Icru 78

 

I    C R  U R E  P   O R T  7   8 

Volume 7 No 2 2007

ISSN 1473-6691

Journal of the ICRU

ICRU REPORT 78 Prescribi Prescr ibing, ng, Rec Recor ordin ding, g, and Rep Reportin orting g Proton-Beam Therapy

 J   o  u r  n  a l    o f    t   h   e I    C R  U V   o l    u m  e 7  N  o 2  2   0   0  7 

OXFORD UNIVERSITY PRESS

INTERNATIONAL COMMISSION ON RADIATION RADIA TION UNITS AND MEASUREMENTS

 

Journal of the ICRU ISSN 1473-6691 Commission Membership

P. M. DeLuca (Chairman) A. Wambersie (Vice Chairman) S. M. Seltzer (Secretary) P. Dawson K. Doi R. A. Gahbauer M. Inokuti D. T. L. Jones H.-G. Menzel B. D. Michael H. G. Paretzke H. Tatsuzaki G. F. Whitmore A. Allisy (Honorary Chairman)

SUBSCRIPTIONS A subscription to Journal of the ICRU comprises 2 issues (Reports). All prices include postage, and for subscribers outside the UK delivery is by Standard Air. Air. Annual Subscription Subscription Rate (Volume 7, 2 issues, 2007)  Institutional

Print edition and site-wide online access: £194/$350/ €291 Print edition edition only: £184/$332/ €276 Site-wide online access only: £184/$332/ €276 Personal

Print only: £100/$180/ €150 Please note: note: £ Sterling or € Euro rates apply in Europe, US$ elsewhere Reports are available for purchase as single issues. Please see the Journal homepage: http://jicru.oxfordjournals.org/, and go to ‘Purchase back reports of the ICRU’. Full prepayment, in the correct currency, currency, is required for all orders. Orders are regardregarded as firm and payments are not refundable. Subscriptions are accepted and entered on a complete volume basis. Claims cannot be considered more than FOUR months after publication or date of order, whichever is later. All subscriptions in Canada are subject to GST. Subscriptions in the EU may be subject to European VAT. If registered, please supply details to avoid unnecessary unnecessary charges. For For subscriptions that include online versions, a proportion of the subscription price may be subject to UK VAT. Personal rate subscriptions are only available if payment is made by personal cheque or credit card and delivery is to a private address. Visit www.jicru.oxfordjournals.org www.jicru.oxfordjournals.org for complete listi ng of previous reports available from Oxford University Press. For further information, information, please contact: contact: Journals Customer Service Department,

Oxford Oxfor d University University Press, Press, Great Clarend Clarendon on Street, Street, Oxfor Oxford d OX2 6DP, UK. Email: Email:  [email protected]. Tel (and answerphone outside normal working  [email protected]. contact: hours): +44 (0)1865 353907. Fax: Fax: +44 (0)1865 353485. In the US, please contact: Journals Customer Service Department, Oxford University Press, 2001 Evans Evans Road, Cary, NC 27513, 27513, USA. Email: [email protected] [email protected]. rg. Tel (and answerphone answerphone outside normal working hours): 800 852 7323 (toll-free in USA/Canada). Fax: Fax: 919 contact: Journals Customer Services, Oxford University 677 1714. In Japan, please contact: University Pres Pr ess, s, 11-11-17 17-5 -5F F, Mu Muko koga gaok oka, a, Bu Bunk nkyo yo-k -ku, u, Tok okyo yo,, 11 1133-00 0023 23,, Ja Japa pan. n. Em Emai ail: l: [email protected]. okudaoup@po. iijnet.or.jp. Tel: Tel: (03) 3813 1461. Fax: Fax: (03) 3818 1522. Methods of payment. Payment should be made: by cheque (to Oxford Oxford University

Press, Cash Press, Cashiers iers Offi Office, ce, Great Clare Clarendon ndon Stree Street, t, Oxfor Oxford, d, OX2 6DP 6DP,, UK); by bank  bank  transfer [to Barclays Bank Bank Plc, Oxford Office, Office, Oxford (bank sort code 20-65-18) (UK); overseas only Swift code BARC GB22 (GB£ Sterling Account no. 70299332, IBAN GB89BARC206518702 GB89BARC20651870299332; 99332; US$ Dollars Account no. 66014600, IBAN GB27BARC20651866014600; EU󲂬 EUR EURO O Account Account no. no. 78923655, 78923655, IBA IBAN N GB16BAR GB16 BARC206 C2065187 518789236 8923655]; 55]; or by credit card (Mastercard (Mastercard,, Vis Visa, a, Switch or American Express).

 Executive Secretary Secretary

P. L. Russell  Assistant Executive Executive Secretary Secretary

L. J. Atwell  Managing Editors: Editors: ICRU Website Website and ICRU News

R. A. Gahbauer D. T. L. Jones  Honorary Editor ICRU ICRU News

H. G. Ebert  Honorary Counsel

W. R. Ney  Journal of the ICRU Scientific Editor  M. Inokuti

Journal of the ICRU is published twice a year by Oxford University University Press, Oxford, UK. Annual subscription price is £184/$332/ €276. Journal of the ICRU is distributed by Mercur Mercury y Internati International, onal, 365 Blair Blair Road, Road, Av Avenel, enel, NJ 07001, 07001, USA. Perio Periodical dicalss postage paid at Rahway, Rahway, NJ and at additional entry points. US Postmaster Postmaster:: send address address changes changes to Journa Journall of the ICRU, ICRU, c/o Mercury Mercury Internatio Inter national, nal, 365 Blair Blair Road, Av Avenel, enel, NJ 07001, 07001, USA. Permissions

For information on how to request permissions to reproduce articles/information from this journal, please visit www.oxfordjournals.org/jnls/permissions. www.oxfordjournals.org/jnls/permissions. Dosage disclaimer

The mention of trade names, commercial products or or organizations, and the inclusion of advertisements in the journal does not imply endorsement by the ICRU, ICRU, the editors, the editorial board, board, Oxford University University Press or the organization to which the authors are affiliated. The editors and publishers have taken all reasonable precautions to verify dosage, the results of experimental work and clinical findings published published in the journal. The ultimate responsibility for the use and dosage of drugs mentioned in the Journal and in interpretation of published material lies with the medical practitioner, andor theomissions editors and for damages arising from any errors in publishers the journal.cannot Pleaseaccept informliability the editors of any errors. Legal Notice

This report was prepared by the International Commission on Radiation Units and Measurements, Measure ments, Inc. (ICRU). (ICRU). The Commission strives to provide accurate, complete and useful information in its reports. However, neither the ICRU, the members of the ICRU, Oxford University Press, other persons contributing to or assisting in the preparation of  this report, nor any person acting on behalf behalf of any of these parties: (a) makes any warranty or representation, representation, express or implied, implied, with respect to the accuracy accuracy,, completen completeness ess or usefulness of the information contained in this report, or that the use of any information, method or process disclosed in this report may not infringe on privately owned owned rights; or (b) assumes any liability liability with respect to the use of, or for damages resulting from the use of any informati information, on, method or process disclosed in this report. © International Commission on Radiation Radiation Units and Measurements 2007 2007 All rights reserved; no part of this publication may be reproduced, stored in a retrieval system, syste m, or transmitted transmitted in any form form or by any means, means, electr electronic, onic, mecha mechanical nical,, photo photo-copying, recording, or otherwise without without prior written permission of the Publishers, or a licence permitting restricted copying issued in the UK by the Copyright Licensing Agency Ltd, 90 Tottenham Tottenham Court Road, Road, London W1P W1P 9HE, or in the USA by the Copyright Copyright Clearance Clearance Center, Center, 222 Rosewood Rosewood Drive, Drive, Danvers, MA 01923. For those in the USA or Canada not registered with CCC, CCC, articles can be obtained by fax in 48 hours by calling: WISE for MedicineTM 1-800-667-WISE. Typeset by Techset Composition (P) Ltd Printed Print ed by Bell & Bain, Glasgo Glasgow, w, UK

 

ICRU REPORT No. 78

PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

THE INTERNATIONAL COMMISSION ON RADIATION UNITS AND MEASUREMENTS Prepared Prepar ed jointly with the INTERNATIONAL INTERNATIONAL ATOMIC ATOMIC ENERGY AGENCY  December 2007

Journal of the ICRU Volume 7 No 2 2007

ISBN 9780199543489 Published by Oxford University Press

 

PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY   Report Committee

D. T. L. Jones (Co-Chair), iThemba LABS, Somerset West, South Africa H. D. Suit (Co-Chair), Massachusetts General Hospital, Boston, Massachusetts, USA   Y.. Akine, Proton Medical Radiation Center,  Y Center, Tsuku Tsukuba, ba, Japan G. Goitein, Paul Scherrer Institute, Villigen, Switzerland M. Goitein, Ankerstrasse 1, Windisch AG, Switzerland N. Kanematsu, National Institute of Radiological Sciences, Chiba, Japan R. L. Maughan, University of Pennsylvania, Philadelphia, Pennsylvania, USA  H. Tatsuzaki, National Institute of Radiological Sciences, Chiba, Japan H. Tsujii, National Institute of Radiological Sciences, Chiba, Japan S. M. Vatnitsky, International Atomic Energy Agency, Vienna, Austria Commission Sponsors

P. M. DeLuca, Jr, University of Wisconsin, Madison, Wisconsin, USA  R. A. Gahbauer, A. James Cancer Hospital, Ohio State University, Columbus, Ohio, USA   A. Wambersie, Universite´  Catholique de Louvain, Brussels, Belgium G. F. Whitmore, Ontario Cancer Institute, Toronto, Ontario, Canada  International Atomic Energy Agency Sponsors

P. Andreo, International Atomic Energy Agency Agency,, Vienna Vienna,, Austria J. H. Hendry, International Atomic Energy Agency, Vienna, Austria Consultants to the Report Committee

G. Coutrakon, Loma Linda University Medical Center, Loma Linda, California, USA   A. Lomax, Paul Scherrer Institute Institute,, Villigen, Switzerland H. Paganetti, Massachusetts General Hospital, Boston, Massachusetts, USA  E. Pedroni, Paul Scherrer Institute, Villigen, Switzerland The Commission wishes to express its appreciation to the individuals involved in the preparation of this repor re port, t, for the time and efforts efforts whi which ch the they y de devot voted ed to th this is tas task k and to ex expr press ess its appr appreci eciat ation ion to th the e organizations with which they are affiliated.  All rights reserved. No part of this book may be reproduced reproduced,, stored in retrieval syst systems ems or transmitte transmitted d in an any y for form m by any mea means, ns, ele electr ctroni onic, c, ele electr ctros osta tatic tic,, mag magnet netic, ic, mec mechan hanica icall ph photo otocop copyin ying, g, re recor cordin ding g or otherwise, without the permission in writing from the publishers. British Library Cataloguing in Publication Data. A Catalogue record of this book is available at the British Library. ISBN 9780199543489

 

 ACKNOWLEDGMENTS  ACKNOWLEDGMENTS The International Commission on Radiation Units and Measurements gratefully acknowledges the generous contributions contributions to the production of the pres present ent report by the following following organiz organizatio ations. ns. Hitachi, Ltd. 7-2-1 Omika-cho Hitachi-shi Ibaraki-ken 319-1221 JAP JAPAN AN

Ion Beam Applications S.A. Chemin du Cyclotron 3 B-1348 Louvain-la-Neuve BELGIUM

Siemens Medical Solutions Henkestrasse 127 D-91052 Erlangen FEDERAL REPUBLIC OF GERMANY 

 

Journal of the ICRU Vol 7 No 2 (2007) Report 78 Oxford University Press

doi:10.1093/jicru/ndm021

PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

CONTENTS

 ACKNOWLEDGM  ACKN OWLEDGMENTS ENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

i

PREF PR EFA ACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1

 ABSTRACT  ABSTRA CT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3

EXE EX ECU CUTI TIVE VE SU SUMM MMAR ARY Y . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5

1

INTR IN TROD ODUC UCTI TION ON . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Princi Principle ples s of proto proton n therap therapy y. . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1.1 Physi Physical cal chara characteris cteristics tics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1.2 1.1 .2 Bio Biologi logical cal effe effects cts  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1.3 1.1 .3 Clin Clinical ical eva evalua luation tionss   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 His History tory of proton proton the therap rapy y  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 Pre Presen sentt sta status tus   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2

11

. .. . . . . .. . . . . .. . . . . .. . . . . . .. . . . . .. . . . .. . .. . .. . .. . . . .. . . . .. . . . .. . . . .. . . . .

11 11 15 15 16 17

RADI RA DIA ATI TION ON BIO BIOL LOG OGY Y CON CONSI SIDE DERA RATI TION ONS S.......................................

21

2.1 Introd Introduct uction ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Micro Microdosim dosimetry etry and and linear linear energy trans transfer fer . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Revie Review w of publi published shed proton RBE values values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  2.3.1 RBE values determined using in vitro and in vivo systems   . . . . . . . . . . . . .  2.3.2 RBE versus depth for 60 – 250 MeV beams  . . . . . . . . . . . . . . . . . . . . . . . . . .  2.3.2.1 60 60– – 85 MeV   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  2.3.2.2 160 160– – 250 MeV   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  2.3.2.3 RBE on the declining distal edge of the SOBP  . . . . . . . . . . . . . . . .

21 21 21 23 24 24 25 25

3

2.4 Use of a gen generic eric RBE valu value e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 Dose specifi specificat cation ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  2.5.1 The RBE-weighted absorbed dose  ( DRBE) . . . . . . . . . . . . . . . . . . . . . . . . . . .

26 27 28

BEAM BE AM DE DELI LIVE VER RY AN AND D PR PROP OPER ERTI TIES ES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

29

3.1 3.2

Proton-therapy Proton-ther apy faci facilities lities  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The treat treatmentment-delive delivery ry sys system tem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  3.2.1 The beam nozzle and snout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  3.2.1.1 Passive beam-delivery techniques  . . . . . . . . . . . . . . . . . . . . . . . . . .  3.2.1.2 Dynamic beam-delivery techniques  . . . . . . . . . . . . . . . . . . . . . . . . .  3.2.1.2.1 Wobbled beams  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  3.2.1.2.2 Repainting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  3.2.2 Patient support and positioning  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  3.2.3 Special treatment techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

29 29 29 29 32 35 35 36 36

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

 3.2.3.1  3.2.3.2

4

Eye treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stereotactic radiosurgery and stereotactic radiotherapy  . . . . . . . . .

36 37

Rotating 3.3  3.2.4 Acceler Acc elerato ators rs  . .gantries . . . . . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..  3.3.1 Linear accelerators  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  3.3.2 Cyclotrons, isochronous cyclotrons, and synchrocyclotrons . . . . . . . . . . . . . .  3.3.3 Synchrotrons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  3.3.4 Typical accelerator operating parameters  . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 The prop properti erties es of prot proton on beams beams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  3.4.1 Proton interactions with matter   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  3.4.2 Definition and specification of beam properties and beam parameters . . . . .  3.4.2.1 Beam properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  3.4.2.2 Beam parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5 Rad Radiat iation ion qua quality lity   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

37 38 39 39 41 42 43 43 43 44 44 46

DOSI DO SIM MET ETR RY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

49

4.1 4.2 4.3 4.4

General consi General considera derations tions   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reference Refer ence dosime dosimetry try with with a Fara Faraday day cup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reference Refer ence dosime dosimetry try with with a calorime calorimeter ter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reference Refer ence dosimetry dosimetry with with ionization ionization chambers chambers having  having   60Co calibration coefficients  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  4.4.1 ICRU proton dosimetry protocol (ICRU 59)   . . . . . . . . . . . . . . . . . . . . . . . . .  4.4.1.1 Physical quantities for ICRU 59  . . . . . . . . . . . . . . . . . . . . . . . . . . .  4.4.2 IAEA proton dosimetry code of practice (TRS 398) . . . . . . . . . . . . . . . . . . . .  4.4.2.1 Physical quantities for TRS 398  . . . . . . . . . . . . . . . . . . . . . . . . . . .  4.4.3 Considerations concerning dry and humid air  . . . . . . . . . . . . . . . . . . . . . . .  4.4.4 The value of w /  e in air for proton beams . . . . . . . . . . . . . . . . . . . . . . . . . . . .  4.4.4.1 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  4.4.4.2 Determination of the w( E)  value  . . . . . . . . . . . . . . . . . . . . . . . . . . .

49 50 54

 4.4.5  4.4.6

66

Comparison of proton dosimetry protocols   . . . . . . . . . . . . . . . . . . . . . . . . . . Relation Rela tion between between absorbed dose to water and air-kerma air-kerma calibration calibration coefficients  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5 Refer Reference ence dosime dosimetry try for for scanned scanned beams. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6 Ioniza Ionizationtion-chamb chamber er dosimetry dosimetry compari comparisons sons  . . . . . . . . . . . . . . . . . . . . . . . . . . 4.7 Bea Beam m mon monito itorin ring  g . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  4.7.1 Ionization chambers  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

57 57 58 59 61 62 64 64 65

68 69 69 70 71

5

 4.7.2 Position and dose uniformity   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  4.7.3 Considerations for scanned beams   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  4.7.4 Secondary-emission monitors  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  4.7.5 Range Range/ /energy measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.8 Rel Relati ative ve dos dosime imetry try   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  4.8.1 Phantom materials  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  4.8.2 Detectors for dose-distribution measurements . . . . . . . . . . . . . . . . . . . . . . . .  4.8.2.1 Single ionization chambers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  4.8.2.2 Silicon diodes   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  4.8.2.3 Radiographic films  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  4.8.2.4 Alanine  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  4.8.2.5 Other detectors   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  4.8.3 Determination of dose distributions  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  4.8.3.1 Range and depth – dose characteristics  . . . . . . . . . . . . . . . . . . . . . .  4.8.3.2 Beam profiles and penumbrae . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

72 73 73 74 74 74 75 76 76 77 77 78 80 80 81

GEOME GEO METRI TRIC C TERM TERMS, S, AND AND DOSE DOSE AN AND D DOSE– DOSE– VO VOLU LUME ME DEF DEFINI INITIO TIONS NS . . . . . . . . . . . . . .

83

5.1

Anatomic Anato mic volumes volumes relat relating ing to to the tumor   . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

83

 

CONTENTS

5.1.1 5.1.1

Gross Gro ss tumo tumorr volume volume (GTV (GTV))  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

83

5.1.2 5.1 Clinical (CTV) . . . CTV  . . .V  .  . 5.1.2. .2.1 1 target The dose dovolume se intent intention ion for fo r. the CT   .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. 5.1.3 Interna Internall target volume (ITV) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.4 Planning target volume (PTV)   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.4.1 5.1. 4.1 Margins for for the different different types types of variations variations and uncertaintie uncertaintiess   . . . 5.1.4. 5.1 .4.2 2 Deli Delinea neating ting the PTV  .  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.4. 5.1 .4.3 3 Mult Multiple iple PTV PTVss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.4.4 5.1. 4.4 Proto Proton-spec n-specific ific issues issues regardin regarding g the PTV  . . . . . . . . . . . . . . . . . . . . 5.2 Anato Anatomic mic volumes volumes relating relating to uninvolv uninvolved ed normal tissues tissues and organs organs   . . . . 5.2.1 5.2 .1 Org Organ an at at risk risk (OA (OAR) R)  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.2 Planning organ at risk volum volumee (PRV) (PRV) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.3 Remain Remaining ing volume at risk (RVR)   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Gen Generi eric c geome geometric tric term terms s  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.1 5.3. 1 Volume of intere interest st (VOI) (VOI)   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

8 4 84 85 85 85 86 86 86 87 87 87 88 88 88

5.3.2 Targ 5.3.2 arget et volum volumee (TV) (TV)  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.3 5.3 .3 Sur Surfac facee of inte interes restt (SOI) (SOI)  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.4 5.3 .4 Po Point int of of intere interest st (POI) (POI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4 Nom Nomenc enclat latur ure e. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.1 5.4 .1 Mul Multipl tiplee GTVs, GTVs, CTVs, CTVs, and and PTVs PTVs  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.1.1 5.4. 1.1 One GTV plus plus a surrounding surrounding volume volume intended intended to rece receive ive the same dose   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.1.2 5.4. 1.2 One GTV plus plus a surrounding surrounding volume volume intended intended to rece receive ive a lower dose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4. 5. 4.1. 1.3 3 No GT GTV  V  .  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.1.4 5.4. 1.4 Two spatially spatially separ separated ated GTVs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.1.5 5.4. 1.5 Two nested GTVs GTVs plus a region of possible possible microscopic microscopic disease disease   . . . 5.4.1.6 5.4. 1.6 Reduce Reduced d tumor dose in a region region closely closely adjacent adjacent to to an OAR. . . . . . 5.4.2 5.4 .2 Mul Multipl tiplee OARs OARs and and PRVs PRVs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.3 5.4 .3 Numb Number er of RVR RVRss  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.4 5.4. 4 Qualific Qualification ation of geomet geometric ric terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.4.1 5.4. 4.1 Tumor umor,, nodal, or metastatic metastatic basis for a target target volume. . . . . . . . . . .

88 88 88 89 89 89 89 89 90 90 91 91 91 91 91

6

5.5 Variat ariation ion of geometry geometry with time. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6 Dose and and dose– dose– vo volum lume e relate related d definit definition ions s. . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6.1 5.6. 1 One-dime One-dimensional nsional dose and dose– volum volumee summariza summarization tion . . . . . . . . . . . . . . . 5.6.1.1 5.6. 1.1 The volume volume receiving receiving at at least a specified specified dose ( V  D) . . . . . . . . . . . . . 5.6.1.2 5.6.1 .2 The least least dose receiv received ed by a specifie specified d volume ( DV ) . . . . . . . . . . . . . 5.6.1. 5.6 .1.3 3 Othe Otherr dose dose meas measure uress  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6.2 Two-dimens wo-dimensional ional dose– volume summari summarization zation   . . . . . . . . . . . . . . . . . . . . . . 5.6.2.1 5.6.2 .1 Two-dimen wo-dimensional sional dose displa displays ys . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6.2. 5.6 .2.2 2 Dose– vo volume lume his histog togram ramss (DVH) (DVH) . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6.2. 5.6 .2.3 3 Dose– ar area ea hist histogr ograms ams (DAH (DAH))  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6.3 5.6 .3 Pre Prescr scribed ibed dos dosee  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6.3.1 5.6.3 .1 Presc Prescribed ribed dose for multiple multiple PTVs   . . . . . . . . . . . . . . . . . . . . . . . . . 5.6.3.2 5.6.3 .2 Presc Prescribed ribed dose dose for multiple multiple treatment treatment segments segments (segment (segment dose) dose)   . . 5.6.4 5.6 .4 Rel Relati ative ve dos dosee  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6.5 Trea reated ted volume   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6.6 5.6 .6 Co Confor nformity mity inde index x (CI) (CI)  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6.7 Irra Irradiated diated volume (at a specified specified dose) dose)  . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

91 92 92 92 93 93 93 93 93 93 93 94 94 94 94 94 94

TREA TR EATM TMEN ENT T PLA PLANN NNIN ING G . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . .

95

6.1 Introd Introduct uction ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 What is is different different about about planning planning proton-b proton-beam eam 6.2.1 Heter Heterogeneit ogeneities ies   . . . . . . . . . . . . . . . . . . . . . . . 6.2.2 Beam-del Beam-delivery ivery techniqu techniques es  . . . . . . . . . . . . . . .

. .. . . . .. . . . therapy? ther apy? . . . . . .. . . . .. . . . . .. . . . . .. .

. .. . . . .. . . . .. .. .. .. .. .. .. . . . .. . . . .. . . . . . . .. . . . . .. . . .

95 95 95 96

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

6.2.3 .3 6.2

Single Sing le bea beams ms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.3.1 Inver 6.2.3.1 Inverse se beam design   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.3.2 6.2. 3.2 Selecti Selection on of beam direc directions tions   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2. 6. 2.3. 3.3 3 Th Thee PT PTV  V   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.3. 6.2 .3.4 4 The pro proton ton RBE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.3.5 6.2. 3.5 The design of beam-modify beam-modifying ing devices devices   . . . . . . . . . . . . . . . . . . . . . . 6.2.3. 6.2 .3.6 6 Rep Repaint ainting ing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.3. 6.2 .3.7 7 Dos Dosee alg algori orithms thms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plans Pl ans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.4. 6.2 .4.1 1 Num Number ber of bea beams ms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.4.2 6.2. 4.2 Intens Intensity-modul ity-modulated ated proto proton n therapy therapy (IMPT) (IMPT) . . . . . . . . . . . . . . . . . . 6.2.4. 6.2 .4.3 3 Ima Imaging ging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.4.4 6.2. 4.4 Posit Positioning ioning accurac accuracy y, immobilization, immobilization, and localization localization   . . . . . . . . . 6.2.4.5 6.2. 4.5 Uncert Uncertainty ainty analys analysis is  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

96 96 96 96 97 97 97 97 97 97 98 98 98 98

6.2.4.6 6.2. 4.6 Target volume size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.5 Qua 6.2.5 Quality lity ass assura urance nce . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3 The pat patien ient’ t’s s anat anatomy omy   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4 Het Hetero erogen geneiti eities es  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.1 6.4 .1 Int Intro roduct duction ion   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.2 6.4. 2 Inter Interactio actions ns of of protons protons in matter matter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.2. 6.4 .2.1 1 Mas Masss thic thickne kness ss   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.2. 6.4 .2.2 2 Ene Energy rgy los losss   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.2.2.1 6.4.2 .2.1 Water-e ater-equivale quivalent nt density . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.2.3 6.4. 2.3 Multiple Coulo Coulomb mb scat scattering tering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.3 6.4. 3 Bulk hetero heterogeneity geneity interse intersecting cting the the full beam beam . . . . . . . . . . . . . . . . . . . . . . . . 6.4.4 6.4. 4 Bulk hetero heterogeneity geneity partiall partially y intersecting intersecting the beam beam . . . . . . . . . . . . . . . . . . . . 6.4.5 6.4. 5 Comp Complexly lexly struc structure tured d heteroge heterogeneities neities   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.6 6.4. 6 Comp Compensati ensation on for heter heterogeneit ogeneities ies  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

99 99 99 100 10 0 100 10 0 100 10 0 100 10 0 100 10 0 100 10 0 101 10 1 101 10 1 102 10 2 102 10 2 103 10 3

6.2. 6. 2.4 4

6.4.6.1

6.5

6.6

Conversion from Conversion from CT Hounsfie Hounsfield ld number to water-eq water-equivalent uivalent density  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.6.1.1 6.4.6 .1.1 Direc Direct-fit t-fit method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.6.1.2 6.4.6 .1.2 Stoich Stoichiometric iometric method   . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.6.1.3 6.4.6 .1.3 Confi Confirmatio rmation n of calibr calibration ation . . . . . . . . . . . . . . . . . . . . . . . 6.4.6. 6.4 .6.1.4 1.4 Acc Accura uracy cy  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.6.2 6.4. 6.2 Design of compen compensator satorss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.6. 6.4 .6.2.1 2.1 Cho Choice ice of rang rangee  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.6.2.2 6.4.6 .2.2 Comp Compensato ensatorr design close close to, and outside, outside, the the projected projected target boundary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.6.2.3 6.4.6 .2.3 Real and virtual compen compensator satorss  . . . . . . . . . . . . . . . . . . . . 6.4.6.2.4 6.4.6 .2.4 The effect effect of an air air gap between between compensa compensator tor and patien patientt   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.6.2.5 6.4.6 .2.5 High-Z heter heterogeneit ogeneities ies   . . . . . . . . . . . . . . . . . . . . . . . . . . Design of indivi individual dual proton beams  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.1 6.5. 1 Comp Compensati ensation on for heter heterogeneit ogeneities ies  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.1.1 6.5. 1.1 Choosi Choosing ng beam direc directions tions  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.1.2 6.5. 1.2 Tangent angential ial irradi irradiation ation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.2 6.5. 2 Categ Categories ories of models models for for dose computa computation tion . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.2.1 6.5. 2.1 Uniform Uniform-intensit -intensity y beam algori algorithms thms   . . . . . . . . . . . . . . . . . . . . . . . . 6.5.2.2 6.5. 2.2 Penci Pencil-beam l-beam algorit algorithms hms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.2.3 6.5. 2.3 Monte Carlo algorit algorithms hms   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.3 6.5. 3 Normaliz Normalization ation and the calcula calculation tion of monitor units . . . . . . . . . . . . . . . . . . . Design of groups groups of beams: beams: the treatme treatment nt plan . . . . . . . . . . . . . . . . . . . . . . . . 6.6.1 6.6. 1 Trea reatment tment goals and constr constraints aints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.6.1.1 6.6. 1.1 Setting goal(s goal(s)) and const constraints raints . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.6.1.2 6.6. 1.2 Estab Establishing lishing a score score combinin combining g target-volum target-volumee and normal-tissue effects  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

103 103 104 10 4 104 10 4 104 10 4 105 10 5 105 10 5 106 10 6 106 106 107 10 7 107 107 107 10 7 108 10 8 108 10 8 108 10 8 108 10 8 108 10 8 109 10 9 109 10 9 109 10 9 110 110 110 110 110

 

CONTENTS

6.6.1.3 .3 Target-v arget-volume olume goals goals and and constraint constraintss . . . . . . . . . . . . . . . . . . . . . . . . 6.6.1 6.6.1.4 6.6.1 .4 NormalNormal-tissue tissue const constraint raintss  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.6.2 Appro Approaches aches to trea treatment tment design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.6.2.1 6.6.2 .1 Uniform Uniform-intensity -intensity radia radiation tion thera therapy py . . . . . . . . . . . . . . . . . . . . . . . . 6.6.2.2 6.6.2 .2 Intensi Intensity-modul ty-modulated ated radiatio radiation n therapy therapy (IMRT) (IMRT) . . . . . . . . . . . . . . . . Plan Pl an as asse sess ssme ment nt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.7.1 Inspec Inspection tion of of the dose distri distribution bution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.7.2 6.7 .2 Clin Clinical ical feas feasibi ibility lity  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.7.3 Dose-su Dose-summarizing mmarizing quantit quantities ies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plan Pl an co comp mpari ariso son n  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.8.1 Plan compar comparison ison by inspect inspection ion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.8.2 Automa Automated ted plan compar comparison ison   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plan Pl an op opti timi miza zati tion on  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

111 111 111 111 112 112 112 113 113 113 114 114 115

6.10 Com Comparison parison of of uniform-inte uniform-intensity nsity versus versus IMPT IMPT treatment treatment plans plans  . . . . . . . . 6.11 6.1 1 Special Special techni techniques ques   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.11.1 6.1 1.1 Intra Intraocular ocular trea treatments tments with protons protons. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.11.1.1 6.1 1.1.1 Model of the patien patientt anatomy anatomy  . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.11.1.2 6.1 1.1.2 Beam simula simulation tion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.11.1.3 6.1 1.1.3 Output of the the planning planning proc process ess . . . . . . . . . . . . . . . . . . . . . . . . . . 6.11.1.4 6.1 1.1.4 Prese Presentatio ntation n of resu results lts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.11.1.5 6.1 1.1.5 Trea reatment tment examp example le   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.11.2 6.1 1.2 Stere Stereotac otactic tic treatme treatments nts with prot protons ons   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.11.2.1 6.1 1.2.1 Immobil Immobilizatio ization n  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.11.2 6.1 1.2.2 .2 Ima Imaging ging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

115 118 118 118 118 118 118 118 119 119 121 12 1

6.7 6. 7

6.8 6. 8

6.9 6. 9

7

6.11.2.3 6.11.2 .3 Plan Planning ning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.11.2.4 6.1 1.2.4 Trea reatment tment   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

121 121 122 12 2

MOTI MO TION ON MA MANA NAGE GEME MEN NT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

123 12 3

7.1 Motion Motion of, and and withi within, n, the the patien patientt  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 Support and immob immobiliza ilization tion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.1 Proto Proton-specific n-specific aspect aspectss of immobili immobilizatio zation n  . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.2 Immobi Immobilizatio lization n techniqu techniques es   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3 Loc Locali alizat zation ion  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3.1 Localiz Localization ation based on skin marks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3.2 Localiz Localization ation based on bony bony anato anatomy my . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3.3 Localiz Localization ation relati relative ve to the the immobilizatio immobilization n device   . . . . . . . . . . . . . . . . . . . . 7.3.4 Localiz Localization ation based based on identificati identification on of target-volum target-volumee markers markers or the tumor itself  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.4 Veri erifica fication tion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.4.1 Verifica erification tion using radiog radiography raphy  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

123 123 123 12 3 123 12 3 124 12 4 124 12 4 124 12 4 124 12 4 125

7.4.2 Verifica erification tion using positron positron emission emission tomogra tomography phy  . . . . . . . . . . . . . . . . . . . . . Organ mo Organ motio tion n. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.5.1 The measur measurement ement of organ organ motion  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.5.2 Organ motion motion in the absence absence of special measur measures es . . . . . . . . . . . . . . . . . . . . . 7.5.3 Organ motion motion under under conditions conditions of respira respiration tion gating gating . . . . . . . . . . . . . . . . . . 7.5.4 Organ motion with tumor tra tracking cking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.6 Comp Compensat ensation ion for for patient patient and organ organ motion motion . . . . . . . . . . . . . . . . . . . . . . . . . . 7.6.1 7.6. 1 Margin Marginss at the peripher periphery y of the CTV CTV or OARs: OARs: lateral lateral margins margins  . . . . . . . . . . 7.6.2 7.6 .2 Mar Margin ginss at the peripher periphery y of the CTV or OARs: OARs: margin margin in depth. . . . . . . . . . 7.6.3 7.6. 3 Dose variatio variation n within the CTV CTV and OARs: OARs: interplay interplay effects . . . . . . . . . . . . . 7.6.3.1 7.6. 3.1 Experim Experimental ental observa observation tion of interpl interplay ay effects effects  . . . . . . . . . . . . . . . . . 7.6.3.2 7.6. 3.2 Repain Repainting ting to reduce reduce the the influence influence of interplay interplay effects   . . . . . . . . . . 7.7 Co Concl nclusi usion on  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

125 126 126 126 12 6 126 126 127 12 7 127 12 7 127 128 128 129 130 130 13 0

7.5 7. 5

125 125 125 12 5 125 12 5

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

8

ESTIMA EST IMATION TION AND AND PRESEN PRESENT TATION OF UNCER UNCERT TAIN AINTY TY IN THE DELIVE DELIVERED RED DOSE DOSE . . . . 8.1 8.2 8.3 8.4

9

131

The inevi inevitabili tability ty of uncer uncertainty tainty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The estim estimation ation of uncert uncertainty ainty   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The prese presentati ntation on of uncerta uncertainty inty. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Recommenda Recom mendations tions for the consider consideration ation and and reporting reporting of uncertainty uncertainty  . . .

131 131 131 13 1 132 13 2 134

QUAL QU ALIT ITY Y ASS SSU URA RANC NCE E.......................................................

135 13 5

9.1 9.2 9.3 9.4

Proton beambeam-deliv delivery ery sys systems tems   . . . . . . . . . . . . . . . . . . . . Patient Pati ent position positioning ing and and immobil immobilizati ization on . . . . . . . . . . . . Trea reatment-p tment-planni lanning ng sys systems tems . . . . . . . . . . . . . . . . . . . . . . Exampl Exa mples es of peri periodi odic c check checks s. . . . . . . . . . . . . . . . . . . . . .

.. . . . .. . . . . .. . . . .. . . .. . . . .. . . . .. . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . .. . . . .

135 135 137 13 7 137 13 7 139 13 9

10 PR PRES ESCR CRIB IBING ING,, RECOR RECORDIN DING, G, AND AND REPO REPORT RTING ING TRE TREA ATM TMEN ENT T .....................

141

10.1 10.1 10.2

Introducti Introd uction on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Generall recommenda Genera recommendations tions for prescrib prescribing, ing, recording recording,, and reporting  reporting   . . . 10.2.1 10.2 .1 The components components of prescr prescribing, ibing, recording recording,, and reporting reporting a  patient’s treatment  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.2.2 10. 2.2 Plan Planning ning aims  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

141 141 141 141 141 141 14 1

10.2.2.1 Specify 10.2.2.1 Specifying ing planning aims . . . . . . . . . . . . . . . . . . . . . 10.2.2.2 10.2. 2.2 NormalNormal-tissue tissue const constraint raintss . . . . . . . . . . . . . . . . . . . . . 10.2.2.3 10.2. 2.3 Selecti Selection on of treatment treatment approac approach: h: the treatme treatment nt plan. 10.3 Presc Prescribin ribing g proton proton-beam -beam thera therapy py  . . . . . . . . . . . . . . . . . . . . . . . . . 10.3.1 10.3 .1 Gener General al approa approaches ches to presc prescribing ribing . . . . . . . . . . . . . . . . . . . . . . 10.3.2 10. 3.2 The pr presc escript ription ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

10.4

10.5

10.6

10.7

. . .. . . . .. . . . .. . . . . .. .. .. .. . . . . .. . . . . . . .. . . . .. .........

141 141 142 14 2 143 14 3 143 14 3 143 14 3 143 14 3

10.3.3 Gener 10.3.3 General al recomme recommendation ndationss for pres prescribing cribing  . . . . . . . . . . . . . . . . . . . . . . . . . 10.3.4 10.3 .4 Appro Approval val of the the prescript prescription ion and technic technical al data   . . . . . . . . . . . . . . . . . . . . Additional Additi onal aspects aspects of prescribi prescribing, ng, recording recording,, and reporting  reporting  .  . . . . . . . . . . . 10.4.1 10.4 .1 Ther Therapy apy equipme equipment nt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.4.2 10.4 .2 Beam-s Beam-shaping haping techniq techniques ues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.4.3 10.4 .3 Techniq echniques ues for dealing dealing with hetero heterogeneitie geneitiess . . . . . . . . . . . . . . . . . . . . . . . . 10.4.4 10. 4.4 Mar Margin ginss  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Recording Recor ding proton proton-beam -beam therap therapy y . . . . .. . . . .. . . . .. . . . .. . . . .. . . . .. . . . . 10.5.1 10.5 .1 The trea treatment tment reco record rd   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.5.2 10.5 .2 The patie patient’s nt’s reco record rd   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reporting Repor ting the the treatment treatment of a single single patient patient . . . . . . . . . . . . . . . . . . . . . . . . . . 10.6.1 10.6 .1 Repor Reporting ting prot proton-beam on-beam thera therapy py  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.6.2 10.6 .2 Pat Patient-spe ient-specific cific repo reports rts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.6.2 10. 6.2.1 .1 The initi initial al medica medicall note note  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.6.2 10. 6.2.2 .2 The com complet pletion ion not notee  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.6.2.3 10.6. 2.3 Summary repo report rt to the the referring referring physic physician ian   . . . . . . . . . . . . . . . . 10.6.2 10. 6.2.4 .4 Case re repor portt  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.6.2.5 10.6. 2.5 Detailed repor reportt to a physician physician. . . . . . . . . . . . . . . . . . . . . . . . . . . Reporting Repor ting proton-b proton-beam eam therapy therapy for a series series of patien patients ts  . . . . . . . . . . . . . . .

143 143 143 14 3 149 14 9 149 14 9 149 14 9 149 14 9 149 14 9 149 14 9 149 14 9 149 14 9 149 14 9 149 14 9 150 15 0 150 15 0 150 15 0 150 15 0 150 15 0 150 15 0 150 15 0

 APPENDIX  APPEND IX A

IMPLEMENTA IMPLEMENT ATION OF THE TRS 398 CODE OF PRA PRACTICE CTICE FOR IONIZA IONIZATION TION CHAM CH AMBE BER R DOSI DOSIME METR TRY Y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 151 1

 A.1

151 151 151 15 1 153 15 3 153 15 3 153 153 15 3

Calibration of ionization chambers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  A.1.1 Calibration in a   60Co beam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  A.2 Reference dosimetry in the user proton beam   . . . . . . . . . . . . . . . . . . . . . . . .  A.2.1 Determination of the absorbed dose to water   . . . . . . . . . . . . . . . . . . . . . . . .  A.2.2 Practical considerations for measurements in the proton beam  . . . . . . . . . .  A.2.3 Corrections for f or influence quantities   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 

CONTENTS

 A.3

 A.2.3.1 Temperature, pressure, and humidity   . . . . . . . . . . . . . . . . . . . . . .  A.2.3.2 Electrometer calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  A.2.3.3 Polarity effect  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  A.2.3.4 Ion recombination   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dosimetry worksheet (IAEA, 2000)  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

153 154 15 4 154 15 4 154 15 4 157 15 7

 APPENDIX B

CLINICAL CLINIC AL EXAMPLE EXAMPLES S .............................................

159

Case Cas e nu numbe mber r B.1 Uve Uveal al melan melanoma oma   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General patient information   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Medical evaluation of presenting lesion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General medical evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment intent   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Patient immobilization and positioning   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment prescription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

160 160 160 16 0 160 16 0 160 16 0 160 16 0 161 16 1 161 16 1 161 16 1

Treatment technique   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Quality assurance   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Equipment  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Total doses delivered  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Patient status at completion of treatment   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Addendum: technical information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Case numb number er B.2 Adenoca Adenocarcin rcinoma oma of pros prostate tate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General patient information   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Medical evaluation of presenting lesion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General medical evaluation Treatment intent   . . . . . . . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Treatment planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Patient immobilization and positioning   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment prescription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment technique   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Quality assurance   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Equipment  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Total doses delivered  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Patient status at completion of treatment   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Addendum: technical information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Case Cas e number number B.3 B.3 Car Carcin cinoma oma of of lung  lung . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General patient information   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Medical evaluation of presenting lesion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

162 162 162 16 2 162 16 2 163 16 3 163 16 3 163 16 3 164 16 4 164 16 4 164 16 4

General medical evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment intent   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Patient immobilization and positioning   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment prescription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment technique   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Quality assurance   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Equipment  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Total doses delivered  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Patient status at completion of treatment   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Addendum: technical information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Case numb number er B.4 B.4 Acous Acoustic tic neuro neuroma ma  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General patient information   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Medical evaluation of presenting lesion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General medical evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment intent   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

168 168 168 16 8 168 16 8 169 16 9 169 16 9 169 16 9 170 17 0 171 17 1 171 17 1 171 17 1 171 17 1 172 17 2 172 17 2 172 17 2 172 17 2 172 17 2

16 164 4 164 16 4 164 16 4 164 16 4 165 16 5 165 16 5 165 16 5 165 16 5 166 16 6 167 16 7 167 16 7 168 16 8 168 16 8 168 16 8

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

Treatment planning  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Patient immobilization and positioning   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment prescription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment technique   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Quality assurance   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Equipment  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Total doses delivered  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Patient status at completion of treatment   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Addendum: technical information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Case numb number er B.5 B.5 Medull Medulloblas oblastoma toma ( pedia pediatric) tric) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General patient information   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Medical evaluation of presenting lesion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

172 172 174 17 4 174 17 4 175 17 5 175 17 5 175 17 5 175 17 5 175 17 5 176 17 6 177 17 7 177 17 7 177 17 7

General medical evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment intent   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment planning  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Patient immobilization and positioning   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment prescription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment technique   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Quality assurance   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Equipment  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Total doses delivered  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Patient status at completion of treatment   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Addendum: technical information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Case numb number er B.6 SkullSkull-base base chordo chordoma ma. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

177 177 177 17 7 177 17 7 178 17 8 179 17 9 181 18 1 181 18 1 181 18 1 181 18 1 182 18 2 183 18 3 185 18 5

General patient information   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Medical evaluation of presenting lesion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General medical evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment intent   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment planning  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Patient immobilization and positioning   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment prescription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment technique   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Quality assurance   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Equipment  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Total doses delivered  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Patient status at completion of treatment   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  Addendum: technical information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

185 185 185 18 5 185 18 5 185 18 5 185 18 5 186 18 6 186 18 6 186 18 6 186 18 6 186 18 6 187 18 7 188 18 8 188 18 8

REFE RE FERE RENC NCES ES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

189 18 9

 

Journal of the ICRU Vol 7 No 2 (2007) Report 78 Oxford University Press

doi:10.1093/jicru/ndn001

THE INTERNATIONAL COMMISSION ON RADIATION UNITS

 AND MEASURE MEASUREMENTS MENTS INTRODUCTION The International Commission on Radiation Units and Mea Measur sureme ements nts (IC (ICRU RU), ), sinc since e its inc incept eption ion in 1925, has had as its principal objective the development of internationally acceptable recommendations regarding: (1) quantities and units of radiation and radioactivity, (2)) pr (2 proc oced edur ures es su suit itab able le fo forr th the e me meas asur urem emen entt and app applic licat ation ion of th these ese qua quanti ntitie tiess in cli clinic nical al radiology and radiobiology, and (3) physical data needed in the application of these proce pr ocedu dures res,, th the e use of whi which ch ten tends ds to ass assur ure e uniformity in reporting. The Commission also considers and makes similar types of recommendations for the radiation protection tio n fiel field. d. In th this is con connec nectio tion, n, its wor work k is car carrie ried d out in clo close se coo cooper perat ation ion wit with h th the e Int Intern ernat ation ional al Commission Commis sion on Radiologi Radiological cal Protect Protection ion (ICRP). POLICY 

is in inad advi visa sabl ble e fr from om a lo long ng-t -ter erm m vi view ewpo poin int; t; it endeav ende avors ors to base its deci decisions sions on the long-range long-range advantages to be expected. The ICR ICRU U inv invite itess and we welco lcomes mes con const struc ructiv tive e comm co mmen ents ts an and d su sugge ggest stio ions ns re rega gard rdin ing g it itss re reccommend omm endat ation ionss and re repor ports. ts. Th These ese ma may y be tr trans ans-mitted to the Chairman. CURRENT PROGRAM The Th e Com ommi miss ssio ion n re reco cogn gniz izes es it itss ob obli liga gati tion on to provid prov ide e gu guid idan ance ce an and d re reco comm mmen enda dati tion onss in th the e area areas of radiati radi ation on therapy , radi radiatio ation n protecti on, and sth the e com compil pilat ation ionther of apy, data da ta import imp ortant antpro totection, these th ese fields, fiel ds, and to sci scient entific ific re resea searc rch h and ind indus ustri trial al appl ap plic ica ati tion onss of ra radi dia ati tion on.. In Incr crea easi sing ngly ly,, th the e Commission is focusing on the problems of protection of the patient and evaluation of image quality in dia diagno gnost stic ic ra radio diolog logy y. Th These ese ac activ tiviti ities es do not diminish the ICRU’s commitment to the provision of a rigorously defined set of quantities and units useful in a very broad range of scientific endeavors. The Commission Commission is currently engaged in the formulation of ICRU reports treati treating ng the following subjects:

The IC The ICRU RU en ende deav avor orss to co coll llec ectt an and d ev eval alua uate te the lat lates estt da data ta and inf inform ormat ation ion per pertin tinent ent to the problems of radiation measurement and dosimetry

 Approaches

and to recommend theuse. most acceptable values and techniques for current The Th e Co Commi mmissi ssion’ on’ss rec recomm ommend endat ation ionss ar are e ke kept pt under continual review in order to keep abreast of  the rapidly expanding uses of radiation. The Th e ICR ICRU U fee feels ls tha thatt it is the respons responsibi ibility lity of  nati na tion onal al or orga gani niza zati tion onss to in intr trod oduc uce e th thei eirr ow own n detailed deta iled tech technical nical proc procedur edures es for the deve developme lopment nt and main maintenan tenance ce of sta standar ndards. ds. How However ever,, it urges thatt all cou tha count ntrie riess adh adhere ere as clo closel sely y as pos possib sible le to the internationally-recommended basic concepts of  radiation quantities and units. The Commission feels that its responsibility lies in de deve velo lopi ping ng a sy syst stem em of qu quan anti titi ties es an and d un unit itss

 ExposuresoftoImage Ionizing Radiation  Assessment Quality in Mammography  Design of a Voxel Phantom for Radiation Protection  Dose and Volume Specifications for Reporting  Intra-Cavity Therapy in Gynecology  Dose Distributions in Normal Tissues Distant from the PTV in Radiati Radiation on Ther Therapy apy  Doses from Cosmic-Ray Exposures of Aircrew  Dosimetry Systems for f or Use in Radiation Processing  Fundamental Quantities and Units  Harmonizatio  Harmon ization n of Repo Reporting rting Pat Patient ient Diagnos Diagnostic tic Doses  Image Quality and Patient Dose in Computed Tomography  Key Data for Measurement Standards in the

having the wide having widest st possi possible ble ran range ge of appl applicabil icability ity.. Situa Sit uatio tions ns ma may y ari arise se fr from om tim time e to tim time e whe when n an expedient solution of a current problem may seem advi ad visa sabl ble. e. Ge Gene nera rall lly y sp spea eaki king ng,, ho howe weve verr, th the e Commission feels that action based on expediency

 Dosimetry of Ionizing Radiation  Prescribin  Prescr ibing, g, Rec Record ording, ing, and Rep Report orting ing Inte Intensit nsityy Modulat  Modu lated ed Phot Photon-B on-Beam eam Radi Radiati ation on The Therap rapy y (IMR (IMRT) T)  Prescribing, Recording, and Reporting Ion-Beam Therapy

to

the

Dosimetry

of

Low-Dose

# International Commission on Radiation Units and Measurements 2007

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

Quantitative Aspects of Bone Densitometry  ROC (Receiver Operator Characteristic) Analysis in  Medical Imaging

International Radiation Protection Association International Union of Pure and Applied Physics Unit Un ited ed Na Nati tion onss Ed Educ ucat atio iona nal, l, Sc Scie ient ntifi ificc an and d Cultural Organization

In add additi ition, on, the ICR ICRU U is eva evalua luatin ting g th the e pos possisibility of expanding its program to encompass nonionizing niz ing ra radia diatio tion, n, par partic ticula ularly rly the qua quant ntiti ities es and units aspects. The Com Commiss mission ion cont continua inually lly revi reviews ews radi radiatio ation n science with the aim of identifying areas where can the development of guidance and recommendations make an important contribution. THE ICRU’S RELATIONSHIP WITH OTHER ORGANIZATIONS In addition to its close relationship with the ICRP, the ICR ICRU U has dev develop eloped ed rel relat ation ionshi ships ps wit with h oth other er organizations interested in the problems of radiation quanti qua ntitie ties, s, uni units, ts, and mea measur sureme ements nts.. Sin Since ce 195 1955, 5, the ICRU has had an official relationship with the Worl orld d Heal Health th Org Organi aniza zatio tion n (WH (WHO), O), whe whereb reby y the ICRU is looked to for primary guidance in matters of radiation units and measurements and, in turn, the WHO assists in the worldwide dissemination of  the Co Commi mmissi ssion’ on’ss rec recomm ommend endat ation ions. s. In 1960 1960,, the ICRU IC RU en ente tere red d int into o con consu sult ltat ativ ive e st stat atus us wit with h th the e Intern Int ernat ation ional al Ato Atomic mic Ene Energy rgy Age Agency ncy (IA (IAEA) EA).. The Comm Co mmiss ission ion ha hass a fo form rmal al re rela latio tions nshi hip p wit with h th the e United Nations Scientific Committee on the Effects of Ato Atomic mic Rad Radiat iation ion (UN (UNSCE SCEAR) AR),, whe whereb reby y ICR ICRU U observe obse rvers rs are inv invite ited d to att attend end ann annual ual UN UNSCE SCEAR AR meetin mee tings. gs. The Co Commi mmissio ssion n and the Int Intern ernati ationa onall Organization Organiz ation for Stan Standardiz dardizatio ation n (ISO) inform informally ally exchange notifications of meetings, and the ICRU is formally designated for liaison with two of the ISO tech te chni nica call co comm mmit itte tees es.. Th The e IC ICRU RU al also so en enjo joys ys a strong rela strong relationsh tionship ip with its siste sisterr organiz organization ation,, the Nati Na tion onal al Co Coun unci cill on Ra Radi diat atio ion n Pr Prot otect ectio ion n an and d Measureme Measu rements nts (NCR (NCRP). P). In essenc essence, e, these organizations were founded concurrently by the same indi viduals.  vidual s. Prese Presently ntly, this long-s long-standi tanding ng rela relationsh tionship ip is formal for mally ly ack ackno nowle wledge dged d by a spe special cial liai liaison son agr agreeeement. men t. The ICR ICRU U also corr correspo esponds nds and exc exchan hanges ges final reports with the following organizations: Bureau International de Me´trologie Le´gale Bureau International des Poids et Mesures European Commission Council for International Organizations of Medical Sciences Food and Agri Agricultu culture re Organ Organizat ization ion of the Unit United ed Nations International Committee of Photobiology International Council for Science International Internation al Electrotechn Electrotechnical ical Commission International Labor Organization International Organization for Medical Physics

The Commission has found its relationship with all of these organizations fruitful and of substantial benefit to the ICRU program. OPERATING FUNDS In recen recentt year years, s, prin principal cipal financial support has been bee n pr provi ovided ded by the Eu Europ ropean ean Co Commi mmissi ssion, on, the National Cancer Institute of the US Department of  Health and Human Services and the International  Atomic Energy Agency Agency.. In addition, during the last 10 years, financial support has been received from the following organizations:  American Association of Physicist Physicistss in Medicine Belgian Nuclear Research Centre Canadian Nuclear Safety Commission Eastman Kodak Company Electricite´  de France Fuji Medical Systems Helmholtz Zentrum Mu ¨ nchen Hitachi, Ltd. International Radiation Protection Association International Society of Radiology Ion Beam Applications, S.A. Italian Radiological Association Japan Industries Association of Radiological Systems Japanese Society of Radiological Technology MDS Nordion National Institute of Standards and Technology Nederlandse Vereniging voor Radiologie Philips Medical Systems, Incorporated Radiation Research Society Radiological Society of North America Siemens Medical Solutions  Varian  V arian Medical Systems In add addition ition to the direct monetary monetary supp support ort pro pro- vided by these organizations, many organizations prov pr ovid ide e in indi dire rect ct su supp ppor ortt fo forr th the e Co Comm mmis issi sion on’s ’s program. This support is provided in many forms, including, among others, subsidies for (1) the time of ind indivi ividua duals ls par partic ticipa ipatin ting g in ICR ICRU U ac activ tiviti ities, es, (2) tr trav avel el cos costs ts inv involv olved ed in ICR ICRU U mee meetin tings, gs, and (3) meeting facilities and services. In recognition of the fact that its work is made possible by the generous support provided by all of  the th e or orga gani niza zati tion onss su supp ppor orti ting ng it itss pr prog ogra ram, m, th the e Commission expresses its deep appreciation. Paul M. DeLuca Chairman, ICRU Madison, Wisconsin, USA 

 

Journal of the ICRU Vol 7 No 2 (2007) Report 78 Oxford University Press

doi:10.1093/jicru/ndm022

PREFACE

The current report on proton therapy is the latest in a series of ICRU reports dealing with prescribing, recording, and reporting of external-beam radiation at ion the therap rapy y. Thi Thiss joi joint nt re repor portt of the ICR ICRU U and the IAEA highlights the long and effective collaboration between the two organizations. Rober Ro bertt R. Wils ilson on was the firs firstt to re recog cogniz nize e the potent pot ential ial adv advant antage age of pr proto otons ns and lig light ht ion ionss for exter ex terna nal-b l-beam eam ra radia diatio tion n (W (Wils ilson, on, 194 1946). 6). In thi thiss benchm ben chmark ark pu publi blica catio tion, n, Wils ilson on out outlin lined ed all the advantages that protons and light ions would offer due in large part to their physical dose distributions.

PET/CT image sets provide complementary physiological information. The Th e use of th these ese dr dram amat atic ic adv advanc ances es to ef effec fectt th the e desired improvements improvements in radia radiation tion therapy requires simila sim ilarr ad advan vances ces in th the e del deliv ivery ery of abs absorb orbed ed dos dose. e. Photon treatment from a narrow range of beam direction ti onss wi with th a li limi mite ted d se sele lect ctio ion n of fie field ld sh shap apes es ca can n never achieve the desired outcome. Only with the use of large numbers of smal small, l, varia variable-i ble-inte ntensit nsity y beam beam-letss del let deliv iver ered ed fr from om mul multip tiple le dir direct ection ionss can thr threeeedimensi dim ensional onal abso absorbed rbed dose dis distrib tributio utions ns appr approac oach h the resolution of the imaging data. This technique is

 At the energies used for therapy therapy,, protons interact almo al mosst ex excl clus usiv ivel ely y wi with th ato tomi micc el elec ectr tron onss in Coulomb Cou lomb forc force e collis collisions ions yield yielding ing near nearly ly str straight aight traj tr ajec ecto tori ries es cu culm lmin inat atin ing g in a ra rapi pid d in incr crea ease se in energy loss near the end of range, and thus forming the so-called Bragg peak. Essentially, no energy is deposited beyond the end of range and there is a  very sharp distal dose fall-off. Despite these significantt ph can physi ysical cal dos dose-d e-dist istrib ribut ution ion adv advant antage ages, s, the init in itia iall cl clin inic ical al us use e of pr prot oton onss was mo mode dest st as accelera acce lerators tors capab capable le of pro producin ducing g 200– 300 MeV MeV/u /u protons were limited to large physics laboratories. Even Ev en so, the cli clinic nical al exp experi erienc ence e gr grew ew sy syst stema ematiticall ca lly y, fa faci cili lita tate ted d by th the e ad adve vent nt of ac acce cele lera rato tors rs suitab sui table le for hos hospit pitalal-bas based ed ins instal talla latio tions ns in the mid 1990s.  Anticipating  Anticipatin g the wider clinical use of protons, the Clinical al Pr Prot oton on ICRU IC RU pu publ blis ishe hed d Rep epor ortt 59 59,,   Clinic  Dosimetry Part I: Beam Production, Beam Delivery and an d Me Meas asur urem emen entt of Ab Abso sorb rbed ed Do Dose se, in 1998. Report Rep ort 59 note noted d tha that, t, while protons protons requ require ire some definit defi nition ionss and ap appr proa oache chess tha thatt emp emphas hasize ize the advant adv antage agess of the ph phys ysica icall dos dose e dis distri tribut bution ion,, for the th e mo most st pa part rt pr prot oton on th ther erap apy y co coul uld d be an and d wa wass deli de live vere red d in a ma mann nner er ty typi pica call of co conv nven enti tion onal al radia ra diatio tion n th thera erapy py.. Rad Radia iatio tion n tr trea eatm tment ent of sol solid id tumors tum ors has ad advan vanced ced dr drama amatic ticall ally y sin since ce ICR ICRU U

called intensity-modulated radiation therapy (IMRT). Protons, by virtue of their physical dose distributions, can pro produc duce e thr three-d ee-dimen imension sional al dose dis distrib tributio utions ns simila sim ilarr to typ typica icall IM IMRT RT tr trea eatm tmen entt re regim gimes es usi using ng beambea m-del deliv ivery ery sch schem emes es les lesss com comple plex x th than an ph photo oton n IMRT. Much of the justification for using protons is the rela relativ tive e ease of cre creati ating ng thr three-d ee-dimen imension sional al dose distributions that are highly conformal to the tumor  volume  vol ume with a conco concomita mitant nt red reducti uction on in tota totall dose dose.. Wit ith h som some e inc incre rease ase in bea beam-d m-deli elive very ry com comple plexit xity y, protons can also be used for IMRT. In this changed environment, environm ent, the specificat specification ion of approp appropriate riate therapeutic volumes for reporting treatments requires new approaches approa ches and more elaborate interpret interpretations. ations. Reco Re cogn gniz izin ing g th thes ese e ad adva vanc nces es an and d th thei eirr im impl pliications, ICRU initiated the creation of the present report rep ort dealing with proton proton ther therapy apy and a rela related ted repor re portt dea dealin ling g wit with h hig high-e h-ener nergy gy ph photo oton n IMR IMRT T. Throughout the process of report creation, the committees worked closely to ensure a consistent and coordinated scientific approach. Compar Com pared ed with hig high-e h-energ nergy y pho photon tonss and ele elecctrons, protons exhibit a modest increase in relative biological effectiveness (RBE) related to their somewhatt high wha higher er line linear ar ene energy rgy tr trans ansfer fer (LE (LET) T) val values. ues. The Th e LE LET T va value luess fo forr sw swift ift pr prot oton onss ar are e sim simila ilarr to phot ph oton ons, s, bu butt in incr creas ease e so some mewh what at ne near ar th the e en end d of 

pub publis hed an Repo Re port rt 59. Advan Ad vances ces image ima ge scienc sci ence e have hav elished had enormous enorm ous impact. impa ct. in Four-di Fou r-dimens mensional ional data, dat a, time plus thr three-d ee-dimens imensional ional highhigh-reso resolutio lution n  volume information, largely from computerized tomography (CT) and magnetic resonance imaging (MRI), (MRI ), pro provide vide finel finely y reso resolved lved ana anatomic tomical al deta detail. il. MRI, MR I, pos positr itronon-emi emissi ssion on tom tomogr ograph aphy y (P (PET ET), ), and

ran range. ge.s,Thu Thus, forponse equival equ doses dos dose distridis tribution but ions, thes, res respon seivalent of ent tissue tis sues sesis and similar sim ilar, , in fact fa ct almost clinically indistinguishable from conventional photo photon n ther therapy apy.. How However ever,, in sever several al locati locations ons in the clinical proton beam,   e.g., on the distal edge of  the th e sp sprrea eadd-ou outt Br Brag agg g pe peak ak (S (SOB OBP) P) an and d in th the e later la teral al pen penumb umbra, ra, slo slow w pr proton otonss cre creat ate e mea mean n LET

# International Commission on Radiation Units and Measurements 2007

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

 values well in excess of 10 keV   mm21, introducing a differ dif ferent ent biol biologi ogical cal res respon ponse se to pr proton otonss com compar pared ed with conve conventiona ntionall low-L low-LET ET radiations. radiations. Signifi Significant cant numbers of these higher LET protons are limited to a smal small-v l-volu olume me irr irradia adiated ted reg region ion.. Whi While le the their ir significance is still a matter of debate from a clinical  viewpoint,  viewp oint, this repor report, t, based on long clinical exper experii-

dose-volu dose-v olume me met metric ricss der deriv ived ed fr from om dos dose-v e-volu olume me histogra hist ograms. ms. Such reporting reporting is essen essential tial when the physicalphy sical-absorb absorbed ed dose dis distribu tribution tion inclu includes des stee steep p dose gradients and the therapeutic treatment plans are ar e fu full lly y th thre reee-di dime mens nsio iona nall in sc scop ope. e. In fa fact ct,, clinically-valid comparisons between different therapeutic regimes,  e.g., protons, heavy ions, and IMRT

ence andRBE biological suggests use ofatauniform uni form value valu eevidence, of 1.1 1.10 0 thr throug oughou houtthe t the treat tre ment volume. This is a core recommendation of this report. This RBE value is assumed to be independentt of dos den dose e fr fract action ionat ation ion,, fr fract action ion size size,, and oth other er routine variations in the clinical treatment strategy.  Although  Althou gh these assum assumptions ptions are not stric strictly tly correc correct, t, they are of modest clinical significance as the RBE  value is near unity. unity. In this report, the product of the absorbed dose and RBE is called the RBE-weighted absorbed dose with wit h the spe specia ciall un unit it of gra gray y (Gy (Gy). ). His Histor torica ically lly,, proton therapy centers have often used the annotation of CGE, GyE, or Gy(E) to indicate the RBE-

with high-energy sophisticated thre th reee-di dime mens nsio iona nallphotons, and an d require even ev en this four fo ur-d -dim imen ensi sion onal al recording and reporting. In an appendix, this report gives several examples of clinical reporting that are compr com preh ehen ensi sive ve,, de deta tail iled ed,, an and d ba base sed d on re recom com-mendations contained in the present report. Subse Su bseque quent nt to th the e pu publi blica catio tion n of IC ICRU RU Re Repor portt 59, several issues surrounding dosimetry protocols and values of various physical parameters arose. In some som e wa ways, ys, pr proto oton n dos dosime imetry try pr proto otocol colss are mor more e simplistic than those for energetic photons or electrons tr ons.. Ma Many ny of the com comple plex x iss issues ues of sec second ondary ary charged-particle transport that affect cavity theory are dram dramatic atically ally red reduced. uced. How However ever,, thes these e issue issuess

weighted absorbed dose,   e.g., ‘. . .the RBE-weighted dose was 70 GyE’. Such modifications of a fundamental or special unit are not acceptable in the SI context. Moreover, the use of the symbol E with the connotat conn otation ion of ‘equiv ‘equivalent alent’’ can easily lead to confusion as to what equivalence is being discussed or specified. In the present report, when indicating the abso ab sorbe rbed d do dose se th the e ex expl plici icitt lan langu guag age e su such ch as ‘t ‘the he  D absorbed dose was 63 Gy’ or ‘ D  63 Gy’ is appropriate. priat e. When RBE-weighte RBE-weighted d dose is expr expressed, essed, the ICRU ICR U rec recomm ommend endss the use of: ‘th ‘the e RBE RBE-w -weigh eighted ted  DRBE  70 Gy (RBE)’. In dose was 70 Gy (RBE)’ or ‘ D each case, the unit is Gy and the RBE weighting is indica ind icated ted by exp explici licitt lan langua guage ge or sub subscri scripti pting ng and by app append ending ing ‘(RB ‘(RBE)’, E)’, whi which ch is clea clearly rly sep separa arated ted from fro m the base uni unit. t. The language language is una unambi mbiguo guous us as to whether physical dose or RBE-weighted dose is used. When possible and important for clarity, both RBE-weighted and physical dose should be specified. This situation highlights the issue that RBE corrections are only one of a variety of modifications to the phys ph ysic ical al do dose se ne need eded ed to ac accou count nt fo forr ot othe herr th ther eraapeutic peu tic fea featur tures es suc such h as dos dose e fra fracti ctiona onation tion or tot total al treatment time if one wishes to refer back to a standard therapeutic regime. Over time, the ICRU has systematically defined  various volumes essential for the reporting and

are replaced with greater uncertainty in three fundamental quantities, proton-stopping power, either relat rel ative ive or abs absolu olute, te, pr proto oton n ra range nge,, and th the e mea mean n proto pr oton n ene energy rgy los losss nee needed ded to cre creat ate e an ion ion-p -pair air,, W  /  e. As dis discus cussed sed in thi thiss rep report ort,, ado adopti ption on of the IAEA IAE A dos dosime imetry try pr proto otocol col de descr scribe ibed d in th the e IAE IAEA  A  Absorbed Dose Techni echnical cal Rep Reports orts Serie Seriess No. 398,   Absorbed  Determination in External Beam Therapy Therapy.. An  International Code of Practice for Dosimetry Based on Standards of Absorbed Dose to Water   (2000) is recomm rec ommend ended. ed. Thi Thiss pr proto otocol col wa wass dev develo eloped ped su subbsequent to the publication of Report 59 and is suitable for use in most proton applications. Besides the dosimetry protocol, the treatment of the  W  /  e  value is discussed in detail. For a proton SOBP of the type used in typical therapeutic applications, a range of  proton pro tonss wit with h dif differ ferent ent ene energi rgies, es,   E, is pr pres esen entt at differe diff erent nt loca location tionss in the SOB SOBP P. Sin Since ce the pro proton ton differential   w( E)/  e   value value is not ind indepe epende ndent nt of   E, W ( E)/  e   is no nott co cons nsta tant nt an and d in incr crea ease sess fo forr lo lowe werr proton pro ton spe speeds. eds. The cho choice ice of the mos mostt app approp ropria riate te W ( E)/  e   value value is a cha challen llenge. ge. Th These ese con consid sidera eration tionss and improved measurements of  W  /  e  based on calorimetric techniques techniques subse subsequent quent to the publication publication of  ICRU Report 59, led to improved recommendations of   W  /  e   for prot proton on clinical dosimetry. These recommendations have been fully adopted in the present

reco recordin rding of exte externalrnal-beam beamReports radiati rad iation on treatm ents, seen, forgexample, in ICRU 50, tre 62,atments and 71., Some volumes relate to the tumor and various critical organs, whereas others relate to the uncertainties of position of anatomic volumes with respect to trea tr eatme tment nt bea beams. ms. Re Recog cogniz nizing ing th the e lim limita itatio tion n of  repo re port rtin ing g do dose se at a po poin int, t, th the e pr pres esen entt re repo port rt recommen reco mmends ds the use, recording, recording, and reporting reporting of 

rep report. ort. Atdose thecan prese pr esent moment mom ent,, with proto pr oton n absolute physica phy sicallabsorbed be nt determined an uncerta unc ertaint inty y of    , 3 pe perc rcen entt fo forr re relev levan antt cli clini nical cal conditions.

  ¼

  ¼

Paul M. DeLuca Jr  Andre´  Wambersie Gordon Whitmore 2

2  

Journal of the ICRU Vol 7 No 2 (2007) Report 78 Oxford University Press

doi:10.1093/jicru/ndm041

 ABSTRACT CT  ABSTRA

Proton ther Proton therapy apy is a fast fast-exp -expandin anding g moda modality lity and the number of new facilities is rapidly increasing. The advantage of proton beams relative to conventional tio nal ph photo oton n bea beams ms for ra radia diatio tion n the thera rapy py lie liess in their the ir sup superi erior or dos dose e dis distri tribut bution ions. s. The effi efficac cacy y of  prot pr oton on th ther erap apy y is we well ll es esta tabl blis ishe hed d fo forr se seve vera rall tumo tu morr ca cate tego gori ries es.. Th The e pr pres esen entt re repo port rt pr prov ovid ides es informat infor mation ion neces necessary sary to sta standard ndardize ize tech technique niquess and an d pr proc oced edur ures es an and d to ha harm rmon oniz ize e th the e cl clin inic ical al descri des cript ption ionss of pr proto oton n tr trea eatme tment ntss wit with h tho those se of  other oth er mod modali alitie ties. s. Th The e con concep cepts ts an and d re recom commen men-dations in other ICRU reports concerning radiation therapy are extended to proton therapy. The topics covered here include the rationale for as we well ll as th the e his histor tory y of pr proto oton n the thera rapy py,, pr proto oton n radiation biology, proton-beam delivery and properties, dosimetry, geometric and dose–volume terms, treatment planning, uncertainties in dose delivery, moti mo tion on ma mana nage geme ment nt,, qu qual alit ity y as assu sura ranc nce, e, an and d prescribing, recording, and reporting treatment. In addi ad diti tion on,, si six x cl clin inic ical al ex exam ampl ples es of pr prot oton on-b -bea eam m

therapy are provided to illustrate the application of  the recommendations contained in this report. Recommen Rec ommendat dations ions in the repo report rt inclu include de the use of a generic relative biological effectiveness (RBE)  value of 1.1 and the adoption of the IAEA TRS (Tec (T echn hnic ical al Rep epor ortt Se Seri ries es)) 39 398 8 co code de of pr prac acti tice ce as th the e st stand andard ard pr proto oton n dos dosime imetry try pr proto otocol col.. Th The e conc co ncep eptt of RB RBEE-we weig ight hted ed ab abso sorb rbed ed do dose se ( DRBE, the product of proton-absorbed dose,   D, and proton RBE) RB E) is int intro rodu duced ced to es estim timat ate e the ph photo oton n dos dose e that would produce the same therapeutic effect as the proton-absorbed dose,   D, given under identical conditions.  As the present report describes in some detail the radi radiobiol obiological ogical,, phy physical, sical, tech technical nical,, trea treatmen tmenttplan pl anni ning ng,, an and d cl clin inic ical al as aspe pect ctss of pr prot oton on be beam am therapy, it should be a useful reference for current prac pr acti titi tion oner erss an and d sh shou ould ld al also so pr prov ovid ide e ne new w an and d potential users, as well as other interested readers, with the basic background to enable them to understand the techniques involved in proton therapy.

# International Commission on Radiation Units and Measurements 2007

 

Journal of the ICRU Vol 7 No 2 (2007) Report 78 Oxford University Press

doi:10.1093/jicru/ndm023

EXECUTIVE SUMMARY  The ra The rati tion onal ale e fo forr us usin ing g pr prot oton onss fo forr ra radi dia ati tion on therapy lies in their physical properties (i.e., nearzero dose distal to the target volume and the resultantt cap tan capabi abilit lity y of con confor formin ming g the pla plann nned ed dos dose e more closely to the specified target volume than is feas fe asib ible le by ph phot oton on te tech chni niqu ques es). ). Th The e bi biol olog ogic ical al effects of proton beams have no known or predicted advantages. The depth–dose curve for a monoenergetic proton beam exhibits a relatively flat low-dose entrance entr ance region (the pla plateau teau)) follo followed wed by a sharp high hi gh-d -dos ose e pe peak ak (t (the he Br Brag agg g pe peak ak), ), ju just st be beyo yond nd whic wh ich h th the e pa part rtic icles les lo lose se th the e re rema main inde derr of th thei eirr energy ene rgy in a fe few w mil millim limete eters. rs. Fo Forr pla planni nning ng of con con- ventional proton-beam therapy therapy,, the distribution of  prot pr oton on en ener ergi gies es is de desi sign gned ed to pr prov ovid ide e a ne near ar-uniform dose across the volume of interest,   i.e., in the spr spread ead-ou -outt Br Bragg agg pea peak k (SO (SOBP BP). ). Th There ere is an additiona addi tionall dosedose-dist distribut ribution ion advan advantage tage of prot protons, ons, viz., fo forr de dept pths hs ty typi pica call lly y up to    17–1 17–18 8 cm in tissu tis sue, e, the la later teral al dos dose e fal fall-o l-off ff is st steep eeper er tha than n for photon beams. This provides an additional gain for the tr trea eatme tment nt of les lesion ionss of int interm ermedi ediat ate e dep depth, th, sited close to critical and radiosensitive structures. The Th e we well ll-d -defi efine ned d li limi mitt of pe pene netr trat atio ion n in ti tiss ssue ue depends depe nds on the incident proton-beam proton-beam energy and the den densit sity y of th the e tis tissue sue tr trans ansite ited. d. To irr irradi adiat ate e deep-seated lesions in adult patients, and depending on the clinical application, proton beam energiess of 200– 250 MeV (co gie (corre rresp spond onding ing to ra range ngess in tissue of 26–38 cm) are required. Proton acceleratorss pr tor prov ovidi iding ng 60– 75 Me MeV V bea beams ms (wi (with th ra range ngess of  3–4.5 cm) are used almost exclusively for the treatment of ocular lesions. The clinical outcomes of proton-beam therapy in terms ter ms of loc local al con contr trol ol con confirm firm its effi effica cacy cy for the trea tr eatme tment nt of sev severa erall typ types es of les lesion ions, s,   e.g., uv uveal eal

non-sm nonsmal alll-ce cell ll lu lung ng ca canc ncer er,, he hepa pati ticc ca canc ncer ers, s, parana par anasal sal sin sinus us car carcin cinoma omas, s, and sol solid id tu tumor morss in pedi pe dia atr tric ic pa pati tien ents ts.. Th The e pr prob obab able le ef effic fica acy of  protonpro ton-beam beam ther therapy apy with respect to othe otherr trea treattment modalities is being assessed by the method of  compar com parat ativ ive e tr trea eatme tment nt pla planni nning ng com combin bined ed wit with h biom bi omat athe hema mati tica call mode mo delin ling. g. Al Alth thou ough gh pr prot oton on therapy is a relatively new external-beam modality, it is a rapidly proliferating field. More than 53 000 patients have been treated to date (December 2007) with great success; the greatest number of patients being those with uveal melanoma, prostate cancer, and beni benign gn intr intracr acrania aniall lesion lesions. s. Prot Proton on ther therapy apy is now firmly established in the radiation oncologists’ armamentarium. However, at present, there are no recognized uniform standards for dose prescription and treatment description for proton therapy. Such stand st andard ardss ar are e ess essent ential ial for the des design ign of cli clinic nical al trials, for the evaluation of clinical data from the  various proton therapy facilities, and for assessing the efficacy of proton radiation therapy relative to photon or other radiation treatments. These include conv co nven enti tion onal al th ther erap apie iess an and d th the e ne newl wly y em emer ergi ging ng techniques techn iques such as intens intensity-mod ity-modulat ulated ed radia radiation tion therapy (IMRT) and tomotherapy. Such comparisons are ar e ess essent entia iall fo forr as asse sess ssing ing th the e effi effica cacy cy of pr prot oton on therap the rapy y for the tr treat eatmen mentt of ne new w site sites. s. Imp Import ortant ant aspects to ensure uniform standards of proton dose presc pr escri ript ption ion in incl clud ude e th the e ad adop opti tion on of a un unifo iform rm dosimetry protocol and the specification of a generic clinical relative biological effectiveness (RBE). The present report provides information necessary to standardize techniques and procedures and to har harmon monize ize th the e cli clinic nical al des descri cripti ptions ons of pr proto oton n trea tr eatm tment entss wit with h tho those se of oth other er mod modali alitie ties, s, re reccommends the adoption of a uniform dosimetry pro-

melanomas melano mas,, cho chordo rdomas mas and cho chondr ndrosa osarc rcoma omass of  the th e sk skul ulll ba base se an and d ax axia iall sp spin ine, e, sm smal alll tu tumo mors rs of 

tocol, pr tocol, propo oposes ses the use of a gen generi ericc cli clinic nical al RB RBE, E, and describes the radiobiological, physical, technical, treatment-planning, and clinical aspects so as

NOTE: Sections of the text of ICRU Reports 50, 59, 62, and 71 (ICRU (IC RU,, 199 1993b; 3b; 1998; 199 1999; 9; 200 2004) 4) and IAE IAEA A Re Repor portt TRS 398 (IAEA, 2000) are quoted verbatim or in summary form in the present report. Figures and tables from these reports have also been be en co copi pied ed or ad adap apte ted. d. Th Thes ese e es estr trac acts ts ar are e no nott al alwa ways ys speci spe cific ficall ally y re refer feren enced ced or ac ackno knowl wledg edged ed but ar are e use used d wi with th permission.

to provide new users with the basic background to enable them to understand the techniques involved in proton therapy. ICRU ICR U Re Repor ports ts 50 and 62 (IC (ICRU RU,, 199 1993b; 3b; 199 1999) 9) described the prescribing, recording, and reporting of photon-beam therapy, whereas ICRU Report 71

# International Commission on Radiation Units and Measurements 2007

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

(ICR (ICRU, U, 2004) desc described ribed elect electronron-beam beam ther therapy apy.. A  similar ICRU report is presently being prepared on confor con formal mal ph photo oton n the therap rapy y and IM IMRT RT.. Th The e defi defi-nitions of geometric and other therapy terms in the present pres ent rep report ort are consi consisten stentt with the defin definition itionss given giv en in th these ese re repor ports. ts. ICR ICRU U Re Repor portt 59 (IC (ICRU RU,, 1998) and IAEA Report TRS 398 (IAEA, 2000) on proton prot on dosim dosimetry etry wer were e consi considere dered d in makin making g recommenda omme ndation tionss rega regardin rding g a unif uniform orm inte internat rnationa ionall proton dosimetry protocol. Throughout the present repo re port rt,, th the e di diff ffer eren ence cess be betw twee een n th the e pr proc oces esse sess involved in proton-beam therapy and conventional therapies are highlighted.

Radiation biology

 Although it has been shown that the RBE (relativ (relative e biological biologi cal eff effect ective ivenes ness) s) rel relat ative ive to hig high-e h-ener nergy gy photons phot ons is sligh slightly tly,, but signi significan ficantly tly,, grea greater ter than unit un ity y in ne near arly ly al alll st stud udies ies,, th the e ra radi diob obio iolog logic ical al charact char acterist eristics ics of pro proton ton beam beamss pro provide vide no know known n clinical advantages. Nevertheless, the statement of  proton doses and the choice of an appropriate RBE  value to relate proton and photon doses are important issues. †

†

†

  Th The e RB RBE E fo forr pr prot oton on be beam amss is de defin fined ed in th the e present report as the ratio of the photon dose to the proton dose required to give the same biological effect under identical irradiation conditions.   RBE is a function of LET (linear energy transfer).. Th fer) The e LE LET T of a mo mono noen ener erge geti ticc 25 250 0 MeV  prot pr oton on be beam am is    8 keV  mm21 fo forr de dept pths hs of  pene pe netr trat atio ion n of 2. 2.5 5 cm to 27 cm cm.. Th Ther ere e is ve very ry slig sl ight ht de depe pend nden ence ce of RB RBE E on LE LET T ov over er th that at range. The LET increases steeply at the end of  range,   i.e., in the Bragg peak, with an accompanying rise in RBE to about 1.5 at its maximum.  RBE determinations have been made in the clinivitro o call en ca ener ergy gy ra rang nge e fo forr a wi wide de va vari riet ety y of   in vitr and   in vivo   systems for diverse end points. The mean me an RB RBE E fr from om th the e pu publ blis ishe hed d st stud udie iess is 1. 1.1. 1.  Available  A vailable data are consistent with a tissueindependent generic RBE value of 1.1.

†

†

  Th Ther ere e ar are e no pr prot oton on RB RBE E de dete term rmin inat atio ions ns fo forr human tissues. Clinical RBE values must therefore be deriv derived ed from labor laborator atory y inve investi stigati gations. ons. No clinical experience has been reported indicating an RBE different different from 1.1. Accor According dingly ly,, the recommendation is that proton radiation therapy be planned using a generic RBE value and that  value be 1.1.  Absorbed dose [in units of gray (Gy)] is a fundamental ment al quan quantity tity in all ther therapeu apeutic tic appl applicat ications, ions, but it is not a sufficient predictor of therapeutic outcome. All relevant treatment parameters such as absorbed dose, fractionation schedule, overall trea tr eatme tment nt tim time, e, bea beam m qua qualit lity y,   etc. sho houl uld d be specifi spe cified ed.. Pr Previ evious ously ly the qu quant antity ity “eq “equiv uivale alent nt dose” with a variety of units [CGE, GyE, Gy(E),  etc.] has been used to describe the product of the prot pr oton on ab abso sorb rbed ed do dose se an and d th the e pr prot oton on RB RBE. E. “Equiva “Equ ivale lent nt do dose se”” is a qu quan anti tity ty re rese serv rved ed fo forr radiati rad iation on prot protectio ection n purp purposes oses whil while e the empl employoyment of modified unit symbols is not permitted ´ s /The in the SI ( Le Syste` me International d’Unite´  International System of Units). In order to overcome com e the these se pr probl oblems ems it is re recom commen mended ded in the present pres ent repo report rt tha thatt the quan quantity tity RBE RBE-wei -weighte ghted d absorbed absor bed dose dose,,   DRBE   [in un unit itss of Gy (R (RBE BE)] )] be used to describe the product of proton absorbed dose,   D, and the proton RBE.   DRBE   [Gy (RB (RBE)] E)] is th ther eref efor ore e th the e ph phot oton on ab abso sorb rbed ed do dose se th tha at would produce the same therapeutic effect as a proto pr oton n abs absorb orbed ed dos dose e of   D   (Gy), (Gy), giv given en un under der identi ide ntical cal con condit dition ions. s. Thr Throug oughou houtt th the e pr prese esent nt report all doses are explicitly given as absorbed dose dos e (Gy (Gy)) or RB RBE-w E-weig eight hted ed abs absorb orbed ed dos dose e [Gy (RBE)].

Beam delivery and properties

 A proton therapy facility consists of three main equipment components, viz. an accelerator, a beam transport system, and a treatment delivery system. In the latter system the proton beam is tailored to suit the lesion being treated. If the aim of the facility is to irradiate safely any lesion in the body of an

†

†

  The The av avail ailabl able e da data ta re revea veall no var varia iatio tion n in RB RBE E with wi th de dept pth h fr from om th the e pl plat atea eau u an and d ac acro ross ss th the e SOBP, excepting for the terminal 5 mm to 10 mm of the SOBP. Over that narrow range, the RBE is some 8 percent to 10 percent higher.   On the declining distal edge of the SOBP sharp relative increments (up to 50 percent) in the RBE have been observed. This results in an effective incr in crea ease se in ran ange ge of  1 mm an and d  2 mm fo forr proton beams in the energy ranges below 75 MeV  and above 150 MeV, respectively.

adult with tumo tumoricid ricidal al doses doses,, pro proton ton ranges of at least 26–38 cm in tissue (corresponding to proton energies of 200–250 MeV) are required. Intensities of the order of (3 to 6)  1010 particles per second (5 nA to 10 nA nA)) ar are e re requi quire red d to del deliv iver er abs absorb orbed ed doses of about 2 Gy uniformly to a target volume of  one liter in one minute. The exact requirements of  beam energy and intensity depend on details of the beam be am-d -del eliv iver ery y sy syst stem em us used ed.. Bo Both th pa pass ssiv ive e an and d acti ac tive ve be beam am-d -del eliv iver ery y sy syst stem emss ar are e em empl ploy oyed ed to deli de live verr th the e re requ quir ired ed do dose se to th the e ta targ rget et.. Pr Prot oton on 6

 

EXECUTIVE SUMMAR SUMMARY  Y 

ranges are adjusted by the use of degraders (with a cyclotron) or by changing the beam energy (with a synchrotron). †

†

†

†

†

 A typical passive beam-modification arrangement employ emp loyss a dou double ble-sc -scat atter tering ing sy syst stem em (to spr spread ead the beam laterally), a rotating variable-thickness propeller (or ridge filter) to spread out the Bragg peak in depth, and a field-specific collimator.  Dynamic  Dyna mic beam beam-sca -scanni nning ng sy syste stems ms can be used to achiev ach ieve e the des desire ired d lat later eral al dos dose e dis distri tribut bution ion at speci sp ecific fic de dept pths hs by ma magn gnet etica icall lly y de defle flect ctin ing g th the e beam bea m ac acros rosss the tar target get.. Sev Severa erall diff differe erent nt tech tech-niques can be used to change the beam penetration depth. The advantages of beam scanning are flexibility (no patient-specific devices required), IMPT (int (i nten ensit sityy-mo modu dula late ted d pr prot oton on th ther erap apy) y) ca can n be undertaken, there and is better dose conformation to the target volume, the background dose to the patient and the activation of beam-line elements are reduced. However, there are specific problems relate rel ated d to pa patie tient nt and organ mot motion ion,, whi which ch can can,, however, be minimized by multiple “repainting” of  the target volume. Scanning beams are not suitable for treating small lesions.  To achieve the potential precision of proton-dose deli de live very ry,, su supp ppor ortt an and d im immo mobi bili liza zati tion on of th the e patient pat ient is of prima primary ry impo importanc rtance. e. How However ever,, the techniques used are, in most cases, very similar to tho those se use used d in con conven ventio tional nal hig high-t h-tech echnol nology ogy therapy.  Dedicated beam-delivery systems are required for special spec ial tre treatm atment ent tech technique niques, s, such as eye treatmentss ( prot ment proton on energ energies ies , 70 MeV MeV), ), an and d st ster ereoeotactic radiosurgery and radiothera radiotherapy py..  A rotating gantry provides flexible beam delivery. Several different types of gantries are in use, and they typically weigh about 200 tons. For accurate beam delivery, delivery, rot rotatio ation n about a mech mechanica anicall isosphe sp here re wi with th a di diam amet eter er of le less ss th than an 2 mm is required.

†

†

electromagnet electroma gnetic ic inter interacti actions ons domi dominat nate e at ther theraapeutic energies.   Many Many defi definit nition ionss of bea beam m pr prope operti rties es and bea beam m parameters are the same for proton beams as for photon phot on beam beams. s. Spec Specific ific prot proton-b on-beam eam defin definition itionss are required for describing scanning beams and depth-dose character characteristics. istics.   The radiation quality of a proton beam is determine mi ned d by th the e en ener ergy gy di dist stri ribu buti tion on,, wh whic ich h ca can n affect aff ect vario various us beam characteris characteristics tics such as the entr en tran ance ce (pla (plate teau au)) do dose se,, di dist stal al do dose se fa fall ll-o -off ff,, penumbra penu mbrae, e, and the rad radiati iation on dose outside the treatment field. The latter is particularly important for the treatment of pediatric cases. The dose outside the treatment field is critically dependent on de detai tails ls of th the e bea beamm-del delive ivery ry sy syst stem em for pa passsive si vely ly mo modi difie fied d be beam ams. s. Th The e do dose se ou outs tsid ide e th the e treatment is typically than an order of  magnitudefield less for scanned more beams.

Dosimetry

 Accurate absorbed-dose determination is a fund Accurate ament ame ntal al pr prer erequ equisi isite te for any ra radia diatio tion n the therap rapy y treatment, as tumor control and normal tissue complic pl icat atio ion n pr prob obab abil ilit ities ies ar are e st stee eep p fu func ncti tion onss of  absorbed dose. Relative accuracy and reproducibility it y of 3 pe perc rcen entt an and d 2 pe perc rcen entt re resp spec ecti tive vely ly,, ar are e desi de sira rabl ble. e. Be Beam am mo moni nito tors rs ar are e re requ quir ired ed in th the e beam-d bea m-deli elive very ry sy syst stem em to con contr trol ol the dos dose e in th the e pati pa tien ent. t. Me Meas asur urem emen entt of do dose se di dist stri ribu buti tion onss is necessary for treatment planning purposes. †

  Since calor calorimete imeters rs are absol absolute ute dosim dosimeters eters they are ar e th the e ins instru trumen ments ts of cho choice ice for det determ ermini ining ng refe re fere renc nce e ab abso sorb rbed ed do dose se in pr prot oton on th ther erap apy y beams. However, they are not suited for routine use, and ionization chambers remain the practical ins instrum truments ents for pro proton ton dosi dosimetry metry. Ther There e are curren cur rently tly no pri primar mary y st stand andard ardss for abs absorb orbed ed

†

†

  Lin Linear ear ac accel celera erator tors, s, cyc cyclot lotron rons, s, sy synch nchroc rocycl yclootrons, and synchrotrons can be used for accelerating protons to the required therapeutic energies. To dat date e linea linearr acce accelera lerators tors have not been used. Both Bot h ro roomom-tem temper perat atur ure e and sup superc ercond onduct ucting ing cyclotrons are in use. Synchrotrons are the most flexib fle xible le ma machi chines nes in ter terms ms of ene energy rgy var varia iatio tion, n, which wh ich can be ac accom compli plishe shed d pul pulse se by pul pulse, se, but beam intensities are limited  Protons lose their energy in a medium primarily thro th roug ugh h in inte tera ract ctio ions ns wi with th at atom omic ic el elec ectr tron ons. s. Beca Be caus use e of th thei eirr la larg rge e ma mass ss rel elat ativ ive e to th the e ele el ect ctrron mas asss, prot oton onss lo lose se on only ly a sma mall ll fraction fra ction of energ energy y in each interaction interaction.. Alth Although ough nucle nu clear ar in inte tera ract ctio ions ns in incr crea ease se wi with th en ener ergy gy,,

†

doses in pro proton ton beams and ionization ionization chamb chambers ers have ha ve to ha have ve cal calibr ibrat ation ion coe coeffic fficien ients ts tr trac aceab eable le   60 to a primary standard in a Co bea eam m.  Alternatively  Alternativ ely,, they can be calibrated with a calorimeter in the user’s proton beam.   Two Two pr proto oton n dos dosime imetry try pr proto otocol colss are in cur curren rentt   viz use, ., IC ICRU RU Rep epor ortt 59 (I (ICR CRU, U, 19 1998 98)) an and d IAEA TRS 398 (IAEA, 2000). It is recommended in the present report that the TRS 398 code of  prac pr acti tice ce be ad adop opte ted d as th the e st stan anda dard rd pr prot oton on dosi do sime metr try y pr prot otoc ocol ol as it is ve very ry si simp mple le to us use, e, harmonizes with other modalities’ codes of practice (also given in TRS 398), which are being uni versally adopted, and has a more robust and rigorous formalism than that of ICRU Report 59.

