy; however, laparoscopic evaluation after completion of therapy failed to reveal any improvement or suppression of active endometriosis with the combination regimen, and the regimen failed to significantly reduce symptoms of pelvic pain. Conversely, Makarainen and colleagues reported that medroxyprogesterone acetate, 100 mg per day, diminished hot flushes and the urinary excretion of calcium in women treated with goserelin acetate, 3.6 mg monthly, for 6 months. Second-look laparoscopy revealed equivalent diminution in extent of endometriosis when compared with the goserelin-progestin placebo group. Norethindrone, a 19-nortestosterone progestin, has been shown to suppress both the painful symptoms of endometriosis and the extent of disease at laparoscopy when used in daily doses of 1.4 mg to 10 mg during GnRH agonist therapy. A recent randomized, double-blind study has demonstrated that GnRH agonist therapy may be safely and effectively extended for up to 1 year in the management of endometriosis-associated pelvic pain when prescribed in conjunction with low-dose sex steroid hormones. Hornstein et al. reported that norethindrone acetate, 5 mg, alone or in combination with conjugated equine estrogens, 0.625 mg daily, from the onset of depot leuprolide acetate therapy alleviated hypoestrogenic symptoms and preserved bone density while resulting in equivalent pain relief to that achieved by the placebo estrogen-progestin patient group. The addition of calcium carbonate and alendronate or etidronate sodium, which are organic bisphosphonates, to the low-dose norethindrone acetate add-back therapy in patients with symptomatic endometriosis receiving prolonged GnRH agonist treatment may further minimize the adverse side effects of hypoestrogenism. Additional controlled studies will better establish the optimal medical management of this condition.

Conservative Surgery Endoscopic assessment of the pelvis allows determination of the appropriate therapy for the patient with endometriosis. Surgery is indicated for correction of pain, infertility, or other symptoms in patients with extensive pelvic endometriosis, or when hormonal manipulation fails to adequately diminish pain symptoms in women with lesser stages of disease (Table 25.7). Surgery is successful in relieving pain in a very high percentage of cases and offers a better prognosis for pregnancy than does endocrine therapy in cases of advanced disease. The surgeon who has mastered the specialized techniques of operative laparoscopy can treat a wide range of pathologic findings at the time of diagnosis. Therapeutic planning depends on many factors, including the age of the patient, her desire for fertility or pain relief, the duration and intensity of her symptoms, the extent of disease, and previous treatments undertaken. Preoperative rectoscopy-sigmoidoscopy and intravenous pyelography are recommended in patients with symptoms suggestive of deeply invasive endometriosis of the posterior cul-de-sac and rectovaginal septum. TABLE 25.7. Treatment of Endometriosis

Desires Childbearing

Childbearing Complete

Stage of Disease

Infertility

Pelvic Pain

Pelvic Pain

Stages I and II

1. Expectant Rx 2. Laparoscopic Rx 2. Medical Rx 3. CSEL ± PSN 3. IVF/ET

1. Laparoscopic Rx 2. Medical Rx 3. CSEL + PSN

1. Laparoscopic Rx 2. Medical Rx 3. TAH ± BSO 3. CSEL ± PSN

Stage III

1. Laparoscopic Rx 1. CSEL ± PSN 2. Medical Rx 3. IVF/ET

1. Laparoscopic Rx 2. CSEL ± PSN 3. Medical Rx

1. Laparoscopic Rx 2. Medical Rx 3. TAH ± BSO 3. CSEL + PSN

Stage IV

1. CSEL + perioperative medical Rx 2. CSEL alone 3. Laparoscopic Rx + postoperative medical RX 4. IVF/ET

1. CSEL + PSN + peripoerative medical Rx 2. Medical Rx 3. Laparoscopic Rx + medical Rx

1. TAH ± BSO 1. CSEL + PSN + medical Rx 1. Laparoscopic Rx + medical Rx

Modified from Wheeler JM. Macro- and microsurgical treatment of endometriosis. In: Thomas E, Rock JA, eds. Modern approaches to endometriosis. Boston: Kluwer Academic Publishers, 1991, with permission.

Rx, treatment; CSEL, conservative surgery for endometriosis at laparotomy; PSN, presacral neurectomy; TAH, total abdominal hysterectomy; BSO, bilateral salpingo-oophorectomy; IVF/ET, in vitro fertilization and embryo transfer; ±, adjunctive treatment option based on individual patient findings.

The decision of whether to perform surgical resection of endometriosis through the laparoscope or open abdomen is not entirely dependent on the stage of disease encountered. Laparoscopy can be considered for all cases unless there is difficulty in establishing the appropriate tissue planes of dissection or unless improved access is necessary for atraumatic manipulation of the involved organs. Specific endoscopic procedures include ablation of endometriotic implants, adhesiolysis, ovarian cystectomy, oophorectomy, and salpingectomy. Although the results and complications are similar, the cost savings with respect to decreased hospital expenses and loss of work time favor laparoscopy over laparotomy when other factors regarding risks and outcome are equal. Laparoscopy provides superior visualization of the posterior cul-de-sac and allows a high degree of magnification of peritoneal surfaces, which aids in the identification of subtle disease. Conservative resection of disease by laparotomy is most valuable in cases of extensive, dense pelvic adhesions or endometriomas greater than 5 cm in diameter. In addition, deep involvement of the rectovaginal septum with fibrotic extension into the perirectal fossa, invasion of the bowel muscularis, and endometriotic infiltration in the region of the uterine vessels and ureter are generally best approached through the open abdomen for all but advanced endoscopic surgeons. The objective of the laparotomy procedure is complete excision of all endometriosis and associated adhesive disease to restore normal functional anatomy of the reproductive tract. The usual surgical approach is through a transverse suprapubic incision. A Maylard incision provides adequate exposure for presacral neurectomy and reconstructive surgery of ovarian endometriomas of almost any size.

Principles of Microsurgery Microsurgical technique, or the philosophy of gentle manipulation of tissue in an attempt to avoid trauma, is the major tenet of pelvic reconstruction. The inflammation, trauma, coagulation, and foreign materials associated with conventional macrosurgical technique lead to tissue ischemia and adhesion formation because of local failure of the intrinsic peritoneal fibrinolytic system. Adhesion formation can be reduced by the application of loupe magnification or use of the operating microscope, reconstruction with fine, nonreactive sutures, precise hemostasis, and continuous irrigation of tissues with warmed lactated Ringer solution, each liter of which is supplemented with 5,000 U of heparin. Nevertheless, there are no definitive data to suggest that use of the particularly costly, ancillary laser and the operating microscope has appreciably improved the reproductive prognosis in the surgical management of endometriosis through laparotomy. Several basic techniques are available for the endoscopic ablation of endometriosis, including excision, coagulation, and vaporization. Coagulation can be achieved by monopolar or bipolar cautery, thermocoagulation, or, in some circumstances, laser, depending on the wavelength of energy applied. The extent of tissue penetration in electrocautery is related to the power and type of current, the duration of application, and the size of the electrode. Less tissue damage is achieved with bipolar than with monopolar cautery. The carbon dioxide (CO 2) laser is more precise than the fiber lasers, although CO2 laser energy is strongly absorbed by water molecules and is rendered ineffective in the presence of blood. Meticulous technique that maintains serosal integrity may reduce the incidence of de novo adhesion formation.

Sites of Conservative Surgery PERITONEUM Small lesions of superficial peritoneal endometriosis less than 5 mm in diameter are easily treated with laser or bipolar coagulation while under a constant stream of irrigation. Deep lesions or more extensive peritoneal disease must be excised with a tissue margin of at least 2 to 4 mm, because, as noted previously, microscopic lesions are commonly present in tissue adjacent to visible implants (Fig. 25.4). Ablation of deep disease by monopolar microdiathermy or CO2 laser vaporization rather than excision of the disease may result in inadequate resection and a greater amount of ischemic damage to the tissue, heightening the propensity toward adhesion formation. Immobilizing adhesions can be merely divided during the preparatory phase of the procedure; precise excision is more easily accomplished after the involved organs are freed. Before dissection of the pelvic side wall, the ureter must be identified and isolated; it frequently is displaced from its normal location by endometriotic adhesive disease. A Lucite, Teflon, or laparoscopic titanium probe can be used to isolate adhesions and protect adjacent structures during separation of the tissue planes. Suture placement can lead to foreign body reaction, tissue anoxia,

and fibrosis and should therefore be avoided. Covering hemostatic, deperitonealized surfaces with an P.613 absorbable, oxidized, regenerated cellulose barrier (Interceed) significantly reduces the incidence, extent, and severity of postsurgical pelvic adhesions, even in patients with severe endometriosis. Alternatively, application of the Gore-Tex surgical membrane has been shown to result in a statistical reduction in adhesion score; this barrier can be removed at the time of a second-look laparoscopic procedure if its presence would impair tuboovarian function. FIGURE 25.4. Excision or CO2 laser vaporization of peritoneal implants. A: Superficial implants are vaporized by use of power densities between 1,000 and 3,000 W/cm2, with a spot

View Figure

size of 0.8 to 1 mm, or they are cauterized with microbipolar forceps. More extensive peritoneal disease is excised. Very large defects can be closed with 5-0 or 6-0 polyglactin or polydioxanone sutures. Adhesion barriers can be placed. B: Endometriosis can be associated with extensive adnexal adhesions. C: Wide adhesion bands can be retracted with a glass rod and completely excised with a monopolar microelectrode.

Estimations of the depth of endometrial implants at the time of laparoscopic resection relate well with histologic measurements. Superficial implants can be destroyed by bipolar cauterization; however, 25% of patients have lesions greater than 5 mm in depth. Deep (greater than 5 mm) and very deep (greater than 10 mm) lesions represent an active form of the disease and occur almost exclusively in patients who complain of pain. The diagnosis of retroperitoneal endometriosis is suggested by preoperative digital rectovaginal palpation and laparoscopic blunt probe palpation. The depth of infiltration of deep lesions appears to correlate poorly with the visible surface area of involvement. The laparoscopic treatment of deep disease is often complicated by the proximity of implants to vital structures such as the ureter, bladder, and vessels. The superficial action of nonvaporizing modalities such as bipolar or thermal coagulation is not sufficient for deep disease. Laparoscopic forceps are used to elevate and isolate the tissue to be excised. Instruments should be placed with care, because surgical manipulation of tissue that will not be resected may result in de novo adhesion formation. The diseased peritoneum may also be separated from underlying tissue by the technique of hydrodissection, which forcefully injects physiologic irrigant retroperitoneally through a small defect created in the peritoneum (Fig. 25.5). This retroperitoneal placement of fluid acts to dissipate CO 2 laser energy and, in so doing, promotes safer dissection or vaporization of the peritoneal surface. Coagulation or vaporization of disease in the ovarian fossa or near the uterosacral ligament should be undertaken only after clear identification of the ureter. Uterine manipulation with a Valtchev retractor may be used while treating lesions of the posterior cul-de-sac.

FIGURE 25.5. Hydrodissection.

View Figure

Dissection of retroperitoneal disease can be facilitated by placing a bougie probe in the rectum and a sponge forceps in the vagina (Fig. 25.6). Traction in either direction opens the rectovaginal and perirectal spaces. Initial dissection of the anterior rectum provides a landmark of the retrovaginal space and permits posterior mobilization of the nodule. Subsequent lateral dissection is performed, followed by anterior dissection, which permits retrieval of the involved tissue. FIGURE 25.6. Laparoscopic therapy for endometriosis. A: Dorsal lithotomy positioning for surgery, with multiple puncture sites marked for placement of ancillary instruments. B: Traction on bougie in rectum and sponge forceps in vagina mobilize rectovaginal and perirectal spaces. Uterine manipulation cannula, bladder drainage, and multiple transabdominal instruments facilitate safe dissection. C: CO2 laser division of endometriosis-associated adhesions extending from lower uterus to rectal serosa.

View Figure

It is difficult to evaluate the depth of tissue damage with electrocauterization; however, laser vaporization allows visualization of the three-dimensional boundaries of every lesion. The laser beam should be applied until the bubbling of retroperitoneal areolar tissue is noted. The zone of thermal necrosis is minimal with the CO 2 laser, particularly when applied in the superpulse mode. In the region of the ureter, urinary bladder, colon, or large blood vessels, a single or repeat pulse mode of 0.05 to 0.1 second allows a depth of penetration of 100 to 200 µm. Irrigation of the pelvis washes off debris and carbon deposition and better exposes the base of the site of laser impact. A 2- to 4-mm clear margin is desired around each lesion treated. Excision of the involved peritoneum is superior to vaporization of implants when the extent of tissue penetration cannot be recognized. Resection of deep posterior cul-de-sac nodules requires great endoscopic expertise. A combined laparoscopic-vaginal approach may be necessary to effectively remove these implants (Fig. 25.7). It is often helpful to have an assistant place his or her fingers deep in the vaginal fornices or a bougie probe in the rectum to indicate the sites of the nodules to ensure their complete removal (Fig. 25.7). The direct palpation made possible through laparotomy may be required to recognize all indurated, deep lesions. A complete bowel preparation is mandatory in all cases of suspected deep endometriosis.

FIGURE 25.7. A: Deep laparoscopic dissection of the rectovaginal space, in combination with colpotomy, for the excision of a large endometriotic nodule of the rectovaginal septum. B: Initial laparoscopic dissection of nodule. C: Completion of dissection by way of colpotomy incision. D: Vaginal suture placement to reapproximate the rectovaginal septal defect.

View Figure

Defects in the peritoneal surface are frequently associated with endometriosis and are most commonly found in the posterior cul-de-sac region. These defects should be explored and ablated even if they appear grossly normal, because of the frequency of microscopic disease. Electrosurgery should be avoided when extensive dissection is performed because it may be associated with widespread thermal damage and difficulty in recognizing tissue planes. Superficial invasion of the muscularis of bowel or bladder can be treated with laser vaporization or endocoagulation because of the precision and lack of penetration of these energy sources. Anterior cul-de-sac treatment should be accompanied by continuous bladder drainage. Tubal endometriosis may distort the normal anatomic relationship of the distal tube to the ovary and in severe cases may cause complete fimbrial obstruction. Short pulses of CO2 laser may be used to vaporize lesions while minimizing thermal damage. Endoscopic adhesiolysis of the distal tube may be accomplished with fine scissors or careful application of laser. Unipolar electrocautery should not be used on this tissue.

OVARY Superficial endometriosis of the ovary usually presents as small, dark, punctate lesions immediately beneath the cortical surface. This disease can be readily treated with laser or bipolar forceps under constant irrigation. Occasionally, however, the small, visible lesion may be merely the tip of a large endometrial cyst. If there is any doubt, the implant should be excised and the ovary explored to determine the extent of disease. Care should be taken to minimize thermal injury to surrounding ovarian tissue. This is particularly important near the fimbria ovarica, because postoperative adhesion formation could compromise distal tubal function. Inability to elevate the ovary is usually a sign of adhesions and endometriotic implants of the inferolateral surface of the ovary and the peritoneum of the ovarian fossa.

Reconstruction by Laparotomy Extensive ovarian endometriosis is often associated with periovarian and peritubal adhesions. These adhesions may become apparent while manipulating the ovary to visualize the lateral surface adjacent to the broad ligament. Filmy adhesions are elevated with delicate tissue forceps and can be resected with fine-needle cautery, a scalpel, or the laser. Care must be taken to maintain the integrity of the ovarian capsule. After the appropriate adhesiolysis is accomplished, the posterior cul-de-sac is packed with moist, lint-free packs, and a silicon surgical platform can be placed to stabilize the adnexa. The ovary should be carefully examined for extent of disease involvement before creation of the initial incision. Peritoneal spillage of the contents of the endometrioma can be avoided by placement of a lint-free pack around the platform. The cortical incision should be made in a way that will preserve the normal anatomic relations of the ovary with the uteroovarian ligament and fimbria ovarica (Fig. 25.8). This is best accomplished by making a shallow longitudinal incision over the endometrioma with the monopolar microneedle, scalpel, or laser. The surgeon should attempt to remove the endometrioma in an intact state; however, if the cyst cavity is

inadvertently entered, an elliptical incision around the site of rupture is useful for exposure. The intact endometrioma can be transfixed with a traction suture of 2-0 nylon to facilitate creation of a cleavage plane between the cyst and normal ovarian tissue. Blunt, curved scissors or a flat probe or knife handle is used for dissection. Particular care must be taken when dissecting the hilar region to maintain hemostasis. An attempt should be made to preserve as much of the normal ovarian cortex as possible; pregnancies have been achieved with only a small fraction of remaining ovary.

View Figure

FIGURE 25.8. Excision of ovarian endometrioma through laparotomy. A: The ovarian cortex is gently incised so that the endometrial cyst is not entered. The incision is made along the longitudinal axis of the ovary. B: The endometrioma is then peeled out with the blunt knife handle. C and D: The ovarian defect is closed with two layers of pursestring sutures of 4-0 or 5-0 absorbable, nonreactive material. E: In the case of a deep defect, a more superficial running suture may be necessary before the cortical edges are approximated with 5-0 nonreactive, delayed-absorbable sutures.

The ovary is reconstructed by placing one or two pursestring sutures of 4-0 or 5-0 polyglactin, polyglycolic acid, or polydioxanone to eliminate the dead space and maximize hemostasis. This is followed by placement of a running subcortical 5-0 suture of the same delayed-absorbable material (Fig. 25.8). In some circumstances, less tissue distortion can be achieved by placing a deep layer of interrupted mattress sutures followed by additional layers of running sutures (Fig. 25.8). Suture on or extruding through the surface should be strictly avoided because of its adhesiogenic properties. After the ovary has been carefully approximated, the posterior surfaces of the uterus and broad ligament are inspected for hemostasis wherever the ovary was previously adherent. Microbipolar cauterization may be necessary. Placement of an adhesion barrier is useful in separating raw peritoneal surfaces during the healing process.

Endoscopic Therapy of Ovarian Endometriosis Surgical treatment of endometriosis less than 4 to 5 cm in diameter can be accomplished with relative ease; however, endoscopic resection of larger lesions may be compromised by the presence of dense, cohesive adhesions and by difficulties removing the entire cyst wall. The endometrioma can be excised in an intact or ruptured state during the laparoscopic procedure. In either case, the technique is initiated by longitudinally incising the cortex overlying the cyst after achieving full mobilization of the ovary by adhesiolysis. The incision is generally made along the inferior pole on the opposite side to the hilus in such a manner as to preserve the apposition of healthy ovarian tissue to the fimbria. The cyst contents are immediately drained with the suction cannula and the cavity is irrigated and inspected for papillary structures or other suspicious features. Very small endometriomas may be effectively treated by electrocoagulation of the mucosal lining. Because carbon dioxide laser is absorbed by fluid, complete ablation of the cyst wall with this energy source may be compromised in an environment rich in blood and hemosiderin. With larger endometriomas, the normal ovarian cortex is stabilized with atraumatic forceps, and the cyst wall is grasped with biopsy forceps and stripped from the bed of normal ovarian tissue (Fig. 25.9). Hydrodissection may facilitate separation of the tissue planes. Remaining fragments of the cyst wall should be vaporized with laser or fulgurated by electrocautery. Hemostasis can be achieved with bipolar cautery.

FIGURE 25.9. Laparoscopic ovarian cystectomy after fenestration of the cyst. A: The cut edges of the ovarian cortex and cyst wall are held and teased apart. B: The cyst wall can be stripped off by twisting it around the grasping forceps. Hydrodissection may be helpful. C: Large defects can be closed with laparoscopic suturing. Most incisions are left to heal by second intention.

View Figure

An alternative technique involves sharp and blunt dissection to remove the cyst in an intact state. Hydrodissection is particularly useful with this approach. The cyst contents are carefully drained in a plastic laparoscopy pouch to facilitate clean removal from the peritoneal cavity. The ovarian defect is usually left to heal spontaneously. Ischemia associated with suture placement can provoke adhesion formation after laparoscopic ovarian reconstruction. Low-power continuous carbon dioxide laser or bipolar coagulation can be applied to the inside wall of the redundant ovarian capsule to cause an inversion of the incised cortex. Most authors have reported excellent results with this no-suture technique. Rarely does the ovarian cortex need to be reapproximated with sutures. If fine, absorbable suture is used, the knot should be placed internally to minimize the possibility of it becoming a nidus for adhesion formation. Fayez and Vogel prospectively evaluated four laparoscopic methods for the treatment of endometriomas. Patients were treated postoperatively with danazol and underwent a second-look laparoscopy 8 weeks after their initial surgery. Complete excision with scissors successfully eliminated recurrence of the cysts, but adnexal adhesions had developed postoperatively in all cases. Mere incision and drainage of the cyst contents, followed by stripping or CO2 laser vaporization of the lining, resulted in adhesion formation in only 25% to 37% of cases, but endometrioma cysts recurred in 21% to 22%. Other authors have used the KTP laser to photocoagulate or remove the cyst lining of large endometriomas and have reported a very low rate of recurrence at 6 months after the procedure. In a prospective study by Beretta and colleagues, patients were randomly allocated at the time of laparoscopy to undergo either cystectomy or drainage of the endometrioma and bipolar coagulation of the inner lining. No preoperative or postoperative adjunctive medical therapies were administered. The excision technique resulted in a lower 24-month cumulative recurrence rate of dysmenorrhea, deep dyspareunia, and nonmenstrual pelvic pain. The median interval between the operation and the recurrence of moderate to severe pelvic pain was longer in the cystectomy group (19 months) versus the drainage and coagulation group (9.5 months). In addition, the 24-month cumulative pregnancy rate was statistically significantly higher in the former group than in the latter group (66.7% versus 23.5%, respectively). If there is evidence of functional destruction of the ovary or if the patient has chronic, incapacitating pelvic pain secondary to ovarian endometriosis and has completed her family, appropriate therapy may consist of oophorectomy. The infundibulopelvic and uteroovarian ligaments can be ligated with Roeder loop suture, bipolar coagulation, or surgical staples before excision of the structure. The ovary is retrieved by morcellation or by posterior colpotomy. This type of surgery must be performed carefully when adnexal adhesions are present to avoid ovarian remnant syndrome. Screening sonography and sometimes CA-125 measurements are recommended before endoscopic therapy of larger endometriomas or adnexal masses in women older than 40 years to identify those with increased risk of a neoplastic process that should be treated by laparotomy.

INTESTINES Intestinal involvement has been estimated to occur in 3% to 15% of women with endometriosis and in up to 50% of patients with severe disease. The most common areas of intestinal involvement are the rectum and rectosigmoid colon, followed by the sigmoid colon, cecum, terminal ileum, proximal colon, and appendix. The incidence of appendiceal endometriosis has been estimated at approximately 0.8% of all appendectomies; 3% to 5% of patients with endometriosis have appendiceal involvement. Symptoms that should arouse suspicion of colorectal involvement include constipation alternating with diarrhea, rectal pain, tenesmus, dyspareunia, and dysmenorrhea. Cyclic rectal bleeding is seen in as many as one third of females with rectosigmoid involvement, but the mucosa is rarely invaded. Small intestine disease accounts for up to 16% of gastrointestinal endometriosis and most often involves the terminal ileum. The most common symptom associated with disease in this location is midabdominal cramping pain. Ten percent of small bowel involvement presents with obstruction requiring surgery. The more common large bowel disease results in clinical obstruction in only 1% of cases. The differential diagnosis of intestinal endometriosis includes primary carcinoma, metastatic carcinoma, diverticulitis, inflammatory bowel disease, irritable bowel syndrome, pelvic inflammatory disease, radiation colitis, and ischemic stricture. Endometrial adenocarcinomas have been reported in the colon and rectum but are exceedingly rare in comparison to the relatively large numbers of patients with colorectal endometriosis. Preoperative or intraoperative rigid sigmoidoscopy may be helpful in ruling out primary colorectal malignancy. An intact mucosa effectively rules out primary colorectal malignancy. The greatest chance of diagnosing colorectal endometriosis occurs when the examination is performed at the time of menstruation. Although endometriosis rarely invades the intestinal mucosa, mucosal distortion is possible secondary to infiltration of the submucosa. Pelvic and rectal pain are the major symptoms that lead to colorectal resection in patients with advanced endometriosis. Bowel resection should be undertaken in the symptomatic patient or when there is a suspicion of malignancy; however, the frequency of such indications is small. In a series authored by Prystowsky and colleagues of 1,573 consecutive patients with endometriosis, only 11 women (0.7%) required bowel resection. Resection is usually undertaken for lesions producing partial obstruction because most of these lesions are fibrotic and unresponsive to hormonal manipulation. Recommended approaches for less extensive lesions include CO 2 laser vaporization of superficial serosal disease of the rectum or large intestine, excision without entering the mucosa, and oophorectomy or induction of hormonal menopause. Although oophorectomy can cause regression of the endometrial nodule, large implants of the bowel can scar and ultimately lead to obstruction. The use of electrocautery or fiber lasers should be avoided, because of their greater risk of causing transmural thermal damage. A full mechanical and antibiotic bowel preparation is carried out preoperatively. In cases of large lesions that encroach on the mucosa, full-thickness excision of involved bowel can be undertaken either by disk excision of small, isolated lesions or by segmental resection for larger lesions. The anastomosis can be hand sewn with a continuous single layer of absorbable monofilament suture or created with surgical staples; however, patients with cul-de-sac disease must be in the lithotomy position to allow transanal placement of the stapler. These procedures have been performed by or with the assistance of general surgeons. Appendectomy should be considered when there is physical evidence of peritonitis, when implants are large and active, when associated adhesive disease to adjacent bowel may result in partial or complete angulation and obstruction, or when the benign nature of the lesion is in doubt. Spontaneous perforation of the appendix due to endometriotic involvement is very rare. The technique of incidental endoscopic appendectomy is similar to that performed through laparotomy, although the stump need not be buried in the cecum. The tip of the appendix is grasped and elevated. The appendiceal vessels are bipolar cauterized or occluded with surgical clips near the base of the appendix before being excised. Two Endoloop ligatures are placed immediately next to each other at the base, and a third Endoloop is then secured approximately 5 mm distal to the first two. The appendix may then be transected between the second and third ligature and placed in a surgical pouch for safe retrieval from the abdominal cavity. Judicious application of bipolar cautery at the stump sterilizes the raw surface of the pedicle without causing damage to the adjacent cecum. Coronado and colleagues reported a complete relief of pelvic symptoms in 49% and an improvement in 39% of patients who underwent full-thickness resection of the colon; 39% of patients in the series achieved a term pregnancy. In a later series by the same colorectal surgeons of 130 patients who underwent aggressive, conservative surgical management for advanced disease, the operative

procedures performed included low anterior resection with anastomosis to the extraperitoneal rectum (n = 109), sigmoid resection (n = 10), disc excision of the rectum (n = 7), ileocecal resection (n = 2), and small bowel resection (n = 2). Twenty-four of 49 patients (49%) who attempted to conceive delivered a viable child. The sequelae of intestinal endometriosis may not appear until the patient is postmenopausal. Although the endometriosis can become inactive, the resulting cicatrization can lead to a decrease in the bowel lumen and to symptoms of obstruction.

URINARY TRACT Endometriosis involving the urinary tract is relatively rare. The spectrum of disease severity varies from incidental findings at laparoscopy, laparotomy, or cystoscopy to more significantly associated hematuria, flank pain, hypertension, and ureteral obstruction. Bladder and ureteral involvement represent 85% and 15% of cases, respectively. Cystoscopy and intravenous pyelography are helpful studies in documenting the extent of disease. Vesical endometriosis can be treated by hormonal suppressive therapy or partial cystectomy. These nodular lesions develop within the muscularis. Extrinsic ureteral compression by endometriosis presents four times more frequently than intrinsic involvement and is most likely to occur in the region of the ovarian fossa. Patients with paracervical and extensive uterosacral ligament disease are also at risk. The preferred treatment for ureteral obstruction is ureterolysis or resection of the involved segment followed by ureteroneocystostomy or ureteroureterostomy. Involvement of peritoneum overlying the ureter is amenable to resection by laparotomy or laparoscopy. An incision is made in normal peritoneum adjacent to the involved area. The inferior margin of the incision is grasped and deviated medially, and the ureter is separated from the peritoneum bluntly or by hydrodissection. The peritoneal lesion can be excised or vaporized. Periureteral vessels must remain intact to prevent ischemia and resultant fistula formation. If the peritoneum is adhesed and the lesion cannot be dissected, the ureter is likely involved in the disease process. Ureteroneocystostomy should be considered.

INCISIONAL SCARS Surgical scars are occasionally the sites of endometriotic implantation. Perineal, vaginal, and vulvar scars, particularly episiotomies, colporrhaphies, and Bartholin gland excisions, are likely areas for involvement by endometriosis. There is often a history of delayed wound healing of the incisional scar infiltrated with endometriosis. These implants typically appear as either deep-lying or subcutaneous nodules infiltrating the fascia and muscle. Bleeding into the tissues at the time of menstruation can cause cyclic local pain, tenderness, and discoloration; however, the nodule may lie too deep for detection of any color change through the skin. If the nodule is superficial, cyclic bleeding or ulceration may be apparent. In most instances, incisional endometriomas have followed surgical procedures that violated the uterine cavity and allowed the endometrium to be transplanted. Wespi and Kletzh?ndler suggested that the frequency might approach 5% among patients having Caesarean section or hysterectomy. Metroplasty and myomectomy also increase the risk of incisional endometriosis. Indeed, endometriosis has been reported along the needle tracts after amniocentesis or saline injection for abortion. Careful flushing and irrigation of the abdomen and of the incision during closure should minimize the chance of contamination when incision into the uterine cavity is required. Episiotomy scars and cervical and vaginal lacerations also serve as implantation sites after delivery. The chance is significantly increased when postpartum curettage is performed. Paull and Tedeschi reported 15 instances in 2,208 deliveries when curettage was carried out, and no instances in 13,800 deliveries without curettage. Management, usually best accomplished by local excision, is both diagnostic and curative. Various hormonal regimens may be appropriate if it is imperative to avoid surgery. However, malignancy can occur in each area of ectopic endometriosis, and histologic confirmation of the tentative diagnosis is recommended.

THORAX Sixty-five cases of thoracic endometriosis were reported by Foster and associates in 1981. Pleural and lung parenchymal disease presented with different clinical findings. Ninety-three percent of women with pleural disease developed pain with right-sided pneumothorax or pleural effusion. Because numerous right diaphragmatic defects were noted in patients with pleural involvement, pleural implants are believed to be secondary to tubal regurgitation and transport of endometrial tissue through the diaphragmatic defects. Other symptoms may include upper quadrant abdominal pain or referred pain to the shoulder. Disease

involving the lung parenchyma produced hemoptysis rather than the pleuritic symptoms. Previous pelvic surgery was more common among women who had parenchymal endometriosis; however, pelvic endometriosis was found more often in those with pleural disease. Catamenial pneumothorax or hemoptysis should alert the physician to the possibility of thoracic endometriosis. The chest roentgenogram is usually of little value in diagnosing this disease; however, cytology, aspiration biopsy, and pleuroscopy may be useful. Massive effusion and bleeding can occur, but this presentation is more commonly associated with a malignancy. GnRH agonist or surgical treatment may be effective in the symptomatic patient. Surgical pleural abrasion may be superior to hormonal treatment in the long-term management of pneumothorax.

Adjunctive Procedures of Conservative Surgery UTERINE SUSPENSION Uterine suspension techniques have been devised to prevent adhesion formation at denuded peritoneal surfaces of the posterior cul-de-sac, uterine serosa, and broad ligament. Elevation of the adnexa can prevent adhesion reformation of the ovary or fallopian tube at a site where existing adhesions have been excised. This procedure may be particularly useful in the case of a posterior or retroflexed uterus. It is indicated in selected cases of dyspareunia after resection of posterior cul-de-sac endometriosis. There is no evidence to suggest that uterine suspension is detrimental to subsequent pregnancies, although it is of unproven efficacy in enhancing fertility. The modified Gilliam procedure offers certain advantages over other uterine suspensions because of its maintenance of normal anatomic relations. Shortening the round ligament through the internal inguinal ring eliminates the opening that is made lateral to the point of the ligament's attachment to the abdominal wall in the Olshausen suspension procedure. When a modified Gilliam suspension is performed, the uterus is elevated and a 2-0 absorbable suture is placed around each round ligament about 3 to 4 cm from its insertion into the uterus (Fig. 25.10). The edge of the rectus fascia is grasped by a Kocher clamp at the level of the anterosuperior spine of the ileum. The adjacent peritoneal edge is grasped with a Kelly clamp. The rectus fascia is separated from the underlying musculature with blunt dissection. A long Kelly clamp is inserted between the fascia and muscle to the level of the inguinal ring, while displacing the peritoneum superiorly. This clamp is inserted through the ring and along the round ligament by gently opening and closing the instrument. The insertion is facilitated by placing traction on the suture to stabilize the round ligament. The peritoneum overlying the ligament is then incised at a point adjacent to the suture, and the suture is grasped by the Kelly clamp. By withdrawing the clamp, the round ligament is brought through the internal ring and outside of the peritoneal cavity; it can then be sutured to the rectus sheath with 2-0 interrupted delayed-absorbable sutures. These sutures must be placed through the round ligament without encircling the ligament and thus occluding its blood supply. This procedure is repeated on the opposite side. FIGURE 25.10. Modified Gilliam suspension. A: A chromic suture is placed around the round ligament about 3 to 4 cm from the uterine cornu. B: The rectus fascia is grasped with Kocher clamps and separated from the belly of the rectus muscle bluntly with the index finger or knife handle. C: The parietal peritoneum is grasped with Kelly forceps. A long Kelly forceps is introduced through the internal inguinal ring as it passes over the belly of the rectus. D: The Kelly clamp is brought through the internal inguinal ring and along the round ligament to a point adjacent to the chromic stay suture. A knife is used to open the peritoneum. The ends of the chromic suture are grasped by the Kelly clamp. E: As traction is applied to the suture, a knuckle of the round ligament passes through the internal ring. F: Three sutures of 2-0 delayed-absorbable or silk suture are placed, fixing the ligament to the rectus fascia in a manner that will not interrupt the blood supply. View Figure

At the end of the suspension, the surgeon's hands should be introduced into the abdomen to ascertain whether there is a loop of round ligament lateral to the point where the ligament has been withdrawn from the peritoneal cavity. If so, this should be corrected to prevent strangulation of the involved segment lying between the ligament and abdominal wall. In addition, the fallopian tube should be inspected to ensure that its course has not been disturbed. This can occur if the traction suture has been placed through a segment of round ligament too close to the uterus. Laparoscopic suspension is possible after placement of a trocar and sheath approximately 5 cm lateral to the midline and 3 cm above the inguinal ligament. The anterior rectus fascia in this site is tagged with suture. The round ligament is grasped at the usual site with laparoscopic forceps to elevate the ligament to the tagged anterior fascia, where it is sutured in place with nonabsorbable suture. The desired positioning of the uterus is confirmed laparoscopically.

PRESACRAL NEURECTOMY Presacral neurectomy, or division of the superior hypogastric plexus, is useful as an adjunctive procedure to eliminate the uterine component of dysmenorrhea that results from endometriosis. Sixty percent to 70% of patients with secondary dysmenorrhea experience complete relief of symptoms. There is no evidence that this procedure enhances fertility. A significantly greater relief of midline pelvic pain is achieved when endometriosis resection is combined with presacral neurectomy, compared with conservative resection alone. In a series by Tjaden and colleagues, all 17 patients undergoing presacral neurectomy noted a complete resolution of midline pelvic pain, and only two of these had a recurrence of pain within the 42-month follow-up period. Endometriosis rarely provokes exclusively midline pelvic pain, however, and lateralizing adnexal pain and deep dyspareunia are not affected by this procedure. Careful patient selection is necessary if the desired outcome is to be achieved. The hypogastric plexus consists of fine strands of nerves embedded in a delicate areolar tissue. The plexus is formed as a continuation of the aortic and inferior mesenteric plexuses and passes over the bifurcation of the aorta. It then continues below the promontory of the sacrum before dividing into the right and left inferior hypogastric nerves. The presacral neurectomy procedure can be performed through a transverse Maylard incision or longitudinal incision that adequately exposes the region of the bifurcation of the aorta (Fig. 25.11). At the time of laparotomy, the descending colon is packed superiorly and to the left to expose the left margin of the hypogastric plexus. The posterior peritoneum overlying the sacrum is elevated and incised with the scalpel. The incision is extended caudally with scissors for about 5 cm to the third or fourth sacral vertebra, and cranially to just below the bifurcation of the aorta. The margin of the posterior peritoneum can be drawn upward and outward by a stay suture or an Allis clamp. A Kitner sponge is then used to dissect the areolar tissue and associated nerve fibers off the posterior aspect of the peritoneal flap. The right ureter is readily visible and can be retracted laterally, and the areolar tissue is dissected from it without disturbing its blood supply. The common iliac artery, which lies just below the ureter, is freed superiorly from the adjacent tissue. A right-angle clamp or probe can be introduced medially next to the promontory to elevate the sheath and allow blunt dissection underneath it. Care must be taken to avoid the middle sacral vessels that may be left intact on the surface of the promontory. Injury to the middle sacral vein can result in significant blood loss. Hemorrhage is controlled with cautery, suture ligation, hot packs, hypogastric vessel ligation, use of an absorbable gelatin sponge (Gelfoam) or microfibular collagen (Avitene), or packing with bone wax.

FIGURE 25.11. Presacral neurectomy. A: Location of incision in relation to anatomic landmarks. A Maylard incision can also be used in some cases. The descending colon is displaced superiorly and to the left for good exposure of the left margin of the hypogastric plexus. B: A Kitner sponge is used to dissect the areolar tissue medially and off the posterior aspect of the peritoneal flap. The right ureter can be identified easily. C: The areolar nerve-bearing tissue is dissected from the peritoneum on the left side, exposing the left internal iliac vessels and superior hemorrhoidal vessels. D: The plexus is isolated and elevated off the sacral promontory. A segment of plexus about 5 cm in length is isolated with 2-0 silk sutures. E: The plexus is excised. Note relation between pedicles of the nerve bundle and adjacent structures. View Figure

The areolar tissue is taken off the left flap of peritoneum until the superior hemorrhoidal vessels are exposed. These vessels should remain on the peritoneum but are bluntly freed from the overlying tissue. By elevating the sheath, several vessels that feed into the left common iliac vein can be identified. These branches are isolated, clamped, and tied as they are visualized. When the plexus has been isolated, a Babcock clamp can be used to elevate the sheath. Two 2-0 absorbable or silk sutures are placed around the proximal and distal aspects of a 5-cm segment of the plexus and are loosely tied. The tonsil clamp is applied to each end of the nerve bundle. As the clamps are removed, the sutures are slipped down over the crushed areas and tied securely. The intervening portion of the plexus is then excised. The procedure is terminated by approximating the peritoneum with absorbable suture. In less than 10% of cases, the pelvic mesocolon is inserted in front of the interiliac trigone and the nerve bundle cannot be reached by simple incision of the peritoneum. In these cases, the chief branches of the inferior mesenteric artery must be moved to the left to expose the triangular space between the two common iliac arteries. Unless there is adequate exposure and meticulous dissection, incomplete resection of the superior hypogastric plexus can occur, resulting in suboptimal denervation. A laparoscopic approach to presacral neurectomy has been described, but its efficacy has yet to be conclusively established. This technique involves insertion of a 10-mm trocar sheath 3 cm above the symphysis pubis and placement of two accessory ports in each iliac fossa. Steep Trendelenburg position with a left lateral tilt is required to allow displacement of the intestines cephalad to expose the bifurcation of the aorta and sacral promontory. The parietal peritoneum overlying the sacral promontory is grasped and elevated to allow a transverse incision midway between the bifurcation of the aorta and the sacral promontory. The presacral nerve is isolated by developing the avascular space between the nerve and right internal iliac artery down to the periosteum. Segments of the superior hypogastric plexus are removed by sharp dissection after diathermy. The entire length of removed nerve plexus should not exceed 3 to 4 cm. Venous bleeding is controlled with bipolar cautery. Meticulous hemostasis must be ensured at the completion of the operation. This technique should only be performed by experienced laparoscopic surgeons because the vascular complications can be serious. In a review of 655 laparoscopic presacral neurectomy procedures, Chen and colleagues reported a 0.6% rate of major complications, including one case of injury of the right internal iliac artery and three cases of chylous ascites. Polan and DeCherney reported that the combination of presacral neurectomy and conservative surgery in women with chronic pelvic pain, endometriosis, and pelvic inflammatory disease increased total postoperative pain relief from 26% to 75%, although only a small number of patients were included in this laparotomy series. In 1986, Lee and co-workers performed presacral neurectomy in 50 women with chronic pelvic pain. Dysmenorrhea resolved in 73% of the cases, dyspareunia lessened in 77%, and acyclic pain improved in 63%. The uterosacral ligaments were resected in half of the subjects in this study, but this did not seem to affect the overall rate of pain relief. In a randomized clinical trial of women with

moderate to severe endometriosis and pelvic pain undergoing conservative surgical therapy, Candiani and colleagues reported a recurrence of midline menstrual pain in 23% of women who underwent presacral neurectomy versus a 42% recurrence in those who did not. This difference reached the limit of statistical significance (p = .06). In a recent uncontrolled laparoscopic study by Nezhat and associates of 100 women subjected to vaporization of endometriosis and presacral neurectomy, the symptoms of pelvic pain, dysmenorrhea, and dyspareunia were reduced by more than 50% in 74, 61, and 55 patients, respectively, over the 1-year follow-up period. The stage of endometriosis did not correlate with the degree of pain improvement achieved. Two common side effects of the presacral neurectomy procedure have been observed. Constipation may require laxatives or stool softeners for a period of 3 to 4 months. The vaginal dryness that develops in as many as 10% to 15% of patients is transient and usually resolves within 6 months. Difficulty with micturition is an infrequent complication that rarely lasts for more than 1 or 2 months. A painless first stage of labor has been reported in women who have undergone presacral neurectomy.

UTERINE NERVE ABLATION The technique of uterosacral neurectomy was initially described by Ruggi in 1899. Later popularized by Doyle, it has since been adapted for performance during laparoscopic procedures for the alleviation of dysmenorrhea. Sympathetic fibers T10 to L1 are contained within the inferior hypogastric plexus and course along the inferior vena cava and sacrum to enter the uterus through the nerves of the uterosacral ligaments and accompanying uterine arteries. The parasympathetic components of the paracervical nerves originate from S1 through S3 or S4, travel within the nervi erigentes, and emerge in the lateral pelvis to form the Frankenh?user ganglia lateral to the cervix. Division of the uterosacral ligaments at a point approximately 1.5 cm distal to the cervix should interrupt many sensory nerve fibers of the cervix and uterine corpus. In general, uterine nerve ablation by laser is preferable to electrocautery because it is less likely to cause undesirable thermal damage. The course of the ureters and adjacent vasculature should be noted before commencement of dissection. The uterosacral ligaments are exposed by manipulating the uterine cannula to anteflex the corpus and by applying pressure to the posterior cervix with an ancillary laparoscopic probe. The initial incision is made on the medial aspect of the ligament at its junction with the uterus. A second incision is made just lateral to the uterosacral ligament and medial to the ureter. The ligament is then grasped with forceps and stretched toward the sidewall. The CO 2 laser is used to vaporize a 2- to 5-cm area of each ligament to a depth of approximately 1 cm. This division should be centered approximately 1.5 cm distal to the cervix. The posterior aspect of the cervix between the insertion of the uterosacral ligaments may be superficially vaporized to interrupt the sensory fibers crossing to the contralateral side. Because extension of the beam too far laterally or posteriorly from the ligament can result in considerable bleeding, the surgeon should have immediate access to bipolar cautery, endocoagulation, or hemostatic clips. P.624 Fiber lasers such as the KTP/532 offer the advantages of increased hemostasis and lack of carbon plume, compared with the CO2 instrument. If bipolar diathermy is used to fulgurate the uterosacral ligament, laparoscopic scissors are used to excise the segment of ligament in question. Few published clinical trials exist that delineate the efficacy of the uterosacral neurectomy technique. Feste reported significant improvement in the symptoms of primary dysmenorrhea or dysmenorrhea associated with endometriosis in 71% of a series of 42 patients. In a similar series of 100 patients by Donnez, 50% experienced complete relief, 41% had mild to moderate relief, and 9% described no relief. Using the carbon dioxide laser, Davis observed a considerable improvement in dysmenorrhea in 135 of 146 women (92%) with endometriosis and an improvement in dyspareunia in 103 of 109 women (94%) with endometriosis who underwent uterine nerve ablation and vaporization of endometriosis. This therapeutic benefit did not seem to differ among revised AFS classification stages of endometriosis. Lichten and Bombard published a randomized, prospective, double-blind study of laparoscopic uterosacral nerve ablation for the treatment of severe or incapacitating dysmenorrhea unresponsive to oral contraceptives and nonsteroidal antiinflammatory agents. None of the control patients noted improvement, whereas 9 of 11 in the treated group had almost complete relief at 3 months, and 5 of 11 described complete relief from dysmenorrhea 1 year after surgery. Patients with endometriosis were not included in this small series. Uterine nerve ablation by laparotomy fell from favor before it was revived as an endoscopic technique. The

potential neurologic, intestinal, orthopedic, and psychological components of pain should be considered before subjecting the patient to a procedure that, although now performed endoscopically, carries some surgical risk, and whose effectiveness has been questioned because of the small number of cases evaluated. Surgical resection of pelvic endometrial implants may be all that is necessary to alleviate discomfort in most endometriosis patients. Vercellini and colleagues could not demonstrate the efficacy of this laparoscopic technique. Complications associated with transection of the uterosacral ligaments include ureteral damage, bowel damage, and postoperative hemorrhage, which, if undetected, may result in death. Uterine prolapse has recently been described as a potential long-term side effect of the procedure.

Second-Look Laparoscopy Second-look laparoscopy has been suggested as an appropriate procedure for additional lysis of pelvic adhesions in patients who have undergone a laparotomy or a laparoscopy for the resection of endometriosis. If scheduled 8 days to 6 weeks after the initial dissection, second-look laparoscopy allows separation of de novo adhesions that are still relatively filmy in consistency. In addition, laparoscopy after pelvic reconstructive surgery provides an opportunity to assess future prognosis for fertility. Early second-look laparoscopy after endoscopic treatment of endometriomas has revealed a recurrence rate of endometriomas of 15% to 20%. Equally significant are the nearly 20% incidence of de novo adhesion formation and the 40% to 82% recurrence rate of dense adhesions. Second-look laparoscopy allows treatment of these findings; however, there is little direct evidence that this secondary surgical procedure will increase the cumulative pregnancy rate.

Surgical Outcomes No classification schedule for endometriosis provides an accurate correlation between extent of disease and pregnancy rate. Nevertheless, point categorization through the revised AFS classification does provide a framework in which to report outcomes of therapy. The crude pregnancy rate after conservative surgery by laparotomy for mild endometriosis is 61%; this approximates the 58% rate derived from an accumulated series of patients with minimal and mild endometriosis treated by CO2 laser laparoscopy (Table 25.8). The results of laparoscopic electrocoagulation of all stages of disease are presented in Table 25.9. Murphy and colleagues, using life-table analysis and the two-parameter exponential method, studied 72 patients with stage I or stage II endometriosis treated by laparoscopic electrocoagulation of endometrial implants. They reported a crude pregnancy rate of 74% for stage I and 57% for stage II during an average follow-up period of 7.9 months; however, the monthly fecundity rates were only 10.3% for stage I and 7.6% for stage II endometriosis. TABLE 25.8. Pregnancy Rates After Laparoscopic CO2 Laser Vaporization of Endometriosisa

Number of Pregnancies/Number Treated (%)

Investigator

Stage I

Stage II

Stage III

Stage IV

Combined

Length of Follow-up (mo)

Daniell and Brown

—

—

—

—

3/10 (30)

5

Daniell and Pittaway

—

—

—

—

3/15 (20)

6

Kelly and Roberts

3/3 (100)

3/7 (43)

—

—

6/10 (60)

6

Chong et al.

21/32 (66)b

—

—

—

21/32 (66)

12

Feste

24/47 (51)

4/6 (66)

2/5 (40)

—

30/58 (52)

12

Martin

7/27 (26)

3/9 (16)

1/4

—

11/50 (22)

9

Martin

25/56 (45)d

22/45 (49)d

9/14 (64)d

—

56/115 (49)

12

Davise

20/31 (65)

—

15/26 (58)

2/7 (29)

37/64 (58)

15

Olive and Martin f

23/59 (39)

22/48 (46)

10/20 (50)

—

55/127 (43)

—

(25)c

Olive and Martin f

23/59 (39)

22/48 (46)

10/20 (50)

—

55/127 (43)

—

Donnezg

26/42 (62)

11/21 (52)

3/7 (43)

—

40/70 (57)

18

Paulsen and Asmerg

109/140 (78)

60/88 (68)

—

—

169/228 (74)

8–32

Gast et al.e

36/70 (51)

12/33 (38)

9/19 (47)

—

50/122 (41)

10

Fayez et al.c

27/38 (71)

33/44 (75)

—

—

60/82 (73)

12

Nezhat et al.g

28/39 (72)

60/86 (70)

45/67 (67)

35/51 (69)

168/243 (69)

—

TOTALS

329/553 (59)

230/397 (58)

94/162 (58)

37/58 (64)

690/1,170 (59)

—

a

The original American Fertility Society classification system is used unless noted.

b

Postoperative danazol for 132 days.

c

Postoperative danazol for 6 months.

d

Three-month course of danazol 6 to 18 months postoperatively if not pregnant.

e

Revised American Fertility Society classification system.

f

Patients treated with either laser laparoscopy only or a combination of laser laparoscopy and preoperative or postoperative danazol.

g Patients with factors other than endometriosis excluded from study.

Cook AS, Rock JA. The role of laparoscopy in the treatment of endometriosis. Fertil Steril 1991;55:663, with permission.

TABLE 25.9. Pregnancy Rates After Laparoscopic Electrocoagulation of Endometriosis

Number of Pregnancies/Number Treated (%)

Investigator

Minimal

Mild

Moderate

Severe

Combined

Length of Follow-up (mo)

Eward

4/7 (57)

10/18 (56)

—

—

14/25 (56)

13

Hasson

0/1 (0)

—

2/2 (100)

4/5 (80)

6/8 (75)

7

Sulewski et al.

—

20/42 (48)

20/58 (35)

—

40/100 (40)

37

Daniell and Pittaway

—

—

—

—

33/60 (55)

—

Reich and McGlynn

—

—

—

—

15/23 (65)

18

Seiler et al.

—

20/45 (44)

—

—

20/45 (44)

7

Nowroozi et al.

—

42/69 (61)

—

—

42/69 (61)

8

Murphy et al.

24/36 (67)

18/36 (50)

2/7 (29)

0/3 (0)

44/82 (54)

8

TOTALS

28/44 (64)

110/210 (52)

24/67 (36)

4/8 (50)

214/412 (52)

—

From Cook AS, Rock JA. The role of laparoscopy in the treatment of endometriosis. Fertil Steril 1991;55:663, with permission.

A recent retrospective study by Tulandi and Al-Took that compared reproductive outcome after treatment of mild endometriosis with laparoscopic excision and electrocoagulation showed no significant difference between the two modalities. The total pregnancy rate was 53.5% in the excision group and 57.1% in the electrosurgery group. The mean interval between surgery and conception was 10.7 months in the electrosurgery group and 13.3 months in the excision group. Excision of tissue may result in more complete removal of infiltrating endometriosis, which should be of particular benefit to patients with deep nodules. Expectant management of mild to moderate endometriosis after diagnosis by laparoscopy yields a crude pregnancy rate of about 50%, which has brought into question whether surgical therapy of lesser stages of disease enhances fertility. In a retrospective study comparing the efficacy of electrosurgical treatment of endometriosis with the efficacy of expectant management in minimal and mild endometriosis-associated infertility, Tulandi and Mouchawar reported that the cumulative probability of conception was significantly higher among patients treated surgically. In a more recent large prospective, multicenter, double-blind, controlled, randomized trial, resection or ablation of endometriosis during diagnostic laparoscopy resulted in a significantly higher fecundity rate after 36 weeks as compared with expectant management (Fig. 25.12). Electrosurgery or laser was used to treat implants and adhesions in this Canadian study by Marcoux and co-workers. FIGURE 25.12. Cumulative probability of a pregnancy carried beyond 20 weeks in the 36 weeks after laparoscopy in women with endometriosis, according to study group. (Marcoux S, Maheux R, Béribé S. Laparoscopic surgery in infertile women with minimal or mild endometriosis. N Engl J Med 1997;337:217.)

View Figure

The performance of a laparotomy to excise minimal or mild stages of endometriosis is not warranted. In a comparison between laparotomy and laparoscopy for treatment of these categories of disease, life-table analysis showed similar pregnancy outcomes. Operative treatment of more extensive disease does offer a greater likelihood of conception than does expectant management, in part because of correction of mechanical factors such as adhesions. The overall crude pregnancy rate reported by various studies of conservative laparotomy for endometriosis that

stratified reproductive results by disease severity was 38%, with a monthly fecundity rate averaging 1.4% to 1.5% (Table 25.10). Laparoscopic treatment of severe endometriosis offers a mean crude pregnancy rate of 47.6%, although data from only a few series have been published (Table 25.8). Hence, expert laser laparoscopists have reported results that appear to be as good as those obtained through the open abdomen, although there are no substantive data for direct comparison of outcomes of the two surgical modalities, and the correct identification and classification of disease may vary between laparotomy and laparoscopy groups. The apparent equivalence of pregnancy rates for all stages of endometriosis after laparoscopic laser vaporization is noteworthy. Life-table analysis demonstrated that pregnancy is most likely to occur during the first 36 months after surgery. Furthermore, the duration of infertility and, perhaps, patient age may have a greater impact on cumulative pregnancy rates than the actual stage (revised AFS stages I through IV) of the disease. TABLE 25.10. Conservative Surgery by Laparotomy for Severe Endometriosis: Effect on Conception According to the Literature

Investigators

Year

Procedures (N)

Pregnancies (N)

Incidence of Pregnancy (%)

Monthly Fecundity Rate

Acosta et al.

1973

39

13

33.3

—

Hammond et al.

1976

2

0

0.0

—

Sadigh et al.

1977

42

20

47.6

—

Garcia and David

1977

49

14

28.6

—

Schenken and Malinak

1978

21

6

28.6

—

Buttram

1979

68

32

47.1

—

Rock et al.

1981

81

39

48.1

0.015

Wheeler and Malinak

1981

119

36

30.3

—

Wheeler and Malinak

1981

119

36

30.3

—

Rantala et al.

1983

46

18

39.1

—

Chong and Baggish

1984

10

3

30.3

—

Gordts et al.

1984

57

20

35.1

—

Olive and Lee

1986

34

10

29.4

0.014

Candiani et al.

1986

8

3

37.5

0.012

Donnez et al.

1987

15

7

46.7

0.026

591

221

37.4

TOTALS

From Candiani GB, Vercellini P, Fedele L, et al. Conservative surgical treatment for severe endometriosis in infertile women: are we making progress? Obstet Gynecol Surv 1991;46:490, with permission.

In a prospective, randomized double-blind, controlled trial of laser laparoscopy in the treatment of pelvic pain associated with minimal to moderate endometriosis, Sutton and associates found that 62.5% of the laser-treated women reported symptom improvement at 6 months, as compared with 22.6% of those treated expectantly. Treatment results were poorest for patients with stage I disease and best for those with stage III disease. Symptom relief continued at 1 year in 90% of those who initially responded. Deeply infiltrative endometriosis is frequently present in stage I disease, and an inadequate depth of incision may contribute to the lack of surgical response in this group. Aggressive and complete excision of deep endometriosis is justified. Resection of deep endometriosis relieved dyspareunia in 40% and dysmenorrhea in 60% of cases. Nezhat and associates noted moderate to complete relief of pain in 162 of 175 women; however, some patients in this series had several surgical interventions. Preliminary analysis of the surgical results in 250 women in whom deep endometriosis had been excised with CO2 laser showed a cure rate of pelvic pain in 70% and a recurrence rate of less than 5% over a 5-year follow-up period. Spontaneous pregnancy rates as high as 60% within 1 year of excision of deep endometriosis have been reported. Rock and colleagues have shown that 13.5% of patients initially treated with conservative surgery required subsequent operative procedures. Wheeler and Malinak noted a cumulative recurrence rate at 3 and 5 years after conservative surgery of 13.5% and 40.3%, respectively. Neither the initial staging nor the ability to conceive after the initial surgery greatly affected the recurrence rates. In a study of patients who underwent laparoscopic cystectomy of ovarian endometriomas of greater than 3 cm in diameter, Busacca and colleagues reported a cumulative rate of ultrasonographic recurrence of 11.7% over 48 months. Laparoscopic excision of ovarian endometriomas by the stripping technique is associated with a lower

reoperation rate than that of fenestration. Redwine reported a cumulative recurrence rate of 19% at 5 years after conservative laparoscopic excision of endometriosis by sharp dissection. A second cytoreductive procedure may benefit some infertile women who have undergone surgery in the past, if assisted reproductive technologies are not pursued. A cumulative pregnancy rate of 31% at 25 months was achieved after a second conservative laparotomy for recurrent endometriosis.

Combination Medical and Surgical Treatment Preoperative and postoperative medical therapies have been proposed as treatment adjuncts to conservative resection of endometriosis to enhance fertility. Preoperative suppression of disease with hormonal agents may facilitate the surgical procedure by reducing tissue vascularity, and the greater ease in tissue dissection may decrease adhesion formation during the postoperative period. The preoperative hormonal agents also eliminate the corpus luteum that might otherwise be mistaken for an endometrioma. However, they may also reduce the size of endometriosis implants, making them less recognizable after short-term drug therapy. In a controlled clinical trial, a 3-month course of GnRH agonist treatment before laparoscopy for endometrioma excision failed to result in a reduction in operative time or recurrence rate of disease during a 1-year follow-up period. Initiation of postoperative medical therapy may inhibit the activity of any residual disease, suppress ovulation, and decrease the possibility of adverse effects of peritoneal spillage of disease at the time of resection. Postoperative medical therapy has a serious drawback, however; the patient is unable to attempt conception for several months. Wheeler and Malinak demonstrated an improved pregnancy rate in patients with severe endometriosis treated with danazol in the immediate postlaparotomy period compared with patients treated with surgery alone. Nevertheless, in Buttram's 1985 series, the conception rate for severe disease treated by conservative surgery followed by a 6-month course of danazol was only 32% (7/22), compared with a 40% rate for surgery alone in the historical control group. Andrews and Larsen have noted that the best chance for postsurgical conception occurs during the first 6 months after conservative surgery by laparotomy. Thus, suppressing ovulation during that critical period may be counterproductive. Preoperative danazol therapy has been shown to slightly improve fecundity rates when compared with rates achieved with surgery alone, although this result has been questioned because of the lack of treatment randomization. The addition of hormonal therapy has not been shown to be beneficial in less advanced stages of disease. Hence, the value of combination therapy in the treatment of endometriosis-associated infertility remains controversial. Because of the lack of substantive data to support its efficacy, a short postoperative course of danazol or GnRH analog of 3 months or less should be considered in infertile patients in cases of residual disease or peritoneal spillage of the contents of an endometrioma. Contemporary management of women with endometriosis-associated pelvic pain involves both surgical and long-term medical therapy. When reductive laparoscopy is followed by a 6-month course of GnRH analog, there is a significant delay in the return of endometriosis symptoms requiring further treatment. In a randomized, prospective study, Hornstein and colleagues found that this interval was more than 24 months in those receiving nafarelin versus 11.7 months in the placebo group. Similarly, postoperative administration of low-dose, cyclic oral contraceptives for 6 months delayed the recurrence of pain symptoms and endometriomas at 12 months, but no significant differences were detected at 24 months or 36 months following laparoscopic excision. A shorter duration of hormonal therapy during the postoperative period may be inadequate in reducing recurrence risk. A 3-month course of nafarelin following surgical therapy of stage III and IV endometriosis was ineffective in reducing pain scores as compared with placebo.

Hysterectomy The number and rate of hysterectomies performed for endometriosis increased steadily from the 1960s to the 1980s, more so than for other diagnoses. The reported rate for 1982 to 1984 was more than double the rate for 1965 to 1967, although the exact reasons for the increase remain uncertain. Endometriosis was the primary indication for 20% of white women and 9% of black women undergoing hysterectomy in the United States from 1988 to 1990. Because of concern over the risk of recurrence even after definitive surgical therapy, bilateral oophorectomy was performed at the time of hysterectomy in 52% of women 44 years of age or younger and in 81% of women 45 years of age or older. Definitive surgery offers prompt, complete, and long-term relief of pain from endometriosis more often than do the various available medical regimens. Most hysterectomies are performed by the abdominal route; in selected cases, laparoscopy might reveal a free cul-de-sac or allow lysis of complicating

adhesions, thus allowing safe vaginal hysterectomy. When the posterior cul-de-sac is obliterated and extensive fibrosis is present deep in the pelvis, subtotal hysterectomy may be indicated. The recurrence of cyclic pain associated with endometriosis after hysterectomy with preservation of normal ovaries has been estimated at 3% to 7%. Nevertheless, in a study of 138 women who underwent hysterectomy with the diagnosis of endometriosis at the Johns Hopkins Hospital, ovarian conservation was associated with a 6.1 times greater risk of development of recurrent pain and an 8.1 times greater risk of reoperation as compared with oophorectomy at the time of hysterectomy. Laparoscopic resection of invasive peritoneal and intestinal disease that persists after castration may result in an improvement in pain symptoms. Hysterectomy does not improve symptoms in 25% of cases of chronic pelvic pain when the uterus is believed to be the source of the pain. Minute, hormonally active ovarian fragments may be detected in women with symptomatic endometriosis, even after total abdominal hysterectomy and bilateral salpingo-oophorectomy. This ovarian remnant syndrome is the result of incomplete excision of cortical tissue at the time of extirpative surgery for endometriosis or pelvic inflammatory disease. Most ovarian remnants are retroperitoneal in location, and they are often densely adherent to pelvic side wall structures, including the ureter, hypogastric vessels, and bladder base. Complete surgical removal may be difficult. Estrogen replacement therapy after total hysterectomy and bilateral oophorectomy is associated with less than a 10% rate of recurrence of endometriosis. A cause-and-effect relation between estrogen replacement and malignancy in endometriosis has not been established, suggesting that progestational agents need not be prescribed together with estrogens after hysterectomy for a diagnosis of endometriosis. However, it may be wise to administer both progestin and estrogen if the disease was incompletely resected or deeply invasive, contained atypical epithelial changes, or is recurrent. Women who begin estrogen replacement therapy immediately after total abdominal hysterectomy and bilateral salpingo-oophorectomy are at no greater risk of recurrent pain than those who delay estrogen therapy for more than 6 weeks postoperatively.

ENDOMETRIOSIS AND ASSISTED REPRODUCTIVE TECHNOLOGIES Part of "Chapter 25 - Endometriosis" If spontaneous conception is not achieved within 3 years of surgical resection of endometriosis or within 1 year of repair of tubal obstruction associated with endometriosis, the odds are poor that it ever will occur. Techniques in assisted reproduction have been widely used during the past decade for the management of endometriosis-associated infertility unresponsive to cytoreductive surgical or hormonal therapy. Endometriosis is the sole identifiable cause of infertility in 25% to 35% of women undergoing in vitro fertilization/embryo transfer (IVF/ET). The responses to gonadotropic stimulation, the numbers of preovulatory oocytes, the fertilization and cleavage rates, and the clinical pregnancy rates associated with endometriosis have been equivalent to rates associated with tubal disease and unexplained infertility. Loh and colleagues found that the ovarian follicular response to gonadotropins after laparoscopic cystectomy for endometriotic cyst was equivalent to that of normal ovaries. Moreover, the stage of disease does not appear to influence clinical outcome. However, one study has reported a significantly higher miscarriage rate with moderate or severe endometriosis; this is perhaps due to poor embryo quality or autoimmune phenomena. The necessity of initial medical or surgical therapy before use of assisted reproductive technologies remains controversial. In a 1986 report by Wardle and co-workers, oocyte fertilization rates were markedly reduced in women with untreated endometriosis as compared with those with tubal infertility and endometriosis who were treated with danazol for 6 to 9 months before the IVF treatment cycle. Dicker and associates noted that 35 women with severe endometriosis who underwent 6 months of ovarian suppression with a GnRH analog had a higher clinical pregnancy rate per cycle and per transfer than did 32 women who received ovarian stimulation for IVF without prior GnRH treatment (per cycle, 25% versus 3.9%; per transfer, 33% versus 5.3%). Recent reports suggested that success with IVF was lower in women with endometriomas and that the spontaneous abortion rate may be higher. Hence, resolution or surgical therapy of the endometriomas before the IVF treatment cycle may enhance the ongoing pregnancy rate. Reports of infected endometriotic cysts secondary to oocyte aspiration for IVF also give support to surgical correction before commencing techniques of assisted reproduction. Gamete intrafallopian transfer (GIFT) may overcome impairment of sperm transport to the fallopian tube, failed ovum capture, or abnormalities in the peritoneal environment associated with endometriosis, although the presence of any anatomic disorders of the fallopian tubes has negative prognostic significance for a successful outcome for this procedure. Hulme and colleagues performed GIFT on 46 infertile patients with minimal to moderately active endometriosis not previously treated by medical or surgical methods. The only prerequisite was patency of at least one fallopian tube. The pregnancy rate per GIFT cycle was 30.5% (18/59), which compared with a clinical pregnancy rate of 25.8% for all patients undergoing the procedure at their unit. One or more endometriomas were aspirated from the ovaries at the time of follicle aspiration in 11 patients; 4 of the 11 achieved live births with GIFT. Nevertheless, Guzick and associates, in a case-control study, found that pelvic endometriosis significantly impaired the efficacy of GIFT. Of 114 laparoscopic oocyte retrievals performed in the endometriosis group, there were 37 pregnancies (32.5%) and 25 deliveries (23.7%); of the 214 retrievals in the control group, there were 101 pregnancies (47.2%) and 76 deliveries (35.5%). Controlled ovarian hyperstimulation (COH) with human menopausal gonadotropins or pure FSH together with intrauterine insemination (IUI) has been proposed as a method to increase cycle fecundity of patients with endometriosis, although few series have been published to date. By increasing the number of oocytes released at the time of ovulation and introducing a high concentration of spermatozoa into the female reproductive tract, the chance for conception is improved merely because of the larger number of gametes available for fertilization. In addition, subtle abnormalities of folliculogenesis, corpus luteum function, tubal motility, or sperm function may be corrected with this therapy. Cycle fecundity rates associated with COH/IUI therapy in patients with endometriosis-associated infertility have ranged from 9% to 13%, although these series did not include a nontreatment control group. One recent prospective randomized study found a higher pregnancy rate with COH/IUI following at least 6 weeks of GnRH agonist suppression in patients with advanced stages of endometriosis. One report has suggested that controlled ovarian hyperstimulation does not enhance fecundity in women with minimal endometriosis undergoing timed IUI. In addition, Fedele and associates reported that superovulation with timed intercourse was not associated with a better cumulative pregnancy rate than expectant management in infertile women with endometriosis stages I and II, although the cycle fecundity rate was improved. However, a more recent randomized, controlled trial of controlled ovarian hyperstimulation (COH) and IUI for infertility associated with stage I and II endometriosis demonstrated a live birth rate of 11% in the treatment group and 2% in the control group. Hence, the clinical history must

be carefully weighed when planning a sequence of therapy for the infertile patient with endometriosis.

ADENOMYOSIS Part of "Chapter 25 - Endometriosis" The disease called adenomyosis is defined as heterotopic endometrial glands and stroma located deep within the myometrium. Adenomyosis can be categorized as diffuse or local in its distribution. Diffuse adenomyosis can be relatively localized but is never encapsulated (Fig. 25.13). The uterus itself is usually mildly enlarged, rarely to more than twice-normal size, and is generally symmetric. Cut sections of the myometrium reveal a coarse trabecular pattern of interlacing musculature and fibrous tissue with small islands of endometrium that are often dark and hemorrhagic. Localized, encapsulated disease of the uterine wall is termed adenomyoma, to distinguish this manifestation of adenomyosis from the more usual diffuse pattern. An adenomyoma is always located mainly within the wall of the uterus but may project into the uterine cavity to become further known as a submucous adenomyoma. This encapsulated, submucous form of adenomyosis disease resembles the leiomyoma. FIGURE 25.13. Uterus showing three types of adenomyomatous growth: encapsulated intramural adenomyoma, submucous adenomyoma, and diffuse adenomyosis of walls.

View Figure

The most widely accepted theory of the origin of adenomyosis is that endometrial tissue within the myometrium is of müllerian origin. Its presence in this location is the result of a direct, downward extension of the endometrium of the uterine cavity. Serial sectioning of tissue has revealed a direct continuity between the basalis portion of the endometrium and the endometrial islands within the areas of adenomyosis. Endometrial extensions sometimes are present through the full thickness of the myometrium to the serosal surface of the uterus. Occasionally, only subserosal adenomyosis is seen. Subserosal adenomyosis is often associated with pelvic endometriosis and may cause the lymphatic spread of endometrial fragments. The intramural islands generally have the same histologic appearance as the basalis of the endometrium (Fig. 25.14) and often respond to estrogen stimulation by demonstrating a proliferative or, occasionally, cystic hyperplastic pattern. Cellular atypia is rare. The effect of progestational agents on the ectopic endometrium is less predictable. Secretory changes in the glands are uncommon except in pregnancy, when a decidual reaction of the stroma is anticipated. Unlike endometriosis, adenomyotic lesions are not characterized by a pronounced hemorrhagic tendency or inflammatory response. In the absence of hormonal stimulation, adenomyosis becomes atrophic. Adenocarcinomas involving adenomyosis are characterized by a history of prior exogenous estrogen use, by low histologic grades, and by an excellent prognosis. Adenomyosis can be definitively diagnosed only through histologic sections of myometrium. The reported incidence of the disease varies widely among institutions, from 8% to 62%, depending on the criteria used for diagnosis and on the thoroughness with which the excised uterine tissue is studied. By tradition, a histologic diagnosis is made when endometrial glands and stroma are found at least 1 low-power field beneath the endomyometrial junction (greater than 4 mm). A more rigid criterion suggested by Benson and Sneeden requires that ectopic endometrium extend into the myometrium P.630

at least 2 low-power fields (8 mm) from the basalis. FIGURE 25.14. Area of adenomyosis. Compact stroma and proliferative, slightly hyperplastic glands surrounded by hypertrophied myometrium.

View Figure

The incidence of adenomyosis peaks in the fifth decade. Infertility is not common, although most patients are multiparous. About 12% have coexisting external endometriosis. Adenomyosis is often discovered incidentally in patients undergoing surgery for uterine leiomyomas.

Symptoms Adenomyosis is often an incidental pathologic finding and may be entirely asymptomatic. Dysmenorrhea is more likely to be reported when glandular invasion exceeds 80% or more of the myometrium. Pain can be severe, cramping, or knifelike. The pattern of dysmenorrhea is likely associated with bleeding episodes within the deep-lying islands of endometrium. Menorrhagia can be a consequence of the increased surface area of the enlarged uterine cavity. In addition, extensive involvement of the myometrium can interfere with the normal contractility of the uterine musculature and can lead to excessive bleeding. Nevertheless, data collected from 1,851 hysterectomies for the prospective, multicenter Collaborative Review of Sterilization study indicate that adenomyosis occurs as often in asymptomatic uteri removed for prolapse (19%) as in uteri removed for excessive bleeding (22%) or pain symptoms (15%).

Pelvic Findings The uterus may be very firm to palpation and is usually enlarged to not more than twice its normal size. As it is classically described, the adenomyotic enlargement occurs in the anteroposterior dimension, a reflection of the more prominent involvement of the posterior uterine wall. In the more common diffuse type of adenomyosis, the uterus is a symmetrically enlarged, globular structure. Encapsulated adenomyomata may cause the uterus to be irregular or asymmetric, much as it is when leiomyomata are present. At times, particularly during menstruation, the enlarged uterus is tender on examination.

Diagnosis Adenomyosis should always be suspected in a woman with dysmenorrhea and menorrhagia of increasing severity her fourth or fifth decade, particularly if the uterus is symmetrically enlarged, firm, and tender. An exact preoperative diagnosis is often difficult to establish because dysfunctional uterine bleeding and multiple small leiomyomas can present in a similar fashion. Gambone and colleagues reported that a presumptive diagnosis of adenomyosis was verified in only 38% of hysterectomy specimens. The diagnosis can be histologically established before hysterectomy only in the rare case in which excessive myometrium is removed during curettage or a polypoid submucous adenomyoma is excised. However, hysteroscopic myometrial biopsy of the posterior uterine wall with use of a 5-mm loop electrode has been shown to effectively establish the diagnosis in women with menorrhagia. Hysterosalpingography of the adenomyomatous uterus with water-based media can occasionally

demonstrate multiple spiculations or tuft defects leading from the uterine cavity to the myometrial wall; however, similar findings can occur in cases of vascular or lymphatic extravasation. Magnetic resonance imaging has proved to be highly accurate for distinguishing adenomyosis from leiomyomata; on T2-weighted images, adenomyosis appears as an ill-defined, relatively homogeneous, low-signal-intensity area embedded with sparse, high-intensity spots. The optimal junctional zone thickness value for establishing the diagnosis of adenomyosis is 12 mm or more. Recent studies have also suggested an important role for transvaginal ultrasound in distinguishing adenomyosis from leiomyomata. By using the diagnostic criterion of the presence of unencapsulated, heterogeneous, myometrial areas within round anechoic areas 1 to 3 mm in diameter, Fedele and colleagues noted a sensitivity of 80%, a specificity of 74%, a negative predictive value of 81%, and a positive predictive value of 73%. Nevertheless, when transvaginal sonography and magnetic resonance imaging have been prospectively compared, the latter has been significantly more accurate in correctly establishing the diagnosis. P.631

Treatment Hormone receptor studies have documented the presence of steroid receptors in adenomyotic foci. Estrogen receptors are more consistently present than are progesterone receptors, which are completely absent in 40% of cases evaluated. Progestins or cyclic estrogen-progestin combination preparations offer little aid in treatment, although one recent report indicated that adenomyosis-associated menorrhagia may be controlled with the insertion of a levonorgestrel-releasing intrauterine device. GnRH agonist therapy for 6 months resulted in the disappearance of pain symptoms and a decline in uterine volume in 65% of cases of biopsy-proven adenomyosis, but the dysmenorrhea and menorrhagia recurred at the end of treatment. Nevertheless, extended, intermittent use of these agonists can effectively relieve pain symptoms while having the significant advantage of preserving fertility between treatments. Curettage does not aid in establishing the diagnosis of adenomyosis and is ineffective as treatment, although it may be required because of abnormal bleeding. The need for surgery, therefore, is based on continued menorrhagia and dysmenorrhea rather than on an estimation of uterine size or even the known presence of adenomyosis or leiomyomata. The definitive treatment for abnormal bleeding caused by adenomyosis is hysterectomy. The vaginal route is preferred if the size of the uterus is appropriate and no other pelvic abnormalities are present. Under certain circumstances, as with a younger patient who wishes to retain her reproductive capability, excision of an encapsulated adenomyoma should be considered instead of hysterectomy. Such situations arise infrequently because adenomyosis is generally diffuse and usually occurs in multiparous women who are no longer interested in childbearing. The precise efficacy of hysteroscopic endometrial resection, laparoscopic myometrial reduction, and myometrial excision as conservative surgical procedures for adenomyosis has yet to be proved. Endometrial ablation is ineffective as treatment for deep, subserosal adenomyosis. Hysterectomy should be considered when preoperative ultrasonography, magnetic resonance imaging, or myometrial biopsy demonstrates deep adenomyosis.

Chapter 26

Surgery for Benign Disease of the Ovary Joseph S. Sanfilippo John A. Rock Significant progress has been made with regard to ovarian reconstruction for benign disease. It has been established that the ovaries and fallopian tubes are sensitive to ischemia from the trauma of surgery; secondary adhesions may develop, and the normal anatomic relationship between fallopian tubes, ovaries, and uterus may be altered. Knowledge regarding anatomy and embryology of the ovaries and other reproductive organs complemented by mastery of the principles and skills of microsurgery are the prerequisites for excellent results following ovarian reconstructive surgery. Embryology and anatomy are addressed in this chapter with emphasis on the importance of the anatomic relations of the ovary to other pelvic organs in the section on the evaluation and management of the adnexal mass. State of the art surgical procedures and techniques devised for the reconstruction of the ovary and for full restoration of normal pelvic anatomy are presented in the context of specific pathology or other abnormal conditions that require surgical intervention. This chapter also focuses on pediatric and adolescent surgical procedures that are performed when ovarian pathology is identified. EMBRYOLOGY GENE EXPRESSION FEMALE FETAL DEVELOPMENT ANATOMY ADNEXAL MASS OVARIOSCOPY ADNEXAL MASS DURING PREGNANCY ULTRASOUND TUMOR MARKERS LAPAROSCOPIC MANAGEMENT OF AN OVARIAN MASS OVARIAN REMNANT SYNDROME RESIDUAL OVARY ADNEXAL TORSION SURGERY OF THE OVARIAN SURFACE RECONSTRUCTION OF THE OVARY RESECTION OF BENIGN CYSTS OVARIAN SURGERY FOR POLYCYSTIC OVARIAN DISEASE PARADOXICAL OOPHORECTOMY OVARIAN TRANSPOSITION BEFORE RADIOTHERAPY LAPAROSCOPIC OVARIAN SURGERY IN THE PEDIATRIC OR ADOLESCENT PATIENT

EMBRYOLOGY Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" Early in gestation, at about 4 to 5 weeks, two gonadal ridges arise in the developing embryo as thickening on the medial aspect of the coelomic cavity adjacent to the mesonephros. These gonadal outgrowths are composed of coelomic epithelium and underlying mesenchyme projecting into the future peritoneal cavity. The epithelial and mesenchymal cells of the gonadal primordia are of mesodermal origin (large, spherical ovoid germ cells that originate extragonadally in the wall of the yolk sac and migrate to the developing gonads). The gonads of the two sexes remain morphologically indistinguishable until the 6th week of gestation. The presumptive ovaries remain undifferentiated until the onset of meiosis at the end of the first trimester. The ovarian cortex is a single germinal epithelium. The tunica albuginea lies beneath the cortex and is composed of connective tissue. The stroma is made up of fibroblasts, smooth muscle, endothelium and interstitial cells, including undifferentiated theca cells and corpora albicans. Sexual differentiation requires initiation by various genes along with a single gene determinant on the Y chromosome (TDF, testis-determining factor), which is necessary for testicular differentiation. In XX individuals (in the absence of a Y chromosome), the bipotential gonad develops into an ovary. The mechanisms responsible for gonadal sex differentiation are largely unknown. Investigators have theorized the presence of a testicular determining factory (H-Y cell-surface antigen on the short arm of the Y chromosome) that is elaborated by a specific gene. Meiosis-inducing and preventing substances, both of which are produced by cells derived from mesonephric structures adjacent to the gonad, are the agents of regulation of ovarian and testicular germ cell differentiation. The balance between these two substances varies between the two sexes and at different stages of development. The meiosis-inducing substance predominates in the fetal ovary. Maternal ovarian hormone production is not required for differentiation of the germ cells or, apparently, for later development of the fetal reproductive tract. Various ultrastructural studies have shown no specific changes in fetal granulosa cells that can be definitely associated with steroid hormone secretion such as is identified in the fetal Leydig cells. Thecal cells play an essential role in steroid synthesis in the adult ovary, but they do not appear until later in gestation and even then retain a relatively undifferentiated appearance. Fetal pituitary gonadotropin production begins as early as 10 weeks' gestation and reaches peak levels at midgestation. Gonadotropins have a major influence on follicular development in the adult ovary, but evidence for a similar function in the fetus is lacking.

GENE EXPRESSION Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" Specific follicular cell receptors bind growth factors, which are locally synthesized with the ultimate effect of intracellular signaling and protein kinase activation. This activity affects transcription of targeted genes. Gene expression is involved in follicle development, ovulation, and corpus luteum and corpus albicans formation. Trans-cription factors include protooncogenes, C-myc, and CCAAT/enhancer binding protein.

FEMALE FETAL DEVELOPMENT Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" During the early prefollicular stage, the ovarian surface cortex is characterized by germ cells and granulosa cells organized in cords and sheets, but the cortex lacks specific conformation. The last distinctive change to occur in the fetal ovary is the onset of meiosis, at the 11th or 12th week of gestation. Meiosis is preceded by differentiation of primitive germ cells into actively dividing mitotic cells called oogonia. The mitotic divisions of the oogonia are associated with complete separation at telophase, leaving the daughter cells connected by intracellular bridges. After a series of mitotic divisions, there is progressive entry into meiosis by cell groups, beginning in the innermost cortex and gradually extending to the periphery. These cells passing through the various stages of the first meiotic prophase are then designated oocytes. By late gestation, all surviving oocytes have advanced to the diplotene stage. Further differentiation of the oocytes is arrested at this stage and does not resume until ovulation begins at menarche, about 12 years later.

Follicular formation begins at 18 to 20 weeks' gestation and continues throughout the remaining weeks of fetal development. All the surviving oocytes are surrounded by adjacent granulosa cells; oocyte and follicular growth are well established by the late fetal and early neonatal period. The constant degeneration and loss of oocytes before their incorporation into the follicles reduces their numbers to only 1 to 2 million (follicles) in the newborn ovary.

ANATOMY Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" The ovary is almond-shaped and is about half the size of a testis. The dimensions of the adult ovary vary from individual to individual but average 3 to 5 cm long, 2 to 3 cm wide, and 1 to 2 cm thick, with a weight of 3 to 8 g. The surface of the ovary is pinkish gray to white. The ovary normally is smooth in childhood, but its surface becomes pitted from follicular maturation and atresia, and the surface of the adult ovary can be markedly wrinkled. The size, shape, and position of the ovary in the pelvis are somewhat variable, and both the consistency and the follicular changes taking place within the ovary vary with the stage of the menstrual cycle. The ovary typically is anchored to the side wall of the pelvis in the shallow peritoneal fossa of Waldeyer formed between the angle of proximity of the ovary to the ureter. This knowledge is important before dissecting the ovary off the pelvic side wall. The ovary is connected to the uterus by the uteroovarian ligament, to the posterior aspect of the broad ligament by the mesovarium ligament, and to the lateral pelvic sidewall by the infundibulopelvic ligament (Fig. 26.1). The mesovarium ligament attaches to the mesentery of the ovary. The other two ligaments are attached at the hilum of the ovary. FIGURE 26.1. Normal anatomy of the ovary. A: Anatomic relations of the uterus, tube, and ovary. B: The infundibulum of the oviduct extends onto the ovary and is attached at its most distal pole (ovarian fimbria). The mesovarian is the mesentery of the ovary. Each ovary is attached at the hilum.