7  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  †

†

†

  Th Ther ere e is li litt ttle le di diff ffer eren ence ce be betw twee een n re refe fere renc nce e dosimetry dosi metry for pass passively ively modified modified beam beamss and for scanned beams.   Parall Parallel-pl el-plate ate ioniz ionizati ation on cham chambers bers are the most common com mon and we wellll-pr prov oven en det detect ectors ors for pr proto oton n beam bea m mon monito itorin ring. g. Non Non-sa -satu tura ratin ting g sec second ondary ary emissi emi ssion on mon monito itors rs (SE (SEM) M) and mu multi ltiwir wire e ion ionizization chambers (MWIC) are also used. MWICs or pixel ionization chamber arrays can be used for monitoring the beam position of scanning beams.   Detectors for relative dose-distribution measurement me ntss in incl clud ude e io ioni niza zati tion on ch cham ambe bers rs,, si sili lico con n diodes dio des,, ra radio diogra graphi phicc film films, s, ala alanin nine, e, dia diamon mond d detectors dete ctors,, rad radiochr iochromic omic film, and gel dete detectors ctors.. Ionization chambers are often the instruments of  choice as they are the simplest to use and do not re requ quir ire e sp spec ecifi ificc Radi cali ca libr brat atio ion n fo for r su such ch rela re lati tive ve measurem meas urements. ents. Radiogra ographic phic films, scintillat scint illator or screens, scre ens, or twotwo-dime dimension nsional al pixe pixell detec detectors tors can be us used ed to me meas asur ure e th the e do dose se di dist stri ribu buti tion onss of  scanned beams.

Geometry and dose-volume definitions

Previous reports (ICRU, 1993b; 1999; 2004) hav Previous have e introduced and defined a variety of terms and acronyms to standardize the identification of a number of vol volum umes es of cli clinic nical al int intere erest st.. The These se inc includ lude e the gros gr osss tu tumo morr vo volu lume me (G (GTV TV), ), th the e cl clin inic ical al ta targ rget et  volume (CTV), the planning target volume (PTV), the organ at risk (OAR), and the planning organ at risk volu volume me (PR (PRV), V),   etc. With the exception of the PTV PT V an and d PR PRV V, the these se ter terms ms rem remain ain una unalte ltered red in the cont context ext of prot proton on beam ther therapy apy.. Inde Indeed ed thei theirr definit defi nition ionss and del deline ineat ation ionss sho should uld be mod modali ality ty

†

†

†

planning. If need planning. needed, ed, addi additiona tionall volu volumes mes shou should ld be used for this purpose.   So Some me no nome menc ncla latu ture ress to st stan anda dard rdiz ize e th the e ge geoomet etri ricc ter erm ms wh when en mult ltip iple le vol olu umes are required for complex situations are suggested in the present report.   The The terms terms:: volu volume me of inte interes restt (VO (VOI), I), surface of  inte in tere rest st (S (SOI OI), ), an and d po poin intt of in inte tere rest st (P (POI OI)) as gene ge neri ricc te term rmss ar are e in intr trod oduc uced ed in th the e pr pres esen entt report.   A nu numb mber er of do dose se an and d do dose se-v -vol olum ume e te term rmss ar are e defined in the present report. These include:

 DV 

 

V  D  D

 

 Dnear-min  Dnear-max

†

the leas leastt dose rece received ived by a volu volume, me,   V , of a specified VOI the lar larges gestt vol volume ume of a spe specifi cified ed VO VOII which re which recei ceive vess mor more e tha than n or equ equal al to the dose,  D   the least least dose dose receive received d by at lea least st 98 percent of a VOI. It equals  D 98%   th the e grea eattes estt dos ose e whic ich h al alll bu butt 2 percent of a VOI receives. It equals  D 2%.

  The term ‘prescribed dose’, or equivalently ‘prescript scr iption ion dos dose’, e’, is defi defined ned dif differ ferent ently ly fro from m pr pree vious ICRU reports. It was previously defined as a goal dose at a specified point ( e.g., the ICRU reference point). It is now proposed that the prescribed scr ibed dose be re refer ferre red d to a spe specifi cified ed vo volum lume. e. For example, one could prescribe that 95 percent of th the e PT PTV V sh shou ould ld re rece ceiv ive e wi with thin in   2 5 an and d þ7 percent of the prescribed dose.

Treatment planning 

independent. †

†

The present report focuses on the main differences between the planning of proton-beam therapy and photon-beam therapy, namely:

 The PTV and PRV serve two functions: they are planning tools, and they are used to report doses. Regardin Rega rding g the first first,, pro protons tons diff differ er fro from m phot photons ons in th that at th the e de desi sign gn of a pr prot oton on be beam am re requ quir ires es different margins lateral to, and distal and proximal to the CTV. As a res result ult,, a dif differ ferent ent PTV in principle, would have to be used for each possible beam bea m dir direct ection ion.. Thi Thiss is cum cumber bersom some, e, and som some e centers design their proton beams without reference en ce to a PT PTV V, bu butt pu pure rely ly ba base sed d on th the e CT CTV  V  using appropriate lateral and distal and proximal margins. Nevertheless, it is recommended in the present report that PTVs be used in proton planning for dose reporting purposes, allowing for all but dis distal tal and proxima proximall mar margin ginss du due e to ra range nge uncertainties.  It is recommended that neither the PTV nor the PRVs PR Vs be co comp mpro romi mise sed d in or orde derr to gu guid ide e do dose se

†

†

 The depth-dose properties of protons and photons are quite different. A monoenergetic proton beam loses energy slowly until near the end of range where the rate of energy loss increases sharply, resulting in the formation of the narrow (several mill mi llim imet eter ers) s) Br Brag agg g pe peak ak,, be beyo yond nd wh whic ich h no appreciab appr eciable le energ energy y is depo deposite sited. d. Con Consequ sequently ently, proto pr otons ns del deliv iver er les lesss int integr egral al dos dose e out outsid side e the targ ta rget et vo volu lume me,, by a fa fact ctor or of 2 to 3, th than an do photons.   Protons scatter within the patient. This has the conseq con sequen uence ce tha thatt th the e wid width th of a pr proto oton-b n-beam eam penum pen umbra bra dep depend endss on dep depth th and th this is mu must st be take ta ken n in into to ac acco coun untt in pl plan anni ning ng pr prot oton on-b -bea eam m therapy ther apy.. Sca Scatteri ttering ng also aff affects ects the beha behavior vior of  proto pr otons ns in re regio gions ns of sha sharpl rply y cha changi nging ng den densit sity y

8  

EXECUTIVE SUMMAR SUMMARY  Y 

†

†

†

(along ong the dir direct ection ion of the pr proto oton n bea beam) m) whe when n (al anatomic displacement occurs, so that additional care is needed when protons are directed nearly tangent to an inhomogeneity such as the external skin surface.   The penetration of a proton beam is very sensitive to inho inhomogen mogeneitie eitiess of tissu tissue e dens density ity with within in the th e pa patie tient nt.. The These se can affect affect the dep depth th of pen pen-etration and the sharpness of the terminal Bragg peak pe ak.. In Inho homo moge gene neit itie iess mu must st th ther eref efor ore e be meas me asur ured ed an and d al allo lowe wed d fo forr. Th Thei eirr in influ fluen ence ce is much greater than it is on photon beam distributions.  The measurement of internal densities, and their conversion to ‘water-equivalent’ densities, is thus pa part rtic icul ular arly ly im impo port ant t entional forr onal fo prot pr oton ons. s.voltage This Th is CT is usually usua lly done using usin grtan conventi conv kilovolt kilo age (co com mputed tom omog ogrraphy hy)) tog oget eth her wi witth a meas me asur ured ed lo look ok-u -up p ta tabl ble e re rela lati ting ng Ho Houn unsfi sfiel eld d numbers numb ers to wa waterter-equiv equivalent alent dens density ity.. How However ever,, CT using megavoltage photons has the potential to min inim imiz ize e th the e ef efffec ectts of ar arti tiffact ctss from embedded metallic objects.   Intensity-modulated radiation therapy (IMRT) is feasible feasi ble with prot protons ons (where it is termed inte intennsity-modu sitymodulate lated d pro proton ton ther therapy apy or IMPT IMPT)) jus justt as it is wi with th x ra rays ys (i (int nten ensi sity ty-m -mod odul ulat ated ed xx-ra ray y thera th erapy py or IMX IMXT) T).. Th The e mai main n dif differ ferenc ence e is tha thatt protons have an additional dimension that can be manipula manipu lated ted (de (depth pth of pen penetr etrat ation ion), ), wh which ich in principle prin ciple incr increases eases the flexi flexibilit bility y and impr improve ovess the likelihood of achieving the desired result.

†

Small reg Small region ionss wit within hin the tar target get vo volum lume e mig might ht either eit her esc escape ape bei being ng irr irradi adiat ated ed by mo movin ving g aw away ay from the pencil beam as it passes through them, or be ov overd erdose osed d by mo movin ving g in syn synchr chrony ony wit with, h, and within, within, a mo movin ving g pen pencil cil bea beam. m. Th This is giv gives es rise to dose fluctuations within the target volume - so-called ‘dose mottle’.   Inte Interpla rplay y effec effects ts cann cannot ot be enti entirely rely elim eliminat inated ed (the (t hey y oc occu curr in IM IMXT XT as wel ell, l, bu butt to a le less sser er exten ex tent), t), but th they ey can be red reduce uced d to a tol tolera erable ble level. lev el. Th The e mos mostt st stra raigh ightf tforw orward ard app appro roac ach h is to unde un dert rtak ake e wh what at is ca call lled ed ‘r ‘rep epai aint ntin ing’. g’. Wit ith h repaintin repa inting, g, one deliv delivers ers the entir entire e dose pat pattern tern multiple times - with a fraction of the dose delivered at each repainting. The individual repainticomp ngs arable need to the be period at timofesmoti greon ate(typ r tically han , oar comparab le to motion (typically few seconds for respiration motion).

Uncertainty

Uncertainty Uncertain ty is inevi inevitable table in plan planning ning and deliv deliverering radiation therapy. In this, protons are no different fr from om oth other er ra radia diatio tions. ns. Ho Howe weve verr, pr proto otons ns do have ha ve ad addi diti tion onal al so sour urce cess of un unce cert rtai aint nty y. Th Thes ese e include the effects of motion, additional dosimetric uncertain unce rtainties ties with protons, protons, and issue issuess of rela relative tive biolog bio logica icall eff effect ective ivenes ness. s. Th The e pr prese esent nt rep report ort re reccommends omme nds the follo following wing for all proc procedur edures es and for each patient: †

  The The sourc sources es of unc uncert ertain ainty ty sho should uld be ana analyz lyzed ed and minimized to the extent possible or

†

Motion management

 As with photons, motion of the patient and his or her internal anatomy must be accounted for in the planning and delivery of therapy. However, because of the sharp lateral and distal dose fall-offs, there are some differences in proton-beam therapy: †

†

†

†

 The motion of regions of inhomogeneous density potent pot ential ially ly up upset setss th the e dos dose e dis distri tribut bution ion,, if it is planned without regard to motion. This has two conseq con sequen uences ces:: 1) th the e eff effect ect of mot motion ion sh shoul ould d be red educ uced ed to th the e ext xten entt po poss ssib ible le th thrrou ough gh,, fo forr example, exam ple, immo immobiliza bilization tion of the pat patient ient and respirat pir ation ion gat gating ing;; 2) the re resid sidual ual unc uncert ertain aintie tiess must be taken into account, for example, through the use of appropriate margins, both lateral and distal to the PTV.   Mot Motio ion n ca caus uses es a sp spec ecifi ificc pr prob oblem lem in sc scan anne neddbeam therapy due to what are termed ‘interplay effects’. These occur when a scanned pencil beam is moved throughout the targe gett volume (“painted”) while the internal anatomy is moving.

†

practicable.   The mag magnit nitud ude e of the rem remain aining ing una unavo voida idable ble uncertainties should be documented.   Trea reatm tment ent pla plann nning ing pr progr ogram amss sho should uld pr prov ovide ide tools to aid in this analysis. In this regard, some appr ap proa oach ches es to th the e di disp spla lay y of un unce cert rtai aint nty y ar are e presented.  For normal reporting purposes, in uncomplicated cases, cas es, it wou would ld be suf suffici ficient ent to re repor portt a gen generi ericc levell of unce leve uncertain rtainty ty,, tailo tailored red when necessary to the ind indivi ividu dual al cir circum cumst stanc ances es of th the e par partic ticula ularr patient.  For cases where exceptionally large uncertainties might exist, the uncertainties in the dose distributions, as well as in the summarizing statistics, should shoul d be est estimat imated ed (tog (together ether with thei theirr corr correesponding confidence levels).

Quality assurance

 A rigorous QA (quality assurance) program is required to ensure reproducible, accurate, and safe fulfi fu lfill llme ment nt of th the e tr trea eatm tmen entt pr pres escr crip ipti tion on.. QA  chec ch ecks ks ar are e of ofte ten n te tech chno nolo logy gy-- an and d eq equi uipm pmen entt9

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

specific and focus principally on various aspects of  dose deliv delivery ery,, pat patient ient posit positionin ioning g and trea treatmen tmentt planning as well as radiation protection. †

†

†

  The positioning and alignment of hardware components in the treatment nozzle are checked by the th e me meas asur urem emen entt of de dept pth h do dose sess an and d la late tera rall beam profiles, the parameters of which should be within wit hin pr prede edeter termin mined ed tol tolera erance nce lim limits its.. Bea Beam m monitors monit ors and the sett settings ings of beam beam-mod -modifica ification tion devices also require routine checking. Additional specifi spe cificc tes tests ts are re requi quired red for sca scann nning ing bea beammdelivery syst systems. ems.   Regular checks of the patient-positioning system with respect to the treatment isocenter and beam axis sho axis should uld be un under dertak taken en to ens ensur ure e th that at the patient can be positioned accurately in the beam.   Co Comp mpre rehe hens nsiv ive e ac acce cept ptan ance ce te test stin ing g an and d QA  program prog ramss are requ required ired for tre treatm atmentent-plan planning ning systems sys tems.. These systems systems shou should ld be valid validate ated d by comparing calculated with measured doses over a range of clinical situations. Regular calibration of  the th e pl plan anni ning ng CT sc scan anne nerr is al also so re requ quir ired ed to ensur ens ure e th the e int integr egrity ity of th the e dos dose e cal calcul culat ation ions. s. Proton Prot on tre treatm atment-p ent-plann lanning ing QA proc procedur edures es are similarr to thos simila those e for conventiona conventionall ther therapies apies and are fully documented in various publications.

proced proc edur ures es af afte terr co comp mple leti tion on of tr trea eatm tmen ent. t. Th The e procedures include the following: †

†

†

†

†

†

†

  Completing the initial medical note in which all relevant clinical information is recorded.   Specifyi Specifying ng the plan planning ning aims,   i.e., det detaili ailing ng all the th e in info form rmat atio ion n ne need eded ed to pl plan an th the e in inte tend nded ed treatment.   Creating an acceptable treatment plan; this may be the result of an iterative process if the initial planning aims cannot be met.   Providin Providing g a tre treatm atment ent pres prescript cription ion acco according rding to the specifications in the treatment plan; this specifies the dose at specified points or within deline neat ated ed vo volu lume mes, s, an and d ho how w th the e pa pati tien entt is to be treated.   Sp Spec ecif ifyin ying g th the e tr trea eatm tmen entt da data ta ne nece cess ssar ary y to execute the treatment plan.  Compiling the treatment record in which all data relevant to the patient’s treatment are recorded.   Writin Writing g rep reports orts of the trea treatmen tment, t, inclu including ding the comp co mple leti tion on no note te,, re repo port( rt(s) s) to re refe ferr rrin ing g ph phys ys-ician( ici an(s) s) and oth other er inv involv olved ed med medica icall per person sonnel nel,, and publication of the results of the treatment.

Six cl Six clin inic ical al ex exam ampl ples es of re repo port rtin ing g pr prot oton on be beam am therapy from several proton treatment facilities are prese pr esent nted ed to ill illus ustra trate te ho how w to int interp erpre rett the con con--

cepts and how to apply the recommendations developed ope d in thi thiss re repor port. t. Th The e cli clinic nical al ex examp amples les ar are e for treatments of uveal melanoma, adenocarcinoma of  the th e pr pros osta tate te,, ca carc rcin inom oma a of th the e lu lung ng,, aco cous usti ticc neurom neu roma a (r (radi adiosu osurgi rgical cal cas case), e), med medull ullobla oblast stoma oma ( pedi pediatr atric ic case) case),, and skul skull-bas l-base e chord chordoma. oma. These clin cl inica icall ex exam ampl ples es sh shou ould ld no nott be in inte terp rpre rete ted d as ICRU ICR U re recom commen menda datio tions ns for sel select ecting ing par partic ticula ularr treatment strategies.

Prescribing, recording, and reporting  treatment

The proc procedur edures es for pre prescribi scribing ng and docu document menting ing patie pa tient nt tr trea eatme tment ntss ar are e an imp import ortant ant pa part rt of the trea tr eatme tment nt pr proce ocess, ss, whi which ch beg begins ins wit with h an ini initia tiall exami ex amina natio tion n and ass assess essmen mentt of the pa patie tient nt and determ det ermina inatio tion n of th the e the thera rapeu peutic tic goa goals ls (r (rad adica icall or pa pall llia iati tive ve), ), an and d co cont ntin inue uess to th the e fo foll llow ow-u -up p

10  

Journal of the ICRU Vol 7 No 2 (2007) Report 78 Oxford University Press

1

doi:10.1093/jicru/ndm024

INTR IN TROD ODUC UCTI TION ON

Many re Many revie views ws of pr proto oton n the therap rapy y ha have ve bee been n pu pubblished over the years. These studies have discussed the th e ph phys ysic ical al,, bi biol olog ogic ical al,, te tech chni nica cal, l, an and d cl clin inic ical al

relatio rela tionsh nship ip bet betwe ween en the inc incide ident nt pr proto oton-b n-beam eam energy ene rgy and its max maximu imum m pen penetr etrat ation ion in tis tissue sue is shown in Fig. 1.3. To be able to irradiate all possible

aspects, as as rning considerations of its and speculati specu lations onswell concernin conce g the futu future re roleefficacy of pro protons tons et al., 19 in ra radia diatio tion n the thera rapy py (Ar (Archa chambe mbeau au   et 1974 74;; Bonnett, 1993; Breuer and Smit, 2000; Chu, 1995a; al.., 198 Fowler Fo wler,, 1981; Goite Goitein, in, 1995; Goite Goitein in   et al 1985; 5; 2002; Graffmann, 1975; Jones, 1999; 2001a; 2001b; 2001c; Koehler and Preston, 1972; Kogelnik, 1997; al.., 19 Millerr, 1995; Mun Mille Munzenr zenrider ider   et al 1981 81;; Na Nahu hum m  et al., 199 1994; 4; Pe Pett ttii and Lennox, Lennox, 199 1994; 4; Ra Raju, ju, 1980; 1994; 1995a; 1995b; 1996; Smith, 2006; Suit, 2002; Suit and Krengli, 1997; Suit and Urie, 1992; Suit

target adult patients, proton ranges of  26–38 volumes cm in in tiss ti ssue ue (c (cor orre resp spon ondi ding ng to prot pr oton on energies of 200–250 MeV) are required. This allows for ene energy rgy los losses ses in bea beam-m m-modi odifyi fying ng dev device ices, s, in diagnost diag nostic ic and dosimetry equipment equipment in the beam line, lin e, and in th the e air gap bet betwe ween en th the e ac accel celer erat ator or  vacuum syste system m and the patient surface. Proton accelerators producing energy beams with energies of 60–75 MeV (with proton ranges of    3.0–4.5 cm in water) are used principally for the treatment of  ocular ocu lar les lesion ionss (Go (Goite itein in and Mil Miller ler,, 198 1983; 3; Goi Goitei tein n

 et al., 1992; 2003; Verhey and Munzenrider, 1982; Wamb ambersi ersie e and Ba Batt tterm ermann ann,, 199 1995; 5; Wamb ambers ersie ie  et al., 1999; 2002; 2004b).

 et al., 1983a).  Accelerated protons are near-monoenerg near-monoenergetic etic and form fo rm a be beam am of sm smal alll (r (rel elat ativ ive e to ty typi pica call ta targ rget et

In the present section first the principles of, and rationale for, proton therapy are summarized. Then follows a brief history of the development and use of proton therapy, after which a description of the present status of this treatment is given. The scope and goals of the present report and its relationship to the existing reports are then discussed.

The ra The rati tion onal ale e fo forr th the e us use e of pr prot oton on be beam amss in radiation therapy is remarkably simple. It is based on th the e ph phys ysica icall cha chara racte cteris ristic ticss of ene energy rgy los losss by protons prot ons when the they y pene penetra trate te mat matter ter,, name namely ly,, (i) prot pr oton onss ha hav ve a fin finit ite e de dept pth h of pe pene netr tra ati tion on in material, the magnitude of which depends on their ene en erg rgy y an and d on th the e den ensi sitty of the irr rra adi dia ated material; (ii) protons exhibit a relatively low ionization density (energy loss per unit path length) at the surface that slowly increases to near the end of  the beam range at which point there is a narrow region of high ionization density, termed the Bragg peak, pea k, wit with h neg neglig ligibl ible e dos dose e dep deposi osited ted be beyo yond nd the peak; and (iii) the dose from a proton beam falls off  sharply both laterally and distally. These character-

 volumes) lateral dimension and angular divergence; gen ce; they nee need d to be mod modifie ified d for pr prac actic tical al use use.. Ther Th ere e ar are e two approa approache chess to for form m a des desire ired d dos dose e dist di stri ribu buti tion on fo forr a pr prot oton on-t -the hera rapy py be beam am,,   viz., passive scattering and modulation (referring to the method meth od of spre spreadin ading g the beam late laterall rally y and with the desired dose distribution in depth), or dynamic scan sc anni ning ng of a pe penc ncil il be beam am bo both th la late tera rall lly y an and d in depth. In pas passiv sive e sca scatte tterin ring, g, th the e la later teral al dos dose e dis distri tri-bution is formed by placing the scattering material in the beam together with an occluding aperture to shape the beam laterally so as to provide a nearuniform dose across the field (Koehler  et al., 1977). The distribution of dose in depth is formed either by la laye yeri ring ng su suit itab ably ly we weigh ighte ted d pr prot oton on be beam amss of  diff di ffer eren entt en ener ergi gies es (r (ran ange ges) s),, us usua uall lly y us usin ing g a  variable-thic  variab le-thickness kness prop propeller eller tha thatt rota rotates tes in the beam (Chu  et al., 1993; Koehler  et al., 1975; Wilson, 1946), or by a ridge filter made up of multiple ridges of vari variabl able-th e-thick icknes nesss abso absorber rber pla placed ced in the bea beam m (Kostjuchenko et al., 2001; Larsson, 1961). The Th e se seco cond nd ap appr proa oach ch to fo form rm th the e do dose se di dist stri ri-buti bu tion on in invo volv lves es sc scan anni ning ng of a pe penc ncil il be beam am bo both th laterally and in depth (by changing its energy), in which case a near-arbitrary distribution of dose is possible poss ible lat lateral erally ly,, and consi considera derable ble dose shap shaping ing

isti is tics cs ar are e il illu lust stra rate ted d in Fi Figs gs 1. 1.1 1 an and d 1. 1.2. 2. Th The e

al.., 19 can ca n be ac achi hiev eved ed in de dept pth h (B (Bac ache herr   et al 1989 89;;

1.1 PRIN PRINCIPL CIPLES ES OF PRO PROTON TON THER THERAPY  APY  1.1.1 Phys Physical ical chara characteris cteristics tics

# International Commission on Radiation Units and Measurements 2007

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

Figure 1.1. Depth– dose curves for a 200 MeV MeV proton beam: both unmodul unm odulate ated d and with a 5 cm spread-out spread-out Bragg peak (SOBP), compared with a 16 MV x-ray beam (for 10    10 cm2 fields). The curves curv es are normaliz normalized ed in eac each h case to 100 at maxi maximum mum dose. (Adapted from Jones, 1999; reproduced with permission).

Pedroni  et al., 1989; 1995; 2005). The lateral distri-

Figure Figur e 1.3 1.3.. Proton Proton ra range nge in wa water ter as a fun functi ction on of en energ ergy y. Compiled from data given in ICRU (1993a).

distribut distri bution ion of ene energi rgies es is des design igned ed to pr prov ovide ide a uniform dose over some depth, the resultant distribution is designated as the spread-out Bragg peak

(SOBP) as shown in Figs 1.1 and 1.4. The SOBP is the segment of the depth–dose curve characterized by an ess essent ential ially ly fla flatt dos dose e re regio gion. n. Thu Thus, s, for each beam be am,, th the e SO SOBP BP is de desi sign gned ed to co cove verr th the e ti tiss ssue ue  volume of interest in depth with virtually no dose at gr grea eate terr de dept pths hs.. La Laye yeri ring ng of pr prot oton on be beam amss of  graded gra ded ene energi rgies es has the eff effect ect of inc incre reasi asing ng the entra ent rance nce dos dose e re relat lativ ive e to a mon monoen oenerg ergeti eticc bea beam. m. This effect increases with the length of the SOBP. Beam scanning can achieve a closer conformation of the high-dose region to the PTV than can passive beam-m bea m-modi odifica ficatio tion n sy syste stems ms (IC (ICRU, RU, 199 1993b; 3b; 1999 1999;; Urie and Goitein, 1989); specifically the dose proximal to the PTV is less in most cases. Beam scanning nin g is the onl only y pr prac actic tical al tec techn hniqu ique e tha thatt ena enable bless

bution for any proton energy is determined by the lateral positions and weights of each pencil beam of  a cho chosen sen ene energy rgy,, and the dis distri tribu butio tion n in dep depth th is determined by weighting the pencil beams at each position within the field. In ei eith ther er me meth thod od,, th the e be beam am ra rang nge e is ch chan ange ged d either by of inserting absorbers the beam path (in the th e ca case se cycl cy clot otro rons ns)) or by in chan ch angin ging g th the e be beam am ener en ergy gy (i (in n th the e ca case se of sy sync nchr hrot otro rons ns). ). Wh When en th the e

Figure Figur e 1.4 1.4.. De Depth–dose pth–dose cur curve ve for a mo monoe noener nerget getic ic 200 Me MeV  V  clinical proton beam (thick line) showing the Bragg peak at the end of the range. The superposition of suitably weighted proton beamss (thi beam (thin n line lines) s) of diff differe erent nt ener energies gies (ranges) (ranges) resu results lts in a spread-o spre ad-out ut Bra Bragg gg peak (SOBP) that pro provides vides a unif uniform orm depth dose do se ov over er th the e ta targ rget et re regi gion on (J (Jon ones es an and d Sc Schr hreu eude derr, 20 2001 01;; reproduced reprodu ced with permission).

Figure Figu re 1. 1.2. 2. Is Isod odos ose e di dist stri ribu buti tion on fo forr a 10 100 0 mm di diam amet eter er unmodulated 200 MeV proton beam. The isodoses in 10 percent stepss fro step from m 10 percent percent to 90 percent percent are shown shown (Jon (Jones, es, 1995; reproduced reprodu ced with permission permission). ).

12  

 INTRODUCTION 

inten intensitysity-modu modulate lated d prot proton on ther therapy apy (IMP (IMPT) T) to be performed. There are further potential advantages in that activation of beam line components and the second sec ondary ary sca scatte ttere red d dos dose, e, pr predo edomin minant antly ly due to neutr neu trons ons,, to th the e pa patie tient nt mig might ht be sub subst stant antial ially ly redu re duce ced d (H (Hal all, l, 20 2006 06), ), an and d fe fewe werr or no pa pati tien enttspec sp ecifi ificc tr trea eatm tmen entt de devi vice cess ar are e re requ quir ired ed.. Th Thes ese e factors are the principal bases for the current trend to employ beam scanning in the clinical situation. However, the effects of anatomical motion are more severe sev ere tha than n wit with h pa passi ssivel vely y mod modifie ified d bea beams, ms, and might result in the creation of regions of high dose outside, outsi de, and low dose insid inside e the target volume, and the th e re redu duct ctio ion n of sc scat atte tere red d do dose se is of ofte ten n no nott th the e deci de cidi ding ng fa fact ctor or in pr pres escr crib ibin ing g th the e tr trea eatm tmen ent. t. Beam-m Bea m-modi odifica ficatio tion n tec techn hniqu iques es ar are e des descri cribed bed in

 Yamada   et al., 2005), and tomotherapy techniques  Yamada (Welsh   et al., 20 2002 02)) is as gr grea eatt fo forr pr prot oton on as fo forr photo ph oton n tr trea eatme tments nts.. Th The e sig signifi nifican cantt ski skin-s n-spar paring ing effect with high-energy photon beams, due to dose build-up build -up resu resultin lting g from elect electroni ronicc dise disequili quilibrium brium near ne ar th the e su surf rfac ace, e, is mu much ch re redu duce ced d wi with th pr prot oton on beam be ams. s. Ho Howe weve verr, fo forr de deep ep le lesi sion onss tr trea eate ted d wi with th multi-fie mult i-field ld tech technique niques, s, skin dose is rar rarely ely a clini clini-cal problem. Forr a giv Fo given en dos dose e to the target target vol volum ume, e, pr proto otons ns deposit depo sit subs substant tantially ially less dose outside the targ target et  volume than do photons. In this situation, the tolerance of patients to proton treatments is increased over that experienced with photon treatments. The resu re sult lt is a de decr crea ease sed d fr freq eque uenc ncy y an and d se seve veri rity ty of  injury to uninvolved normal tissues for a specified

more detail in Section 3. Figure 1.5 emphasizes the dose sparing achieved proximally and distally to the target volume by a single singl e ran range-m ge-modula odulated ted prot proton on beam in comp compariarison so n wi with th a si sing ngle le hi high gh-e -ene nerg rgy y xx-ra ray y be beam am.. Th The e clin cl inic ical al adva van nta tage gess of the prot oton on bea eam m ar are e obvious: in strong contrast to photon beams, there is a ve very ry li litt ttle le do dose se di dist stal al to th the e SO SOBP BP (w (whi hich ch encloses the target volume) and, for all except the most mo st su supe perfi rfici cial al le lesi sion ons, s, th ther ere e is a lo lowe werr do dose se proxi pr oxima mall to the tar target get vo volum lume. e. Th The e exc excess ess dos dose e from fr om ph phot oton onss fo forr ea each ch be beam am pa path th is sh sho own in black for emphasis. These dose-distribution advantages of protons are obtained for all techniques of  dose delivery. Importantly, the flexibility in choosing the number of beams, beam orientation, beam weigh we ightin ting, g, bea beam m sca scann nning ing,, the use of int intens ensity ity-modulation techniques (Hong   et al., 2005; Pedroni  et al., 1995), four-dimensional image-guided treatet al., 20 ment me nt (t (tra rack ckin ing g or ga gate ted) d) (S (She heng ng   et 2005 05;;

dose to th dose the e ta targ rget et.. Al Alte tern rnat ativ ivel ely y, on one e ca can n ta take ke advantage of the improved dose distribution of the proto pr oton n irr irradi adiat ation ion to inc increa rease se the dos dose e del deliv ivere ered d to th the e ta targ rget et vo volu lume me an and d he henc nce e to in incr crea ease se th the e probability of local control of the tumor for a given complicat comp lication ion prob probabilit ability y. In acco accordan rdance ce with thes these e predictions, the efficacy of proton therapy has been demonstrated for several tumor types, as shown in Table 1.1. Further, for patients whose tumors are to be treated by combined radiation and chemotherapy, the les lesser ser ra radia diatio tion n dos dose e to che chemot mother herapy apy-sensitive normal tissues may permit an increment in dru drug g dos dose e and and,, hen hence, ce, aug augmen mentt re respo sponse nse fr freequency and/or duration. Tissu Ti ssue e inh inhomo omogen geneit eity y has a gr grea eater ter eff effect ect on proton dose distributions than on photon dose distribution tribu tions. s. Whil While e plan planning ning for prot proton on ther therapy apy,, the density of tissue along the proton path must be precisely cis ely det determ ermine ined d and ac accou counte nted d for in ord order er to obtain obta in the requ required ired proton energy dis distribu tribution tion to ach chie iev ve th the e pl plan anne ned d do dose se di disstr trib ibut utio ion n in th the e pati pa tien ent. t. Fai ailu lurre to al allo low w fo forr a zo zone ne of hi high gher er density could result in a near-zero dose in a distal segmentt of th segmen the e tar target get volume volume du due e to the reduced reduced rang ra nge e of th the e pr prot oton ons. s. In co cont ntra rast st,, fo forr ph phot oton ons, s, because of their different energy loss processes, an increased density would cause only a modest lowering of the dose distal to the higher density region (Fig. 1.6). Conversely, neglecting to account for an air cavity upstream of the target volume would, for proton pro ton beams, result in a high dose being depo deposited sited in distal normal structures, while only a modestly incre inc rease ased d dos dose e wou would ld be dep deposi osited ted in th the e cas case e of  photon beams. Proto Pr oton n bea beams ms ha have ve a sha sharp rp lat latera erall pen penum umbr bra, a, but that sharpness decreases with increasing beam energy ene rgy and and,, hen hence, ce, dep depth th of pen penetr etrat ation ion (Jo (Jones nes,, 1995; Larsson, 1961). The width of the penumbra (80– (8 0– 20 pe perc rcen entt is isod odos ose e le leve vels ls)) is na narr rrow ower er fo forr proton than for photon beams for penetrations up

Figure Figu re 1.5. Depth– dose curves for a modu modulat lated ed pro proton ton beam an and d a hi high gh-e -ene nerg rgy y black phot ph oton on beam be am no norm rmal aliz ized ed to th the e sa same me maximum dose. The areas highlight the regions where the photon photo n dose dra drastic stically ally exceeds exceeds the pro proton ton dose (ada (adapted pted from Fig. 1.1).

13  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  Table 1.1. Selected clinical results of proton and high-technology photon radiation therapy. therapy.

Primary tumor

Proton ( þ photon) dosea,  D RBE [Gy (RBE)]

Uveal melanoma

Photon doseb,  D   (G (Gy) y)

Numbe berr of  patients

Local control

Reference

70 in 5 fractions

2069

95 % at 15 y

60 in 4 fractions 60 – 70 in 4 fractions

2435 1406 158

92 % at 10 y 96 % at 5 y 98 % at 33 m

Gragoudas  et al. (2002b) Egger et al.  (2001) Dendale  et al.  (2006) Dieckman  et al.  (2006)

200

98 % at 10 y

Rosenberg  et al.

60 –7 –70 in 5 fractions Chondrosarcoma

72 in    38 fractions

(1999)

Chordoma

67 in    34 fractions

26

92 % at 36 m

Noel  et al.  (2004)

69 in    36 fractions 67 in 37 fractions

132 100 45

44 % at 10 y 54 % at 4 y 50 % at 5 y

Terahara  et al.  (1999) Noel  et al.  (2005) Debus  et al.  (2000)

67 in 37 fractions Pros Pr osta tate te (TI (TIII–TIV) II–TIV)

67.2 ver 67.2 versus sus 75. 75.6 6 (Phase III trial)

202

60 versus 77 %c at 8 y

Ship Sh iple ley y  et al.  (1995)

Prostate (TIa – TII)

74 in 37 fractions

1255

73 %c at 8 y

Slater  et al.  (2004)

Pros Pr osta tate te (T (TI–TI I–TII) I)

70.2 ve 70.2 vers rsus us 79 79.2 .2 (Phase III trial)

393

61.4 versus 80.4 % at 5 y

Zietman  et al.  (2005)

255

78 %c,d at 8 y

Zelefsky et al.  (2006)

29 (TI) 39 (TI TII) I)

87 % at 3 y 39 % at 3 y

Bush  et al.  (2004a)

40

85 % at 2 y

Hoyer et al.  (2006)

23

18 % survival at 3 y

Fitzek et al.  (1999)

23

93 % at 5 y

Pommier  et al.  (2006)

85

62 % at 5 y

Hoppe  et al.  (2007)

Prostate (TIc – TIIc) Lung (TI – TII)

81 in 45 fractions 51 – 60 in 10 fractions

Lung (TI)

45 in 3 fractions e

Glioblastoma multiforme

90 in 5 fractions

 Adenocystic carcin  Adenocystic carcinoma oma of paranasal sinus

76 in    38 fractions

Paranasal sinus

63 in    32 fractions

 Axial skeleton chondrosarcoma

72.2 in 40 fractions

6

100 % at 5 y

Hug  et al.  (1995)

Liver: primary

72 in 16 fractions 63 in 15 fractions

162 34

87 % at 5 y 75 % at 2 y

Chiba  et al.  (2005) Bush  et al.  (2004b)

26 in 1 fraction 37 in 3 fractions 25 in 5 fractions

28

82 % at 24 m

Wulf  et   al.  (2006)

45

82 % at 2 y

Mendez-Romero  et al.

30 in 10 fractions 60 in 3 fractionsf 

lesions 21

93 % at 18 m

(2006) Kavanaugh  et al. (2006)

Liver: primary and metastatic

a

Proton Pro ton ( þ phot photon) on) RBE-weigh RBE-weighted ted dose dose.. Pro Protons tons were delivered delivered alon alone e or in com combina bination tion with phot photons. ons. A pro proton ton RBE of 1.10 was employed. Mean or median doses are quoted as stated in the original papers. Neither the proton ( þ photon) or the photon series are notated for combined surgery or chemotherapy. chemotherapy. b Photons were delivered delivered by either IMRT or a stereotactic technique. technique. Mean or median doses are quoted as stated in the original papers. c Biochemical-relapse free. d Intermediate risk group. e Two fractions per day. f  Stereotactic body radiotherapy.

to int interm ermed ediat iate e dep depths ths,,    17–18 cm cm.. Fi Figu gure re 1. 1.7 7 show sh owss an ex exam ampl ple e of pe penu numb mbra ra wi widt dths hs ve vers rsus us

 18

cm. For both these beams, the penumbrae are narr na rrow ower er th than an fo forr th the e   60Co be beam am.. For a gi give ven n

depth of penetration for a 200 MeV proton beam, a 60 Co beam, and an 8 MV linear accelerator beam. For these conditions, the proton-beam penumbra is narr na rrow ower er th than an th the e 8 MV be beam am up to a de dept pth h of 

incident-beam energy, energy, the proton-beam penumbra is wi wide dest st in th the e Br Brag aggg-pe peak ak re regi gion on,, wh wher ere e th the e prot pr oton on en ener ergy gy is le leas astt (s (see ee Fi Fig. g. 1. 1.2) 2).. Be Beca caus use e prim pr imary ary pr prot oton onss ar are e sc scat atte tere red d ou outt of th the e be beam am,, 14

 

 INTRODUCTION 

Figure 1.7. Measured Figure Measured late lateral ral penu penumbr mbrae ae (80– 20 perc percent ent dose leve le vels ls)) as a fu func ncti tion on of de dept pth h fo forr a 20 200 0 Me MeV V pr prot oton on be beam am compared with other radiation beams. For these conditions the penumbr penu mbrae ae for 8 MV x ray rayss are narrower narrower than for protons at depths   . 18 cm. (Fi (Figur gure e cou courte rtesy sy of D.T.L. Jon Jones, es, iTh iThem emba ba Labora Labo ratory tory for Acce Acceler lerator ator-Base -Based d Scie Sciences nces,, Some Somerset rset West est,, South Africa.)

fra fr act ctio iona nati tion on sche sc heme mess empl em ploy oyed ed clin cl inic ical ally ly..  Although none of the available RBE values are based bas ed on hum human an tis tissue sue re respo sponse nse,, th the e rec recomm ommenendation in the present report is that a RBE of 1.10  i.e .e., a value that is be employed as a generic value,  i independent of dose, tissue type, fractionation,   etc. (Section 2).

Figure 1.6. Effe Figure Effect ct of a 3 cm thick slab of bone placed placed in: a   60Co beam (a); and in the spread-out Bragg peak (SOBP) of a proton beam bea m (b) (b).. Th The e   60Co be beam am is re redu duce ced d in in inte tens nsit ity y bu butt st stil illl penetrates deeply, whereas the penetration of the proton beam is reduced, but the magnitude of the dose in the high-dose region is unaffected (Goitein, 1982b; reproduced with permission).

1.1.3 Clin Clinical ical evalua evaluations tions

 As noted earlier earlier,, the available clinical outcome data da ta (se (see e Tabl able e 1.1 1.1)) sup suppor portt th the e con conten tentio tion n th that at proto pr oton n the thera rapy py is eff effect ective ive for the tr trea eatm tment ent of a number of types of lesions. This is evident in the local control results achieved for chondrosarcoma of  the skull base, uveal melanoma (the results given

proton beams have progressively lower central axis Bragg Br agg pea peaks ks as the bea beam m dia diamet meter er is dec decrea reased sed al.., 199 below bel ow abo about ut 10 mm (Ho (Hong ng   et al 1996; 6; Lar Larsso sson, n, 1967). 196 7). Th This is eff effect ect is ill illus ustr trat ated ed in Fig Fig.. 1.8 1.8.. Thi Thiss series ser ies of cur curves ves wa wass der derive ived d by cal calcul culat ation ion,, sup sup-ported port ed by a de dept pth– h– do dose se me meas asur urem emen entt fo forr th the e 2.4 mm radius beam. 1.1.2 1.1 .2

Biolog Bio logica icall eff effects ects

The biological effects of proton beams in the therapeutic energy range have been extensively studied in vario various us expe experime rimental ntal sett settings ings and revi review ewed ed by al..   (2002) Paganetti   et al (2002) and Skar Skarsgard sgard (1998) and are ar e dis discus cussed sed in det detail ail in Sec Sectio tion n 2. The These se ex exper per-iments have employed a wide variety of biological vitro o), res systems syste ms (both   in vivo   and   in vitr respon ponse se end points, point s, pro proton ton energ energies, ies, dose leve levels, ls, and fra fraction ction-ation protocols. The relative biological effectiveness vivo o   systems (RBE) val (RBE) values ues for the div divers erse e   in viv systems for proton energies between 60 and 250 MeV are consist si sten entt wi with th a me mean an RB RBE E of 1. 1.1 1 (s (see ee Se Sect ctio ion n 2) 2).. This is applicable for both acute- and late-res late -respond ponding ing tissu tissues es and for the range of dose/ 

Figure 1.8. Dept Figure Depth– h– dose curves curves for a 160 MeV MeV proton proton beam as a function func tion of coll collima imator tor rad radius, ius, calculated calculated usin using g a penc pencil il beam algorithm (Hong  et al., 1996; reproduced with permission).

15  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

were re ob obta tain ined ed fo forr le lesi sion onss of a wi wide de sp spec ectr trum um of  we sizes siz es and sit sites es wit within hin th the e glo globe) be),, hep hepat atoce ocellu llular lar carcin car cinoma oma,, st stage age T1 non non-sm -small all-ce -cell ll lun lung g can cancer cer,, parana par anasal sal sin sinus us car carcin cinoma oma,, and sar sarcom comas as of the

184 in 184 in.. syn ynch chro rocy cycl clot otro ron n at th the e Un Univ iver ersi sity ty of  al..   (1952) repo California, Berkeley Berkeley.. Tobias   et al reported rted LD50   values values for who wholele-bod body y irr irradia adiatio tion n of mic mice e by 340 34 0 Me MeV V pr prot oton onss an and d fo forr th the e ra ratt pi pitu tuit itar ary y by

pa a asa asa s us ca c o a, a, a d sa co as o t e axia ax iall sp spin ine. e. Th The e ev eval alua uati tion on of a ga gain in in lo loca call control cont rol prob probabilit ability y by charg charged-p ed-particl article e rad radiati iation on ther th erap apy y at lo long ng fo foll llow ow-u -up p ha hass be been en as asse sess ssed ed in Phas Ph ase e II IIII cl clin inica icall tr tria ials ls fo forr on only ly tw two o tu tumo morr ca cattegories,   viz., uveal melanomas [helium ions versus

3 0 eV p ot oto o s a d o t e at p tu ta y by al..   (1958; 190 MeV deu deuter terons ons.. Lar Larsso sson n   et al (1958; 195 1959), 9), et al.   (196 et al.   (1960) Leksell   et (1960) 0),, an and d Re Rexed xed   et describe desc ribed d rad radiat iation ion inj injury ury to the spi spinal nal cord and brai br ain n of exp xper erim imen enta tall an anim imal alss by hi high gh-e -ene nerg rgy y al..   (1975) proton proto n beams. Robe Robertson rtson   et al (1975) rep report orted ed a

125

I brachytherapy] (Char  et al., 1993) and prostate cancer (Shipley   et al., 1995; Zietman   et al., 2005). In the case of prostate cancer each of the two trials was designed to assess local control probability for two proton dose levels used as boost doses. Effect Eff ective ive and alt altern ernat ative ive form formss of eva evalua luatio tion n of  the probable efficacy of proton-beam therapy are the comparison of treatment plans in terms of dose distributions tribut ions and possib possibly ly includ including ing bioma biomathema thematicalticalmode mo dell pr pred edict ictio ions ns of tu tumo morr co cont ntro roll pr proba obabil bilit itie iess (TCP)) and normal tissue compli (TCP complicati cation on proba probabilitie bilitiess al.., 20 al.., (NTCP) (NT CP) (Baum (Baumert ert   et al 2001 01,, 20 2004 04;; Bo Bols lsii   et al al.., 2001 al.., 19 2003;; Dea 2003 Deasy sy   et al 2001;; Drz Drzyma ymala la   et al 1991 91;; Fogliata  et al., 2002; Goitein and Schultheiss, 1985; Hsiung-Stripp   et al., 2001; Lin   et al., 2000; Lomax  et al., 2003; Miralbell  et al., 2002; Mock  et al., 2004; Nahum and Glimelius, 2001; Nahum and Sanchez-Nieto, 2001; Niemierko and Goitein, 1993; al.., 20 al.., 20 Stavrev   et al 2001 01;; Su Suit it   et al 2003 03;; van Lu Luij ijk k  et al., 2003; Weber   et al., 2004; Zurlo   et al., 2000). Further references and a summary table are given in Glimel Glimelius ius et al.  (2005).

Robert Wilson first proposed the use of protons and

comprehens compreh ensiv ive e RBE stu study dy of mam mammali malian an cel cells ls   in vitro   using loss of colony-forming ability as the end point. Subsequently Subsequently,, ther there e has been exten extensive sive laborator ora tory y res resear earch ch to det determ ermine ine the biol biologic ogical al eff effecectivene tiv eness ss of pr proto otons ns rel relat ative ive to pho photon ton irr irradi adiati ation on (Paganetti et al., 2002: Skarsgard, 1998). This is considered further in Section 2. The first use of proton beams in the management of human patie ien nts was at the Uni niv ver ersi sitty of  California, Berkeley. This work was directed principally at the suppression of hormone production by the pit pituit uitary ary gla gland nd by ve very ry hig high-d h-dose ose and hig highly hly localized proton irradiation. The rationale was that a hi high gh pr prop opor orti tion on of hu huma man n br brea east st ca canc ncer erss ar are e hormon hor mone e dep depend endent ent.. Ext Extens ensive ive ani anima mall st studi udies es using usi ng pr proto oton n and deu deuter teron on bea beams ms pr prece eceded ded the clinical work. The first ‘patient’ was a dog with an extensiv exte nsive e ulcer ulcerate ated d brea breast st tumo tumorr. Tobias and coworkers irradiated the pituitary gland with deuterons (Lawrence, 1957; Tobias  et al., 1954; 1955) and found a substantial reduction in size of the lesion. On th the e ba basi siss of th this is go good od re resp spon onse se an and d pr prio iorr demo de mons nstr trat ated ed ma mark rked ed su supp ppre ress ssion ion of ho horm rmon onal al levels following high-dose irradiation of the rat pituitary itar y, the Ber Berkel keley ey team dec decide ided d to pro procee ceed d wit with h patient treatment. They commenced a Phase-I clinical tria triall of dos dose e esca escala lation tion of pro proton ton irradiat irradiation ion of 

heavierAtions radiation therapy in 1946 (Wilson, 1946). thatfor time, accelerators capable of generating pr proton oton beams with sufficien sufficientt ene energy rgy for pr proton oton radiation therapy were under construction. He noted that the large mass of the proton would cause it to travel in a nearly straight path through tissue and that the energy deposition pattern of a proton beam would produce high radiation doses near the end of  the range in a relatively narrow region, referred to as the Bragg peak. He pr propo oposed sed the irr irradia adiatio tion n of  localized regions within the body with proton beams, thereby providing significant sparing of surrounding tissues. tissue s. He also expla explained ined how rota rotating ting modulator modulator propel pro pellers lers cou could ld spr spread ead the Br Bragg agg pea peak k ov over er lar large ge

theiation pituitary women with carcinoma. rad radiat ion wasofadm admini iniste stered red breast in thr three ee fract fr action ionss The per week for    2 weeks; the total dose was 140–300 Gy. The Th e cr cross ossfir fire e te tech chni niqu que e wa wass em empl ploy oyed ed,, in wh which ich multiple small, unmodulated 340 MeV proton beams intersected at the pituitary and exited the skull,  i.e., the pituitary was irradiated by the plateau region of  each eac h bea beam. m. Th The e firs firstt pa patie tient nt was treated treated in 1954 1954,, only 8 years after the publication of Wilson’s paper (Wilso (W ilson, n, 194 1946). 6). Sev Severa erall of the 26 pa patie tients nts exp experierienced en ced go good od cli clini nica call re resp spon onse sess (L (Law awre renc nce, e, 19 1957 57;; al.., 19 Tobias   et al 1958 58). ). In 19 1957 57,, th the e ac accel celer erat ator or wa wass upgraded and the Berkeley team commenced a longterm te rm cl clin inica icall an and d lab labor orat ator ory y inv inves estig tigat atio ion n wi with th

targets, tran targets, transmissio smission n ioniz ionization ation chambers could be employ emp loyed ed to mon monito itorr pa patie tient nt dos dose, e, and ion ioniza ization tion chambe cha mbers rs cou could ld pr provi ovide de abs absolu olute te dos dose-c e-calib alibra ration tion coefficients for the beam monitors. The Th e fir firsst stu tudi dies es of th the e bi biol olog ogic ical al ef effe fect ctss of  charg ch arged ed-p -part artic icle le be beam amss we were re co cond nduc ucte ted d on th the e

910 MeV helium helium ions and the then n wit with h hea heavie vierr ion ions, s,  e.g., C and Ne (Castro   et al., 1980). The treatment facilities were closed in 1992. The Th e firs firstt pr proto oton n cli clinic nical al tr trea eatm tment entss in Eu Europ rope e were we re con condu ducte cted d in 195 1957 7 usi using ng the 185 Me MeV V sy synnchro ch rocy cycl clot otro ron n at th the e Un Univ iver ersi sity ty of Up Upps psal ala, a,

1.2 HIST HISTOR ORY Y OF PRO PROTON TON THER THERAPY  APY 

16  

 INTRODUCTION 

Sweden. Pilot studies of the efficacy of fractionated irradiat irra diation ion (between 1 and 10 fra fraction ctions) s) of malig malig-nant tumors were undertaken. A few patients with glioblastoma multiforme and tumors of the uterine cervix cer vix,, nas nasoph ophary arynx, nx, hea head d and nec neck, k, and oth other er site si tess we were re tr trea eate ted. d. Fo Forr se seve vera rall pa pati tien ents ts in th the e study, the tumor responses were judged to be good (Falkmer  et al., 1962). In the late 1950s, a program

et al., 19 required SOBP (Koeh required (Koehler ler   et 1975 75). ). La Late terr, in 1977, they described a double-scattering technique to provide uniform lateral proton-dose distributions (Koehler   et al., 1977) 1977).. Techn echniques iques for dis distal-e tal-edge dge al.., 19 compensa comp ensation tion (Goit (Goitein, ein, 1978a 1978a;; Urie   et al 1984 84;; Wagner, 1982) were also developed at the Harvard Cyclo lottron Labo borrato torry. These latter beammodific mod ificati ation on sy syst stems ems and tec techni hnique quess or vari varian ants ts

of st stere ereota otacti cticc rad radios iosurg urgery ery wit with h nar narro row w pr proto oton n beams for the treatment of Parkinson’s disease and intr in tra act ctab able le pa pain in was al also so st star arte ted d in Up Upps psal ala a (Graffman  et al., 1985; Larsson  et al., 1963). Following the change from proton to helium ion beams bea ms by the Ber Berke kele ley y pr progr ogram am in 195 1957, 7, pr proto oton n therapy recommenced in the USA in 1961 on the 160 16 0 MeV sync nch hrocy cycl clot otrron at th the e Ha Harv rva ard University Cyclotron Laboratory. Kjellberg and coworkers work ers unde undertook rtook singl single-do e-dose se ste stereot reotacti acticc tre treatatment me ntss of pi pitu tuit itar ary y ad aden enom omas as an and d in intr trac acra rani nial al   et al. al . tumors tumo rs (Kje (Kjellberg llberg,, 1979; Kjel Kjellberg lberg , 196 1962a; 2a; 1962b; 1968) and later proton treatment of arterio venous malformations (AVM) (Kjellberg, 1979; Kjellberg   et al., 1983). They used irradiation techniqu ni ques es in wh whic ich h th the e Br Brag agg g pe peak akss of mu mult ltip iple le narrow beams stopped in the lesion. Good success rates were realized for the ablation of pituitary adenomas and AVMs (Kjellberg, 1979). Proton therapy began in Russia (Dubna) in 1967, in Ja Japa pan n (C (Chi hiba ba)) in 19 1979 79,, an and d in th the e So Sout uthe hern rn Hemisphere (Somerset West, South Africa) in 1993 (Sisterso (Sis terson, n, 2005) 2005).. All the early pro protonton-trea treatmen tmentt facilities were established in the existing accelerator labor laborato atories. ries. iThemba LABS (Sou (South th Afri Africa) ca) is initio io the only acce accelera lerator tor labor laborato atory ry desi designed gned   ab init for particle therapy. The first hospital-based facility beca be came me op oper erat atio iona nall in 19 1990 90 at th the e Lo Loma ma Li Lind nda a Universit Univ ersity y Med Medical ical Cen Center ter,, CA CA,, and inclu included ded the first prot proton-b on-beam eam isoce isocentric ntric gant gantries ries (Co (Coutra utrakon kon  et al., 1994; Slater  et al., 1995). In 19 1973 73,, Su Suit it an and d co co-w -wor ork ker erss in inssti titu tute ted d a program prog ram at the Harva Harvard rd Cycl Cyclotron otron Laboratory Laboratory to evalua eva luate te th the e effi effica cacy cy of hig highly hly fr frac actio tiona nated ted rad radiiati tion on th ther erap apy y of ma mali lign gnan antt tu tumo mors rs.. Th The e fir firsst pati pa tien ents ts we were re tr trea eate ted d ma main inly ly fo forr sa sarc rcom omas as at several sites, with special emphasis on sarcomas of  the skull base (Suit and Goitein, 1974; Suit   et al., 1975; 197 5; 198 1982). 2). The These se ini initia tiall st studi udies es we were re ra rapid pidly ly expa pan nde ded d to tu tum mor orss at se sev ver era al ot oth her si sittes es.. Foll Fo llow owin ing g in init itia iall ex expe peri rime ment ntss on mo monk nkey ey ey eyes es

thereof are still widely used in broad-beam energymodula mod ulated ted pr proton oton beam the therap rapy y. The use of rid ridge ge filters to form SOBPs was first described by Larsson (1961), and these filters are used at several centers. Magnet Mag netic ic bea beam m sca scann nning ing wa wass als also o firs firstt use used d in Uppsala for broad-beam SOBP treatments with the 187 MeV proton beam (Larsson, 1961). Kanai  et al. (1980; 1983) used scanned beams with the 70 MeV  proto pr oton n bea beam m in Ch Chiba iba,, Jap Japan an.. Ext Extens ensive ive cli clinic nical al use of high-energy scanned proton beams has been undertaken at the Paul Scherrer Institute, Villigen, al.., 19 Switzerla Swit zerland nd (Pe (Pedron dronii   et al 1995 95;; 20 2005 05). ). Su Such ch al.., 200 beams bea ms are re requi quire red d for IM IMPT PT (C (Cell ella a   et al 2001; 1; Lomax, Lom ax, 199 1999; 9; Lom Lomax ax   et al., 20 2001 01;; Oe Oelf lfk ke an and d Bortfeld, 2001). Detailed Deta iled dosi dosimet metric ric meth methods ods for pro proton ton the therap rapy y al..   (1979), were wer e firs firstt form formula ulated ted by Verhe erhey y   et al (1979), and techniques techniqu es for accura accurate te patien patientt positionin positioning g (V (Verhey erhey  et al., 1982) were developed at the Harvard Cyclotron Laboratory Labora tory.. Goitein developed the first threethree-dimendimensional treatment-planning systems used in radiation therapy therap y, one specifically for eye treatments treatments (Goitein and an d Mi Mille llerr, 198 1983) 3) and an anoth other er for gen gener eral al us use. e. He also als o int intro rodu duced ced dos dose– e– vo volum lume e hi hist stogr ogram amss (DV (DVHs Hs), ), digitally reconstructed reconstructed radiographs (DRRs), and the display of uncertainty bands around isodose contours (Goitein (Goi tein,, 1978 1978a; a; 1980 1980;; 1982 1982a; a; Goit Goitein ein and Abr Abrams ams,, 1983; Goitein  et al., 1983b).

(Constable  et al., 1976), the first program involving the treatment of uveal melanomas was initiated in 197 19 76 at the Har arva varrd Cycl clot otrron Lab abo orato tory ry (Gragoudas  et al., 1977; 1980; 2002a; 2002b). In 1975, Koehler  et al.  first described the detailed design of a rotating propeller of variable thickness whic wh ich, h, wh when en pl plac aced ed in th the e be beam am,, ac achi hiev eved ed th the e

comparable for proton therapy and high-technology photo ph oton n the thera rapy py for th those ose cas cases es for whi which ch sim simila ilarr RBE-wei RBE -weighte ghted d absorb absorbed ed dose dosess wer were e admi administ nistered ered.. Proton therapy achieved no gain in the long-term loca lo call co cont ntro roll in th the e tr trea eatm tmen entt of pa pati tien ents ts wi with th glioblastoma multiforme. Of the 23 patients in the study st udy, onl only y one pa patie tient nt sur surviv vived ed 5 ye years ars (F (Fitz itzek ek

1.3

 A summary of the clinical outcomes of proton and ph photo oton n ra radia diatio tion n th thera erapy py for sel select ected ed tum tumor or type ty pess is pr pres esen ente ted d in Tab able le 1. 1.1. 1. Th Thes ese e re resu sult lts, s, togeth tog ether er wit with h the phy physic sical al ra ratio tiona nale le for pr proto oton n ther th erap apy y, ha have ve ge gene nera rall lly y be been en ac acce cept pted ed as ve very ry positive and constitute the bases for the substantial and increasing interest in the use of proton beams forr ra fo radi diat atio ion n tr trea eatm tmen entt of pa pati tien ents ts wi with th bo both th beni be nign gn an and d ma mali lign gnan antt le lesi sion ons. s. Lo Loca call co cont ntro roll is

17  

PRESENT PRESE NT ST STA ATUS

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  Table 1.2. Proton therapy facilities (December 2007).

Institute

Location

Crocker Nuclear Laboratory (operated by the University of 

Davis, CA, USA

C

60

3.1

Horizontal

1994

920

Clatterbridge, UK

C

62

3.3

Horizontal

1989

1 701

Nice, France

C

65

3.6

Horizontal

1991

3 129

Chiba, Japan

C

70

4.1

Vertical

1979 – 2002

145

Catania, Italy

C

70

4.1

Horizontal

2002

151 15

 Villigen,  Villigen, Switzerland  Vancouver  V ancouver,, Canada Berlin, Germany Louvain-la-Neuve, Belgium

C

72

4.3

Horizontal

1984

4 875

C

72

4.3

Horizontal

1995

130

C C

72 90

4.3 6.4

Horizontal Horizontal

1998 1991 – 1993

1 014 21

Harvard Cyclotr Cyclotron on Laboratory The Svedberg Laboratory The Svedberg Laboratory Institute for Theoretical Theore tical and Experimental Experiment al Physics iThemba Laboratory for Accelerator Based Sciences Midwest Proton Radiotherapy Institute

Cambridge, MA, USA  Uppsala, Sweden (1) Uppsala, Sweden (2) Moscow, Russia

SC

160

17.7

Horizontal

1961 – 2002

9 116

SC

185

22.8

Horizontal

1957 – 1976

73

SC

200

26.0

Horizontal

1989

738

S

200

26.0

Horizontal

1969

4 024

Somerset West, South Africa

C

200

26.0

Horizontal

1993

500

Bloomington, IN, Bloomington, USA (1)

C

200

26.0

Horizontal

1993 – 1999

34

Midwest Proton Radiotherapy Institute Centre de Protonthe´ rapie de l’Institut Curie Joint Institute for Nuclear Research Joint Institute for Nuclear Research Wakasa Wan Energy Research Center Paul Scherrer Institute Hyogo Ion Beam Medical Center

Bloomington, IN, Bloomington, USA (2)

C

200

26.0

Iso, Horiz

2004

3 79 37

Orsay, France (1)

SC

200

26.0

Horizontal

1991

3 766

Dubna, Russia (1)

SC#

200

26.0

Horizontal

1967 – 1996

124

Dubna, Russia (2)

SC#

200

26.0

Horizontal

1999

402

Tsuruga, Japan

S

200

26.0

Vert, Horiz

2002

49

 Villigen,  Villigen, Switzerland Switzerlan d (1) Nishi-Harima, Japan

C#*

230

32.9

Isocentric

1996

320

S

230

32.9

Is, Ve, Ho, 45º

2001

1 658

Jacksonville, FL, Jacksonville, USA 

C

230

32.9

Iso, Horiz

2006

15

Kashiwa, Japan

C

235

34.2

Iso, Horiz

1998

552

Boston, MA, USA

C

235

34.2

Iso, Horiz

2001

2 710

California, San Francisc Francisco) o) Clatterbridge Centre for Oncology Centre  Antoine-Lacassagne  Antoine-L acassagne National Institute of  Radiologicall Sciences Radiologica Centro di Adro Terapia e Applicazioni Nucleari Avanzate Paul Scherrer Institute Tri-University Meson Factory Hahn-Meitner-Institut Universite´   Catholique de Louvain

University of Florida Proton Therapy Institute Nationall Cancer Nationa Center Francis H Burr Proton Therapy Center

Accelerator

Max. clinical energy MeV 

Beam Bea m dir direct ectio ion n

First First treatment

Patients treated

Continued

18  

 INTRODUCTION  Table 1.2. Continued

Institute

Location

Accelerator

Shizuoka Cancer Center Wanjie Proton Therapy Center National Cancer Center Proton Medical Research Center M D Anderson Cancer Center Paul Scherrer Institute Loma Linda University Universi ty Medical Center Proton Medical Research Center Lawrence Berkeley National Laboratory

Mishima, Japan

S

235

34.2

Iso, Horiz

2003

570

Wanjie, China

C

235

34.2

Iso, Horiz

2004

537

Ilsan, South Korea

C

235

34.2

Iso (2), Horiz(1)

2007

Tsukuba, Japan Tsukuba, (1) Houston, TX, USA

S#

250

37.9

Vert, Horiz

S

250

37.9

sC*

250

S

 Villigen,  Villigen, Switzerland Switzerlan d (2) Loma Linda, CA, USA  Tsukuba, Japan Tsukuba, (2) Berkeley, CA, USA

Petersburg Nuclear Petersburg Physics Institute

St. Petersburg, Russia

Max. clinical energy MeV 

Beam Bea m dir direct ection ion

Firstt Firs treatment

Patients treated

1983 – 2000

700

Iso, Horiz

2006

527

37.9

Iso (2), Horiz(1)

2007

270

43.2

Iso, Horiz

1990

11 414

S

270

43.2

Isocentric

2001

1 188

SC

340

63.3

Horizontal

1954 – 1957

30

1 000

325.4

Horizontal

1975

1 327

1954–200 19 54–2007 7

53 43 439 9

S

Total number of patients

 Accelerators: C Cyclotr  Accelerators: Cyclotron; on; SC Synchrocyclotron; S Synchrotron Beam directions: Is(o) Isocentric; Ve(rt)  V  Vertical; ertical; Ho(riz) Horizon Horizontal; tal; 45º Inclined at 45º to horizontal Symbols:  # degraded beam; * scanned beam Facilities with energies less than 100 MeV (above the horizontal line) are used principally for the treatment of ocular lesions ¼

¼

¼

¼

¼

¼

¼

¼

¼

 et al., 19 1999 99), ), fo foll llow owin ing g tw twic ice e da dail ily y de deli live very ry of  RBE-w RB E-weig eighte hted d dos doses es of 1.8 Gy (R (RBE BE)) to a tot total al of  90 Gy (RBE) in 5 weeks.  A critical requiremen requirementt is the demonstr demonstration ation of an

currently 26 active clinical facilities. As noted above, all the early facilities were based on accelerators in physics phy sics labor laborator atories. ies. Recen Recently tly,, comme commercial rcial prot proton on accelera accel erators tors and ancilla ancillary ry equip equipment ment hav have e become

increment tumor-control probability for a similar or reduced in normal tissue-control probability. When considering local control, attention must be paid to the length of follow-up, and the assessment of the freq fr eque uenc ncy y an and d se seve veri rity ty of ra radi diat atio ion n in inju jury ry.. No attempt has been made in the latter regard in the present report. It is not feasible at present because of lac lack k of lon long-t g-term erm fol follo low-u w-up p da data, ta, esp especi eciall ally y in the case of high-technology photon treatments. Tabl able e 1.2 pr prov ovid ides es a fu full ll lis listt of pr prot oton on-t -the hera rapy py faci fa cilit litie iess at wh whic ich h pa pati tien ents ts ha have ve be been en tr trea eate ted, d, including the number of patients treated at the time of writ writing ing.. Th Thirty irty-th -three ree cen center terss ha have ve bee been n es estabtablished. lish ed. Fo Four ur hav have e bee been n clo closed sed (th (those ose at Ber Berkel keley ey,,

available, most aregan now hospitalbas ased ed an and dandfe fea atu turrenew 360 36 0facilities -rot -r otat atin ing g gantr tries ies. . Th The e grow gr owth th of pr prot oton on-t -the hera rapy py fa faci cilit litie iess is sh show own n in Fig.. 1.9 Fig 1.9.. The sites of the cur curre rently ntly active active cen center terss

CA, Cambridge, MA, Louvain-la-Neuve, and Chiba). Several Seve ral of the older centers have undergone major upgr up grad ades es of th thei eirr or orig igin inal al fa faci cili liti ties es (U (Upp ppsa sala la,, Dubn Du bna, a, Bl Bloo oomi ming ngto ton, n, IN IN,, an and d Tsu suku kuba ba). ). In th the e latt la tter er ca case se,, a co comp mple lete tely ly ne new w acc ccel eler era ato torr wa wass

8

installed. In addition, a new dedicated cyclotron for proton therapy has been commissioned at the Paul Sche Sc herr rrer er In Inssti titu tute te (P (PSI SI), ), Vil illi lige gen. n. Th Ther ere e ar are e

Figure 1.9. Number of operating operating proton therapy facilities given at 5 year intervals from 1950 to 2005. The number for 2007 is also shown. The curve is drawn to guide the eye.

19  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  Table 1.3 Dedicated proton therapy facilities under construction or funded (December 2007).

Institute

Location

Wesdeutsche Protontherapiezentrum Roberts Proton Therapy Center Oklahoma ProCure Treatment Center Centre de Protonthe´ rapie de l’Institut Curie Hampton Universi University ty Proton Beam Therapy Center Sino-Japanese Friendship Hospital Rinecker Proton Therapy Center

Essen, Germany Philadelphia, PA, USA  Oklahoma City, OK, USA  Orsay, France (2)

Max. clinical energy (MeV)

Range in water (g cm22)

Beam di dirrecti tio on

Schedu Sche dulled first treatment

C

230

32.9

Iso (3), Horiz (1)

2009?

C

230

32.9

Iso (4), Horiz (1)

2009?

Ca

230

32.9

2009?

C

230

32.9

Iso (1), Horiz (1), Dual fixed (2) Iso (1), Horiz (4)

Hampton, VA, USA 

C

230

32.9

Iso (4), Horiz (1)

2010?

Beijing, China

C

235

34.2

Iso (1), Horiz (1)

2007?

sCa

250

37.9

Iso (4), Horiz (1)

2007?

Munich, Germany

Accelerator

2010?

 Accelerators: C, cyclotron  Accelerators: cyclotron;; sC, supercon superconducting ducting cyclotron. Beam directio directions: ns: Iso, isocentri isocentric; c; Horiz, horizontal; Dual fixed, two fixed beam directions directio ns intersectin intersecting g at single isocenter isocenter.. a Scanned beam.

are: Europe, 8; USA, 6; Japan, 5; Russia, 3; South  Africa,, 1; Canad  Africa Canada, a, 1; Chin China, a, 1; and South Korea Korea,, 1. The accelerators used in these active centers are cyclot cyc lotro rons ns in 15 cen cente ters rs,, sy sync nchr hroc ocyc yclot lotro rons ns in 3 centers, cent ers, and syn synchr chrotr otrons ons in 8 cen center ters. s. Of the 15 cyclotrons, 7 produce beams with energies between 60 an and d 72 Me MeV V and ar are e us used ed pr princ incip ipal ally ly fo forr th the e trea tr eatm tmen entt of oc ocul ular ar le lesi sion ons. s. A fu furt rth her se sev ven protonpro ton-the therap rapy y fa facili cilities ties ar are e cur curren rently tly und under er con con-struct str uction ion or ful fully ly fun funded ded (T (Table able 1.3) 1.3).. In add additio ition, n, there are six new carbon-ion therapy facilities under developmen deve lopmentt (Heide (Heidelberg lberg and Marbu Marburg, rg, German Germany; y; Pavia, Pa via, Ita Italy; ly; Lan Lanzho zhou, u, Ch China ina;; Wien iener er Neu Neusta stadt, dt,  Austria;  Aus tria; and Maebash Maebashi, i, Japan) that will also provide therapeutic proton beams.  As shown in Table 1.2, more than 53 000 patients are recorded recorded as havi having ng been treated treated with prot protons ons up to December 2007. The growth in patient treatments to December 2003 is shown in Fig. 1.10. As shown, a substantial fraction of proton treatments up to the then n had been for ey eye e les lesion ionss (41 percent percent)) and an d fo forr pr pros osta tate te ca canc ncer er (1 (16 6 pe perc rcen ent) t).. Al Alth thou ough gh prot pr oton on th ther erap apy y is a re rela lati tive vely ly ne new w te tele leth ther erap apy y moda mo dali lity ty,, it is cl clea earl rly y in a ra rapi pid d gr grow owth th ph phas ase. e.

Proton Prot on th ther erap apy y is ju judg dged ed li like kely ly to be beco come me an increasingly important component of the radiation oncology armamentarium.

Figure 1.10. Number of patients treated treated annually with protons up to 2003. The data for 1957, 1963, 1968, 1973, and 1978 are annual averages for periods 1954–1960, 1961–1965, 1966– 1970, 1971–1975, andthe 1976–1980, respectively.

20  

Journal of the ICRU Vol 7 No 2 (2007) Report 78 Oxford University Press

2

doi:10.1093/jicru/ndm025

RADIATION RADIA TION BIOL BIOLOGY OGY CON CONSIDE SIDERA RATIONS TIONS

2.1 INTR INTRODUC ODUCTION TION The rationale for proton therapy relates primarily to improved physical dose selectivity. Little clinical benefit is to be expected from selective radiobiological eff effect ects. s. Nev Nevert erthel heless ess,, the re rela lativ tive e bio biolog logica icall effect eff ective ivenes nesss (R (RBE BE)) of pr proto otons ns rel relat ative ive to hig highhenergy (1–30 MeV) photons is significantly greater than This poses two questions:unity. what is thefact bestimmediately RBE estimate to be used in any calcul calculati ation on rela relating ting proton and photon doses doses,, and an d ho how w sh shou ould ld pr prot oton on do dose sess be pr pres escr crib ibed ed,, recorded, and reported? Both of these subjects are discussed in the present section beginning with a discussion of the role of microdosimetry as a predictor of RBE. This will be followed by an evaluation of the available  in vitro   and   in vivo  information on proton RBE values, which leads to a suggestion to use a generic RBE for all therapeutic applications of protons. Finally, the section ends with a discussion si on of th the e im impa pact ct of RB RBE E on th the e is issu sue e of do dose se specification in proton therapy.

2.2 MICR MICRODOS ODOSIMET IMETR RY AND LIN LINEAR EAR ENE ENERGY  RGY  TRANSFER The micr microdos odosimet imetric ric char charact acteris eristics tics of the pro proton ton beam have a strong influence on the RBE for each biological system. Linear energy transfer (LET) has been adopted as an approximation to the microdosimetricc quan metri quantity tity   y¯  ¯ D   (dos (dose e me mean an li line neal al en ener ergy gy)) (ICR (I CRU U 19 1977 77;; 19 1989 89). ). LE LET T is a de desc scri ript ptor or of th the e energy transferred from the beam to the irradiated material, per unit of particle path length (in units

of th the e SO SOBP BP. Th Ther ere e th the e ab abso sorp rpti tion on of en ener ergy gy becomes becom es incr increasin easingly gly rap rapid. id. This is illus illustra trated ted in Fig. 2.2, which presents presents the cumu cumulati lative ve fra fraction ctional al 21 absorption of energy per unit of LET (in keV  m  m m ) from an energ energy-mod y-modulat ulated ed 250 MeV prot proton on beam at four depths,   viz., 2.5, 14.5 and 27.5 cm, and on the th e di dist stal al ed edge ge of an 8 cm SO SOBP BP.1  As discussed earlier, there are very modest differences in energy absorbed per unit LET over the extended range in dept de pth h fr from om 2. 2.5 5 to 27 27.5 .5 cm cm.. Ap Appr prox oxim imat atel ely y 50 perc pe rcen entt of th the e do dose se is ab absor sorbe bed d at 3– 4 ke keV  V mm21. However, at the distal edge of the SOBP, the curve is displaced sharply to the right and 50 percent of the dose is absorbed at  18 keV  mm21. This is associated with an increased RBE,  vide infra.  The consequence of both LET and RBE increasing as dose is decreasing on the distal edge of the Bragg peak is to extend the bio biologi logicall cally y eff effect ective ive ran range ge of the pro proton ton bea beam m by    2 mm for 160–250 MeV beams and  1 mm for 60–85 MeV proton beams,  vide infra. Line Li near ar en ener ergy gy tr tran ansf sfer er ha hass pr prov oven en us usef eful ul in understanding the variation in biological effectiveness of different segments of the range of a particularr pa la part rtic icle le be beam am in an in indi divi vidu dual al bi biol olog ogic ical al system sy stem.. How However ever,, LET does not pred predict ict biolog biological ical respo re sponse nse wit with h hig high h ac accur curac acy y for dif differ ferent ent cel celll or et al al.. tissue tis sue sy syst stems ems.. Fo Forr ex examp ample, le, Wey eyra rathe therr   et (1999) (19 99) det determ ermine ined d cel celll sur surviv vival al cur curve vess and th the e ratio of   a-values in   12C and x-ray beams for exponentia nen tiall ph phase ase V79 V79,, CHO CHO-K1 -K1,, and xrs xrs5 5 cel celll lin lines es growing as monolayers. They observed substantial differences in the ratio of the   a-values versus LET rela re latio tions ns for th the e thr three ee cel celll lin lines, es, as ill illus ustr trat ated ed in Fig. 2.3.

of keV   mm21, for example example). ). Fo Forr pr proto otons, ns, th the e LE LET T increases incr eases slowly along the part particle icle path and then quit qu ite e ra rapi pidl dly y at th the e en end d of th the e pa part rtic icle le ra rang nge, e, resulting in the formation of the Bragg peak (see Fig. 1.1). The variation in LET with depth in monoenergetic energ etic and energ energy-mod y-modulat ulated ed pro proton ton beams is illustrated in Figs 2.1a and 2.1b, respectively. The rate of absorption of energy from an energymodula mod ulated ted pr proto oton n bea beam, m, pr produ oducin cing g a spr spread ead-ou -outt Bragg peak (SOBP), varies only to a modest degree with depth from the entrance to near the beginning

2.3 REV REVIEW IEW OF PUBLIS PUBLISHED HED PRO PROTON TON RBE VALU ALUES ES The pr The pres esen entt re repo port rt is co conc ncer erne ned d wi with th pr prot oton onss of cl clin inic ical al ut util ilit ity y in ra radi diat atio ion n on onco colo logy gy,,   viz., 60–250 Me MeV V be beam ams. s. RB RBE E va valu lues es fo forr lo loww-en ener ergy gy protons,   viz.,   , 10 MeV, exhibit significant energy depe de pend nden ence ce an and d ar are e su subs bsta tant ntia iall lly y hi high gher er th than an 1

Calculated from Coutrakon  et al.  (1997).

# International Commission on Radiation Units and Measurements 2007

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

Figure 2.2. Plot of the cum Figure cumula ulativ tive e ener energy gy absor absorbed bed per unit lineal energy from a 250 MeV proton beam with an 8 cm SOBP at dep depth thss of 2.5 2.5,, 14. 14.5, 5, 27. 27.5 5 cm and at the dis distal tal edge of the terminal term inal Bragg peak peak.. Adapt Adapted ed fro from m Cou Coutra trakon kon,, Loma Linda al.. Universi Univ ersity ty Medi Medical cal Cen Center ter,, usin using g data fro from m Cou Coutra trakon kon   et al (1997). Reproduced with permission.

Figure Figur e 2.1 2.1.. (a) Mi Micr crodo odosim simetr etric ic (li (linea neall ene energy rgy,   y) sp spec ectr tra a measu me asure red d at dif differ ferent ent dep depths ths in a 200 Me MeV V mo monoe noene nerge rgetic tic proton beam,  viz., on the entrance plateau, on the Bragg peak, and on the declining distal edge of the Bragg peak (Binns  et al., 1993; reproduced with permission permission). ). (b) Microdosimetric (lineal energy,   y) spe spectr ctra a mea measur sured ed in a 90 MeV en energ ergy-m y-modu odula lated ted ´   Cathol prot pr oton on be beam am at th the e Un Univ iver ersi site te Catholiqu ique e de Lou Louvai vain. n. Measure Meas urement mentss wer were e perf performe ormed d in the init initial ial pla plateau teau (yell (yellow) ow),,

Figure Figur e 2.3 2.3.. Ra Ratio tio of   a-valu -values es for   12C ion onss an and d x rays as a functi fun ction on of LET for thr three ee cel celll lin lines. es. Th The e ra range nge in LET was was obtained by use of   12C ion track segment irradiation (Weyrather  et al., 1999; reproduced with permission).

RB RBEs Es of cli clinic nical al pr proto oton n bea beams ms (Be (Belli lli   et al., 199 1993; 3; Cox et al., 1977; Raju, 1995b). There is no proposed clinical utility of beams of this low energy. RBE RB E val values ues ha have ve bee been n det determ ermine ined d for a br broad oad array of mammalian cell lines  in vitro  using loss of 

and at the proximal (red), middle (green), and distal (blue) parts of the SOBP, SOBP, as indicate indicated d on the schema at the top of the figure. The y  spectrum for   60Co is given for comparison. Compared with 60 Co, the four proton spectra are slightly shifted toward high   y  values, which might explain the 10 percent relative relative differenc difference e in

colony-fo colony -formi rming ng cap capac acity ity as the end endpoi point nt an and d for in vi vivo vo   experime experimental ntal sys systems tems emplo employing ying div diverse erse celll and tis cel tissue sue re respo sponse nsess as end endpoi point nts. s. Fo Forr the these se experiments, cells or tissues were positioned in the mid-point midpoint of the SOB SOBP P. The refe referenc rence e radi radiati ations ons were   60Co or other high-energy photon beams. The resu re sult ltan antt RB RBE E va valu lues es ha have ve be been en co coll llat ated ed an and d reviewed by Paganetti   et al.  (2002) and are pooled here with the more recent values by Kagawa  et al. (2002) (20 02).. In th the e ass assess essmen mentt of the these se ex exper perime imenta ntall data, special attention was given to the relationship between RBE and dose. For higher LET radiations,  e.g., fast neutrons, RBE has been shown to increase inversely with dose, especially for doses less than about 10 Gy (Field, 1977).

RBE. In addition, there is a progressive shift, with depth, of the proton spectra toward higher  y  values that could be responsible for the slight additional RBE increase of 5–10 percent at the end of the SOBP compared compared with the initial initial pla plateau teau and othe otherr shallower depths in the proton beam. The right ordinate is the biological weighting function that expresses the RBE variation as a fu func ncti tion on of   y   for for th the e ca case se of in inte test stin inal al cr cryp yptt ce cell ll regener rege neratio ation. n. Onl Only y a smal smalll prop proporti ortion on of the pro proton ton spec spectra tra overlap ove rlapss with the asce ascendin nding g part (RBE . 1) of th the e bio biolog logica icall weighting function (as indicated by the pink oval). Adapted from Gueulette,, Universi Gueulette Universite te´  Catholique de Louvain, redrawn from the data of Gueulette  et al.  (2004), Loncol   et al.  (1994), and Menzel  et al.  (1990). Reproduced with permission.

22  

 RADIATION  RADIA TION BIOLOG BIOLOGY Y CONSIDERA CONSIDERATIONS TIONS

determined ined usin using  g   in vitro 2.3.1 RBE values determ and   in vivo  systems

 All known published RBE values at all dose levels for mammalian cell lines studied   in vitro  in proton beams in the clinical energy range are presented in Figs Fi gs 2. 2.4a–c. 4a–c. Th The e ma majo jori rity ty of va valu lues es ar are e fo forr V7 V79 9 2 cells. The mean RBE is 1.19 (1.13–1.24). There is no apparent increase in RBE as dose is decreased to   , 10 Gy Gy.. RBE valu values es are pre present sented ed sepa separat rately ely for V79 (Fig. 2.4 2.4b) b) and non-V79 non-V79 cel cells ls (Fi (Fig. g. 2.4 2.4c). c). The mean RBE value for the V79 cells is 1.24 (1.04 to 1.44), while for the non-V79 cells it is 1.12 (0.98 to 1. 1.22 22). ). No Note te th tha at V7 V79 9 ce cell llss ex exhi hibi bitt a la larg rger er shou sh ould lder er on su surv rviv ival al cu curv rves es th than an do ot othe herr ce cell ll lines. RBE values for a variety of   in vivo   systems are pres pr esen ente ted d in Fi Fig. g. 2. 2.5a 5a.. Th The e an anim imal al ti tiss ssue uess an and d organs org ans st stud udied ied inc includ lude e ski skin n (a (acut cute e re reac actio tion n and late la te con contr tract action ion), ), lun lung, g, jej jejuna unall cry crypts pts,, tai tail, l, ver ver-tebrae, tebr ae, testis, lens, and bone marr marrow ow.. Mean RBE forr al fo alll pr prot oton on en ener ergi gies es,, do dose se le leve vels ls,, an and d ti tiss ssue ue syst sy stems ems is 1.1 1.10 0 (1. (1.09– 09– 1.1 1.12). 2). This val value ue is sig signifi nifi-vitro ro cant ca ntly ly le less ss th than an th the e me mean an va valu lue e fo forr al alll   in vit RBE values. Also, for the   in vivo   studies, there is no ev evid iden ence ce of an in incr crea ease se in RB RBE E as do dose se is reduced from 10 to 0.7 Gy. Furthermore, there is no evid ev iden entt di diff ffer eren ence ce in RB RBE E am amon ong g th the e ti tiss ssue ue systems investigated at any dose level. As discussed earl ea rlie ierr, se seve vera rall of th the e RB RBE E va valu lues es we were re   , 1.0. Figure 2.5b and c shows the RBE values for acutereacting and late-rea late-reacting cting tissues, respectively. No significant difference was detected. There are three determinations of RBE for  in vivo syste sy stems ms at the mid mid-SO -SOBP BP pos positio itions ns of 60– 85 MeV  al..   (1996) determ proton proto n beams. Gueulette   et al determined ined the th e RB RBE E of mou mouse se je jeju juna nall cr cryp yptt cel celll su survi rviva vall fo forr

single-dos singledose e irr irradi adiati ation on by an 85 MeV pr proto oton n beam   60  versuss Co photons. The resultant RBE was 1.083  versu at a photon-absorbed dose of 13.3 Gy resulting in 20 surviving crypts per intestinal circumference. Ando  et al.   (1985) (1985) emplo employed yed inac inactivat tivation ion of mouse fibrosarc sa rcom oma a mi micr croo-col colon ony y fo form rmin ing g cel cells ls in lu lung ng an and d measured RBE values for absorbed doses of 5–6 Gy in a 70 MeV beam. The RBE values in three experimen im ents ts we were re 1. 1.01 01,, 1. 1.14, 14, an and d 1.0 1.02. 2. Fo Forr th the e se secon cond d and third experiments, assays were also performed using a 250 MeV beam. The RBE values were 1.09 and 0.99, respectively respectively..4 Nemoto   et al.   (1998) found the th e RB RBE E fo forr 80 Me MeV V prot oton onss to be 1. 1.20 205 at a

Figure Figur e 2.4 2.4.. (a) RB RBE E ve versu rsuss abs absorb orbed ed dos dose e for all cel celll li lines nes irradiated   in vitro  and using colony formation as the endpoint (Paganetti   et al., 2002; reproduced with permission). (b, c) RBE  versus dose for V79 cells (b) and for non-V79 cells (c) irradia irradiated ted in vit vitro ro, bot both h usi using ng col colony ony for forma matio tion n as the end endpoi point nt.. Da Data ta derived from (a) (Paganetti  et al., 2002).

photon-absorbed dose of 8.5 Gy for acute skin reactions. They also reported the RBE value to be 1.15, 1.24 1. 24,, an and d 1. 1.15 15 at ph phot oton on-a -abs bsor orbe bed d do dose sess of 33. 33.7, 7, 20.7, and 13.0 Gy, respectively. All these values are consistent with an average RBE of 1.1.

2

The confi confidenc dence e band bandss giv given en her here e and thr through oughout out the pres present ent section are the 95 percent limits unless otherwise stated. 3 This value was significantly higher than 1.00 ( p 0.005). 4 Confidence limits for the RBE values were not given and could not be computed from the paper. 5 95 percent confidence limits were given as 1.06–1.62. ¼

23  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

late-radiation injury. RBE values were determined for 200 MeV proton irradiation of the lungs of Balb/c mice using LD50   (the dose lethal to 50 percent of  animals) anima ls) as the endp endpoint oint (Gue (Gueulett ulette e   et al., 2000). Radiation was delivered using a 3 cm high beam in 1, 3, and 10 dose fractions, with 12 h between fractions; the mice were positioned in the mid-portion of a 7 cm SOB SOBP P. RB RBE E val values ues for th the e 1010-fr frac actio tion n schedu sch edule le we were re 0.8 0.86, 6, 0.9 0.95, 5, 1.0 1.05, 5, and 1.02 for LD50  values at 180, 210, 240, and 270 d, respectively. The absorbed doses per fraction for the latter schedule were 2.3–3. 3–3.7 7 Gy (60Co) an and d 2. 2.0–3. 0–3.2 2 Gy (prot (pr oton on). ). Th The e me mean an RB RBE E fo forr th the e 10 10-f -fra ract ctio ion n irradiati irrad iation on was significantly significantly low lower er than the mean RBE of th the e si sin ngl glee-fr fra act ctio ion n as assa say ys ( p , 0.05).  ed 60 the leg al.MeV Tatsuzaki   (1991) (199protons 1) irr irradi adiat ated legss ofinICR mice using  et 250 or Co photons 10 fraction fra ctionss ove overr 11 d and scor scored ed skin cont contrac raction tion at 250 d. The RBE was found to be 1.02–1.03.6 There are no repo reported rted RBE dete determina rminations tions of late reactions in central nervous system tissues after proton irradiation. vitro o   and In summary, the available data on   in vitr in vivo   systems (including acute- and late-reacting tissues) tissu es) are consi consisten stentt with a tissu tissue-ind e-indepen ependent dent mean me an RB RBE E va valu lue e of 1. 1.10 10.. Fur urth ther er,, th ther ere e is no   in vivo vi vo sugges sug gestio tion n fr from om st stud udies ies on   syst systems ems of  an inc increa rease se in RB RBE E as abs absorb orbed ed dos dose e is re redu duced ced to , 3– 4 Gy Gy.. Th This is fin findi ding ng fo forr pr prot oton onss of no do dose se

dependenc depend ence e of RB RBE E is in con contra trast st wit with h th those ose for fast fa st ne neut utro rons ns,, wh wher ere e RB RBE E in incr crea ease sess st stee eepl ply y as abso ab sorb rbed ed do dose se is de decr crea ease sed d be belo low w 4 Gy (F (Fie ield ld,,

1977). The fa 1977). fact ct tha thatt th the e pr proto oton n RB RBE E da data ta do not conform to the expectations from the neutron data may, in part, reflect the difficulty in demonstrating the modest changes in relatively small RBE values. The avai available lable experimenta experimentall stu studies dies (Fig. 2.5a 2.5a)) do not extend into the absorbed dose range ( , 1.0 Gy) where whe re hyp hypers ersens ensiti itivit vity y ha hass bee been n dem demons onstr trat ated ed (Chandna  et al., 2002). 2.3.2 RBE ver 2.3.2 versus sus dep depth th for for 60– 250 MeV  beams Figure Figu re 2. 2.5. 5. (a) (a) Al Alll RB RBE E ve vers rsus us do dose se va valu lues es fo forr ac acut utee- an and d late-reacting late-r eacting experimental animal-tissue systems. The tissues studied include jejunal crypt cells, lung, skin (acute reaction and late contraction), contraction), vertebr vertebral al growth growth,, bone marrow, marrow, testis, lens, and tumor (Paganetti  et al., 2002; reproduced with permission). (b,, c) RB (b RBE E ve vers rsus us ab abso sorb rbed ed do dose se fo forr ac acut ute e (b (b)) an and d la late te (c (c)) reactions reacti ons in experime experimental ntal animal-tissue systems. Data derived from (a) (Paganetti   et al., 2002), supplemented with data from Kagawa  et al.  (2002).

 A clinically important radiobiological question is the magnitude of change, if any, of RBE along the plateau, across the SOBP, and on the distal portion of the terminal Bragg peak.  2.3.2.1

60 – 85 MeV 

For proton beams of this energy range, the SOBPs vitro ro are    14– 30 mm in dep depth. th. Eig Eight ht rep reporte orted d   in vit measur mea sureme ements nts of RB RBE E valu values es ve versus rsus depth in the SOBP SO BP fo forr 60– 85 Me MeV V be beam amss ar are e su summ mmar ariz ized ed in

Special Speci al men mentio tion n is mad made e of the findings findings of two studies stu dies on late late-rea -reaction ction endpoints. endpoints. Thes These e are the only published published RBE values for late tissue inju injury ry,, des esp pit ite e the hi high gh cl clin inic ical al in intter eres estt in ri risk sk of 

6

No confidence limits were provided for these RBE values.

24  

 RADIATION  RADIA TION BIOLOG BIOLOGY Y CONSIDERA CONSIDERATIONS TIONS

Tabl able e 2. 2.1. 1. Th Thes ese e re resu sult ltss ar are e co cons nsis iste tent nt wi with th an increase in RBE along the 14–30 mm SOBP of 60– 85 MeV proton beams, but with clear variability in the magnitud magnitude e of the change change in RBE between between the different studies. The very detailed study of Wouters  et al.  (1996) reported an increase of 7 percent across the 24 mm SOBP. Bett Bettega ega   et al.  (2000) determined the th e RB RBE E at 3–12 mm fr from om th the e beg egin inni ning ng of a 14 mm SO SOBP BP of a 65 MeV be beam am.. For on one e ce celll system, there was a 16 percent relative increase and in th the e se seco cond nd ce cell ll sy sysste tem m a 6 pe perrce cent nt rel ela ati tive ve decrease in RBE over those 9 mm. Blomquist   et al. (199 (1 993) 3) fo foun und d no in incr crea ease se in RB RBE E be betw twee een n 2 an and d 17 mm along a 17 mm SOBP for the V79–379A cell line. lin e. Fr From om fiv five e st stud udies ies,, th ther ere e wa wass evi evide denc nce e of an incre inc reas ase e in th the e RB RBE E ov over er th thes ese e sh shor ortt le leng ngth thss of  SOBP,   viz., relative increments of 7–23 percent. In one study, it was shown that there was no change in RBE values at 9 and 14 mm along a 14 mm SOBP (Matsumura  et al., 1999).  2.3.2.2

RBEs ve RBEs versu rsuss dep depth th com comput puted ed for sur surviv viving ing fr frac ac-tions tio ns (S (SF) F) of 0.1 and th their eir 95 per percen centt con confide fidence nce 11 limits lim its ar are e pr prese esent nted ed in Fig Fig.. 2.7 2.7.. Th There ere we were re no incre inc rease asess in RB RBE E wit with h dep depth th fr from om th the e ent entran rance ce through the SOBP until the final 1 cm proximal to the end of range for the 155 and 200 MeV beams, viz., at 12.2 and 19.9 cm, respectively. This increase was significant for the 155 MeV beam. In contrast, for the 250 MeV beam, there there wa wass no inc increa rease se in RBE RB E wi with th de dept pth h in incl clud udin ing g th the e po poin intt at ab abou outt 1.4 cm pr proxi oximal mal to the fina finall Br Bragg agg pea peak, k,   i.e., at 30.0 cm depth. The endpoint range was 31.4 cm. In a subsequent experiment, Robertson extended these studie stu diess to exa examin mine e RBE at 5 mm bey beyond ond the mid mid-portion of the final peak. His finding was that the ratio of the RBE at the declining edge point to that at the entrance for the 200MeV beam was    1. Raju  et al.   (1978) (1978) inves investiga tigated ted the resp response onse of V79 cells and found that the surviving fractions at absorbed doses of 3, 4, and 8 Gy were constant over the full depth of the 9 cm SOBP of a 160 MeV beam.