View Figure

The ovary, like the testis, migrates downward from high in the abdomen during embryonic life. The infundibulum of the fallopian tube extends onto the ovary and is attached to it at its most distal pole by a structure called the fimbriaovarica. The relation of the ovary to the fimbria ovarica and to the uteroovarian ligament is crucial, and they should be carefully maintained during ovarian reconstruction. During embryogenesis, the ovary may assume an unusual appearance (i.e., it may be septate) or assume an unusual position (Fig. 26.2). An accessory ovary (Fig. 26.2A) contains ovarian tissue and usually is close to or is connected to a normally placed ovary. An accessory ovary also may be attached to the broad, uteroovarian or infundibulopelvic ligaments. Unlike the accessory ovary, a supernumerary ovary (Fig. 26.2B) must have an independent embryologic origin. It may develop from a primordium such as arrested migrating gonadocytes. A supernumerary ovary consists of typical ovarian tissue but has no direct or ligamentous connection with a normally placed ovary. A supernumerary ovary is thus a true third ovary that has independent function and is located at some distance from a normally placed ovary. Ovarian malposition (Fig. 26.2C) also may occur when the ovary fails to descend into the pelvis to assume its normal location. In ovarian malposition, the ovary is attached as it should be to the uterus by the uteroovarian ligament and to the fallopian tube by the fimbria ovarica, but it may lie adjacent to the liver or spleen. The ovary is elongated and may measure up to 15 cm in length. The fallopian tube attaching to such a malpositioned ovary may be 20 to 26 cm long, almost twice its normal length.

FIGURE 26.2. Ovarian anomalies. A: Accessory. B: Supernumerary. C: Malpositioned.

View Figure

The normal ovary has a surface covering composed of a single layer of flattened, germinal epithelial cells. This layer is contiguous at the ovarian hilum, with the peritoneal epithelium of the posterior leaf of the broad ligament. Beneath the germinal epithelium is a second, strong layer of condensed ovarian stroma that forms a fibrous capsule called the tunica albuginea. The area through which the vessels and nerves enter and exit is called the hilum of the ovary. Immediately around the hilum and extending into the substance of the ovary is an area known as the medulla, which is covered by the cortex. The medulla is composed of fibrous tissue unlike the condensed stroma of the ovarian cortex. The medulla contains no follicles; it has only blood vessels and the remnants of the tubular structure that would have developed into a testis (i.e., the rete ovarii) had the fetus been male. The ovarian artery arises on each side of the abdominal aorta just below the renal arteries. The artery descends from the aorta and crosses the ureter obliquely to enter the infundibulopelvic ligament on its course to the ovary. When it reaches the broad ligament, the ovarian artery branches also to the fallopian tube and ovary before it finally anastomoses directly with the uterine artery to form a continuous arcade in the broad ligament. The ovarian veins are situated mainly in the mesosalpinx, where they give rise to the pampiniform plexus. At the outer end of the broad ligament, this plexus coalesces to form a single, large ovarian vein. The ovarian vein accompanies the ovarian artery to terminate in the inferior vena cava on the right and the renal vein on the left. The lymphatic vessels of the ovary drain in three directions. The main group accompanies the ovarian vessels in the infundibulopelvic ligament and eventually reaches the periaortic nodes in the vicinity of the kidney. Other lymphatic channels communicate with channels of the opposite ovary by crossing the fundus of the uterus through the ovarian ligament. Some channels drain through the ovarian and round ligaments into the superficial inguinal lymph nodes in the groin. The ovary is supplied by both motor and sensory parasympathetic and sympathetic nerves, which accompany the ovarian vessels from the abdomen as they pass into the infundibulopelvic ligament to reach the hilum of the ovary. The segmented nerves supply the ovary from T-10 and T-11.

ADNEXAL MASS Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" The uterine adnexa (gynecologic origin) consist of the ovaries, the fallopian tubes, and the uterine ligaments. Although adnexal pathology often involves one of these structures, contiguous tissue of nongynecologic origin also may be involved. The bimanual examination is the most practical method of screening for an adnexal mass. When such a mass is found, its initial characteristics should be carefully described so that any subsequent change can be appreciated and the nature of the mass can be better ascertained. The description should include location, size (in centimeters), consistency, shape, mobility, tenderness, bilaterality, and associated findings (e.g., fever, ascites). Adjunctive diagnostic techniques such as sonography, magnetic resonance imaging (MRI), and computed tomography (CT) may help delineate the nature of adnexal enlargement. Pelvic ultrasonography is an accurate means of determining the location, size, extent, and consistency of pelvic masses and is also useful for detecting obstructive uropathy, ascites, and metastasis. Other, more specialized diagnostic procedures also may be necessary for the evaluation of an adnexal mass (Table 26.1). TABLE 26.1. Special Diagnostic Procedures for the Evaluation of an Adnexal Mass

Nonoperative Noninvasive

Abdominal and pelvic radiography

Barium enema

Excretory urography

Gastrointestinal series with small bowel follow-through

Computed tomography scan

Magnetic resonance imaging

?-hCG

CA-125

Nonoperative Invasive

Culdocentesis

Culdocentesis

Pelvic arteriography

Operative Noninvasive

Abdominal and pelvic examination under anesthesia

Operative Invasive

Culdoscopy

Laparoscopy

Exploratory posterior colpotomy

Exploratory laparotomy

CT can detect and precisely measure pelvic masses with a diameter of 2 cm or more. It has been particularly useful in gynecologic oncology because it helps define the extent of paracervical and parametrial involvement and allows a reasonable determination of the resectability of malignant neoplasms. MRI has surpassed CT in the precision of measurement of adnexal masses. MRI also allows a clear definition of the relationship of adjacent organs.

OVARIOSCOPY Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" Evaluation and management has taken on a new parameter–ovarioscopy. This surgical technique provides endocystic visualization within the ovary. The surgeon can evaluate the cyst wall and any anatomic abnormality related to its blood supply. Ovarioscopy has proven to be one other means of further enhancing diagnostic acumen to distinguish benign from frankly malignant as well as borderline lesions. The procedure was performed in a series of 68 patients with unilateral or bilateral adnexal masses without clinical or sonographic evidence of suspected malignancy. Tumor markers [cancer antigen 125 (CA-125), carcinoembryonic antigen, and alpha fetoprotein] were obtained preoperatively. Intraoperative endocystic ovarioscopic evaluation and ovarioscopic-directed biopsy were performed before laparoscopy. In comparison with tumor markers and transvaginal ultrasonography, ovarioscopy proved to have the greatest specificity for detecting benign ovarian cysts. The positive predictive value of the technique was 50% in comparison to 5% for tumor markers and 6% for transvaginal ultrasound. Although the data are preliminary, ovarioscopy appears to be one possible additional step in distinguishing benign from malignant ovarian masses. Although every adnexal mass requires individual evaluation and management, it is possible to make a number of useful general recommendations. Expectant management is justified only when an asymptomatic, physiologic cyst is suspected. Most cysts greater than 6 cm in diameter require a thorough evaluation. Imaging techniques are invaluable for characterizing the nature of the adnexal enlargement, but these procedures do not replace a careful medical history and thorough physical and pelvic examination. In a study conducted by Timmerman and co-workers, assessment was made of the use of both ultrasound and circulating levels of CA-125 antigen. Multivariate logistic regression analysis algorithms were used to distinguish benign adnexal masses from a malignant process. Transvaginal ultrasonography with color Doppler imaging was recorded in the 191 patients evaluated, ages 18 to 93 years. Of interest, 26.7% of the cohort of patients studied had malignant tumors. The authors believed that regression analysis could be used to accurately discriminate malignant from benign adnexal masses preoperatively. An intriguing aspect of ultrasound assessment is the prediction of malignancy in adnexal masses using an artificial neural network. Taylor and colleagues reported generating a neural network algorithm that enables computing of a probability of malignancy score for preoperative discrimination between malignant and benign adnexal masses. A retrospective analysis that included training in artificial neural network assessing transvaginal B-mode ultrasonography and color Doppler imaging was determined. The variables that were put into the artificial neural network included age, menopausal status, maximum diameter of the neoplasm, tumor volume, and papillary projections. The results identified four primary variables that were most effective in distinguishing benign versus malignant processes. These variables included age, time-average maximum velocity, papillary projection score, and maximum tumor diameter. The authors concluded that artificial neural networks are a useful clinical parameter to distinguish benign from malignant ovarian masses. Surgery ultimately may be necessary to determine the nature of the adnexal mass. Laparoscopy may be useful to exclude benign ovarian or nonovarian neoplasms. Indications for visualization of an adnexal mass with laparoscopy or exploratory laparotomy include the following: Ovarian mass greater than 6 cm in diameter Adnexal mass greater than 10 cm in diameter Any mass first developing after menopause Failure to discover the nature of the mass (e.g., leiomyoma) with radiologic or sonographic imaging techniques One of the major goals of the evaluation of the adnexal mass is to rule out malignancy. There is an age-dependent risk for a malignant adnexal mass. The incidence of malignant neoplasm increases significantly after age 50 years. Increased size of the adnexal mass is associated with an increased risk of malignancy. Granberg and colleagues found that less than 1% of masses smaller than 5 cm were malignant, less than 11% of masses 5 to 10 cm were malignant, and 72% of masses larger than 10 cm were malignant. Sassone and associates, in an evaluation of women of all ages (mean age, 41 years) by transvaginal sonography, found that 3% of masses smaller than 5 cm and 7% of masses 5 to 10 cm were malignant; the incidence of malignancy for masses larger than 10 cm was 13%. Endometriosis is a common cause of an adnexal mass. An endometrial cyst of the ovary may develop into an endometrioma. Leakage of blood from the cyst may cause peritoneal irritation, pelvic adhesions, and

pelvic organ fixation. Tuboovarian inflammatory complex usually is the result of incompletely treated or unresolved subacute, chronic pelvic inflammatory disease (PID) in the walled-off area surrounding the pelvic structure. A hydrosalpinx may be unilateral or bilateral and is primarily a sequela of acute or chronic PID. A hydrosalpinx usually is asymptomatic; however, it may be associated with chronic pelvic pain, dyspareunia, and a sense of pelvic pressure. Ectopic pregnancy may be the cause of a potentially serious, nonneoplastic, nonovarian mass. The physician should have a high index of suspicion in the case of any patient with irregular bleeding, pain, and an adnexal mass. More than 50% of women with tubal pregnancy have no palpable adnexal mass, and an adnexal mass is an unusual finding if the pregnancy is early. Uterine leiomyomata cause nodularity and consequent irregular conformation of the uterus. The uterus may become enlarged and may present as an abdominal mass. The inability to distinguish a leiomyoma from an ovarian tumor on pelvic examination is an indication for further diagnostic evaluation. Adnexal enlargement may be the result of carcinoma of the rectum, appendix, or bladder. Patients present with a variety of symptoms according to the organ involved. A complete and thorough evaluation is necessary to fully delineate the cause of a neoplasm. A barium enema before surgery for women older than 40 years of age with a left adnexal mass is recommended to address the possibility of cancer of the rectosigmoid. An adnexal mass may be noted in cases of acute abdomen. The differential diagnosis should include adnexal torsion, ruptured hemorrhagic cyst, degenerating leiomyomata, ectopic pregnancy, unruptured tuboovarian abscess, acute appendicitis with or without abscess formation, and diverticular disease of the sigmoid colon. A careful history, pelvic examination, and appropriate imaging studies often allow a prompt diagnosis.

ADNEXAL MASS DURING PREGNANCY Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" The incidence of adnexal mass in pregnancy requiring surgical intervention has been reported to occur in 1 in 81 to 2,500 pregnancies. When an adnexal mass is noted incidentally on ultrasound during pregnancy, the majority of small, simple cysts do not pose a risk to the pregnancy. Furthermore, most large or sonographically complex masses spontaneously resolve as reported by Bernhard and colleagues. This study evaluated 18,391 ultrasound studies done in an obstetric population for which 432 women were identified with an adnexal mass. The rate of incident of adnexal masses was 2.3% in the pregnant population evaluated. In addition, the rate of torsion of the adnexal mass was 1% and the rate of malignancy was also reported as 1%. Before operative intervention, a complete assessment of the fetus, including ultrasound to rule out a lethal anomaly and to document cardiac activity, is in order. The optimal time for elective surgery is during the second trimester. The patient should be informed of the increased risk of preterm labor and delivery. The patient should be placed in the left lateral tilt position to avoid inferior vena cava compression and associated uteroplacental insufficiency. Postoperatively, the fetus should be placed on continuous fetal heart rate monitoring. The most effective approach in management of adnexal masses during pregnancy remains a point of controversy (i.e., laparoscopy versus laparotomy). In a series of 88 pregnant women who underwent 93 surgical procedures for suspected adnexal pathology, laparoscopy was performed during the first trimester in 39 patients. The remaining 54 patients underwent laparotomy, 25 during the first trimester and 29 during the second trimester. Neither intraoperative nor postoperative internal complications were reported in the series. Five of 39 women undergoing the first trimester surgery had a spontaneous abortion. During the first trimester, a Veress needle was used for insufflation and the procedure was in essence conducted in a manner virtually identical to that in the nonpregnant state (i.e., closed laparoscopy). It was concluded that laparoscopic gynecologic surgery is safe during pregnancy when conducted in the first trimester.

ULTRASOUND Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" Ultrasound is also useful in predicting malignancy (Table 26.2). Many authors have identified characteristic features of benign versus malignant neoplasms. Collated data from studies of ultrasound accuracy in the prediction of malignancy have an average positive predictive value of 74% and an average sensitivity of 88% (Table 26.3). TABLE 26.2. Ultrasound Characteristics of the Ovary

Benign Pattern

Simple cyst without internal echoes

Simple cyst with scattered echoes

Polycystic echoes

Polycystic echoes with thick septum

Sessile or polypoid smooth mural echoes

Central dense round echoes

Thin or thick multiple linear echoes

Thin or thick multiple linear echoes with dense part

Malignant Pattern

Cystic echoes with papillary or indented mural part

Polycystic echoes with irregular thick septum and solid part

Solid pattern (>50%) heterogeneous component with irregular cystic part

Completely solid with homogeneous component

Low impedance to flow (color Doppler)

TABLE 26.3. Ultrasound Accuracy in Prediction of Malignancy

Author

Patients (n)

Malignancy Prevalence

Positive Predictive Value (%)

Negative Predictive Value (%)

Sensitivity

Specificity

Kobayashi et al. 1976

406

15

31

93

71

73

Meire et al. 1978

51

35

83

91

83

91

Pussell, 1980

25

48

83

91

83

84

Herrmann et al. 1987

241

21

75

95

82

93

Finkler et al. 1988

102

36

88

81

62

95

Benacerraf et al. 1990

100

30

72

91

80

87

Granberg et al. 1989

180

21.5

74

95

82

92

Sassone

143

10

87

100

100

83

Sassone et al. 1991

143

10

87

100

100

83

Weiner and co-workers have used transvaginal color flow imaging before exploratory surgery to study the impedance to blood flow in women with an adnexal mass. Intramural blood vessels consistently demonstrated low impedance to flow with a pulsatility index less than 1:16 in women with malignant tumors. The sensitivity and specificity of the preoperative pulsatility index in detecting malignant ovarian tumors were 94% and 97%, respectively. Kurjak and colleagues found that vessels with a low resistance index near the center of the mass or within papules or septa were highly correlated with malignancy. Therefore, transvaginal color flow imaging may be a useful clinical tool in the preoperative evaluation of ovarian masses. Doppler resistance index has been used as a “vascular” scoring system. Color Doppler ultrasonography appears to be a reliable method in presurgically evaluating ovarian neoplasms. Transvaginal color Doppler sonography has identified the following parameters as useful in determining malignant versus benign ovarian masses. The parameters include the number of vessels detected in each tumor, tumor vessel location (central versus peripheral), peak systolic velocity, lowest resistance index, mean resistance index, lower pulsatility index, and mean pulsatility index. Color Doppler signals were detected in 100% of malignant masses and 75% of benign masses, with the difference being statistically significant as reported by Alcazar and associates. Tumor vessel location appears to be central in virtually all malignant masses. Overall the receiver operating characteristic curves generated can be used to predict malignant processes. The lowest resistance index was associated with the majority of malignant tumors. Three-dimensional ultrasonographic technology has been used to evaluate adnexal masses. Images are dissected in XYZ planes and can be focused especially on areas suggestive of malignancy. Three-dimensional ultrasonography facilitates real-time analysis of acquired image data and allows reassessment of the findings at the time of the original ultrasound. Three-dimensional transvaginal ultrasonographic technology appears to enhance and facilitate morphologic assessment of benign as well as malignant ovarian masses.

TUMOR MARKERS Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" Tumor markers are substances that are identified in higher than normal amounts in blood, urine, or body tissues of patients with specific malignancies. The tumor marker is produced by the tumor per se or as a response to the presence of cancer. They are not unique to malignant processes and can be elevated with benign conditions. Tumor markers are not elevated in every patient with malignancy, especially in the early stages of the disease. Many (tumor markers) are not specific for a particular type of cancer; therefore, there are limitations to the use of tumor markers. CA-125 is a tumor-associated antigen to an antibody expressed by about 80% of patients with epithelial ovarian cancer. It can be increased by nongynecologic malignancies with involvement of the pleura or peritoneum and by benign conditions that result in ascites. Because of the many medical diagnoses that give false-positive CA-125 results, CA-125 cannot be used for general population screening for ovarian cancer in either premenopausal or postmenopausal women. However, in menopausal women who present with a pelvic mass, CA-125 can help differentiate benign from malignant masses. Because menopausal women have fewer gynecologic diseases that give false elevation of CA-125, the test is more sensitive and specific in this age group. Several authors have demonstrated that a panel of assays can improve both sensitivity and specificity in the detection of ovarian malignancies. For example, Soper and associates demonstrated 100% specificity and predictive value for CA-125 with TAG 72 or CA-15-3. Table 26.4 provides specific markers and their clinical application. TABLE 26.4. Tumor Markers-Adnexal Masses

Marker

Comments

CA-125

80% nonmucinous ovarian carcinomas have elevation of CA-125. Decreasing levels generally indicate response to therapy. Used to identify recurrences.

CEA

Primary use is to monitor recurrence of colon cancer. Oncofetal antigen-Ag complex glycoprotein, 20,000 d associated with plasma membrane of tumor cells. Increased with ovarian cancer and with melanoma, breast, pancreatic, stomach, cervical, bladder, kidney, thyroid, and liver cancer. Inflammatory bowel disease and smoking elevates CEA.

cMyc

Amplified in 30%–50% of ovarian tumors. The protein is simultaneously overexpressed.

CMycRA

Associated with aneuploidy in ovarian malignant cell progression.

BRCA-1

Associated with mutations of breast tumor related antigen. BRCA-1 tumor suppressor gene has been identified; 63% risk of developing ovarian cancer with positive BRCA-1 gene.

LAPAROSCOPIC MANAGEMENT OF AN OVARIAN MASS Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" The pelvic (adnexal) mass may be of gynecologic or nongynecologic origin (Table 26.5). Specific clinical findings are helpful to differentiate a malignant from a benign neoplasm (Table 26.6). It is important to establish whether the mass is of ovarian origin and to understand that a mass causing an ovary to enlarge to greater than 6 cm in diameter should be considered potentially malignant until proved otherwise. The most common ovarian mass is the physiologic ovarian cyst, which is caused by failure of a follicle to rupture or to regress. Physiologic ovarian cysts normally are less than 6 cm in diameter, smooth, mobile, and slightly tender to palpation. They usually contain straw-colored fluid and may be associated with menstrual irregularity. Physiologic ovarian cysts smaller than 6 cm usually regress by absorption of the fluid or spontaneous rupture. The premenopausal patient may be managed conservatively over two menstrual cycles. If regression fails to occur over two periods of observation or if enlargement is noted, reassessment is indicated. TABLE 26.5. Classification of the Adnexal Mass

Gynecologic Origin

Nongynecologic Origin

Nonneoplastic

Nonneoplastic

Ovarian

Appendiceal abscess

Physiologic cysts

Diverticulosis

Follicular

Adhesions of bowel and omentum

Corpus luteum

Theca-lutein cyst

Peritoneal cyst

Luteoma of pregnancy

Feces in rectosigmoid

Polycystic ovaries

Urine in bladder

Inflammatory cysts

Pelvic kidney

Urachal cyst

Nonovarian

Ectopic Pregnancy

Anterior sacral meningocele

Neoplastic

Congenital anomalies

Carcinoma

Embryologic remnants

Sigmoid

Tubal

Cecum

Pyosalpinx

Hydrosalpinx

Bladder

Appendix

Retroperitoneal neoplasm

Presacral teratoma

Neoplastic

Ovarian

Nonovarian

Leiomyomata

Paraovarian cyst

Endometrial carcinoma

Tubal carcinoma

Adapted from Hall DJ, Hurt WG. The adnexal mass. J Fam Pract 1982;14:135, with permission.

Adapted from Hall DJ, Hurt WG. The adnexal mass. J Fam Pract 1982;14:135, with permission.

TABLE 26.6. Clinical Findings Suggesting Benign or Malignant Adnexal Mass

Benign

Malignant

Unilateral

Bilateral

Cystic

Solid

Mobile

Fixed

Smooth

Irregular

No ascites

Ascites

Slow growth

Rapid growth

Young patient

Older patient

Oral contraceptives have been suggested as an alternative treatment for functional cysts. The combination-type oral contraceptives send negative feedback to the pituitary gland to decrease gonadotropin stimulation of the ovary, which causes regression of the cyst. Steinkampf and colleagues noted that the rate of disappearance of functional ovarian cysts was not affected by estrogen-progestin treatment; nevertheless, a patient taking oral contraceptives with an adnexal mass should be thoroughly investigated. Failure of the corpus luteum to regress (in the nonpregnant patient) may cause development of a corpus luteum cyst. The size of the corpus luteum cyst varies according to the amount of blood contained within the cyst. A large corpus luteum may rupture and cause intraperitoneal hemorrhage. Amenorrhea or irregular uterine bleeding may accompany the development of a corpus luteum cyst. A sensitive pregnancy test, ultrasonography, and laparoscopy can be used to differentiate an ectopic pregnancy from a persistent corpus luteum. A theca-lutein cyst, which may be associated with gestational trophoblastic disease or pregnancy, is the result of luteinization of the ovary by human chorionic gonadotropin (hCG). Many of these cysts are bilateral and multicystic. A reduction in hCG levels usually leads to their spontaneous regression.

Polycystic ovarian disease is associated with bilaterally enlarged ovaries with a smooth surface. The ovaries contain multiple follicular cysts; many patients are obese and hirsute and have accompanying anovulation. The clinical findings listed in Table 26.3, Table 26.4, Table 26.5 and Table 26.6 are often helpful in differentiating a malignant from a benign neoplasm. All ovarian neoplasms larger than 6 cm in diameter or with a solid component should undergo investigation. The postmenopausal ovary is usually small and nonpalpable. Enlargement of the postmenopausal ovary requires immediate investigation. Symptoms of ovarian neoplasms usually depend on their size, rate of growth, and position in the pelvis or abdomen. Symptoms may include vague lower abdominal fullness or pressure discomfort. Larger masses rise out of the true pelvis and may cause abdominal enlargement with varicosities and edema of the lower extremities. Most ovarian neoplasms are asymptomatic until they enlarge or involve adjacent organs and structures. Congenital anomalies of the müllerian system and vestigial remnants of the wolffian system are of gynecologic, if not strictly ovarian, origin. Müllerian anomalies should be considered in the differential diagnosis of an adnexal mass. Uterine anomalies usually are associated with cyclic pain from development of hematometra, whereas an enlarged paraovarian cyst may be asymptomatic.

OVARIAN REMNANT SYNDROME Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" The ovarian remnant occurs in patients who have had previous oophorectomy with or without hysterectomy. The symptomatic patient may present with or without a palpable mass or with a palpable pelvic mass but no symptoms. Pathologic investigation confirms the presence of ovarian tissue when there should be none. The ovarian remnant syndrome differs from the residual ovarian syndrome in that, with the latter, the ovary is purposely saved and a pathologic process subsequently develops in the ovary. The ovarian remnant syndrome follows oophorectomy. Minke and associates demonstrated that devascularization of ovarian tissue can occur with reimplanting on intact or abraded peritoneal surfaces, where it may resume endocrine function. Thus, the authors suggest that great care should be exercised to remove all ovarian tissue, particularly when oophorectomy is performed through the laparoscope. Ultrasonography remains a valuable tool in establishing the diagnosis of ovarian remnant syndrome. The use of both transabdominal sonography and transvaginal sonography with use of color Doppler identification of the mass acquire information with respect to both arterial and venous flow. This facilitates identification of ovarian tissue. With respect to diagnosis, use of gonadotropin-releasing hormone (GnRH) agonist stimulation test has specific utility. The patient often presents with chronic pelvic pain with or without a pelvic mass. Use of GnRH agonist stimulation test allows identification of the presence of functioning ovarian tissue in association with ovarian remnant syndrome. Associated chronic pelvic pain frequently responds to suppressive therapy. Initially, the gonadotropin flare results in increased production of estradiol and allows confirmation of the diagnosis. As treatment is continued, the GnRH agonist often proves efficacious in relief of pelvic pain. Symmonds and Petit identified three major factors that may complicate the initial surgery and make it difficult or impossible for the surgeon to ascertain whether all ovarian tissue has been removed: increased pelvic vascularity, which renders hemostasis difficult; adhesions, which distort the anatomy and make dissection difficult; and neoplasms, which also distort the anatomy. The most common preexisting disease is endometriosis, followed in frequency by PID. Patients with ovarian remnant syndrome often present with both pelvic pain and a mass. The quality of the pain varies, often cyclically, and ranges from a sensation of pressure or dull aching to a severe stabbing pain. The clinical diagnosis of ovarian remnant syndrome can be difficult. A finding of premenopausal levels of follicle-stimulating hormone (FSH) may facilitate the diagnosis. Sonography (especially vaginal) may be of some value, and a CT scan or MRI may be useful for defining the physical relation of the ovarian remnant to surrounding structures. The treatment of choice is adequate excision of the ovarian remnant with removal of contiguous adherent tissue such as pelvic peritoneum, bowel serosa, the underlying involved ligament, and alveolar and vascular tissues (Fig. 26.3). Excision of ovarian tissue may require a retroperitoneal dissection to define the relation of the ureter to the bowel and ovary. Special care should be taken to carefully define all anatomic relations before extirpation of the remnant.

FIGURE 26.3. A: Not infrequently, the ovarian remnant may adhere to the bowel and the pelvic side wall peritoneum. B: The ureter must be visualized and its relation to the bowel and ovarian remnant established. This may require development of the pararectal and rectovaginal spaces.

View Figure

Laparoscopic excision of ovarian remnant ovaries is feasible. A laparoscopic retroperitoneal approach that allows dissection of the course of the ureters with coagulation and dissection of the infundibulopelvic ligament and the uterine vessels can be accomplished. Surgeons with appropriate laparoscopic skills can consider this surgical approach. There is potential for ureteral injury as well as cystotomy and bowel injury.

RESIDUAL OVARY Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" Based on the clinical circumstance, the gynecologic surgeon should consider the value of ovarian conservation at the time of hysterectomy for benign disease. Some authors have noted the incidence of malignant neoplasm in retained ovaries as a reason for prophylactic oophorectomy, and others have noted the presence of “residual ovary syndrome,” characterized by either recurrent pelvic pain or a persistent pelvic mass (Fig. 26.4). However, Funt followed up 992 patients after conservation of one or both ovaries at the time of hysterectomy and reported that none developed ovarian malignancy and only 1.4% required subsequent surgical intervention for adnexal pathology. The benefits of preserved ovarian function thus appear to substantially outweigh the risk of subsequent ovarian pathology requiring further surgery. Before surgery, the gynecologic surgeon should discuss the various risks and benefits of castration and should encourage the patient to participate in any decision concerning the fate of her ovaries. FIGURE 26.4. Abdominal sonogram showing a residual ovary with presumed follicular activity.

View Figure

GnRH agonists have been used to assess response of residual ovaries with chronic pelvic pain followed by surgical intervention to remove the residual ovarian tissue. Resolution of pelvic pain in six treated patients occurred with the analog (GnRH agonist) and persisted with surgical extirpation of the ovarian tissue. Suppression of ovarian function by GnRH agonists allows differentiation of pelvic pain caused by residual ovary from other sources and thus should be a prerequisite to surgical intervention. In a retrospective report of 20 years' experience with residual ovary syndrome in which 2,561 hysterectomies were performed, the incidence of residual ovary syndrome was 2.85%. Thus, 1 in 35 women who undergo hysterectomy become symptomatic, that is, they experience pelvic pain often with the presence of a benign cyst. Patients should be counseled preoperatively with respect to the potential for residual ovary syndrome when the initial surgical intervention is anticipated. In addition to chronic pelvic pain, a pelvic mass and dyspareunia include the “cluster of symptoms” that can occur in patients who have undergone previous hysterectomy.

ADNEXAL TORSION Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" Torsion of the adnexa is an infrequent cause of pain in the lower abdomen. However, torsion is a common gynecologic surgical emergency, with a prevalence of 2.7%. Treatment of adnexal torsion is considered an emergency because peritonitis and death can result. Any portion of the adnexa (tube or ovary) may undergo torsion. It may occur in neoplastic ovaries or as a consequence of hyperstimulation. The clinical findings of torsion are usually nonspecific. For this reason, delays in diagnosis and surgical intervention may be substantial. The classic presentation is the acute onset of abdominal pain with clinical evidence of peritonitis and an adnexal mass. However, according to Bayer and Wiskind, the presenting findings in most patients are nonspecific and unimpressive. Torsion is more likely to occur during ovulation or as a premenstrual event associated with increased pelvic congestion; the authors found no correlation between the phase of the cycle and the onset of the symptoms. Historically, the adnexa usually were removed because some authors suggested that untwisting the adnexa could increase the risk of thromboembolism and infection. There is growing evidence that unwinding the involved adnexa to observe for tissue reperfusion and viability is safe. Nevertheless, a significant delay in surgical intervention may result in irreversible necrosis requiring removal of the tube, ovary, or both. The laparoscopic management of adnexal torsion has been increasing in efficacy. Mage and colleagues found that unwinding the adnexa was possible in most patients in their series and no further intervention was required. Likewise, Shalev and Peleg demonstrated that laparoscopic detorsion of the adnexa is safe and reliable as a primary treatment of this condition. Thus, the weight of evidence warrants conservation of the adnexa, if there is evidence of reperfusion and if significant delay has not resulted in irreversible tissue necrosis. In most instances, detorsion may be accomplished through the laparoscope.

SURGERY OF THE OVARIAN SURFACE Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" Surgery to remove adhesions or endometriosis from the ovarian surface is not unusual. De novo adhesions or adhesions between the medial surface of the ovary and the broad ligament may be filmy and vascular (Fig. 26.5A), and may be excised by fine electrocautery or vaporized with the use of a laser (Fig. 26.5B). More extensive adhesions that completely cover the ovarian surface may be thick and avascular (Fig. 26.5C and Fig. 26.5D). The plane of dissection between the broad ligament or pelvic side wall and the adherent ovarian surface must be developed with care so as not to remove or damage the peritoneum while excising the adhesion (Fig. 26.5D). FIGURE 26.5. Ovarian adhesions. A: Filmy adhesions between the medial aspect of the ovary and the pelvic side wall. B: These may be removed with laser or fine electrocautery with use of a quartz or glass rod, respectively, as a backstop. C: The ovary may be enveloped by adhesions. D: Care should be taken to tent up the adhesions so that the peritoneum is not damaged or incised.

View Figure

The removal of multiple, small adhesions distributed over the ovarian surface once coagulated can be gently removed from the ovary without trauma to the ovarian cortex. Alternatively, such adhesions can be coagulated by monopolar cautery and then removed. If the lateral aspect of the ovary is densely adherent to the broad ligament, it may be necessary to dissect the ovary free. Some cases require that a large area of the side wall or the broad ligament be denuded; reperitonealization can be accomplished with 7-0 fine, nonreactive suture material. Small endometrial implants can be fulgurated or vaporized. The resulting small ovarian defect usually does not require closure. Care should be taken to ensure that the endometriosis is superficial and that the implant is not actually the tip of a large endometrioma within the substance of the ovary.

RECONSTRUCTION OF THE OVARY Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" Before ovarian reconstruction is begun, it is important that proper mobilization of the ovary be accomplished for reestablishment of normal anatomic relations. Once complete excision of the involved ovarian pathology has been accomplished, the main objectives of ovarian reconstruction are atraumatic closure of the stroma and cortex and prevention of adhesion formation. The principles for accomplishing this goal include gentle tissue handling, hemostasis, the use of fine (and ideally minimally reactive) suture material, and an effort to “bury knots” for the prevention of adhesions. The restoration of normal-appearing anatomy is the most logical approach for creating maximal ovarian surface, which facilitates ovum pickup by the fallopian tube. Controversy continues as to whether it is more appropriate to completely excise or lyse paraovarian and peritubal adhesions. The most important aspect of ovarian reconstruction is reapproximation of the cortex with the use of atraumatic techniques, including fine absorbable suture material. Large amounts of intraabdominal lavage should be used, ideally with a physiologic substance such as lactated Ringer's solution. Every effort should be made to remove all blood from the peritoneal cavity, preferably with the patient taken out of the Trendelenburg position. The approach to resection of an ovarian cyst should be planned so as to minimize adhesion formation. The incidence of de novo adhesion formation appears to be decreased when the initial approach is through laparoscopy rather than laparotomy. The Operative Laparoscopy Study Group assessed the issue of frequency and severity of adhesion reformation and of de novo adhesions after operative laparoscopy. In a multicenter collaborative approach that included early second-look intervention, 68 patients underwent operative laparoscopic procedures, including adhesiolysis as well as ovarian cystectomy. The scoring of adhesions noted during the second-look laparoscopy occurred at nine sites (each ovary, each fallopian tube, omentum, cul-de-sac, pelvic side wall, and large and small bowel). The study concluded that adhesion reformation is a frequent occurrence and that de novo adhesion formation occurred less frequently after initial operative laparoscopy. A number of agents have been advocated for preventing adhesions, including oxidized regenerated cellulose [Interceed (TC7), Johnson & Johnson Medical, Arlington, TX], which is an absorbable barrier that promotes reepithelialization of the affected area. Pagidas and Tulandi compared Interceed with Ringer's lactate solution for adhesion prevention. Ringer's lactate solution was as effective as Interceed in decreasing adhesion formation. Haney and colleagues compared oxidized regenerated cellulose with expanded polytetrafluoroethylene (Gore-Tex surgical membrane). The results indicated that expanded polytetrafluoroethylene was associated with fewer postsurgical adhesions. Other agents include sodium hyaluronate carboxymethyl cellulose (Seprafilm).

Functional Ovarian Cysts Physiologic cyst enlargement of the ovary may occur as a sequela of failure of either follicular rupture or corpus luteum regression. The latter is termed Halban's syndrome. The former has been associated with luteinized unruptured follicle syndrome in which “intraovarian ovulation” is thought to occur; this is a diagnosis usually established with ultrasound. In general, functional ovarian cysts regress spontaneously; however, they may persist and become symptomatic, reaching dimensions as large as 10 cm in diameter. The obvious and most feasible approach is observation, because most such cysts are self-limited. The cyst, however, may prove to be a source of continued pelvic pain or may adhere to the posterior broad ligament, producing persistent symptoms. The potential for adnexal torsion always exists with an ovarian cyst.

RESECTION OF BENIGN CYSTS Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" Surgical intervention often is initiated with a laparoscopic approach, which permits completion of fenestration of the nonneoplastic ovarian cyst. The cyst lining is stripped from the remaining “normal ovary,” and ovarian reconstruction takes place. In several series of laparoscopic management of ovarian cysts, a simple follicular or luteal cyst was identified in most patients evaluated for pelvic pain. In a series by Kleppinger, 31 of 64 ovarian cysts were noted to fall into this category.

Surgical Techniques

Laparotomy An elliptic incision is made through the thin ovarian cortex of a benign cyst (Fig. 26.6). The end of the knife handle is then inserted and a plane developed over the cyst wall. Alternatively, fine-needle electrocautery can be used to develop a plane, and microsurgical scissors can P.651 be used to separate the cyst wall from the ovarian cortex. Low-power magnification (i.e., surgical loupes) often assists the surgeon in identifying the correct plane between the cyst wall and the ovarian parenchyma. After the cyst wall has been completely separated from its adherent attachments to the thin ovarian cortex, it can be shelled out without rupture. However, even with the gentlest technique, rupture can occur because of the friability of the cyst wall. Before the cyst is shelled out, it is important to pack the cul-de-sac with moist, lint-free pads so that, if rupture does occur, spillage does not contaminate the pelvic cavity. After the cyst has been removed, the dead space can be obliterated with a pursestring suture of 7-0 nonreactive material. Alternatively, 5-0 nonreactive vertical mattress sutures or figure-of-eight, or both, can be placed to approximate the lateral walls of the ovary. The ovarian surface is then neatly reapproximated with a subcortical running suture of 7-0 nonreactive material (Fig. 26.6D). If the cortex is quite friable, it may be necessary to place interrupted 7-0 sutures to achieve adequate approximation. Some authors advocate leaving the ovary open after cystectomy. To date, there have been no controlled trials evaluating postoperative adhesion formation when the incised ovarian surface is or is not reapproximated. FIGURE 26.6. Resection of benign cyst. A: Thin-walled ovarian cyst. B: An incision is made through the cortex. C: A plane is developed by the use of blunt dissection. The inner ovarian stroma may be approximated with a pursestring suture of 5-0 nonreactive material. D: The ovarian cortex is approximated with 7-0 nonreactive suture material.

View Figure

In some instances, there is excessive redundant thin cortex, which may present a special problem in ovarian reconstruction. The amount of cortex removed depends on the position of the cyst as well as its overall size. Careful assessment of the ovary is necessary before the initial incision is made. The incision in the ovarian cortex should allow a symmetric reconstruction. The redundant cortex can be removed and the dead space obliterated with an internal closure, with care taken that suture material does not penetrate the ovarian cortex. This prevents ischemia and adhesion formation. The infolding technique recommended by Kistner and Patton may result in anatomic distortion and puckering of the ovarian cortex. The “baseball” closure allows careful approximation or cortical edges when redundancy is noted (Fig. 26.7). FIGURE 26.7. Closure of the ovary with a baseball stitch.

View Figure

Van der Watt reported concern over ovarian surgery. Of 36 young women operated on for an ovarian cyst, 45% were noted subsequently to be infertile. The author conveyed the importance of not interfering with functional cysts in “normal ovaries,” because resulting adhesion formation could compromise fertility. It was advocated that benign ovarian cysts should not be removed at the time of surgery for other indications unless they are sufficiently large to interfere with tubal function or cause discomfort to the patient.

Laparoscopy Specific skill levels reflecting both the degree of operator expertise and appropriate instrumentation provide clinicians with four levels of training. Level I stands for equipment needs and potential surgical procedures for basic operative laparoscopy, including such entities as diagnostic laparoscopy, tubal sterilization, lysis of filmy adhesions, and biopsy. Level II reflects the clinician's ability to perform linear salpingostomy for ectopic pregnancy, salpingectomy, lysis of vascular adhesions, and elimination of endometriotic implants. Level III includes the ability to perform salpingo-oophorectomy, lysis of extensive adhesions (including bowel adhesions), ovarian cystectomy, appendectomy, myomectomy, laparoscopic-assisted hysterectomy, and neosalpingostomy, as well as the ability to treat tuboovarian abscess and uterine suspension. Level IV includes bowel resection, anastomosis, pelvic lymphadenectomy, presacral neurectomy, tubal reanastomosis, and excision of deep, infiltrating vaginal, paravaginal, and rectal endometriosis. A number of principles should be followed as surgeons proceed with the correction of pelvic abnormalities that are amenable to a laparoscopic approach. The first is to restore normal anatomy. Once the ovary is stabilized, ideally with an atraumatic forceps, an appropriately planned ovarian incision can be made to correct the pathology encountered. Every effort should be made not to spill the contents. Large amounts of irrigation solution should be used. When necessary, hydrostatic pressure (aqua

dissection) facilitates removal of the ovarian cyst lining from the cortex. In some instances, the cyst wall can be stripped (Fig. 26.8), electrocoagulated, or vaporized. The ovarian incision can then be either left open to heal by primary intention or reapproximated with sutures with either extracorporal or intracorporal suture-tying techniques. When this procedure is completed, the pelvis is irrigated with large amounts of irrigation solution (Ringer's lactate), and the patient is taken out of the Trendelenburg position to facilitate removal of any blood products that remain in the peritoneal cavity. FIGURE 26.8. Removal of a small ovarian endometrial cyst through the laparoscope. A: After incision of the ovarian cortex, the contents of the endometrioma are removed with suction and irrigation. B: The plane between the ovary and the cyst wall is developed by using traction and twisting the forceps clockwise. C: The endometrial cyst wall is grasped with forceps. D: The cyst wall separates from the ovarian tissue by use of a twisting motion. The ovarian defect may be left open to heal by secondary intention or may be closed with vertical mattress sutures. E: The cyst wall is removed.

View Figure

One alternative to suturing is to reapproximate incised segments of ovarian cortex with the use of bipolar coagulation to provide coaptation of the incised segment of the ovary. There is continued debate regarding the use of adhesion prevention materials. A number of potential pitfalls continue to be of concern in the laparoscopic approach to ovarian lesions. These have been addressed by Seltzer and include the following: The potential for disruption of an ovarian malignancy Whether observation-recommended surgical intervention would be the most feasible alternative Potential for increased duration of the surgical procedure if done endoscopically Total cost Potential for incomplete resection of an ovarian lesion laparoscopically Education and credentialing of the gynecologic surgeon Overlooking the ultimate surgical goal One can view laparoscopic approach to the adnexal mass based on age. Specifically, in the pediatric patient, problems such as torsion, hemorrhagic cysts, benign neoplasm (e.g., teratoma), as well as oophorectomy have been reportedly addressed via the laparoscope. One advantage over laparotomy is the ability to better visualize the entire lower abdomen and pelvis including the opposite ovary. In the adult, depending on the clinical circumstance, cyst aspiration, cystectomy, or oophorectomy can be accomplished laparoscopically. Concern is expressed for an ovarian neoplasm subsequently noted to be malignant. In a countrywide survey in Austria, Wenzl and colleagues reported on 54,198 laparoscopies; 16,601 were performed for adnexal masses and 108 cases of ovarian tumors were subsequently found to be malignant. Of the 108 cases, 20 were managed laparoscopically, 22 by immediate lapa-rotomy, and the rest by delayed laparotomy (3 to 1,415 days). The authors concluded that laparoscopic surgery with the finding of an ovarian malignancy is rare: 0.65% of all endoscopic surgical procedures. If a malignancy is identified, laparotomy is recommended for optimal staging and treatment.

OVARIAN SURGERY FOR POLYCYSTIC OVARIAN DISEASE Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" Signs and symptoms of polycystic ovarian (PCO) syndrome begin at puberty. PCO is a sign, not a diagnosis. It is usually accompanied by a degree of hirsutism, infertility, and, in most cases, oligoovulation. The polycystic ovary may result from a virilizing ovarian or adrenal neoplasm or from congenital adrenal hyperplasia, or it may result from suboptimal hypothalamic-pituitary function at puberty. The exact mechanism for the development of ovulatory failure has been attributed to androgen overproduction and its effect on the hypothalamic-pituitary ovarian axis. Stein and Leventhal, during the period 1902 to 1935, noted that a group of women had evidence for, what is currently called polycystic ovaries, at the time of laparotomy. Specifically, in 1935, Stein and Leventhal reported seven patients with the hallmarks of PCO. The histologic findings in a polycystic ovary cover a broad spectrum, ranging from the typical Stein-Leventhal type of polycystic ovary with a large number of follicular cysts and few atretic cysts in which there is marked stromal hyperplasia and hyperthecosis, to a smaller ovary with a few follicular cysts and atretic follicles. The polycystic ovary may exhibit microscopic islands of luteinized thecal cells scattered in the stroma, but usually there is a thickened, fibrosed tunica with a large number of cystic follicles beneath this thickened capsule. There are several hypotheses regarding the mechanism by which wedge resection of the polycystic ovary resolves ovulatory failure. The theory stating that the fibrous capsule acts as a mechanical barrier to the ovulatory follicle has been refuted. Evidence against this theory consists of the observation that if one ovary is removed, ovulation occurs from the other ovary. In addition, the use of clomiphene citrate results in ovulation through an intact capsule. Some have stated that neonatal androgens may cause an abnormal hypothalamic-pituitary axis, resulting in abnormal gonadal patterns. This theory is not widely accepted. Neonatal androgen treatment in rats is associated with masculinization of the hypothalamus and with ovulatory failure with polycystic ovaries. The most popular theory explaining how wedge resection results in the resumption of ovulatory cycles notes that the removal of androgen-secreting stroma and theca reduces the amount of abnormal steroid production in the ovary. After wedge resection, there is usually a decrease in the mean level of 17?-hydroxyprogesterone, dehydroepiandrosterone, androstenedione, and testosterone, as well as a transitory decrease in estradiol. This reduction in the steroidogenesis of androgens, allowing normalization of the luteinizing hormone (LH): follicle stimulating hormone (LH:FSH) ratio, results in the resumption of ovulatory cycles. Ovarian renin-angiotensin activity is enhanced with PCO. This system—renin-angiotensin—remains unaltered following ovarian electrocautery (i.e., ovarian drilling), even though serum levels of LH, testosterone, and androstenedione decline. Sex hormone–binding globulin (SHBG) concentrations following electrocautery with PCO have been evaluated. Whereas there were significant decreases in serum androgens and gonadotropins, the concentration of SHBG increased in the serum. Gjonnaess has reported that there is no change with respect to dehydroepiandrosterone sulfate (DHEAS) with ovarian drilling. This is indicative of neural alteration in the pituitary-adrenal axis in comparison to the pituitary-ovarian axis. There is some debate as to the amount of ovarian mass that should be removed at the time of wedge resection. Halbe and co-workers attempted to clarify this question by removing different amounts of ovarian cortex and medulla from a random selection of patients with polycystic ovarian disease. Thirty-eight of 62 patients were interested in conception. The 38 patients were divided into three groups, the first of which underwent removal of not more than one fifth of the original ovarian size. The second group had one third of the ovarian mass removed, and the third group had one half to three fourths of the original ovarian size reduced. The resumption of ovulatory cycles was recorded at 53%, 71%, and 91%, respectively. The authors concluded that the best ovulatory rate and the best pregnancy rate resulted after removal of at least half of the ovarian medulla.

Indications The introduction of clomiphene citrate has changed the management of polycystic ovarian disease in patients who desire pregnancy. Johnson reported ovulation in 359 of 436 patients with Stein-Leventhal syndrome with the first course of therapy. An additional 58 patients ovulated after two or more cycles. The conception rate with clomiphene citrate (50% to 60%) is, however, less than 86% pregnancy rate reported

by Stein and Leventhal for wedge resection. The incidence of post-clomiphene citrate birth defects (3.1%) is not increased over commonly quoted rates for populations at large. Some patients may not want to accept the risks of multiple births or hyperstimulation with pure FSH or FSH and LH ovulation induction. The need for wedge resection of the polycystic ovary is much lower with the development of the newer reproductive technologies. Antidiabetic agents have been advocated to reduce insulin resistance with PCO. Metformin has been shown to decrease insulin levels with resultant diminishing of circulating androgens. Hirsutism often improves. Metformin may enhance the efficacy of clomiphene and gonadotropin therapy with PCO. Metformin may also promote weight loss. Baseline and periodic liver function tests are recommended. Metformin is contraindicated with renal or hepatic disease. Patients have shown a response at dosages of 500 mg three times per day.

Surgical Technique of Laparoscopic Treatment of Polycystic Ovaries The laparoscopic approach incorporates the use of monopolar cautery with a needlepoint applicator, laser, or bipolar cautery to drill holes several millimeters apart through the ovarian cortex (Fig. 26.9). Care should be exercised to avoid the hilum because bleeding could result if it is penetrated. It is important to achieve hemostasis over the drilled areas. FIGURE 26.9. Laser drilling of ovary for surgical treatment of polycystic ovarian disease.

View Figure

The ovarian drilling technique is performed with a 10-mm laparoscope coupled with a CO2 laser. A 5-mm second puncture is placed suprapubically, through which suction irrigation or grasping of tissues can be performed. All visible subcapsular follicles are vaporized, and a 2- to 4-mm diameter crater is made randomly in the ovarian stroma. Hemostasis is accomplished with bipolar forceps. Ovarian coagulation can be accomplished with unipolar punch biopsy forceps or a needle electrode. The power setting is 20 to 30 W in a cutting mode. The cortex is usually penetrated at 10 to 15 sites for a depth of 3 to 5 mm. Caution is exercised to minimize thermal damage. Smaller ovaries may require fewer cauterization sites. There are no randomized controlled studies addressing the efficacy of the laparoscopic approach to ovarian drilling. Twenty-seven studies were evaluated by Donesky and Adashi and involved a total of 729 patients. The ovulation rate was 84.2% and the pregnancy rate was 55.7%. These authors emphasized that well-designed studies are needed in this area, which would encompass the PCO population proposed for laparoscopic drilling. This cohort of patients would require a well-documented clinical and biochemical finding of PCO, documented long-standing infertility (2 years or more), evidence for failure of clomiphene citrate, absence of correction of other infertility factors, randomization into a treatment group, and standardized documented follow-up, with particular attention to postovulatory patterns.

Complications The major concern is that of adhesion formation after either wedge resection or laparoscopic drilling. Toaff and associates noted extensive peritubular and periovarian adhesions in a small series (seven) of patients who did not conceive after bilateral wedge resection. One other concern is that of bilateral ovarian atrophy, as a reflection of aggressive ovarian resection. This is a rare complication of the procedure. Thus, iatrogenic consequences of the surgical approaches must be discussed with the patient preoperatively. There is continued controversy as to whether a bilateral wedge resection approach results in suboptimal pregnancy rates because of postoperative adhesion formation. The principles of microsurgery must be reemphasized (e.g., gentle tissue handling, precise hemostasis, keeping tissues moist).

PARADOXICAL OOPHORECTOMY Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" Paradoxical oophorectomy is the removal of severely pathologic adnexa to improve fertility in patients with strictly unilateral tubal disease. Consideration of removal of the contralateral ovary when there is single tubal patency (i.e., paradoxical oophorectomy) has perhaps taken on a new perspective with the advent of assisted reproduc-tive technology (ART). From a historical point of view, Jeffcoate advocated that a patient with one functional tube would benefit from paradoxical oophorectomy, thus ensuring that ovulation would occur repeatedly on the appropriate side. Scott and co-workers reported a series of 24 patients with unilateral tubal patency diagnosed by retrograde injection at laparotomy. Contralateral oophorectomy or salpingo-oophorectomy was performed on all patients, and 16 women subsequently had 21 pregnancies, for a pregnancy rate of 67%. The authors suggested that the frequency with which transperitoneal migration occurs may be a factor. Hallet noted that one in five tubal ectopic pregnancies has a corpus luteum. On the contralateral side, Jansen noted an intrauterine pregnancy rate of 18.7% (n = 91), contrasted with bilateral salpingostomy for hydrosalpinges in the presence of only one ovary wherein the pregnancy rate was 43.8% (n = 16). With unilateral salpingostomy or bilateral division of adhesions, pregnancy rates were comparable to those after bilateral salpingolysis. The mean surgery-pregnancy interval was longer after unilateral salpingostomy (104 weeks) than after bilateral salpingolysis (45 weeks). The author suggested that salpingo-oophorectomy may be preferable to salpingoneostomy for unilateral hydrosalpinx. Perhaps the major concern is for the patient who presents with tubal ectopic gestation in which the opposite (i.e., normal-appearing) adnexa appears to be unaffected. The paradoxical salpingo-oophorectomy approach has been advocated with this circumstance by Scott and co-workers. It has been advocated to wait at least 2 years after diagnostic laparoscopy reveals extensive unilateral disease before proceeding with paradoxical oophorectomy. Randomized, carefully controlled clinical trials are necessary to further evaluate the efficacy of paradoxical oophorectomy. The risks and benefits must be carefully considered both preoperatively and intraoperatively, especially if the patient is a candidate for ART. There is clear evidence that increased numbers of ova can be recovered when both ovaries are in situ. Increased pregnancy success after superovulation is a reflection of the number of ovaries (one versus two)—the total number of follicles available for stimulation.

Laparoscopic Oophorectomy The general principles of laparoscopic oophorectomy include placing the patient in the Trendelenburg position, with appropriate planning of ports for the proposed procedure, and planning for removal of the affected adnexa. Pelvic washings and the use of frozen section may be germane to the task at hand. After restoration of normal anatomy and adhesiolysis as indicated, the adnexa are gently placed on stretch. They are approached from either the infundibulopelvic ligament or the insertion of the round ligament. Regardless of the approach chosen, identification of the ureter is mandatory. The infundibulopelvic ligament is identified, ligated with a loop ligature, or coagulated. Use of a suture ligament is an appropriate alternative to coagulation. The broad ligament is incised, beginning at the round ligament, and further dissection is performed with an irrigating dissecting probe into the retroperitoneal space. Every effort must be made to completely remove all ovarian tissue to prevent ovarian remnant syndrome.

Tissue Removal Once the adnexa have been completely freed, if benign disease is extremely likely, desiccation of the tissue and thus segmental removal is appropriate. However, if there is concern for the pathology, use of either an endoscopic pouch or a culpotomy incision is appropriate. In this circumstance, every effort is made to remove the ovary intact. Careful inspection of the operative site and a check for any bleeding are recommended. In addition, the end-point pressure of CO2 insufflation should be reduced with suctioning of some of the CO2 to check for any tamponade effect. As with all laparoscopic procedures, the patient should be monitored carefully after surgery for any signs of intraperitoneal bleeding.

OVARIAN TRANSPOSITION BEFORE RADIOTHERAPY Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" Lemevel and co-workers reported laparoscopic transposition in a patient being treated for Hodgkin's disease before receiving radiotherapy. The ovaries were laparoscopically suspended out of the field of radiation. Iatrogenic menopause did not occur in the four patients for which this was reported. Other authors have reported similar recommendations. Fertility outcome following ovarian transposition and pelvic irradiation for pelvic cancer has been addressed in a total of 37 consecutive cases by Morice and colleagues. Patients were treated for clear cell adenocarcinoma of the vagina or cervix, ovarian dysgerminoma, and sarcoma. The pregnancy rate was 15% (4/27) in patients attempting pregnancy with clear cell adenocarcinoma of the vagina or cervix. In the dysgerminoma and sarcoma group, 80% pregnancy occurred (8/10). Thus, the prognosis for future fertility following ovarian transposition and irradiation should be considered for discussion in selected patients before radiotherapy.

LAPAROSCOPIC OVARIAN SURGERY IN THE PEDIATRIC OR ADOLESCENT PATIENT Part of "Chapter 26 - Surgery for Benign Disease of the Ovary" A number of gynecologic problems from the neonatal period through adolescence can be addressed laparoscopically. Entities such as ovarian torsion, acute pelvic inflammatory disease (diagnosis and treatment complementing antimicrobial therapy), torsion, and benign neoplasms must be considered in this age group. In addition, gonadectomy for problems such as male pseudohermaphroditism is amenable to the laparoscopic approach. Follicular cysts appear to be of particular concern in both the pediatric and adolescent patient because they present with abdominal pain. An abdominal or pelvic mass can be identified on physical examination. The clinician must always keep in mind the importance of appropriate preoperative assessment (with the use of ultrasound and other clinical parameters) in deciding which patients are candidates for operative intervention. Depending on the clinical circumstance, a conservative approach in this age group is advocated; appropriate concern must be given (especially in the pediatric patient) to the potential for malignancy with ovarian masses, particularly if a solid component is identified. The literature attests to operative intervention of ovarian cysts that appear to be nonfunctional. One of these was reported in a 1.5-month-old infant in whom ultrasound showed evidence of an ovarian mass. At laparotomy, the mass proved to be consistent with right ovarian torsion and necrosis, and it required adnexectomy. In one other reported case, a follicular-appearing cyst seen on ultrasound was associated with rapidly progressive virilization and a markedly elevated plasma testosterone level (289 ng/dL); histologic evaluation identified a granulosa cell tumor with mild luteinization. The feasibility of aspirating an ovarian cyst continues to be controversial. In two case reports, laparoscopic puncture and aspiration of a malignant ovarian cyst was performed. Preoperative ultrasound indicated that the involved adnexal mass had a benign nature. Cytologically negative fluid was obtained from the aspirate. Eight weeks after operation, extensive disseminated ovarian carcinoma was noted at laparotomy. Endoscopic surgery continues to broaden its horizon with expansion into laparoscopic surgical care beginning with the neonate. The use of 2-mm laparoscopes with the addition of 2- to 3-mm instrumentation has facilitated the diagnostic and therapeutic aspects of laparoscopy in this age group. Miniaturized video camera systems are necessary when one uses the 2-mm laparoscope telescope. When a decision is made to proceed with laparoscopy in a neonate or infant, general endotracheal anesthesia is used. Ideally, prophylactic antibiotics are administered preoperatively. The stomach is emptied with a suction catheter when the patient is asleep, and the bladder is emptied by use of the Crede maneuver. The abdominal wall is thinner and more elastic in the child than in the adolescent or adult. One must take this into consideration when introducing instrumentation, because it may be easier to insufflate in the subcutaneous space of the child than in the subcutaneous space of an adult. A Veress needle can be used with insufflation of carbon dioxide at 0.5 L/min. In infants, the end point of peritoneal distending pressure should be set at 6 to 8 mm Hg; in the pediatric patient, 8 to 10 mm Hg; in the older child or adolescent, 10 to 12 mm Hg. After surgery, the trocar sites can be sutured in children, whereas in neonates and infants the use of Steri-Strips (3M, St. Paul, MN) or other wound closure bandages is usually adequate for reapproximation of the incised skin edges. Waldschmidt and Schier reported a series of 136 laparoscopic surgical procedures in neonates and infants. The most frequent indications were lysis of adhesions, abdominal cysts and neoplasms, gonadectomy, appendectomy, and cholecystectomy. A 1,400-g preterm infant was the only one in the series who suffered a complication (hernia at the incision site). Thus, adnexal pathology in this age group appears to be amenable to a laparoscopic approach. Procedures such as transposition of an ovary before radiotherapy, bilateral gonadal excision in a male pseudohermaphrodite (i.e., Y-bearing chromosomal analysis), adnexal torsion, suspected salpingitis, and endometriosis all have been identified in this age group. The authors concluded that because the morbidity is low and recovery is likely, the laparoscopic approach should be considered. Although certain procedures in the child do not differ significantly from those in the adult, the early diagnosis of ovarian pathology (e.g., adnexal torsion) can result in a significant advantage in terms of managing the patient and preserving ovarian tissue.

Chapter 27

Persistent or Chronic Pelvic Pain John F. Steege Chronic pelvic pain (CPP) is defined as pain present either intermittently or continuously for 6 months or more. Although much work remains to be done concerning the epidemiology of the disorder, preliminary surveys suggest that it may affect 5% to 15% of women at some time in their lives, predominantly during their reproductive years. The gynecologist confronts this problem on a regular basis. Many organic and functional disorders of the reproductive, gastrointestinal, urinary tract, and musculoskeletal systems may contribute nociceptive stimuli to the problem. In addition, by dysregulation of neurologic pathways, pain can become an illness in itself when it lasts longer than 4 to 6 months. This chapter reviews the common structural and functional components of CPP, describes the criteria for a chronic pain syndrome (CPS), and offer a theoretical model to explain the evolution of chronic pain over time. DEFINITION HISTORY CONTRIBUTIONS OF TISSUE DAMAGE TO PAIN PSYCHOLOGIC FACTORS DIAGNOSTIC STRATEGIES MANAGEMENT THE EVOLUTION OF A CHRONIC PAIN SYNDROME

DEFINITION Part of "Chapter 27 - Persistent or Chronic Pelvic Pain" Many pain disorders are deemed “chronic” by virtue of their duration, that is, 6 months or longer. In practice, this usually connotes decreased function (perhaps even disability) as well as progressive behavioral and affective changes. Although this definition is useful as a benchmark, some victims of chronic pain may develop these characteristics sooner than 6 months into the problem, and others may endure severe pain for years while remaining surprisingly functional and well adjusted. As suggested later, it is therefore perhaps useful to employ additional behavioral and psychometric criteria to the chronologic definition of “chronic” in order to focus diagnostic and treatment efforts.

HISTORY Part of "Chapter 27 - Persistent or Chronic Pelvic Pain" Over the past 50 years, the study of CPP has gone through significant changes in approach. Investigations undertaken before the development of laparoscopy focused on correlations between pelvic pain and psychologic distress. (These studies included an extensive literature dealing with the psychologic dimensions of dysmenorrhea, much of which has received less attention since the discovery of prostaglandins.) This approach was prompted by the fact that the two most common organic contributors to CPP—endometriosis and adhesions—can be difficult to diagnose by history and physical examination alone. In the absence of palpable pathology, the gynecologist of the 1950s and 1960s was understandably reluctant to subject a patient to laparotomy to investigate pain. During this era, the prevailing cartesian theory of pain perception suggested that pain should be somewhat proportional to the degree of tissue damage found. If the pathology was not big enough to palpate, it was seldom operated upon. Although this was sufficient to explain most acute pain, the cartesian model fails to explain the majority of chronic pain disorders, in gynecology as well as other areas of medicine. The gate control theory, promulgated by Melzack and Wall in 1965, allowed integration of physical and psychologic parameters, and explains how chronic pain can be quite different from acute pain. The model also suggests that information flows in two directions regarding pain: (1) nociceptive signals from peripheral tissue ascend through the spinal cord to higher centers, and (2) central centers can modulate, via descending signals altering spinal cord neurotransmitter and interneuron activity, the transmission of these nociceptive signals from the periphery. Deterioration of these regulatory processes were thought to potentially account for development of chronic pain states by allowing too many peripheral signals through the spinal cord “gates.” While these changes in pain theory were stimulating the field of pain research, gynecologists were busy developing laparoscopy. Previously cherished myths soon fell by the wayside, for example, endometriosis is seldom found in adolescents or African-Americans. With these observations came the hope that laparoscopic and medical treatment of this pathology would fix CPP. Reports of CPP from that era focused on “laparoscopy-negative” patients; indeed, some pelvic pain clinics required a negative laparoscopy as an entry criterion, implying that if some pathology were found, it must be a “real” cause for pain. Subsequent experience has shown that, even though treatment of laparoscopically diagnosed pathology is often helpful, the clinical reality is more complex: In many instances, the organic pathology found at laparoscopy may be incidental, and not related to the pain. In those with pathology that does contribute to nociception, the pain experienced by the patient may be the sum of this contribution plus signals from some or all of the disorders listed in Table 27.1. TABLE 27.1. Possible Causes of Pelvic Pain in Laparoscopy-Negative Patients

Gastrointestinal

Constipation

Irritable bowel syndrome

Inflammatory bowel disease

Diverticulitis

Urinary

Urethral syndrome

Interstitial cystitis

Musculoskeletal or Neurologic

Pelvic floor tension myalgia

Piriformis syndrome

Nerve entrapment

Ventral hernia

Rectus tendon strain

Myofascial pain

Back or pelvic postural changes

Gynecologic

Pelvic vascular congestion

Cervical stenosis

From Steege JF, Stout AL, Somkuti SG. Chronic pelvic pain: toward an integrative model. Obstet Gynecol Surv 1993;48:95, with permission.

Consider the research of the 1980s that documented a distressingly high prevalence of physical and sexual abuse. Epidemiologic surveys of community samples revealed that as many as 25% to 30% of adult women reported having experienced sexual abuse during childhood. Studies of women attending pelvic pain clinics, especially those based in psychiatric settings, showed that up to 60% of these women had been abused. These observations led to the speculation that the experience of abuse may make a person more vulnerable to the development of CPP or perhaps be a specific cause for pain. Currently, preliminary studies using positron-emission tomography and functional magnetic resonance imaging (MRI) methods suggest that the experience of abuse may indeed leave its neurophysiologic footprints: stressful stimuli produce different central response patterns in abused vs nonabused subjects. However, detection of abuse in a patient's history seems to prompt referral for tertiary care, whereas symptom levels in various pain disorders [e.g., irritable bowel syndrome (IBS)] are similar in referred and nonreferred patients. The presence of an abuse history means that the evaluating health professionals need to take this into account, but it does not necessarily imply that the abuse is causally involved in the development of the pain. The impact of a history of abuse on the outcome of treatment of organic pathology should be the subject of further research. Finally, even the gate control theory is held to be inadequate to explain clinical manifestations of pain. The neuromatrix theory, described by Melzack, is an expansion that includes the notion of neuroplasticity, among other elements. The concept of neuroplasticity suggests that experience can change the neurophysiologic behavior of the central nervous system in a manner that influences the subsequent processing of nociceptive stimuli. It may explain the apparent development of pain responses to stimuli usually thought of as nonpainful (allodynia), as well as exaggerated responses to painful stimuli (hyperalgesia). Every practicing gynecologist has seen patients whose pain responses seem out of proportion to the pathology found. This may reflect the emotional meaning of the problem for the patient, as well as past or present emotional trauma, but it may also be the result of nociceptive mechanisms not yet understood (e.g., the mechanism of pain from endometriosis) or the result of sensitization of spinal cord interneurons that have become pain generators as a result of being on the receiving end of peripheral nociceptive stimuli for prolonged periods. The positive side of the neuroplasticity concept is that, perhaps, given enough time and the right treatment, even seemingly intractable chronic pain problems may ameliorate to the point of allowing substantially improved function. This chapter reviews the various organic, psychologic, and physiologic factors that contribute to CPP, outlines methods of evaluation, gives suggestions for treatment, and finishes with a discussion of an integrated model of the development of CPP.

CONTRIBUTIONS OF TISSUE DAMAGE TO PAIN Part of "Chapter 27 - Persistent or Chronic Pelvic Pain"

Endometriosis A recent review summarized laparoscopic findings from 2,615 patients in 15 studies (nine retrospective, six prospective). Endometriosis was found in 2% to 51% of patients, suggesting that referral biases lead to very skewed samples. Clearly, not every woman with pain has endometriosis, nor does every woman with endometriosis have pain, although women with the disease had pain more often than those without it. A number of previous studies of CPP either described only patients without organic laparoscopic findings or stratified patients according to the presence or absence of physical pathology. The description of atypical (nonpigmented) endometriosis by Jansen in 1986 calls these classifications into question. Laparoscopy studies published before that time reported that 11% of women with CPP had endometriosis, whereas three similarly conducted studies published since 1986 reported a 41% prevalence of endometriosis in women undergoing laparoscopy for CPP. The pre-1986 literature on pelvic pain must be reevaluated with this information in mind. Many studies may have included women with endometriosis in the anatomically normal group, thus generating erroneous conclusions about the entirely psychogenic nature of their pain. The first symptom of significant endometriosis is often increased dysmenorrhea alone, but other pain often develops, and its duration and severity often progress as the disease advances, to the point that it can be present almost constantly. Again, the severity of the pain correlates poorly with the amount of diffuse peritoneal disease, but may vary more directly with deeply infiltrative cul-de-sac disease. Fear of worsened pain, impaired fertility, or recurrent disease after treatment can increase pain levels. Of the many women upon whom we have performed laparoscopy for recurrent pain following complete hysterectomy and adnexectomy for endometriosis, only a small minority (3% to 5%) proved to have recurrent disease. Most cases of postoperative pain have been attributable to a combination of postoperative adhesions, fear of recurrent disease, and functional problems such as levator spasm and IBS. Hysterectomy and bilateral salpingo-oophorectomy relieve endometriosis-related pain in more than 90% of cases. When one or both ovaries are preserved, the recurrence rate is estimated to be 30% or higher, but this figure includes endometriosis of all stages. Some practitioners believe that if disease does not involve the ovaries, surgical excision of the disease, with ovarian conservation, results in an acceptable 5% to 10% recurrence rate over the subsequent 5 years. Further investigation of this unorthodox approach is warranted.

Pelvic Adhesions In a review by Steege and colleagues, 6% to 55% of the 2,615 patients who underwent laparoscopy for pelvic pain had pelvic adhesions. As is the case for endometriosis, the site of adhesive disease correlated well with the site of pain, but the intensity of pain was unrelated to the extent of adhesions present. Nociceptive signals from the damaged tissue seem subject to modulation at the spinal cord level and interpretation in higher centers. Adhesions are most likely stable anatomically a few months after injury (e.g., surgery, infection), but the intensity of associated pain can progress. After about 6 months of pain experience, complex relationships among physical, emotional, and cognitive factors can exist that require comprehensive treatment.

Pelvic Support Although problems with pelvic relaxation are common in women in their sixth or seventh decade of life, most patients at pain clinics are in their third or fourth decade. The implication is that cases of pain associated with pelvic support problems constitute the minority of pelvic pain problems. Pelvic relaxation usually leads to complaints of heaviness, pressure, dropping sensations, or aching. In attempting to hold in prolapsing organs, the patient may be tensing the levator plate and contributing to tenderness during daily activities and intercourse. Fear of (or actual) loss of urinary control during coitus

can add to the discomfort by impairing physiologic sexual response. Excess mobility of the pelvic organs (universal joint or Allen-Masters syndrome) attributed to childbirth or other traumatic causes of ligament (especially broad ligament) tears in uterine supports has been implicated in CPP. Such highly subjective physical examination findings are difficult to document rigorously, and because surgery has been the traditional primary treatment, controlled studies of the association of pain with ligamentous tears have not been possible. Uterine retroversion is another controversial potential etiology for CPP, particularly in the form of deep dyspareunia. Uncontrolled clinical series of uterine suspension procedures for pelvic pain suggest that the most frequent scenario involves a combination of uterine retroversion (most often innocent by itself) with new intrinsic uterine pathology (e.g., adenomyosis, myomas) or nearby cul-de-sac or adnexal pathology (e.g., endometriosis, adhesions, ovary prolapsed into the cul-de-sac). On occasion, these anatomic circumstances combined with a new partner of more generous penile dimensions produce new deep dyspareunia.

Pelvic Congestion Overfilling (congestion) of the pelvic venous system has been implicated as a cause of dull chronic aching pain that usually is worse at the end of the day after prolonged standing, premenstrually, and after coitus not accompanied by orgasm. The pain is occasionally unilateral, is usually present in multiparous women, and is likely due to anatomic causes, at least in part. The high comorbidity with psychologic distress indicates that psychophysiologic factors may also contribute. Medroxyprogesterone can help (by eliminating the menstrual cycle), but the best long-term results have followed psychotherapy in conjunction with this medication. In any case, good radiologic studies in women who are awake during the procedure have documented increased pelvic venous diameter in those affected. Veins cannot be reliably evaluated during laparoscopy due to confounding effects of position, fluid load, etc.

Residual Ovary When the uterus has been removed, with or without removal of one ovary, the remaining ovary or ovaries can become symptomatic in 1% to 4% of women. In many instances, benign functional ovarian cysts can form and can be transiently symptomatic. Pain from the ovary can be increased by confinement of the ovary within postoperative adhesions, rupture or leakage of a cyst prompting additional adhesion formation, or attachment of the ovary to the sigmoid colon or vaginal apex by postoperative adhesions. In the case of attachment to the vaginal apex, deep dyspareunia can result when the vaginal apex is struck. When this pain causes diminished sexual response, loss of the normal vaginal apex expansion and elongation that is part of sexual response can leave the ovary or ovaries closer to the introitus, thus aggravating the deep dyspareunia.

Ovarian Remnant A more difficult situation can develop if a small fragment of ovarian tissue is left behind during attempted oophorectomy. In most instances, this happens when extensive pelvic adhesive disease or endometriosis made the dissection difficult. Within 1 to 3 years of the attempted oophorectomy, continued follicle-stimulating hormone (FSH) stimulation will result in growth of the ovarian fragment, often producing an intermittently symptomatic pelvic mass located along the course of the ovarian vascular supply. This problem is uncommon, but not rare, as implied by early case series. The prevalence of asymptomatic ovarian remnants is unknown. As in the case of the residual ovary, the remnant can produce dyspareunia if it is located close to the vaginal apex.

Musculoskeletal Problems Musculoskeletal changes can become involved with CPP, either as the primary problem or as a secondary reaction to the pelvic pain. Dysmenorrhea can be referred to the midline of the low back, especially when

the uterus is retroverted. Pain can also be referred to the midline of the low back in the presence of cul-de-sac endometriosis. An ovary fixed to the pelvic sidewall can refer pain to the ipsilateral low back. The muscular problem that most often produces pelvic pain is pelvic floor tension myalgia. Intermittent or constant painful contraction of the levator plate can be present as a primary psychophysiologic problem, but contraction is more often a reaction to some other source of pain. Even when the primary source of pain is successfully treated, the reactive muscle contraction can persist as a learned response, in much the same way that vaginal introital muscle spasm (vaginismus) can persist after transient but repeated painful vaginal events. Lumbar musculature can become tender as a primary problem or in reaction to subtle changes in posture and motion. Trigger points can be present in the low back and gluteal areas, in the muscles best inspected by pelvic examination (e.g., levator plate, piriformis, obturator internus). The piriformis muscle warrants additional mention because it is seldom appreciated as a possible source of pain. This muscle is an external rotator of the leg, and rotation against resistance can allow detection of tender spasm of the muscle during the pelvic examination. The sciatic nerve can traverse the belly of the piriformis as a normal anatomic variant, producing symptoms similar to sciatica when the muscle is in spasm.

Myofascial Pain Focal lower quadrant abdominal wall pain can be produced by entrapment of the genitofemoral and ilioinguinal nerves, as described by Applegate. Such entrapment appears most often after Pfannenstiel abdominal incisions. Slocumb has advanced the theory that myofascial trigger points account for a large fraction of CPP. In my experience, abdominal wall components may be the primary cause for pain in some cases, but are more often a later reaction to the long duration of pain from some other source. Reiter and Gambone reported that 14% of 122 laparoscopy-negative women had myofascial pain probably related to a previous surgical incision.

Medical Comorbidity The cause of chronic lower abdominal or pelvic pain often involves nongynecologic systems (Table 27.1). A careful history and close physical examination of gastrointestinal, urologic, musculoskeletal, and neurologic systems are needed to evaluate the contributions of nongynecologic systems to CPP. Most of the available literature examines these problems of other systems independently of each other and without reference to their relevance to CPP or to the overall prevalence of these disorders in CPP. The gastrointestinal system is perhaps the most common nongynecologic source of pelvic pain. Constipation and IBS occur most frequently, although inflammatory bowel disease and diverticulitis can at times present with pain alone. Women with IBS may have increased relaxin levels (produced by a dysfunctional corpus luteum) as one of many possible contributing factors. Treatment with a gonadotropin-releasing hormone (GnRH) agonist may reduce symptoms of IBS. Urologic problems, which are less easily confused with gynecologic disorders, are perhaps second in terms of prevalence. The urethral syndrome (frequency, urgency, and dysuria in the absence of bacteriuria), interstitial cystitis, and bladder spasms are all accompanied by significant anxiety and depression symptoms. The symptoms of these three disorders are very similar to those in a population of gynecologic CPP patients. A history (whether pain occurs during micturition, daily activities, or coitus) does not always reveal the involved system, but careful pelvic examination with stepwise gentle palpation of the urethra, bladder base, and bladder may help the physician identify from the patient's response the site of the pain she is experiencing. Many patients do not experience the problems described here in pure form, but rather in varying degrees of intensity, with varying contributions to an individual's total discomfort. Close attention to such nuances of detail is warranted both in clinical management and in published reports.

PSYCHOLOGIC FACTORS Part of "Chapter 27 - Persistent or Chronic Pelvic Pain"

Personality The links between chronic pain and individual psychology and personality style have been sought after and discussed in the psychiatric literature for many years. Some early reports implied that women who complained of CPP had a high prevalence of feminine identity problems related to conflicts about adult sexuality, psychiatric disturbance characterized by mixed character disorder with predominant schizoid features, high neuroticism, and unsatisfactory relationships. Although these initial studies were an important beginning, the high prevalence of psychopathology in some reported samples did not seem applicable to significant numbers of CPP patients seen in practice. The findings are difficult to interpret, partly because there is a lack of clarity concerning the operational definition of CPP that was used. Biases in patient selection and interviewer information, inadequate control groups, and the absence of diagnostic laparoscopy also contribute to the confusion. Despite these shortcomings, it seems apparent that disorders of personality, especially borderline personality, are overrepresented both in the general population of severe chronic pain patients and in the population of pelvic pain patients. In primary care, such patients usually are seen less often. In any case, a label of personality disorder should not be applied indiscriminately to every angry patient by her frustrated physician. People who have difficulties maintaining satisfactory relationships and function in life, even when these difficulties are caused in part by subsyndromal personality problems, can be more vulnerable to nociceptive signals from tissue damaged by endometriosis, infection, or surgery. Unmet dependency needs may lead them to seek external solutions such as medications and further surgery, rather than to rely on their own undeveloped coping skills.

Depression Focusing specifically on a CPP sample that had been evaluated by diagnostic laparoscopy, Walker and associates found that women with CPP (with and without positive laparoscopic findings) met criteria for lifetime major depression, current major depression, lifetime substance abuse, adult sexual dysfunction, and somatization more often than did control subjects. Stout and Steege found that 59% of 294 women seeking evaluation at a pelvic pain clinic scored in the depressed range (greater than 16) on the Center for Epidemiologic Studies Depression Scale at the time of their initial visit. Slocumb and colleagues reported that patients with an abdominal pelvic pain syndrome scored higher as a group on scales of anxiety, depression, anger-hostility, and somatization on the Hopkins Symptom Checklist; however, 56% of the total sample scored within the normal range on all scales. Because no study of CPP has assessed its association with depression over time, no statement can be made as to whether depressive symptoms are a predisposing factor leading to, or a reaction to, the pain condition. There seem to be two distinct groups of CPP patients: one in which pain and depression are common final presentations reached by a number of pathways and another in which depression develops in reaction to pain, as is the case with many other acute and chronic medical diseases.

History of Sexual Abuse Women seeking treatment for CPP have a high prevalence of sexual trauma in their personal histories. In Reiter's study of 106 women with CPP, 48% had a history of major psychosexual trauma (molestation, incest, or rape), compared with 6.5% of 92 pain-free control subjects presenting for annual routine gynecologic examination (p < .001). The high prevalence of reports of psychosexual trauma elicited from pelvic pain patients supports the hypothesis that pelvic pain is specifically and psychodynamically related to sexual abuse. However, Rapkin did not find a higher prevalence of childhood or adult sexual abuse in a group of women with CPP compared with women with chronic pain in other locations, although women with CPP reported a higher incidence of childhood sexual abuse. These findings argue against a unique relation between sexual abuse and CPP and suggest that abusive experiences promote the chronicity of

many different painful conditions. Morrison also reported an association between sexual abuse and a wide variety of pathologic conditions. Jamieson and Steege, in a survey of 581 women seen in primary care practices, found that 28% of women reported having been sexually abused as children, and 26% as adults. In this study, those abused only in childhood did not have an increased prevalence of CPP or other pain disorders, whereas those who had suffered abuse both as children and adults did. When such a history is documented, the clinician and patient together must judge whether the feelings surrounding these events are intense enough to intrude upon the present. If so, psychotherapeutic help may be indicated. If not, although the memories may be painful, further emotional work on this area may not be beneficial. The treatment literature on the sequelae of abuse (and their treatment) is disappointing, especially when the abuse occurred in the distant past. In either case, it is difficult to judge whether these events are directly relevant to the present pain and hence demand attention, or whether they contribute to a psychologically vulnerable substrate acted upon by subsequent physical and emotional events. In these circumstances, it may be worthwhile to suggest further mental health evaluation as an exploratory measure, being careful not to imply that the patient is being referred because the physician is certain that the abuse is related to the development of the pain.

Sexual Dysfunction In clinical practice, women presenting with CPP often report a high incidence of marital distress and sexual dysfunction, particularly dyspareunia. Stout and Steege found that 56% of 220 married women scored in the maritally distressed range (less than 100) on the Locke-Wallace Marital Adjustment Scale at the time of initial visit. A high level of marital distress has also been reported in other chronic pain patients and their spouses. Although some women report satisfactory sexual functioning before the onset of pain symptoms, others appear to have long-standing impairments in sexual response.

DIAGNOSTIC STRATEGIES Part of "Chapter 27 - Persistent or Chronic Pelvic Pain"

Recognizing a Chronic Pain Syndrome Many women can experience pain for longer than 6 months without becoming debilitated; although their pain is chronic, such women are not described as having a CPS. The following are the common clinical hallmarks of true CPS: Duration of 6 months or longer Incomplete relief by most previous treatments Significantly impaired physical function at home or at work Signs of depression (sleep disturbance, weight loss, loss of appetite) Pain out of proportion to pathology Altered family roles Of the signs of depression, sleep disturbance is usually the first to appear. Careful questioning is needed to distinguish awakening caused by pain from awakening that just happens. In the true vegetative sign, the person usually cannot get back to sleep even if pain is relieved (by medication or other means). The alteration of family roles is perhaps the most important of those mentioned. This includes changed responsibilities for household, children, finances, and so forth. Initially intended as helpful, such changes may in the long run diminish the patient's self-esteem and progressively reduce her family's interactions with her to little more than checking on her pain. Over time, this covertly reinforces the complaint of pain and imparts to it unintended value as a major means of maintaining communication within the family.

Simultaneous Medical and Psychosocial Evaluation When the aforementioned markers of CPS are present, one should surrender the need to immediately discover how much of the pain problem is physical and how much is psychologic. Rather than guess, it is useful to ask two separate questions: Is there physical disease that requires medical or surgical treatment? Is there emotional or psychological distress that requires treatment? It is useful to directly state that the precise connection between these two cannot be measured; this can help diminish the patient's fear that she will be told “it's all in her head.” The patient can then be more open to sharing her personal and emotional concerns. If this statement is made early in the evaluation, before all physical evaluations have been carried out, the patient is likely to be less defensive. At this stage, a mental health consultant has a better chance of developing rapport with the patient and will be a more helpful collaborator when needed.

History Taking The site, duration, pattern during activities, relation to position changes, and association with bodily functions are all important elements of pain. For example, pain that is focal and positional is more often associated with adhesive disease; pain that is absent in the morning but worsens progressively during the day may be associated with pelvic congestion. The patterns of pain associated with endometriosis are discussed earlier. The chronology of the pain is critical. As CPS develops, pain can be present over a progressively larger area despite stable organic pathology. Interpreting this as the breakdown or wearing out of physiologic systems that deal with pain signals has some biologic validity and may make sense to the patient. From a cognitive perspective, it is invaluable to discern the patient's and her family's ideas about the causes of and future for her pain. Fears of cancer can be discovered even if this diagnosis was never even remotely considered by the clinician. Less dramatic, but equally powerful, attributions of cause can emerge, such as pelvic infection due to sexual acts remote in time, arguments with a spouse, divine retribution, and so forth.

Physical Examination Guiding a patient through contraction-relaxation sequences of the abdominal, thigh, and vaginal introital muscles can reduce the discomfort of the examination and can indicate the patient's degree of control over muscle tension. Single-digit palpation of the levator plate and piriformis muscles can elicit tenderness compatible with the label of pelvic floor tension myalgia. This condition is often present as a sequel to some other pelvic pain, but it can become a problem in itself. Discomfort is usually felt as pelvic pressure and radiation pain to the sacrum, near the insertions of the levator plate muscles. Adnexal thickening and mobility, pelvic relaxation, coccygeal tenderness, and foci of pain that reproduce dyspareunia should be noted. Gentle palpation with a cotton swab can detect areas of sensitivity compatible with vestibulitis in the introitus or trigger points higher in the vagina. On occasion, gentle fingertip palpation of the abdominal wall can detect such trigger points in the musculature.

Laboratory Tests

Imaging Studies In the case of CPS, it has already been established that intensity of pain does not correlate well with extent of organic pathology. It follows that if the physical examination is relatively benign and is not severely limited by body habitus, extensive imaging usually adds little to the database needed before laparoscopy is performed. This is especially true in the case of organ-specific studies (intravenous pyelography, barium enema, colonoscopy) in the absence of symptoms or signs pointing to a specific organ system (e.g., blood in the stools). If the patient has had multiple previous surgeries, high-resolution studies such as MRI and computed tomography scans are often misleading because of postoperative artifact. “Cystic masses” seen on such studies often prove to be nothing more than pockets of adhesions or peritoneal inclusion cysts.

Blood Studies Relatively few hematologic or chemical measures are of use in diagnosing CPP. An elevated leukocyte count and erythrocyte sedimentation rate may make the clinician suspect chronic pelvic inflammatory disease even when cervical cultures are negative for the most common sexually transmitted diseases. The cancer antigen-125 test is positive in advanced endometriosis but is not sufficiently sensitive to detect early stage disease or reliably monitor its response to treatment. If any remnant ovarian tissue is present, FSH and estradiol levels remain in premenopausal ranges in almost all instances. Replacement estrogen therapy should be withdrawn 3 weeks before these levels are measured.

Anesthetic Blocks Injection of small volumes of a local anesthetic, 1 to 5 mL of 1% lidocaine or 0.5% bupivacaine, blocks pain from either an entrapped segmental nerve (e.g., ilioinguinal) or an abdominal wall trigger point. In the latter case, such blocks can be therapeutic as well as diagnostic. Many anesthesia pain clinics administer epidural or spinal anesthetics to distinguish pain arising from peripheral organs from pain that has become completely central in origin. In some instances it is useful to attempt transvaginal blocks with the same local anesthetics. For example, the (posthysterectomy) vaginal apex may be focally intrinsically tender, suggesting a trigger point or other focal nerve root irritation. Again, local blocks can be therapeutic as well as diagnostic. A series of three or four blocks administered 1 to 2 weeks apart may give relief for months to years in some instances. In most cases, a history and careful routine physical examination distinguishes central from lateral sources of pelvic pain. When this discrimination is difficult, it may be useful to administer a transvaginal uterosacral block (blocking most uterine innervation) and then repeat the pelvic examination. When this relieves the pain, the pain can be assumed to arise from the uterus, but if the pain is not relieved, one cannot

distinguish a failed block from pain of nonuterine origin.

Psychologic Tests and Interviews To distinguish physical from psychologic causes for pain, many studies of CPP have used traditional psychologic instruments that were developed to measure general psychopathology or personality factors. In some studies, more abnormalities are detected in women without physical pathology at laparoscopy. In other papers, women with organic disease who have had pain for a long time appear equally distressed in their questionnaire responses. These psychometric instruments generally have little face value for chronic pain patients, and their use often confirms the patient's fears that the health care provider thinks she is “crazy” or that the pain is “all in her head.” Once again, the question of whether the emotional distress identified by these instruments is an antecedent to or a consequence of persistent pain remains unanswered. A more complete review of the psychometric instruments that have been used to study CPP is presented elsewhere. Psychometric tests are most useful when they are interpreted by a psychologist who has interviewed the patient, and they serve best as a means to better understand the patient's strengths and weaknesses, rather than as a means to decide who needs surgery.

Laparoscopy Great strides have been made in operative laparoscopy in the past 2 decades. New techniques and new terminology (e.g., pelviscopy) imply new “magic” to the physician and public alike. Laparoscopy should be liberally performed for diagnostic purposes, and ablation/excision of endometriosis and lysis of adhesions are no doubt useful procedures. However, the premature surgical procedure errs in the notion that pain is “hard-wired” to pelvic pathology. The available evidence clearly argues against this. When a CPS is clinically evident, results of laparoscopic treatment alone, despite comparable pathology, are much less impressive. For a patient with the clinical markers of CPS listed earlier, the complete workup as described should be performed before laparoscopy. In some puzzling cases in recent years, we have performed laparoscopy under local anesthesia to “pain map” the pelvis. A 2-mm laparoscope and a small suprapubic probe are placed with the use of short-acting intravenous analgesia (remifentanyl), and local lidocaine 1%. Having been oriented to the procedure beforehand, as each organ is touched, she is asked (1) “Does this give you the pain you get?” and (2) “Please rate the pain.” She responds with a number picked from a scale of 1 to 10, with 10 signifying the “worst pain you could imagine.” This technique may make it possible to determine whether visualized pathology is causing nociceptive signals or to discriminate visceral from somatic pain. It is possible in some cases to block the superior hypogastric plexus during pain mapping to better predict benefit from presacral neurectomy. In this approach, mapping is done before and after injecting 10 mL of 1% lidocaine just underneath the peritoneum over the sacrum, using a 7-inch, 22-gauge spinal needle.

MANAGEMENT Part of "Chapter 27 - Persistent or Chronic Pelvic Pain" Physicians use specific treatments in chronic pain depending on the model of pain perception that they follow. The surgically oriented gynecologist often tacitly follows the cartesian model, attempting to eliminate organic tissue damage to diminish pain proportionately. The behaviorist, cognitive therapist, and insight-oriented psychotherapist use approaches consistent with each one's basic therapeutic orientation, whereas advocates of the gate control theory use medications and other treatments that make sense based on that theory. The following represents an eclectic treatment approach. The treatment literature is considerably more sparse than is the literature exploring the etiology of CPP.

General Principles A complete evaluation of CPS often reveals a number of contributing factors, such as bladder irritability, irregular bowel function, poor posture, and emotional and relationship stresses, in addition to laparoscopically visualized pathology. Treating each component sequentially is common practice but often ends in frustration because each treatment addresses only a part of the problem. Simultaneous treatments often begin with disquieting multiple drug therapy but allow better relief. Close follow-up at regularly scheduled visits allows gradual tapering of medications over time. Planned visits also provide support and a coping mechanism for the patient. When the patient is essentially required to feel worse in order to be seen again, the pain may be tacitly reinforced.

Medication Use Analgesics Advocates of the operant conditioning model suggest that analgesics be taken continuously, in a non—pain-contingent fashion. This eliminates the need to demonstrate pain behaviors or voice pain complaints to justify the use of medication, thus eliminating the tendency of medication to act as a reinforcer of pain behaviors. This approach is benign enough when relatively nontoxic and nonaddicting medications such as acetaminophen are used, but it presents potential hazards when drugs such as the following are used: nonsteroidal antiinflammatory drugs (NSAIDs; gastric irritation, renal damage); aspirin or NSAIDs in combination with the milder narcotics, such as codeine, oxycodone, and pentazocine (constipation, sedation, habituation); and pure narcotics (addiction, diminished analgesic potency over time). When well tolerated, all three types of drugs can serve well in appropriate patients (without histories of substance abuse). Indeed, contrary to common perception, there is support for the notion that chronic low-dose opiate therapy may allow good return of function without adverse side effects in those who have failed intensive pain clinic treatments.

Antidepressants This class of drugs, particularly the tricyclic antidepressants, can potentiate the effects of analgesics in CPS, even when given at doses less than those usually used in the treatment of depression. New agents such as fluoxetine (Prozac) show promise and have a low level of side effects. Few controlled trials of antidepressants have been carried out in CPP patients.

Anxiolytics Anxiolytic drugs are certainly widely prescribed by gynecologists, although it is uncertain how often they are given for pain. In one study, alprazolam, a triazolobenzodiazepine with mixed anxiolytic and antidepressant effects, had a surprising degree of analgesic effect in moderate to high doses in patients with chronic pain of malignant origin and concomitant mood changes or anxiety. These patients were

already receiving narcotics, which may suggest that alprazolam potentiates the analgesic effect of narcotics. Their role in conjunction with nonnarcotic analgesics is uncertain, and the addiction potential is obvious.

Other Medications Nonanalgesic and nonpsychotropic drugs also have potential roles in the treatment of specific pelvic conditions. For example, medroxyprogesterone acetate treatment sufficient to suppress ovarian function may reduce the diameter of engorged pelvic veins and thus reduce the discomfort of pelvic congestion. However, the longest lasting relief was observed when psychotherapy was used as well. The use of GnRH agonists has been recommended to distinguish gynecologic from nongynecologic sources of pain; however, these agents also relieve the symptoms of IBS, probably by reducing serum relaxin levels. When the differential diagnosis includes ovarian remnant syndrome, residual ovary syndrome, or any other disorder influenced by the menstrual cycle, the impact of GnRH agonists on pelvic pain must be interpreted with caution in anyone with symptoms at all compatible with bowel dysfunction. In addition, pain threshold has been shown to be lower premenstrually, even in asymptomatic women. The impact of the menstrual cycle itself in chronic pain patients has not been well explored, but it seems likely that it may impart some cyclicity even to conditions unrelated to the reproductive tract. Cyclicity of symptoms must therefore be interpreted with caution, and the obliteration of symptoms or of their cyclicity by pharmacologically obliterating the menstrual cycle does not demonstrate a gynecologic cause. To address the most common clinical circumstance: relieving pain with a GnRH agonist does not prove that the pain is due to endometriosis.

Surgery Two basic surgical approaches have been used to treat CPP: removing pelvic organs and treating visible disease while leaving the pelvic organs in place. Only the former approach has been evaluated for efficacy to any degree. In the United States, about 12% of hysterectomies are performed with pelvic pain as the primary indication. An additional 6.1% are performed for endometriosis or adenomyosis, and 5.1% are performed for pelvic inflammatory disease; no doubt many in these two categories also involve complaints of pain. In about one third of hysterectomies performed for pain, no pathology is found. Despite the frequency of pain as an indication for this procedure, data regarding efficacy are surprisingly sparse. One report notes relief in 78% of women after hysterectomy for pelvic pain of uterine etiology (women with adnexal or other pelvic disease were excluded). However, the presence or absence of uterine pathology (adenomyosis or leiomyomata) had no bearing on whether or not pain was relieved. It cannot be determined from the report (Table 27.1) whether failure to obtain relief was due to the presence of other pelvic conditions, postoperative adhesion formation, or psychologic reasons. Symptom substitution was not evaluated in the report. Relief of pain by removal of the normal uterus is even more puzzling; apparently this was not accounted for by concomitant procedures performed for deficient pelvic support. Some of these cases may have involved pelvic congestion or perhaps other mechanisms even less well understood. A 22% failure rate emphasizes the need for careful preoperative evaluation of all potentially contributing factors, both physical and emotional. According to two recent reports, hysterectomy performed in primary care settings was very effective for the treatment of CPP. In a prospective observational study of private practices in Maine, Carlson and associates reported that at a 1-year follow-up, satisfaction with the outcome of surgical treatment was much higher than satisfaction with the outcome of medical therapy. However, about one third of women improved substantially on medical therapy, and perhaps women were more likely to undergo operation when organic pathology was demonstrable. In the Maryland Women's Health Study, 1,299 women were interviewed at length before hysterectomy for benign disease and at 3, 6, 12, and 24 months after surgery. In more than 90% of cases, the procedure was well tolerated and did not result in postoperative depression or a decline in sexual functioning. In the subset of women with pain as the primary indication for surgery, relief of pain occurred in more than 80%, indicating that the clinicians involved generally employed good judgment and technique. In general, women with preoperative depression or sexual dysfunction did not fare as well as their less symptomatic counterparts. Further analyses of these data are underway.

Adnexal and other intrapelvic diseases, usually endometriosis or adhesions due to either postoperative changes or chronic pelvic inflammatory disease, have been treated by both laparotomy and laparoscopy. Use of the CO2 laser during laparotomy does not improve the results of standard infertility surgical techniques as evaluated by second-look laparoscopy and by pregnancy rate. Laparoscopy is probably superior to laparotomy for treating adhesive disease. In the rabbit model, infliction of injuries with the laser by way of laparotomy resulted in adhesion formation, but no adhesions formed when identical damage was caused by laparoscopy. In the rabbit and the human, adhesiolysis is more effective with laparoscopy than with laparotomy. However, none of the large clinical trials of laparoscopic adhesiolysis evaluate the effects of the procedures on pain. Although second-look verification of effective adhesiolysis is lacking, results of several studies on the treatment of pelvic adhesions for the relief of pain are encouraging. Even when treated by laparotomy, with the use of infertility techniques, 28 of 42 (65%) patients reported cure or improvement of pain. In a sample of mostly primary care patients, 84% of 65 patients had relief of pain after laser laparoscopic adhesiolysis with follow-up intervals of 1 to 5 years. In a sample of 42 patients with more severe adhesive disease, Daniell reported improvement in 67% at a 4-month follow-up. In Sutton's large series, 85% had pain relief at 1 year. Steege and Stout reported that 15 of 20 (75%) patients without a CPS who were undergoing laser laparoscopic adhesiolysis had good relief of pain at a follow-up 6 to 12 months after surgery. However, if a CPS was present, only 4 of 10 (40%) patients with equivalent adhesive disease obtained relief. The greater the emotional and behavioral disability, the greater the need for combined medical, surgical, and mental health management. Several studies support an association between pelvic pain and endometriosis, although a strict quantitative relation is lacking. In terms of surgical treatment of pain from endometriosis, laser ablation results in relief of pain in about 60% of women at 1 year follow-up, compared with 96% treated by laparoscopic excision of the disease. A randomized trial comparing these two methods has not been done. The benefits of surgical excision may be prolonged by subsequent medical therapy; studies comparing the relative benefits of various medical therapies (continuous oral contraceptives, continuous progestins, GnRH agonist with/without add-back estrogens) following surgical treatment have not been performed. The more economical and less physiologically intrusive approach would seem to favor sex steroids over GnRH agonists. Presacral neurectomy, as an adjunct to surgical excision of endometriosis, has been evaluated for its effect on pelvic pain. In a retrospective sample of 71 women undergoing conservative resection of endometriosis by way of laparotomy, 35 (50%) who also had presacral neurectomy enjoyed significantly greater improvement in both dysmenorrhea and dyspareunia, with 75% to 95% obtaining improvement. Two subsequent retrospective reports noted that similar percentages (about 75%) of women obtained pain relief after endometriosis surgery that included presacral neurectomy, compared with about 25% who obtained relief without neurectomy. Two studies on the treatment of presacral neurectomy as adjunctive therapy along with conservative resection for stage III-IV endometriosis reached different conclusions about the additive value of presacral neurectomy. Tjaden and associates performed conservative resection by way of laparotomy and randomized eight women to receive or not receive presacral neurectomy. All the neurectomized women improved dramatically, whereas the others did not. The institutional review board mandated stopping the study because statistical significance had been reached. The authors then reported an open clinical series in which 15 of 17 women improved substantially after conservative resection with presacral neurectomy. Candiani and colleagues reached a different conclusion after randomizing 71 women to receive or not receive presacral neurectomy along with conservative resection of advanced endometriosis. They found that central dysmenorrhea improved slightly more in neurectomized women, but daily pain and dyspareunia were not improved by adding presacral neurectomy to the procedure. The literature is therefore inconclusive regarding the value of the neurectomy procedure. Laparoscopic treatment of even severe endometriosis has a growing number of advocates. Again, the focus has been on fertility rates, but one prospective report notes prolonged pain relief in two thirds of patients. The ability to perform presacral neurectomy by way of laparoscopy should not provoke wide adoption of the procedure without careful consideration of these data. An ovarian remnant should be removed if it is persistently symptomatic despite all reasonable attempts at medical suppression, and if menopause cannot be expected in the patient's near future. Whether performed by laparoscopy or laparotomy, the dissection should be detailed and should include all the peritoneum surrounding the mass. The ureter and pelvic side wall vessels should be exposed and carefully freed from the specimen. When a GnRH agonist has been used preoperatively for symptom control or to distinguish the relative contributions made by the remnant and other pelvic pathology, such as

adhesions, the remnant tissue may become so small as to make it difficult to identify. Hence, if a palpable (or ultrasonically visible) mass disappears after GnRH agonist treatment, it may be wise to allow time for it to regrow before pursuing surgical excision. When the remnant is small, some surgeons have stimulated the remnant with clomiphene citrate to make it easier to find. Finally, the peritoneal windows syndrome must be mentioned. Openings in the peritoneum covering the posterior broad ligament and the cul-de-sac of Douglas, or peritoneal “windows,” have been noted since the 1950s and were associated with endometriosis by Chatman. Excision and closure of the windows have been anecdotally reported, but the value of these operations in treating pain associated with the disease remains to be demonstrated. It is this author's practice to excise the peritoneum involved with the window, but not to stitch the area closed. Microscopic endometriosis is found in the vast majority of cases.

Alternative Treatments Biofeedback, transcutaneous electric nerve stimulation units, relaxation training, and individual and couples counseling all have their appropriate roles in individual cases, but none is so clearly applicable or effective that its automatic use is supported in cases of CPP. In keeping with the approach outlined in the earlier discussion of medications, problems requiring counseling should be addressed but should be treated as issues separate from the discussion of appropriate surgical approaches. Cause-and-effect relations are difficult to demonstrate even in retrospect when psychologic assistance works.

Management Overview The most effective clinical approach requires simultaneous treatment of as many factors as possible: anatomic, musculoskeletal, functional bowel and bladder, psychologic, and so forth. Patient and physician must contract for the long term and work from a rehabilitation perspective, rather than hope that the latest single addition to the treatment will prove to be the answer. The physician, to prevent frustration and feelings of defeat, must often play the role of helping to manage and relieve the pain while helping to maximize function, even when pain persists. To the surgically trained gynecologist who prefers a clear-cut single answer to a clinical problem, this can be the most difficult part of dealing with the problem of CPP.

THE EVOLUTION OF A CHRONIC PAIN SYNDROME Part of "Chapter 27 - Persistent or Chronic Pelvic Pain" As is apparent from this discussion, CPP is a heterogeneous problem, not a single diagnosis, and no single etiologic hypothesis is clearly supported. Most of the hypotheses reviewed here have some credible evidence supporting them; none have been sufficiently validated. Psychologic and neurologic mechanisms are proposed here to explain how the evolution of chronic pain may occur, regardless of the particular tissue damage or functional disorder that may first have provided nociceptive stimuli. We suggest the following elements (Fig. 27.1): biologic events sufficient to initiate nociception, alterations of lifestyles and relations over time, anxiety and affective disorders, and a circular interaction (vicious cycle) among these elements.

Biologic Events Sufficient to Initiate Nociception Sexually transmitted diseases, endometriosis, recurrent bladder and vaginal infections, primary or secondary functional dyspareunia, alterations of bowel habit, pelvic congestion, and gynecologic or other abdominal surgeries (Table 27.1) may contribute individually or in combination.

Alteration of Lifestyle and Relations Physical activities at home and recreational pursuits can suffer. Believing that rest usually helps in the treatment of most causes of acute pain, the patient may assume that the same applies to chronic pain and may thus restrict herself more than actual discomfort dictates. Family members start to regard the patient as sick and leave her out of many activities, thus reducing her roles within the family structure. With time, concern for and discussion of her pain can become the family's major pattern of communication with the pain victim. If sexual intimacy has been the major means of emotional sharing and smoothing over of differences and if this intimacy is reduced, then the altered pattern of interactions may take hold more quickly.

Anxiety and Affective Disorders Depression can occur as a cumulative result of the disability suffered, or the pain can bring on an episode of depression in a patient already biologically vulnerable. The observation most relevant here is that pain patients with a family history of depression can derive the most benefit from tricyclic antidepressants.

The Vicious Cycle Diminished activity, altered family roles and social supports, anxiety, and affective disturbances can influence nociception by a variety of central pathways, ultimately altering spinal cord “gating” of nociceptive signals. Cognitions about the pain can play an additional role. Several important modifying influences can be present in addition to these major pathways (Fig. 27.2). Incest and other forms of sexual abuse have attracted the most attention as possible forerunners of CPP. However, CPP is clearly not a unique or specific sequel to sexual abuse, and a large proportion of CPP patients have not been abused in this manner. Victims of sexual abuse have many negative emotional sequelae; pain problems often occur after abuse, but they are not necessarily directly caused by the abuse. Sexual abuse perpetrated by a family member is a clear indication of a psychologically detrimental environment of rearing that might have led to disturbances in character and personality development perhaps equal in importance to the terrible trauma of the abuse itself. Learning more about the factors that can mitigate the impact of sexual abuse in general will facilitate our understanding of the psychologic vulnerabilities that can make a particular person more susceptible to development of CPP or other chronic pain disorders. A genetic predisposition to depression also allows the vicious cycle to become easily established and

strengthened over time. Antidepressant medications play an important role in the overall therapeutic plan in such cases. Several authors have suggested that the concept of perceived control best explains the development of affective changes accompanying chronic pain, regardless of the location of the pain. The individual who sees herself as having little control over the physical and emotional events affecting her may be most vulnerable to development of a CPS. It may be reasonable to consider this variable as a culmination of the effects of affective change, activity, family roles, sexual dysfunction, and previous victimization experiences. The longer that pain has been a part of the person's life, and the more psychologic vulnerabilities she carries forward to the present, the less likely it is that any treatment of the tissue damage itself will be effective in relieving pain and restoring physical and emotional function. However, as treatment studies show, the organic contribution to chronic pain can seldom be dismissed entirely. The more difficult task is the selection of an efficient and cost-effective combination of treatment approaches aimed at the most important factors acting in the present.

Chapter 28

Pelvic Inflammatory Disease Mark G. Martens Pelvic inflammatory disease (PID) is one of the most serious infections facing women today. Untreated or unsuccessfully treated women may suffer life-threatening consequences, and even adequately treated women are at much higher risk for potentially serious sequelae. PID is a spectrum of diseases initially involving the cervix, uterus, and fallopian tubes. Acute PID, the acute clinical syndrome, is most often attributed to an ascending spread of microorganisms from the vagina and endocervix to the endometrium, fallopian tubes, and contiguous structures. The terms acute PID and acute salpingitis are often used interchangeably, but PID is not limited to tubal infection only. Recently, a more descriptive term to differentiate the severity and extent of various forms of PID was introduced by Hemsell and colleagues, termed upper genital tract infection (UGTI). This is differentiated from lower genital tract infection (LGTI), because response to treatment appears to be different in these two entities, which were previously grouped as inpatient or outpatient treatment regimens, sometimes at a conflict with diagnostic or severity of illness realities. Sexually transmitted diseases (STDs) have been reported at epidemic proportions in the United States. Since the mid-1960s, the incidence of gonorrhea and chlamydia infections has been estimated at more than 2 million and 4 million cases, respectively, each year. For women, acute PID is the most common and important complication of STDs. Bell and Holmes estimated that 1 million women a year are treated for acute salpingitis in the United States. About 250,000 to 300,000 women are hospitalized each year with a diagnosis of salpingitis or PID. The disease generates nearly 2.5 million visits to physicians, and an estimated 150,000 surgical procedures are performed for complications every year. According to Westrom, the direct and indirect costs of PID and its sequelae totaled 4 billion dollars in the United States this past decade. In terms of overall incidence, acute PID occurs in about 1% to 2% of young, sexually active women each year. PID is the most common serious infection in women age 16 to 25 years, and the resultant morbidity exceeds that produced by all other infections combined for this age group. ETIOLOGY RISK FACTORS DIAGNOSIS SEQUELAE MORTALITY TREATMENT MANAGEMENT OF SEX PARTNERS HIV INFECTION SURGICAL MANAGEMENT POSTERIOR COLPOTOMY RUPTURED PELVIC ABSCESS PRIMARY OVARIAN ABSCESS SURGERY FOR CHRONIC PELVIC INFLAMMATORY DISEASE PELVIC TUBERCULOSIS

ETIOLOGY Part of "Chapter 28 - Pelvic Inflammatory Disease" Although certain geographic areas with near epidemic rates of STDs have Neisseria gonorrhoeae as a common cause of PID, many cases of acute PID are the result of a polymicrobial infection caused by organisms ascending from the vagina and cervix to infect the mucosa of the endometrium and fallopian tube. About 85% of cases are naturally occurring infections in sexually active females of reproductive age. The remaining 15% of infections occur after procedures that break the cervical mucous barrier, such as placement of an intrauterine device (IUD), endometrial biopsy, or uterine curettage, which allow the vaginal flora to colonize the upper genital tract. In the United States, nontuberculous acute PID was traditionally separated into gonococcal and nongonococcal disease, depending on the isolation of N. gonorrhoeae from the endocervix. A variety of organisms can be isolated from the endocervix, and it is difficult to determine which of these organisms are PID-related and which are normal cervicovaginal flora. Upper genital tract organisms are probably more indicative of the causative organisms but are difficult to obtain. Bacterial organisms cultured directly from tubal fluid commonly include N. gonorrhoeae, Chlamydia trachomatis, endogenous aerobic and anaerobic bacteria, and genital Mycoplasma species. Laparoscopic studies have demonstrated a correlation of no more than 50% between endocervical and tubal cultures, but the presence of N. gonorrhoeae is usually considered an important causative factor. However, endocervical gonorrhea does not necessarily indicate its sole pathogenic nature in all cases. Direct fallopian tube cultures have demonstrated that tubal infections are often polymicrobial. The type and number of species vary depending on the stage of the disease. Gonorrhea, for example, is often cultured from the cervix during the first 24 to 48 hours of the disease but is often absent later. Similarly, fewer organisms are cultured late in the disease, and anaerobic bacteria such as Prevotella, Bacteroides, Peptococcus, and Peptostreptococcus species tend to predominate. Whether these anaerobes play a causative role or increase in number and frequency as a result of the acute inflammatory response is uncertain. Sweet has summarized the literature by stating that in approximately one third of women with PID, N. gonorrhoeae is the only organism recovered by direct tubal or cul-de-sac culture. One third have a culture positive for N. gonorrhoeae plus a mixture of endogenous aerobic and anaerobic flora, and the remaining one third have only aerobic and anaerobic organisms. Chow and colleagues and Monif and colleagues have postulated that the gonococcus initiates acute PID and produces tissue damage. This damage changes the local environment, which in turn allows anaerobic and aerobic organisms from the vaginal and cervical flora to invade the upper genital tract. Eschenbach and Sweet have suggested that not all PID follows gonococcal infection, and that acute PID initially may also solely have a polymicrobial etiology. According to Sweet and Gibbs, about 20% of all women with salpingitis have tubal cultures positive for C. trachomatis. N. gonorrhoeae and C. trachomatis are found in the same individual 25% to 40% of the time. Scandinavian studies by Eilard and co-workers have reported the recovery of C. trachomatis from the cervix in 22% to 47% of women with acute PID. C. trachomatis by itself produces a mild form of salpingitis with an insidious onset. In contrast to gonorrhea, Chlamydia can remain in the fallopian tubes for months or years after initial colonization of the upper genital tract. Svensson and colleagues found that women with C. trachomatis infection at laparoscopy had the most severe fallopian tube involvement, probably because of its clinically silent or minimally symptomatic nature, which results in difficult or delayed diagnosis and therefore delayed or absent treatment. The two major sequelae of acute PID are tubal infertility and ectopic pregnancy. These have been strongly associated with prior chlamydial infection as a consequence of intratubal and peritubal adhesions. Although C. trachomatis is generally believed to be one of the most common causes of PID, along with N. gonorrhoeae, its etiologic role is very different. N. gonorrhoeae is a gram-negative diplococcus with rapid growth due to a growth cycle of about 20 to 40 minutes. This results in a great increase in the number of organisms once N. gonorrhoeae reaches an area such as the endometrium or fallopian tube, where growth is relatively unimpeded. This rapid increase in the number of gram-negative bacteria usually results in a rapid and intense inflammatory response by the woman's host defenses. The response to this rapid bacterial growth is proliferation and aggregation of white blood cells and their inflammatory products. Migration of this bacterial and leukocytic mixture through the fallopian tube to the ovary and peritoneal cavity, and back to the cervix and vagina, causes the symptoms that are pathognomonic of acute PID. C. trachomatis, however, is a slow-growing intracellular organism. Its lack of mitochondria results in its obligatory intracellular existence and also causes its growth cycle to be extremely slow compared with N. gonorrhoeae and nonintracellular microorganisms. The growth cycle of Chlamydia is 48 to 72 hours; therefore, several weeks to months are required for the growth to reach numbers sufficient to cause acute symptoms, if at all. Its slow growth does not induce a rapid or violent inflammatory response. This explains

the slow and insidious nature of the symptoms of acute C. trachomatis infections. However, because of its intracellular growth cycle, the release of the elementary bodies (its infectious vehicle) occurs by rupture of the cell that it has invaded. The repeated occurrence of elementary body infection of susceptible cells, and their subsequent destruction by rupture, is the major mechanism by which C. trachomatis causes disease in acute pelvic infections. Also, because of its slow growth and lack of acute inflammatory response and clinical symptoms, treatment is often delayed or not started at all, adding to the extended tissue destruction and PID sequelae. The lack of acute symptoms does not lessen the importance of Chlamydia as a PID pathogen. Not only does the tissue destruction result in severe complications such as ectopic pregnancy and infertility, but also the tissue damage provides fertile ground for the growth of secondarily infecting aerobic and anaerobic bacteria. This necrotic tissue is an excellent growth medium, and the epithelial damage enhances the breakdown of the surface defense mechanisms. The importance of Chlamy-dia was documented during the 1980s when treatment of acute PID was initially believed to be successfully accomplished with regimens not active against C. trachomatis. However, although success was evident with short-term follow-up, long-term follow-up demonstrated that treatment of C.trachomatis was necessary. PID regimens without Chlamydia coverage resulted in an increased incidence of long-term complications such as abscesses and chronic pelvic pain, with resultant increased surgical intervention. Therefore, modern treatment of PID includes C. trachomatis coverage, even though it may not be the cause of the acute symptoms. Nongonococcal infections are believed to be the result of acute bacterial infections or possibly a preceding or previously treated C. trachomatis UGTI. This is confirmed by the high incidence of Chlamydia antibodies in patients with acute PID, ectopic pregnancy, and infertility. In addition to N. gonorrhoeae, Chlamydia, and aerobic and anaerobic bacteria, other microorganisms have been implicated as etiologic agents in acute salpingitis. The genital tract mycoplasmas, Mycoplasma hominis and Ureaplasma urealyticum, have also been suggested as causal agents in acute salpingitis. However, their role remains controversial. Cervical cultures positive for both M. hominis and U. urealyticum have been recovered from women with PID. However, the rate of isolation is about 75%, which is not statistically different from that of women who are sexually active but without PID (baseline rate of about 50%), as found by Lemeke and Lsonka.

RISK FACTORS Part of "Chapter 28 - Pelvic Inflammatory Disease" Several factors that predispose to the development of acute PID have been identified. Risk factors are important considerations in both the clinical management and prevention of UGTIs. There is a strong correlation between exposure to STDs and PID. In the United States, recent studies have confirmed this association with the recovery of N. gonorrhoeae or C. trachomatis in about 50% of patients hospitalized with acute PID. Age at first intercourse, frequency of intercourse, number of sexual partners, and marital status are all associated with the frequency of exposure to STDs and thus are associated with PID. Women with multiple partners have an increased risk (four to six times normal) for development of acute salpingitis, compared with women who have monogamous sexual relations. The incidence of acute PID decreases with advancing age. Adolescent females are at significant risk for development of acute salpingitis. Westrom reported that nearly 70% of females with PID were younger than 25 years of age, 33% experienced their first infection before the age of 19, and 75% were nulliparous. The risk for development of acute PID in a sexually active adolescent female patient was 1:8, whereas the risk was 1:80 for a sexually active woman 24 years of age or older. Several reasons have been suggested for this increased risk. The two microorganisms most commonly considered to be the inciting agents in cases of PID, N. gonorrhoeae and C. trachomatis, have a predilection for columnar epithelium. As suggested by Schaefer and by Sweet and colleagues, cervical columnar epithelium is exposed to a greater extent in younger individuals, and recedes into the cervical canal with increasing age. Clinical and laboratory studies have documented that the use of contraceptives changes the relative risk for development of PID. Multiple case-controlled studies have shown an increased risk of acute PID in women who wear an IUD. It has been estimated that IUD users have a threefold to fivefold increased risk for development of acute PID. In a report by Tatum and colleagues of animal model investigations, it was suggested that multifilament strings may be a major contributing factor in the increased risk of PID. Barrier methods of contraception (condoms, diaphragms, and spermicidal preparations) are effective both as mechanical obstructive devices and as chemical barriers. A nearly 60% decrease in the risk of PID has been demonstrated among women using a barrier method of contraception. Nonoxynol 9, the material in spermicidal preparations, is both bactericidal and viricidal. Laboratory tests have demonstrated that nonoxynol 9 kills N. gonorrhoeae, genital Mycoplasma species, Trichomonas vaginalis, Treponema pallidum, herpes simplex virus, and human immunodeficiency virus (HIV), but has been removed from many vaginal products due to safety warnings. Oral contraceptives have also been shown to reduce the risk of acute PID. The mechanism for such protection remains speculative. The thicker cervical mucus produced by the progestin component of oral contraceptives is believed to inhibit sperm and bacterial penetration into the upper genital tract. The decrease in duration of menstrual flow accompanying oral contraceptive use theoretically creates a shorter interval for bacterial colonization. Svensson and co-workers reported that, in addition to protecting against PID, the use of oral contraceptive pills was associated with a better prognosis for future fertility than was seen in women with acute PID using other contraceptive methods or no contraceptive methods. Surgical procedures of the female genital tract also place the patient at risk for PID. About 15% of pelvic infections occur after procedures that break the cervical mucous barrier, allowing for colonization of the upper genital tract. Eschenbach and Holmes reported that these procedures include endometrial biopsy, curettage, IUD insertion, hysteroscopy, and hysterosalpingography. The incidence of UGTI associated with first-trimester abortions is about 1 in 200 cases. Recent practice has emphasized the use of prophylactic antibiotics in high-risk cases to attempt to decrease the incidence of iatrogenic acute PID. Acute salpingitis occurring in a woman with a previous tubal ligation was once believed to be rare. Phillips and D'Abling reported that acute PID developed in the proximal stump of previously ligated fallopian tubes in 1 of 450 women hospitalized for acute salpingitis. However, many cases may be undiagnosed because of the absence of peritoneal signs. Previous acute PID is also a risk factor for future episodes of the disease. Another acute tubal infection develops in about 25% of women who have had acute PID. The exact mechanism for this increased susceptibility has not been determined, but it may be loss of the natural protective mechanisms of the fallopian tube against microorganisms. This increased risk may be related to the sexual habits of the woman involved, such as reinfection from an untreated male partner or genital tract damage from the initial infection. Eschenbach has documented that more than 80% of male contacts are not treated.

DIAGNOSIS Part of "Chapter 28 - Pelvic Inflammatory Disease" Acute PID presents with a broad spectrum of clinical symptoms. The differential diagnosis of acute PID includes acute appendicitis, endometriosis, torsion or rupture of an adnexal mass, ectopic pregnancy, and lower genital tract infection. Common clinical manifestations include lower abdominal pain, cervical motion tenderness, and adnexal tenderness and may include fever, cervical discharge, and leukocytosis. Historically, the diagnosis of acute PID was not established unless the patient had the triad of lower abdominal and pelvic pain, fever, and leukocytosis. Jacobson and Westrom have shown that all three are present in only 15% to 30% of actual PID cases. In addition, about 50% of patients initially present with a normal temperature and white blood cell (WBC) count. Pain in the lower abdomen and pelvis is by far the most common symptom of acute PID. It occurs in more than 90% of patients at initial presentation. The pain is usually described as constant and dull and is accentuated by motion and sexual activity. Generally, the pain is of recent onset, usually less than 7 days. About 75% of patients with PID have an associated endocervical infection and coexistent purulent vaginal discharge. Nausea and vomiting are comparably late symptoms in the course of the disease. Abnormal vaginal bleeding, especially menorrhagia, or spotting, is noted in about 40% of patients. The Centers for Disease Control and Prevention (CDC) have established the criteria for making the diagnosis of salpingitis based on clinical grounds. The most recent update includes only selected tender signs as required for a diagnosis and eliminates leukocytosis and fever as essential criteria (Table 28.1). TABLE 28.1. Criteria for the Diagnosis of Acute Salpingitis Minimum Criteria Empirical treatment of PID should be initiated in sexually active young women and others at risk for STDs if the following minimum criteria are present and no other cause for the illness can be identified: Uterine/adnexal tenderness or Cervical motion tenderness More Elaborate Criteria More elaborate diagnostic evaluation often is needed, because incorrect diagnosis and management might cause unnecessary morbidity. These additional criteria may be used to enhance the specificity of the minimum criteria. Additional criteria that support a diagnosis of PID include the following: Routine Criteria for Diagnosing PID: Oral temperature >38.3°C (>101°F) Abnormal cervical or vaginal mucopurulent discharge Elevated erythrocyte sedimentation rate Elevated C-reactive protein Laboratory documentation of cervical infection with Neisseria gonorrhoeae or Chlamydia trachomatis Presence of WBCs on saline microscopy of vaginal secretions Specific Criteria for Diagnosing PID: Histopathologic evidence of endometritis on endometrial biopsy Transvaginal sonography or MRI scan showing thickened fluid-filled tubes with or without free pelvic fluid or tuboovarian complex Laparoscopic abnormalities consistent with PID MRI, magnetic resonance imaging; PID, pelvic inflammatory disease; STD, sexually transmitted disease. to-T1-28to-T1-28 Perihepatic inflammation and adhesions, more commonly known as the Fitz-Hugh-Curtis syndrome, develop in 1% to 10% of patients with acute PID. Signs and symptoms include right upper quadrant pain, pleuritic pain, and tenderness in the right upper quadrant when the liver is palpated. Usually the symptoms and signs of this syndrome are preceded by the clinical onset of acute PID. The condition is often mistakenly diagnosed as either acute cholecystitis or pneumonia. Fitz-Hugh-Curtis syndrome is believed to develop from vascular or transperitoneal dissemination of either N. gonorrhoeae or C. trachomatis to produce the perihepatic inflammation. Other organisms may be involved, but limited data exist on their causality. Jacobson and Westrom attempted to correlate the clinical diagnosis of acute salpingitis with laparoscopic visualization. Of 814 women in whom laparoscopy was performed for presumed acute PID, 512 (65%) had

visual evidence of salpingitis; 184 (23%) had normal visual findings; and 98 (12%) had other pelvic pathology. Because of the positive clinical findings, many of the patients with normal findings were suspected to have early PID with endometritis and endosalpingitis without visual evidence of tubal or pelvic damage. Thus, laparoscopy is limited as a method of diagnosing the early stages of PID, but it is important to rule out non-PID surgical emergencies such as appendicitis, and other entities requiring different treatment modalities, such as endometriosis. Despite these shortcomings of early diagnosis, laparoscopic visualization of the pelvis is still the most accurate method of confirming the diagnosis of acute PID. However, it is logistically and economically impractical for all patients suspected of having acute PID to undergo diagnostic laparoscopy in the United States. Therefore, the diagnosis of most episodes of acute PID is often made on the basis of clinical history and physical examination. Although it is suggested that laparoscopy be offered to all patients with an uncertain diagnosis, it is strongly indicated for patients who are not responding to therapy, in an effort to confirm the diagnosis, obtain cultures from the cul-de-sac or fallopian tubes, and drain pus if necessary. In summary, laparoscopic studies have shown the following: The clinical diagnosis of acute PID may be inaccurate. Acute PID is sometimes found in patients undergoing laparoscopy for other causes of pelvic pain. Laparoscopy is a relatively safe method for making the visual diagnosis of the latter stages of PID, and thus assessing future fertility prognosis and planning. Laparoscopy is an excellent means of obtaining cultures directly from the tube. The appearance of the pelvic organs can vary from erythematous, indurated, edematous oviducts, to pockets of purulent material, to a large pyosalpinx or tuboovarian abscess. However, although no disease may be evident in early stages, it is imperative to render treatment to all stages to avoid long-term sequelae. Other less invasive methods of diagnosis have been suggested for verifying a clinical diagnosis of acute PID. Endometrial biopsy is one alternative to laparoscopy. Paavonen and associates reported a 90% correlation between histologic endometritis and laparoscopically confirmed salpingitis. However, results may be delayed up to 2 to 3 days, making its clinical applicability limited. Ultrasonography is of limited value for patients with mild or moderate pelvic PID. Thus, the routine use of sonography in patients with acute salpingitis does not appear to be indicated. Ultrasound is helpful in distinguishing an adnexal mass, especially in patients who demonstrate a lack of response to antimicrobial therapy in the initial 48 to 72 hours of therapy. Sonohysterography, an ultrasound examination using the instillation of saline to better define pelvic structures, is not indicated at this time for patients suspected of having PID, because no studies have been performed to demonstrate its safety in the event that pathogens are dispersed into the upper genital tract in the process of instilling the saline. Culdocentesis, with evidence of purulent peritoneal fluid, is helpful in the diagnosis of acute PID. With acute PID, the WBC count of peritoneal fluid is greater than 30,000 cells/mL, compared with a WBC count of 1,000 cells/mL in women without peritoneal inflammation. However, other infections, such as appendicitis and diverticulitis, among others, can also cause purulent pelvic fluid and a false diagnosis of PID. Laboratory tests can be obtained, but their results lack sufficient sensitivity and specificity to make them an important factor in establishing the diagnosis. Leukocytosis is not a reliable indicator of acute PID, nor does it accurately correlate with the severity of tubal inflammation or need for hospitalization. Less than 50% of women with acute PID have a WBC count greater than 10,000 cells/mL. Similarly, the erythrocyte sedimentation rate (ESR), which for years was a laboratory test for women with acute PID, is nonspecific and is a crude indicator of severity of disease. The ESR is elevated higher than 15 mm/hr in about 75% of women with laparoscopically confirmed acute salpingitis. However, 53% of women with pelvic pain and normal-appearing pelvic organs have an elevated ESR. Plasma proteins, such asC-reactive protein and antichymotrypsin, have been studied to determine whether they help in the diagnosis of acute PID. They have been found to be more sensitive than the ESR. Other investigators have found that decreased or absent isoamylase in peritoneal fluid in cases of acute PID is the best nonculture laboratory test for the disease. The major disadvantages of this test are that it requires several hours to complete and that peritoneal fluid must be obtained. Other evaluation has revealed various inflammatory cytokines to be associated with pelvic infections; however, these tests are not commercially available to a useful extent. Because most cases of UGTI are associated with, and preceded by, lower genital tract infection, examination of the endocervix for inflammation, Gram stain, and culture for both N. gonorrhoeae and C. trachomatis are all important for proper evaluation. A negative Gram-stained smear of the endocervix does not rule out upper tract infection. However, other studies have found that acute PID is rare without a concomitant increase in inflammatory cells in the vagina and the cervix.

SEQUELAE Part of "Chapter 28 - Pelvic Inflammatory Disease"

Infertility One fourth of all women who have had acute salpingitis experience one or more long-term sequelae. The most common is involuntary infertility, which occurs in about 20% of patients. PID ranks as one of the major causes of infertility. Before antibiotic therapy, 50% to 70% of women who had experienced UGTIs were sterile. The sequelae of infections vary from a patent oviduct, to peritubular and periovarian adhesions that may interfere with ovum pickup, to complete tubal obstruction. The infertility rate increases directly with the number of episodes of acute pelvic infection. Also, women with mild disease are seven times less likely to suffer tubal obstruction than women with severe PID.

Ectopic Pregnancy The number of ectopic pregnancies has doubled over the past 10 years. This increased rate is directly proportional to the increase in cases of STD and acute PID. The chance of ectopic pregnancy is increased sixfold to 10-fold in patients with a previous episode of acute salpingitis. Pathologic studies estimate that at least 50% of ectopic pregnancies occur in fallopian tubes damaged by previous salpingitis. The mechanism for the increased rate is believed to be interference of ovum transport through the tube or entrapment of the ovum secondary to microscopic tubal damage.

Chronic Pelvic Pain The chance that chronic pelvic pain will develop in a woman after acute salpingitis is four times that of control subjects without pelvic infection (20% versus 5%). Chronic pelvic pain can be caused by a hydrosalpinx. A hydrosalpinx is presumably the end-stage development of a pyosalpinx. The pain can also be related to adhesions surrounding the ovary. All patients with chronic pelvic pain believed to be caused by acute PID should undergo laparoscopy or laparotomy to establish the cause of the chronic pain and rule out other diseases such as endometriosis, which require different treatment. A tuboovarian complex is a collection of pus within an anatomic space created by adherence to adjacent organs. The incidence of true adnexal abscess is about 10% in women with acute PID. Landers and Sweet noted a 20% rate of early treatment failure (after 48 to 72 hours) of antibiotic therapy as a result of persistent pain or enlargement of a tuboovarian abscess or complex. In addition, according to Landers and Sweet, 31% required an operation several weeks to months after their acute infections for persistent disease or pain.

MORTALITY Part of "Chapter 28 - Pelvic Inflammatory Disease"

Before antibiotic therapy, the mortality rate associated with acute PID was 1%. Most of these deaths resulted from rupture of tuboovarian abscesses. Today, death associated with PID is rare, but the mortality rate can still be as high as 5% to 10% for ruptured tuboovarian abscesses, even with modern medical and operative therapy, mostly the result of subsequent development of adult respiratory distress syndrome (ARDS), a condition often associated with serious infection.

TREATMENT Part of "Chapter 28 - Pelvic Inflammatory Disease" The therapeutic goals in the management of acute PID include both elimination of the acute infection and symptoms and prevention of long-term sequelae such as infertility, ectopic pregnancy, chronic pelvic pain, and the residue of infection. Antibiotic treatment should be started as soon as cultures have been obtained and diagnosis is confirmed or strongly suspected. Treatment is based on the consensus that PID is polymicrobial in cause. Empirical antibiotic protocols should cover a wide range of bacteria, including N. gonorrhoeae, C. trachomatis, anaerobic rods and cocci, gram-negative aerobic rods, gram-positive aerobes, and Mycoplasma species. Despite general agreement that broad-spectrum therapy is appropriate, questions persist regarding optimal therapeutic regimens. Controversy has arisen over the issue of outpatient treatment with oral antibiotics versus inpatient treatment with parenteral antibiotics. There are no data available to evaluate the efficacy of hospital versus ambulatory management of acute PID. In the United States, three of four women with acute pelvic infection are treated as outpatients for their disease. In Scandinavia, which has a different health care system, most women are treated as inpatients. In 2002, the CDC published recommended treatment guidelines for outpatient management of acute PID (Table 28.2). Some of the treatment regimens are based on the controversial premise that it may be adequate to cover just a few of the major etiologic agents (N. gonorrhoeae and C. trachomatis) involved in acute salpingitis. As a result, studies have documented a 10% to 20% treatment failure rate for women receiving oral antibiotics as outpatients compared with a 5% to 10% failure rate for women receiving intravenous antibiotics as inpatients, where broader coverage is used. The inclusion of the quinolone arm, ofloxacin, and levofloxacin in the outpatient treatment regimen does permit broader coverage of pathogens, but this still may not be adequate for serious disease. It is important to reevaluate patients within 48 to 72 hours of initiating outpatient therapy to determine the response of the disease. If a poor response has been obtained, the patient should be hospitalized with parenteral antibiotics in the hope of preventing or limiting the sequelae of PID. TABLE 28.2. CDC-Recommended Treatment Regimens for Outpatient Therapy of Acute Pelvic Inflammatory Disease Regimen A Ofloxacin 400 mg orally twice a day or levofloxacin 500 mg orally once daily for 14 days, with or without metronidazole 500 mg orally twice a day for 14 days. Oral ofloxacin has been investigated as a single agent in two well-designed clinical trials, and it is effective against both Neisseria gonorrhoeae and Chlamdyia trachomatis. Despite the results of these trials, ofloxacin's lack of anaerobic coverage is a concern; the addition of metronidazole provides this coverage. Regimen B Ceftriaxone 250 mg IM once, or cefoxitin 2 g IM plus probenecid 1 g orally in a single dose concurrently once, or other parenteral third-generation cephalosporin (e.g., ceftizoxime or cefotaxime), plus doxycycline 100 mg orally twice a day for 14 days with or without metronidazole 500 mg orally twice a day for 14 days. The optimal choice of a cephalosporin for Regimen B is unclear; although cefoxitin has better anaerobic coverage, ceftriaxone has better coverage against N. gonorrhoeae. Clinical trials have demonstrated that a single dose of cefoxitin is effective in obtaining short-term clinical response in women who have pelvic inflammatory disease (PID); however, the theoretical limitations in its coverage of anaerobes may require the addition of metronidazole. The metronidazole also effectively treats BV, which also is frequently associated with BV. No data have been published regarding the use of oral cephalosporins for the treatment of PID. Alternative Oral Regimens Information regarding other outpatient regimens is limited, but one other regimen has undergone at least one clinical trial and has broad-spectrum coverage. Amoxicillin/clavulanic acid plus doxycycline was effective in obtaining short-term clinical response in a single clinical trial; however, gastrointestinal symptoms might limit the overall success of this regimen. Several recent investigations have evaluated the use of azithromycin in the treatment of upper reproductive tract infections; however, the data are insufficient to recommend this agent as a component of any of the treatment regimens for PID. to-T2-28to-T2-28 Ideally, every woman with acute PID should be hospitalized for the first few days for parenteral antibiotic

treatment. Because this may not be practical because of limited economic or physical facility resources, the clinician who diagnoses acute salpingitis in the office or emergency department is faced with the question of which patient to hospitalize. Indications for the hospitalization of patients with acute salpingitis are also defined by the CDC in the 2002 guidelines (Table 28.3). However, it is suggested that all adolescents with salpingitis be hospitalized because of their high noncompliance rate and to optimize treatment to prevent damage to the reproductive tract, which could affect future fertility. TABLE 28.3. Criteria for Hospitalization of Patients With Acute Pelvic Inflammatory Disease The following criteria for hospitalization are based on observational data and theoretical concerns: Surgical emergencies such as appendicitis cannot be excluded. The patient is pregnant. The patient does not respond clinically to oral antimicrobial therapy. The patient is unable to follow or tolerate an outpatient oral regimen. The patient has severe illness, nausea and vomiting, or high fever. The patient has a tuboovarian abscess to-T3-28to-T3-28 Another indication for hospitalization is the presence of an adnexal or pelvic abscess. Outpatient therapy may not provide antibiotic levels high enough to penetrate an abscess, and rupture of the abscess may have serious consequences. Women in whom the definitive diagnosis of acute PID is in question should also be hospitalized, and diagnostic measures should be instituted. As previously stated, at least 10% of all patients have other serious diagnoses, such as acute appendicitis, ectopic pregnancy, or adnexal torsion, and these should be ruled out. Patients with serious illness, patients with nausea and vomiting, patients who are unable to follow or tolerate outpatient therapy, and patients with a previously failed outpatient oral regimen also should be hospitalized and given parenteral antibiotics. The 2002 CDC guidelines for inpatient treatment of acute PID describe two regimens (Table 28.4). Regimen A is a combination of oral or parenteral doxycycline plus intravenous cefoxitin or cefotetan. Other third-generation cephalosporins can be substituted, such as ceftizoxime (Cefizox) or cefotaxime (Claforan). All of these agents are effective against penicillinase-producing N. gonorrhoeae, Peptostreptococcus and other anaerobic species and Escherichia coli, and other aerobic (facultative) species. Ceftriaxone is recommended by the CDC; however, its poor anaerobic activity and lack of trials do not make it an acceptable alternative for several investigators. Doxycycline can be given intravenously if the patient is unable to tolerate oral therapy, but it must be infused very slowly to prevent pain and sclerosis of the vein. Oral doxycycline has been demonstrated to be equally effective because of the slow growth cycle of Chlamydia and the requirement of prolonged treatment. A possible disadvantage of the cephalosporin-doxycycline combination is that these two antibiotics may be less than ideal for anaerobic infections or for a pelvic abscess. Regimen B is a combination of clindamycin and an aminoglycoside (gentamicin). This combination provides excellent activity against anaerobes, gram-negative aerobes, and gram-positive aerobes. Historically, it has been the preferred regimen for patients with an abscess, IUD-related infections, or pelvic infections after a diagnostic or operative procedure. However, there are few data to prove that it is significantly more effective than the cephalosporin regimens. A possible disadvantage of regimen B is that it may not provide optimal activity against C. trachomatis and N. gonorrhoeae. Clindamycin in high doses (900 mg in 8 hours) has good activity against Chlamydia, and in vitro studies by Martens and colleagues have demonstrated effectiveness against only 90% of C. trachomatis strains. Doxycycline is believed to be the most effective chlamydial agent, according to in vitro testing, and is often used for at least 7 days to complete treatment when the patient is switched from parenteral to post-hospitalization therapy. Also, the CDC recommendation of once daily dosing for gentamicin is not based on any data on PID patients, and should be used only if indicated for renal considerations. TABLE 28.4. CDC-Recommended Treatment Regimens for Inpatient Therapy of Acute Pelvic Inflammatory Disease

Regimen A Cefoxitin, 2 g intravenously every 6 hours or Cefotetan, 2 g intravenously every 12 hours plus Doxycycline, 100 mg intravenously or orally every 12 hours Note: This regimen should be continued for at least 24 hours after the patient demonstrates substantial clinical improvement, after which doxycycline, 100 mg orally two times a day, should be continued for a total of 14 days. Doxycycline administered orally has bioavailability similar to that of the intravenous formulation and may be administered if normal gastrointestinal function is present. When tuboovarian abscess is present, many health care providers use clindamycin or metronidazole with doxycycline for continued therapy rather than doxycycline alone, because it provides more effective anaerobic coverage. Other cephalosporins (ceftizoxime, cefotaxime, and ceftriaxone) may be effective therapy for pelvic inflammatory disease (PID) as replacements for cefotetan or cefoxitin. Regimen B Clindamycin, 900 mg intravenously every 8 hours plus Gentamicin, loading dose intravenously or intramuscularly (2 mg/kg of body weight), followed by a maintenance dose (1.5 mg/kg) every 8 hours (single daily dosing may be substituted) Note: This regimen should be continued for at least 24 hours after the patient demonstrates substantial clinical improvement and then followed with doxycycline, 100 mg orally two times a day, or clindamycin, 450 mg orally four times a day, to complete a total of 14 days of therapy. When tuboovarian abscess is present, many health care providers use clindamycin for continued therapy rather than doxycycline because it provides more effective anaerobic coverage. Single daily dosing of gentamicin has not been evaluated for PID, but it has been efficacious in other analogous situations. Alternative Parenteral Regimens Limited data support the use of other parenteral regimens, but the following three regimens have been investigated in at least one clinical trial, and they have broad-spectrum coverage. Ofloxacin 400 mg IV every 12 hours, or levofloxacin 500 mg IV once daily, with or without metronidazole 500 mg IV every 8 hours, or Ampicillin/sulbactam 3 g IV every 6 hours, plus doxycycline 100 mg IV or orally every 12 hours Ampicillin/sulbactam plus doxycycline has good coverage against Chlamydia trachomatis, Neisseria gonorrhoeae, and anaerobes, and appears to be effective for patients who have tuboovarian abscess. IV ofloxacin has been investigated as a single agent; however, because of concerns regarding the anaerobic coverage, metronidazole may be included. to-T4-28to-T4-28 Each regimen stresses two concepts: the polymicrobial etiology of acute pelvic infection and the necessity of protecting against C. trachomatis and N. gonorrhoeae. With both protocols, the CDC recommends a minimum of at least 24 hours of intravenous treatment after clinical improvement. Both protocols also require completion of a 14-day course of oral antibiotics (doxycycline or clindamycin) to eradicate slow-growing organisms such as C. trachomatis. Alternative inpatient parenteral regimens are included in the 2002 CDC PID guidelines. While the CDC lists only the ?-lactamase inhibitor combination ampicillin-sulbactam (Unasyn), piperacillin-tazobactam has been demonstrated by Hemsell, Sweet and colleagues, and others to have excellent in vitro and in vivo activity against PID and its pathogens (Table 28.4). Management of acute PID should include treatment of the male partner and education for the prevention of reinfection, including the use of proper contraception. The importance of treating sexual partners cannot be overstressed. Eschenbach reported that 25% of gonococcal PID patients were readmitted to the hospital within 10 weeks of the initial treatment. A study of gonococcal PID noted that 13% of male partners screened were asymptomatic urethral carriers and that even higher rates for C. trachomatis were present. These partners should be treated with one of the regimens for uncomplicated gonorrhoeae and chlamydial infection (i.e., ceftriaxone, 125 mg intramuscularly, followed by oral doxycycline, 100 mg twice a day for 7 days, oral azithromycin in 1 g or ofloxacin 300 mg b.i.d. for 7 days orally). Women with acute PID often return to the same social situations they were in before treatment. Treating sexual partners and educating patients with regard to contraception would decrease the incidence of recurrent infections and hopefully affect the often poor prognosis for future fertility.

Following are the 2002 CDC guidelines for the management of male partners of women with PID.

MANAGEMENT OF SEX PARTNERS Part of "Chapter 28 - Pelvic Inflammatory Disease" Evaluation and treatment of sex partners of women from the previous 60 days who have PID is imperative because of the risk of reinfection and the likelihood of urethral gonococcal or chlamydial infection of the partner. Sex partners should be treated empirically with regimens effective against both of these infections, regardless of the apparent etiology of PID or pathogens isolated from the infected woman. Even in clinical settings in which only women are seen, special arrangements should be made to provide care for male sex partners of women with PID. When this is not feasible, health care providers should ensure that sex partners are appropriately referred for treatment.

HIV INFECTION Part of "Chapter 28 - Pelvic Inflammatory Disease" Treatment of HIV-infected patients diagnosed with PID has also been addressed by the CDC. Differences in the clinical manifestations of PID between HIV-infected women and uninfected women have not been described clearly. In recent studies, HIV-infected women with PID had similar symptoms to non–HIV-infected patients with PID, but they were more likely to have a tuboovarian abscess. HIV-infected women in whom PID develops should be managed aggressively but appear to respond equally well to antibiotic regimens. Hospitalization and inpatient therapy with one of the intravenous antimicrobial regimens is recommended by several experts, but is no longer considered mandatory by the CDC. Higher HPV infections and cytologic abnormalities were noted in HIV-positive PID patients and should be evaluated fully.

SURGICAL MANAGEMENT Part of "Chapter 28 - Pelvic Inflammatory Disease" Laparotomy should generally be reserved for patients with surgical emergencies such as ruptured abscesses or definitive treatment of failed medical management. Laparoscopy, however, is an underused but usually helpful procedure for diagnosis, prognosis, and possibly treatment of PID. Laparoscopic evaluation should be considered in all patients with a differential diagnosis of PID and without laparoscopic surgery contraindications. Laparoscopy is important not only to diagnose PID but also to rule out surgical emergencies, such as appendicitis and ruptured abscesses. It also prevents inappropriate management of patients with noninfectious problems, such as endometriosis. These patients need additional surgical and medical management, not antibiotic therapy. In addition, evaluation of the extent of the inflammatory process in confirmed PID is helpful in establishing a prognosis and further management plan, if initial treatment fails. Patients with evidence of current or previous abscesses have a higher failure rate with antibiotic therapy. Also, treatment of unilateral abscesses may necessitate surgical management to avoid the spread of the infection to the other, perhaps less damaged, tube and ovary. Cultures obtained from the peritubal region or from the peritoneal cavity can also be helpful for identifying organisms resistant to initial management. This has become increasingly important in light of the increasing rate of clindamycin-resistant anaerobes and the elimination of metronidazole from the CDC-recommended inpatient guidelines in the 1980s. Laparoscopic management of PID that appears helpful includes copious drainage of the pelvis with normal saline or preferably Ringer's solution. Antibiotic inclusion in the lavage fluid has not been demonstrated to be helpful to date. Laparoscopic manipulation or drainage of documented pelvic abscesses has been attempted by several investigators. Henry-Suchet and associates reported the successful use of laparoscopy to diagnose and drain tuboovarian abscesses in 50 women. Adhesions were lysed, and the abscesses were drained through the laparoscope. All patients received intravenous antibiotics. Forty-five of the 50 (90%) patients were cured. Reich and McGlynn had a similar experience in 25 women with pelvic abscesses treated laparoscopically. Four of seven women desiring pregnancy conceived, and two women had unplanned pregnancies. However, the diagnosis of abscesses is not uniform in these studies. Also, it is of concern that similar results will not necessarily be demonstrated in less experienced hands. Anatomically, drainage of abscesses within the pelvic cavity by laparoscope will not drain the entire abscess contents out of the pelvic cavity, and despite how extensive the lavage or laparoscopic removal is, pus and bacteria will be spilled and exposed to the pelvic cavity. This is contrary to the natural defense mechanism of the body of isolating and containing the inflammation-causing organisms within an abscess. Therefore, laparoscopic drainage of pelvic abscesses should be undertaken only by experienced laparoscopic surgeons, and with the patient's full understanding of all other options. Laparotomy with extensive pelvic surgery was often recommended in the past, before the development of broad-spectrum antibiotics. If a patient has been hospitalized on several occasions for acute exacerbation of PID with bilateral tuboovarian abscesses to the point where the future surgical risk increases significantly, definitive surgical intervention may be indicated. The operation should be done when the infection is quiescent, if possible. The surgery may still be difficult, but there will be fewer complications than when patients are operated on in the acute phase of the infection. The timing of the operative intervention is important. There should be complete absorption of the inflammatory exudate surrounding the focus of the infection, as seen radiologically. Bimanual pelvic examination should be possible without producing a marked or persistent febrile response. It has been suggested that definitive surgery be delayed for 2 to 3 months after the recent exacerbation for more complete resolution of the infection. Ideally, the patient should have a normal ESR, WBC count, and hematocrit, and relatively nontender pelvic organs, except with motion. Kaplan and associates recommended more aggressive management in patients who exhibit either no clinical response or only partial response after 24 to 72 hours. Their approach included a total abdominal hysterectomy and bilateral salpingo-oophorectomy and was thought to reduce the protracted period of intensive medical therapy in a group of patients who would eventually require surgery. They noted that conservative management of their cases usually resulted in protracted periods of intensive care and repeated hospital admissions, and rarely in subsequent pregnancies. However, the early surgical intervention of Kaplan and colleagues was associated with six incidences of injury to the bowel and additional postoperative complications. Unfortunately, patients with acute pelvic abscess are frequently young, and future childbearing is often desired, even though it may be impossible for most patients. Conservation of ovarian function for these young women is an important benefit of medical management. Some differences in the percentage of patients responding to conservative management in different

studies and different geographic locations might be explained by differences in the predominant microorganisms causing the infection at these locations and their sensitivity to the antibiotics used. Older studies of the management of patients with pelvic abscess, which emphasized the early use of surgery, are no longer pertinent, because modern antibiotic drugs were not available then. Collins and Jansen in 1959 had an early failure rate of 10% for conservative medical therapy. However, 113 of their 174 patients required later surgery, which resulted in a late failure rate of 65%. Ginsburg and associates reviewed cases of 160 patients treated for tuboovarian abscess during the years 1969 to 1979. The early failure rate with broad-spectrum antibiotics was 31%, whereas the late failure rate was 21%. In an average follow-up period of 25.5 months, 48% did not require later surgery. Subsequent reports by Hager and by Landers and Sweet support conservative management. When conservative management fails and a pelvic abscess is noted dissecting the rectovaginal septum, drainage by way of colpotomy may be possible.

POSTERIOR COLPOTOMY Part of "Chapter 28 - Pelvic Inflammatory Disease" In a classic article, Wharton described various techniques of vaginal drainage of pelvic abscess. Today, posterior colpotomy is done to evacuate pus and to establish drainage from a pelvic abscess that presents in the cul-de-sac. There are three requirements for colpotomy drainage of a pelvic abscess. The abscess must be midline or nearly so. The abscess should be adherent to the cul-de-sac peritoneum and should dissect the rectovaginal septum to assure the surgeon that the drainage will be extraperitoneal and that pus will not be disseminated transperitoneally. The abscess should be cystic or fluctuant to ensure adequate drainage. Occasionally, a cul-de-sac abscess can be successfully drained without dissecting the septum. However, the serosal surface of the abscess should be adherent to the cul-de-sac peritoneum. Ultrasonography may be helpful in locating the pockets of pus. After adequate anesthesia, the patient is placed in the lithotomy position. It is essential that a thorough examination of the pelvis be performed under anesthesia so that the operator knows the size and position of the mass that is to be drained. After preparation and draping in the dorsal lithotomy position, the posterior lip of the cervix is grasped with a tenaculum and drawn down and forward. The vaginal mucosa of the posterior vaginal fornix is incised just below the reflection of the vaginal mucosa onto the cervix, and the transverse incision is widened with a pair of long scissors (Fig. 28.1A). The incision must be large enough to allow adequate exploration and drainage of the abscess cavity with the index finger. The cul-de-sac peritoneum and abscess wall are punctured with a long Kelly clamp (Fig. 28.1B). As the abscess wall is perforated, there is a definite sensation of puncturing a cystic cavity. If blood or pus is present, this is soon seen in the upper vagina. The jaws of the clamp are spread, and the flow of liquid from the cul-de-sac is increased. A sample of the purulent exudate is sent to the microbiology laboratory for appropriate culture and sensitivity. Collection of the specimen anaerobically with a capped syringe with rapid transport to the laboratory allows the more fastidious flora to be defined. A direct smear for Gram stain is also made from the pus and examined for predominating organisms. FIGURE 28.1. Posterior colpotomy. A: A transverse incision is made through the vaginal mucosa at the junction of the posterior vaginal fornix with the cervix. B: A Kelly clamp is thrust through the abscess wall.

View Figure

There may be more than one compartment in an abscess cavity (Fig. 28.2). It is desirable to insert an index finger in the cavity and explore. Fibrous adhesions within the cavity can be gently broken. If another abscess wall is felt, it can often be cautiously and safely punctured under the guidance of a finger. Exploration and manipulation should be done carefully to avoid intraperitoneal rupture of the abscess or perforation of the bowel. To allow adequate drainage, the vaginal incision should be at least 2 cm wide. If

pus has been obtained, one or two drains are inserted into the abscess cavity and anchored with fine absorbable suture to permit easy removal. Penrose or closed suction drainage systems can be used. These are left for several days or longer. Wharton has emphasized the importance of prolonged drainage. A suture or two may be required to control bleeding from the vaginal mucosa. However, if a mushroom (Malecot) catheter is used for drainage, it should be removed in 48 to 72 hours to prevent significant fibrosis that could hinder removal. FIGURE 28.2. Pus may be contained within the tuboovarian abscess and within other pockets in the pelvic cavity.

View Figure

Patients in whom the abscess does not meet the criteria for colpotomy drainage often require laparotomy and direct drainage. Transabdominal or transvaginal drainage has been attempted to avoid the expense and complications of laparotomy and for patients in whom laparotomy is contraindicated. Experience with percutaneous drainage of intraabdominal and pelvic abscesses under ultrasonographic or computed tomographic (CT) guidance has been reported by Olak and associates, and by others. Worthen and Gunning used percutaneous catheter drainage of 11 abscesses in nine patients and achieved a cure rate of 77%. Two patients required surgical intervention subsequently. In 19 patients, simple percutaneous aspiration of 23 abscesses was successful with a 94% cure rate. The attempt at aspiration failed in seven patients (Fig. 28.3). The Grady Memorial Hospital experience, as reported by Tyrrel and associates, is similar. CT-guided percutaneous drainage in eight patients with tuboovarian abscess resulted in recovery without surgery in seven. One patient had marked clinical improvement but still required a posterior colpotomy. No complications occurred. Loy and associates have reported that the simultaneous use of real-time pelvic ultrasonography can facilitate transvaginal drainage of a pelvic abscess. If patients do not respond to intravenous antibiotics and percutaneous drainage or aspiration, surgical intervention is required.

FIGURE 28.3. Transabdominal needle aspiration of a pelvic abscess under guidance of computed tomography. Drainage tube is also placed.

View Figure

The long-term effects of pus and organisms released into the pelvis from the puncture site are unknown. However, short-term success rates are good, and surgical drainage of acute abscess is a basic principle. Therefore, needle drainage can be considered with proper patient selection and appropriate informed consent, which includes other management options. If exploratory laparotomy is necessary, the patient can be positioned in Allen universal stirrups. A lower abdominal transverse Maylard incision is ideal because it affords good exposure to the lateral adnexal pelvic organs and pelvic side walls. Pelvic adhesions should be released and the bowel should be packed off before the pelvic dissection commences. During the dissection, free pus is often spilled, and the upper abdomen should be isolated from this, if possible. When a ruptured abscess is encountered, the exudate is collected and sent immediately to the laboratory for anaerobic and aerobic cultures and antimicrobial sensitivity studies. The easiest way to obtain the material for anaerobic culture is simply to collect it in an airtight syringe and to submit a small piece of the abscess wall in an airtight container. The easiest place to begin the dissection is in the round ligament, which is the most consistently available and identifiable landmark. Following the round ligament medially always leads to the uterine corpus. Variations in the usual technique for the operation may be required because of extensive disease, dense adhesions, indurated and edematous tissue, and distorted anatomy. For example, it is sometimes convenient to perform the central dissection first (i.e., a subtotal hysterectomy). This allows more space and adequate exposure to perform the required adnexal surgery. Tuboovarian inflammatory masses may be found densely adherent in the cul-de-sac to the uterus, to the posterior surface of the broad ligament, and to the lateral pelvic side wall. There is risk of injury to the ureters, sigmoid, rectum, and small intestines. The method of dissection used depends on the nature of the adhesions. Soft, fresh adhesions can be broken gently and easily with finger dissection. Dense fibrotic adhesions must be carefully dissected and cut with scissors. The dissection can be especially difficult and risky if pelvic tissues are intensely indurated, as in ligneous pelvic cellulitis. If the infundibulopelvic ligament can be clamped, cut, and securely ligated, one can gain access to the lateral retroperitoneal space and identify the ureter. This facilitates a safe dissection of the abscess wall away from the ureter. In cases with extensive disease involving one or both adnexa, the use of preoperative ureteral catheterization may be helpful in identifying the location of the pelvic ureters. With tuboovarian abscess, the anatomic limits of the ovary may be difficult to define. If the ovary is to be removed, it should be removed completely to prevent subsequent development of ovarian remnant syndrome. When both adnexa must be removed, a hysterectomy should usually be performed. In some cases, only a subtotal hysterectomy is feasible. The cervix can usually be excised after removal of the adnexa and the uterine corpus. The vaginal vault is left open for drainage. A Penrose drain can be inserted and then removed several days later. Suspension of the vaginal vault and reperitonization of the pelvis are accomplished in the usual manner, if possible. A routine closure of the abdominal incision is performed. Jackson-Pratt suction drains are often placed above the fascia and brought out through a separate incision. Because the patient has been placed in the Allen universal stirrups for laparotomy, ureteral integrity can be confirmed as discussed elsewhere. Five milliliters of indigo carmine is given intravenously and a cystoscope is placed in the bladder. Blue dye can then be seen flowing from each ureteral orifice.

In the past, it was standard practice to do a bilateral salpingo-oophorectomy in almost all patients who had a laparotomy for acute pelvic abscess. This practice was based on the belief that the disease is almost always severe in both adnexa. Recent studies have suggested that as many as 25% to 50% of patients will have a relatively normal tube and ovary on one side. This may be especially true of patients whose infection is associated with IUD use. Golde and associates reported that 37 of 85 patients (44%) with tuboovarian abscesses confirmed at operation had unilateral abscesses; 20 were using an IUD. The studies of Landers and Sweet, Hager and Majmudar, and Ginsburg and co-workers also found a higher percentage of unilateral adnexal disease than was previously reported. In light of these findings, conservative adnexal surgery should be performed if possible. We have no hesitation in leaving a relatively normal tube and ovary at the time of hysterectomy with removal of the opposite adnexa for acute pelvic abscess. When the uterus is removed and the continuity between the conserved tube and the lower genital tract is interrupted, there is little risk of a new infection. If a strictly unilateral pelvic abscess is found at laparotomy, removal of the affected tube and ovary only, leaving in the uterus and the opposite adnexa, is acceptable in a patient who wishes to preserve fertility. However, in vitro fertilization techniques may be required to accomplish pregnancy. Such a patient does have a risk of recurrent tuboovarian abscess. It is especially important that her sexual partner be examined and receive treatment when indicated. In recent years there have been advances in reproductive technology that allow infertile patients to conceive and carry pregnancies to term under the most extraordinary circumstances. It has been possible, for example, to accomplish a successful pregnancy in a woman who has a uterus but no ovaries by instillation of a donated fertilized ovum into a suitably prepared uterus. Such a sophisticated procedure is not available to a large number of patients. However, mostly for medicolegal reasons, the option of leaving in the uterus when lateral salpingo-oophorectomy is to be performed should be discussed with the patient, especially if she is young and nulliparous. In summary, patients with an acute pelvic abscess should be hospitalized for treatment with parenteral broad-spectrum antibiotics. Surgery is indicated if the diagnosis is uncertain, if intraperitoneal rupture is diagnosed or suspected, or if the patient fails to respond to medical management.

RUPTURED PELVIC ABSCESS Part of "Chapter 28 - Pelvic Inflammatory Disease" A tuboovarian or pelvic abscess can rupture spontaneously into the rectum or sigmoid colon, into the bladder, or into the free peritoneal cavity. A pelvic abscess almost never ruptures spontaneously into the vagina unless the patient had a previous posterior colpotomy for drainage of an abscess. Under these circumstances, a recurrent pelvic abscess can dissect along the tract of the previous posterior colpotomy incision and drain spontaneously through the vagina. Spontaneous drainage through the rectum or sigmoid colon usually occurs in a patient whose abscess is too high to drain with a posterior colpotomy. In other words, although the abscess is fluctuant and midline, it is not yet dissecting the rectovaginal septum. While waiting for the abscess to come down, a sudden unexpected improvement in the patient's condition is noted, and she will confirm that pus has begun to drain through the anus. Further improvement in her condition usually occurs. A posterior colpotomy is not needed and, indeed, is contraindicated because doing so could cause a rectovaginal fistula to form. Spontaneous drainage through the bladder is rare. It occurs most commonly in elderly women with chronic abscesses developing from ruptured sigmoid diverticula. Only rarely does a chronic tuboovarian or pelvic abscess rupture and drain through the bladder, causing secondary infection of the bladder. When the abscess is removed with laparotomy, a defect in the bladder wall is noted. The indurated tissue around the defect should be removed and the defect closed with 3-0 delayed-absorbable suture in two layers. A Foley catheter can be left in place for 10 to 14 days while healing of the bladder wall takes place. Of all the complications that can result from PID, intraabdominal rupture of a tuboovarian abscess is the most life-threatening. Mortality from this complication is due to septic shock and the complications of generalized peritonitis, and the mortality rate can approach up to 10% in patients with warm shock. Abscesses can rupture spontaneously, after bimanual examination or accidental trauma. Bacteriologic study of the contents of the abscess has historically been unrewarding; a specific organism has been isolated in less than 50% of cases. The gonococcus is rarely identified in a pelvic abscess. Careful aerobic and anaerobic cultures often demonstrate the presence of a mixed infection that includes anaerobic organisms. McNamara and Mead reviewed the results of three separate studies that demonstrated 31 positive isolates of anaerobes in 30 patients with a pelvic abscess. Landers and Sweet have also confirmed similar findings in their series.

Diagnosis of Ruptured Tuboovarian Abscess The major clinical symptom of ruptured tuboovarian abscess is acute, progressive pelvic pain that is usually so severe that the patient can accurately identify the time and place of its occurrence. In the series from the Johns Hopkins Hospital reported by Vermeeren and Te Linde, the average age of patients with a ruptured tuboovarian abscess was 33 years, which is at least 10 years older than the average age of patients with acute PID. About 2% of these patients are postmenopausal. To our knowledge, only two cases of ruptured tuboovarian abscess in a pregnant patient have been reported. Often, there is a history of recurrent attacks of PID, with a sudden increase in the severity and extent of abdominal pain during a recent exacerbation of infection. On examination, the patient appears seriously ill and dehydrated, with rapid, shallow respirations. The abdomen is distended and quiet, with diminished or absent bowel sounds. Signs of generalized peritonitis, direct and rebound tenderness, muscle rigidity, and shifting dullness may be noted. A pelvic mass is palpable in more than 50% of cases. Tachycardia is common. Shock can be present or can develop while the patient is under observation. It is due to accumulation of fluids in peripheral tissues and later failure of compensatory vasoconstrictor mechanisms. The patient's temperature is usually greater than 101°F, but it can also be normal and even subnormal late in the course. The leukocyte count is likely to be more than 15,000, but it also can be normal. Severe leukopenia is an ominous sign. A culdocentesis is a valuable diagnostic aid and was positive for purulent material in 70% of the cases in the Mickal and Sellmann series. An abdominal radiograph usually shows a paralytic ileus, sometimes evidence of free fluid in the peritoneal cavity, and atelectasis in the lung bases.

Treatment of Ruptured Tuboovarian Abscess The longer the delay in the operative treatment of ruptured tuboovarian abscess, the greater the primary

mortality rate. In the series by Vermeeren and Te Linde from the Johns Hopkins Hospital, death occurred less than 90 hours after the time of rupture in 88% of fatal cases, both operative and nonoperative. As time passes after rupture of a tuboovarian abscess, septic peritonitis becomes more severe and generalized. The passage of time allows the development of septic shock from greater absorption of bacteria and bacterial endotoxins, and secretion of great quantities of fluid into the peritoneal cavity across inflamed peritoneal surfaces. Fluid shifts from the intravascular compartment to interstitial spaces as a result of the increased vascular permeability of the inflamed peritoneal membrane. This leads to hypovolemia, decreased cardiac output, decreased central venous pressure, hypotension, vasoconstriction, increased peripheral resistance, decreased tissue perfusion, metabolic acidosis, ARDS, decreased renal glomerular perfusion and filtration with decreased urine flow, severe hypoxemia, multiple organ system failure, and death. The prompt diagnosis and treatment of intraperitoneal rupture of a tuboovarian abscess is essential to minimize the risk of mortality of generalized peritonitis. The treatment of patients with ruptured tuboovarian abscess can be divided into three phases: preoperative, operative, and postoperative.

Preoperative Phase Operation should be undertaken after rapid but adequate preoperative preparation. The patient should be typed and crossmatched with 2 to 4 units of packed red blood cells. Monitoring of central venous pressure is essential for proper evaluation of the hemodynamics of this condition because many patients are dehydrated, in shock, and anemic. Swan-Ganz catheter placement may be preferable because it allows pulmonary capillary wedge pressure and pulmonary artery pressure determinations that are helpful in assessing the adequacy of fluid replacement and in detecting fluid overload. Variable amounts of fluid, sometimes tremendous amounts, are lost into the peritoneal cavity and intestinal tract because of peritonitis. Emergency blood chemistry determinations (e.g., serum electrolytes, creatinine, glucose, bilirubin, and alkaline phosphatase) are obtained, and intravenous fluids, preferably Ringer's lactate, are started immediately. Crystalloid solutions for fluid volume resuscitation are preferred for most patients with septic peritonitis. It may be advantageous to use partial colloid resuscitation in some patients with evidence of cardiopulmonary dysfunction because a smaller total volume is required. An excess of intravenous crystalloid solution may result in fluid overload. Vigorous broad-spectrum intravenous antibiotic therapy should be instituted. An indwelling urethral catheter is used to monitor fluid intake with hourly urine output. Generally, it is advantageous to insert a Cantor or Miller-Abbott intestinal tube before operation to decompress the distended bowel. Combating shock is a primary concern throughout treatment. Clinical assessment of respiratory function should be made. A distended tender abdomen may cause rapid, shallow respirations and use of accessory muscles for ventilation. Arterial blood gases may indicate mild hypoxemia, in which case oxygen should be administered and ventilator support provided. Blood transfusion should be started before surgery. When the patient has been properly prepared, immediate surgery should be undertaken. The results of treatment are better if major metabolic and hemodynamic problems are corrected before operation, but one cannot waste time in treating a critically ill patient with septic peritonitis.