160 – 250 MeV 

Robertson  et al.  (1975) employed the H4 (rat hepatoma to ma)) ce cell ll li line ne wi with th 16 160 0 Me MeV V pr prot oton onss an and d   60Co photon pho tonss in the their ir det determ ermina inatio tion n of th the e RB RBE E as a 7 function func tion of posi position tion in a 5 cm SOBP. The They y foun found d that th at th the e me mean an of 78 RB RBE E va valu lues es de dete term rmin ined ed at

 2.3.2.3  SOBP

RBE on the declining distal edge of the

The studies in a 5 cm SOBP of a 160 MeV beam by al..   (1975) Robertson   et al (1975) inclu included ded meas measurem urement ent of  the variation (in    0.5 mm intervals) of RBE from

absorbed doses of   3 Gy for positions proximal to the final Bragg peak was 1.00 + 0.01.8  Additionally  Additionally,, forr 10 fo 102 2 de dete term rmin inat atio ions ns at ab abso sorb rbed ed do dose sess of 2– 3 Gy Gy,, th the e me mean an RBE was    1.0 1.0.. Fu Furth rther er,, RB RBE E  values were not significantly different from 1.0 at five positions over the 3 mm proximal to the center of the final Bragg peak. That is, there was no evidence for an RBE above 1.0 over the entire SOBP. The Th e va valu lues es fo forr th the e di dist stal al 3 mm an and d be beyo yond nd ar are e shown in Fig. 2.6a. Slabbert   et al.   (1994) reported RBE values for V79 cells irradiated by a 200 MeV  proto pr oton n bea beam m re relat lativ ive e to a   60Co bea beam. m. Th They ey mad made e measur mea sureme ements nts for abs absorb orbed ed dos doses es bet betwe ween en 2 and 12 Gy at depths of 43.5 (initial plateau) and 141.5,

3 mm before, to 6.5 mm beyond, the midpoint of the distal dis tal Bra Bragg gg peak contributing contributing to the SOB SOBP P. The RBE incr increased eased from    1.0–1.4 fr from om the mid midpoi point nt to 6 mm beyond the distal Bragg peak (Fig. 2.6a). No RBE variation was observed in the 3 mm proximal to this midpoint. In Fig. 2.6b, the absorbed dose curve is compared with the exper experiment imentally ally determ determined ined RBERBE-weigh weighted ted absorbe abs orbed d dos dose e cur curve ve (ab (absorb sorbed ed dose  experimentally tal ly det determ ermine ined d RBE RBE). ). In the reg region ion betw between een    1 and 4 mm beyon ond d th the e dis ista tall Br Bra agg pea eak, k, th the e RBE-weighted absorbed dose is    8 percent greater than the RBERBE-weigh weighted ted absorbed dose in the SOBP. SOBP.

167.5, and 191.5 mm in a 7 cm SOBP; the 191.5 mm depth was at 1 cm proximal to the distal peak. The resu resultan ltantt mean RBE valu values es wer were e 1.00, 1.04, 1.07, and 1.16,9 respectively. Coutrakon   et al.   (1997) (1997) empl employed oyed V79 cells and 10 determined   a   and   b   values at 2.5 cm inter intervals vals from fr om th the e en entr tran ance ce to th the e di dist stal al po port rtio ion n of 8 cm SOBPs SOB Ps of 155 155,, 200 200,, an and d 250 Me MeV V pr proto oton n bea beams. ms.

The effect of the rather increment in eRBE is primar pri marily ily to extend ext end the large RBE-w RB E-weigh eighted ted dos dose curve cur ve along the declining edge of the final Bragg peak by  2 mm as the relative dose decreases from   100–85 percent and by  1 mm at the 50 perce percent nt dose level. al..   (1994 Slabbert   et al (1994)) det determ ermine ined d the RBE for monolayers of V79 cells at the distal end and on the decli de clini ning ng ed edge ge of th the e SO SOBP BP in a 200 Me MeV V bea beam. m. Incr In crea ease sed d RB RBE E va valu lues es of 1. 1.3, 3, 1. 1.4, 4, an and d 1. 1.5 5 wer ere e obtained at the end of the SOBP and at the 35 and 32 percent isodose levels, respectively. In contrast to

7

The endpoint endpoint for the RBE deter determin minati ations ons was SF (sur (survivi viving ng fraction) 0.1. 8 The standard error. 9 These were not significantly   . 1.00. 10 a   and  b   are parameters in the linear quadratic model for cell survival. ¼

11

RBE values were computed as the ratios of dose to yield an SF of 0.1 at the surface, to the dose to yield the same SF at each specified depth.

25  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  Table 2.1.   In vitro  RBE values versus depth in the SOBP of 60–85 MeV proton beams.

Energy (MeV)

Reference beam

85 70 67 65 65

 

60

 

60

65 65

 

137

 

137

65

     

Co Co 60 Co 60 Co 60 Co Cs Cs

Proton beam plateau

Cel elll li line ne

End En d po poin intt

SOBP SOBP width (mm)

RBE

SOBP proximal

SOBP middle

SOBP distal

RBE ratio distal/proximal

Reference

CHO V79 V79 379A SCC25 SCC25 progenyb CHO V79

SF 0.5 SF 0.5 SF SF 0.5 SF at 2 Gy SF at 2 Gy

30 24 20 14 14

0.94 1.28 1.63 1.05 1.13

1.22 1.30 1.63

1.16 1.38 1.63 1.22 1.05

1.23 1.07 1.00 1.16 0.94

Gueulette  et al.  (1996) Wouters  et al.  (1996) Blomquist  et al.  (1993)a Bettega  et al.  (2000) Bettega  et al.  (2000)

SF SF SF S F

0.67 0.5

18 14

1.19

1.23 1.35

1.37 1.34

1.15 1.00

Tang  et al.  (1997) Matsumura  et al.  (1999)

CAL4

SF

0.01

28

1.09

1.12

1.27

1.16

Courdi  et al.  (1994)

¼

¼

¼

¼

¼

¼

SF, surviving fraction.a In this paper Blomquist  et al.   (1993) also reported that the RBE values for the LS-174T cells tended towards higher values with dose,  viz.  the RBEs were 1.21, 1.23, and 1.26 for SFs of 0.5, 0.12, and 0.001, respectively. This was the opposite of  the results for the V79 cells,  viz . the RBE for those SFs were 1.63, 1.28, and 1.15, respectively. b These experiments were performed on the progeny of irradiated SCC25 cells.

Robertson   et al.’s (197 (1975) 5) exp experim eriment ent,, the there re was an

vivo   systems, and then ‘validated’ clinically. Clinical

increase in RBE in the distal half of the SOBP. Bettega   et al.  (2000), using a 65 MeV beam and human SCC25 cells, determined RBE at 15.6, 25.0,

experience experie nce has not bee been n int interpr erpreted eted as ind indica icatin ting g that the RBE is different from 1.10. The literature on RBE values indicates that there

27.2, 27.8 the latter two positions were and at the 91mm anddepths; 52 percent depth–dose levels, respe re specti ctivel vely y, on the dec declin lining ing edg edge e of the SO SOBP BP. The RBE values obtained were 1.05, 1.22, 1.39, and 2.05 2. 05,, re resp spec ecti tive vely ly.. Th Thes ese e hi high gher er RB RBEs Es on th the e decl de clin inin ing g ed edge ge of th the e SO SOBP BP ha hav ve th the e ef effe fect ct of  extending the RBE-weighted range by  1 mm. Paganetti and Goitein (2000), using a biomathematic ma tical al mod model, el, con conclu cluded ded tha thatt th the e ext extens ension ion in depth of the RBE-weighted range is a function of  the proton energy and energy spread.

  in vivo  systems that are no experimental data from support the use of a tissue-, dose-, dose per fraction-, or en energ ergyy- (6 (60– 0– 25 250 0 Me MeV) V) sp spec ecific ific pr prot oton on RB RBE E fo forr murine tissues.  Accordingly  Accord ingly, the use of a generi genericc RBE in proton pro ton-r -radi adiat ation ion the therap rapy y is jud judged ged to be clin clinical ically ly appr ap prop opri riat ate e an and d is th ther eref efor ore e re reco comm mmen ende ded. d. A  generic proton RBE means the use of a single value, independent of the tissue irradiated, dose per fraction, total dose, proton energy, and position on the physical depth–dose curve up to the midpoint of the term te rmin ina al Bragg pea eak. k. The RBE value to be employed should be that which best fits pooled RBE  values from in vivo  studies,  e.g., 1.1. This applies to

2.4 USE OF A GENER GENERIC IC RBE RBE VALU ALUE E There are no proton RBE values based on humantissue response response data data,, despi despite te clinical experience experience of  the treatment of more than 50 000 patients. All of  the available proton RBE values on tissue systems with use useful fully ly nar narro row w confi confiden dence ce limi limits ts ha have ve been deri de rive ved d fr from om da data ta on la labo bora rato tory ry ex expe peri rime ment ntal al syst sy stem ems. s. Es Esti tima mate tess of RB RBE E re requ quir ire e da data ta fr from om a range of doses for both the photon and proton arms delivered to relatively homogeneous subjects treated in a defi defined ned protoco protocol. l. Thi Thiss con condit dition ion is not really really feasible in clinical medicine but is readily achieved in lab labora orator tory y exp experim eriment ents. s. Acc Accord ording ingly ly,, the RB RBE E  value to be applie applied d clinical clinically ly must be deriv derived ed from laboratory-based investigations, preferably using   in

all tissues in the direct beam path. In ad addi diti tion on to inc incre reas asin ing g th the e le leve vell of th the e RB RBEEweight wei ghted ed abso absorbed rbed dos dose e (re (rela lativ tive e to the abs absorbe orbed d dose), the application of a generic RBE of 1.1 to a comp co mput uted ed ab abso sorb rbed ed de dept pth– h– do dose se cu curv rve e sl slig ight htly ly increases incre ases the range of the RBE-weighted RBE-weighted absorbed depth dose. While the use of a generic value of 1.1 to convert the th e ab absor sorbe bed d do dose se to a RB RBEE-we weig ight hted ed ab abso sorb rbed ed dose do se in a pr prot oton on be beam am is re reco comm mmen ende ded d in th the e present pres ent rep report, ort, seve several ral expe experimen rimental tal dat data a sugge suggest st thatt the RBE might increas tha increase e by 5 –10 per percen centt in the deepest part of the SOBP relative to the middle of th the e SOB SOBP P. Th There ere is als also o evi eviden dence ce (se (see e Sec Sectio tion n 2.3.2. 2.3 .2.3) 3) tha thatt the RB RBE E doe doess inc incre rease ase sig signifi nifican cantly tly 26

 

 RADIATION  RADIA TION BIOLOG BIOLOGY Y CONSIDERA CONSIDERATIONS TIONS

Figure re 2.6. RBE determina determination tion for H4 cell cellss at SF 0.1  (surviving fraction of 0.1) in an unmodulated 160 MeV proton beam and in the Figu distal region of a 5 cm SOBP produced by the same beam. The measurements cover a range from 3 mm proximal to 6.5 mm distal to the Bragg peak (Robertson  et al., 1975; reproduced with permission). (a) RBE as a function of depth normalized to the RBE at the midpoint of the distal Bragg peak contributing to the SOBP. Open circle represents RBE for unmodulated beam. Filled circle represents RBE for

modulated beam. (b) Variation with depth of the absorbed dose and the RBE weighted absorbed dose. The two curves are normalized to their values at the depth of the midpoint of the distal Bragg peak. Open circle represents RBE-weighted absorbed dose. Filled circle represents absorbed dose.

over the ini over initia tiall fe few w mil millim limete eters rs of the dec declin lining ing edge of the SOBP relative to the RBE at the depth

common practice to report, in addition to absorbed dose, dos e, the ‘eq ‘equiv uivale alent nt’’ or ‘co ‘cobal balt-e t-equi quival valent ent’’ dos dose. e.

of distal peak. This yields an increase in thethe range ra nge ofBragg the th e RB RBE-w E-weig eighte hted d abs absorb orbed ed dos dose e by 1–2 mm (see Fig. 2.6). These effects might need to be considered in treatment planning, especially for single-field treatments and when organs at risk are located at these positions.

 Absorbed dose is a fundam fundamental ental quantity used in all ther th erap apeu euti ticc ap appl plica icati tions ons of ion ioniz izin ing g ra radi diat ation ion.. Measu Me asure reme ment nt an and d re repo port rtin ing g of abs absorb orbed ed do dose se is crucial to the understanding of any radiation effects

This ‘equivalent’ definedRBE. as theItproduct of  the absorbed dosedose andwas the proton thus represents the photon dose that would give the same therapeutic effect as the actual proton dose, assuming all irradiation conditions including the number of fr frac actio tions ns and ov over erall all tr trea eatme tment nt tim time, e, ar are e the same for both radiation qualities. The use of the term ‘equivalent dose’ as defined abov ab ove e ca cann nnot ot be re reco comm mmen ende ded d fo forr th ther erap apeu euti ticc applicat appl ications ions.. The term ‘equi ‘equivalen valentt dose’ has pre pre- viously been defined for radiation protection purposess (ICR pose (ICRP P, 1991 1991)) and is alre already ady used in seve several ral importan impo rtantt nat national ional and inter internat nationa ionall regu regulato latory ry document docu ments. s. Mor Moreov eover er,, the term ‘equi ‘equivalen valent’ t’ could could,,

(BIPM, 2006; ICRU, 1993b; 1998). As a general recommendation and in line with ICRU Reports 50, 62, and an d 71 (I (ICR CRU, U, 19 1993 93b; b; 199 1999; 9; 20 2004) 04),, th the e ab absor sorbe bed d dose should be specified at a certain number of relevantt poi evan points nts and and/or /or in spe specific cific vol volume umes. s. How Howeve everr, absorbed dose alone is not a sufficient predictor of  thera the rapeu peutic tic out outcome come.. The There refor fore, e, all rel relevan evantt par par-ameters such as absorbed dose, fractionation, overall time, tim e, and rad radiat iation ion qua qualit lity y sho should uld be rep report orted. ed. In radiation radia tion ther therapy apy,, when compar comparing, ing, combin combining, ing, or exchanging excha nging inform information ation for trea treatments tments performed under und er dif differ ferent ent con condit ditions ions,, we weigh ighting ting of abs absorbe orbed d dose do se is of ofte ten n us usef eful ul an and d so some meti time mess ne nece cess ssar ary y (Gre´ goire et al., 2004; Wambersie et al., 2004a; 2006). To ac accou count nt for th the e fa fact ct tha thatt pr proto otons ns ex exhib hibit it a RBE significantly different from unity, it has been

in any case, be misleading to the extent that the equiva equ ivalen lence ce is onl only y rel relat ative ive to pho photon tonss del deliv ivere ered d under un der th the e sam same e con condit dition ionss as the pr proto otons. ns. Whe When n comparing comp aring two pro proton ton trea treatmen tments ts deliv delivered ered with differen diff erentt fra fraction ctionatio ation n sche schemes, mes, the ther therapeu apeutic tic effects of equal ‘equivalent’ doses could be different if the respective reference photon treatments were delivered with a different fractionation scheme. It has been common practice to report ‘equivalent dose’ in units of ‘gray equivalent’ or ‘cobalt gray equivalen al ent’ t’ us usin ing g sy symb mbol olss su such ch as CG CGE, E, Gy GyE, E, or Gy Gy(E (E). ). ´ s /   Syste te`me Inte Interna rnation tional al d’Un d’Unite ite´  How Ho wev ever er,, th the e SI ( Sys International System of Units) does not permit the use of arbi arbitr trary ary uni units ts nor nor the add additi ition on of wor words, ds, sub subscr script ipts, s, asterisks, etc. to the unit; hence, the use of CGE, GyE, or Gy(E) is not recommended (BIPM, 2006).

2.5 DOSE SPE SPECIF CIFICA ICATION TION

27  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  †

†

 Symbol   D   represe represents nts the pro proton ton absor absorbed bed dose (in Gy).  Symbol  D RBE  (in Gy) is the RBE-weighted proton absorb abs orbed ed dos dose e and is the dose of ph photo otons ns that would produce the same therapeutic effect as a proton-absorbed proton-abso rbed dose,   D, giv given en und under er ide ident ntical ical circumstances.

Because RBE is dimen Because dimensionles sionless, s,   D   and   DRBE   are both expressed in Gy. The use of a single unit for two quantities may be seen as a source of confusion. To avoid this confusion it is recommended that the quantity quanti ty,, either   D   or   DRBE, be alw alway ayss ex expl plic icit itly ly specified. For example, one can write ‘the absorbed dose to the PTV was 63 Gy (or  D 63 Gy) and/or ¼

PTV 

the RBE RBE-w -weigh eighted ted abso absorbed rbed dose to the PTV was 70 Gy (o (orr   DRBE,PTV  70 Gy Gy)) as assu sumi ming ng an RB RBE E of  1.1’. To further reduce the possibility of confusion, it is rec ecom omm men end ded th tha at th the e qu quan anti tity ty   DRBE   be expres exp ressed sed in Gy Gy,, foll follow owed ed by a spa space ce and the par par-entheti enth etical cal des descrip criptor tor ‘(R ‘(RBE BE)’. )’. The RB RBE-w E-weig eighte hted d  DRBE absor ab sorbe bed d do dose se sp spec ecifi ifica cati tion on wo woul uld d re read ad ‘ D 70 Gy (RBE)’. This notation will be used throughout the remainder of the present report. Other sections ¼

¼

of thetities present will elaborate on aE-weig variety of  quantiti quan es suc such hreport as   DRBE,PT RBE-w eighte hted d RBE,PTV V,98%, an RB absorbed absorbe d dose encompassing encompassing 98 percen percentt of a PTV. In every case, the quantities   D98%   or   DRBE,98%, representing absorbed or RBE-weighted absorbed doses, respectively, and the chosen RBE value used should be clearly specified. specified. In su summ mmar ary y, th the e co conc ncep epts ts of ab abso sorb rbed ed an and d RBE-wei RBE -weighte ghted d absor absorbed bed doses serv serve e diff differen erentt pur pur-poses. Absorbed dose is a physical quantity derived from fr om me meas asur urem emen entt or ca calc lcul ula ati tion on,, wh wher erea eass RBE-weighted dose is a biologically weighted quantity designed to define doses of protons that would prod pr oduc uce e id iden enti tica call bi biol olog ogic ical al ef effe fect ctss as do dose sess of 

Figure 2.7. RBE ver Figure versus sus depth (sol (solid id cir circles cles)) for 155, 200, and 250 MeV proton beams (see text). The data are normalized to an RBE of 1.0 at 2.5 cm depth for the 250 MeV beam. The circles are  values determine determined d from V79 cell survival curves. The triangles are ar e va valu lues es de deri rive ved d fr from om mi micr crod odos osim imet etry ry me meas asur urem emen ents ts (Coutrakon  et al., 1997; reproduced with permission).

2.5.1 The RBE-w RBE-weighted eighted absorb absorbed ed dose  D

( RBE) To replace the term equivalent dose and its units Gy(E), Gy(E ), GyE, or CGE CGE,, the quan quantity tity RBE-weighte RBE-weighted d absorbed absor bed dose,   DRBE,   will will be use used d in the pr prese esent nt repo port rt to des esig ign nate th the e prod odu uct of the tot otal al proton-absorbed dose,  D , and the proton RBE, with respect resp ect to phot photons ons delivered delivered und under er the same conditi di tion ons. s. Th The e sym ymbo boll   dRBE   wil will denote the RBERB E-we weigh ighted ted abs absorb orbed ed dos dose e per fr frac actio tion. n. Th The e special name of the unit of both absorbed dose and RBE-weighted dose is gray (Gy). In th the e ca case se of pr prot oton ons, s, wh wher ere e us use e of a ge gene neri ricc RBE of 1.1 is recommended:  DRBE  ¼ 1:1  D:

 

photons if administered under identical conditions.  As such, the absorbed dose will have a primary role in dosimetry protocols and a prominent role in any clinical protocol and final report. The RBE-weighted dose is better suited to a comparison of the effects of phot photon on vers versus us prot proton on ther therapy apy,, for the selection of appr appropria opriate te pro proton ton dose doses, s, and the predictio pred iction n of ther therapeu apeutic tic outco outcomes mes based on pre pre- vious experience with photons. Whether the quantities absorbed dose and/or RBE-weighted absorbed dose do se sh shou ould ld be us used ed in cl clin inic ical al pr prac acti tice ce in th the e different steps of treatment preparation and planning procedures is a matter of experience and local policy. It is, however, important and obligatory that the quantities involved be clearly specified to avoid any risk of confusion.

ð2:1Þ

28  

Journal of the ICRU Vol 7 No 2 (2007) Report 78 Oxford University Press

doi:10.1093/jicru/ndm026

3

BEAM BE AM DEL DELIV IVER ERY Y AND PR PROP OPER ERTI TIES ES

3.1

PROTON-THERA PROT ON-THERAPY PY FAC ACILITIES ILITIES

delivered within a reasonable time. Typically beam 11

11

 A typic typical al prot proton-th on-therapy erapy facil facility ity,, shown in Fig. 3.1, comprises several main components: (i) an accelerator wit with h an ene energy rgy-se -selec lectio tion n sy syst stem em to pr produ oduce ce energetic energ etic prot protons, ons, (ii) a beam beam-tra -transpor nsportt sys system tem to steerr the beam to the trea stee treatmen tment-de t-deliver livery y sy syste stem, m, and (ii (iii) i) a tr trea eatme tmentnt-del deliv ivery ery sy syst stem. em. Th The e la latte tterr compri com prises ses sev severa erall sub subsy syst stems ems an and d ma may y inc includ lude e some so me or al alll of th the e fo foll llo owi wing ng:: a ga gant ntry ry,, a be beam am nozzle, a snout, a volume-tracking and beam-gating device dev ice,, an and d a pa patie tientnt-pos positi itioni oning ng and imm immobi obiliz liz-ation system. The final compon one ent of the proton-t prot on-thera herapy py fac facility ility is a shiel shielded ded enclo enclosure sure to sepa se para rate te th the e acc ccel eler erat ator or an and d be beam am-t -tra rans nspo port rt syst sy stem em fr from om th the e tr trea eatm tmen entt ro room oms, s, in or orde derr to prote pr otect ct pa patie tients nts and to all allow ow per person sonnel nel to mo move ve freely between treatment rooms while the beam is in use within adjacent restricted areas. The accelerator era tor and its ene energy rgy-se -selec lectio tion, n, bea beam-t m-tra ransp nsport ort,, and treatment-delivery systems need to interact to prod pr oduc uce e th the e de desi sire red d tr trea eatm tmen ent. t. Fo Forr ex exam ampl ple, e, energy ene rgy cha change ngess are necessar necessary y to sca scan n the beam thro th roug ugh h th the e ta targ rget et vo volu lume me,, wh whic ich h ma may y re requ quir ire e chan ch ange gess in th the e ac acce cele lera rato torr cy cycl cle e or an en ener ergy gy-selection select ion sys system, tem, the beam beam-tra -transpo nsport rt sy system stem and the beam nozzle and snout. From the perspective of  the pa patie tient nt tr trea eatm tment ent,, th the e tr trea eatme tment nt-de -deliv livery ery sy syst stem em faci is the most t efore, signifi sig nifican t com compon ponent entsecti of the overall ove rall facility lity.. mos Therefor Ther e, cant in the present pres ent section, on, the pat patient ient supp support, ort, immo immobiliz bilizatio ation, n, and posit positionioning in g syst stem em ar are e co cons nsid ider ered ed fir firsst, fo foll llow owed ed by disc di scus ussi sion on of th the e ga gant ntry ry or fix fixed ed be beam am li line ne,, th the e beam be am-t -tra rans nspo port rt sy syst stem em,, th the e ac acce cele lera rato torr, an and d energy-selection requiremen requirements. ts. The overall aim of the facility is to deliver therapeut pe utic ic dos oses es of prot oton on bea eam ms to tu tum mor sit ites es anywhe any where re in the hu human man bod body y. Th This is aim re requi quire ress a pr proto oton n bea beam m of su suffic fficien ientt ene energy rgy to pen penetr etrat ate e pastt the cen pas center terlin line e in the thickes thickestt reg region ion of the body (i.e., in the pelvis) pelvis) pot potent ential ially ly at an obl obliqu ique e angle.. In pra angle practice ctice,, the beam pene penetra tration tion must be 26–38 cm in hu huma man n ti tiss ssue ue,, wh whic ich h req equi uire ress an accelerator capable of producing proton energies of  200– 250 MeV. In add additi ition, on, the bea beam m mu must st ha have ve sufficient intensity to allow therapeutic doses to be

intensiti intens ities es of bet betwe ween en 1.8  10 and an d 3. 3.6 6  10 part pa rtic icle less pe perr mi minu nute te ar are e req equi uirred if do dose sess of  21 2 Gy min are to be delivered uniformly to target  volumes of one liter liter.. The exact energy and intensity requ re quir irem emen ents ts de depe pend nd cr crit itic ical ally ly on th the e mo mode de of  beam delivery (either scattering or scanning) that is actually used.

3.2

THE TREA TREATMENT TMENT-DELIVE -DELIVERY RY SYSTEM

The tre treatme atment-d nt-deliv elivery ery sy system stem comp comprises rises seve several ral major maj or sub subsy syst stems ems:: th the e bea beam m noz nozzle zle,, sno snout, ut, th the e patient pat ient supp support, ort, immo immobiliza bilization, tion, and posit positionin ioning g system, and the gantry. 3.2.1 3.2. 1

The bea beam m nozzle nozzle and snou snoutt

Beam-delivery Beam-deli very techn techniques iques are comm commonly only cat categoregorized ize d as pas passiv sive e or dyn dynami amic. c. Th This is ca categ tegori oriza zatio tion n refers to the method used to spread out the beam laterally. The spreading out of the beam in depth is often done dynamically (i.e., in a time sequence of  steps) or passively using a ridge filter. The spreading of the beam in depth is used with both scattering systems systems ( pass passive ive beam delivery) delivery) and uniform scanning scan ning sy system stemss (dyn (dynamic amic beam deliv delivery). ery). Thes These e beam-s bea m-spr pread eading ing dev device icess are inc incorp orpor orat ated ed in the nozzle and sno nozzle snout ut,, tog togeth ether er wit with h pa patie tientnt-spe specifi cificc beam-modifying devices.  3.2.1.1

Passive beam-delivery techniques

Passive beam deliv Passive delivery ery invo involves lves ‘sca ‘scatter ttered ed beams beams’’ and is a met method hod of ac achie hievin ving g a spa spatia tially lly un unifo iform rm dose distribution distribution by scat scatterin tering g and degrading degrading the primary proton beam in a set of distributed absorbers be rs to cr crea eate te th the e be beam am di diam amet eter er,, ma maxi ximu mum m ener en ergy gy, an and d en ener ergy gy sp spre read ad ne need eded ed to de deli live verr unif un ifor orm m do dose se to th the e ta targ rget et at al alll de dept pths hs.. An examp ex ample le of suc such h a sy syst stem em is a ro rota tatin ting g pr prope opelle llerr with varia variable-t ble-thickn hickness ess blade blades, s, as first discussed discussed by Wilson (1946) in combination with a separated al.., pair pa ir of sc scat atte tere rers rs (I (ICR CRU, U, 19 1998 98;; Ko Koeh ehle lerr   et al 1977). The field shape is determined by a block or aperture, the shape of which is determined by the

# International Commission on Radiation Units and Measurements 2007

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

Figure 3.1. ProtonFigure Proton-ther therapy apy fac facilit ility y lay layout out at Mass Massach achuset usetts ts Gene General ral Hospi Hospital tal (Fr (Franci anciss H. Burr Pro Proton ton The Therap rapy y Cen Center) ter).. For a synchrotron-based facility, the layout is similar. Facilities can have as many as six treatment rooms. (Adapted from Goitein  et al., 2002; reproduced reprodu ced with permission permission.) .)

Figure 3.2. Schemat Schematic ic diagram of the components of the passive scattering scattering nozzle used for proton-therapy proton-therapy treatment at Loma Linda University Medical Center. MWIC, multi-wire ionization chamber; SEM, secondary-electron monitor (Moyers, 1999; reproduced with permission).

projected target volume. The major components of  a passive scattering system are shown in Fig. 3.2.

and the patient and would result in a low efficiency of the beam in use. Double scattering by a pair of 

Unifo Un iformi rmity ty of int intens ensity ity ov over er the use useful ful cr cross oss-section sectio n of the beam can be obt obtain ained ed by sel select ecting ing only onl y th the e cen centr tral al por portio tion n of the Gau Gaussi ssian an dis distri tri-buti bu tion on of a si sing ngly ly sc scat atte tere red d be beam am.. Th This is wo woul uld d require a large drift distance between the scatterer

separa sepa rate ted d sc scat atte tere rers rs ca can n in incr crea ease se th the e ar area ea of  unifor uni form m flu fluenc ence e in th the e cen centra trall are area a at the tr trea eattment position (Koehler  et al., 1977). With this technique,, deve nique developed loped at Harva Harvard rd Univ Universit ersity y Cycl Cyclotron otron Labor Lab orat atory ory,, the firs firstt sca scatte tterin ring g foi foill re resul sults ts in a 30

 

 BEAM DELIVER DELIVERY Y AND PROPERTIES

Gauss Gaussian ian dose dis distribu tribution tion at the trea treatmen tmentt position with a high intensity in the center of the field.  A second composite scatterer scatterer,, placed between the

different thickness sectors. The resulting dose distribution can be made essentially uniform in depth over a distance determined by the thickness of the

first scatterer and the treatment position, typically has an occluding ring on the beam axis followed by a thin, high- Z  scatterer ( Z  is the atomic number of  the element). This combination reduces the dose in the th e ce cent nter er an and d us uses es th the e sc sca att tter ered ed pr prot oton onss to increase the dose outside the center. This technique

 volume to be irradiat irradiated. ed. Together with collimation of th the e be beam am,, th this is re resu sult ltss in a si sing ngle le-fi -fiel eld d do dose se distri dis tribut bution ion tha thatt can be con confor formed med to the dis distal tal surface of the target volume contour by the use of a patie pa tientnt- and bea beamm-spe specifi cificc com compen pensa sator tor (IC (ICRU RU,, 1998). A limitation of a passive scattering system is

results in a larger homogeneous, circular dose distribution at a specific distance from the two scattere te rers rs an and d a mo more re ef effic ficie ient nt us use e of th the e pa part rtic icle les, s, althou alt hough gh a lar large ge dri drift ft dis distan tance ce is st still ill nec necess essary ary.. The homo homogeneo geneous us dose distribution distribution must be substantially larger in each direction than the target  volume, and collimators (blocks or apertures) are needed nee ded to sha shape pe th the e pr proto oton n fiel field d to th the e pr proje ojecte cted d target cross-section. A drift distance of at least 3 m is rec recomm ommend ended ed bet betwe ween en sca scatt ttere ererr and pa patie tient. nt. This is not a problem for a fixed, horizontal beam line li ne,, bu butt wo woul uld d re resu sult lt in a la larg rge e di diam amet eter er fo forr a gantry gantr y beambeam-deliv delivery ery sy system stem when the scat scattere terers rs are placed after the last bending magnet. Smaller drift distances of 1.5–2 m would yield an undesirable abl e inc incre rease ase in sur surfa face ce dos dose e re relat lativ ive e to dos dose e at depth dept h (Rab (Rabin, in, 1987). The gant gantry ry desig designed ned for the Loma Lom a Lin Linda da fa facil cility ity ha hass a dia diamet meter er of 12 m, but econom eco nomize izess on the cos costt of shi shield elding ing by arr arran angin ging g most of the magnets in one plane (ICRU, 1998).  A refinement of the double-scat double-scattering tering technique uses a contoured second scatterer made of a combi Z   materia nati na tion on of a hi high gh-- Z   and and a lo loww- Z material. l. The high- Z   scatterer provides the main scattering and is combined with a plastic counterpart, thinner at the th e ce cent nter er an and d wi with th in incr crea easi sing ng th thic ickn knes esss at increasing radii to ensure the same energy loss of 

that the compensator is designed to shape the dose distri dis tribut bution ion to th the e dis distal tal sur surfa face ce of the pla planni nning ng targett volu targe volume me (PT (PTV V, defin defined ed in Sect Section ion 5.1.4 5.1.4)) and, hence, this shape is also imposed on the proximal PTV surface, so that one cannot avoid exposure of   volumes of normal tissue in the proximal region to approximately the full target dose. For large, irregular target volumes, the unnecessary exposure of  normal norm al tiss tissues ues adjacent adjacent to the target volume can be reduced with the use of a dynamic multi-leaf collimator in combination with a compensating bolus and an d a ste tepp-wi wise se red educ ucti tion on in th the e ran ange ge of th the e protons (Chu  et al., 1993; ICRU, 1998). The field-shaping device and compensator should be as close to the patient’s skin as possible to minimize miz e th the e sca scatte tterin ring, g, whi which ch deg degra rades des the la later teral al penumbra (LP). In order to achieve this, these component pon entss ar are e mou mounte nted d so tha thatt the they y ma may y be mo moved ved along the central beam axis, toward or away from the th e pa patie tient nt.. The These se re retr trac actab table le com compon ponent entss ar are e mounted on a movable snout, which is carried on the nozzle, where the fixed components of the last part of the beam line are supported. In a pa pass ssiv ive e syst stem em,, be beam am mo modu dula lati tion on is a prob pr oble lem m be beca caus use e a sp spec ecifi ificc ra rang nge e mo modu dula lato torr is requ re quir ired ed fo forr ea each ch en ener ergy gy an and d ea each ch sp spre read ad-o -out ut Brag Br agg g pe peak ak (S (SOB OBP) P) le leng ngth th,, if va vari riat atio ion n in do dose se

the protons over the entire scatterer  et al. (Grussell , 1994). If, in surface additionoftothe using a contoured scatterer, scatterer, the scattering and range-shifting elements are placed far upstream of the patient, a sharper dose fall-off and a higher beam-usage efficiency of    46 percent result. Beam spreading can also als o be ac achie hieve ved d by pa passi ssive ve mag magnet netic ic dis disper persio sion n into a circular or linear shape (Blosser  et al., 1991; ICRU, 1998). In order to achieve optimum beam shaping with pass pa ssiv ive e sca cattter erin ing g te tech chn niq ique ues, s, mos ostt be beam am-modifying devices are patient- and/or field-specific, as shown schematically in Fig. 3.3. However, it is cust cu stom omar ary y to us use e a di disc scre rete te se sett of no nonn-pa pati tien entt-

th thro roug ugho hout the th e SO SOBP BP is to mod be ulator mini mi nimi mize zed. d.els This Th is requi re quire res s ut a lar large ge numbe nu mber r of modula tor wheels whe or ridge filters, and in practice the dose uniformity in the SOBP may be dictated by the variety of beam modula mod ulator torss av avail ailabl able. e. In mod modern ern hig high-p h-pat atien ienttthrou th roughp ghput ut pr proto oton-t n-ther herap apy y fa facil ciliti ities, es, effi efficie cient nt soluti sol utions ons ar are e re requi quired red to th the e pr probl oblem em of pa patie tientntspec sp ecifi ificc be beam am mo modi difie fiers rs fo forr be beam am mo modu dula lati tion on without sacrificing dose uniformity. Precise control of the beam intensity and beam-on time with fast switch swi tching ing,, suc such h as is ac achie hievab vable le in cyc cyclot lotro rons ns by contr con troll olling ing th the e bea beam m at th the e ion sou source rce,, or du durin ring g the th e fir first st fe few w tu turn rnss of acc ccel eler erat atio ion, n, pr prov ovid ide e a method met hod for bea beamm-int intens ensity ity mod modula ulatio tion. n. In pri prinn-

specifi spec ificc ra rang nge e sh shif ifte ters rs an and d mo modu dula lato tors rs.. Ra Rang nge e modulation is achieved by varying the thickness of  absorber absor ber mat material erial traversed traversed by the pro protons tons.. This can be don done e spa spatia tially lly by usi using ng an abs absorb orber er pla plate te with ridge-shaped elevations, called a ‘ridge filter’, or in time, with a rotating absorber propeller with

ciple, cipl e, a si sing ngle le mo modu dula lato torr wh whee eell ca can n be us used ed to spread out the Bragg peak to any desired extent if  the beam can be switched on and off in synchrony with wi th se sele lect cted ed se segm gmen ents ts of a mo modu dula lato torr wh whee eell and an d if th the e be beam am in inte tens nsit ity y ca can n be co cont ntro roll lled ed.. In synchrotrons, using radiofreque radiofrequency ncy (RF)-driven 31

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

Figure 3.3. Schemati Schematicc diagram showing how patient-specific beam-modifying beam-modifying devices (range shifter, shifter, range modulator, modulator, apertur aperture, e, and bolus) are used to shape a passively scattered proton beam to conform to the tumor position, dimensions, and distal edge.

extraction, the beam can be switched on and off in  200–250  m s within a single spill, which may be suffici suf ficient ent for ac accur curat ate e sy synch nchron rony y (Hi (Hira ramot moto o and Nish Ni shi, i, 19 1992 92). ). In ad addi diti tion on,, rec ecen entt ad adva vanc nces es in beam-inten beamintensity sity control migh mightt allo allow w a redu reduction ction in

One disadvantage of beams that feature the use of a double-scattering system is that they have considerably sider ably larger penumbra penumbra due to the large effective tiv e sou source rce siz size e pr produ oduced ced by th the e two sca scatte ttere rers. rs. This Th is is on one e re reas ason on wh why y dy dyna nami micc be beam am-d -del eliv iver ery y

the number number of mod modula ulator tor whe wheels els to be use used d wit with h synch sy nchro rotr tron on bea beams ms (se (see e Sec Sectio tion n 3.2 3.2.1. .1.2 2 for mor more e details of beam intensity control). In principle, range shifters should not be necessary in scattering systems with either a synchrotron (where it should be relatively easy to vary the exact energy ene rgy fr from om pu pulse lse to pu pulse lse), ), or wit with h a cyc cyclot lotron ron (where (wh ere ene energy rgy sel select ection ion is av avail ailabl able). e). The use of  comput com puterer-con contr troll olled ed mu multi lti-le -leaf af col collim limat ators ors can further furt her simp simplify lify the situ situatio ation n by elimi eliminat nating ing the need for individualized blocks or apertures for each field fiel d tr trea eated ted.. Th Thus, us, the nu numbe mberr of ind indivi ividua dually lly manu ma nufa fact ctur ured ed pa pati tien entt-sp spec ecifi ificc be beam am mo modi difie fiers rs requi re quired red in pas passiv sive e sca scatt tteri ering ng can be re reduc duced ed to one per field per patient (a set of compensators to shape the dose distribution to the distal surface of  the PTV). The pass passive ive beam beam-disp -dispersal ersal tech technique niques, s, which simultaneously irradiate the entire target volume, have hav e many advantages advantages such as saf safety ety,, simp simplicit licity y, and a lower sensitivity to the time structure of the proton beam than any of the dynamic techniques. However How ever,, passi passive ve scat scatterin tering g tech techniqu niques es tend to be sensitive to variations in beam position relative to the th e sc scat atte tere rerr. Fo Forr mo moni nito tori ring ng an and d do dosi sime metr try y, passiv pas sive e bea beam m spr spread eading ing re resul sults ts in les lesss st strin ringen gentt cond co ndit itio ions ns on ti time me an and d sp spat atia iall re reso solu luti tion on.. Th The e

techniques (providing wobbled and scanned beams) are thought to be advantageous.

 3.2.1.2

Dynami Dyna micc be beam am sc scan anni ning ng is a ti time me-d -dep epen ende dent nt method of achieving a desired dose distribution by magnetica magn etically lly mov moving ing the beam across the targ target et cross cr oss-se -secti ction on whi while le dyn dynam amica ically lly cha changi nging ng th the e energy of the beam and, consequently, the depth of pene penetra tration tion (ICRU, (ICRU, 1998 1998). ). A scan scanned ned beam is comp co mpos osed ed of a nu numb mber er of fin finit ite e pe penc ncil il be beam ams, s, delivered in time sequence, which together result in the desired three-dimensional dose distribution. Ther Th ere e ar are e tw two o mai main n ap appr proa oache chess to del delive iverin ring g a scann sca nned ed bea beam, m, tog togeth ether er wit with h an int interm ermedi ediat ate e  variant: (1) Discrete scanning (spot or voxel scanning) is a method in which the dose is delivered by pencil beam be amss ap appl plied ied in di disc scre rete te st step eps. s. Af Afte terr ea each ch penc pe ncil il be beam am is de deli live vere red, d, th the e pr prot oton on so sour urce ce is int interr errup upted ted,, the bea beam-s m-stee teerin ring g ele elemen ments ts (mag (m agne nets ts and/ an d/or or mech me chan anic ical al posi po siti tion onin ing g devi de vices ces)) ar are e ch chan ange ged d to de deli live verr pr prot oton onss at a differen diff erentt posit position ion and/ and/or or ener energy gy,, and the beam

reduced flexibility in shaping the dose distribution in three dimensions is less important for small or regu re gula larl rly y sh shap aped ed ta targ rget et vol olum umes es.. Ho How wev ever er,, dynamic techniques can reduce the dose to normal tissue tis sue and ma may y be pr prefe eferr rred ed in som some e sit situa uatio tions ns (ICRU, 1998).

is th then en tu turn rned ed ba back ck on unt ntil il th the e des esir ired ed number of protons has been delivered. While it is not essential, the Bragg peaks of the pencil beam be amss ar are e ge gene nera rall lly y de deli live vere red d in a re regu gula larr Cartesian Cart esian grid for reas reasons ons of techn technical ical simp simplilicity ci ty.. Th Thei eirr sp spac acin ing g is ch chos osen en to av avoi oid d no nonnuniformity (ripple) in the dose profile. 32

 

Dynamic beam-delivery techniques

 BEAM DELIVER DELIVERY Y AND PROPERTIES

(2)) Cont (2 Contin inuo uous us sc scan anni ning ng (r (ras aste terr sc scan anni ning ng)) is a meth me thod od in wh whic ich h a pe penc ncil il be beam am of pr prot oton onss is scanned continuously across the cross-section of  the beam in a raster pattern. Variation in intensity as a function of beam position is achieved by continuous control of the proton-beam intensity and/or the scanning speed. Once one ‘layer’ of protons of a particular energy has been laid down do wn,, th the e pr prot oton on so sour urce ce is in inte terr rrup upte ted, d, th the e beam be am en ener ergy gy is ch chan ange ged d by a us usua uall lly y sm smal alll incre inc remen ment, t, and th the e bea beam m is the then n tur turned ned back on to irradiate the next layer. (3) QuasiQuasi-discr discrete ete scan scanning, ning, a poss possible ible varia variant nt of  discrete scanning, which has been used with a carbon-ion beam, is a method in which the ion source is not turned off during the move to the next pencil beam position but is allowed to continue to deliver particles between grid points of  the scan (Haberer   et al., 1993). The dose thus delivered deliv ered between between grid points is acco accounte unted d for in the dose delivered by the next pencil beam. This approach works well only when the time to move to a new position is small compared with the dwell time and the pencil beams are delivered in a spatially contiguous manner.

Figure 3.4. Schemat Schematic ic diagram showing how a variable-en variable-energy ergy pencil beam can be spot scanned in three dimensions through a targett volu targe volume me to pro produce duce a dose distributi distribution on that conf conforms orms to the shape of the target volume.

 volume with a voxel size of 5  5  5 mm 3 needs  10 000 spots and as the treatment time should be

In order to prevent unintended ripple in the dose profile in discrete and quasi-discrete scanning, the inter-pencil beam spacing of near-Gaussian profile beams needs to be    80 percent of the pencil beam’s full fu ll wid idtth at hal alff max axim imu um (fw (fwh hm). In th the e carb ca rbon on-i -ion on be beam am im impl plem emen enta tati tion on de desc scri ribe bed d in Haberer et al.  (1993), the spacing was chosen to be  20 percent of the pencil beam’s fwhm, making the beam delivery much less sensitive to possible fluctuations in the pencil-beam positions. The advantage advantage of dyna dynamic mic beam shap shaping ing using beam scanning is that it can overcome all the problem bl emss ass ssoc ocia iate ted d wit ith h cu cussto tom m-ma mad de be beam am-modify mod ifying ing dev device icess by pot potent ential ially ly con contr troll olling ing the posi po siti tion on of a pe penc ncil il be beam am pr prec ecis isel ely y wi with thin in th the e patient using electronic or electronic/electromechanical cont control. rol. Mor More e impo important rtantly ly,, dyna dynamic mic meth methods ods can be use used d to imp implem lement ent int intens ensity ity-mo -modu dulat lated ed proton therapy (IMPT) to better conform dose distribut tri bution ionss to the pr proxi oximal mal sur surfa face ce of the PT PTV V as well as the distal surf surface, ace, while avo avoidin iding g norm normalaltiss ti ssue ue st stru ruct ctur ures es.. Se Seve vera rall dy dyna nami micc me meth thod odss of 

comparabl compar able e to tr trea eatm tment ent tim times es for ph photo otons, ns,   e.g., several minutes, it requires either fast ramping of  the th e ma magne gnett cur curren rentt or a fas fastt swi switch tching ing ma magne gnet, t, depending on the time structure of the beam of the accel ac celera erator tor (IC (ICRU RU,, 199 1998). 8). Dis Discre crete te vo voxel xel or spo spott scanning scan ning for char charged-p ged-particl article e ther therapy apy is pre present sently ly in us use e at th the e Pau aull Sc Sche herr rrer er In Inst stit itut ute e (P (PSI SI)) fo forr proton-beam scanning (Pedroni  et al., 1995) and at ´´r Schw the th e Ges Gesell ellsha shaft ft fu Schwerion erionenfo enforschu rschung ng (GSI (GSI)) et al., 19 wit ith h a   12C io ion n be beam am (H (Hab aber erer er   et 1993 93). ). Scan Sc anni ning ng te tech chni niqu ques es ar are e un unde derr de deve velo lopm pmen entt al.., 20 in Up Upps psal ala a (L (Lor orin in   et al 2000 00)) and by maj ajor or proton-therapy proton-t herapy accelerat accelerator or vendors. Cont Co ntinu inuous ous sca scann nning ing,, firs firstt use used d at Up Uppsa psala la in 1960s (Graffm (Graffman an   et al., 1985) and later developed at Berkeley for heavy ions (Chu   et al., 1989), is a flexible technique to yield large homogeneous dose distributions for different shapes. The scanning is done on a rectangular grid with a higher scanning frequ fr equenc ency y in one dir direct ection ion and a lo lowe werr sca scann nning ing freq fr eque uenc ncy y in th the e di dire rect ctio ion n pe perp rpen endi dicu cula larr to it it.. Rectangular fields of different shapes and sizes can be scanned in this way, giving a field shape more closel clo sely y rel relat ated ed to th the e tar target get vol volum ume e pr proje ojecti ction, on, therefore, reducing the beam particle losses. Up to 40  40 cm2 fiel fields ds cou could ld be sca scann nned ed at Be Berke rkele ley y with wi th sc scan anni ning ng ra rate tess of 40 an and d 1 Hz fo forr th the e tw two o

 varying complexity can be used to raster scan or scan sc an a pen enci cill bea eam m th thrroug ugho hou ut th the e PTV as described above. The principle of spot scanning is illustrated schematically in Fig. 3.4. Discre Dis crete te sca scanni nning ng dep deposi osits ts dos dose e in a vox voxel el and switch swi tches es th the e bea beam m off during during th the e cha change nge of par par-ameters for the next voxel (Kanai  et al., 1980). As a threethr ee-dim dimens ension ional al con confor formal mal tr trea eatme tment nt of a 1 l

axes, respectivel respect ively ytailored . The late lateral shape e of thecollimatarget targ et  volume was still byral anshap individual tor (ICRU, 1998). In synchrotrons, pulse-to-pulse energy variation is possible. In cyclotrons, which operate at a fixed energy ene rgy, ene energy rgy var varia iatio tion n re requi quire ress ene energy rgy deg degra ra-dation dat ion by a varia variable-t ble-thickn hickness ess absor absorber ber.. This pro pro-duces a large energy spread in the beam that must

33  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

be limited by magnetic energy selection; this technique uses a magnet to spread the degraded beam energy ene rgy spe spectr ctrum um ac acro ross ss an ene energy rgy-d -defin efining ing sli slit. t. The slit sel select ectss the require required d ene energy rgy ra range nge and a second sec ond mag magnet net re refoc focuse usess the bea beam m int into o the bea beam m trans tr anspor portt sy syst stem. em. Fo Forr the lo lowes westt ene energi rgies, es, thi thiss

have slow pulse-repetition rate (Fig. 3.5b) that can extend treatment times if multiple scanning of dose  voxels (‘repainting’) is required in order to ensure thatt dos tha doses es ar are e del deliv iver ered ed thr throug oughou houtt th the e tar target get  volume with sufficient uniformity (better than + 2.5 percent) or to overcome the effects of intra-

process proce ss is ve very ry ine ineffic fficien ientt an and d a lar large ge amo amoun untt of  beam is lost on the energy-defining slits leading to activation and shielding problems. Scan Sc anni ning ng,, wh whet ethe herr di disc scre rete te or co cont ntin inuo uous us,, requires range control and a controllable and sufficien ci entl tly y in inte tens nse e be beam am ou outp tput ut in or orde derr to de deli live verr wellwe ll-defi defined ned dos dose e thr throug oughou houtt th the e tar target get vo volum lume e safe sa fely ly,, wi with th su suffi ffici cien entt ac accu cura racy cy,, an and d wi with thin in a reasonable time. The beam intensity as a function of time for cyclotrons and synchrotrons is shown in Fig. 3.5. Cyclotrons have a continuous and inherently higher beam output (Fig. 3.5a). Synchrotrons

fraction tumor motion. This is also a problem with scann sca nned ed bea beams ms pr produ oduced ced in cyc cyclot lotron rons, s, alt althou hough gh the th e co cont ntin inuo uous us be beam am ou outp tput ut fr from om a cy cycl clot otro ron n (i.e., higher duty cycle) gives a two or three times high hi gher er do dose se ra rate te.. Fo Forr sy sync nchr hrot otro rons ns,, th the e ra radi dioofrequency-driven extraction (RFDE) system used in the Tsuk sukub uba a fa facil cility ity has imp impro rove ved d the con contr trol ol of  beam extraction,  i.e., the ability to switch the beam on and off multiple times on a 200–250  m s timescale, during a single spill stretched up to 5–7 s in length (Fig. 3.5c and d). It effectively increases the beam duty cycle from 83 to 87 percent in

Figure 3.5. Beam intensity as a function of time for (a) an isochro isochronous nous cyclotron and (b), (c), and (d) for a synchro synchrotron. tron. (a) The cyclotron gives a continuous output stable to within a few percent; the beam can be switched on and off in   50 ms. (b) The synchrotron produces pulsed output; with conventional extraction (magnetic deflection) pulses (spills) may be typically 0.2–1 s long with a repetition rate of 

between    0.5 and 2 Hz. (c) In a synchrotron with RF-driven extraction (RFDE), the beam can be switched on and off in    200 ms, at any time within the spill envelope, producing a large number of short pulses within a single spill. (d) In RFDE, the extraction RF amplitude may be adjusted to control the extracted beam intensity, allowing a more uniform beam intensity during the spill (‘flat-top’). In RFDE, the spill may be stretched to be up to 5–7 s in length.

34  

 BEAM DELIVER DELIVERY Y AND PROPERTIES

synchro hrotron tronss (Hir (Hiramot amoto o and Nish Nishi, i, 1992) 1992).. Ras Raster ter sync scann sca nning ing has th the e fur furth ther er re requi quirem rement ent tha thatt the beam intensity be stable and known. In cyclotrons, fluctuations in beam intensity can be controlled at, or close to, the ion source with a fast time response

technique reduces the sensitivity of the beam uniformit for mity y to the time st struc ructur ture e of the beam pu pulse lse,, especially when beam gating is used for respiratory synchronization. synchron ization. This single-ring wobbling system is widely used in Japanese particle-therapy centers.

 m s) feedb (10–50 ack loop. Recent develop development s in sync sy nchr hrot otro ron nfeedback beam be am ext xtra ract ctio ion n en enab able le ments a mo more re uniform extraction of the beam during a single spill (Fig. 3.5d). Reproducible beam pulses of this type might allow for continuous scanning.

Wob obbl blin ing gthose uses us es beam am-m -mod odif ifyi ying ng scattering; tech te chni niqu ques es similar to used be for passive beam the techniques include fixed-range modulation and may inclu include de pat patientient- and beambeam-spec specific ific modi modifying fying device dev ices. s. As wit with h pas passiv sive e sca scatte tterin ring, g, dos dose e con confor for-mation at the proximal edge of the target volume can be impr improve oved d usin using g a dyna dynamic mic mult multi-lea i-leaff colli colli-mator mat or in comb combinat ination ion with a comp compensa ensating ting bolus and step-wise reduction in the range of the protons (Chu  et al., 1989).

 3.2.1.2.1 Wobbled beams.  A ‘wobbled beam’ is a scanned scann ed bea beam m wh whose ose fin finite ite pen pencil cil bea beams ms ha have ve a large lateral size that is greater than that of a pristine tin e pen pencil cil bea beam m (wh (which ich typ typica ically lly has a Gau Gaussi ssian an lateral profile with a fwhm of 5–10 mm) by a factor of 2 or 3, but nevertheless is small compared with the field width. width. A sca scatte ttere rerr is use used d to ad adjus justt the lateral size of the pencil beam. This technique has

 3.2.1.2.2 Repainting.  The rationale for “repainting,’   i.e., th the e de deliv livery ery du durin ring g sca scann nnin ing g of at le leas astt some so me of th the e pe penc ncil il be beam amss mu mult ltip iple le ti time mes, s, is di diss-

been con been consid sider ered ed onl only y for th the e del deliv ivery ery of un unifo iformrmintensity inten sity dose dis distribu tribution tions. s. Wobble obbled d beams hav have e the advantage over scattered beams in that: (i) it is estimated that protons are used two to three times more efficiently; efficiently; (ii) cons consequen equently tly a low lower er neut neutron ron backg ba ckgro round und is ex expec pected ted to be pr produ oduced ced;; and (ii (iii) i) almost the identical equipment as scanned beams is use used, d, th thus us all allow owing ing for th the e imp implem lement entat ation ion of  fulll sca ful scann nning ing cap capabi abilit lity y. Ho Howe weve verr, suc such h bea beams ms usually require the use of a patient collimator and compensator. In pri princi nciple ple,, wob wobble bled d bea beams ms cou could ld be del deliv ivere ered d using a sequence of static pencil beams. However, in practice, they have been implemented using continuous scanning, for which a variety of scan patterns is poss possible. ible. ‘Wobblin ‘Wobbling’ g’ was developed developed at the Lawrenc Law rence e Berk Berkeley eley Nat National ional Labo Laborat ratory ory for hea heavy vy al.., 19 ion bea beams ms at th the e BE BEV VAL ALA AC (Ch (Chu u   et al 1985 85;; 1989). By wobbling, the particles of a beam pulse are smeared out on rings by the use of a pair of  dipole magnets with fields which vary sinusoidally with time, with a phase difference of 90 8. Several ring ri ngss of di diff ffer eren entt rad adii ii an and d do dose sess ar are e ad adde ded, d, depend dep ending ing on th the e des desire ired d fiel field d siz size, e, to obt obtain ain fla flatt field fie ldss of up to 30 cm di diam amet eter er wi with th   , 5 per percen centt dose variation. variation. This technique technique econo economize mizess on the use us e of pa part rtic icles les,, as do does es do doub uble le sc scat atte teri ring ng,, bu butt

cussed in Section 7.6.3. In the limit of scanning that is either very much faster or very much slower than the mot motion ion und under er con consid sider erati ation on rep repain aintin ting g is not required. For repainting to be most effective in deli vering  verin g a unifo uniform rm dose to a given volume, the volum volume e must be repainted multiple times within the period of the organ motion, or the repainting should occur on a time scale greater than the period of the motion and out of synchrony with that motion. Presently proposed commercial systems use spot scanni sca nning ng or con contin tinuou uouss sca scanni nning, ng, but ha have ve rel relaatively tiv ely slo slow w ene energy rgy-sw -switc itchin hing g tim times. es. In a typ typica icall comm co mmer erci cial al reg egim ime e fo forr sp spot ot sc scan anni ning ng a cu cubi bicc  volume of 1 liter, a proton pencil beam with a Gaussian profile of 10 mm fwhm (s 4.5 mm) may be us used ed to la late tera rall lly y sc scan an an ar area ea of 10  10 cm2 using a 15  15 array of spots at a single depth. If  depth layers are 5 mm apart, a total of 4 500 spots are ar e re requ quir ired ed to sc scan an th the e vo volu lume me on once ce.. In bo both th a synchrotron and a cyclotron, the position of an indi vidual spot can be verified and the specified dose delivered in about 5 ms with between 1 s and 1.5 s required to change energy. Thus, 1.1 s are required to sc scan an a si sing ngle le la laye yerr, an and d th the e to tota tall ti time me fo forr a single scan of all layers is about 45 s. Continuous scanning offers the advantage that there is no dead time betw between een spot spotss assoc associate iated d with spot spot-pos -position ition

strongly depends stable pulse intensities larg la rge e nu numb mber erss ofonpu puls lses es ar are e us used ed forr ea fo each chunless ring ri ng..

adj adjus ustme tment nttimes and rema verifi ve rificat , but while wh iletor, energyene rgyswitching swit ching remain incation aion, significan signi ficant t fac factor , thes these e

¼

This system was in routine use for heavy ion radio therapy at Berkeley from 1985 to 1992 and could equally well be used for protons (ICRU, 1998). The original wobbling technique has been simplified by al..   (1999) Kanai   et al (1999) at the Na Natio tional nal Ins Instit titute ute for Radiological Sciences, Chiba, Japan. The simplified technique uses a single wobbling ring with a strongerr sc ge scat atte tere rerr, ra rath ther er th than an mu mult ltip iple le ri ring ngs; s; th this is

efficie effic ienc ncy y ad adva vant ntag ages es ar are e li limi mite ted d to ab abou outt 25 percen per cent. t. Th The e dis distal tal la layer yerss can be re repai painte nted d mor more e frequently than the proximal layers, since the proximal layers receive dose from protons in the plateau region of the distal pencil-beam Bragg peaks, thus, reducing the number of repaintings required. al..   (2004) Lomax   et al (2004) hav have e docu document mented ed the Pau Paull Scherrer Institute (PSI) experience in proton-beam 35

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

spo spott sca scanni nning ng usi using ng a cyc cyclot lotro ron n wit with h an ene energy rgy-selector system. In this system, lateral scanning is achieved by a combination of magnetic scanning and patient pat ient-cou -couch ch moti motion. on. The PSI pro proton ton penci pencill beam

have so far bee have been n the mai main n tec techn hniqu ique e emp employ loyed. ed. ConeCo ne-bea beam m CT CT,, con conve venti ntiona onall CT CT,, or ult ultra rasou sound nd equipment available in the treatment room is also used. The patient can be repositioned if necessary;

¼

s has Gaussian profile of 85mm 3.5 mm), and a beam spots are spaced mmfwhm apart (orthogonal to the beam direction and 4.5 mm apart in depth. With thiss arr thi arrang angeme ement nt app appro roxim ximat ately ely 10 000 sp spots ots ar are e required to irradiate a volume of 1 liter and approximately 3 000 spots can be delivered per minute. In analyzing the spot weights and energies for practical proton pro ton field fields, s, Loma Lomax x   et al.   (2004) (2004) hav have e sho shown wn tha thatt there is a relatively broad spread of low- and highweigh we ighted ted Br Bragg agg pea peaks ks ov over er all ene energy rgy st step eps, s,   i.e., high-weighted Bragg peaks are not concentrated in a single distal layer. This indicates that there is only a limited relationship between the pencil-beam weighting and depth of penetration, which implies that sys-

tematically tematical ly rep repainti ainting ng usin using g more dis distaltal-lay layer er than proximal-layer repaintings, as suggested above, may not be sufficient to overcome the effects of motion in practice. Thus effective repainting of moving targets in times less than two or three minutes is not feasible at present.  At the Paul Scherrer Institute continuous scanning strategies are under development (Meer  et al., 2006; Pedroni, E., Personal communication, 2007), which use fast scanning (1 cm s21 to 2 cm s21) combined bin ed wit with h bea beam m int intens ensity ity mod modula ulatio tion n at th the e ion source and a newly-developed fast energy-switching system (switching time  90 ms). Such innovations shou sh ould ld al allo low w a 1 li lite terr vo volu lume me to be sc scan anne ned d in approxima appr oximately tely 5 s. Thus Thus,, with the dev developm elopment ent of  more advanced scanning and repainting strategies it should be possible to adequately over-paint large volume moving targets in times of one minute or less. Other Oth er met method hodss for ov overc ercomi oming ng tar target get mot motion ion,, such su ch as mo moti tion on tr trac acki king ng an and d be beam am ga gati ting ng ar are e described in Sections 7.5 and 7.6. 3.2.2 3.2. 2

Patien Pa tientt support support and positi positioning  oning 

The function of the patient support and immobiliz-

a trea tr eatm tmen enttvides couc co uch h chai ch air r wi with th six si x de degr eessthe of  freedom free dom provides pro theormaxi maximum mum flexibilit flexi bility ygree for accurate repositioning of the patient.  An alternativ alternative e method of patient positioning is the use of modern couches that feature easily remo vable, precision-mount precision-mounted ed couch tops. By using multiple couch tops or whole whole-body -body pods, patients patients can be pre-positioned in a positioning suite outside the treatment room. Using a specially designed transporter, the couch top or pod is moved into the treatment room, where it is fixed to the couch base and accu ac cura rate tely ly po posi siti tion oned ed in th the e tr trea eatm tmen entt be beam am (Pedroni  et al., 1995). The purpose of such systems is to increase the efficiency and patient throughput.

3.2.3 3.2. 3

Special Speci al trea treatmen tmentt techniq techniques ues

The first applications of charged particles in radiatio at ion n th ther erap apy y we were re th the e us use e of fin finel ely y co coll llim imat ated ed beams bea ms to tr trea eatt sma small ll we wellll-defi defined ned les lesion ionss to hig high h dose do se in a si sing ngle le or fe few w fr frac acti tion onss us usin ing g th the e te tech ch-niquess of ste nique stereot reotacti acticc surg surgery ery.. Thes These e speci special al tech tech-niqu ni ques es ar are e st stil illl us used ed wit with h pr prot oton on be beam amss fo forr th the e trea tr eatm tmen entt of ma mali lign gnan antt an and d be beni nign gn di dise seas ases es,, including inclu ding uvea uveall melan melanomas, omas, pitu pituitary itary aden adenomas, omas, brain tumors, and arteriovenous malformations.

 3.2.3.1

Eye treatments

Proton beam Proton beamss for ocular tumo tumorr trea treatmen tmentt requ require ire energy of about 70 MeV,  i.e., a penetration in water of 4 cm (Go (Goite itein in   et al., 198 1983b) 3b),, alt althou hough gh a bea beam m with energy as low as 62 MeV has been used. The method met hodss for tr trea eatin ting g ocu ocular lar les lesion ionss wit with h pr proto oton n beams wer were e deve develope loped d at Mass Massach achusett usettss Gener General al Hospit Hos pital al and the Har Harvar vard d Cyc Cyclot lotron ron Lab Labora orator tory y (Goitein and Miller, 1983; Gragoudas   et al., 1977).

ation system is to hold the patient in a stable position during treatment. The patient support can be a couch or a chair. Several institutions use patient couch tops or chairs mounted on robotic arms (Noel  et al., 2003). Immobilization devices are used to ‘fix’ the patient patient to the tr trea eatme tment nt cou couch ch or cha chair ir (se (see e Section 3.2.3). Once On ce th the e im immo mobi bili lize zed d pa pati tien entt ha hass be been en po posi si-tioned using lasers and skin marks, the position of  the th e tr trea eatm tmen entt vo volu lume me re rela lati tive ve to th the e be beam am is checked using a suitable imaging method immediately ate ly prior to tre treatm atment. ent. Orth Orthogona ogonall rad radiogr iographs aphs

 A typical proton-beam irradiation arrangement for the treatment of ocular lesions is shown in Fig. 3.6. Patien Pa tients ts are commonly treated treated in the seated position, and a custom-made facemask and bite block are used for immobilization. The eye is fixated on an ex exter terna nall lig light ht sou source rce that can be adj adjus usted ted to control the direction of the gaze during the treatment. Tantalum rings are usually sutured around the perimeter of the tumor and are used for radiologica log icall set up; oth otherw erwise ise,, th the e set up is ac achie hieved ved using a light field projected through the treatment aperture on to the eye (Munzenrider, 1999). 36

 

 BEAM DELIVER DELIVERY Y AND PROPERTIES

Figure 3.6. A typical proton-beam arrangement arrangement for eye irradiations as was installed at the Hahn-Meit Hahn-Meitner ner Institute, Berlin. The beam enters from the left. After a 100 mm Kapton foil (1), which acts as a vacuum exit window, a computer-controlled variable range shifter (2) and a range modulator (3) are mounted close together. The beam passes a collimator (4), a segmented ionization chamber (5), and two transmission ion chambers (6). Directly in front of the patient is mounted the nozzle (7),  i.e., a pipe that can hold a collimator, an apertur aper ture, e, a comp compensa ensator tor,, or a phan phantom tom for dosim dosimetry etry experimen experiments. ts. An on-l on-line ine x-ray imaging imaging sys system tem (8 and 9) is moun mounted ted from the ceiling ceil ing and can be rem remov oved ed duri during ng tre treatm atment. ent. The pat patient ient sits in a chai chairr with multiple multiple degrees degrees of fre freedom edom (10). (Adapted (Adapted fro from m Paganetti, 1998; reproduced with permission.)

 3.2.3.2 Stereotactic radiosurgery and stereotactic radiotherapy

The techn techniques iques of ster stereota eotactic ctic radi radiosurg osurgery ery (SRS (SRS), ), in which the head is held rigidly in a metal ring attached directly to the patient’s skull and aligned with a radiation source, were first investigated by et al., 19 Leksel Lek selll and coco-wor worke kers rs (L (Lars arsson son   et 1958 58;; Leksell, 1951). Tumors were irradiated with crossfired proton beams using the plateau region of the Bragg Br agg cur curve ve to del delive iverr dos dose e to th the e tar target get volume, volume,

patients to be treated supine while stereotactically alig al igne ned d wi with th a fix fixed ed,, ho hori rizo zont ntal al pr prot oton on be beam am (Chapman  et al., 1993).  Another alignment technique employs an automatic positioning and monitoring system that uses real-time stereophotogrammetry. In this technique, reflective fiducials are placed on the surface of the patient’s molded facemask, which is rigidly fixed to the treatment chair. These fiducials are detected in thre th ree e dim dimens ension ionss by vid video eo cam camera erass and the da data ta

are used to control the motion of the computerized adju ad just stab able le ch chai airr wi with th fiv five e de degr gree eess of fr free eedo dom m (Jones et al., 1995). The intr introduc oduction tion of prot proton-t on-thera herapy py cent centers ers with rotational gantries allows for the supine stereotacticc tr ti trea eatm tmen entt of pa pati tien ents ts fr from om mu mult ltip iple le no nonncoplanar copla nar dir directio ections, ns, using comb combinat inations ions of couch and gantry angles, without the need for specialized patient-alignment apparatus of the type shown in Fig.. 3.7 Fig 3.7.. A mor more e det detail ailed ed des descri cripti ption on of av avail ailabl able e immobilization techniques is given in Section 7.2.2.

rather than using the Bragg peak. Modern Mode rn ster stereota eotactic ctic ther therapy apy uses mult multi-slic i-slice e CT images for three-dimensional planning. Treatments are delivered in a single fraction or sometimes in three or four fractions in SRS, or with conventional fraction fra ctionati ation on in ster stereota eotactic ctic rad radioth iotherap erapy y. For use in mu mult ltif ifra ract ctio ion n tr trea eatm tmen ents ts,, mo more re co conv nven enie ient nt methods for immobilization and repositioning have been developed, developed, one of which uses stai stainles nless-st s-steel eel microsphere fiducials implanted in the skull (Gall  et al., 1993b). The microspheres can be detected on the th e CT used used fo forr tr trea eatm tmen entt pl plan anni ning ng an and d on th the e radi ra diog ogra raph phss us used ed fo forr pa pati tien entt po posi siti tion onin ing. g. Th This is syst sy stem em wa wass us used ed wi with th a so soph phis isti tica cate ted d pa pati tien enttpositioning device, shown in Fig. 3.7, which allows

3.2.4 3.2. 4

Rotati Rot ating ng gant gantries ries

For the grea For greates testt fle flexib xibili ility ty in bea beam m del deliv ivery ery,, th the e pass pa ssiv ive e sc scat atte teri ring ng no nozz zzle le or sc scan anni ning ng sy syst stem em 37

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

3.3 ACCE CCELER LERA ATOR TORS S Proto Pr oton n ac accel celera erator torss for the thera rapy py nee need d to pr produ oduce ce

Figure Figur e 3.7 3.7.. An app appar arat atus us for Ste Stere reoT oTac actic tic Ali Alignm gnment ent for Radiosur Radi osurgery gery (STAR). (STAR). The pat patient ient’s ’s head is held in a fixa fixation tion ring and can be mov moved ed into different different positions positions rel relati ative ve to the al.., 20 fixed fix ed pr prot oton on be beam am (H (Har arsh sh   et al 2002 02;; re repr prod oduc uced ed wi with th permission).

sh shou ould ld be mo moun unte ted d on a ga gant ntry ry th that at al allo lows ws th the e beam bea m to rot rotat ate e thr throug ough h a 3608   around around the trea treattment me nt co couc uch h as sh show own n in Fi Fig. g. 3. 3.8. 8. Th The e si simp mple lest st gan ga ntry req equ uir ire es ma magn gnet etss th tha at can be bend nd the maximum-e maxim um-energy nergy proton beam thr through ough a total of  1808  bend; one 458  bend followed by a 135 8  bend. In addition, some quad addition, quadrupol rupole e focus focusing ing eleme elements nts are requi re quired red bet betwe ween en th these ese mag magnet nets. s. Th The e mag magnet netss weigh up to 100 Mg (  100 tons) and must be supported so that the intersection of the central beam axis and the axis of rotation of the gantry is contained within a 1 or 2 mm diameter sphere as the

beams wit beams with h suf suffici ficient ent ene energy rgy to re reac ach h th the e dis distal tal edges edg es of the deepest deepest tumors tumors and with bea beam m cur cur-rents adequate to achieve treatment times comparable abl e wit with h (or bet better ter tha than) n) the con conven ventio tional nal x-r x-ray ay treatment facilities for the range of field sizes and doses used in radiotherapy (ICRU, 1998). Incident beam energy of 215 MeV is required required to obtain beam penetration to a depth of 30 g cm22 on the patient surfa sur face. ce. Thi Thiss re requi quire ress a bea beam m ene energy rgy eme emergi rging ng from fro m an ac accel celera erator tor of bet betwe ween en 225 and 250 MeV  depe de pend ndin ing g on th the e be beam am-m -mod odif ifyi ying ng el elem emen ents ts required for the treatment. To achieve dose rates of  2 Gy min21 in a vol olu ume of 100 000 0 cm3, a beam current of    8 nA, equivalent to   . 5  1010 protons perr se pe seco cond nd ex extr trac acte ted d fr from om th the e acc ccel eler erat ator or,, is required. Parti Pa rticle cle ac accel celera erator torss ar are e nor normal mally ly bui built lt wit with h either a straight or a circular arrangement of the component comp onents. s. To prod produce uce high high-ener -energy gy prot protons, ons, it is

gantry rota gantry rotates. tes. Fur Further thermore more,, the gant gantry ry need needss to be    5 m in radius to accommodate the scattering or scanning nozzle. This requires a rigid structure of    10 m dia diamet meter er,, 10 m len length gth and wit with h a tot total al weigh we ight, t, inc includ luding ing the mag magnet nets, s, of up to 200 Mg (  200 tons tons). ). Gant Gantry ry desig designs ns hav have e been developed developed to red reduce uce the siz size e re requi quire remen ments ts of the shi shield elded ed room ro om.. Ko Koeh ehle lerr (1 (198 987) 7) pr prop opos osed ed a ‘c ‘cor orks kscr crew ew’’ bending magnet arrangement (Fig. 3.8), which can reduce the overall length of the gantry. At the Paul Scherr Sch errer er Ins Instit titute ute (P (PSI) SI),, a com compa pact ct gan gantry try wa wass created by mounting the treatment couch eccentricall ca lly y on th the e ga gant ntry ry st stru ruct ctur ure, e, th thus us al allo lowi wing ng a reduction redu ction in the ove overall rall diam diameter eter (Pe (Pedron dronii   et al., 1995). 1995 ). Most recently recently,, it has been proposed proposed at PSI that a gantry allowing only 190 8  rotation should be used us ed in co conj njun unct ctio ion n wi with th a 18 180 08   couch rota rotation tion,, thus reducing reducing the room size requ requirem irements ents significant ca ntly ly an and d al allo lowi wing ng ea easy sy ac acce cess ss to th the e pa pati tien entt (Pedroni  et al., 2004).

Figure 3.8. Mode Figure Modell of a cork corkscr screwew-type type gantry, gantry, view viewed ed fro from m the rear (Moyers and Lesyna, 2004; reproduced with permission).

38  

 BEAM DELIVER DELIVERY Y AND PROPERTIES

necessary to accelerate them in stages. Rather than accel ac celera eratin ting g th the e pa parti rticle cless thr throug ough h a sin single gle lar large ge potential, poten tial, the simp simplest lest arra arrangem ngement ent for ach achieving ieving thiss is the lin thi linear ear ac accel celer erat ator or,, in wh which ich a nu numbe mberr of ac accel celera eratin ting g ca cavit vities ies,, dri drive ven n by an RF po powe werr supply, are connected in series along a linear path. In a lin linear ear ac accel celera erator tor,, ea each ch par partic ticle le gen genera erally lly trav tr avers erses es th the e ac accel celer erat ation ion ca cavit vity y onl only y onc once e and and,, conseq con sequen uently tly, lin linear ear ac accel celer erat ators ors for pr produ oducin cing g high hi gh-e -ene nerg rgy y pr prot oton onss ar are e lo long ng.. Th The e pr prob oble lem m of  length len gth is ov over ercom come e in cyc cyclic lic ac accel celera erator torss us using ing a magn ma gnet et to co cons nstr trai ain n th the e pa part rtic icle less to mo move ve in a closed path and traverse a single RF-powered accelerat er atin ing g st stru ruct ctur ure e mu mult ltip iple le ti time mes. s. Th The e or orig igin inal al cyclic accelerator accelerator was the cyclo cyclotron tron developed developed by

protons from 70 to 250 MeV and RF power systems from medical electron linacs, the facility would be better bet ter ada adapt pted ed for hos hospit pital al ins instal talla latio tion n and its price could be reduced (ICRU, 1998).

Lawrence and co-workers incles the 1930s. In a cyc cyclot lotron ron, , the pa parti rticle s early are acc accele elera rated ted in the gap between the pole pieces of a large magnet using usi ng a fixe fixed d mag magnet netic ic fiel field d and a fixe fixed d RF. The partic par ticles les st start art at the cen center ter of th the e mag magnet net wit with h zero ze ro en ener ergy gy an and d sp spir iral al ou outw twar ards ds as th they ey ga gain in energy. They are extracted from the accelerator at the per periph iphery ery of th the e mag magnet netic ic fiel field d by defl deflect ecting ing them from the circular path into a beam-transport system. In a later development of the cyclic accelerator, the synchrotron, particles from a low-energy accelera acce lerator tor are inje injected cted into a fixed fixed-rad -radius ius ring of  magnet mag nets; s; ea each ch par partic ticle le tr trav avers erses es the sam same e pa path th repeate repe atedly dly durin during g the acce accelera leration tion cycle, with an

acce accelera by a high high-vol tageped high-freq highfrequenc y rodes electric lerated fieldted between two-voltage D-shaped D-sha hollow hollo wuency electrod elect es (dee (d ees) s),, wh whic ich h ar are e su supp ppor orte ted d in a va vacu cuum um ta tank nk betwe bet ween en the pol poles es of a lar large ge ele electr ctroma omagne gnet. t. Th The e particles move in the gap between the pole pieces of the magnet. In classical cyclotrons, the magnetic field and the frequency of the accelerating RF field are ar e co cons nsta tant nt.. In pr pra act ctic ice, e, a sm smal alll ga gap p is le left ft between the opposing edges of the electrodes and a source of ions is located in this gap at the center of  the magnet pole pieces as shown in Fig. 3.9. A low velocity ion emerging from the ion source is

increasing magnetic field keeping the particles in a fixed orbit.  At present, the only accelerat accelerators ors being used in

3.3.2 Cyclotro Cyclotrons, ns, isochron isochronous ous cyclotro cyclotrons, ns, and synchro synchrocyclotro cyclotrons ns

In th thes ese e ma mach chin ines es,, th the e acc ccel eler era ati tion on pr proc oces esss depe de pend ndss on a ma magn gnet etic ic re reso sona nanc nce e co cond ndit itio ion, n, in which the reso resonant nant frequency frequency of the RF acceleraacceleration tio n vol voltag tage e is det determ ermine ined d by the mag magnet netic ic fiel field d and the cha charge rge-to -to-m -mass ass ra ratio tio of the par partic ticle. le. As mentio men tioned ned abo above ve,, th the e par partic ticles les in a cyc cyclot lotron ron ar are e

dedicated hosp dedicated hospitalital-based based prot proton-t on-thera herapy py fac faciliti ilities es are either cyclotrons or synchrotrons. 3.3.1 3.3. 1

Linear Linea r acc acceler elerato ators rs

Linear acce Linear accelera lerators tors are typic typically ally char charact acterize erized d by high hig h ene energy rgy con consum sumpti ption on and a ve very ry hig high h bea beam m intensity inten sity,, which could prod produce uce a pote potentia ntiall saf safety ety probl pr oblem. em. Ne Never verth thele eless, ss, a fe few w aut author horss ha have ve pr prooposed this type of accelerator for proton radiotheral.., 19 al.., 19 apy (Bo (Boyd yd   et al 1982 82;; Ha Hamm mm   et al 199 91) 1).. A   versatile proton linear accelera accelerator tor was designed at Los Alamos for pion therapy and was subsequently calculated to be able to produce proton beam cural.., rent re ntss of up to 10 100 0  m A at 650 MeV (Boyd   et al 1982).. Prot 1982) Protons ons of    200 MeV energy energy cou could ld be pr prooduced with an accelerator length of 40 m. Commercial klystrons from radar equipment could be us used ed to pr prod oduc uce e RF po powe werr fo forr acc ccel eler erat atio ion n (ICRU, 1998).  A version of a linac more tailored to the needs of  prot pr oton on th ther erap apy y in te term rmss of en ener ergy gy an and d be beam am current curr ent was prop proposed osed by Hamm   et al.   (1991). (1991). By using sideside-coup coupled led lina linacc secti sections ons for acce accelera lerating ting

Figure 3.9. Schema Schematic tic diagram showing the main components of  a classical cyclotron and the path of the accelerated ions.

39  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

ac accel celera erated ted to towa ward rd the ele electr ctrode ode and ent enters ers the electric field-free space within the hollow electrode, continuing its path on a semi-circular arc. When it reaches the gap between the electrodes, it receives a fu furth rther er ac accel celera eratio tion n bec becaus ause e of th the e sy synch nchron rony y between the RF and the particles’ orbital frequency.  As the particle gains energy crossing the electrode gap, it spirals outward until it reaches a maximum energy ene rgy det determ ermine ined d by th the e ra radiu diuss of th the e mag magnet net pole pieces.  A classical cyclotron can be used to accelerat accelerate e proto pr otons ns to onl only y 10– 15 MeV, wh which ich is mu much ch low lower er than the energ energy y requ required ired for ther therapy apy.. Abo Above ve this limit lim it the re rela lativ tivist istic ic mas masss inc incre rease ase of th the e pr proto oton n

Therapy Center in Munich as dedicated proton-t prot on-thera herapy py faci facilitie lities. s. The magn magnet et is 3.09 m in diameter, 1.65 m high, and weighs 80 Mg (90 tons). The cryogenic magnet’s coil is cooled by four cryocooler coo ler uni units, ts, each wit with h a ra rated ted cooling cooling po powe werr of  1.5 1. 5 W (t (tot otal al 6 W) W).. Th The e co coil ilss ar are e co cont ntai aine ned d in a sealed cryostat cryostat and rema remain in sealed and cold when the th e ma magn gnet et po pole le ca caps ps ar are e ra rais ised ed fo forr se serv rvic ice e or repair rep air.. The use of lo low-m w-main ainten tenanc ance e cry cryo-c o-cool ooler er units considerably simplifies cryostat operation and maintenance. The magnet requires 2.5 W of cooling to maintain its operating temperature; therefore, at any an y ti time me tw two o cr cryo yo-c -coo oole lers rs ca can n be sh shut ut do down wn fo forr maintenance, if necessary.

causes the resonant condition to fail. problem of the re rela lativ tivis istic tic ma mass ss inc incre rease ase can The be ov over ercom come e using an isochronous cyclotron, in which the relati vistic mass increase is compensat compensated ed by increasing the magn magnetic etic field with rad radius, ius, thus main maintaini taining ng the reso resonanc nance e cond condition ition,,   i.e., th ther ere e is a co cons nsta tant nt orbital frequency. When the magnetic field strength increases with radius, axial defocusing of the beam occurs occ urs.. Th This is wou would ld cau cause se th the e bea beam m to st stra ray y fr from om the median median pla plane ne and to st strik rike e the magnet magnet pol pole e pieces pie ces.. Co Comp mpens ensat ation ion for thi thiss de defoc focusi using ng can be achieved by introducing an azimuthal variation in the mag magnet netic ic fiel field d (Th (Thoma omas, s, 193 1938): 8): the mag magnet netic ic field is allowed to increase with radius, and vertical

The capa capabilit y ofallows accurat accu rate e mach machinin ining g intri intricate ly shaped polebility pieces the construction ofcately large isochrono isoch ronous us cyclot cyclotrons rons,, which hav have e disp displaced laced the synchrocyclotron as the accelerator of choice in the energy range from    100 MeV to 1 GeV. In the synchrocyclot chro cyclotron, ron, also called a freq frequenc uency-mod y-modula ulated ted cyclot cyc lotro ron, n, the inc incre rease ase in the rel relat ativi ivist stic ic ma mass ss of 

focusing focusi ng is obt obtain ained ed by inc includ luding ing ra radia diall or spi spira rall hills hi lls an and d va vall lley eyss on th the e po pole le pi piec eces es in or orde derr to crea cr eate te al alte tern rnat ate e hi high gh-- an and d lo loww-fie field ld se sect ctor ors, s, as shown in Fig. 3.10. In this design the dees (within which the charged particles move in circular orbits in perpendicular magnetic fields) may be in three or four sections. sections. On crossing the field boundaries boundaries betw be twee een n th thes ese e ma magn gnet etic ic fie field ld se sect ctor ors, s, th the e io ions ns experience focusing forces that restore them to the median med ian pla plane. ne. Iso Isochr chrono onous us cyc cyclot lotro rons ns ar are e fixe fixeddfield fiel d and fixe fixed-f d-freq requen uency cy ac accel celer erat ators ors;; mod modern ern machining techniques allow the magnetic-field profiles to be engineered with high precision ensuring a high degree of isochronicity with no necessity for magnet trim coils.  An isochronous cyclotron using a high field stre st rengt ngth h in a nar narro row w pol pole e gap has bee been n des design igned ed (Beeckman  et al., 1991), which reduces the magnet weight and energy consumption. A 230 MeV cyclotron built to this design is used at Massachusetts General Hospital for proton therapy (ICRU, 1998). The magnet is 4.34 m in diameter, 2.1 m high, and weighs 180 Mg (200 tons). Total magnet power consumption is 220 kW.  A 250 MeV isochronous cyclotron, which uses a superc sup ercond onduct ucting ing mag magnet net,, has bee been n des design igned ed by Blosser et al.  (1993). Machines based on this design are ar e ins instal talled led at the PS PSII and th the e Rin Rineck ecker er Pr Proto oton n

Figure 3.10. Cut-aw Cut-away ay diagram of a 250 MeV superconducting superconducting cyclotron showing: (1) the magnet yoke, (2) the upper pole piece, (3) the lower pole piece, (4) the spiral hill on the magnet pole piece, (5) the spiral shaped RF dee positioned in the pole piece  valley,, and (6) the magnet coil. The outer diameter of the  valley mag agn net yok oke e is 3.2 m. The cons nstr truc ucti tion on of mode derrn room-tem room -temper peratu ature re isoc isochro hronous nous cycl cyclotr otrons ons foll follows ows a simi similar lar design, but significantly larger (Blosser   et al., 1993; reproduced with permission permission). ).

40  

DELIVERY Y AND PROPERTIES  BEAM DELIVER

Figure 3.11. A schema schematic tic view of the Loma Linda University University Medical Center 250 MeV MeV proto proton-th n-therapy erapy synchrotro synchrotron n show showing ing the major components of the accelerator. The outer diameter of the synchrotron ring is 6.71 m (Coutrakon  et al., 1994; reproduced with permission).

the prot the oton onss wi with th en ener ergy gy is co com mpen ensa sate ted d by a decr de crea ease se in th the e fr freq eque uenc ncy y of th the e acc ccel eler era ati ting ng  voltage,  voltag e, while the magnet magnetic ic field str strength ength remain remainss constant. const ant. This accel accelerat eration ion metho method d requi requires res pulsed operation, since one proton bunch must exit the synchrocyc chr ocyclotr lotrons ons befo before re the acc acceler elerat ation ion of the nex nextt bunch begins begins.. Synchr Synchrocyclot ocyclotrons rons with conve convention ntional al magn ma gnets ets ar are e lar large ge in siz size e an and d ma mass ss.. In pr prin incip ciple, le, they the y can be tun tuned ed for different different energies, energies, but cur cur-rently ren tly ope opera ratin ting g ma machin chines es of thi thiss typ type e ha have ve fixe fixed d energies. The extraction efficiency of the synchrocyclo cl otr tron on is of th the e or orde derr of 90 per erce cent nt an and d th the e

bending magnets, is increased with increasing speed of th the e par arti ticl cle e an and d is cl clo ose sely ly co coup uple led d wi with th an incr in crea ease se in th the e ma magn gnet etic ic fie field ld in th the e be bend ndin ing g magnets. The protons can be extracted at any energy by either single turn extraction or by slow extraction to achieve longer pulses. Both room-temperature and super sup ercon condu ducti cting ng pr proto oton n sy synch nchro rotr trons ons ha have ve bee been n 2 considered as well as H synchrotr synchrotrons. ons. Only roomtemperature designs will be discussed below (ICRU, 1998 19 98). ). Th The e la layo yout ut of th the e co comp mpon onen ents ts of a ty typi pica call synchrotron is shown in Fig. 3.11. Low-energy particles are injected into the accel-

extracted extra cted pr proton oton-be -beam am int intens ensity ity is norm normally ally mor more e than adequate for therapy. The synchrocyclotron is consi con side dere red d a ve very ry re reli liab able le ac accel celer erat ator or,, wi with th no stringe str ingent nt tol tolera erance ncess on the magn magnetic etic field shap shape. e. This assertion is verified by the experience with the Harvard Har vard syn synchr chrocyc ocyclotr lotron on tha thatt wa wass ope opera rated ted for proton therapy with a very high reliability between 1961 and 2002 (ICRU, 1998; Sisterson  et al., 1991).

erator ring. To ach erator achieve ieve acce accelera leration tion,, the magn magnetic etic field and the frequency of the accelerating electric field fiel d mu must st be inc incre rease ased d in sy synch nchron rony y. Bec Becaus ause e of  the finite time required to cycle the magnets, synchrotro chr otrons ns pro produce duce a puls pulsed ed outp output. ut. Typical ypically ly,, the beam acce accelera leration tion cycle take takess    200 ms to 1 s and beam extraction occurs over a similar period. The pulse pu lse re repet petiti ition on ra rate te is the there refor fore e typ typica ically lly 0.5– 2 Hz. The use of numerous small magnets around the th e acc ccel eler era ato torr ri ring ng,, po posi siti tion oned ed so th tha at th the e magnetic-field radial vectors alternate in direction between successive magnets, creates net radial and  vertical focusing in a small magnet aperture furth fur ther er re redu ducin cing g mag magnet net cos cost. t. Th This is pri princi nciple ple of  alternating gradient focusing (Courant  et al., 1952)

3.3.3 3.3. 3

Synchrot Synch rotron rons s

In a synchrotron, bending magnets keep a bunch of  protons in a fixed orbit during the acceleration cycle. The Th e fr freq eque uenc ncy y of th the e hi high gh-v -volt oltage age ac acce celer lerat atio ion n syste sy stem, m, ins instal talled led in str straig aight ht sect sections ions betw between een the 41  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  operating parameters of some accelerators in use in proton-therapy facilities. Table 3.1 Typical operating

Parameter

Cyclotrons

Synchrotrons

MGH

PSI

LLUMC

PMRC

Magnet ring or magnet max. diameter (m) Magnet weight (tons) or number of magnets in ring Energy at extraction (MeV) Beam current (nA) or particles per pulse ( pp ppp) Pulse repetition rate (s)

4.34 165 tons 230 300 nA CW

3.198 90 tons 250 500 nA CW

6.71 8 magnets 70 to 260 3.4  1010 ppp 2.2 s cycle

Extraction system RF cavity frequency (MHz) Field strength (T) Hill  Valley  V alley maximum  Average  A verage power consumption (kW) Ion source or injector type

Electrostatic deflection 70 kV  106.1   2.9   0.9 446 Cold cathode

Lambertson magnet 0.974 to 9.713

RF voltage (kV)

130 peak

Electrostatic deflection 74  3.6  2.0 350 Cold Cathode 100 peak

7.00 – 7.82 6 magnets 70 to 250 7.5  1010 ppp 2 s to 7 s spill length 0.2 s to 0.5 s between spills RFDE 1.5 to 2.0

1.52 350 2 MeV RFQ

1.814 ? 7 MeV linac

0.3

1.3

MGH, Massachusetts General Hospital, Boston, MA, USA; PSI, Paul Scherrer Institute, Villigen, Switzerland; LLUMC, Loma Linda University Medical Center, Loma Linda, CA, USA; PMRC, Proton Medical Research Center, Tsukuba, Japan; CW, continuous wave.

has enabled the construction of very large synchrotron tr onss fo forr hi high gh-e -ene nerg rgy y ph phys ysic icss re rese sear arch ch.. Wh When en applied to the relatively low energy, 250–300 MeV,

Synchrot Sync hrotrons rons with conv conventi entional onal magn magnets ets usin using g H2 ions have been proposed by Martin (1987) and at ITEP in Moscow (Khoroshkov   et al., 1991), and

machines required for proton therapy, the alternating in g gr grad adien ient t pr prin inci cipl ple e re resu sult lts s in re redu duce ced d si size ze,, weight, and total cost of the synchrotron. The first hospital-based proton radiotherapy facility it y wa wass in inst stal alle led d at th the e Lo Loma ma Li Lind nda a Un Univ ivers ersity ity Medical Center and is based on a room-temperature et al., 19 synchr syn chrotr otron on (C (Cole ole   et 1987 87). ). Pu Purp rpos osee-bu buil iltt hospital-base hospit al-based d partic particle-ther le-therapy apy synch synchrotr rotrons ons hav have e also bee een n in inssta tallle led d at Tsu suku kuba ba Univ iver ersi sitty (Fukumoto et al., 1989) and at the Hyogo Ion Beam al.., 200 Medicall Cen Medica Center ter,, Japan (Kaga (Kagawa wa   et al 2002) 2).. Th The e   12 latt la tter er ac accel celer erat ator or ca can n al also so pr prod oduc uce e a C be beam am.. More Mor e rec recent ently ly,, suc such h syn synchr chrotr otrons ons ha have ve also been inst in stal alle led d at th the e Wak akasa asa Wan En Ener ergy gy Re Rese sear arch ch

are consider considered for hosp hospital ital use inargument northern nort hern   et al.ed Italybeing (Amaldi , 1994). The primary for the construction of an H 2 machine is the simplicity of beam extraction extraction,, whic which h is acco accomplis mplished hed by strip st rippin ping g th the e ele electr ctrons ons off the H2 io ions ns in a ve very ry thin thi n foi foill tar target get.. Thi Thiss lea leads ds to a ver very y sma small ll bea beam m divergenc dive rgence, e, perm permittin itting g tra transpo nsport rt of the ext extrac racted ted beam be am wit with h na narr rrow ow-g -gap ap ma magn gnet ets. s. Ho Howe weve verr, H2 mach ma chin ines es ha have ve st stri ring ngen entt re requ quir irem emen ents ts fo forr th the e  vacuum in the accelerat accelerator or,, due to the possibility of  electron stripping by residual gas molecules in the beam pipe. Magnetic electron stripping is another potential problem for the H2 machine, resulting in a maximum allowable field strength of slightly over

Center Cent er,, Tsu suru ruga ga,, Jap Japan an,, Th The e Sh Shizu izuok oka a Ca Cance ncerr Cent Ce nter er,, Mi Mish shim ima, a, Jap Japan an,, an and d th the e M D An Ande derso rson n Cancer Center, Houston, TX, USA. The Th e sy synch nchro rotr tron on is hig highly hly fle flexib xible le in ter terms ms of  energy variation. With a synchrotron, it is feasible to us use e en ener ergy gy va vari riat atio ion n of th the e be beam am in inst stea ead d of  range ran ge shift shifting ing with a varia variable ble thick thickness ness absorber to mo modu dula late te th the e en ener ergy gy fo forr de dept pth h co cont ntro roll of th the e proton beam. A possible limitation for this accelerator is the maximum current per bunch that can be extracted. This is due to space-charge effects that depend dep end on the inj inject ection ion ene energy rgy. Th The e inj inject ector or at Loma Lo ma Li Lind nda a is an RF RF-q -qua uadr drup upol ole e (R (RFQ FQ)) li line near ar accel ac celera erator tor (li (lina nac) c) (Ka (Kapch pchins inskii kii and Tepl eplya yako kov v, 1970) of 2 MeV, whereas at Tsukuba, a 7 MeV RFQ linac is used (ICRU, 1998).

0.5 T, thus requ requiring iring a large diam diameter eter acce accelera lerator tor ring ri ng.. Th The e av aver erag age e be beam am cu curr rren entt co coul uld d al also so be rather low for clinical use (ICRU, 1998). 3.3.4 Typic Typical al acceler accelerator ator operat operating  ing  parameters

 At prese present, nt, isochr isochronous onous cyclotr cyclotrons ons or slowslow-cycling cycling synchr syn chrotr otrons ons are com common monly ly used for pr proto oton-b n-beam eam ther th erap apy y. Bo Both th of th these ese ty type pess of ac acce celer lerat ator orss can pro pr ovi vid de pr prot oton on be bea ams th tha at ar are e well su suit ited ed to applica app licatio tions ns in the clin clinical ical env enviro ironm nment ent,, and are comme com merc rcial ially ly av avai ailab lable le.. Th The e ty typic pical al op oper erat atin ing g para pa rame mete ters rs of som some e is isoch ochro rono nous us cyc cyclot lotro rons ns an and d slow-cycling synchrotrons for use in proton-therapy facilities are listed in Table 3.1. For a proton-therapy 42

 

 BEAM DELIVER DELIVERY Y AND PROPERTIES

facility capable of treating deep-seated tumors, energies in the range of 225– 250 MeV are are req requir uired. ed. If  ranges in excess of 37 cm are required due to highly oblique obli que beam ent entries ries,, or if pr proto oton n rad radiogr iograph aphy y is planned, energies of    300 MeV are needed. All the machi ma chine ness de desc scrib ribed ed in Tab able le 3. 3.1 1 op oper erat ate e at we well ll below 300 MeV.

3.4 THE PROP PROPERT ERTIES IES OF PRO PROTON TON BEA BEAMS MS 3.4.1 Pro 3.4.1 Proton ton inter interact actions ions with with matter matter Prot Pr oton on in inte tera ract ctio ions ns wi with th ma matt tter er ha have ve be been en described in detail in ICRU (1998) and only some

proportional to the square of its velocity. Thus, as a part pa rtic icle le tr trav avers erses es ma matt tter er,, its ra rate te of en ener ergy gy los losss increases until, at very low velocities, it captures an electron, decreasing its effective charge. The average rate ra te of ene energy rgy loss consequen consequently tly dec decrea reases ses at ver very y low velocities. The foregoing explains the formation of the Bragg dose peak (ICRU, 1998).  At increasing energy energy,, nuclear interactio interactions ns (Chadwick   et al., 1999; ICRU, 2000) become more importan impo rtantt (Lai (Laitano tano   et al., 19 1996 96;; Me Medi din, n, 19 1997 97;; Medin Medi n and And Andreo, reo, 1997a; Pag Paganett anetti, i, 2002 2002;; Wroe  et al., 2005). In the therapeutic energy range, the proba pr obabil bility ity of nu nucle clear ar ev event entss is sma small ll com compar pared ed with wi th th the e pr prob obab abil ilit ity y of el elec ectr tron on in inte tera ract ctio ions ns,, although each nuclear reaction can transfer a sig-

of the main features will be repeated here. Protons lose their energy in a medium primarily throug thr ough h num numer erous ous ele electr ctroma omagne gnetic tic int intera eracti ctions ons with atomic elect electrons rons (ICRU, 1993a; 1998 1998). ). The They y have a mass that is large compared with the mass of electrons, hence they lose only a small fraction of  their energy in each interaction (at most 4 m /  M  0.0022, where   m  is the electron mass and   M  is   is the proton mass) and they are deflected by only small angles in each interaction.

nifi ifica can nt por orttio ion n of the pr prot oton on en ener ergy gy to the medium. medi um. Nuc Nuclear lear inte interact ractions ions essen essentially tially rem remove ove primary protons from the beam and result in the product pro duction ion of secon secondary dary part particles icles.. Thes These e part particles icles may be important from the biological point of view because of their higher RBE values (ICRU, 1998).  Among the nuclear processes that occur occur,, neutron production has a non-negligible impact due to the produ pr oducti ction on of hea heavy vy cha charge rged d pa parti rticle cless gen gener erat ated ed by subsequent neutron interactions. In addition to

The proton electronic  S ( E)/ mass r,  is defined r  material, as stopping power in a

th the e po pote tent ntia ialprotons l bi biol olog ogic ical al ef effe fect ct du due e tothenu nucl clea earr secondaries, slowing down near Bragg peak may also be expected to produce an enhanced biolog bio logica icall eff effect ect (IC (ICRU RU,, 199 1998; 8; Pa Pagan ganett etti, i, 200 2002; 2; Paganetti and Goitein, 2000) (see Section 2).

¼

 Sð EÞ r 

¼

     1

r 

  d E d x

;

 

ð3 1Þ :

where  S  is the linear stopping power (note that the r  is used interchangeably), d E nomenclature   s  S / r  is the mean energy lost by a proton in electronic collisions while traversing a distance d x   in a material of density   r . Frequently, the unit MeV cm2 g21 is used, where the mass thickness, or areal density, (in g cm22) is defined as the product of density   r  and an d ab abso sorb rber er th thic ickn knes esss [r d x   in Eq Eq.. (3 (3.1 .1)] )].. Th The e ¼

3.4.2 Defin 3.4.2 Definitio ition n and specifi specificat cation ion of of beam beam properties and beam parameters

In practice, a number of definitions of the proton beam characterist characteristics ics are requ required ired.. Some of thes these e relate to the beam models used in treatment planning nin g and oth others ers rel relat ate e to the ph physi ysical cal cha chara racte cterr-

quantity   S / r    ( s) de depe pend ndss on th the e ph phys ysic ical al an and d chemical chemi cal compo composition sition and dens density ity of the material material and on the energy of the proton. In the conti continuou nuous-slo s-slowingwing-dow down n appr approxim oximatio ation n (csda), the range  R  of a particle is given by ¼

 R ¼

ð 

½ S  Sð EÞ1 d E

:

 

istics of the beam as determined by measurement. To dr draw aw a dis distin tincti ction on bet betwe ween en the these se set setss of defi defi-nitions, the following section discusses ‘Beam properties’ that relate to the treatment-planning system and ‘B ‘Bea eam m par ara ame metter ers’ s’ that ar are e defi efin ned fo forr measurementt purposes. measuremen From Fr om th the e per perspe specti ctive ve of tr trea eatm tment ent pla plann nning ing,, a ‘beam’’ cons ‘beam consists ists of dire directed cted rad radiati iation on char characte acterized rized by a number of physical factors that identify: the type and quality of the radiation source; the direction ti on fr from om wh whic ich h ra radi diat atio ion n co come mess an and d th the e po poin intt within the patient to which it is directed; and the ‘sha ‘s hape pe’’ (l (lat ater eral al to th the e be beam am an and, d, in th the e ca case se of  protons, in depth) of the high-dose region (treated  volume). At any point, the directional properties of  a beam may be characterized by its angular divergence and angular emittance (Section 3.4.2.1).

ð3 2Þ :

Because each particle experiences a different set of  interactions, a group of particles of the same initial energy has a distribution of energies after traversing a thi thickn ckness ess of abso absorber rber (en (energ ergy y st strag raggli gling) ng) and a resulting distribution of depths at which the particles stop sto p (ra (range nge str stragg aggling ling)) (IC (ICRU RU,, 199 1998). 8). The ra rate te of  energy loss of a charged particle is proportional to the square of its charge and approximately inversely 43  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

The ph The phys ysic ical al ch char arac acte teri rist stic icss of th the e cl clin inic ical ally ly acceptab acce ptable le trea treatmen tmentt beam beamss prod produced uced by passi passive ve scattering or dynamic methods must be defined in term te rmss of a se sett of be beam am pa para rame mete ters rs.. Th The e be beam am characteristics should be defined in terms of a set of mea measur surabl able e and cal calcul culabl able e par parame ameter ters, s, whi which ch can be use used d to ass assess ess the cli clinic nical al ac accep ceptab tabili ility ty of  the beam beams, s, to benc benchmark hmark the tre treatm atmentent-plann planning ing algorithms used to calculate patient isodose distributions and for routine quality-assurance purposes. These The se pa param ramete eters rs dif differ fer for bea beams ms pr produ oduced ced by

given poi given point nt (on (on,, sa say y, the pa patie tient’ nt’ss ski skin n sur surfa face) ce).. Owing to mult multiple iple scat scatterin tering g of charg charged ed part particles icles within withi n mat matter ter,, all pro proton ton beam beamss acqu acquire ire angu angular lar ‘confusion’ at depth within the patient and this is a major contribution to the beam penumbra.  Infinitesimal pencil beam. An ‘i ‘infi nfini nite tesi sima mall pencil beam’ is a beam which, at the place that it is incide inc ident nt on th the e pa patie tient, nt, has infi infinit nitesi esimal mal la later teral al extent ext ent,, infi infinit nitesi esimal mal ang angula ularr emi emitta ttance nce,, and is monoenergetic.  Finite pencil beam.  A ‘finite pencil beam’ is analo-

passive scattering or by dynamic methods, and are discussed in Section 3.4.2.2.  3.4.2.1

gous to an infinitesimal pencil beam except that, at the place that it is incident on the patient, it has: finite finit e (but (but,, nev neverth ertheless eless,, smal small) l) lat latera erall ext extensi ension, on, finite (but, nevertheless, small) angular emittance, and finite (but, nevertheless, small) energy spread.  Pencil beam.   The ge gene nera rall te term rm ‘p ‘pen enci cill be beam am’’ may ma y be em empl ploy oyed ed wh when en no di dist stin inct ction ion be betw twee een n infinitesi infin itesimal mal and finite initi initial al cond condition itionss need needss to be made.

Beam properties

 Field. The ‘field’ of a beam is the area of intersection of the beam with a plane normal to the beam direct dir ection ion and at any dis distan tance ce fr from om the eff effect ectiv ive e source. Thus, it is a two-dimensional structure and it is, for ex examp ample, le, pos possib sible le to re refer fer to the fluence fluence distribution within a field (often termed the beam’s intensity profile). Uniform Uni form-int -intensi ensity ty bea beam. m.   A ‘unif ‘uniform-i orm-inten ntensity sity

 3.4.2.2

Beam parameters

be beam am’ ’ iswith used us ednearto uniformit irra ir radi diat ate e ya volu lume me with wi thin in sity the th e patient pat ient near-unif ormity . Avo uniform-i unif orm-inten ntensity beam can be produced by either passive or dynamic beam-delivery techniques.  Intensity-modulated beam. Inten Intensitysity-modu modulate lated d thera the rapy py is a tec techn hniqu ique e of tre treat atmen mentt in whi which ch the indivi ind ividua duall bea beams ms do not irr irradi adiat ate e the PT PTV V un uniiformly—indeed are likely to do so in a highly nonuniform manner. The intent of intensity-modulated therapy is that the set of beams that constitute the plan generate a conformal and usually uniformlytreated volume within a PTV that has concavities, and/or and/ or ach achieve ieve ‘conf ‘conformal ormal avo avoidan idance’ ce’ of part particula icularr organs org ans.. Th The e bea beams ms ar are, e, the then, n, ter termed med ‘in ‘inten tensit sityy-

 Passively scattered beams.   For For passively scattered beams, the parameter set should include the profile i.e. of the beam in depth ( , the profile of the SOBP), the th e la late tera rall pr profi ofile le of th the e be beam am,, an and d da data ta on th the e beam penet penetra ration tion and width width,, flat flatness ness,, sym symmetry metry, lateral penumbra and distal dose fall-off. The variatio at ion n in th thes ese e pa para rame mete ters rs wi with th be beam am en ener ergy gy,, length len gth and dep depth th of th the e SOB SOBP P, and the fiel field d siz size e should be documented. Gall  et al.  (1993a) defined a set of beam parameters that has been widely used and is ill illus ustr trat ated ed in Fig Fig.. 3.1 3.12. 2. Mo More re re recen cently tly,, an alternative set of beam parameters for characterizing in g th the e SO SOBP BP (F (Fig ig.. 3. 3.13 13)) ha hass be been en pr prop opos osed ed (Gottschalk, 2003).

modulated beams’ (to contrast them with uniform-intensity beams). In proton-beam therapy, an intensity-modulated beam is generally produced using some method of beam scanning.  Angular divergence and angular emittance. There are two distinct aspects to the angular distribut utio ion n of th the e pa part rtic icle less in a be beam am.. The firs rst, t, ‘angular ‘angu lar dive divergenc rgence’, e’, desc describes ribes the corre correlat lation ion of  the th e me mean an di dire rect ctio ion n of th the e pa part rtic icle less in th the e be beam am with wit h the their ir pos positi ition on wit within hin the fiel field. d. Th This is eff effect ect is typically due to the spreading out of a beam that emanates from a source, which is a finite distance from the patient. It gives rise, for example, to the invers inv erse-s e-squa quare re eff effect ect in wh which ich dos dose e fal falls ls off for purely geometric reasons as the square of the distanc ta nce e fr from om th the e so sour urce ce.. On th the e ot othe herr ha hand nd,, th the e ‘angul ‘an gular ar emi emitt ttanc ance’ e’ des descri cribes bes the dis distri tribut bution ion of  particle directions around their mean direction at a

al.. The Th e pa param ramete eterr defi definit nition ionss giv given en by Gal Galll   et al were as follows:  Depth of penetration   (d090), or   range   in Gal Gall’s l’s 22 nomenc nom encla latur ture, e, is defi defined ned as the depth depth (in g cm ) along the beam central axis in water to the distal 90 percent point of the maximum dose value.  Distal-dose fall off   (DDF) (DDF) is defi defined ned as the distanc ta nce e (in g cm22) in which the dose, measured in water along the beam central axis, decreases from 80 to 20 percent of the maximum dose value.  SOBP length   (m090) is defined as the distance in water between the distal and proximal 90 percent points of the maximum dose value.  Lateral penumbra   (LP) (LP) is de defin fined ed at a gi give ven n depth as the distance (in mm) in which the dose, measured along the line perpendicular to the beam central axis, decreases from 80 to 20 percent of the maximum dose value at that depth. 44

 

DELIVERY Y AND PROPERTIES  BEAM DELIVER

Figure 3.12. (a) Depth– dose curve for an SOBP and (b) a lat Figure latera erall prot proton-b on-beam eam profile profile in the Bragg peak show showing ing some para paramete meters rs commonly used to characterize the proton dose distribution as defined by Gall  et al.  (1993).

 Field size   is de defin fined ed as th the e di dist stan ance ce (i (in n mm mm)) between the 50 percent points of the maximum dose  value, measured along the line perpendicular to the beam central axis, on the isocenter plane in air.

Target (or treatment) length  is defined as the distance bet tance betwe ween en two DDF len length gthss (2  DDF) prox proxiimall to th ma the e di dist stal al 50 pe perc rcen entt is isod odos ose e le leve vell of th the e SOBP, and one DDF length (1  DDF) distal to the

45  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

inter ersec secti tions ons of th the e bes bestt-fit fit po polyn lynomi omials als pr prov ovid ide e a int robu ro bust st de defin finit ition ion of tw two o po point intss an and d bas based ed on th these ese points the depth at full dose (d100) and the modulation

at full dose (m100) are defined. The procedure can be used us ed to defi efine ne th the e 10 100 0 pe perrce cent nt dos ose e le lev vel el,, th the e range-equivalen rangeequivalentt depth ( d80), the gradient of the DFF, the slope of the SOBP ‘flat top’, the entrance dose and the goodness of the fit (rms). Gottschalk argues that the param parameters eters   d100   and   m100   are more conven convenient ient definitions definiti ons of pene penetra tration tion and ext extent ent of mod modula ulation tion 0 than the commonly used definitions of   d090   and   m90 (see Figs 3.12 and 3.13). He argues that d090 is an arbitrary choice and if the intention is to take the target  volume  vol ume to full full dose then then d100 is more appropriate. The term ‘range’ has often been used loosely to defin de fine e be beam am pe pene netr trat atio ion. n. Th The e us use e of th this is te term rm should be restricted to describe calculated or tabulated csda particle ranges as derived from Eq. (3.2) above and as discussed in ICRU Report 49 (ICRU, 1993a). The depth of penetration may be defined in sever sev eral al dif differ ferent ent wa ways ys (d80,   d 90, or   d100) as dis is-cussed cus sed abo above ve and as ill illus ustr trat ated ed in Fig Figss 3.1 3.12 2 and 3.13. It is important, in reporting the depth of penetra et rati tion on of a pr prot oton on be beam am,, to ca care refu full lly y sp spec ecif ify y which of these definitions is being used. It is rec recomm ommend ended ed th that at the dos dose-d e-dist istrib ributi ution on parameterization of Gall   et al.   (1993a) be used to characterize lateral-dose distributions, and that the paramete para meteriza rization tion of Gott Gottschal schalk k (200 (2003) 3) be used to characterize depth–dose distributions.

Figure 3.13 Figure 3.13.. Model SOBP showin showing g the quan quantiti tities es defin defined ed by Gottschalk,   d100   and   m100, and the commonl commonly y used quantities quantities d090,  m 090, and  d 80.

proxim prox imal al 90 pe perc rcen entt is isod odos ose e le leve vell of th the e SO SOBP BP (Fig. 3.12a). Target (or treatment) width  is defined as the distance between LP  LP) from 50 percen per cent t iso isodos dose etwo levels lev els widths of the (2 later la teralal-bea beam m the profil pr ofile e (Fig. 3.12b). The following additional parameters can be used to define the dose uniformity: lateral symmetry and lateral flatness.  Lateral flatness  (in percent) is defined as



  dlpmax  dlpmin  F lp lp  ¼ dlpmax þ dlpmin



 100

 

;

 Dynamically scanned beams.   For scan scanned ned beam beams, s, the chara characterizati cterization on paramete parameters rs defined above are not appropriate. Scattered-beam data are generally collec col lecte ted d in an ex exten tended ded wa water ter ph phan antom tom usi using ng a small-vo small -volume lume ionization ionization cham chamber ber or diod diode. e. Wi With th a spot-scanned beam, the integrated-beam dose distribution is produced by superimposing a large number of in indi divi vidu dual al pe penc ncil il be beam ams. s. Mo More re im impo port rtan antl tly y, trea tr eatme tment nt-pl -plan annin ning g sy syst stems ems for sca scann nned ed pr proto oton n beams perform their calculations by summing indi vidual pencil beams. Thus, beam-scannin beam-scanning g sys systems tems require the measurement of depth–dose curves and later la teral al pr profil ofiles es of fini finite te pen pencil cil bea beams ms,, re requi quirin ring g small-field dosimetry equipment with a good spatial resolution (Pedroni  et al. , 2005). If th the e dos dose e dis distri tribu butio tion n in a th three ree-d -dime imensi nsiona onall  volume resulting from beam scanning is to be measured, then some multi-channel or integrating dete de tect ctor or is re requ quir ired ed.. Th Thes ese e me meas asur urem emen ents ts ar are e discussed in more detail in Section 4.

ð3 3Þ :

where   dlp max   and   dlp min   are are th the e max maximu imum m and minimum absorbed dose values in the beam profile measured in the target width.  Lateral symmetry  (in percent) is defined as



  D1  D2  Slp  ¼  D1 þ D2



 100

;

 

ð3 4Þ :

where  D 1  and  D 2  are the integrated absorbed doses in each half of the field about the central axis. In th the e SO SOB BP par aram amet eter eriz iza ati tion on pr prop opos osed ed by Gottschalk (see Fig. 3.13), the terms depth and modulation are used in preference to depth of penetration and an d ta targe rgett (or tr trea eatm tmen ent) t) len lengt gth h de defin fined ed abo above ve.. Gott Go ttsch schalk alk us uses es a mo mode del-i l-ind ndep epen ende dent nt an analy alysis sis of  SOB OBP P data to det ete erm rmin ine e bea eam m par ara ameters (Gottschalk (Gottsc halk,, 2003 2003). ). The beam is div divided ided into thr three ee separate segments along the depth axis, namely proximal rise, flat top, and distal drop, and each segment is fitted with a separate polynomial. The break points betw be twee een n se segm gmen ents ts ar are e pa para rame mete ters rs of th the e fit in addition to three sets of polynomial coefficients. The

3.5 RADI RADIA ATION QUAL QUALITY  ITY  The radiation quality of the beam as it enters the patient is determined by the energy distribution of  46

 

DELIVERY Y AND PROPERTIES  BEAM DELIVER

the scattered beam, or of the pencil beam in scanning applications, and the degree of scattering that the beam has experienced in transit through beamline li ne co comp mpon onen ents ts an and d be beam am-m -mod odif ifyi ying ng de devi vice ces. s. Radiation quality can affect the distal fall-off of the Bragg Bra gg peak peak,, the late lateralral-beam beam penu penumbra mbra,, and the low-le low -leve vell dos dose e to th the e pa patie tient nt out outsid side e the field at large lar ge off off-ax -axis is dis distan tances ces,, inc includ luding ing the sec second ondary ary

interaction interac tionss with material material in the trea treatmen tmentt head or nozzle (scatterer, range shifter, range modulator, collimator aperture, and beam-shaping bolus), and to a ver ery y mu much ch le less sser er ex exte tent nt fr from om th thos ose e fr from om proton interactions with nuclei of patient tissue in the th e be beam am pa path th.. In dy dyna nami micc be beam am de deli live very ry,, th the e neutron dose is produced almost entirely by proton intera int eracti ctions ons wit with h nuc nuclei lei of pa patie tient nt tis tissu sue e in th the e

neutron and photon doses. The mode of beam productio duc tion n and de deliv livery ery can con consid sidera erably bly af affec fectt th the e radiati radi ation on quali quality ty.. Secon Secondary dary radi radiati ation on can also be produced in the patient.  An experiment experimental al study of the factors affecting proton-b prot on-beam eam penu penumbra mbra has been made by Oozee Oozeerr  et al.  (1997) for a passively scattered proton beam. They found that for the system under investigation the beam pen penumb umbra ra as it ent enters ers a ph phant antom om (t (the he ‘entrance’ penumbra) is little affected by the range shifter but is influenced by the final collimator, the bolus thickness, and the air gap between the bolus and the entrance to the phantom. At a given depth in the phantom, the LP depends on the ‘entrance’ penu pe numb mbra ra,, th the e ra rang nge, e, an and d th the e de dept pth. h. As de dept pth h incr in crea ease ses, s, th the e fr fra act ctio iona nall co cont ntri ribu buti tion on of th the e ‘ent ‘e ntra ranc nce’ e’ pe penu numb mbra ra to th the e to tota tall pe penu numb mbra ra decreases. Paganetti and co-workers used Monte Carlo calculat cul ation ionss to inv inves estig tigat ate e the infl influen uence ce of var variou iouss beambea m-mod modify ifying ing dev device icess on the bea beam m cha chara racte cterristics for an ocular melanoma treatment beam line, but these methods methods ar are e als also o we well ll sui suited ted to ap appli pli-catio ca tions ns for the tr trea eatme tment nt of dee deepp-sea seated ted tu tumor morss (Paganetti, 1998; Paganetti and Goitein, 2000). The effect eff ectss of dif differ ferent ent bea beamm-lin line e arr arrang angeme ements nts on beam penumbra, the depth–dose distributions and

beam path. Hence, the secondary neutron scattering in the treatment head can theoretically lead to an incr increased eased risk of rad radiati iation-in on-induce duced d secon secondary dary cancerr for scat cance scattere tered d prot proton on beam beamss comp compared ared with scanned proton beams. However, it must be noted that th at:: (i) th there ere is lim limite ited d cli clinic nical al ex exper perien ience ce wit with h scanned beams, which have only been used at one cent ce nter er;; (i (ii) i) th ther ere e is no pu publ blis ishe hed d ev evid iden ence ce of  increased risk of secondary cancers in patients who have undergone proton therapy; and (iii) the available abl e da data ta on dos doses es out outsid side e the treatm treatment ent fiel field d in prot pr oton on th ther erap apy y ha have ve la larg rge e un unce cert rtai aint ntie iess an and d (where comparable) are not in good agreement. The ene energy rgy dis distri tribut bution ion of the sec second ondary ary neu neu-trons tr ons con contri tribut buting ing sig signifi nifican cantly tly to dos dose e is in the range from zero to the maximum initial energy of  the pro proton ton beam. The rad radiati iation-w on-weigh eighting ting fac factors, tors, or qu qual alit ity y fa fact ctor ors, s, of ne neut utro rons ns,, as fu func ncti tion on of  neutron energy, recommended by the International Comm Co mmiss ission ion on Rad Radiol iologi ogical cal Pr Prote otecti ction on (IC (ICRP RP, 1991) are: 5, 10, 20, 10, and 5 for energies of   , 10, 10–100 ke keV V, 10 100 0 ke keV V to 2 Me MeV V, 2– 20 Me MeV V, an and d . 20 MeV, respectively. It is suggested that 10 is a reasonabl reas onable e quali quality ty fact factor or to be used for scat scattere tered d radi ra diat atio ion n fr from om 15 150 0 to 25 250 0 Me MeV V pr prot oton on be beam ams. s. These quality factors are used only for calculating the th e ris risk k of sec second ondary ary can cancer cerss and should should not be

the distal of the Braggtopeak can bevariation studied. There are fall-off two contributions the slight in relative biological effectiveness (RBE) within the irra ir radi diat ated ed vo volu lume me:: th the e fir first st is th that at th the e li line near ar ener en ergy gy tr tran ansf sfer er (L (LET ET)) of th the e pr prim imar ary y pr prot oton onss increases as they slow down and only becomes significant at the end of their range; the second contributi bu tion on is fr from om se seco cond ndar ary y pa part rtic icle less (c (cha harg rged ed particles and neutrons), particularly those originating fr from om non non-el -elas astic tic nu nucle clear ar int intera eracti ctions ons,, whi which ch can have high LET and can affect all parts of the Bragg curve. These small RBE variations are consist si sten entt wi with th a ge gene neri ricc RB RBE E va valu lue e of 1. 1.1 1 an and d ar are e usually ignored.

used for calculations. Ther Th ere e therapeutic are ar e li limi mite ted ddose data da ta on se seco cond ndar ary y do dose sess to proton therapy patients outside the treatment field. al..   (2002) Binn Bi nnss an and d Ho Houg ugh h (1 (199 997) 7) an and d Yan   et al made mad e mea measur sureme ements nts in 200 and 160 MeV bea beams, ms, respe re specti ctivel vely y. Bin Binns ns and Hou Hough gh mea measur sured ed a dos dose e equiva equ ivalen lentt of 33– 80 mS mSv v per treatme treatment nt Gy at dis dis-tances of 50–120 cm off-axis of a prototype beam designed for treating small intracranial lesions and skul sk ulll-ba base se tu tumo mors rs at th the e Na Nati tion onal al Ac Acce cele lera rato torr Centre (now iThemba LABS). To reduce this backgroun gr ound d add additi itiona onall shi shield elding ing wa wass ins instal talled led in th the e beam line prior to comm commission issioning ing for clini clinical cal use.  Yan  Y an   et al.  measured dose equivalents in the range

In add additi ition on to af affec fectin ting g the dos dose e dis distri tribut bution ion within with in th the e ta targ rget et vo volu lume me,, se seco cond ndar ary y pa part rtic icle les, s, especially espec ially neut neutrons rons from nuclear inter interacti actions, ons, can deliver dose outside the target volume, both laterally and beyond the distal edge of the Bragg peak. In pass passive ive beam deliv delivery ery,, the secon secondary dary neut neutron ron dose dos e res result ultss fro from m neu neutr trons ons pr produ oduced ced by pr proto oton n

of     1– 15 mS mSv v pe perr tr trea eatm tmen entt Gy ar arou ound nd th the e passive scattering nozzle at the Harvard Cyclotron Laborat Labo ratory ory at a dis distance tance of 50 cm. Rec Recently ently, Wroe  et al.   (2007 (2007)) det determ ermine ined d the dos dose e equ equiva ivalen lents ts outside outs ide a 225 MeV prot proton on ther therapy apy pro prosta state te trea treattment field at the Loma Linda University Medical Cente Ce nterr. Th The e dos dose e equ equiva ivalen lents ts on th the e sur surfa face ce of a 47

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

phantom decreased from 3.9 to 0.18 mSv per treatment Gy as the distance from the proton field edge incr in crea ease sed d fr from om 2. 2.5 5 to 60 cm cm.. Da Data ta fr from om Mo Mont nte e Carlo calculations are also available (Agosteo  et al., 1998; Fontenot   et al., 2005; Jiang   et al., 2005; Polf  and Newhauser, 2005; Polf   et al., 2005). The data al..   show from fr om Po Polf  lf   et al show goo good d agr agreem eement ent wit with h the

equivalent during proton therapy of the mother in the th e pa pass ssiv ivel ely y sc scat atte tere red d 20 205 5 Me MeV V be beam am at th the e Midwe Mid west st Pr Proto oton n Ra Radio diothe thera rapy py Ins Insti titu tute. te. Th They ey measured fetal dose equivalents of between (0.10 and 0. 0.26) 26) an and d (0. (0.025 025 an and d 0.4 0.450) 50) mSv pe perr tr trea eattment Gy, respectively. The excess risk to the fetus of rad radiat iationion-indu induced ced canc cancer er dea death th in the firs firstt 10

measured data of Yan  et al.  (2002), with the Monte Carlo Ca rlo cal calcul culat ation ionss of the dos dose e equ equiva ivalen lentt bei being ng within a factor of 2–3 of the measurements. Jiang  et al.   (2005) (2005) cal calcul culat ated ed org organan-spe specifi cificc pa patie tientnteffect eff ective ive dose due to seco secondar ndary y neu neutro trons ns for two simulated treatment plans in the lung and paranasall sinu nasa sinus. s. The tota totall who whole-bo le-body dy effe effectiv ctive e dose wass cal wa calcul culat ated ed to be 0.1 0.162 62 and 0.0 0.0266 266 Sv Sv,, wit with h 0.058 0.05 8 and 0.00 0.0042 42 Sv Sv cont contribu ribution tionss fro from m neu neutro trons ns prod pr oduce uced d int intern ernall ally y in th the e pa pati tient ent,, for the lun lung g and pa para rana nasal sal sin sinus us tr trea eatm tmen ents, ts, re resp spect ectiv ively ely.. The work of Fon Fonteno tenott   et al.   (2005) (2005) show showed ed tha thatt Mont Mo nte e Ca Carl rlo o si simu mula lati tion onss gi giv ve va valu lues es of th the e dosedo se-equ equiva ivalen lentt 2.6 tim times es th those ose fr from om mea measur sureements. ment s. Roy and San Sandiso dison n (200 (2004) 4) and Meso Mesolara larass  et al.  (2006) focused on determining the fetal dose

years of life was estimated to be 17.4 percent per 10 000 children. Thes Th ese e da data ta sh show ow th that at th the e sc scat atte tere red d ne neut utro ron n dose do se fo forr a pa pass ssiv ivel ely y sc scat atte tere red d be beam am is hi high ghly ly depe de pend nden entt on th the e en ener ergy gy an and d on th the e sp spec ecifi ificc desi de sign gn fe fea atu ture ress of th the e be beam am li line ne an and d ma mach chin ine e head. Monte Carlo simulations differ substantially from meas measurem urements ents,, which hav have e rela relative tively ly large uncert unc ertain aintie ties. s. The mea measur sureme ement ntss of Sch Schnei neider der  et al.   (2002b) (2002b) show that with scan scanned ned beam beamss the neutro neu tron n dos dose e equ equiva ivalen lentt out outsid side e the tr trea eatme tment nt field fie ld is le less ss th than an fo forr mo most st pa pass ssiv ivel ely y sc scat atte tere red d beam be ams. s. Ho Howe weve verr, th the e re rece cent nt me meas asur urem emen ents ts by al..   (2006) Mesolaras   et al (2006) in a pas passiv sively ely sca scatte ttere red d beam be am ar are e su subs bsta tant ntia iall lly y lo low wer th than an th thos ose e by Schneider  et al.  (2002b).

48  

Journal of the ICRU Vol 7 No 2 (2007) Report 78 Oxford University Press

4

DOSI DO SIME METR TRY  Y 

4.1

GENERAL GENERA L CONSI CONSIDERA DERATIONS TIONS

 Absorbed-dose  Absorbeddose deter determina mination tion mus mustt be accu accurat rate e and reproducible as tumor-contr tumor-control ol and normal-tissue complication probabilities are steep functions of absorbed dose. Dosimetry Dosimetry techn techniques iques at any facility must be consistent with those at other facilities if clinical data are to be compared. At the level of one standard deviation (1 SD), relative accuracy of 3 percent is desirable, abl e, alt althou hough gh 5 per percen centt is oft often en ac accep cepted ted,, whi while le relat rel ativ ive e pr preci ecisio sion n (r (repr eprodu oducib cibili ility) ty) of 2 per percen centt is required. If the latter cannot be achieved in clinical practice, the cause should be investigated. The determinat min ation ion of ran range ge is an imp import ortant ant issue in pr proto oton n therapy ther apy and is addressed addressed in Sections Sections 4.7 and 4.8. The specifications for the design of a dosimeter depend on the requirements for:

doi:10.1093/jicru/ndm027

ionization chambers calibr ionization calibrated ated with Far Farada aday y cups have giv gi ven prob oble lem matic resu sullts, yiel yi eld ding proton-absorbed doses that have been found in some cases to be in excess of 10 percent lower than doses determined with ionization chambers calibrated with a calorimeter or with a   60Co beam (Delacroix   et al., 1997; Palmans and Vynckier, 2002; Vatnitsky  et al., 1996b; Vynckier Vynckier,, 2004). However, However, the probl problems ems can be reduced by careful design, and accurate measureal.., 19 ments men ts ar are e ach achiev ievabl able e (Gru (Grusel selll   et al 1995; 95; Jo Jone ness

Calori Calo rime mete ters rs ar are e ab abso solu lute te do dosi sime mete ters rs as th the ey measure directly the energy deposited in a sensing elemen ele mentt by a tem temper perat atur ure e rise rise.. No ra radia diatio tion n cal caliibration is necessary and the measurements have a relative rela tively ly small unce uncertain rtainty ty (Lai (Laitano tano,, 1997 1997;; 1998; Seuntjens and DuSautoy, 2002). They have been recommended ommen ded as the instruments instruments of choice for determindetermining in g th the e ab abso sorb rbed ed do dose se in pr prot oton on be beam amss un unde derr referenc refe rence e condi conditions tions (AAP (AAPM, M, 1986; Vyncki ynckier er   et al., 1991; 199 1; 199 1994). 4). Ho Howe weve verr, the they y ar are e not com commer mercia cially lly avail av ailabl able e and are cum cumber bersom some e and dif difficu ficult lt to use routinely routi nely.. Nev Neverthe ertheless, less, they are findi finding ng incr increasin easing g use us e in al alll ty type pess of th ther erap apy y be beam amss (I (ICR CRU, U, 20 2001; 01; Palmans  et al., 2004; Seuntjens and DuSautoy, 2002). Fara Fa raday day cups can also be use used d at high ene energie rgiess

 et al., 1999). Nevertheless, the use of Faraday cups for clinical clinical pro proton ton dos dose e mea measur sureme ements nts is not rec rec-ommended (see Section 4.2). However, as mentioned in Section 4.7, multilayer Faraday cups (MLFCs) are  very usefu usefull instr instrument umentss for making in-be in-beam am rang range e al.., 199 measureme measu rements nts (Gott (Gottschalk schalk   et al 1999; 9; Pa Pagan ganett ettii and Gottschalk, 2003). Carbon activation can also be used for proto pr oton-b n-beam eam dos dosime imetry try. Thi Thiss tec techni hnique que is bas based ed on the ac activ tivat ation ion of   11C (hal (half-lif f-life e 20.4 min) via the   12C( p,pn)11C rea eact ctio ion n (Cumm mmin ing, g, 19 196 63; Kostjuc Kos tjuchenk henko o and Nich Nichipor iporov ov,, 1993; Lars Larsson son and Sarby, 1987; Larsson  et al., 1965; Nichiporov, 2003; Nichiporov  et al., 2004). It is essentially an off-line method involving irradiation of a carbon sample in the th e be beam am an and d su subs bseq eque uent nt 4p b – g   coincidence   11 counting. C emits a positron ( b) (with maximum energy ene rgy of 968 ke keV) V) tha thatt ann annihi ihilat lates es wit with h an ele elecctron tr on,, th the e re resu sult lt of wh whic ich h is th the e em emis issi sion on of tw two o oppositel oppo sitely y dire directed cted 0.51 0.511 1 MeV quan quanta. ta. The count count-ing system requires the detection of the three particl ti cles es (a   b   and and two 0.5 .51 11 MeV qua uan nta ta)) in coin co inci cide denc nce. e. Th The e te tech chni niqu que e is no nott su suit itab able le fo forr routine applications as the experimental set-up is quite complicated, but it can be used for independent veri verificat fication ion of ioniz ionizatio ation-ch n-chambe amberr dosim dosimetry etry..  A recent study showed relativ relative e agreement of 1.7

al.., 19 (ICRU, 199 (ICRU, 1998; 8; Lai Laitan tano, o, 199 1997; 7; Verh erhey ey   et al 1979 79). ). Howe Ho weve verr, th their eir us use e is ba base sed d on flu fluen ence ce de dete term rmiinations in air and requires accurate knowledge of the effective area, energy spectrum of the proton beam, and knowledge of secondary-particle production. The

percentand between doseschambers measured(Nichiporo using carbon acti vation ionization (Nichiporov v, 2003). The pr prac actic tical al ins instru trumen ments ts for abs absorb orbeded-dos dose e dete de term rmin inat atio ion n ar are e io ioni niza zati tion on ch cham ambe bers rs (B (Boa oag, g, 1966; 196 6; Ch Chu, u, 199 1995b) 5b),, wh which ich ar are e re readi adily ly av avail ailabl able, e,

1. the accuracy of the absorbed-dose determination; 2. the sensitivity of the measuring system; 3. the energy dependence of the dosimeter response; 4. the spatial resolution.

¼

measurements have to be converted to absorbed dose using appr appropria opriate te stop stoppingping-pow power er rati ratios. os. Altho Although ugh reas re ason onab able le ag agrree eeme ment nt ha hass be been en fo foun und d in so some me et al., 20 stud st udies ies ( e.g., Ne Newha whause userr   et 2002 02a; a; 20 2002 02b) b),,

cheap, robust, easy to use, and require little ancil lary lar y equ equipm ipment ent.. Ho Howe weve verr, the there re ar are e no pri primar mary y sta tand ndar ards ds fo forr pr prot oton on be beam ams, s, an and d io ioni niza zati tion on chambers have to be calibrated, either by means of 

# International Commission on Radiation Units and Measurements 2007

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

a cal calorim orimete eterr in th the e use user’s r’s pr proto oton n bea beam, m, or ha have ve calibration coefficients traceable to a primary standard   60Co beam (Andreo, 2002; ICRU, 2001). In the latter case, the knowledge of physical data and constants is required to determine appropriate correc-

been ma been made de (M (Med edin in,, 19 1997 97;; Me Medi din n an and d An Andr dreo eo,, 1997b). These latter calculations were based on the ICRU (ICRU, 1993a) stopping powers, but included the con contri tribut bution ionss fro from m sec second ondary ary ele electr ctrons ons and nonelastic nuclear collisions.

tion factors for for proton-therapy the use in clinical proton beams. Protocols dosimetry have been published by the American Association of Physicists in Med edic icin ine e (A (AAP APM M) (A (AA APM, 19 1986 86)) an and d th the e European Clinical Heavy Particle Dosimetry Group al.., 19 (ECHED) (ECH ED) (V (Vynckier ynckier,, 1995; Vynckier   et al 1991 91;; 1994)) and we 1994 were re con consol solida idated ted int into o ICR ICRU U Re Repor portt 59 (ICRU, 1998; Jones, 2001d). In ICRU Report 59, it is reco re comm mmen ende ded d th that at th the e re refe fere renc nce e ab abso sorbe rbedd-do dose se measurements in the clinical situation be made with cylindr cyli ndrica icall ioni ioniza zatio tion n cham chamber berss ha havin ving g   60Co cal caliibratio bra tion n coe coeffic fficien ients ts tr trace aceabl able e to sta standa ndards rds lab labora ora-tories. The calibration coefficients should be checked, prefer pre ferabl ably y with a calo calorim rimete eterr (IC (ICRU RU,, 199 1998). 8). The

The Th e de dens itie iess an and drelevant atom at omic ic toco comp mpos osit itio ions ns of a  variety of nsit materials proton dosimetry are given in Table 4.1 (ICRU, 1984; 1993a). Proton mass total stopping powers (including nuclear stopping power) and continuous slowing-down approximation (csda) ranges for these materials are given in Tables 4.2 and 4.3, respectively (ICRU, 1993a). The contribution of nuclear stopping power is negligible gib le in the the therap rapeut eutic ic pr proto oton n ene energy rgy ra range nge and 23 amounts to   , 10 of the electronic stopping power . 1 MeV (ICRU, 1993a). In the pre present sent section, reference reference dosi dosimetry metry with Fara Fa rada day y cup cupss and cal calori orimet meters ers is firs firstt dis discu cusse ssed. d. Farada Far aday-cup y-cup dosimetry dosimetry is not reco recommen mmended ded and is

calibra calib rati tion on co coef effic ficien ients ts can be gi give ven n in te term rmss of  airr ke ai kerm rma, a, ex expo posu sure re,, or ab abso sorb rbed ed do dose se to wa wate terr. The Th e la latt tter er is pr pref efer errred as th the e un unce cert rtai aint ntie iess in the cha chambe mber-d r-depen ependen dentt fa facto ctors rs use used d to con conver vertt the measurements to absorbed dose are less, and the formalism formal ism is simpl simpler er and easier to interp interpret ret (And (Andreo, reo, al.., 199 1992;; Hoh 1992 Hohlfie lfield, ld, 198 1988; 8; Med Medin in   et al 1995; 5; Ro Roger gers, s, 1992). 1992 ). Re Recen cently tly the Int Interna ernatio tional nal Ato Atomic mic Ene Energy rgy  Agency  Agenc y (IAEA (IAEA)) has publis published hed (T (Technic echnical al Rep Reports orts Series—TRS 398) a code of practice for radiotherapy dosi do sime metr try y (i (inc nclu lud din ing g pr prot oton on th ther erap apy) y) ba base sed d solely sole ly on abs absorbe orbed d dos dose e to wa water ter sta standa ndards rds usi using ng 60 Co-calibrated ionization chambers (IAEA, 2000). Comparisons of dose measurements among different centers are important for establishing uniform stand st andard ardss and for ve verif rifyin ying g the int integr egrity ity of the al.., 20 al.., dosimetry (Fukum (Fukumura ura   et al 2002 02;; Jo Jone ness   et al 1992; Kacperek  et al., 2002; Vatnitsky  et al., 1996b; 1999b). Such studies are essential in order to establish the relative doses delivered if pooling or comparing par ing cli clinic nical al re resul sults ts fr from om dif differ ferent ent ins instit titute utes. s. Compari Com parisons sons are espec especially ially impo important rtant for unco unconn ventional beams such as protons, and have revealed probl pr oblems ems wit with h do dosim simetr etry y st stra rateg tegies ies (V (Vat atnit nitsky sky  et al., 1996b) (see Sections 4.2 and 4.6). If a single dosimetry dosim etry prot protocol ocol is used used,, inter inter-ins -institu titutiona tionall conet al., sisten sis tency cy is usu usuall ally y ex excel cellen lentt (V (Vat atnit nitsky sky   et

only included here for completeness. The two recent protocols (IAEA, 2000; ICRU, 1998) for ionization-chamber dosimetry in proton beams, based on   60Co calib calibrat rations ions tra traceabl ceable e to sta standard ndardss labor laboraatories (IAEA, 2000; ICRU, 2001), are then described and compared. The importance of dry versus humid (ambie (am bient) nt) air in int interp erpre retin ting g ion ioniza izatio tion-c n-cham hamber ber response is also considered. The energy required to form an ion pair in air (w-value) for protons contributes potentially the largest uncertainty to the determina mi nati tion on of ab abso sorb rbed ed do dose se,, an and d is co cons nsid ider ered ed in Sectio Sec tion n 4.4 4.4.4. .4. Be Becau cause se it is re recom commen mended ded th that at the IAEA protocol (IAEA, 2000) be adopted, its features are covered in full. Implementation details are given in Ap Appen pendix dix A. A workshee worksheett for det determ ermini ining ng the absorbed dose to water in a proton beam using this protocol prot ocol is also provided provided in App Appendix endix A. Dosim Dosimetry etry comparison comp arisons, s, beam monit monitoring oring,, and rela relative tive dosim dosim-etry are also described described in some detail. The descriptions descriptions in the present section concern mainly the dosimetry of static (passively modified) proton beams, but in many cases apply equally to the dos dosime imetry try of ac activ tive e (sc (scan anned ned)) bea beams. ms. Sp Speci ecial al requir req uireme ements nts for sca scann nned ed bea beams ms are elu elucid cidat ated ed wher wh ere e ap appl plic icab able le.. Th The e ma main in di diff ffer eren ence cess in th the e dosi do sime metr try y of pa pass ssiv ive e an and d ac acti tive ve sy syst stem emss ar are e in beam monitoring and relative dosimetry.

1999b) and for is independent of what are thebased calibration coefficients that specific protocol on.

Medin Me din and And Andre reo o (19 (1992) 92) re revie viewe wed d and eva evaluluated th ated the e st stop oppi ping ng-p -pow ower er ra rati tios os of in inte tere rest st fo forr prot pr oton on-b -bea eam m do dosi sime metr try y in us use e at th tha at ti time me.. Subsequently, new mass total stopping-power data were published by the ICRU (ICRU, 1993a). More recently rece ntly,, calcu calculati lations ons of the wa water-t ter-to-air o-air stop stopping ping-powe po werr ra ratio tios, s, wh which ich ar are e spe specifi cificall cally y re requi quire red d for ionizati ioniz ation on cham chamber-b ber-based ased pro proton ton dosim dosimetry etry,, hav have e

4.2 REF REFERE ERENCE NCE DOSI DOSIMET METRY RY WITH A  FARADAY CUP  A Far Farada aday y cup is a device that can be used to determine the number of protons in a beam. Protons that reach the thick absorber inside the Faraday cup produce a net charge proportional to the number of  protons. proto ns. The electric electrically ally insula insulated ted and condu conducting cting 50

 

 DOSIMETRY   DOSIMETR Y  Table 4.1. Densit Table Density y and composition of mater materials ials for which proton stopping stopping powe powers rs (Table 4.2) and rang ranges es (Table 4.3) are given (ICRU, 1984; 1993a).

Material

Density,  r  (g cm23)

Atomi Ato micc com compos positi ition/ on/fr frac actio tion n by mas masss H

Water TE gas (propane)a Polystyrene PMMA ICRU muscle Graphite  Air,, drya (n  Air (nea earr se sea a le leve vel) l)  A-150 plastic

1.0000 1.8263  1023 1.0600 1.1900 1.0400 2.2200 1.20 1. 2048 484 4  1023 1.1270

0.111894 0.102672 0.077418 0.080538 0.101997

0.101327

C

N

O

Others

0.568940 0.922582 0.599848 0.123000 1.000000 0.00 0. 0001 0124 24 0.775501

0.035022

0.888106 0.293366

0.035000

0.319614 0.729003 0.011000 (Na, Mg, P, S, K)

0.7552 0.75 5267 67 0.035057

0.2317 0.23 1781 81 0. 0.01 0128 2827 27 (A (Ar) r) 0.052316 0.017422 (F) 0.018378 (Ca)

a

 At standard tempera temperature ture and pressur pressure. e.

proton-beam absorber must be thick enough to stop all primary protons and proton-produced secondary-

fluen ence ce me mea asu surrem emen entt fol ollo low wed by a char charge ge measu mea surem rement ent wit with h an ioniza ionizatio tion n chambe chamberr at the the

charged particles in the A potential error in this measurement canabsorber. occur because of the collection of seco seconda ndaryry-cha charge rged d par particl ticles es such as spa spalla lla-tion products and electrons produced upstream that can ca n ad add d to to,, or su subt btra ract ct fr from om,, th the e ch char arge ge du due e to primary protons. Charged particles generated in the sensing absorber, usually electrons, can escape the absorber and modify the response. Appropriate use of thi thin n ent entra rance nce foi foils, ls, va vacuu cuum m env enviro irons, ns, and tr trapapping electromagnetic fields surrounding the sensing absorber minimizes the effects due to the escape of  al.., 19 secondary second ary electr electrons ons (Camb (Cambria ria   et al 1997 97;; Jo Jone ness  et al., 1999; Kacperek and Bonnett, 1990; Lin  et al., al.., 197 al.., 19 1994; Verhey   et al 1979; 9; Vync ynckier kier   et al 1984 84;;

sam same point. poi nt.   E If and the th e is proto pr oton n bea beam m is mon monoen oenerg ergeti eticc withe energy of small enough cross-section to be com comple pletel tely y ac accep cepted ted by the apertur aperture e of th the e cup, then the dose to water  D w  is given by

Ziegler  et al., 1996). Another problem concerns energetic get ic pr proto otons ns and oth other er hea heavie vierr cha charged rged par particl ticles es produced near the periphery of the absorber by fast neutrons neutr ons tha thatt are gener generated ated by prot proton on inter interaction actionss in the upstream portion of the sensing absorber. The use of non-hydrogenous high- Z   absorbers with concomitantly comita ntly small partic particle-pr le-producti oduction on cross cross-secti -sections ons minimizes this effect (ICRU, 1998). Faraday-cup-based dosimetry is more sensitive to the ene energy rgy dis distri tribut bution ion of the pr proto oton n bea beam m th than an ionization chambers or calorimeters and, therefore, caution is recommended when using fluence deter-

water (in MeV cm21. g ),   Dw   is then expressed in Gy (monitor unit) When Wh en th the e ca cali libr brat atin ing g be beam am is co comp mpos osed ed of a mixtu mix ture re of pr proto otons ns of var variou iouss ene energi rgies, es, the ma mass ss electronic stopping power in Equation (4.1) must be replaced with an integral over the proton spectrum:

10

 Dw  ¼ ð1:602  10

  ð Þ

Þ

  N    S  E a r 

;

 

ð4:1Þ

w

where 1 eV  where eV  1.602  10219 J,   N   is the number of  protons per monitor unit collected in the Faraday cup,   a   is th the e ef effe fect ctiv ive e ar area ea of th the e be beam am (i (in n cm2; assuming assu ming unif uniformit ormity), y), [ S( E)/ r  r ]w   is the mas masss ele elecctronic stopping power of the protons of energy   E   in ¼

 2

21

10

 Dw  ¼ ð1:602  10

ð 

 ð Þ

Þ F E ð EÞ

  S  E r 

d E;   ð4:2Þ

w

where   FE   ( E) is the fluence of protons of energy   E per monitor unit.

mination as a basis for clinical dosimetry. A fluence measur mea sureme ement nt in a pr proto oton n bea beam m of kno known wn flu fluenc ence e distribution (as a function of energy) can be used to calibrate a transfer ionization chamber for use in a clinical beam. If one assumes that a transmission ioniza ion izatio tion n cha chambe mberr is use used d to mon monito itorr the bea beam m 1 fluence, the dose to water  Dw  can be obtained by a

The mass stopping power to be used depends critically cal ly on accura urately tely known kno wn of beam energy ene rgy. scattered The pr preesence of aacc small admixture low-energy proton pr otonss can lead to sign significa ificant nt err errors ors in abso absorbed rbed-dose determ determinati ination on (V (Verhe erhey y   et al., 197 1979) 9).. Mo Mont nte e Carlo calculations may be helpful in estimating the effect of low-en low-energy ergy contam contaminant inantss on the calibr calibration ation..  A monoen monoenergetic ergetic prot proton on beam const constructed ructed withou withoutt collimators has been reported to be capable of avoiding the prod production uction of low-e low-energy nergy scattered scattered proto protons ns al.., 19 (Grusell   et al 1995 95). ). Th The e ef effec fectt of nu nucle clear ar in inte terractions is to increase the apparent mean deposited

1

For consistency consistency the dosim dosimetry etry nome nomencla nclatur ture e of IAEA Rep Report ort TRS 398 (IAE (IAEA, A, 2000) is used throughou throughoutt the present present repo report. rt. Specifically, the use of italics in some subscripts of symbols for dosimetric quantities follows IAEA Report TRS 398.

51  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  Table 4.2. Proton mass total stopping powers in various materials (ICRU, 1993a).

Energy,  E   (MeV (MeV))

1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.50 4.00 4.50 5.00 5.50 6 6..0 50 0 7.00 7.50 8.00 8.50 9.00 9.50 10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 30.0 35.0 40.0 45.0 50.0 55.0 60.0 65.0 70.0 75.0 80.0 85.0

r  (MeV Proton Pro ton mass tota totall stoppi stopping ng pow power er,,  S tot / r    (MeV cm2 g21)

Wate terr

TE ga gass (p (prropa pan ne)

Pol oly ysty tyrrene

PMM PM MA

ICRU IC RU musc scle le

Graph phit ite e

Air (d (dry ry))

A-15 A150 0 pl plas asti ticc

260.8 222.9 195.7 174.9 158.6 145.4 134.4 125.1 117.2 104.2 94.04 85.86 79.11 73.43

27 2 73.7 233.4 23 204.5 20 182.6 18 165.3 16 151.4 15 139.8 13 130.0 13 121.7 12 108.0 10 97.38 97 88.80 88 81.74 81 75.80 75

257.7 220.5 193.7 173.3 157.2 144.0 133.2 123.9 116.1 103.2 93.06 84.92 78.20 72.54

253.2 216.8 190.5 170.5 154.6 141.8 131.1 122.1 114.3 101.7 91.79 83.79 77.19 71.64

25 2 58.5 221.0 22 194.0 19 173.5 17 157.3 15 144.2 14 133.3 13 124.1 12 1 16.2 11 103.4 10 93.29 93 85.17 85 78.47 78 72.84 72

229.7 196.8 173.1 155.1 140.9 129.3 119.6 111.5 104.5 93.01 83.99 76.71 70.69 65.62

222.9 191.2 168.3 150.9 137.1 125.8 116.5 108.6 101.8 90.68 81.97 74.92 69.09 64.17

268.1 229.0 200.9 179.6 162.7 149.1 137.7 128.2 120.0 106.6 96.10 87.66 80.70 74.85

6 68 4..5 38 8 60.71 57.47 54.60 52.02 49.69 47.59 45.67 38.15 32.92 29.05 26.07 23.69 21.75 20.13 18.76 16.56 14.88 13.54 12.45 11.54 10.78 10.13 9.559 9.063 8.625 8.236

70 7 60 66 6..7 34 7 62.55 62 59.18 59 56.19 56 53.51 53 51.10 51 48.91 48 46.92 46 39.14 39 33.73 33 29.74 29 26.67 26 24.22 24 22.22 22 20.56 20 19.15 19 16.90 16 15.17 15 13.80 13 12.68 12 11.76 11 10.98 10 10.31 10 9.727 9. 9.221 9. 8.774 8. 8.376 8.

6 67 3..7 52 6 59.91 56.70 53.84 51.29 48.99 46.90 45.00 37.56 32.39 28.57 25.62 23.28 21.36 19.77 18.42 16.26 14.60 13.28 12.21 11.32 10.57 9.926 9.369 8.882 8.452 8.070

6 66 2..9 80 0 59.21 56.05 53.24 50.73 48.45 46.40 44.52 37.19 32.08 28.31 25.39 23.08 21.18 19.61 18.27 16.13 14.49 13.18 12.12 11.24 10.50 9.858 9.306 8.823 8.397 8.018

68 6 68 63 3..0 82 6 60.22 60 57.00 57 54.15 54 51.59 51 49.28 49 47.19 47 45.29 45 37.84 37 32.64 32 28.81 28 25.85 25 23.49 23 21.56 21 19.96 19 18.60 18 16.42 16 14.75 14 13.42 13 12.34 12 11.45 11 10.69 10 10.04 9.477 9. 8.986 8. 8.552 8. 8.166 8.

6 51 7..3 50 5 54.28 51.39 48.81 46.52 44.44 42.56 40.84 34.13 29.45 25.99 23.32 21.19 19.46 18.01 16.78 14.82 13.31 12.11 11.14 10.33 9.645 9.060 8.553 8.109 7.717 7.368

5 59 6..9 37 3 53.15 50.33 47.83 45.59 43.57 41.73 40.06 33.51 28.94 25.55 22.94 20.85 19.15 17.73 16.53 14.60 13.12 11.94 10.99 10.19 9.517 8.942 8.443 8.006 7.620 7.277

6 69 5..8 56 5 61.78 58.46 55.51 52.87 50.49 48.34 46.37 38.70 33.36 29.42 26.38 23.96 21.99 20.34 18.95 16.73 15.02 13.66 12.56 11.64 10.87 10.21 9.634 9.133 8.691 8.297

90.0 95.0 100 125 150

7.888 7.573 7.289 6.192 5.445

8..020 8 7.699 7. 7.409 7. 6.290 6. 5.528 5.

7.728 7.420 7.140 6.064 5.331

7.678 7.372 7.095 6.027 5.300

7..820 7 7.509 7. 7.227 7. 6.139 6. 5.398 5.

7.056 6.775 6.520 5.538 4.868

6.970 6.693 6.443 5.475 4.816

7.945 7.628 7.341 6.233 5.479

175 200 225 250 275 300 350 400 450 500

4.903 4.492 4.170 3.911 3.698 3.520 3.241 3.032 2.871 2.743

4..976 4 4.558 4. 4.230 4. 3.966 3. 3.749 3. 3.568 3. 3.284 3. 3.072 3. 2.908 2. 2.778 2.

4.800 4.397 4.081 3.827 3.618 3.444 3.170 2.966 2.808 2.683

4.772 4.372 4.058 3.806 3.599 3.426 3.154 2.951 2.794 2.670

4..861 4 4.453 4. 4.134 4. 3.877 3. 3.666 3. 3.490 3. 3.213 3. 3.006 3. 2.846 2. 2.720 2.

4.382 4.014 3.724 3.492 3.300 3.140 2.889 2.700 2.554 2.438

4.338 3.976 3.691 3.462 3.275 3.118 2.871 2.687 2.544 2.431

4.932 4.518 4.193 3.932 3.717 3.538 3.257 3.046 2.884 2.755

52  

 DOSIMETRY   DOSIMETR Y  Table 4.3. Proton continuous slowing down approximation (csda) ranges in various materials (ICRU, 1993a).

Energy,  E   (MeV (MeV))

Proton Pro ton csda ran range, ge, r 0 (g cm22) Wate terr

TE ga gass (p (prropa pane ne))

Pol oly ysty tyrrene

PMMA

ICR IC RU musc sclle

Grap aph hit ite e

Air (d (drry)

A-1 -150 50 pl plas asttic

1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.50 4.00 4.50 5.00

0.00246 0.00350 0.00470 0.00605 0.00756 0.00920 0.0110 0.0129 0.0150 0.0195 0.0246 0.0302 0.0362

0.00227 0.00326 0.00441 0.00571 0.00715 0 .00873 0. 0.0105 0.0123 0.0143 0.0187 0.0236 0.0289 0.0348

0.00244 0.00349 0.00470 0.00607 0.00759 0.00925 0.0111 0.0130 0.0151 0.0197 0.0248 0.0304 0.0366

0.00247 0.00354 0.00477 0.00616 0.00770 0.00939 0.0112 0.0132 0.0153 0.0200 0.0252 0.0309 0.0371

0.00247 0.00352 0.00473 0.00609 0.00761 0 .00927 0. 0.0111 0.0130 0.0151 0.0197 0.0248 0.0304 0.0365

0.00275 0.00393 0.00529 0.00682 0.00851 0.0104 0.0124 0.0145 0.0169 0.0219 0.0276 0.0338 0.0406

0.00287 0.00408 0.00548 0.00705 0.00879 0.0107 0.0128 0.0150 0.0174 0.0226 0.0284 0.0348 0.0417

0.00230 0.00331 0.00448 0.00580 0.00726 0.00887 0.0106 0.0125 0.0145 0.0189 0.0239 0.0294 0.0353

5.50 6.00 6.50 7.00 7.50 8.00 8.50 9.00 9.50 10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 30.0 35.0 40.0 45.0 50.0 55.0 60.0

0.0428 0.0498 0.0574 0.0654 0.0738 0.0828 0.0922 0.102 0.112 0.123 0.183 0.254 0.335 0.426 0.527 0.637 0.757 0.885 1.17 1.49 1.84 2.23 2.64 3.09

0.0412 0.0480 0.0553 0.0631 0.0713 0.0800 0.0891 0.0986 0.109 0.119 0.178 0.247 0.326 0.415 0.513 0.621 0.738 0.864 1.14 1.46 1.80 2.18 2.59 3.03

0.0432 0.0503 0.0580 0.0661 0.0747 0.0837 0.0932 0.103 0.114 0.125 0.186 0.258 0.340 0.432 0.535 0.647 0.769 0.900 1.19 1.52 1.87 2.27 2.69 3.15

0.0438 0.0510 0.0588 0.0670 0.0756 0.0848 0.0944 0.105 0.115 0.126 0.188 0.260 0.344 0.437 0.540 0.654 0.776 0.909 1.20 1.53 1.89 2.29 2.72 3.18

0.0431 0.0502 0.0578 0.0659 0.0744 0.0834 0.0929 0.103 0.113 0.124 0.185 0.256 0.338 0.430 0.531 0.642 0.763 0.893 1.18 1.50 1.86 2.25 2.67 3.12

0.0480 0.0559 0.0643 0.0733 0.0827 0.0927 0.103 0.114 0.126 0.138 0.205 0.284 0.375 0.476 0.589 0.712 0.846 0.990 1.31 1.66 2.06 2.49 2.96 3.46

0.0493 0.0573 0.0659 0.0751 0.0847 0.0949 0.106 0.117 0.129 0.141 0.209 0.290 0.382 0.486 0.600 0.725 0.861 1.01 1.33 1.69 2.09 2.53 3.00 3.51

0.0417 0.0487 0.0561 0.0639 0.0722 0.0810 0.0902 0.0999 0.110 0.121 0.180 0.250 0.330 0.420 0.519 0.628 0.747 0.874 1.16 1.47 1.82 2.20 2.62 3.06

65.0 70.0 75.0 80.0 85.0 90.0 95.0 100 125 150 175 200 225 250 275 300 350 400 450 500

3.57 4.08 4.62 5.18 5.78 6.40 7.05 7.72 11.5 15.8 20.6 26.0 31.7 37.9 44.5 51.5 66.3 82.3 99.2 117

3.50 4.00 4.53 5.09 5.67 6.28 6.92 7.58 11.3 15.5 20.3 25.5 31.2 37.4 43.8 50.7 65.3 81.1 97.8 115

3.64 4.16 4.71 5.28 5.89 6.52 7.18 7.87 11.7 16.1 21.1 26.5 32.4 38.7 45.5 52.6 67.7 84.0 101 120

3.67 4.19 4.74 5.32 5.93 6.57 7.24 7.93 11.8 16.2 21.2 26.7 32.6 39.0 45.7 52.9 68.1 84.5 102 120

3.60 4.12 4.66 5.23 5.83 6.45 7.11 7.78 11.6 15.9 20.8 26.2 32.0 38.3 44.9 51.9 66.9 83.0 100 118

3.99 4.56 5.16 5.79 6.46 7.15 7.87 8.63 12.8 17.6 23.1 29.0 35.5 42.4 49.8 57.6 74.2 92.1 111 131

4.05 4.63 5.24 5.88 6.55 7.25 7.98 8.74 13.0 17.9 23.3 29.4 35.9 42.9 50.3 58.2 74.9 92.9 112 132

3.54 4.04 4.58 5.14 5.73 6.34 6.98 7.65 11.4 15.7 20.5 25.8 31.5 37.7 44.2 51.1 65.9 81.8 98.7 116

53  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

energy per proton. This increases the predicted dose to water per proton, depending on proton energy, by severa sev erall per percent cent (Se (Seltz ltzer er,, 1993 1993). ). The unc uncert ertaint ainties ies introduced by nuclear effects and by the sensitivity of th the e ca cali libr brat atio ion n to th the e en ener ergy gy an and d ty type pe of th the e beam particles combine to make ionization-chamber calibration calibr ationss based on fluence potentially potentially less accurate than those based on calorimetry and   60Co. Several studies in clinical proton beams, comparing in g do dose sess de dedu duce ced d di dire rect ctly ly fr from om Far arad aday ay-c -cup up measure meas uremen ments ts or with ion ioniza izatio tion n cham chambers bers cal caliibrat br ated ed wi with th Fa Fara rada day y cup cups, s, an and d wit with h io ioniz nizat ation ion cham ch ambe bers rs ca calib libra rate ted d wi with th cal calori orime mete ters rs or   60Co beams, have shown consis consistently tently large rela relative tive discrepan cre pancies cies (5– 20 per percent cent), ), with the Far Farada aday-c y-cup up dose determinat ions usually being lower low er (Cambria (Cambr ia,  et al.,determ   et al.ly 1997; 199 7; inations Cutton Cut tone e usual , 1999 1999: : Dela Delacro croix ix   et al. 1997; Gall   et al., 1994; Jones   et al., 1994 1994a; a; 1994 1994b; b; al.., 1994 al.., 19 Schreuder   et al 1994;; Vatn atnits itsky ky   et al 1996 96b; b;  Verhey  V erhey et al., 1979). The magnitudes of these discrepancie pan ciess dep depend end to som some e ext extent ent on whic which h pro protoco tocols ls and physical constants were used to determine the absorbed absorbe d dose with ioniza ionization tion chambers havin having g   60Co calibration coefficients, and on the stopping powers used for the Far Faradayaday-cup cup measur measurement ements. s. How However ever,,  variations  variat ions in these latt latter er values do not accou account nt for the large differences in absorbed dose that are frequentl que ntly y obse observe rved. d. Bet Better ter rel relati ative ve agr agreeme eement nt ( , 4 percent) was found in some studies (Grusell   et al., al.., 19 1995; Jon 1995; Jones es   et al 1999 99;; Ka Kacp cper erek ek an and d Bo Bonn nnet ett, t, et al., 19 1990;; Ka 1990 Kacper cperek ek   et 1991; 91; Ne Newh whau ause serr   et al., 2002a; 2002b). These latter results can possibly be

4.3 REF REFERE ERENCE NCE DOSI DOSIMET METRY RY WITH A  CALORIMETER Unlike Unlik e mea measur sureme ements nts bas based ed up upon on th the e pr produ oducts cts produced prod uced by the inter interacti action on of ioniz ionizing ing radi radiati ation on with wit h ma matte tterr,   e.g., io ioni niza zati tion ons, s, a ca calo lori rime metr tric ic meas me asur urem emen entt is a di dire rect ct de dete term rmin inat atio ion n of th the e energy imparted to a sensing element as indicated by the temperature change. Under the assumption tha th at al alll th the e de depo posi site ted d en ener ergy gy is th ther erma mali lize zed, d, absorbed dose may thus be directly determined in a calorimetric measurement. In principle, no knowledge ed ge of an any y ra radi diat atio ion n pa para rame mete ters rs is re requ quir ired ed.. Ideally Ideal ly,, absor absorbedbed-dose dose stan standard dardss for ther therapeu apeutic tic prot pr oton on be beam amss sh shou ould ld be ca calo lori rime mete ters rs.. A ca calo lori ri-metr me tric ic ab abso sorb rbed ed-d -dos ose e de dete term rmin inat atio ion n ca can n al also so provide an independent confirmation of ionizationchamber chamb er dete determin rminatio ations. ns. Detai Detailed led discu discussions ssions of  calorimetri calori metricc techn techniques iques are given in Att Attix ix (1986 (1986), ), Daures   et al.   (1994), (1994), Dome Domen n (198 (1987), 7), Pal Palmans mans and al..   (2004),  Vynckier  V ynckier (2002), Palmans   et al (2004), Ros Rosss and Klassen (1996), Schulz  et al.  (1987), Seuntjens and DuSautoy (2002), and Seuntjens  et al.  (1994). Calorimetric Calorim etric determ determinati ination on invol involves ves the measur measureement me nt of th the e te temp mper era atu turre ch cha ang nge e in a ma mass ss of  material resulting from energy imparted by ionizing radiat rad iation ion.. The tem temper peratu ature re chan change ge is typ typical ically ly of  the th e or ord der of th the e 1023 8C Gy21. Th Therm ermis isto tors rs ar are e metal-oxide semi-conductors with a negative temperature coefficient, which can be used in conjunction with a Wheatstone bridge to conveniently determine this th is te temp mper era atu ture re ch chan ange ge wi with th gr grea eatt pr prec ecis isio ion n

(Domen, 1987; ICRU, 1998; Kubo and Brown, 1984; 1986; Kubo  et al., 1989; Laughlin and Jenna, 1966). Many Ma ny ca calo lori rime mete ters rs op oper era ate in an ad adia iaba bati ticc manner. One or more insulating jackets surround a core of material. In an adiabatic or isothermal operation mode, the core and jacket are maintained at equal equ al tem temper perat atur ure. e. If a tem temper perat atur ure e dif differ ferenc ence e exist exi sts, s, the then n the cal calori orimet meter er re respo sponse nse nee needs ds to be correc cor rected ted for hea heatt tr trans ansfer fer bet betwe ween en the cor core e and  jacket. The net core temperat temperature ure change, relative to the pre pre-irra -irradiat diation ion perio period, d, is then proportiona proportionall to

ascribed to improved Faraday-cup construction, and more accurate determination of the beam area and correction for secondary-particle absorption. Doses have been determined in a 250 MeV clinical beam at the Paul Scherrer Institute (PSI) with a Fa Fara rada day y cup an and d an ion ioniza izatio tion n cha chambe mberr usi using ng penc pe ncil il be beam amss (u (use sed d fo forr be beam am sc scan anni ning ng). ). For resid re sidual ual ene energi rgies es bet betwe ween en 138 and 214 MeV, the ratio ra tio of th the e dos doses es mea measur sured ed wit with h the ion ioniza izatio tion n chamber to those measured with the Faraday cup al.., 20  varied from 0.978 to 1.008 (Cora (Coray y   et al 2002 02;; Pedroni  et al., 2005). This illustrates that with care relatively good agreement between the two dosimetry etr y sy syst stems ems can be obt obtain ained. ed. Nev Nevert erthel heless ess,, the Faraday-cup primary monitor at PSI is corrected to the ionization-chamber dose values. In vi view ew of th the e la larg rge e di diff ffer eren ence cess fr freq eque uent ntly ly observed obser ved,, but not suffi sufficient ciently ly unde underst rstood, ood, betw between een absorbed absor bed doses obtained using Far Farada aday y cups and 60 Co-c Co -cal alib ibra rate ted d io ioni niza zati tion on ch cham ambe bers rs or ca calo loririmeters met ers,, the sol sole e use of suc such h flu fluenc ence-b e-base ased d tec techhniques is not currently recommended for proton-therapy dosimetry.

the energy deposited by the incident ionizing radiation at ion.. To ens ensur ure e th that at the sig signal nal der deriv ives es fro from m the material of interest, even for large imparted energy densit den sity y, a cor core e mas masss of sev sever eral al gr grams ams is usu usuall ally y employ emp loyed. ed. The imp impart arted ed ene energy rgy per uni unitt mas masss is then representative of the average absorbed dose to the th e co core re ma mate teri rial al.. Co Conv nver ersi sion on of th the e me meas asur ured ed temperature change to energy imparted is accomplished by knowledge of the mass of the core, combined bin ed eit either her wit with h cal calibr ibrat ation ion of th the e tem temper perat atur ure e response of the core with resistive heating, or with knowledge of the specific heat of the core material.  Adiabatic calorimeters usually use homogeneous 54

 

 DOSIMETRY   DOSIMETR Y 

core and jacket materials such as graphite (Domen and Lamperti, 1974; Palmans  et al., 2004) or A-150 al.., 19 tissue-eq tissu e-equival uivalent ent plas plastic tic (Del (Delacro acroix ix   et al 1997 97;; McDonald et al., 1976; Verhey  et al., 1979).  Although calorimetry avoids difficulties associated with determination or knowledge of radiationspecific speci fic par paramet ameters, ers, tech technical nical prob problems lems can limit the ac accur curac acy y of th the e mea measur sureme ement. nt. Pe Perha rhaps, ps, the most significant is the conversion of the deposited energy into non-thermal processes such as chemical reactions. These reactions may create or absorb heat,   i.e., they ca can n be ei eith ther er exot oth her erm mic or end en doth ther erm mic ic.. On One e ca can n de defin fine e a hea eatt exc xces esss (exothermic case) or heat defect (endothermic case)

course, this latter factor does not apply if a water calorimeter is used. The heat defect  T D  is given by

as fraction of energy imparted creates or the absorb abs orbs s hea heat, t, the respec res pectiv tively ely, , in the that calori cal orimet meter er material. Calorimet Calo rimeters ers pro provide vide the mos mostt dire direct ct meth method od of  absorbedabsor bed-dose dose dete determina rmination tion and are an exce excellent llent choice cho ice for a pri primar mary y st stand andard ard.. Ho Howe weve verr, cal calori ori-meters are not generally available at proton facilitiess and are mor tie more e dif difficu ficult lt to use than ion ioniza izatio tion n chambers, and therefore they have not been chosen as routine reference dosimeters for proton therapy. They are, however, in use as reference dosimeters at na natio tional nal and int intern ernat ation ional al st stand andard ardss lab labor oraatories. torie s. Appr Appropria opriate te exam examples ples of calor calorimet imeters ers hav have e been bee n con const struc ructed ted of gr graph aphite ite (M (McDo cDonal nald, d, 198 1987; 7;

fo for rsorp heat he at tran ansf sfer (con onve vect ctio n and an d ceco cond nduc tion on), ), abso ab rpti tion on tr and an d ersc sca a(c tter tt erin ing g ion (pres (pr esen ence ofucti othe ot her r material mat erials), s), and dose grad gradient ient (diff (differen erence ce betw between een measu mea sured red dos dose e and dos dose e at ref refer erenc ence e poi point nt). ). Fo Forr calorimetr calor imetry y with solid mat materials erials,, conv convectiv ective e hea heatt transfer does not apply. The core can be thermally insulated by a vacuum or air cavity, and a correction factor is required to account for the presence of  the cavity (Bou (Boutillon tillon,, 1989; McEwen McEwen and Duane, 2000; Palmans and Vynckier, 2002; Seuntjens and DuSautoy, 2002). If conduction and convection heat losses are minimal, the temperature rise is directly related to the energy imparted per unit mass near the measurement measurement point point.. Any possible hea heatt defe defect ct

T D  ¼

 Ea  Eh ;  Ea

 

ð4:4Þ

where   Ea   is the ene energy rgy imp impart arted ed to th the e ma mater terial ial and   Eh  is the energy appearing as heat. A positive heatt def hea defect ect mea means ns tha thatt som some e dep deposi osited ted ene energy rgy is lost in rear rearrang rangemen ementt of the lattice structure structure and the temperature rise is therefore too small. The factors   ki   for calorimetry include corrections

McEwen McEw en and Duane, 2000 2000;; Pal Palmans mans   et al., 2004),  A-150 tissue-equivalent (TE) plastic (Caumes  et al., 1984; Delacroix  et al., 1997; McDonald and Domen, 1986), and water (Brede  et al., 2000; Domen, 1980; 1994; Schulz   et al., 1987; Seuntjens and Palmans, 1999). The use of an ice calorimeter has even been reported repo rted in Ros Rosser ser (1994 (1994). ). When available, available, calori calori-meters met ers sho should uld be use used d as pri primar mary y st stand andard ardss or or,, alternatively, to confirm the proton calibration coefficient of the reference ionization chamber (ICRU, 1998).  A calorimeter can be used to determine the absorbed dose to water  D w  in a proton beam as

where   DT  is the temperature rise due to radiation (in K),   c   the specific heat of the sensitive element (in J kg21 K 21),   T D   is the heat defect defect (or excess) excess) due to deposited energy which does not result in a temperature increase (see below),   ki  is the product of sev sever eral al cor corre recti ction on fa facto ctors rs (e (excl xclud uding ing the hea heatt defe de fect ct)) th that at de depe pend nd on th the e co cons nstr truc ucti tion on of th the e calorimeter and its core material (see below), and sw,cal   is the ra ratio tio of the mas masss ele electr ctroni onicc st stopp opping ing

has to be considered and taken into account.  As mentioned above, the value of   DT   is normally determin dete rmined ed with the help of a Whea Wheatsto tstone ne bridg bridge e that is used to measure the change in resistance of  a the thermi rmist stor or in the therma rmall con conta tact ct wit with h the cal calori ori-meter met er.. Th The e spe specifi cificc hea heatt   c   can can be me meas asur ured ed by passin pas sing g an ac accur curat ately ely kno known wn cur curren rentt th throu rough gh a heating resistor that is in thermal contact with the calo ca lori rime mete terr, an and d me meas asur urin ing g th the e te temp mper erat atur ure e change chan ge for a kno known wn amou amount nt of dissi dissipat pated ed energy. In the case of water,   c  is well known (Wagner and Pruss, 2002). These techniques are discussed elsewhere (Domen, 1987; McDonald and Domen, 1986; Seuntj Seu ntjens ens and DuS DuSaut autoy oy,, 200 2002). 2). Th The e mas masss ele elecctronic tro nic stop stopping ping powers needed are energy depe depenndent, den t, bu butt the ra ratio tio req requir uired ed for the cal calori orimet metric ric dose determination in water is only weakly dependent on proton energy for energies above    1 MeV  (ICRU, 1998). Properties of common materials used for calorimetry are given in Table 4.4. Graphite calorimeters have a longer history than water calorimeters as national laboratory absorbeddose standards (Chauvenet   et al., 1997; DuSautoy, al.., 19 1996;; Guerr 1996 Guerra a   et al 1996 96;; Mc McEw Ewen en an and d Du Duan ane, e, 2000), 200 0), and th the e hea heatt def defect ect and con convec vectio tion n eff effect ectss are assumed to be negligible (ICRU, 1998; Schulz  et al., 199 1990; 0; Seu Seuntj ntjens ens and DuS DuSaut autoy oy,, 200 2002). 2). In

powe po werr of wa wate terr to th the e ca calo lori rime mete terr ma mate teri rial al.. Of 

addition, the specific heat of graphite is a factor of 

 Dw  ¼ DTcð1 þ T D Þ1 ki sw;cal ;

 

ð4:3Þ

55  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  Table 4.4. Properties Table Properties of mater materials ials used for calor calorimetry imetry (adapted from Palma Palmans ns and Vynckie ynckier r, 2002; reproduced reproduced with permission).

Property

Water (48C)

Graphite

A-150

Heat defect (T D) (%) Specific heat (c) (J kg21 K 21) Thermal diffusivity (a) (m2 s21) Temperature rise per unit dose ( DT  /  D) (mK Gy21)

0.0 + 0.5 4205 1.44  1027 0.24

0.0 + 0.3 710 0.80  1024 1.41

4.0 + 1.5 1720 2.72  1027 0.58

6 lo low wer th than an th tha at of wat ater er,, gi givi ving ng ri rise se to an increa inc reased sed sig signal nal-to -to-no -noise ise ra ratio tio of pot potent ential ially ly the same magnitude compared with water calorimetry. Difficu Dif ficulti lties es wit with h bot both h wa water ter (R (Ross oss and Kla Klasse ssen, n, 1996) 199 6) an and d gra graphi phite te (M (McEw cEwen en an and d Dua Duane, ne, 200 2000; 0; Seuntj Seu ntjens ens and DuS DuSaut autoy oy,, 200 2002) 2) cal calori orimet metry ry are understood and have largely been solved. However, the th e us use e of a gr grap aphi hite te ca calo lori rime mete terr to de deri rive ve th the e absorbed dose to water requires knowledge of con version factors that can increase the uncertainty uncertainty.. Wate terr ca calo lori rime mete ters rs ha have ve be been en de deve velo lope ped d to prov pr ovide ide a mor more e dir direct ect det determ ermina inatio tion n of abs absorb orbed ed dose do se to wate terr (D (Dom omen en,, 19 1980 80;; 19 1994 94;; Ros osss an and d

 et al., 1991) for low-LET beams. The latter result is consistent with later measurements by Brede  et al. (1994; 1997), Palmans   et al.   (1996), and Seuntjens  et al.   (1994 (1994). ). It is re reco comm mmen ende ded d th that at th the e he heat at defectt be dete defec determine rmined d for each wa waterter-calori calorimetri metricc system used. The heat defect for A-150 plastic is usually taken al.., 199 as 0.0 0.04 4 + 0.015 (Br (Brede ede   et al 1997; 7; Fle Flemin ming g and Glass, 1969; McDonald and Goodman, 1982; Sa¨bel  et al., 197 al.., 199 1973; 3; Sch Schulz ulz   et al 1990). 0). Bec Becaus ause e AA-150 150 plastic is a mixture of several materials, the heat defect may be sensitive to the manufacturing and curing curin g pro process cess (Br (Brede ede   et al., 1997) as well as the

et al., 19 Klassen, Klasse n, 199 1996; 6; Se Seunt untjen jenss   et 1994 94). ). Th Thes ese e device dev icess ar are e us usual ually ly ope opera rated ted in a non non-ad -adiab iabat atic ic manner, and the temperature in a small region (the

radiatio radia tion n his histor tory y of the sam sample ple.. Th There erefor fore, e, it is desirable to measure the heat defect of individual batch ba tches es of AA-150 150 pri prior or to any dos dosime imetry try inv inves estiti-

‘co ‘core’ re’)) of the water wa ter cal calorim orimete eterrelement surrou sur roundi nding ng the thermistor temperature-sensing is directly measured measu red.. For wa water ter calor calorimet imeters ers oper operated ated near room ro om tem temper perat atur ure, e, con convec vectio tion n cur curren rents ts can be a recurring problem. Several Sev eral wat water-b er-based ased port portable able calor calorimet imeters ers hav have e been developed for absorbed-dose determinations in  various therapy (including proton) beams (Brede  et al., 2000; 2006; Medin   et al., 2005; 2006; Schulz  et al., 1987; 1991; 1992). Palmans  et al.  (2004) have also recen recently tly dev develope eloped d a port portable able grap graphite hite-base -based d instrume ins trument nt with which meas measurem urements ents wer were e made in a lowlow-energ energy y (62 MeV) ocula ocular-tr r-treat eatment ment prot proton on beam line.

gations. Using low-energy protons, deuterons, and  et al.  (1997) a-particles of different energies, Brede have demonstrated an LET dependence of the heat defectt in A-15 defec A-150 0 plas plastic tic and pur pure e wat water er.. How However ever,, forr lo fo loww-LE LET T ra radi dia ati tion onss su such ch as hi high gh-e -ene nerg rgy y protons, this effect is negligible. The uncertainties of the mass electronic stoppingpowe po werr ra ratio tio and the hea heatt def defect ect giv give e a com combin bined ed relative uncertainty of     2 – 3 percent percent in calori calorimetri metricc dose dos e det determ ermina inatio tions ns in A-1 A-150 50 pla plast stic; ic; but in the case of graphite or water, that uncertainty can be reduce red uced d du due e to a sma smalle llerr unc uncert ertain ainty ty in the heat defect (Palmans   et al., 1996; Schulz   et al., 1987). Typical uncertainties for dose determination with a

 Although the heat defect in water is sensitive to the amount of dissolved gases and absorbed impurities rit ies,, dos dose e ra rate, te, and ac accum cumula ulated ted dos dose e (IC (ICRU RU,, al.., 198 1998; 199 8; Kla Klasse ssen n and Ro Ross, ss, 199 1991; 1; Ro Ross ss   et al 1989), 9), for practical irradiation conditions, the heat defect due to diss dissolve olved d impu impuritie ritiess is appr approxima oximately tely constant. sta nt. Inv Investi estigat gations ions by Dome Domen n (1994 (1994)) and Schu Schulz lz  et al.  (1992) indicate that by using nitrogen-purged high-resistivity water, the heat defect can be made negl ne glig igib ibly ly sm smal alll fo forr lo low w li line near ar en energ ergy y tr tran ansf sfer er (LET (L ET)) be beam ams. s. Si Simi mila larr re resu sult ltss ar are e fo foun und d wi with th argonarg on- and hyd hydro rogen gen-sa -satur turat ated ed wa water ter (P (Palm almans ans et al., 1996; and an d Vyn ynck ckie ierr, 20 2002 02;; Pa Palm lman anss   et Seuntjens et al., 1994). For pure (hypoxic) water at 48C, the heat defect is assumed to be zero (Schulz

water calorimeter in a high-energy photon beam are shown in Table 4.5 (Seuntjens and DuSautoy, 2002). Simi Si mila larr un unce cert rtai aint ntie iess ar are e ap appl plic icab able le to pr prot oton on beams. Several Sev eral absor absorbedbed-dose dose meas measurem urements ents hav have e been made with calori calorimete meters rs in clini clinical cal prot proton on beam beams. s. Most of these measurements involved water calorimeters (Brede  et al., 1999; 2006; Hashemian  et al., 2003; Jones   et al., 1999; Medin   et al., 2005; 2006; Palmans  et al., 1996; Seuntjens  et al., 1994; Schulz  et al., 1992; Siebers  et al., 1995), while two involved  A-150 plastic calorimeters (Delacroix   et al., 199 1997; 7;  Verhey  V erhey   et al., 197 1979), 9), and one inv involv olved ed a gr graph aphite ite et al., 200 calorimete calori meterr (Pa (Palman lmanss   et 004 4). The dos oses es measur mea sured ed we were re com compa pared red wit with h tho those se mea measur sured ed 56

 

 DOSIMETRY   DOSIMETR Y  Table 4.5. Typical relative standard uncertainties,   uc, for an ab absor sorbe bed d dos dose e to wa wate ter r ( Dw) dete determi rminat nation ion wit with h a wate wa ter r (pur (pure e or sa satu tura rate ted d wi with th H2) ca calo lori rime mete ter r in a high-ener highenergy gy photon beam (adapt (adapted ed from Seuntjens Seuntjens and DuSautoy, 2002; reproduced with permission).

Quantity

 

Thermistor calibration Repeatability Specific heat capacity Conduction heat loss correction Field perturbation correction Profile uniformity Positioning Water density Heat defect Combined relative standard uncertainty in  D w

uc  (%)

 

0.20 0.15 0.05 0.15 0.02 0.02 0.10 0.02 0.30 0.43

us usin ing g ot othe herr in inst stru rume ment nts. s. The The pr prot oton on en ener ergi gies es requ re quir ired ed to cr crea eate te an io ion n pa pair ir in ai airr (w-values)

The Bragg–Gray theory was developed in order to cal calcul culat ate e th the e abs absorb orbed ed dos dose e to a med medium ium from ionization measurements in a gas cavity placed in the medium. The following two conditions need to be satisfied in order for the theory to hold (ICRU, 1977; 1989): 1. A ca cavi vity ty in th the e me meas asur urin ing g me medi dium um do does es no nott pertur per turb b the cha charge rgedd-par partic ticle le fiel field, d,   i.e., the charged-particle energy fluence in the cavity is identical to that in the medium, and the cavity has no effect on this energy fluence distribution. 2. Th The e en ener ergy gy de depo posi site ted d in th the e ca cavi vity ty is du due e entirely to the charged particles crossing it. Underr th Unde the e ab abo ove co cond ndit itio ions ns,, th the e rat atio io of th the e abso ab sorb rbed ed do dose se in th the e ad adja jace cent nt me medi dium um to th the e absorbed dose in the cavity is equal to the ratio of  the mass collision electronic stopping powers of the

respective materials.

could be ded could deduce uced d by com compar pariso ison n of cal calori orimet metric ric with ionization-chamber dose determinations.