Operative Phase The anesthetic of choice depends on the preference and experience of the anesthesiologist and the medical condition of the patient. The operation should be performed as rapidly as possible. Because speed as well as access to the upper abdomen may be required, a lower midline incision should be used. It can be quickly extended above the umbilicus if necessary. The patient should not be put in the Trendelenburg position until the abdomen is packed off, and no more of a dependent position should be used than is needed to prevent further dissemination of pus into the upper abdomen. When the abdomen is opened, any odor that is present should be noted. An unpleasant putrid odor is indicative of infection with anaerobic organisms. Pus from the abdomen should be collected correctly for both aerobic and anaerobic culture and for Gram stain, and be promptly transported to the laboratory. Organisms grown should be tested for sensitivity to various antibiotics. The operation of choice is removal of the free pus, together with the abscess, the uterus, the tubes, and usually the ovaries. Only occasionally is it possible to leave an ovary in a patient with a ruptured pelvic

abscess. If rupture has occurred from a strictly unilateral tuboovarian abscess, with a relatively normal tube and ovary on the opposite side, a unilateral salpingo-oophorectomy can be performed, especially if the patient is young. However, the risk of a recurrent abscess in the opposite tube and ovary is high if the uterus is also left in place. When the uterus is removed along with the tuboovarian abscess, the risk of recurrent abscess in the opposite adnexa is reduced. When hysterectomy is performed, usually a total hysterectomy can be done. However, even in the best surgical hands, a subtotal hysterectomy is faster than a total one and is sometimes justified. It is probable that the mortality rate would be increased if total hysterectomies were always performed. Although we believe firmly in total hysterectomy, we do not believe in performing it when the danger of total hysterectomy exceeds the danger from a retained cervix. Except in the young patient, it is better to remove the corpus than to perform a unilateral adnexectomy alone. Furthermore, the opposite adnexa is significantly involved in most patients, and subsequent operation may be necessary if conservation of one side is practiced, as was required in 35% of Pedowitz and Bloomfield's cases. This is contrary to what has been described earlier in the surgical treatment of an unruptured abscess, because the risk of incomplete eradication of the immediate infection in an acutely ill patient with rupture, peritonitis, and possibly septic shock is much too risky; therefore, definitive surgical treatment is usually recommended in severely ill patients with ruptured abscess. The technical performance of the procedure may be difficult, but is similar to that described earlier for laparotomy followed by failed colpotomy drainage or suspected rupture. Anatomy is distorted, dependable landmarks are obscured, and tissues are thick, edematous, friable, and inflamed. Loops of densely adherent intestine must be separated carefully to avoid injury. Injury to the serosa of distended bowel occurs commonly and sometimes requires repair. An entry into the lumen of the bowel must be recognized and repaired. The most dependable anatomic landmark is the round ligament. Followed medially, it always leads to the uterine corpus. Retroperitoneal planes of dissection can be used to advantage in identifying the ureters and removing inflammatory adnexal masses. Otherwise, it is likely that fragments of ovary will be left behind, which can subsequently cause signs and symptoms of the ovarian remnant syndrome. As much of the remaining abscess wall as possible should be removed without causing unnecessary additional bleeding. Pieces of the abscess wall can be left adherent to the pelvic side wall and cul-de-sac. Oozing of blood from all dissected tissue has been likened to “cinder bed bleeding” and is difficult to control. In 1977, Rivlin and Hunt used conservative pelvic surgery combined with intraoperative and postoperative peritoneal lavage with antibiotics in 113 women with generalized peritonitis caused by a ruptured tuboovarian abscess. The uterus, ovaries, and tubes were retained whenever possible. Either one or both of the adnexa were retained in whole or in part, and hysterectomy was performed in only four cases. All loculations of pus were opened, and aggressive lavage of the peritoneal cavity with gentamicin was carried out for several days postoperatively. The mortality rate was 7.1%, and further surgery was required in only 17.5%. Before the incision is closed, the abdominal cavity should be irrigated with copious quantities of sterile saline to remove remaining bacteria and debris. When generalized septic peritonitis is also present, large volumes of warm saline should also be used to irrigate the upper abdomen. There is always some fear of dissemination of the infection by copious irrigation. However, this disadvantage is far outweighed by the benefit of diluting and removing bacteria and necrotic debris. We do not add antiseptics or antibiotics to the irrigating solution. If hemostasis is poor or if considerable necrotic material is left behind, there may be some benefit from peritoneal drainage with closed suction catheters. Closed suction drains can be placed through a separate stab wound in the abdominal wall, through the cul-de-sac, or through the vaginal vault when a total hysterectomy has been done, but the drainage of free peritoneal exudate in the upper abdomen is of no therapeutic value. The upper abdomen should be carefully explored for collections of pus in the subdiaphragmatic and subhepatic regions. If an upper abdominal abscess is found, it may be necessary to place a closed suction drain into the abscess cavity through the upper abdominal wall. The abdominal incision is closed with a Smead-Jones technique or with a continuous suture taking large bites of tissue. A monofilament suture of polypropylene or nylon should be used. Retention sutures can be placed but are not usually necessary. The incision should be irrigated with warm saline. When there has been gross contamination of the incision, the subcutaneous fat and skin should be left open and packed lightly with gauze soaked in an antibiotic solution. The wound is repacked daily and inspected. In 4 to 5 days, if the tissues are healthy, the incision is closed secondarily with sutures. Alternatively, the edges can be drawn together with sterile adhesive strips.

Postoperative Phase Postoperative care should consider shock, infection, ileus, and fluid imbalances. Complications of the late postoperative period include pelvic and abdominal abscesses, intestinal obstruction, intestinal fistulas, incisional breakdown with or without evisceration, pulmonary embolus, continued sepsis, and disseminated intravascular coagulation. Serious medical diseases such as uncontrolled diabetes or renal or pulmonary failure (ARDS) further complicate recovery from this potentially lethal disease. Septic shock should be combated with blood (when indicated for a hemoglobin less than 7.0 g), Ringer's lactate, respiratory support, and, if necessary, vasoactive substances. Infection is controlled by the continued aggressive use of broad-spectrum intravenous antibiotics until the patient can take antibiotics orally. When the results of the antibiotic sensitivity studies on the operative specimen are available, a change to more effective agents should be considered, but only if the patient shows evidence of continued sepsis. Antibiotics should not necessarily be changed on the basis of sensitivity studies if the patient is improving clinically. Sometimes the patient's condition improves initially only to show signs of recurring intraabdominal infection the second week after operation. Under these circumstances, it is appropriate to change antibiotics. Antibiotics should be continued until the patient is afebrile with only a mild leukocytosis and is able to eat a regular diet. A long period of treatment with antibiotics may result in complications such as pseudomembranous enterocolitis. The semi-Fowler position may help prevent subphrenic and subdiaphragmatic abscess formation. Patients with signs of continued intraabdominal sepsis should have CT scans to identify collections of pus. If found, CT-directed drainage may be possible. Constant intestinal suction by means of a long intestinal tube is a very important feature of postoperative care. A dynamic ileus persists postoperatively for a variable period and is best treated with the long intestinal tube until there is evidence of peristalsis and the patient is passing flatus. Close attention to fluid balance and blood chemistry determinations is mandatory. Frequently, patients with ruptured tuboovarian abscess have poor kidney function. The fluid output and serum creatinine should be followed closely. The results of the preceding therapeutic measure have been gratifying. At Grady Memorial Hospital, the mortality rate for this formerly lethal disease is 3.5%.

PRIMARY OVARIAN ABSCESS Part of "Chapter 28 - Pelvic Inflammatory Disease" A primary ovarian abscess is an entity distinctly different from tuboovarian abscess. A tuboovarian abscess is one in which the abscess wall is composed of fallopian tube and ovarian parenchyma. A primary ovarian abscess, on the other hand, is one in which the infection occurs in the parenchyma of the ovary. Unlike tuboovarian abscess, it is an unusual condition. Interest in primary ovarian abscess was stimulated by the 1964 report of Willson and Black. According to a review by Wetchler and Dunn, 120 cases had been reported by 1985. Although bacteria can gain access to the ovarian parenchyma by hematogenous or lymphatic spread, it is probable that most primary ovarian abscesses occur because bacteria present around the ovary gain access to the parenchyma through a break in the ovarian capsule. The capsule can be broken naturally by ovulation or it can be broken by a surgical procedure. Bacteria come from the fallopian tube, from the vagina during or after hysterectomy, from intrauterine infection associated with an IUD, or from appendicitis, diverticulitis, or any other condition that is associated with peritonitis. A primary ovarian abscess is usually unilateral. However, its occurrence simultaneously in both ovaries and during pregnancy seems to support the occasional hematogenous or lymphatic spread, or both. Primary ovarian abscess has been reported secondary to infections at distant sites (tonsillitis, typhoid, parotitis, and tuberculosis). A mixed flora of anaerobic and aerobic bacteria is usually present. Actinomyces israelii with sulfur granules has also been identified in a few cases. Diagnosis of an unruptured primary ovarian abscess can be difficult because of the variable clinical presentation. Lower abdominal pain and fever are usually present. Lower abdominal and pelvic tenderness and an adnexal mass may be present, but the pelvic examination is sometimes not helpful. Although an event predisposing to primary ovarian abscess (e.g., surgery, IUD use, appendicitis, or systemic infection) may be uncovered in the history, the event is sometimes remote. Ultrasonography and CT can be helpful in identifying an abscess cavity. When the ovarian abscess ruptures, the clinical picture is much the same as in ruptured tuboovarian abscess, with abdominal distention, direct and rebound tenderness, ileus, and sometimes shock. The patient appears gravely ill, and the need for immediate surgery is usually obvious. The management of patients with primary ovarian abscess is similar to the management of patients with acute tuboovarian abscess. If the abscess is not ruptured, medical management with antibiotics for both anaerobic and aerobic organisms plus supportive care is indicated. A failure to respond or deterioration in the patient's condition suggests alteration in antibiotic coverage or possible exploratory surgery, or both, to remove the abscess. Ruptured ovarian abscess requires immediate laparotomy after a brief but intense effort to stabilize the patient and start antibiotic therapy. At operation, only the affected ovary need be removed. The tubes and the uterus can be conserved. If both ovaries are involved, they should be removed. For a patient who is not interested in conception in the future, the uterus and both tubes can also be removed. If the patient is interested in pregnancy, the uterus and fallopian tubes can be left in place for possible implantation of a donated egg in the future.

SURGERY FOR CHRONIC PELVIC INFLAMMATORY DISEASE Part of "Chapter 28 - Pelvic Inflammatory Disease" Although the gonococcus may be responsible for initiating acute salpingitis, which is short-lived, the residual chronic salpingitis is usually due to secondary invaders, both aerobic and anaerobic, or perhaps to an initial infection with C. trachomatis. As a result of the initial infection or from subsequent secondary exacerbations, the fimbria can become occluded and the tubes bound to the ovaries with adhesions. In addition, the bowel can become adherent to the broad ligament and the adnexal structures, and the fascia and loose connective tissue of the broad ligament can be converted into an indurated, brawny structure typical of ligneous induration. This can extend to include tissues beneath the peritoneum on the lateral pelvic side wall, where ligneous pelvic cellulitis can cause ureteral obstruction. If the chronic infection persists, serious effusion from the inflammatory process within the endosalpinx produces a hydrosalpinx that can ignite periodically with secondary subacute pelvic infection or can progress to produce a pyosalpinx and tuboovarian abscess. If the subacute infection is left untreated or is treated inadequately, spontaneous intraabdominal rupture or leakage of an old tuboovarian abscess can occur. In a review of this subject, Heaton and Ledger identified this problem principally in premenopausal women, with only 1.7% of patients with a tuboovarian abscess being postmenopausal. The signs and symptoms of chronic PID that most often require surgical treatment include severe, persistent, progressive pelvic pain, usually bilateral, although occasionally localized in one of the lower abdominal quadrants; repeated exacerbations of PID requiring multiple hospitalizations and recurrent medical treatment; progressive enlargement of a tuboovarian inflammatory mass, especially if it cannot be distinguished from a neoplastic tumor of the ovary; severe dyspareunia related to the chronic pelvic infection; and bilateral ureteral obstruction from ligneous cellulitis. It was formerly accepted that a history of previous colpotomy for drainage of a pelvic abscess was sufficient reason in itself to justify definitive abdominal surgery later for removal of the uterus and adnexa. We have seen several patients who have become pregnant after posterior colpotomy for drainage of a cul-de-sac abscess and who have remained relatively free of symptoms for long periods. Today, previous posterior colpotomy for pelvic abscess drainage is not a sufficient indication by itself for definitive abdominal surgery.

Selection of Operation The final decision regarding the proper operation for the surgical management of chronic PID is usually made with the abdomen open. Consideration must be given not only to the pathologic lesions found at operation but also to the patient's age, parity, desire for children, previous history of pelvic disease, and other associated pelvic disease and symptoms. Because a knowledge of all these is essential to the best surgical judgment, the operator should be thoroughly familiar with the patient, her history, and her desires. In the surgical management of chronic PID, the question of removal or retention of the ovary at the time of hysterectomy and salpingectomy has been left open to conjecture and individual surgical opinion in most instances. This question was the subject of a study by Weiner and Wallach of the ovarian histology in ovaries removed from patients with PID. In 40 consecutive women who underwent oophorectomy during surgical treatment of PID, nearly 50% of the removed ovaries were free of inflammatory disease and demonstrated normal follicular activity. The study concluded that ovarian histology was usually normal among patients who gave no history of dysfunctional uterine bleeding. Therefore, the menstrual history of such patients should be helpful in the decision regarding ovarian conservation or ablation. Kirtley and Benigno have reviewed our experience with ovarian conservation at the time of surgery for PID. In this series, 98 (82%) patients who required surgery had a total abdominal hysterectomy and bilateral salpingo-oophorectomy. In 22 patients (18%), either part or all of an ovary was retained. Of the 22 patients, 15 were available for follow-up hormonal assays. The mean follow-up time was 58 months. Cyclic ovarian function was confirmed in all but two patients. In the two patients with ovarian failure, other significant disease processes were also present. No patient suffered a complication as a result of adnexal conservation. We believe that normal ovarian tissue should be conserved at the time of definitive surgery for PID. A small hydrosalpinx on the same side as the normal ovary can also be left in place so that ovarian blood supply is not disturbed during an attempt to remove the tube. The release of peritubal adhesions in mild chronic PID is indicated occasionally in women in whom future childbearing is desired, as long as the tubes can be shown to be patent, usually by transfundal chromotubation after the lower uterine isthmus is occluded by a Ziegler clamp. This type of procedure

provides the most rewarding pregnancy rate of all types of tubal reconstructive surgery. More often, one tube is hopelessly closed and the opposite tube is patent after release of adhesions. In such a case, unilateral salpingectomy may be required if reconstructive surgery is not possible. Many other procedures are available in the treatment of this disease, including salpingo-oophorectomy with or without hysterectomy (Fig. 28.4 and Fig. 28.5). FIGURE 28.4. Total abdominal hysterectomy and unilateral salpingo-oophorectomy from extensive chronic salpingo-oophoritis. A small hydrosalpinx on the opposite side can be left in to preserve blood supply to the ovary.

View Figure

FIGURE 28.5. When significant chronic pelvic inflammatory disease involves only one adnexa and preservation of uterine function is indicated, a unilateral salpingo-oophorectomy can be performed.

View Figure

In most instances of surgery for chronic PID, total abdominal hysterectomy and bilateral salpingo-oophorectomy are necessary to remove the primary tubal pathology because of inflammatory damage of both tubes and ovaries. Total abdominal hysterectomy and bilateral salpingo-oophorectomy (Fig. 28.6) have been performed for severe actinomycosis infection. FIGURE 28.6. Total abdominal hysterectomy and bilateral salpingo-oophorectomy for severe pelvic actinomycosis.

View Figure

If the uterus is removed and an ovary is preserved, it may be preferable to leave the entire adnexa in place in the absence of active tubal infection rather than compromise the venous drainage or the arterial blood supply to the ovary, with subsequent cystic changes that may require an additional operative procedure later. Once the continuity of the tubal lumen from the uterine cavity is broken, the chronically inflamed tube does not usually produce subsequent symptoms, as shown by Falk in his series of cases with interstitial tubal resection. When it is considered advisable to remove both adnexa because of the extent of the tuboovarian disease, a total hysterectomy is also advisable unless the uterus is hopelessly encased in pelvic scar tissue and densely adherent to the pelvic viscera. Usually the uterus can be removed without difficulty, thus providing an easier opportunity to peritonize the operative site and avoid additional postoperative adhesions. However, there is a place for mature surgical judgment in this instance, and

discretion dictates whether a subtotal hysterectomy rather than a total hysterectomy is surgically advisable. In the optimum case, especially in a young woman who wishes to establish or maintain the possibility of future fertility, conservative surgery may be desirable, with the hope that pregnancy can be accomplished through in vitro fertilization techniques. In this situation, the uterus and one adnexa should be conserved, and the ovary should be positioned in the pelvis so an ovum can be harvested later through the laparoscope or through the vagina. As mentioned earlier, if the patient wishes, the uterus can be left in place even though both tubes and ovaries have been removed.

Salpingectomy for Chronic Salpingitis At the time of surgery for the treatment of chronic PID, every effort should be made to retain uninvolved organs. Unilateral salpingectomy should be considered when the oviduct is hopelessly destroyed by the disease process and presents as a large hydrosalpinx. P.692 The abdomen is entered through a transverse Maylard incision. The adhesions binding the tube are cut and the tube is freed. It is held by a Kelly clamp placed on the mesosalpinx just beneath the fimbriated end. The mesosalpinx is then clamped and cut, with a succession of small bites taken as close to the tube as possible (Fig. 28.7A). FIGURE 28.7. Mesosalpinx is clamped with multiple Kelly clamps and cut. Dotted lines indicate cornual excision, which is elective. B: Cornual wound is closed with 2-0 delayed-absorbable suture. C: Mesosalpinx vessels are transfixed. D: Mattress suture is placed to cover operative area. E: Round ligament and broad ligament cover operative area.

View Figure

Keeping the operative trauma as far as possible from the ovary that is to be retained lessens the danger of imperiling its blood supply. Experience has shown that the ovary whose tube has been removed is more apt to become cystic than the ovary whose tube has been left undisturbed. Therefore, it seems logical to interfere as little as possible with the blood supply of the ovary by hugging the tube closely when excising it. The tube is excised at the uterine cornu in a wedge-shaped manner, as indicated in Figure 28.7B. A wide, figure-of-eight 2-0 delayed-absorbable suture is placed in the cornu before the wedge is excised and is tightened as the interstitial portion of the tube is removed. If there is palpable extension of the inflammation at the uterine cornu (so-called salpingitis isthmica nodosa), the wedge may be large. The wound in the uterus is closed with one or more 2-0 delayed-absorbable figure-of-eight sutures (Fig. 28.7B). The vessels in the mesosalpinx are ligated with transfixion 3-0 delayed-absorbable sutures. The advantage of the transfixion suture is that it does not slip off the tissue when tied as the clamp is withdrawn (Fig. 28.7C). A mattress suture of 3-0 delayed-absorbable material is used to bring the broad and round ligaments over the cornual wound (Fig. 28.7D). This suture passes just beneath the round ligament, so that the ligament is not strangulated when the suture is drawn tight. When this suture is tied, the cornual wound is covered with the broad ligament, and the uterus is suspended to some extent in a manner similar to that used in

the Coffey suspension. Usually, a second mattress or interrupted suture is necessary to cover the mesosalpinx completely, as shown in Figure 28.7E.

Salpingo-oophorectomy for Chronic Salpingitis As in salpingectomy, the abdomen is entered through a transverse Maylard incision. The chronic tuboovarian inflammatory mass is first dissected free and the infundibulopelvic ligament is identified. It is doubly clamped with Ochsner clamps, and a third clamp is applied to control back-bleeding (Fig. 28.8A). The ureter must be identified before the infundibulopelvic ligament is clamped, cut, and ligated. FIGURE 28.8. Salpingo-oophorectomy.A: The infundibulopelvic ligament is doubly clamped. Another clamp is placed to control back-bleeding. Dotted line indicates incision.

View Figure

After the infundibulopelvic ligament is cut and ligated, the remainder of the broad ligament attachment of the tube and the ovary is clamped, cut, and ligated. The uterine end of the tube and the ovarian ligament are excised from the uterus in a wedge-shaped manner. The ascending uterine vessels are ligated just below the cornual wound, and the cornual incision is closed with a 2-0 delayed-absorbable figure-of-eight suture (Fig. 28.8B). FIGURE 28.8. Salpingo-oophorectomy. B: A suture has been placed to ligate the ascending uterine vessels just below the cornual incision. The cornual incision is closed with a figure-of-eight suture of 2-0 delayed-absorbable material.

View Figure

The infundibulopelvic ligament is doubly ligated with 2-0 delayed-absorbable sutures, and the vessels in the broad ligament are ligated with 3-0 delayed-absorbable sutures. The cornual wound is peritonized, and the uterus is suspended to some degree by bringing the round and the broad ligaments over the uterine cornu with a mattress suture of 2-0 delayed-absorbable material, as shown in Figure 28.8C. An attempt should be made to remove the tuboovarian inflammatory complex completely. If a fragment of ovary is left attached to the lateral pelvic peritoneum or the broad ligament, the ovarian remnant syndrome may develop later. To prevent this, a retroperitoneal approach may be required. FIGURE 28.8. Salpingo-oophorectomy.C: The infundibulopelvic ligament and the rest of the broad ligament vessels have been ligated. The cornual wound is covered with the round and the broad ligament using a mattress suture of 2-0 delayed-absorbable material.

View Figure

P.693 P.694

Identification of the Ureter Identification of the course of the ureter in a pelvis in which the anatomy has become obliterated as a result of PID is one of the most important responsibilities of the gynecologist. In the surgical treatment of this disease, one may find a tuboovarian inflammatory mass that is located between the leaves of the broad ligament and extends to the lateral pelvic wall. It is not uncommon for the ligneous induration of the thickened parietal peritoneum to obscure completely the location and course of the pelvic ureter so that dissection of the diseased adnexa produces a surgical risk to the urinary tract, requiring great technical skill to avoid ureteral injury. Knowledge of the normal anatomic location of the pelvic ureters is essential so that these vital structures can be identified before an attempt is made to remove the adnexal masses. Division of the round ligament allows access to the lateral pelvic wall beneath the peritoneum. After the round ligament is divided, the peritoneum is incised inferiorly toward the internal cervical os and superiorly just lateral to the infundibulopelvic ligament. The peritoneum is easily reflected medially away from the pelvic side wall with finger dissection, and the ureter is identified. It remains attached to the peritoneum. If there is difficulty with this procedure, the ureter can usually be identified as it crosses over the common iliac artery just above its bifurcation, and it can be traced downward. Such patients may have a preoperative ureteral catheterization when there is clinical evidence of large, adherent adnexal masses. However, if such an anatomic problem is encountered at the time of laparotomy, an incision can be made in the dome of the bladder that allows the passage of ureteral catheters. If the patient has been positioned in the Allen universal stirrups for operation, intraoperative cystoscopy with passage of ureteral stents is easily accomplished. At the end of the operation, 5 mL of indigo carmine is given intravenously. With a cystoscope in the bladder, the dye can be seen effusing from both ureteral orifices, confirming that the ureters have not been injured or compromised. P.695

Drainage at Laparotomy for Pelvic Abscess Views on drainage at laparotomy for PID have changed during the past several years. Whereas drainage was an everyday occurrence in gynecologic operating rooms in the 1950s, it is used only occasionally today. Several factors are responsible for this change. Operations for acute and subacute PID are avoided by aggressive medical management; hence, pus is encountered less frequently. Even when small pockets of pus are encountered, experience has shown that the pus can be suctioned away, the peritoneum irrigated thoroughly with saline, and the abdomen closed without drainage. Antibiotic therapy has also reduced drainage. The operator is justified in depending on postoperative antibiotics to combat the infection. When an abscess is densely adherent to the bowel wall or the region of the ureter, thorough removal of all the abscess wall may result in bleeding and damage to a viscus. In such cases, small portions of the necrotic abscess wall can be left in situ and a closed suction drain placed against the area. The ideal exit for a drain is through the cul-de-sac, as shown in Figure 28.9. Sometimes the cul-de-sac is completely obliterated by adhesions between the anterior rectal wall and the cervix. In such instances, use of the posterior vaginal fornix for drainage may not be feasible. When drainage is indicated under such circumstances, it should be done through a small stab wound in whichever lower quadrant is most directly above the point to be drained (Fig. 28.10). We dislike drainage through the primary incision because of the danger of hernia formation and incisional infection. When the large bowel has been entered accidentally and a perfectly satisfactory closure has been effected, the abdomen is closed without drainage and the patient is placed on antibiotic therapy. If the condition of the large bowel wall is such that satisfactory closure cannot be done, temporary colostomy above the injury may be preferable to the risk of an abscess and an intestinal fistula. FIGURE 28.9. Drainage of pelvis through the cul-de-sac. A long Kelly clamp is inserted in the vagina and opened slightly as the posterior vaginal fornix is pushed upward. The scalpel incises between the jaws of the clamp. The drain is clamped and withdrawn through the vagina.

View Figure

FIGURE 28.10. Bilateral transabdominal drainage through stab wounds. A closed suction drainage system is preferred.

View Figure

When pus is spilled and gross contamination of the operative field is present, a closed suction drainage system P.696

P.697 can also be installed in the incision above the fascia with exit through a separate stab wound. This is especially important in diabetic and obese patients. Alternatively, the incision above the fascial closure can be packed open for a few days and closed secondarily to avoid incisional breakdown from infection.

PELVIC TUBERCULOSIS Part of "Chapter 28 - Pelvic Inflammatory Disease" Tuberculosis of the upper genital tract is a rare disease in the United States. However, it is a frequent cause of chronic PID and infertility in other parts of the world. For various reasons, the incidence of tuberculosis is again increasing in the United States. Therefore, cases of tuberculosis-associated PID may also become more evident. It should always be suspected in immigrants, especially those from Asia, the Middle East, and Latin America, and in patients with HIV. Pelvic tuberculosis is produced primarily by either Mycobacterium tuberculosis or Mycobacterium bovis. The primary site of infection for tuberculosis is usually the lung, with lymphatic spread from the Ghon complex to regional lymph nodes at the hilum usually occurring within 1 to 2 years. More rapid spread is due to hematogenous spread, which results in miliary disease often within the first year. The fallopian tubes are the predominant site of pelvic tuberculosis, but the bacilli also spread to the endometrium and occasionally the ovaries. No location in the body is immune to the development of metastatic foci of infection. Tuberculosis of the bone, meninges, kidney, epididymis, fallopian tubes, and other sites can develop. At some sites of miliary spread, the lesions can remain quiescent for long periods before reactivation and further spread of the disease. Direct extension from one organ or system to an adjacent organ or system can also occur. Organs of the female reproductive tract are usually infected by hematogenous miliary spread from a primary pulmonary lesion, by hematogenous spread from a secondary miliary site, by lymphatic spread from a primary pulmonary site to intestinal lymph nodes and then to the pelvis, or by direct extension from adjacent abdominal organs (small intestines, appendix, rectum, bladder) that are the site of tuberculous infection. Fistulas between the intestinal tract and the fallopian tubes have been reported with pelvic tuberculosis. A venereal transmission of the disease has been reported, with primary genital infection in the woman occurring after coitus with a sexual partner who had tuberculosis of the genitourinary tract. According to Sutherland and MacFarlane, it is not possible to prove conclusively that genitourinary tuberculosis in the male can be transmitted to the female through sexual intercourse. Because it has been shown that M. tuberculosis is present in the sperm of men with urogenital tuberculosis, the possibility of transmission to the pelvic organs of the female through intercourse must be accepted. Sutherland presents five cases in which sexual transmission of genitourinary tuberculosis from male to female presumably occurred. However, of 128 husbands of women with genital tuberculosis, only five (3.9%) were found to have active genitourinary tuberculosis. When tuberculosis of the vulva, vagina, and cervix is present without evidence of tuberculosis elsewhere in the body, venereal transmission should be suspected.

Pathology of Pelvic Tuberculosis Both fallopian tubes are involved in almost all patients with pelvic tuberculosis. About one half of patients with tuberculous salpingitis have tuberculous endometritis. P.698 Tuberculosis of the cervix is present in 5% of cases. The vagina and vulva are rarely involved. At operation, one may find evidence of generalized tuberculous peritonitis with small, grayish white tubercles covering all peritoneal surfaces of the abdominal and pelvic organs. The mucosa of the fallopian tubes may not be involved in generalized serosal tuberculous infection. At a later stage of infection, tuberculous salpingitis may grossly resemble other forms of PID involving the adnexa. Unless tubercles are seen, the diagnosis may not be apparent until microscopic sections are examined by the pathologist. A large pyosalpinx may contain the caseous material of a tuberculous infection but may also contain the purulent exudate of a secondary infection with other common organisms. Tubercles form in the lining of the tube. Some have caseation at the center, with giant cells and epithelioid cells. A proliferation of the mucosal lining of the fallopian tube may resemble a primary tubal carcinoma microscopically and may be confusing to the pathologist. Tuberculous peritonitis is commonly associated with tuberculosis of the pelvis. Clinically, tuberculous peritonitis can be divided into two groups. In “wet” peritonitis there is an outpouring of straw-colored fluid into the peritoneal cavity, producing ascites. The peritoneum of the parietal wall and viscera is covered with innumerable small tubercles (Fig. 28.11). The tubes, in addition to being covered with miliary tubercles on the serosal surface, are usually slightly enlarged and distended. In contrast to other forms of salpingitis, the fimbriae may be patent. Within the tubal wall and tubal mucosa, the histology is typical of tuberculosis, with tubercle formation, multinucleated giant cells, and epithelioid reaction (Fig. 28.12). In advanced cases, frank caseation is present. This pattern is usually associated with hematogenous spread

of the tuberculous organism to the peritoneal surfaces and the pelvic organs. FIGURE 28.11. Typical specimen of tuberculosis of the reproductive organs as part of generalized tuberculous peritonitis.

View Figure

FIGURE 28.12. Tuberculosis of the fallopian tube. Note the multinucleated giant cells.

View Figure

Another type of tuberculous peritonitis encountered in women is the “dry” or adhesive type. Bowel adheres to bowel by innumerable dense adhesions that blend with the musculature. The muscle of the bowel is often invaded to some degree by the tuberculous process. Separation of these adhesions is extremely difficult surgically, and accidental injury to the bowel is common. The pelvic organs show evidence of tuberculous salpingitis with enlargement of the tubes, and occasionally pyosalpinges and even tuboovarian abscess formation. Tuberculous involvement of the myometrium is rare. Tuberculous endometritis, however, is common, occurring in 60% to 70% of women with pelvic tuberculosis. Microscopically, tubercles are seen scattered throughout the endometrium, but they may be scanty. Tubercles are often seen in the endometrium removed by curettage in the premenstrual phase and are usually located in the endometrium adjacent to the tubal ostia. Apparently, the uterine cavity is protected from advanced tuberculous infection by the cyclic shedding of endometrial tissue in the reproductive years. Even in advanced pelvic tuberculous infections, evidence of caseation, fibrosis, and calcification are rarely seen in the uterine cavity. Occasionally, the endometrial cavity is obliterated by extensive adhesions. Total destruction of the endometrium can result

in amenorrhea. Tuberculous pyometra can also develop, especially in postmenopausal women with an occluded internal cervical os. P.699 Tuberculous lesions of the cervix are rare. They can be either ulcerative or exophytic and can resemble a primary cervical malignancy or granuloma inguinale of the cervix. When there is a tuberculous lesion of the cervix, the cervical biopsy often reveals tubercles. A tuberculous infection of the ovary usually involves only the surface of the ovary and represents simply an extension of infection from the peritoneal cavity and the adjacent fallopian tubes. The infection is usually limited to a perioophoritis. Extension of the tuberculous infection to the ovarian parenchyma can be prevented by the tunica albuginea. Tuberculous caseation can be found within the ovarian parenchyma, although this is uncommon. Presumably, it occurs as a result of hematogenous spread to the ovarian parenchyma rather than by direct extension through the ovarian capsule. However, a break in the tunica caused by ovulation may also allow the tubercular bacilli to gain access to the ovarian parenchyma. The ovaries are involved in about 25% of cases of pelvic tuberculosis. It is uncommon for tuberculosis to involve the vulva and vagina. It is seen in only 2% of patients with pelvic tuberculosis. The gross appearance may be ulcerative with multiple sinuses, it may be hypertrophic with elephantiasis, or it may be similar to that of carcinoma. Throughout the pelvic organs, the microscopic picture is similar, with tubercles of granulomatous inflammation, Langhans' giant cells, epithelioid cells, and central caseation associated with chronic inflammation. With special stains, acid-fast bacilli can be demonstrated on careful microscopic examination of the tubercles.

Clinical Features of Pelvic Tuberculosis Pelvic tuberculosis occurs most often in patients between the ages of 20 and 40 years. The age of patients with gynecologic tuberculosis has changed in recent years; the proportion of patients older than 40 years of age is now much higher than it was in the past. Falk and associates found that the incidence of pelvic tuberculosis in postmenopausal Swedish women is increasing. This was also the opinion of Sutherland, who reported an investigation from Glasgow in which 26 of 701 patients (3.7%) with proven gynecologic tuberculosis were postmenopausal. The most common clinical symptoms of pelvic tuberculosis include pelvic pain, general malaise, menstrual irregularity, and infertility. Brown and associates found that menstrual irregularity occurred in nearly 50% of patients, whereas amenorrhea or oligomenorrhea was present in 27%. A low-grade fever that on occasion can produce a fulminating septic course is noted in most cases of active or subacute disease. The failure of fever to subside with high doses of broad-spectrum antibiotics is a classic feature of pelvic tuberculosis. A clinical course that is refractory to antibiotic therapy for the usual PID should always alert the clinician to the possibility of tuberculosis. Among patients with pulmonary tuberculosis, the incidence of pelvic tuberculosis generally varies between 10% and 20%. Falk and associates noted that 38% of women with genital tuberculosis had previously had tuberculosis in other organs, usually the lungs. Often, the patient's clinical course is that of a chronic indolent illness.

Diagnosis of Pelvic Tuberculosis The clinical symptoms and signs of pelvic tuberculosis should direct the clinician to the diagnosis. The disease is so uncommon that it is seldom encountered in the gynecologist's usual practice; therefore, the clinical index of suspicion is generally low. In many cases, the clinical presentation is obscure and the diagnosis is delayed. Howard Kelly once said that when competent gynecologists disagree about the diagnosis of an obscure pelvic condition, it usually is diagnosed as either an old ectopic pregnancy or pelvic tuberculosis. More than two thirds of the cases are diagnosed at the time of laparotomy performed for some other indication, or at the time of investigation for infertility or abnormal uterine bleeding. The most common symptom is infertility, and the second most common symptom is lower abdominal and pelvic pain. Some patients are completely asymptomatic and are found to have pelvic tuberculosis during examination for other disorders such as infertility. A dilatation and curettage or endometrial biopsy is diagnostic in some cases, especially if performed in the late premenstrual phase of the menstrual cycle. In addition to standard microscopic sections, the specimen can P.700

be examined by fluorescent antibody technique. Acid-fast staining of tissue or culture of menstrual blood is effective in detecting the organism in about 10% of cases, according to Overbeck. Guinea pig inoculation with menstrual blood may be even more effective. The menstrual blood can be collected in a cervical cap. The culture or inoculation can be repeated many times before a positive result is obtained. Acid-fast stains of tissue suspected of tuberculous infection are important to confirm the diagnosis. Because some acid-fast bacilli are not tuberculous bacilli, it is important to obtain a positive culture whenever possible. A negative evaluation of the endometrium does not rule out pelvic tuberculosis, because the disease can be present in the fallopian tubes without tuberculous endometritis in 30% to 40% of cases. On pelvic examination, bilateral adnexal tenderness is the rule. The tenderness is usually less marked than with acute gonococcal or streptococcal infections. Occasionally, a large tuberculous tuboovarian abscess is palpated on pelvic examination and even felt through the abdominal wall. The classic doughy feel of the broad ligament suggests a tuberculous inflammatory disease that is produced by a combination of thickening of the broad ligament, adherent bowel, and some ascitic fluid. On occasion, cul-de-sac nodules representing tubercles on the serosal surfaces of pelvic organs can be felt. The clinical detection of ascites is the strongest evidence obtainable in favor of pelvic tuberculosis. It was present in one fifth of the cases reported by Brown and associates. However, other causes of ascites must be considered, including ovarian carcinoma and cirrhosis of the liver. In differentiating tuberculous salpingitis from neisserial infections, the finding of a virginal outlet in the presence of obvious tubal inflammation should lend strength to the diagnosis of pelvic tuberculosis. The diagnosis of tuberculosis cannot be made with certainty from a hysterosalpingogram, but it may be helpful. The radiographic criteria for a suspicion of pelvic tuberculosis by hysterosalpingogram have been described by Klein and associates as follows: calcified lymph nodes or smaller, irregular calcifications in the adnexal areas; obstruction of the fallopian tube in the zone of transition between the isthmus and the ampulla; multiple constrictions along the course of the fallopian tube; endometrial adhesions or deformity or obliteration of the endometrial cavity in the absence of a history of curettage or abortion; and vascular or lymphatic extravasation of contrast material. Although a conclusive diagnosis of pelvic tuberculosis can be made only from a positive culture, these authors conclude that hysterosalpingography is a useful aid, especially in patients who are asymptomatic except for infertility. When the diagnosis of pelvic tuberculosis cannot be made in other ways, laparoscopy has been used. Because numerous adhesions may be present, making the introduction of the trocar hazardous, we believe that laparoscopy should be used with particular care. If possible, biopsy specimens of tubal fimbriae or other suspicious areas should be examined histologically or cultured to confirm the diagnosis. In addition to disclosing numerous adhesions, laparoscopy may reveal widespread miliary tubercles involving the omentum and peritoneal surfaces. Matted adnexal masses may be seen. Microscopic examination of peritoneal fluid shows a predominance of lymphocytes. Vaginal cytology is of limited value in diagnosing tuberculosis. The cytologist must be familiar with the morphology of epithelioid cells in the vaginal smear. Only in cases of tuberculosis of the cervix may cytology be helpful. Patients with pelvic tuberculosis should also have an examination and special diagnostic procedures to rule out tuberculous infections in the upper genital tract. Chest radiograph, tuberculin skin test, pelvic ultrasonography, intravenous pyelogram, and urine, gastric, and sputum cultures for M. tuberculosis should be done. In some patients, exploratory laparotomy is needed to make the diagnosis.

Treatment of Pelvic Tuberculosis Before the advent of antituberculous drug therapy, surgery was often used in the treatment of pelvic tuberculosis. Primary surgical treatment was technically difficult, sometimes ineffective, and associated with a high risk of fistula formation and persistent draining sinuses. With the advent of effective drug therapy, the surgical treatment for genital tuberculosis has been restricted to specific indications. Beginning with streptomycin more than 30 years ago, and later isoniazid and para-aminosalicylic acid, it became evident that many cases of pelvic tuberculosis could be cured or controlled with antituberculous drug therapy. There have been major advances in the antibiotic treatment of this disease, including the use of isoniazid with rifampin, with or without ethambutol, given sometimes for a period of 2 years or longer. Sutherland analyzed the results obtained with various drug schedules. The drugs that have been used to treat tuberculosis are isoniazid, rifampin, streptomycin, ethambutol, and pyrazinamide. Isoniazid and rifampin are the most effective and have the lowest toxicity. They should be the foundation of most drug regimens. The addition of ethambutol may not be of benefit, at least not in pulmonary tuberculosis. Severe and sometimes fatal hepatitis, which can develop even after months of treatment, has been associated with isoniazid therapy. The risk of developing hepatitis increases with age and with the daily

consumption of alcohol. Liver function studies should be done before treatment is started, and patients should be carefully monitored with liver function studies throughout the course of therapy and later. The regimen options and dosage recommendations of the American Thoracic Society and the CDC from 1993 for the treatment of tuberculosis are given in Table 28.5 and Table 28.6. TABLE 28.5. Regimen Options for the Initial Treatment of Tuberculosis

Option 1

Option 2

Option 3

Administer daily INH, RIF, and PZA for 8 wk, followed by 16 wk of INH and RIF daily or two to three times per week.a In areas

Administer daily INH, RIF, PZA, and SM or EMB for 2 wk, then administer the same drugs two times per weeka for 6 wk (by DOT).

Treat by DOT, three times per weeka with

where the INH resistance rate is not documented as less than 4%, EMB or SM should be added to the initial regimen until susceptibility to INH and RIF is demonstrated. Continue treatment for at least 6 mo and 3 mo beyond culture conversion. Consult a tuberculosis medical expert if the patient is symptomatic or smear- or culture-positive after 3 mo.

Next, administer INH and RIF two times per week for 16 wk (by DOT). Consult a tuberculosis medical expert if the patient is symptomatic or smear- or culture-positive after 3 mo.

M, RIF, PZA, and EMB or SM for 6 mo.b Consult a tuberculosis medical expert if the patient is symptomatic or smear- or culture-positive after 3 mo.

INH, isomazid; RIF, rifampin; PZA, pyrazinamide; EMB, ethambutol hydrochloride; SM, streptomycin sulfate; DOT, directly observed therapy.

a All regimens administered two times a week or three times a week should be monitored by DOT for the duration of therapy.

b The strongest evidence from clinical trials is the effectiveness of all four drugs administered for the full 6 months. There is weaker evidence that SM can be discontinued after 4 months if the isolate is susceptible to all drugs. The evidence for stopping PZA before the end of 6 months is equivocal for the three times a week regimen, and there is no evidence on the effectiveness of this regimen with EMB for less than full 6 months.

TABLE 28.6. Dosage Recommendations for the Initial Treatment of Tuberculosis

Dosage

Drug

Daily

Two Times a Week

Three Times a Week

Isoniazid

5 mg/kg

15 mg/kg

15 mg/kg

Max 300 mg

Max 900 mg

Max 900 mg

10 mg/kg

10 mg/kg

10 mg/kg

Max 600 mg

Max 600 mg

Max 600 mg

15–30 mg/kg

50–70 mg/kg

50–70 mg/kg

Max 2 g

Max 4 g

Max 3 g

5–25 mg/kg

50 mg/kg

25–30 mg/kg

Max 2.5 g

Max 2.5 g

Max 2.5 g

15 mg/kg

25–30 mg/kg

25–30 mg/kg

Max 1 g

Max 1.5 g

Max 1 g

Rifampin

Pyrazinamide

Ethambutol hydrochloride

Streptomycin sulfate

The therapeutic success of modern antituberculous drug treatment regimens is difficult to assess in view of the limited number of cases available in the literature. The cure rate varies in the literature from 65% to 95%. Kardos removed the fallopian tubes from 168 patients P.701 after medical treatment for 10 months and found active tuberculosis in 35% of the surgical specimens. The experience of Sutherland suggests, however, that the results of treatment may be improved with newer drugs. The patients under treatment must be followed up closely for evidence of regression or remission of the pelvic tuberculosis. Only about 50% of patients with genital tuberculosis have the disease in the endometrial cavity; therefore, repeat endometrial biopsies and culture of menstrual egress provides only limited diagnostic information. The progress of the disease can be monitored closely by evaluating the size of adnexal masses with pelvic examinations and ultrasonography, as well as tracking the ESR, WBC count, and temperature response. Prolonged follow-up is probably indicated in all cases, because recurrence of the tuberculous pelvic lesion 5 years and even later after the end of drug treatment has occasionally been found. Surgery in the management of patients with pelvic tuberculosis should be reserved for specific indications, as outlined by Schaefer and by Sutherland. In general, surgery is reserved for those patients who have failed to respond to an adequate trial of medical therapy. Our indications for the surgical treatment of pelvic tuberculosis include the following: Persistence or enlargement of an adnexal mass after 4 to 6 months of antituberculous antibiotic therapy. The rare possibility of an ovarian tumor must always be considered, even though pelvic tuberculosis is also present. In a 1980 report by Sutherland, the persistence or development of substantial pelvic masses was the indication for surgery in 36 of 91 women with proven tuberculosis of the genital tract treated by surgery. Pelvic ultrasonography should be useful in following the response of adnexal masses to treatment. P.702 Persistence of pelvic pain or recurrence of pelvic pain while on medical therapy. In Sutherland's report, 40 of 91 patients were operated on because of pain. Primary unresponsiveness of the tuberculous infection to antibiotic therapy, as shown by persistent spiking temperature, leukocytosis, elevated ESR, and evidence on biopsy specimens of continued endometrial infection. Of the 91 women in Sutherland's report, 10 were operated on because of persistence of endometrial tuberculosis. Difficulty in obtaining patient cooperation for continued long-term therapy. In these cases, we are accustomed to giving a brief course of streptomycin, 0.5 g every 12 hours intramuscularly for 1 week before surgery, to perform definitive surgery, and then we give 0.5 g every 24 hours in the postoperative period for 2 weeks. A persistent effort should be made to obtain the patient's cooperation for continued antituberculous therapy postoperatively. It is advisable to continue treatment for a year or longer. Isoniazid and rifampin should be used if possible. A common reason for failure of treatment is a tendency for the physician to discontinue drugs after only a few months because the patient appears well. The preferred surgical treatment includes total abdominal hysterectomy and bilateral salpingo-oophorectomy. The nature of this inflammatory disease may make this operative procedure technically difficult, with an increased risk of injury to bowel and bladder. Consequently, in the event of a frozen pelvis from pelvic tuberculosis, it is occasionally necessary to perform only a subtotal abdominal hysterectomy and adnexectomy. Adhesions, which are invariably present and usually widespread, may make the dissection more difficult and injury more likely. However, it is usually possible to do this operation without a high incidence of bowel fistulas and other significant complications. Sutherland reported the results of surgery in 77 patients operated on while antituberculous therapy was administered. There were no deaths, no fistulas, and few late complications. For young patients who are eager to attempt future childbearing, conservative adnexectomy should be carried out only if it is possible to do so after the extent of the adnexal disease is carefully evaluated and is found to be minimal. It is unwise for the surgeon to be committed to a specific operative procedure before the time of surgery, because conservative pelvic surgery for tuberculosis may constitute poor surgical judgment once the operative findings are known. The patient should be forewarned that conservative surgery will be performed only if the disease is minimal and such surgery is considered medically advisable. Conservation of an ovary at the time of operation for pelvic tuberculosis is occasionally possible if the ovary is involved only on the surface. However, if one finds gross evidence of ovarian enlargement or

other gross evidence of infection deep in the ovarian parenchyma, the ovary should be removed. Bisection of ovaries to assess the presence of disease deep in the ovarian parenchyma is not advisable. Reactivation of silent pelvic tuberculosis after tubal reconstructive surgery has been reported by Ballon and associates and by others. We believe that reconstructive tubal surgery has no place in the management of patients whose infertility is the result of bilateral tubal obstruction from tuberculous salpingitis.

Pregnancy After Pelvic Tuberculosis It is evident from the literature, including the studies of both Schaefer and Sutherland, that only about 5% of patients with genital tuberculosis are capable of becoming pregnant, and only 2% carry a pregnancy to term. It is also evident that in the presence of tuberculous tuboovarian abscesses, pregnancy is extremely rare, and conservative surgery for the purpose of preserving fertility is unwarranted. Only when there is minimal pelvic disease without adnexal masses should conservative surgery be considered.

Chapter 29 Surgery for Anomalies of the Müllerian Ducts John A. Rock Lesley L. Breech Maldevelopment of the müllerian ducts occurs in a variety of forms, and each anomaly is distinctive. Nevertheless, some generalizations can be made. Classifications of vaginal anomalies based on certain anatomic findings are useful in organizing the type of malformation, but there usually are exceptions to each rule. Thus, what appears, after a preliminary diagnostic evaluation, to be an apparently isolated vaginal malformation may be found later to be associated with a uterine or renal anomaly. A comprehensive preoperative evaluation of patients with suspected malformations of the müllerian ducts is essential, but a clear understanding of the particular anomaly may not be established until the time of surgical correction. Reproductive surgeons must therefore be equally skilled in both uterine and vaginal reconstruction. The patient with a uterovaginal anomaly often relies entirely on her physician to clarify the reproductive consequences associated with her diagnosis. The physician can help to allay her anxieties by making a prompt evaluation and giving a full and accurate description of the reproductive implications or the obstetric consequences of her particular uterovaginal anomaly. CLASSIFICATION OF UTEROVAGINAL ANOMALIES EMBRYOLOGY CONGENITAL ABSENCE OF THE MULLERIAN DUCTS METHODS OF CREATING A VAGINA DISORDERS OF VERTICAL FUSION DISORDERS OF LATERAL FUSION UNUSUAL CONFIGURATIONS OF VERTICAL-LATERAL FUSION DEFECTS

CLASSIFICATION OF UTEROVAGINAL ANOMALIES Part of "Chapter 29 - Surgery for Anomalies of the Müllerian Ducts" Classifications of uterovaginal anomalies originally were organized on the basis of clinical findings. Our improved understanding of the embryologic development of most uterovaginal anomalies has enabled categorization on this basis. The 1988 American Fertility Society (AFS) classification of müllerian anomalies (Table 29.1) offers an alternative based on the degree of failure of normal uterine development. Anomalies are grouped according to similarities of clinical manifestations, treatment, and prognosis for fetal salvage. The AFS classification system is weighted primarily toward disorders of lateral fusion and does not include associated vaginal anomalies, although the scheme does allow the user to describe anomalies involving the vagina, tubes, and urinary tract as associated malformations. TABLE 29.1. American Fertility Society Classification of Müllerian Anomaliesa

Classification

Anomaly

Class I

Segmental, müllerian agenesis–hypoplasia

A. Vaginal

B. Cervical

C. Fundal

D. Tubal

E. Combined anomalies

Class II

Unicornuate

A. Communicating

B. Noncommunicating

C. No cavity

C. No cavity

D. No horn

Class III

Didelphus

Class IV

Bicornuate

A. Complete (division down to internal os)

B. Partial

Class V

Septate

A. Complete (septum to internal os)

B. Partial

Class VI

Arcuate

Class VII

Diethylstilbestrol related

a This classification allows the user to indicate the malformation type and provides additional findings to describe associated variations involving the vagina, cervix, tubes (right, left), and kidneys (right, left).

Adapted from the American Fertility Society. Classification of müllerian anomalies. Fertil Steril 1988;49:944.

No classification of müllerian maldevelopment can focus entirely on the uterus, however. The vagina is often involved, and sometimes the tubes are involved as well. This discussion follows a suggested modification of the AFS classification of uterovaginal anomalies (Table 29.2) P.706 that comprises four groups based on embryologic considerations.

TABLE 29.2. American Fertility Society Classification of Uterovaginal Anomalies

CLASS I. DYSGENESIS OF THE MULLERIAN DUCTS

CLASS II. DISORDERS OF VERTICAL FUSION OF THE MULLERIAN DUCTS

A. Transverse Vaginal Septum Obstructed Unobstructed

B. Cervical Agenesis or Dysgenesis

CLASS III. DISORDERS OF LATERAL FUSION OF THE MULLERIAN DUCTS

A. Asymmetric-Obstructed Disorder of Uterus or Vagina Usually Associated With Ipsilateral Renal Agenesis Unicornuate uterus with a noncommunicating rudimentary anlage or horn Unilateral obstruction of a cavity of a double uterus Unilateral vaginal obstruction associated with double uterus

B. Symmetric-Unobstructed Didelphic uterus a. Complete longitudinal vaginal septum b. Partial longitudinal vaginal septum c. No longitudinal vaginal septum Septate uterus a. Complete Complete longitudinal vaginal septum Partial longitudinal vaginal septum No longitudinal vaginal septum b. Partial Complete longitudinal vaginal septum Partial longitudinal vaginal septum No longitudinal vaginal septum Bicornuate uterus a. Complete Complete longitudinal vaginal septum Partial longitudinal vaginal septum No longitudinal vaginal septum b. Partial Complete longitudinal vaginal septum Partial longitudinal vaginal septum No longitudinal vaginal septum T-shaped uterine cavity (diethylstilbestrol related) Unicornuate uterus

Partial longitudinal vaginal septum No longitudinal vaginal septum b. Partial Complete longitudinal vaginal septum Partial longitudinal vaginal septum No longitudinal vaginal septum T-shaped uterine cavity (diethylstilbestrol related) Unicornuate uterus a. With a rudimentary horn With endometrial cavity a) Communicating b) Noncommunicating Without endometrial cavity b. Without a rudimentary horn

CLASS IV. UNUSUAL CONFIGURATIONS OF VERTICAL-LATERAL FUSION DEFECTS

Modified from the American Fertility Society. Classification of müllerian anomalies. Fertil Steril 1988;49:944.

Class I. Dysgenesis of the Müllerian Ducts Dysgenesis of the müllerian ducts, which includes agenesis of the uterus and vagina (the Mayer-Rokitansky-Küster-Hauser syndrome), is an impairment of the reproductive system characterized by no reproductive potential other than that achieved by in vitro fertilization in a host uterus.

Class II. Disorders of Vertical Fusion of the Müllerian Ducts Disorders of vertical fusion can be considered to represent faults in the junction between the down-growing müllerian ducts (müllerian tubercle) and the up-growing derivative of the urogenital sinus. Typically, these disorders are characterized by an atretic portion of vagina that can be quite thick, extending through more than half the distance of the vagina, or it can be quite thin and limited to a small obstructing membrane. Regardless of the length of the septum, a disorder of vertical fusion should be regarded as a transverse vaginal septum and classified as either obstructed or unobstructed. The so-called partial vaginal agenesis with uterus and cervix present is probably a misnomer for a large segment of atretic vagina. Cervical agenesis or dysgenesis P.707 is also included in the group of disorders of vertical fusion.

Class III. Disorders of Lateral Fusion of the Müllerian Ducts Disorders of lateral fusion of the two müllerian ducts can be symmetric-unobstructed, as with the double vagina, or asymmetric-obstructed, as with unilateral vaginal obstruction. Obstructions associated with disorders of lateral fusion are particularly noteworthy in that they are observed clinically only as unilateral obstructions that almost invariably are associated with absence of the ipsilateral kidney. Bilateral obstruction is thought to be associated with bilateral kidney agenesis and subsequent nonviability of the developing embryo. The three varieties of asymmetric obstruction with ipsilateral renal agenesis are as follows: Unicornuate uterus with a noncommunicating horn that contains menstruating endometrium Unilateral obstruction of a cavity of a double uterus Unilateral vaginal obstruction The five groups of symmetric-unobstructed disorders of lateral fusion are as follows: The didelphic uterus

The septate uterus The bicornuate uterus The T-shaped uterine cavity, which may be hypoplastic and irregular, and which is associated with diethylstilbestrol (DES) exposure in utero The unicornuate uterus with or without a rudimentary horn The first three groups are types of double uteri; differentiation between a septate uterus (second group) and a bicornuate uterus (third group) requires visualization of the fundus. The septum within the septate uterus is complete or partial. When the septum is complete, there inevitably are two cervices with a longitudinal vaginal septum that can extend to the introitus or partially down the vagina. The bicornuate uterus also can have a partial or almost complete separation of the uterine cavities. The term arcuate uterus is used primarily by radiologists to refer to a slight septum in the uterine fundus that forms no clear separation of the uterine cavities. This type of uterus is usually included in the category of partial septate uterus. The unicornuate uterus may have an attached horn with a cavity that communicates with the unicornuate uterus, or there may be no uterine horn or a uterine horn with no cavity. Some debate has focused on whether the unicornuate uterus with a communicating horn can represent a hypoplastic side of a bicornuate uterus.

Class IV. Unusual Configurations of Vertical-Lateral Fusion Defects This final category includes combinations of uterovaginal anomalies and other disorders. Unusual uterovaginal configurations have been described that do not fit a particular category, and vertical and lateral fusion disorders can coexist. Unusual configurations of vertical-lateral fusion defects can be seen with abnormalities of the lower urinary tract. Singh and co-workers have described a patient who was noted to have a persistent hymen and a longitudinal vaginal septum with a didelphic uterus. The patient was noted also to have a double urethra and bladder and left renal agenesis. Obstructive lesions require immediate attention to relieve retrograde flow of trapped mucus and menstrual blood and increasing pressure on surrounding organs and structures. When no obstruction is present, attention may not be required immediately, but it will always be required eventually to establish or improve reproductive or coital function.

EMBRYOLOGY Part of "Chapter 29 - Surgery for Anomalies of the Müllerian Ducts" The reproductive organs in the female (and in the male) consist of external genitalia, gonads, and an internal duct system between the two. These three components originate embryologically from different primordia and in close association with the urinary system and hindgut. Thus, the developmental history is complex (Fig. 29.1 and Fig. 29.2). Even in the 3.5- to 4-mm embryo, it is possible to recognize the bilateral thickenings of the coelomic epithelium known as the gonadal ridges medial to the mesonephros (primitive kidney) in the dorsum of the coelomic cavity. At about the sixth week of gestation, in the 17- to 20-mm embryo, the gonad can be distinguished as either a testis or an ovary.

View Figure

FIGURE 29.1. Diagrammatic representation of the development of the female reproductive organs and structures in early embryogenesis. A: At the 18- to 20-somite stage (fourth week), the gonadal ridges have not yet begun to form. B: In the 3.5- to 4-mm embryo (fifth week), the gonadal ridges can be recognized as thickenings of the coelomic cavity just medial to the mesonephric tubules. (Gonadal differentiation into either testis or ovary does not occur until the sixth week of development.) The allantoenteric diverticulum is joined caudally to the dilated cloaca. C: and D: The genital tubercle and labial folds form in the region just anterior to the cloaca. The cloaca later divides into the ventral urogenital sinus and the dorsal rectum. The development of the urinary system closely parallels that of the reproductive system. The nonfunctioning pronephric tubules shown in (A) develop to form the mesonephric ducts shown in (B) and (C). The permanent kidneys eventually develop from the metanephric tissue, and the urinary collecting system develops from the metanephric ducts. The paramesonephric (müllerian) ducts are apparent by the 12- to 14-mm stage (D). (Their subsequent development is illustrated in Figure 29.2.)

View Figure

FIGURE 29.2. Further development of the paramesonephric (müllerian) ducts and the urogenital sinus. A: Early development of the paramesonephric ducts. The cranial ends of the paramesonephric ducts develop first. These ends remain open to form the fimbriated ends of the fallopian tubes. The paramesonephric ducts grow caudally and cross the mesonephric ducts ventrally. B: Eventually, they fuse together to form the uterovaginal canal. C: Further caudal development brings this structure into contact with the wall of the urogenital sinus, producing the müllerian tubercle. The caudal ends of the fused paramesonephric ducts form the uterine corpus and cervix. Together with the urogenital sinus, they also form the vagina. The cranial point of fusion of the paramesonephric ducts marks the location of the future uterine fundus. The fallopian tubes form from the unfused cranial parts of the paramesonephric (müllerian) ducts. The proliferation of the lining of the uterovaginal canal above the upward growth of the sinovaginal bulb from below (D) forms the vaginal plate (E), which later becomes canalized to leave an open vaginal canal. Thus, the vagina is of composite origin. The mesonephric ducts in the female degenerate but can persist into adult life as Gartner's ducts.

In the female, the labia minora and majora develop from the labioscrotal folds, which are ectodermal in origin. The phallic portion of the urogenital sinus gives rise to the urethra. The müllerian (paramesonephric) duct system is stimulated to develop preferentially over the wolffian (mesonephric) duct system, which regresses in early female fetal life. The cranial parts of the wolffian ducts can persist as the epo?phoron of the ovarian hilum; the caudal parts can persist as Gartner's ducts. The müllerian ducts persist and attain complete development to form the fallopian tubes, the uterine corpus and cervix, and a portion of the vagina.

Origin of the Müllerian Ducts About 37 days after fertilization, the müllerian ducts first appear lateral to each wolffian duct as invaginations of the dorsal coelomic epithelium. The site of origin of the invaginations remains open and ultimately forms the fimbriated ends of the fallopian tubes. At their point of origin, each of the müllerian ducts forms a solid bud. Each bud penetrates the mesenchyme lateral and parallel to each wolffian duct. As the solid buds elongate, a lumen appears in the cranial part, beginning at each coelomic opening. The lumina extend gradually to the caudal growing tips of the ducts. Eventually, the caudal end of each müllerian duct crosses the ventral aspect of the wolffian duct. The paired müllerian ducts continue to grow in a medial and caudal direction until they eventually meet in the midline and become fused together in the urogenital septum. A septum between the two müllerian ducts gradually disappears, leaving a single uterovaginal canal lined with cuboidal epithelium. Failure of reabsorption of this septum can result in a septate uterus. The most cranial parts of the müllerian ducts remain separate and form the fallopian tubes. The caudal segments of the müllerian ducts fuse to form the uterus and part of the vagina. The cranial point of fusion is the site of the future fundus of the uterus. Variations in this site of fusion can result in an arcuate or bicornuate uterus. Complete failure of fusion can result in a didelphic uterus.

Development of the Vagina The vagina is formed from the lower end of the uterovaginal canal, which developed from the müllerian

ducts and the urogenital sinus (Fig. 29.2). The point of contact between the two is the müllerian tubercle. A solid vaginal cord results from proliferation of the cells at the caudal tip of the fused müllerian ducts. The cord gradually elongates to meet the bilateral endodermal evaginations (sinovaginal bulbs) from the posterior aspect of the urogenital sinus below. These sinovaginal bulbs extend cranially to fuse with the caudal end of the vaginal cord, forming the vaginal plate. Subsequent canalization of the vaginal cord occurs, followed by epithelialization with cells derived mostly from endoderm of the urogenital sinus. Recent proposals hold that only the upper one third of the vagina is formed from the müllerian ducts and that the lower vagina develops from the vaginal plate of the urogenital sinus. Recent studies also suggest that the vaginal canal is actually open and connected to a patent uterus and tubes, even in early embryonic life, and that the vagina does not form and later become canalized from an epithelial cord of squamous cells growing upward from the urogenital sinus. Most investigators now suggest that the vagina develops under the influence of the müllerian ducts and estrogenic stimulation. There is general agreement that the vagina is a composite formed partly from the müllerian ducts and partly from the urogenital sinus. At about the 20th week, the cervix takes form as a result of condensation of stromal cells at a specific site around the fused müllerian ducts. The mesenchyme surrounding the müllerian ducts becomes condensed early in embryonic development and eventually forms the musculature of the female genital tract. The hymen is the embryologic septum between the sinovaginal bulbs above and the urogenital sinus proper below. It is lined by an internal layer of vaginal epithelium and an external layer of epithelium derived from the urogenital sinus (both of endodermal origin), with mesoderm between the two. It is not derived from the müllerian ducts.

Anomalies in Organogenesis of the Vagina Anomalies in the organogenesis of the vagina are easily understood. If there is failure in the development of the müllerian ducts at any time between their origin from the coelomic epithelium at 5 weeks of embryonic age and their fusion with the urogenital sinus at 8 weeks, the sinovaginal bulbs will fail to proliferate from the urogenital sinus and the uterus and vagina will fail to develop. Congenital absence of the uterus and the vagina, known as the Mayer-Rokitansky-Küster-Hauser syndrome, is the most common clinical example of this anomaly.

Transverse Vaginal Septum A transverse vaginal septum can develop at any location in the vagina but is more common in the upper vagina at the point of junction between the vaginal plate and the caudal end of the fused müllerian ducts. This defect presumably is caused by failure of absorption of the tissue that separates the two or by failure of complete fusion of the two embryologic components of the vagina. A large segment of vagina can be atretic. In past reviews, this has been termed partial vaginal agenesis with a uterus present. Elucidation of the cause of a high transverse vaginal septum is more difficult. A local abnormality of the vaginal mesoderm or failure of canalization of the epithelial vaginal plate can provide the answer, but why the abnormality should occur at this particular site is not evident. The proportion of the vagina originating from the urogenital sinus can at times be considerably more than one fifth, and a high transverse vaginal septum thus may represent the junction of an abnormally long urogenital sinus contribution and a short müllerian portion. Alternatively, the high transverse septum could be the sequela of a local infection of the septum at the end of the vagina. Septa in other areas of the vagina are unexplained by this theory, which has not gained widespread acceptance.

Disorders of Ineffective Suppression of Müllerian Ducts When abnormal gonadal development is caused by ineffective suppression of the müllerian ducts, ambiguous external genitalia frequently are accompanied by a small rudimentary uterus or a partially developed vagina. Additionally, when there is a genetic loss of cytoplasmic receptor proteins within androgenic target cells, such as occurs in the androgen insensitivity syndrome (formerly called testicular feminization syndrome), the vagina is incompletely developed because the existing male gonads suppress the development of the müllerian ducts. Because these genetically male patients are seen clinically as phenotypic XY females without a completely formed vagina, it is important that a vagina be surgically constructed so that these patients may have satisfactory sexual function in their female gender role. Congenital rectovaginal fistula, imperforate (covered) anus, hypospadias, and other anatomic variants of cloacal dysgenesis also can occur. These anomalies can be P.711

associated with maldevelopment of the müllerian and mesonephric duct derivatives.

Müllerian Duct Abnormalities Abnormalities in the formation or fusion of the müllerian ducts can result in a variety of anomalies of the uterus and vagina—single, multiple, combined, or separate. Just as the entirely separate origin of the ovaries from the gonadal ridges accounts for the infrequent association of uterovaginal anomalies with ovarian anomalies (see Chapter 26), so do the close developmental relationships of the müllerian and wolffian ducts explain the frequency with which anomalies of the female genital system and urinary tract are associated. Failure of development of a müllerian duct is likewise associated with failure of development of a ureteric bud from the caudal end of the wolffian duct. Thus, the entire kidney can be absent on the side ipsilateral to the agenesis of a müllerian duct. Depending on the timing of the teratogenic influence, renal units can be absent, fused, or in unusual locations in the pelvis. Ureters can be duplicated or can open in unusual places such as the vagina or uterus. Jones and Rock have pointed out that failure of lateral fusion of the müllerian ducts with unilateral obstruction is associated consistently with absence of the kidney on the side with obstruction. Bilateral obstruction has not been observed clinically, presumably because it would be associated with bilateral renal agenesis, a condition that would not allow the embryo to develop. According to Thompson and Lynn, 40% of female patients with congenital absence of the kidney are found to have associated genital anomalies. Much investigation has been undertaken to determine a genetic relationship in the development of disorders of the müllerian ducts. Familial aggregates of the most common disorders of the müllerian differentiation are best explained on the basis of polygenic or multifactorial inheritance. No information exists on the number and chromosomal location of responsible genes. Single mutant genes are responsible for the McKusick-Kaufman syndrome and the hand-foot-genital syndrome. Hand-foot-genital syndrome is a rare, dominantly inherited condition that affects both the distal limbs and the genitourinary tract. A nonsense mutation of the HOXA13 gene has been identified in several families. HOX gene mutations have been reported in several families with multiple müllerian abnormalities. Reproductive abnormalities involving the uterus and vagina may also be associated with other more complex malformation syndromes, in which the molecular basis of many of the syndromes remains unknown.

METHODS OF CREATING A VAGINA Part of "Chapter 29 - Surgery for Anomalies of the Müllerian Ducts" There is no unanimity of opinion regarding the correct approach to the problem of vaginal agenesis (Table 29.3). With the development of the Ingram method for vaginal dilatation, fewer patients require surgical vaginoplasty. The role of tissue expanders in vaginoplasty has been reviewed by Patil and Hixon. Labial expansion with an expander having a capacity of 250 mL provides a flap 10 cm long and 8 cm wide with a 4-cm projection. Thus, well-vascularized flaps can be available to provide an outlet for stenosis-free vaginoplasty. This approach has been suggested to maximize the success of surgical vaginoplasty. A review of the methods devised for the formation of a vagina follows. The editors of this book have found the modified McIndoe technique to give the most consistently satisfactory results. TABLE 29.3. Classification of Methods to Form a New Vagina

NONSURGICAL (INTERMITTENT PRESSURE ON THE PERINEUM)

Active Dilatation

Passive Dilatation

SURGICAL

Without Use of Abdominal Contents

Without cavity dissection

Vulvovaginoplasty

Constant pressure (Vecchietti)

No attempt to line cavity (now unacceptable)

Lining cavity with grafts

Split-thickness skin grafts (McIndoe operation)

Dermis grafts

Dermis grafts

Amnion homografts

Lining cavity with flaps

Musculocutaneous flaps

Fasciocutaneous flaps

Subcutaneous pedicled skin flaps

Labial skin flaps (can be created with tissue expander)

Penoplasty (transsexualism)

With Use of Abdominal Contents (Cavity Lining With)

Peritoneum

Free intestinal graft

Pedicled intestine

Nonsurgical Methods In 1938, Frank described a method of creating an artificial vagina without operation. In 1940, he reported remarkably satisfactory results in eight patients treated with this method. His follow-up study showed that a vagina formed in this manner remained permanent in depth and caliber, even in patients who neglected dilatation for more than 1 year. It has been emphasized that the pelvic floor itself is embryologically deficient in some patients. Indeed, the ease with which some patients are able to create a vagina with intercourse alone or with other intermittent pressure techniques can be explained on this basis. Five patients were reported to have developed enteroceles, one after coitus alone, three after a Williams vulvovaginoplasty, and three after a McIndoe operation. This complication can develop when the vaginal mucosa is brought in close proximity to the pelvic peritoneum, but a relative embryologic weakness or an absence of endopelvic fascia can also contribute to this complication. Rock, Reeves, and associates at the Johns Hopkins Hospital reported that an initial trial of vaginal dilatation was successful in 9 of 21 patients. Prompted by the rewarding results of Broadbent and Woolf, Ingram has described a passive dilatation technique of creating a new vagina. Instructing his patients in the insertion of dilators (Fig. 29.5) specially

designed for use with a bicycle seat stool, Ingram was able to produce satisfactory vaginal depth and coital function in 10 of 12 cases of vaginal agenesis and 32 of 40 cases of various types of stenosis. FIGURE 29.5. Vaginal dilators for use in Ingram passive dilatation technique to create a new vagina. The set consists of 19 dilators of increasing length and width. (Courtesy of Faulkner Plastics, Tampa, FL)

View Figure

P.716 The Ingram technique for passive dilatation has several advantages. The patient is not required to press the dilator against the vaginal pouch. A series of graduated Lucite dilators slowly and evenly dilate the neovaginal space. The patient should be carefully instructed in the use of dilators, as recommended by Ingram, beginning with the smallest dilator. The patient is shown and instructed with the use of a mirror how to place a dilator against the introital dimple. The dilator may be held in place with a supportive undergarment and regular clothing worn over this. The patient is shown how to sit on a racing type bicycle seat that is placed on a stool 24 inches above the floor. She is instructed to sit leaning slightly forward with the dilator in place for at least 2 hours per day at intervals of 15 to 30 minutes. Follow-up is usually at monthly intervals and the patient can expect to graduate to the next size larger dilator about every month. An attempt at sexual intercourse may be suggested after the use of the largest dilator for 1 or 2 months. Continued dilatation is recommended if intercourse is infrequent. In our experience, functional success rates are outstanding. Rock and Roberts reported the largest series of vaginal agenesis patients who used the Ingram method of dilatation to create a neovagina. The records of 51 patients with müllerian agenesis were reviewed: 37 patients attempted vaginal dilatation and 14 young women underwent a surgical intervention. Functional success was defined as satisfactorily achieving intercourse or accepting the largest dilator without discomfort in the clinic visit. All patients were followed up for at least 2 years and for an average of 9.25 years. Functional success was achieved in 91.9% of those who attempted dilatation (Table 29.4). Thus, passive dilatation should be suggested as an initial therapy for vaginal creation. If dilatation is unsuccessful, operative vaginoplasty is indicated. TABLE 29.4. Outcomes of Patients With Vaginal Agenesis Who Attempted Dilatation

Patients

Totals

Percent

Successful dilatation

34/37

91.9%a

Failed dilatation

3/37

8.1%

a p < .001.

Modified from Roberts CP, Haber MJ, Rock JA. Vaginal creation for muüllerian agenesis. Am J Obstet Gynecol 2001;185:1349.

Surgical Methods During the past 3 decades, experience has proved the Abbe-Wharton-McIndoe procedure (more popularly called the McIndoe operation) for dealing with complete absence of a vagina to be generally superior to others in most cases. In special circumstances, alternative methods of creation of a neovagina may be indicated.

Historical Development of Surgical Procedures In 1907, Baldwin used a double loop of ileum to line a space dissected between the rectum and bladder, leaving the mesentery connected to the bowel. The continuity of the intestinal tract was reestablished by an end-to-end anastomosis. He reported that the new vagina was absolutely normal in every way. In 1910, Popaw constructed a vagina using a portion of the rectum that was moved anteriorly. This operation was modified by Schubert in 1911. The rectum was severed above the anal sphincter and moved anteriorly to serve as the vagina. The sigmoid was sutured to the anus to reestablish the continuity of the intestinal tract. Both operations had soberingly high morbidity and mortality rates, and their popularity declined. Today, segments of sigmoid are used most often to create a vaginal pouch or extend vaginal length in patients who have lost vaginal function as a result of extensive surgery or irradiation for pelvic malignancy. Some patients who are treated for multiple genitourinary or gastrointestinal abnormalities may be treated with a bowel vaginoplasty during a combined procedure. P.717 Less formidable procedures involving dissection of a space between the bladder and rectum and lining of this space with flaps of skin from the labia or inner thighs also were tried. Marked scarring resulted, and hair usually grew in the vagina. Extensive plastic procedures to construct a vagina are no longer necessary or desirable and have been discarded in favor of safer procedures unless there is the problem of maintaining a vaginal canal after an extensive exenterative operation for pelvic malignancy. In this case, the physician may want to consider using the gracilis myocutaneous flap technique described by McCraw

and associates in 1976.

The Abbe-Wharton-McIndoe Operation The operation most popular today for creating a new vagina began with simple surgical attempts to create a space between the bladder and the rectum. These early attempts were often made in patients with cryptomenorrhea. However, such a space usually would constrict because the surgeon would fail to recognize the importance of prolonged continuous dilatation until the constrictive phase of healing was complete. At the Johns Hopkins Hospital in 1938, Wharton combined an adequate dissection of the vaginal space with continuous dilatation by a balsa form that was covered with a thin rubber sheath and was left in the space. He did not use a split-thickness skin graft. Instead, he based his operation on the principle that the vaginal epithelium has remarkable powers of proliferation and in a relatively short time will cover the raw surface. Recalling that a similar process occurs in the fetus when the epithelium of the sinovaginal bulbs and the urogenital sinus form the vaginal canal, Wharton merely applied this same principle in the adult. This simple procedure is entirely satisfactory as long as the space is kept dilated long enough to allow the epithelium to grow in. Occasionally, however, even after several years, the vault of the vagina remains without epithelial covering. Coital bleeding and leukorrhea result from the persistent granulation tissue, and there is a tendency for vaginas constructed by this method to be constricted by scarring in the upper portion. In Counseller's 1948 report from the Mayo Clinic of 100 operations to construct a new vagina, 14 were performed by Wharton's method, with excellent results in all 14 patients. It was stated that the disadvantages of persistent granulation tissue with bleeding and leukorrhea were of no consequence. This has not been the experience of the editors of this book. When inlay skin grafts were first used to construct a new vagina, the results were poor because the necessity for dilatation of the new vagina again was not recognized. Severe contraction, uncontrolled by continuous or intermittent dilatation, almost invariably spoiled the results. Although Heppner, Abbe, and others preceded him by many years in using a skin-covered prosthesis in neovaginal construction, it was Sir Archibald McIndoe, at the Queen Victoria Hospital in England, who popularized the method and gave it substantial clinical trial. He emphasized the three important principles used today in successful operations for vaginal agenesis: Dissection of an adequate space between the rectum and the bladder Inlay split-thickness skin grafting The cardinal principle of continuous and prolonged dilatation during the contractile phase of healing Other tissues such as amnion and peritoneum have been used to line the new vaginal space, but they have not had substantial success. However, Tancer and associates reported good results with human amnion. Karjalainen and associates stated that a more physiologic result was achieved with an amnion graft than with a skin graft. Nevertheless, concerns about the transmission of human immunodeficiency virus with human amnion now limit this option.

TECHNIQUE OF ABBE-WHARTON-MCINDOE OPERATION Taking the Graft After a careful pelvic examination is performed under anesthesia to verify previous findings, the patient is positioned for taking a skin graft from the buttocks. For cosmetic reasons, the graft should not be taken from the thigh or hip unless for some reason it cannot be obtained from the buttocks. Patients may be asked to sunbathe in a brief bathing suit before coming to the hospital so that its outline can be seen; an attempt should be made to take the graft from both buttocks within these borders. The quality of the graft determines to a great extent the success of the operation. We have found the Padgett electrodermatome to be the most satisfactory instrument for taking the graft. With relatively little experience and practice, the gynecologic surgeon can successfully cut a graft of controlled width and thickness (Fig. 29.6). The instrument is set and checked for taking a graft approximately 0.018 inch thick and 8 to 9 cm wide. The total graft length should be 16 to 20 cm. If the entire graft cannot be taken from one buttock, then a graft 8 to 10 cm long is needed from each buttock.

FIGURE 29.6. Section of split-thickness skin grafts. Grafts should be uniform in thickness. The Padgett electrodermatome is set to take a graft approximately 0.018 inch thick. A graft that is slightly thick is better than a thin graft.

View Figure

The skin of the donor site is prepared with an antiseptic solution (povidone-iodine), which is then thoroughly washed away. The skin is then lubricated with mineral oil as assistants steady and stretch the skin tight. P.718 Considerable pressure should be applied uniformly across the dermatome blade. The thickness of the graft must have minimal variation. A graft that is a little too thick is better than one that is a little too thin. There should be no breaks in the continuity of the graft. The graft is placed between two layers of moist gauze and the donor sites are dressed. The donor site is soaked with a dilute solution of epinephrine for hemostasis, and a sterile dressing is applied. A pressure dressing is then placed over the site; this dressing can be removed on the seventh postoperative day. The sterile dressing dries in place over the donor site and ultimately will fall off by itself. Moistened areas on the dressing can be dried with cool air. If there is separation and evidence of some superficial infection, then merbromin can be applied to these areas.

Creating the Neovaginal Space The patient is placed in the lithotomy position and a transverse incision is made through the mucosa of the vaginal vestibule (Fig. 29.7A). The space between the urethra and bladder anteriorly and the rectum posteriorly is dissected until the undersurface of the peritoneum is reached. This step may be safer with a catheter in the urethra and sometimes a finger in the rectum to guide the dissection in the proper plane. After incising the mucosa of the vaginal vestibule transversely, the physician often is able to create a channel on each side of a median raphe (Fig. 29.7B), starting with blunt dissection and then dilating each channel with Hegar dilators or with finger dissection. In some instances, it may be necessary to develop the neovaginal space by dissecting laterally and bringing the fingers toward the midline. The median raphe is then divided, thus joining the two channels. This maneuver is helpful in dissecting an adequate space without causing injury to surrounding structures.

FIGURE 29.7. The McIndoe procedure. A: A transverse incision is made in the apex of the vaginal dimple. B: A channel can usually be dissected on each side of the median raphe. The median raphe is then divided. Careful dissection prevents injury to the bladder and rectum. C: A space between the urethra and bladder anteriorly and the rectum posteriorly is dissected until the undersurface of the peritoneum is reached. Incision of the medial margin of the puborectalis muscles will enlarge the vagina laterally.

View Figure

To avoid subsequent narrowing of the vagina at the level of the urogenital diaphragm, it may be helpful to incise the margin of the puborectalis muscles bilaterally along the midportion of the medial margin (Fig. 29.7C). Although useful in all circumstances, incision of the puborectalis muscle is more important in cases of androgen P.719 insensitivity syndrome with android pelvis, in which the levator muscles are more taut against the pelvic diaphragm, than in cases of gynecoid pelvis. Incision of the puborectalis muscle causes no difficulty with fecal incontinence, significantly improves the ease with which the vaginal form can be inserted into the canal in the postoperative period, and has eliminated the problem of contracture of the upper vagina caused by a poorly applied form. The dissection should be carried as high as possible without entering the peritoneal cavity and without cleaning away all tissue beneath the peritoneum. A split-thickness skin graft will not take well when applied against a base of thin peritoneum. All bleeders should be ligated by clamping and tying them with very fine sutures. It is essential that the vaginal cavity be dry to prevent bleeding beneath the graft. Bleeding causes the graft to separate from its bed, resulting in the inevitable failure of the graft to implant in that area and in local graft necrosis.

Preparing the Vaginal Form Early skin grafts were formed over balsa, which has the advantages of being an inexpensive, easily available, light wood that can be sterilized without difficulty. It also can be whittled easily in the operating room to a proper shape to fit the new vaginal space. However, uneven pressure from the form can cause a skin graft to slough in places, and pressure spots are associated with an increased risk of fistula formation. The Counseller-Flor modification of the McIndoe technique (Fig. 29.8) uses, instead of the rigid balsa form, a foam rubber mold shaped for the vaginal cavity from a foam rubber block and covered with a condom. The foam rubber is gas sterilized in blocks measuring approximately 10 × 10 × 20 cm. The block is shaped with scissors to approximately twice the desired size, compressed into a condom, and placed into the neovagina (Fig. 29.8A, Fig. 29.8B and Fig. 29.8C). The form is left in place for 20 to 30 seconds with the condom open to allow the foam rubber to expand and conform to the neovaginal space (Fig. 29.8D). The condom is then closed and the form is withdrawn. The external end is tied with 2-0 silk, and an additional condom is placed over the form and tied securely (Fig. 29.8E and Fig. 29.8F).

FIGURE 29.8. Counseller-Flor modification of the McIndoe technique. A: A form is cut from a foam rubber block. B: A condom is placed over the form. C: The form is compressed and placed into the vagina. D: Air is allowed to expand the foam rubber, which accommodates to the neovaginal space. The condom is closed and the form removed. A second condom is placed over the form (E) and tied securely (F). G: The graft is then sewn over the form with interrupted 5-0 nonreactive sutures. H: The undersurfaces of the sutured edges of the graft are exteriorized. The vaginal form is ready for insertion into the neovagina.

View Figure

Sewing the Graft Over the Vaginal Form The skin graft is then placed over the form and its undersurface exteriorized and sewn over the form with interrupted vertical mattress 5-0 nonreactive sutures (Fig. 29.8G and Fig. 29.8H). Where the graft is approximated, the undersurfaces of the sutured edges are also exteriorized. The graft should not be “meshed” to make it stretch farther, and the edges of the graft should be approximated meticulously around the form without gaps. Granulation tissue develops at any place where the form is not covered with skin. Contraction usually occurs where granulation tissue forms. After the form has been placed in the neovaginal space, the edges of the graft are sutured to the skin edge with 5-0 nonreactive absorbable sutures, with sufficient space left between sutures for drainage to occur. The physician must be careful not to have the form so large that it causes undue pressure on the urethra or rectum. A balsa form should have a groove to accommodate the urethra. With a foam rubber form, this is unnecessary. A suprapubic silicone catheter is placed in the bladder for drainage. If the labia are of sufficient length, then the form can be held in place by suturing the labia together with two or three nonreactive sutures.

Replacing with a New Form After 7 to 10 days, the form is removed and the vaginal cavity is irrigated with warm saline solution and inspected. This is usually performed with mild sedation and without an anesthetic. The cavity should be inspected carefully to determine whether the graft has taken satisfactorily in all areas of the new vagina. Any undue pressure by the form should be noted and corrected. It is especially important that there not be too much pressure superiorly against the peritoneum of the cul-de-sac. Such a constant upward pressure could result in weakness with subsequent enterocele formation. The new vaginal cavity must be inspected frequently to detect and to prevent pressure necrosis of the skin graft. The patient is given instructions on daily removal and reinsertion of the form and is taught how to administer a low-pressure douche of clear warm water. She is advised to remove the form at the time of urination and defecation, but otherwise to wear it continuously for 6 weeks. A neoprene form, which is much easier to remove and keep clean than a foam rubber form, is substituted for the original form in 6 weeks. A new form is molded with a sterile sheath cover (condom) to fit the size of the vaginal canal. The patient is instructed to use the form during the night for the following 12 months. If there has been no change in the caliber of the vagina by that time, then it is unlikely to occur later, and insertion of the form at night can be done intermittently until coitus is a frequent occurrence. However, if there is the slightest difficulty in inserting the form, then the patient should be advised to use the form continuously again. Most patients are able to maintain the form in place simply by wearing a panty girdle and perineal pad. Douches are advisable during residual vaginal healing and discharge.

RESULTS AND COMPLICATIONS Results with the McIndoe operation have improved over the years. Recently reported percentages of satisfactory results have ranged from 80% to 100%. The serious complications formerly associated with the McIndoe operation have been significantly reduced by improvements in technique and greater experience. Serious complications do still occur, however, including a 4% postoperative fistula rate (urethrovaginal, vesicovaginal, and rectovaginal), postoperative infection, and intraoperative and postoperative hemorrhage. Failure of graft take is also still reported as an occasional complication. Failure of graft take often leads to the development of granulation tissue, which might require reoperation, curettement of the granulation tissue down to a healthy base, and even regrafting. Minor granulation can be treated with silver nitrate application. The functional result is more P.720 important than the anatomic result in evaluating the success of this operation. Although a vagina of only 4 cm is adequate for some couples, in most instances a vagina smaller than 4 cm causes major problems. The postoperative results have improved significantly since the balsa vaginal form was replaced by the foam rubber form. Between 1950 and 1989, the McIndoe operation was performed on 94 patients at the Johns Hopkins Hospital. During these 39 years, 83% of the 94 patients had a 100% take of the graft; in only 3 cases was there a significant area over which the graft failed. Urethrovaginal fistula has become even more infrequent since the introduction of the suprapubic catheter and the foam rubber form. The catheter is removed when the patient is voiding well and has no residual urine. In general, the patient is able to void without difficulty within the first few days of the procedure. Prophylactic broad-spectrum antibiotics started within 12 P.721 hours of surgery and continued for 7 days are of definite value in reducing the incidence of graft failures from infection in the operative site. Because of the excellent results obtained after a modified McIndoe vaginoplasty, this operation is recommended as the procedure of choice for women unable or unwilling to obtain a neovagina with dilatation methods. Women with a flat perineum with no dimple or pouch have no alternative other than the McIndoe vaginoplasty to obtain a neovagina for comfortable sexual relations. It is important that a McIndoe operation be performed correctly the first time. If the vagina becomes constricted because of granulation tissue formation, injury to adjacent structures, or failure to use the form properly, then subsequent attempts to create a satisfactory vagina are more difficult. The first operation has the best chance of success. Ozek, like many other surgeons, modified the McIndoe procedure by describing an X-type perineal incision and the use of a perforated vaginal mold during the postoperative period. He postulated that this incision minimized stricture at the vaginal introitus and provided greater ease of dissection of the vaginal cavity. He reinforced that the overall procedure is simple with a generally uneventful postoperative course. Complications included infection, failure of skin graft take, stress urinary incontinence, partial graft loss, and vaginal stricture. All were treated satisfactorily except the patient with stress urinary incontinence. Despite any minor modifications of the McIndoe vaginoplasty, the essential components of dissection of an adequate space, split-thickness skin grafting, and continuous dilatation during the contractile phase of healing remain unchanged. Recent reviews continue to support the safety and efficacy of the procedure. Hojsgaard and Villadsen reported 26 patients who underwent vaginoplasty, 18 of whom had Rokitansky syndrome. All patients were recorded as having a satisfactory result with complete graft take, adequate vaginal dimensions, and no strictures or fistulas giving symptoms. Complete take was achieved in 33% of patients within a week postoperatively, and after one further grafting procedure, an additional 38% had complete take. The intraoperative and early postoperative complications were perforation of the rectum in one patient (3.8%) and postoperative bleeding in three patients (11.5%). The late complications were vaginal stricture in three patients (11.5%), urethrovaginal fistula in two patients (7.7%), and rectovaginal fistula in one patient (3.8%). Alessandrescu and colleagues described the surgical management of 201 cases of Rokitansky syndrome. The surgeon substituted a modified transverse perineal incision and a perforated, rigid plastic mold. Intraoperative and postoperative complications consisted of two rectal perforations (1%), eight graft infections (4%), and 11 infections of graft site origin (5.5%). Sexual satisfaction was investigated with both objective and subjective criteria. Among the 201 cases, 83.6% had anatomic results evaluated as “good,” 10% as “satisfactory,” and 6.5% as “unsatisfactory.” More than 71% of patients rated their sexual life as “good” or “satisfactory” and reported that they had been able to

experience orgasms related to vaginal intercourse. Twenty-three percent reported the ability to have sexual intercourse but had no ability to achieve orgasm and only 5% percent expressed dissatisfaction with their sexual performance. Strickland and colleagues reported on the coital satisfaction, perception of vaginal competence, and impact on lifestyle of adult women undergoing vaginoplasty as adolescents. Ten of 22 women responded to a questionnaire at a median of 18 years (range = 5–13 years) following surgical intervention with a McIndoe vaginoplasty. All of the women had sexual experience and 80% were sexually active at the time of evaluation. The most frequent difficulty reported was vaginal dryness and lack of lubrication with sexual intercourse. Ninety percent of the subjects expressed satisfaction that sexual ability was acceptable. This experience also supports the role of the McIndoe vaginoplasty in providing young women with vaginal agenesis long-term coital ability and minimal disabilities.

Development of Malignancies At least 10 case reports exist of malignant disease developing in a vagina created by various techniques; these reports were reviewed by Gallup, Castle, and Stock. The authors reported a patient who was initially treated for intraepithelial malignancy by total vaginectomy combined with a split-thickness skin graft vaginoplasty to reconstruct a functional vagina. The authors noted a lesion in her vaginal apex 7 years later. These findings suggest that epithelium transplanted to the vagina can assume the oncogenic potential of the lower reproductive tract. It is therefore important that patients have long-term follow-up examinations after split-thickness skin graft vaginoplasty.

The Williams Vulvovaginoplasty Construction of a perineal bridge to help contain the vaginal mold was a routine part of the operation described by McIndoe, but it was not adopted subsequently by others. However, Williams described a similar vulvovaginoplasty procedure in 1964 and advised that it could be used to create a vaginal canal. In 1976, he reported that the procedure was unsuccessful in only 1 of 52 patients. Feroze and co-workers reported that the anatomic results were good in 22 of 26 patients. According to these authors, the advantages of the Williams operation are its technical simplicity, its absence of serious local complications even when performed as a repeat procedure, the ease of postoperative care, the absence of postoperative pain, the speed of recovery, the possible elimination of dilators and consequent applicability to patients who do not intend to have regular intercourse in the near future, and the higher success rates of primary and repeat procedures. The technique is not applicable to patients with poorly developed labia. It does result in an unusual angle of the vaginal canal, which is reported to straighten to a more normal direction with intercourse. If a very high perineum is created, urine can momentarily collect in the pouch after urination, giving the impression of postvoid incontinence. Failure of the P.722 suture line to heal by primary intention results in a large area of granulation tissue and most likely an unsatisfactory result. Williams believes that if the urethral meatus is patulous, a vulvovaginoplasty should not be performed because the urethra might be stretched further by coitus. He suggests that varying deficiencies in muscular and fascial tissue can explain why some patients with uterovaginal agenesis are able to develop a satisfactory vaginal canal with simple intermittent pressure with coitus, whereas others are prone to develop enteroceles. The technique of vulvovaginoplasty described by Williams is as follows (Fig. 29.9). A horseshoe-shaped incision is made in the vulva to extend across the perineum and up the medial side of the labia to the level of the external urethral meatus. The success of the operation depends on the appropriation of sufficient skin to line the new vagina. For this reason, the initial mucosal incisions are made as close to the hairline as possible and approximately 4 cm from the midline. After complete mobilization, the inner skin margins are sutured together with knots tied inside the vaginal lumen. A second layer of sutures approximates subcutaneous fat and perineal muscles for support. Finally, the external skin margins are approximated with interrupted sutures. If the procedure is performed properly, it should be possible to insert two fingers into the pouch to a depth of 3 cm. An indwelling bladder catheter is used. The patient is confined to bed for 1 week to avoid tension on the suture line. Examinations are avoided for 6 weeks, at which time the patient is instructed in the use of dilators.

FIGURE 29.9. The Williams vulvovaginoplasty. A: through C: No. 3-0 polyglycolic acid sutures can be used throughout to close both inner and outer skin margins and the tissue between. D: The entrance to the pouch should not cover the external urethral meatus.

View Figure

Capraro and Gallego have advised a modification of the Williams technique. They make the U-shaped incision in the skin of the labia majora at the level of the urethra or even lower, claiming that this modification results in a vulva more normal to sight and touch and still P.723 satisfactory for intercourse, and that it avoids trapping stagnant urine in the vaginal pouch. Other modifica-tions have been made by Feroze and associates and by Creatsas. The Williams vulvovaginoplasty is a useful operation and should certainly be considered the operation of choice for patients needing a follow-up to an unsatisfactory McIndoe operation or a supplement to a small vagina resulting from extensive surgery or radiation therapy. Rarely does a patient with a solitary kidney low in the pelvis not have room for dissection of an adequate vaginal space.

Alternative Techniques Several authors, including Adamyan and Soong and Templeman and colleagues, have described the laparoscopic use of the peritoneum to create a neovagina in patients with vaginal agenesis. Adamyan and Soong reported a group of 45 patients without significant postoperative complications. The most common postoperative problem involved the formation of granulation tissue at the vaginal vault. Templeman and others described the laparoscopic mobilization of peritoneum for the creation of a neovagina in only one patient. The peritoneum was grasped through a perineal dissection and sutured to the introitus. A pursestring closure was placed at the apex. Stenting of the neovagina was continued for 3 months postoperatively followed by rigid dilator use. At 9-month follow-up evaluation, an 8- by 2-cm vagina was described, with squamous epithelialization present. Both groups describe the technique as safe and efficient, producing a neovagina with apical granulation tissue as the only complication. The Vecchietti operation was first described in 1965 by Giuseppe Vecchietti. He subsequently reported his cumulative 14-year experience in 1979 and 1980. Veronikus and colleagues reviewed the use of the technique and described a laparoscopic modification that uses cystoscopy to confirm bladder integrity. The Vecchietti procedure is a surgical technique for the treatment of vaginal agenesis that constructs a dilatation-type neovagina in 7 to 9 days. The procedure uses specialized equipment including a traction device, a ligature carrier, and an acrylic shaped olive. The process is in two steps, with essential operative and postoperative components. The operative phase involves positioning the olive at the perineum and the traction sutures extraperitoneally. Classically performed through a Pfannenstiel incision; the ligature carrier introduces the suture into a newly dissected vesicorectal space. The olive is threaded with suture at the perineum and the suture is reintroduced at the abdomen. The suture is then guided lateral to the rectus muscles bilaterally in a subperitoneal fashion and advanced along the sidewall. The traction device, which provides constant traction on the olive, is positioned on the abdomen. During the postoperative invagination phase, the neovagina is created by applying constant traction to the olive. The process

reportedly occurs at a rate of 1.0 to 1.5 cm per day, developing a 10- to 12-cm vagina in 7 to 9 days. Patients are instructed on the use of a vaginal obturator to be used as an outpatient. Borruto reported on Vecchietti's personal series of vaginal agenesis patients, comprising 522 consecutive patients. The surgical complications included one bladder and one rectal puncture with the ligature carrier and three cases of vaginal vault bleeding. At 100% follow-up at 1 month, dyspareunia was initially reported to be 12%, but resolved in all cases by 3 months. There were no reported failures of the neovaginal construction with 1- and 2-year follow-up of 70% and 30%, respectively. Modifications of the Vecchietti approach include the use of laparoscopy and elimination of the dissection of the vesicorectal space. The first description of a laparoscopic modification was published by Gauwerky and colleagues. The vesicorectal space was dissected laparoscopically. The threads of the olive device were positioned using a probe introduced into the abdomen through the perineum. Six small abdominal incisions were used for laparoscopic instruments and the traction springs of the specialized device. In 1995, Laffarque and others described a laparoscopic intervention, creating a neovagina in three patients without dissection of the vesicorectal space. At completion of the procedure, cystoscopy was used to confirm bladder integrity. Some experts believe the theoretical risk of bladder or rectal perforation without the dissection of the vesicorectal space is unacceptably high. Fedele and others modified the approach to use a combined laparoscopic-ultrasonographic technique. The ultrasound assists in identifying the space of connective tissue between the bladder and rectum. The operating time for this modified procedure was only 40 minutes. After 10 days, the patient engaged in sexual intercourse. One-month evaluation confirmed a 12-cm vaginal length. Long-term follow-up outcomes are not available. Bowel vaginoplasty is a well-known alternative for creation of a neovagina. The Ruge procedure and others are characterized by the formation of a neovagina using sigmoid colon grafts. Advocates propose that scar formation and vaginal stenosis occur less often than with other procedures; however, the disadvantage is the necessity of an abdominal laparotomy. Ota and colleagues reported a laparoscopic-assisted Ruge procedure. Mesenteric dissection and sigmoid resection were performed laparoscopically. A 3.5-cm incision was used for appropriate bowel suturing. The segment of sigmoid colon was mobilized and brought to the introitus. The serosal layer of the pediculate end was stabilized to pelvic peritoneum. The patient remained hospitalized for 14 days. The benefit of this modification is certainly the accomplishment of a difficult surgical procedure endoscopically. Other advantages include the functional, ample vaginal length (12 cm) without postoperative dilatation. Disadvantages include the extended postoperative hospitalization period and the small number of patients evaluated. Makinoda and colleagues reported a nongrafting method of vaginal creation. This group reported 18 women who underwent a two-step protocol. The initial step used noninvasive dilatation using a vaginal mold based P.724 on the technique of Frank. The second step was a surgical procedure via a perineal approach. The apex of the dilated vaginal space was incised and further dissection between the bladder and rectum was carried to the peritoneal cavity. After peritoneal perforation, the uterine structures, when present, were pulled down and sutured to the newly created vaginal space. A firm vaginal mold was inserted and recommended for use for 6 months postoperatively. The authors propose a benefit of avoidance of grafting. The time course for success with the initial dilatation step may be unacceptably long in many patients (mean = 10.90 ± 9.8 months). No significant surgical complications were reported despite the theoretical risk of ureteral contortion and kinking when pulling the rudimentary uterine structures inferiorly. During the follow-up period, shrinkage of the vaginal length and diameter was noted in some patients who had been noncompliant with the mold or without coitus. The authors noted a minimal vaginal length of 5 cm in a patient who was noncompliant with the mold and nonsexually active. The authors dispute the necessity of any lining of the neovaginal space. In their experience, significant narrowing or contraction of the margins did not occur. They maintain that the vaginal space is maintained by suturing the muscular buds of the uterus to the pressure-created neovaginal space. A spatial W-plasty technique using a full-thickness unilateral groin graft has been described in a limited number of patients by Chen and others. The authors advocate an earlier intervention with the premise that a full-thickness graft may grow as the patient grows. This is recommended to eliminate psychological issues in patients who would be treated during the late teens, when sexual identity may be forming.

Acquired Vaginal Insufficiency Unusual types of infection and atrophy can rarely cause closure of part of the vagina, but acquired vaginal inadequacy most often is the result of treatment of various gynecologic malignancies with surgery or

radiation, or a combination of both. Restoration and maintenance of vaginal function are important elements of the treatment plan for such malignancies, especially when the patient is young and otherwise healthy. The techniques of vaginal reconstruction in gynecologic oncology have been reviewed by Magrina and Masterson, by Pratt, and by McCraw and associates.

DISORDERS OF VERTICAL FUSION Part of "Chapter 29 - Surgery for Anomalies of the Müllerian Ducts" The problems associated with vertical fusion include transverse vaginal septum with or without obstruction. Although imperforate hymen is a vertical fusion problem, the hymen is not a derivative of the müllerian ducts; therefore, this condition is discussed elsewhere (see Chapter 34).

Transverse Vaginal Septum No reliable epidemiologic data exist regarding the incidence of transverse vaginal septum. Reported incidences vary from 1 in 2,100 to 1 in 72,000. It is probably less common than congenital absence of the vagina and uterus. It has been diagnosed in newborns, infants, and older adolescent girls. Its etiology is unknown, although McKusick has suggested that some and perhaps most cases are the result of a female sex-limited autosomal recessive transmission. There is a developmental defect in vaginal embryogenesis that leads to an incomplete fusion between the müllerian duct component and the urogenital sinus component of the vagina. The incomplete vertical fusion results in a transverse vaginal septum (AFS class IIA) that varies in thickness and can be located at almost any level in the vagina (Fig. 29.10). Lodi has reported that 46% occur in the upper vagina, 40% in the midvagina, and 14% in the lower vagina. Rock, Zacur, and associates have noted septa in the upper, middle, and lower thirds of the vagina in 46%, 35%, and 19% of patients, respectively. In general, the thicker septum is noted to be more common closer to the uterine cervix. In contrast to congenital absence of the müllerian ducts, the transverse vaginal septum is associated with few urologic or other anomalies. Imperforate anus and bicornuate uterus can be found, as reported by Mandell and colleagues. The lower surface of the transverse septum is always covered by squamous epithelium. The upper surface can be covered by glandular epithelium, which is likely to be transformed into squamous epithelium by a metaplastic process after correction of the obstruction. FIGURE 29.10. Positions of septum responsible for complete vaginal obstruction. High (A), mid (B)), and low (C) transverse vaginal septa. Note the position of the hematocolpos. Lower vaginal septa allow more blood to accumulate in the upper vagina. The vaginal mass shown in (C) is more accessible through rectovaginal examination.

View Figure

In neonates and young infants, imperforate transverse vaginal septum with obstruction can lead to serious and life-threatening problems caused by the compression of surrounding organs by fluid that has collected above the septum. The fluid undoubtedly comes from endocervical glands and müllerian glandular epithelium in the upper vagina that have been stimulated by the placental transfer of maternal estrogen. Continued fluid collection in infants, even after the first year, has been reported; thus, the possibility of a fistula between the upper vagina and the urinary tract should be considered. The distended upper vagina creates a large pelvic and lower abdominal mass that can displace the bladder anteriorly, displace the ureters laterally with hydroureters and hydronephrosis, compress the rectum with associated obstipation and even intestinal obstruction, and limit diaphragmatic excursion to indirectly compress the vena cava and produce cardiorespiratory failure. Fatalities have been reported. The hydrocolpos develops along the axis of the upper vagina and therefore may not necessarily cause the outlet or perineum to bulge when there is compression of the mass from above. After careful preoperative radiologic and endoscopic investigations of the infant, the septum should be removed through a perineal approach. Bilateral Schuchardt incisions may be required to ensure that the septum has been removed. Because of the

subsequent tendency for vaginal stenosis and reaccumulation of the fluid in the upper vagina, follow-up studies to assess the recurrence of urinary obstruction are important. Vaginal reconstruction may be required in later years to allow satisfactory menstruation and coitus. P.725 A hematocolpos may not develop until puberty. Symptoms include cyclic lower abdominal pain, no visible menstrual discharge, and gradual development of a central lower abdominal and pelvic mass. Sometimes a small tract opens in the septum, some menstrual blood escapes periodically, and symptoms are variable. A septum large enough to allow pregnancy to occur can still cause dystocia during labor. Cyclic hematuria may be present if a communication between the bladder and upper vagina exists. The pelvic organs of a woman with a transverse vaginal septum are shown in Figure 29.11. The woman developed severe cyclic pain at the time of onset of menstruation, but there was no external bleeding until menstrual blood finally began to flow through the small sinus. Pelvic examination per rectum revealed a cervix and a normal-sized corpus. The ovaries were palpable but adherent, probably because of organized blood from hematosalpinx and hematoperitoneum. Remarkably, the woman had little dysmenorrhea after beginning to menstruate externally. Coitus was fairly satisfactory before surgical correction, but the shortness of the vagina was something of a handicap. The obstructing membrane was excised and an anastomosis of the upper and lower vagina was performed. FIGURE 29.11. Surgical correction of transverse vaginal septum. A: The upper end of a short vagina. The small sinus tract opening, through which the patient menstruated, is shown. The line of incision is drawn through the mucous membrane between the vaginal dimple and the sinus. B: Areolar tissue is dissected through to the pocket of mucosa that covered the cervix. The mucosa is incised. C: An anastomosis is made between the lower vagina and the upper vagina. D: Completed vagina. It is slightly shorter than normal but of normal caliber.

View Figure

The findings of 26 patients with complete transverse vaginal septum reported from the Johns Hopkins Hospital by Rock, Zacur, and colleagues have shown that associated congenital anomalies include urinary tract anomalies, coarctation of the aorta, atrial septal defect, and malformations of the lumbar spine. Vaginal patency and coital function were successfully established in all patients, and 7 of 19 patients attempting pregnancy eventually had children. The incidence of endometriosis and spontaneous abortion was high. A lower pregnancy rate and more extensive endometriosis were present when the transverse septum was located high in the vagina, suggesting that retrograde flow through the uterus and fallopian tubes occurs earlier in these patients. More extensive dissection between the bladder and rectum was required to identify the upper vagina when the septum was thick and high. Exploratory laparotomy was necessary in five patients to guide a probe through the uterine fundus and cervix and to assist in locating a high hematocolpos.

Surgical Technique for a Transverse Vaginal Septum A transverse incision is made through the vault of the short vagina (Fig. 29.11A). A probe is introduced through the septum after a portion of the barrier has been separated by sharp and blunt dissection. The physician usually finds some areolar tissue in dissecting the space between the vagina and the rectum. Palpation of a urethral catheter anteriorly and insertion of a double-gloved finger along the anterior wall of the rectum posteriorly provides the proper surgical guidelines so that

P.726 the bladder and rectum can be avoided during this blind procedure. After the dissection is continued for a short distance, the cervix can usually be palpated, and continuity can be established with the upper segment of the vagina (Fig. 29.11B and Fig. 29.11C). The lateral margins of the excised septum are extended widely by sharp knife dissection to avoid postoperative stricture formation. The edges of the upper and the lower vaginal mucosa are undermined and mobilized enough to permit anastomosis with the use of interrupted delayed-absorbable sutures (Fig. 29.11C). Figure 29.11D shows the completed anastomosis with a vagina that is of normal caliber but has a length slightly shorter than average. A soft foam rubber vaginal form covered with a sterile latex sheath can be placed in the vagina and removed in 10 days for evaluation of the healing process. The form can be worn for 4 to 6 weeks until complete healing has occurred. After this, coitus is permitted. If the patient is not sexually active, then vaginal dilatation may be necessary to maintain established patency. Alternatively, a silicone elastomer (Silastic) vaginal form can be inserted at night until the constrictive phase of healing is complete.

HIGH TRANSVERSE VAGINAL SEPTUM If the length of the obstructing transverse vaginal septum is such that reanastomosis of the upper and lower vagina is impossible, as is the case with a high transverse vaginal septum, in which a significant portion of vagina is atretic, then a space is created between the rectum and bladder to permit identification of the obstructed vagina (Fig. 29.12). The mass that has resulted from accumulated menstrual blood must be distinguished from the bladder anteriorly P.727 and the rectum posteriorly, a process that is facilitated by the mass itself. When differentiation is impossible, however, exploratory laparotomy can be performed. During this procedure, a probe is passed through the fundus of the uterus to tent out the vaginal septum and enable the surgeon to excise it from below and resect it safely. FIGURE 29.12. Correction of an atretic vagina. A: A large portion of atretic vagina is palpated with two fingers. Once the vaginal space is developed, it may be necessary to open the abdomen via laparotomy and pass a probe through to the uterine fundus (B) to tent out the septum, which may then be safely excised. C: An acrylic resin (Lucite) form is then placed into the vagina and secured with rubber straps.

View Figure

In most surgical procedures to remove the high transverse vaginal septum, the obstructing membrane can be readily identified (Fig. 29.13), after which the operator can probe the mass with an aspirating needle to identify old menstrual blood. The upper vagina is then opened and the septum excised. Because the distance between the septum and the upper vagina is too great to permit an anastomosis, an indwelling acrylic resin (Lucite) form, consisting of a bulbous end and a channel through which menstrual blood can drain, is placed into the vagina and anchored with a retaining harness. The bulbous end of the form, in most instances, is retained in the upper vagina and should be left in place for 4 to 6 months while epithelialization is accomplished. After its removal, vaginal dilatation should be practiced on a daily basis for 2 to 4 months to prevent contracture of the space. It is essential to the success of the operation that the new space not become constricted; to avoid constriction, the form must be worn for many months

during the constrictive phase of healing. As an alternative to the Lucite form, the physician can consider using a split-thickness graft to bridge the gap. The graft is usually sutured in situ in the vagina rather than sutured to a form. An ingenious but rather complicated Z-plasty method of bridging the gap has been described by Garcia and by Musset. A simpler flap method was described by Brenner and associates.

View Figure

FIGURE 29.13. A high transverse vaginal septum. A: The neovaginal space is dissected, revealing a high obstructing vaginal membrane. B: This can be palpated with the middle finger. C: A needle is then placed into the mass. D: The incision is made with a sharp knife, and considerable bleeding can occur. E: The septum is excised. F: The septum is removed. G: After the septum is removed, the wall of the septum is oversewn with interrupted sutures of 2-0 chromic catgut. H: Because the distance between the septum and the upper vagina is too great to allow anastomosis, an acrylic resin (Lucite) form is placed in the vagina so that epithelialization can occur over the form while vaginal patency is maintained. The form, in place, is fitted with a plastic retainer. Rubber straps can be placed through the retainer and attached to a waist belt to allow constant upper pressure so that the form is retained in the upper vagina. Modification of this method includes a small adapter to allow drainage through the acrylic resin (Lucite) form, preventing the accumulation of old blood and mucus in the upper vagina. (From Rock J. Anomalous development of the vagina. Semin Reprod Endocrinol 1986;4:24, with permission.)

A transverse vaginal septum diagnosed after the onset of puberty presents numerous problems. Often, a large segment of the vagina is absent, making anastomosis of the upper and lower segments difficult. Furthermore, postoperative vaginal dilatation is necessary to prevent stenosis at the anastomosis site. Poor compliance with dilatation in a poorly motivated pubertal patient is always a concern. However, rarely is the surgeon able to delay vaginoplasty until the patient is more mature because of increasingly severe cyclic abdominal pain caused by the hematocolpos. Thus, a difficult vaginoplasty can have less than optimal results. Hurst and Rock have described an alternative approach to maximize surgical resection and anastomosis in women with a high transverse vaginal septum. Aspiration of the hematocolpos under ultrasound guidance was necessary to relieve the acute pain and delay surgery. Continuous oral contraceptives were used to delay recurrence of hematocolpos. Most important, vaginal dilatation was used to lengthen the lower vaginal segment to facilitate resection and reanastomosis (Fig. 29.14). In all three patients, the approach was successful.

FIGURE 29.14. A: High transverse vaginal septum demonstrating a small hematocolpos and hematometra. Upper to lower vaginal anastomosis at this stage can result in stenosis at the anastomosed site. B: Vaginal depth is increased with passive dilatation using progressively larger dilators. C: A primary upper to lower vaginal anastomosis can be performed easily after dilatation.

View Figure

P.728 P.729

Congenital Absence or Dysgenesis of the Cervix Agenesis or atresia of the cervix (AFS class IIB) is a relatively infrequent müllerian anomaly. When this anomaly does occur, it is often in association with absence of a portion or all of the vagina. In many cases of cervical agenesis or atresia, retention of menstrual blood initiates symptoms of cyclic lower abdominal pain without menstrual flow, causing the patient to seek gynecologic evaluation and care. In past times, diagnosis was suspected on the basis of a history and physical findings but was not proved until the time of surgery. Today, diagnosis of cervical agenesis or atresia is still usually difficult before operation, but the possibility of making a correct diagnosis before surgery does exist, with the help of modern diagnostic tools. Early diagnosis offers significant advantages in patient care, the most important of which is effective presurgical planning and preparation.

Diagnosis of Cervical Dysgenesis Patients with congenital absence of the cervix present a diagnostic challenge. Patients with cervical aplasia with a functioning midline uterine corpus have aplasia of the lower two thirds of the vagina with an upper vaginal pouch. Similarly, some patients have a considerable atretic segment of vagina and an upper vaginal pouch with a properly developed uterine cervix and corpus above. Differentiation of these two müllerian anomalies is essential. Ultrasonography may be helpful. Valdes and associates have reported the use of preoperative ultrasonography in the evaluation of two patients with atresia of the vagina and cervix. Magnetic resonance imaging P.730 (MRI) has been found to be helpful in confirming this diagnosis, as reported by Markham and associates. The lower uterine segment and cervical tissue can be carefully examined (Fig. 29.15). With cervical dysgenesis there is no vaginal dilatation with the accumulation of blood, as seen with a high transverse vaginal septum. Both ultrasonography and MRI are most helpful when they are correlated with the findings of a careful pelvic examination under anesthesia.

FIGURE 29.15. Magnetic resonance T1-weighted image showing atretic segment of distal cervix. The tip of an atretic cervix is shown. No vagina is noted.

View Figure

Anatomic Variations of Congenital Cervical Anomalies Two basic categories of cervical anomalies have been observed in several configurations. Patients exhibiting the first type, cervical aplasia, lack a uterine cervix (Fig. 29.16A), and the lower uterine segment narrows to terminate in a peritoneal sleeve at a point well above the normal communication with the vaginal apex. The second type, cervical dysgenesis, can be described as four subtypes: Intact cervical body with obstruction of the cervical os (the cervix is usually well formed, but a portion of the endocervical lumen is obliterated) (Fig. 29.16C) FIGURE 29.16. Congenital cervical anomalies. A: The fundus of the uterus is noted without a cervix. B: The cervical body consists of a fibrous band of variable length and diameter that can contain endocervical glands. C: The cervical body is intact with obstruction of the cervical os. Variable portions of the cervical lumen are obliterated. D: Stricture of the midportion of the cervix, which is hypoplastic with a bulbous tip. No cervical lumen is identified. E: Cervical fragmentation in which portions of the cervix are noted with no connection to the uterine body. Hypoplasia of the uterine cavity can be associated with cervical cord fragmentation.

View Figure

Cervical body consisting of a fibrous band of variable length and diameter (endocervical glands may be noted on pathologic examination) (Fig. 29.16B) Stricture of the midportion of the cervix (which is hypoplastic with a bulbous tip and no identifiable cervical lumen) (Fig. 29.16D) Fragmentation of the cervix (with portions that can be palpated below the fundus and that are not connected to the lower uterine segment) (Fig. 29.16E)

Associated anomalies of the urinary tract are rare, but they do occur. Variable portions of the vagina can be atretic. Cervical obstruction is most often associated with a vagina of normal length.

Treatment When both the vagina and cervix are absent and a functioning uterine corpus is present, it is difficult to obtain a satisfactory fistulous tract through which menstruation can occur. Many methods have been tried, most of them involving creation of a passage through the dense fibrous tissue between the uterine cavity and the vagina and placement of a stent to keep the tract open. Occasional successes in maintaining an open passageway and normal cyclic menstruation have been reported, but endocervical glands do not develop, and there is no way to compensate for the absence of the cervical mucus, which plays an important role in sperm transport. Even though cyclic ovulatory periods can be achieved in a few patients, pregnancy is unlikely. Eventually the uterovaginal tract closes from constriction by fibrous tissue. Endometriosis can develop along the tract. Endometriosis also can develop in ovaries and other pelvic sites because of retrograde menstruation. Recurrent and severe pelvic infection is a common problem and may require total hysterectomy and removal of both ovaries. As in vitro fertilization procedures began to offer the possibility for a host uterus to carry a pregnancy to term, procedures to establish a fistulous tract were abandoned. Nevertheless, Cukier and associates in 1986 reported treating a patient with congenital absence of the cervix by construction of a splint that extended into the neocervical canal such that a split-thickness skin graft could actually be placed within the endocervical canal. This patient has continued to menstruate without difficulty, although pregnancy has not been accomplished. Many authors have recommended hysterectomy as an initial procedure for a patient with a functioning uterine corpus and congenital absence of the cervix and vagina. A hysterectomy eliminates much needless suffering from associated problems such as cryptomenorrhea, sepsis, endometriosis, and multiple operations. If the hysterectomy is performed soon enough, before the problems become great, it may be possible to conserve the ovaries and their useful functions. The reconstructive surgeon should be prepared to perform a vaginoplasty with use of a split-thickness graft if hysterectomy is performed, particularly if there has been a vaginal dissection. If the neovaginal space is allowed to close and scar, then future operations to develop an adequate neovagina are associated with increased risks of graft failure and fistula formation. Despite the overall poor results from reconstruction for congenital absence of both the cervix and the vagina, clinical experience suggests that cannulization procedures can be worthwhile for a few carefully selected cases with adequate stroma to allow a cervicovaginal anastomosis. If a long segment of cervix is fibrous cord, a cervical grafting technique may be required. If a fragmented cervix is noted, then hysterectomy is usually warranted. Those few patients who have achieved a pregnancy after P.731 P.732 cervical reconstruction have had a well-formed cervical body. Anecdotal case reports occasionally appear in the literature confirming the necessity of palpable cervical tissue. Letterie described the development of a cervico-vaginal tract in an adolescent patient with a core of cervical tissue. The tract has remained patent for menstrual flow for 2 years; however, pregnancy has not yet been attempted. Data published by the senior author regarding the long-term follow-up of 21 patients with abnormal cervical development support the success of cannulization in only selected patients with sufficient rudimentary cervical tissue. All of the patients with fragmentation of the cervix (n = 4) eventually underwent hysterectomy. Only those patients with a well-formed cervical body, with at least a palpable cord or only distal obstruction, achieved successful surgical outcomes (4/7 patients). Only one patient, with distal obstruction, treated with cannulization using a full-thickness skin graft, achieved pregnancy.

DISORDERS OF LATERAL FUSION Part of "Chapter 29 - Surgery for Anomalies of the Müllerian Ducts" Failures of lateral fusion of the two müllerian ducts cause vaginal anomalies that are grouped as obstructed or unobstructed.

The Unobstructed Double Uterus (Bicornuate, Septate, or Didelphic Uterus) Complete failure of medial fusion of the two müllerian ducts can result in complete duplication of the vagina, cervix, and uterus. Partial failure of fusion can result in a single vagina with a single or duplicate cervix and complete or partial duplication of the uterine corpus. A failure of absorption of the uterine septum between the two fused müllerian ducts causes the septum to persist inside the uterus to a variable extent while the external appearance remains that of a single uterus. The septum can be so complete that it divides both the uterine cavity and endocervical canal into two equal or unequal components. More often, incomplete disappearance of the septum leaves only the upper uterine cavities divided. Each of these and a variety of other forms of double uteri have their own individual features of clinical significance. When no obstruction is present, surgical reconstruction is performed primarily because of difficulties with reproduction. Some aspects of lateral fusion disorders remain controversial because information is still inaccurate or incomplete. Many reports are based on small samples of selected patients, patients who have been diagnosed as having one anomaly or another based on incomplete data, and patients who have received unification operations without preliminary studies to rule out other causes of reproductive difficulty. A comparison of results from one series to the next is difficult because authors have used different classifications based on a variety of embryologic, anatomic, physiologic, functional, and radiologic considerations. Unknown numbers of uterine anomalies may have escaped detection because reproductive performance is generally acceptable and gynecologic difficulties do not necessarily occur. The müllerian ducts undergo multiple steps in development including caudal, medial growth followed by fusion and later resorption of the remaining septum. Apoptosis has been proposed as a mechanism by which the septum regresses. Bcl-2, a protein involved in regulating apoptosis, was found to be absent from the septa of several uteri. The absence of this critical protein may play a pivotal role in the persistence of the septum and lateral fusion disorders.

Historical Development of Surgical Procedures Ruge, in 1882, first reported excision of a uterine septum in a woman who had suffered two pregnancy losses. The woman subsequently carried a pregnancy to term. Paul Strassmann of Berlin and later Erwin Strassmann, his son, were strong advocates of uterine unification operations. The studies of Jones and Jones have contributed greatly to modern understanding of the management of uterine anomalies. Their studies began with a report in 1953 of a series that was started in 1936. Updates have been published from time to time. Wheeless, Rock, Andrews, and others have joined in these reports.

Diagnosis of Uterine Anomalies If a uterine anomaly is associated with obstruction of menstrual flow, then it causes symptoms that will come to the attention of the gynecologist shortly after menarche. Unobstructed uterine anomalies are diagnosed later in a variety of circumstances. Young girls may notice difficulty in using tampons or later difficulty in coitus if a longitudinal vaginal septum is present. This can lead to the diagnosis of an associated uterine anomaly. A patient with an anomalous upper urinary tract on intravenous pyelogram may be found to have a uterine anomaly on gynecologic evaluation. A uterine anomaly is occasionally found when a patient complains of dysmenorrhea or menorrhagia or when a dilatation and curettage (D&C) is performed for abortion or some other indication. A palpable mass may be a uterine anomaly but should be confirmed as such by ultrasonography, hysterography, or laparoscopy. Woelfer and colleagues recently described the use of three-dimensional ultrasonography in screening for congenital uterine anomalies. During an investigation of the correlation of uterine anomalies with obstetric complications, the authors assessed the potential value of three-dimensional ultrasound for screening. More than 100 women with uterine anomalies were identified. Seventy-two arcuate uteri, 29 septate, and five bicornuate uteri were described. The authors emphasized how the three-dimensional ultrasound may overcome the limitations of conventional two-dimensional ultrasonography in providing a coronal view of the uterus, thus differentiating between arcuate, bicornuate, and subseptate uteri. This technique remains investigational.

Semmens has pointed out that the diagnosis of a uterine anomaly can also be made from astute observation P.733 of an abnormal uterine contour during pregnancy, either in the antepartum period or at the time of abdominal or vaginal delivery. The abnormal contour is caused by a combination of fetal malpresentation and an anomalous uterus. An anomalous uterus can also be diagnosed when a pregnancy occurs despite the presence of an intrauterine contraceptive device. Persistent postmenopausal bleeding despite recent D&C can lead to a diagnosis of an anomalous uterus. Sometimes the diagnosis is made as an incidental finding at laparotomy. However, most uterine anomalies are diagnosed after hysterosalpingography to evaluate infertility or reproductive loss, usually from repeated spontaneous abortion.

Uterine Anomalies and Reproductive History Although some uterine anomalies can cause infertility, most patients with uterine anomalies are able to conceive without difficulty. There is no question that uterine anomalies can be associated with perfectly normal reproductive performance. Overall, however, the incidences of spontaneous abortion, premature birth, fetal loss, mal-presentation, and Caesarean section are clearly increased when a uterine anomaly is present. It is impossible to predict which patients with uterine anomalies will have these problems.

Etiology of Reproductive Failure The etiology of reproductive failure in patients with uterine anomalies remains unclear. Mahgoub believes that the presence of a uterine septum can lead to abortion because of diminished intrauterine space for fetal growth or because of implantation of the placenta on a poorly vascularized septum. Mizuno and associates have attached importance to the inadequacy of vascularization of the uterine septum. Associated cervical incompetence, luteal phase insufficiency, and distortion of the uterine milieu have all been implicated in the etiology of increased reproductive loss. However, it is as yet unexplained why some patients with a uterine anomaly have normal reproductive function, whereas others abort early in pregnancy. Interestingly, it has been reported that the chance for a liveborn child increases with each pregnancy loss. It is unknown whether this apparent “conditioning” of the uterus is due to better vascularization, better myometrial stretching and accommodation, or some other factor. A medical history of three or more episodes of spontaneous abortion or premature labor merits hysterosalpingography to determine whether structural abnormalities of the uterus are present. An abnormality is found in about 10% of such cases. Among chronic early second-trimester aborters, the incidence may be higher. The etiology of spontaneous abortion is complex, and a complete workup should be done even when an anomalous uterus has been found. A careful history should include a detailed discussion of each previous pregnancy loss and inquiry into DES exposure or other drug or chemical toxicity, specific medical illnesses, and exposure to contagious diseases. A family history should emphasize reproductive failures among family members of both the patient and the husband. Specific medical diseases such as thyroid disease, diabetes mellitus, renal disease, and systemic lupus erythematosus should be ruled out. The possibility of infection by such agents as Neisseria gonorrhoeae, Chlamydia, Mycoplasma, Toxoplasma, and Listeria should be considered. Chromosome analyses should be done. Abnormalities in aborted tissue are found in more than 50% of spontaneous abortions, and abnormalities appear in up to one fourth of couples with a history of habitual abortion. Identifying such couples makes it possible to offer genetic counseling for subsequent pregnancies. Uterine leiomyomas, especially lower uterine segment and submucous leiomyomas, can cause spontaneous abortion. Basal body temperature charts, serum progesterone determinations, and endometrial biopsies timed in the luteal phase help determine the presence of luteal phase deficiency. The cervix should be studied for incompetence. Couples with multiple etiologies for reproductive loss should have all other problems corrected before metroplasty is considered. Indeed, correcting other factors first may correct the problem of reproductive loss without metroplasty. In 1977, Rock and Jones reported on seven patients who had anomalous uterine development and extrauterine factors in the etiology of their reproductive loss. These patients had already had 16 pregnancies, 5 (29%) of which resulted in a liveborn child. After therapy to correct the extrauterine factor, the success rate increased to 71%. Stoot and Mastboom reported an impressive increase in reproductive performance among uterine anomaly patients by simple improvement of abnormal carbohydrate metabolism.

Hysterographic Studies

Proper technique during the performance of hysterosalpingography to diagnose uterine anomalies is important. The hysterogram must be taken at right angles to the axis of the uterus for a true assessment of the deformity to be made. The study is best done under fluoroscopy. A septate uterus cannot be distinguished from a bicornuate uterus by hysterogram alone (Fig. 29.17). The external uterine configuration also cannot usually be determined by pelvic examination alone, but some idea of the configuration can be obtained by ultrasonography. McDonough and Tho have suggested the use of double-contour pelvic pneumoperitoneum-hysterographic studies for precise identification of müllerian malformations. Of course, laparoscopy is even more certain. If the uterine corpus has not been previously visualized, the physician must be prepared to correct either anomaly (i.e., obstructed or unobstructed), depending on the findings at laparotomy. FIGURE 29.17. A: A hysterogram of a double uterus. A bicornuate uterus (B) and a septate uterus (C) are types of double uteri. Visualization of the fundus is required to determine the type of uterus.

View Figure

Additional Testing A complete investigation should also include an assessment of tubal patency and an intravenous pyelogram. A variety of upper urinary tract anomalies are seen, including absence of one kidney, horseshoe kidney, pelvic kidney, duplication of the collecting system, and ectopically P.734 located ureteral orifices. The lower urinary tract (bladder and urethra) is much less often anomalous.

The Double Uterus and Obstetric Outcome The percentage of full-term pregnancies with various types of double uteri in an unselected series of women who have not been operated on is unknown. For all types combined, it is probably approximately 25%. In patients selected for operation, it probably increases from approximately 5% to 10% to approximately 80% to 90%. Because patients with uterine anomalies who have relatively normal obstetric histories cannot be identified, there is confusion in the literature about which anomalies are more often associated with obstetric difficulties and which are relatively benign in their effect. Special diagnostic procedures to detect uterine anomalies are not usually performed before reproductive performance is tested. A didelphic uterus is the exception. This anomaly can be diagnosed easily on routine pelvic examination by identification of two complete cervices and perhaps also a longitudinal vaginal septum. A study by Heinonen in Finland of 182 women with uterine anomalies indicated that pregnancies in the septate uterus had a better fetal survival rate (86%) than they did in the complete bicornuate uterus (50%) or in the unicornuate uterus (40%). These findings differ from prevailing opinions that the septate uterus is associated with the highest reproductive loss, as proposed by Jones and Jones. A recent report by Woefler and others supports Jones and Jones' opinions by noting that women with a septate uterus had a significantly higher proportion of first trimester loss than women with a normal uterus. In 1968, Capraro and colleagues reported on 85 patients with uterine anomalies seen between 1962 and 1966. One uterine anomaly was seen for every 645 admissions (0.145%). Metroplasty was considered

necessary in only 14 (16%) of these 85 cases. According to Jones and Jones, only one third of patients with a double uterus have important reproductive problems. In most instances, the presence of a double uterus is not in itself an indication for metroplasty. In 1980, Jewelewicz and co-workers estimated the spontaneous abortion rate to be 33.8% in women with a bicornuate uterus, 22.2% in those with a septate uterus, and 34.6% in those with a unicornuate uterus. More recently, Ludmir and associates reported that high-risk obstetric intervention did not significantly increase the fetal survival rate for uncorrected uterine anomalies. Capraro and associates found a preoperative fetal salvage rate of 33.3% for the septate uterus, 10% for the bicornuate uterus, and 0% for the didelphic uterus. Postoperatively, the fetal salvage rate was 100% for the bicornuate uterus, 80% for the septate uterus, and 66% for the didelphic uterus. The report gives the improved salvage figures, compared with several previous studies, after abdominal metroplasty. Ravasia and others described the incidence of uterine rupture in a cohort of women with müllerian duct anomalies who attempted vaginal birth after prior Caesarean delivery (VBAC). Of the 1,813 patients who attempted P.735 VBAC between 1992 and 1997, only 25 patients with known müllerian duct anomalies attempted a trial of labor. This included 14 patients with a bicornuate uterus, five with a septate uterus, four with a unicornuate uterus, and two with uterine didelphys. Uterine rupture was diagnosed in two patients with müllerian anomalies. The authors proposed several mechanisms for the greater incidence of uterine rupture in this population: abnormal development of the lower uterine segment, previous scar similar to a vertical or classic incision, and the possibility of abnormal traction on the uterine scar during labor.

THE DIDELPHIC UTERUS A didelphic uterus with two hemicorpora is easily diagnosed because all patients have two hemicervices visible on speculum examination, and most, if not all, have a longitudinal sagittal vaginal septum. In the series reported by Heinonen and associates, all 21 patients with a didelphic uterus had a vaginal septum. Conversely, a patient with a longitudinal vaginal septum usually has a didelphic uterus. The indication for uterine unification is related to the role of this anomaly as an etiologic factor in reproductive loss. Of all the uterine anomalies (except arcuate uterus), the didelphic uterus is associated with the best possibility of a successful pregnancy. However, there is still some increase in perinatal mortality, premature birth, breech presentation, and Caesarean section for delivery. Heinonen and associates reported a fetal survival rate of 64% without metroplasty. Musich and Behrman stated that the didelphic uterus offers the best chance for a successful pregnancy (57%) and should not be considered an appropriate indication for metroplasty. However, W. S. Jones considered the didelphic uterus to give the worst obstetric outcome. In the opinion of the editors of this book, a unification operation for a didelphic uterus is not often indicated, and the results may be disappointing, especially when an attempt is made to unify the cervix. Not only is this procedure technically difficult in a complete didelphic anomaly, but it can also result in cervical incompetence or cervical stenosis.

THE SEPTATE UTERUS Most patients who are evaluated for repeated abortion and who are found to have a uterine anomaly have a septate uterus. A few have other anomalies, mostly the bicornuate uterus. In our experience, fetal survival rates are higher after septate uterus repair than after other repairs. In 1977, Rock and Jones reported on 43 patients with septate uteri selected for Jones metroplasty at the Johns Hopkins Hospital. Of these 43 patients, 95% became pregnant postoperatively, 73% carried to term, and 77% delivered a liveborn child. Similarly, hysteroscopic metroplasty for the septate uterus provides a substantial improvement in obstetric outcome. Data obtained from retrospective studies suggest that hysteroscopic metroplasty is associated with favorable outcomes, with a pregnancy rate of approximately 80% and a miscarriage rate of only approximately 15%. Recently, the histologic features of the septum in this abnormal uterus have been described. Dabirashrafi and colleagues noted less connective tissue in uterine septa. Poor decidualization and placentation were suggested as a cause. Finally, the AFS class VA uterus (a double cervix and uterine cavity with a single fundus) can result from a rotation abnormality during the descent of the müllerian ducts. Among the reported cases of the septate uterus, the incidence of the complete septum involving the cervix varies from 4% to 29%. If the dextrorotating müllerian ducts overrotate, the senior author theorizes (J. A. Rock, personal observations, 1991) that the septum fails to absorb after fusion of the ducts. In virtually every patient with a complete septate uterus, the left cervix is higher than the right. In one patient, one cervix has been noted above the

other (Fig. 29.18). This rotation abnormality may be a factor associated with lack of absorption of the uterine septum in these patients. FIGURE 29.18. A double uterus with two cervices and a single fundus (class V). Note that the left cervix is positioned over the right cervix. This rotation abnormality may be a factor associated with a lack of absorption of the uterine septum.

View Figure

Uterine Anomalies and Menstrual Difficulties Dysmenorrhea and abnormal and heavy menstrual bleeding have been reported to occur more frequently with any form of double uterus and to be relieved after unification operations. Capraro and associates reported several cases in which dysmenorrhea was cured by metroplasty. Erwin Strassmann also believed that all cases of dysmenorrhea and menorrhagia associated with uterine anomalies were relieved by unification of the two uterine cavities. Generally, however, dysmenorrhea and menorrhagia are inappropriate indications for uterine unification, and the operation should not be performed solely for these reasons.

Uterine Anomalies and Infertility Opinions differ considerably in terms of whether infertility is a proper indication for metroplasty. Erwin Strassmann stated that primary infertility could be cured in 60% of patients with uterine anomaly if all other causes of P.736 infertility were excluded. Strassmann reported eight metroplasties for primary sterility that yielded nine pregnancies and seven liveborn children, although the number of patients who conceived was not given. Similar reports of small numbers of patients can be found throughout the literature. Heinonen and Pystynen indicated that uterine anomalies are rarely the reason for infertility. Nonuterine causes of infertility must be ruled out before metroplasty, as a last resort, is considered. Certainly, a full infertility investigation to rule out other causes should be completed before the anomalous uterus is blamed. Even when no other cause for infertility is found, if the uterus is septate or bicornuate, then there may not be any proper indication for metroplasty. This question of when to perform metroplasty simply has not yet been answered. The decision is difficult, and becomes even more difficult, when the opportunity for metroplasty presents itself because a septate or bicornuate uterus requires laparotomy for some other reason, such as endometriosis or tubal occlusion.

Surgical Technique for Uterine Unification Traditionally, the septate uterus has been unified with either the Jones or the Tompkins procedure. Clinical reports by Chervenak and Neuwirth, Daly, Walters, and co-workers, DeCherney and associates, and Israel and March have favorably compared hysteroscopic or resectoscopic incision of a uterine septum with the more traditional transabdominal approach. Term pregnancy rates after these procedures have approached 80% to 85%. Several attempts may be necessary to incise a wide septum, although the septum usually can be incised completely at the first operation.

TRANSCERVICAL LYSIS OF THE UTERINE SEPTUM

Abdominal metroplasty for transfundal incision or for excision of the septum associated with the septate uterus generally has been abandoned. With hysteroscopic scissors, the procedure can be tedious, especially with a large, broad septum. Although the hysteroscope and scissors are still used for cutting the septum, the resectoscope has been found to be comparable. The optics are excellent, and the septum can be electrosurgically incised with little difficulty. Laser-assisted procedures have also been described. Before transcervical lysis of a uterine septum, a gonadotropin-releasing hormone agonist may be given for 2 months to reduce the amount of endometrium that can obscure the surgeon's view during the procedure. Many authors do not consider routine preoperative preparation of the endometrium essential and may only use medications in procedures involving exceptionally wide septa or complete septa that involve the lower one third of the uterine cavity or the cervical canal. If medical preparation is not used, surgical intervention should be scheduled during the early proliferative phase of the cycle to avoid bleeding and impaired visualization from a vascular endometrium associated with the secretory phase. Transcervical lysis is usually performed in conjunction with laparoscopy under general endotracheal anesthesia. The uterine cavity is distended with dextran 70 (Hyskon) by way of the resectoscope, which is inserted into the cervix. The septum is then electrosurgically incised by advancing the cutting loupe, using the trigger mechanism of the resectoscope. The uterine septum is incised until the tubal ostia are visualized and there is no appreciable evidence of the septum. The procedure is performed under simultaneous laparoscopy to limit the risks of uterine perforation. The laparoscopic light can be turned off so that the light from the hysteroscope can be clearly visualized through the fundus. Most patients can be discharged within 4 hours of the procedure. There is no role for placement of a postoperative intrauterine device. The benefit of routine procedure-related antibiotic therapy has not been well supported with evidence; however, it is recommended to administer antibiotics before the procedure and to continue for 5 days after surgery to limit the risks of infection. If excessive bleeding occurs after the procedure, a Foley catheter should be placed in the uterine cavity for tamponade and removed in 4 to 6 hours. Hormonal therapy is the most commonly used postoperative treatment regimen. The aim of the treatment is the promotion of rapid epithelialization. Dabirashrafi and colleagues reported that estrogen therapy did not appear to demonstrate a benefit. Further evidence is necessary before dismissing the current trend of postoperative estrogen therapy. Transcervical lysis also can be performed to repair a complete septate uterus (i.e., a single fundus with two cavities and two cervices). In this instance, a no. 8 Foley catheter is inserted into one cervix and indigo carmine is injected into the cavity. The other cavity is distended with dextran 70 (Hyskon) by way of the resectoscope. The septum is electrosurgically incised at a point above the internal cervical os until the Foley catheter is visualized. The septum is then incised in a superior direction until the tubal ostium is visualized and there is no appreciable septum (Fig. 29.19). FIGURE 29.19. Resectoscopic metroplasty. A: A Foley catheter is placed in one cavity of a complete septate uterus (American Fertility Society class VA uterus). The resectoscope is inserted in the opposite cavity, and the septum is incised until the Foley is visualized. The septum can be easily incised with the resectoscope until both internal os are visible. B: A septate uterus with a single cervix. The septum can be incised with the straight loupe of the resectoscope.

View Figure

After transcervical lysis of a uterine septum, a 2-month delay before attempting pregnancy is suggested to allow complete resorption of the septum. Delivery may be vaginal. The Jones procedure is used to repair a septate uterus when a particularly broad septum cannot be easily incised with the resectoscope. The Strassmann procedure is used for unification of a bicornuate uterus. The safety and efficacy of

hysteroscopic resection of the uterine septum in patients with a class Va septate uterus has been demonstrated by the senior author. Historically, case reports such as that of Hundley and colleagues, was the only source of information about this interesting variant; however, one of the largest populations of patients with a complete septum was reported in 1999 by Roberts and Rock. The patients underwent hysteroscopic metroplasty with preservation of the cervical portion of the septum. With the exception of one case of pulmonary edema, no significant intraoperative or postoperative complications were reported. Postoperative hysteroscopy revealed only minor fundal septal remnants without clinical significance. P.737

THE MODIFIED JONES METROPLASTY In the modified Jones unification operation (Fig. 29.20), the abdomen is generally opened through a transverse incision. If only the unification operation is planned, then a Pfannenstiel incision is permissible. The pelvic viscera are inspected. The septate uterus may demonstrate a median raphe across the fundus, but it is surprising how often the corpus looks normal. To facilitate manipulation, a traction suture of heavy silk is placed through the top of the septum. This suture is removed from the site when the septum is excised. FIGURE 29.20. The modified Jones metroplasty. See the text for a full description of the various steps in the operative repair of a septate uterus by excision of a wedge. The modified Jones metroplasty. See the text for a full description of the various steps in the operative repair of a septate uterus by excision of a wedge.

View Figure

No attempt is made to stain the uterine cavity with methylene blue. Normal unstained endometrial tissue can be easily differentiated from the myometrium. There are essentially two methods to control bleeding during this procedure. In the first, a tourniquet is applied at the junction of the lower uterine segment and cervix by inserting a 0.5-inch Penrose drain through an avascular space in the broad ligaments just lateral to the uterine vessels on each side. The tourniquet is placed around the lower uterine segment and is tied anterior to the uterus. Because the uterine corpus receives a significant blood supply through the ovarian arteries, tourniquets should also be tied around the infundibulopelvic ligaments on each side, using the same hole in the broad ligament. All tourniquets must be tied tightly enough to occlude both the arterial supply to and the venous drainage from the uterus. If only the venous drainage is occluded, then the corpus becomes engorged and congested and bleeding is increased. If the arterial supply is occluded, then the uterus blanches and the bleeding is minimal. A sterile Doptone can be used to establish disappearance of uterine artery pulsations. Hypotensive anesthetic techniques used in conjunction with the tourniquets allows a uterine unification operation to be accomplished with negligible blood loss. The alternative method for hemostasis uses up to 20 units of vasopressin that is diluted in 20 mL of saline and injected into the anterior and posterior walls of the uterus before the incision is made. The uterine septum should be surgically excised as a wedge (Fig. 29.20D). The incisions begin at the fundus of the uterus. The approach to the endometrial cavity should be handled carefully so that it is not transected (Fig. 29.20E). The original incisions at the top of the fundus are usually within 1 cm, and

sometimes even less, of the insertion of the fallopian tubes. If the incision is directed toward the apex of the wedge, however, there seems to be little danger of transecting the tube across its interstitial transit in the myometrium. After the wedge has been removed, the uterus is closed in three layers with interrupted stitches; 2-0 nonreactive suture on an atraumatic tapered needle is convenient. Two sizes of needles are needed: a half-inch needle for the inner and intermediate layers and a large needle (three fourths half-round) for the outer muscular layer. The inner layer of stitches must include about one third of the thickness of the myometrium, because the endometrium alone is too delicate to hold a suture and will be cut through. The inner sutures should be placed through the endometrium and the myometrium in such a way that the knot is tied within the endometrial cavity (Fig. 29.20G and Fig. 29.20H). While the suture is being tied, the two lateral halves of the uterus should be pressed together P.738 both manually and with the guy sutures to relieve tension on the suture line and to reduce the possibility of cutting through. These sutures are placed alternately, first anterior and then posterior. After the first few stitches are placed and before the first layer is completed, the second layer can be started to reduce tension. As the operation proceeds, the third layer of stitches is begun in the serosa both anteriorly and posteriorly (Fig. 29.20I, Fig. 29.20J and Fig. 29.20K). Finer, nonreactive suture material can be used to approximate the serosal edges of the uterus more precisely to prevent adhesion formation to the suture line (Fig. 29.20K and Fig. 29.20L). By the conclusion of the operation, the uterus appears near normal in configuration. The striking feature is usually the proximity of the insertions of the fallopian tubes. Special care must be exercised not to obstruct the interstitial portions of the fallopian tubes while placing the fundal myometrial and serosal sutures. The final size of the uterine cavity seems to be relatively unimportant to reproductive capability; uterine symmetry appears to be a more important factor. Often the constructed cavity is quite small compared with the normal uterus. Whether the surgeon removes the septum with the Jones procedure or lyses the septum transcervically, postoperative hysterogram films often show small dog-ears that are leftover tags from the original bifid condition of the uterus. Such dog-ears do not seem to interfere with function, although a postoperative roentgenogram cannot be considered normal in the sense that it does not have the appearance of a normal endometrial cavity after such an operation. If a double cervix is present, the physician should not attempt to unify the cervix because an incompetent cervical os will result. To allow the uterine incision the best possible opportunity to heal, a delay of 4 to 6 months in attempting pregnancy is advised after abdominal metroplasty. P.739

THE JONES METROPLASTY VERSUS THE TOMPKINS PROCEDURE The technique of modified Jones metroplasty is a compromise between the classic Jones metroplasty and the Tompkins metroplasty. In the Jones operation, the entire septum is removed. In the Tompkins operation, a single median incision divides the uterine corpus and septum in half. The incision is carried inferiorly until the endometrial cavity is reached. Each lateral septal half is then incised to within 1 cm of the tubes. No septal tissue is removed. The myometrium is reapproximated, taking care not to place sutures too close to the interstitial portion of the tubes. Proponents of the Tompkins technique suggest that it is simpler than the classic Jones procedure, that it conserves all myometrial tissue and leaves the uterotubal junction in a more normal and lateral position, and that it provides better results than the Jones metroplasty. Good results with the Tompkins technique have been reported by McShane, Reilly, and Schiff.

THE WEDGE METROPLASTY VERSUS TRANSCERVICAL LYSIS There are obvious advantages to a transcervical incision of a uterine septum for patients with a septate uterus. Morbidity is decreased after the procedure and delivery can be vaginal. Term pregnancy rates are comparable to those after abdominal metroplasty for repeated pregnancy wastage. Most of the septa associated with a septate uterus can be cut through the cervix by way of the

hysteroscope or the resectoscope. Nevertheless, cases of broad uterine P.740 septum can benefit from the wedge metroplasty, and reconstructive surgeons should be knowledgeable in its performance.

THE STRASSMANN METROPLASTY The Strassmann procedure is not easily adapted to the septate uterus, but it is the procedure of choice for unification of the two endometrial cavities of an externally divided uterus, both bicornuate and didelphic (Fig. 29.21). A bicornuate uterus cannot be repaired through transcervical lysis because perforation will result. When there has been failure of fusion of the two müllerian ducts, inspection of the pelvic cavity often reveals a broad peritoneal band that lies in the middle between the two lateral hemicorpora. This rectovesical ligament is attached anteriorly to the bladder, folds over and is attached between the uterine cornua, continues posteriorly in the cul-de-sac, and ends with its attachment to the anterior wall of the sigmoid and rectum. It is not invariably present, but when it is, its potential significance in the etiology of the anomaly, possibly by preventing the two müllerian ducts from joining, must be considered. This rectovesical ligament must be removed before a unification procedure can be performed (Fig. 29.21A). FIGURE 29.21. The Strassmann metroplasty with modification. A: If a rectovesical ligament is found, it should be removed. B: An incision is made on the medial side of each hemicorpus and carried deep enough to enter the uterine cavity. The edges of the myometrium will evert to face the opposite side. C: and D: The myometrium is approximated by use of interrupted vertical figure-of-eight 3-0 polyglycolic acid sutures. One should avoid placing sutures too close to the interstitial portion of the fallopian tubes. E: A continuous 5-0 polyglycolic acid subserosal suture is used as a final layer. Tourniquets are removed, and defects in the broad ligament are closed.

View Figure

For hemostasis, tourniquets are used in a manner similar to that described for the modified Jones procedure. The two uterine cornua are incised on their median sides in their longitudinal axes, deeply enough to expose the uterine cavities (Fig. 29.21B). Superiorly, the incision must not be too close to the interstitial portion of the fallopian tubes. Inferiorly, the incision is carried far enough to join the two sides into a single endocervical canal. If it appears that a deeper incision will compromise the competence of the cervix, then a double cervical canal can be left. If the cervix is already duplex, then it should not be joined. As the incision in the myometrium releases the internal stresses in the walls of the hemicorpora, each one everts and is perfectly positioned for apposition, almost as if the original intention in embryologic development is finally to be realized. The suture technique for joining the two sides (Fig. 29.21C, Fig. 29.21D and Fig. 29.21E) is exactly the same as for the modified Jones procedure. The suture line in the uterine corpus should be observed for several minutes to determine the adequacy of hemostasis. Occasionally it is necessary to place one or two extra sutures to control bleeding. A uterine suspension can be performed as necessary. However, in the event of pregnancy, the shortened round ligaments can produce symptoms from an enlarging uterus. Presacral neurectomy in association with uterine unification should be considered only in patients with severe midline dysmenorrhea. The cervix should be dilated to ensure proper drainage from the uterine cavity. This can be accomplished transvaginally after the abdominal procedure or from above by inserting a dilator through the cervical canal into the vagina to be removed later. The operative technique should always be consistent with the goal of maintaining or enhancing fertility and

possibly achieving a successful pregnancy. Tissue surfaces should be kept moist throughout the procedure, and instruments should be selected and used in such a way that tissue damage is minimized. Abdominal packs should be placed in plastic bags to avoid adhesions, or no-lint laparotomy pads can be used. Talc should be carefully washed from gloves, and meticulous aseptic technique should be used. The appendix should not be removed. Lactated Ringer solution containing heparin and corticosteroid can be used for peritoneal lavage throughout the procedure.

Cervical Incompetence Associated with a Double Uterus When a patient with an anomalous uterus, with or without unification, becomes pregnant, she must be watched closely for evidence of cervical incompetence, especially if a history of previous reproductive loss suggests cervical incompetence. Heinonen and associates improved fetal survival rate from 57% to 92% by cervical cerclage. Cerclage was used mostly in patients with a partial bicornuate uterus. In these patients, the fetal salvage rate was improved from 53% before cerclage to 100% afterward. Prematurity also was decreased, from 53% to 3%. The authors stress that cervical incompetence, not the uterine anomaly, is the proper indication for cerclage in these patients. However, the frequency with which these problems are found together suggests the importance of doing a careful evaluation for both problems. Some reproductive losses from a uterine anomaly might be prevented by cerclage of an incompetent cervix during metroplasty. However, routine cerclage at the time of metroplasty is not recommended. Attempts to unify a double cervix or a septate cervix also are not recommended because of the possibility of causing cervical incompetence. However, a double or septate cervix can adversely affect the outcome of delivery if vaginal delivery is attempted, and delivery should be by Caesarean section if it appears that the cervix will cause dystocia.

Mode of Delivery After Metroplasty The scar formed in the myometrium after unification is as strong as, if not stronger than, the scar formed after Caesarean section. The biologic conditions under which healing occurs are entirely different in these two situations. Endomyometritis is a common complication after Caesarean section but is not a complication of uterine unification. Of 71 known pregnancies in Strassmann's collected series reported in 1952, 61 were delivered vaginally. There were no cases of uterine rupture during pregnancy or delivery. Despite evidence that the uterine scar heals securely after unification operations, our policy is to recommend delivery by elective Caesarean section in all patients who have undergone abdominal metroplasty. Patients can deliver vaginally after a metroplasty by hysteroscope or resectoscope.

Diethylstilbestrol-Related Uterine Anomalies Exposure of the female fetus to DES can cause significant anomalous development of the uterus, as reported P.741 P.742 by Kaufman and associates and by Haney and colleagues. The T-shaped uterus is the variant most commonly seen. It is associated with an increased rate of spontaneous abortions, preterm deliveries, and ectopic pregnancies. Nagel and Malo determined the feasibility of correcting the uterine malformations seen in DES-exposed women by incising constriction rings and septa. Their goal was to incise the irregular uterine walls until the cavity assumed a smooth, straight line from the lower uterine cavity to the uterine tubal ostium. Their results suggested that metroplasty can decrease pregnancy loss but does not enhance fertility. The editors of this book suggest that the rare patient can benefit from a uterine reconstructive procedure, but that most will not. Surgeons may never develop a large series to document efficacy of surgical outcomes because patients with this anomaly will eventually age beyond reproductive years, and some latitude is required in cases that might possibly benefit from metroplasty. DES-exposed patients must be monitored closely for evidence of dilatation and effacement of the cervix early in pregnancy. Cervical cerclage may be indicated in some patients.

Unicornuate Uterus A unicornuate uterus can be present alone or with a rudimentary horn or bulb on the opposite side. In a series reported by Heinonen and associates, 11 of 13 patients with a unicornuate uterus had a rudimentary horn, and two did not. The rudimentary anlage (uterine muscle bundle or bulb) can

communicate directly with the unicornuate uterus. In some instances, there is no cavity within the anlage or there is no rudimentary horn. Most rudimentary horns are noncommunicating (90% according to O'Leary and O'Leary). The two sides may be connected by a fibromuscular band, or there may be no connection and no communication between the two uterine cavities. Fedele and associates have found sonography useful in determining the presence of not only a rudimentary horn but also a cavity within.

Associated Anomalies Urinary tract anomalies are often associated with a unicornuate uterus. On the side opposite the unicornuate uterus there may be a horseshoe or a pelvic kidney, or the kidney may be hypoplastic or absent. This is especially true if there is associated müllerian duct obstruction. When all müllerian duct derivatives and the kidney are absent on one side, this implies failure of development of the entire urogenital ridge, including the genital ridge where the ovary forms. In addition, the ovary may be malpositioned (Fig. 29.22). Rock, Parmley, and associates reported a unilateral ovary located above the pelvic brim in four cases of uterine anomalies. The orifice of the müllerian duct develops at about the level of the fourth thoracic vertebra (T4) in the embryo. The tip subsequently migrates along the course of the müllerian duct into the pelvis. The orifice of the duct or the fimbriated end of the tube comes to lie in the pelvis as a result of differential growth of the fetus. The subsequent differential growth is retarded so that the portion of the urogenital ridge that gives rise to both the gonad and tube does not displace into the pelvis. Malpositions of the ovary and tube are the result. FIGURE 29.22. A unicornuate uterus associated with ovarian malposition on the left. Note that the ovary and the tube are slightly above the pelvic brim. In this instance, the ovary measured 6 inches long.

View Figure

Reproductive Performance According to Heinonen and associates, the unicornuate uterus carries the poorest fetal survival rate (40%) of all uterine anomalies. In 1957, Jones reported similar findings. The abnormal shape, the insufficient muscular mass of the uterus, and the reduced uterine volume and inability to expand may explain the poor obstetric outcome. Moutos and colleagues compared the reproductive performance of the unicornuate uterus with that of the didelphic uterus. Twenty of the 29 women with a unicornuate uterus produced a total of 40 pregnancies, whereas 13 women with a didelphic uterus produced a total of 28 pregnancies. The percentages of pregnancies resulting in preterm delivery, term delivery, and living children were similar in both groups. The authors concluded that reproductive performance of the unicornuate uterus was not different from that of the didelphic uterus, that it is uncommon for either malformation to be a primary cause of infertility, and that there is insufficient information to support recommendation of placement of a cervical cerclage in the absence of cervical incompetence. Thus, there is no evidence that uterine reconstruction should be performed for patients with a unicornuate (or a didelphic) uterus. Because most cases of unicornuate uterus have a noncommunicating rudimentary uterine horn on the opposite side, there is danger of pregnancy in the rudimentary horn from transperitoneal migration of

sperm or ovum from the opposite side. According to Holden and Hart, approximately 350 cases of pregnancy in a rudimentary horn have been reported since the original case report by Mauriceau in 1669. O'Leary and O'Leary found the corpus luteum on the side contralateral to the rudimentary horn containing a pregnancy in 8% of cases. Signs and symptoms of an ectopic pregnancy develop with eventual rupture of the horn if the pregnancy is not detected early. Rupture through the wall of the vascular rudimentary horn is associated with sudden and severe intraperitoneal hemorrhage and shock. Death can occur in a few minutes. It is surprising that the current mortality rate has decreased to 5%. Very little, if anything, can be done to improve the reproductive performance of patients with a unicornuate uterus. The physician should observe closely for cervical incompetence and perform cerclage as indicated. Andrews and Jones have suggested that removal of the rudimentary uterine horn may improve the chances of a successful pregnancy, but the experience is too small to support a definite recommendation. Cases of asymmetric development of the unicornuate uterus with an opposing rudimentary uterine horn are not amenable to unification. P.743

Longitudinal Vaginal Septum Failure of fusion of the lower müllerian ducts that form the vagina can result in a vagina with a longitudinal septum. The septum can be partial or complete. Young patients have difficulty using tampons. In cases of didelphic uterus with a longitudinal vaginal septum, one uterine hemicorpus is usually better developed than the other. If intercourse consistently occurs on the vaginal side connected to the uterine hemicorpus that is less well developed, then infertility or repeated abortion could result. For these reasons, the septum should be removed (when the patient is not pregnant) unless there is a contraindication. This can usually be accomplished easily with reasonable precautions against injury to the urethra, bladder, and rectum. Haddad and colleagues reported their experience over a 24-year period with management of the longitudinal vaginal septum. The retrospective review of 202 patient charts described a complete septum (extending from cervix to introitus) in 45.6% of patients, high partial in 36.1%, and a medium or low partial, involving only the distal vagina, in 18.3%. Uterine malformations were noted in 87.8% of cases. The frequency of uterine malformations was 99.4% in cases of complete or partial high septum and 30.3% in cases of partial medium or low septum. The most common malformation was class Va complete septate uterus in 59.5% of malformations, followed by class III didelphys uterus (24.3%) and class Vb partial septate uterus (15%). Section or resection was performed in 201 cases. Bladder injury in one patient was the only reported complication. As highlighted by the high prevalence of associated uterine malformations in this review, management should always include an assessment of uterine anatomy.

Asymmetric Obstruction of the Uterus or Vagina Unicornuate Uterus and Noncommunicating Uterine Anlage Containing Functional Endometrium If one müllerian duct develops normally while the opposite müllerian duct fails to develop or develops incompletely, then a relatively normal unicornuate uterus is found on one side and the cervix, musculature, uterine cavity, endometrium, fallopian tube, blood supply, and ligamentous attachments are absent or hypoplastic to a varying degree on the other side. Obstruction to menstruation P.744 can also occur to varying degrees on the improperly developed side. For example, if a rudimentary uterine horn does not communicate externally but does have an endometrium-lined uterine cavity, then clear symptoms of obstructed menstruation may begin soon after menarche, and severe dysmenorrhea will be present. Cryptomenorrhea can be overlooked as the diagnosis because there is cyclic menstruation from the opposite side. It is important to make the diagnosis as soon as possible, because if the lumen of the tube communicates with the P.745 endometrial cavity of the rudimentary uterus, then retrograde menstruation and pelvic endometriosis will develop, and reproductive potential can be destroyed. During the operation illustrated in Figure 29.23, which was performed to remove an obstructed rudimentary uterine horn, the fallopian tube was obstructed and retrograde menstruation was impossible. Occasionally, the fallopian tube connected to the

rudimentary uterine horn may not be patent because of incomplete development. FIGURE 29.23. A: A noncommunicating rudimentary horn with functional endometrium that contains menstrual blood under pressure. Note the congenital abnormality of the fallopian tube, which prevented retrograde menstruation. B: The same patient after excision of the rudimentary horn.

View Figure

Unilateral Obstruction of a Cavity of a Double Uterus Another example of a rare obstructed lateral fusion problem is the complete septum between two uterine cavities illustrated in Figure 29.24. One cavity communicated with a cervix and the other did not. This could represent an example of unilateral failure of cervical development. The patient complained of incapacitating dysmenorrhea that appeared shortly after the menarche and lasted 5 days. A tense, cystic mass was palpable in the right half of the pelvis. The operation, described originally by Jones in the second edition of this book, consisted of making an incision through the anterior wall of the cystic right portion of the uterus. It was found to contain old menstrual blood. The entire septum was excised and the uterus was reconstructed by anastomosis of the two cavities. A continuous lock-stitch was reinforced by interrupted myometrial sutures, and the plastic reconstruction of the uterus was completed by a third layer of interrupted sutures uniting myometrium and serosa.

FIGURE 29.24. A: A double uterus seen at operation. Hematometra in the right uterine cavity (inset), which does not communicate with the other cavity or the cervical canal. B: The septum of the double uterus has been excised and anastomosis is performed to unite the two cavities. C: Anastomosis is completed. The small incision in the left uterine cavity was made before the septum was removed for the purpose of orientation.

View Figure

Sanders and colleagues described several cases in which the role of interventional radiology was crucial in the management of obstructive anomalies. The report described the drainage of a noncommunicating right uterine cavity distended with blood in a unicornuate uterus in a 14-year-old patient. Adequate access was established by using ultrasound-guided needle aspiration followed by a hysteroscopic excision. The assistance of interventional radiologic procedures, including percutaneous drainage and dilatation of small maldeveloped areas, may allow access to areas otherwise inaccessible by conventional mechanisms and assist in preserving reproductive function. P.746

Double Uterus with Obstructed Hemivagina and Ipsilateral Renal Agenesis The unique clinical syndrome consisting of a double uterus, obstruction of the vagina (unilateral, partial, or complete), and ipsilateral renal agenesis is rare. The renal agenesis (mesonephric involution) on the side of the obstructed vagina associated with a double uterus and double cervix is suggestive of an embryologic arrest at 8 weeks of pregnancy that simultaneously affects the müllerian and metanephric ducts. The exact cause is unknown.

Diagnostic Groups Clinical symptoms vary depending on the uterovaginal relations in individual cases, but the syndrome can be described generally in three groups. Group 1 patients have complete unilateral vaginal obstruction without uterine communication, resulting in a paravaginal mass and symptoms of severe dysmenorrhea and lower abdominal pain. Menses are regular. Group 2 patients have an incomplete unilateral vaginal obstruction without uterine communication. The presenting symptoms are lower abdominal pain, severe dysmenorrhea, excessive foul mucopurulent discharge, and, in some instances, intermenstrual P.747 bleeding. Group 3 patients have complete vaginal obstruction with a laterally communicating double uterus. They have a paravaginal mass, lower abdominal pain, and dysmenorrhea. Menses are regular. A 10-year review of patients with this anomaly was published by Phupong and colleagues. Most patients presented with dysmenorrhea (73%), or a pelvic or paravaginal mass (71%). The right uterus and vagina were affected in 63.5% of patients. Because menses in patients with this syndrome are rarely irregular, the possibility of this syndrome as a diagnosis can easily be overlooked. A careful pelvic examination is necessary to make the correct diagnosis. MRI can identify the obstructed vagina, double uterus, and absence of a kidney on the side of

the obstruction (Fig. 29.25), but it may not be helpful if there is incomplete vaginal obstruction or a uterine communication. FIGURE 29.25. A double uterus with unilateral complete vaginal obstruction and ipsilateral renal agenesis. Magnetic resonance imaging reveals the left hematocolpos, both uteri, and absence of the left kidney on the side of the vaginal obstruction.

View Figure

Complete unilateral vaginal obstruction (group 1) can go unrecognized for a number of years after the onset of menses. The vagina is quite distensible and can accommodate a large amount of accumulated blood in the obstructed side. There is sufficient absorption of menstrual blood between periods so that each subsequent flow can add to the increments of accumulated blood without pain. Nevertheless, once retrograde menstruation occurs, endometriosis invariably is the result.

Surgical Treatment Careful excision of the vaginal septum is the treatment of choice for a unilateral vaginal obstruction. Prophylactic antibiotics should be administered before surgery. After opening the vaginal pouch, the surgeon should use suction and lavage to remove the pooled blood and mucus. Phupong and colleagues' review also confirms the successful use of this primary therapy in 84.3% of patients. Haddad and colleagues reported a similar experience in a report describing patient management over a 27-year period. Excision of the vaginal septum was successful in 88% of patients with complete excision in one procedure in 92% of those patients. In cases of pyocolpos or hematocolpos, distention and stretching of the septal tissue may increase the risk of inadequate resection and possible postoperative stenosis; the authors found the use of a two-step graduated resection advantageous to ensure adequate resection. A limited resection (3 cm) was performed to allow adequate drainage, followed by a return to the operating room in approximately 1 month to remove any remaining septum. Because the obstructing septum is usually thick, removal can be difficult. Clamps should be used to isolate a generous vaginal pedicle while the suture is being tied in place to prevent slippage of tissue. Such pedicles generally retract during healing, and formation of a vaginal stenosis is avoided. In most instances, surgery is restricted to excision of the septum, and abdominal exploration is unnecessary. Uterine reconstruction is not indicated for cases of lateral communication of the uterine horns. Some authors have reported the use of hemi-hysterectomy in patients with a high, thick-walled obstruction, massive ovarian involvement, endometriosis, or adenomyosis; however, this is generally not recommended in young patients.

Reproductive Performance Reproductive performance for patients with this disorder is usually consistent with that of patients with a double uterus unless the delay in diagnosis and resection of the obstructing septum has been sufficient to destroy tubal function or to cause the development of endometriosis. P.748 Haddad and colleagues' review was notable for a predominance of pregnancies (80%) in the contralateral endometrial cavity.

UNUSUAL CONFIGURATIONS OF VERTICAL-LATERAL FUSION DEFECTS Part of "Chapter 29 - Surgery for Anomalies of the Müllerian Ducts" Unusual configurations of both vertical and lateral fusion defects may occur simultaneously. Figure 29.26 depicts the radiographic evaluation (MRI) of a young woman in whom cryptomenorrhea developed above a transverse vaginal septum. The MRI study depicting the hematocolpos also suggests the longitudinal vaginal septum. Incision, drainage, and resection of the transverse vaginal septum allowed appropriate evaluation of the more proximal müllerian anatomy. The artist's depiction in Figure 29.26 demonstrates the unusual constellation of a didelphys uterus with an intrauterine communication of the cavities and a longitudinal vaginal septum. This type of atypical combination occurs frequently enough to emphasize the importance of proper delineation of individual anatomy preoperatively for proper surgical preparation. FIGURE 29.26. An unusual combination of both vertical and lateral fusion defects. A: A didelphys uterus with an intrauterine communication and a longitudinal vaginal septum are present proximal to a transverse vaginal septum. B: The magnetic resonance image demonstrates the presence of both septa.

View Figure

Müllerian duct anomalies can occur in association with a variety of other problems. For example, Stanton reported that in a series of 70 patients with bladder exstrophy, 30 (43%) had reproductive tract abnormalities. He suggested that the true figures were actually higher. Müllerian abnormalities included absence of the vagina; septate vagina; unicornuate, bicornuate, and didelphic uterus; and absent uterus. Fewer müllerian anomalies are seen with epispadias. Jones investigated anomalies of the external genitalia and vagina in 30 patients with bladder exstrophy seen at the Johns Hopkins Hospital and suggested operative techniques for correction of these anomalies. Techniques for the management of other gynecologic and obstetric problems (especially uterine prolapse) also have been discussed by Weed and McKee and by Blakeley and Mills. A number of other rare combinations of congenital malformations of the vagina and perineum have been found in association with uterine anomalies. Their surgical correction, especially in children, is reported by Hendren and Donahoe and by others. Several authors have considered the uterovaginal anomalies that occur in association with multiple other gastrointestinal and genitourinary abnormalities. Goh described an infant girl with complete duplication of the bladder, urethra, uterus, and vagina associated with a urogenital sinus and an anterior ectopic anus. Gastol and Magalhaes also described children with complete duplication of the bladder, urethra, vagina, and uterus. These complex anomalies include significantly more defects than lateral fusion concerns in the müllerian ducts. These cases emphasize the variable anatomy in this rare group of anomalies and that a substantial effort should be placed on defining anatomy before surgical exploration and management. Sheldon and others reviewed 13 consecutive cases of vaginal reconstruction in pediatric patients with multisystem anomalies. The review emphasized several important principles involved in the surgical management: (1) all anticipated perineal reconstruction should be performed in a single stage, (2) urethral catheterization has an important role, (3) urinary reconstruction is often intimately involved in the vaginal reconstruction, (4)

avoidance of overlapping suture lines is essential for optimal healing, (5) maximum growth potential of the neovagina should be considered, and (6) meticulous follow-up of proper routine dilatation of the neovagina should be expected. Coordinated reconstruction of all organ systems is especially important in these complex cases. P.749 Müllerian duct anomalies are seen with the McKusick-Kaufman syndrome, an autosomal recessive disorder. Other clinical findings reported with this syndrome include hydrometrocolpos, postaxial polydactyly, syndactyly, congenital heart disease, intravaginal displacement of the urethral meatus, and anorectal anomalies. In 1982, Jabs and co-workers added the most recent unusual case to the few cases previously reported in the literature. Müllerian duct anomalies may also affect the development of the fallopian tube. Although extremely rare, episodes of unilateral or bilateral absence of the fallopian tube have been reported. Of the less than 10 cases in the literature, Eustace reported two of the described cases. He hypothesized that compromise of the local blood supply to the caudal aspect of the Müllerian duct was a more likely cause than a fusion disorder. This situation could affect fallopian tube development to a variable extent with even some effect on ovarian development.