4.4.1 ICRU proton dosim dosimetry etry protoc protocol ol (ICRU 59)

4.4 REF REFERE ERENCE NCE DOSI DOSIMET METRY RY WITH IONIZATION CHAMBERS HAVING   60Co CALIBRATION COEFFICIENTS

The ICRU dosimetry protocol (ICRU, 1998), which

When pr When proto oton n abs absorb orbed ed dos dose e is det determ ermine ined, d, it is usuall usu ally y mea measur sured ed in som some e ma mater terial ial th that at dif differ ferss from fr om th the e ma mate teri rial al of in inte terres est. t. For ex exam ampl ple, e, meas me asur urem emen ents ts ar are e us usua uall lly y ma made de in a wate terr phantom using a gas-filled ionization chamber constructed with walls of some other material such as plast pla stic ic or gr graph aphite ite The absorbed absorbed dose to wa water ter is then inferred inferred from the res respons ponse e of the dosi dosimeter meter,, which whi ch is not composed composed of wa water ter.. In th this is cas case, e, the ioni io niza zati tion on pr prod oduc uced ed in th the e fil filli ling ng ga gass mu must st be rela re late ted d to th the e ab abso sorb rbed ed do dose se in th the e ma mate teri rial al of  interest. Interactions in the chamber walls and sur-

is restated here, recommends that ambientambi ent-air-fi air-filled lled cylin cylindrical drical ioniz ionizatio ation n chamb chambers ers with graphite or A-150 TE plastic walls with traceable   60Co calibration factors be used. The emphasis is on prescribing the techniques for absorbed-dose measurem meas urements ents with ioniz ionizati ation on cham chambers bers hav having ing air-kerm airkerma a calibr calibratio ation n coeffi coefficient cients. s. Cha Chambers mbers with 3  volumes of 0.5 cm or gr grea eater ter should should be use used d for large beams (  5 cm diameter). For smaller beams, cham ch ambe bers rs wi with th vo volu lume mess of    0.1 cm3 sho should uld be used. Absorbed dose should be measured in a water phant ph antom om or in oth other er ma mater terial ialss tha thatt ar are e clo close se to tiss ti ssue ue in el elec ectr tron on de dens nsit ity y. Wh When en me meas asur urin ing g in materials other than water, the depth of measure-

rounding round ing med media ia pr produ oduce ce lit little tle dir direct ect res respon ponse se in the gas. The relationship between absorbed dose in the gas and absorbed dose in the medium of interest is determined by the ratio of the mass electronic stop st oppi ping ng po powe wers rs fo forr th the e tw two o ma mate teri rial alss (I (ICR CRU, U, 1998). The response of a radiation detector to a proton beam be am is al also so in influ fluen ence ced d by th the e ge geom omet etry ry of th the e detector. This is because the energy deposited by a particle in the detector depends on both the effective tiv e st stopp opping ing po powe werr of the med medium ium and the pa path th len le ngt gth h of th the e part rtic icle le in the med ediu ium m. As an example, exam ple, the resp response onse of cylin cylindrical drical and para parallel llel plat pl ate e io ioni niza zati tion on ch cham ambe bers rs of th the e sa same me vo volu lume me would be expected to differ because of differences in mean path length (Bichsel, 1995).

ment should be scaled to the equivalent depth in water using measured equivalent depths (Schneider   et al., 2002a) or the csda ranges given in ICRU (1993a). The effective beam energy should be determined from the residual csda range, which is defined for this purpose as the distance in water between the point of measurement and the practical range,  i.e., the depth beyond the Bragg peak at which the dose falls to 10 percent of its maximum  value (see Fig. 4.1). The preferred reference point forr ca fo cali libr brat atio ion n is th the e ce cent nter er of th the e mo modu dula late ted d spread-out Bragg peak (SOBP). The absorbed absorbed dose to water water in a pr proto oton n bea beam, m,  Dw,p, mea measur sured ed wit with h an amb ambien ientt air air-fil -filled led ion ioniza iza-tion chamber having an air-kerma calibration coefficientt [assu ficien [assuming ming constant constant values for ( sw,air) p   and 57

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

atten tenua uatio tion n and sca scatte tterr in th the e wa wall ll and bui buildld-up up at cap in the   60Co calibration beam,   Aion  is the correction factor for ion recombination in the gas of the ionization chamber when exposed in air in the   60Co calibrat calib ration ion beam,   swall,air   is th the e mea ean n ratio of  resstr re tric icte ted d ma mass ss sto topp ppin ing g po pow wer erss of th the e wal alll material mat erial to air for the secondary secondary electrons electrons gener gener-ated by the   60Co calibration photons, ( m en / r  r)   air, wall is the mean ratio of air-to-wall mass energy absorption tio n coe coeffi fficien cients ts for th the e cal calibr ibrat ation ion   60Co gam gamma ma rays,  K hum  is the correction factor to account for the hum difference in response between ambient air and dry air,, (sw,air) p   is th air the e ra ratio tio of wa water ter-to -to-ai -airr mas masss ele elecctronic stopping powers in the proton beam, ( wair /  e) p is th the e me mean an en ener ergy gy to fo form rm an io ion n pa pair ir in th the e

 e)c ionizatio ioniza tion-c n-cham hamber ber gas for pr proto otons, ns, and (W air air /  is the mean energy to form an ion pair in the th e io ioni niza zati tion on-c -cha hamb mber er ga gass fo forr th the e ca cali libr brat atio ion n 60 Co beam (see Section 4.4.4 below for full discus w  and  W ). sion and definitions of  w ). If an exposure calibration factor ( N  X ) is used,   N  K  can be calculated from

 

 N  K   ¼  N  X 

W air air  e

ð1  gÞ1 ð2:58 c

 104 C R1 kg1 Þ:

Figure Figu re 4. 4.1. 1. (a (a)) Pe Perrce cent ntag age e de dept pth–dos h–dose e di dist stri ribu buti tion on fo forr a 200 MeV proton beam, illustrating the ‘plateau’ region and the Bragg Br agg pea peak; k; (b) per percen centag tage e dep depth–dose th–dose dis distri tribu butio tion n for a modu mo dula late ted d pr prot oton on be beam am.. In Indi dica cate ted d on th the e fig figur ure e ar are e th the e reference depth   zref   (middle of the SOBP), the residual range at  zref  used to specify the quality of the beam  R res and the practical range  R p. Adapted from IAEA (2000).

 

ð4:8Þ

If the ionization chamber is calibrated in terms of abs absorb orbed ed dos dose e to wa water ter in th the e cal calibr ibrat ation ion   60Co beam ( N  D,w,  D,w,c c), the dose to water,  D w,p, is then given (ICRU, 1998) by corr  Dw; p  p  ¼  M  p   N  D;w;c k p ;

 

ð4:9Þ

 

ð4:10Þ

(wair /  e) p] is given (ICRU, 1998) by where corr  Dw; p  p  ¼  M  p   N  D; g  g C p ;

ð4:5Þ

 

k p  ¼

  N  K ð1  gÞ Awall Aion  g  ¼ swall ;air ðm  =r Þ  N  D; g hum ; en air;wall K hum

C p  ¼ ðsw;air Þ p

ðwair = eÞr  ; ðW air air = eÞc

   

ð4:6Þ

ðsw;air Þ p ðwair = eÞ p : ðsw;air Þc ðW air air = eÞc

The errors and omissions in the ICRU 59 dosimetry protocol are addressed in Section 4.4.5.

ð4:7Þ  4.4.1.1

where   M  pcorr is th the e pr prod oduc uctt of th the e el elec ectr trom omet eter er reading   M   corrected corrected for ion reco recombina mbination tion,, temp temp-erature, pressure, and all the factors that produce a modified modi fied resp response onse relative relative to the calib calibrat ration ion conditions diti ons (IAE (IAEA, A, 1997a 1997a), ),   N  D,g   is the absor absorbedbed-dose dose calibration coefficient,   N  K    is the air air-k -kerm erma a cal calii  60 brat br ation ion coe coeffi fficie cient nt for Co un unco corr rrec ecte ted d fo forr io ion n recombination,  g  is the fraction of kinetic energy of  secondary charged particles that is lost to radiative processes proc esses (bremsstra (bremsstrahlun hlung) g) in air [0.0 [0.003 03 for   60Co (Boutillon, 1987)],   Awall   is the correction factor for

Physical quantities for ICRU 59

The large largest st unce uncertain rtainty ty in absor absorbedbed-dose dose deter determimination is in the value of ( wair /  e) p. This issue is discussed fully in Section 4.4.4. AAPM (AAPM, 1986) and ECHED (Vynckier  et al., 1991; 1994) used two different values: 34.3 J C 21 and 35.2 J C21, respectively. A compromise value of (34.8 + 0.7) J C21 for ambient air was recommended in ICRU Report 59 (ICRU (IC RU,, 199 1998). 8). Thi Thiss val value ue is con consis sisten tentt wit with h bot both h dire di rect ct me meas asur urem emen ents ts an and d th thos ose e in infe ferr rred ed fr from om comparison between ionization chamber and calori e)c, a value of   metric met ric mea measur sureme ement nts. s. Fo Forr (W air air /  58

 

 DOSIMETRY   DOSIMETR Y 

(33.97 + 0. 0.05 05)) J C21 for dry air is rec recomm ommend ended ed (Bout (Bo utill illon on and Pe Perr rroch oche-R e-Roux oux,, 198 1987; 7; CC CCME MERI, RI, et al., 19 1985; 198 5; Ni Niat atel el   et 1985 85). ). Ho Howe weve verr, if hu humi mid d (ambie (am bient) nt) air is use used d in the ion ioniza izatio tion n cha chambe mberr   60 in th the e Co bea beam, m, th this is val value ue sho should uld be re redu duced ced to 33. 3.7 77 J C21 (I (ICR CRU, U, 19 1998 98;; Ro Roge gers rs an and d Ro Ross ss,, al.., 19 1988; 198 8; Sch Schulz ulz   et al 1986 86)) wi with th an un unce cert rtai aint nty y 21   + of  0.0 .05 5J C (0.1 (0 .15 5 pe perc rcen ent) t) (B (Bou outi till llon on an and d Perr Pe rroch oche-R e-Roux oux,, 198 1987). 7). Wh When en us using ing hu humid mid air in the ionization chamber in the   60Co beam, a correction factor of   K hum 0.997 (CCMERI, 1977; ICRU, hum ¼

beam of qu beam qual alit ity y   Q0   and and in th the e ab abse senc nce e of th the e chamber is given by  Dw;Q0   ¼  M Q0 N  D;w;Q0 :

 

ð4:11Þ

The ab The abso sorb rbed ed do dose se to wa wate terr at th the e re refe fere renc nce e depth   zref   in water, in a proton beam of quality   Q and in the absence of the chamber, is given by  Dw;Q  ¼  M Q N  D;w;Q0 kQ;Q0 ;

 

ð4:12Þ

1979; Schulz   et al., 1986) [Equation (4.6)] must be  e)c  for humid air is applied. The ratio ( wair /  e) p  /( W air air /  1.031 + 2.0 per percen centt (IC (ICRU RU,, 199 1998; 8; Jon Jones, es, 200 2001d) 1d).. The rat ratio io of the water-towater-to-air air mass electronic electronic stop stop-ping pi ng po pow wer erss fo forr pr prot oton onss is ta tak ken fr from om IC ICRU RU (1993a) and can be considered constant within 1.2 percent (ICRU, 1998) in the energy range of interest here. A value of 1.133 can be assumed (Jones, 2001d). The values of the other physical quantities and an d th thei eirr rel ela ati tiv ve un unce cert rtai aint ntie iess ar are e gi give ven n in Tab able le 4.6 (J (Jo one nes, s, 200 001 1d; Vync ncki kier er,, 19 199 95) 5).. Quan Qu anti titi ties es no nott gi give ven n in th the e ta tabl ble e ( e.g.,   Aion) ar are e assumed to have values of 1 with negligible uncer-

where   M  is the reading of the dosimeter with the refer re ferenc ence e poi point nt of the cha chambe mberr pos positi itione oned d at   zref , corrected for the influence quantities pressure and temper tem perat atur ure, e, ele electr ctrome ometer ter cal calibr ibrat ation ion,, pol polari arity ty effect and ion recombination (IAEA, 1997a; 2000),  N  D,w, the e cal calibr ibrat ation ion coe coeffi fficien cientt in ter terms ms of   D,w,Q Q0 is th absorb abs orbed ed dos dose e to wa water ter for the dosimete dosimeterr at th the e 60 reference quality   Q0   ( Co),   kQ,Q0 is th the e cha chambe mberr specific factor that corrects for differences between the th e ref refere erence nce bea beam m qua qualit lity y   Q0   and and th the e ac actu tual al quality being used   Q.  Reference conditions for the determ det ermina inatio tion n of abs absorb orbed ed dos dose e in pr proto oton n bea beams ms are ar e giv given en in Tabl able e 4.7 4.7.. Wat ater er is re recom commen mended ded as

tai tainti nties. For rsorb ioniza ion izatio tion n cha chambe rs, cal calibr ated ede in term te rms ses.of Fo abso ab rbed ed dose do se to mbers wat ater er, the th eibrat valu va lue of  (sw,air)c  is taken as 1.134 (ICRU, 1998).

th the e re refer ferenc ence e and medium med ium for the det determ ermina inatio tion n of  absorbed dose for beam-quality measurements. The bea beam-q m-qual uality ity cor correc rectio tion n fa facto ctorr   kQ,Q0 is defined as the ratio, at the qualities   Q   and   Q0, of  the th e cal calibr ibrat ation ion coe coeffic fficien ients ts in ter terms ms of abs absorb orbed ed dose to water of the ionization chamber:

4.4.2 IAEA proton proton dosime dosimetry try code code of prac practice tice (TRS 398)

This code of practice (IAEA, 2000) is based solely upon upo n th the e use of ion ioniza izatio tion n cha chambe mbers rs ha havin ving g cal caliibration bra tion coefficients coefficients spec specified ified in terms of abso absorbed rbed dose do se to wa wate terr in a re refe fere renc nce e be beam am of qu qual alit ity y   Q0 (60Co). The absorbed dose to water at the reference depth,   zref , in wa wate terr fo forr a re refe fere renc nce e ca cali libr brat atio ion n

kQ;Q0   ¼

  N  D;w;Q   Dw;Q = M Q ¼ :  N  D;w;Q0  Dw;Q0 = M Q0

 

ð4:13Þ

Fie ield ld siz ize es sh shou ould ld be 10  10 cm2 an and d the phantom should extend at least 5 cm beyond all four side si dess of the field size em emplo ploye yed d at the dep depth th of 

Table 4.6. Physical Table Physical quant quantities ities and rela relative tive standard unce uncertain rtainties ties for absorb absorbed ed dose deter determina minations tions using A-150/air A-150/air ionization ioniz ation chamb chambers ers with   60Co air-kerma calibration coefficients [Jones (2001d), using data from AAPM (1983), Gastorf  et al.  (1986), IAEA (1997a), Nath and Schulz (1981), Vynckier (1995), and Vynckier  et al.  (1991, 1994)].

Quantity

Symbol

Value

  M  Electrometer reading Product of electrometer reading correction factors   N  K   Air-kerma  Air-kerm a calibration coefficient  Absorption and scatterin scattering g correc correction tion factor in A-150 in calibra calibration tion  60 Co beam   Awall   swall, g Mean ratio of mass electronic stopping powers in calibration    g 60 Co beam (A-150/air) Mean ratio of mass energy absorption coefficients in calibration [(m en / r  r)  air,wall ]c 60 Co beam (air/A-150) Mean ra ratio tio of mass elec electro tronic nic stop stopping ping pow powers ers in pro proton ton beam (wa (water/ ter/air) air) (sw,air) p  

 

0.983–0. 0.98 3–0.99 992 2 1.145

0.1 0.2 1.0 0.1 0. 1 0.1

0.906

0.1

1.133a

1.2

 

 

(wair /  e) p /(W air /  e)c   1.031 K hum   0.997 hum

Ratio of en Rat ener ergy gy req equ uir ired ed to prod odu uce an ion pa paiir in amb ambie ien nt ai airr   Humidity correction factor Combined relative standard uncertainty

Relative standard uncertainty uncertaint y (%)

2.0 0.1 2. 6

a

 Assuming a constan constantt value.

59  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  Table 4.7. Refe Table Referenc rence e condi conditions tions for the deter determinat mination ion of  absorbed dose in proton beams (IAEA, 2000).

Quantity

Phantom material

Reference value or reference characteristics Water 22

Th The e pr proto otonn-bea beam m qua qualit lity y ind index ex is the residu residual al 22 range   Rres   ( g cm ) in water and is defined (Fig. 4.1) as  Rres  ¼  R p   zref :

 

ð4:14Þ

where   R p   is the practical range (g cm22), namely namely,,

Chamber type

For  R res  0.5 g cm , cylindrical and plane parallel chambers For R res , 0.5 g cm22, plane parallel chambers

the depth of the 10 percent level of the peak dose on the distal edge of the Bragg peak, and  z ref  is the refere ref erence nce dep depth th for the mea measur sureme ement nt (m (midd iddle le of  SOBP). Reference conditions for the determination of proton-beam quality are given in Table 4.8. al..   (1996 Laitano   et al (1996)) and Lai Laitan tano o and Ro Roset setti ti (2000) have calculated stopping powers in water and TE materials averaged averaged over the proton energy spectra at va vari riou ouss po posi siti tion onss in un unmo modu dula late ted d an and d ra rang ngeemodulated proton beams of the same maximum energies (50–300 MeV). They found relative differences of  up to 20 percent between   absolute   stopping powers calc ca lcul ulat ated ed at th the e sa same me de dept pth h in mo modu dula late ted d an and d unmo un modu dula late ted d bea beams. ms. Th They ey al also so fo foun und d neg neglig ligibl ible e differences in the case of stopping-power ratios, which have hav e litt little le ener energy gy dep depende endence. nce. Med Medin in and And Andreo reo

Measurement depth, z ref    Middle of the SOBPa Measurement Reference point of the For plane-parallel chambers, on the chamber inner surface of the window at its centres For cylindrical chambers, on the central axis at the center of the cavity volume Position of the reference For plane-parallel and cylindrical point of the chamber chambers, at the point of  measurement measure ment depth z ref  SSD Clinical treatment distance Field size at the phantom 10  10 cm2, or that used for surface normalization of the output factors whichever is larger For small field applications ( i.e., eye treatments), treatm ents), 10  10 cm2 or the largest field clinically available

(1992) and Palmans and Verhaegen (1998) calculated the dif differ ference encess in sto stoppi pping-p ng-pow ower er ra ratios tios betw between een a modula mod ulated ted beam and a mon monoene oenerget rgetic ic pr proton oton beam with the same residual range. The effects are small and can safely be neglected (Palmans and Vynckier, 2002). The use of residual range or energy is therefore an approp appropriate riate beam-quality specifier for ionizat ionizationionchamber-based proton dosimetry. Ideally   kQ,Q0 should be obtained from Eq. (4.13) by direct measurement of the absorbed dose at the qualities   Q   and   Q0. However, as no primary standar ard ds of ab abssor orbe bed d do dose se to water fo forr prot oto ons   k are ar e av avail ailabl able, e, all val values ues of  Q,Q0 ar are e der derive ived d by

a

The reference depth can be chosen in the ‘plateau region’, at a depth of 3 g cm22, for clinical applications with monoenergetic proton beams ( e.g., for plateau irradiations).

measurement and also extend to at least 5 g cm 22 beyo be yond nd th the e ma maxi ximu mum m de dept pth h of me meas asur urem emen ent. t. Kanai   et al.  (2004) have assessed these field- and phantom-size requirements in a 235 MeV clinical beam be am wi with th a 6 cm SO SOBP BP. Th They ey fo foun und d th that at th the e lateral dimensions of the water phantom need not exceed field Indeed, there was discrepancy ofthe only 0.5 size. percent if a phantom of 5a  5 cm c m2 cros cr osss-se sect ctio ion n is used in a 10  10 cm2 field. Nevertheless, the 5 cm margin should be adhered to as these latter measurements were made under specific conditions conditions.. For ho hori rizo zont ntal al be beam ams, s, th the e wa wate terr ph phan anto tom m entrance window should be made of 2–5 mm thick poly(methyl methacrylate) (PMMA) or polystyrene. To calcu calculate late the wat water-eq er-equiva uivalent lent thic thickness kness,, mass dens de nsit itie iess of 1. 1.19 19 an and d 1. 1.06 06 g cm23, resp respectiv ectively ely (ICR (I CRU, U, 19 1984 84), ), sh shou ould ld be us used ed (I (IAE AEA, A, 20 2000 00). ). If  requi re quired red,, a clo closese-fit fittin ting g wa water terpr proofi oofing ng co cove verr of  PMMA PMM A with a thick thickness ness not exceeding exceeding 1 mm shou should ld

Table 4.8. Refe Table Referen rence ce condit conditions ions for the determ determinat ination ion of  proton beam quality: residual range ( Rres) (IAEA, 2000).

Quantity

Pha han nto tom m mate terria iall Chamber Reference point of the chamber

Water Cylindrical and plane parallel For plane-parallel chambers, on the inner surface of the window at its center For cylindrical chambers, on the central axis at the center of the cavity volume Position of the reference For plane-parallel and cylindrica cylindricall point of the chamber chambers, at the point of interest SSD Clinical treatment distance Field size at the phantom 10 cm  10 cm surface For small field applications ( i.e., eye treatments), 10 cm 10 cm or the largest field clinically available

be used. There should be an air gap of 0.1–0.3 mm between the chamber and the cover to allow the air pressure and temperature in the chamber to equilibrate (IAEA, 2000).

60  

 DOSIMETRY   DOSIMETR Y 

cal calcul culat ation ion (IA (IAEA, EA, 200 2000). 0). Th These ese cal calcul culat ation ionss are   60 based bas ed on Co gamma rad radiati iation on as the refe referen rencece Q  et al. beam quality 0  (Andreo, 1992; Medin , 1995):

Reference value or reference characteristics

kQ;Q0   ¼

ðsw;air ÞQ ðwair = eÞQ  pQ ; ðsw;air ÞQ0 ðW air air = eÞQ0  pQ0

 

ð4:15Þ

where   sw,air   is th the e wa water ter-to -to-ai -airr st stopp opping ing-p -pow ower er ratio,   w  /  e   and  W   /  e  are the differential and inteair air gral mean energies, energ ies, respectiv resp ectively ely,, exp expende ended d in dry air per ion pair formed (see Section 4.4.4 for full discussion discu ssion and defin definition itionss of   w   and   W ), ),   p   is th the e product of ionization-chamber perturbation factors, which includes all departures from the behavior of  an ideal Bra Bragg– gg– Gra Gray y dete detector: ctor:  p ¼  pcav pcel pdis pwall ;

 

Figure 4.2. Water-to ater-to-air -air stopping-power stopping-power ratios for proton beams as a function of beam quality index,   Rres, calculated using Eq. (4.1 (4 .17) 7).. Th The e cu curv rve e is a fit (I (IAE AEA, A, 20 2000 00)) to mo mono noen ener erge geti ticc stopping stop ping-po -power wer rat ratios ios (Med (Medin in and Andr Andreo, eo, 1997 1997a; a; 1997 1997b). b). The data include the tra transpo nsport rt of seco secondar ndary y elec electro trons ns and nucl nuclear ear inelast inel astic ic pro processe cesses, s, and the basic stopping stopping pow powers ers are taken from ICRU (1993a).

ð4:16Þ

where   pcav   is the factor that corrects the response of an ionization chamber for effects related to the air cavity, predominantly the in-scattering of electrons and heavy charged particles that makes the

1997a; 1997 1997a; 1997b). b). The basic prot proton on stop stopping ping powers are ar e ta tak ken from IC ICR RU (1 (199 993a 3a)). The PETRA  stopping-power ratios include the transport of secondary electrons and heavy charged particles produced in nuclear non-elastic processes, which is not the case for the ICRU stopping powers. The statistical uncertainty of (sw,air)Q  is estimated to be 0.2 percent per cent (Med (Medin in and And Andreo, reo, 1997b). The rela relative tive uncer un certai tainty nty of th the e st stopp opping ing-po -powe werr ra ratio tioss at th the e refer re ferenc ence e dep depth th in a cli clinic nical al pr proto oton n bea beam m is es estitimate ma ted d to be 1 pe perc rcen ent. t. Th The e va valu lue e of (sw,air)Q0 for 60 Co is taken as 1.133 + 0.5 percent (IAEA, 1997a; al.. 1997b) 199 7b),, whi which ch wa wass cal calcul culat ated ed by An Andr dreo eo   et al (1986) using Monte Carlo methods and the mono-

particle partic le flue fluence nce ins inside ide the cav cavity ity dif differ ferent ent fr from om thatt in the medium tha medium in the absence absence of the cavity cavity,,  pcel   is the factor factor th that at cor correc rects ts the re respo sponse nse of an ionization chamber for the effect of the central electrode trod e (lac (lack k of air equiv equivalence alence)) duri during ng in-p in-phant hantom om meas me asur urem emen ents ts (for pl plan anee-pa para rall llel el ioni io niza zati tion on chambers,   pcel   is not appl applicable icable), ),   pdis   is the factor that accounts for the effect of replacing a volume of  water with the detector cavity when the reference point of the chamber is taken to be at the chamber center—it is the alternative to the use of an effective ti ve po poin intt of me meas asur urem emen entt of th the e ch cham ambe berr (for plane-par plan e-parallel allel ioniz ionizati ation on cham chambers, bers,   pdis   is is not applicable), and   pwall  is the factor that corrects the respon resp onse se of an io ioni niza zati tion on ch cham ambe berr fo forr th the e no nonnwater wa ter equivale equivalence nce of th the e cha chambe mberr wa wall ll and any waterproofing waterpr oofing material.  4.4.2.1

energetic electron data tabulated in ICRU ICR U (19 (1984) 84).. An stopping-power additi add itiona onall un uncer certai tainty nty of + 0.1 percent is assigned to account for spectrum differences among   60Co beams. To assess (wair /  e)Q   for dry air in TRS 398 (IAEA, 2000), only the values of ( W air /  e)Q  and (wair /  e)Q  given in ICR ICRU U (19 (1998) 98) wer were e cons conside idered red (Se (Secti ction on 4.4. 4.4.4). 4). A  proc pr oced edur ure e us usin ing g we weig ight hted ed me medi dian ans, s, ta taki king ng in into to accou ac count nt the sta statis tistic tical al unc uncert ertaint ainty y of eac each h val value ue (Mu ¨ ller ller,, 200 2000a; 0a; 2000 2000b), b), yiel yielded ded the valu value e of (wair /   e)Q 34.23 J C21, with a rel relat ativ ive e st stand andard ard unc uncerertainty tai nty of 0.4 per percent cent.. The consisten consistency cy of the da data ta used in the evalua evaluation tion (Mu ¨ ller ller,, 2000a) is discu discussed ssed  60  W air  e in Jones (2006). The value of  W   /   for dry air in Co air

Physical quantities for TRS 398

For proton dosimetry dosimetry,, the calculated beam-quality corre cor recti ction on fa facto ctors rs giv given en in TR TRS S 398 (IA (IAEA EA,, 200 2000) 0)   60 are based on a calibration in Co. The values for (sw,air)Q   are describe described d as a fun functi ction on of the proton proton beam-quality specifier  R res (Fig. 4.2):

ðsw;air ÞQ  ¼ a þ bRres þ

 c ;  Rres

 

¼

ð4:17Þ

2

25

is 33.97 J C 1 wit with h a rel relat ative ive unc uncert ertain ainty ty of   +0.2 percent, although the uncertainty quoted in the original publication (Boutillon and Perroche-Roux, 1987) is  + 0.05 J C21 ( +0.15 percent). The range of values of the ionization-chamber perturbation factors   p   and their rela relative tive uncerta uncertainties inties

where   a 1.137,   b 2 4.265  10 , and   c 1.840  1023 (IAEA, 2000). This equation is derived as a fit to the monoenergetic get ic st stopp opping ing-p -pow ower er ra ratio tioss cal calcul culat ated ed usi using ng the Mont Mo nte e Ca Carl rlo o co code de PE PETR TRA A (M (Med edin in an and d An Andr dreo eo,, ¼

¼

¼

61  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

Table 4.9. Range of perturbation factors and relative standard uncertainties for cylindrical and plane-parallel ionization chambers (IAEA, 2000).

Factor

Beam quality Q0  / c  ( 60Co)

 

 Value  V alue

 pcav  pcela

 

 pdisa  pwall

 

1.000 0.993b 1.000 0.981– 0.98 1– 0.99 0.993 3 (cc) 1.000 (ppc) 0.977–1.013c (cc) 0.989– 0.98 9– 1.02 1.024 4 ( ppc)

Q /  p  (protons)

Relative standard Relative uncertainty uncertain ty (%)  

 Value  V alue

Relative standard Relative uncertainty uncertain ty (%)

0.1 0.2

1.000 1.000

0.3 0.4

0.3 0.2 0.5 1.5

1.000

0.2

1.000

0.6

,

cc, cylindrica cylindricall chamber; ppc, plane-par plane-parallel allel chamber. a Not applicable to plane-parallel chambers. b For cylindrical chambers with 1 mm diameter aluminum central electrodes. c Includes a 0.5 mm thick PMMA water waterproofing proofing sleeve.

in   60Co and proton beams are given in Table 4.9. All

w-values related to humid (ambient) air or dry air

the factors for proton beams are taken as 1.000, but uncertainties are assigned to these values. Calculat Calc ulated ed valu values es of   kQ,Q0 for proton beams, for  various cylindrical and plane-paralle plane-parallell ionization cham ch ambe bers rs,, an and d fo forr se sele lect cted ed va valu lues es of th the e be beam am quality   Rres   are given in Table 4.10 (IAEA, 2000).  Values  V alues of    kQ,Q0 fo forr so some me co comm mmon on io ioni niza zati tion on chambers are plotted in Fig. 4.3. In Table 4.11, the uncertainty estimates are summarized and show a combined standard relative uncertainty in   kQ,Q0 for proton beams of 1.7 and 2.1 percent for cylindrical and plan plane-pa e-parall rallel el ioniz ionizatio ation n chamb chambers, ers, resp respectectively ivel y. The large largest st comp componen onents ts of this unce uncertain rtainty ty  s w,air and  p wall  for ionization are the uncertainties of  s

condi con dittio ion ns an and d the humid idit ity y co corr rrec ecti tio on fo forr air-kermaair-k erma-based based dete determina rmination tionss of absor absorbed bed dose to wate terr. In IC ICRU RU 59 (I (ICR CRU, U, 19 1998 98), ), th the e rec ec-  60  e ommended ommen ded value of   W air  /    in the Co calibration air beam is (33.77 + 0.05) J C21 (ICR (ICRU, U, 1998; Schu Schulz lz  et al., 1986), corresponding to humid air, which is  e   value of (33.97 + 0.05) J C21 for based on a   W air air /  dry air (Boutillon and Perroche-Roux, 1987). Note that th at th the e un unce cert rtai aint ntie iess ar are e in inco corr rrec ectl tly y gi give ven n in 21 ICRU 59 (ICRU, 1998) as 0.15 J C . The   W-   or   w-value (see Section 4.4.4) for air at the sta standar ndard d temp tempera eratur ture e and pres pressure sure (STP) of  208C, 101.325 kPa and 50 percent relative humidity is 0. 0.6 6 pe perc rcen entt lo lowe werr th than an th that at fo forr dr dry y ai airr at th the e same sa me te temp mper erat atur ure e an and d pr pres essu sure re (I (ICR CRU, U, 19 1979 79;; Niatel Nia tel,, 196 1969). 9). It is ass assume umed d tha thatt thi thiss ra ratio tio is the same for all radiations. Thus, for the same amount of energy available for creating charge, 0.6 percent more mor e cha charge rge will be cre creat ated ed in air at 50 per percen centt relative humidity than in dry air at STP. However, the re respo sponse nse of a ca cavit vity y ion ioniza izatio tion n cha chambe mberr wil willl also als o de depen pend d on the degree degree to whi which ch the incident incident charged particles deposit energy,   i.e., on the stopping power, which is different for humid air than for dry air. For the secon secondary dary electrons electrons gene generat rated ed   60 in a Co be beam am,, th the e co comb mbin ined ed ef effe fect ct is th that at 0. 0.3 3 percent more charge is created in air at STP and 50

chambers in the prot chambers proton on beam. Estimated Estimated rela relativ tive e uncertain uncer tainties ties in absor absorbedbed-dose dose dete determin rminatio ations ns are given in Table 4.12. The combined relative standard uncertainties for cylindrical and plane-parallel ionization chambers at the reference depth for clinical proton beams are 2.0 and 2.3 percent, respectively.  Values  V alues of   kQ,Q0 or   N  D,w,Q   can can be ob obta tain ined ed fo forr differ dif ferent ent ion ioniza izatio tion n cha chambe mbers rs in pr proto oton n bea beams ms by ‘cr ‘cross oss-ca -calib libra ratio tion’ n’ tec techn hniqu iques. es. The res respec pectiv tive e  values for a reference ionization chamber are calculated or derived from calorimetric measurements in the use user’s r’s pr proto oton n bea beam. m. By com compar paring ing mea measur sureement me ntss at th the e re refe fere renc nce e po poin int, t, va valu lues es fo forr ot othe herr al.., 20 cham ch ambe bers rs ca can n be de dedu duce ced d (K (Kan anai ai   et al 2004 04;;  Vatnitsky  V atnitsky et al., 1996a; 2002).

  60 tive humidity perc percent ent relative humidit y than in dry factor air at K  STP. Thus, forrela Co, the humidity correction hum hum thatt cor tha correc rects ts the am ambie bient nt air ion ioniza izatio tion n cur curren rentt  I hum hum   to the dry-air ionization current   I dry dry   is 0.997 (ICRU, 1979; 1998; Schulz  et al., 1986). Usually standards laboratories provide air-kerma ( N K  K)   calibration coefficients for ionization chambers

4.4.3 Cons Considera iderations tions conce concerning rning dry and humid air al..   (2000) Medin   et al (2000) have have dis discus cussed sed in det detail ail th the e imp im por orta tan nce of the co con nsis sisten ency cy in the use of  62  

 DOSIMETRY   DOSIMETR Y  Table 4.10. 4.10. Calculated values of  k  k Q,Q0 for proton beams, for various cylindrical and plane-parallel ionization chambers as a function of beam qualit quality y  R res  (IAEA, 2000).

Ionization chamber typea

Beam quality,  R res (g cm22) 0.25 0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

7.5

10

15

20

30

Cylindrical chambers

Capint Capi ntec ec PR PR-0 -05P 5P mi mini ni Capi Ca pint ntec ec PR PR-0 -05 5 mi mini ni Capint Cap intec ec PRPR-06C 06C/G /G Fa Farm rmer er Exrrad Ex adin in A2 Sp Spok okas as Exrrad Ex adin in T2 Sp Spok okas as Exrrad Ex adin in A1 mi mini ni Sh Shon onka ka Exrrad Ex adin in T1 mi mini ni Sh Shon onka ka Exrrad Ex adin in A1 A12 2 Far arme merr Farr Wes Fa estt Tec ech h IC IC-1 -18 8 FZH TK 01 Nucl Nu clea earr As Asso socc 30 30-7 -750 50 Nucl Nu clea earr As Asso socc 30 30-7 -749 49 Nucl Nu clea earr As Asso socc 30 30-7 -744 44 Nucl Nu clea earr As Asso socc 30 30-7 -716 16 Nuclear Assoc 30-753 Farmer shortened Nuclea Nuc learr Ass Assoc oc 3030-751 751 Fa Farm rmer er Nuclea Nuc learr Ass Assoc oc 3030-752 752 Fa Farm rmer er NE 2515 NE 2515/3 NE 2577 NE 25 2505 05 Far arme merr NE 25 2505 05/A /A Far arme merr NE 25 2505 05/3 /3,, 3A Fa Farm rmer er NE 25 2505 05/3 /3,, 3B Fa Farm rmer er NE 25 2571 71 Far arme merr NE 25 2581 81 Far arme merr NE 256 2561/2 1/261 611 1 Sec Sec.. Sta Standa ndard rd PTW PT W 23 2332 323 3 mi micr cro o PTW PT W 23 2333 331 1 ri rigi gid d PTW PT W 23 2333 332 2 ri rigi gid d PTW PT W 233 3333 33 PTW PT W 30 3000 001/ 1/30 3001 010 0 Fa Farm rmer er

– – – – – – – – – – – – – – –

1.046 1.04 6 1. 1.04 045 5 1. 1.04 044 4 1. 1.04 044 4 1. 1.04 044 4 1.04 1. 046 6 1. 1.04 045 5 1. 1.04 044 4 1. 1.04 044 4 1. 1.04 044 4 1.038 1.0 38 1.0 1.037 37 1.0 1.036 36 1.0 1.036 36 1.0 1.036 36 1.05 1. 057 7 1. 1.05 055 5 1. 1.05 054 4 1. 1.05 054 4 1. 1.05 054 4 1.02 1. 020 0 1. 1.01 018 8 1. 1.01 018 8 1. 1.01 018 8 1. 1.01 017 7 1.04 1. 045 5 1. 1.04 043 3 1. 1.04 043 3 1. 1.04 042 2 1. 1.04 042 2 1.00 1. 009 9 1. 1.00 007 7 1. 1.00 007 7 1. 1.00 006 6 1. 1.00 006 6 1.04 1. 043 3 1. 1.04 042 2 1. 1.04 041 1 1. 1.04 041 1 1. 1.04 041 1 1.00 1. 007 7 1. 1.00 006 6 1. 1.00 005 5 1. 1.00 005 5 1. 1.00 005 5 1.032 1.031 1.030 1.030 1.030 1.03 1. 037 7 1. 1.03 035 5 1. 1.03 034 4 1. 1.03 034 4 1. 1.03 034 4 1.04 1. 041 1 1. 1.03 039 9 1. 1.03 039 9 1. 1.03 038 8 1. 1.03 038 8 1.04 1. 041 1 1. 1.03 039 9 1. 1.03 039 9 1. 1.03 038 8 1. 1.03 038 8 1.04 1. 041 1 1. 1.03 039 9 1. 1.03 039 9 1. 1.03 038 8 1. 1.03 038 8 1.041 1.040 1.040 1.039 1.039 1.039 1.038

1.043 1.04 3 1. 1.04 043 3 1.04 1. 043 3 1. 1.04 043 3 1.036 1.0 36 1.0 1.035 35 1.05 1. 054 4 1. 1.05 054 4 1.01 1. 017 7 1. 1.01 017 7 1.04 1. 042 2 1. 1.04 042 2 1.00 1. 006 6 1. 1.00 006 6 1.04 1. 041 1 1. 1.04 040 0 1.00 1. 004 4 1. 1.00 004 4 1.030 1.029 1.03 1. 034 4 1. 1.03 034 4 1.03 1. 038 8 1. 1.03 038 8 1.03 1. 038 8 1. 1.03 038 8 1.03 1. 038 8 1. 1.03 038 8 1.038 1.038 1.038

– – – – – – – – – – – – – – – – –

1.037 1.0 1.037 1.036 36 1.0 1.035 35 1.0 1.035 35 1.044 1.0 44 1.0 1.042 42 1.0 1.041 41 1.0 1.041 41 1.033 1.032 1.031 1.031 1.043 1.041 1.041 1.040 1.043 1.041 1.041 1.040 1.03 1. 033 3 1. 1.03 032 2 1. 1.03 031 1 1. 1.03 031 1 1.02 1. 021 1 1. 1.01 019 9 1. 1.01 019 9 1. 1.01 018 8 1.04 1. 043 3 1. 1.04 041 1 1. 1.04 041 1 1. 1.04 040 0 1.02 1. 025 5 1. 1.02 023 3 1. 1.02 023 3 1. 1.02 022 2 1.04 1. 043 3 1. 1.04 041 1 1. 1.04 041 1 1. 1.04 040 0 1.02 1. 020 0 1. 1.01 018 8 1. 1.01 017 7 1. 1.01 017 7 1.040 1.0 40 1.0 1.038 38 1.0 1.038 38 1.0 1.037 37 1.02 1. 027 7 1. 1.02 025 5 1. 1.02 025 5 1. 1.02 025 5 1.03 1. 037 7 1. 1.03 035 5 1. 1.03 034 4 1. 1.03 034 4 1.03 1. 031 1 1. 1.02 029 9 1. 1.02 028 8 1. 1.02 028 8 1.0 1. 033 1.0 .031 31 1.0 .031 31 1.0 .03 30 1.03 1. 033 3 1. 1.03 031 1 1. 1.03 031 1 1. 1.03 030 0

PTW 30 PTW 3000 002/ 2/30 3001 011 1 Fa Farm rmer er PTW PT W 30 3000 004/ 4/30 3001 012 2 Fa Farm rmer er PTW PT W 30 3000 006/ 6/30 3001 013 3 Fa Farm rmer er PTW PT W 31 3100 002 2 fle flexi xibl ble e PTW PT W 31 3100 003 3 fle flexi xibl ble e PTW PT W 31 3100 006 6 Pi PinP nPoi oint nt PTW PT W 31 3101 014 4 Pi PinP nPoi oint nt SNC SN C 10 1007 0700 00-0 -0 Fa Farm rmer er SNC SN C 10 1007 0700 00-1 -1 Fa Farm rmer er  Victoreen  Victor een Radocon III 550  Victoreen  Victor een Radocon II 555  Victoreen  Victor een 30-348  Victoreen  Victor een 30-351  Victoreen  Victor een 30-349  Victoreen  Victor een 30-361 Scdx-Wellho¨ fe ferr CC CC01 01 ¨ Scdx-Wellhofe ferr CC CC04 04/I /IC0 C04 4 Scdx-Wellho¨ fer CC CC08/ 08/IC IC05/ 05/IC0 IC06 6 Scdx-Wellho¨ fer CC CC13/ 13/IC IC10/ 10/IC1 IC15 5 Scdx-Wellho¨ fe ferr CC CC25 25/I /IC2 C25 5 Scdx-Wellho¨ fer FC23-C/IC28 Farmer shortened Scdx-Wellho¨ fer FC65FC65-P/IC P/IC69 69 Farme Farmerr Scdx-Wellho¨ fer FC65FC65-G/IC G/IC70 70 Farme Farmerr

– – – – – – – – – – – – – – – – – – – – –

1.036 1.03 6 1. 1.03 035 5 1. 1.03 034 4 1. 1.03 034 4 1. 1.03 034 4 1.04 1. 044 4 1. 1.04 042 2 1. 1.04 041 1 1. 1.04 041 1 1. 1.04 041 1 1.03 1. 033 3 1. 1.03 032 2 1. 1.03 031 1 1. 1.03 031 1 1. 1.03 031 1 1.03 1. 032 2 1. 1.03 030 0 1. 1.02 029 9 1. 1.02 029 9 1. 1.02 029 9 1.03 1. 032 2 1. 1.03 030 0 1. 1.02 029 9 1. 1.02 029 9 1. 1.02 029 9 1.02 1. 027 7 1. 1.02 025 5 1. 1.02 025 5 1. 1.02 024 4 1. 1.02 024 4 1.02 1. 028 8 1. 1.02 026 6 1. 1.02 025 5 1. 1.02 025 5 1. 1.02 025 5 1.03 1. 033 3 1. 1.03 031 1 1. 1.03 030 0 1. 1.03 030 0 1. 1.03 030 0 1.04 1. 044 4 1. 1.04 042 2 1. 1.04 042 2 1. 1.04 042 2 1. 1.04 041 1 1.031 1.030 1.029 1.029 1.028 1.014 1.012 1.012 1.011 1.011 1.023 1.022 1.021 1.021 1.021 1.026 1.024 1.023 1.023 1.023 1.030 1.028 1.027 1.027 1.027 1.023 1.021 1.020 1.020 1.020 1.04 1. 042 2 1. 1.04 040 0 1. 1.04 040 0 1. 1.04 040 0 1. 1.03 039 9 1.03 1. 037 7 1. 1.03 035 5 1. 1.03 035 5 1. 1.03 034 4 1. 1.03 034 4 1.041 1.0 41 1.03 1.039 9 1.03 1.039 9 1.03 1.039 9 1.03 1.038 8 1.041 1.0 41 1.03 1.039 9 1.03 1.039 9 1.03 1.039 9 1.03 1.038 8 1.04 1. 041 1 1. 1.03 039 9 1. 1.03 039 9 1. 1.03 039 9 1. 1.03 038 8 1.042 1.040 1.040 1.039 1.039 1.039 1.039

– –

1.037 1 1.036 .036 1.035 1.035 1.0 1.035 35 1.035 1.035 1.03 1.035 5 1.034 1. 1.034 034 1.034 1.034 1.03 1.034 4 1.034 1.034 1.034 1.034 1.034 1.0 1.033 33 1.033 1.033 1.044 1.042 1.042 1. 1.041 041 1.041 1.041 1.0 1.041 41 1.041 1.041 1.041 1.041 1.041 1.0 1.041 41 1.040 1.040 1.04 1.040 0 1.040 1. 1.040 040 1.040 1.040 1.03 1.039 9

1.035 1.0 1.035 1.035 35 1.041 1.0 41 1.0 1.041 41 1.031 1.031 1.040 1.040 1.040 1.040 1.03 1. 031 1 1. 1.03 031 1 1.01 1. 018 8 1. 1.01 018 8 1.04 1. 040 0 1. 1.04 040 0 1.02 1. 022 2 1. 1.02 022 2 1.04 1. 040 0 1. 1.04 040 0 1.01 1. 017 7 1. 1.01 017 7 1.037 1.0 37 1.0 1.037 37 1.02 1. 024 4 1. 1.02 024 4 1.03 1. 034 4 1. 1.03 034 4 1.02 1. 028 8 1. 1.02 028 8 1.0 .030 30 1. 1.03 030 0 1.03 1. 030 0 1. 1.03 030 0

1.034 1.034 1.041 1.0 41 1.030 1.040 1.040 1.03 1. 030 0 1.01 1. 018 8 1.04 1. 040 0 1.02 1. 022 2 1.04 1. 040 0 1.01 1. 017 7 1.037 1.0 37 1.02 1. 024 4 1.03 1. 034 4 1.02 1. 028 8 1.0 .03 30 1.03 1. 030 0

1.034 1.03 4 1. 1.03 033 3 1.04 1. 041 1 1. 1.04 041 1 1.03 1. 030 0 1. 1.03 030 0 1.02 1. 029 9 1. 1.02 029 9 1.02 1. 029 9 1. 1.02 029 9 1.02 1. 024 4 1. 1.02 024 4 1.02 1. 025 5 1. 1.02 025 5 1.03 1. 030 0 1. 1.03 030 0 1.04 1. 041 1 1. 1.04 041 1 1.028 1.028 1.011 1.011 1.020 1.020 1.023 1.023 1.027 1.027 1.020 1.020 1.03 1. 039 9 1. 1.03 039 9 1.03 1. 034 4 1. 1.03 034 4 1.038 1.0 1.038 1.038 38 1.038 1.0 1.038 1.038 38 1.03 1. 038 8 1. 1.03 038 8 1.039 1.039 1.039

1.043 1.04 3 1. 1.04 043 3 1.04 1. 043 3 1. 1.04 043 3 1.035 1.0 35 1.0 1.035 35 1.05 1. 054 4 1. 1.05 054 4 1.01 1. 017 7 1. 1.01 017 7 1.04 1. 042 2 1. 1.04 042 2 1.00 1. 006 6 1. 1.00 006 6 1.04 1. 040 0 1. 1.04 040 0 1.00 1. 004 4 1. 1.00 004 4 1.029 1.029 1.03 1. 034 4 1. 1.03 033 3 1.03 1. 038 8 1. 1.03 038 8 1.03 1. 038 8 1. 1.03 038 8 1.03 1. 038 8 1. 1.03 038 8 1.038 1.038 1.038

1.043 1.04 3 1. 1.04 043 3 1.04 1. 043 3 1. 1.04 043 3 1.035 1.0 35 1.0 1.035 35 1.05 1. 054 4 1. 1.05 053 3 1.01 1. 017 7 1. 1.01 017 7 1.04 1. 042 2 1. 1.04 042 2 1.00 1. 006 6 1. 1.00 005 5 1.04 1. 040 0 1. 1.04 040 0 1.00 1. 004 4 1. 1.00 004 4 1.029 1.029 1.03 1. 033 3 1. 1.03 033 3 1.03 1. 038 8 1. 1.03 037 7 1.03 1. 038 8 1. 1.03 037 7 1.03 1. 038 8 1. 1.03 037 7 1.038 1.038 1.038

1.043 1.04 3 1. 1.04 043 3 1. 1.04 042 2 1. 1.04 042 2 1.04 1. 043 3 1. 1.04 043 3 1. 1.04 042 2 1. 1.04 042 2 1.035 1.0 35 1.0 1.035 35 1.0 1.034 34 1.0 1.034 34 1.05 1. 053 3 1. 1.05 053 3 1. 1.05 053 3 1. 1.05 052 2 1.01 1. 017 7 1. 1.01 016 6 1. 1.01 016 6 1. 1.01 016 6 1.04 1. 042 2 1. 1.04 041 1 1. 1.04 041 1 1. 1.04 041 1 1.00 1. 005 5 1. 1.00 005 5 1. 1.00 005 5 1. 1.00 004 4 1.04 1. 040 0 1. 1.04 040 0 1. 1.03 039 9 1. 1.03 039 9 1.00 1. 004 4 1. 1.00 003 3 1. 1.00 003 3 1. 1.00 003 3 1.029 1.029 1.028 1.028 1.03 1. 033 3 1. 1.03 033 3 1. 1.03 033 3 1. 1.03 032 2 1.03 1. 037 7 1. 1.03 037 7 1. 1.03 037 7 1. 1.03 036 6 1.03 1. 037 7 1. 1.03 037 7 1. 1.03 037 7 1. 1.03 036 6 1.03 1. 037 7 1. 1.03 037 7 1. 1.03 037 7 1. 1.03 036 6 1.038 1.037 1.037 1.037 1.037 1.037

1.034 1.034 1.041 1.0 41 1.030 1.040 1.040 1.03 1. 030 0 1.01 1. 018 8 1.04 1. 040 0 1.02 1. 022 2 1.04 1. 040 0 1.01 1. 017 7 1.037 1.0 37 1.02 1. 024 4 1.03 1. 033 3 1.02 1. 027 7 1.03 1. 030 0 1.03 1. 030 0

1.034 1.034 1.040 1.0 40 1.030 1.040 1.040 1.03 1. 030 0 1.01 1. 018 8 1.04 1. 040 0 1.02 1. 022 2 1.04 1. 040 0 1.01 1. 016 6 1.037 1.0 37 1.02 1. 024 4 1.03 1. 033 3 1.02 1. 027 7 1.03 1. 030 0 1.03 1. 030 0

1.034 1.0 1.034 1.034 34 1.0 1.033 33 1.0 1.033 33 1.040 1.0 40 1.0 1.040 40 1.0 1.040 40 1.0 1.039 39 1.030 1.030 1.029 1.029 1.039 1.039 1.039 1.038 1.039 1.039 1.039 1.038 1.03 1. 030 0 1. 1.03 030 0 1. 1.02 029 9 1. 1.02 029 9 1.01 1. 017 7 1. 1.01 017 7 1. 1.01 017 7 1. 1.01 016 6 1.03 1. 039 9 1. 1.03 039 9 1. 1.03 039 9 1. 1.03 038 8 1.02 1. 021 1 1. 1.02 021 1 1. 1.02 021 1 1. 1.02 020 0 1.03 1. 039 9 1. 1.03 039 9 1. 1.03 039 9 1. 1.03 038 8 1.01 1. 016 6 1. 1.01 016 6 1. 1.01 016 6 1. 1.01 015 5 1.036 1.0 36 1.0 1.036 36 1.0 1.036 36 1.0 1.036 36 1.02 1. 024 4 1. 1.02 023 3 1. 1.02 023 3 1. 1.02 023 3 1.03 1. 033 3 1. 1.03 033 3 1. 1.03 033 3 1. 1.03 032 2 1.02 1. 027 7 1. 1.02 027 7 1. 1.02 027 7 1. 1.02 026 6 1.0 .029 29 1. 1.0 029 1.0 .029 29 1.0 .028 28 1.029 1.0 29 1. 1.02 029 9 1. 1.02 029 9 1. 1.02 028 8

1.034 1.034 1.040 1.0 40 1.030 1.040 1.040 1.03 1. 030 0 1.01 1. 018 8 1.04 1. 040 0 1.02 1. 022 2 1.04 1. 040 0 1.01 1. 016 6 1.037 1.0 37 1.02 1. 024 4 1.03 1. 033 3 1.02 1. 027 7 1.03 1. 030 0 1.03 1. 030 0

1.033 1.03 3 1. 1.03 033 3 1.04 1. 041 1 1. 1.04 041 1 1.03 1. 030 0 1. 1.03 030 0 1.02 1. 029 9 1. 1.02 029 9 1.02 1. 029 9 1. 1.02 029 9 1.02 1. 024 4 1. 1.02 024 4 1.02 1. 025 5 1. 1.02 025 5 1.03 1. 030 0 1. 1.03 030 0 1.04 1. 041 1 1. 1.04 041 1 1.028 1.028 1.011 1.011 1.020 1.020 1.023 1.022 1.027 1.026 1.020 1.020 1.03 1. 039 9 1. 1.03 039 9 1.03 1. 034 4 1. 1.03 034 4 1.038 1.0 38 1.0 1.038 38 1.038 1.0 38 1.0 1.038 38 1.03 1. 038 8 1. 1.03 038 8 1.038 1.038 1.038

1.034 1.034 1.040 1.0 40 1.030 1.039 1.039 1.03 1. 030 0 1.01 1. 018 8 1.03 1. 039 9 1.02 1. 021 1 1.03 1. 039 9 1.01 1. 016 6 1.037 1.0 37 1.02 1. 024 4 1.03 1. 033 3 1.02 1. 027 7 1.02 1. 029 9 1.02 1. 029 9

1.033 1.03 3 1. 1.03 033 3 1.04 1. 040 0 1. 1.04 040 0 1.03 1. 030 0 1. 1.03 030 0 1.02 1. 029 9 1. 1.02 028 8 1.02 1. 029 9 1. 1.02 028 8 1.02 1. 024 4 1. 1.02 024 4 1.02 1. 024 4 1. 1.02 024 4 1.03 1. 030 0 1. 1.02 029 9 1.04 1. 041 1 1. 1.04 041 1 1.028 1.028 1.011 1.011 1.020 1.020 1.022 1.022 1.026 1.026 1.020 1.019 1.03 1. 039 9 1. 1.03 039 9 1.03 1. 034 4 1. 1.03 034 4 1.038 1.0 38 1.0 1.038 38 1.038 1.0 38 1.0 1.038 38 1.03 1. 038 8 1. 1.03 038 8 1.038 1.038 1.038

1.0 1 .033 33 1. 1.03 033 3 1.040 1.0 40 1. 1.04 040 0 1.030 1.0 30 1. 1.02 029 9 1.02 1. 028 8 1. 1.02 028 8 1.02 1. 028 8 1. 1.02 028 8 1.02 1. 023 3 1. 1.02 023 3 1.02 1. 024 4 1. 1.02 024 4 1.02 1. 029 9 1. 1.02 029 9 1.04 1. 041 1 1. 1.04 040 0 1.028 1.027 1.010 1.010 1.020 1.020 1.022 1.022 1.026 1.026 1.019 1.019 1.03 1. 039 9 1. 1.03 038 8 1.03 1. 033 3 1. 1.03 033 3 1.038 1.03 8 1.03 1.037 7 1.038 1.03 8 1.03 1.037 7 1.03 1. 038 8 1. 1.03 037 7 1.038 1.038 1.038

1.032 1.03 2 1. 1.03 032 2 1.04 1. 040 0 1. 1.03 039 9 1.02 1. 029 9 1. 1.02 029 9 1.02 1. 028 8 1. 1.02 027 7 1.02 1. 028 8 1. 1.02 027 7 1.02 1. 023 3 1. 1.02 022 2 1.02 1. 024 4 1. 1.02 023 3 1.02 1. 029 9 1. 1.02 028 8 1.04 1. 040 0 1. 1.04 040 0 1.027 1.027 1.010 1.010 1.019 1.019 1.022 1.021 1.026 1.025 1.019 1.018 1.03 1. 038 8 1. 1.03 038 8 1.03 1. 033 3 1. 1.03 032 2 1.037 1.03 7 1.03 1.037 7 1.037 1.03 7 1.03 1.037 7 1.03 1. 037 7 1. 1.03 037 7 1.038 1.037 1.037

Continued

63  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  Table 4.10. Continued

Ionization chamber typea

Beam quality,  R res (g cm22) 0.25 0.5

1

1.5

0.995 1.02 1. 029 9 1.00 1. 000 0 1.01 1. 014 4 0.99 0. 994 4 1.00 1. 009 9 1.008

0.990 1.02 1. 024 4 0.989 1.02 1. 024 4 0.99 0. 995 5 0. 0.99 994 4 1.00 1. 009 9 1. 1.00 008 8 0.98 0. 989 9 0. 0.98 989 9 1.0 .004 04 1. 1.00 003 3 1.003 1.002

2

2.5

3

3.5

4

4.5

5

7.5

10

15

20

30

0.989 1.02 1. 023 3 0.99 0. 994 4 1.00 1. 007 7 0.98 0. 988 8 1.0 .00 02 1.001

0.989 1.02 1. 023 3 0.99 0. 993 3 1.00 1. 007 7 0.98 0. 988 8 1.00 1. 002 2 1.001

0.988 1.02 1. 022 2 0.99 0. 993 3 1.00 1. 007 7 0.98 0. 987 7 1.00 1. 002 2 1.001

0.988 1.02 1. 022 2 0.99 0. 993 3 1.00 1. 007 7 0.98 0. 987 7 1.00 1. 002 2 1.001

0.988 1.02 1. 022 2 0.99 0. 993 3 1.00 1. 007 7 0.98 0. 987 7 1.00 1. 002 2 1.001

0.988 1.02 1. 022 2 0.987 1.02 1. 021 1 0.99 0. 993 3 0. 0.99 992 2 1.00 1. 007 7 1. 1.00 006 6 0.98 0. 987 7 0. 0.98 986 6 1.0 .002 02 1. 1.00 001 1 1.001 1.000

Plane-parallel chambers

 Attix RMI 449 Capi Ca pint ntec ec PS PS-0 -033 33 Exrrad Ex adin in P1 P11 1 Holt Ho lt (M (Mem emor oria ial) l) NACP NA CP/C /Cal alca cam m Mark rkus us Roos

0.992 1.02 1. 026 6 0.99 0. 997 7 1.01 1. 010 0 0.99 0. 991 1 1.00 1. 005 5 1.004

0.989 1.02 1. 023 3 0.989 1.02 1. 023 3 0.989 1.02 1. 023 3 0.989 1.02 1. 023 3 0.989 1.02 1. 023 3 0.99 0. 994 4 0. 0.99 994 4 0. 0.99 994 4 0. 0.99 994 4 0. 0.99 994 4 1.00 1. 008 8 1. 1.00 008 8 1. 1.00 008 8 1. 1.00 008 8 1. 1.00 007 7 0.98 0. 988 8 0. 0.98 988 8 0. 0.98 988 8 0. 0.98 988 8 0. 0.98 988 8 1.0 .003 03 1.0 .003 03 1.0 .003 03 1.0 .003 03 1. 1.00 002 2 1.002 1.002 1.002 1.002 1.001

a

Some of the chambers listed in this table fail to meet some of the main requirements of the code of practice (IAEA, 2000). However, they have been included in this table because of their current clinical use.

fil fille led d wi with th ai airr at 50 pe perc rcen entt re rela lati tive ve hu humi midi dity ty.. Medin et al.  (2000) have drawn attention to the fact thatt Eq tha Eqs. s. (4. (4.5)– 5)– (4. (4.7) 7) use used d her here e (IC (ICRU RU,, 199 1998) 8) for the determination of absorbed dose to water, using an   N  K  calibration coefficient that is referred to air at 50 per percent cent rela relativ tive e humi humidity dity,, cont contain ain phy physical sical cons co nsta tant ntss th that at ar are e al alll re refe ferr rred ed to dr dry y ai airr. Th The e humi hu midi dity ty co corr rrec ecti tion on fa fact ctor or in th the e pr prot oton on be beam am cancel can cels, s, at lea least st to firs firstt ord order er,, wit with h th the e hu humid midity ity   60 correction factor in Co. Con Conseque sequently ntly,, the use of  the humidity correction factor in Equation (4.6) is not needed. The approximate cancellation of humidity correction ti on fa fact ctor orss is th the e rea easo son n wh why y IA IAEA EA co code dess of  practice (IAEA, 1997a; 1997b; 2000) do not include a hu humi midi dity ty co corr rrec ecti tion on in th the e ch cham ambe berr fa fact ctor or

 N  D  D,air ,air  when   N  K   refers to air at 50 percent relative humidity humi dity.. When   N  K    is referred to dry air, as al alre read ady y st stat ated ed,,   K hum Equa uati tion on (4 (4.6 .6)) is hum   in Eq omitted omitt ed and the only remaining remaining humid humidity ity corr correcection is for the ionization current measured in the proton beam. Furth Fu rther er det detail ailed ed inf inform ormat ation ion on th the e eff effect ect of  air hu humid midity ity in ion ioniza izatio tionn-cha chambe mberr dos dosime imetry try in photon pho ton and ele electr ctron on bea beams ms is giv given en by Mij Mijnhe nheer er and an d Wil illi liam amss (1 (198 985) 5) an and d by Rog oger erss an and d Ros osss (1988). 4.4.4 4.4. 4 Th The e val value ue of   w / e  in air for proton beams

For th the e pu purp rpos oses es of si simp mpli lici city ty in th the e pr pres esen entt section, the subscripts c and p are used to denote the th e re refe fere renc nce e (60Co Co)) and pr proto otonn-bea beam m qua qualit lities ies,, respectiv resp ectively ely,, even thou though gh the corre correspon sponding ding subscripts  Q 0  and  Q  are used in IAEA (2000).  w  or  W  (or  w /  e  and  W  /  e) is required Knowledge of  w forr co fo con nvers rsio ion n of the ch char arge ge co coll llec ectted in an ionization chamber to deposited energy. The value of (wair /  e) is an imp import ortant ant fa facto ctorr in det determ ermini ining ng the th e ab abso sorb rbed ed do dose se wi with th io ioni niza zati tion on ch cham ambe bers rs in proto pr oton-t n-ther herapy apy bea beams. ms. As des descri cribed bed bel below ow,, the  various dosimetry protocols recommend the use of  significant signi ficantly ly diff differen erentt (wair /  e) valu values. es. Mos Mostt impo imporrtantly tan tly, th the e val values ues re recom commen mended ded in th the e two mos mostt recentt prot recen protocols ocols,, ICRU 59 (ICR (ICRU, U, 1998) and TRS 398 (IAEA, 2000) differ by more than 2 percent.  4.4.4.1

Definitions

The value of   W ( E) (in J C21) for charged particles of energy  E  in a gas is the mean energy required to

create an elect create electronron-ion ion pair by an ioniz ionizing ing parti particle cle that th at im impa part rtss al alll it itss en ener ergy gy to th the e ga gas. s. An Anot othe herr interpretation of   W ( E) arises from recognizing that  E/W(E)   is th the e mea mean n nu numbe mberr of ion pairs formed formed

 k Q,Q0 for various cylindrical and Figure 4.3. Calcula Calculated ted values of  k plane-parallel ionization chambers commonly used for reference dosime dos imetry try, as a fun functi ction on of pr proto oton n bea beam m qua qualit lity y   Q   ( Rres).  Adapted from IAEA (2000).

64  

 DOSIMETRY   DOSIMETR Y  Table 4.1 Table 4.11. 1. Estima Estimated ted rel relativ ative e stand standard ard unce uncertain rtainties, ties,   uc, of the calcula calculated ted values for   kQ,Q0 for proton beams (IAEA, 2000).

Chamber type

Cylindrical

Plane parallel

Component

Protons,  u c  (%)

sw,air    s w,air   to beam quality  Assignment of  s W air  e   air /   pcav    pdis    pwall    pcel   Combined relative standard uncertainty in  k Q,Q0

1.0 0.3 0.4 0.3 0.2 0.6 0.4 –

  60

Co þ protons,  u c   (%) Pr Proto otons, ns, u c  (%)

1.1 0.4 0.5 0.3 0.4 0.8 0.5 1.7

1.0 0.3 0.4 0.3 0.2 0.6 – –

  60

Co þ protons,  u c  (%)

1.1 0.4 0.5 0.3 0.3 1.6 – 2.1

when en pa part rtic icle less of en ener ergy gy   E   dissi dissipa pate te all th their eir wh energy in the gas. For indirectly ionizing radiation such as photons or neutrons, or when directly ionizing particles dis-

than the dim imen ensi sion onss of a typ ypic ical al io ioni niz zati tio on chamber cham ber.. Values of   w( E) and   W ( E) ar are e la larg rgel ely y independent of energy for particle speeds that are well we ll in ex exce cess ss of ou oute terr-or orbi bita tall el elec ectr tron on sp spee eeds ds..

sipate sipa te al alll th thei eirr en ener ergy gy in th the e ga gas, s,   W ( E) is the correct dosimetric conversion coefficient from ionization to energy imparted to the gas. Secondary particles of all energies less than that of the indirectly ionizing radiation are generated and interact with the gas and make   W ( E) the proper choice (ICRU, 1998; Verhey and Lyman, 1992).  As protons, even for range-modula range-modulated ted beams, lose only a fraction of their energy in traversing the gas, the proper conversion coefficient is the differential ent ial val value, ue,   w( E). He Here re   D E / w( E) is the mean numb nu mber er of io ion n pa pair irss fo form rmed ed wh when en a pa part rtic icle le of   E  in the gas. Use of  w  w  is approenergy  E  expends  D  D E pria pr iate te ev even en fo forr pr prot oton onss wi with th en ener ergi gies es as lo low w as

When  w ( E) is constant,  W ( E)

500 keV, as their range in air at standard temperature tur e an and d pr press essur ure e is   . 1 cm, signi significan ficantly tly large largerr

 4.4.4.2

¼

w( E) (ICRU, 1998).

Determination of the w(E) value

 Valuess of   w   (w /  e) in air for protons recommended in  Value the th e di diff ffer eren entt do dosi sime metr try y pr prot otoc ocol olss ar are e gi give ven n in Table 4.13. Also given are the corresponding dry-air or humid-a humid-air ir W c-values for  60Co used in the respective protocols. As can be seen, the w p-values recommended differ diff er both in abso absolut lute e valu values es and in quot quoted ed unce uncerrtaint ta inties ies,, an and d in wh whet ether her th the e val values ues re refer fer to dr dry y or humid hum id air air.. Fu Furth rthermo ermore, re, the   w p-va -valu lues es giv given en ar are e derived deri ved fro from m a vari variety ety of sou sources rces.. Spe Specific cifically ally, the recom re comme mend nded ed val value uess in th the e mo most st re recen centt pr prot otoco ocols ls (IAEA, 2000; ICRU, 1998) differ significantly: ICRU Report Re port 59 (IC (ICRU, RU, 1998) rec recomme ommends nds a   w p-val -value ue of 

Table 4.12. Estimated typical relative standard uncertainty of the absorbed dose to water, D w,p, at the reference depth in water for a clinical proton beam, based on a chamber calibration in   60Co (IAEA, 2000).

Physical quantity or procedure

Step 1: Standards laboratory  N  D,w   calibration of secondary standard at SSDL Long-term stability of secondary standard  N  D,w   calibration of the user dosimeter at the standards laboratory   Combined uncertainty in Step 1 Step 2: User proton beam Long term stability of the user dosimeter

Relative standard uncertainty (%) Cylindrical

Plane parallel

0.5 0.1 0.4 0.6

0.5 0.1 0.4 0.6

0.3

0.4

Establishment of reference conditions Correction Corre ction for influenc influence e quantities k i   Beam quality correction  k Q,Q0   Combined uncertainty in Step 2 Combined relative standard uncertainty in  D w,Q  (Steps 1 1 2)

0.4 0.4 1.7 1.9 2. 0

0.4 0.5 2.0 2.2 2.3

SSDL, Secondary Standard Dosimetry Laboratory. If the calibration of the user dosimeter is performed at a Primary Standard Dosimetry Laboratory (PSDL) then the combined uncertainty in Step 1 is lower. The combined standard uncertainty in  D w,Q  should be adjusted accordingly.

65  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  Table 4.13. 4.13. Recommended w p-values in air for protons for different dosimetry protocols. The corresponding  W c  values in air for   60Co are shown in brackets.

Protocol  AAPM ECHED ICRU 59

 

w p  ( W c) (J C21)

Source of  w  p

   

34.3 + 1.4  (33.73 + 0.04)a (hum (humid id air) 35.2 + 1.4  (33.97 + 0.07) (dry air) 34.8 + 0.7  (33.77 + 0.05)c (hum (humid id air)

Measured Measure d valu value e (V (Verhe erhey y et al.,1979)b Measured value (ICRU, 1979) Compro Com promise mise betw between een valu values es meas measure ured d directly and those inferred from comparisonss with calorimetry comparison Statisti Stat istical cal anal analysis ysis usin using g wei weighted ghted medi medians ans (Mu¨ ller, 2000 2000a) a)

 

IAEA TRS 398 398   34.23 + 0.13  (33.97 + 0.07 0.07)) (dry air)

 

Reference  AAPM (1986) Vynckier  et al.  (1991; 1994) ICRU (1998)

IAEA (2000)

a

From Verhey et al.  (1979).  Verhey et al.  (1979) reference Myers (1968) in which a value of (36.0  +  1.1 %) J C21 is actually given (Hayakawa and Schechtman,  Verhey 1988). c The value of (33.97 + 0.15) J C21 given in ICRU (1998) is incorrect. b

(34.8 + 0.7) J C21 in humid air, whereas IAEA Report TRS 398 (IAEA, 2000) recommends a  wQ(w p)-value of 

Subsequently Subsequent ly,, Medin   et al.   (2006) (2006) determ determined ined absorbed absorbe d doses with a sealed water calorimeter calorimeter oper-

(34.23 + 0.13) J C in dry air. The ICRU value was a compromise among values measur mea sured ed dir direct ectly ly and tho those se obt obtain ained ed fr from om com com-pariso par isons ns of ion ioniza izatio tion n cha chambe mberr and cal calori orimet meter er measurements (ICRU, 1998). The IAEA value was obtained by performing a statistical analysis using weighed medians (Mu ¨ ller ller,, 2000a; 2000b) on a modified ve versi rsion on of the sam same e ori origin ginal al da datas taset et (IA (IAEA, EA, 2000). The ICRU value converted to the equivalent  value for dry air becomes 35.0 J C21, which is 2.3 percent perc ent higher than the IAEA valu value. e. Jones (2006) undertook a comprehensive analysis  e) p   an of al alll ava vail ilab able le (W air and (wair /  e) p   data data fo forr air /  proto pr otons ns wit with h ene energi rgies es    1 MeV. Ap Apart art fr from om the w-values derived from the comparison of ionometric and calorimetric measurements [Brede   et al., 2006 al.., 199 (the (th e sam same e da data ta ar are e giv given en in Br Brede ede   et al 1999); 9); al.., 199 al.., 20 Delacroix   et al 1997; 7; Has Hashem hemian ian   et al 2003 03;; al.., 199 al.., 19 Palmans   et al 1996; 6; 200 2004; 4; Sch Schulz ulz   et al 1992 92;; Siebers   et al., 199 1995] 5] and one oth other er mea measur sureme ement nt (Moyers   et al., 200 000 0), non one e of th the e data wer ere e obtained under clinical conditions. The calorimetric techn tec hniqu ique e als also o giv gives es a   w-va -value lue th that at is av aver erage aged d over all charged particles in the beam at the point of mea measur sureme ement nt (th (there erefor fore, e, ind indica icatin ting g the use of  stopping powers that include the effects of second-

ated at 48C and compared ated compared the results results with those obtaine obta ined d usin using g ion ioniza izatio tion n cham chamber berss and the IAE IAEA  A  TRS-398 code of practice. Measurements were made in a 175 MeV clinical proton beam at a depth corresponding to a residual range of 14.7 cm and energy of 144 MeV. MeV. A value for ( wair /  e) p of 33.6 + 0.6 J C21 (1 SD) was deriv derived ed using ionization-cham ionization-chamber ber pertu perturrbation factors of 1.00. Inclusion of this datum in the calorimetric-derived   w-value analysis did not affect the th e con concl clus usio ion n by Jo Jone ness (2 (200 006) 6) th that at th the e TR TRS S 39 398 8  value should contin continue ue to be used. The data used in the analysis are shown in Fig. 4.4.

21

4.4.5 Comparison Comp arison of proton dosime dosimetry try protocols

 As discussed above, one of the main differences betw be twee een n th the e tw two o pr prot otoc ocol olss (I (IAE AEA, A, 20 2000 00;; IC ICRU RU,, 1998 19 98)) li lies es in th the e   wair /  e   value value rec recomm ommend ended ed for

ary ele electr ctrons ons and nu nucle clear ar re recoi coils) ls).. Fu Furth rtherm ermor ore, e, sim simila r ex exper perime imenta techni tec hnique ques sesha have vevebee been nobably used use d andilar evalua eva luatio tions ns ntal of l un uncer certai tainti nties have ha proba pr bly been bee n mor more e un unifo iforml rmly y ass assess essed. ed. In add additi ition, on, the uncertain unce rtainties ties are rela relative tively ly smal smalll compa compared red with otherr measu othe measurem rements ents in the clini clinical cal energ energy y ran range. ge. Jones Jon es (20 (2006) 06),, the there refor fore, e, re recom commen mended ded th that at the w-v -val alue ue in ai airr fo forr cl clin inic ical al pr prot oton on do dosi sime metr try y be deri de riv ved on only ly fr from om ev eval alua uati tion on of ca calo lori rime metr tric ic data. Analysis of these data gave a mean value for (wair /  e) p   that that is co cons nsis iste tent nt wi with th th the e TR TRS S 39 3988recommended value of (34.2 + 0.1) J C21.

Figure 4.4. Value Figure aluess of   w /  e   in dry air for pro protons tons deduced deduced from comp co mpar aris ison on of ioni io niza zati tion on cham ch ambe berr and an d calo ca lori rime mete terr measurements. The thick dotted line is the recommended value (34.2 J C21). [Ad [Adap apted ted fr from om Jon Jones es (20 (2006) 06);; re repr produ oduced ced wi with th permission.]

66  

 DOSIMETRY   DOSIMETR Y 

relativ rela tive e inc incre rease ase in abs absorb orbed ed dos dose e of 0.3 per percen cent. t. This issue is discussed fully by Medin  et al.  (2000). ICRU ICR U 59 doe doess not take into ac accou count nt the possipossi-

Figure 4.5. Water-to ater-to-air -air stopping-power stopping-power ratios for protons as a function func tion of res residua iduall ran range ge for the TRS 398 (IAEA, 2000) and ICRU 59 (ICRU, 1998) dosimetry protocols.

 w air /  e  for protons to  W air  e  for protons. The ratios of  w air /  60 Co are 1.0305 (humid air) and 1.0077 (dry air) for the ICRU 59 and TRS 398 prot protocols ocols,, resp respectiv ectively ely,, i.e., the relative TRS 398 value is 2.3 percent lower than the ICRU 59 value. Both protocols protocols use prot proton on stop stopping ping pow powers ers give given n in IC ICRU RU Re Repo port rt 49 (I (ICR CRU, U, 19 1993 93a), a), bu butt TR TRS S 39 398 8 includes includ es second secondary ary electr electron on tran transport sport and nucle nuclear ar interactions (Medin and Andreo, 1997b). The respective wa water-t ter-to-air o-air sto stoppin pping-po g-power wer ra ratios tios are show shown n in Fig. 4.5. The relative TRS 398 (IAEA, 2000) waterto-air stopping-power ratios are at most 0.6 percent higherr (Med highe (Medin in   et al., 2000) than the corre correspon sponding ding ICRU values (ICRU, 1993a; 1998). TRS 398 provides a formula, Eq. (4.17), to calculate   sw,air  for protons. Chamber Cham ber pertu perturbat rbation ion fact factors ors and unce uncertain rtainties ties in 60 Co and in proton beams are specified in TRS 398, whil wh ile e IC ICR RU 59 doe oess not sp spec ecif ify y an any y ch cham ambe berr perturbat pertu rbation ion fact factors ors in its dosedose-to-w to-water ater calibr calibrati ation on formalism. As shown in Table 4.9, the perturbation factors given in TRS 398 can be significantly different fr from om un unit ity y in   60Co for a var variet iety y of ion ioniza izatio tion n

bility ofng thecon ionization-chamber wall ma and build-up cap bei being const struc ructed ted of dif differ ferent ent mater terial ials. s. In such cases, the relative calculated correction factor for the non non-ai -air-e r-equi quival valenc ence e of the cha chambe mberr wa wall ll   60 and build-up cap during calibration in Co (IAEA, 1997 19 97a; a; 20 2000 00)) ca can n be up to 1. 1.1 1 pe perc rcen entt hi high gher er   et al. al . (Medin , 200 2000) 0) tha than n tho those se val values ues cal calcul culat ated ed accor ac cordin ding g to IC ICRU RU 59 (IC (ICRU RU,, 199 1998). 8). The   swall,air sto topp ppin ingg-po pow wer rati tios os ar are e us used ed to ca calc lcul ulat ate e   60 wall-at wal l-attenua tenuation tion corre correction ction fact factors ors in Co. Th The e rati ra tios os gi give ven n in IC ICRU RU 59 ar are e th the e sa same me as in th the e  AAPM TG-21 photon and electron dosimetry protocoll (A co (AAP APM, M, 19 1983 83). ). Th The e IA IAEA EA us uses es Mo Mont nte e Ca Carl rlo o calculated data from Andreo   et al.  (1986) based on data gi data give ven n in IC ICRU RU (1 (198 984) 4).. Th The e re rela lati tive ve IA IAEA  EA  graphite grap hite-to-a -to-air-s ir-stopp topping ing pow power er rat ratio io is 0.8 perc percent ent al.., 20 higher high er (Med (Medin in   et al 2000 00)) th than an gi give ven n in IC ICRU RU (1998). For A-150 plastic, the relative difference is 0.3 percent (Medin  et al., 2000). Overall relative uncertainties in proton-absorbeddose do se de dete term rmin inat atio ions ns ar are e 2. 2.6 6 pe perc rcen entt (I (ICR CRU U 59 59)) (Table 4.6) and 2.0 percent (TRS 398) (Table 4.12) for cylindrical cylind rical ioniza ionization tion chamb chambers. ers. The rela relative tive rat ratios ios of pro protonton-abs absorbe orbed d dose dosess mea measur sured ed usi using ng the TRS 398 formalism to those measured using the ICRU 59 formalism differ by up to 3.1 percent depending on chamber type and residual range (Fukumura   et al., al.., 2000; Vatni al.., 200 2002; Medin   et al atnitsky tsky   et al 2002). 2). These Th ese di diff ffer eren ence cess can lar largel gely y be as ascri cribe bed d to th the e differ dif ferent ent pr proto oton n (wair /  e) va valu lues es use sed d an and d to th the e chamber chamb er pertu perturbati rbation on fact factors ors not being appli applied ed for 60 Co calibrations in the ICRU 59 protocol. The main featur fea tures es of the two pro protoc tocols ols (IA (IAEA, EA, 2000; ICR ICRU, U, 1998) are summarized in Table 4.14. It is recommended that the TRS 398 code of practice ti ce (I (IAE AEA, A, 20 2000 00)) be ad adop opte ted d as th the e sta tand ndar ard d proton dosimetry protocol: it is simple to use; it pro-

chambers (IAEA, 2000; Medin  et al., 2000). The total perturbation factor in   60Co can differ from unity by al.., 200 up to 3 pe perc rcen entt (M (Med edin in   et al 2000). 0). Cha Chambe mberrperturbation factors in proton beams are assumed to be un unit ity y in TR TRS S 39 398. 8. Mon onte te Ca Carl rlo o si simu mula lati tion onss (Palm (P almans ans and Verh erhaeg aegen, en, 199 1998; 8; Verh erhaeg aegen en and Palmans, 2001), analytical calculations (Medin  et al., 2006), and expe experimen rimental tal measu measureme rements nts (Pa (Palmans lmans  et al., 199 1999; 9; 200 2001; 1; 200 2002a) 2a) sup suppor portt thi thiss ass assum umpti ption on within uncertainties of     1 percent.  As discussed in Section 4.4.3, ICRU 59 includes a humi hu midi dity ty co corr rrec ecti tion on fa fact ctor or in ca calc lcul ulat atin ing g th the e cham ch ambe berr re resp spon onse se,, bu butt th this is is pr prob oblem lemat atic ical al as the phy physic sical al con const stant antss re refer fer to val values ues for dry air air.. The fac factor tor   K hum 0.997, 0.9 97, Eq. (4. (4.6), 6), should should not be hum used us ed.. Ho How wev ever er,, it itss us use e wo woul uld d on only ly res esul ultt in a

 vides tabulated beam-quality correction factors for a wi wide de ra rang nge e of co comm mmon on cy cyli lind ndri rica call an and d pl plan aneeparallel ionization chambers; it provides a formula for calculating   sw,air   for proton beams in terms of  th the e monize beam-q bea m-qual ity param par eter rcols (resi (r dual al range) ran ge); ; itl harmon har izes s uality with wit h th the e amete proto pr otocol sesidu for conve con venti ntiona onal radioth rad iotherap erapy y and heavy-ion heavy-ion beams (also give given n in TRS 398), which are being adopted in many institutes; the uncertainties in the dose determinations are ar e les less; s; mor more e re recen centt and ac accur curat ate e phy physic sical al con con-stants are used; and the formalism is more robust and rigorous than that of ICRU Report 59 (Medin  et al., 2000 2000;; Vatn atnitsk itsky y and Andreo, 2002). As discussed in Section 4.4.4, the recommended   w-value for protons in dry air remains 34.2 J C21 with a relative standard uncertainty of 0.4 percent.

¼

67  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  Table 4.14. Compar Table Comparison ison of ICRU 59 (ICRU, 1998) 1998) dosimetry protocol and TRS 398 (IAEA, 2000) dosimetry dosimetry code of pra practice ctice for clinical proton beams.

Feature/quantity

ICRU 59 (ICRU, 1998)

TRS 398 (IAEA, 2000)

Ionization chamber

Cylindrical

Wall material Gas filling Chamber volume

Graphite or A-150 plastic Ambient air 3 . 0.5 cm for beams   5 cm diameter  0.1 cm3 for beams   , 5 cm diameter   – Water 60   Co   Primarily  N  K   K , also  N  X   X ,  N  D,w  D,w Residual range (to distal 10 % level) Wate terr (o (orr oth ther er mate teri rial alss whic ich h matc tch h ele lect ctrron density of water)

Cylindrical and plane-parallel ( Rres  0.5 g cm22) Plane-parallel Plane-par allel ( Rres , 0.5 g cm22) Graphite for cylindrical chambers Ambient air –

Water proof sleeve Dose specification Calibration Calibra tion beam Calibration coefficient Beam quality Phan anto tom m ma mate terria iall

 

 

Reference point for measurem measurement ent Field size SSD Stopping powers (wair /  e) p  (J C21) (wair /  e)60Co (J C 21) (sw,air)60Co Chamber perturbation factors Relative uncertainty in absorbed dose (1s)

Middle of SOBP – – ICRU (ICRU, 1993a) 34.8 + 0.7 (ambient air) 33.77 + 0.05 (ambient air) 1.134 No 2.6 % (Jones, 2001d)

4.4.6 Rela Relation tion betwe between en absorbe absorbed d dose dose to water and air-kerma calibration coefficients

The re The rela lati tion on be betw twee een n th the e   N  K – N  D  D,air ,air   formalism [used in IAEA (1997a) and in ICRU (1998)] and the  N  D,w  formalism [used in IAEA (2000)] is relatively straight str aightforw forward ard (IAE (IAEA, A, 2000 2000). ). The quan quantity tity   N  D  D,air ,air is given (IAEA, 2000) by

 1 mm thick PMMA  Water 60 Co N  D,w  only Residual range (to distal 10 % level ) Water Middle of SOBP22 Depth of 3 g cm for plateau irradiations irradiations 10  10 cm2 Clinical treatment distance PETRA (Medin and Andreo, 1997b) 34.23 + 0.13 (dry air) 33.97 + 0.07 (dry air) 1.133 Yes 2.0 % (cylindrical chambers) 2.3 % (plane-parallel chambers)

beam qu beam qual alit ity y. Th The e fa fact ctor or   N  D then al allo lows ws  D,air ,air   then the th e de dete term rmin inat atio ion n of th the e me mean an ab abso sorb rbed ed do dose se with wi thin in th the e ai airr ca cavi vity ty fo forr th the e us user er-p -pro roto tonn-be beam am quality   Q:  Dair;Q  ¼  M Q N  D;air:

 

ð4:19Þ

 N  D;air  ¼  N  K ð1  gÞkatt km kcel ;

 

The absorbed dose to water   Dw,Q  at the reference point in a phantom at the center of the chamber is obta ob tain ined ed fr from om th the e do dose se to ai airr us usin ing g th the e Br Brag agg– g– Gray principle as

ð4:18Þ

where   N  K  is the air-kerma calibration coefficient,   g is the fraction 0.003 (Boutillon, 1987) of the kinetic energy of secondary-charged particles that is lost in radiative radiativ e processes (bremsst (bremsstrahlung), rahlung),   katt   is th the e factor that corrects for the absorption and scattering in the walls of an ionization chamber irradiated in the calibration   60Co beam,   km   is the factor that corrects correc ts for the non-air equivalence equivalence of the ioniza ionizationtionchamber wall and buildup-cap materials in the calibration   60Co beam, and  k cel is the factor that corrects for the non-air equivalence of the central electrode of  a cyl cylind indrica ricall ion ioniza izatio tion n cha chambe mberr in the cali calibra bration tion 60 Co bea beam m (IA (IAEA EA,, 199 1997a; 7a; 199 1997b) 7b).. Not Note e tha thatt   N  D  D,air ,air correspond corres pondss to   N gas AAPM PM pro protoc tocol ol (AA (AAPM PM,, gas   of the AA 1983). It is ass assum umed ed th that at the   N  D factor deriv derived ed at  D,air ,air   factor   60 the Co quality quality,,   Q0, is al also so va vali lid d fo forr th the e us user er--

 Dw;Q  ¼  M Q N  D;air ðsw;air ÞQ pQ ;

 

ð4:20Þ

where   M Q   is the dos osim imet eter er rea eadi ding ng in th the e proton pro ton bea beam m corr correcte ected d for infl influen uence ce qua quanti ntitie ties, s, ( sw,air)Q   is th the e pr prot oton on wa wate terr-to to-a -air ir st stop oppi ping ng-power ratio,   pQ   is the overall perturbation factor of th the e io ion niz iza ati tion on ch cham ambe berr fo forr in in-p -pha hant ntom om measur mea sureme ements nts in the pro proton ton bea beam m (se (see e Sec Section tion 4.4.2). Compa Co mparin ring g Eqs Eqs.. (4. (4.11 11)) and (4. (4.20) 20) for the sam same e beam quality  Q 0, one obtains  N  D;w;Q0   ¼  N  D;air ðsw;air ÞQ0  pQ0 :

 

ð4:21Þ

68  

 DOSIMETRY   DOSIMETR Y 

 Alternatively ely,, from Eqs. (4.16), (4.18), and (4.20)  Alternativ one obtains in expanded form  N  D;w;Q0   ¼ ½ N   N  K ð1  gÞkatt km kcel Q0 ðsw;air ÞQ0

 ½ p  pcav pcel pdis pwall Q0 ;

 

ð4:22Þ

The sy The symb mbol olss an and d th thei eirr me mean anin ing g co corr rres espo pond nd to thos th ose e gi give ven n in IA IAEA EA (1 (199 997b 7b)) an and d IA IAEA EA (2 (200 000) 0).. Detail Det ailss of the re requi quire red d st stopp opping ing-po -powe werr da data ta and pert pe rtur urba bati tion on co corr rrec ecti tion on fa fact ctor orss ar are e gi giv ven in Sectio Sec tions ns 4.4 4.4.2 .2 and 4.4 4.4.2. .2.1. 1. Fa Facto ctors rs rel relat ated ed to the determina deter mination tion of    N  D can be fo foun und d in IA IAE EA   D,air ,air   can (1997a; 1997b).

4.5 REFERE REFERENCE NCE DOSIMETR DOSIMETRY Y FOR SCANN SCANNED ED BEAMS  Although the first proton-beam scanning syste system m was developed in the late 1970s on a 70 MeV beam (Kanai   et al., 1980; 1983), practical clinical experience with such systems has been acquired only on the 270 MeV proton beam at the PSI (Bacher  et al., al.., 20 1989 19 89;; Lo Loma max, x, 19 1999 99;; Lo Loma max x   et al 2001 01;; 20 2004 04;; al.., 19 Pedroni   et al 1990 90;; 19 1995 95;; 20 2005 05). ). Th Ther ere e ar are e fe few w differenc diffe rences es in prin principle ciple betw between een refe referenc rence e dosim dosim-etry et ry in sc scan anne ned d be beam amss an and d pa pass ssiv ivel ely y sc scat atte tere red d beams, and the calibration techniques are similar (Coray et al., 2002; Pedroni  et al., 2005). The dose delivered is calculated

by

the

calculat calcul ation ion mod model el is oft often en bas based ed on dos dose e dis distri tri-bution measurements, which include the effects of  nuclear interaction interactions. s. Ref efer eren ence ce do dosi sime metr try y is ba base sed d on th the e us use e of    60 cylindrica cylin dricall ioniz ionizatio ation n cham chambers bers calib calibrat rated ed in Co using usi ng th the e TR TRS S 398 cod code e of pr prac actic tice e (IA (IAEA, EA, 2000). 2000). 3 For the ca cali libr bra atio ion n, a 10  10  10 cm water  volume (‘dose box’) is irradiated to give a homogeneous dose distribution to an assumed dose of 1 Gy. A reference ionization chamber is placed in the water wa ter phantom phantom and the dose is mea measur sured ed at th the e center of the field at a residual range of 5 g cm 22.  At PSI, the ratio between these measuremen measurements ts and those determined with a Faraday cup varies from 0.978 at a proton energy of 138 MeV to 1.008 at a proton energy of 214 MeV. These values are used to correct corr ect the dose model in the trea treatmen tment-pl t-plannin anning g system (Pedroni  et al., 2005).

4.6 IONIZA IONIZATION-CHA TION-CHAMBER MBER DOSIMETR DOSIMETRY  Y  COMPARISONS The IC The ICRU RU ha hass lo long ng re recog cogni nize zed d th the e im impo port rtan ance ce of  dosimetry dosim etry compa comparisons risons for non-co non-conven nventiona tionall ther therapy apy beams. In the early 1970s, it sponsored an Intern Int ernat ation ional al Neu Neutro tron n Dos Dosime imetry try Int Interc ercomp ompari arison son (IND (I NDI) I) (I (ICR CRU, U, 19 1978 78), ), wh whic ich h fo follo llowe wed d an ea earl rlie ierr European Neutron Dosimetry Intercomparison Project (ENDIP) (Broerse  et al., 1978) and several small-scale comparisons compa risons (Almond and Sma Smathers, thers, 1977; Broe Broerse rse

 et al., 1979). Fourteen groups participated in the INDI and the res result ultss hig highlig hlighte hted d man many y sho shortc rtcomi omings ngs in neutron neut ron dosim dosimetry etry pra practices ctices.. No dosim dosimetry etry prot protocols ocols were we re av avai aila labl ble e at th that at ti time me an and d mo most st gr grou oups ps us used ed ‘in-h ‘in -hou ouse se’’ me meth thod ods. s. Th Thes ese e co comp mpar ariso isons ns le led d to th the e deve de velo lopm pmen entt of th the e Am Amer erica ican n (A (AAP APM, M, 19 1980) 80) an and d al.., 198 European Euro pean (Bro (Broerse erse   et al 1981) 1) neu neutr tron on dos dosime imetry try proto pr otocols cols and fina finally lly to the publicat publication ion of a uni uniform form protocol (Mijnheer   et al., 198 1987b) 7b),, whi which ch rem remain ainss the accepted standard 20 years later. This latter protocol wass en wa endo dors rsed ed by th the e IC ICRU RU (I (ICR CRU, U, 19 1989 89). ). The sit situa uatio tion n wit with h pr proto oton n dos dosime imetry try is ra rath ther er differen diff erent. t. Prot Protocols ocols hav have e exis existed ted for 15 year yearss and al.., 199 more mor e (AAP (AAPM, M, 1986 1986;; Vyncki ynckier er   et al 1991; 1; 199 1994). 4). However How ever,, no prot proton on abso absorbedrbed-dose dose sta standar ndards ds are yet available, and dose comparisons are extremely useful in confirming the integrity of the dosimetry techniques used at different institutes. Dosi Do sime metr try y co comp mpar aris ison onss be betw twee een n io ioni niza zati tion on-chamb cha mber er mea measur sureme ements nts,, and Fa Fara rada day-c y-cup up and calori cal orimet metric ric mea measur sureme ements nts ha have ve bee been n dis discus cussed sed above abo ve (S (Sect ection ionss 4.2 and 4.3 4.3,, res respec pectiv tively ely). ). Ma Many ny compariso comp arisons ns invo involving lving ioniz ionizatio ation n cham chambers bers (bot (both h cylind cyl indric rical al and pla planene-par parall allel) el) ha have ve bee been n und underertaken in the last 15 years or so by various groups

treatment planning system. Because the scanning treatment system does not require patient-specific hardware, the scanned beam is quite clean (i.e., it does not contai con tain n par partic ticles les sca scatte ttere red d fro from m col collim limat ators ors and other oth er bea beam-m m-modi odifyi fying ng ele elemen ments) ts).. Th There erefor fore, e, an absolute absol ute dose model based on calcu calculate lated d stop stopping ping powers (see Section 3) and the number of protons delive del ivere red d du durin ring g the sca scan n can be use used. d. Ho Howe weve verr, the th e mo mode deli ling ng of nu nucl clea earr in inte tera ract ctio ions ns mu must st be included inclu ded in the dosedose-calcu calculati lation on algori algorithm thm;; other other-wise the calculated dose distributions do not agree al.., 20 with measu measureme rements nts (Pe (Pedro droni ni   et al 2005 05). ). Th The e impo im port rtan ance ce of ta taki king ng nu nucl clea earr in inte tera ract ctio ions ns in into to acco ac coun unt, t, in incl clud udin ing g th the e tr tran ansp sport ort of se seco cond ndar ary y proto pr otons, ns, wa wass emp emphas hasize ized d by Med Medin in and An Andr dreo eo (1997a (19 97a). ). Sp Speci ecifica fically lly,, con consid sidera eratio tions ns of nuc nuclea learr interactions are required to account for the attenuatio at ion n of th the e pr prim imar ary y pr prot oton on flu fluen ence ce an and d fo forr th the e beam halo around the primary pencil beam. If they are not included relative differences between calculation and measurement of the order of 5 percent are found. The number of protons delivered is controll tr olled ed in pr pract actice ice by the bea beam-fl m-fluen uence ce mon monito itors rs calibrated calibr ated again against st Far Faraday aday-cup -cup meas measurem urements ents.. In the th e ca case se of pa pass ssiv ivel ely y mo modi difie fied d be beam ams, s, th the e do dose se 69  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  Table 4.15 Table 4.15.. Prot Proton on dosimet dosimetry ry compar comparisons isons with ioniz ionization ation chambers. chambers. The relative relative maximum measured measured dose differences differences are given in the last column.

Reference

Protocol

Number of chambers

Beam energy (MeV)

Max. difference (%)

Kacperek et al.   (1991) Jones  et al.   (1992) Jones  et al.   (1994a) Schreuder  et al.   (1994) Medin  et al.   (1995) Jones (1996) Palmans  et al.   (1996)  Vatnitsky  V atnitsky et al.   (1996b) Hiraoka  et al.   (1997) Cuttone  et al.   (1999)  Vatnitsky  V atnitsky et al.   (1999b) Nohtomi  et al.   (2001)

ECHED ECHED ECHED AAPM/ECHED TRS 277 ECHED ECHED EC AAPM/ECHED ECHED ICRU 59 IC ICRU 59

8 7 7 7 7 6 10 23 5 3 11 5

60 60 – 80 200 135 – 185 170 58 – 168 85 100 – 250 70 28 – 62 155 250

4.5 2.5 1.8 1.4 1.5 2.3 1.2 5.8 0.8 2.1 2.9 1.5

Palmans  et al.   (2001) Fukumura  et al.   (2002) Kacperek et al.   (2002)  Vatnitsky  V atnitsky et al.   (2002)

TRS 398 TRS 398 ICRU 59 ICRU 59/TRS 398

17 8 10 6

75 150 63 100 – 155

1.5 0.9 3.2 3.1

 AAPM, AAPM (1986); ECHED, Vynckier Vynckier et al.  (1991; 1994); TRS 277, IAEA (1997a); ICRU 59, ICRU (1998); TRS 398, IAEA (2000).

in di diff ffer eren entt cl clin inic ical al be beam ams. s. A su summ mmar ary y of th the e details of these studies is given in Table 4.15. The results are very satisfactory with a few exceptions. In mos mostt cas cases, es, th the e rel relat ativ ive e max maximu imum m dif differ ferenc ences es    between measured doses are less than 2 percent.

dose an dose and d flu fluen ence ce in th the e pa pati tien ent. t. Pa Para rall llel el-p -pla late te ioni io niza zati tion on ch cham ambe bers rs (B (Boa oag, g, 19 1966 66)) ar are e th the e mo most st common and well-proven detectors for this purpose. Their The ir lo low w mas mass, s, hig high h sen sensit sitivi ivity ty,, lar large ge dyn dynami amicc range, ran ge, and eas ease e of con const struc ructio tion n ma make ke the them m we well ll

The standard standard deviations deviations of the means are    0.3–

suited to the task of dose monitoring and control.

0.5 of the max maximu imum m dif differ ferenc ences. es. The max maximu imum m relat re lativ ive e dif differ ferenc ence e of 5.8 per percen centt (s (stan tandar dard d dev deviiati tion on of th the e me mean an:: 1. 1.8 8 pe perc rcen ent) t) oc occu curr rred ed in a large-scale study in which two different dosimetry protocols were used. It is clear from these data that proton dosimetry is on a rela relative tively ly good footing. However However,, comp compariarison studies are still extremely useful, especially for new ne w fa facil ciliti ities, es, but als also o for ex exist isting ing fa facil ciliti ities, es, in order to detect and eliminate any possible systematic errors occurring in the dosimetry process. They also als o ser serve ve as an ind indepe epende ndent nt che check ck on th the e ent entir ire e dosimetry chain. This chain starts at the standards labor lab orat atory ory and end endss at the del deliv ivery ery of abs absorb orbed ed dose to a reference point in a phantom in the user’s clinical beam. Such studies are also important for inst in stit itut utes es th that at wi wish sh to po pool ol or co comp mpar are e cl clin inica icall data. da ta. Cl Clear early ly,, the ado adopti ption on of a sin single gle dos dosime imetry try proto pr otocol col wil willl go a lon long g wa way y to towa ward rd st stand andard ardizi izing ng clinical proton dosimetry and reducing the magnitude of the differences differences in dose determinat determinations ions by different centers.

Beam monitors in beam-delivery systems (nozzles)

Other type Other typess of moni monitors tors inclu include de gas-s gas-scinti cintillat llation ion counters (Coutr (Coutrakon akon   et al., 1991a) and secondaryemis em issi sion on mo moni nito tors rs (S (SEM EMs) s) (S (Sec ecti tion on 4. 4.7. 7.4) 4).. Fo Forr safe sa fety ty re reas ason ons, s, at le leas astt tw two o in inde depe pend nden entt do dose se moni mo nito tors rs,, wh whic ich h ar are e ca cali libr brat ated ed da dail ily y, ar are e re reccommended. ommen ded. For pass passively ively modified beam beamss empl employoying scattering systems for lateral beam spreading, beam-centering systems (including segmented ionization izat ion cham chambers) bers),, whic which h are capable of dete detecting cting misa mi sali lign gnme ment ntss be betw twee een n th the e ce cent ntra rall ax axes es of th the e beam and the scattering devices, are required. Dose-moni Dosemonitorin toring g dete detectors ctors can eithe eitherr inte intercep rceptt the ent entir ire e bea beam m ar area ea or jus justt th the e cen centr tral al por portio tion. n. The former requires larger detectors and measurements are more reliable as they are less dependent on bea beam m ali alignm gnment ent var varia iatio tions. ns. Bea Beam m st steer eering ing is much mu ch mo more re se sens nsit itiv ive e in do doub uble le-- th than an in si sing ngle le-scatterer beam-modification syste systems. ms. Sealed ionization chambers filled with argon gas have ha ve bee been n use used d in con conven ventio tional nal ele electr ctron on lin linear ear accelerators for many decades and have a long lifetime before repl replacem acement ent or rep repair air is neces necessary sary.. If  the th e ac acce cele lera rato torr us used ed is ca capa pabl ble e of hi high gh in inst stan an-taneou tan eouss int intens ensiti ities, es, spe specia ciall car care e in the des design ign of  the th e mo moni nito tori ring ng sy syst stem em is re requ quir ired ed (C (Cou outr trak akon on  et al., 1991b). Positive-ion collection times as little as 20 ms ca can n be ac achi hiev eved ed,, an and d th the e ca capa pabi bili lity ty of 

and elsewhere are required for accurate control of 

handli han dling ng hig high h ins instan tantan taneou eouss int intens ensiti ities es can be

4.7 BEA BEAM M MON MONITO ITORING RING

70  

 DOSIMETRY   DOSIMETR Y 

ac acco comm mmod odat ated ed us usin ing g ga gase sess su such ch as he heli lium um in which ions have higher drift velocities. Proton nozzles can be divided into two categories: ones that use flat or contoured scatterers (passive systems) (Grusell   et al., 1994; Koehler  et al., 1977; Nauraye   et al., 1995) and ones that use magnetic defle de flect ctio ion n (s (sca cann nnin ing g or ac acti tive ve sy syst stem ems) s) (K (Kan anai ai  et al., 198 al.., 19 1980; 0; 198 1983; 3; Pe Pedr droni oni   et al 1995 95;; 20 2005 05)) to spre sp read ad th the e be beam am la late tera rall lly y (s (see ee Se Sect ctio ion n 3 fo forr details). deta ils). Scan Scanning ning sys systems tems add addi addition tional al requ requireirements men ts for th the e bea beam m mon monito itors rs and are dis discus cussed sed separ sep arat ately ely her here e (se (see e Se Secti ction on 4.7 4.7.3) .3).. Fo Forr exa exampl mple, e, scanning systems require devices capable of moni-

 A solution is to perform upstre upstream am range measurements men ts in rea reall tim time e (du (durin ring g pa patie tient nt tre treat atmen ment) t) by locating suitable detectors just outside the periphery of the beam and upstream of a fixed or variable patient pat ient collimator collimator.. This requires requires tha thatt some of the beam strike the range detector either continuously or, in the case of scanned beams, periodically as the beam bea m sw sweep eepss ov over er th the e ra range nge det detect ector or.. Re Rela lativ tive e preci pr ecisio sion n of ran range ge mea measu surem rement entss sho should uld be les lesss than th an 0.5 mm mm.. If th this is lim limit it is exc exceed eeded, ed, th the e cau cause se ( e.g., beam beam-ener -energy gy drift drifts, s, mov movemen ementt of beam beam-line -line compon com ponent ents) s) sh shoul ould d be inv invest estiga igated ted.. Som Some e re reccommendations for energy/range monitoring during

to tori and d meter) cont co ntro roll llin ing gving the th posi po siti tion on of det a ectors smal sm alll ( ring 1ng cm an diamet dia er) movin mo g ebea beam. m. Such Suc h detect ors need to have fast response times (,, 1 ms) in order to tr trac ack k the re realal-tim time e mot motion ion of the bea beam m acr across oss the target. The requirement for detector bandwidth depe de pend ndss on th the e sc scan anni ning ng sp spee eed d in th the e pa pati tien ent, t, which can be quite high, typically 1 beam diameter per millisecond. Detectors such as strip or segmented ionization chambers (Cirio   et al., 2004) can be effect eff ective ive,, and mu multi ltiwir wire e pr propo oporti rtiona onall cha chambe mbers rs

patient treatments are discussed in Section 4.7.5. 4.7.1 Ioniz Ionization ation chamb chambers ers

Paralle Para llel-p l-pla late te ion ioniza izatio tion n cha chambe mbers rs ar are e mos mostt fr freequently selected as integral dose monitors. Air- or argo ar gonn-fil fille led d ch cham ambe bers rs ar are e ty typi pica call ch choi oice cess fo forr passiv pas sive e bea beam-s m-spr pread eading ing sy syst stems ems.. An ano anodede-totocathode spacing, or gap thickness, of 3–10 mm has been used for therapy applications. For monoenergetic protons protons cros crossing sing a gas-fi gas-filled lled gap with negli-

(MWPCs) (MWPC s) (C (Char harpak pak,, 197 1970; 0; Sa Sauli uli,, 197 1977) 7) can als also o be used. Energy or range measurements made anywhere upstr ups tream eam of the tr trea eatme tment nt pos positi ition on mu must st be cal calii-

gible ener energy gy loss (and therefore therefore cons constant tant stopping stopping power), the signal strength can be calculated from the charge per proton,  Q , produced in the gas:

brated brat ed ag agai ains nstt ra rang nge e me meas asur urem emen ents ts ma made de in a water phantom at the treatment position, as these latter measurements pertain to the beam actually entering the patient. This is particularly important because the range attenuation of nozzle materials can be accurately predicted only if adequate knowledge of the material compositions, densities, thicknesses,  etc., is available. Furthermore, it is possible thatt com tha compon ponent entss ma may y ha have ve fal fallen len out of, or int into, o, the bea beam m lin line. e. A sui suitab table le det detect ector or for up upst stre ream am measurements is a multilayer Faraday cup (MLFC) that is constructed of alternate plates of metal and et al., 1999; insula ins ulatin ting g ma mater terial ial (Go (Gott ttsch schalk alk   et Paganett Paga nettii and Gott Gottscha schalk, lk, 2003; Schr Schreude euderr   et al., 2001). 200 1). The met metal al pla plates tes ar are e con connec nected ted to cha charge rge integr int egrat ators ors,, the sig signa nals ls fr from om wh which ich ess essent ential ially ly prov pr ovide ide a cha charge rgedd-par partic ticle le flu fluenc ence e dis distri tribut bution ion (spectrum (spec trum)) from whic which h the effec effective tive prot proton-b on-beam eam energy can be obtained. Alternatively, a multiplate ioniza ion izatio tion n cha chambe mberr can be use used d to det determ ermine ine the depth–dose distribution from which the range can be determined (Siebers  et al., 1988). Clearly Clea rly,, energ energy/ra y/range nge monit monitoring oring is need needed ed for both active and passive systems, but the accuracy and th the e fr frequ equenc ency y of th the e ene energy rgy mea measur sureme ements nts requir req uired ed ar are e mu much ch hig higher her for sca scanni nning ng sy syst stems ems because different ranges must be delivered sequentially in layers and with high accuracy to generate the required dose distribution in the target volume.

Q¼

 Sgas tð1:602  1019 Þ ; ðw= eÞ p

ð4:23Þ

 

where  S gas  ( 2 d E /d x) is the linear proton stopping power in the gas filling,   t  is the gap thickness, and w p   is the pr proto oton n ene energy rgy requir required ed to for form m an ion pair in the gas. For the gases discussed here, the proton stopping powe po werr dec decrea reases ses app approx roxima imatel tely y inv invers ersely ely wit with h energy ene rgy (1/  E) in th the e th ther erap apy y en ener ergy gy ra rang nge. e. Th The e charge produced per proton is 2.5 times larger at 70 MeV than at 250 MeV (ICRU, 1993a). Clearly, a larger lar ger gap thi thickn ckness ess wil willl inc increa rease se the cha charge rge col col-lected per unit dose but will also result in longer ionio n-co coll llec ecti tion on ti time mess an and d po pote tent ntia iall lly y in incr crea ease sed d detector-saturation effects due to ion recombination at high intensities. Typi ypical cal ele electr ctric ic fiel field d st stre rengt ngths hs of 3– 5 kV cm21 can ca n be us used ed fo forr ai airr or ar argon gon be befo fore re br brea eakd kdow own n 21 occurs, whereas 1–2 kV cm are ar e ty typi pica call lly y max axim imu um fiel eld d str tren engt gth hs fo forr hel eliu ium m-fi -fill lled ed chamb cha mbers ers.. Th The e goa goall of cha chambe mberr des design ign is to ke keep ep the th e co coll llec ecti tion on ef effic ficie ienc ncy y at 10 100 0 pe perc rcen entt fo forr th the e highest intensities that occur during normal operation at ion.. Wh When en des design igning ing cha chambe mbers rs for use in hig highhintens int ensity ity bea beams, ms, hel helium ium wil willl be sup superi erior or to bot both h air and argon (Table 4.16) owing to the fewer ion pairs produced per proton in the gas and the lower ¼

71  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  Table 4.16. Data for a parallel-plate ionization chamber with a 5 mm gap filled with air, argon, and helium at STP. STP.

Gas

Field strength (kV cm21)

Positive ion velocity (cm ms21)

Positive ion Positive collection time (ms)

w   (e (eV) V)

No. of io ion n pa pair irss pe perr pr prot oton on

 Ep

 Air  Argon Helium

4 4 2

5 6.8 20

100 74 25

34.2a 26.6b 45.2b

¼

70 MeV 

 

160 460 18

Ep

¼

230 MeV 

64 190 7

a

IAEA (2000). ICRU (1979).

b

ion-collection times (Boag and Wilson, 1952). Once the chamber is built, ion-collection efficiency as a functi fun ction on of hig high h vo volta ltage ge sho should uld be mea measur sured ed to ensu en sure re 10 100 0 pe perc rcen entt co coll llec ecti tion on ef effic ficie ienc ncy y at th the e hig igh hes estt in intten ensi siti ties es.. A gu guar ard d ri rin ng sh shou ould ld be

mixture in order to maintain a known   w-value for the ionization chambers. 4.7.2 Posi Position tion and dose unifo uniformity rmity

included in the construction included construction to mini minimize mize leakage curren cur rents. ts. To all allow ow app appro ropri priat ate e cor correc rectio tions ns to be made ma de,, a se sepa para rate te te temp mper erat atur ure e se sens nsor or ma may y be required in the nozzle as the temperature here may differ from that in the room or in the phantom in which the doses are measured. For sc scan anne ned d pr prot oton on be beam ams, s, th the e io ioni niza zati tion on

In par parall allelel-pla plate te ion ioniza izatio tion n cha chambe mbers, rs, one of the electrode planes is held at high voltage, either positive or negative, whereas the other plane is used to collect ions of one polarity. Without segmentation of  the col collec lectio tion n ele electr ctrode ode,, the ent entire ire bea beam m is int interercepted on a single foil and a signal proportional to the total inte integra grated ted intensity intensity or dose is reco recorded. rded.

chambers must detect the ionization in times short comp co mpar ared ed wi with th th the e sw swee eep p sp spee eed d of th the e be beam am.. Typ ypic ical al be beam am sw swee eep p sp spee eeds ds ar are e 1– 3 cm ms21. Therefor Ther efore, e, subsub-milli millisecond second inte integra gration tion time timess are required. For typical ionization chamber gaps of 3– 5 mm and electric field strengths of 3–5 kV cm 21, helium or argon should be used to keep positive-ion-collection positive-ion-co llection times   , 1 ms (th (the e ele electr ctrononcollec col lectio tion n tim times es ar are e sev severa erall ord orders ers of mag magnit nitude ude faster and comprise only 50 percent of the signal). Table 4.16 shows data for a chamber with a 5 mm gap filled with air, argon, and helium. Higher electric fields and smaller gaps will decrease collection times proportionately and reduce ion-recombination effects. Helium-filled chambers are capable of operating at ing in hig higher her int intens ensity ity bea beams ms and ha have ve fas faster ter response times. If a no nobl ble e ga gass is us used ed fo forr fil filli ling ng,, ga gass pu puri rity ty bec ecom omes es im imp por orta tan nt (IC ICR RU, 19 1979 79;; Je Jessse an and d Sadauskis Sada uskis,, 1952; 1953). The so-ca so-called lled Jesse effec effectt results in a large (up to 50 percent) decrease in the w- values (and therefore increases in the number of  ion pairs formed) at relatively small concentrations (  0.1 percent) of impurities in the noble filling gas. This is due to metastable excited states of the noble gas producing ionization in the contaminant gas by energy ene rgy tr trans ansfer fer.. Med Mediat iation ion of th the e unc uncert ertain ainty ty in ionization due to the Jesse effect is possible by purpose po sely ly ad addi ding ng en enou ough gh co cont ntam amin inan antt to th the e ga gass

Often the elect electrode rode is segme segmented nted,,   e.g., int into o qua quaddrants, to determine the position of the beam relative to the center of the ion chamber, or equiva equ ivalen lently tly, th the e cen centr tral al axi axiss of the noz nozzle zle.. The mass of the ion chambers can be kept low by constructing the electrodes from very thin foils with a micros mic roscop copic ic la laye yerr of gol gold d (or oth other er met metal) al) eva evapor por-ated onto the surface. Standard photo-etching techniques niq ues can be use used d to seg segmen mentt the electrod electrode e int into o quadrants, concentric rings, or large square arrays. In the case of passively scattered beams, it can be useful to place two quadrant-segmented chambers in the nozzle, one before and one after the scattering system, to monitor the position and inclination of the axis of the beam relative to the central axis of the nozzle. When deviations deviations of the beam posit position ion or ang angle le exc exceed eed th the e spe specifi cified ed tol toler eranc ance, e, the bea beam m mustt be inh mus inhibi ibited ted to av avoid oid large flu fluenc ence e or dos dose e non-uniformity in the patient. Alternatively, appropriate signals must be fed back to steering magnets to correct automatically the beam position or beam trajectory. Fast Fa st-r -recy ecycli cling ng int integr egrat ators ors (Go (Gotts ttscha chalk, lk, 198 1983; 3;   et al. Renner , 1989) are often used to digitize the charge collected on each quadrant and to calculate center cen tering ing inf inform ormat ation ion.. Ad Addit dition ionall ally y, ion ioniza izatio tion n chamb cha mbers ers wit with h lar large ge arr array ayss of squ square are seg segmen ments ts (from (fr om 20  20 up to 80  80 squ squar ares) es) ha have ve bee been n cons co nsid ider ered ed fo forr a si sing ngle le pl plan anar ar el elec ectr trod ode e (4 (40 0 2 40 cm area) yielding high resolution when placed 72

 

 DOSIMETRY   DOSIMETR Y 

within 0.5 m or so from the patient. Recent results (Cirio   et al., 2004) have demonstrated the efficacy of a two-dimensional pixel ionization chamber with a 32  32 matrix of 1024 cylindrical ionization cells arranged in a 24  24 cm2 area with 1 ms digitization at ion and re reado adout ut tim times. es. The tis tissue sue-eq -equiv uivale alent nt thickn thi ckness ess of th the e det detect ector or is onl only y 1 mm. For pas pas-sively modified beams, response uniformity can be determine deter mined d with respect to know known n beam profiles, profiles, while the absolute dose response for any pixel can be obtained using a standard calibrated cylindrical ioniza ion izatio tion n cha chambe mberr un under der re refer ferenc ence e con condit dition ionss (Amerio   et al., 2004) 2004).. For scanning scanning beams, dete detecctors with good spatial resolution are also attractive beca be caus use e th the ey ca can n de dete tect ct an and d re reco cord rd fu full ll tw twoo-

about 10 ms. Int about Integr egral al bea beam m int intens ensity ity at an any y pos pos-itio it ion n is mo moni nito tore red d by pa para rall llel el-p -pla late te io ioni niza zati tion on chamb cha mbers ers wit with h re respo sponse nse tim times es of    , 100 m s (L (Lin in  et al., 1994; Pedroni  et al., 1995). In addition to their use as integral monitors with all the ano anode de wir wires es con connec nected ted tog togeth ether er (ab (abov ove), e), MWPCs can also be used for monitoring the position of scanning beams. These chambers use thin (25–50 mm di diam amet eter er)) tu tung ngst sten en or co copp pper er wi wire ress with a 1–2 mm pitch and two thin kapton foils for contai con tainin ning g the gas gas.. The MW MWPC PCss ha have ve bee been n bui built lt with wit h 2 mm spa spatia tiall res resolu olutio tion n and a 15 kHz data acquisit acq uisition ion rat rate e (Bru (Brusasco sasco   et al., 2000) at the Gesellschaft fu ¨ r Schwerionen Schwerionenforschung forschung (GSI). Two of these MWPCs are part of the beam-monitoring

dimensional dose information at or near the target with real-time tracking of the beam movement.

system and provide active feedback to the scanning magn ma gnet etss to en ensu sure re co corr rrec ectt po posi siti tion onin ing g of th the e carbon-ion beam. The cathode planes are made of  closely-spaced wires strung on an epoxy frame. The disadvan disa dvantages tages of MWP MWPCs Cs are that they only give the th e pr proje ojecti ction on of th the e bea beam m sha shape pe ont onto o the   x- or  y-a -axi xiss an and d no nott th the e tr true ue tw twoo-di dime mens nsio iona nall be beam am shape. Nevertheless, GSI and other institutions are pursuing this option because of their ease of construction, high data-acquisition rate capability, and

4.7.3 Cons Considera iderations tions for scann scanned ed beams beams

In addi addition tion to ach achievin ieving g real real-tim -time e tra tracking cking of the motion of pencil beams across the patient, special attention must be given to the location and mass of  the monitors that may contribute to beam enlargementt fr men from om mul multip tiple le Co Coulo ulomb mb sca scatte tterin ring. g. In the energy range 70–250 MeV, even thin (25  mm) foils of kapton or titanium (used for vacuum windows) can con contri tribut bute e to a sig signifi nifican cantt inc incre rease ase in bea beam m size at the patient if they are placed at a distance of several meters from the target. In addition, 3 m of air in the nozzle will also have noticeable effects on th the e be beam am si size ze.. Var ario ious us so solu luti tion onss ha have ve be been en applied to reduce these problems. In general, it is best to keep all beam monitors within a distance of  70–100 cm fr from om th the e pa pati tien entt an and d us use e th the e lo lowe west st mass ma ss of ma mate teri rial al po poss ssib ible le fo forr co cons nstr truc ucti ting ng th the e chamber. Gold-plated kapton foil parallel-plate ionization chambers are ideal for this purpose, but the detect det ector or siz size e re requi quire remen ments ts ar are e lar large ge (u (up p to 40 

high reso resolutio lution. n. Alte Alterna rnative tively ly,, twotwo-dimen dimensiona sionall pixel pix el det detect ectors ors,, lik like e th the e one des descri cribed bed in Se Secti ction on 4.7.2,, or tra 4.7.2 track ck dete detectors ctors (Sunaga   et al., 1988) can give good position information for scanned beams.

40 cm2) an and d ne nece cess ssit itat ate e hi high gh st stre retc tchi hing ng fo forc rces es when fixing the foils to a frame in order to keep electrostatic bulging to a minimum. In such cases, the det detect ector or re respo sponse nse can be cal calibr ibrat ated ed at man many y diff di ffer eren entt re regi gion onss of th the e de dete tect ctor or (S (Sau auli li,, 19 1977 77). ). Similarly, a multiwire proportional counter (MWPC) (Charpak, 1970; Sauli, 1977, 1992) operated in ioncollection mode can be used as an integral intensity monit mon itor or wh when en fil fille led d wit with h a no nobl ble e ga gas, s, an and d all the anod an ode e wir wires es ar are e con conne nect cted ed to toge geth ther er to fo form rm on one e readout channel through a recycling integrator. Up to double the signal strength can then be achieved for the sam same e ano anodede-toto-ca catho thode de spa spacin cing g as with the parallel-plate chamber described above.  At the PSI, a 1l volume of tissue is typically filled with 10 000 ‘spots’ with a spacing of 5 mm in 2 min min,, giv giving ing a mea mean n tr trea eatm tment ent tim time e per spo spott of 

1 MHz, but they have at least three orders of magnitude fewer ions produced per passing proton than do ion ioniza izatio tion n cha chambe mbers. rs. The These se det detect ectors ors can be calibrated in the same way as ionization chambers. Their advantage is that they cannot saturate when the th e int intens ensiti ities es bec become ome ac accid cident entall ally y too hig high h an and d there th erefor fore e th they ey pr prov ovide ide an add additi itiona onall mea measur sure e of  safety saf ety tha thatt ioniz ionizati ation on cham chambers bers cann cannot. ot. How However ever,, beca be caus use e th they ey ar are e lo loca cate ted d in va vacu cuum um,, th they ey ar are e usuall usu ally y sev severa erall met meters ers fr from om th the e pa patie tient nt an and d can add ad d si sign gnifi ifica cant ntly ly to be beam am en enla larg rgem emen ent. t. Th This is becomes becom es prob problema lematic tic for scann scanning ing beams beams,, but not for passive beams. More Mo re re rece cent ntly ly,, th the e tr tren end d ha hass be been en no nott to us use e SEMs, but rather to find a safe way for ionization chambers to inhibit the beam if intensities increase into int o the sa satur turat ation ion re regio gion. n. Not Note e th that at if pas passiv sive e

4.7.4 Secon Secondary-em dary-emissio ission n monit monitors ors

Second Seco ndar aryy-em emis issi sion on moni mo nito tors rs (Chu (C hu,, 1995 19 95b; b; Karzmark, 1964; Tautfest and Fechter, 1955) have been bee n use used d at sev sever eral al fa facil ciliti ities. es. Th These ese det detect ectors ors consist of several parallel thin-metal foils mounted in a vacuum. The foils emit electrons from the surfaces fa ces wh when en ene energe rgetic tic bea beams ms pas passs thr throug ough h th them. em. The Th e ba band ndwi widt dth h of th thes ese e de dete tect ctor orss is we well ll ab abov ove e

73  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

bea beam m spr spread eading ing is emp employ loyed ed do downs wnstr tream eam of th the e ionizati ioniz ation on cham chamber ber in ques question tion,, the dose rat rates es or intensities are much lower than a similar chamber located upstream of the beam-spreading devices for the same intensity accelerator beam. If the associated electronics,   e.g., recy recycling cling integrators, integrators, can be used to detect a dose rate that produces saturation, then the beam might be inhibited before an unsafe condition prevails. In a rare case with an electron-therapy linear accelerator, the monitor ionizat iz atio ion n ch cham ambe bers rs fa fail iled ed un unde derr ex extr trem emel ely y hi high gh

4.8 REL RELA ATIVE DOSI DOSIMET METR RY  Calibrations of proton-therapy systems are carried out in reference conditions as described in Sections 4.2–4.5. Relative dose measurements or measurement me ntss in no nonn-re refe fere renc nce e co cond ndit itio ions ns re requ quir ire e no detector dete ctor calibration calibration other than the veri verificat fication ion of  the th e li line near arit ity y of re resp spon onse se wi with thin in th the e as assu sume med d dynami dyn amicc ra range nge of th the e mea measur sureme ement nt con condit dition ions. s. Rela Re lati tive ve do dose se me meas asur urem emen ents ts ar are e em empl ploy oyed ed fo forr routine rout ine daily clini clinical cal phy physics sics acti activitie vities, s, beambeam-line line

commissioning, commission ing, and collec collecting ting dat data a for trea treatmen tmenttplan pl anni ning ng sy syst stem ems, s, qu qual alit ity y as assu sura ranc nce, e, an and d fo forr research and development. Measurem Meas urements ents made duri during ng the comm commission issioning ing of proton-beam-delivery systems and computerized treatmen trea tment-pl t-plannin anning g sy system stemss inclu include de mapp mapping ing of  the clinical radiation fields involving unmodulated beam bea m dep depth– th– dos dose e cur curves ves and la later teral al pr profil ofiles es and dose dos e dis distri tribut bution ionss of pe penci ncill bea beams ms use used d for sca scannning. nin g. Fr From om the these se mea measur sureme ement nts, s, fun fundam dament ental al beam characteristics such as beam range and symmetry can be determined. These characteristics are necessary for the design and control of the beamdelivery system and for fitting the data for dose calculat cul ation ion alg algori orithm thms. s. Fo Forr ex examp ample, le, un unmod modula ulated ted depth– dep th– dos dose e da data ta ar are e use used d to des design ign range range mod moduulators; depth–dose distributions for different beam cross-sections, including pencil beams for scanning and an d la late tera rall pr profi ofile les, s, ar are e us used ed as in inpu putt da data ta fo forr trea tr eatm tmen entt-pl plan anni ning ng sy syst stem ems. s. On Once ce th the e be beam am-delive del ivery ry sy syst stem em is rea ready dy for pa patie tient nt tr treat eatmen ment, t, clinic cli nical al bea beam m fiel field-m d-mapp apping ing mea measur sureme ements nts ar are e made for the range of treatment conditions. Proton patient portals require individual physical calibration measurements of the output factors for computation of the treatment dose-monitor settings from fr om a ca cali libr brat atio ion n mo mode del. l. Th The e ou outp tput ut fa fact ctor or is dete de term rmin ined ed as th the e ra rati tio o of co corr rrec ecte ted d do dosi sime mete terr read re adin ings gs at th the e re refe fere renc nce e de dept pth h   zref    measured under a given set of non-reference conditions relative to that measured measured unde underr refe referenc rence e cond condition itionss (as given in Table 4.7). Individual patient calibrations for passive and active beam-delivery systems are ar e usu usuall ally y per perfor formed med wit with h a dos dosime imetry try sy syst stem em having a single ionization chamber, or an array of  ionization chambers with a known dose-calibration coefficient coeffi cient rela relative tive to the pro proton ton fac facility’ ility’ss prima primary ry dose sta standar ndard. d. Kooy   et al.   (2003) derived a model that th at pr pred edic icte ted d ou outp tput ut fa fact ctor orss fo forr SO SOBP BP pr prot oton on fields, which agreed with measurements within 2.9 percent.

intensities before any shut off signal could be generated. Each facility must ensure that such a scenario ari o is not pos possib sible le wit with h pr proto oton-t n-ther herapy apy bea beams. ms. Finally, the ionization chamber should be tested for the high high-inte -intensit nsity y cond condition ition used to abort patient patient treatment.

4.7.5 Range/ Range/energy energy measur measurements ements

Energy measurements can be made in the accelerator, in the beam-transport system, or in the treatmentt noz men nozzle zle.. At Lom Loma a Lin Linda da Un Unive iversi rsity ty Me Medic dical al Center, for example, the beam position and sity monitors in the synchrotron are used to intencalculate proton velocity from the frequency and radial position. This information is used to inhibit beam deli de live very ry if th the e de deri rive ved d en ener ergy gy va valu lue e fr from om th this is measurem measu rement ent exce exceeds eds a toler tolerance ance of   + 0.1 MeV. MeV. It is mea measur sured ed sev sever eral al tim times es thr throug oughou houtt ea each ch 0.2 s beam spill. For a cyclotron that generates fixed proton energies gi es be betw twee een n 20 200 0 an and d 25 250 0 Me MeV V, it is mo most st co conn venient to measure the energy by verifying that the th e co corr rrec ectt am amou ount nt of ran ange ge sh shif ifte terr ha hass be been en inserted in the beam for a treatment field. It is also possible to monitor the field strength of a bending magnet in the beam line after the energy has been select sel ected. ed. In th this is cas case, e, the magneti magneticc fiel field d is con con-strained to be a certain value consistent with the intended inte nded proton ener energy gy deliv delivered ered.. Only one magnetic field strength (within some tolerance) will be acceptable for the beam to be enabled. A beam with the incorrect energy entering the bending magnet field will be bent at the wrong angle, and therefore will wi ll be lo lost st in th the e be beam am-t -tra rans nspo port rt sy syst stem em or be steer st eered ed int into o the noz nozzle zle at an inc incorr orrect ect pos positi ition. on. Either condition will induce an aborted treatment.  A third scenario for beam-energy monitoring is to include a range detector in the nozzle, but at a location just outside the active field and upstream of a pass passive ive collimator collimator,, which can dete detect ct part particles icles outside outs ide the beam apert aperture. ure. Mult Multipla iplate te ioniz ionizatio ation n chambers or MLFCs can be used for this purpose (see the beginning of Section 4.7).

4.8.1 4.8 .1

Phantom Pha ntom ma materi terials als

The ir The irra radi diat atio ion n medi medium um fo forr cl clin inic ical al do dosi sime metr try y should shoul d simu simulate late the patient patient as closel closely y as possi possible ble 74

 

 DOSIMETRY   DOSIMETR Y 

in ter terms ms of ph phant antom om com compos positi ition, on, loc locat ation ion,, and extent. exte nt. Wate aterr is the sta standar ndard d ref referen erence ce medi medium. um. The use of plastic phantoms is strongly discouraged because, in general, they are responsible for discrepanc pa ncie iess in th the e de dete term rmin inat atio ion n of ab abso sorb rbed ed do dose se (IAEA, 2000). Plastic phantoms should not be used for ref refere erence nce dos dosime imetry try in pr proto oton n bea beams ms bec becaus ause e

charac char acte teri rist stic icss fo forr plastic, and water.

PMMA PM MA,,

poly po lyst styr yren ene, e,

4.8.2 Detecto Detectors rs for dose-d dose-distri istributio bution n measurements

A-15 A150 0

the flue fluence nce cor correc rectio tion n fa facto ctors, rs,   hpl, for sc sca ali lin ng absorbed dose to water in plastic to absorbed dose to wa water ter in wa water ter at the sam same e wa water ter-eq -equiv uivale alent nt depth are not very well known. The differences in proton fluence distributions are almost entirely due to di diff ffer eren ence cess in th the e no nonn-el elas asti ticc nu nucl clea earr cr cros ossssections secti ons betw between een the plas plastic tic mat materials erials and wa water ter.. Neverthe Nev ertheless, less, when accu accurat rate e cham chamber ber posit positionin ioning g in wa wate terr is no nott po poss ssib ible le or wh when en no wa wate terp rpro roof  of  chamber is available, their use is permitted for the measureme measu rement nt of dept depth h – dose distribution distributionss for lowenergy ene rgy pr proto oton n bea beams ms (be (below low    100 Me MeV) V) (IA (IAEA, EA, 2000). It can be assumed that   hpl   has a cons constant tant  value of unity at all depths for such low-energy broad beams (Palmans  et al., 2002b), as the contributions butio ns from non-elastic non-elastic inte interac ractions tions to the tota totall

Detectors Detect ors emp employ loyed ed for re relat lativ ive e dos dosime imetry try mu must st have hav e the appr appropria opriate te sensi sensitivit tivity y, energ energy y inde indepenpendence, response linearity, and spatial resolution for each clinical dosimetry dosimetry task. Depending Depending upon the task, tas k, ion ioniza izatio tion n cha chambe mbers, rs, sil silico icon n dio diodes des,, ra radio dio-graphic films, diamond detectors, gels, scintillators, thermoluminescence dosimeters (TLDs), and radiochromic films can be employed (Chu   et al., 1993). The time structure of the beam must also be consider sid ered ed for cli clinic nicalal-dos dosimet imetry ry mea measur sureme ement nts. s. A  detector must dwell at the same location for many beam be am cy cycl cles es to ob obta tain in re repr prod oduc ucib ible le re resu sult lts. s. Exam Ex ampl ples es of cy cycl clic ic be beha havi vior or th that at mu must st be co connsider sid ered ed inc includ lude e the ac accel celera erator tor du duty ty cyc cycle, le, tim time e structu str ucture re with within in a trea treatmen tmentt beam pulse, puls pulsed ed irradia irra diation tion of a rota rotating ting modu modulat lator or prop propeller eller,, and

,

do dose se are e apply very ve ryysm smal l ( her 1 per percen cent). Note te cil that th at thiss thi does doe s ar not appl toall higher hig ener energy gyt).orNo pencil pen beams al.., 200 al.., 20 (Palmans   et al 2002b; 2b; Sch Schnei neider der   et al 2002 02a) a) where corrections of 2–5 percent have been found. The de dens nsit ity y of the pl pla astic ic,,   r pl, should be meas me asur ured ed fo forr th the e ba batc tch h of pl plas asti ticc in us use e ra rath ther er than th an usi sin ng a nom omin ina al va valu lue e for th the e pla lasstic type ty pe.. Ea Each ch me meas asur urem emen entt de dept pth h in pl plas asti ticc   zpl (expressed in g cm22) must also be scaled (IAEA, 2000) to give the corresponding depth in water   zw (in g cm22) by  zw  ¼  zpl cpl ;

 

dynamic beam spreading. Dosimetry measurement syst sy stems ems wit with h mu multi ltiple ple det detect ectors ors in one one-, -, tw two-, o-, or thre th ree-d e-dime imensi nsiona onall arr array ayss sa save ve tim time e in mea measur sureement me ntss of do dose se di disstr trib ibut utio ions ns,, pa part rtic icul ular arly ly in dyna dy nami micc sy syst stem ems. s. It sh shou ould ld al also so be no note ted d th that at many relative dose measurements employ a separate at e det detect ector or as a ref refere erence nce mon monito itorr to cor correc rectt for dose-rate dose-rat e variations. The Th e me meas asur urem emen ents ts us used ed fo forr co cont ntro roll llin ing g th the e dose and the shape of the dose distribution delivered with scanning beams differ from those with passive scattering only from the point of view of  the th e lon longer ger tim time e nee neede ded d for th the e mea measur sureme ement nts. s. With a dynamic method one needs to accumulate

ð4:24Þ

the time time-vary -varying ing sign signal al of the dosi dosimete meterr pla placed ced at a fixed point in a phantom over the full time of a sc scan an.. Ob Obvi viou ousl sly y, me meas asur urem emen ents ts ba base sed d on dose rates using moving detectors cannot be used in sc scan anne ned d be beam ams. s. Th The e si simu mult ltan aneo eous us us use e of  several dosimeters is recommended. These should be dis distr tribu ibuted ted eit either her at di discr scret ete e pos positi ition onss in a  volume or aligned along arrays ( for measuring dose do se pr profil ofiles) es).. Alt Altern ernat ativ ively ely,, two two-d -dime imensi nsion onal al dosimeters such as films, scintillating screens, or gel do dosim simete eters rs can be em emplo ploye yed. d. Th The e cho choice ice of  ionization chambers is a compromise between the sensitiv sens itive e vol volume ume and spa spatial tial res resolut olution. ion. Sma Smaller ller chambers accumulate a relatively high nois no isee-to to-s -sig igna nall ra rati tio o ov over er th the e du dura rati tion on of th the e scan sc an.. Io Ioni niza zati tion on ch cham ambe bers rs wi with th vo volu lume mess of   0.1 cm3, which give a relative dose precision of  a few percent, can be used, but smaller chambers are not viable.

where   cpl   is a dep depthth-sca scalin ling g fa facto ctorr. Fo Forr pr proto oton n beams,   cpl   can can be cal calcul culat ated, ed, to a goo good d app approx roxiimatio ma tion, n, as the ra ratio tio of csd csda a ra range ngess in gra grams ms per squar squ are e cen centim timete eterr (T (Tabl able e 4.3 4.3)) (IC (ICRU RU,, 199 1993a) 3a) in water and in plastic. For example, the depth-sc dept h-scaling aling fact factor or   cpl   has has va valu lues es of 0. 0.97 974 4 fo forr PMMA and 0.981 for clear polystyrene for nominal densit den sities ies of 1.1 1.190 90 and 1.0 1.060 60 g cm22, res respecti pectively vely (IAE (I AEA, A, 20 2000 00). ). If a pl plas asti ticc ph phan anto tom m is us used ed to measur mea sure e the bea beam-q m-qua ualit lity y ind index, ex, th the e mea measur sured ed quantity is the residual range in the plastic,  R res,pl. The residual range,  R res, in water is obtained using the scaling factor   cpl   given in Eq. (4.24). Palmans and Verhaegen (1997) have performed Monte Carlo calcu cal cula lati tion onss of de dept pth– h– do dose se di dist stri ribu buti tion onss fo forr 50– 250 MeV pro proton ton beam beams. s. Allo Allowin wing g for the dif differ ferent ent mater ma terial ial com compos positi itions ons the they y fou found nd exc excelle ellent nt agr agreeeement between the proton depth –d – dose 75  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

 4.8.2.1

Single ionization chambers

Output Outp ut fact factors ors are meas measured ured at refe referenc rence e dept depths hs

 Additionally, because of lattice damage (Knoll,  Additionally, 1989 19 89), ), th the e se sens nsit itiv ivit ity y pe perr un unit it ab abso sorb rbed ed do dose se varies with the magnitude of previous exposure.

 varies with the magnitude of previous exposure. Such lattice damage is dependent upon the type of  part pa rtic icle le pr prod oduc ucin ing g th the e de defe fect cts, s, wi with th gr grea eate terr damage resulting from more massive particles. The relative damage from equal doses of 8 MV x rays, 20 MeV el ele ect ctrron ons, s, an and d 70 MeV prot oto ons was measured, and a relative damage ratio of 1:20:40, respectiv resp ectively ely,, was esta establish blished ed (Rik (Rikner ner,, 1983) 1983).. For 70 MeV proton bombardment, the damage from a 10 kG kGy y ex expo posu sure re re redu duce cess th the e se sens nsit itiv ivit ity y to 30 percen per centt of th the e val value ue pri prior or to irr irradi adiat ation ion (IC (ICRU RU,, 1998). Newhause Newhauserr   et al.   (2002a) (2002a) show showed ed tha thatt the sensit sen sitivit ivity y of a dio diode de use used d ro rout utine inely ly in an ocu ocular lar proton beam line decreased to    40 percent because of ra radia diatio tion n da damag mage e ac accum cumula ulated ted ov over er a 5 ye year ar period. When Si diodes are used for absorbed-dose determina mi nati tion onss fo forr pr prot oton ons, s, an en ener ergy gy-d -dep epen ende dent nt res esp pons nse, e, di difffe ferren entt fr from om th tha at of io ion niz iza ati tion on chambe cha mbers, rs, is oft often en see seen. n. Ho Howe weve verr, thi thiss eff effect ect is largel lar gely y eli elimin minat ated ed whe when n HiHi-p p Si dio diodes des ar are e use used d (see below). Koehler (1967) and Raju (1966) found a dis iscr crep epa anc ncy y bet etw ween dio iode de res esp pons nse e an and d gasga s-iion oniz iza ati tio on cham ch ambe berr resp spon onse se nea earr the th e Bragg-pe Brag g-peak ak regi region. on. Such resu results lts are consi consisten stently tly obse ob serv rved ed wi with th bo both th mo mono noen ener erge geti ticc an and d SO SOBP BP proton beams (Coutrakon  et al., 1991b; Onori  et al., et al., 19 2000; Schr Schreude euderr   et 1997 97). ). Fi Figu gure re 4. 4.6 6 sh show owss plots of depth–dose curves measured with a diode compared comp ared with thos those e meas measured ured with other detectors for monoenergetic and SOBP beams (Schreuder  et al., 1997). The large difference in the diode response compared with the other detectors is not explained explained by diff differen erences ces in stop stopping ping power

(middl (mid dle e of SO SOBP BP or in th the e pl plat atea eau) u).. Fo Forr la larg rger er fields, fiel ds, th the e dos dose e dis distri tribut bution ionss ar are e un unifo iform rm at th the e 3 referenc refe rence e posi positions tions and larger volume (0.5– (0.5 – 1 cm ) cylindrical ionization chambers are typically used. Depe De pend ndin ing g on th the e be beam am ran ange ge an and d fie field ld si size ze,, smaller chambers might have to be used. The use of extrapolation ionization chambers  et al., 1998). hasparallel-plate also been reported (Zankowski For dose do se-d -dis istr trib ibut utio ion n meas me asur urem emen ents ts,, the th e chamber must have good resolution in the direction of the distribution being measured because of the shar sh arp p di dist stal al an and d la late tera rall do dose se gr grad adien ients ts (M (Mob obit it  et al., 20 2000 00). ). Par aral alle lell-pl plat ate e ch cham ambe bers rs ar are e re reccommended for depth–dose measurements in large fields, wheras small (mini) cylindrical chambers (or other detectors) detectors) are reco recommen mmended ded for dept depth h – dose measur mea sureme ements nts in sma small ll fiel fields ds and for all la later teral al dose-profile measurement measurements. s. There are often significant discrepancies between measur mea sured ed Br Bragg agg cur curve vess and Mon Monte te Car Carlo lo cal calcuculations owing to the complexities of beam transport and an d of ac acco coun unti ting ng fo forr nu nucl clea earr in inte tera ract ctio ions ns an and d detector dete ctor geome geometry try (Boo (Boon, n, 1998) 1998).. This disc discrepa repancy ncy increases with detector resolution and is especially significan signi ficantt for prot proton on ener energies gies   , 100 MeV. Bichsel (1995) (19 95) use used d an ana analyt lytic ic mod model el to int interp erpre rett the these se effect eff ects. s. Ul Ulmer mer and Kai Kaissl ssl (20 (2003) 03) ha have ve app applie lied d a deconvolution model to depth–dose measurements with wit h dif differ ferent ent det detect ectors ors at 80– 180 Me MeV V, and th the e results show excellent agreement with Monte Carlo calculations.

 4.8.2.2

Silicon diodes

Semiconductor diodes have routinely been used for absorb abs orbeded-dos dose e mea measur sureme ement nts. s. Bec Becaus ause e of th their eir 3 small volume, typically    0.1 mm , excellent spatial reso re solu luti tion on wi with th go good od se sens nsit itiv ivit ity y is ach chie iev ved ed.. Because of these advantageous features, Si diodes have ha ve be been en wi wide dely ly us used ed in ra radi diat atio ion n do dosi sime metr try y et al., 20 et al., 20 (Bjo¨ rk   et 2000 00;; Bu Bucc ccio ioli lini ni   et 2003 03;; Gulbr Gul brand andsen sen and Mad Madsen sen,, 196 1962; 2; Koe Koehle hlerr, 196 1967; 7; Raju, 1966; Rikner, 1985; Smith  et al., 1977; Trump and Pinkerton, 1967; Wilkins  et al., 1997). In most cases, Si diodes are operated without external bias, in th the e so so-c -cal alle led d ph phot otov ovol olta taic ic mo mode de,, wh wher ere e th the e intrin int rinsic sic dep deplet letion ion reg region ionss ar are e use used d to pr produ oduce ce

between betwee n air and Si. Th This is wou would ld pr produ oduce ce a cor correc rec-tion incr increasin easing g the diff differen erence. ce. Col Columna umnarr reco recombimbinati na tion on ma may y co cont ntri ribu bute te to th the e ob obse serv rved ed ef effe fect cts. s. Sil ilic icon on di dio ode dess als lso o exh xhib ibit it a tem emp per era atur ureedepend dep endent ent res respon ponse se (Ko (Koehl ehler er,, 196 1967; 7; Ne Newha whause userr  et al., 2002a). This should be taken into account to achieve reproducible measuremen measurements. ts. et al.   (199 Case   et (1994) 4) mo mode deled led th the e re resp spon onse se of  thimble and parallel-plate ionization chambers and silicon silico n diod diodes es usin using g a sto stochas chastic tic prot proton on tran transport sport technique. techn ique. The They y demo demonst nstrat rated ed geome geometry-in try-induce duced d effect eff ectss in the re respo sponse nse of the thi thimbl mble e ion ioniza izatio tion n chamber that they claim might explain part of the differ dif ferenc ence e bet betwe ween en the re respo sponse nsess of dio diodes des and

charge flow (Klevenhage gen n, 1977a; 1977b; Maruhashi, 1977).  As the charge flow is by impurity carrier in the diod di ode e ju junc ncti tion on,, ver ery y hi high gh do dose se ra rate tes, s, ev even en of  lowlo w-LE LET T ra radi diat atio ion, n, pr prod oduc uce e a no nonl nlin inea earr do dose se response in n-Si diodes (Rikner and Grusell, 1987).

thimble thimbl e cha chambe mbers rs nea nearr the dis distal tal por portio tion n of the Bragg Br agg pea peak. k. Geo Geomet metric ric eff effect ectss wou would ld not ex expla plain in the diff differen erence ce in resp response onse betw between een plan plane-pa e-paralle rallell ionization chambers and silicon diodes. The use of  Hi-p Si diodes reduces these effects and the depth– dose dos e dis distri tribu butio tions ns in pr proto oton n bea beams ms are fou found nd to 76

 

 DOSIMETRY   DOSIMETR Y 

 et al., 2001; 2002). The calculation is based on the kno kn own en ener ergy gy di disstr trib ibut utio ion n an and d a me meas asur ured ed dens de nsit ityy-flu fluen ence ce ca cali libr brat atio ion n cu curv rve e of th the e fil film. m.  Although film does not necessarily give a response linear in absorbed dose, the simplicity and convenienc ie nce e of fil film m ma make kess it a ve very ry us usef eful ul me medi dium um fo forr study st udying ing th the e cha changi nging ng flue fluence nce dis distri tribut bution ion of a proton beam as it passes through matter.

correspo corre spond nd clo closel sely y to the dis distri tribut bution ionss obt obtain ained ed with plan plane-par e-parallel allel ioniz ionizati ation on chamb chambers ers (Gru (Grusell sell al.., 2002; Vatn and Med Medin, in, 200 2000; 0; Pa Pacil cilio io   et al atnitsky itsky  et al., 1999a). However, the sensitivity of this type of diod diode e decr decreases eases rapidly with accumulated accumulated dose. Grus Gr usel elll an and d Me Medi din n (2 (200 000) 0) sh show owed ed th that at Hi Hi-p -p Si detect det ectors ors giv give e a sig signal nal tha thatt is pr prop oport ortion ional al to th the e ionization density in the silicon crystal in all parts of the Bragg curve. This is in contrast to detectors based on n-type silicon, or on low-resistivity p-type silicon. After pre-irradiation, these latter detectors exhibit recombination effects that are related to the proton stopping power, yielding an increase in the detector signal per unit dose with increasing LET. This effect leads to an over-response in the Bragg peak, which increases gradually with the accumulated detector dose. Whatev Wha tever er the exp explana lanation, tion, silicon diod diodes es do not all giv give e a res respon ponse se th that at cor corres respon ponds ds to abs absorb orbed ed dose in tissue as measured by ionization chambers at pro proton ton energ energies ies   ,  20 MeV. To interpret their resp re spon onse se as ab abso sorb rbed ed do dose se to ti tiss ssue ue req equi uirres knowledg kno wledge e of the prot proton-fl on-fluence uence energy spec spectrum trum at ea each ch po poin intt of me meas asur urem emen ent, t, pa part rtic icul ular arly ly fo forr protons prot ons with energ energies ies   ,  20 MeV. Hence, caution is adv advise ised d wh when en usi using ng sil silico icon n dio diodes des for ene energy rgy-dependen depe ndentt pro proton ton dosim dosimetry etry.. How However ever,, the they y can safely be used to measure dose distributions at constant energy ( e.g., lateral beam profiles) as shown in Fig. 4.7 (Schreuder  et al., 1997).

 4.8.2.4

Radiog Rad iogra raphi phic-fi c-film lm dos dosime imetry try can be a ve very ry con con- venient method of measuring relativ relative e proton doses.

Small tis Small tissue sue-li -like ke int integr egrat ating ing dos dosime imeter terss all allow ow a more mor e pr preci ecise se det determ ermina inatio tion n of abs absorb orbed ed dos dose e at high hi gh sp spat atia iall re reso solu luti tion on th than an ma many ny ot othe herr do dosi si-meters. The crystalline amino acid,   L-a-alanine , is a good example (Bartolotta  et al., 1984; 1990; 1993; Ciesielski and Wielopolski, 1994; Waligo´rski   et al., 1989; Wieser  et al., 1993). Alanine is a solid hydrocarbon car bon in mic micro rocry cryst stall alline ine for form m tha that, t, whe when n bom bom-barded by ionizing radiation, produces free radicals that th at can sub subseq sequen uently tly be qua quanti ntified fied by ele electr ctron on spin resonance (ESR) spectroscopy. Similar to TLD dosimeters, alanine acts as a single-target system, exhib ex hibiti iting ng a lin linear ear re respo sponse nse wit with h abs absorb orbed ed dos dose e until saturation occurs. Regulla and Deffner (1982) repor re portt a tis tissue sue-li -like ke abs absorb orbeded-dos dose e res respon ponse se to a wid ide e ran ange ge of phot oton on en ener ergi gies es.. For 20 200 0 mg samp sa mple les, s, th they ey ob obse serv rved ed a li line near ar ab abso sorb rbed ed-d -dos ose e resp re spon onse se ra rang nge e fr from om 1 to 105 Gy Gy.. As wit with h oth other er solid-state solid-sta te dosimeters, the absorbed-dose response depends depe nds upon LET LET,,   i.e., the mic micros roscop copic ic spa spatia tiall pattern of energy depositions by ionizing radiation. Hans Ha nsen en an and d Ol Olse sen n (1 (198 985) 5) re repo port rt in inte tegr gral al ES ESR R respon res ponse se valu values es rel relat ativ ive e to gam gammama-ra ray y or elec electr tron on bombardment for 16 and 6 MeV protons and 20 MeV 

The response is based on the formation of a latent late nt image imag e of in film microscop micr oscopic ic silve silver r hali halide de crys crystals tals (grai (gr ains ns)) di disp sper erse sed d un unif ifor orml mly y on a ge gela lati tin n ba base se (emulsion (emu lsion). ). The dev developm elopment ent proc process ess red reduces uces the affected grains to silver, whereas the fixing process removes unirradiated grains. The resulting optical densit den sity y is pr propo oporti rtiona onall to th the e flu fluenc ence e of par partic ticles les passing through the emulsion (Dudley, 1966). As a result res ult,, cha change ngess in opt optica icall den densit sity y ac acros rosss the film can only be interpreted as changes in dose if the ener en ergy gy sp spec ectr trum um of th the e pr prot oton onss is co cons nsta tant nt.. There The refor fore, e, film filmss can be con conven venien iently tly and sa safel fely y used to measure the distributions perpendicular to the proton-beam direction, but not along the beam

a-particles of 1.00, 0.86, and 0.58, respectively. For these the se bom bombar bardme dment nt con condit dition ions, s, the av avera erage ge mas masss electronic stopping powers in alanine were 38, 119, and 534 MeV MeV cm2 g21, respe respective ctively ly.. While the relative tiv e sen sensit sitivi ivity ty per uni unitt abso absorbe rbed d dos dose e var varied ied wit with h stoppi sto pping ng po power wer,, the obs observ erved ed line linearit arity y and sat satururatio at ion n de depe pend nden ence ce wa wass in inde depe pend nden entt of st stop oppi ping ng power,   i.e., the fre free-r e-radi adical cal pro produc ductio tion n was dep depenendent upon stopping power as a simple scaling factor. Low-LET radiation results in the production of free radicals, which are stable at room temperature. Howe Ho weve verr, for hig high-L h-LET ET ra radia diatio tion, n, sig signifi nifican cantt fading occurs that depends upon the total absorbed dose delivered. Hansen and Olsen (1989) reported

axis to mea axis measur sure e dep depth– th– dos dose e dis distri tribut bution ions. s. In an intermediate situation in which film is to be used to verify the dose distribution of irregularly shaped  volumes, and where there is a mixture of protons with wit h dif differ ferent ent ene energi rgies, es, an exp expect ected ed film den densit sity y distri dis tribut bution ion sho should uld be cal calcul culat ated ed and com compar pared ed with the meas measured ured dose dis distribu tribution tion (Spi (Spielber elberger ger

2.5, 3.2, and 3.7 percent fading at 1000 h following exposure to 16, 6, and 1 MeV protons, respectively. Even Eve n for lowlow-LET LET radiatio radiation, n, Regu Regulla lla and Deffn Deffner er (198 (1 982) 2) re repo port rted ed 10 pe perc rcen entt fa fadi ding ng at el elev evat ated ed storage temperatures of 708C, as well as increased yields for irradiations at elevated temperatures, 20 percent at 908C and 105 Gy. Bradshaw  et al.  (1962)

 4.8.2.3

Radiographic films

77  

Alanine

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

and Ebert   et al.   (1965) (1965) comp compared ared the resp response onse of  alan al anin ine e po powd wder er in lo loww-en ener ergy gy ( , 8 MeV MeV)) pr proto oton n   60 beams with the response in Co. The Th e use of the alanine alanine dos dosime imeter ter for absorbed absorbed-dose do se de dete term rmin inat atio ions ns in cl clin inic ical al pr prot oton on be beam amss requires consideration of the proton energy-fluence spectr spe ctrum um and con contr trol ol of th the e env envir ironm onment ent dur during ing

parallelparall el-pla plate te ion ioniza izatio tion n cha chambe mbers rs in a 62 MeV  clin cl inic ical al pr prot oton on be beam am.. In bo both th in inve vessti tiga gati tion ons, s, anomalous doses were measured with the alanine detectors distal to the Bragg peak and could not be satisfactorily explained. In a dosimetry comparison in lowlow-ener energy gy prot proton on ther therapy apy beam beams, s, the rela relative tive difference between alanine and ionization-chamber

and subsequent to bombardment. Nichiporov   et al. (1995) investigated the dose response of alanine in a 169 MeV proton beam using ESR spect spectrosc roscopy opy.. They found a linear response ( +2 percent) in the dose range from 50 to 300 Gy and a relative interdetect det ector or res respon ponse se of    +1.5 per percen cent. t. Dep Depth– th– dos dose e meas me asur urem emen ents ts at a re resi sidu dual al pr prot oton on en ener ergy gy of  50 MeV were indirectly compared with ionization-chamber measurements and found to be al..   (1996) in goo good d agr agreem eement ent.. Fa Fatti ttiben bene e   et al (1996) and et al.   (199 Onori   et (1997) 7) fo foun und d go good od ag agre reem emen entt betwe bet ween en dep depth th dos doses es mea measur sured ed wit with h ala alanin nine e and

dose dos e det determ ermina inatio tions ns wa wass   +2 per percen centt (C (Cutt uttone one  et al., 1999 1999). ). Rec Recently ently Palmans Palmans (2003 (2003)) unde undertook rtook Monte Carlo simulations of the depth–dose curves in alanine pellets in a 60 MeV ocular proton beam. The Th e re resu sult ltss sh show owed ed an un unde dere rest stim imat atio ion n of th the e Bragg peak intensities for both unmodulated (up to 30 pe perc rcen ent) t) an and d mo modu dula late ted d (u (up p to 15 pe perc rcen ent) t) beams.

 4.8.2.5

Other detectors

Other types of dosimeters such as diamond detectors and TLDs have been used for measuring relative doses in a phantom. The potential advantages of di diam amon ond d de dete tect ctor orss fo forr re rela lati tive ve do dosi sime metr try y in rad adio iotther era apy have bee een n wel elll docu do cum men entted al.., 20 al.., 20 (Bucciolini   et al 2003 03;; De An Ange geli liss   et al 2002 02;; et al al.., 19 ¨ rvinen, Khrunov   et 1990 90;; Vat atni nits tsky ky an and d Ja 1993; Vatnitsky   et al., 1993) 1993).. The They y are radiationradiationresistant, practically TE, have high sensitivity and stability stab ility,, and are rela relative tively ly small small.. Depe Dependin nding g on the typ type e of det detect ector or use used, d, rel relat ative ive dos dose e mea measur sureements men ts ha have ve sho shown wn goo good d agr agreem eement ent wit with h tho those se of  ionizatio ioniz ation-ch n-chamber amber meas measurem urements ents in high high-ener -energy gy (  200 MeV) proton beams (Schreuder   et al., 1997;  Vatnitsky  V atnitsky   et al., 1993 1993;; 1995; 1999a). However However,, the

Figure 4.6. Com Figure Compari parison son of dept depth h doses measured measured in a 200 MeV  proton beam with a diode, diamond detector, and three different ionizat ioni zation ion cham chamber berss (Ma (Markus rkus par paralle allell plat plate, e, Exr Exradin adin 0.5 cm3 3 cylindrical, and Mini, which is a 0.007 cm specially construc constructed ted cylindri cyli ndrical cal ioni ionizat zation ion cham chamber) ber).. (a) For an unmo unmodula dulated ted beam for which which onl only y the BraggBragg-pea peak k re regio gion n is sho shown wn and (b) for a beam with a 5 cm SOBP. In both cases, the data are normalized to the same entrance dose level where the measured dose levels are independent of detector resolution (Schreuder (Schreuder   et al., 1997; reproduced reprodu ced with permission permission). ).

Figure 4.7. Compari Comparison son of beam profiles measured in a 200 MeV  proton beam with a diode, diamond detector, and three different ionization chambers (see Fig. 4.6) in a 3.5 cm diameter field at a dept de pth h of 5 cm in wate terr ( Ep  167 Me MeV). V). Th The e Ma Marku rkuss an and d Exra Ex radi din n T2 ch cham ambe berr me meas asur urem emen ents ts ar are e on only ly sh show own n fo forr comparison, and should not be used to measure beam profiles. The data are normalized to 100 on the central axis (Schreuder  et al., 1997; reproduced with permission).

78  

 DOSIMETRY   DOSIMETR Y 

result ultss obt obtain ained ed in low low-en -energ ergy y bea beams ms ha have ve bee been n res problematic (Onori  et al., 2000; Pacilio  et al., 2002). Several experimental investigations of the usefulness ne ss of TL TLDs Ds,, ma main inly ly Li LiF F (T (TLD LD-1 -100 00), ), fo forr ap appl pliicati ca tion onss in cl clin inic ical al pr prot oton on do dosi sime metr try y ha have ve be been en al.., 199 al.., undertaken (Bilski   et al 1997; 7; Buc Buccio ciolin linii   et al

 et al., 1997 1997). ). Two disa disadvan dvantages tages of rad radiochr iochromic omic film have been its high cost and its low sensitivity. Howe Ho weve verr, gi give ven n it itss ea ease se of pr proc ocur urem emen ent, t, lo long ng storage life, and high spatial resolution, its use is increasin incr easing, g, while efforts to impr improve ove its sensi sensitivit tivity y are ar e bei being ng mad made e (Ge (Geso so   et al., 2004; Soares, Soares, 2006 2006), ),

1999; 2000; Carlsson and Carlsson, 1970; Fattibene  et al., 199 1996; 6; Lon Loncol col   et al., 1996; Vatn atnitsky itsky   et al., 1995; 1999a). 1999a). Thei Theirr usef usefulnes ulnesss is limit limited ed by thei theirr relat rel ative ively ly lar large ge wa water ter-eq -equiv uivale alent nt thi thickn ckness ess and LET-dependent LET -dependent response. However However,, measuremen measurements ts have ha ve sh show own n go good od ag agre reem emen entt wi with th io ioni niza zati tion on-chambe cha mberr mea measur sureme ement nts, s, par partic ticula ularly rly in ra range nge-modula mod ulated ted bea beams ms in whi which ch th the e LE LET T var varia iatio tion n is al.., 1996; Vatn al.., diluted (Fatt (Fattibene ibene   et al atnitsk itsky y   et al 1995). 199 5). The mai main n app applica licatio tion n for TL TLDs Ds in pr proto oton n therapy is probably for   in vivo   dosimetry. Diamond detectors and TLDs should only be used for relative proton prot on dosim dosimetry etry once thei theirr char characte acterist ristics ics hav have e been check checked ed again against st the char charact acteris eristics tics of appr approopriate ionization chambers in the clinical beam in which they will be used. The advent of scanning proton beams, as well as intensity-modulated radiation therapy, has brought more demands for dosimeters that can make twoand thr threeee-dim dimens ension ional al det determ ermina inatio tions ns.. Un Unlik like e passive beam-delivery systems, in which the entire target volume is continuously filled to uniform dose through thr oughout out the irra irradiat diation, ion, scan scanned ned beam beamss mus mustt sequ se quen enti tial ally ly de deli live verr la laye yers rs of do dose se at di disc scre rete te prot pr oton on en ener ergi gies es to fil filll th the e ta targ rget et vol olum ume e to a unif un ifor orm, m, or pr pres escr crib ibed ed (i (in n th the e ca case se of in inte tens nsit ity y modula mod ulatio tion) n) dos dose. e. Fo Forr th these ese tr trea eatme tment nt pr presc escrip rip--

while che while cheape aperr film is bec becomi oming ng av avail ailabl able. e. A ne new w rad adio ioch chrrom omic ic film (Ga GafC fCh hrom omic ic EBT) wit ith h et al., 20 improv impr oved ed char charact acterist eristics ics (Che (Cheung ung   et 200 05; Sankar  et al., 2006) is being used in clinical dosimetry (Zeidan  et al., 2006). Two-dimensional dosimetry with high spatial resolution can be performed with a scintillating screen  viewed through a 458   mirror mirror by a charg charge-cou e-coupled pled al.., device dev ice (C (CCD) CD) cam camer era a (B (Boon oon,, 199 1998; 8; Boo Boon n   et al et al., 19 1998; 199 8; 200 2000; 0; Co Coutr utrak akon on   et 1990 90;; Ry Ryne neve veld ld,, 1998). The same CCD system has also proved to be extr ex trem emel ely y us usef eful ul to ve veri rify fy th the e do dose se de deli live very ry in al.., 20 scanned scan ned beam beamss (Lom (Lomax ax   et al 2004 04). ). It ca can n al also so provide the necessary data to describe the dose distribution of the elemental scanned beam (the physical pencil beam). These data are required as input for the dose calculation calculation of exte extended nded scanned scanned dose fields fiel ds in tr trea eatme tment nt pla plann nning ing.. Ho Howe weve verr, the CC CCD D hass th ha the e pr prob oble lem m of ha havi ving ng a sm smal alll am amou ount nt of  quenching of the signal in the Bragg-peak region. This can be accounted for in the calculation of the dose dos e re respo sponse nse,, but it can be als also o be cor corre recte cted d by choosing a proper mixture of the phosphors (Safai  et al., 2004).  Another class of detectors is liquid and gel detectors to rs.. On One e of th the e fir first st ty type pes, s, Fr Fric icke ke do dosi sime mete ters rs,, reli re lies es on th the e co conv nver ersi sion on of Fe2þ to Fe3þ in an

tions, the entire all energies be de deli live vere red d fo forr treatment each ea ch do dose sewith meas me asur urem emen entt inmust the th e target tar get vo volum lume. e. Fo Forr dis discr crete ete dos dosime imeter ters, s, or eve even n line li near ar io ionn-ch cham ambe berr ar arra rays ys,, th this is pr proc oces esss ca can n be quite time-consuming. One of the twotwo-dimen dimensiona sionall dosim dosimeters eters showing promi pr omise se is ra radio diochr chromi omicc film ( e.g., Gaf GafChr Chromic omicTM MD-55 MD-5 5 film) (Bu (Butson tson   et al., 2001 2001;; Niro Nirooman omand-Ra d-Rad d  et al., 1998) that gives more accurate depth–dose profiles than radiographic film because of reduced satur sa turat ation ion eff effect ectss in the hig high-L h-LET ET region region of the Brag Br agg g pe peak ak.. Th The e fil film m ca can n be sa sand ndwi wich ched ed,, fo forr exam ex ampl ple, e, be betw twee een n po poly lyst styr yren ene e bl bloc ocks ks wi with th th the e beam entering through the edge of the film. This

aqueous ferrous sulfate solution byde irradiation. rela re lativ tive e dos dose e can be obt obtain ained ed by deter termin mining ing The the th e numbe nu mberr of fer ferric ric ion ionss pr produ oduced ced,, usi using ng an opt optica icall absorption spectrophotometer with ultraviolet light at 30 304 4 nm (A (Att ttix ix,, 19 1986 86). ). Wit ith h Fr Fric icke ke ge gels ls,, th the e chan ch ange ge in param amag agne neti ticc prop oper erttie iess ca can n be measu mea sured red usi using ng MR MRII or opt optica icall tec techn hniqu iques. es. Th The e former can provide three-dimensional images of the dose do se di dist stri ribu buti tion ons. s. Th The e MR MRII of ge gels ls co cont ntai aini ning ng FeSO4   has bee been n inv inves estig tigat ated ed by sev sever eral al aut author horss (Gore   et al., 198 1984; 4; Haz Hazle le   et al., 1991; Maryanski Maryanski  et al., 19 1994 94;; Pod odgo gors rsak ak an and d Sc Schr hrei eine nerr, 19 1992 92). ).  A major limitation of Fricke gel system systemss is the continual tinu al post post-irra -irradiat diation ion diffu diffusion sion of ions, resu resulting lting

technique can give a two-d technique two-dimen imensiona sionall dose distributi bu tion on th thro roug ugh h th the e ta targ rget et vo volu lume me th that at sh show owss distal-ed dist al-edge ge boun boundaries daries quit quite e well. Afte Afterr optic optically ally scanning the film along the depth axis and comparing with ionization-chamber data, saturation in the

in a blurred dose distribution. Polym Po lymer er gel dos dosime imeter terss ar are e als also o und under er ex exper per-imentat imen tation ion as thre three-di e-dimensi mensional onal dosim dosimeters eters,,   e.g., TM   et al. BANG gel (Ram (Ramm m , 2000 2000;; Uusi Uusi-Sim -Simola ola   et. al., 2003). One of the promising ones, such as the

Bragg-pe Bragg -peak ak re regio gion n app appear earss to be   , 10 per percen centt of  the peak dose (Daftari   et al., 1999; Fidanzio   et al., 2002; Luch Luchin in   et al., 2000 2000;; Nich Nichipor iporov ov   et al., 1995; Piermattei   et al., 2000; Vatnitsky, 1997; Vatnitsky

so-called MAGIC (methacrylic and ascorbic acid in gelatin initiated by copper) dosemeter (Fong   et al., 2001)) can be manu 2001 manufact facture ured d and used mor more e easil easily y than th an ot othe herr ge gels ls an and d ca can n be im imag aged ed by op opti tica call 79

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

scanning or by MRI. In addition, the gel does not need to be stored in an oxygen-free environment as many ma ny other other ge gels ls do do.. Ge Gels ls ca can n al also so be us used ed wi with th scanning beams, but the experience is very limited. Gels have the advantage of giving a representation of the dose in the entire three-dimensional volume, but the precision of the data is still less than that obtained with a CCD-based system. However, there is only limited practical and theoretical knowledge of gel dosimetry in proton and ion beams (Gustavsson   et al., 2004; Hilts   et al., 2000; Jirasek and Duzenli, 2002; Ramm   et al., 2000). Presently, there ther e are mean meanss ava available ilable for veri verifying fying scanningscanningbeam be am do dose se de deli live very ry wi with th su suffi ffici cien entt pr prec ecis isio ion n to ensure that the system is safe for patient trea tr eattme ment nts. s. Ho How wev ever er,, si sin nce the en ener erg gy (or LET)-de LET )-depend pendent ent resp responses onses of gel dosi dosimete meters rs hav have e et al., not bee been n we well ll doc docume umente nted d (Gu (Gust stavs avsson son   et 2004), their clinical use has thus far been limited.  A new type of radiochrom radiochromic ic gel material (PRESAGETM) has rec recent ently ly bee been n des descri cribed bed (Gu (Guo o  et al., 200 2006). 6). Th This is ma mater terial ial has pot potent ential ial adv advanantages over other gel dosemeters, including insensitivity tiv ity to oxy oxygen gen and ame amenab nabili ility ty to ma machi chinin ning g to arbitrary shapes and sizes, without the need for a container.  An example of a large multiarray detector for three-dimensional dosimetry is a 400 scintillator/  fiber system developed at PSI (Safai   et al. , 2004). The Th e sc scin inti till llat ator orss (s (sen ensi siti tive ve vo volu lume me of 5 mm3) are ar e op opti tica call lly y co coup uple led d to th the e fib fiber er li ligh ghtt gu guid ides es (2 mm dia diamet meter er by 1 m lon long). g). Th The e de detec tector torss ar are e embedde embe dded d at diff differen erentt dept depths hs in a poly polyethy ethylene lene phantom pha ntom.. The ligh lightt sign signals als fro from m the scint scintilla illators tors are ar e ca carr rrie ied d by the bu bun ndle of fibe bers rs and are mapp ma pped ed on a CC CCD D by me mean anss of a le lens ns sy syst stem em.. Oth Ot her no nove vell de deve velo lop pme ment ntss fo forr qu quas asii-th thrree ee-dimensio dime nsional nal dose ver verifica ification tion inv involv olve e the use of  stac st acks ks of two two-d -dime imens nsion ional al arr array ayss of ion ioniz izat ation ion chambers cham bers inte interlea rleaved ved with plas plastic tic slabs (Am (Amerio erio  et al. , 200 al.., 19 2004; 4; Br Brusa usasco sco   et al 1997 97;; 20 2000 00;; Ci Ciri rio o  et al. , 2004).

4.8.3 Determi Determinatio nation n of dose distr distributio ibutions ns

The description of the changing energy and spatial char ch ara act cter eris isti tics cs of a pr prot oton on be beam am as it pa pass sses es

 4.8.3.1

Range and depth – dose characteristics

 A critical aspect of proton treatm treatment ent is the possibili bi lity ty of sto topp ppin ing g th the e be beam am at a sp spec ecifi ificc po poin intt within the patient. Accurate control of the stopping point depends on knowledge of the beam range in water wa ter and of the wa water ter-eq -equiv uivale alent nt pa path th len length gthss (WEPLs) of materials placed in the beam path and of the tis tissue suess tr trav avers ersed. ed. Th The e cen centr tralal-axi axiss bea beam m range in water is measured for the beam energies and range absorbers that are to be used clinically. Depth–dose distributions should be determined for a selection of energy, range modulation, field size, and other treatment parameters that might affect the distributions. The sharpness of the distal-dose falll off dep fal depend endss on th the e inc incide identnt-bea beam m ene energy rgy and energy spread of the accelerated beam, and on the rang ra nge e st stra ragg ggli ling ng pr prod oduc uced ed by ab abso sorb rber erss in th the e beam-delivery system and by the patient. For measureme rements nts of chambers depth dept h – dose distributions, distribution s, the usemeasu of plane-parallel is recommended. In prin principle, ciple, a measu measured red dept depth-io h-ioniza nization tion dist distriribution should be converted to a depth–dose distributi bu tion on,, usi sin ng the de dept pth h dep epen ende denc nce e of th the e stopping-pow stoppin g-power er ratio,   sw,air. Th This is is ac achi hiev eved ed by multip mul tiplyi lying ng th the e mea measu sure red d ion ioniza izati tion on cha charge rge or curren cur rentt at eac each h de depth pth   z   by the sto stoppin pping-po g-power wer ratio,   sw,air, an and d the per pertur turba batio tion n fa facto ctorr at th that at depth. Values for   sw,air  as a function of   Rres  can be calculated from Eq. (4.17). Perturbation factors are assu as sume med d to ha have ve a va valu lue e of un unit ity y (s (see ee Se Sect ctio ion n 4.4.2.1) 4.4. 2.1).. For pra practi ctical cal pur purpose poses, s, thes these e conv conversio ersion n factors are not usually applied, as   sw,air  varies by less than   + 0.4 percent from the median value for resi re sidu dual al pr prot oton on ra rang nges es in wa wate terr fr from om 0. 0.25 25 to 30.00 cm (IAEA, 2000). The influence of ion recombina bi nati tion on an and d po pola lari rity ty ef effe fect ctss on th the e de dep pth th-ionization distribution should be investigated and taken tak en int into o ac accou count nt if th ther ere e is a var varia iatio tion n wit with h depth. If the field size for which measurements are to be performed is smaller than twice the diameter of the cavity of the plane-parallel chamber, then a detector with a better spatial resolution ( e.g ., minichamber, diode, or diamond) is recommended (see below). The resulting distribution distribution must also be con verted, if necessary, using the appropriate appropriate stoppingstoppingpower pow er ra ratios tios ( e.g., wat water-to er-to-air -air,, wate water-to-s r-to-silicon, ilicon, or water wa ter-to -to-gr -graph aphite ite). ). Fo Forr the lat latter ter,, the nece necessa ssary ry

stoppi stop ping ng-p -pow ower er va valu lues es ca can n be fo foun und d in IC ICRU RU (1993a). The suitability of such detectors for depth– dose measurements should be verified by test comparisons with a plane-parallel chamber at a larger field size. For scanning beams, measurement times should be long enough compared with the scanning cycl cy cle e of th the e fie field ld in or orde derr to yi yiel eld d re repr prod oduc ucib ible le readings.

through throug h ma matte tterr is cri critic tical al for pr predi edicti cting ng the dos dose e deposition depo sition of the prot protons. ons. The rela relativ tive e dosim dosimeters eters described above can be very helpful in determining these characteristics. The determination and parameterization of the dose distributions in matter is impo im port rtan antt fo forr th the e acc ccur urac acy y of do dose se-c -cal alcu cula lati tion on algo al gori rith thms ms us used ed in co comp mput uter eriz ized ed tr trea eatm tmen enttplanning syst systems. ems. 80  

 DOSIMETRY   DOSIMETR Y 

 Accurate ate depth– depth–dose dose measu measureme rements nts must be  Accur made throughout the Bragg peak region, including the sharp distal fall-off region near the end of the range. ran ge. Th This is req requir uires es a det detecto ectorr wit with h goo good d spa spatia tiall

calibration. The depth–dose curve near the surface of treatment fields can be affected by protons scattered from field-shaping apertures. Aperture scatter effects (van Luijk  et al., 2001) can also be more pro-

resolution and with a dose response that is practically independent of the variation of proton energy with depth. As mentioned above parallel-plate ionizal.., 20 ation ati on cha chambe mbers rs (P (Palm almans ans   et al 2002 02a) a) ar are e re reccomm om men ende ded d fo forr de dept pth– h– do dose se me meas asur urem emen ents ts,, bu butt smallsma ll-vol volume ume thi thimbl mble e ion ioniza izatio tion n cha chambe mbers rs (mi (mini ni al.., 19 chambers) chambe rs) (Schr (Schreuder euder   et al 1997 97), ), an and d dia diamo mond nd et al., 20 detect det ectors ors (P (Paci acilio lio   et 2002 02;; Vatn tnit itsk sky y an and d Ja¨ rvinen, 1993; Vatnitsky  et al., 1999a) can be used. Caut Ca utio ion n is ca call lled ed fo forr wh when en us usin ing g si sili lico con n di diod odes es since, as was noted above, they have been observed to ha hav ve up to 10 pe perrce cent nt hi high gher er re resp spon onse se th than an paralle par allel-p l-plat late e ion ioniza izatio tion n cha chambe mbers rs in the Br Bragg agg-peak region (Koehl (Koehler er,, 1967; Raju, 1966; Schreuder Schreuder

nounced for small and irregularly-shaped fields.

 et al., 199 1997). 7). Ho Howev wever er,, the use of sui suitab table le dio diodes des (and the avoidance of n-doped types) can give good agreement agreem ent with ioniza ionization-ch tion-chamber amber measur measurement ementss (Grusell and Medin, 2000). When using a diamond detect det ector or,, the res respon ponse se shou should ld be corr correct ected ed for the dose-rate effects (Pacilio   et al., 2002). Radiographic Radiographic film sh shou ould ld no nott be us used ed fo forr de dept pth– h– do dose se me meas asur ureements because of its significant variation in response per unit dose as a funct function ion of depth in a proton beam. beam.  Accurate  Accur ate depth– dose measurements measurements with radiochroradiochromic films can be made if the films are placed perpendicular to the beam axis at different depths. Wat ater er-e -equ quiv ival alen entt pa path th len lengt gths hs fo forr ma mate teri rials als placed in the beam path are measured, and, for purpose po sess of tr trea eatm tmen entt pl plan anni ning ng,, th the e re rela lati tion onsh ship ip between WEPL (or proton stopping power) and CT number (based on rela relative tive x-ray attenuation attenuation coefficients) for body tissues is established (Schaffner and Pedroni, 1998; Schneider  et al., 1996; Szymanowski and an d Oe Oelf lfke ke,, 20 2003 03). ). Th The e di dist stri ribu buti tion on of do dose se as a function of depth, determined as a basic characteristic of the beam, is also required data for treatment plan ann ning. The WEPLs fo forr materi rial alss can be measur mea sured ed by sub submer merging ging sam sample pless of kno known wn thi thickckness in a water phantom and measuring the effect on beam range. The WEPLs for materials used for

 4.8.3.2

Beam profiles and penumbrae

For pr For proto oton-t n-tre reatm atment ent bea beams ms tha thatt are pr produ oduced ced by passive passi ve beam-s beam-shapin haping g (scatt (scattering) ering) techn techniques, iques, the lateral uniformity of proton-treatment fields should be comp compar arable able or sup superi erior or to tha thatt of con conven vention tional al phot ph oton on an and d el elec ectr tron on fie field lds. s. Un Unif ifor ormi mity ty ma may y be expressed in terms of field symmetry for points equidistan dis tantt fr from om the beam cent central ral axis and a fla flatne tness ss  variation  varia tion over some design designated ated portio portion n ( e.g., 80 percent) of the field area. Uniformity characteristics should be measured at several depths for a variety of  treatment-planning dose distributions. The shape of the dose dis distri tribut bution ion at the lateral lateral field fie ld ed edge ge is ex extr trem emel ely y im impo port rtan antt in pl plan anni ning ng proton-bea prot on-beam m trea treatment tments. s. Field place placement ment in proxi proxi-mity to radiation-sensitive normal tissue depends on accurate knowledge of the penumbra as well as consideratio sider ation n of the uncertainly uncertainly in pati patient ent alignment. Penumbra widths are defined (see Section 3.4.2.2) as the distance separating stated dose levels ( e.g., 80– 20 per percent cent of the central central axis dose at that depth). depth). Penumbra characteristics, which can be determined by beam profile scans, will depend on the design of  the bea beam-d m-deliv elivery ery sy syste stem m and will var vary y wit with h mos mostt treatment parameters, including beam energy, range modula mod ulatio tion, n, dep depth th in the pa patie tient, nt, and coll collima imator tor-to-skin distance (Urie  et al., 1986b). These variations should sho uld be acc accura urately tely mea measur sured ed and rep repro roduc duced ed in the trea treatment tment-plan -planning ning sys system. tem. Oozer   et al.   (1997) have developed a model for the lateral penumbrae in water for a 200 MeV clinical proton beam based on beam-profile measurements with a silicon diode. The results show that the lateral penumbrae are essentially independent of beam modulation and collimator diameter. The model has been integrated into a treatment-planning treatment-plann ing system. Lateral-uniformity measurements should be made

patient-specific devices, such as range shifters and

in water or in a water-equivalent phantom with a

ti tissu ssue e com compe pens nsat ator ors, s,for need nee d to be de dete term rmin ined ed.. In addition, the WEPLs tissue-substitute materials should be determined and correlated with observed CT numbers for the same materials. The shape of the cen centra tral-a l-axis xis dep depth– th– dos dose e cur curve ve  varies with field size (Larsson, (Larsson, 1967; Vatni Vatnitsky tsky et al., 1999a) 199 9a) and can dif differ fer sign signific ificant antly ly for sma small ll fiel fields ds and at loc locati ations ons wit within hin irr irregu egular larly ly sha shaped ped fiel fields ds where the lateral extent is less than a few penumbra widths. An understanding of this variation is importantt fo an forr tr trea eatm tmen entt pl plan anni ning ng an and d tr trea eatm tmen entt-fie field ld

detector spatial resolution in scanning the scan direction.having Energyhigh independence for lateral is not as imp import ortant ant as for depth scanning. scanning. Silicon Silicon diod di odes es,, di diam amon ond d de dete tect ctor ors, s, an and d sm smal alll ion ioniz izat atio ion n chambers are useful. Radiographic film placed in a phanto pha ntom m per perpen pendicu dicular lar to the beam dir directi ection on pr proo vides results similar to the other detectors. The film shou sh ould ld be sc scan anne ned d wit with h a sy syst stem em th that at ha hass goo good d spatial spat ial resol resolution ution and a film-d film-density ensity response correctio rec tion n sho should uld be mad made. e. Rad Radioc iochro hromic mic film filmss als also o give very high spatial resolution. 81

 

Journal of the ICRU Vol 7 No 2 (2007) Report 78 Oxford University Press

doi:10.1093/jicru/ndm028

5 GEO GEOMET METRIC RIC TE TERMS RMS,, AND AND DOSE DOSE AND DOS DOSE E – VO VOLUM LUME E DEFINITIONS In pr prev evio ious us IC ICRU RU re repo port rtss (I (ICR CRU, U, 19 1993 93b; b; 19 1999 99;; 2004), a number of important concepts concerning anatomic volumes of interest were introduced and defined. defin ed. Among these were: for tumo tumors, rs, the gros grosss target tar get vo volum lume e (GT (GTV) V),, th the e cli clinic nical al tar target get vo volum lume e

the body outline) which are not properly part of the  volume of interes interest. t. All geometric objects should rela re late te to th the e pa patie tient nt’s ’s an anat atomy omy und under er tr trea eatm tment ent condit con dition ions, s, and at one poi point nt in tim time e (se (see e Se Secti ction on 5.5). In the present report, the term   delineation   is

(CTV), and the planning target volume (PTV); and, for nor normal mal tissues, tissues, the con concep cepts ts of org organ an at ris risk k (OAR) and planning organ at risk volume (PRV). Excep Ex ceptt for the PT PTV V and PR PRV V, th these ese de defini finitio tions ns apply to all modalities equally. Both in their delineation and in their use, it is irrelevant as to whether photons, phot ons, elec electro trons, ns, pro proton tons, s, or any othe otherr rad radiat iation ion are to be employed. Indeed, it is important not to let knowledge of the likely radiation modality affect how these the se vo volum lumes es ar are e del deline ineat ated ed sin since ce (i) th their eir de defifinitions are modality independent; and (ii) one might wish wi sh to co comb mbin ine e or co comp mpar are e or re retr tros ospe pect ctiv ivel ely y analyze treatment plans for more than one modality. Becau Bec ause se the defi definit nition ionss of geo geomet metric ric (an (and d dos dosii-

used to des used descri cribe be the ide ident ntific ificati ation on of the spa spatia tiall exte ex tent nt of an ob obje ject ct,, re rega gard rdle less ss of wh whet ethe herr it is identified by a manual process of drawing contours on an image, or by an automatic feature extraction technique, or by any other method.

metric) terms are very largely the same for protons as for all other modalities of radiation therapy, the definit defi nition ionss and ex expla plana natio tions ns of pr previ evious ous re repor ports ts remain rema in large largely ly valid for prot proton-be on-beam am ther therapy apy.. For thiss re thi reaso ason, n, and to ha harmo rmoniz nize e the pr prese esent nt re repor portt with wi th th the e pr prec eced edin ing g re repo port rts, s, po port rtio ions ns of IC ICRU RU Reports Rep orts 71 (ICR (ICRU, U, 2004) have been repr reproduce oduced d in the present section. The reader is referred to this report rep ort for cla clarif rifyin ying g dia diagra grams ms and exp explan lanat ation ionss

5.1 ANA ANATOMIC TOMIC VOLUM VOLUMES ES RELA RELATING TING TO THE TUMOR The vo The volum lumes es rel relat ating ing to the tu tumor mor ar are e sch schema ematitically represented in Fig. 5.1. 5.1.1 5.1 .1

Gross Gro ss tumo tumor r volum volume e (GTV) (GTV)

The GTV is the gro gross ss pal palpab pable, le, vis visibl ible, e, or cli clinic nicall ally y demonstrable location and extent of the malignant (or othe ot herw rwis ise) e) gr grow owth th.. It ca can n co cons nsis istt of pr prim imar ary y tu tumo morr an and, d, if pr prese esent nt,, met metas asta tatic tic lym lympha phaden denopa opathy thy, or oth other er metast met astase ases. s. In the GTV, the tum tumor or cell den densit sity y is alway alw ayss hig high h ( 106 mm23). Hen Hence, ce, an ade adequa quate te dos dose e must mu st be de deliv liver ered ed to th the e wh whol ole e GT GTV V fo forr ra radi dica call th ther erap apy y. There Th ere is no GTVaf GTVafter ter com comple plete te sur surgic gical al res resect ection ion.. The Th e sh shap ape, e, si size ze,, an and d lo loca cati tion on of a GT GTV V ca can n be determ det ermine ined d by cli clinic nical al ex exami amina natio tion n ( e.g., ins inspec pec--

that are omitted in the present section. Each of the above-mentioned volumes is discussed below.. The GTV and CTV are purely oncologica below oncologicall concept ce ptss an and d ar are e in inde depe pend nden entt of an any y th ther erap apeu euti ticc approa app roach. ch. The They y rep repres resent ent vo volum lumes es with kno known wn or suspected suspe cted tumor involvement. involvement. The OAR represents represents normal nor mal tiss tissues. ues. The PTV and and PR PRV V are pur purely ely geometric met ric con conce cept pts, s, wh whic ich h do no nott ne neces cessa saril rily y cor corre re-spond to tissue or organ borders. The definitions of  these concepts and explanations are given in ICRU Report Re port 62 (IC (ICRU RU,, 1999 1999)) and can be ext extend ended ed and appl ap plied ied to all fo form rmss of ex exte tern rnal al-b -beam eam th ther erap apy y, including conformal therapy (ICRU, 2004).  Volumes  V olumes should be carefully delineated; for example, exam ple, contours should not undu unduly ly ext extend end into spaces spa ces (su (such ch as air cavities cavities or the region region out outsid side e

tion, pal tion, palpa patio tion, n, and end endosc oscopy opy), ), and and/or /or var variou iouss  e.g imaging techniques ( ., x ray, CT, digital radiography,, ultr phy ultrasono asonograp graphy hy,, MRI MRI,, magn magnetic etic reso resonanc nance e spectros spec troscopy copy [MR [MRS], S], PET PET,, and othe otherr rad radionuc ionuclide lide imaging methods). The methods used to determine the GTV should meet the requirements for staging the th e tum tumor or ac accor cordin ding g to th the e cli clinic nical al TN TNM M (U (UICC ICC,, 1997) and AJCC (AJCC, 2002) systems. The GTV (primary tumor, metastatic lymphadenopathy, and other known metastases) can appear different in size and shape, sometimes significantly, depending on which examination technique is used for evaluation. The radiation oncologists should in each case indicate which technique has been used for the eva evalua luatio tion n and for th the e del deline ineat ation ion of th the e GTV (see Section 5.4.4).