Chapter 30 Leiomyomata Uteri and Myomectomy Lesley L. Breech John A. Rock Leiomyomata are the most common tumors of the uterus and the female pelvis. This chapter discusses the pathologic and clinical features of uterine leiomyomata, the choice of treatment, and the indications and techniques for myomectomy. Hysterectomy is sometimes required for the management of leiomyomata. It is also performed for many other indications, but leiomyomata uteri is the most common indication for hysterectomy. Refer to Chapter 31 for a complete discussion of hysterectomy. Advances in gynecologic surgery just 100 years ago finally brought this common and sometimes fatal disease of women under reasonable control. Before the 20th century, no effective treatment was available. Uterine leiomyomata often grew to enormous size and caused great suffering from bleeding, pain, and emaciation (Fig. 30.1). Death from this benign disease was not uncommon. Progress in gynecologic surgery and anesthesia finally allowed the safe removal of these tumors by skilled gynecologic surgeons. FIGURE 30.1. The patient is thin and emaciated, the outline of the ribs being prominent. Such advanced and neglected cases of multiple uterine leiomyomata are rarely seen today.

View Figure

No one played a more important role in this endeavor than Drs. Kelly and Cullen. Working together at the Johns Hopkins Hospital, they gradually developed surgical techniques that were successful in preventing and controlling intraoperative hemorrhage. Several illustrations from their magnificent treatise, Myomata of the Uterus, published in 1907, are included in this chapter. In the preface, Cullen wrote: It was my good fortune to come to Baltimore in 1891, shortly after the hospital opened. At that time many cases of myoma were considered inoperable, and even when hysterectomy was undertaken it was only in the cases in which a stout rubber ligature could be temporarily tied around the cervix and when, as happened in some cases, this ligature slipped, alarming hemorrhage followed. Then came the systematic controlling of each of the cardinal vessels; later the bisection, and finally the transverse severance of the cervix as a preliminary feature of the operation in exceptionally difficult cases, until at present a myomatous uterus that cannot be removed is almost unheard of. I have watched the gradual simplication of the surgical procedures with the greatest interest. Many American surgeons have had much to do with the wonderful advance in this direction, but I know of no other man, either here or abroad, who has done as much toward this advancement as Howard A. Kelly. The mortality rate for 1,373 operations performed for uterine leiomyomata at the Johns Hopkins Hospital between 1889 and 1906 was 5.75%; it was less than 1% for 238 operations performed between 1906 and 1909. In 55 patients, no operation was attempted because of P.754 refusal or the patient's weakened condition. Among these patients, 21 deaths occurred in the hospital. Death from uterine leiomyomata is a rare occurrence today. The almost total elimination of mortality caused by this tumor represents a major milestone in the health care of women. During the past century, hysterectomy and myomectomy by the traditional and classic techniques have been the main treatment for women with uterine leiomyomata and significant symptoms, and they continue to be so today. Each year in the United States, about 175,000 hysterectomies are performed with uterine

leiomyomata as the primary indication. However, this traditional management will change in the future, for several reasons: Concern regarding the increasing costs of health care has focused on the need to use effective but less expensive methods of management of uterine leiomyomata. Advances in surgical technology now allow certain patients to be treated with new, minimally invasive techniques, including laparoscopic hysterectomy, laparoscopic-assisted vaginal hysterectomy, laparoscopic myomectomy, laparoscopic myoma coagulation (myolysis), and hysteroscopic resection of submucous myomata. Under proper circumstances, these procedures can be safe, effective, and less costly, but more time is needed before they will be available to a greater number of patients in need. Interest in nonsurgical management also appears to be increasing with more data available regarding minimally invasive procedures including uterine artery embolization. This procedure is emerging from an investigational realm to common clinical practice. As more long-term data become available, outcomes and prognosis may be more clearly delineated. A medical approach to the management of patients with leiomyomata is now available. Gonadotropin-releasing hormone (GnRH) analogs, administered for 4 to 6 months, cause most uterine leiomyomata to shrink. However, the myomata regain their original size several months after the GnRH analog is discontinued. This medical regimen has been useful as an adjunct to surgical management. Women who become symptomatic with leiomyomata just before menopause can be treated temporarily with GnRH analogs and can possibly avoid surgical therapy. Uterine leiomyoma is a major public health and women's health care problem. Society has a legitimate reason for interest and concern and has questioned the advisability of hysterectomy for the management of most cases of uterine leiomyomata. Many women insist on the preservation of uterine function for future childbearing, and sometimes even when future childbearing is not desired or not likely to occur. There will probably be a greater emphasis on expectant management, medical management, minimally invasive surgical procedures, and conservational management of uterine leiomyomata in the future. In the future, the traditional and classic techniques of hysterectomy and myomectomy will be required less often for patients with symptomatic leiomyomata. At present, however, these operations are still appropriate in many situations. ETIOLOGY, PATHOLOGY, AND GROWTH CHARACTERISTICS OF UTERINE LEIOMYOMATA CLINICAL FEATURES OF UTERINE LEIOMYOMATA CHOICE OF TREATMENT FOR UTERINE LEIOMYOMATA

ETIOLOGY, PATHOLOGY, AND GROWTH CHARACTERISTICS OF UTERINE LEIOMYOMATA Part of "Chapter 30 - Leiomyomata Uteri and Myomectomy" A leiomyoma is a benign tumor composed mainly of smooth muscle cells but containing varying amounts of fibrous connective tissue. The tumor is well circumscribed but not encapsulated. Various terms are used to refer to the tumor, such as fibromyoma, myofibroma, leiomyofibroma, fibroleiomyoma, myoma, fibroma, and fibroid. The latter designation is the one most commonly used, but it is the least accurate and acceptable. The term leiomyoma is a reasonably accurate one that emphasizes the origin of this tumor from smooth muscle cells and the predominance of the smooth muscle component. P.755 The tissue culture work of Miller and Ludovici suggested an origin from smooth muscle cells. The studies of Townsend and associates suggest a unicellular origin for leiomyomata. Leiomyomata are the most common tumors of the uterus and female pelvis. It is impossible to determine their true incidence accurately, although the frequently quoted incidence of 50% found at postmortem examinations seems reasonable. Leiomyomata are responsible for about one third of all hospital admissions to gynecology services. It is well recognized that the incidence is much higher in black women than in white. In a careful study of leiomyomata among women in Augusta, GA, Torpin and associates found the incidence among black women to be three and one third times that among white women. There is no explanation for this racial difference. Leiomyomata also are larger and occur at a younger age in black women. In our institution, a large degenerating intramural leiomyoma was removed from a 1-year-old black girl; the tumor had enlarged the uterus to the level of the umbilicus. In black women, leiomyomata are not uncommon before 30 years of age. However, they are uncommon in either race before 20 years of age. Patients with uterine leiomyomata often have a positive family history of uterine leiomyomata. This suggests the presence of a gene encoding for their development. About 40% to 50% of leiomyomas show karyotypically detectable chromosomal abnormalities that are both nonrandom and tumor-specific. Identified chromosomal abnormalities include t(12;14)(q15;q23-24), del(7)(q22q32), rearrangements involving 6p21, 10q, trisomy 12, and deletions of 3q. Interestingly, a recent study of 217 myomas found a positive correlation between the presence of a cytogenetic abnormality and the anatomic location of the myoma. In this study by Brosens and colleagues, submucous myomas were consistently shown to have fewer cytogenetic abnormalities when compared with intramural and subserous lesions (12% versus 35% and 29%, respectively). An increased prevalence in certain races, twin studies indicating higher correlation with hysterectomy in monozygotic twins, and increased incidence in first-degree relatives all seem to support an inherited predisposition. The true genetic contribution to the development of uterine leiomyoma remains to be defined. Most of the data concerning the incidence of uterine leiomyomata are based on gross examination of the uterus, routine pathology reports, or the clinical diagnosis of uterine leiomyomata. Cramer and Patel subjected 100 uteri to gross serial sectioning at 2-mm intervals. They found 649 leiomyomata, roughly threefold the number identified by routine pathologic examination. Admittedly, some were only a few millimeters in diameter, but all were grossly visible. In 48 uteri with no mention of leiomyomata in the routine report, 27 were found to have small tumors. The incidence of leiomyomata was the same in premenopausal and postmenopausal uteri, although the average number of leiomyomata and the average size of the largest leiomyoma were greater in the premenopausal women. This work has important implications for future epidemiologic studies. It also suggests that it is almost never possible to surgically remove all leiomyomata when a myomectomy is performed. The growth of leiomyomata is dependent on estrogen production. The tumors thrive during the years of greatest ovarian activity. Continuous estrogen secretion, especially when uninterrupted by pregnancy and lactation, is thought to be the most important underlying risk factor in the development of myomata. After menopause, with regression of ovarian estrogen secretion, growth of leiomyomata usually ceases. Actual regression in the tumor size may occur. There are rare instances, however, of postmenopausal growth of be-nign leiomyomata, suggesting the possibility of postmenopausal estrogen production either in the ovary or elsewhere. Postmenopausal ovarian cortical stromal hyperplasia may be associated with an increase in estrogen secretion by the ovary. The postmenopausal ovarian stroma in a variety of presumably inactive

ovarian tumors, including mucinous cysts and Brenner tumors, can also produce estrogen. When a central pelvic tumor presumed to be uterine leiomyomata enlarges after menopause, one should think of the possibility of malignant change in the leiomyoma itself or in the adjacent myometrium, or of the growth of a new pelvic tumor of extrauterine origin. Older nulliparous women have an increased risk of developing leiomyomata. However, in multiparous women, the relative risk decreases with each pregnancy. A woman who has had five term pregnancies has only one fifth the risk of a nulliparous woman of developing myomata. The risk is reduced in women who smoke and is increased in obese women; this is possibly related to the conversion of androgens to estrogen by fat aromatase. The observation that leiomyomata may show significant enlargement during pregnancy provides further clinical evidence of the relation of estrogen and progesterone to the growth of these tumors. However, a better blood supply during pregnancy might also encourage their growth. In a prospective ultrasonographic study of 29 pregnant patients with uterine leiomyomata, Aharoni and associates found no evidence of enlargement of the myomata in 78%. Lev-Toaff and colleagues also confirmed that some leiomyomata do enlarge during pregnancy in response to estrogen and progesterone. In the initial two decades following the introduction of oral contraceptives containing high-dose estrogen, there was a striking increase in the occurrence of large leiomyomata among young women of all racial backgrounds who took these pills. Although the growth of uterine leiomyomata is not invariably stimulated, oral contraceptives containing high-dose estrogen should not be prescribed for women with these tumors. Oral contraceptives with low-dose estrogen are less likely to stimulate growth. According to Parazzini and associates, there is no significant relation between the occurrence or growth of leiomyomata and the newer oral contraceptives that contain much smaller amounts of estrogens P.756 and progestins, and some believe that the risk of developing myomata is reduced with these low-dose pills. Scientific investigators have been intrigued by the observation that leiomyomata develop during the reproductive years, sometimes grow during pregnancy, and regress after menopause. Nelson, Lipschutz, and others have produced multiple leiomyomata artificially on the serosal surface of the uterus and other peritoneal surfaces in guinea pigs given prolonged estrogen injections. Spellacy and co-workers found that levels of plasma estradiol were the same in patients with and without leiomyomata. However, Wilson and associates found a significantly higher concentration of estrogen receptors in leiomyomata than in myometrium. Farber and colleagues reported that these tumors bind about 20% more estradiol per milligram of cytoplasmic protein than does the normal myometrium of the same organ. This observation was not uniformly true for all leiomyomata, suggesting that different cellular components with a leiomyoma may be associated with different biologic activity. Otubu and co-workers found the concentration of estradiol to be significantly higher in leiomyomata than in normal myometrium, especially in the proliferative phase of the menstrual cycle. Soules and McCarty reported that leiomyomata had more estrogen receptors than did normal uterine tissues in the first phase (days 1 through 9) and in the second phase (days 10 through 18) of the menstrual cycle. Gabb and Stone found that the ability to convert estradiol to estrone was similar in leiomyomata and myometrium. However, Pollow and associates found the conversion of estradiol into estrone to be significantly lower in leiomyomata than in myometrium. This difference in conversion rate could result in a relative accumulation of estrogen in a leiomyoma, causing a hyperestrogenic state within the tumor and surrounding tissues. The enzyme 17?-hydroxy dehydrogenase accelerates the conversion of estradiol to estrone. Leiomyomata have a low concentration of 17?-hydroxy dehydrogenase, which results in a relative accumulation of estradiol in leiomyomatous tissue. These findings may explain the myometrial hypertrophy that is invariably present with leiomyomata. Other abnormalities in endocrine function have also been suggested. Ylikorkala and colleagues found that pituitary function may be abnormal in women with leiomyomata. Patients with leiomyomata had a low follicle-stimulating hormone level and a diminished follicle-stimulating hormone response to pituitary GnRH. There was an excessive prolactin response to thyrotropin-releasing hormone. Spellacy found that the peak levels of human growth hormone reached during a hypoglycemic test were twice as high in patients with leiomyomata as in the control group. Reddy and Rose suggested the possibility that 5?-reduced androgens may play a role in the pathophysiology of uterine leiomyomata, because a significant increase in 5?-reductase has been found in leiomyoma tissue as compared with myometrium and endometrium. Influenced by the experimental investigations of Lipschutz and associates, Goodman in 1946 treated patients with uterine leiomyomata with progesterone and noted a decrease in tumor size in all patients. However, Segaloff and colleagues reported no effect in their study. Goldzieher and co-workers

produced histologic evidence of extensive degenerative changes in leiomyomata by administering high-dose progestin therapy (medrogestone in high doses for 21 days). Filiceri and associates have reported the regres-sion of a uterine leiomyoma after long-term administration of a long-acting luteinizing hormone-releasing hormone agonist given to suppress ovarian estrogensecretion. Coutinho successfully used a potent 19-norsteroid antiestrogen–antiprogesterone to treat excessive uterine bleeding in 16 patients with uterine leiomyomata. A reduction in the size of the tumors was noted. Although the exact etiology of uterine leiomyomata is not known, the puzzle may be solved bit by bit by the research of Kornyei and colleagues, Wilson and co-workers, Tamaya and associates, Buchi and Keller, Sadan and colleagues, and others who continue to investigate estrogen and progesterone as possible growth factors. Although some data are conflicting, evidence suggests that both estrogen and progesterone are involved in the growth of uterine leiomyomata. The possibility that progesterone may play a role in the growth of leiomyomata is suggested by the work of Kawaguchi and co-workers, who found a higher mitotic count in leiomyomata obtained in the proliferative phase of the menstrual cycle. Anderson and associates have shown that medroxy-progesterone acetate, a progestin, causes a decrease in connexin-43 messenger ribonucleic acid levels in primary cultures of human myometrium and leiomyoma. Connexin-43 is a gap junction protein whose formation is stimulated by 17?-estradiol. According to the research data of Brandon and colleagues, progesterone receptor messenger ribonucleic acid is overexpressed in uterine leiomyomata, compared with normal adjacent myometrium, suggesting that amplified progesterone-mediated signaling is instrumental in the abnormal growth of these tumors. It is possible that the increased amount of progesterone receptor is caused by an alteration of estrogen or estrogen receptors in leiomyomata. The work of Kastner and co-workers and Nardulli and associates has demonstrated that progesterone receptor expression is regulated by estrogen. Research in recent years has also focused on polypeptide growth factors in the stimulation of growth of leiomyomata. Polypeptide growth factors that have been investigated include epidermal growth factor, transforming growth factor-alpha, insulin-like growth factor (IGF), and fibroblast growth factor. Other growth factors may also be involved. Polypeptide growth factor research has been performed by Goustin and colleagues, by Hoffmann and co-workers, by Lumsden and associates, and by others. A brief review of this research has been written by Vollenhoven and associates, who have been involved in the study of IGFs in uterine leiomyomata. Results from a study by Strawn and colleagues demonstrate that IGF-I stimulates leiomyoma growth in P.757 a dose-related manner over that of normal myometrial tissue in monolayer culture. This stimulatory effect, in the absence of sex steroid hormones or other growth factors, provides additional support that IGF-I may play an important direct role in the pathogenesis of these tumors, possibly by modulating the response of these tumors to various levels of sex steroids. Dawood and Kahn-Dawood were unable to find any significant elevation in peripheral levels of serum IGF-I in nonpregnant premenopausal women with uterine leiomyomata of 14 weeks' gestational size. The authors state, “Nevertheless, the finding does not detract from the potential paracrine or autocrine role that IGF-I produced by leiomyoma cells may have either on the growth of its own or adjacent myomas or on the vascular supply and blood flow of the uterus and myomas.” Rein and co-workers have proposed a hypothesis to explain the pathogenesis of myomata. This hypothesis suggests a critical role for progesterone in their growth. They state: The initiation and growth of myomas likely involves a multistep cascade of separate tumor initiators and promotors. The initial neoplastic transformation of the normal myocyte involves somatic mutations. Although the initiators of the somatic mutations remain unclear, the mitogenic effect of progesterone may enhance the propagation of somatic mutations. Myoma proliferation is the result of clonal expansion and likely involves the complex interactions of estrogen, progesterone, and local growth factors. Estrogen and progesterone appear equally important as promotors of myoma growth. To treat patients with uterine leiomyomata properly, the gynecologic surgeon must be familiar with their pathology, growth characteristics, and clinical features. Leiomyomata may be single, but most are multiple. They develop most commonly in the uterine corpus and much less often in the cervix. They may develop in the round ligaments, but this is rare. Because they arise in the myometrium, they are all interstitial or intramural in the beginning. As they enlarge, they can remain intramural, but growth often extends in an internal or external direction. Thus, the tumor can eventually become subserous or submucous in location. A subserous tumor can become pedunculated and occasionally parasitic, receiving its blood supply from another source, usually the omentum. A submucous tumor can also become pedunculated and may gradually dilate the endocervical canal and protrude through the cervical os. Indeed, a submucous myoma

may descend through the vagina. Rarely, chronic uterine inversion results if the prolapsing submucous leiomyoma is attached to the top of the endometrial cavity and pulls the uterine fundus downward through the cervix. In general, subserous leiomyomata contain more fibrous tissue than submucous leiomyomata. However, submucous leiomyomata contain more smooth muscle tissue than subserous leiomyomata. Sarcomatous change is more common in submucous tumors. The typical uterine leiomyoma is a firm multinodular structure of variable size. The largest tumor, reported by Hunt in 1888, weighed more than 65 kg. Tumors of 4 to 5 kg are not rare, but most are smaller. At the operating table, leiomyomata appear as nodular tumors of different sizes that distort the uterus in various ways, depending on their size, location, and direction of growth. Growth between the leaves of the broad ligament and origin from the cervix may make surgical removal difficult. Subserous and subserous pedunculated tumors, as well as intraligamentous tumors, may create problems in diagnosis because they are difficult to distinguish from tumors arising from the adnexal organs (Fig. 30.2). When tumors cause symmetric enlargement of the uterus, they may be mistaken for a pregnant uterus on bimanual examination. FIGURE 30.2. The uterine corpus is almost completely replaced by small and large myomas in intramural, subserous, and submucous positions. Some are pedunculated. A pedunculated submucous myoma is dilating the endocervical canal. A pedunculated subserous myoma is adjacent to the left ovary and will interfere with its palpation.

View Figure

The “normal” intramural leiomyoma on section protrudes from the surrounding compressed myometrium. Ordinarily there is a clear distinction between the myoma and the myometrium so that dissection between the P.758 two is easy to accomplish. Myomata usually can be removed from surrounding myometrium with ease. Although these tumors are not encapsulated, a clear distinction can usually be made between a myoma and the myometrium that surrounds it. The cut surface appears as glistening pinkish white and gray. It is firm, and there is a whorl-like arrangement of the muscle and the fibrous tissue. In contrast to this typical appearance, the myometrium may be thickened by a diffuse, ill-defined nodularity of smooth muscle. This so-called diffuse leiomyomatosis usually involves all parts of the myometrium and causes symmetric enlargement of the uterine corpus. The nodules of smooth muscle are not distinct, contain little collagen, and merge with one another and the surrounding hypertrophied myometrium. The extracellular matrix of leiomyomata is composed mostly of collagen but also contains proteoglycans and fibronectin. According to Fujita, myomata contain 50% more collagen than does normal myometrium, and the ratio of collagen type I to collagen type III is increased in myomata. Proteoglycans provide hydrated spaces between myoma cells. Fibronectin is a glycoprotein that mediates adhesion between myoma cells and extracellular matrix. The most common change in leiomyomata is hyaline degeneration. The cut surface of a hyalinized area is smooth and homogeneous and does not show the whorl-like arrangement of the rest of the leiomyoma. Almost all leiomyomata, except the smallest, have scattered areas of hyaline degeneration. Eventually these may become liquefied and form cystic cavities filled with clear liquid or gelatinous material (Fig. 30.3). Sometimes the P.759 cystic change is so great that the leiomyoma becomes a mere shell and is truly a cystic tumor. Softness of a tumor does not necessarily indicate cystic degeneration. Fleshy leiomyomata may be equally soft.

FIGURE 30.3. Multiple leiomyomata are present. A large subserous myoma has undergone partial cystic degeneration.

View Figure

Over time, with continued diminished blood supply and ischemic necrosis of tissue, calcium phosphates and carbonates are deposited in myomata. Their presence is evidence of a continuum of degenerative changes. The calcium may be deposited in varying amounts. If it is deposited at the periphery of the tumor, the leiomyoma may resemble a calcified cyst. Other calcified leiomyomata may show an irregular or diffuse distribution throughout with a honeycomb or mulberry appearance. When the degenerative change is advanced, the leiomyoma may become solidly calcified. Such calcified tumors have been called “wombstones.” Calcified leiomyomata are seen most often in elderly women, in black women, and in women who have pedunculated subserous tumors. They are easily seen radiographically (Fig. 30.4). FIGURE 30.4. An abdominal radiograph shows typical calcification in a leiomyoma.

View Figure

Leiomyomata may undergo changes as a result of infection. Submucous leiomyomata are most commonly infected when they protrude into the uterine cavity, or especially into the vagina (Fig. 30.5). The pedunculated submucous leiomyoma thins out the endometrium as it grows inward, and eventually the surface becomes ulcerated and infected (Fig. 30.6). An intramural leiomyoma in an involuting puerperal uterus can also become infected when endometritis is present. Microscopic abscesses can be found, and

gross abscesses occasionally occur, particularly if the leiomyoma descends as low as the cervical canal. Such infections are usually streptococcal and may be virulent. Bacteroides fragilis infections also occur. Parametritis, peritonitis, and even septicemia may result. FIGURE 30.5. A large submucous pedunculated myoma has dilated the cervix and is now located in the vagina. Its pedicle is attached inside the uterine cavity. Morcellation of the myoma performed transvaginally allows clamping and ligation of the pedicle.

View Figure

FIGURE 30.6. Pedunculated submucous myoma showing necrosis and ulceration.

View Figure

Necrosis of a leiomyoma is caused by interference with its blood supply. Occasionally a pedunculated subserous leiomyoma twists, and if an operation is not done immediately, infarction results. Necrosis sometimes occurs in the center of a large tumor simply as a result of poor circulation. Necrotic leiomyomata are dark and hemorrhagic in the interior. Eventually the tissue breaks down completely. So-called red or carneous degeneration is seen occasionally, especially in association with pregnancy. This condition is thought to result from poor P.760 circulation of blood through a rapidly growing tumor. Thrombosis and extravasation of blood into the myoma tissue are responsible for the reddish discoloration (Fig. 30.7).

FIGURE 30.7. Degenerating leiomyoma showing carneous discoloration caused by thrombosis and extravasation of blood into the myoma tissue. A Dalkon shield can be seen in the endometrial cavity. See color version of figure.

View Figure

A subserous and especially a subserous pedunculated myoma may gradually outgrow its blood supply (Fig. 30.8). To keep the myoma tissue from undergoing complete ischemic necrosis, the omentum becomes adherent to the peritoneal surface of a pedunculated subserous myoma and provides whatever blood supply is needed. Eventually, the pedicle may disappear or twist, and the myoma will become completely free from the uterus, wander in the upper abdomen, and receive its “parasitic” blood supply from the omentum and other sources. FIGURE 30.8. A subserous pedunculated myoma receives tenuous blood supply through its uterine pedicle. Such a myoma may wander in the upper abdomen and eventually receive its blood supply from other sources. It may also twist on its pedicle and undergo infarction.

View Figure

On occasion, fat occurs in leiomyomata as true fatty degeneration. The cut surface may have a yellowish discoloration. Infrequently, a deposit of true fat may form a fibrolipoma; however, the presence of fat in a leiomyoma is rare. Indeed, if fat is seen grossly or microscopically in a curettage specimen, one should not assume that it represents fatty degeneration of a leiomyoma. One should assume that the uterus has been perforated and that fragments of fat have been curetted from the mesentery or omentum. The most important, but rare, change in a leiomyoma is sarcomatous degeneration. There is much variation in the reported incidence of sarcoma in leiomyomata. The incidence given by Novak is 0.7%.

However, a review of 13,000 myomata by Montague and associates at the Johns Hopkins Hospital revealed 38 cases of malignant change, the incidence of sarcoma thus being 0.29%. Corscaden and Singh indicated by their study that the true incidence of sarcoma developing within uterine leiomyomata is no higher than 0.13% and is probably as low as 0.04%. It should be remembered that because most women with uterine leiomyomata do not undergo surgical removal, the true incidence of sarcoma in leiomyomata is probably much lower than 1 per 1,000 (0.1%). After hysterectomy in 1,429 patients with presumed benign leiomyomata, the histologic diagnosis of leiomyosarcoma was made in seven (0.49%), according to a study by Leibsohn and co-workers. There was no evidence of malignancy in the endometrial sampling of any of these seven patients, and the diagnosis was suspected intraoperatively in only three. Uterine weights ranged from 120 to 1,100 g. In a woman between 41 and 50 years of age with presumed symptomatic leiomyomata, there is a 1 in 112 chance of a leiomyosarcoma being present, according to these authors. This information has important implications in the consideration of conservative or delayed treatment for these women. Parker and associates found that the total incidence of uterine sarcomas (leiomyosarcoma, endometrial stromal sarcoma, and mixed mesodermal tumor) among patients operated on for presumed benign uterine leiomyoma is lower (0.23%) than the 0.49% reported by Leibsohn. P.761 The difficulty in defining the true incidence of sarcomatous change is understandable if one is familiar with the histology of leiomyomata. Abundantly cellular leiomyomata are relatively common, and at first glance they suggest sarcoma; however, they lack a significant number of mitotic figures, and patients from whom such tumors are removed all remain well. Misinterpretation of the histologic picture of this type of cellular leiomyoma undoubtedly accounts for the increased incidence of leiomyosarcoma reported by some. When cutting leiomyomata in the operating room, the surgeon finds that sarcomatous areas have a somewhat characteristic appearance, although the histologic diagnosis certainly cannot be made by gross examination. A sarcoma is likely to occur in a rather large leiomyoma and toward the center of the tumor, where the blood supply is poorest. Instead of being firm fibrous tissue that grates when scraped with a knife blade, the tissue is soft and homogeneous, and is described as resembling raw pork. Later, as necrosis of the malignant tissue occurs, it becomes more friable and hemorrhagic. It has been difficult to understand uterine leiomyosarcoma because pathologists do not agree on the criteria necessary for diagnosis. Some pathologists rely on the mitotic count. All tumors with less than five mitotic figures per 10 high-power fields are considered benign. All tumors with more than 10 mitotic figures per 10 high-power fields are called malignant. Those in between can be called “smooth muscle tumors of uncertain malignant potential.” Other pathologists believe the mitotic count may have some significance but choose to rely instead on the presence of nuclear hyperchromatism, nuclear pleomorphism, or giant cells and other bizarre cell forms to make the diagnosis. Corscaden and Singh believe that no combination of histologic features is reliable and that only smooth muscle tumors that metastasize or recur are definitely malignant. We believe that all of these features should be taken into consideration for diagnosis and prognosis. When the tumor is confined to the uterus, both mitotic grade and histologic grade are important in the diagnosis and prognosis. A poor prognosis is associated with high mitotic counts and extremely atypical and anaplastic cytologic features. Bell and colleagues at Stanford University Medical Center assessed a variety of histopathologic features of 213 problematic smooth muscle neoplasms for which there were at least 2 years of clinical follow-up data. From the wide variety of light-microscopic features assessed, the important predictors that emerged were mitotic index, the degree of cytologic atypia, and the presence or absence of coagulative tumor cell necrosis. Previously, the mitotic index was relied on exclusively to determine whether a uterine smooth muscle tumor was benign or malignant, but currently an approach is used that incorporates additional histopathologic features. A normal chromosome complement (46,XX) was observed by Meloni and co-workers in about 50% of leiomyoma cases. About 50% showed clonal abnormalities, such as those of chromosomes 1, 7, and 13, and t(12;14). Interstitial deletions of chromosome 7 were the ones most often involved, suggesting that this abnormality may be of primary importance in the cellular proliferation of leiomyomata. A relation between more aggressive histology and chromosomal abnormalities was also suggested. Tumors that show obvious evidence of blood vessel invasion or spread to contiguous organs are rarely cured. The extent of the disease at the time of initial diagnosis is of even greater significance. In other words, when the diagnosis is suspected for the first time by the pathologist when he examines routine sections from a uterine leiomyoma, the patient almost always survives. However, if the diagnosis is made preoperatively by the gynecologist or is suspected during the operative procedure because of invasion of

surrounding organs, the prognosis is grave. The management of leiomyosarcoma is discussed in Chapter 47. An unusual atypical smooth muscle tumor was first described in the stomach by Martin and associates in 1960. Variously called bizarre leiomyoma, leiomyoblastoma, clear-cell leiomyoma, and plexiform tumorlet, these atypical smooth tumors probably all belong together. The term epithelioid leiomyoma was adopted by the World Health Organization. Kurman and Norris have proposed that this term be used for all atypical leiomyomata. Histologically, the characteristic feature is the mixture of rounded polygonal cells and multinucleated giant cells present in epithelioid clear-cell and plexiform patterns. Clinically, in the uterus most of these tumors are benign. They may rarely exhibit malignant potential. Malignancy is difficult to predict from histologic criteria because some metastases occurred from tumors that demonstrated very few mitoses. Kurman and Norris have suggested, however, that epithelioid neoplasms having more than five mitotic figures per 10 high-power fields should be called epithelioid leiomyosarcomas and that the term epithelioid leiomyoma should be applied when there is a lower level of mitotic activity. Although combination therapy (surgery plus radiation therapy or chemotherapy) may not be indicated for a patient with an epithelioid leiomyoma, follow-up should be considered essential, as emphasized by Klunder and colleagues. An unusual benign form of leiomyomata uteri, intravenous leiomyomatosis, was first recognized at the turn of the century and has been reported sporadically since then. Before 1982, about 50 cases had been reported, according to Bahary and co-workers. Probably at least that many have been reported since then. Marshall and Morris presented the first detailed report of this entity in the American literature in 1959. The characteristic feature of this peculiar smooth muscle tumor is the extension of the polypoid intravascular projections into the veins of the parametrium and broad ligaments. Although there may be some difficulty in distinguishing such lesions from low-grade sarcoma, they are distinctly different histologically from stromatosis uteri because the intravenous plugs are mainly smooth muscle in origin. In 1966, Edwards and Peacock collected 32 cases of intravenous leiomyomatosis, including two cases of their P.762 own, and reviewed the clinical experience with this condition. In approximately 50% of the cases, the intravenous tumor was confined to the parametrium; in 75%, it extended no further than the veins of the broad ligament. The observations of Edwards and Peacock suggest that the severed intravenous extensions are probably incapable of independent parasitic existence and remain dormant after removal of the uterus. However, the cases presented by Bahary and associates tend to refute this idea. Total surgical excision of the tumor should be attempted for successful therapy. Some patients have survived for many years after incomplete resection of the tumor. A review of 14 cases of this rare uterine tumor from the file of the Armed Forces Institute of Pathology has been reported by Norris and Parmley. In this series, two of three patients with incomplete resection had a recurrence; the recurrent tumor was excised surgically, and the patients were alive and free of disease 5 and 11 years after operation. The authors concluded that this tumor behaves clinically like a benign neoplasm, although its wormlike extensions may involve uterine, vaginal, ovarian, and iliac veins. The uterine veins in the broad ligaments are the most common sites of extension. The mitotic index is quite low, with the most active lesions showing only one mitosis per 15 high-power fields. The material from the Armed Forces Institute of Pathology provides histologic evidence consistent with both theories of origin of intravenous leiomyomatosis, namely, that it may be the result of unusual vascular invasion from a leiomyoma or may arise de novo from the wall of veins within the myometrium. Extension of benign leiomyomatosis up the vena cava and into the right atrium has been reported in several cases, with a fatal outcome in some. Before 1994, approximately 27 cases of intravenous leiomyomatosis extending to the heart were reported. Several recent cases requiring open-heart surgery to remove the intracardiac tumor thrombosis have been successful and without recurrence. All reported cases occurred in women. Tierney and colleagues reported that substantial quantities of cytoplasmic estradiol and progesterone receptors were found in the right atrial tumor removed from a patient with intravenous leiomyomatosis. Their patient was treated with the antiestrogen tamoxifen because of residual tumor in the vena cava that could be estrogen dependent. Irey and Norris have presented evidence that female reproductive steroids can produce intimal proliferation of veins in predisposed persons. Interestingly, of the 30 patients with leiomyomata and leiomyosarcomas of the vena cava reviewed by Wray and Dawkins, 80% were female. Both intravenous leiomyomatosis and benign metastasizing leiomyoma have been reported to metastasize to the lung. As suggested by Banner and co-workers, by Horstmann and associates, and by Evans and colleagues, oophorectomy may be indicated in patients with these conditions, again because of the possibility that these tumors may be estrogen dependent or that

estrogens may have the ability to stimulate their development, whether in a uterine or extrauterine location and whether they appear to be endothelial or mesenchymal in origin. The possibility of metastases from a histologically benign uterine leiomyoma has been discussed by Idelson and Davids and by Clark and Weed. When such a case occurs, it is usually settled by finding a sarcomatous component in the leiomyoma or by finding evidence of intravenous leiomyomatosis. However, more than 12 cases have now been reported in which a benign uterine leiomyoma metastasized. Idelson and Davids' case showed metastases to the aortic lymph nodes. The patient reported by Cramer and associates had metastatic tumor to the omentum, ovary, periaortic lymph node, and lung. In each location, the histology and estrogen receptor content of the tumor resembled those of a benign leiomyoma. The recommended treatment consists of surgical removal with castration and little or no estrogen replacement. Leiomyomatosis peritonealis disseminata is sometimes confused with intravenous leiomyomatosis. However, only subperitoneal surfaces of the uterus and other pelvic and abdominal viscera are involved with leiomyomatosis peritonealis disseminata, and invasion of the lumen of blood vessels does not occur. Only about 15 cases have been reported, according to Pearce. All occurred in patients in the reproductive years who often had large uterine leiomyomata and were usually pregnant or taking oral contraceptives. The condition is likely to be confused with a disseminated intraabdominal malignancy, but it is entirely benign histologically and clinically. Parmley and colleagues have demonstrated the histologic similarities between this peritoneal lesion and the decidual change of the mesothelium in the pelvis, and they propose that the condition represents a benign reparative process in which fibroblasts replace soft peritoneal decidua. They suggest that this fibrocytic reaction occurs during pregnancy and especially in the postpartum period, resulting in nodules with a pseudoleiomyomatous pattern. Similar findings have been noted in patients with endometriosis treated with prolonged Enovid therapy. These findings indicate that prolonged and continuous stimulation of subperitoneal decidua by either endogenous or exogenous estrogen and progesterone is important in the pathogenesis of this condition. Parmley and co-workers suggest that the condition is more appropriately called disseminated fibrosing deciduosis. Goldberg and associates, on the other hand, on the basis of electron microscopy studies, believe that the tumors arise from smooth muscles of small blood vessels. This has been confirmed by Ceccacci and colleagues. It has been possible to show a continuum from fibroblastic cells through myofibroblasts to leiomyocytes. Although the cell of origin of this tumor is still controversial, the tumor is benign and the acceptable treatment to date is total abdominal hysterectomy and bilateral salpingo-oophorectomy. If this tumor occurs in the omentum, an omentectomy should also be performed to define more clearly the histologic nature of the lesion. In attempting to distinguish between benign and malignant disease in a patient with uterine leiomyomata who also has unusual clinical findings, it is appropriate to keep the entities mentioned earlier (intravenous leiomyomatosis, P.763 atypical bizarre leiomyoma, benign metastasizing leiomyoma, and disseminated intraperitoneal leiomyomatosis) in mind. Although they all have features similar to those of malignant disease, they are almost always benign and amenable to treatment. One should also remember that benign uterine leiomyomata have been associated with pseudo-Meigs' syndrome in a few cases. Meigs reported five cases in 1954. In these cases, the ascites did not reappear after removal of the uterine leiomyomata. There is a high frequency of endometrial hyperplasia when the uterus contains leiomyomata. Degligdish and Loewenthal reported that cystic glandular hyperplasia is often found in the endometrium at the margin of the leiomyoma. Yamamoto and co-workers have reported high concentrations of estrone and estrone sulfatase activity in the endometrium overlying a myoma. They suggest that the local hyperestrogenism in the endometrium overlying a leiomyoma may assist in the genesis or enlargement of these tumors. Gynecologic surgeons are especially concerned about the vascularity of individual leiomyomata and about the blood flow to the uterus in the presence of multiple and sometimes very large leiomyomata. These considerations are pertinent when surgery, especially myomectomy, is contemplated. According to Vollenhoven and associates, the vascularization of leiomyomata was studied by Vasserman and colleagues, and the findings were presented to the World Congress of Gynecology and Obstetrics in 1988. Using femoral arteriography, selective intraoperative angiography, radiography, and injection of surgical specimens, these investigators showed that leiomyomata have a rich vascular supply, including blood lakes within tumors. They found more than one nutrient vessel per myoma. Venous channels were predominantly peripheral, whereas the arterial supply was both internal and peripheral. Farrer-Brown and co-workers, using radiologic methods, demonstrated that myomata in various locations within the

myometrium can cause congestion and dilatation of endometrial venous plexuses by obstructing venous return. These obstructions can result in ectasia of endometrial and myometrial venules (Fig. 30.9). The degree of vascularity of leiomyomata was also studied by Karlsson and Persson. Vascularity varied from many, to few, to no intrinsic vessels demonstrable. Generally, the sum of the width of the uterine arteries increases with the size of the uterus, but the diameter of the two sides sometimes differs markedly. A rich vascularity was found in 22 of 34 uteri with leiomyomata, but with increasing size there is a tendency to less vascularity. In none of five cases with very large (20 cm or more) leiomyomata uteri P.764 was the vascularity rich. The intrinsic vessels were few in two cases and absent in three cases.

View Figure

FIGURE 30.9. Sonographic images of a 48-year-old patient with a history of pelvic pain. Both transabdominal and transvaginal images were obtained on cycle day 8. Longitudinal (A) and transversely oriented (B) transvaginal images reveal a suspicious endometrial cavity, suggesting a thickened lining, endometrial polyp, or a submucosal myoma. C and D: Fluid-enhanced sonohysterographic studies clearly demonstrate a mass lesion and not generalized thickening of the endometrial lining. The patient underwent a hysteroscopy and uterine curettage, which revealed a histologically confirmed endometrial polyp. (Images courtesy of Jeff Dicke, M.D.)

The total blood flow in a myomatous uterus is greater than the total blood flow in a normal uterus. Blood flow in a myomatous uterus, gram for gram, is less than blood flow in a normal uterus. Blood flow in myomata is reduced compared with blood flow in normal myometrium in the same uterus. Blood flow within a myoma varies from one location to another.

CLINICAL FEATURES OF UTERINE LEIOMYOMATA Part of "Chapter 30 - Leiomyomata Uteri and Myomectomy"

Asymptomatic Leiomyomata Most leiomyomata are asymptomatic. Untold numbers of such symptomless leiomyomata are removed surgically by either hysterectomy or myomectomy when they would have been better left undisturbed. The incidence of malignancy in leiomyomata is less than 0.1%, which is lower than the operative mortality rate of hysterectomy in the average hospital; therefore, unless there is some reason to suspect malignant change, the danger of the operation for asymptomatic leiomyomata may exceed the danger of malignancy. A history of rapid growth, however, particularly postmenopausal growth, does indicate removal, even when the tumor produces no symptoms. Signs of rapid enlargement are important in all patients but are even more ominous in older patients. In younger patients, the most common reason for rapid enlargement of a uterus with leiomyomata is pregnancy. If pregnancy can be ruled out, a leiomyosarcoma may be suspected but is rarely found. Small leiomyomata that are asymptomatic need only to be observed from time to time, with pelvic examinations perhaps every 6 to 12 months and pelvic ultrasonography when indicated. In the beginning, frequent examination may be indicated to determine the growth rate. Such tumors may remain remarkably constant in size for years. If small leiomyomata are discovered late in menstrual life, it is unusual for symptoms to appear or for surgical treatment to be required. Larger tumors can also be watched safely, but if a policy of watchful waiting is adopted, one should be very sure of the nature of the tumors. If there is uncertainty of the uterine or ovarian origin of a tumor, as may well be the case when the tumor fills the whole pelvis or when a pedunculated tumor is felt in the adnexal region (Fig. 30.10), special diagnostic procedures may be indicated. Pelvic examination by an experienced gynecologist can usually clear up the uncertainty. In difficult cases, an examination under anesthesia may be necessary. Laparoscopy may be of great value in determining the nature of an adnexal mass. Before invasive techniques are used, however, noninvasive special diagnostic procedures should be done. These include radiographic studies of the abdomen and pelvis, ultrasonography (US), and computed tomography (CT). The characteristic calcification in a leiomyoma may be seen on radiographs. The US and CT features of uterine leiomyomata have been well described. However, mistakes in the interpretation can still be made. Tada and associates reported that 5% of patients given the diagnosis of uterine leiomyomata by CT actually had an ovarian tumor at operation. Therefore, if uncertainty about the diagnosis persists, laparoscopy or laparotomy should still be performed. FIGURE 30.10. Although this central pelvic mass may feel like a multiple leiomyomatous uterus on bimanual pelvic examination, it is actually a bilateral ovarian malignancy. Differentiation between these two diagnoses may require special diagnostic procedures.

View Figure

When large asymptomatic leiomyomata occur in premenopausal women who have had their families or in whom future childbearing is not important, a recommendation for removal may be made. It is impossible to

predict which patients will become symptomatic in the remaining years before menopause. However, such tumors, with additional years to grow, are likely to require surgical removal eventually. Therefore, it is better to remove them when the patient is a good operative risk and when conservation of normal ovaries with a good blood supply can be easily accomplished. Such tumors should usually be 12 to 14 weeks in gestational size or larger. Depending on a variety of factors, either myomectomy or hysterectomy can be recommended to the patient. GnRH agonists may be useful in women approaching menopause to control symptoms or asymptomatic uterine myoma growth until menopause. The regrowth of tumors after the cessation of treatment limits the usefulness of these agents, however. Nakamura and Yoshimura reported their experience with GnRH agonists in the treatment of uterine leiomyomata in perimenopausal women. One third of patients reached menopause after 16 weeks of treatment, thus avoiding the need for surgery. Reiter and colleagues studied 93 consecutive patients undergoing hysterectomy for leiomyomata. When the uterus was larger than 12 weeks' gestational size, there was no increased incidence of surgical complications compared with women with smaller uteri. On the basis of this small series, the authors concluded that hysterectomy need not be recommended to women with large asymptomatic uterine leiomyomata to avoid a possible increased risk of surgical complications. P.765 There is no uniform size of an asymptomatic leiomyomatous uterus that can be used as an indication for hysterectomy or myomectomy. When size is the only significant indication for surgery in an asymptomatic patient, the location of the tumors is more important than the total uterine mass. When the leiomyomata are located in the cornual area or in the lateral wall of the uterus and obscure the anatomy of the adnexa and broad ligament, the risk of error in the early recognition of an ovarian tumor is greater. In such cases, one must carefully weigh the advantages and disadvantages of the conservative approach to the management of uterine leiomyomata. When adnexal tumors are present, it is critical that the origin of these tumors be confirmed. The diagnostic studies mentioned earlier should be performed to establish clearly that the tumors are of uterine origin before a decision is made to follow up the patient rather than operate. It is unacceptable to wait to see whether an adnexal tumor enlarges before identifying the site of origin of the mass as either uterine or ovarian. Ovarian carcinoma remains the most lethal disease of the female reproductive tract and the most difficult to diagnose early. Every diagnostic and therapeutic effort must be made to avoid errors in the clinical evaluation of pelvic neoplasms (Fig. 30.10). In women who are approaching menopause, relatively large uterine leiomyomata can be kept under observation with the knowledge that after menopause they will not increase in size and may actually regress somewhat. Still, one must be certain that the entire central pelvic mass is a leiomyomatous uterus. Patient management is largely dependent on knowledge of the exact location and size of leiomyomas. Imaging modalities play an important role in determining patient management, especially when differentiating a benign leiomyoma from other pathologic conditions that may require different therapies. Uterine size as an indication for surgical intervention in women with leiomyomata has been thoughtfully discussed by Friedman and Haas. These authors point out that many gynecologists advocate surgical removal of leiomyomata when the uterus reaches 12 weeks' gestational size or greater, regardless of the presence or absence of significant symptoms. The reasons given for surgical intervention include the following: The inability to accurately assess the ovaries by examination The possible malignancy of the pelvic mass The potential for compromise of adjacent organ function if the mass continues to enlarge The greater risk of surgical complications if the mass grows to a larger size The potential for better fertility if myomectomy is performed when the uterus is smaller The possibility of continued growth of uterine leiomyomata if hormone replacement therapy is given after menopause Friedman and Haas find very little in the literature to support these indications for surgical intervention and believe the availability of modern high-resolution US and magnetic resonance imaging (MRI) allows for expectant management in many patients with large asymptomatic uterine leiomyomata. They prefer to give primary consideration to the presence and severity of myoma-related symptoms in deciding whether surgical intervention is indicated. We believe that such a course of expectant management is appropriate only when there is certainty regarding the benignity of the central pelvic mass and all of its components, and when it is possible to get the patient to return for periodic assessment of gynecologic symptoms and findings on pelvic examination. Repeat MRI may also be required occasionally. If one elects to observe a patient with a relatively large asymptomatic uterine leiomyoma, it is a good rule to obtain an excretory urogram or renal ultrasound. Everett and Sturgis showed many years ago that

sometimes there is evidence of ureteral compression at the pelvic brim so that hydroureter and hydronephrosis develop (Fig. 30.11). It is usually the symmetrically enlarged uterus with intramural leiomyomata that extends near or above the umbilicus and rests on the pelvic brim that compresses the ureters, in the same way as a symmetrically enlarged gravid uterus. The process is usually slow and painless even when moderate to severe hydronephrosis has occurred. Pyelographic evidence of kidney damage may be the determining factor in a decision to operate on a patient with an entirely asymptomatic leiomyoma. The irregularly and asymmetrically enlarged uterus with subserous tumors usually does not produce pressure on the ureters. FIGURE 30.11. Bilateral ureteral obstruction and dilatation from pressure of large leiomyomata.

View Figure

P.766 After menopause, asymptomatic leiomyomata generally should be left undisturbed. Again, the gynecologist must be absolutely certain that an ovarian neoplasm can be ruled out. In the postmenopausal years, shrinkage of myomata and the myometrium occurs. However, the myometrial shrinkage may be disproportionately greater than the myoma shrinkage. Therefore, a myoma in an intramural location before menopause may become a submucous myoma after menopause and then become symptomatic for the first time, usually with postmenopausal bleeding. In menopausal women, the appearance of even the slightest trace of vaginal bleeding should make one suspect cervical or endometrial malignancy or the possibility of sarcomatous change in the leiomyoma or elsewhere in the uterus (Fig. 30.12 and Fig. 30.13). Careful pelvic examination, Papanicolaou smear, and evaluation of the cervix by colposcopy or biopsy, pelvic US, fractional curettage, and perhaps hysteroscopy should be done. If the bleeding remains unexplained and the presence of atrophic vaginitis or the use of exogenous estrogens has been excluded, the leiomyomatous uterus should be removed because of the danger of sarcomatous change or other significant problems.

FIGURE 30.12. Adenocarcinoma of the endometrium is present in a symmetrically enlarged leiomyomatous uterus.

View Figure

FIGURE 30.13. This specimen shows multiple uterine leiomyomata but also shows chronic pelvic inflammatory disease and cervical carcinoma. Patients with uterine leiomyomata may have abnormal bleeding, but a coexisting cervical carcinoma may also cause bleeding.

View Figure

Transabdominal and endovaginal US are the standard imaging modalities for the detection of leiomyomas. Pelvic US by transvaginal and transabdominal techniques is most useful because of good patient tolerance, relatively low cost, availability, and accuracy when performed by well-trained and experienced ultrasonographers (Fig. 30.14). Ultrasonography is the most cost-effective screening mechanism for uterine masses suggestive of myomas. Sonographic criteria for diagnosis have been well described. Generally, abdominal US is unable to detect myomas less than 2 cm in diameter. Endovaginal probes have allowed for improved visualization of both the uterus and adnexa. With higher frequencies, sensitivity in the detection of small myomas has substantially increased. In a series evaluated by Fedele and colleagues using endovaginal ultrasound before hysterectomy, submucous leiomyomas were identified with a sensitivity of 100%. Difficulties may arise, however, if myomas are small or pedunculated, patients are obese, or the uterus is retroverted.

FIGURE 30.14. An ultrasonography study shows a cervical myoma and a very early intrauterine pregnancy. See color version of figure.

View Figure

Transvaginal fluid-enhanced vaginal probe sonography (sonohysterography) is a useful technique to assess myomata that distort the endometrial cavity. This technique has little to no complications and is generally well tolerated with only mild cramping described by patients. The limitation of detection of leiomyomas with this modality is 0.5 cm diameter. In a study by Hoetzinger, the majority of intrauterine myomata (14 of 16, or 88%) P.767 were detected by sonohysterography. Ultrasound transducing catheters have been suggested as a potential tool to supplement abdominal and endovaginal sonography. Three-dimensional data display has recently undergone development and application in sonography (Fig. 30.9). There is no technique that reliably identifies a leiomyosarcoma. Plain abdominal or pelvic radiographs and hysterosalpingography are older, standard techniques that are still useful in assessing uterine size, calcification in myomata, intrauterine filling defects caused by submucous myomata, and tubal patency. These techniques, combined with US, are the most useful for assessing patients with a central pelvic mass thought to be a leiomyomatous uterus. CA-125 levels may be elevated in women with uterine leiomyomata, but the levels are generally lower than those in patients with ovarian cancer.

Symptomatic Leiomyomata Less than 50% of patients with uterine leiomyomata have symptoms. Symptoms may be single or multiple and depend on the location, size, and number of tumors present. A clinical and pathologic study of 298 patients with uterine leiomyomata by Persaud and Arjoon revealed no significant relation between the presenting symptoms and the presence of degenerative changes in the tumors. Some form of degeneration was demonstrated in 65% of the specimens, with hyaline degeneration accounting for 63% of all types of degeneration. Hyaline degeneration produces no characteristic symptoms. Symptoms, especially pain and fever, may be present in some patients with red degeneration of a leiomyoma during pregnancy, with torsion and infarction of a subserous pedunculated leiomyoma, or with an infected leiomyoma. A discussion of the signs and symptoms caused by uterine leiomyomata follows.

Abnormal Bleeding It is surprising but not unusual that even patients with large uterine leiomyomata may have a history of normal menstruation. Such patients should be questioned carefully about recent slight increases in the amount, duration, and frequency of menstruation. Some patients with a history of normal menstruation are found to have iron-deficiency anemia from a gradual increase in menstrual blood loss that even the patient has not recognized. If a case of uterine leiomyomata is to be followed, the patient should be asked to monitor her menstrual blood loss carefully and should be given instructions to keep a menstrual calendar and monthly record of the number of pads or tampons used each day. A more objective measurement of the amount of menstrual blood loss using the method of Hallberg and Nilsson may be helpful in doubtful cases. Iron depletion may not be evident by laboratory determination unless one performs an iron stain of the bone marrow or serum ferritin levels. In the early months of increased menstrual blood loss, the

hemoglobin and hematocrit values are normal. Heavy menstruation does not cause anemia until iron stores are first depleted. Abnormal bleeding occurs in about one third of patients with symptomatic uterine leiomyomata and commonly indicates that treatment is necessary. The menstrual flow is usually heavy (menorrhagia), but it can also be prolonged (metrorrhagia) or both heavy and prolonged (menometrorrhagia). Abnormal bleeding may be associated with submucous, intramural, and subserous tumors, but there is a distinct clinical impression that bleeding is more common and more severe in the presence of submucous tumors. The submucous leiomyoma bleeds freely at menstruation and may also bleed between periods as a result of passive congestion, necrosis, and ulceration of the endometrial surface over the tumor and ulceration of the contralateral uterine surface. If the submucous myoma is pedunculated, there is usually a constant, thin, blood-tinged discharge in addition to the menorrhagia. An intramural tumor that is just beginning to encroach on the uterine cavity can also be responsible for menorrhagia. Intramural leiomyomata near the serosal surface and pedunculated subserous tumors can also be associated with abnormal bleeding. When bleeding occurs with such tumors, however, one should search for some other lesion to account for it. The mere presence of leiomyomata in a woman who has abnormal uterine bleeding is not proof that the leiomyomata are causing the bleeding. This fact is important, particularly when there is intermenstrual bleeding. When a patient with leiomyomata has intermenstrual bleeding, it is a rule on our service to examine and study the cervix carefully with special diagnostic procedures and to sample and evaluate the uterine cavity before we proceed with treatment of the leiomyomata. If an endometrial or cervical malignancy is detected, the treatment of the leiomyomata will need to be altered. There are several mechanisms by which leiomyomata can cause abnormal bleeding, although a single specific mechanism may not be apparent in a particular patient. According to Sehgal and Haskins the surface area of the endometrial cavity in a normal uterus is 15 cm 2. The surface area of the endometrial cavity in the presence of leiomyomata may exceed 200 cm2 . These authors demonstrated a correlation between the severity of the bleeding and the area of endometrial surface. In addition to an increased surface area from which to bleed, the endometrium may demonstrate local hyperestrogenism in areas immediately adjacent to submucous tumors, and endometrial hyperplasia and endometrial polyps are commonly found. Deligdish and Loewenthal noted a broad spectrum of histologic abnormalities in the endometrium associated with leiomyomata, ranging from atrophy to hyperplasia. Thinning and ulceration of the endometrial surface may be present over large submucous tumors; smaller ones may show slight thinning without ulceration. The presence of leiomyomata may interfere with myometrial contractility as well as contractility of the spiral arterioles in the basalis portion of the endometrium. Miller and Ludovici suggested that anovulation and dysfunctional uterine bleeding are more common in the presence of uterine leiomyomata. Sampson in 1913 was the first to study the blood supply of uterine leiomyomata and its effect on uterine P.768 bleeding. More recent studies have been performed by Faulkner and by Farrer-Brown and associates. The most prominent and important change is the presence of endometrial venule ectasia. Tumors that are strategically located in the myometrium may cause obstruction and proximal congestion of veins in the myometrium and endometrium. Thrombosis and sloughing of these large dilated venous channels within the endometrium produce heavy bleeding (Fig. 30.15).

FIGURE 30.15. Dilated endometrial venous space communicating with a grossly enlarged vessel in the inner myometrium of a uterus with submucous leiomyomata. (From Farrer-Brown G, Beilby JO, Tarbit MH. Venous changes in the endometrium of myomatous uteri. Obstet Gynecol 1971;83:743.)

View Figure

Makarainen and Ylikorkala have presented evidence that further supports the concept that prostanoids play a role in primary menorrhagia. They found that the production of 6-keto-prostaglandin F 1 alpha (6-keto-PGF1?), a metabolite of prostacyclin (PGI 2 ), and thromboxane B2 (TXB2), a metabolite of thromboxane A2 (TXA2 ), was normal in menorrhagic endometrium. However, the balance between TXA2 and PGI2 shifted to a relative TXA2 deficiency and was negatively related to blood loss in patients with menorrhagia. Although ibuprofen decreased the blood loss in patients with primary menorrhagia, it failed to reduce myoma-associated menorrhagia. The authors suggest that uterine factors other than prostanoids are more important in causing menorrhagia associated with uterine leiomyomata. In most cases, when bleeding occurs postmenopausally and leiomyomata are discovered on bimanual examination, the bleeding is due to some other factor, such as cervical or endometrial abnormalities, atrophic vaginitis, or exogenous estrogen, and the leiomyomata are purely incidental. However, the postmenopausal leiomyoma can be responsible for the bleeding. As stated earlier, leiomyomata that do not bleed during the menstrual life of the patient have been found to migrate to a submucous position in later years. This occurs because after menopause the myometrium atrophies and the uterine wall becomes thinner. Leiomyomata also shrink somewhat, but not as much as the surrounding myometrium. Thus, a leiomyoma that was intramural before menopause may work itself into a submucous position after menopause, become ulcerated, and bleed. Postmenopausal growth of uterine leiomyomata may indicate malignant change, especially if associated with postmenopausal bleeding. We have rarely observed postmenopausal growth in a leiomyoma without finding malignancy in the tumor; whenever there is enlargement of the leiomyoma after menopause, one should seriously consider the possibility of sarcomatous change and remove the leiomyoma. Patients with heavy menstruation and uterine leiomyomata should be evaluated for the presence of submucous myomata. Even patients without palpable evidence of uterine leiomyomata or uterine enlargement who have heavy menstruation should be evaluated for the presence of submucous myomata. When endometrial curettage is performed, irregularity of the uterine cavity may suggest the presence of a submucous myoma. However, a submucous myoma may not be detected with the curette. An accurate diagnosis is more likely to be made by hysterosalpingography, conventional transvaginal or transabdominal US, sonohysterography, MRI, or hysteroscopy. Cincinelli and colleagues reported their experience with transabdominal sonohysterography, a technique that involves transabdominal ultrasonographic scanning while 30 mL of sterile isotonic saline is slowly injected into the uterine cavity. According to these investigators, this technique provided the most accurate evaluation of the size of submucous myomata, intracavitary and intramural growth, and location within the uterine cavity, with sensitivity, specificity, and predictive values of 100%.

Pressure

Evidence of pressure on nearby pelvic viscera may be an indication for treatment. The urinary bladder suffers most often from such pressure, giving rise to urgency and frequency of urination and sometimes even urinary incontinence (Fig. 30.16). Although this symptom is common with large leiomyomata, one frequently finds the pelvis filled with leiomyomata when there is no urinary frequency. Occasionally, acute retention of urine or overflow incontinence results from a leiomyoma and necessitates surgical intervention. These effects can occur as a result of rapid interior growth of the leiomyoma with compression of the urethra and bladder neck against the pubic bone. More often, a tumor the size of a 3-month pregnancy may become incarcerated in the cul-de-sac, wedging the cervix forward against the urethra and obstructing the flow of urine through the urethra. A large pedunculated submucous tumor may fill and distend the vagina and press the urethra against the symphysis, causing urinary retention. FIGURE 30.16. Cystogram and intravenous pyelogram showing distortion of the bladder by pressure from a leiomyoma.

View Figure

P.769 As pointed out by Mattingly, one can expect to encounter women who have uterine leiomyomata of significant size and in addition have protrusion of the bladder base and posterior urethra through a widened levator muscle hiatus and a weakened urogenital diaphragm. Both conditions are relatively common. In addition to the usual symptoms produced by the leiomyomata, socially disabling stress urinary incontinence may be present. When the anterior wall of the uterus is greatly distorted by the presence of these tumors, pressure against the bladder can cause urinary frequency. If anatomic pressure equalization incontinence is also present, it may be aggravated by the increased intravesical pressure caused by the leiomyomata. However, the presence of anatomic stress urinary incontinence has no etiologic relation to the uterine enlargement caused by the leiomyomata. Silent ureteral obstruction from pressure against the pelvic brim is an uncommon complication of uterine enlargement caused by multiple large leiomyomata. Such an asymptomatic anatomic change occurs more often with a symmetrically enlarged leiomyomatous uterus that becomes large enough to fill the pelvis and compress the ureter against the pelvic side walls (Fig. 30.11). Although an infrequent complication, the obstruction can occur in either ureter, depending on the location of the uterine tumors. If there has been no infection or parenchymal damage to the kidney, this anatomic alteration is completely reversible with removal of the uterus and relief of the pressure against the ureter. However, if urinary tract obstruction from leiomyomata has been neglected, uremia may result. Removal of the tumor and relief of obstruction are necessary to restore kidney function. Chronic bladder neck obstruction from uterine leiomyomata can be so severe as to cause a remarkable increase in the thickness of the bladder wall and enlargement of the bladder resembling that seen in men with urethral obstruction from prostatic enlargement. Indeed, in these neglected cases, the bladder may fill the entire lower abdominal wall so that an incision above the umbilicus is required to enter the peritoneal cavity to remove the tumor without injury to the bladder. The bowel is less apt to show symptoms from pressure than is the bladder, but constipation can be caused and aggravated by pressure of leiomyomata against the rectum. The small intestines can become

entwined with subserous pedunculated tumors, causing intermittent intestinal obstruction.

Pain Abdominal and pelvic pain or discomfort, a feeling of heaviness in the pelvis, and dyspareunia are present in about one third of patients with symptomatic uterine leiomyomata and may be an appropriate reason for operative intervention. There are several causes of pain with leiomyomata. However, the usual hyaline or cystic degeneration of these tumors does not produce symptoms. In rare instances, pedunculated subserous leiomyomata twist and give rise to a clinical picture of acute abdominal pain, much like that seen with a twisted ovarian tumor. These pedunculated tumors twist more often during pregnancy and after menopause. Acute carneous or red degeneration of a leiomyoma can occur at any period of reproductive life, although pain from this form of degeneration is more common during pregnancy. Dysmenorrhea, acquired in the fourth or fifth decade, may be the outstanding symptom of the growth of leiomyomata. A common symptom complex resulting from leiomyomata at this time of life is menstrual pain coupled with increased menstrual flow. Diffuse adenomyosis can also cause these symptoms, and the differentiation of this condition from a symmetrically enlarged intramural leiomyoma may be extremely difficult and may require MRI. The differentiation is purely academic, for in either case surgery is indicated if the symptoms are of sufficient severity. Patients who have uterine leiomyomata and pain may have concomitant pelvic disease such as ovarian pathology, pelvic inflammatory disease, tubal pregnancy, endometriosis, or urinary tract or intestinal pathology, including appendicitis. One must be careful to rule out other pathology that may be obscured by uterine leiomyomata.

Abdominal Distortion Distortion of the normal abdominal wall contour due to large tumors may justify their removal. Tumors of such P.770 size often give rise to other symptoms also, so there is ample reason for surgical interference. However, when no other symptoms are present, one may recommend removal of the tumors if the abdominal distortion is of such a magnitude as to be embarrassing to the patient.

Rapid Growth Evidence of rapid growth of uterine leiomyomata, as observed by the same examiner over time or as confirmed by US, is an indication for surgical intervention. Such rapid growth in a premenopausal patient is only rarely due to sarcoma. Parker and others reviewed the medical records of 1,332 women admitted for surgical management of uterine leiomyoma. They actually found no correlation between rapid growth and the presence of uterine sarcoma. It may be due to pregnancy or to the use of oral contraceptives containing large amounts of estrogens. In the latter case, these drugs should be discontinued and an alternative method of contraception prescribed. In the postmenopausal patient, however, growth of a uterine leiomyoma is highly suggestive of a malignancy. The malignancy may be a sarcomatous change in the leiomyoma itself, a sarcoma or carcinoma in the endometrium causing uterine enlargement, or an ovarian neoplasm whose estrogen secretion is stimulating enlargement of the leiomyoma or whose growth may be mistaken for rapid enlargement of uterine leiomyomata. Although malignancy is not invariably found, the chances in its favor are so great that one must proceed on the assumption that it exists and must perform dilatation and curettage followed by removal of the enlarged uterus. Rapid growth of a leiomyomatous uterus is difficult to define in exact terms. Buttram and Reiter have arbitrarily defined it as a gain of 6 weeks or more in gestational size within a year or less. Although this definition could apply in premenopausal women, it might be disastrous to wait for this amount of growth in a postmenopausal woman. It is important to have a definite method of documenting uterine size at periodic intervals. Repeated sounding of the uterine cavity may be of some benefit, although leiomyomatous growth is not always accompanied by concomitant enlargement of the uterine cavity. It is important to document the size of specific leiomyomata or the total uterine size in terms of centimeters or grams of uterine weight rather than in terms of gestational size of the uterus, although the latter method has become quite popular. Changes in a patient's weight can make evaluation of growth more difficult. A uterine leiomyoma can erroneously appear to be growing in a patient who is dieting and losing weight when actually the tumor is just felt more easily. Conversely, in a patient who is gaining weight rapidly, the tumor will be more difficult to feel and may appear to be getting smaller. Ultrasonography is a much more objective way of establishing the size of a uterine leiomyoma in the beginning and, when indicated, of

evaluating its rate of growth. There is a need for more information about the natural growth patterns of myomata before and after menopause. Although leiomyomata can increase dramatically in size during pregnancy, usually there is no appreciable growth. Winer-Muram and co-workers studied 89 pregnant women with uterine leiomyomata documented by US examination. In 83 of the patients, there was no demonstrable increase in the size of the leiomyomata. In six patients there was an increase in size of up to 4 cm. Those myomata that increase in size during pregnancy will decrease in size a few weeks after the pregnancy is over.

Spontaneous Abortion and Other Pregnancy-Related Problems Uterine leiomyomata are associated with a significantly increased risk of spontaneous abortion. In a collected series of patients undergoing myomectomy, Buttram and Reiter reported that 41% had spontaneous abortions. This rate was reduced to 19% after myomectomy. Various mechanisms have been proposed to explain the occurrence of spontaneous abortion from uteri with leiomyomata. These include disturbances in uterine blood flow, alterations in blood supply to the endometrium, uterine irritability, rapid growth or degeneration of leiomyomata during pregnancy, difficulty in enlargement of the uterine cavity to accommodate for the growth of the fetus and placenta, and interference with proper implantation and placental growth by poorly developed endometrium or by subjacent leiomyomata. Implantation in a thin, poorly vascularized endometrium over a submucous leiomyoma is doomed to failure, because proper growth and development of the embryo and placenta are impossible (Fig. 30.17). Matsunaga and Shiota found a twofold increase in the number of malformed embryos recovered from patients with uterine leiomyomata having artificial termination of pregnancy. Uterine leiomyomata may be associated with premature delivery, stillbirth, and interstitial pregnancy, as in the case reported by Starks, although we are unaware of good statistics regarding these associations. Muram and associates have followed patients with leiomyomata through pregnancy with US. When a leiomyoma was in close proximity to the placental site, an increased incidence of pregnancy-related complications was seen. These were mainly bleeding complications, but pain, premature delivery, and postpartum hemorrhage also occurred. Exacoustos and Rosati reviewed the US scans of 12,708 pregnant patients. Four hundred ninety-two patients had myomata. A statistically significant increased incidence of threatened abortion, threatened preterm delivery, abruptio placentae, and pelvic pain was observed in patients with myomata. Abruptio placentae was particularly evident in women with myoma volumes greater than 200 cm 3, submucosal location, or superimposition of the placenta. The authors suggest that US findings make it possible to identify women at risk for myoma-related complications of pregnancy. Factors responsible for spontaneous abortion in patients without uterine leiomyomata may also be responsible for spontaneous abortion in patients with leiomyomata. FIGURE 30.17. When the placenta is implanted over a myoma in the uterine wall, the blood supply to the fetus may be tenuous.

View Figure

Occasionally, pregnancy causes a remarkable growth of leiomyomata in the same way that the myometrium P.771 undergoes hypertrophy in pregnancy. Red or carneous degeneration of leiomyomata during pregnancy is

associated with pain, tenderness over the tumor, low-grade fever, and leukocytosis. Management should be expectant with analgesic medications and bed rest. If premature uterine contractions occur, B-mimetics can be given. Pain usually subsides within a few days. Operation is not indicated unless it is necessary to rule out other problems that require surgery for relief, because differentiation from appendicitis, placental abruption, twisted ovarian cyst, and other problems may be difficult. After delivery, leiomyomata involute and generally return to their prepregnancy size by the third postpartum month. Torsion with infarction of subserous pedunculated leiomyomata is more common in pregnancy. A leiomyoma may interfere with labor and delivery by causing an abnormal presentation, by causing dysfunctional labor, or by obstructing the pelvis. A submucous leiomyoma in the lower uterine segment may entrap the placenta, necessitating manual removal. Indeed, furious postpartum hemorrhage can result if a submucous leiomyoma is disturbed at delivery or during exploration of the uterine cavity. Immediate hysterectomy may be necessary to control the bleeding. Most patients with uterine leiomyomata have no difficulty conceiving and carry their pregnancies to term without complications. The only problem encountered may be a difficulty in estimating gestational age from uterine size because of the presence of leiomyomata.

Infertility When asymptomatic leiomyomata are discovered in young women, the question of how these tumors relate to sterility and pregnancy usually arises. A number of factors may be responsible for infertility in a patient with uterine leiomyomata. Anovulatory cycles may occur more commonly. There may be interference with sperm transport caused by distortion and an increased surface area within the uterine cavity, impingement of leiomyomata on the endocervical canal or interstitial portion of the fallopian tube, or interference with prostaglandin-induced uterine contractions, which are thought to enhance sperm migration. Endometrial changes (atrophy, ulceration, focal hyperplasia, and polyps), vascular alterations (venous congestion, venule ectasia, impaired blood flow), and enlargement of the uterine cavity may be present. Because uterine leiomyomata occur in later reproductive years, relatively greater difficulty accomplishing conception can be expected in older couples. The finding of small leiomyomata in sterile women is not an indication for immediate myomectomy. Quite often, an infertile patient with uterine leiomyomata is found to have some other cause of infertility. Tubal inflammatory disease with associated pelvic adhesions is especially common in patients with uterine leiomyomata. Both marital partners should have a complete infertility investigation, and the leiomyomata should be disregarded for a while. The ultimate decision regarding disposal of the tumors depends on their size and location. Usually, small subserous leiomyomata are not considered a factor in infertility. Even if the woman fails to become pregnant, removal of small subserous leiomyomata is not justified. When leiomyomata are intramural or submucous and of significant size, they may well be factors causing the infertility, and a myomectomy may be rewarded with a subsequent pregnancy. When an unsuspected asymptomatic leiomyomatous uterus of significant size is found in a woman who is planning to become pregnant in the future, great tact is required in describing the problem to the patient. The best surgical and obstetric judgment is needed to make a proper recommendation. Should the patient be discouraged from attempting pregnancy because the risk of complications may be increased? Should a myomectomy be advised before pregnancy is attempted, with the knowledge that postmyomectomy adhesions may cause P.772 infertility? Such questions cannot be answered in a stereotypical manner. Each case presents its own problems, and the answers depend on the patient's age, her general physical health, her pelvic findings, and, most important, her own desires. All must be considered before a final recommendation can be made. In general terms, under these circumstances, an attempt to become pregnant will be rewarded with a satisfactory outcome in most cases. If pregnancy does not occur or is not successful, a myomectomy may be advised, but one must keep in mind that all causes of infertility, spontaneous abortion, and other pregnancy-related problems must also be investigated in patients with uterine leiomyomata; uterine leiomyomata represent an infrequent cause of infertility. Eldar-Geva and colleagues performed a retrospective review of the treatment outcome of 106 assisted reproductive technology cycles in 88 patients with uterine myomata (subserosal, intramuscular without cavity distortion, and submucosal). Patients underwent controlled ovarian hyperstimulation and advanced reproductive technology. Not surprisingly, pregnancy (30.1%) and implantation (15.7%) rates were significantly lower in women with submucosal myomas; however, both pregnancy (16.4%) and implantation (6.4%) rates were also significantly lower in women with intramural myomas. In some

advanced assisted reproductive technology patients, this information may influence the decision for surgical intervention regardless of menstrual pattern. A review of information about infertility and uterine leiomyomata was published by Wallach and Vu, by Vercellini and colleagues, and by Verkauf.

Miscellaneous Signs and Symptoms A variety of other unusual problems may be associated with uterine leiomyomata and may require treatment. Ascites and uterine inversion have already been mentioned. Sudden intraperitoneal hemorrhage can result from rupture of a dilated vein beneath the serosal surface of a subserous leiomyoma. Although leiomyomata are more often associated with iron-deficiency anemia from chronic uterine blood loss, occasionally patients present with polycythemia. Islands of extramedullary erythropoiesis have been found in leiomyomata. Arteriovenous shunts within the tumors have been found and may be etiologically important in polycythemia. If the tumor obstructs the ureters and causes back pressure on the renal parenchyma, erythropoiesis can be stimulated. Weiss and co-workers and other investigators have found marked erythropoietin activity within uterine leiomyomata. The polycythemia in these cases is cured by hysterectomy.

CHOICE OF TREATMENT FOR UTERINE LEIOMYOMATA Part of "Chapter 30 - Leiomyomata Uteri and Myomectomy" Six hundred fifty thousand hysterectomies are performed annually in the United States. About 175,000 are performed with uterine leiomyomata as the primary indication. Nearly 17,000 myomectomies are performed each year, and it is believed that this number is increasing substantially. There are no statistics to indicate the number of hysteroscopic and laparoscopic myomectomies performed each year. Effective medical therapies are available to use as adjuncts to surgical treatment. Abnormal bleeding from a leiomyomatous uterus can be controlled by ovarian irradiation if the patient is not a suitable surgical candidate. However, surgery is the preferred method of therapy for many reasons, especially because there are so few patients whose medical condition cannot be improved sufficiently to allow surgery. Hysterectomy (abdominal, vaginal, and laparoscopy assisted) is discussed in Chapter 31. In this chapter, surgical techniques that allow conservation of uterine function are discussed, as are medical therapies that can be used as adjuncts to surgical therapy.

Medical Management of Uterine Leiomyomata Most (70% to 80%) uterine leiomyomata are asymptomatic and are discovered initially during a routine pelvic examination. Such patients require an explanation and reassurance and reexamination at periodic intervals. An initial baseline pelvic US examination or MRI study may be indicated for comparison with future examinations and to evaluate the adnexa if the ovaries cannot be felt on pelvic examination. An experienced pelvic examiner can be fairly certain that a central pelvic mass is a leiomyomatous uterus. However, pelvic US examinations and repeat pelvic examinations can add to the certainty of the diagnosis. If the diagnosis remains doubtful, however, visualization of the mass, usually by laparoscopy, is indicated. Patients with an asymptomatic central pelvic mass should be followed up with periodic pelvic examination only when the mass is benign, usually a leiomyomatous uterus. Otherwise, expectant management is not appropriate. Effective medical treatment that is likely to result in the permanent cure of uterine leiomyomata is not yet available. Surgical excision by a variety of techniques remains the most effective and widely used method of management for patients with significant symptoms. Medical therapies are available as an adjunct to surgical treatment or as a temporary substitute for definitive surgical treatment. Hormonal therapy for the management of uterine leiomyomata has been the subject of investigation for many years. There is no support for the use of danazol or progestins in view of the disappointing results reported. Antiprogestin therapy with mifepristone (RU486) for 3 months has been shown by Murphy and colleagues to decrease leiomyoma volume by an average of 49%, with a variation of 0% to 87%. The immunoreactivity of progesterone but not estrogen receptors in the myoma and myometrial tissue was decreased significantly by RU486 treatment, suggesting that regression of these tumors may be attained through a direct antiprogesterone effect. All patients became amenorrheic. Side effects were mild, and bone density was not diminished. An effective dose P.773 to cause a clinically significant (50%) decrease in leiomyoma volume appears to be 25 mg daily. Additional experience is needed to further evaluate these promising results. Reinsch and associates have demonstrated that RU486 and leuprolide acetate are both effective in decreasing blood flow to the uterus. It is suggested that a decrease in uterine artery blood flow may provide a mechanism for a decrease in uterine size. Gestrinone, a synthetic derivative of ethinyl-nor-testosterone with antiestrogen and antiprogesterone properties, has been shown by Coutinho and associates to induce regression of leiomyomata. The treatment lasted 6 months to 1 year. The best results were obtained when the drug was administered intravaginally. Even the regression of large leiomyomata lasted up to a year after treatment. Side effects, though mildly androgenic, were well tolerated. Many studies have been performed to investigate the treatment of patients with uterine leiomyomata with GnRH analogs. GnRH analogs bind to GnRH receptors, resulting in a biphasic response: a temporary increase in the levels of gonadotropins and gonadal steroids (agonist phase) is followed by chronic suppression of gonadotropin and gonadal steroid secretion (desensitization phase). In 1 to 3 weeks, a profound hypogonadotropic hypogonadal state begins and exists as long as the treatment lasts, but it is

promptly reversed when treatment is discontinued. GnRH agonist treatment results in “medical oophorectomy” and “medical menopause” and is associated with the usual symptoms of a profound hypogonadal state (e.g., hot flashes, insomnia, mood lability, headaches, vaginal dryness, arthralgias, and myalgias). According to Friedman and associates, these adverse effects of treatment are self-limited and disappear within 3 to 6 months of cessation of GnRH agonist treatment. Dawood and colleagues describe a significant reduction in trabecular bone density after 24 weeks of GnRH agonist treatment that may not be completely reversible when treatment is discontinued. A mean reduction in trabecular bone density of 1% per month occurs in women treated for 6 months. Some of this bone loss may be permanent, but some is reversible. Friedman and colleagues state that the average reduction in uterine and myoma volume is 40% to 50% after 3 to 6 months of GnRH agonist treatment; this is generally confirmed by others. Most of the response occurs in the first 12 weeks, and it is variable and unpredictable. According to the analysis by these investigators, 4% of patients had an increase in uterine volume ranging from 0.1% to 25%; 24% had decreases in uterine volume ranging from 0.1% to 25%; 51% had decreases in uterine volume ranging from 25.1% to 50%; and 21% had decreases in uterine volume greater than 50%. No factors were found to predict the degree of uterine shrinkage. There were negative correlations with body weight, pretreatment uterine volume, age, height, and serum estradiol concentration. It is commonly thought that GnRH analogs affect leiomyomata by reducing vascularity and the individual cell size within the tumor. The biochemical changes in leiomyomata obtained from women treated with the GnRH agonist leuprolide acetate depot for 3 months were studied by Rein and co-workers. The concentrations of amino acids contained in collagen were significantly greater in uterine myomata from treated patients than in myomata from placebo-treated controls. These investigators suggest that the reduction in uterine myoma volume associated with GnRH agonist therapy is due primarily to alterations in the extracellular matrix rather than to a reduction in the number or volume of cells in the myoma. Because uterine leiomyomata are hormone-sensitive neoplasms that can be stimulated to grow by estrogen, some clinicians have been reluctant to prescribe oral contraceptive pills in patients with leiomyomata. However, Friedman and Thomas and others have demonstrated conclusively that oral contraceptives containing 30 to 35 µg of ethinyl estradiol do not cause uterine leiomyomata to increase in size. Therefore, low-dose contraceptives can be used to manage menorrhagia in patients with uterine leiomyomata. Friedman and Thomas demonstrated a significant decrease in the mean duration of menstrual flow and a significant increase in hematocrit values in response to low-dose oral contraceptives in patients with uterine leiomyomata. When myoma-associated menorrhagia is more severe, GnRH agonist and iron treatment may be more effective than oral contraceptives. In about two thirds of patients, GnRH agonist treatment induces amenorrhea. Most of the remaining patients experience very light, irregular vaginal bleeding or spotting, according to Friedman. A combination of menstrual suppression and iron therapy allows correction of iron deficiency and iron-deficiency anemia during a 6-month treatment period. Ovulatory menses resume 3 to 24 weeks after the last depot GnRH agonist injection. Stovall and associates reported that a GnRH agonist plus iron was more effective than iron alone in treating anemia in patients with leiomyomata and in alleviating menorrhagia. With such effective treatment now available, there is rarely a need to use blood transfusions to correct anemia caused by myoma-associated menorrhagia. Only patients with significant symptoms from severe anemia may require transfusion. Medical therapy may also be used transiently before surgery. By initiating the medication preoperatively, the maximum decrease in myoma volume may play a role in determining the route of surgery. If a hysterectomy is planned, the pharmacologic effect may facilitate a vaginal hysterectomy when the uterus is of borderline size. Vercellini and colleagues performed a multicenter, prospective, randomized, controlled study to assess if this shrinkage may increase the likelihood of a vaginal procedure. One hundred and twenty-seven premenopausal women with uterine volumes of 12 to 16 weeks were enrolled. After examination and disposition for an abdominal or vaginal hysterectomy, patients were randomized for GnRH therapy. Clinical assessment after the treatment course showed that abdominal hysterectomy P.774 was no longer indicated in 25 of 53 (47%) patients. No appreciable difference was found between the groups in postoperative complications. These findings are consistent with previously published studies, as well. GnRH agonist treatment alone cannot be given for periods longer than 6 months. A prolonged hypoestrogenic state is undesirable for a number of reasons, the most important being the loss of trabecular bone. If there are circumstances that require that GnRH treatment be extended beyond 6

months, consideration should be given to adding low-dose steroids after 3 months of GnRH therapy. The usual postmenopausal estrogen-progestin replacement regimen can be prescribed without interfering with the reduction in uterine volume anticipated. Loss of trabecular bone may not be as great. By adding estrogen-progestin replacement to GnRH agonist therapy, the adverse effects of a prolonged hypoestrogenic state may be prevented, and treatment with GnRH agonists may be prolonged. Friedman and colleagues treated 51 premenopausal women with large, symptomatic myomata with leuprolide acetate depot for 104 weeks. After the first 12 weeks, 0.75 mg of estropiptate plus 0.7 mg of norethindrone were added on days 1 through 14 each month. Menorrhagia and other symptoms of uterine leiomyomata were controlled successfully. Hemoglobin and hematocrit levels increased. Symptoms of hypoestrogenism (hot flashes, vaginal dryness) were decreased significantly. Bone density decreased in the first 12 weeks, but only a small additional decrease occurred between weeks 12 and 52. The use of GnRH analogs in the medical management of uterine leiomyomata is an emerging issue. How valuable it will be remains to be seen. Additional information and experience will define its use more exactly. For example, it may be possible for patients with symptomatic uterine leiomyomata who are approach-ing menopause to be managed medically through menopause without having a hysterectomy. It may be possible to improve fertility in some patients with uterine leiomyomata by treatment with GnRH analogs without myomectomy. With additional data, these and other questions can be answered.