# International Commission on Radiation Units and Measurements 2007

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

of a cancer is performed, one often finds sub-clinical extensions around the GTV,  i.e., individual malignant cells cells, , sma small cell clusters clus ters, , or micro-e o-exten which h cannot can not bellde detec tected ted clinic cli nicall ally ymicr . Th The e xtension CTV CT Vsions, is s, a whic tissu tis sue e  volume  volu me tha thatt cont contains ains the GTV GTV(s) (s) and and/or /or sub sub-cli -clinica nicall malignant disease at a certain probability level. This  volume  vol ume thu thuss has to be tre treat ated ed ade adequa quately tely. Lik Like e the GTV, the CTV is a clinical-anatomical concept. Delineation of the CTV is based (i) on the available data on the probability of (sub-clinical) malignan antt ce cell llss out utsi sid de th the e GT GTV V; and (i (ii) i) on th the e  judgement of the radiation oncologist. The relevant data dat a to consi consider der concern the prob probabili ability ty of micr microoscopicc exte scopi extensio nsion n at diffe different rent distances distances arou around nd the GTV, and the probability of sub-clinical invasion of  region reg ional al lym lymph ph nod nodes es or oth other er tis tissue sues. s. Th The e CT CTV  V 

Figure 5.1. Illu Figure Illustr stratio ation n of the volumes volumes and margins relating relating to the definition of the target volume.

There are at lea There least st fou fourr re reaso asons ns to des descri cribe be and repor re portt th the e GTV in a com comple plete te and ac accur curat ate e wa way y. First, as indicated above, it is required for staging,  e.g., ac accor cordin ding g to the TN TNM M sy syst stem. em. Se Secon condly dly,, an adequate dose must be delivered to the whole GTV  in or orde derr to ob obta tain in lo loca call tu tumo morr co cont ntro rol. l. Th Thir irdl dly y, evaluati evalu ation on of the GTV regression regression is need needed ed when rede re defin finin ing g th the e PT PTV V (S (Sec ecti tion on 5. 5.1. 1.4) 4) du duri ring ng th the e course of treatment. Fourthly, changes in the GTV  duri du ring ng tr trea eatm tmen entt ca can n be on one e of th the e pr pred edic icti tive ve  values for response to treatme treatment. nt.  A GTV can be confined to only part of an organ, can inv involv olve e a who whole le org organ, an, or can ex exten tend d out outsid side e

delineation should be based on knowledge of pathways of tumor infiltration in three dimensions. Ther Th ere e mi migh ghtt be no ma macr cros osco copi picc di dise seas ase, e, an and d hence no GTV, after gross resection or, occasionally, after chemotherapy as the gross tumor might then no lo long nger er be ev evid iden entt or ev even en pr pres esen ent. t. In su such ch a case, only a CTV, and not a GTV, would be defined. The delineation of the CTV is a required part of  the treatment prescription. If the CTV is identical with the GTV, then its specific delineation could be useful, but it would be sufficient to state that the CTV and GTV GTV are identi identical. cal. The CTV includes the gross lesion (GTV) and the suspected sub-clinical extension of the tumor. After the planned dose to the CTV has been delivered, an additi add itiona onall or ‘bo ‘boos ost’ t’ dos dose e can be adm admini iniste stered red to the GTV. GTV. In some circumstanc circumstances, es, the CTV can be coincident with the GTV, for example, in the case of  a benig benign n tumo tumorr or a wel well-enc l-encapsu apsulate lated d malig malignant nant

the normal borders of the organ or tissue involved. There Th ere can be mor more e th than an one GTV. Thi Thiss wil willl be the case, for example, when there is more than one  volume containing gross palpable, visible, or clinically demo demonst nstrabl rable e malig malignant nant gro growth, wth, and thes these e  volumes are spatially separated separated.. A possible nomenclature for such cases is discussed in Section 5.4.1.  A GTV can change with time because of tumor shrinkag shri nkage, e, tum tumor or gro growth, wth, or the therap rapeuti euticc int interv ervenention.. In tha tion thatt case case,, sev severa erall GTV GTVss can be deli delinea neated ted,, and an d sh shou ould ld be qu qual alifi ified ed,, fo forr ex exam ampl ple, e, as in GT GTV  V  (initial), (ini tial), GTV [ DRBE . 30 Gy (R (RBE BE)] )],, GT GTV V [ DRBE . 50 Gy (RBE)], etc. (see Section 5.4.4). The delineation of the GTV is a required part of  the treatment prescription, except in the event that no gross disease is present.

tumor. Nevertheless, for record-keeping reasons, it is preferable to define a CTV in such a case. 5.1.2. 5.1 .2.1 1

The dose dose intent intention ion for for the CTV  CTV 

 Additional volumes with presumed sub-clinical spread spr ead can als also o be con consid sidere ered d for th thera erapy py,,   e.g., regional lymph nodes or metastases. They are also defined defi ned as CT CTVs. Vs. Th Thus, us, two typ types es of sub sub-cl -clini inical cal disease (adjacent to the GTV or at a distance,   e.g., lymph nodes) can be envisaged. The prescription is then based on the probability that there are cancer cells in some anatomically definable tissues/organs, even though they cannot be detected with presentday techniques: techniques: the they y are sub-clinical. sub-clinical. For prescription of treatment, these sub-clinical deposits (or their probability of existence) can be described in terms of  frequency of risk for later detectable manifestations if not tr treat eated ed ade adequa quatel tely y in the subsub-clin clinical ical st stage age.. The estimate of that probability is based on clinical experience exper ience from adequ adequately ately docum documented ented trea treatment tment

5.1.2 Clini Clinical cal target volum volume e (CTV) (CTV)

Macroscopically Macroscopi cally,, tumor tumorss might seem rela relatively tively well delineated. However, when microscopic examination 84  

GEOMETRIC TERMS, AND DOSE AND DOSE–VOLUME DEFINITIONS

and follow-up (Gre´goire et al., 2000; Martinez-Monge  et al. , 1999, for head and neck tumors). 5.1.3 5.1 .3

Intern Int ernal al target target volu volume me (ITV) (ITV)

The PTV, as discussed in Section 5.1.4, allows for two components of uncertainty, namely namely,, internal uncertainties tain ties ( e.g., phy physiolo siologic gic mov movemen ements ts and varia variation tionss in size, shape, and position of the CTV within the patient) and uncertainties in factors external to the patient ( e.g., set-up uncertainties). The volume that includes the CTV plus an allowance for the internal compon com ponent ent of unc uncert ertai ainty nty is ter termed med th the e int intern ernal al targett volu targe volume me (ITV (ITV). ). The margin between between the ITV  and the CTV is termed the internal margin and is more fully described in Section 5.1.4.1. Delineation of  the ITV is considered optional. 5.1.4 Planni Planning ng target volum volume e (PTV)

The co The conc ncep eptt of PT PTV V was in intr trod oduc uced ed in IC ICRU RU Report 50 (ICRU, 1993b). The PTV is a geometrical concept, introduced for treatment planning. It surround rou ndss the CT CTV V wit with h ad addit dition ional al ma margi rgins ns to com com-pens pe nsat ate e fo forr di diff ffer eren entt ty type pess of va vari ria ati tion onss an and d uncertainties of beams relative to the CTV. The PTV has two functions: †

 It is a tool that can be used to help in the selection of the appropriate beam sizes and beam arrangements to ensure that the prescribed dose is actually al ly de deli live verred to al alll pa part rtss of th the e CT CTV V wi with th a

Beam Be amss ca can n be de desi sign gned ed di dire rect ctly ly fo forr th the e CT CTV V, taking into account the need for internal and external mar margin ginss wit within hin th the e ap apert ertur ure e des design ign,, wit withou houtt referenc refe rence e to a PTV. This is part particula icularly rly nat natura urall in the th e ca case se of pr prot oton on (a (and nd ot othe herr ch char arge ged d pa part rtic icle) le) beams. beam s. Nev Neverthe ertheless, less, PTVs mus mustt be defin defined ed since they th ey ar are e re requi quired red for re repor portin ting g pu purpo rposes ses.. Th These ese matters are further discussed in Section 5.1.4.4. Each Ea ch CT CTV V, and thu thuss GTV, mus mustt ha have ve a cor corre re-spon sp ondi ding ng PT PTV V. Th The e de deli line neat atio ion n of th the e PT PTV V is a required part of the treatment prescription. 5.1.4.1 Margins for the differe 5.1.4.1 different nt types types of  variations and uncertainties

 As discussed in Sectio Section n 7, a variety of techni techniques ques are availa av ailable ble for pa patien tientt imm immobi obiliza lization tion,, ass assess essmen ment, t, and reduction of organ motion. However, even once the appropriate techniques have been implemented, there always remains some degree of residual motion and som some e unc uncerta ertaint inties ies abou aboutt whe where re the pa patien tient, t, GTV, CTV, and OARs are located relative to the treatment equipment. These uncertainties must be taken into account in planning the treatment, and constitute the basis for determining the PTV. To avoid significant deviation from the prescribed dose in any part of the CTV(s), one must add margins to the CTV(s) CTV(s) to compe compensa nsate te for variatio variations ns and unceruncertainti tai nties es (i) in pos positi ition, on, size, and shape of the CT CTV V, and (ii) in patient-beam patient-beam positioning, positioning, both duri during ng a given radiation treatment fraction and between successiv ces sive e fr frac actio tions. ns. To fa facili cilita tate te dis discus cussio sion, n, and in

†

specified probability, qualified by the need to keep the th e ri risk sk of da dama mage ge to ad adja jacen centt no norm rmal al ti tiss ssue uess below a clinic clinically ally accep acceptable table level.   It It is of us use e in pr pres escr crip ipti tion on an and d re repo port rtin ing. g. Fo Forr example, a radiation oncologist will usually prescribe dose in terms of the coverage of the PTV:  e.g., ‘at least 95 percent of the PTV must receive 70 Gy (R (RBE BE)’. )’. [T [This his ex examp ample le can be mor more e con con-cisely written as ‘V  D 70Gy(RBE)(PTV)  95 percent’ (see Section 5.6.1.1.)]. It should be noted that the estimated distribution of dose within the PTV is an under underestim estimate ate of the dose distribution distribution in the CTV. This results because the CTV can remain well within the the PTV and hence not experience experience the reduced dose often delivered at the PTV boundary. In addition, the dose distribution within the CTV  is, if anything, an over-estimate at each point of  the dose the CTV will actually receive because the CTV can well move close to the PTV boundary and hen he nce ca can n rec ecei eiv ve a lo low wer per erip iph her era al dos ose. e. Therefore, the computed dose distributions to the PTVand PTV and CTV bracket bracket the dose at the specified confidence fidenc e level that the conten contents ts of the CTV are likely to receive, and they can be used to estimate the range of possible doses at that confidence level.

keeping with previous reports (ICRU, 1993b; 1999), these will be referred to as the internal margin (IM) and the set-up margin (SM), respectively.  Internal margin. This is the margin that must be added to the CTV to compensate for expected physiological movements and variations in size, shape, and position of the CTV during therapy. The IM is often asymm as ymmetr etric ic ar aroun ound d the CTV. Tech echniq niques ues suc such h as breath gating or tumor tracking exist to reduce the IM (see Section 7.5). As mentioned above, the volume encomp enc ompass assed ed wit within hin the IM is ter termed med the ITV. In

¼

practice, it might not be necessary to explicitly delineate the ITV, but the IM (as well as the SM) must be taken into account when delineating the PTV.  Set-up margin. To account specifically for uncertainties tain ties (ina (inaccur ccuracie aciess and lack of repr reproduc oducibilit ibility) y) in patient positioning and alignment of the therapeut pe utic ic be beam amss du duri ring ng tr trea eatm tmen entt pl plan anni ning ng an and d thro th rough ugh all tr trea eatm tment ent ses sessio sions, ns, an SM for ea each ch beam is needed. The uncertainties vary with different anatomical directions, and thus the size of such margins depends on the selection of beam geometries. The inaccuracies depend on such factors as: †

 variations in patient positioning;

85  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  †

†

†

  mech mechani anical cal un uncer certai tainty nty of the equ equipm ipment ent ( e.g., sagging of gantry, collimators, and couch);   transfer set-up errors from CT and simulator to the treatment unit;  human factors.

These uncertainties can also vary from center to center, and within a given center from machine to mach ma chin ine. e. Th The e us use e of pa pati tien entt im immo mobi bili liza zati tion on devices, devic es, the appl applicat ication ion of qual quality ity assur assurance ance programs, and the skill and experience of the radiographers/rad phers /radiothe iotherap rapists ists are impo important rtant and mus mustt be taken into account. The use of different record and  verification systems (in real time or not) can also be import imp ortant ant and sig signifi nifican cantly tly alt alter er the siz size e of the required set-up margins. 5.1.4. 5.1 .4.2 2

Delinea Deli neating ting the PTV 

The pr proce ocedur dure e for arr arrivi iving ng at an ov over erall all mar margin gin,, comb co mbin inin ing g th the e in indi divi vidu dual al co cont ntri ribu buti tion onss to th the e uncertainty, is discussed in Section 7.6. The needed margins are likely to differ in different directions.  A generally satisfact satisfactory ory approach is to assess them explicitly in the six cardinal directions: AP, PA, left, right, cephalad, and caudad. The PTV can then be del elin ine eated by exp xpan and din ing g the CTV by tho hose se margin mar gins. s. Th This is pr proce ocess ss can be per perfor formed med usi using ng a

the higher or lower doses are nested within the parent PTV. This situation can arise under two circumstances: (a) One wishes to ‘paint’ the dose non-uniformly within the tumor,   e.g., to give a higher dose to a region within the tumor that is judged to be of higher radio-resistance than the rest of  the tumor. (b) The tumor lies close to a sensitive uninvolved organ org an and therefo therefore re a sec sectio tion n of th the e tar target get  volume must receive a lower dose than that deli de live vere red d to th the e re rest st of th the e ta targ rget et vo volu lume me.. Nest Ne sted ed PT PTVs Vs mi migh ghtt th then en be ne need eded ed if gu guiidance is required for planning purposes. In general, delineation of the PTV margin should not be compromised even if it overlaps with another PTV, an OAR, or a PRV. In the case of 2(a), if the dose gradient within the target volume can be designed by an automated optimization scheme, there might be no need to define nested PTVs as the computer algorithm itself might determine automatically how the dos dose e sho should uld vary wit within hin the out outerm ermost ost PTV. However, it should be noted in the prescription that the PTV is intended to receive a non-uniform dose. If  nested nes ted PTVs ar are e nee needed ded for the purposes purposes of 2(a) 2(a),,

then the n the oute outermos rmostt sho should uld be del deline ineate ated d with without out

computer-based expansion tool. Manual expansion is ted tediou iouss and pr prone one to err error or,, as the out out-of -of-pl -plane ane dimension is not easily appreciated while delineating the PTV in any given plane.

comprom comp romise ises s due to overla ov erlappi pping ng on with wit h nei neighb ghbori oring ng  volum  volumes es of interest (VOIs, see Secti Section 5.3.1). 5.1.4.4 5.1.4 .4

Proton-spec Proto n-specific ific issues issues regarding regarding the PTV  PTV 

(1) When there is more than one CTV. For example, there ther e migh mightt be two spatially spatially sepa separat rated ed GTV GTVs, s, which then have two spatially separated CTVs associated with them, which, in turn, need separate PTVs. (2) When it is desired to irradiate a target volume

The ma The mate teri rial al in Se Sect ctio ion n 5. 5.1. 1.4 4 up to th this is po poin intt would wou ld app apply ly equ equall ally y to any ra radia diatio tion n mod modali ality ty.. However, there are some differences in the way the PTV is used in pro protonton-beam beam therapy. For photon beams, the PTV is primarily used to determine the lateral beam margins. In the case of protons (and other oth er cha charge rged-p d-part articl icle e bea beams) ms),, in add additi ition on to the lateral margins, some margin in depth must be left to al allo low w fo forr un unce cert rtai aint ntie iess in th the e kn know owle ledg dge e of  where the distal (say) 90 percent isodose would fall. The be bea am en ener ergy gy (i.e., pe pene netr trat atio ion) n) sh shou ould ld be chosen such that the CTV is within the irradiated  volume taking into account   both  motion   and  range unce un cert rtai aint ntie ies. s. Th Thus us,, fo forr pr prot oton ons, s, th the e la late tera rall margins and the margins in depth (relative to the proximal and distal tumor surfaces) solve different problems and will virtually always be numerically differen diffe rent. t. As a consequence, consequence, for each beam orien orien-tatio ta tion n bei being ng con consid sider ered, ed, one wou would, ld, in pri princi nciple ple,, need a separate PTV with different margins laterally and along the direction of each beam.  An alterna alternative tive appr approach oach is to determ determine ine the beam parameters using the CTV, CTV, rather than the PTV, and

non-uniformly. In this case, the PTVs receiving

to pla place ce the burd burden en of add adding ing app approp ropria riate te la later teral al

5.1.4. 5.1 .4.3 3

Multipl Mul tiplee PTV PTVss

The intention is usually to deliver an as-uni asuniform form-as -as-po -possi ssible ble dos dose e to one spe specifi cificc CT CTV V, subject only to the restrictions of physical achievability it y an and d th the e ne need ed to ba bala lanc nce e tu tumo morr co cont ntro roll wi with th spar sp arin ing g of ad adja jacen centt no norma rmall ti tiss ssue ues. s. In th that at cas case, e, there will be only one corresponding PTV. However, there are cases in which one specifically wishes to prescribe a non-uniform to the target for volume(s). It is recommended that dose the prescription this be accomplished by defining multiple-nested GTVs and/  or CTVs, each with their own PTV to each of which a uniform uni form dos dose e is pr presc escribe ribed d (see Section Section 5.4. 5.4.1). 1). The circumstan circum stances ces in which more than one PTV might be needed include the following:

86  

GEOMETRIC TERMS, AND DOSE AND DOSE–VOLUME DEFINITIONS

and range margins to the computer algorithm. That is, both lateral and depth margins are computed in designing each beam. In the case of scattered-beam treatment trea tments, s, the lateral margins would be design designed ed into int o the ape apertu rture re in the beam beam’s’s-ey eye e vie view w, and the depth margins would be designed into the compensator.. For scann sator scanned ed beams, and intens intensity-m ity-modula odulated ted proton pro ton the therap rapy y (IM (IMPT) PT) in gene general ral,, the these se mar margin ginss would influence which pencil beams would be used, and each one’s depth of penetration. It is required that the dose distribution within the ‘PTV ‘P TV’’ be re reco cord rded ed an and d re rep por orte ted. d. Th This is wo woul uld d be unworka unw orkable ble if the there re were a sep separ arate ate PTV for each bea beam employ emp ed,gins and imposs imp ible if sep separ ate lat latera eral andmdep depth th loyed, margin mar s wer were e ossibl built bui lt einto thearate compute comp uter’s r’sl beambea m-de desi sign gn al algor gorith ithm. m. It is th ther erefo efore re pr prop opos osed ed that, in proton therapy, the PTV be defined relative to th the e CT CTV V on th the e ba basi siss of la late tera rall un uncer certa tain inti ties es alone. An adjustment must then be made within the beambea m-de desi sign gn alg algor orit ithm hm to ta take ke in into to ac accou count nt th the e differences, differ ences, if any any,, betw between een the margins needed to account accou nt for uncer uncertainti tainties es along the beam direction (i.e., range uncertainties) and those included in the

Figure 5.2. Illu Figure Illustr strati ation on of the volumes volumes and margins relating relating to the definition of organs at risk.

tubular organs such as the rectum. In this case, the extent of delineation should be carefully reported;  e.g., ‘th ‘the e sp spina inall cor cord d was del deline ineat ated ed   + x  c  cm m fro from m the cephalad and caudad edges of the PTV’, or ‘the rectum was delineated up to the sigmoid flexure’. The delineation of tubular organs, or organs with

cavities, for which the organ wall is of interest ( e.g., the bladder wall), poses significant technical difficultie cul ties. s. If the organ wa wall ll is thi thick ck (r (rela elativ tive e to th the e  voxel size in the imaging study), both the outer and

so-defined PTV (i.e., based on lateral uncertainties). 5.2 ANA ANATOMIC TOMIC VOLUM VOLUMES ES RELA RELATING TING TO TO UNINVOLVED UNINVOL VED NORMAL TISSUES  AND ORGANS Just as for vol Just volum umes es ass associ ociate ated d wit with h tum tumors ors,, the delineat delin eation ion of norm normal al tiss tissues ues and organs that are  judged to be uninvolv uninvolved ed is no different for protons than for any other radiation modality. The volumes relat rel ating ing to the uni uninv nvolv olved ed nor normal mal tis tissue suess (O (OAR AR and PRV) are schematically represented in Fig. 5.2.

the inner wall can be delineated, and the dose statistic is ticss for the differen difference ce in th these ese vo volum lumes es can be computed and reported. If the organ wall is thin, this th is st stra rateg tegy y is pr prone one to err error or.. The bes bestt sol soluti ution on then th en is to defi define ne a ‘su ‘surfa rface ce of int inter eres est’ t’ (SO (SOI) I) (se (see e Section 5.3.3) and compute and report the dose statistics for that surface. A dose–area histogram (see Section 5.6.2.3) would be useful for this purpose.

5.2.1 5.2 .1

5.2.2 Plann Planning ing organ organ at at risk volum volume e (PRV) (PRV)

Organ Orga n at at risk risk (OA (OAR) R)

Normal tiss Normal tissues ues and orga organs ns who whose se rad radiat iation ion sen sensisitivity can significantly influence treatment planning and/ an d/or or pr presc escrib ribed ed dos dose, e, te terme rmed d OA OAR, R, sh shou ould ld be delineated. It is de desir sirab able le th that at th the e en enti tire rety ty of al alll OA OARs Rs be imaged and delineated, even if this involves imaging parts of the OAR that are not expected to be included in the treatment beams. This is useful in order to be able to characterize irradiated parts of the OAR in terms of their fractional volume—a quantity that is often useful in satisfying prescription constraints or estimating normal-tissue complication probabilities. When Wh en it is no nott pr prac acti tical cal or cli clini nical cally ly des desir irab able le to image the full organ, its full volume can, for purposes of estimating fractional volumes only, be estimated. If the relative volume is based on such an estimate, it should be clearly clearly so stated. stated.

In order to ensure with a high probability that adequate sparing of OAR will actually be achieved, an integrated margin has to be added to the OAR, using th the e sam same e pri princi plessgy of with IM and PTV PT V. This leads, innciple analogy analo the SM PTV,astofor thethe concept of PRV. The PR PRV V is a tool designed to help treatment planning and evaluation. The design of the necessary margins follows the same principles as for the design of a PTV, and these are discu discussed ssed in Sect Section ion 5.1.4, above, abo ve, and in Secti Section on 6.3. For the PRV, the uncertainties that need to be taken into account include:

†

 movements of the OAR during treatments;  variations in patient positioning;   mech mechani anical cal unc uncert ertain ainty ty of th the e equ equipm ipment ent ( e.g., sagging of gantry, collimators, and couch);   transfer set-up errors from CT and simulator to

†

the treatment  human factors.unit;

† † †

Often Oft en itl isor infeas inf easibl ible echto as image ima gee in their the entire ent irety ty cylind cyli ndri rica cal orga gans ns such su the th spin sp inal alir co cord rd or 87  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

Each institute should evaluate its own geometrical uncertainties and variations as these depend on clinical clinic al pro procedur cedures, es, leve levell of skill skill,,   etc. Eval Evaluat uation ion should be representative of routine clinical practice. Uncertainties and variations should be investigated forr ea fo each ch pa pati tien entt gr grou oup p ( e.g., se sett-up up er erro ror) r),, bu butt shou sh ould ld al also so be co cons nsid ider ered ed fo forr ea each ch in indi divi vidu dual al patie pa tient nt ( e.g., re respi spira rator tory y mot motion ion). ). In dec decidi iding ng on the margin to be allo allowed wed between between OAR and PR PRV V, all uncertainties should be taken into account. For reporting, it is recommended that, as for the PTV, the PRV be described by including the size of 

such a si such situ tuat atio ion, n, it is pe perm rmiss issib ible le to re refe ferr to th the e general term ‘target volume’. It would not be correct to state, say, ‘a mean dose of 75 Gy (RBE) was delivered to the target volume’; rather, one should use the specific term, as in ‘a mean dose of 75 Gy (RBE) was delivered to the PTV’. However, it would be permissible ibl e to st stat ate, e, for exa examp mple, le, tha thatt ‘al ‘alll tar target get vol volum umes es must be delineated before planning can proceed’.

th the e ctio comb co mbin ed marg rgin ins satio of the e OA OAR inV di diff ffer eren ent dire di rect ions ns..ined The Th ema deli de line neat ion n th of the th eR PR PRV marg ma rgin int should not be compromised even if it overlaps with another PRV, PTV, or OAR.

planar surface,  e.g., to report the dose to the skin, to the surface of the spinal cord, or the dose distribution receiv rec eived ed by the surface surface of the rectum rectum (se (see e Sect Section ion 5.2.1). 5.2. 1). In suc such h a cir circum cumst stanc ance, e, the surface surface can be designated a surface of interest (SOI). An SOI has no thickness, although in practice it is likely to be rep-

5.2.3 Rema Remaining ining volu volume me at at risk (RV (RVR) R)

5.3.3 5.3 .3

Surfac Sur face e of inter interes estt (SOI)

It is sometimes necessary to specify what happens on the th e su surf rfa ace of a VOI or on a sp spec ecifi ified ed cu curv rved ed or

resented by the voxels that lie closest to it.

The volume that is (i) within the imaged region of  the pa patie tient nt,, and (ii (ii)) out outsid side e all del deline ineat ated ed OA OARs Rs and CT CTVs Vs sho should uld be ide ident ntifie ified d as the ‘re ‘rema maini ining ng  volume at risk’ (R (RVR). VR). Doses to the RV RVR R should be reported in addition to the doses to specifically-delineated nea ted volumes of inter interest est (Sec (Section tion 5.3.1) in order to ensure that attention is paid to all tissues, and not just a sel select ected ed su subse bsett of the them. m. Fo Forr exa exampl mple, e, there the re cou could ld be uns unsusp uspect ected ed re regio gions ns of hig high h dos dose e within the patient that would go undetected if the RVR were not explicitly evaluated. In addition, the dose dos e to the RVR can be use usefu full in est estima imatin ting g the risk of late effects such as carcinogenesis.

5.3.4 5.3 .4

It is often useful to designate points in space (not necess nec essari arily ly wit within hin the pa patie tient nt vo volum lume, e, and not necess nec essari arily ly re repr prese esenti nting ng ana anatom tomic ic lan landma dmarks rks). ). When needed, these can be referred to as points of  interest (POIs). Examples POIs  Internal reference pointsof. such These Th ese areare: anatom ana tomica icall landmarks ( e.g., bony structures, gas-filled cavities, or surgical clips), which can be used for localization of the GTV, CTV, and OARs and for accurate setup at the imaging unit, simulator, and treatment unit. Often Oft en sep separ arat ate e ref refere erence nce poi points nts mu must st be use used d for different beams and if there is more than one GTV  or CTV.  External reference points. These are palpable or  visible points located on or near the surface of the body bod y or on th the e sur surfa face ce of imm immobi obiliz lizat ation ion dev device icess tha th at fit cl clos osel ely y to the bod ody y co con ntou ourr ( e.g., face masks, mask s, bite blocks, and shel shells). ls). As exte external rnal refer-

5.3 GEN GENERIC ERIC GEOM GEOMETRI ETRIC C TERM TERMS S 5.3.1 Volume of inter interest est (VOI) (VOI)

The VOI is a generic term that can be used to refer to any volume that needs to be identified identified.. The GTV, PTV, PTV, and OAR are examples of specifically-named VOIs. Sometime timess reasons, the delineation delineat of volu volumes mes useful   e.g.,ion forSome technical to identify oneisorusefu morel  volumes surroundin surrounding g the PTV to guide dose fall-off  in an opt optimi imiza zatio tion n pr proce ocedu dure; re; or to ide identi ntify fy sub sub-regio re gions ns of a PT PTV V to all allow ow the pa patch tching ing of pr proto oton n beams (i.e., the treatment of one part of a target  volume by one beam and the remaining part(s) by another beam or beams). The term ‘VOI’ should be used use d for the these se vo volum lumes, es, qua qualifi lified ed by its use use,,   e.g., ‘VOI (posterior patch for PTV-T1)’. 5.3.2 5.3 .2

Point Poi nt of of inter interes estt (POI) (POI)

ence po ence poin ints ts,, on one e ca can n al also so us use e sk skin in ma mark rkin ings gs or alignment align ment tat tattoos toos that are rep reprodu roducibly cibly rela related ted to the body as a whole ( e.g., skeletal structures).  ICRU reference point. A particular point of interest is the so-called ICRU reference point introduced in IC ICRU RU Re Repo port rt 50 (I (ICR CRU, U, 19 1993 93b) b).. It is a po poin intt selected according to the following general criteria: (1) the dose at the point should be clinically clinically relevant; relevant; (2) the point should be easy to define in a clear and unambiguous way; (3) the point should be selected where the dose can be accurately determined; (4) the point should be selected in a region where there is no steep dose gradient. Its use is discussed in Section 5.6.3.

Targ arget et volu volume me (TV)

The term ‘target volume’ has largely been replaced by the more specific terms GTV, CTV, and PTV. However, it is sometimes necessary to refer to the general category of VOIs that are associated with the tumor. In 88  

GEOMETRIC TERMS, AND DOSE AND DOSE–VOLUME DEFINITIONS

5.4 NOM NOMENC ENCLA LATUR TURE E Often th Often there ere ar are e mu multi ltiple ple tar target get vol volum umes es an and, d, in such situ situatio ations, ns, a nomen nomenclat clature ure (nam (naming ing conv convenention) is need needed. ed. It shou should ld be unam unambiguo biguous, us, clear clear,, and, where possible, provide insight into the nature of the object being named. Previous ICRU reports ( e.g., ICRU 1993b; 1999; 2004) 200 4) ha have ve add addres ressed sed the iss issue ue of tar target get vo volum lume e nami na ming ng co conv nven enti tion ons. s. Ho Howe weve verr, be beca caus use e of th the e increasing complexity of dose prescription in proton beams and other modern therapies, a further clarificatio fica tion n is giv given en her here, e, pri primar marily ily in ord order er to fa facil ciliitate the distinction between multiple-nested target  volumes and multiple spatially-sepa spatially-separated rated target

but th but the e mo more re co comp mple lex x si situ tuat atio ions ns de desc scri ribed bed in Sections 5.4.1.3–5.4.1.6 are often encountered. 5.4.1.1 5.4.1. 1 One GTV GTV plus a surro surroundi unding ng volume volume intended to receive the same dose

In this case, ther there e is a single GTV, GTV, and a single CTV  and a single PTV are associated associated with it as illustrated illustrated in Fig. 5.3a. The CTV can be identical with the GTV  or it can inc includ lude e allo allowa wance nce for un undete detecte cted d dis diseas ease e that is intended to receive the same dose as the GTV. 5.4.1.2 5.4.1. 2 One GTV GTV plus a surro surroundi unding ng volume volume intended to receive a lower dose

 At first glance, the case of a single GTV plus a

 volumes. It is recommended that: †

†

 volume of possib  volume possible le microscopic microscopic ( i.e., not grossly obser vable) diseas disease e appea appears rs the same as the prece preceding ding example. However, in this case, the outer region is intended to receive a lower dose, perhaps because it has only a small probability of harboring microscopic disease. For example, it might be intended to deliver a dose of, say, 70 Gy (RBE) to the inner region and a dose of 48 Gy (RBE) to the outer region. The inner region that is intended to receive the higher dose will include the GTV, and it might also

  nested target volumes be labeled with the extensions ‘1’, ‘2’,  etc., with the outermost volume being design des ignate ated d ‘1’ ( e.g., GTV GTV-1, -1, GTV GTV-2, -2,   . . ., or GTV 1, GTV 2,  . . .);   spa spatia tially lly-se -sepa para rated ted tar target get vo volum lumes es be lab labele eled d with wit h th the e ex exten tensio sions ns ‘A ‘A’, ’, ‘B’ ‘B’,,   etc. ( e.g., GTV GTV-A, -A,  . . . GTV-B,  . . ., or GTV  , GTV  , )  A  B  Alternatively  Alternativ ely,,

for

nested target

volumes, a

dose-based nomenclature can be employed whereby each ea ch ta targ rget et vo volu lume me is qu qual alifi ified ed by a no nomi mina nall  e.g. RBERB E-we weig ight hted ed ab abso sorb rbed ed do dose se [ , GT GTV V-4 -48 8 Gy (RBE), (RB E), GTV GTV-70 Gy (RBE), (RBE),   . . .   or GTV 48Gy 4 8Gy (RBE) , GTV 70Gy The e us use e of th this is al alte tern rnat ativ ive e 70Gy (RBE),   . . .]. Th nomenclat nomen clature ure has the virtu virtue e tha thatt it conv conveys eys more directly dire ctly infor informat mation ion usef useful ul for plan planning ning,, becau because se different doses are generally planned for different target tar get vo volum lumes. es. Ho Howe weve verr, as tar target get vo volum lumes es are generally created before planning the treatment, it hass th ha the e di disa sadv dvan anta tage ge th that at th the e ta targ rget et vo volu lume mess would ideally need to be renamed should the iterative planning process (see Section 10.2.2) require a change cha nge in the tar target get vol volum ume e dos dose(s e(s). ). Ad Addit dition ional al qual qu alifi ifier erss can be ad add ded to thes ese e nam ame es, as described in Section 5.4.4. Because the naming of the various VOIs can be rathe ra therr com comple plex, x, it is re recom commen mended ded th that at,, for ea each ch patie pa tient, nt, a dr draw awing ing be pr prov ovide ided d to doc docum ument ent the relation rela tionss betw between een the defin defined ed VO VOIs Is (unl (unless ess these relationships are self-evident). 5.4.1 5.4 .1

include an allo include allowan wance ce for und undetect etected ed gros grosss dise disease ase that th at is int intend ended ed to re recei ceive ve th the e sam same e dos dose e as the GTV GT V. Th The e ent entire irety ty of thi thiss inn inner er re regio gion n wou would ld be defined by a CTV, and the outer region intended to receive a lower dose would be defined by a different, larger CTV. The formal way of describing this situation is to identify the outer region as CTV-1, and to iden identify tify the regi region on inclu including ding the GTV and and possib pos sible le und undete etecte cted d gro gross ss dis diseas ease e as CT CTV V-2. Th The e GTV can be identified as GTV-2 as it is associated with wit h CT CTV V-2. Ea Each ch of the these se CT CTVs Vs wou would ld th then en be associated with a corresponding PTV. Often in such a situation, however, no undetected gross disease is suspected; CTV-2 is coincident with GTV and the higher dose is intended intended to be deliv delivered ered to the GTV without extensions. In such a situation, a shorth sho rthand and pre prescr script iption ion suc such h as ‘a RB RBE-w E-weigh eighted ted dose of 70 Gy (RBE) to the GTV and 48 Gy (RBE) to the CTV is to be delivered’ delivered’ would would be permissible. permissible. 5.4. 5. 4.1. 1.3 3

No GTV 

Figure 5.3c illustrates the nomenclature in the case in which there is a CTV, but no underlying GTV, in one location. This would be the case, for example, if  the surgical bed after complete resection were to be irradiated.

Multip Mul tiple le GTVs, GTVs, CTVs CTVs,, and PTV PTVs s

The examples that follow illustrate the nomenclature tur e and are schem schematic atically ally illustrate illustrated d in Fig. 5.3 (both alphanumeric- and dose-based nomenclatures are shown in Fig. 5.3, but, for simplicity, only the alphanum alph anumericeric-based based nomen nomenclat clature ure is used in the following sub-sections). Of these examples, those of  Sections 5.4.1.1 and 5.4.1.2 are the most common,

5.4.1.4 Two spatially 5.4.1.4 spatially separ separated ated GTVs Figur Fig ure e 5.3 5.3d d ill illus ustra trates tes th the e cas case e of two spa spatia tially lly-separated GTVs: as, for example, in the case of a 89

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

Figure 5.3. Sever Figure Several al scen scenario arioss sche schemat matical ically ly illu illustr strati ating ng nam naming ing conv conventi entions ons for GTVs GTVs,, CTVs CTVs,, and PTVs PTVs.. Both nume numerica ricall and dose-based nomenclatures are illustrated. (a) Single GTV; (b) a GTV and surrounding region of possible microscopic ( i.e., sub-clinical) disease; (c) a region of suspected disease without gross tumor ( e.g., a surgical bed); (d) two spatially separated GTVs; (e) two nested GTVss with the inner one inte GTV intended nded to rec receive eive a boos boostt dose dose,, plus a surr surround ounding ing region region of possible disease; disease; ( f) a neig neighbor hboring ing OAR requiring a reduced dose to the tumor in its neighborhood, plus a surrounding region of possible disease. The PRV surrounding the OAR is not shown for reasons of clarity. The RBE-weighted absorbed doses shown in this figure are for purposes of illustration only and would be replaced by the doses relevant to the particular case.

5.4.1.5 Two nested 5.4.1.5 nested GTVs plus plus a region region of possible possible microscopic disease

head and neck tumor with a palpable neck node. In such su ch a ca case se,, th the e GT GTVs Vs ar are e na name med d GT GTV V-A -A,, an and d GTV-B, respectively, and their corresponding CTVs and PTV PTVss are label labeled ed CTV CTV-A, -A, CTV CTV-B, -B, PTV PTV-A, -A, and PTV-B, PTV -B, respectively, respectively, with the obvious extension to the case of more than two GTVs. It can be useful, for planning purposes, to delineate the union between two or more VOIs so that the dose

Figure 5.3e illustrates the nomenclature in the case in wh whic ich h th ther ere e ar are e tw two o ne nest sted ed GT GTVs Vs,, wi with th th the e inner one intended to receive a boost dose, plus a surrounding region of possible microscopic disease. In all all,, th then, en, th there ere are th three ree re regio gions ns int intend ended ed to receive three different dose levels. Formally, a CTV  is ass associ ociat ated ed wit with h ea each ch of th the e thr three ee re regio gions, ns, but practica pra ctically lly,, CTV CTVss are likely to be coinc coincident ident with the GTVs for the two inner regions. In describing such suc h a sit situa uatio tion, n, it wou would ld be per permis missib sible le to omi omitt

distribution to the two becase planned and evaluated together. An volumes example can is the in which two CTVs, CTV-A, and CTV-B, overlap and the same dose is desired to be given to each. The union of two such volumes can be referred to as CTV-AB. 90  

GEOMETRIC TERMS, AND DOSE AND DOSE–VOLUME DEFINITIONS

mention of the two inner CTVs (data not shown in Fig. 5.3e). All nested GTVs should have at least an associate assoc iated d PTV. The exte extension nsion to mult multiple iple inne innerr GTVs GT Vs re rece ceiv ivin ing g bo boos ostt do dose sess wo woul uld d re requ quir ire e th the e identi ide ntifica ficatio tion n of tar target get vo volum lumes es wit with h na names mes lik like e GTV-3A, GTV-3B,   etc., toget together her with thei theirr assoc associi-

 volume to be treated is derived by synthesizing information from several studies with which several GTVs GT Vs ha have ve bee been n del deline ineat ated, ed, sa say y, GTV GTV(PE (PET) T) and GTV(cont GTV (contras rastt CT), then the volu volume me to be trea treated ted should be termed simply GTV. If there were two or more spatially separated GTVs, then they could be

ated CTVs and PTVs.

qualified as GTV A(CT), GTV B(PET),  etc .

5.4.1.6 Reduc 5.4.1.6 Reduced ed tumor tumor dose dose in a region region closely closely adjacent to an OAR

5.4.4.1 Tumor 5.4.4.1 umor,, nodal, or or metastatic metastatic basis basis for a target volume

Figure 5.3f illustrates the case in which a reduced dose is to be given to the part of the target volume that abuts a sensitive OAR and so is divided into two pa parts rts wit with h dif differ ferent ent dos dose e pr presc escrip riptio tions ns (se (see e Section 5.1.4.3). This situation is topologically identical to the boost example illustrated in Fig. 5.3e. It differs only in that the region to receive the ‘boost’ dose is much larger.

There is a value, as emphasized in ICRU Report 71 (ICRU, 2004), in indicating whether a given GTV or CTV CT V re repr prese esent ntss th the e pr prim imar ary y tu tumo morr, or gr gross ossly ly involved nodes, or grossly visible metastases. As that report stated ‘Adding the letters T, N, or M to identify the volumes might better clarify their clinical significance can ce (com (compar pared ed to the ide identi ntifica ficatio tion n by num numbers bers only, as used in ICRU Reports 50 and 62)’. For this purpos pur pose, e, tha thatt rep report ort pr propo oposed sed the nom nomenc enclat latur ure e GTV-T, CTV-T, CTV-N,   etc., and CTV CTV-N1, -N1, CTV CTV-N2, -N2,  etc. in the case of multiple spatially-separated CTVs. Whil Wh ile e th this is no nome menc ncla latu ture re is un unam ambi bigu guou ous, s, an and d rema re main inss en enti tire rely ly va valid lid,, th the e ge gene nera rall met metho hod d fo forr

5.4.2 5.4 .2

Multip Mul tiple le OARs OARs and PR PRVs Vs

Usually, several OARs are delineated. These could, in principle, be designated OAR-1, OAR-2, etc., and their corresponding PRVs can be named PRV-1, PRV-2,  etc. However, it is much more helpful to name the OAR accor ac cordin ding g to th the e an anat atomi omicc ent entit ity y it re repr pres esent ents, s, in which case one can have OARs such as ‘prostate’, ‘left kidney’,   etc., wit with h th the e cor corre resp spond ondin ing g PR PRVs Vs bei being ng named PRV (prostate), PRV (left kidney), etc. 5.4.3 5.4 .3

in incl clud uding ing qual qu ifyin ying g in info form ation ionthepr pres esen ente d in Section 5.4.4 isalif preferred. Itrmat allows use ofted names such as CTV(N), CTV-1(N), PTV  D 70 Gy (RBE)  DRBE (T), etc., and leaves room for further qualifying informatio ma tion n suc such h as GTV [M, PET, 50 Gy (RB (RBE)] E)] for a region reg ion of met metas asta tatic tic spr spread ead ide identi ntified fied by PET and intended to receive a dose of 50 Gy (RBE). ¼

Number Num ber of RV RVRs Rs

Multiple RVRs do not exist. As the RVR represents the volume volume wit within hin the pa patie tient nt in whi which ch no ana ana-tomic structure has been delineated, it follows that there can be only one RVR.

5.5

VARIA ARIATION TION OF GEOMETR GEOMETRY Y WITH TIME

 A pa patien tient’s t’s geom geometry etry vari varies es with time time.. Thu Thus, s, the regions reg ions asso associa ciated ted with VOIs, AOIs, and PO POIs Is vary with time. The variation can be intra-fractional and rapid ra pid,, as wit with h re resp spir irat ation ion or he heart art be beat at.. It can be

5.4.4 Quali Qualificati fication on of geometri geometric c terms

Not inf infreq requen uently tly, one wis wishes hes to ass associ ociat ate e one or more pieces of qual qualifyin ifying g info informat rmation ion with a geometricc term. For example, metri example, ther there e migh mightt be seve several ral imagin ima ging g st stud udies, ies, ea each ch one of whi which ch ind indica icates tes an apparent appa rently ly diff differen erentt exte extent nt of disea disease, se, and, based on th thes ese e st stud udie ies, s, a de deci cisi sion on mu must st be ma made de as to what GTV should be treated. In such a case, one would identify a GTV for each imaging study, qualified by the type of study, plus an overall GTV used to plan the treatment. It is recommended that qualifying information be placed within parentheses immediately following the name of the geometric term and that sufficient informati ma tion on be in incl clud uded ed as is ne need eded ed fo forr cl clar arit ity y. For example, one might write: GTV(PET), GTV(contrast CT), GTV(T2-weighted MRI), etc. No recommendation is made as to the specific form of the qualifying information. However, if the meaning of the nomenclature used could be unclear, it should be explained. If the

inter-fractional and slow, as with tumor regression or weight loss. In any event, any representation of geometr me try y mu must st be sp spec ecifi ificc to an id iden enti tifie fied d ti time me.. If  stud st udies ies at mu mult ltipl iple e ti times mes ( e.g., at dif differ feren entt br brea eath th phas ph ases) es) ar are e av avail ailabl able, e, th the e VOIs det determ ermin ined ed fr from om these studies should be qualified ( e.g., by the phase of  breathing). The relationship between objects of interest, es t, suc such h as im image agess or VO VOIs, Is,   etc., asso associa ciated ted with diffe di ffere rent nt tim times es sh shoul ould d be re refe ferr rred ed to on one e an anoth other er through thr ough a mat mathem hemati atical cal tra transfo nsforma rmation tion such as is accomplished by image-registration techniques. 5.6 DOSE AND DOSE– DOSE–VOL VOLUME UME REL RELA ATED DEFINITIONS The RBE-weighted absorbed dose to water should be computed and reported. The ultimate description of the dose del deliv ivere ered d to the patient patient is th the e thr threeeedimension dime nsional al dis distribu tribution tion of absol absolute ute (not rela relativ tive) e) 91

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

dose, e, sup superi erimpo mposed sed upo upon n a map of the pa patie tient nt’s ’s dos anatomy ( e.g., a CT scan). However, such a descrip-

estimation estima tion of thes these e qua quantit ntities ies does not imp impose ose any additio add itional nal comp computa utation tional al burd burden; en; both can be rea read d

tion is complex and voluminous; a threedimension dime nsional al absor absorbedbed-dose dose dist distribut ribution ion combi combined ned with wit h th three ree-d -dime imensi nsiona onall ana anatom tomic ic inf inform ormat ation ion is impo im poss ssibl ible e to ap appr prec ecia iate te in a st stat atic ic,, an and d he henc nce e repo re port rtab able le,, vi view ew.. For th this is re reas ason on,, a nu numb mber er of  approa app roache chess for phy physic sical al dos dose e sum summar mariza izatio tion n are used.. These inclu used include de oneone-dime dimension nsional al par paramete ameters rs such as the mean dose delivered to a VOI; and twodimens dim ension ional al dis distri tribut bution ions, s, suc such h as dos dose e –volu –volume me histograms (DVHs). In addition, biophysical models such su ch as eq equi uiva vale lent nt un unif ifor orm m do dose se (E (EUD UD), ), tu tumo morr control probability (TCP), and normal tissue complication probability (NTCP) are sometimes employed. The Th e de desc scri ript ptio ion n of do dose se an and d vol olum ume e po pose sess a problem of units, as they can be expressed as absolute ut e or rel relat ative ive qua quant ntitie ities. s. Th The e ter term m ‘ab ‘absol solute ute’’ or ‘relative’ should be used to indicate which of these is int intend ended ed whe whenev never er con confus fusion ion cou could ld ari arise. se. Th The e units of any quantity should always be stated. The use of ‘ml’ or ‘Gy (RBE)’, for absolute volumes or

from a DVH. Some examples are shown in Fig. 5.4. 5.6.1.1 The volume 5.6.1.1 volume receiving receiving at least least a specified dose (V  D )

The quantity  V  D  D  is the largest volume of a specified  VOI that receives a dose more than or equal to the RBE-weighted dose,   DRBE. Bo RBE-weighted Both th th the e vo volu lume me an and d the dose can be in absolute or relative units. Which is in inte tend nded ed is ma made de cl clea earr by th the e ad addi diti tion on of th the e appropriate units. For example, †

†

†

  V 70Gy 142 cm3 means ‘142 cm3 of the VOI 70Gy (RBE) receives at least 70 Gy (RBE)’.   V 70Gy 80 percent means ‘80 percent of the 70Gy (RBE)  VOI receives at least 70 Gy (RBE)’.   V 90% 142 cm3 means ‘142 cm3 of the VOI receives 90% at least 90 percent of the prescribed dose’. ¼

¼

  ¼

For relative volumes, the reference volume should be identified. Usually, it will be the entire volume of  the VOI, either as imaged or, if not fully imaged, as estimated (see Section 5.2.1). For relative doses, the refere ref erence nce dos dose e is tak taken en to be the pr presc escrib ribed ed dos dose e (see Section 5.6.3) unless otherwise stated.

RBE RBE-we -weighte ighted de absor absorbed doses, respectiv resp ely,,esand the per percen centag tage sign sig n bed (%) dose for s, relat re lativ ive eectively volum vo lumes or doses provides a sufficient clarification of whether the quantity is absolute or relative (as indicated in Section 5.6.1.1, the value to which the relative dose is normalized should be stated).

5.6.1.2 The least 5.6.1.2 least dose dose receive received d by a specified specified volume (DV  )

5.6.1 One 5.6.1 One-di -dimen mensio sional nal dose dose and and dose– dose– volume summarization

The qua quanti ntity ty   DV    is th the e le leas astt do dose se re rece ceiv ived ed by a  volume,  volum e, V , of a specified VOI. Expressed another way, the valu value e   DV    indica indicates tes tha thatt a vol volum ume e   V   of a VOI receives at least a dose equal to   DV . Both the dose and the volume can be in absolute or relative units.

The pre present sent rep report ort intr introdu oduces ces a nom nomencl enclatu ature re for dose– do se– vo volu lume me sp speci ecific ficat ation ionss (V  D   and   DV , de defin fined ed immediately immedia tely below). Prov Provided ided DVHs are available, the

Figure 5.4. A typical dose– dose – volume histogram histogram display, display, illustrating illustrating the use of dual axes (relative and absolute values) for both the dose and  volume dimensions  volume dimensions (unfortuna (unfortunately tely, most commerc commercial ial treatment treatment planning planning systems systems do not support this feature), feature), and how a numb number er of volu volume me and dose statistics can be read off from a DVH.  D 50% is identical to  D mean, and  D min and  D max are identical to  D 100% and  D 0%, respectively.

92  

GEOMETRIC TERMS, AND DOSE AND DOSE–VOLUME DEFINITIONS

Which is intended is made clear by the addition of  the appropriate appropriate unit to numbe numbers. rs. For exam example, ple, †

†

†

well-cho wellchosen sen two two-d -dime imensi nsiona onall sec sectio tions ns can giv give e a good goo d app apprec recia iatio tion n of th the e thr threeee-dim dimens ension ional al dos dose e distribution. Of particular help is the inspection of  orthogonal views, such as transverse, sagittal, and coronal views, side by side.

  D142ml 70 Gy (RBE) me mean anss ‘a ‘att le leas astt 70 Gy (RBE) is delivered to 142 ml of the VOI’.   D80% 70 Gy (RBE) means ‘at least 70 Gy (RBE) is delivered to 80 percent of the VOI’.   D142ml 90 percent means ‘at least 90 percent of  ¼

¼

5.6.2. 5.6 .2.2 2

¼

 A cumulativ cumulative e DVH is a graph of the volume of a specified VOI that receives at least a given dose as a function of that dose,   i.e., of   V  D   versus   D   (Chen, 1988; Drzymala  et al., 1991; Shipley  et al., 1979). A  differential DVH is an alternative, but less widely used, representation of the same information. It is a his histog togra ram m of th the e vo volum lume e re recei ceivin ving g a giv given en dos dose e within with in a spec specified ified,, gener generally ally small, dose interval, as a function of that dose. Both forms of DVH can employ either relative or absolute volumes, and relative or absolute doses. It is recommended that two axes be drawn for volume and two for dose dose,, givi giving ng res respec pectiv tively ely,, the rel relati ative ve and absolute values of these variables, as illustrated in Fig. 5.4. The DVH is a helpful quantitative tool to summarize the dose distribution to a given VOI. It suffers from the disadvantage that all spatial information is lost. It is recommended that DVHs be prepared for all specifically identified anatomic VOIs of  clinical interest, and for the RVR (see Section 5.2.3). For th the e RVR VR,, as it ca can n be a la larg rge e vol olum ume e of  perhaps several liters, and as hot spots can cover only a relatively small volume, it is recommended that its DVH, if dual axes are not available for the display, use the absolute volume and dose, not the relative volume or dose.

the prescribed dose is delivered to 142 ml of the  VOI’. Rela elati tive ve do doses ses an and d vo volu lume mess sh shou ould ld fo follo llow w th the e same rules as for  V  D  D  as described in Section 5.6.1.1. 5.6.1. 5.6 .1.3 3

Otherr dose Othe dose meas measure uress

There ar There are e a nu numbe mberr of oth other er dos dose-s e-spec pecify ifying ing par par-ameters, generally specific to a particular VOI ( e.g., the th e PT PTV) V) th that at can be us used ed to gu guid ide e or re repo port rt a treatment. These are the following: †

†

†

†

†

†

  Dmean: the mean dos dose e del delive ivered red to the specified specified  VOI. D average  can also be used.   Dmedian: the median dose delivered to the specified

 VOI. This param parameter eter is used primar primarily ily because of  the fact that it is easy to estimate, as it can be read dirrect di ctly ly fr from om a DV DVH H as   D50%   (see (see Fi Fig. g. 5. 5.4) 4).. However, if one wishes to use a single number to characterize the dose distribution, there are radiobiologi bio logical cal arg argume uments nts in fa favor vor of usi using ng the mea mean n dose dos e for tha thatt pur purpose pose (Brahme, (Brahme, 198 1984). 4). Thi Thiss dos dose e statistic is not as readily available as it cannot be read off directly from a DVH (it can, though, be straightforwardly calculated from a DVH).   Dmin: the minimum dose delivered to the specified VOI.  D min  is equivalent to  D 100%.   Dnear-min: the use of   D98%   is som someti etimes mes used to indica ind icate te the nea near-m r-mini inimu mum m dos dose e bec becaus ause e art artiifacts fa cts in th the e cal calcul culat ation ion or dis displa play y pr proce ocess ss can yield misleadingly low values of   D100%.   D98%   can be identified as the ‘near-minimum’ dose.   Dmax: the maximum dose delivered to the specified VOI.  D max  is equivalent to  D 0%.   Dnear-max: the use of   D2%   is som someti etimes mes use used d to indica ind icate te th the e nea near-m r-maxi aximu mum m dos dose e bec becaus ause e art artiifacts fa cts in th the e cal calcul culat ation ion or dis displa play y pr proce ocess ss can yield misleadingly misleadingly high value valuess of   D0%.   D2%   can be identified as the ‘near-maximum’ dose.

5.6.2. 5.6 .2.3 3

Dose– ar area ea hist histogr ograms ams (DAH (DAH))

By an anal alog ogy y wi with th DV DVHs Hs,, do dose–are se–area a hi hist stog ogra rams ms (DAHs) can be computed. A cumulative DAH is a graph of the area of a specified SOI that receives at least a given dose as a function of that dose. 5.6.3 5.6 .3

Presc Pr escrib ribed ed dos dose e

The ‘prescribed dose’ (or, equivalently, the ‘prescription dose’ ) is a dose whose value can be used as a reference when prescribing a treatment, and when evalua eva luatin ting g th the e dos dose e dis distri tribu butio tion n ac achie hieve ved. d. It is  valuable, for example, as the reference dose in presenting relative dose distributions, or in computing rela re lativ tive e dos doses es for a DV DVH H (se (see e Sec Sectio tion n 5.6 5.6.2. .2.2), 2), in which whi ch cas case, e, a re relat lativ ive e dos dose e of 100 per percen centt wou would ld

5.6.2 Two-dim wo-dimension ensional al dose– volum volume e summarization 5.6.2.1 Two-dimen 5.6.2.1 wo-dimensional sional dose displa displays ys The dis displa play y of dos dose e on a two two-di -dimen mensio sional nal pla plane ne with wi thin in th the e th thre reee-di dime mens nsio iona nall pa pati tien entt is a ve very ry common way of inspecting a three-dimensional distribution. However, a single image provides only a sampling, rather than a summarization, of the dose distribution. Nevertheless, the inspection of several

correspond to the prescription dose. In pr previ evious ous IC ICRU RU Re Repor ports ts (IC (ICRU RU 199 1993b; 3b; 199 1999; 9; 2004), the prescribed dose was represented by the ‘ICR ‘I CRU U Re Refe fere renc nce e Do Dose se’, ’, wh whic ich h is th the e do dose se to be delivered deliv ered to the so-ca so-called lled ‘ICRU Reference Reference Poi Point.’ nt.’ This Th is us usag age e ha had d th the e gr grea eatt be bene nefit fit of si simp mpli lici city ty.. However, modern radiotherapy has moved towards 93

 

Dose– vo volume lume his histog togram ramss (DVH) (DVH)

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

a volumetric approach in prescribing and assessing dose, largely because it is uncommon to deliver a comple com pletel tely y un unifo iform rm dos dose e to th the e ent entir ire e PT PTV V. Fo Forr example, it is common to use a dose–volume prescription,   e.g., tha thatt th the e ent entire ire PTV is re requi quire red d to receive at least 95 percent of the stated prescription dose.. The prescription dose prescription dose, ther therefor efore, e, is a refe referr-

5.6.5 5.6 .5

ence dos ence ose e that ca can n be th the e dos ose e at th the e IC ICR RU Reference Point, but is not necessarily so.

 volume. The Th e tr trea eate ted d vo volu lume me is th the e ti tiss ssue ue vo volu lume me th that at (according to the approved treatment plan) receives at le leas astt th the e RB RBEE-we weig ight hted ed do dose se se sele lect cted ed as th the e minimum dose to the PTV (or some specified percentage of the PTV), and specified by the radiation oncology team as appropriate to achieve tumor eradication or palliation, within the bounds of acceptable complications. In proton therapy,   D98%  could be selected to determine the treated volume. The treated volume is the volume enclosed by the isodos iso dose e sur surfa face ce cor corre respo spond nding ing to tha thatt dos dose e lev level. el. When reported, the value of the isodose selected to define defin e the treated treated volu volume, me, should be quote quoted d rela rela--

5.6.3.1 5.6.3 .1

Prescribed Pres cribed dose for multiple PTVs

It is not uncommon to have nested PTVs (see, for example, Fig. 5.3b and d–f). In this case, the physician will wish to prescribe a different dose level for each PTV. For example, one might wish to prescribe 48 Gy (RB (RBE) E) for PT PTV V-1, 70 Gy (RB (RBE) E) for PT PTV V-2, and 77 Gy (RBE) for a small boost region, PTV-3. In such a case, one can speak of a ‘prescribed dose for the particular volume’ and write it ‘prescribed dose (PTV-1)’, ‘prescribed dose (PTV-2)’,  etc . 5.6.3.2 Pres 5.6.3.2 Prescribed cribed dose for for multiple multiple trea treatment tment segmentss (segme segment (segment nt dose)

In th the e ex exam ampl ple e pr pres esen ente ted d in Se Sect ction ion 5. 5.6.3 6.3.1 .1,, th the e three PTVs could be irradiated simultaneously; for exampl exa mple, e, by emp employi loying ng non non-un -unifo iform rm fiel fields ds as with IMPT IM PT.. Fr Freq eque uent ntly ly,, ho howe weve verr, wh what at is do done ne is to deli de liv ver th the e tr trea eatm tmen entt in th thre ree e se segm gmen ents ts [s [see ee Tab able le 10 10.1 .1 (n (not ote e 8) 8)]] wi with th,, fo forr ex exam ampl ple, e, th the e fir first st segm se gmen entt pr pres escr crib ibed ed to de deli liv ver 48 Gy (R (RBE BE)) to PTV-1; the second segment prescribed to deliver an addi ad diti tion onal al 22 Gy (R (RBE BE)) to PT PTV V-2 -2;; an and d th the e th third ird segment a further 7 Gy (RBE) to PTV-3. In such a situation, one can speak of the ‘segment dose’. This is the incremental dose delivered in a given segment of th the e tr trea eatm tmen ent. t. In th the e ex exam ampl ple e ju just st gi give ven, n, th the e segm se gmen entt do dose se fo forr se segm gmen entt on one e wo woul uld d be 48 Gy (RBE (R BE); ); fo forr se segm gmen entt tw two, o, 22 Gy (R (RBE BE); ); an and d fo forr segment three, 7 Gy (RBE). The segment dose can be used as a reference dose in the prescription. For exampl exa mple, e, in the cur curren rentt cas case e one migh mightt pr prescr escribe ibe the dose for segment two as ‘PTV-2 shall receive at least 95 percent of the prescribed segment dose’.

Trea reated ted vol volume ume

Because of the limitations Because limitations of the irradiation irradiation techniques, nique s, the volu volume me rece receiving iving the pre prescribe scribed d dose might not match the PTV; it might be larger (sometimes tim es mu much ch lar larger ger)) and in gen gener eral al mor more e sim simply ply shap sh aped ed.. Th This is le lead adss to th the e co conc ncep eptt of tr trea eate ted d

tive to the prescribed dose (see Section 5.6.3) or in absolute terms. It is imp import ortant ant to ide ident ntify ify th the e sha shape, pe, size, and positi pos ition on of th the e tr trea eated ted volume volume in rel relat ation ion to the PTV for different reasons. One reason is to evaluate causes for local recurrences recurrences (inside or outs outside ide the treated volume). Another reason is to evaluate and interpret side effects. 5.6.6 5.6 .6

Conform Con formity ity ind index ex (CI (CI))

The conformity index (CI) is defined as the ratio of  the treated volume to the PTV. Ideally, the treated  volume should totally encompass the PTV. If this is not the case, the percentage of the PTV included in the treated volume should be reported. The CI can be used as part of the optimization procedure, as was proposed by Kno¨o¨ s   et al.  (1998) and van van’t ’t Rie Riett   et al.   (1997). (1997). How However ever,, it is reco recoggnize ni zed d th that at wh when en op opti timi mizi zing ng th the e CI (a (ass cl clos ose e as possible possi ble to unity unity), ), othe otherr optim optimiza ization tion para parameter meterss might deteriorate,   e.g., th the e si size ze of th the e ir irra radi diat ated ed  volume or the dose homogeneity in the PT V. 5.6.7 Irradi Irradiated ated volume volume (at a speci specified fied dose) dose)

5.6.4 5.6 .4

Relati Rel ative ve dos dose e

For all forms of dose reporting, if the dose is pre-

The irra The irradi diat ated ed vol olum ume e is th the e tiss tissue ue vol olum ume e that th at re rece ceiv ives es a do dose se co cons nsid ider ered ed si sign gnifi ifica cant nt in

sented as a relative rather than absolute value, a reference dose (to which the dose is considered relative) must be chosen. It is recommended that the reference dose be specifically identified, unless it is comp co mple lete tely ly cl clea earr fr from om th the e co cont ntex ext. t. It wo woul uld d be usual usu al tha thatt the referenc reference e dos dose e wou would ld be the prescribed scribe d dose, although although in some circumstance circumstancess the segment dose might be used.

relation to norma relation normal-tis l-tissue sue damag damage. e. The dose leve levell should shoul d be exp explicitl licitly y sta stated, ted,   e.g., irradiated volume (V  D 25Gy (RBE)). If the irradiated volume is reported, the significantt dos can dose e mus mustt be ex expli plicit citly ly ex expre presse ssed d eit either her in absolute absolu te valu values es or rela relativ tive e to the pres prescribe cribed d dose to the PT PTV V. The irr irrad adiat iated ed vo volum lume e dep depend endss on the treatment technique used. ¼

94  

Journal of the ICRU Vol 7 No 2 (2007) Report 78 Oxford University Press

6

doi:10.1093/jicru/ndm029

TREA TRE ATME TMENT NT PLA PLANNI NNING NG

6.1 INTR INTRODUC ODUCTION TION Tre reat atmen mentt pla planni nning ng is the pr proce ocess ss of sim simula ulatin ting g a nu numb mber er of de deli live very ry st stra rate tegi gies es fo forr a ra radi diat atio ion n trea tr eatm tmen entt an and d ch choo oosi sing ng th the e be best st on one e to us use e fo forr treatment. The simulation of the patient is based on a rec econ onsstr tru uct ctio ion n of the pati tien entt’s nor orm mal ana an ato tomy my an and d tu tumo mor( r(s) s) de deri riv ved fr from om im imag agin ing g stud st udie iess su supp pplem lemen ente ted d by de deli line neat atio ion n of ta targ rget et  volumes and organs at risk (OARs). A  plan  consists of an en ense semb mble le of be beam ams, s, to toge geth ther er wi with th th thei eirr wei eigh ghti ting ng fa fact ctor ors. s. Th The e be beam am pr prop oper erti ties es an and d weights can be generated manually, automatically, or semi-automatically. The dose within the patient can be cal calcul culat ated ed for any arr arrang angeme ement nt of bea beams ms using physical models of the beam properties. The resulting dose distributions within the patient can then the n be com compar pared ed am among ong riv rival al pla plans. ns. Th The e jud judgegement me nt of wh whic ich h pl plan an is be best st is co comp mple lex, x, an and d th the e experience of the treatment planner and radiation oncologist figures prominently. As described below, thiss ju thi judge dgemen mentt can be aid aided ed by met method hodss of dos dose e displa dis play y and com compar pariso ison, n, of dos dose e sum summa mariz rizat ation ion,, and an d by co comp mput uta ati tion on of se sev ver eral al mo mode dell-ba base sed d meas me asur ures es of do dose se im impa pact ct su such ch as tu tumo morr co cont ntro roll probability (TCP), normal-tissue complication probability (NTCP), and equivalent uniform dose (EUD) (see (se e Sec Sectio tion n 6.7 6.7.3) .3).. In int intens ensity ity-mo -modul dulat ated ed rad radiiation atio n ther therapy apy (IMR (IMRT), T), the gener generatio ation n and evaluatio at ion n of pl plan anss is is,, be beca caus use e of th the e co comp mput utat atio iona nall burden bur den,, nec necess essari arily ly per perfor formed med aut automa omatic ticall ally y by computer, often with manual iteration of the treatment aims at improving the plan further. In pla planni nning ng tr treat eatmen ments ts (u (unif niform orm int intens ensity ity or intens int ensity ity mod modula ulated ted)) wit with h pr proto otons ns as com compar pared ed with conventional photons, the features in common far out outwe weigh igh tho those se tha thatt ar are e dif differ ferent ent.. In Sec Sectio tion n

6.2 WHA WHAT T IS DIFF DIFFEREN ERENT T ABO ABOUT UT PLAN PLANNIN NING G PROTON-BEAM THERAPY? The dif The differ ferenc ences es bet betwe ween en pla planni nning ng pr proto oton-b n-beam eam 1 therapy and photon-beam therapy derive from the differences in the physics of protons and photons, namely, †

†

†

  th tha at th the e pe pene netr tra ati tion on of pr prot oton onss is str tron ongl gly y aff ffec ecte ted d by th the e na natu turre ( e.g., den ensi sitty) of the tissues tissu es thr through ough which the they y pass, while phot photons ons are much less affected (density changes generally give rise to only small intensity changes, except forr th fo the e lu lung ng). ). Th Ther eref efor ore, e, he hete tero roge gene neit ities ies ar are e much mu ch mor more e imp import ortant ant in pr proto oton-b n-beam eam the therap rapy y than in photon-beam therapy;   the apparatus apparatus for prot proton-b on-beam eam deliv delivery ery is diffe diffe-rent, and its details affect the dose distributions.

6.2.1 6.2. 1

Heterog Hete rogenei eneities ties

Because of the influence of heterogeneities, a map of het hetero erogen geneit eities ies alo along ng the bea beam m pa path th mus mustt be made and compensated for (to the extent feasible); final fin ally ly,, th the e do dose se di dist stri ribu buti tion onss mu must st re refle flect ct th the e remainin rem aining g effec effects ts of the hete heterogen rogeneitie eities. s. The map of hete heteroge rogeneiti neities es is built up from fine-resolut fine-resolution ion compu com puted ted tom tomogr ograp aphy hy (C (CT) T) ima images ges con conve verte rted d to water wa ter-eq -equiv uivalen alentt den densit sities ies (se (see e Sec Sectio tion n 6.4 6.4.6. .6.1) 1) in or orde derr to co comp mput ute e th thre reee-di dime mens nsio iona nall do dose se distributions. The resulting requirements for planning protonbeam therapy imply the following: †

6.2, som 6.2, some e of the mor more e imp import ortant ant dif differ ferenc ences es are briefly bri efly des descri cribed bed—t —the hey y ar are e tr trea eated ted mor more e fu fully lly in the th e fo foll llow owin ing g su subs bsec ecti tion onss or el else sewh wher ere e in th the e present report. The report concerns itself primarily with pro protonton-beam beam ther therapy apy.. How However ever,, the pre present sent

  th tha at pr prot oton onss ha hav ve a fin finit ite e an and d co cont ntro roll llab able le (through choice of energy) penetration in depth;

†

 to ascertain the CT Hounsfield number to waterequiva equ ivalen lentt den densit sity y con conve versi rsion on tab table le (S (Sect ection ion 6.4.6.1);  to compensate, either physically or virtually, for hetero het erogen geneit eities ies,, inc includ luding ing met metall allic ic imp implan lants ts

section is equally relevant to light-ion ( e.g., carbon or neo neon) n) the thera rapy py,, ap apart art fr from om Sec Sectio tion n 6.2 6.2.3. .3.4, 4, in which whi ch rel relat ativ ive e bio biolog logica icall eff effect ectiv ivene eness ss (RB (RBE) E) is discussed.

when present (Section 6.4.6.2); 1

J. Adam Adams, s, M. Moy Moyers, ers, P. P. Pet Petti, ti, S. Rose Rosentha nthal, l, B. Scha Schaffne ffnerr,  A. N. Schreude Schreuder, r, and L. Verhey contribut contributed ed significant significantly ly to this section.

# International Commission on Radiation Units and Measurements 2007

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  †

†

†

 to be aware of, and mitigate the effect of, possible hotho t-an andd-co cold ld sp spot otss du due e to la late tera rall sc scat atte teri ring ng effects (Section 6.4.4);   to tak take e int into o ac accou count nt un uncer certai tainti nties es ass associ ociat ated ed with possible misa misalignm lignment ent of the comp compensa ensator tor with the patient’s tumor, organs, and tissues;  to take into account uncertainties in proton-beam penetra pene tration tion.. For example, example, it is comm common on pra practice ctice to avoid, when possible using beam directions for which there would be a tight margin between the planni pla nning ng tar target get vo volum lume e (PT (PTV) V) an and d a sen sensit sitive ive structure lying distal to it ( e.g., the spinal cord). One center has a rule that at most one of three beams may allow a tight margin in depth.

6.2.2 6.2. 2

Beam-del Beam -deliver ivery y tech technique niques s

In pr proto oton-b n-beam eam the therap rapy y, a num number ber of dif differ ferent ent beam-shaping and delivery techniques can select be used, and these techniques techniqu es str strongly ongly affect the selection ion of bea beams ms an and d th their eir re resul sultin ting g dos dose e dis distri tribut bution ions. s. The pla planni nning ng sof softw twar are e mu must st the there refor fore e be abl able e to simula sim ulate te all tec techn hniqu iques es of pr proto oton-b n-beam eam del deliv ivery ery avai av aila labl ble e to th the e us user er.. Fo Forr ex exam ampl ple, e, it mi migh ghtt be required requ ired to comp compute ute the dose distributi distributions ons of the following:

6.2.3.2 6.2.3 .2

The fin The finit ite e pe pene netr trat atio ion n of pr prot oton onss al allo lows ws ‘d ‘dis ista tall blocking’—the analogy in depth of the lateral blocking pr prov ovide ided d by an ape apertu rture. re. As a con conseq sequen uence, ce, with protons, one only needs to be concerned about entra ent rance nce tis tissue sues, s, ex excep ceptt wh when en th the e bea beam m abu abuts ts a critical sensitive structure. With photons, however, one needs to be concerned about both entrance and exit ex it ti tiss ssue uess in ch choo oosi sing ng a be beam am di dire rect ctio ion. n. As a result, a wider range of desirable beam directions is usually available to the planner of proton-beam therapy. In this connection, the use of non-coplanar beam bea m dir direct ection ions, s, suc such h as ve verte rtex x bea beams ms in tr trea eattments of the brain and base of skull, is feasible and often found to be quite advantageous. With photons, in part because of the need to employ near-opposing (or otherwise paired) beam directions, non-coplanar beam arrangements are unusual. ‘good’therapy beam directions more necessary in Picking proton-beam because ofisthe following: †

† †

  Sca Scatte ttere red d bea beams ms

†

 Scanned beams (continuo (continuous discrete) us or

†

 Wobbled beams (a special case of  beam scanning, using relatively wide finite pencil beams)

6.2.3 6.2 .3 6.2.3. 6.2 .3.1 1

Generall Gener ally y des design igned ed to pr produ oduce ce a near-uniform near-uni form dose distribu distribution tion within the target volume for each beam Can either produce a near-uniform dose distribution more usually, highly non-uniform non-unifo rmor, dose distribution distribu tionawithin the target volume for each beam—and are thus suitable for use in intensitymodulated proton therapy (IMPT) Generally Genera lly producin producing g a near-un near-uniform iform dose within the target volume for each beam

Single Sin gle bea beams ms Invers Inv ersee beam beam desi design gn

 A fundamenta fundamentall difference between planning with

Selection Selecti on of beam direc directions tions

†

  the the desir desire e to avo avoid, id, if possi possible, ble, beam dir directio ections ns that pass through complex or high- Z  heterogeneities, or that lie tangent to a tissue–air interface (see Section 6.5.1). If the latter is una unavoid voidable, able, the use of a few beams only slightly separated in angl an gle e ca can n mi miti tiga gate te th the e do dose se pe pert rtur urba bati tion onss (Goitein, 1977);   the desire to angle a beam so as to achieve the maximum maxim um spa spatial tial separ separati ation on betw between een the PTV  and distal critical OARs;   the desire, due to lack of skin sparing, to avoid

superficial or shallow sensitive structures. The consequence is that, in contrast with IMRT with x rays (IMXT), where one is often able to use equall equ ally-s y-spa paced ced ang angles les for th the e bea beams, ms, one wou would ld select the angles more judiciously for protons. 6.2. 6. 2.3. 3.3 3

Thee PT Th PTV  V 

The PTV is a volume that is based on an underlyin ly ing g cl clin inic ical al ta targ rget et vol olum ume e (C (CTV TV)) to wh whic ich h margin mar ginss ar are e ad added ded to ac accou count nt for int intern ernal al an and d external uncertainties in the location of the CTV  relative rela tive to the rad radiat iation ion beam beam(s). (s). With pro proton tons, s, differe diff erent nt marg margins ins are gene generall rally y req require uired d in the

protons proto ns and ph photo otons ns is tha that, t, in cur curren rentt pr prac actic tice, e, proton prot on plan planning ning has aspec aspects ts tha thatt are   inverse. For exam ex ampl ple, e, on th the e ba basi siss of th the e kn know owle ledg dge e of th the e targ ta rget et vo volu lume me an and d no norm rmal al an anat atom omy y, us usua uall lly y a one-pass calculation is made of the beam settings ( e.g., max maximu imum m ra range nge)) and of the real or vir virtu tual al compen com pensa sator tor des design ign (wh (which ich pr prov ovide ide the des desir ired ed three-d thr ee-dimens imensional ional shap shape e of the dose dis distribu tribution tion,, both laterally and in depth).

depth dim depth dimens ension ion tha than n in the la later teral al dir direct ection ions. s. The consequence is that, if the PTV is to be used to es esta tabli blish sh th the e dos dose e mar margin ginss abo about ut th the e CT CTV V in all di dire recti ctions ons,, a dif differ ferent ent PT PTV V wou would ld ha have ve to be des esig ign ned for ea eacch be bea am dir irec ecttion on.. In som ome e centers, cent ers, ther therefor efore, e, the beam is desi designed gned relative relative to th the e CT CTV V an and d th the e im impl plem emen enta tati tion on of do dose se margins is built into the beam-design algorithm. However, it is recommended that a PTV always be 96

 

TREATMENT TREA TMENT PLANNING

de deli line neat ated ed as it is ne need eded ed fo forr do dose se reco record rdin ing g purposes. purp oses. These mat matters ters are disc discusse ussed d in detail detail in Section 5.1.4.4.

6.2.3. 6.2 .3.4 4

The pro proton ton RBE

 As discussed in Section 2, the RBE-weig RBE-weighted hted absorb abs orbed ed dos dose e sho should uld be use used d in tr trea eatm tment ent pla plannning ni ng.. If If,, as is us usua uall lly y th the e ca case se,, th the e tr trea eatm tmen enttplanning plan ning program program does not tak take e into account account the distinction between absorbed dose and RBE-weighted absorbed dose, the planner must do so. Unless otherwise stated, all references here to dose refer to RBE-weighted absorbed dose.

6.2.3.5 6.2.3 .5

†

Repain Rep ainting ting

Motio Mo tion n of th the e pa patie tient’ nt’ss int intern ernal al org organs ans and tim time e  variation of the beam delivery in IMPT can give rise to local dose fluctuations (dose mottle) due to the so-ca so-called lled inte interpla rplay y effe effects cts (see Section 7.6.3 7.6.3). ). Beam Be am ga gati ting ng on th the e re resp spir irat ator ory y cy cycl cle e or tu tumo morr tracking (see Section 7.5) can reduce this effect, but will not eliminate it. The solution is to repaint the target applying the same beam multiple times over many man y re respi spira ratio tion n cyc cycles les whi while le re reduc ducing ing th the e dos dose e perr ap pe appl plic icat atio ion n pr prop opor orti tion onat atel ely y (s (see ee Se Sect ctio ion n 7.6.3.2). 7.6.3 .2). This is a techn technicall ically y deman demanding ding requirerequirement, but is considered a necessity wherever organ moti mo tion on is ap appr prec ecia iabl ble e ( e.g., in the thorax and abdomen).

The design of beam-modif beam-modifying ying devices

The design and pla placemen cementt of the aperture aperture (bloc (block) k) and comp compensa ensator tor is mor more e comp complica licated ted for pro protons tons than for photons. First, it must be ascertained that the design of these devices is achievable in practice ( e.g., that there is not too abrupt a change in compensator thickness, nor too small a feature in the aperture periphery). Secondly, particular attention must be paid to the distance of closest approach of  the aperture aperture and the comp compensa ensator tor to the patient, patient, because of the following: †

6.2.3. 6.2 .3.6 6

  an ai airr ga gap p be betw twee een n th the e co comp mpen ensa sato torr (or (or an any y upst up stre ream am de degr grad ader er su such ch as a ra rang nge e sh shif ifte ter) r) increases thereso beam penumbra and leadsfor to less satisfa sat isfactory ctory resolution lution in comp compensa ensating ting fine heterogene heter ogeneities. ities. Thus, this distance distance need needss to be minimi min imized zed,, ev even en to th the e poi point nt of des design igning ing the compensator so that its downstream face can be placed pla ced in con conta tact ct wit with h th the e pa patie tient. nt. Ho Howe weve verr, mechanical interferences between the nozzle and the th e pa patie tient nt can for force ce und undesi esirab rably ly big air gap gaps, s, even to the point of making some beam directions infeasible.   fo forr re reas ason onss of ed edge ge sh shar arpn pnes ess, s, es espe peci cial ally ly in scatter scat tered ed beam beamss usin using g dou double ble scat scatteri tering ng tha thatt

6.2.3. 6.2 .3.7 7

Dose alg algorit orithms hms

The do The dose se al algo gori rith thms ms in pr prot oton on-b -bea eam m th ther erap apy y requ re quir ire e mo more re de deta tail iled ed kn know owle ledg dge e of th the e be beam am-delivery deliv ery techn techniques iques than is gene generally rally required required in photon therapy, with the exception of multi-leaf collima li mato torr se sett ttin ings gs in IM IMXT XT, wh whic ich h sh shar are e th this is problem. Dose computations also tend to be more demanding of computer performance, especially for IMPT wher wh ere e th ther ere e is ab abou outt a 30 30-f -fol old d in incr crea ease se in th the e numb nu mber er of pe penc ncil il-b -bea eam m wei eigh ghts ts th that at mu must st be adjusted and optimized.

6.2. 6. 2.4 4 6.2.4. 6.2 .4.1 1

Plan Pl ans s Number Numb er of bea beams ms

While, generally, several beams are employed in a patient’s treatment, a single proton beam can constitu st itute te a sa satis tisfa facto ctory ry tr trea eatm tment ent pla plan n for a dee deeppseated sea ted target ( e.g., a para-vertebral sarcoma) as it can give a lower entrance than target dose and no exit dose. How However ever,, for a rela relative tively ly shall shallowow-lying lying site sit e suc such h as a sa sacr cral al cho chordo rdoma ma tr trea eated ted from th the e

leads to a large effective source size, one wants the aperture as close to the patient as possible.

posterior poster ior,, th the e ski skin n dos dose e due to pr proto otons ns wou would ld be  very close to 100 percent and hence, some photon

 As with the compensator, mechanical interference en cess ca can n pr prev even entt th this is.. On th the e ot othe herr ha hand nd,, ther th ere e is a sm small all bu butt not negligib negligible le bea beam m con con-tamin ta minat ation ion th that at com comes es fr from om pr prot otons ons st strik riking ing the aperture on or near its edge, leading to a source sour ce of low low-en -energy ergy protons, protons, an effe effect ct not too dissimil diss imilar ar to the elec electro tron n cont contamin aminati ation on whic which h comes com es fr from om a blo blocki cking ng tr tray ay in a ph photo oton n bea beam. m. The Th e ape apertu rture re nee needs ds to be su suffic fficien iently tly dis dista tant nt from fr om th the e pa patie tient nt’s ’s ski skin n su surfa rface ce to all allow ow the these se low-energy protons to diffuse out.

component is often employed. Typica ypically lly,, few fewer er prot proton on than phot photon on beams are needed nee ded to ac achie hieve ve a sa satis tisfa facto ctory ry tr trea eatm tment ent pla plan n (Lomax  et al., 2004). Also, whereas photons tend to necessita neces sitate te nearnear-oppo opposed sed beam beamss to comp compensa ensate te for the th e ex expon ponent ential ial dos dose e fal fall-o l-off, ff, or a pa pair ir of we wedge dged d fields fiel ds som some e 90   apar apart, t, it is no nott ne nece cess ssar ary y to us use e widely separated proton beams; they need only sufficie fic ient nt an angu gula larr se sepa para rati tion on th that at no ap appr prec ecia iabl ble e overlap of the beams on the skin occurs (Rutz and Lomax, 2005). 8

97  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

6.2.4.2

Intensity-modulated Intensity -modulated proton therapy (IMPT)

Proton therapy has long used an elementary form of IMRT through the use of beam patching. A pair of beams is designed, with one beam ‘filling in’ a portio por tion n of th the e tar target get volume volume and the oth other er bea beam m intent int ention ionall ally y blo block cked ed out out.. Thi Thiss pr proce ocess ss is eas easy y to conceive, and can be done manually or using computer tools. There is no analogy of this technique in conventi conv entional onal phot photon-t on-thera herapy py plan planning, ning, exce except pt in fully automated IMXT. IMPT is a completely natural extension of broadbeam bea m pr proto oton n the thera rapy py,, jus justt as IM IMXT XT is a na natu tura rall extension of conventional three-dimensional conformal radiation therapy (3DCRT). IMPT is computationally more demanding than IMXT, as one has to consid con sider er a who whole le add additi itiona onall dim dimens ension ion of ra range nge  variability  variability, identical. , 6.2.4. 6.2 .4.3 3

but

otherwise

the

principles

are

Imaging Ima ging

While all modalities of radiation therapy make use of mo mode dern rn im imag agin ing g te tech chni niqu ques es,, hi hissto tori rica call lly y proton-b prot on-beam eam ther therapy apy has part particula icularly rly emph emphasize asized d the use of imaging. This is for three main reasons: †

†

†

  prot proton on ther therapy apy,, like mode modern rn conf conformal ormal phot photon on therap the rapy y, off offers ers add additi itiona onall pos possib sibilit ilities ies for con con-formal for mality ity, the del deline ineat ation ion of tar target get vol volume umess in three-dimensional space is particularly crucial;   heterogeneities; CT imaging imaging is gener generally ally essential essential for mana managing ging   proton-beam therapy tends to base patient alignment on bony anatomy, and this must be located by imaging.

In the cou course rse of pla planni nning ng a pa patie tient’ nt’ss th thera erapy py,, a number of different imaging procedures are likely to be employed, both for diagnostic purposes and,

In all forms of therapy, in order to transfer the plan to the patient, the positioning of the patient must be the same for the plan design and the treatment. In proton-beam therapy, owing largely to the tend te nden ency cy to us use e bo bony ny la land ndma mark rkss or ph phys ysic icia ianninsert ins erted ed fid fiduci ucial al mar marke kers rs for pa patie tient nt (an (and d hen hence ce CTV) CT V) lo loca cali liza zati tion on (s (see ee Se Sect ctio ion n 7. 7.3. 3.2) 2),, DR DRRs Rs (digitally-reconstructed (digitally-recons tructed radiographs) are generally need ne eded ed so th that at th the e se setu tup p ra radi diog ogra raph phss ob obta tain ined ed from a pair of x-ray tubes located in the treatment room can be verified. 6.2.4.4 Posit 6.2.4.4 Positioning ioning accura accuracy cy,, immobiliza immobilization, tion, and localization

Proton-beam Proton-be am ther therapy apy gener generally ally emph emphasiz asizes es accu accu-racy ra cy of bea beam m pla placem cement ent to a gr grea eater ter extent extent tha than n con conven ventio tional nal ofpho photon ton thera the py, , oton althou alt gh modern mod tech te chni niqu ques es conf co nfor orma mal lrapy phot ph onhough ther th erap apy y ern are ar e tending toward achieving comparable accuracy. The need nee d for accura accuracy cy der derive ivess fr from om the eff effort ort to tak take e full advantage of the superior conformality of dose possible with protons, and the need for good registrat tr ation ion bet betwe ween en the re real al or vir virtua tuall com compen pensa sator tor and an d any het eter erog ogen enei eitty wit ith hin the patie ient nt.. Placem Pla cement ent ac accur curac acies ies of fr from om 2 to 1 mm, or eve even n less, les s, are rou routin tinely ely ac achie hieved ved.. It sho should uld be emp emphahasized siz ed th that at th this is ac accur curac acy y oft often en has less to do wit with h target tar get-v -volu olume me con confor forma matio tion n (th (the e tar target get vo volum lumes es often cannot be defined so precisely) than with the confor con formal mal av avoid oidanc ance e of nea nearby rby sen sensit sitive ive nor normal mal structur struc tures, es, the loc locat ation ion,, and the ex exten tentt of whi which ch can be more accurately determined. To achieve such accuracies, excellent immobilization and accurate localization of the patient relative to the treatment equipment is required. The latter is usually accomplished by the localization of bony landm lan dmark arkss as see seen n in ort orthog hogona onall ra radio diogra graphs phs of  diagnostic quality. Also, a strategy for dealing with the remaining uncertainties needs to be developed

with appropriate image registration, for volume of  interest (VOI) delineation. However, given the need to evaluate heterogeneities, one CT scan sequence, taken tak en with the immo immobiliz bilized ed pat patient ient in trea treatmen tmentt position, is usually required. This sequence is often

and implemented. These matters are discussed in Sections 7 and 8. 6.2.4.5 6.2.4 .5

planningproton CT . Because referred toused as the the planning CT is to  compute path lengths for use in dose calculations, the calibration of the CT unit un it is of gr grea eate terr im impo port rtan ance ce in pr prot oton on-b -bea eam m therapy than in photon-beam therapy. Because the presence of contrast material can lead to incorrect estimates of the proton path lengths, it is common to perform the planning CT study without contrast media. med ia. Al Alter terna nativ tively ely,, the Hou Hounsfi nsfield eld nu numbe mbers rs (Sect (S ection ion 6.4 6.4.6. .6.1) 1) in reg region ionss con contai tainin ning g con contr tras astt media might have to be altered to conform to the expected values in that region.

Uncertainty Uncert ainty analys analysis is

Th The e hasize anal an alys ysis ofchun unce cert rtai ntie iessgly hassinte ha tend nded edn-beam toeam be emphas emp ized dismu much more mor eaint stron st rongly proto pr oton-b ther th erap apy y th than an ha hass be been en do done ne in ph phot oton on-b -bea eam m ther th erap apy y. Th The e ma mana nage geme ment nt of un unce cert rtai aint nty y pe perrmeates the entire planning process in proton-beam therapy.  An example of the use of uncertainty analysis is in the design of compensators and the calculation of the max maximu imum m pr proto oton n ra range nge (en (energ ergy) y) for ea each ch beam be am.. Th Thes ese e ma matt tter erss ar are e di disc scus usse sed d in Se Sect ctio ion n 6.4.6.2. Another example is in the management of  abutt abu tting ing pr proto oton n bea beams, ms, suc such h as in the pa patch tching ing 98

 

TREATMENT TREA TMENT PLANNING

tec techn hniqu ique e des descri cribed bed in Sec Sectio tion n 6.2 6.2.4. .4.2. 2. Bec Becaus ause e there are uncertainties in the depth of penetration of proton beams (as well as lateral placement me nt), ), be beam am ‘f ‘fea eath ther erin ing’ g’ inistheir ofte of ten n em empl ploy oyed ed,, whereby wher eby the spat spatial ial location of the abutment abutment (the patch line) is varied in space every treatment day over a cycle of three to five fractions. Beam feathering is emp employ loyed ed for blu blurri rring ng the lateral lateral edges of  abutting photon beams too, but there is no analogy with the depth feathering of protons. 6.2.4.6 6.2.4 .6

Target volume size

Proton-b Proto n-beam eam the therap rapy y is som someti etimes mes inc incorr orrect ectly ly thought of as being useful mainly for small tumors. Experience suggests that the dose-sparing possible with protons is likely to be most valuable for large target tar get vo volum lumes es for whi which ch spa sparin ring g the re remai mainin ning g  volume is likely to be particularly valuable. In this cont co ntex ext, t, ‘l ‘lar arge ge’’ re refe fers rs to th the e si size ze of th the e ta targ rget et  volume relative to the body compartment of which the target volume is part. Planners must therefore be prepared to plan large as well as small volumes, and an d to de deve velo lop p te tech chni niqu ques es fo forr ti tiss ssue ue sp spar arin ing g in such instances. 6.2.5 6.2. 5

Quality Qua lity assur assurance ance

 Variations in dose with depth can be large for  Variations proto pr oton n bea beams, ms, and the there refor fore e qua qualit lity y ass assur uranc ance e tool to olss ar are e ge gene nerral ally ly mo more re co comp mple lex x th than an th thos ose e needed for photon beams. They have to confirm the machine mach ine perf performan ormance ce in thr three ee dime dimension nsions, s, rat rather her than th an in tw two o di dime mens nsio ions ns.. Es Espe peci cial ally ly in sc scan anne ned d beam be ams, s, fo forr wh whic ich h it ma may y ta take ke ma many ny mi minu nute tess to

the location of nearb nearby y unin uninvolv volved ed norm normal al tissu tissues. es. Such studies are the following: †

† † † †

  volu volumet metric ric CT (i.e., mult multiple iple thin slice slices)—w s)—with ith and/or without contrast;  magnetic resonance imaging (MRI);  positron emission tomography (PET);  ultrasound (for some specialized applications);  other studies appropriate to the situation.