Vaginal Myomectomy In 1845, Atlee performed the first successful vaginal myomectomy on a patient with a submucous pedunculated myoma. When a submucous myoma becomes pedunculated within the uterine cavity, there is a natural tendency for the uterus to try to expel it through the endocervical canal. Eventually, the cervix dilates. Even very large submucous pedunculated myomata can be delivered gradually through a markedly dilated cervix. Because adequate blood circulation through a long pedicle is difficult to maintain, the myoma becomes necrotic and infected (Fig. 30.5 and Fig. 30.6). Patients complain of cramping lower abdominal pain; pressure and heaviness in the pelvis; a thin, bloody, foul discharge; difficulty with urination; and other symptoms. Episodes of profuse vaginal hemorrhage can occur. Such large submucous myomata may resemble a fetal head. After satisfactory preoperative preparation, including broad-spectrum antibiotics and correction of anemia, vaginal myomectomy should be performed in the operating room. Morcellation may be required to remove very large tumors in many small pieces. Usually there is very little bleeding. One should avoid too much downward traction on the tumor because the uterine fundus may invert. Eventually, the pedicle is identified. It should be clamped and ligated as high as possible within the uterine cavity. If ligation of the pedicle is not possible, the clamps can be left in place and safely removed 48 hours later. Smaller submucous pedunculated myomata can be diagnosed by hysteroscopy, by hysterosalpingography, by US, or at the time of dilation and curettage. They can also be felt on digital exploration through a slightly dilated external cervical os. If the myoma can be grasped with an instrument (ring forceps, Allis clamp, and so forth), it can be removed by twisting it free of its attachment. A tonsil snare can also be used. Bleeding is usually minimal. If brisk bleeding does occur, a 26-French, 30-mL Foley catheter can be inserted through the cervix and inflated for tamponade. If necessary, the cervix can be sutured around the catheter to hold it in place. To gain access to submucous pedunculated myomata that are higher in the endocervical canal or uterine cavity, special procedures are required. An attempt at hysteroscopic removal may be successful. Alternatively, the cervix can be dilated with instruments or with Laminaria japonica, as described by Goldrath. Dührssen incisions can be made in the cervix. Also, the cervix can be incised by having the surgeon perform a vaginal hysterotomy (Fig. 30.18A and Fig. 30.18B). After the bladder is advanced, the cervix is dilated and an anterior midline incision is made in the cervix high enough to identify the myoma. The pedicle of the myoma is ligated as high as possible (Fig. 30.18C and Fig. 30.18D). The incision in the cervix is repaired with 2-0 interrupted delayed-absorbable sutures. The vaginal mucosa is reapproximated with 3-0 delayed-absorbable sutures (Fig. 30.18E and Fig. 30.18F).

FIGURE 30.18. Transvaginal removal of a pedunculated submucous myoma that presents itself at the external cervical os. A: Sagittal view of uterus, demonstrating the location of the myoma originating on the posterior wall of the fundus just above the cervix. B: A transverse incision made anteriorly through the vaginal mucosa at the cervicovaginal junction. C: After the bladder is advanced bluntly, the cervix is incised anteriorly in the midline. D: The myoma and its pedicle are exposed, and the pedicle is suture ligated for hemostasis. E: After the myoma is excised, the cervix is reapproximated with interrupted 0-0 absorbable, nonreactive sutures. F: The overlying vaginal mucosa is sutured with interrupted 3-0 absorbable sutures.

View Figure

A submucous pedunculated myoma may be solitary, and there may not be other myomata in the uterus. In fact, many patients exhibit no evidence of uterine enlargement. After a successful vaginal myomectomy, most patients are asymptomatic and menstruate normally. A few even become pregnant and deliver vaginally without difficulty. Cervical incompetence has been reported. Hysterectomy or myomectomy is required only for those few patients who have multiple leiomyomata and continue to have significant symptoms. Excellent results of vaginal removal of submucous pedunculated myomata in 151 patients were reported by Goldrath. Ben-Baruch and co-workers also achieved excellent results in 43 of 46 women in whom vaginal myomectomy was attempted. Vaginal myomectomy is recommended P.775 P.776 as the most appropriate initial treatment for pedunculated submucous myomata. Vaginal myomectomy is traditionally used for submucosal myomas; however, it has been described for other myomas. Davies and colleagues reported a prospective study regarding the safety and efficacy of excision of intramural and subserosal leiomyoma by a vaginal route. Preoperative criteria included (1) uterine size less than or equal to 16 weeks' gestation, (2) good uterine mobility, (3) adequate vaginal access, (4) the presence of intramural or subserosal myomas, and (5) the absence of adnexal pathology. Essentially, an open abdominal myomectomy technique was performed through an anterior or posterior colpotomy. The uterus was manipulated to bring the myoma into the colpotomy. The management of 35 women was described. The mean number of myomas removed was 2.5 per patient with mean mass of 113.8 g. Three patients (8.6%) required conversion to a laparotomy. Neither mean blood loss nor length of hospital stay was improved. Additionally, four (11.4%) patients developed pelvic hematomas postoperatively. At this time this procedure does not seem to provide inherent benefits over an open abdominal myomectomy or a laparoscopic approach. With further review, better outcome data may demonstrate the advantages of this technique. The tissue removed at vaginal myomectomy must be submitted for pathologic examination to rule out malignancy.

Hysteroscopic Resection of Submucous Myomata Hysteroscopic resection of submucous myomata was first reported by Neuwirth and Amin in 1976 and was reported again by Neuwirth in 1978. A urologic resectoscope was used. In 1981, Goldrath and associates used “photocoagulation” of the endometrium with the neodymium:yttrium-aluminum-garnet (Nd:YAG) laser to treat patients with menorrhagia. Many subsequent reports by Derman and associates, Donnez and colleagues, Goldenberg and co-workers, Corson, Indman, Hallez, Baggish and associates, Wamsteker

and colleagues, and others have confirmed the advantages of hysteroscopic treatment of menorrhagia in women with and without submucous leiomyomata. The menorrhagia associated with submucous myomata can sometimes be managed with oral contraceptives as long as the bleeding is not too severe. A favorable response can also be expected with GnRH analogs, but the menorrhagia usually reappears when the treatment is discontinued. Friedman has reported three cases of severe menorrhagia with resultant anemia requiring transfusions in women with submucous leiomyomata treated with leuprolide acetate. Both oral contraceptives and GnRH analogs are counterproductive in women who are seeking relief from infertility. The uterine cavity can be curetted several times, but the benefit of this procedure is temporary at best. When hysteroscopic resection of submucous myomata is performed, menorrhagia can be controlled in more than 90% of patients. According to Indman, the mean number of pads used during the heaviest day of menses decreased from 17.8 before treatment to 6.8 after treatment in women undergoing myoma resection only, and from 21.4 to 1.7 pads per day in women whose treatment also included endometrial ablation. Dysmenorrhea was also reduced significantly. Forty-eight of 51 women (94%) with uterine leiomyomata who were seen with menorrhagia were able to avoid major gynecologic surgery for up to 5 years of follow-up. In the report of 156 patients by Derman and associates, 91.3% of patients did not require further surgery after 6 years of follow-up, and 83.9% did not require further surgery after 9 years of follow-up. Further review in recent papers supports Derman's results. Magos and colleagues performed a prospective observational study to identify factors that influence outcomes of hysteroscopic myo-mectomies by following up patients for almost 8 years. One hundred and twenty-two patients enrolled in the study, and results suggest that hysteroscopic myomectomy is successful in treating menstrual symptoms in four of five cases. In addition, statistical analysis demonstrated that outcome is significantly better when the uterus is only slightly enlarged and if the myoma is mainly submucous in nature. If endometrial ablation is performed with myoma resection, pregnancy is not likely to occur subsequently. Without simultaneous endometrial ablation, 21 patients became pregnant after myoma resection only, with 18 infants delivered in the series reported by Derman and colleagues. The pregnancy rates among women who wished to conceive varied between 47% and 66% in several reports. These rates are comparable to those reported for abdominal myomectomy. Donnez and co-workers used a biodegradable GnRH agonist (Zoladex Implant ICI) preoperatively in a series of 60 women with large submucous myomata. Submucous myomectomy by hysteroscopy and Nd:YAG laser was easily performed. In 12 patients, the procedure was accomplished in two stages. Perino and associates used GnRH agonists in 58 women with submucous leiomyomata diagnosed during investigation for infertility or menstrual disorder. There was a significant reduction in operating time, intraoperative bleeding, infusion volume, and failure rate in the treated group compared with the control subjects. Myoma size is reduced and hemoglobin concentration is restored to normal preoperatively. When submucous myomata extend deeply into the myometrium, it may not be possible to perform a complete resection for obvious technical reasons. However, it should be possible to remove most irregularities in the uterine cavity and to restore the contour of the cavity to almost normal in most cases. According to Wamsteker and colleagues, hysteroscopic resection of submucous myomata with more than 50% intramural extension should be performed only in selected cases. Repeat procedures may be needed in cases of initial incomplete resection. To avoid the possibility of inadvertent uterine perforation or to allow its prompt diagnosis if it occurs, hysteroscopic P.777 resection is usually performed under laparoscopic guidance. However, as reported by Sullivan and co-workers, laparoscopy may be insufficient to evaluate fully the possible sequelae of uterine preformation. Laparotomy may be necessary to assess the pelvic viscera fully. Letterie and Kramer were able to safely substitute intraoperative transabdominal ultrasonographic guidance for laparoscopy. In their opinion, operative hysteroscopy with intraoperative ultrasonographic guidance provide an accurate and precise method to monitor intrauterine surgery, and it can be used to enhance the performance of hysteroscopic myomectomy and endometrial resection. Intraoperative US guidance provided sufficient details of the relation between the hysteroscope and the myoma and uterine walls to gauge the depth of resection and prevent uterine perforation. The success and safety of the procedure depend on the experience and skill of the operator. During hysteroscopic resection, vascular spaces are opened in the endometrium and myometrium. Large volumes of fluid are instilled into the uterine cavity. Fluid balance must be monitored carefully by the surgeon and the anesthesiologist to avoid fluid overload.

All tissue must be submitted for pathologic examination. Among 92 patients undergoing hysteroscopic resection in the series reported by Corson and Brooks, two cases of leiomyosarcoma were diagnosed. Leiomyosarcoma is said to be more common in submucous leiomyomata than in intramural or subserous leiomyomata. As time passes after hysteroscopic resection of submucous myomata, the possibility of recurrent problems increases because of regrowth of myomata. However, this is no more likely to occur than it is with standard abdominal myomectomy. The experience of many investigators has demonstrated that hysteroscopic management of menorrhagia in patients with submucous leiomyomata is a reasonable alternative to classic surgical hysterectomy or myomectomy. The occasional psychological problems and complications of hysterectomy are avoided. An abdominal incision is avoided, there is less discomfort, the procedure can often be performed in an outpatient setting, and the patient can usually resume normal activity after a very brief recovery period. Hysteroscopic resection of a small submucous myoma is illustrated in Figure 30.19. Details of the indications, technique, complications, and results of hysteroscopic resection are also provided in Chapter 17. FIGURE 30.19. Hysteroscopic removal of a submucous myoma. A: After insertion of the resectoscope, the submucous myoma is removed by progressive shaving. The loop of the resectoscope is placed at the most distant portion of the myoma, and the current is applied as the resectoscope is drawn toward the surgeon. Pressure is exerted by the loop against the myoma with each stroke. B: A grasping forceps is used to twist off the remaining tissue once the size has been appreciably reduced.

View Figure

Laparoscopic Myomectomy When abdominal myomectomy is indicated, the laparoscopic approach can be offered as an alternative to the standard “open” abdominal myomectomy in selected patients. However, this procedure is appropriate in very few patients for several reasons. First, myomectomy is indicated in infertility patients only if there is significant distortion of the uterine wall or endometrial cavity or if there is obstruction or distortion of the fallopian tubes by myomata. Second, myomectomy is indicated in patients who wish to retain their uterus only if the myomata are large or are significantly symptomatic with menorrhagia. In both circumstances, the myomata are likely to be multiple and large, and laparoscopic myomectomy is rarely the most appropriate procedure for removal. There are limitations to laparoscopic myomectomy, and these are mostly technical. Myomata in certain locations P.778 are difficult to remove. When myomata are large or multiple, or both, operative time and blood loss may be unacceptable. When myomata are embedded deeply in the myometrium, proper repair of the uterine wall may be difficult or impossible, and uterine rupture may occur in a subsequent pregnancy. Retrieval of the resected myomata from the peritoneal cavity can also pose problems. Large myomata must be morcellated into smaller pieces for retrieval. Retrieval through the posterior vaginal fornix or though the abdominal wall requires separate additional incisions, which somewhat defeats the idea of a minimally invasive procedure. Only very skillful laparoscopists should attempt extensive myomectomy through the laparoscope. According to Mais and colleagues, operation time for myomectomy was significantly longer for laparoscopy than for laparotomy when more than four myomata had to be removed and the largest myoma was greater than 6 cm. Dubuisson and co-workers also reinforce the difficulty of the technique by

reporting conversion to laparotomy at a rate of 7.5% (93.7% due to operative difficulties) and a complication rate of 3.8%. These authors echo similar intraoperative concerns: (1) the location of the hysterotomy, (2) the type of hysterotomy, (3) the uterine suture, and (4) removal of the myoma. In addition, they report that one third of patients developed adhesions at the uterine scar. Several technical innovations have been developed to facilitate laparoscopic myomectomy. Electrosurgical and laser techniques are used in ingenious ways. Special traction devices, including corkscrews of various sizes, are required. The operator must be able to provide hemostasis using monopolar cutting current and bipolar forceps. Aquadissection can be used to establish planes for dissection between myomata and the surrounding myo-metrium. Special techniques of approximating myo-metrium with larger curved needles are used with extracorporeal suture tying. Autologous blood donation with intraoperative transfusion when necessary reduces the risk of homologous transfusion. Larger myomata can be removed vaginally with morcellation through a posterior colpotomy incision. In cases of myomas of extreme size, Pelosi and colleagues proposed the use of hand-assisted laparoscopy to avoid a laparotomy. This technique allows the insertion of a hand into the abdomen to assist in dissection. This is accomplished through a glove-sized incision at laparoscopy, while preserving the pneumoperitoneum. A cylindrical serrated morcellator can also be used to convert smaller myomata to small strips of tissue, which can then be removed abdominally through the trocar sleeve or through a minilaparotomy incision. Retrieval of all bits and pieces of myoma tissue from the peritoneal cavity can be a tedious challenge. Hirai and colleagues from Japan described a microwave coagulator and electromechanical tissue borer to minimize invasion of the myometrium and abdominal wall. The proposed advantage of this technique is that, by morcellating the tissue before removal from the uterus, less myometrial trauma is sustained. Horizontal and perpendicular blades at the tip rotate and hollow out the myoma, allowing large myomas to be removed through a small uterine incision. The authors described the use in five patients with four of the five having myomas weighing less than 170 g. The blood loss and operating time were not substantially different than with conventional abdominal procedures. Long-term data regarding myometrial strength over time and pregnancy outcomes are not yet available. More experience with this procedure is necessary to determine its role in myomectomy. Another technical innovation called myolysis has been described by Goldfarb and is based on earlier experience in Europe. Either Nd:YAG laser or bipolar needles are used laparoscopically to penetrate the myomata at multiple sites at a 90-degree angle to the uterus. In response to treatment, the myomata ultimately atrophy. The technique is based on the theory that the coagulating effects of lasers or the bipolar needle can necrose myometrial stroma, denature protein, destroy vascularity, and result in substantial shrinkage of myomas when deprived of their blood supply. Goldfarb advises treatment with GnRH agonists before surgery. The ideal candidates for myolysis are perimenopausal women who have symptomatic leiomyomata measuring 3 to 10 cm or uterine size less than 14 weeks' gestation. Goldfarb combines myolysis with endometrial ablation in patients with symptomatic myomas with persistent uterine bleeding. The addition of myolysis to endometrial ablation increased the rate of postsurgical amenorrhea from 36.5% to 57% and second procedures, including hysterectomy, were reduced from 38% to 12.5%. Goldfarb described significant adhesions at follow-up laparoscopy in patients treated with the Nd:YAG laser technique due to excessive serosal injury from multiple punctures. A circumferential technique was later developed to destroy vasculature instead of the myomatous tissue. The devascularized myoma becomes cyanotic, loses viability, and fibroses. Phillips and colleagues reported on women who underwent elective diagnostic laparoscopy to evaluate adhesions associated with previously performed myolysis. Mean adhesion score was only 1.15 ± 0.6 on a scale of 10. Zreik and colleagues at Yale University modified the myolysis procedure to include cryotechnology to “freeze” uterine leiomyomas. The technique, cryomyolysis, was described in a prospective pilot study of 14 patients. All patients were pretreated with GnRH agonist therapy for 3 months. Thirteen of the 14 endoscopic procedures were performed by laparoscopy and the remaining one by hysteroscopic visualization. Cryoprobe placement was verified and freezing was performed at an internal probe temperature of -180°C until the ice ball encompassed the entire fibroid or reached maximum size. A thaw cycle was then performed, followed by one more freeze-thaw cycle. A hollow track remained within the frozen myoma after removal of the cryoprobe. MRI studies were used to assess uterine and myoma size. The uterus enlarged by 22% after discontinuation of the GnRH therapy. Myoma volume decreased by 6% over 4 months postoperatively, with some patients having a decrease of more than 50%. Four of six women who underwent second-look office laparoscopy had adhesion formation at freezing sites. The authors attributed risk P.779

and severity of adhesion formation to the number of punctures with the cryoprobe. The role of this therapy in conservative treatment of uterine myomata remains to be defined. Hysteroscopic myomectomy and endometrial resection can be performed simultaneously if submucous myomata are present. In more than 300 myolysis procedures, the author reported minimal morbidity with a 30% to 50% reduction in myoma size beyond the reduction achieved with GnRH agonist treatment. No regrowth occurred after several years of follow-up, even after estrogen replacement therapy. Bipolar coagulation myolysis may be less likely to cause damage to the uterine serosa and less likely to cause adhesion formation postoperatively. According to Goldfarb, “As a same-day procedure, myoma coagulation appears to be an extremely safe alternative to hysterectomy, allowing the patient to avoid major surgery and its subsequent recovery time, while providing an alternative solution for patients with symptomatic leiomyomas.” Nezhat and co-workers used a combination of laparoscopy and minilaparotomy to perform myomectomy in 57 women with uteri at 8 to 26 weeks in gestational size. In this laparoscopically assisted myomectomy procedure, the myomata were removed and the uterus repaired through the minilaparotomy incision. It was technically less difficult than laparoscopic myomectomy and allowed better closure of the uterine defects. This technique may be preferable in the case of large myomas in that it is easier to achieve conventional multilayer suturing and easier to extract myomas. A significant disadvantage of myomectomy is the risk of postoperative pelvic adhesions. The adhesions may adversely affect fertility, give rise to pain, and increase the risk of ectopic pregnancy or even intestinal obstruction. Several studies have demonstrated that the risk of postoperative adhesions decreases when a laparoscopic approach is used in lieu of an open abdominal approach. Literature review demonstrates that the average rate of postoperative adhesions after laparoscopic myomectomy is 41% versus more than 90% after a myomectomy via laparotomy. Dubuisson and colleagues assessed adhesion formation after laparoscopic myomectomy in a prospective manner. Forty-five patients underwent a second look after laparoscopic myomectomy. Seventy-two sites were evaluated. The overall rate of postoperative adhesions was 35.6% per patient. The rate of adhesions per myomectomy site was 16.7%. The rate of adhesions on the adnexa was 24.4%. Associations with the occurrence of adnexal adhesions included an additional surgical procedure carried out at the same time, the existence of adhesions before the operation, and posterior location of the myoma. Several factors may increase the risk of postoperative adhesion formation after a laparoscopic myomectomy. Recognition of these factors may be helpful in limiting adhesion formation. The use of uterine suture appears to increase the risk of uterine adhesions. In some studies, the frequency doubled after suturing. The suture induces local tissue ischemia with inflammatory changes, which slow the healing process and induce the formation of adhesions. Contradictory data have been published regarding adhesion formation and the use of bipolar coagulation during a laparoscopic procedure. The location of the myoma also affects adhesion formation. Adhesions are more likely to form when the myomectomy site is located on the posterior uterine wall. During laparoscopy, a uterine incision must be made over each individual myoma. With laparotomy, a single anterior uterine incision may be used for polymyomectomy even when posterior myomas are present. The prior existence of pelvic adhesions significantly increases the risk of postoperative adnexal adhesions, but has not been shown to affect adhesions at the myomectomy site. Laparoscopic myomectomy is further discussed in Chapter 16.

Abdominal Myomectomy The first successful abdominal myomectomy was performed in the United States by the Atlee brothers, Washington and John, in 1844. The first abdominal multiple myomectomy was performed by W. Alexander of Liverpool in 1898. In the early part of the 20th century, the technique of abdominal myomectomy was refined by many notable gynecologic surgeons, including Kelly, Cullen, Mayo, Rubin, Bonney, and others. The procedure did not gain popularity until the middle of the 20th century. The incidence of complications, including hemorrhage, infection, and postoperative intestinal obstruction from adhesions, was considered to be too high. Advances in surgical techniques to control intraoperative bleeding during myomectomy, along with advances in anesthesia, blood transfusion therapy, and GnRH analogs, have made myomectomy a safe alternative to hysterectomy in women with symptomatic leiomyomata. The number of myomectomies performed in the United States is increasing. Because myomectomy is rarely an emergency, time is available to prepare the patient for surgery. It is important that she be properly informed of the reasons myomectomy has been recommended. She should

understand the nature of the procedure so she can know what to expect and what is expected of her. It is especially important that she be informed of the possibility that intraoperative findings may contraindicate myomectomy and require that hysterectomy be performed instead. For example, myomectomy may not be technically feasible if diffuse leiomyomatosis is found. The technical challenge of removing a large cervical myoma can also preclude myomectomy. A preoperative hysterosalpingogram may indicate distortion of the fallopian tubes or uterine cavity, findings that are important in planning the technique of myomectomy. An assessment of fallopian tube patency is helpful in predicting fertility. If the tubes are occluded, however, myomectomy is not necessarily contraindicated. According to Seoud and associates, myomectomy does not interfere with in vitro fertilization performance P.780 in relation to overall and ongoing pregnancy rates. The patient whose tubes are occluded should understand that fertility may not be established by myomectomy, and assisted reproductive technologies may still be required after myomectomy. Tubal reconstruction procedures are uniformly unrewarding when performed at the same time multiple myomectomy is done. Indeed, tubal reconstruction may not always be necessary to establish tubal patency. In a report by Lev-Toaff and associates, nonfilling of the fallopian tubes was present on the preoperative hysterosalpingogram unilaterally in two patients and bilaterally in another two. In all four patients, tubal patency was shown after myomectomy. In the experience of these authors, hysterosalpingography before myomectomy can assist the gynecologic surgeon in planning the surgical approach by showing the presence, size, and location of submucous leiomyomata and concomitant tubal disease. Imaging modalities such as transabdominal and transvaginal US and MRI play an important role in the management of patients with leiomyomata, especially those patients who are being prepared for myomectomy. As explained by Mayer and Shipilov, US is the preferred method for screening and initial evaluation of the pelvis. In many cases, it is the only imaging study necessary. There are special cases for which US cannot provide all the diagnostic information required. In a study by Schwartz and colleagues, US results were inconclusive in 20% of cases and did not yield a definitive diagnosis in 59% of cases. MRI was more definitive in all cases (Fig 30.20). The preoperative diagnosis of submucous leiomyomata by MRI may allow hysteroscopic resection and avoid abdominal myomectomy in some cases. Differentiation between uterine leiomyomata and adnexal pathology is more accurate with MRI and thus avoids the need for laparoscopy or laparotomy in some cases (Fig. 30.21). MRI studies can differentiate between uterine leiomyomata, diffuse and localized adenomyosis, and diffuse leiomyomatosis. MRI is the most accurate imaging technique for the detection and localization of leiomyomata (Fig 30.22). Hricak and co-workers were able to identify accurately by MRI all subserosal (9 of 9), all intramural (37), and 10 of 11 submucosal leiomyomata. Myomata as small as 0.3 cm can be detected. Various degrees of cellularity, degeneration, necrosis, and calcification can be identified by MRI, and sarcomatous change can be suspected. MRI provides imaging planes that are not available on transabdominal or transvaginal US, a feature that permits better visualization of the more lateral and posterior areas of the pelvis. MRI is the P.781 most accurate method for preoperative localization of leiomyomata and surgical planning for myomectomy. Given the greater costs of MRI, it should be used judiciously. However, as noted by Mayer and Shipilov, the effective cost differential between MRI and US is decreasing.

FIGURE 30.20. A: The ultrasound evaluation of this patient with a very large leiomyoma was not helpful in delineating the leiomyoma from the adnexa. B: The magnetic resonance imaging study demonstrated multiple leiomyomas distinctly separate from the adnexa bilaterally. The ovaries are seen bilaterally with multiple cysts. (Image courtesy of Deborah Baumgarten, M.D.)

View Figure

FIGURE 30.21. On magnetic resonance imaging, a large anterior, pedunculated leiomyoma is shown as a separate entity from the ovary in this patient. The ovary is displaced superiorly. (Image courtesy of Deborah Baumgarten, M.D.)

View Figure

FIGURE 30.22. The location of this small posterior intramural leiomyoma is clearly delineated in this magnetic resonance image. The cervical canal is also easily visible. (Image courtesy of Deborah Baumgarten, M.D.)

View Figure

As discussed by Wiskind and Thompson, one of the most serious risks of surgical bleeding during myomectomy is the risk associated with homologous blood transfusion. The first rule in reducing or eliminating the need for transfusion is to bring the patient to the operating room with the highest possible hemoglobin and hematocrit level. About 30% of myomectomy patients have associated menorrhagia. These small repeated menstrual hemorrhages deplete the body's iron stores over time and eventually result in iron-deficiency anemia of various degrees of severity. Patients scheduled for myomectomy benefit from oral iron supplementation. In a study by Thompson, the liberal use of oral iron therapy preoperatively was shown to decrease the number of blood transfusions on the gynecologic surgical service at the Johns Hopkins Hospital. A blood transfusion is seldom necessary to correct iron-deficiency anemia in a gynecologic patient. Blood transfusions should generally be reserved for patients with hypovolemic shock or aregenerative forms of anemia. In most other circumstances, elective surgery should be delayed until the anemia has been corrected by oral iron supplementation. Occasionally, patients with a myomatous uterus have iron-deficiency anemia due to menstrual bleeding that is too heavy or too continuous to allow a response to oral iron therapy. In this situation, it may be beneficial to induce amenorrhea with hormonal therapy to allow the anemia to be corrected more expeditiously. Amenorrhea can be induced with progestational agents such as norethindrone or medroxyprogesterone acetate, with danazol, or with GnRH agonists. Several studies have demonstrated a significant increase in hemoglobin and hematocrit values in patients with leiomyomata treated preoperatively for 8 to 24 weeks with GnRH analogs compared with matched control groups. Friedman and colleagues also found a significant increase in serum iron and total iron-binding capacity in a study group treated with the GnRH agonist leuprolide acetate. In some patients, oral iron was also given. In an evaluation of 265 patients, GnRH agonists plus iron were more effective than iron alone in treating the anemia of patients with uterine leiomyomata, according to Stovall and co-workers. In a double-blind, placebo-controlled, multicenter study, Friedman and associates reported resolution of menorrhagia in 97% of uterine leiomyomata patients treated with GnRH agonists. Preoperative treatment with GnRH analogs can actually reduce the operative blood loss during myomectomy, according to studies by Friedman and colleagues, Andreyko and co-workers, Moghissi, and others. In a prospective, randomized study of 50 patients undergoing hysterectomy for symptomatic leiomyomata, Stovall and colleagues found a significant decrease in operative blood loss between those patients who received 2 months of leuprolide acetate treatment preoperatively and matched control subjects. An elegant study by Friedman and associates demonstrated a significant decrease in operative blood loss during myomectomy between patients with pretreatment uterine volumes greater than 600 cm 3 who were treated with depot leuprolide acetate for 12 weeks preoperatively and a matched control group. However, there was no significant difference in blood loss between the two groups when patients with smaller uterine volumes (150 to 600 cm 3) were included in the analysis. It has been suggested that the

hypoestrogenic environment caused by GnRH analog therapy reduces the vascular supply to uterine leiomyomata. However, even in patients not treated with GnRH agonists, blood flow has been observed to be lower in myomata and adjacent tissue. The advantages and disadvantages of the preoperative collection of autologous blood are discussed in Chapter 18. Intraoperative autotransfusion and normo-volemic hemodilution are also discussed. These techniques of reducing or avoiding the risk of homologous blood transfusion are discussed in detail by Wiskind and Thompson. Perioperative antimicrobial prophylaxis is indicated with myomectomy. It is preferable to perform the operation in the follicular phase of the menstrual cycle. This avoids the chance of encountering an unknown or unsuspected pregnancy and reduces the problems encountered P.782 when a fresh corpus luteum is inadvertently traumatized. After induction of anesthesia, the patient is placed in Allen universal stirrups, the bladder is emptied, and a careful pelvic examination, including a rectovaginal-abdominal bimanual examination, is performed under anesthesia. Preparation and draping are done to allow access to the vagina and cervix in case it is necessary to place an instrument through the cervix and into the endometrial cavity during the procedure. Cervical dilatation should be done to facilitate postoperative drainage from the endometrial cavity, especially for cases in which the endometrial cavity has been entered during the myomectomy. Many of the general principles of pelvic surgery are applicable to myomectomy. Perhaps the most important of these is optimum exposure at the operative site. This is accomplished primarily by an adequate incision, but there must also be proper retraction, good lighting, and able assistants. Although a Pfannenstiel incision is considered adequate for myomectomy on a small uterus, we prefer the Maylard incision for larger uteri, even those that exceed a size equivalent to a 12-week pregnancy. A Maylard incision provides excellent exposure throughout the pelvis. Because it is a transverse incision, it is stronger and provides better cosmesis than a vertical midline incision. A Bookwalter retractor optimizes exposure of the operative site. A Pfannenstiel incision can be used for removal of a small solitary myoma. The importance of adequate exposure cannot be overemphasized. With proper exposure, operative time can be shortened and surgical bleeding can be more easily identified and controlled. Limited exposure may lengthen operative time, increase the risk of inadvertent injury to other pelvic structures, and force abandonment of a myomectomy in favor of a hysterectomy in especially difficult cases. After the peritoneal cavity is entered, the abdomen is explored as usual. Adhesions in the pelvis must be carefully released or excised so that the intestines can be placed in the upper abdomen and held there with packs. The operation is performed according to microsurgical techniques and principles. For example, the laparotomy packs that are used to hold the intestines in the upper abdomen are placed in plastic bags to reduce the microscopic trauma to peritoneal surfaces caused by regularlaparotomy packs. Lintless laparotomy packs are preferred. Several laparotomy packs in plastic bags can be used to fill the cul-de-sac, thus elevating and stabilizing the uterus for easier access. Visualization of the operative site can be improved by the liberal use of suction to remove blood from the field. Suction should be used instead of sponges because it allows for a more accurate determination of blood loss and is less traumatic to tissues. The operative field is kept moist and free of clots with a solution of lactated Ringer solution containing heparin. Very fine instruments and sutures are used when possible, and tissue is handled gently to avoid unnecessary trauma to serosal surfaces. Traumatic instrumentation (e.g., uterine elevators with teeth, Kocher clamps, or any instrument on the uterine serosa) must be avoided. Sutures on serosal surfaces should be of a fine absorbable nonreactive material. Running suture lines are preferable to avoid extra knot volume, which may contribute to adhesion formation. If pelvic adhesions develop after myomectomy, future fertility may be adversely affected. Performing the operation in a way that minimizes adhesion formation greatly improves the possibility of a successful result. At this point in the operative procedure, one should pause and evaluate the size, location, and number of myomata present. Special note should be made of their proximity to the endocervical canal, uterine vessels, and fallopian tubes. One must decide if myomectomy is still feasible, how the leiomyomata will be removed (and in what sequence), and how the uterus will be reconstructed. The conservation of uterine function with myomectomy requires control of bleeding from uterine incisions and myoma beds. Contrary to hysterectomy for leiomyomata, conservation of the uterus requires that the blood supply to the uterus through the uterine and ovarian vessels remain intact. Removing multiple myomata embedded deeply in a vascular myometrium can result in considerable blood loss. Proper application of special techniques to limit blood loss can allow multiple myomectomies even in uteri up to

20 weeks' pregnancy size if satisfactory reconstruction is possible. Controlled hypotensive anesthesia has become a useful adjunct to decrease surgical bleeding in selected patients. The main mechanism in the control of operative field bleeding with hypotensive anesthesia is the reduction of venous tone. This can be accomplished by specific vasodilating agents, such as nitroglycerin or sodium nitroprusside, epidural or spinal anesthesia, some inhalation anesthetic agents, and ganglionic blockade, to achieve and maintain a target mean blood pressure of 60 mm Hg. Our experience with this technique has been favorable. Venous bleeding can be further reduced if the patient is placed in a moderate Trendelenburg position. This facilitates venous drainage from the lower extremities and pelvis by gravity and may further reduce the blood pressure at the operative site. Induced hypotension is contraindicated in patients with cerebrovascular disease, myocardial ischemia, peripheral vascular disease, severe renal or hepatic disease, and hypovolemia. None of these contraindications is seen very often in myomectomy patients. An anesthesiologist experienced with the technique is an essential requirement. The decision to use hypotension should be made jointly by the surgeon and the anesthesiologist. It is essential that the blood pressure be returned to normal before closure of the incision to ensure that adequate surgical hemostasis has been established. Early proponents of myomectomy focused on methods to temporarily occlude uterine blood flow to control hemorrhage and provide a bloodless operative field. One of the earliest methods was simply to have an assistant grasp the broad ligaments firmly with each hand during myomectomy to impede blood flow through the uterine P.783 vessels. In the 1920s, Victor Bonney introduced a specially designed clamp that was placed around the uterine vessels and the round ligaments. The ovarian vessels were occluded with ring forceps. Using this technique, he was able on one occasion to remove more than 200 myomata from a single uterus. Rubin, in 1938, was the first to use an elastic rubber tourniquet through the broad ligament, encircling the cervix and occluding the uterine vessels during myomectomy. Rubber-shod clamps applied to the broad ligaments have also been used to occlude the uterine vessels and control bleeding. Gynecologic surgeons do not often have the opportunity to use tourniquets to control bleeding; however, a myomectomy is particularly suited to their use. We prefer to use tourniquets fashioned in the manner of a Rumel-type tourniquet, which is used by vascular, thoracic, and trauma surgeons to occlude major vessels. Initially, a small hole is made in an avascular space in the broad ligament on either side of the uterine isthmus just lateral to the uterine vessels. A 5-French pediatric feeding tube is looped around the upper cervix through the holes in the broad ligament, and the two ends of the tube are then threaded through a 4-inch length of 35-French Malecot catheter and held with a clamp. A loop tourniquet can then be placed around each infundibulopelvic ligament through the same holes in the broad ligaments (Fig. 30.23A). FIGURE 30.23. A: Through a small hole in the broad ligament on each side of the uterus, a Rumel tourniquet is placed around the lower uterus and around each infundibulopelvic ligament. B: When the tourniquets are tightened sufficiently, the blood flow to the uterus stops. The absence of arterial pulsations can be determined with the sterile Doptone.

View Figure

As the tourniquets are being placed, controlled hypotension is induced by the anesthesiology staff. Before the tourniquets are tightened, the location of the uterine arterial blood flow is identified with a sterile Doptone. When everything is in readiness and the plan of operation P.784 has been selected by the surgical team, the tourniquets are snugged down and tightened progressively until the uterine arterial flow is no longer audible with the Doptone (Fig. 30.23B). It is very important that the arterial blood flow be occluded. If the venous flow is occluded while the arterial flow remains intact, blood loss could actually be increased with the tourniquets. The mean blood pressure should be reduced to the target hypotensive level (about 60 mm Hg) before the tourniquets are tightened. The higher the blood pressure, the tighter the tourniquets must be to occlude the uterine circulation. With the combination of properly applied tourniquets and controlled hypotensive anesthesia, the entire circulation to the uterus can be occluded. The myomectomy can then be performed in a bloodless field, greatly facilitating complete removal of all tumors and a neat reconstruction of the uterus (Fig. 30.24 and Fig. 30.25). Occasionally, a large cervical or broad ligament myoma prevents placement of the tourniquets. In this situation, the offending tumor should be removed first, the defects repaired when feasible, and the tourniquets applied for the remainder of the multiple myomectomy. FIGURE 30.24. With tourniquets properly secured, myomectomy can be performed with minimal blood loss. See color version of figure.

View Figure

FIGURE 30.25. The use of tourniquets to control bleeding facilitates closure of the myoma bed and uterine incision. See color version of figure.

View Figure

Once the tourniquets are tightened in place, the myomectomy should proceed expeditiously to prevent ischemic damage to the uterus, tubes, and ovaries. The length of time the pelvic structures can be without blood flow before irreversible damage occurs is unknown. We generally do not release the tourniquets until the myomectomy is complete (usually within an hour) and have not experienced any adverse events. Lock also agrees that intermittent release of the tourniquets is unnecessary. However, because the potential for injury does exist, tourniquet time should be monitored and kept to a minimum. The

tourniquets should not be tightened until the surgical team is ready to perform the myomectomy. Intermittent release of the tourniquets should be considered if the operating time becomes excessive. We usually P.785 release the tourniquets around the ovarian vessels as the uterine serosa is being closed to restore circulation to the tubes and ovaries and to restore some collateral flow to the uterus. After reconstruction of the uterus is complete, the determination of adequate hemostasis in the uterus cannot be made until all the tourniquets are released and the blood pressure has returned to normal. Sometimes additional sutures are required for hemostasis. Before the abdomen is closed, the small holes in the broad ligament are repaired with figure-of-eight sutures. Tourniquets are not necessary for every myomectomy, especially when the tumors are small or pedunculated. However, they are safe and inexpensive to use and can be of great benefit when large or multiple intramural myomata must be removed. A criticism of uterine tourniquets is that they are traumatic to pelvic structures. Our experience, to the contrary, is that soft plastic tubes used as tourniquets are quite atraumatic. No injuries attributable to these tourniquets have occurred in several hundred cases. An alternative to the use of tourniquets to control bleeding during myomectomy is the local injection of vasoconstrictive agents. Perhaps the most commonly used agent is vasopressin, a synthetic derivative of the antidiuretic hormone from the posterior lobe of the pituitary gland. In addition to this antidiuretic effect, vasopressin induces smooth muscle contraction of the gastrointestinal tract and the vascular bed. In particular, it has been found to have a potent vasoconstrictor effect on the nonpregnant uterus when injected locally. It has a plasma half-life of 10 to 20 minutes and has been used effectively as a hemostatic agent during myomectomy. Pharmacologic vasoconstriction can be accomplished with vasopressin (antidiuretic hormone), 20 U (Pitressin). Twenty units of vasopressin are diluted in 20 mL of normal saline and injected into the superficial myometrium and serosa overlying the myoma. The effect usually lasts for 30 minutes. Dillon reported that with the use of vasopressin, 72% of patients requiring myomectomy did not need blood replacement, compared with control subjects. Frederick and co-workers noted significantly less blood loss compared with an untreated group. Ginsburg and associates compared vasopressin with mechanical vascular occlusion and found that there were no demonstrable differences in blood loss, morbidity, or transfusion requirements between the two techniques. A favorable experience with vasopressin has also been reported by Semm and Mettler. The weight of evidence in current clinical investigation indicates that vasopressin is as effective as mechanical vascular occlusion in controlling blood loss with myomectomy. Nevertheless, careful dissection around myomata and prompt suturing with exertion of direct pressure to bleeding vessels by the operative assistant are necessary to minimize blood loss. Care should be taken to avoid injecting the solution directly into a vascular channel, and no more than 30 mL per patient is recommended because of potential side effects. Vasopressin should not be used in patients with vascular disease, especially disease of the coronary arteries. Inadvertent intravascular injection can cause anginal pain; larger doses can cause myocardial infarction. Water intoxication can also occur as a result of the antidiuretic effect of vasopressin. This effect is potentiated in patients taking tricyclic antidepressants. Although late postoperative bleeding does occur with the use of vasopressin, it is not a common complication. Arterial bleeding masked by vasopressin still requires suture ligation. Because of the short half-life of vasopressin, the hemostatic effect is observed only for 20 to 30 minutes and should be over before the incisional closure is started. However, some do claim that vasopressin simply delays bleeding, gives a false sense of security, and is not particularly effective for larger myomata and very extensive myomectomies. For a variety of reasons, epinephrine as a vasoconstrictive agent is not recommended for use in gynecologic surgery. Since its introduction into clinical practice in 1972, the CO2 laser has been touted as a tool that increases surgical precision and decreases bleeding, tissue injury, and adhesion formation. The laser can be used to make a single uterine incision through which multiple myomata are removed. An elliptical incision can also be made around the base of larger myomata to facilitate their removal. Myomata less than 1 cm in diameter can be vaporized directly with the laser, which destroys tissue by vaporizing cellular water. Despite the favorable results reported by Weather, Reyniak and Corenthal, McLaughlin, and Starks, we believe that there is no clear advantage in using the CO 2 laser for abdominal myomectomy, especially considering the added cost to the patient. Although the methods described earlier to control bleeding during myomectomy are helpful, they cannot

substitute for good surgical techniques. Adherence to basic principles is essential for good results. Perhaps the most important of these is careful planning of the uterine incisions. Only a minimal number of incisions should be made. If possible, removal of all leiomyomata should be accomplished through a single incision in the anterior uterine corpus, and in the midline when feasible to avoid the vascular areas of the uterus and broad ligaments laterally. Even intramural leiomyomata in the posterior uterine wall can be removed through anterior incisions. Incisions in the posterior uterine wall may be necessary, however, if posterior subserous tumors are being removed. If posterior uterine incisions are made, adhesions are more likely to develop and will likely involve the tubes and ovaries as well. As many tumors as possible should be removed through a single incision. Methods of removing myomata through a single anterior incision have been described by Bonney. The linear or elliptic incision should usually be over the largest myoma. It should be carried through the superficial myometrium directly into the underlying myoma. The myoma is then grasped with a double-tooth tenaculum or a large Lahey thyroid clamp for traction. The plane of cleavage between the myoma and the surrounding myometrium is easily identified. Sometimes in P.786 patients who have been treated with GnRH analogs, the plane of cleavage may seem less distinct. Sharp dissection with the scalpel or Metzenbaum scissors, or blunt dissection with the finger or knife handle, is required to enucleate the myoma from its bed. Sometimes the myoma is larger than expected. It may then be necessary to enlarge the incision or to remove the tumor by morcellation. Other adjacent tumors should be removed through the same incision. Any entry in the endometrial cavity should be noted, and a special attempt should be made to close it with sutures placed in the underlying supporting myometrium. Examples of the step-by-step planning and P.787 performance of a multiple myomectomy are illustrated in Figure 30.26 and Figure 30.27.

View Figure

FIGURE 30.26. The sequence of steps in a multiple myomectomy is shown in these illustrations. A: Through a transverse Maylard incision, tourniquets are placed to occlude the uterine and ovarian artery flow. Through a single incision in the anterior myometrium, a large anterior myoma is removed first. All other myomas are removed through this incision. B: A smaller intramural myoma is removed through the same incision. C: To avoid making a separate incision in the posterior uterine wall, a large posterior myoma is removed through the uterine cavity. After an incision has been made through the anterior endometrium, an incision is made in the posterior endometrium directly over the posterior myoma. D: The myoma in the posterior uterine wall is dissected from its bed and removed through the uterine cavity. An incision in the posterior uterine serosa is thus avoided. E: Multiple sutures (2-0 delayed-absorbable) are used to close the defect in the posterior uterine wall first. Incisions in the uterine cavity are closed. Then the defects in the anterior uterine wall are closed. F: Trimming excess myometrium from the anterior uterine wall allows a better approximation of the myometrium. The edges of the serosa are closed with a continuous “baseball” stitch with 4-0 delayed-absorbable sutures.

FIGURE 30.27. Techniques of multiple myomectomy. A: A vertical incision is made over a myoma on the anterior surface of the fundus as close to the midline as possible. Many myomas can be removed through this single incision. B: The incision is extended into the substance of the myoma. C: By incising progressively deeper into the myoma, the surgeon can identify and bluntly dissect the plane between myoma capsule and myometrium. D: Sharp dissection may be necessary to separate the myoma from its capsule at its base. E: After the removal of as many myomas as possible, the remaining cavity is obliterated and hemostasis secured. F: Multiple rows of nonreactive interrupted absorbable suture material are used to close the cavity. G: When the “dead space” has been obliterated, the serosa is closed with a continuous “baseball” suture of 5-0 or 6-0 nonreactive absorbable material. H: This type of closure approximates the serosal edges. View Figure

The muscle fibers and blood vessels surrounding a myoma are compressed by its growth. This compression of surrounding tissue forms a pseudocapsule around the myoma. No large blood vessels enter the myoma, and there is no vascular pedicle. If the dissection can be carried out between the myoma and the pseudocapsule, blood loss can be minimized. If blood vessels are cut or left on the surface of the myoma, it usually means that the dissection has been carried out in an improper plane. Dissection in the proper plane may be more difficult if the patient received GnRH analog therapy preoperatively. P.788 Several ingenious techniques for removing leiomyomata and for repairing defects have been described. For example, Bonney's hood can be used to remove a large leiomyoma in the uterine fundus. The myoma is first exposed through an elliptic incision made transversely across the anterior fundus, taking care to avoid the interstitial portion of the fallopian tube on each side (Fig. 30.28A and Fig. 30.28B). After the primary tumor is removed (Fig. 30.28C), other leiomyomata can also be removed through the same incision. Excess myometrium can be trimmed away (30.28D). Interrupted sutures obliterate the dead space, approximate the myometrium, and accomplish satisfactory hemostasis. The sutures are placed in such a way that the posterior flap of myometrium is folded over the anterior uterine wall and sutured in place, thus fashioning Bonney's hood (Fig. 30.28D).

FIGURE 30.28. Technique of myomectomy using an anterior hood incision as described by Bonney. A: A transverse incision is made in the anterior fundal wall over the myoma. B: Sagittal view of the location of the incision. C and D: Using blunt and sharp dissection, the surgeon enucleates the myoma from its bed. Excess hypertrophied myometrium may be trimmed and removed before closure of myometrium and serosa.

View Figure

Meticulous closure of defects from the enucleated myomata is essential to maintain hemostasis postoperatively, but this should be deferred until all the tumors are removed. Hypertrophy of the normal myometrium is always present with uterine leiomyomata. Some of this hypertrophied myometrium is considered excess and can be trimmed to facilitate a more normal reconstruction of the uterus. Involution of the myometrial hypertrophy is expected to occur in the first few months after myomectomy. Therefore, only a small amount of normal myometrial P.789 tissue should be removed. In reconstructing the uterus, the surgeon should refer to fixed points such as the attachments of the round ligaments and fallopian tubes on each side of the corpus. Symmetric reconstruction is preferred but is not always possible. The myoma beds are usually closed with interrupted figure-of-eight or mattress 2-0 delayed-absorbable sutures. Large defects can be closed initially with a pursestring suture to obliterate the dead space. Several layers of sutures may be required. One must be careful to avoid occlusion of the uterine vessels, the endocervical canal, or the interstitial portion of the fallopian tubes. Transfundal ortranscervical chromotubation to test fallopian tube patency after uterine reconstruction is complete is not usually possible because of leakage of the dye through the myometrial incisions. In closing a myomectomy incision, the security of the closure comes from sutures placed in the myometrium. If possible, these sutures and knots must not be exposed. In Figure 30.26E and Figure 30.26F, several techniques of closing myometrial defects are illustrated. The serosal edge of the uterine incision should be carefully approximated with a continuous 5-0 or 4-0 delayed-absorbable “baseball” stitch. The tourniquets are removed, the hypotensive anesthesia is reversed, and the uterus is carefully inspected for evidence of bleeding. Additional sutures are sometimes required. If a uterine suspension is needed, a modified Coffey or modified Gilliam technique along with uterosacral ligament plication is used. Adhesion prevention can also be achieved by the use of absorbable or nonabsorbable barriers. The absorbable barrier Interceed (oxidized regenerated cellulose) can be placed over the uterine corpus to protect the tubes and ovaries from denuded peritoneal surfaces and uterine incision. Alternatively, a nonabsorbable barrier, Gore-Tex (polytetrafluoroethylene surgical membrane), can be sutured over the uterine incisions P.790 with 7-0 absorbable minimal reactive sutures. The use of Gore-Tex has been associated with a reduction in new adhesion formation. Diamond and the Seprafilm Adhesion Study Group assessed the efficacy of Seprafilm (HAL-F) Bioresorbable Membrane (sodium hyaluronate and carboxymethylcellulose) in reducing the incidence, severity, extent, and area of uterine adhesions after myomectomy. This prospective,

randomized, blinded, multicenter study involved an independent gynecologic surgeon's review of each patient's second-look laparoscopy. One hundred and twenty-seven women undergoing uterine myomectomy with at least one posterior uterine incision were randomized to treatment with Seprafilm or no treatment at the completion of the myomectomy. All indices, including incidence, severity, and extent of adhesions, were decreased in the treatment group. This suggests that newer barriers may also have a role in adhesion prevention. Free grafts of peritoneum or omentum should not be used to cover uterine incisions. Second-look laparoscopy may be indicated in patients with multiple incisions or in those with posteriorly located incisions adjacent to the adnexa. Early adhesions can be easily lysed, and an additional barrier membrane can be placed. The role of second-look laparoscopy is not well defined, and conflicting studies can be found in the literature regarding the efficacy of this procedure. A comprehensive review of methods to prevent adhesion formation in gynecologic surgery has been published by Damario and Rock and by diZerega.

Results of Myomectomy An extensive multiple myomectomy is a major operation with the potential for a higher morbidity than that found with hysterectomy. The major immediate postoperative complications after myomectomy are febrile morbidity and intraperitoneal bleeding. Postoperative febrile morbidity may be related to extensive tissue trauma or to infection for a variety of reasons. Perioperative antibiotics are routinely given, but antibiotics are not usually continued beyond the day of operation. Any evidence of infection in the recovery period should be treated vigorously and promptly because infection in the operative site may be adhesiogenic and may have devastating effects on future fertility. Unfortunately, subclinical infection in the operative site may not be recognized and therefore may not be treated, but it can also have adverse effects on fertility because of de novo adhesion formation. For these reasons, meticulous and sterile surgical technique during myomectomy must be impeccable. Intraperitoneal bleeding after myomectomy is usually due to failure to achieve hemostasis of the myometrial vessels during closure of the myoma beds and uterine incisions. Although we do use a heparin solution (5,000 U of heparin per 1,000 mL of lactated Ringer solution) for irrigation during myomectomy, there is no evidence to suggest that this contributes to occult intraperitoneal bleeding. The diagnosis of intraperitoneal bleeding in the postoperative patient may be difficult. The vital signs can remain stable for several hours before rapidly deteriorating. Peritoneal signs are often subtle and may be masked by incisional pain and analgesic medications. In addition, the peritoneal cavity has an enormous capacity for accommodating occult blood loss. Indeed, as much as 3,000 mL of blood can be shed into the peritoneal cavity with only a 1-cm increase in the abdominal radius. Therefore, patients must be carefully monitored for the first 24 hours after myomectomy. Vital signs are routinely checked every 15 minutes for the first 2 hours after surgery, then every 30 minutes until stable. Subsequently, they are monitored every 2 to 4 hours for the first 24 hours postoperatively. A hematocrit is usually performed 6 hours after the operation is completed and again on the first postoperative morning. It can also be performed whenever there is a suspicion of intraperitoneal bleeding, anemia, or hypovolemia. Any sign of restlessness, tachycardia, or tachypnea may be an indication of blood loss, especially when associated with hypotension. When occult postoperative intraperitoneal bleeding is suspected, peritoneal lavage can be a valuable diagnostic tool. If the lavage solution yields a red blood cell count of 100,000/mm 3, intraperitoneal bleeding is likely and reexploration is indicated without delay. Lavage is unnecessary when the diagnosis of intraperitoneal bleeding is unequivocal and associated with definite hypovo-lemia. In this situation, immediate return to the operating room for reexploration is indicated. Postoperative bleeding after myomectomy can be devastating. Intraoperative control of bleeding during an extensive multiple myomectomy often requires that the uterine blood flow be impeded with tourniquets, clamps, or the local injection of vasoconstrictive agents. However, the demonstration of adequate surgical hemostasis in the uterus cannot be made until the uterine circulation has been fully restored. Assiduous attention to this principle intraoperatively prevents postmyomectomy bleeding in almost all cases. Reports by Smith and Uhlir, Rosenfield, and LaMorte and associates indicate that the morbidity of myomectomy is no greater than the morbidity of hysterectomy. Verkauf reviewed current published reports and found that the operative risk of myomectomy does not exceed that of hysterectomy. One case of disseminated intravascular coagulation, hemolytic anemia, and acute renal failure associated with extensive multiple myomectomy was reported by Sacks and Hoyne. Myomectomy has an excellent record in reducing heavy menstruation in patients with a complaint of

menorrhagia. In more than 80% of patients, menorrhagia is cured or significantly improved. Pelvic pain and discomfort and dysmenorrhea can also be relieved, but the results are not as dramatic because leiomyomata are often associated with other gynecologic diseases (e.g., endometriosis and pelvic inflammatory disease) that can also cause pelvic pain. The impact of abdominal myomectomy on infertility is difficult to assess. Other factors besides leiomyomata P.791 may be present to a varying degree. The extent to which the uterine cavity or the fallopian tubes are distorted also varies. The percentage of patients in each series who wish to conceive after myomectomy is not the same. There is also considerable variation in the surgical technique and skill of the gynecologic surgeon. Prospective, randomized, controlled studies are lacking. These and other factors make it difficult to assess the impact of abdominal myomectomy on infertility. There are a number of published reports regarding women who experience recurrent pregnancy wastage or prior infertility with another cause, and who undergo myomectomy. According to Verkauf's review, conception occurs in more than half of such women who were not previously pregnant. A review of nine abdominal myomectomy reports by Vercellini and colleagues indicates similar results. Two hundred thirty-four women similarly evaluated and treated by myomectomy were included in Verkauf's review. Of patients with leiomyoma-associated infertility without other apparent cause, 59.5% conceived. Among patients who had additional treated infertility factors, 50% conceived after myomectomy. Eighty-six percent of pregnancies occurred within 2 years of the operation. The conception rate among women older than age 35 may not be as good. Also, the postmyomectomy conception rate may be lower when the uterus is greater than 12 weeks' gestational size and when more than four myomata are removed. When abdominal myomectomy includes removal of submucous myomata, Garcia and Tureck report that 53% of patients attempting to establish a pregnancy conceive. Both Li and colleagues and Vercellini and co-workers published retrospective reviews supporting excellent conception and pregnancy rates following abdominal myomectomies. Vercellini's work suggests that certainly age at the time of the procedure is important as it was one of three independent variables (age, duration of infertility before surgery, and the presence of other infertility factors) associated with postoperative cumulative conception rate. Buttram and Reiter reviewed 1,914 cases of myomectomy and compared preoperative and postoperative abortion rates. The reduction in abortion rate after myomectomy from 41% to 19% suggests improvement in reproductive salvage through the use of this procedure. According to Verkauf's review, 69.2% of women with previous recurrent pregnancy wastage conceived after myomectomy and had a reduction in fetal loss. An ultrasonographic study of uterine remodeling after conservative myomectomy was reported by Beyth and associates. There was a gradual decrease in uterine volume in all patients during the 6 months after myomectomy, with the most remarkable decrease occurring in the first 2 to 3 months. Presumably, this represents an involution of myometrial hypertrophy and postoperative healing of uterine incisions. We recommend that all patients use local methods of contraception (diaphragm, condoms, and spermicidal jelly or foam) for at least 3 months to avoid conception until the myomectomy incisions are healed. Finally, there is the matter of recurrence of myomata after myomectomy. In Verkauf's review, leiomyomata recurred in 7.5% of patients, and 6.8% required reoperation. Most recurrences appeared more than 3 years after myomectomy, thus allowing sufficient time for conception to occur before recurrence. Friedman and associates investigated a concern that GnRH agonist–induced myoma shrinkage would make some small intramural and submucosal tumors “invisible” at myomectomy, causing early “recurrence” of leiomyomata once gonadal suppression ceased and estrogen production returned. In their study, there was no difference in myoma recurrence between women pretreated with GnRH agonists (67%) and those treated with placebo (56%) 27 to 38 months after myomectomy. Their myoma recurrence rate of 61% is much higher than that previously reported in combined myomectomy series. The authors believe that this discrepancy is most likely due to the use of high-resolution US to diagnose small myomata that would otherwise be missed on bimanual examination. Matta and colleagues reported that after GnRH analog treatment, the US outline of some myomata was lost or obscured. Such myomata are probably more difficult to identify and remove with myomectomy and may be more likely to reappear when GnRH analog treatment is discontinued after myomectomy.

Embolotherapy With current technology progressing toward less invasive therapies, the minimally invasive procedure of uterine artery embolization (UAE) is gaining popularity. This procedure can potentially obviate the need for surgical procedures in patients who suffer from symptomatic leiomyomas.

In the female genital tract, embolotherapy for control of hemorrhage from malignancy was first reported in the late 1970s. In 1980, Pais and colleagues described successful embolization for postpartum hemorrhage. Many more reports followed in the mid 1980s. In the early 1990s, Ravina and co-workers began utilizing embolotherapy as a preoperative maneuver to decrease intraoperative blood loss during surgery for myomas. The protocol generally included embolization about 24 hours before the surgery; however, some occurred a few days or weeks before surgery. Such an improvement in symptoms occurred that many surgeries were canceled altogether. This serendipitous discovery led to the performance of UAE as a primary procedure. UAE for leiomyomas was first performed in the United States by Goodwin and colleagues in 1995. Since then, several large series have been reported and experience continues to grow. UAE is appropriate for patients with symptomatic leiomyomas and preference for other medical or surgical treatment. Clinical findings, therapeutic goals, and overall medical conditions factor into the decision making. Several concerns have developed in treating patients who may desire future conception. Hypotheses include reduced fertility as a consequence of injury to the uterus or ovaries, placental insufficiency resulting from inadequate P.792 blood flow through the uterus, or uterine rupture during pregnancy from UAE induced myoma necrosis. In practice, there have been no reports of decreased fertility after occlusion of the major uterine vessels either by surgical or radiologic technique. In addition, there are numerous reports of successful pregnancies after these procedures. Amenorrhea has been reported in 1% to 2% of patients after UAE; however, some authors have attributed this to the coincidental onset of menopause. Contraindications to this procedure include pregnancy, active pelvic infection, severe contrast medium allergy, arteriovenous malformations, desire for future pregnancy, and undiagnosed pelvic mass. The technique is generally preceded by preprocedural testing and patients are pretreated with intravenous antibiotics. Many perform preliminary arteriographic mapping of the pelvis. A review by Hutchins and Worthington-Kirsch provides an excellent description of the procedure. The technical success rate is consistently reported in the 96% to 98% range with experienced teams. Eighty percent to 90% of embolized patients have reported improvements in menorrhagia, bulk related symptoms, or both. Reduction in overall uterine volume peaks at more than 60% by 6 to 9 months after the procedure. Individual myomas show average volume reductions of 60% to 65%. Forty percent of patients develop a syndrome of fever and malaise in the first 10 to 14 days after UAE. This is also associated with leukocytosis. This entity is well described as postembolization syndrome. It is typically self-limited and resolves in 3 to 5 days and rarely requires treatment except antipyretics. Other complications include those that may be attributed to the angiographic component and target or nontarget organ embolization. Groin infections, groin bleeding or hematoma, contrast-induced renal damage, and vascular damage may be attributed to the angiographic component. Uterine infection or perforation, sexual dysfunction, and myoma sloughing may be attributed to the target organ effects. Reported nontarget organ embolization complications include ovarian sequelae, sciatic nerve effects, and gluteal muscle pain. Current experience confirms a major complication rate of less than 1%.

Chapter 31 Hysterectomy Howard W. Jones III Hysterectomy is the most common operation performed by the gynecologist, and it is the second most common major surgical procedure done in the United States. Only cesarean section is more common. There are many indications for hysterectomy and the uterus can be removed using any of a variety of techniques and approaches, including abdominal, vaginal, or laparoscopic. The gynecologic surgeon should be not only technically adept at these various procedures, but also should use history, physical examination, and discussion with the patient to match the surgical procedure to the patient in order to obtain the most satisfactory outcome. HISTORY INDICATIONS FOR HYSTERECTOMY PREOPERATIVE COUNSELING PREPARATION FOR HYSTERECTOMY THE CHOICE OF APPROACH: ABDOMINAL, VAGINAL, OR LAPAROSCOPIC TOTAL ABDOMINAL HYSTERECTOMY: SURGICAL TECHNIQUE SUBTOTAL ABDOMINAL HYSTERECTOMY VAGINAL HYSTERECTOMY POSTOPERATIVE CARE COMPLICATIONS

HISTORY Part of "Chapter 31 - Hysterectomy" The history of hysterectomy is long and varied. Although significant advances in the technique of hysterectomy did not occur until the 19th century, earlier attempts are known. Some references to hysterectomy even date to the 5th century BC, in the time of Hippocrates. The earliest attempts at removal of the uterus were made vaginally for indications of uterine prolapse or uterine inversion. By the 16th century AD, a number of hysterectomies already had been done in Europe, including Italy, Germany, and Spain. In 1600, Schenck of Grabenberg cataloged 26 cases of vaginal hysterectomy. Vaginal hysterectomies were done sporadically through the 17th and 18th centuries. In 1810, Wrisberg presented a paper to the Vienna Royal Academy of Medicine recommending vaginal hysterectomy for uterine cancer. Three years later, the German surgeon, Langenbeck, successfully performed a vaginal hysterectomy for uterine cancer. The first vaginal hysterectomy performed in the United States was in 1829 by John Collins Warren at Harvard University; however, the patient expired on the fourth postoperative day. Three years following Warren's attempt in Pittsburgh, Herman and Werneberg successfully performed a vaginal hysterectomy for uterine cancer. By the late 19th century, techniques for vaginal hysterectomy were systematically studied and developed by Czerny, Billroth, Mikulicz, Schroeder, Kocher, Teuffel, and Spencer Wells. The earliest abdominal hysterectomy attempts usually involved uterine leiomyomas that had been misdiagnosed as ovarian cysts. In early 19th century, laparotomy for ovarian cysts still was considered dangerous, despite initial successes by McDowell in the United States and Emiliami in Europe in 1815. Abdominal hysterectomy for any reason was considered impossible to accomplish successfully. Many of the earliest myomectomies involved pedunculated tumors. Washington L. Atlee of Lancaster, PA performed the first successful abdominal myomectomy in 1844; although in a series of 125 surgeries, he did not attempt to remove the uterus. The first abdominal hysterectomy was attempted by Langenbeck in 1825. The 7-minute operation for advanced cervical cancer resulted in the patient's demise several hours later. Abdominal section was commonly P.800 complicated by postoperative hemorrhage that was often lethal. In the mid-19th century, Manchester, England surgeon, A.M. Heath, was the first to ligate the uterine arteries, but it would be nearly 50 years before his practice became more common. Successful surgery depends on control of bleeding, infection, and pain. Ligatures were known to be used to clamp bleeding vessels as early as 1090 and artery forceps were invented in the mid-16th century by Ambroise Pare. However, information regarding the pathophysiology of hemorrhage, shock, and blood transfusions was not available until the 20th century. The importance of infection control was first recognized by Austrian Ignaz Semmelweiss in his work with childbed fever. His 1840s work was furthered by Joseph Lister in the 1860s, and aided by notable discoveries by Louis Pasteur and Robert Koch. American Crawford W. Long first used ether as anesthesia in 1842, and Scotsman Sir James Y. Simpson initiated use of chloroform in his obstetric practice. Although the mentioned advances were made to decrease surgical and postoperative bleeding, it was not until 1864 that much success in this area was achieved. Frenchman Koeberle introduced his method of securing the large vascular pedicle of the lower uterus with his tool, the serrenoeud. This ligature en masse around the lower uterus with the corpus amputated above was the usual technique of controlling bleeding with hysterectomy in the earliest years. The stump thus formed was such a large mass of tissue that it could not always be safely returned to the peritoneal cavity owing to risk of intraperitoneal bleeding; often the stump was fixed extraperitoneally in the incision so that it could be clamped later if necessary. W.A. Freund of Germany further refined hysterectomy techniques in 1878 using anesthesia, antiseptic technique, Trendelenburg position, and ligature around ligaments and major vessels. The bladder was dissected from the uterus and the cardinal and uterosacral ligaments were detached; the pelvic peritoneum then was closed. Late in the 19th century, further refinements were made to abdominal hysterectomy by the Johns Hopkins Hospital, where they reduced their mortality to 5.9%. In the early decades of the 20th century, hysterectomy became more commonly used as treatment for gynecologic disease and symptoms. Gynecology as a specialty was developing and little else but surgery was available to gynecologists to help their patients. Major discoveries and concepts of reproductive organ physiology and pathology were just beginning. As surgery became safer, gynecologists concentrated on developing newer surgical procedures. Estrogen and progesterone were not discovered until the late

1920s and early 1930s. As gynecology matured as a specialty in medicine, knowledge of reproductive organ function and disease became more complete; special and more accurate diagnostic techniques were developed (e.g., the excretory urogram was developed in 1923); and effective nonsurgical methods of therapy were discovered. In the modern practice of gynecology, appropriate use of this knowledge and advanced modern diagnostic technologies allow more correct choices of treatment for complicated medical diseases. With proper use of blood transfusions and antibiotics, and with improvements in anesthetic techniques, a hysterectomy can be done fairly safely by the skillful gynecologic surgeon. Mortality from hysterectomy in most medical centers is one to two per 1,000. It is possible to report no mortalities in a series of several thousand hysterectomies. However, morbidity continues to plague the procedure. Complications may occur in as many as 25% of patients undergoing vaginal hysterectomy and 50% of those undergoing abdominal hysterectomy. Some complications may be serious (e.g., infection, hemorrhage, urinary tract and intestinal tract injury, and pulmonary embolus). It should be emphasized that the ability to perform several hundred hysterectomies with a low mortality and morbidity rate, although extremely desirable, is not ipso facto evidence that gynecologic surgery is being practiced correctly. Along with low morbidity and mortality rates, the physician also must be certain that only patients with proper indications are chosen for surgical treatment. To formulate proper indications for hysterectomy, the surgeon must have a thorough understanding of the physiology and pathology of the female reproductive organs, clinical manifestations of pelvic disease, and normal and abnormal psychosocial-sexual development. This knowledge and understanding is the absolute foundation on which the pyramid of successful practice of gynecologic surgery is built. The surgeon must use this knowledge and his or her own wisdom to determine if a patient is a candidate for surgical therapy, remembering that it is only the occasional gynecologic patient who requires surgery for relief. If the patient requires surgical intervention, she must be properly prepared for surgery, and the surgeon must properly perform the surgery and provide proper postoperative care. In the practice of gynecologic surgery, most mistakes are made by those physicians who may know how, but may not understand when and why. Referring to this point, the late Richard W. Te Linde, professor of gynecology at the Johns Hopkins University and the original author of this text, said: The ease with which the average hysterectomy may be done has proven both a blessing and a curse to womankind. There is no doubt that a hysterectomy done with proper indications may restore a woman to health and even save her life. However, in the practice of gynecology, one has ample opportunity to observe countless women who have been advised to have hysterectomies without proper indications.… I am inclined to believe that the greatest single factor in promoting unnecessary hysterectomies is a lack of understanding of gynecologic pathology. The greatest need today among those who are performing pelvic surgery is a better knowledge of gynecologic pathology. As Te Linde stated, many women have benefited from hysterectomy. However, considerable concern has been expressed that this same operation is overused and not always performed for proper indications. The medical profession has a strong tradition of accountability for P.801 the quality of patient care, evident in daily rounds and conferences, chart reviews, audits, hospital tissue and credential committees, specialty and subspecialty board certification and recertification, specialty society activities, rules and regulations of the Joint Commission on Accreditation of Hospitals, licensure, continuing medical education requirements, and so forth. Nevertheless, accusations of unnecessary operations, including unnecessary hysterectomies, still arise from unions, insurance companies, consumer groups, media, state and federal agencies, the US Congress, and members of the medical professions. A review by the Centers for Disease Control and Prevention (CDC) estimated that from 1988 to 1990, 1.7 million US women had a hysterectomy, an overall rate of 56.8 procedures per 10,000 women 15 years old and older. Strongly linked to age, the rate peaked at 100.5 hysterectomies per 10,000 women 30 to 54 years old. The 30- to 54-year-old age range accounted for 74% of all hysterectomies performed in the US. Approximately 20% of women 40 years old had had a hysterectomy, and 37% of women 65 years old. These numbers have declined since the 1970s. Authors of the CDC paper, Wilcox and colleagues, account for the shift in numbers with second opinions for surgery, quality assurance programs, higher levels of education, media campaigns, and alternatives to hysterectomy, such as medical therapies and more conservative surgical procedures. As we have seen in this brief historical review, hysterectomy has developed and evolved in the past 150 years from an extremely dangerous and infrequently performed operation that was almost abandoned to an important and major therapeutic modality that can save life and improve health assuming that proper

patient selection, preparation, and skillful performance are employed. Nevertheless, because of significant rates of mortality and morbidity in all age and diagnosis groups, hysterectomy cannot be considered a low-risk operation to be used for treating relatively minor gynecologic symptoms or disease. However, as demonstrated by Carlson and colleagues in the Maine Women Health Study, hysterectomy is highly effective for relief of symptoms associated with common nonmalignant gynecologic conditions. In this study, symptom relief following hysterectomy is associated with a marked improvement in the quality of life. A study by Clarke and associates yielded similar findings.

INDICATIONS FOR HYSTERECTOMY Part of "Chapter 31 - Hysterectomy" Because hysterectomy is the second most common surgical procedure performed in the United States, regulatory agencies, insurance companies, and the general public have studied its appropriate indications very extensively. In some cases, a pathologically normal uterus may be removed as part of an operation to treat stress urinary incontinence with pelvic relaxation, whereas in other instances hysterectomy is done as part of the surgery for endometriosis or in association with ovarian cancer. Many times, there is no histopathologic or anatomic abnormality of the uterus; some critics have been quick to label these procedures “unnecessary hysterectomies.” Although it is undoubtedly true that some hysterectomies are done that are not indicated and others are performed for marginal indications, most are done for sound indications to relieve present symptoms, prevent future symptoms, or save life. Surgeons need to constantly review their practices to be sure that they conform to currently accepted guidelines for hysterectomy. Although some tried and true indications such as endometrial cancer have been unwavering, other indications, such as sterilization or even menorrhagia and dysmenorrhea, remain highly controversial. The more common indications for hysterectomy are listed in Table 31.1; these and others are discussed in detail on the following pages. TABLE 31.1. Indications for Hysterectomy

Benign Disease

Malignant Disease

Abnormal bleeding

Cervical intraepithelial neoplasm

Leiomyoma

Invasive cervical cancer

Adenomyosis

Atypical endometrial hyperplasia

Endometriosis

Endometrial cancer

Pelvic organ prolapse

Ovarian cancer

Pelvic inflammatory disease

Fallopian tube cancer

Chronic pelvic pain

Gestational trophoblastic tumors

Pregnancy related conditions

Miscellaneous

Miscellaneous

Abnormal Uterine Bleeding Abnormal uterine bleeding can be caused by anatomic or pathologic abnormalities such as leiomyomas; polyps or cervical cancer; or hormonal imbalances leading to heavy, prolonged, or irregular uterine bleeding. In the absence of tumor, infection, pregnancy, or endometriosis, this type of bleeding generally is referred to as dysfunctional uterine bleeding. Although it was a common indication for hysterectomy in the past, today these symptoms usually can be managed by hormonal manipulation. The cause of bleeding can be evaluated by careful history, pelvic examination, endometrial biopsy, hysteroscopy, dilatation and curettage, and pelvic ultrasound. In women more than 40 years old or those with persistent bleeding, an endometrial biopsy or D&C is necessary to rule out endometrial hyperplasia or cancer. Pinion and colleagues conducted a randomized trial of hysterectomy, endometrial laser ablation, and transcervical endometrial resection for dysfunctional uterine bleeding. Hysteroscopic endometrial ablation was superior to hysterectomy in terms of operative complications and postoperative recovery. Satisfaction after hysterectomy P.802 was significantly higher, but between 70% and 90% of women were satisfied with the outcome of hysteroscopic surgery. According to these authors, hysteroscopic surgery can be recommended as an alternative to hysterectomy for dysfunctional uterine bleeding. According to Brooks and associates, greater familiarity with the technique of resectoscopic endometrial ablation, improved patient selection for the procedure, and the use of appropriate pharmacotherapy for suppressing endometrial growth prior to ablation probably substantially improve the rate of success, reduce procedural costs, and further enhance the cost advantage of this procedure over hysterectomy. Transvaginal endometrial resection requires advanced hysteroscopic training and experience, but hysteroscopic rollerball ablation and, more recently, microwave and thermal endometrial ablation techniques have been introduced that require less technical skill. In a prospective, randomized trial from Denmark, Boujida et al. found similar and significantly reduced bleeding at both 2- and 5-year follow-up in women treated with either endometrial resection or rollerball ablation. In this series, 15 of the patients required a second treatment for persistent bleeding, and 15% had undergone hysterectomy for bleeding and/or pain. Bain et al. compared the results of endometrial resection with microwave endometrial ablation in a prospective, randomized group of 249 women followed for 2 years. Average age was 42 years, and approximately 45% reported complete amenorrhea, whereas another 45% had decreased bleeding. About 20% reported similar or increased dysmenorrhea, and 12% of the women in each group underwent hysterectomy within 2 years of initial treatment. In this and other similar studies, persistent bleeding, dysmenorrhea, and pelvic pain resulted in diminished quality of life scores compared to the general, age-matched population. Alexander et al. reported higher levels of satisfaction and a better overall quality of life in women managed by hysterectomy compared to those treated by ablation or resection in the follow-up of Pinion's randomized study. Approximately 20% to 30% of hysterectomies are performed with dysfunctional uterine bleeding as the primary indication. Hysterectomy is not indicated unless the bleeding is recurrent, severe, and unresponsive to hormonal therapy and endometrial curettage on several occasions. Endometrial resection or ablation should be considered, especially for the patient 40 to 50 years of age. When hysterectomy is necessary, usually it can be done vaginally unless contraindicated. Recalcitrant dysfunctional uterine bleeding is not an indication for bilateral oophorectomy at the time of hysterectomy. Hysterectomy is very rarely required to treat dysfunctional uterine bleeding in young women.

Uterine Leiomyomas Uterine leiomyomas are the most common reason for hysterectomy, accounting for approximately 30% of the indication for this procedure. Abdominal or vaginal hysterectomy may be indicated if the myomas are associated with discomfort, urinary frequency or obstruction, menorrhagia, metrorrhagia, or if they are significantly increasing in size. Uterine leiomyomas are common benign tumors of the myometrium that

usually do not cause symptoms or require removal. The patient with small, asymptomatic uterine leiomyomas may be followed with periodic pelvic examinations. If the uterus is not too large, then a conservative approach with uterine curettage rather than hysterectomy should be tried first. If a submucous myoma is found by hysterography, hysteroscopy, or uterine curettage, then it is unlikely that conservative measures will relieve the patient's bleeding. Depending on the circumstances, a submucous myoma may be removed with the hysteroscope, by myomectomy, or with hysterectomy. If symptoms do not appear before menopause, then it is unlikely that a hysterectomy will be needed after menopause. No new myomas appear after menopause, and myomas already present cease to grow and usually become smaller. Management of uterine leiomyomas is not always surgical. Hormonal management with gonadotropin-releasing hormone (GnRH) analogs can be used alone or in conjunction with conservative surgery or even hysterectomy. Transcervical hysteroscopic myomectomy has been used for small submucosal myomas that cause abnormal bleeding. Recently, electrosurgical ablation and arterial embolization have been reported to provide good results in patients with larger intramural myomas. These techniques are discussed in Chapter 30. It is possible for a leiomyomatous uterus to be rather large and still cause no symptoms. When it reaches the size of a 12- to 14-week pregnant uterus, a recommendation for removal may be made based on size alone. However, Reiter and colleagues stated that hysterectomy need not be recommended to asymptomatic women with larger uteri. Yet, Hillis and associates present clear evidence that when the uterine weight exceeds 500 g, there is an increased risk of complications from abdominal hysterectomy. When deciding not to operate but to follow such a patient, the physician must be extremely careful that all components of the pelvic mass indeed are benign. There is always the possibility that one or more of the components of the mass is an ovarian tumor, the removal of which should not be delayed. Some idea of the various components of the mass may be obtained from pelvic ultrasound or computed tomography (CT). Although an abdominal approach usually is preferred when hysterectomy is indicated for large leiomyomas, many gynecologists still prefer vaginal hysterectomy with morcellation. Rapid enlargement of a leiomyomatous uterus may indicate sarcomatous change. However, according to Parker and associates, of 371 women operated on for rapid growth of the uterus, only one patient was found to have sarcoma. Among 198 patients who met a definition of rapid growth (an increase by 6 weeks gestational size over 1 year), two were diagnosed with endometrial stromal sarcoma. The incidence of leiomyosarcoma in patients presumed to have benign leiomyomata is approximately 0.2% to 0.7%. Leibsohn and colleagues, reviewed P.803 1,429 patients who underwent hysterectomy for presumed benign leiomyomata. A histologic diagnosis of leiomyosarcoma was made in seven (0.49%). The possibility that a benign leiomyoma may at some time undergo sarcomatous change is not an indication for hysterectomy. In young women, rapid growth of a leiomyomatous uterus is more commonly an indication of associated pregnancy. In this situation, a pregnancy test is indicated before hysterectomy is recommended. A pedunculated myoma protruding through the cervical os always is infected and should be removed by vaginal myomectomy to prevent unavoidable bacterial contamination of the peritoneal cavity with abdominal removal. This may be done by passing a pretied laparoscopic Endoloop around the pedunculated myoma and cinching it down before cutting the pedicle. Several weeks later, hysterectomy may be done with less risk of infection.

Adenomyosis Adenomyosis is an uncommon indication for hysterectomy. Dysmenorrhea, menorrhagia, uterine enlargement, and tenderness may indicate adenomyosis. Depending on the severity of symptoms, hysterectomy may be indicated after failure of response to more conservative measures, such as hormonal therapy and uterine curettage. The diagnosis may be suggested by magnetic resonance imaging, ultrasound, or CT scan, and core needle biopsy can be done with guidance under local anesthesia to confirm the diagnosis preoperatively. In many cases, a preoperative diagnosis of adenomyosis may not be confirmed by pathologic study of the removed uterus, and minimal adenomyosis found by careful pathologic study is probably of no clinical significance. In a study of validation of hysterectomy indications by Gambone and colleagues, adenomyosis represented the least verified indication, with only 38% of pathology reports providing confirmation. These authors stated that adenomyosis should no longer be considered a reliable indication for hysterectomy and recommended that patients suspected of having adenomyosis should be considered under other categories (i.e., chronic

pelvic pain or recurrent uterine bleeding). We do not agree that adenomyosis should no longer be mentioned as an appropriate indication for hysterectomy, but do agree that it should be an uncommon indication. Patients with dysmenorrhea, heavy, prolonged menstrual flow and an enlarged, tender uterus will experience a dramatic relief of symptoms from hysterectomy. Most should probably undergo attempts at hormonal management and D&C prior to hysterectomy. Ultrasound-directed core needle biopsy of the myometrium often provides a tissue diagnosis prior to hysterectomy.

Symptomatic Vaginal Relaxation, Uterine Descensus, and Prolapse Symptomatic descent or prolapse of the uterus, usually associated with symptomatic vaginal wall relaxation (i.e., cystourethrocele, enterocele, and rectocele), with or without stress urinary incontinence, is a common indication for vaginal hysterectomy with restoration of normal anatomy and proper support to the vaginal outlet, vaginal walls, and vaginal vault, accounting for approximately 15% of hysterectomies. It should be emphasized that this operation is indicated only if symptoms are severe enough to justify the risk involved; if conservative measures, such as Kegel exercises or vaginal estrogen, have failed to give sufficient relief of symptoms; and, most of all, if the patient is asking for relief. The use of pessaries to manage pelvic relaxation is discussed in Chapter 35. An operation to correct asymptomatic and mild to moderate anatomic relaxation of the vaginal walls and uterine descensus rarely is indicated. Too many surgeries are performed for this indication in asymptomatic patients. Hysterectomy is not a necessary part of surgical procedures to correct stress urinary incontinence. For example, if a suprapubic colpourethropexy is indicated to correct stress urinary incontinence, then hysterectomy is not indicated unless other conditions are present. Vaginal repair of a large symptomatic cystocele and enterocele usually is more successful if vaginal hysterectomy also is done at the same time.

Obstetric Problems Hysterectomy in obstetric practice is discussed in Chapter 32. Uncontrollable postpartum hemorrhage; uterine rupture; uterine inversion; placenta accreta; and interstitial, abdominal, or cervical pregnancy are obstetric catastrophes that may require abdominal hysterectomy to prevent death from hemorrhage. Septic abortion usually is treated successfully with antibiotics and evacuation of retained infected placental and fetal parts. Few patients are unresponsive to medical therapy and curettage. However, occasionally, the physician also may need to do an abdominal hysterectomy when there is severe infectious septic shock, peritonitis, and impairment of renal function perhaps associated with infection with clostridia microorganisms.

Associated Pelvic Conditions In the circumstances discussed in the sections that follow, the uterus may be excised as part of an operation to remove the diseased adnexa even though there is no pathology in the uterus itself. In the past, hysterectomy was recommended if the tubes or ovaries were irreparably damaged or required removal, rendering conception impossible. However, modern assisted reproductive technology has changed our thinking in this area, and the possibilities, risks, benefits, and costs should be discussed with these patients and their partners before surgery so they may make an informed decision about hysterectomy at the time of adnexal surgery for inflammatory disease, ectopic pregnancy, or neoplasm.

Pelvic Inflammatory Disease As discussed in Chapter 28, an acute exacerbation of chronic pelvic inflammatory disease (PID) with bilateral P.804 tuboovarian abscesses usually can be treated successfully with conservative medical management, including antibiotics for polymicrobial infection and drainage of pus when this is indicated by either posterior colpotomy or CT-guided needle aspiration. In our experience, at least 80% of patients respond to this management. Surgery may be recommended for those patients who do not respond, patients with a ruptured tuboovarian abscess, or patients with significant symptoms of chronic PID. Usually, a bilateral salpingo-oophorectomy is necessary because extensive adnexal disease is found bilaterally. Occasionally, it is possible to conserve the function of one ovary. Hysterectomy usually is recommended in women who are not interested in taking advantage of the new reproductive technologies for achieving pregnancy. For

those who are, the uterus may be left in place, even though both tubes and ovaries have been removed. Chronic pelvic pain often occurs in women with PID, and this may be more likely if an ovary or the uterus is left behind.

Pelvic Endometriosis Extensive pelvic endometriosis unresponsive to hormonal therapy may require surgery for relief. Young women who have not completed childbearing should have conservative surgery, including removal of all visible and palpable endometriosis, possible presacral neurectomy, and possible uterine suspension. For older women with no desire for pregnancy or for patients whose symptoms and pelvic findings of endometriosis have recurred following a previous conservative operation (or operations), a more extensive operation usually is indicated. Under these circumstances, abdominal hysterectomy with bilateral salpingo-oophorectomy is the procedure of choice unless the patient insists on the uterus remaining for the purpose of using new reproductive technologies to achieve conception. The management of pelvic endometriosis is discussed in Chapter 25.

Ectopic Pregnancy Certain ectopic pregnancies may require hysterectomy. A hysterectomy may be required for cervical pregnancies, interstitial pregnancies, or abdominal pregnancies when the placenta cannot be removed without removing the uterus. It also may be permissible to perform a hysterectomy in a patient who has had repeated tubal pregnancies that have completely destroyed both fallopian tubes to the point that their function cannot be restored and whose condition during surgery is completely stable. Again, it is necessary for the patient to give specific permission for hysterectomy after being fully informed of all options for future reproduction. The surgical management of ectopic pregnancies is discussed in Chapter 22.

Neoplastic Diseases Cervical Intraepithelial Neoplasia Under most circumstances, cervical intraepithelial neoplasia may be adequately treated by a loop electrode excision procedure (LEEP), cryosurgery, laser surgery, or cervical conization (see Chapter 45). Sometimes hysterectomy is recommended for patients who do not desire more children and have other indications for hysterectomy. Vaginal hysterectomy is preferred unless contraindicated by adnexal disease, uterine enlargement, or other circumstances. A vaginal cuff should be removed with the uterus when the carcinoma in situ extends close to or on to the adjacent vaginal fornix.

Early Invasive Cervical Cancer A simple, total hysterectomy is an adequate operation for early invasive cervical cancer if an adequate cervical conization examination shows superficial, microscopic stromal invasion limited to a depth of ?3 mm or a lateral spread of 0.001). CIN developed in nine women in the placebo group, whereas there was no instance of CIN among the vaccinated women.

Cytologic and Colposcopic Changes As viral replication takes place in the epithelium, it is accompanied by alterations that are often clinically evident and by changes in the cytology of the epithelial cells that allow such cells to be detected cytologically and histologically. Clinicians have observed that the earliest and most severe changes occur adjacent to the SCJ on the cervix. Anatomic studies have shown that the location of the SCJ is not fixed, but it changes during a woman's life. Under the influence of the vaginal pH, columnar cells on the exocervix are transformed into squamous cells by a process called metaplasia. There are three times in a woman's life that metaplasia is particularly active—in the embryologic development of the vagina and cervix, at puberty, and after the first pregnancy. During puberty and after pregnancy, the columnar epithelium of the endocervix is everted out onto the exocervix or portio of the cervix and metaplasia is stimulated. The columnar cells, starting with the epithelium on the tips of the glandular papillae, are changed into squamous epithelium. This process can be observed with the colposcope and the area of active metaplasia just lateral to the SCJ where this metaplastic change occurs is called the transformation zone—columnar epithelium is transformed into metaplastic squamous epithelia Figure 45.2. This epithelium can be recognized as a circumferential, pale white, translucent ring around the SCJ. FIGURE 45.2. Histologic section of squamous metaplasia. Immature squamous metaplasia may be acetowhite colposcopically and may be confused with cervical intraepithelial neoplasia histologically. Although the metaplastic cells are immature, they are not pleomorphic or cytologically atypical. The growth pattern and cytology are regular, and there is a lack of the features that characterize human papillomavirus–related lesions. This section is typical of the epithelium from the immature transformation zone.

View Figure

These metaplastic cells are very active and may be infected by HPV in the vagina. As described earlier, certain oncogenic HPV types may infect and transform the epithelial cells and the increased mitotic activity and DNA density can be identified as “white epithelium” following application of 3% to 5% acetic acid to

the cervix. As CIN develops, vascular changes occur and these may be recognized as “punctation” or “mosaic.” If the lesion progresses, bizarre vascular changes can be seen. These “atypical vessels” have grossly abnormal architecture and are colposcopic hallmarks of invasive cancer. Women whose initial low-grade CIN lesions are caused by HPVs with a low oncogenic risk or no oncogenic risk either maintain an infected epithelium with persistent disease for long periods of time or, in most instances, seem to resolve the HPV infection over time without treatment. Lesions that are caused by HPV types with an intermediate or high oncogenic risk may undergo remission spontaneously, persist, or, in a minority of cases, progress. It is not well understood what influences a lesion to undergo remission, persist, or progress, but indirect clinical evidence suggests that HPV type, host immunologic factors, and cofactors such as other STDs may play a role.

Cellular Transformation In patients whose lesions progress from a low-grade CIN to a high-grade CIN, the progression is accompanied by alterations in the virus-host interaction. These interactions appear to be critical to the transformation of normal epithelial cells to neoplastic cells. The key alteration seems to be the integration of the viral DNA into host chromosomes. Because the episomal (nonchromosomal) virus that is found in latently or productively infected cells is in a circular form and because the DNA in the host chromosomes is linear, a change in the physical state of the virus must take place before the viral DNA can be integrated into the host chromosome. What initiates this change and the mechanism by which it occurs is not yet understood, but the circular strand of viral DNA is opened and the linear DNA can be spliced into the host DNA. It can then use the protein production apparatus of the transformed host cell. Disruption of the viral genome at the E1-E2 ORFs appears to inactivate E2 expression. Because the E2-encoded proteins regulate the transcription of the early region and ORFs, the loss of E 2 control allows for the unregulated or overexpression of the E 6- and E7 -encoded proteins. The E6 and E7 proteins are highly oncogenic. They appear to disrupt the normal process of cellular replicative control by binding to key proteins that help regulate cell division. The E6 protein binds to the p53 protein and causes its inactivation and more rapid degradation. The p53 protein is an important negative growth regulator that plays a key role in the control of cell division. The E7 protein binds to cyclin A p107 and the retinoblastoma gene product, p105RB. The p105RB proteins are responsible for controlling the progression of cells from the G1 phase of the cell cycle to the S phase. In tissue culture models, E6 - and E7-encoded proteins have been shown to be capable of transforming normal epithelial cells to neoplastic cells, and it has further been reported that both E6 and E7 expression are necessary to maintain the transformed state. The oncogenic proteins encoded for by these viral genes appear to complement each other and are only weakly active when present alone. HPV viruses with a high oncogenic potential (e.g., HPV types 16 and 18) produce E6- and E7-encoded proteins that have a transforming potency greater than that of proteins produced by HPVs with an intermediate oncogenic risk. The E6 - and E7-encoded protein produced by HPV types with little or no oncogenic risk (6 and 11) seem to be incapable of transforming cells in vitro. This feature probably accounts for the differences in risk rates between the viral subtypes (Table 45.3). The importance of E6 and E7 in maintaining the transformed cell state has been demonstrated by inserting an antisense message for these ORFs in tissue culture cells. Such insertion is capable of reversing the transformed phenotype. Simple deregulation of the mitotic process appears to be insufficient for producing neoplastic transformation. Recent data suggest that mutations in other cellular genes whose proteins are important in regulating the mitotic process are required for transformation to occur. It is now evident that the process of malignant transformation usually involves five to six such mutations. The p53 antioncogene is responsible for ascertaining that the genome is intact after DNA replication before the cell is allowed to divide. If DNA coding errors are found, they are repaired and the cell is allowed to progress to and complete cell division, or, if repair is not possible, the cells are not allowed to divide and instead a process of programmed cell death (apoptosis) is initiated. E6 and E7 oncogenic proteins are produced, which bind to and inactivate p53 so that it can no longer carry out the function of surveillance and quality control over DNA replication. Thus, the epithelial cells, even with chromosomal mutations are able to complete repeated mitotic cycles. It is probably that the accumulation of mutated genes is required for the cervical epithelial cell to become fully transformed. These alterations that occur at the molecular level result in changes in the chromosome

number, in mitotic abnormalities, and in morphologically identifiable abnormalities in the chromosomes and eventually result in the production of an aneuploid cell population. This process of malignant transformation probably starts at the point at which HPV is integrated into the host epithelial cell chromosomes.

The Role of Immunosuppression in HPV-Related Lesions It is clear from multiple clinical observations that men and women who are immunosuppressed have a higher risk of developing HPV-related lesions of the anogenital tract. It is probably because of relative immunosuppression that some pregnant women develop vulvar condylomata that then undergo remission postpartum without any treatment having been administered. Similarly, women who undergo renal transplantation or women with chronic immunosuppression have a high risk of developing HPV-related lesions and a high risk of developing lower genital tract cancers. These data support the concept of HPV latency and argue for the strong role of an intact immune system in controlling HPV infections. More recent observations of men and women who are immunosuppressed secondary to infection with HIV confirm the importance of immunocompetence in suppressing latent HPV infection and in facilitating the acquisition of HPV infection. Not only are HPV-related lesions more common in immunosuppressed individuals, but they tend to be larger, more often multicentric, and more likely to recur after treatment. A number of studies of women who have acquired immunodeficiency syndrome (AIDS) have confirmed that this group of individuals is at increased risk for developing HPV-related lesions. In one of the first studies of women with AIDS, it was reported that in a group enrolled in a methadone maintenance program in New York, 14 of 33 (42%) HIV-infected women with AIDS had abnormal Papanicolaou tests compared with 3 of 18 (17%) HIV-infected women without AIDS and 6 of 65 (13%) HIV seronegative women. In a more recent study, the prevalence of biopsy-proven CIN was five times higher in HIV seropositive women and increased progressively as the CD4 T-lymphocyte count decreased. These observations support the model proposed earlier by Stafl and Mattingly in which the clinical development of invasive cervical cancer involves a struggle between the oncogenic effects of the HPV infection and the immunologic defense mechanisms of the woman.

SCREENING FOR CERVICAL NEOPLASIA Part of "Chapter 45 - Cervical Cancer Precursors and Their Management"

Cervical Cytology After the introduction of cervical cytology for cervical cancer screening more than 50 years ago, multiple screening programs from all parts of the world have reported decreased rates of invasive cervical cancer and decreased death rates from a malignancy that had previously been the number one worldwide cause of cancer deaths in women. Yet, even today cervical cancer remains one of the leading causes of death for women in developing nations where Papanicolaou smear screening programs are nonexistent. The cervical screening program in the Province of British Columbia, Canada, is one of many examples of how effective cervical cytology screening can be. Since the implementation of a population-based Papanicolaou smear screening program in 1949, the incidence rate of invasive cervical cancer has decreased from more than 30 per 100,000 women to less than 5, and the mortality rate has decreased from 12 to about 3 per 100,000 (Fig. 45.3). These excellent results have been obtained through a combination of a population-based program with a central registry that notifies women every 3 years that their screening visit is due and follows up to be sure they do not default (a “call and recall” system), an excellent central cytology laboratory and well-trained clinicians to obtain the smears, and a network of colposcopy follow-up clinics where women with abnormal results are evaluated and treated. The program is well organized with quality control built in at every level and funding for this preventive health service and any follow up and treatment required is universally provided to all residents of the province. All of these elements are required to achieve excellent results from a screening program. FIGURE 45.3. Incidence and mortality rate in women older than 20 years. British Columbia, 1988. (From Benedet JL, Anderson GH, Matisic JP. A comprehensive program for cervical cancer detection and management. Am J Obstet Gynecol 1992;166:1254–1259.)

View Figure

Despite the effectiveness of this and other cervical cytology screening programs, there are several limitations of Papanicolaou smear screening. A single Pap smear has a sensitivity of only about 50% to 60%. This means that a single test will not detect a cervical lesion in many women. However, even with this limited sensitivity, if three consecutive tests are negative, there is less than a 1% chance that the patient will have a cervical abnormality. A false-negative Papanicolaou smear may result from either screening or interpretation problems. Screening problems include lesions that do not shed cells or that are not sampled by the clinician. Or, sometimes, the diagnostic cells are not transferred from the spatula or collection device to the glass slide. Rarely, the slide preparation or staining is unsatisfactory. In other patients, problems with interpretation include failure to identify abnormal cells or misinterpretation of cells that are diagnosed as reactive or metaplastic when a dysplastic lesion exists. Various studies have shown that women who are diagnosed with invasive cervical cancer after a reportedly “negative” Papanicolaou smear most often have abnormal cells on review of their slides. The diagnostic cells may be few in number or obscured by blood or inflammatory changes. It is unfortunate that litigation over the “missed diagnosis” of cervical neoplasia is such a common problem for cytology laboratories in the United States.

The threat of a law suit may cause cytologists and cytopathologists to “overcall” a diagnosis and the gynecologist to recommend colposcopy and possibly cervical biopsy in a patient with any hint of abnormality. This not only increases the anxiety and morbidity associated with cervical cancer screening but it also increases the costs.

Liquid-Based, Thin-Layer Cytology To decrease the false-negative rate of cervical cytology, attempts have been made to improve both specimen collection and quality and to reduce errors of interpretation. Over the past several years, several liquid-based techniques have been approved by the Food and Drug Administration in the United States. These techniques differ from the conventional method of Papanicolaou smear collection in several ways. Once the clinician obtains a scraping of the SCJ and transformation zone area of the exocervix, the spatula and brush are dipped and agitated in a small bottle of fixative solution to elute the cells rather than being smeared on a glass slide. The small bottle is labeled and sent to the cytology laboratory rather than sending a slide. Once in the lab, a machine prepares a slide containing about 40,000 representative epithelial cells in a thin layer. The slide is then stained with the conventional Papanicolaou stain and reviewed by the cytologists and cytopathologist. Several studies have shown that more diagnostic cervical cells are collected by this liquid-based technique and that the slides prepared from this sample provide a “cleaner” appearance with less clumping, blood, and inflammatory changes, thus providing a better specimen and hopefully a better interpretation. In a review, Dunton found that thin-layer, liquid-based cytology resulted in significantly fewer unsatisfactory and “satisfactory, but limited by” specimens and that more patients were diagnosed with both low- and high-grade lesion. The agency for Healthcare Research and Quality evaluated the literature available and concluded, “The relative true-positive rate for ThinPrep compared to conventional cytology is 1.13, suggesting a modestly higher sensitivity with ThinPrep, the relative false-positive rate is 1.12 suggesting a modestly lower specificity.” This new technology costs more money than the conventional Papanicolaou smear, but theoretical analysis suggests that these costs are balanced by the possibility of less frequent screening intervals made possible by a more sensitive test. Liquid-based cytology reduces the rate of false-negative results from both screening and interpretation errors.

Computer-Assisted Diagnosis It has long been believed that optical scanning by computer could be used for Papanicolaou smear interpretation but the differences in staining and the overlap of cells has made practical application very difficult. However, in the past several years, two different computer-based systems for cervical cytology use have been approved by the Food and Drug Administration. There are relatively few data published in the one system, which is commercially available in the United States. These techniques have been used mostly for quality control to identify slides that have been read as normal by cytotechnology screeners but that have cellular characteristics recognized by the computer as suspicious. In a primary screening trial, the AutoPap system was used to grade cervical cytology and the results compared with conventional screening. The computer “graded” each slide based on a number of characteristics and the 25% lowest scoring, “most normal” slides were diagnosed as within normal limits and received no further review. The other 75% were manually screened and the computer was used to identify 15% of the most suspicious for a second, manual rescreening. More than 25,000 slides were analyzed in this side-by-side comparison trial. Of the 70 cases of HSIL or greater, the computer-assisted arm identified 68 (including both invasive cancers) whereas the conventional screening arm identified 65 of the 70 patients with high-grade lesions. There was no statistically significant difference between these two arms. There is good evidence to suggest that computer-assisted diagnosis is even better when liquid-based, thin-layer slides are used because the background is much cleaner and there are fewer cell clumps with diagnostic cells obscuring one another. Although the initial cost of a computer-based system is large, considerable savings should be realized by the around-the-clock work and a reduced need for cytotechnologists who could concentrate on the diagnostic evaluation of high-risk slides identified by the computer on primary screening. More clinical experience with these computer techniques is needed but early results are promising.

Human Papillomavirus Testing With the knowledge that cervical neoplasia is almost always (possibly always) associated with HPV, it was suggested that HPV testing could be used to screen women for cervical neoplasia. HPV testing in the past has been inaccurate and the complex laboratory techniques are not conducive to large volume clinical work. However, with the newer second-generation hybrid capture techniques, these problems seem to have been solved. Current technology uses DNA hybridization and quantification by a chemiluminescence reaction to identify the presence of any of 11 different, high- and intermediate-risk HPV. In a trial of 1,119 high-risk women in Costs Rica, Schiffman and colleagues demonstrated an 88% sensitivity to detect high-grade lesions compared with a Papanicolaou test sensitivity of 77.7%. Twelve percent of the women screened were HPV positive and were referred for colposcopy as compared with a 6.9% referral rate with the less sensitive Papanicolaou smear screening. In a similar trial from South Africa, Wright and colleagues instructed patients on self-collection of vaginal samples and were able to detect high-risk HPV DNA in 66% of the patients with high-grade lesions. This was almost identical to the 68% pickup rate of abnormal cytology collected at the time of a pelvic examination by a clinician. High-risk HPV was detected in 84% of the specimens obtained by the clinician during a pelvic examination. In high-risk, sexually active populations, the problem of “false positive” HPV tests may be difficult because a certain number of women with a subclinical HPV infection clear the virus without cervical neoplasia ever developing. In the South African trial, 15% of the women who tested positive for HPV had no cervical lesion. Nevertheless, HPV testing is easy to perform and is a relatively inexpensive test that can be automated in the laboratory and requires no interpretation (a problem with the Papanicolaou test). These advantages must be weighted against its lack of specificity for high-grade lesions; however, it may be of particular benefit in resource-poor settings where cytology laboratories are not available. At the present time, HPV testing is being more widely used as a triage step in women with ASCs on Papanicolaou tests. This is discussed later in this chapter.

EVALUATION OF ABNORMAL CYTOLOGY Part of "Chapter 45 - Cervical Cancer Precursors and Their Management" Following the announcement of the 2001 Bethesda System terminology for reporting the results of cervical cytology, the American Society of Colposcopy and Cervical Pathology sponsored a 3-day consensus conference to develop management guidelines. These are very detailed and carefully crafted to include almost all situations and were published by Wright and colleagues in 2002. It is important for the clinician to distinguish between cytologic or Papanicolaou smear abnormalities and biopsy or histologic abnormalities. In this section, the evaluation of cytologic abnormalities are discussed; in the following section, the management of patients with biopsy-proven abnormalities is presented. In both sections, high-grade abnormalities are discussed first because they are the least controversial and most important. Treatment rarely is initiated based on the results of cervical cytology alone. The Papanicolaou smear result, colposcopic impression, and biopsy diagnosis should all be in general agreement, although it is common that they may not agree precisely on the grade of abnormality.

High-Grade Squamous Intraepithelial Lesions Approximately 0.5% of all Papanicolaou smears are diagnosed as HSIL (Fig. 45.4). When these patients are evaluated, more than 70% will have biopsy-proven moderate dysplasia (CIN 2) or worse, and 1% to 2% will have an invasive cancer. Because high-grade abnormalities may progress to invasive cancer or, indeed, be associated with an already existing invasive cancer, women with an HSIL Papanicolaou test should be referred for colposcopic evaluation. FIGURE 45.4. Papanicolaou smear showing high-grade cervical intraepithelial neoplasia. Normal superficial cells contain small, pyknotic, regular nuclei. Abnormal cells exfoliated from high-grade lesions are of the neoplastic basal and parabasal types, have an altered nuclear:cytoplasmic ratio, and contain very hyperchromatic, irregularly shaped, pleomorphic nuclei with a dense, abnormal chromatin pattern.

View Figure

In many cases, biopsy confirmation of the lesion will be done by loop electrode excision procedure (LEEP) of the whole transformation zone which serves as both a diagnostic and a treatment procedure. This is particularly important in large, high-grade lesions because the colposcopic diagnosis of microinvasive cancer is difficult. Unless excision and histologic study of the whole transformation zone is done, microinvasive cancer may be missed. If a woman with an HSIL Papanicolaou smear does not have a high-grade lesion identified on colposcopy, an endocervical curettage (ECC) should be done. If the origin of the HSIL Papanicolaou finding is still not identified, the clinician should request a review of the cytology, ECC, and any cervical biopsies. If the cytopathologist still believes that a high-grade lesion exists that has not been identified by biopsy, then the cervix and vagina should be reexamined and a LEEP or cone biopsy done, especially if the SCJ cannot be visualized.

Low-Grade Squamous Epithelial Lesion Of all Papanicolaou smear diagnoses, about 1.5% to 2.0% are LSIL. Between 15% and 30% of these women have moderate dysplasia or worse, and 1 to 2 per 1,000 have invasive cancer (Fig. 45.5).

Although the risk of a significant lesion is lower in these women, referral for colposcopic evaluation is still recommended. FIGURE 45.5. Papanicolaou smear showing low-grade cervical intraepithelial neoplasia. A cluster of superficial squamous cells containing hyperchromatic, atypical nuclei is seen. Several cells are binucleated. The cells with perinuclear clearing and nuclear atypia are referred to as koilocytes. These changes are characteristic of a productive human papillomavirus infection of the genital tract mucosa.

View Figure

There are many treatment options for women with mild dysplasia (CIN 1) on colposcopically directed biopsy. But because of the wide possibility of diagnosis in women with an LSIL Papanicolaou finding, colposcopic evaluation and biopsy confirmation of the diagnosis is indicated. In a patient with no visible lesion on the exocervix and vagina to explain the abnormal Pap test, an endocervical curettage is indicated. When no lesion to explain the LSIL Papanicolaou finding can be found, the evaluation can be repeated in 4 to 6 months, a LEEP could be done, or consultation with a more experienced colposcopist could be requested. In postmenopausal women, atrophy can cause minor cellular atypia, which can produce an LSIL Papanicolaou finding, so treatment with topical estrogen for 2 months may be helpful before repeating the cytologic study.

Atypical Squamous Cells A diagnosis of ASC is the most common abnormal Papanicolaou diagnosis, occurring in 2.5% to 5.0% of all Papanicolaou smears. Although high-grade lesions are present in only 5% to 10% of these patients and because of the relatively large number of women with these ASC diagnoses, more women with a biopsy diagnosis of moderate dysplasia or worse present with an ASC Papanicolaou than any other Papanicolaou diagnosis, including HSIL. Even so, immediate referral to colposcopy for all these patients is not cost effective because there are so many women with no significant cervical abnormalities. Another option is to bring these patients back for repeat Papanicolaou testing in 4 to 6 months. This can lead to a delay in diagnosis, and it is not unusual for 20% or more of these patients to be lost to follow-up, producing an even greater delay. It is probably not of great importance if a progression from mild to severe dysplasia occurs during this follow-up or even lost to follow-up delay because treatment of various degrees of intraepithelial neoplasia is similar, but occasionally invasive cancer develops during this period. It has been estimated that there are about 1.25 million women each year in the United States with an ASC Papanicolaou finding. Even if the risk of cervical cancer is very low, with such a large number of women at risk, invasive cancer will already have developed or will develop during a follow-up management protocol. The third management option for women with an ASC-US Papanicolaou finding is triage by HPV testing. Several large prospective studies have demonstrated the effectiveness of this approach. In a National Cancer Institute–sponsored randomized trial, Schiffman and colleagues randomized a group of 4,500 women with ASC-US Papanicolaou findings to immediate colposcopy, follow-up every 6 months until two consecutive ASC-US Paps, or HPV triage. All women were eventually colposcoped and treated by LEEP at 24 months if they had not already been diagnosed and treated for CIN earlier. In this study, HPV testing identified 95% of the patients with CIN 3 or greater and required colposcopy in only 55% of the patients. The follow-up arm was able to identify only 85% of the patients with a high-grade lesion and referred 60% for colposcopy. Even though the immediate colposcopy group should diagnose almost 100% of the CIN 3 lesions, it is very expensive in the United States where colposcopy charges are three to four times greater than an HPV test.

Patients with a Papanicolaou diagnosis of ASC-H should be referred for colposcopy because they have a 15% to 35% risk of having moderate dysplasia or worse. Even if colposcopic evaluation is negative, including an ECC, these patients need to be followed up carefully. The value of HPV testing in these patients is unproven; such patients usually test positive for high-risk HPV.

Glandular Abnormalities In the 2001 Bethesda System, there are three categories plus adenocarcinoma under the classification of glandular cell abnormalities. These include atypical glandular cells (AGC), atypical glandular cells—favor neoplastic, and endocervical adenocarcinoma in situ (AIS). The cytologist should state if possible whether the glandular cells are endocervical, endometrial, or not otherwise specified. While it may be helpful for the clinician in the future, the experience with the different categories is still limited and it remains to be proven whether cytology can reliably differentiate between glandular dysplasia and AIS—and, indeed, between AIS and invasive adenocarcinoma. Therefore, all patients with abnormal glandular cells of Papanicolaou smear should be referred for further evaluation. Follow-up studies show that 35% to 50% of these patients have a lesion and that most have squamous not glandular CIN (Fig. 45.6). A few have endocervical glandular dysplasia or AIS and a few have endometrial hyperplasia or adenocarcinoma. Endocervical glandular neoplasia is also HPV related so that it is often accompanied by a squamous dysplasia. For this reason, the diagnosis of moderate squamous dysplasia on the exocervix in a woman with an AGC Papanicolaou smear does not rule out an accompanying glandular lesion so that the workup for a possible glandular abnormality needs to be pursued. The association of HPV, especially type 18, suggests that HPV testing might also serve as a useful triage step in the evaluation of women with AGC findings, but this concept has not yet been proven clinically. FIGURE 45.6. Papanicolaou smear showing adenocarcinoma in situ. A cluster of atypical endocervical cells with a feathered border and cytologically atypical nuclei characteristic of adenocarcinoma in situ is seen. Cells exfoliated from adenocarcinoma in situ tend to occur in clumps, which should raise the suspicion that a glandular neoplasm may be present.

View Figure

Evaluation of the patient with an AGC Papanicolaou finding usually involves a colposcopic examination and an endocervical curettage, whether or not the SCJ can be visualized. If the ECC is positive for a glandular abnormality, a cone biopsy is indicated. Even though some experts are able to adequately evaluate endocervical glandular lesions with a LEEP specimen, we find that a LEEP often provides a specimen that is not deep enough and the deep endocervical margin is positive for glandular dysplasia or AIS. We recommend a traditional, deep, wide, cold knife cone biopsy under anesthesia in the operating room for the evaluation of glandular abnormalities. If the endocervical curettage is negative and the colposcopic examination is normal, the Papanicolaou diagnosis should be reviewed. Management options include a repeat Papanicolaou with or without colposcopy and ECC in 4 to 6 months or possible cone biopsy if suspicion of a cervical lesion is high. These general guidelines should be helpful in most cases, but there are many special situations, including the evaluation of pregnant women, adolescents, and postmenopausal patients, in which alternative management plans may be more suitable. More extensive management guidelines are available in the JAMA article by Wright and colleagues or on the American Society for Colposcopy and Cervical Pathology Website: www.asccp.org.

Abnormal Papanicolaou Finding in Pregnancy The pregnant woman with an abnormal Papanicolaou finding generally is evaluated by colposcopic examination and directed biopsy if a high-grade lesion is suspected. The goal of evaluation of the pregnant patient with CIN is to rule out invasive cancer. CIN is not treated in pregnancy but is followed up until the postpartum period when the patient usually is reevaluated and managed as indicated by the biopsy results and her social situation. During pregnancy, most clinicians do a cervical biopsy to confirm the diagnosis only if a high-grade lesion is suspected, although some recommend a confirming biopsy in all patients with a visible lesion. The pregnant cervix is very vascular especially in the later trimester, so a biopsy may bleed profusely and cervical trauma can initiate uterine contractions. An endocervical curettage is contraindicated in pregnancy. Rarely, a cone biopsy is required in pregnancy to rule out invasive cancer when colposcopy with directed biopsy does not eliminate that possibility. This can be a difficult and potentially bloody procedure and should be undertaken by an experienced gynecologist only after all other avenues to rule out invasion have been exhausted.

MANAGEMENT OF CERVICAL INTRAEPITHELIAL NEOPLASIA Part of "Chapter 45 - Cervical Cancer Precursors and Their Management" The management of CIN or cervical dysplasia is almost always based on a colposcopically directed biopsy of the lesion rather than just the Papanicolaou or cytologic diagnosis or even the colposcopic impression. If there is a significant discrepancy between these diagnoses, the clinician should review the whole diagnostic evaluation to be sure cervical cancer is not being undertreated on the one hand or benign cervical inflammation or metaplasia is not being overtreated on the other hand. In addition to the biopsy diagnosis of the cervical lesion, the social situation of the patient often plays an important part in management decisions. For example, a 32-year-old mother of three with a three-quadrant CIN 1 lesion should probably be treated with LEEP or cryotherapy, whereas the same lesion in a 21-year-old unmarried student might well be watched for 12 months or so in hope that it will regress and no cervix will have to be removed or destroyed.

Treatment Indications It is difficult, if not impossible, to write out a complete set of guidelines that encompasses every clinical situation. The preceding sections on the natural history of the transformation zone and CIN need to be applied to the specific patient and the various management options reviewed in order to select the best plan for that patient. In many cases, the patient may be helpful in the decision to treat or follow a lesion. Insurance coverage, school vacations, and equipment availability could all influence the choice or timing of treatment.

CIN 1—Mild Dysplasia Reviews by Oster and more recently by Nuovo and colleagues have shown that regression occurs with no treatment in about 50% of women with a cervical biopsy showing CIN 1 during 12 to 24 months of follow-up (Table 45.3). Although good data do not exist, larger lesions and lesions that have persisted for a longer time are probably less likely to regress spontaneously. Because the treatment options are the same and chance of complete cure approaches 100% for CIN 1 and CIN 2 or 3, the consequences of progression are minimal. It is, therefore, reasonable to manage a patient with CIN 1 by observation alone. Follow-up without biopsy-proven CIN 1 is not recommended because cytology and colposcopy are not sufficiently accurate to confidently rule out a more severe lesion (Fig. 45.7). The only two risks to this approach are that invasive cancer already exists and has been missed on the diagnostic evaluation of this patient or that the patient becomes lost to follow-up and does not reappear until her condition has progressed to invasive cancer. These risks are minimal, but in almost all series in which women with low-grade cervical abnormalities have been followed up with no treatment, invasive cancer has eventually been diagnosed in a few patients.

FIGURE 45.7. Histologic section of low-grade cervical intraepithelial neoplasia. A productive human papillomavirus infection in the cervical epithelium is characterized by cytologic atypia and hyperchromasia (heteroploidy), binucleated cells (lack of cytokinesis), and koilocytosis (binding of filaggrin). No other agent produces these changes, which are diagnostic.

View Figure

For this reason and because treatment is relatively easy with limited patient inconvenience, side effects, complications, and costs, CIN 1 lesions in women who have completed their families who have three- or four-quadrant lesions or who have had a lesion for 12 months or longer should be considered for treatment. These low-grade abnormalities have cure rates of 90% or better with LEEP, cryotherapy, or CO2 laser vaporization. Traditional cold knife cone biopsy is probably associated with more morbidity and cost than is warranted by CIN 1. Although patients with “atypical metaplasia” or chronic cervicitis have been treated with cautery or cryotherapy in the past, these are not generally indications for treatment under modern guidelines.

CIN 2 or 3—Moderate or Severe Dysplasia, Carcinoma in Situ Patients with a biopsy diagnosis of CIN 2 or 3 have a less than 50% chance of regression and a significant chance of progression to carcinoma in situ (Fig. 45.8), so treatment is usually recommended. Pregnant women and, occasionally, very young women or those who wish to retain their fertility and have been previously treated for CIN may be followed up conservatively. FIGURE 45.8. Histologic section of high-grade cervical intraepithelial neoplasia. This lesion has some features of low-grade cervical intraepithelial neoplasia but also contains abnormal mitotic figures—a diagnostic feature of an aneuploid, high-grade, human papillomavirus–related lesion.

View Figure

Although treatment of these lesions can be successful using either ablative or excisional techniques, high-grade lesions bordering on invasive cancer are best managed by excision, which makes a tissue

specimen available for examination by the pathologist. In several large series of LEEP specimens, a few invasive cancers that had been unsuspected were diagnosed on histopathologic examination of the whole transformation zone. Cryotherapy or laser vaporization of such lesions would be inappropriate and could result in a deeply invasive cancer going untreated for several months or even several years. A cold knife cone biopsy in the operating room under anesthesia is indicated when cytology, colposcopy, or directed punch biopsy suggests superficially invasive cancer. Negative margins and a large specimen in one piece that can be well oriented is essential for an accurate and complete diagnosis of microinvasive cervical cancer. A final diagnosis of microinvasive cervical cancer cannot be made on punch biopsy or endocervical curettage. It is necessary to examine tissue on all sides of the microinvasive focus to be sure more deeply invasive disease does not exist adjacent to the focal area sampled by punch biopsy or ECC (Fig. 45.9). FIGURE 45.9. Histologic section of microinvasive squamous cell carcinoma. In microinvasive cancer, tongues of neoplastic epithelium extend through the plane of the basal lamina to split collagen bundles as they penetrate the cervical stroma, creating an irregular outline. Early invasion is commonly accompanied by a local inflammatory response.

View Figure

Adenocarcinoma in Situ or Glandular Dysplasia These diagnoses also require a cone or large biopsy of the endocervical epithelium and stroma with negative margins for diagnosis. An endocervical curettage or punch biopsy are inadequate to rule out a more extensive, possibly invasive, lesion. As previously noted, I favor cold knife cone biopsy for the evaluation of glandular lesions. The standard of care for AIS of the cervix in the United States is still simple hysterectomy, although there is some increasing interest in the use of cone biopsy alone for these patients. If cone biopsy is used as definitive management of AIS, the margins should be negative and enough sections of the cervix need to have been studied to be confident that an early invasive adenocarcinoma does not coexist. Glandular dysplasia can also be managed by cone biopsy if the margins are clear, but there are few data on this subject.

Treatment Techniques In general, treatment techniques can be divided into ablative and excisional. Destructive methods such as cryotherapy, CO2 laser ablation, and electrocautery rely on an accurate colposcopically directed biopsy diagnosis because no tissue specimen is provided by the treatment procedure. On the other hand, punch biopsy, LEEP, and cone biopsy all provide a specimen so these techniques are both diagnostic and therapeutic.

Punch Biopsy as Therapy In most developed countries, women are highly screened; invasive cancers are uncommon, particularly in younger age groups; high-grade CIN is relatively uncommon; and most lesions that are detected are

low-grade CINs found as incident cases. Because most of the cases of CIN are low-grade lesions that have developed between screening visits, the lesions tend to be small, to be located at the SCJ, and to occur in younger, often nulliparous women. For such patients, simple removal of the lesion with one or two punch biopsies or a small, segmental resection with an electrosurgical loop is adequate. Theoretically, the entire transformation zone should be destroyed or removed, but in some patients, these small biopsies clear the abnormalities.

Cryotherapy Cryosurgery is a destructive technique that was introduced to gynecologists in the late 1960s to treat CIN. The cryosurgical instruments use nitrous oxide or carbon dioxide as a refrigerant to lower the temperature of the tissue below -22°C and to produce cell death by intracellular and extracellular water crystallization. The refrigerant is applied to the cervix with a cryoprobe, which is placed in contact with the cervical epithelium. As the gas is circulated through the cryoprobe, it withdraws heat from the cervix until freezing temperatures occur. The cervix and cryoprobe generally reach a steady state after about 3 minutes of freezing, at which time the amount of heat brought to the cervix by the vascular supply balances the amount of heat withdrawn from the cervix by the evaporating cryogenic gas. Tens of thousands of patients have been treated with cryotherapy, which has proved to be a predictable, reliable treatment technique with limited side effects and morbidity. Cryotherapy is used principally in patients whose lesions are confined to the exocervix because the depth of cryodestruction seldom exceeds 5 mm. The cure rates are dependent on the size of the lesion. They generally average about 90% for lesions involving one quadrant, but only 75% for lesions involving three or more quadrants. The cryocurability is not related to the histologic grade of the lesion, except as histologic grade influences lesional size and distribution. Opinions differ as to whether a single freeze-thaw cycle or a double freeze-thaw cycle is the optimal approach to cryotherapeutic treatment, but there appear to be no differences in cure rates between the two approaches. The most important end point in cryotherapy is that the ice ball should extend well past the edge of the lesion. A minimum of 5 mm margin is recommended. The incidental, small CIN lesions can be treated adequately by cryotherapy with a single cryoprobe application. When multiple probe placements are required to cover the lesion, cryotherapy may not be the best treatment. It is generally useful to apply acetic acid and Lugol's to the cervix before cryotherapy to reestablish the size and distribution of the lesion. Patients treated with cryotherapy may experience minor cramping during the procedure but otherwise the treatment is not painful. Thus, it is generally not necessary to anesthetize the cervix or give the patient premedication. After cryosurgery, about 20% of patients experience a profuse watery discharge that may require wearing a pad. Some patients experience light spotting, particularly 12 to 15 days after cryotherapy when the eschar begins to separate from the treated areas. Long-term complications of cryotherapy are minimal and consist principally of cervical stenosis or narrowing in 1% to 4% of patients. Several studies have reported that there is no change in fertility and that pregnancy complications seldom occur, except in patients who have excessive cervical fibrosis and in whom cervical dystocia may be a problem. Postcryotherapeutic healing is rapid (it is generally completed within 12 weeks) and usually restores the SCJ to be coincident with the external os. It is important, however, that the flat or shallow cone cryoprobe be used for treatment, because cryoprobes that have deep endocervical extensions may lead to cervical stenosis. In most patients, the SCJ will not be visible after the cervix has healed following cryotherapy. Cryosurgery is a safe, useful, inexpensive procedure that should be considered for any patient (particularly young and nulliparous patients) with a low-grade lesion confined to one or two quadrants of the exocervix.

Carbon Dioxide Laser The CO2 laser is another useful treatment modality for the management of CIN lesions. This laser is an instrument that produces a high-intensity columnar beam of light that can be concentrated to a small spot where it produces a small footprint and a high-power density. This concentration of energy vaporizes the tissue by rapidly boiling the intracellular water, causing the cells to explode. The CO2 laser can be used to ablate tissue or to cut, depending on the spot size and the power density. When the energy is

concentrated on an extremely small spot, the light beam penetrates the tissue deeply and rapidly. It can be used as a knife to excise a lesion, although this requires considerable skill and experience as well as an expensive laser. Surgical excision using the CO2 laser is not often used for CIN these days. When the laser beam is defocused to spread the energy over a larger area but while it still retains sufficient power density to produce cellular destruction, it can be used to evaporate tissues fairly rapidly. The CO 2 laser is used to ablate CIN in patients whose lesion is confined to the portio and in whom invasive cancer has been ruled out by colposcopically directed biopsies and endocervical curettage. The CO2 laser has the following advantages: it can be used for either cutting or ablation; the healing after laser surgery is rapid; and there are minimal side effects, including limited vaginal discharge. Patients treated with the CO 2 laser generally have less cervical narrowing and a lower stenosis rate than those treated with cryotherapy. There is no diminution of fertility, and there are no obstetric complications after laser treatment. The major disadvantages of the laser are (1) its acquisition expense, (2) the time required to acquire and maintain laser skills, particularly for laser conization, and (3) the relatively high maintenance cost of the instrument. Use of the CO2 laser for the management of CIN has decreased dramatically since the introduction of loop electrosurgical excision.

Loop Electrosurgical Excision LEEP was introduced in the United Kingdom as large loop excision of the transformation zone after the modification of techniques originally developed in France. LEEP was introduced in North America in the early 1990s and, as in Europe, rapidly became the procedure of choice for the management of CIN. LEEP takes advantage of the properties of modern, solid-state, electrosurgical generators coupled with loop electrodes, made with a thin stainless steel or tungsten wire, to excise areas of CIN (Fig. 45.10). It is also possible to electrocoagulate tissue with a 5-mm ball electrode. LEEP is used under colposcopic control and has the advantage of being a diagnostic and therapeutic procedure at the same time. The fact that a tissue specimen is provided makes the procedure particularly useful in ruling out early invasive cancer and in identifying unsuspected AIS. Many authors have reported that 4 to 6 per 1,000 cases diagnosed by colposcopy, punch biopsies, and endocervical curettage as having CIN in fact have invasive cancer or AIS in the specimen removed by LEEP.

View Figure

FIGURE 45.10. Diagram of approach to loop excision of small portio lesion, which can be removed by a single pass (A) of the 2-cm × 7-mm exoloop. B: Note the shallow dish configuration of the removed tissue. After the exocervical excision is completed, the canal is recolposcoped after additional acetic acid is applied. If residual disease is found in the canal, it can be removed by the endoloop (C). (From Wright TC, Richart RM, Ferenczy A. Electrosurgery for HPV-related diseases of the anogenital tract. New York: ArthurVision, 1992.)

Because LEEP is easy to learn, easy to teach, and easy to apply, it can be used to treat patients with high-grade CIN without the disadvantages and great cost of cold knife conization (Fig. 45.11). The major advantage of LEEP over conventional cold knife conization is that the procedure can be performed under colposcopic control and the margins can be examined colposcopically after the initial excisions are completed. This gives the operator the opportunity to reexcise additional tissue if it is found colposcopically that residual CIN remains. For LEEP to be used effectively, the extent of excision and the choice of electrodes must be tailored to the size and distribution of the lesion. When disease extends into the canal, however, a loop endocervical excision procedure can be performed with use of a 1- ×1-cm loop electrode. If the lesion is present on both the portio and the canal, a “reverse cowboy hat” combined endocervical-portio excision is appropriate. The preferred electrode for the portio excision should not

exceed 2.5 cm in width and 7 mm in depth. The use of extremely large electrodes may lead to inadvertent cervical amputations and significant reproductive and obstetric morbidity. The side effects and complications of LEEP are almost identical to those of the CO2 laser. FIGURE 45.11. Loop electrosurgical excision procedure. Endosurgical loop is passed through a Lugol-stained cervix from right to left. Note cut surface of cervical stroma from the 12-o'clock to the 6-o'clock position. Also note lack of charring so commonly seen with laser ablation.

View Figure

Although low-grade lesions can readily be treated with cryotherapy, many gynecologists choose LEEP excision instead because of its ease of use, its limited morbidity, and the fact that it provides a specimen. Caution is advised in choosing LEEP for small lesions in women who have only small areas of CIN in a small, nulliparous cervix because it is possible to produce irreversible cervical damage. With few exceptions, all high-grade lesions should be removed with use of an excisional, rather than an ablative, procedure.

Cervical Conization Traditional cold knife conization has been used successfully for generations to excise lesions that extend into the endocervical canal to rule out invasive cancer. The cone is generally planned to be both diagnostic and therapeutic, and the technique is widely understood and practiced (Fig. 45.12). We recommend cone biopsy for endocervical glandular lesions and for suspected microinvasion because accurate diagnosis of these lesions requires a large tissue specimen with negative margins. The cone usually is performed in an operating room under anesthesia and, thus, it is expensive. Because of the size of the specimen excised, bleeding and other sequelae such as cervical stenosis and infection are more common.

FIGURE 45.12. A: Mattress suture is placed as in Sturmdorf tracheloplasty. B: Method of action of suture in drawing the flap into the canal. C: The lower flap has been pulled into position. D: Anterior and posterior flaps have been drawn into the canal. Lateral mucosa wounds are being sutured.

View Figure

See and Treat Because LEEP is a simple procedure that can be used for both diagnosis and treatment during a single office visit, its use has been recommended in the so-called “see and treat” mode. With this application, patients who are detected as having an abnormal Papanicolaou smear are seen for colposcopy, and when a clear-cut colposcopically visible lesion is present, the lesion is removed in its entirety with use of the loop electrode. The patient does not first have to undergo cervical biopsies and an endocervical curettage. The advantages to this approach are that the procedure is done in one office visit, thus reducing patient anxiety by eliminating the need to wait for a diagnosis to be rendered and to then return for additional treatment, and, if used appropriately, it may save time and money. The disadvantage to this approach is that the colposcopic appearance of squamous metaplasia, repair, or other non–HPV-related changes may mimic CIN, and some patients who do not have a CIN lesion may have benign epithelium excised inappropriately. For see-and-treat to be used effectively and appropriately, the colposcopist must be certain that the colposcopic changes that are seen are sufficiently characteristic that a diagnosis of an HPV-related lesion can be assured. Under these circumstances, see-and-treat may be the procedure of choice. Spitzer and colleagues have shown that in an inner-city clinic, the LEEP see-and-treat approach is highly beneficial and provides the opportunity to treat patients with significant lesions who might otherwise not return for their scheduled therapeutic follow-up visits.

Chapter 46

Cancer of the Cervix Dennis S. Chi Nadeem R. Abu-Rustum William J. Hoskins In the first half of the 20th century, more women died from cervical cancer in the United States than from any other cancer. However, with the introduction of the Papanicolaou (Pap) smear in the 1940s, early detection and treatment of preinvasive disease became possible. Consequently, both the incidence and mortality rates owing to invasive cervical cancer in the United States declined approximately 75% by the end of the 20th century. The American Cancer Society estimates that in the year 2002, 13,000 women will be diagnosed with cervical cancer and 4,100 will succumb to the disease. Based on these figures, cervical cancer currently ranks as the third most common female genital tract malignancy in the United States (behind uterine corpus and ovarian cancer); it is the third most common cause of gynecologic cancer death. Worldwide, however, over 370,000 cases are diagnosed annually, leading to approximately 190,000 deaths. This makes cervical cancer not only the most common gynecologic malignancy, but also the third most frequently diagnosed cancer in women (behind breast and colorectal cancer). In general terms, the disease is much more common in developing countries. Overall, 78% of cases occur in these areas. In developing countries, cervical cancer accounts for 15% of female malignancies, carrying a lifetime risk of about 3%. In contrast, in developed countries, cervical cancer accounts for only 4.4% of female malignancies, with a lifetime risk of 1.1%. The highest incidence rates are observed in Latin America, the Caribbean, sub-Saharan Africa, and Southern and Southeast Asia. In developed countries, the incidence rates generally are low, with age-standardized rates less than 14 per 100,000. Very low rates also are observed in China and Western Asia. This geographical disparity is felt to be related to the presence or absence of effective screening programs, because epidemiologic and biologic studies have not shown significant differences in tumor biology in countries with high rates of cervical cancer. EPIDEMIOLOGY FACTORS ASSOCIATED WITH CERVICAL CANCER DEVELOPMENT STAGING OF CERVICAL CANCER HISTOPATHOLOGY CLINICAL PRESENTATION PROGNOSTIC FACTORS PRETREATMENT EVALUATION SURGICAL TREATMENT OF EARLY-STAGE CERVICAL CANCER PERTINENT PELVIC ANATOMY CONCEPT OF RADICAL ABDOMINAL HYSTERECTOMY AND BILATERAL PELVIC LYMPHADENECTOMY SURGICAL TECHNIQUE LAPAROSCOPICALLY ASSISTED RADICAL VAGINAL HYSTERECTOMY WITH PELVIC AND AORTIC LYMPHADENECTOMY LAPAROSCOPICALLY ASSISTED RADICAL VAGINAL TRACHELECTOMY WITH PELVIC AND AORTIC LYMPHADENECTOMY LAPAROSCOPIC RADICAL HYSTERECTOMY WITH PELVIC AND AORTIC LYMPHADENECTOMY LAPAROSCOPIC SURGICAL STAGING PATHOLOGIC EXAMINATION OF THE OPERATIVE SPECIMEN POSTOPERATIVE COMPLICATIONS ADJUVANT THERAPY IN CONJUNCTION WITH RADICAL SURGERY FOLLOW-UP AFTER RADICAL SURGERY FOR CERVICAL CANCER RADIATION THERAPY SUMMARY

EPIDEMIOLOGY Part of "Chapter 46 - Cancer of the Cervix" In the United States, the peak age of developing cervical cancer is 47 years, but approximately 47% of women with invasive cervical cancer are less than 35 years old at diagnosis. Excluding childhood cancers and lymphomas, patients with cervical cancer die younger than women with any other type of cancer (average, 60 years). Older women (>65 years) account for 10% of patients with cervical cancer. Although these older patients represent only 10% of all cases, they are more likely to die of the disease owing to their more advanced stage at diagnosis. Cervical cancer primarily affects women from lower socioeconomic classes and those with poor access to routine medical care. The incidence of cervical cancer in White women is about 7.6 per 100,000 women; for Black women, it is 12.0 per 100,000 women. Other ethnic groups in the United States also have higher incidence rates of invasive cervical cancer than White women. For example, Hispanic, Korean, and Alaskan native women have about twice the incidence rates of White women. Vietnamese women have five times the incidence of White women. On the other hand, Japanese and Chinese women residing in the United States have a slightly lower incidence than their White counterparts. Although there are marked geographical variations in mortality rates for cervical cancer in the United States, sparse data on cervical cancer mortality rates exists by geographic region for Black women. For White women, elevated mortality rates are seen in the South, across Appalachia, in parts of the Midwestern states, and in the upper Northeast (Fig. 46.1). These high mortality rates may be attributed in part to rural and socioeconomic factors related to access to health care. FIGURE 46.1. Cancer mortality rates by state economic area (age-adjusted 1970 US population) cervix uteri: white females, 1970–1994. (From: Devesa SS, Grauman DJ, Blot WJ, et al. Atlas of cancer mortality in the United States, 1950–1994. Bethesda, MD: National Institutes of Health, National Cancer Institute, 1999; NIH Publ No. 99-4564:217, with permission.)

View Figure

Black women with cervical cancer are approximately three times more likely to die from their disease as White women, with mortality rates of 5.7 and 2.2 per 100,000 women, respectively. This “mortality gap” persists when mortality is controlled for demographic characteristics, prognostic factors, and type of treatment. The most recent data on cervical cancer from the American Cancer Society report overall 5-year survivals of 72% for White women compared to 58% for Black women. The 5-year survivals for White women with localized, regional, and distant disease are 92%, 50%, and 15%, respectively. For Black women, the reported 5-year survivals are 86%, 37%, and 7%, respectively. These data suggest that improvements in screening alone may not close the mortality gap, and more data are needed to explain these survival differences between Black and White women. However, equal access to gynecologic oncologists recently has been shown to eliminate disparities in stage-adjusted survival among minority women treated at adjacent public and university hospitals.

FACTORS ASSOCIATED WITH CERVICAL CANCER DEVELOPMENT Part of "Chapter 46 - Cancer of the Cervix" A large number of epidemiologic studies have evaluated factors associated with the development of cervical cancer. Although the risk factors are similar for both invasive cervical cancer and its precursor lesions, the association of the risk factors generally is much stronger with invasive disease. The major risk factors found in most studies are markers of sexual behavior, such as number of sexual partners, age at first intercourse, and history of sexually transmitted disease. Corbett and Crompton reported on a patient who developed invasive squamous-cell carcinoma in only one cervix of a didelphic uterus. A longitudinal septum divided the vagina into two sides. The cancer was found on the side where the vagina was of normal caliber. No cancer was found on the side where the vagina was too small to permit intercourse. This case and other studies, which demonstrated that the disease was much more frequent in married women than celibate women, that it was more common in prostitutes and incarcerated women than the general population, and that it was highly associated with a woman's lifetime number of male sexual partners, all argue for venereal transmission. The fact that a woman's risk was increased not only by her total number of sexual partners, but also by the total number of sexual partners that her male partner or partners had, further strengthens the epidemiologic evidence that cervical cancer behaves as a sexually transmitted disease. With so much evidence suggesting that the development of cervical cancer is related to the sexual transmission of one or more agents, investigators have analyzed virtually every known sexually transmitted disease to determine the causative agent. The most popular candidates for almost a decade were the herpes simplex viruses. However, the most recent molecular research indicates that the human papillomavirus (HPV) is the probable causative agent. Papillomaviruses are members of the A genus of the family Papovaviridae. This family of viruses includes the papillomaviruses, polyoma virus, and SV40 virus. They are all double-stranded DNA tumor viruses. More than 100 types of HPV have been characterized. Thirty-one types of HPV that infect the anogenital tract have been described. Studies that use sensitive molecular methods to detect and type HPV in cervical lesions now find that almost all cervical cancers and their precursors are associated with HPV DNA. Bosch and colleagues analyzed over 900 cervical cancer specimens collected from 32 hospitals in 22 countries. HPV DNA was detected in 93% of the tumors, with no significant variation in HPV positivity among countries. Either HPV type 16 or 18 was found in 64% of specimens. Subsequently, Walboomers et al. reanalyzed the HPV-negative cases from the former study. Of the 55 cases that were reanalyzed, 40 were found to have HPV DNA on more extensive testing. Combining these data with those reported by Bosch et al., and excluding inadequate specimens, the authors reported that 99.7% of the cervical carcinomas contained HPV DNA. Numerous other studies have demonstrated the strong and consistent association between specific types of HPV DNA and invasive cervical cancer and that exposure to HPV precedes the development of cervical disease. Based on the wide body of evidence, the International Agency for Research on Cancer (IARC) of the World Health Organization (WHO) has classified HPV types 16 and 18 as carcinogens in humans. Discussion of the proposed mechanism by which HPV infection leads to cervical carcinoma is beyond the scope of this chapter; the interested reader is referred to excellent reviews of the topic. The majority of women who are infected with HPV never develop cervical dysplasia or cervical cancer. Immunologic mechanisms are thought to clear the virus in these women. It is not known why some women are able to clear the virus, whereas others go on to develop cervical cancer. Additional cofactors most likely contribute to the process; however, the exact mechanism is yet to be established. Risk factors for the development of cervical cancer, in addition to infection with the HPV virus, are listed in Table 46.1. TABLE 46.1. Risk Factors for the Development of Cervical Cancer

Low socioeconomic status

Race (e.g., Black, Hispanic, Vietnamese)

Multiple sexual partners

Early age at first coitus

History of sexually transmitted disease

Infection with human papillomavirus

Cigarette smoking

Immunosuppression

Population studies of women with invasive cervical carcinoma have demonstrated that early age of onset of sexual activity also plays a role in the later development of the disease. It is postulated that during the time of menarche in early reproductive life, the transformation zone of the cervix is more susceptible to oncogenic agents, such as HPV. Women who begin sexual activity before 16 years of age or who are sexually active within 1 year of beginning menses are at a particularly high risk of developing cervical carcinoma. In a review of the literature, Szarewski and Cuzick concluded that a positive association between cigarette smoking and the development of cervical cancer had been reported by the majority of studies designed to address this question. Proposed mechanisms include diminished immune function secondary to a systemic effect of cigarette smoke and its byproducts or a local effect of tobacco-specific carcinogens. The use of oral contraceptives also may play a role in the development of invasive cervical carcinoma, although this theory is more controversial. Because women who use oral contraceptives may be more sexually active than those who do not, this may represent a confounding factor rather than a true independent risk factor. Immunosuppression has been demonstrated to be a risk factor for the development of cervical dysplasia and cancer. Porreco and colleagues reported a relative risk of 13.6 for the development of cervical carcinoma in situ for renal transplant recipients (who generally receive immunosuppressive agents) compared with women in the general population. Women who are infected with the human immunodeficiency virus (HIV) have an elevated prevalence of HPV infections, persistent HPV, and cervical dysplasia compared to women not infected with HIV. Although the exact magnitude of the increased risk of developing invasive cervical cancer among HIV-infected women compared to uninfected women is unclear, Maiman et al. have reported that cervical cancers that develop in HIV-infected women

act aggressively, respond poorly to standard forms of therapy, and are associated with a poor prognosis. In 1993, the Centers for Disease Control and Prevention designated invasive cervical cancer as an AIDS case-defining illness. However, the role of screening for HIV infection in women with newly diagnosed cervical cancer is not universally accepted and the yield may be extremely low, suggesting that decisions about screening for HIV in women with cervical cancer should be based on local prevalence rates. Such screening may be of benefit because the institution of highly active antiretroviral therapy yields longer survival and diminished morbidity, and improvement of immune status through antiretroviral therapy conceivably could improve cervical cancer outcome. The threshold prevalence of HIV infection among women with cervical cancer that justifies HIV screening is a clinical judgment that may be determined best in discussions between patients and treating physicians.

STAGING OF CERVICAL CANCER Part of "Chapter 46 - Cancer of the Cervix" In 1937, the Health Organization of the League of Nations adopted a clinical classification system for cervical cancer. Cervical cancer was the first cancer to be so classified. In 1950, this classification was modified to include preinvasive (in situ) cervical cancer, which was designated stage 0. New recommendations for the clinical classification of carcinoma of the cervix were adopted by the General Assembly of the International Federation of Gynecology and Obstetrics (FIGO) in 1961, and several other modifications have been made since then. The general use of this classification abroad and in the United States has been extremely helpful in reporting and comparing results of various modalities of therapy. Descriptions of the clinical stages in carcinoma of the cervix uteri as updated by FIGO in 1995 appear in Table 46.2. Although the TNM (tumor, regional nodes, and metastasis) staging system is included for completeness, most clinicians use the FIGO (International Federation of Gynecology and Obstetrics) staging. TABLE 46.2. Staging of Carcinoma of the Cervix Uteri (FIGO, 1995)

TNM Classification

Primary Tumor (T)

FIGO Classification

Definition

TX

C

Primary tumor cannot be assessed

T0

C

No evidence of primary tumor

Tis

0

Carcinoma in situ, intraepithelial carcinoma

T1

I

Cervical carcinoma confined to cervix (extension to the corpus should be disregarded)

T1a

IA

Invasive carcinoma, diagnosed microscopically only. All gross lesions, even with superficial invasion, are stage IB cancers. Invasion is limited to measured stromal invasion with maximum depth of 5 mm and maximum width of 7 mm.a

T1a1

IA1

Minimal microscopically evident stromal invasion.

T1a1

IA1

Minimal microscopically evident stromal invasion. Measured stromal invasion with maximum depth of 3 mm and maximum width of 7 mm

T1a2

IA2

Measured stromal invasion with depth from 3B5 mm and maximum width of 7 mm

T1b

IB

Clinical lesions confined to the cervix or preclinical lesions >stage IA

IB1

Clinical lesions no >4 cm in size

IB2

Clinical lesions >4 cm in size

T2

II

Cervical carcinoma invades beyond the uterus but not to the pelvic wall or to the lower third of the vagina

T2a

IIA

No obvious parametrial invasion

T2b

IIB

Obvious parametrial invasion

T3

III

Extends to the pelvic wall or involves lower third of the vagina or causes hydronephrosis or nonfunctioning kidney

T3a

IIIA

Tumor involves the lower third of the vagina. No extension to the pelvic wall

T3b

IIIB

Tumor extends to the pelvic wall or causes hydronephrosis or nonfunctioning kidney

T4

IV

Carcinoma extends beyond the true pelvis or has clinically involved the mucosa of the bladder or rectum. A bullous edema as such does not permit a case to be alotted to stage IV

T4a

IVA

Spread of the growth to adjacent organs

T4b

IVB

Spread to distant organs

a

The depth of invasion should not be more than 5 mm taken from the base of the epithelium, either surface or glandular, from which it originates. Vascular space involvement, either venous or lymphatic, should not alter the staging.

FIGO, International Federation of Gynecology and Obstetrics; TNM, tumor, regional nodes, and metastasis.

Stages II, III, and IV have remained essentially unchanged through the various modifications (Fig. 46.2). The major redefinition and refinements have occurred in stage I disease. Microinvasive (stage IA) carcinoma has been subdivided into stage IA1 and IA2 based on the depth of cervical stromal invasion by carcinoma. Stage IB has been subdivided into stage IB1 and IB2 based on the size of the clinical lesion. FIGURE 46.2. FIGO classification of carcinoma of the cervix. A: Stage Ib: Carcinoma confined to the cervix, exophytic. B: Stage Ib: Carcinoma is confined to the cervix, “cauliflower” lesion. C: Stage Ib: Bulky endocervical barrel-shaped lesion. D: Stage IIa: Carcinoma extends into the upper vagina or fornix. E: Stage IIb: Carcinoma extends into the parametrium but does not extend to pelvic wall. F: Stage IIIa: Carcinoma involves the anterior vaginal wall, extending to the lower third. G: Stage IIIb: The parametrium is infiltrated, and the carcinoma extends to the pelvic wall. H: Stage IVa: The bladder base or rectum is involved.

View Figure

HISTOPATHOLOGY Part of "Chapter 46 - Cancer of the Cervix" The principal histologic type of invasive cervical cancer, occurring in about 80% to 90% of cases, is the squamous (epidermoid) lesion. In 1923, Martzloff classified these squamous tumors into three main histologic subtypes and grades. Grade 1 tumors contain well-differentiated spinal cells, keratin, and squamous pearls (Fig. 46.3A). Grade 2 tumors, the most common, are predominantly composed of transitional cells of the large-cell nonkeratinizing type (see Fig. 46.3B). Grade 3 tumors, the least common, are poorly differentiated small basal-cell-type tumors (see Fig. 46.3C). The classification of Martzloff did not prove to be clinically useful, mainly because biopsies taken from different areas of the same tumor often show different degrees of differentiation and different predominant cell types. Martzloff's work did stimulate Broders, Wentz and Reagan, and others to continue to categorize the histologic types and degree of differentiation of squamous-cell cervical tumors and to study their clinical behavior and response to treatment. The histologic classification of squamous-cell tumor types introduced in 1959 by Wentz and Reagan sometimes is used in pathology reports. However, Willen and coworkers were unable to confirm a predictive value for survival from the Wentz-Reagan classification. Similarly, most recent studies, including those by the Gynecologic Oncology Group (GOG), have shown the use of grading of squamous carcinomas to be of little predictive value. FIGURE 46.3. A: Grade 1: Well-differentiated epidermoid carcinoma of the cervix. High-power view of spinal cell type. The tumor cells contain abundant keratin that forms epithelial pearls. B: Grade 2: Moderately differentiated epidermoid carcinoma of the cervix, transitional cell type. The tumor cells are characterized by a moderate amount of cytoplasm but are without pearl formation. Extensive pleomorphism and mitosis are evident. The tumor is frequently classified as being of large cell, nonkeratinizing type. C: Grade 3: Poorly differentiated epidermoid carcinoma of the cervix, fat spindle or basal cell type. The tumor cells have little cytoplasm, numerous mitoses, and no keratin or epithelial pearls.

View Figure

A rare form of squamous-cell cancer of the cervix is a verrucous carcinoma. It is a very well-differentiated squamous-cell carcinoma with extensive keratinization that usually presents as a large bulky tumor of the cervix and often is confused with giant condylomas, such as those seen on the vulva. There is a sharp line between the tumor and underlying cervical stroma. Verrucous carcinoma has been shown to be associated with HPV infection. Although metastatic disease is rare, this tumor has been said to become more virulent if treated with irradiation. Goldberger and coworkers reported an unusually aggressive verrucous carcinoma of the cervix. According to deJesus and coworkers, at least 49 cases of this tumor have been reported in the female genital tract, sometimes as verrucous carcinoma and sometimes as squamous papillary tumor. Adenocarcinomas of the cervix are becoming more common, especially in younger women. In a review of the Surveillance, Epidemiology, and End Results (SEER) Cancer Incidence Public-Use database from 1973 to 1996, Smith and colleagues reported that although the age-adjusted incidence rates per 100,000 for all invasive cervical cancers and squamous-cell cancers decreased by 37% and 42%, respectively, the rates for adenocarcinoma of the cervix actually increased 29% during the study period. These results suggest that current screening practices may be insufficient in detecting a significant proportion of adenocarcinoma precursor lesions. About one-half of cervical adenocarcinomas are exophytic, usually polypoid or papillary; others diffusely

enlarge or ulcerate the cervix. Approximately 15% of patients have no visible lesion because the carcinoma is within the endocervical canal. Even without visible signs or symptoms, the lesion may infiltrate deeply into the cervix. Drescher and colleagues reported a higher frequency of uterine corpus invasion, nodal metastasis, and ascites in 26 patients with cervical adenocarcinoma compared with 139 cases of squamous-cell carcinoma. More recent studies have reached contradictory conclusions regarding the prognostic significance of this histology. In addition to pure (endocervical) adenocarcinoma (Fig. 46.4), cervical adenocarcinomas can exhibit a variety of patterns and can be composed of diverse cell types. Other histologic patterns include adenoma malignum, endometrioid, clear-cell, serous, and mesonephric. Different histologic patterns and cell types often appear in the same cervical tumor. Because mixtures are common, the designation of tumor type is based on the predominant component. If a second type composes 20% or more of the tumor, the lesion is designated as a “mixed-cell type.” Not infrequently, an adenocarcinoma and squamous-cell carcinoma coexist in the same tumor and these lesions are referred to as adenosquamous carcinomas. The so-called glassy cell adenocarcinoma of the cervix is rare and considered a variant of poorly differentiated adenosquamous carcinoma. It is known to be especially aggressive, with frequent early distant metastasis. Clear-cell adenocarcinoma of the cervix can occur in the presence or absence of intrauterine exposure to diethylstilbestrol. Saigo and coworkers found that the endometrioid pattern was associated with a more favorable prognosis than any other histologic type of cervical adenocarcinoma; however, other authors believe that the subpatterns have no prognostic significance. FIGURE 46.4. Adenocarcinoma of the cervix.

View Figure

The early classifications of squamous-cell carci-noma of the cervix proposed by Martzloff and others divided these tumors into three categories: keratinizing squamous-cell carcinoma, large-cell nonkeratinizing squamous-cell carcinoma, and small-cell carcinoma. Over the years, however, it became apparent that the group designated as “small-cell carcinoma” was composed of a heterogeneous group of tumors, many of which displayed neuroendocrine differentiation. Recent changes in the nomenclature have led to the subdivision of these neuroendocrine tumors into typical carcinoid, atypical carcinoid, large-cell neuroendocrine carcinoma, and small-cell carcinoma. Typical carcinoid and atypical carcinoid tumors are rare in the cervix; therefore, their clinical and pathologic features have not been well characterized. Large-cell neuroendocrine and small-cell carcinomas are highly aggressive neoplasms, with a propensity to metastasize early and widely. Usual methods of therapy are not effective for these histologic types. Various cervical sarcomas have been described by Rotmensch and coworkers. These tumors constitute less than 0.5% of all cervical cancers and include adenosarcomas, leiomyosarcomas, carcinosarcomas, and rhabdomyosarcomas. It is extremely rare for a lymphoma to develop primarily in the cervix, but lymphoma in the cervix is more likely to represent evidence of generalized lymphomatous disease.

CLINICAL PRESENTATION Part of "Chapter 46 - Cancer of the Cervix" Invasive cervical cancer is more likely than its intraepithelial precursors to cause symptoms such as abnormal vaginal bleeding (menorrhagia, metrorrhagia, postcoital bleeding, or postmenopausal bleeding). Many patients have a profuse and often malodorous discharge, especially when the disease is advanced. Thus, any patient with abnormal vaginal bleeding or discharge should have a complete pelvic examination, including a speculum examination with visualization of the cervix. Failure to examine the cervix in a patient with abnormal vaginal bleeding or discharge could result in failure to diagnose cervical cancer. Pain is not a common complaint in patients with cervical cancer unless the disease is advanced. In more advanced stages, patients may complain of bladder and rectal symptoms. When the disease involves lumbosacral and sciatic nerve roots and the lateral pelvic sidewall, chronic boring pelvic bone pain radiating down the leg can be excruciating and indicative of advanced disease. Edema of the lower extremities likewise indicates tumor obstruction of lymphatic and/or venous drainage. Ascites is uncommon in cervical cancer. Unfortunately, the physician cannot rely on the presence of symptoms to lead to a diagnosis of early carcinoma of the cervix. Many women remain without symptoms for many months. It is known that one third of patients with advanced stage III and IV disease have had symptoms for less than 3 months. The only way to diagnose cervical cancer in the earliest possible stages is to routinely apply special diagnostic procedures to large groups of women with and without gynecologic symptoms. This means screening the adult female population with Pap smears. Invasive cervical lesions can be exophytic, infiltrative, ulcerative, or occult. The size of the visible lesion on the cervix may not correlate well with the extent or depth of invasion (Fig. 46.5). FIGURE 46.5. Squamous cell carcinoma, cervix uteri, FIGO stage IIA.

View Figure

An everted exophytic carcinomatous growth may be friable. Bits of tissue may break off on the examining fingers. On inspection, the friable exophytic cancer shows a rough, granular bleeding surface that can be sloughing and infected, with a foul-smelling discharge. A tumor that develops beneath the mucosa of the exocervix and infiltrates the cervical stroma usually causes cervical enlargement. The surface of the cervix may feel smooth, but the cervical consistency to palpation is firm or nodular. It is characteristic of cervical cancers that develop in the endocervical canal to cause cervical enlargement and a firm cervical consistency before breaking through the mucosa of the exocervix to cause a lesion. This also is characteristic of many cervical cancers that develop in postmenopausal women. In fact, it is possible, although uncommon, for a cervical cancer that is developing high in the endocervical canal to invade the parametrial tissue and even obstruct the ureters before causing a visible cervical lesion. An ulcerative lesion can look like a fairly clean punched-out ulcer, but more commonly it is an irregular crater with a necrotic bleeding base and a foul-smelling discharge. Any grossly visible lesion of the cervix should be considered suspicious for cancer, and biopsy should be performed. Good visualization with a speculum and adequate illumination are essential. A Pap smear should be taken even though it can be less accurate in the presence of a grossly visible cervical lesion. Colposcopic examination is neither needed nor particularly effective for a gross cervical lesion, but can be helpful when there is a small surface lesion to identify the most abnormal area for directed biopsies. The

primary benefit of colposcopy is in visualizing noninvasive, precursor, or minimally invasive lesions that cannot he visualized without magnification. Cervical biopsy techniques are discussed in Chapter 45. Biopsies can be undertaken with any of a number of special instruments; the Kevorkian (Fig. 46.6), Younge, or Gaylor biopsy forceps are particularly functional for taking an adequate biopsy specimen. It is important to obtain a specimen where frank stromal invasion can be demonstrated, not from the exophytic portion where no benign stroma is present. Surgical conization under anesthesia is unnecessary when a grossly visible lesion is present; indeed, it is contraindicated in the presence of a gross lesion because it has the potential of delaying the initiation of definitive treatment. Iodine (Schiller) staining can be used to demarcate the vaginal margins of a neoplastic area from adjacent normal epithelium. All of these procedures, as well as the popular loop diathermy conizations, can be done in the outpatient setting. It is rarely necessary to take the patient to the operating room to diagnose cervical cancer. If an anesthetic is deemed necessary for some other reason, a careful pelvic examination under anesthesia, biopsy of any vaginal lesions, cervical and uterine sounding, cystoscopy and proctoscopy (if warranted), and even uterine curettage can be done. Useful information to plan treatment thus can be obtained. FIGURE 46.6. Kevorkian square-jawed cervical biopsy forceps.

View Figure

PROGNOSTIC FACTORS Part of "Chapter 46 - Cancer of the Cervix" Several factors have been reported to affect prognosis in cervical cancer. However, the most important determinant of prognosis remains FIGO clinical stage. Based on studies by Piver, van Nagell, Delgado, and numerous others, which demonstrated the prognostic significance of depth of cervical stromal invasion and tumor size in early-stage disease, FIGO incorporated these factors into its current clinical staging system. The 5-year survival rates by stage as reported by FIGO in the most recent report are shown in Table 46.3. TABLE 46.3. 5-Year Survival by FIGO Stage

Stage

Number of Patients

5-Year Survival Rate (%)

Ia1

787

94.6

Ia2

313

92.6

IB1

986

90.4

IB2

440

79.8

IIA

993

76.0

IIB

2,775

73.3

IIIA

131

50.5

IIIB

2,271

46.4

IVA

258

29.6

IVB

196

22.0

FIGO, International Federation of Gynecology and Obstetrics.

P.1383 In addition to FIGO stage, other reported prognostic factors include endometrial cavity extension, regional (pelvic) and distant (paraaortic) lymph node metastases, histologic tumor grade, and lymphovascular space invasion (LVSI). Uterine corpus and endometrial cavity extension from primary cervical cancer was originally thought to be a poor-prognosis factor. The original League of Nations classification of 1937 included such lesions in the stage II category. Over the years, classification of the disease based on extension to the uterine corpus or endometrial cavity has gradually been discounted. These lesions, classified as stage IC in 1950, were included in the broad category of stage IB in 1961. Despite the change in classification, we still have some concern about patients whose tumor extends into the corpus of the uterus. Evidence from Washington University reported by Perez and coworkers showed that tumor extension into the endometrial cavity lowers the 5-year survival rate of stage IB and IIA lesions by 10% to 20%. Lesions that involve the corpus also were found to have a twofold greater incidence of distant metastases when compared with lesions without corpus extension. Similar observations have been reported by Prempree and coworkers from 82 cases of stage I and II disease with endometrial extension. The absolute 5-year cure rates of 68% for stage I and 62% for II disease with endometrial extension reflects the higher risk of metastases: 20% for stage I cases and 24% for stage II cases. Such reports must be studied with consideration of the difficulty of establishing a diagnosis of endometrial extension by using microscopic study of endometrial curettage specimens. Frequently, the curettage specimen is contaminated by the cervical tumor, making it difficult to be certain about endometrial extension. Treatment planning is not altered by such observations, nor is a fractional curettage recommended as part of pretreatment evaluation. Lymph node metastases, either regional (pelvic) or to higher level (common iliac and paraaortic) lymph nodes, have proved to be one of the most reliable prognostic factors for patients with cervical cancer (Fig. 46.7). The frequency of metastases to pelvic lymph nodes is about 0% to 0.5% for patients with stage IA1; 7% to 9% for stage IA2; 12% to 20% for stage IB; 20% to 38% for stage IIA; 16% to 36% for stage IIB; >35% for stage III; and >50% for stage IV. Preoperative detection of positive pelvic and paraaortic lymph nodes is unreliable, even with newer radiologic imaging techniques. Detection of positive lymph nodes is more accurately determined when lymphadenectomy is used in preoperative staging or treatment. Although Kolstad showed that when intraoperative lymphography is used, 15% to 25% more patients with stage IB disease are found to have positive regional lymph nodes, this technique is rarely used outside the research setting. FIGURE 46.7. Squamous cell carcinoma, cervix uteri, FIGO stage IIa, with gross metastatic tumor in a right parametrial lymph node.

View Figure

When patients with stage IB cervical cancer are primarily treated with radical hysterectomy and pelvic lymphadenectomy, the 5-year cure rate is about 90%, if there are no lymph node metastases. However, if

metastatic disease to lymph nodes is found, the 5-year cure rate falls to about 65%. The number of positive nodes also influences prognosis. In a review of the literature, Hoskins reported an 83% survival rate for patients with stage IB and IIA disease who had negative lymph nodes at the time of radical hysterectomy and pelvic lymphadenectomy. The survival rate decreased to 57% in patients with one to two positive nodes and 31% in those with greater than three positive nodes. Metastatic disease to paraaortic lymph nodes occurs in 4% to 7% of patients with stage I disease, 15% to 20% with stage II disease, 25% to 30% with stage III disease, and 30% to 50% with stage IVA disease. Most studies confirm that metastases to paraaortic nodes occur more frequently when positive pelvic nodes also are present. The rarity of patients found to have positive paraaortic nodes when pelvic nodes are negative raises the question: How well sampled and/or sectioned were the pelvic nodes? Even with extended-field radiation therapy, the 5-year survival rate for patients with metastases to the paraaortic nodes is only about 25% to 35%. Histologic tumor grade has been reported to affect prognosis. Early studies by Chung and coworkers and van Nagell and colleagues demonstrated a poorer prognosis among patients with poorly differentiated tumors. However, more recent studies by Zaino and the GOG have shown the grading of squamous tumors to be of little predictive value in cervical carcinoma. Shingleton and Orr reviewed nine publications that reported on 3,761 patients with predominantly squamous-cell carcinoma. Twenty-eight different factors were evaluated for prognostic significance. On multivariate analysis, tumor volume, lymph node metastasis, parametrial invasion, and LVSI were found to be significant independent prognostic factors but patient age and tumor grade were not. On the other hand, tumor differentiation may have a significant prognostic role in adenocarcinoma of the cervix. Shingleton and Orr also reviewed eight studies containing 577 patients with adenocarcinoma of the cervix. As with squamous-cell carcinoma, the strongest independent prognostic variables were tumor size and nodal metastasis. However, unlike with squamous-cell carcinoma, tumor grade appeared to have prognostic significance. Several investigators, including Swan and Roddick, Wheeless and Graham, and Julian and coworkers, have drawn attention to the fact that when there is a mixture of adenocarcinomatous and squamous elements—so-called adenosquamous tumors—the prognosis is poor and the incidence of pelvic lymph node metastases is high. Histologic combinations should be considered when comparing the prognoses of adenocarcinoma and squamous cancers of the cervix. The literature is mixed on the overall issue of whether adenocarcinoma in general and adenosquamous cancers in specific are more virulent and less curable than their squamous counterparts. Stehman and colleagues performed a multivariate analysis of prognostic variables for 626 patients with locally advanced cervical carcinoma treated with radiation therapy on three GOG protocols. Histologic cell type was not found to be a significant prognostic factor. A national pattern of care and evaluation study of the American College of Surgeons also failed to report statistically different 5-year survival rates for squamous and adenocarcinoma, regardless of type of therapy chosen. In a GOG prospective study of 645 patients with stage IB squamous-cell carcinoma of the cervix treated with radical hysterectomy and pelvic lymphadenectomy, Delgado et al. identified three independent risk factors in relation to disease-free survival: the depth of invasion, the size of the tumor, and LVSI. The disease-free interval was 89% for those patients without LVSI compared to 77% for those found to have LVSI. Although not all studies have found LVSI to be an independent prognostic factor, as stated, in a review of nine studies (including the study by Delgado et al.) containing 3,761 patients, Shingleton and Orr found LVSI to be a significant independent prognostic factor on multivariate analysis. The incidence of LVSI in early-stage lesions varies widely, depending on multiple factors, including the number of sections of the cervix prepared, the depth of stromal invasion, and the interest of the examining pathologist. Observations from Austria and Germany indicate that there is considerable variation in the frequency with which LVSI is recognized by the pathologist. In a combined study of over 1,000 patients at three different reference centers (Graz, Munich, and Erlangen), Burghardt et al. reported the frequency with which LVSI was identified ranged from 9% in Munich, where only blood vessel involvement was so classified, to 43% in Graz, where it was classified as capillary-like space involvement. At the third center, Erlangen, the corresponding value was intermediate at 23%. Such variations in histopathologic criteria may well contribute to some of the controversy that exists regarding the prognostic significance of LVSI. To this point, this discussion of prognostic factors has included only anatomic and morphologic factors. Peipert and associates emphasize that cancer, including cervical cancer, has both form and function. Accordingly, other clinical variables, such as patient's symptoms, symptom severity, and comorbidity, affect the survival rate of patients with invasive cervical cancer. Unless these variables are suitably

included, prognostic estimates based on morphology alone are imprecise, and therapeutic evaluations can be misleading. According to Rutledge and associates, there is no consistent effect of age on survival rate in patients treated for cervical cancer. Younger patients with early-stage disease seemed to survive longer than older patients, but the tendency reversed when disease was advanced.

PRETREATMENT EVALUATION Part of "Chapter 46 - Cancer of the Cervix" When a diagnosis of invasive cervical cancer has been established histologically, the clinician should perform an evaluation of all pelvic organs to determine whether the tumor is confined to the cervix or has extended to the adjacent vagina, parametrium, endometrial cavity, bladder, ureters, or rectum. According to the FIGO guidelines for clinical staging, diagnostic studies may include intravenous urography (IVU), cystoscopic examination of the bladder and urethra, a proctosigmoidoscopic study, a barium enema (BE), and in the case of early-stage disease, a colposcopic study of the vagina and the vaginal fornices. Colposcopic findings may be used for assigning a stage to the tumor (for instance, FIGO stage IIA), but the results must be confirmed by biopsy. Chest radiographs and electrocardiographic studies are used to determine cardiopulmonary disease, particularly in the older patient. Pulmonary function studies can be important, especially for evaluating patients who are candidates for extensive surgery. When studies detect ureteral obstruction, a tumor is classified as a stage IIIB lesion, regardless of the size of the primary lesion. Ureteral obstruction, either hydronephrosis or nonfunction of the kidney, is well established as an indicator of poor prognosis, as recognized in the FIGO classification. Retrograde pyelography can be performed after the ureteral obstruction is located for further evaluation; however, it is not routinely recommended. Kidney function studies such as serum creatinine and creatinine clearance provide important baseline information before treatment; complete urinalysis is useful for detecting the presence of albumin or white and red blood cells and renal tubular casts. In women with bulky or advanced-stage tumors, the bladder mucosa also should be inspected cystoscopically for possible bullous edema, which indicates lymphatic obstruction within the bladder wall. Evidence of tumor in the bladder must be confirmed by biopsy before the lesion can be classified as stage IVA. Rectal mucosal lesions also require a biopsy via proctosigmoidoscopy, because they can be related to an inflammatory process rather than to the cervical tumor. A pelvic examination must be performed as part of the staging process, and it may be necessary to have the patient completely relaxed by general anesthesia. In up to 20% of patients, the initial clinical classification of the disease has proved to be incorrect at the time of pelvic examination under anesthesia. Such an examination can reveal a more advanced stage of the disease than was originally found; additional biopsies (if indicated) or fractional curettage can be done as well as colposcopy, cystoscopy, and proctosigmoidoscopy. In today's health care climate, the cost of a separate examination under anesthesia should be reserved for only the most problematic cases. Pretreatment pedal lymphangiography was used in the past to detect pelvic and paraaortic lymph node metastases, but the procedure is tedious and associated with many false-negative and -positive findings. When compared with lymphadenectomy, positive lymphangiograms have an accuracy rate of