However, to plan radiation therapy, it is importantt th an tha at th the e in info form rma ati tion on ab abou outt th the e pa pati tien ent’ t’ss anatomy ana tomy be spa spatiall tially y accu accurat rate e with respect respect to the way wa y th the e pa patie tient nt wil willl be pos positi itione oned d for the thera rapy py.. Forr pr Fo proto oton-b n-beam eam the thera rapy py,, the map mappin ping g of tis tissue sue densities densit ies mus mustt als also o be ac accur curat ate. e. In gen gener eral, al, th this is mean me anss th that at a pl plan anni ning ng CT [a se seri ries es of CT scans scans take ta ken n wi with th th the e pa pati tien entt in th the e im immo mobi bili liza zati tion on device(s) devic e(s) to be empl employed oyed,, pref preferab erably ly with without out contrast tra st medi medium] um] is need needed. ed. Info Informat rmation ion from other imaging studies may then be superimposed on the plann pla nning ing-CT -CT study study th throu rough gh som some e for form m of ima image ge fusion (either by automatic means, or by a trained observer). This information is further elaborated by deline del ineat ation ion of VO VOIs Is re rela latin ting g to bot both h tum tumor or an and d normal tissues. All together, these build up a threedimensional model of the patient, which is spatially accurat accu rate e and pro provides vides a suffi sufficient ciently ly accu accurat rate e map of the densities of all tissues that may potentially be traversed by protons. In order to maximize resolutio ut ion, n, th the e sc scan anne ner’s r’s fie field ld of vi view ew sh shou ould ld be th the e minimum mini mum possi possible ble tha thatt neve neverthe rtheless less encom encompass passes es any material that could potentially be traversed by the proton beams, including the skin surface where

deliver the entire dose, it is too time-consuming to determine a dose distribution by making measurements at discrete points, and it is not possible to make measurements with a moving detector. Some form of mult multi-di i-dimensi mensional onal detector (at leas leastt twodimens dim ension ional, al, and ide ideall ally y, thr threeee-dim dimens ension ional) al) is therefore required (see Sections 4 and 9). The pro process cess of plan planning ning proton-beam proton-beam therapy is

any beam might enter. In th the e del deline ineat ation ion of nor normal mal tis tissue sues, s, all OA OARs Rs that could potentially be even partially intersected by any radiation beam should be fully imaged and deli de line neat ated ed.. In ad addi diti tion on,, th the e vo volu lume me wi with thin in th the e pati pa tien entt th that at ex excl clud udes es an any y de deli line neat ated ed PR PRVs Vs (o (orr OARs, if PRVs are not delineated) and the PTV(s) should sho uld be ide ident ntifie ified d as the ‘re ‘remai mainin ning g vol volume ume at

no now w des descri cribed bed in gr grea eater ter de detai taill in the fol follow lowing ing sections.

risk’ (RVR). The RVR is of interest when evaluating plan pl ans, s, as it it, , to too, o, co cons nsti titu tute tess ir irra radi diat ated ed norm no rmal al tissue tissu e and should be exam examined ined for regi regions ons within it receiving undesirably large doses that might otherwise go undetected. In som some e spe specia ciall cas cases, es, such as the pla planni nning ng of  proton-beam therapy for ocular tumors (see Section 6.11.1) or the treatment of very superficial disease withou wit houtt sig signifi nifican cantt het hetero erogen geneit eities ies in the bea beam m path(s), a CT study may not be required. Nevertheless, an accurate three-dimensional model of the relevant portion of the patient’s anatomy is essential for planning.

6.3

THE PATIENT’S ANA ANATOMY  TOMY 

Even befor Even before e the deci decision sion to use rad radiati iation on ther therapy apy for all or pa part rt of a pa patie tient nt’s ’s treatm treatment ent has bee been n taken, a number of imaging studies will have been performed with primarily diagnostic intent. These images are vital, too, for the planning of the radiation treatment as they give essential information as to the anatomic site and extent of disease and 99  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

6.4 HET HETERO EROGENE GENEITIE ITIES S

is termed as ‘mass thickness’ or ‘areal density’ and

6.4.1 6.4. 1

ha has s nth the eint unit un itsstion of gram ams sbiliti per pe ressq squa uare cent ce ntim imet er. When Whe inter erac actio n gr proba pr obabil ities are ar ereex expr press essed edeter in. unit un itss of ma mass ss th thic ickn knes ess, s, ra rath ther er th than an ph phys ysic ical al length, dependence on density is largely removed. The mass thickness of an object composed of slabs of dif differ ferent ent ma mater terial ialss alo along ng dir direct ection ionss nor norma mall to the slabs is the sum of the mass thicknesses of the individual slabs.

Introduct Intr oduction ion

In hete heteroge rogeneous neous medi media, a, prot proton-be on-beam am dose dis distritributions are more complex, and often substantially more so, than in a homogeneous medium such as water wa ter.. Heter Heterogenei ogeneities ties caus cause e two main effe effects cts on protons as they penetrate the medium: alterations in the range and alterations in the extent of lateral scattering. In what follows, the dosimetric impacts of th these ese two eff effect ects, s, rel relat ative ive to wha whatt occ occurs urs in a homo ho moge gene neou ouss me medi dium um (u (usu sual ally ly wa wate ter) r),, an and d fo forr beams bea ms wit with h un unifo iformrm-int intens ensity ity spr spread ead-ou -outt Br Bragg agg peaks (SOBPs), are discussed. Given Give n the pre presence sence of heter heterogene ogeneities ities pro proximal ximal to, or within, the target volume, it is necessary to take them into account when designing the treatment beams. This could invo involve lve calculating calculating thei theirr influence and, with very few exceptions, the design of some form of beam modification to correct for the range modification produced by the heterogeneities. This is usually done by means of a real or virtual compensator (as discussed below). Three Thr ee situa situations tions are discu discussed ssed in the follo following wing sections: (i) a block of material of density different from fr om the re rest st of the med medium ium tha thatt int inter ercep cepts ts the enti en tire re be beam am;; (i (ii) i) a bl bloc ock k of ma mate teri rial al of de dens nsit ity y differ dif ferent ent fr from om the res restt of th the e med medium ium tha thatt int interercepts only part of the beam in the lateral direction; and (iii) complexly structured heterogeneities.

6.4.2. 6.4 .2.2 2

Energy Ene rgy loss

When pen When penetr etrat ating ing ma matte tterr, pr proto otons ns los lose e ene energy rgy larg la rgel ely y du due e to Co Coul ulom omb b in inte tera ract ctio ions ns wi with th th the e atomic electrons. For sections of material thin relative to the proton theenergy mass stopping  S r,  describes  / r  the range, average loss of a power, proton per unit mass thickness. It has a strong nonlinear dependen depe ndence ce on the prot proton on energ energy y, and is linea linearly rly dependent on the ratio   Z /  A. Thus, the energy loss of a proton beam of a given energy in a section of  mate ma teri rial al of a gi give ven n ma mass ss th thic ickn knes esss wi will ll be ve very ry simi si mila larr fo forr al alll ma mate teri rial alss wi with th th the e ex exce cept ptio ion n of  highly high ly hydr hydrogen ogenous ous mat materials erials (whe (where re   Z /  A . 0.5) and elements of very high  Z  (where  Z /  A , 0.5). The finite range of protons is due to their almost cont co ntin inuo uous us lo loss ss of en ener ergy gy as th they ey tr tra ave vers rse e th the e material. mat erial. This allo allows ws the comp computat utation ion of the continuous tinu ous slow slowinging-down down appr approxim oximatio ation n (see Sect Section ion 3.4.1) range of a proton of given energy by the integrat gr atio ion n of th the e re reci cipr proc ocal al of th the e st stop oppi ping ng po powe werr

6.4.2 6.4. 2

along its entire path. (In practice, there is a distribution of penetrations due to statistical fluctuations in the energy-loss process.) As the range is inversely related to the stopping power, the penetration of a pr proto oton n bea beam m of a giv given en ene energy rgy,, exp expre resse ssed d in terms of mass thickness, is nearly independent of  the th e ma mate teri rial al tr trav aver erse sed d (e (exc xcep eptt ne near ar th the e en end d of  range, where range straggling becomes important).

Intera Inte ractio ctions ns of proton protons s in matter matter

The ph The phys ysic icss of th the e in inte tera ract ctio ions ns of pr prot oton onss wi with th matter has been discussed in detail in ICRU (1998) and briefly in Section 3.4.1 of the present report. Here He re,, so some me ge gene nera rall fe feat atur ures es of th the e tw two o la larg rges estt effe ef fect ctss ar are e re reca call lled ed,, na name mely ly,, (i (i)) lo loss ss of en ener ergy gy thro th roug ugh h in inte tera ract ctio ion n of pr prot oton onss wi with th el elec ectr tron ons, s, which whi ch re resul sults ts in pr proto otons ns ha havin ving g a fini finite te ra range nge in 2 matter ; an and d (i (ii) i) mu mult ltip iple le sc scat atte teri ring ng of pr prot oton onss through thr ough Cou Coulomb lomb inter interact actions ions with nucle nuclei, i, which cause cau se a pr proto oton’ n’ss pa path th to dev devia iate te fr from om a st strai raight ght line. The description here is confined to protons in the th e th ther erap apeu euti ticc en ener ergy gy ra rang nge e fr from om ab abou outt 30 to 250 Me MeV V; at oth other er,, par partic ticula ularly rly low lower er,, ene energi rgies es some of the statements below no longer apply. 6.4.2. 6.4 .2.1 1

6.4.2.2.1 Water-e 6.4.2.2.1 ater-equivalen quivalentt density density..   To To ta take ke in into to acco ac coun untt th the e de depe pend nden ence ce of ra rang nge e an and d st stop oppi ping ng powe po werr on th the e co comp mposi ositi tion on of th the e ma mate teri rial al,, th the e concep con ceptt of ‘wa ‘water ter-eq -equiv uivale alent nt den densit sity’ y’ is use used d in heavy charged-particle radiation therapy. This can be defined as follows: let the thickness of a block of  some material of interest and the energy loss in the block blo ck be re repr prese esente nted d by   Dt   and   D E, respectively. Then, the water-equivalent density of the material of interest is equal to the density of a block of (fictional) material of the same elemental composition as water and of the same thickness,   Dt, that produce du cess th the e sa same me en energ ergy y lo loss ss,,   D E, of th the e pr prot oton onss passing through it. The waterwater-equivalent equivalent density density,,   r eq eq, can be estimated from range-energy tables by comparing the energy

Masss thi Mas thickn ckness ess

It is co comm mmon on to ch char arac acte teri rize ze a sa samp mple le of ho homo mo-geneou gen eouss ma mater terial ial by th the e pr produ oduct ct of the length length of  the sample multiplied by its density. This quantity 2

By ‘r ‘rang ange’ e’ is me meant ant the dep depth th to the distal distal 80 per percen centt dos dose e level,   d80, in ac accor cordan dance ce wi with th the no nomen mencla clatu ture re of Sec Sectio tion n 3.4.2.2.

100  

TREATMENT TREA TMENT PLANNING

los losss per unit mas masss thi thickn ckness ess ( i.e., the mass stopstopping pin g po powe wer) r) of th the e ma mater terial ial in qu ques estio tion n wit with h the energy loss per unit mass thickness of water,   r water water (at an energy in about the mid-range of the energies of therapeutic interest). Or, it can be obtained by me meas asur urin ing g th the e ch chan ange ge in re resi sidu dual al ra rang nge e in water,   D R   (a po posi siti tive ve va valu lue e of   D R   implie impliess an incre inc rease ase in ran range) ge),, of pr proto otons ns pas passin sing g thr throug ough h a water tank with and without a physical thickness, t, of th the e ma mate teri rial al in qu ques esti tion on in inse sert rted ed in into to th the e wate wa terr ta tank nk in th the e pa path th of th the e pr prot oton ons, s, in wh whic ich h case one has r eq  ¼   1 

  D R

   t

 r water  ¼ 1 

  D R ;

 

  ð6 1Þ

5 percent,  q   varies as 0

q 20

/

   r t  X 0

 r t

  Z 07  Z  A :

;

 

ð6 2Þ :

where   r t   is the mas masss th thick icknes nesss of the sca scatte tterin ring g material and  X 0  is the characteristic length, termed the rad radiati iation on lengt length h (Mu (Musta stafa fa and Jack Jackson, son, 1981; 22 Tsai, 1974) of the material (in g cm ). The exponent of   Z  was obtained empirically by a fit to tabulated radiation-length data. Equation (6.2) implies that, in terms of mass thickness, a high- Z  material scatters more strongly than a low- Z   material, and in terms of thickness, the dependence is more like

:

t

where the mass densities,   r , are assumed to be in units of g cm23 and   D R  and   t  are in the same units ( e.g., cm) cm).. Usi Using ng the con concep ceptt of wa water ter-eq -equiv uivale alent nt dens de nsit ity y, on one e ma may y, in pr prac acti tica call ca case ses, s, ma make ke th the e approximation that one can substitute for a given bulk sample of hete heteroge rogeneou neouss mate material rial of varyi varying ng chemical composition and mass density a geometrically identical sample that is considered to be composed of material of the same chemical composition as water, and in which the actual mass density is ever ev eryw ywhe here re re repl plac aced ed by th the e wa wate terr-eq equi uiva vale lent nt

a power of 1.7. (For this reason, high- Z  materials are preferentially used when it is desired to spread out a proton beam laterally.) 6.4.3 Bulk hete 6.4.3 heterog rogenei eneity ty inters intersecti ecting  ng  the full beam

The in The influ fluen ence ce of in inte terp rpos osin ing g a un unif ifor orm m sl slab ab of  material of composition other than that of the surrounding medium is entirely different for a proton beam than for a photon beam. x rays experience an inte in tens nsit ity y ch chan ange ge in th the e sh shad adow ow of su such ch a sl slab ab heterogeneity; protons suffer virtually no change in

intensity; only a change of penetration. This diffe-

density.. This appr density approxim oximatio ation n will gener generally ally allo allow w sufficiently accurate estimates of energy losses, and hence penetration depth, but it will not give correct results for other important effects such as multiple Coulomb scattering.

6.4.2.3 6.4.2 .3

rence is illustrated schematically in Fig. 1.6. That is, the eff effect ect of int interp erposi osing ng a sla slab b of ma mater terial ial of  compos com positi ition on oth other er th than an th that at of the sur surrou roundi nding ng medi me dium um is ma main inly ly to sh shor orte ten n or le leng ngth then en th the e proton pro ton depth– dose distribution distribution,, but not aff affect ect its shap sh ape e or th the e in inte tens nsit ity y in th the e hi high gh-d -dos ose e re regi gion on distal dis tal to the het heter eroge ogenei neity ty.. Ind Indeed eed,, plo plotte tted d as a functi fun ction on of mas masss thi thickn ckness ess alo along ng th the e bea beam m pa path, th, the dose distribution would be little affected by the interposition of the slab. This is the case whether the th e be beam am is ne near ar-m -mon onoe oene nerg rget etic ic or in invo volv lves es a mixture of energies as is the case, for example, in an SOBP. The change in the range,   D R, of a beam in such a

Multiple Coulo Coulomb mb scatt scattering ering

In addition to losing energy as they traverse matter thro th roug ugh h in inte tera ract ctio ions ns wi with th or orbi biti ting ng el elec ectr tron ons, s, protons prot ons also expe experienc rience e num numerou erouss Cou Coulomb lomb inte interractions with the charged nuclei of the atoms in the mater ma terial ial (se (see e Sec Sectio tion n 3.4 3.4.1) .1).. Ea Each ch of th these ese int intereractions results in a usually very small deflection of  the pro projecti jectile le pro proton. ton. How However ever,, thes these e inte interac raction tionss accumulate and result in the finite deflection of a proton from a straight path. An infinitesimal pencil beam of protons will be increasingly spread out in depth by multiple Coulomb scattering.  A near-monoenerget near-monoenergetic ic proton beam travers traversing ing a thickn thi ckness ess of ma mater terial ial sma small ll re relat lativ ive e to its ra range nge will be scattered with an approximately Gaussian distri dis tribut bution ion of ang angles les for wh which ich sig sigma ma (s (stan tandar dard d deviation) is termed as the characteristic scattering contr tras astt to the ene energy rgy-lo -loss ss pr proce ocess, ss, angle,   q0. In con the scatter scattering ing of a pr proto oton n bea beam m has a sig signifi nifican cantt depend dep endenc ence e on the che chemic mical al com compos positi ition on of the

situat situ atio ion n (m (mea easu sure red d in un unit itss of le leng ngth th an and d no nott water-equivalent density) is altered by an amount given by D R

¼

medium tðr eq    r slab eq   Þ

r water

;

 

ð6 3Þ :

where   t  is the physical thickness of the interposed slab slab,   r eq is the wa water ter-eq -equiv uivale alent nt den densit sity y of th the e eq medium interposed slab, and  r eq is the water-equivalent eq dens de nsit ity y of th the e su surr rrou ound ndin ing g me medi dium um.. Th The e ma mass ss densit den sity y of wa water ter,,   r water water, is taken to be unity when whe n den densit sities ies ar are e ex expr press essed ed in uni units ts of g cm23.

material. The relation is complicated but, to within

If th the e su surr rrou ound ndin ing g me medi dium um is wa wate terr, or ha hass a 101

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

wa wate terr-eq equi uiva vale lent nt de dens nsit ity y of 1, th this is re redu duce cess to Eq. (6.1). These relations hold equally when the interposed slab sl ab re repl plac aces es th the e en enti tire re su surr rrou ound ndin ing g me medi dium um.. They describe the change in range (relative to the range ra nge in wa water ter)) whe when n th the e (ho (homog mogeno enous) us) med medium ium hass a wa ha wate terr-eq equi uiva vale lent nt de dens nsit ity y di diff ffer eren entt fr from om water. The Th e pr prot oton on be beam am wi will ll al also so be af affe fect cted ed at it itss edges—i.e., the beam penumbra will be affected by the interposed material, as the penumbra is largely caused cau sed by ups upstr tream eam mul multip tiple le sca scatte tterin ring g and thi thiss scattering, as indicated above, is dependant on the

example, when a layer of tissue of only one-half the thickness of the tissue in a tissue–air interface is interp int erpose osed d abo above ve the int interf erface ace,, th the e dos dose e per pertu turral.., batio ba tion n is re reduc duced ed to    12 per percent cent (Goit (Goitein ein   et al 1978). If one side of the interface is not air, but rather the interface is between the two materials of differentt sca ren scatte tterin ring g po powe wers, rs, the then n th the e per pertur turba batio tion n of  fluence, and hence dose, is much reduced,   e.g., in the case of a bone–tissue interface, from 50 to    9 percent (Goitein, 1978b).

chemi chemical cal comp compositio n of the inte interpose rposed dare material. mat erial. However, whenosition the materials involved limited to those found in the human body, this is a small effect and is often ignored in practice.

In practice, the patient usually presents a complex patte pa ttern rn of het hetero erogen geneit eities ies;; thi thiss is per perhap hapss mos mostt extre ext reme me in the re regio gion n of th the e bas base e of skull where where protons may be directed along extended bone surfaces, or through a complex bone–tissue–air structure like the petrous ridge or paranasal sinuses. In cons co nseq eque uenc nce, e, a co comp mple lex x co comb mbin inat atio ion n of ra rang ngeepenetrat penet ration ion pert perturba urbation tionss and scat scatterin tering-in g-induced duced dose nonnon-unif uniformiti ormities es tak takes es pla place. ce. The resu results lts of 

6.4.4 Bulk hete 6.4.4 heterog rogenei eneity ty partia partially lly intersecting the beam

When a slab of material of a mass density different from that of the surrounding medium is interposed

6.4.5

Complexly Complex ly structur structured ed hetero heterogeneities geneities

such complex situations are very hard to calculate analytical analy tically ly,, alth although ough the prec precedin eding g discu discussion ssion of  bulk bul k het heter eroge ogenei neitie tiess giv gives es som some e ins insigh ightt int into o the

over on over only ly pa part rt of th the e be beam am cr cros osss-se sect ction ion,, th then en,, away aw ay fr from om the int interf erfac ace e bet betwe ween en th the e two med media, ia, the beam pen penetr etrat ation ion is alt altere ered d in the sha shado dow w of  the th e he hete tero roge gene neit ity y ju just st as fo forr th the e ca case se of a fu full lly y intersecting heterogeneity, and is unchanged in the regi gion on no nott sha had dowed by the he hetter ero oge gen nei eity ty. However, near and in the shadow of the interface regio re gion, n, an add additi itiona onall and qui quite te dif differ ferent ent eff effect ect tak ta kes pla lace ce be beca caus use e of the dif iffe ferren ence ce in the strengths of multiple scattering in the two adjacent material mat erials. s. Nam Namely ely,, a dose enhancement enhancement (hot spot) occu oc curs rs on the lo loww-d den ensi sity ty si sid de, an and d a dos ose e reduc re ductio tion n (co (cold ld spo spot) t) occ occurs urs on the hig highh-den densit sity y side (Goit (Goitein, ein, 1978 1978b; b; Goite Goitein in   et al., 1978). This is illus ill ustr trat ated ed in Fig Fig.. 6.1 6.1a a for the ex extre treme me cas case e of a para pa rall llel el be beam am of pr prot oton onss im impi ping ngin ing g on an ai air– r– tissue tissu e inter interface face.. Pro Protons tons reach point P from both the unscattered beam impinging on the air side (I), and from protons initially impinging on the tissue side (II), but scattered scattered lat lateral erally ly tow toward ard the point P. As a con conseq sequen uence, ce, poi point nt P re recei ceive vess a gr grea eater ter fluence, and hence higher dose, than if the tissue heterogene heter ogeneity ity were absen absent. t. Poi Point nt Q, on the other hand, receives no dose from protons impinging on the air side, as air scatters them hardly at all, and a reduced dose from protons initially impinging on the tissue side, because a portion is scattered out to the th e ai airr si side de.. As a co cons nseq eque uenc nce e of th thes ese e ef effe fect cts, s, dose perturbations can be as high as 50 percent as illustrated in Fig. 6.1b. The perturbation is substantially modified if the

Figure 6.1. Illustration Illustration of edge-scattering edge-scattering effect (Goitein, 1978b; repr re produ oduced ced wit with h permis permissi sion) on).. Schema Sch ematic tic re rende nderi ring ng of the

bea beam md has signifi nifican cant angula ang ular rlyin confus con ion, ,rial such suc woul wo uld be sig induce indu ced d tby over ov erly ing g fusion mate ma teri al.. hFas or

geometry (a) and calculated fluence distribution along the dotted line of the upper figure (b). See Section 6.4.4 for discussion.

102  

TREATMENT TREA TMENT PLANNING

ex exten tentt of the pos possib sible le dos dose e per pertur turba batio tions. ns. Mon Monte te Carlo Car lo cal calcul culat ation ionss (se (see e Sec Sectio tion n 6.5 6.5.2. .2.3) 3) ar are e pr preesently the only way to obtain a reliable estimate in the case of highly complex geometries. Figure 6.2 shows the degradation of the terminal reg egio ion n of an SOBP mea eassur ured ed aft fter er pas assa sage ge through: (i) a relatively homogeneous region in the brai br ain; n; an and d (i (ii) i) th thro roug ugh h hi high ghly ly he hete tero roge gene neou ouss regions of the base of the skull. These data demonstra st rate te ho how w th the e di dist stal al po port rtio ion n of th the e do dose se di dist stri ri-bution butio n can be subs substant tantially ially affected affected by a comp complex lex heterogeneity. The distal fall-off of the SOBP is not

6.4.6

Compensation Compensa tion for hetero heterogeneities geneities

6.4.6.1 Conv 6.4.6.1 Conversion ersion from CT Hounsfield Hounsfield number to water-equivalent density

The advent of CT was, and remains, critical for the development of proton-beam therapy as it provided for the firs firstt tim time e a spa spatia tially lly ac accur curat ate e map mappin ping g of  the patient’s anatomy together with a quantitative measu mea surem rement ent of tis tissue sue pr prope operti rties. es. All CT val values ues based on x-ray tomography are quoted in terms of  Hounsfield numbers,  H , and these measure a quantity related to the ratio of linear x-ray attenuation tissue

sim simply ply shi shifte fted dsteep in ra range nge..less Rathe Ra therr, its slo slope pe is sub sub-stantially less and regular.  An uncertainty analysis, as seen in Goitein (1985) (19 85) and in Sec Sectio tion n 8, can establis establish h the confidence den ce lim limits its for th the e dos dose e dis distri tribu butio tion. n. Fig Figur ure e 6.3 shows an example of the computed bounds for the penetration of a beam passing through the base of

coeffi coefficient s in the medi medium, um,   m mwatercients . Specifically,

, and in wa water ter,,

  mtissue  H   ¼ ¼ 1000  1 mwater





;

 

ð6 4Þ :

penetration of a beam passing through the base of  skull sku ll com compar pared ed wit with h the mea measur sured ed pen penetr etrat ation ion al.., 198 (Urie   et al 1986a) 6a).. Suc Such h an ana analy lysis sis all allow owss th the e design of beams that are assured (at a stated confidence level) to cover the target volume to full dose, with wi th th the e pr pric ice e be bein ing g th that at di dist stal al no norm rmal al ti tiss ssue uess receive a greater-than-desired dose.

The quantity   H  varies from   2 1000 for air to 0 for wate terr an and d po posi siti tive ve va valu lues es fo forr ma mate teri rial alss wi with th grea gr eate terr at atte tenu nuat atio ion n th than an wa wate terr. Li Line near ar xx-ra ray y attenua att enuation tion coeffi coefficient cientss are func functions tions of elect electron ron densit den sity y, at atomi omicc num number ber,,   Z, and ato tom mic ma mass ss number,   A, of the material. This implies that they

Figure 6.2. Measure Measurement ment of the distal portion of the dose distribution of a spread-out Bragg peak (Urie  et al., 1986a; reproduced with permission). Upper left: experimental setup with the proton beam directed from the right and depth–dose measurements made in a water tank placed distal to the water-filled skull. Upper right: magnified portion of a radiograph showing the three locations for which depth–dose curves are shown. Lower panel: depth–dose curves in the shadow of point A (left graph), point B (middle graph), and point C (right graph). Point A has relatively few heterogeneities in the proton-beam path; protons passing through point C go through highly heterogeneous material, and point B is intermediate. The dotted curves show the distal portion of the depth–dose curves with the skull removed (i.e., no heterogeneities).

103  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

Figure re 6.3. Calc Calcula ulated ted 90 per percent cent isodose isodose line distal distal to the skull of Fig. 6.2 sho showin wing g the nominal nominal and upper (maximum) (maximum) and low lower er Figu (minimum) bounds at the 85 percent confidence level (Urie   et al., 1986a; reproduced with permission). The dotted lines labeled A, B,

and C correspond to the points with the same labels as in Fig. 6.2. The range of uncertainty is much less for the point A, as one would expect, because the proton beam traverses a relatively homogeneous section.

have a di hav diff ffer eren entt de depe pend nden ence ce on th the e ch chem emic ical al composition of the medium than do proton-stopping powers. pow ers. The They y are also, at typical diagn diagnosti osticc x-ra x-ray y energies, fairly strong functions of the x-ray energy or energy spectrum, which means that the conversion si on fr from om th the e Ho Houn unsfi sfiel eld d nu numb mber er to th the e wa wate terrequiva equ ivalen lentt den densit sity y mus mustt be es estab tablis lished hed for ea each ch

6.4.6. 6.4. 6.1. 1.2 2 St Stoi oich chiom iometr etric ic meth me thod od..   A method proposed by Schneider  et al.  (1996) is based on the chemic che mical al com compos positi ition on of th the e tes testt ma mater terial ials. s. The stop st oppi ping ng po powe werr, an and d he henc nce e wa wate terr-eq equi uiva vale lent nt density, is determined using a simplified version of  the Bethe–Bloch formula. The Hounsfield number is assumed to be represented by an equation with

particular scanner and radiographic technique. The Th e the theore oretic tical al re relat lation ionshi ship p bet betwe ween en pr proto otonnstop st oppi ping ng po powe wers rs an and d li line near ar xx-ra ray y at atte tenu nuat atio ion n coefficient coeffi cientss is comp complica licated ted and, in pra practice ctice,, is not evaluate evalu ated d anal analytica ytically lly.. Rat Rather her,, the rela relation tionship ship between the Hounsfield number and the mass stopping power of protons is derived from experiment, princi pri ncipal pally ly usi using ng one of two met method hods: s: dir direct ect-fit -fit method and stoichiometric method.

three terms corre three correspon sponding ding to phot photoelect oelectric ric effec effect, t, coherent scattering, and Compton scattering. Each term has a different   Z  dependence and includes a multiplicative constant. The goal of the calibration is to fit the equatio equation n to th the e mea measur sureme ement ntss of the Hounsfi Hou nsfield eld num number bers, s, mad made e in th the e CT scanner scanner of  concern, for a large variety of tissue-equivalent test mater ma terial ialss of kno known wn che chemic mical al com compos positio ition n and to deduce the values of the three constants from the fit.. Gi fit Give ven n th the e co cons nsta tant nts, s, it is th then en po poss ssib ible le to pred pr edic ictt th the e Ho Houn unsfi sfiel eld d nu numb mber er fo forr an any y ot othe herr material, including tissue, of known chemical constitution. This is done for a wide variety of actual tissue tis suess and the res result ultss plo plotte tted d on a cha chart rt of the

6.4.6.1.1 6.4.6. 1.1 Dir Direct ect-fit -fit met method hod..   In th the e si simp mple lerr an and d more mor e wid widely ely use used d app approa roach, ch, mea measur sureme ements nts are taken on each of a wide variety of tissue-equivalent mater ma terial ials, s, and pos possib sibly ly ac actu tual al tis tissue sue sam sample ples, s, of  two quantities, namely: the Hounsfield number (by meas me asur urem emen entt in a CT sc scan anne ner) r) an and d th the e wa wate terrequiva equ ivalen lentt den densit sity y (by mea measur sureme ement nt in a pr proto oton n beam be am). ). Wh When en pl plot otte ted d ag agai ains nstt on one e an anot othe herr on a scatter plot, these data are approximately bi-linear and an d ma may y be fit by a pa pair ir of st stra raig ight ht li line ness (C (Che hen n  et al., 1979). Typically, one line covers the interval from   2 10 1000 00 to   þ 50 Ho Houn unsfi sfiel eld d nu numb mber erss an and d the second line covers Houn Hounsfield sfield numbers above þ 50. Th The e for former mer range range co cove vers rs the soft tis tissue sues; s; in the la latte tterr ra range nge Hou Hounsfi nsfield eld val values ues   . 50 may be assumed to be caused by a mixture of compact bone and sof softt tis tissue sue.. Th This is met method hod yie yields lds a cal calibr ibrat ation ion curve cur ve such as that shown shown in Fig. 6.4 6.4a. a. A sim simila ilarr appr ap proa oach ch,, bu butt us usin ing g a se sett of th thre ree e co conn nnec ecte ted d straight str aight lines, has also been follo followed wed (Kanematsu (Kanematsu  et al., 2003).

water-eq water -equiv uivale alent nt den densit sity y ve versu rsuss th the e Hou Hounsfi nsfield eld numb nu mber er.. Th Thes ese e da data ta po poin ints ts ar are e th then en fit fitte ted d by a series of straight lines, each of which extends over a limited range of Hounsfield numbers, and these lines lin es pr prov ovide ide the fin final al con conve versi rsion. on. Su Such ch a cha chart rt is sh show own n in Fi Fig. g. 6. 6.4b 4b.. Th This is ap appr proa oach ch ha hass be been en  verified for a number of biological tissue samples (Schaffner and Pedroni, 1998). 6.4.6.1.3 Con 6.4.6.1.3 Confirma firmation tion of calibr calibrati ation. on. Confirmation of the calibration process can be obtained by means of pr proto oton n ra radio diogra graphy phy, in whi which ch th the e int integr egrat ated ed water-eq wat er-equivale uivalent nt dens density ity alon along g any pat path h thr through ough the patient can be measured, or by proton CT, in which case the local water-equivalent density at all points within the imaged volume can be measured al.., (Schne (Sc hneide iderr and Pe Pedr droni oni,, 199 1995; 5; Sch Schnei neider der   et al 104

 

TREATMENT TREATMENT PLANNING

Figure 6.4. Conver Conversion sion of Hounsfie Hounsfield ld numbers to water water-equivalen -equivalentt density. density. (a) Direct-fi Direct-fitt method (Chen  et al., 1979). Calibrations of two different scanners are shown. Here the results are fit with a polynomial curve. Usually, two straight lines are fit, one line to the region between   2 1000 and   þ 50 Hounsfield numbers (in this figure the ‘CT number’ is equal to half of a Hounsfield number), and the other line to the region of higher Hounsfield numbers. (b) Calibration curve using the stoichiometric method (Schaffner and Pedroni, 1998).  A magnified portion of the graph in the region of Hounsfield numbers close to zero (scaled to 1000 in this figure) is shown in the inset. Five line segments are used to fit the data. Reproduced with permission.

6.4.6.2 6.4.6 .2

2005). In the latter case, the data could be used as input to the planning process. In the former case, the data are restricted to confirmation of the planning process.

Design of compen compensator satorss

For the high-dose For high-dose volume volume to con confor form m to the distal distal surface of the target over the entire field, some way of tailoring the proton penetration over the field area is required. The required penetration at any point in the th e fie field ld wi will ll de depe pend nd on th the e loc locat atio ion n of th the e ski skin n surface, the distal target surface, and the intervening heterogeneit heter ogeneities. ies. For scattered scattered beams, the penetration penetration is tailored by interposing a   compensator, a piece of  low- Z  material (often PMMA), the thickness of which  varies across across the field, in the path of the beam. In th the e si simp mple lest st ap appr proa oach ch,, on one e co comp mput utes es th the e 3 radiological path lengths of protons over the entire field fie ld an and d th then en de desi sign gnss a va vari riab able le-t -thi hick ckne ness ss

6.4.6.1.4 Accur 6.4.6.1.4 Accuracy acy..   Typically Typically,, the conversion from the th e Ho Houn unsfi sfiel eld d nu numb mber er to wate terr-eq equi uiva vale lent nt density dens ity perm permits its    1– 2 pe perc rcen entt acc ccur ura acy (1 SD SD)) in ca calc lcul ula atin ing g th the e ef effe fect ctiv ive e ran ange ge of nea earrmonoen mon oenerg ergeti eticc pr proto otons ns wit within hin the pa patie tient. nt. Thi Thiss leads lea ds to a 1– 2 mm unc uncert ertain ainty ty in com comput puting ing the effective range of a proton that penetrates some 10 cm into the patient. A more complete account may be fo foun und d in Sc Scha haff ffne nerr an and d Ped edro roni ni (1 (199 998) 8) an and d Schneider   et al.   (1996). (1996). In addi addition, tion, uncer uncertaint tainties ies in th the e CT mea measur sureme ements nts th thems emselv elves, es, inc includ luding ing effects due to beam-hardening effects, can add an additiona addi tionall 1 – 2 perc percent ent unce uncertain rtainty ty (Co (Const nstanti antinou nou  et al., 1992).

3

Radiologi Radi ological cal pat path h leng length th is the inte integra grall alon along g a str straigh aightt line with res espe pecct to th the e pa path th le len ngt gth h of any mate teri rial al of  waterwa ter-equi equivalen valentt dens density ity betw between een the sour source ce and the dis distal tal point of interest.

105  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

compen pensa sator tor tha thatt is th thin in whe where re the ra radio diologi logical cal com path length is large, and thick where it is small, in such a mann manner er tha thatt the radiological radiological pat path h leng lengths ths along alo ng all ra rays ys bet betwe ween en the sou source rce and the dis distal tal target tar get sur surfa face, ce, inc includ luding ing the con contri tribut bution ion of the compensator, are equal. Figure 6.5a shows such a

setup uncer setup uncertain tainties, ties, pat patient ient moti motion, on, ran range ge uncer uncer-taint tai nties ies du due e to CT calibra calibratio tion n err errors ors and un uncer cer-tainty in the Hounsfield number to water-equivalent density calibration,   etc. The extent of this margin mustt be bas mus based, ed, just as for lateral lateral mar margin gins, s, on an analys ana lysis is of th these ese unc uncert ertain aintie ties. s. Wh When en a for formal mal

compensator (see also Section 3.2.1.1 and Fig. 3.3). The Th e ab abo ove ap appr proa oach ch is de defic ficie ient nt in se sev ver eral al regar re gards. ds. Fir First st,, it doe doess not allow for pos possib sible le mis mis-regi re gist stra rati tion on be betw twee een n th the e co comp mpen ensa sato torr an and d th the e patient,   e.g., du due e to pa pati tien entt or or orga gan n mo moti tion on or setup setu p error error.. Seco Secondly ndly,, it assu assumes mes tha thatt the pro protons tons trav tr avel el alo along ng st stra raigh ightt lin lines es fan fannin ning g out fr from om th the e source of protons. That is, it ignores multiple scattering ter ing,, wh which ich cr creat eates, es, in eff effect ect,, a dis distri tribut bution ion of  possible ranges for protons initially directed along a given line. A solution has been proposed to allow for these two effects (Urie   et al., 1984), namely, to exp xpan and d or op open en up th the e co comp mpen ensa sato torr in su such ch a manner that its thickness at any point in the field is the least thickness of the un-expanded compensator sa tor wit within hin som some e defi defined ned ra radiu dius— s—whi which ch is typ typiically at least a few millimeters. This is illustrated

uncert unce rtai aint nty y an anal alys ysis is ha hass be been en pe perf rfor orme med d (s (see ee Section 8), the prescription is straightforward: one sets the maximum energy so that, for example, the 90 per percen centt iso isodos dose e sur surfa face ce jus justt co cove vers rs th the e tar target get  volume.

in Fig. Fi g. is 6.5b 6. 5b.. Th The cons co nseq eque nce e of coverage, the th e ex expa pans nsio ion n process that thee target isuenc assured even in the face of setup error, patient motion, and scattering effects. However, on the downside, the beam will tend to overshoot in some places, thus increasing the dose to distal normal tissues.

target boundary and moving radially outwards, the compensator is not allowed to change thickness by more than a defined amount, typically, so that no gradie gra dient nt in th the e com compen pensat sator or is st steep eeper er tha than n 60 (with respect to the direction of the beam axis) in that region. In general, the beam extends somewhat beyond the projected target volume (often taken to be the CTV) CT V) in or orde derr to pr prov ovid ide e a la late tera rall ma marg rgin in.. Th The e method meth od of desig designing ning comp compensa ensators tors outl outlined ined abo above ve does not work outside the projected target volume. There, in practice, one sets the compensator thickness equal to that of the closest point for which the compensator thickness can be calculated. This step is taken after the tapering process has been done.

6.4.6. 6.4. 6.2 2.2 Com ompe pens nsa ato torr de desi sign gn cl clos osee to to,, an and d outside, the proj projected ected target boundary. boundary.   When When the target tar get vo volum lume e is re relat lativ ively ely sph spheri erical cal,, the abo above ve design approach tends to lead to rather steep gradients in the thickness of the compensator near the edge of the projected target boundary because the target surface also has a steep gradient. This also lead le adss to co comp mpen ensa sato tors rs be bein ing g qu quit ite e th thic ick. k. On One e method used to reduce these effects is to taper the comp co mpen ensa sato torr in a re regi gion on ne near ar it itss ed edge ge.. Th That at is is,, starting at, say, 1 cm from the edge of the projected

8

6.4.6.2.1 6.4.6. 2.1 Cho Choice ice of rang range. e.   The extent in depth of  the th e hi hig ghh-d dos ose e reg egio ion n is det eter erm min ine ed by the maximum proton-beam energy. That energy could, in pr prin inci cipl ple, e, be se sett so th tha at th the e di dissta tal, l, sa say y, 90 percent isodose surface just conforms to the distal target surface. However, such a prescription leaves no room for uncer uncertaint tainty y. Jus Justt as is done laterally laterally,, one needs to allow some distal margin to allow for

Figure 6.5. Desi Figure Design gn of a comp compensa ensator tor (here labeled ‘bolus’) based on com computi puting ng the radiologic radiological al pat path h leng length th of pro protons tons along ra rays ys emanati eman ating ng fro from m the source source and stopping stopping at the dis distal tal surface surface of the targe targett (Uri (Urie e   et al., 1984; reproduced with permission). (a) The calculation ignoring multiple-scattering effects and possible misregistration due to setup error and patient motion. (b) The technique of  expanding the compensator so that the thickness of the new compensator at any point is replaced by the least thickness of the original compensator within a prescribed radius of the point.

106  

TREATMENT TREA TMENT PLANNING

6.4. 4.6. 6.2. 2.3 3 Re Real al an and d vir virtu tual al co comp mpen ensa sato tors rs..   The 6. descripti descr iption on of comp compensa ensator tor design has, up to this

be termed a  bolus-compensator.) Since the compensator induces multiple scattering of the protons in

point, assumed that the compensator is a physical device dev ice,, suc such h as tha thatt dep depict icted ed in Fig Fig.. 6.5 6.5.. Th This is is, indeed, the case for scattered and wobbled proton beam be ams. s. Wh When en a sc scan anne ned d pe penc ncil il be beam am is us used ed to form the trea treatmen tmentt beam beam,, a phy physical sical compensator compensator could still be used. However, in practice, the variation at ion in pr proto oton-b n-beam eam pen penetr etrat ation ion is ach achiev ieved ed by upstream changes in the pencil-beam energies and henc he nce e pe pene netr trat atio ions ns.. Th The e co comp mpen ensa sato torr de desi sign gn described above should then be considered as a prescri sc ript ptio ion n of th the e wa way y in wh whic ich h th the e pe penc ncil il-b -bea eam m ener en ergy gy ne need edss to be mo modi difie fied d ov over er th the e fie field ld ar area ea.. That is, the design is that of a virtual compensator, rather than an actual physical device. One advantage of virtual compensation using a scanned beam is that one has improved control over the proximal exte ex tent nt of th the e tr trea eate ted d vo volu lume me,, an and d ca can n th ther ereb eby y better spare proximal normal tissues (Goitein and Chen, 1983).

the be beam am,, the air gap wil illl res esu ult in bo botth an increase of the beam penumbra and a blurring of  the compensator’s fine structure within the patient. The Th e la larg rger er th the e ai airr ga gap, p, th the e gr grea eate terr ar are e th thes ese e effe ef fect cts. s. Th This is pr prob oble lem m ha hass be been en qu quan anti tita tati tive vely ly explored (Sisterson  et al., 1989; Urie  et al., 1986b). The amount of comp compensa ensator tor expansion expansion shou should ld be tailored to allow for beam spreading with depth, as well as misregistration effects. Obviously, one part of the strategy is to have the compensator induce as little scattering as possible by fabricating it from a low- Z  material such as plastic. One impo important rtant advantage advantage of penci pencil-bea l-beam m scan scan-ning is that the air gap is less important because the th e cha change nge in bea beam m pen penetr etrat ation ion ma may y be ind induce uced d not by a phy physical sical compensator compensator (whi (which ch intr introduc oduces es sca sc att tter er), ), bu butt up upst stre ream am by ch chan ange gess in th the e be beam am energy ene rgy, whi which ch pr prov ovide ide vir virtua tuall com compen pensat sation ion as mentioned above.

6.4. 6. 4.6. 6.2. 2.4 4 Th Thee ef effe fect ct of an air gap be betw twee een n co comm pensator and patient.   In th the e des design ign pr proce ocess ss re reppresented in Fig. 6.5, no account is made of the gap between betw een a phy physical sical compensator compensator and the pat patient. ient. Indeed, the algorithm for designing a compensator implies that the compensator is in contact with the patient surface. However, in practice, an air gap is left between the compensator and the patient’s skin surface. (Schemes have been proposed to shape the downstream face of compensators so that they can conform to, and be placed in juxtaposition with, the patient’ pat ient’ss skin surf surface. ace. Such a comp compensa ensator tor would

6.4.6.2.5 High-Z heterog 6.4.6.2.5 heterogeneitie eneities. s.   The appr approach oaches es to the comp compensa ensation tion for hete heteroge rogeneitie neitiess descr described ibed above work well for most tissues normally found in the hum human an body body.. Not infr infreque equently ntly,, how however ever,, one has to contend with man-made structures such as titanium rods used for vertebral body stabilization. The first approach to deal with such dense high- Z heterogeneities is to seek beam directions that do not require the beam to traverse the heterogeneity in order to cover the target. The use of IMPT can be helpful in this regard, allowing conformal avoidance anc e of the st struc ructur ture. e. Ho Howe wever ver,, it is not alw alway ayss

Figure 6.6. Schemati Schematicc depiction of the situation when a near-tangential near-tangential beam is employed. (a) Correct alignment of compensator (blue) and target volume (red) within a patient section (fawn). The dotted black lines depict the projection of the beam collimators into the patient. The continuous black line depicts the 90 percent isodose line that covers the target volume as intended. The very steep sides of  the compensator are due to a combination of the shape of the target volume and of the patient’s external contour. (b) The patient is shifted left. Now, the thickness from the skin surface to the intended distal 90 percent dose is reduced, so there is overshoot of the dose on the right-hand side. (c) The patient is rotated clockwise. Now, the thickness from the skin surface to the intended distal 90 percent dose is increased, so there is undershoot of the dose on the right-hand side.

107  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

possible to avoid having protons traverse the structure. In that case there are a number of problems: (i) the measured Hounsfield numbers may saturate and not be linear with absorption; (ii) owing to the high- Z   of the ma mater terial ial,, the Hou Hounsfi nsfield eld nu numbe mberrto-wate to-w ater-equ r-equivalen ivalent-de t-densit nsity y conv conversio ersion n schem schemes es break down; and (iii) the structure usually induces heavy hea vy art artifa ifacts cts in the CT da data ta out outsid side e the st struc ruc-ture, making the computation of radiological path length len gthss que quest stion ionabl able, e, eve even n whe when n the ra rays ys do not passs thr pas throug ough h the st struc ructu ture. re. Fo Forr the las lastt of the these se probl pr oblems ems,, the there re ex exis istt pos post-p t-proc rocess essing ing art artifa ifactctreduction techniques, and these can be used with advant adv antage age.. Als Also, o, the re regio gion n of hea heavie viest st art artifa ifacts cts can ca n be ed edit ited ed an and d th the e Ho Houn unsfi sfiel eld d nu numb mber erss ar are e replaced by the expected values for the tissue type in qu quest estion ion.. Sin Since ce th the e com compos positi ition on of the het heter eroogeneou gen eouss st struc ructu ture re is kno known, wn, it is pos possib sible le to edi editt the CT data and substitute a computed number for the Hounsfield numbers within the structure such thatt the Houn tha Hounsfield sfield numb number-to er-to-wa -water-e ter-equiv quivalent alent-dens de nsit ity y co conv nver ersi sion on sc sche heme me us used ed do does es co corr rrec ectl tly y predict the energy loss of protons passing through the structure. However, because the high- Z  of the heterogeneity induce ind ucess sca scatte tterin ring g not han handle dled d by con conven ventio tional nal dose-computation techniques, it is highly desirable to use Mon Monte te Ca Carlo rlo dos dose e com comput putat ation ion tec techn hniqu iques es (see Section 6.5.2.3) to evaluate this situation.

the beam in line with a substantial heterogeneity ( e.g., the base of skull) or through very complexly structured regions ( e.g., the petrous ridge). This is done delib delibera erately tely by expe experienc rienced ed trea treatmen tmentt plan plan-ners. However, the methods are also being investigated gat ed to pr prov ovide ide qu quant antita itativ tive e mea measur sures es to hel help p asse as sess ss wh when en a be beam am is di dire rect cted ed th thro roug ugh h a ba bad d region (see Section 6.2.3.2). 6.5.1.2 6.5.1 .2

External skin surface is a special type of tangential heterogeneity. It can be hazardous to use a beam direction that requires one edge of the beam to be near-tangent to the patient’s outer contour, as illustrated in Fig. 6.6. This figure shows situations in which the whole patient is shifted, but the patient’s anat an atom omy y is st stab able le.. Fi Figu gure re 6. 6.6a 6a sh show owss an id idea eall alig al ignm nmen entt of su such ch a be beam am.. Fi Figu gurre 6. 6.6b 6b an and d c demonstr demon strates ates how norm normal-ti al-tissuessue-spar sparing ing and/ and/or or the tar target get-v -volu olume me co cove verag rage e det deteri eriora orates tes as the pati pa tien entt is sh shif ifte ted, d, or rot ota ate ted. d. Ho How wev ever er,, it is common for the skin and underlying tissue to be differently distorted from day to day, owing to variable pressure from the immobilizing or supporting materials, for example. Such distortions may alter the ski skin-t n-to-t o-targ argetet-vol volum ume e dis distan tances ces,, and hen hence ce affect the dose distribution. Great care needs to be taken take n eith either er to avo avoid id tang tangentia entiall irra irradiat diation ion of the patient’s surface or to control it carefully when it cannot be avoided.

6.5 DESI DESIGN GN OF INDI INDIVIDU VIDUAL AL PRO PROTON TON BEAMS

6.5.2 Cat 6.5.2 Categor egories ies of of models models for dose dose computation

 A number of aspects of beam design in proton-beam therapy have already been touched on in Sections 6.2.2, 6.2.3, and 6.4, and will not be further elaborated here.

There primary of doseare in athree proton beam, models namely,for the calculation †

6.5.1

Compensation Compensa tion for heterog heterogeneities eneities †

The Th e de desi sign gn of co comp mpen ensa sato tors rs,, wh whet ethe herr re real al or  virtual, has been fully described in Section 6.4.6.2. It tak takes es int into o ac accou count nt the eff effect ectss of com compen pensa sator tor misregistration due to patient and/or organ motion or set setup up err error or,, and of mu multi ltiple ple sca scatte tterin ring g of the protons. This is necessary because lateral misregistration can, unlike the situation with x-ray beams, lead to changes in the designed beam penetration (Tatsuzaki and Urie, 1991; Urie  et al., 1984). 6.5.1.1 6.5.1 .1

†

  unif uniform-i orm-inten ntensity sity beam algor algorithm ithmss (whi (which ch can only be used for scattered and wobbled beams);   penci pencil-b l-beam eam alg algori orith thms ms (wh (which ich can be use used d for any type of beam);  Monte Carlo calculations (which can be used for any type of beam).

These are described4 here in the order of increasing accuracy.  All these calculation methods rely on having a thre th reee-di dime mens nsio iona nall de desc scri ript ptio ion n of th the e pa pati tien ent’ t’ss ana an ato tom my in an and d nea earr th the e reg egio ion n of in intter eres estt, together toget her with a map of the water-equi water-equivalen valentt densities. The accuracy of this three-dimensional information, mat ion, which may be comp compromis romised, ed, for example, example, by art artifa ifacts cts in the CT ima images ges or the presenc presence e of  contra con trast st ma mater terial ial,, wil willl af affec fectt th the e ac accur curac acy y of the

Choosing Choosi ng beam direct directions ions

In all radiation modalities, it is common to select the direction of a beam so as to entirely or partially avoid specific OARs. In charged-particle therapy, it is also desirable to avoid beam directions that bring

4

 Adapted from Lomax and Goitein (1997).

108  

Tangenti angential al irradi irradiation ation

TREATMENT TREA TMENT PLANNING

calculations calculati ons (see Secti Section on 8 regar regarding ding unce uncertain rtainty ty analysis). 6.5.2.1 6.5.2 .1

Uniform-intensit Uniform -intensity y beam algorith algorithms ms

UniformUnifor m-int intens ensity ity bea beam m alg algori orithm thmss pr provi ovide de the simplest, simpl est, fas fastest test,, and leas leastt accu accurat rate e appr approach oach to es esti tima mati ting ng do dose The Th e in inpu put t imenta data da ta lly to mea the th e sured dose do se calcul cal culat ation ion are ar ese..eit either her exper ex perime ntally measur ed data, or numerical fits to experimentally measured data da ta (of a var variet iety y of ene energi rgies, es, field siz sizes, es, and so fort fo rth) h),, wh whic ich h de desc scri ribe be th the e do dose se di dist stri ribu buti tion on of   various uniform-inten uniform-intensity sity beams irradiatin irradiating g a flatsurfaced water-equivalent phantom. Typically ypically,, for prot pr oton on-b -bea eam m th ther erap apy y, th thes ese e ma may y co cons nsis istt of a dept de pth– h– do dose se cu curv rve e fo forr an SO SOBP BP to toge geth ther er wi with th lateral dose profiles at representative depths. From these data, the dose at any point in the distribution can be derived by calculating its water-equivalent depth (the integral in depth of the water-equivalent densit den sities ies,, fr from om th the e pa patie tient nt sur surfa face ce do down wn to the point of inte interes rest), t), usually perf performed ormed using a ra rayytracing tra cing algor algorithm, ithm, and inte interpola rpolating ting the res resultin ulting g dose do se fr from om th the e me meas asur ured ed de dept pth– h– do dose se da data ta.. Th The e lateral late ral fall-off fall-off due to collim collimatio ation n is usua usually lly fit by an error function in which the 50 percent level aligns with the projected edge of the collimator. The standard deviation of this error function is derived from measured lateral profiles in water (Petti, 1992). UniformUnif orm-inten intensity sity beam algor algorithm ithmss can be used for both scattered and wobbled beams. In the latter case, a wobbled beam can be assumed to be equi valent to a scattered beam having the same beam-uniformity properties. 6.5.2.2 6.5.2 .2

Pencil-beam Penci l-beam algorit algorithms hms

First,, bec First becaus ause e pen pencil cil bea beams ms pa passi ssing ng out outsid side e of a collim col limat ator or ape apertu rture re ar are e rej reject ected, ed, the pen penumb umbral ral effect of the collimator is automatically calculated. In add additi ition, on, inh inhomo omogen genous ous bea beam m int intens ensiti ities es can easily be modeled, which becomes important when intensity inte nsity-mod -modulat ulated ed beams are to be desig designed ned or their dose distributions to be calculated. However, perhaps the greatest advantage of the pencil-beam approaches is that they more accurately model the effects of heterogeneities on the incident beam. 6.5.2.3 6.5.2 .3

Monte Carlo algori algorithms thms

The most accurate, and hence desirable, dose-estimation algorithms are Monte Carlo models (Jiang and Paganetti, 2004; Paganetti   et al., 2005; Petti, 1996; 1997; Tourovsky   et al., 2005). In this approach appr oach,, indiv individua iduall pro protons tons are tra tracked cked as the they y penetra pene trate te thr through ough the pat patient ient (and in some cases through thr ough the beambeam-deliv delivery ery hard hardwar ware e as well) and inte in tera ract ct wi with th th the e ma mate teri rial al th thro roug ugh h wh whic ich h th they ey pass. The likelihood of an interaction, and its consequ se quen ence ces, s, is sa samp mple led d us usin ing g ra rand ndom om nu numb mber ers, s, from the best ava available ilable pro probabil bability ity dis distribu tribution tions. s. Such random random walk walkss thr through ough the mat material erialss in the beam be am li line ne an and d in th the e pa pati tien entt ar are e fo foll llow owed ed fo forr numerous proton interaction histories. The secondary particles produced in the interactions are also further traced through the material they encounter.  As well as estimatin estimating g the Coulomb interaction interactionss that lead to energy loss and scattering of protons, Monte Mont e Carlo algorithms algorithms perm permit it est estimat imation ion of the nuclear interactions. These are responsible for the

 Although uniform-inten uniform-intensity sity beam algorithms are fast, the estimation of the effects of complex heterogeneit gen eities ies on the fin final al dos dose e dis distri tribu butio tion n is ra rathe therr inaccurate. More accurate modeling of dose can be achi ac hiev eved ed us usin ing g pe penc ncil il-b -beam eam al algo gori rith thms ms (H (Hon ong g  et al., 1996; Pedroni  et al., 2005; Petti, 1992; Scheib and Pedroni, 1992). Typically, the incident beam is modele mod eled d usi using ng a num number ber of clo closel sely y spa spaced ced fin finite ite penc pe ncil il be bea ams ms,, wit ith h ea each ch pe penc ncil il bea eam m bei ein ng assigned assig ned a weig weight ht that is dire directly ctly proportional proportional to the particle fluence of the beam for the pencil’s position ti on.. Ea Each ch pen pencil cil bea beam m br broad oadens ens bec becau ause se of mul mul--

loss of primary protons along the beam path [  20 percent loss over the proton range for a 160 MeV  beam (ICRU, 1998)]; for the production of secondary pr proto otons ns th that at pr produ oduce ce a dos dose e hal halo o ar aroun ound d th the e beam bea m pa path th;; for the pr produ oducti ction on of neu neutr trons ons th that at largely escape from the patient, but may have some importance as regards somatic effects; and for the produ pr oducti ction on of hea heavy vy sec second ondary ary nu nucle cleii ( e.g., alp alpha ha particles, deuterons, and other nuclear fragments) that th at de depo posi sitt do dose se lo loca call lly y ne near ar th the e si site te of th the e nuclear interaction. The latter are heavily ionizing particles that are responsible in part for the elevated RBE observed throughout an SOBP and within the plateau region. The estimation of the dose con-

tiple Coulomb scattering within the patient, and its later la teral al sha shape pe can be mod modeled eled usi using ng mea measur sured ed or calculated data, as can its depth–dose distribution in wa water ter.. Th The e re resul sultan tantt dos dose e at any point is th then en computed by summing the contributions from each of th the e pen pencil cil bea beams, ms, wit with h ea each ch cal calcul culat ated ed poi point nt taken tak en to be at its ac actu tual al wa water ter-eq -equiv uivale alent nt dep depth. th.

tribution from nuclear interactions is an important elemen ele mentt of abs absolu olute te dos dosime imetry try in pla planni nning ng pr proogram gr ams, s, es espe peci cial ally ly fo forr sc scan anne ned d pe penc ncil il be beam amss (Pedroni  et al., 2005). To obtain sufficient statistical accuracy for useful dose dist distribut ributions ions in pra practica cticall situ situatio ations ns [ e.g.,    2 percen per centt st stat atis istic tical al ac accur curac acy y (1 SD) wit within hin ea each ch

dose dos e accu accumul mulat ation ion vo voxel xel of   

Such Suc h app approa roache chess ha have ve a nu numbe mberr of adv advant antage ages. s.

2 mm on a side

109  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

thr throug oughou houtt a 1 l vo volum lume], e], tens of mil millio lions ns of his his-tories tor ies usu usuall ally y mu must st be tr trac aced. ed. Su Such ch cal calcul culat ation ionss can take hours or ev even en da days ys to pr proce ocess. ss. For thi thiss reas re ason on,, so some me Mo Mont nte e Ca Carl rlo o co code dess in inte tend nded ed fo forr proton-beam therapy simulation just follow analytically the primary protons and model contributions

by all OARs of concern. More rarely, the prescription tio n ma may y inc includ lude e goa goals ls for TC TCPs Ps,, NT NTCP CPs, s, and and/or /or EUDs (see Section 6.7.3).  A ‘treatmen ‘treatmentt plan’, or simply ‘plan’, consists of  two tw o el elem emen ents ts.. Fi Firs rst, t, it sp spec ecifi ifies es a nu numb mber er of  beams, each with all its defining parameters such

from secondary secondary part particles. icles. In this case, calcu calculati lation on times tim es can be sub subst stant antial ially ly re redu duced ced by one to two al.., 20 orderss of magn order magnitud itude e (T (Touro ourovsky vsky   et al 2005 05). ). As Mont Mo nte e Ca Carl rlo o me meth thod odss mo mode dell th the e fu fund ndam amen enta tall physics phy sics of pro proton ton inte interac raction tions, s, the resu resultin lting g dose distri dis tribut bution ionss can be con consid sider ered ed to con const stitu itute te th the e gold standard against which other dosecalculational algorithms should be compared. With the next generation of workstations and PCs, and with the aid of variance-reduction techniques, calculat cul ation ion tim times es mig might ht be re redu duced ced su suffic fficien iently tly to make this approach applicable on a routine basis. One way to reduce the long computational time is to us use e a fa fast ster er al algo gori rith thm m in th the e pr proc oces esss of pl plan an developme deve lopment nt and optimizatio optimization, n, and then to use a Mont Mo nte e Ca Carl rlo o al algo gori rith thm m fo forr rec ecom ompu puti ting ng,, an and d perhaps fine-tuning, the final plan.

as mod modali ality ty (prot (protons ons,, ph photo otons ns of a giv given en ene energy rgy,,  etc.; pl plan anss wi with th mi mixe xedd-mo moda dalit lity y be beam amss ar are, e, of  cour co urse se,, en enti tire rely ly po poss ssib ible le), ), ty type pe (b (bro road ad be beam am or scan sc anne ned d be beam am), ), si size ze;; op open en-fi -fiel eld d do dose se pr profi ofile le(s (s); ); direction, and energy, together with the weighting of ea each ch un unifo iformrm-int intens ensity ity bea beam m or or,, for sca scanne nned d beams, of all pencil beams. Secondly, it involves a simu si mula lati tion on of th the e do dose se di dist stri ribu buti tion on wi with thin in th the e patie pa tient nt tha thatt wou would ld res result ult fr from om the ap appli plica catio tion n of  thos th ose e be beam ams. s. In ad addi diti tion on,, th the e pl plan an de defin fines es th the e fracti fra ctiona onatio tion n sch scheme eme (th (the e lis listt of bea beams ms and the dose dos e per bea beam m to be del delive ivere red d in ea each ch tr trea eatme tment nt sessio ses sion n and the tim time e seq sequen uence ce of the sessions sessions)) to be used. Sometimes a treatment is given in more than one segment,   e.g., a fir firsst se segm gmen entt th that at co cov ver erss th the e primar pri mary y tum tumor or and pot potent ential ially ly inv involv olved ed lym lymph ph nodes, and a second segment that covers just the primary tumor and takes it to a higher dose. Each segment will then be represented by an approved plan, pla n, and the tr trea eatme tment nt wil willl con consis sistt of the set of  such plans, which may be combined into an overall plan for purposes of assessing the treatment as a whole.

6.5.3 Norm 6.5.3 Normaliz alizatio ation n and the the calcula calculation tion of monitor units

Ideally,, the dose dis Ideally distribu tribution tion produced produced by a planning calculation provides a three-dimensional map of the absolute dose throughout the patient volume that th at wo woul uld d re resu sult lt fr from om a sp spec ecifi ified ed be beam am-i -inp nput ut intensity ( e.g., monitor units for a scattered beam, and an d pe penc ncil il be beam am flu fluen ence cess fo forr a sc scan anne ned d be beam am)) without with out furt further her norma normaliza lization. tion. This is not alwa always ys the case, however. Some algorithms provide only a relative dose distribution. This is often normalized to some well-defined point (such as the ICRU reference enc e poi point) nt),, wh which ich is sai said d to re recei ceive ve ‘10 ‘100 0 per percen centt dose’, but it can be normalized in many other ways, for example, as 5 percent more than the minimum dose received by 95 percent of the PTV. While there are ar e man many y pos possib sibili ilitie tiess for es estab tablis lishin hing g the bea beam m normal nor maliza izatio tion n (se (see e Sec Sectio tion n 5.6 5.6.4) .4),, a nom nomina inall 100 percent value is generally not the maximum dose delivered.

6.6 DESI DESIGN GN OF GRO GROUPS UPS OF BEAMS BEAMS:: THE TREATMENT PLAN Before the planning Before planning of a radi radiati ation on trea treatmen tmentt can begin, the planning aims must be established (see

6.6.1

Treatment Tr eatment goals and constr constraints aints

6.6.1.1 Setting goal(s 6.6.1.1 goal(s)) and const constraints raints Generally, in this approach, a goal for the target volume dose is established, and a plan is sought thatt mee tha meets ts tha thatt goa goal, l, sub subjec jectt to the re requi quirem rement ent thatt it doe tha doess not violate violate one or mor more e con const stra raint intss usua us uall lly y pl plac aced ed on th the e do dose se de deli live vere red d to no norm rmal al tissues. This is the traditional approach to manual trea tr eatm tmen entt pla lann nnin ing g, an and d it is us used ed in so som me approa app roache chess to opt optimi imized zed pla planni nning. ng. If the goa goall for the tar target get vo volum lume e is to max maximi imize ze som some e qua quanti ntity ty such as the median PTV dose, then this approach guarantees that at least one constraint will be only  just met. This is because, if one were not, the

pla planne nnerr the or weights optimi opt imiza zatio tion n pr progr ogram ambeams would wou lduntil simply sim ply increase of some or all one constraint is just at its limit.

Section 10); generally they include instructions as to the dose which is desired to be received by the target volume(s) (and, often, the desired dose uniformit for mity y wit within hin th the e tum tumor) or) tog togeth ether er wit with h dos dose e or dose–volume constraints on the dose to be received

6.6.1.2 Estab 6.6.1.2 Establishing lishing a score score combini combining ng target-volume and normal-tissue effects

In this approach, the score is a single number composed pos ed of a num number ber of com compon ponent entss th that at rep repre resen sentt 110

 

TREATMENT TREA TMENT PLANNING

the plan’s impact on the target volume or a normal tiss ti ssue ue (o (orr pe perh rhap apss an anot othe herr fa fact ctor or su such ch as so some me measure of plan complexity), each one weighted in some so me ma mann nner er by an ‘i ‘imp mpor orta tanc nce e fa fact ctor’ or’.. Th The e planne pla nnerr or opt optimi imiza zatio tion n alg algori orith thm m the then n see seeks ks to maximize the plan’s score. This approach is used in many ma ny ap appr proa oach ches es to op opti timi mize zed d pl plan anni ning ng.. Th The e factors that go into the score may be comprised of  physical quantities, or quantities estimated by biophysical phy sical models, or some combination. combination. There are many ma ny po poss ssib ible le sc scor orin ing g sc sche heme mess as re revi view ewed ed by Bortfeld (2003) and by York (2003). 6.6.1.3 6.6.1 .3

Targetarget-volume volume goals and constra constraints ints

There Ther e is a wi wide de va vari riet ety y of way ayss to sp spec ecif ify y th the e target tar get-vo -volum lume e dos dose e re requi quire remen ments, ts, inc includ luding ing the following: †

†

†

†

†

†

 all of the PTV receive some specified fraction ( i.e., 95 percent) of the prescribed dose;  at least 95 percent of the PTV receives the prescribe scr ibed d dos dose—a e—and nd it mig might ht fu furth rther er be spe specifi cified ed that the regions of lower dose should be in the PTV PT V per periph iphery ery (so th that at the CT CTV V is min minima imally lly affected);   the mean dose to the PTV be equal to the prescribed dose;   the ICRU reference point receives the prescribed dose;   the prescribed dose be the highest possible dose thatt can be deliv tha delivered ered to the PTV, PTV, subj subject ect to the normal-tissue constraint constraints; s;  a biological model’s estimate of the target-volume response resp onse ( e.g., EU EUD D or TC TCP— P—see see Sec Sectio tion n 6.7 6.7.3) .3) be maximized.

In addition, a constraint might be placed on the target-vo targe t-volume lume dose homo homogenei geneity ty,, for exam example, ple, the dose everywhere within the PTV be within defined limits,   e.g., within   2 5 and   þ 7 percent of the prescribed dose. 6.6.1.4 6.6.1 .4

Normal-tissue Normal -tissue cons constraint traintss

Normal-tissue constraints tend to be either dose or dose–volume constraints. For example: †

 the near-maximum ( D2%) dose to a specified OAR

6.6.2 6.6. 2

Approac Appr oaches hes to to treat treatment ment desig design n

There are two rather different approaches used for the planning and delivery of radiation. 6.6.2.1 6.6.2 .1

Uniform-intensity Uniform -intensity radia radiation tion thera therapy py

In radiation therapy with uniform-intensity beams, treatment is delivered using a number of  uniformunif orm-inten intensity sity beams beams,, gener generally ally cros crosss firing on the th e ta targ rget et vol olum ume( e(s) s),, an and d ea each ch de deli live veri ring ng a clos cl osee-to to-u -uni nifo form rm do dose se to th the e ta targ rget et vo volu lume me.. (I (In n photon therapy, in contrast, there is an unavoidable gradient of dose with depth.) The main art of treatment planning is to choose a set of parameters for each beam which, so far as is possible, lead to an overall dose distribution that spares adjacent critical normal tissues from receiving more dose than the treatment aims will allow, while whi le del delive iverin ring g an ade adequa quate te dos dose e to th the e tu tumor mor.. The Th e di dire rect ctio ion n of in inci cide denc nce e of ea each ch be beam am is on one e import imp ortant ant element element in thi thiss and is oft often en cho chosen sen to avoid av oid criti critical cal stru structur ctures es to the extent possible. So far as beam intensities are concerned, in uniform-intensity-beam radiation therapy, one has only to adjust the weights of the individual beams. In manual optimization, the design of the individual bea beams ms is bas based ed on th the e pla plann nner’s er’s ex exper perien ience; ce; computer tools such as the possibility of designing an aperture or the settings of a multi-leaf collimator aut automa omatic ticall ally y, on the bas basis is of th the e bea beam’ m’s-e s-eye ye  view (BEV) of the PTV, can greatly assist in beam design (Goitein  et al., 1983b). In pro proton-b ton-beam eam therapy ther apy, there ther e to is avoid an addi additiona tionall consideration, namely, the, desire metallic objects ( e.g., metal posts needed for spinal column stabi st abiliz lizat ation ion or hip pr prost osthes heses) es) tha thatt mig might ht ha have ve been implanted into the patient. These can perturb a beam’s dose distribution to a very great degree, which is hard to estimate (see Section 6.4.6.2). The  judgement as to whether a plan is satisfactory if on one e or mo more re be beam amss tr trav aver erse se su such ch a me meta tall llic ic object obj ect is dif difficu ficult. lt. It dep depend endss on ex exper perien ience, ce, an and d can be made much more accurate and objective by Monte Mont e Carlo simulations. simulations. To the ext extent ent possi possible, ble, beams are chosen for which the metallic object is near ne ar th the e di dist stal al en end d of th the e be beam am,, or pr pref efer erab ably ly beyond it.

†

†

†

not exceed some stated value;  the mean or median dose to a specified OAR not exceed some stated value;   the dose,   D, received by some specified fractional  volume,   V , of a sp spec ecifi ified ed OA OAR R no nott ex exce ceed ed so some me stated value ( DV   some value);  a biological model’s estimate of the response of a normal tissue ( e.g., EUD or NT NTCP— CP—see see Section 6.7.3) not exceed some stated value.

The development of a treatment plan involves a trialtria l-an andd-er erro rorr pr proc oces esss in wh whic ich h a se sett of be beam amss (includin (inc luding g their weig weights) hts) are prop proposed osed,, evalu evaluated ated as to th their eir res result ulting ing dos dose e dis distri tribut bution ion,, and the then, n, if nec necess essary ary,, mod modifie ified. d. Thi Thiss pr proce ocess ss is re repea peated ted until a satisfactory plan is arrived at. In unifo un iformrm-int intens ensity ity-be -beam am ra radia diatio tion n th ther erapy apy, the tria tr iall-an andd-er erro rorr pr proc oces esss is us usua uall lly y un unde dert rtak aken en manually (in which case only a limited number of  111 11 1

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

riv rival al pla plans ns can be con consid sider ered) ed).. Ho Howe weve verr, it ma may y also be done auto automat maticall ically y or semisemi-auto automat matically ically, employing the same optimization techniques as are used in IMRT, as described below.

6.6.2.2 Intens 6.6.2.2 Intensity-modu ity-modulated lated radia radiation tion therap therapy y (IMRT)

In IM IMRT RT,, th the e tr trea eatm tmen entt em empl ploy oyss a nu numb mber er of  beams bea ms wh which ich tog togeth ether er re resul sultt in a nea near-u r-unif niform orm dose to the PTV, but each of whic which h deliv delivers ers a nonunif un ifor orm m do dose se wi with thin in th the e PT PTV V. To di dist stin ingu guis ish h betw be twee een n IM IMRT RT, in ge gene nera ral, l, an and d th the e ra radi diat atio ion n modality-specific implementations of it, the following terms and abbreviations are useful:

6.7 PLA PLAN N ASS ASSESS ESSMENT MENT Once a treatment plan has been decided, its impact on the pa patie tient nt mu must st be ass assess essed ed (Go (Goite itein, in, 199 1992). 2). There are four basic elements in this: inspection of  the resulting absorbed-dose distribution overlaid on the pat patient ient’s ’s ana anatomy tomy,, asses assessmen smentt of the clini clinical cal feasib fea sibili ility ty of del delive iverin ring g the pla plan, n, re revie view w of dos doseesumm su mmar ariz izin ing g qu quan anti titi ties es,, an and d re revi view ew of ot othe herr measures measu res of rad radiati iation on effe effect. ct. It shou should ld be emph emphaasize si zed d th tha at th ther ere e is al almo most st no noth thin ing g sp spec ecifi ificc to proton-be prot on-beam am ther therapy apy in this proc process. ess. Com Compar parable able techniques are needed to assess treatment plans no matter what radiation modality they involve.

6.7.1 6.7. 1 IMRT IMXT IMPT

Intensity-modu Intensitymodulat lated ed tre treatm atment ent usin using g any form form(s) (s) of  radiation. IntensityInten sity-modu modulat lated ed tre treatm atment ent usin using g x-r x-ray ay (photon) (photon) beams specifically. IntensityInten sity-modu modulat lated ed tre treatm atment ent usin using g pro proton ton beam beamss specifically.

The dose dis distri tribut bution ion in any pla plane ne (gene (general rally ly a transverse, sagittal, or coronal section) can be displayed in a number of ways: †

†

If intensity-modulated beams of different modalitiess we itie were re co comb mbin ined ed in a si sing ngle le tr trea eatm tmen ent, t, th the e result would be described as IMRT. In IMRT, the number of variables is too great for manual optimization to be viable since, in addition to al alll th the e va vari riab able less ne need eded ed to de defin fine e be beam amss in unifor uni form-i m-int ntens ensity ity-be -beam am ra radia diatio tion n the therap rapy y, th the e intensity profiles of the pencil beams that comprise each ea ch be beam am mu must st al also so be de dete term rmin ined ed,, an and d th this is usua us uall lly y in invo volv lves es th thou ousa sand ndss of ad addi diti tion onal al pa parrameters. amet ers. As a conse consequen quence, ce, comp computer uter automation automation is essential in IMRT (see Section 6.9). This has the benefit that it is possible to generate and assess a  very large number of rival plans. In IMPT IMPT,, there are ar e ev even en mo more re va vari riab able less th than an in IM IMXT XT,, wh whic ich h permit per mit bot both h dep depth th and la later teral al mod modula ulatio tion. n. Th This is makes the computational task even more demanding

Inspectio Inspe ction n of the the dose dist distribu ribution tion

†

  as as iso isodos dose e con contou tours rs ov overl erlaid aid on out outlin lines es of the patient pat ient’s ’s ana anatomy tomy and tum tumor or (with (without out imagi imaging ng data);  as isodose contours overlaid on CT or MRI data; or  as a color wash wash overlaid overlaid on CT or MRI data. data.

Of these these,, the color-wash color-wash display gives the mos mostt immediate qualitative impression of the dose distribution but ion and its imp impac actt on the pa patie tient nt.. Ho Howe weve verr, unless efforts are made to divide the dose-to-color map int into o dis distin tingui guisha shable ble ban bands, ds, thi thiss dis displa play y is somewha some whatt less easy to asses assesss quan quantita titative tively ly than the display of isodose contours. It is ver ery y use sefu full to be ab able le to vie iew w, sim imul ul-taneously on the screen, three orthogonal sections through thr ough the thr three-d ee-dimens imensional ional dat data a (whil (while e bein being g able ab le to in inte tera ract ctiv ivel ely y ad adju just st th thei eirr in inte ters rsec ecti tion on point), and to rapidly scan through a sequence of  such su ch im imag ages es.. A to tool ol th that at al allo lows ws a us user er to mo move ve intera int eracti ctive vely ly a poi point nter er aro aroun und d in any ima image ge and

than in IMXT. With so man many y pot potent ential ial var variab iables les,, th the e com compu pu--

display displa y the dos dose e (id (ideal eally ly wit with h its un uncer certai tainty nty)) at thatt po tha point int is als also o ve very ry hel helpfu pfull for the pur purpos pose e of 

tational task of choosing the best value for each of  them the m is ve very ry dem demand anding ing.. Whe When n all pos possib sible le var variiable ab less ar are e in incl clud uded ed in th the e co comp mput uter er se sear arch ch,, th the e process is said to be   automatic. However, in order to reduce the computational task to a viable level, it is common to choose a number of the variables manually. These often include choice of the number of be beam ams, s, th the e mo moda dali lity ty of ea each ch be beam am,, an and d th the e placement and direction of each beam. When some  variables are chosen manually manually,, the process is said to be  semi-automatic.

exploring a dose distribution in detail. Ther Th ere e ar are e re real ally ly no ve very ry sa sati tisf sfac acto tory ry wa ways ys of  gaining an impression of the full three-dimensional dose distribution in a single view. Dose bands, dose clouds clo uds,, and dose on the surfac surface e of VO VOIs Is ha have ve all been investigated as possible approaches.  Although there are several other quantities that assist in the evaluation of plans (described in the immediately following sections), a final judgement about a plan should never be made without carefully inspecting its dose distribution. 112

 

TREATMENT TREA TMENT PLANNING

6.7. 6.7.2 2

Clinical Clini cal feas feasibili ibility ty

 A plan might appear to make sense, but may be impr im pra act ctic ical al or un unde desi sira rabl ble e in pr pra act ctic ice. e. Th The e numb nu mber er of be beam amss mi migh ghtt be de deem emed ed ex exce cess ssiv ive, e, some beam angulations may be difficult to deliver the time to deliver the plan may be excessive, the patie pa tient’ nt’ss con condit dition ion ma may y con const stra rain in th the e pla plan n in a numb nu mber er of wa ways ys,,   etc. Th The e re reso solu luti tion on of th thes ese e larg la rgel ely y su subj bjec ecti tive ve de deci cisi sion onss ca can n be af affe fect cted ed by institutional policies. 6.7.3 6.7. 3

Dose-summ Dose -summariz arizing ing quan quantitie tities s

In addition to graphical displays of the RBE-weighted absorbed dose distribution, the computer can present the numerical value of the dose or any quantity derived from the dose distribution. Dose-sum Dose-s ummar marizi izing ng par param amete eters rs ar are e des descri cribed bed in Section 5.6. Examples are the following: † †

†

†

†

 the dose at each of several points of interest;   the mean mean,, medi median, an, minim minimum, um, or nearnear-mini minimum mum and/ an d/or or ma maxi ximu mum m do dose se [ Dmean,   D50%   ( Dmedian),  D98%   ( Dnear-min),   D2%   ( Dnear-max), resp respectiv ectively] ely] to appropriate VOIs (such as the CTV, PTV, or OAR or PRV);   the volume, or relative volume, of a VOI receiving at lea least st a spe specifi cified ed dos dose, e,   D, rep represe resented nted as V  D  D;  the least dose received by a given volume, or relative tiv e vo volum lume, e,   V , of a VOI, represented as  DV ;   dose– dose– vo volum lume e his histog togra rams ms (ei (eithe therr cum cumula ulativ tive e or differential).

 A very useful capability is to present such dose-summarizing quantities in juxtaposition with the val values ues spe specifi cified ed in the tr trea eatm tment ent aim aims. s. Fo Forr exam ex ampl ple, e, if a co cons nstr trai aint nt on an OAR is th tha at at most mo st 20 pe perc rcen entt of th the e OA OAR R re rece ceiv ives es a do dose se in

†

†

†

  the tumo tumorr contr control ol pro probabil bability ity (TC (TCP) P) (Nie (Niemierk mierko o and Goitein, 1993; Webb, 1994; York, 2003);   th the e no norm rmal al-t -tis issu sue e co comp mpli lica cati tion on pr prob obab abil ilit ity y (NTCP) of ea (NT each ch of sev eve eral nor orm mal tis issu sues es al.., 19 (Lyman (L yman,, 1985; Schu Schulthei ltheiss ss   et al 1983 83;; Yor ork, k, 2003);   the pr proba obabil bility ity of unc uncomp omplic licat ated ed loc local al con contr trol ol ( pre prefera ferably bly usin using g impo importan rtance ce weig weightin hting g fact factors) ors) (York, 2003).

 All of these models involve assumptions (either explicit or implicit) about the response to radiation of tu tumor morss or nor normal mal tis tissue suess tha thatt are unlikely unlikely to fully represent their actual behavior. Moreover, the exper ex perime imenta ntall and cli clinic nical al da data ta on dos dose– e– vol volume ume effect eff ectss ar are e ra rath ther er spa sparse rse and th the e cli clinic nical al da data, ta, in particular, have substantial uncertainties. The validity of all of these models is therefore in question. Neverth Nev ertheless, eless, the they y offer helpf helpful ul aids aids,, part particular icularly ly in comparing plans (see Section 6.8), provided they are ar e us used ed ca caut utio ious usly ly an and d un unde ders rsto tood od on only ly to support, and not to replace, expert judgement. The plan’s ‘score’ is a summarizing quantity that attempts to quantitatively represent the quality of  the plan by a single number, always subject to its meeting any constraints. This is of very limited use for the assessment of a single plan, but is useful in comparing a few plans, and is essential in computer optimization of very many plans—as discussed in the following section.

6.8 PLA PLAN N COM COMP PARIS ARISON ON One im One impo port rtan antt go goal al of pl plan an as asse sess ssme ment nt is to support the comparison of alternative plans for a given patient (Goitein, 1992). This may be done for the th e pur purpo pose se of dec decidi iding ng wh which ich of sev sever eral al pos possib sible le plans should be used to treat a given patient, or for

comparing rival approaches, including rival modalities ( e.g., protons versus photons), using a specific patient as a test case. One can ha hardl rdly y com compar pare e two (or mor more) e) obj object ectss withou wit houtt som some e mea measur sure(s e(s)) of ea each. ch. The pri princi ncipal pal tools too ls for pla plan n com compar pariso ison n ha have ve alr alread eady y bee been n pr preesented in the previous section on plan assessment. They may be used, however, in two rather different ways wa ys,, on one e su subj bjec ecti tive ve an and d th the e ot othe herr ob obje ject ctiv ive. e. Objective is meant a process that does not rely on human judgement and, if repeated, gives the same result.

excesss of 50 Gy exces Gy,, the then n ju juxta xtapos positi ition on of th the e ac actua tuall  value of    D20%   toget togethe herr wi with th th the e de desi sire red d va valu lue e of 50 Gy di dire rect ctly ly in indi dica cate tess wh whet ethe herr or no nott th the e cons co nstr trai aint nt wa wass me met. t. Eq Equa uall lly y, if se seve vera rall do dose– se–  volume constrain constraints ts (or dose constrain constraints ts such as co cons nstr trai aint ntss on   Dmax,   Dmin, or   Dmedian) are superimposed on a DVH, the planner can immediately see whether any are violated and by how much. In recent years, there has been much effort put into developing measures of the impact of radiation on both tumors and normal tissues. These include models to predict the following: †

5

The Th e equ equiva ivale lent nt uni unifor form m dos dose e is tha thatt dos dose e wh which ich,, if app applie lied d uniformly to the entire volume of the tumor or a normal tissue, is ju judg dged ed on th the e ba basi siss of a si simp mple le bi biop ophy hysi sica call mo mode dell to be equivalent to the non-uniform dose distribution in terms of its clinical consequen consequences. ces.

 the equivalent uniform dose (EUD) 5 (Niemierko, 199 997 7) fo forr bot oth h tum umor or an and d nor orma mall tis issu sues es (Bortfeld, 2003); 113

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

6.8. 6.8.1 1

Plan compa compariso rison n by inspe inspection ction

The com compar pariso ison n of pla plans ns th thro rough ugh ins inspec pectio tion n by humans is a somewhat subjective matter. That is, two experts can disagree on which of a given pair of pla plans ns is the bet better ter.. The comparis comparison on inv involv olves es a multi mu ltipli plicit city y of fa facto ctors. rs. Exp Expert ertss can dif differ fer on th the e relative importance of different aspects of the plans ( e.g., adequacy of tumor coverage versus avoidance of nor normal mal tissues) tissues),, and can dis disagr agree ee on th the e mos mostt important impo rtant measures measures to weig weight ht ( e.g.,   Dmin   or   Dmean of the tumor). Nevertheless, in the end, an observer must be able to rank a pair of plans, with one of  them the m bei being ng jud judged ged as bei being ng re relat lativ ively ely ‘be ‘bett tter’ er’ or ‘worse’ than the other, or perhaps considering them as equal equally ly desir desirable able clini clinically cally.. The obser observer ver mus mustt also be able to judge whether a plan is adequate, or whether an alternative must be sought.  Although plan comparison might be subjective, it is greatly facilitated by the side-by-side comparison of ob obje ject ctiv ive e me meas asur ures es,, su such ch as th thos ose e al alre read ady y descri des cribed bed for pla plan n ana analys lysis. is. Th Thus, us, any pla platfo tform rm that is intended to assist the planner in comparing plans should, for two or more plans, be able to do the following: †

  display, side-by-side, the dose distributions overlaid over the identical anatomic image ( i.e., the same slice) and to advance the slice being viewed rapidly in all panels;

to assess whether an element of a plan satisfies the treatment aims. The first and the last in the above list of capabilities are probably themight most allow usefulthe in comparing plans. Optional features planner to: †

†

†

  tabu tabulate late,, sideside-by-si by-side, de, the EUD for both tumor and normal tissues;   tabula tabulate, te, sideside-by-sid by-side, e, compu computat tation ion of the TCP TCP;; and  tabulate, side-by-side, computation of the NTCP of each of several normal tissues.

 A recent approa approach ch to planning is the use of the so-called Pareto optimization (Bortfeld  et al., 2004). Thiss inv Thi involv olves es mod modific ificat ation ion,, nei neithe therr of th the e usu usual al input inp ut var variab iables les ( e.g., gan gantry try ang angle, le, bea beam m we weigh ight, t,  etc.), no norr of im impo port rtan ance ce fa fact ctor orss bu but, t, rat athe herr, of  meas me asur ures es of ou outp tput ut qu quan anti titi ties es su such ch as EU EUD D or TCP/NTCP, which quantify the effect of the plan on the VOI (these measures may be called the ‘output  variables’). In this approa approach, ch, a vast number of  plans are generated by computer. Plans for which impro imp rove vemen mentt of one of the out outpu putt var variab iables les wil willl inevit ine vitabl ably y re resul sultt in the worsenin worsening g of at lea least st one other output variable are said to lie on the Pareto front. fro nt. The pla plann nner er vie views ws an int inter erac activ tive e dis displa play y thatt sho tha shows ws th the e val values ues of all the output output var variab iables les for the currently selected plan. He or she can then increa inc rease se or dec decre rease ase the val value ue of any one out outpu putt

†

  of tabulate, the computed dose at each severalside-by-side, points of interest, preferably with con†

fidence limits;   tab tabula ulate, te, sid side-b e-by-s y-side ide,, qua quant ntiti ities es suc such h as the mean,, medi mean median, an, (nea (near-) r-) mini minimum, mum, and/ and/or or (near (near-) -) maximu max imum m dos dose e ( D , D , D , D ,

 variable—thus to interactively, a neighboringthe plan on the Pareto front,moving and view, consequen qu ence cess fo forr all ot oth her out utp put va vari ria abl bles es.. For example, one might reduce the NTCP for the spinal cord and see what impact that would have on the

mean

†

†

†

†

median

near min

TCP and the NTCPs of other critical structures. In this manner, the user can choose a plan that represe re sent ntss hi hiss or he herr be best st ju judg dgem emen entt of th the e mo most st accep ac ceptab table le pla plan n amo among ng th those ose lyi lying ng on th the e Pa Paret reto o fron fr ont. t. Th This is ap appr proa oach ch in no wa way y el elim imin inat ates es th the e problem of subjectivity in the comparison of plans. However How ever,, it has two att attrac ractive tive char characte acterist ristics: ics: it cons co nstr trai ains ns th the e us user er to vi view ew on only ly a pr prod oduc ucti tive ve subset of possible plans, and it allows the user to

min

 Dnear-max,  D max, respectively) to any VOI (such as the CTV, PTV, or OAR or PR PRV), V), pre prefera ferably bly with confidence limits;   tabula tabulate, te, side side-by-s -by-side, ide, the comp computed uted volu volume, me, or rela re lati tive ve vo volu lume me,, of a VOI re rece ceiv ivin ing g at le leas astt a specified dose,   D   (represented as   V  D), preferably with confidence limits;   ta tabu bula late te,, si side de-b -byy-si side de,, th the e mi mini nimu mum m do dose se

received by a given volume, or relative volume,  V , of a VOI, represented as   DV , preferably with confidence limits;   ove overla rlay y on th the e sam same e plo plot, t, for each of the plans plans bein be ing g co comp mpar ared ed,, th the e DV DVHs Hs of an any y gi give ven n VO VOII (either cumulative or differential); and   pr pres esen entt es esti tima mate tess of th the e un unce cert rtai aint ntie iess in th the e plans (at a minimum, where these differ between plans).

make adjustments in the space of clinically meaningful variables. 6.8.2 6.8. 2

Automate Autom ated d plan plan comp comparis arison on

It is no now w pos possib sible le (an (and, d, for IMR IMRT T, nec necessa essary) ry) for a computer to generate literally thousands of plans for a single patient. The selection of the most desirable of these is simpl simply y not possi possible ble through through human inspe inspecction ti on of ea each ch on one, e, as it wo woul uld d be to too o bu burd rden enso some me.. Therefore, some automated approach is necessary. In order for a computer to rank plans, it must have a unique numerical measure of plan goodness.

If a quantity has been specified in the treatment aims, it is very useful to indicate its desired value in the tabulation of the data being compared, so as 114  

TREATMENT TREA TMENT PLANNING

That is, in spite of what has just been said about

The constraints and goals are given in the treat-

plan evaluati evalu ation onbe being multi-fa i-faceted ceted matter, matter , the computer must able ato mult reduce several measures to one number; it is termed as the plan’s ‘score’ (see Sectio Sec tion n 6.6 6.6.1. .1.2). 2). With ithout out suc such h a sco score re,, it can cannot not rank any two plans. And, without a choice of the better of any two plans, it cannot determine which of many plans is the best according to the chosen method met hod of sco scorin ring. g. Onc Once e an obj object ectiv ive e met method hod of  scoring is decided upon, the process of determining the most acceptable of the plans is trivial: one must merely identify the plan with the highest score(s). Obviously, being forced to use a single number to  judge plans might not be a clinical advantage as the sco score re ma may y not fai faithf thfull ully y refl reflect ect a giv given en pla plann-

men ment aims aim s and repres rep resent ent th the eThe clinic cli nician ian’s ’s ins instru trucctionst for the planning process. constraints may be ‘hard’ constraints, in which case a minor violation is considered to be as bad as a severe violation, or they may be ‘soft’, in which case a small violation tio n ma may y be all allow owed, ed, if nec necess essary ary.. Alt Althou hough gh th the e process of having a computer look for the best plan is ful fully ly aut automa omatic tic,, it is not uncommon uncommon to mod modify ify the constraints or goals and try again, if the resulting plan is deemed by inspection to be unsatisfactory to ry. Th Thus us,, ev even en th the e au auto toma mate ted d ap appr proa oach ch to trea tr eatm tmen entt pl plan anni ning ng us usua uall lly y ha hass a ma manu nual al (i.e., user-guided) element to it.  A good optimization algorithm should be insensi-

ner’s ju ner’s judg dgem emen ent. t. It do does es ha have ve th the e ad adva vant ntag age, e, howe ho weve verr, of all allow owing ing a lar larger ger nu numbe mberr of pla plans ns to be assessed than would be possible using manual-assessmentt techniques. manual-assessmen The heart of the comparison of plans by computer, then, is the development of a satisfactory algorithm for computing a plan’s score. A number of such algorithms have been developed and used (Bortfeld, 2003;  York,  Y ork, 2003). The rationale rationale for some of them has more to do with computational simplicity than with their fidelity fidelit y to the way an expert expert would judge judge a plan.

tive to the st tive start arting ing val values ues use used d for the var variab iables les being adjusted. This is the case for some optimization algorithms being used in practice, but not for all. In any event, variables that are not adjusted in the th e ite itera rativ tive e pr proce ocess ss do nee need d to be spe specifi cified ed wit with h some so me ca care re as th they ey wi will ll re rema main in in th the e fin final al pl plan an.. Thus Th us,, fo forr re reas ason onss of co comp mput utat atio iona nall sp spee eed, d, it is comm co mmon on no nott to vary th the e nu numb mber er or th the e an angl gle e of  entry of the beams. In that event, if one wishes to take advantage of the geometry, the beams must be pre-selected using expert judgement. Optimi Opt imiza zatio tion n tec techn hniqu iques es ha have ve bee been n re revie viewe wed d by Ce Cens nsor or (2 (200 003) 3).. A nu numb mber er of ma math them emat atic ical al appr ap proa oach ches es ha hass be been en us used ed,, in incl clud udin ing g si simp mple lex x methods meth ods (Dant (Dantzig, zig, 1963 1963;; Nied Nieder er and Mead, 1965) 1965),,  various methods of conjugate-g conjugate-gradient radient search, and

6.9 PLA PLAN N OPTI OPTIMIZA MIZATION TION The Th e pr proc oces esss of it iter erat ativ ivel ely y ge gene nera rati ting ng an and d th then en

automatic automa ticall ally y ass assess essing ing a lar large ge nu numbe mberr of pla plans ns and cho choosi osing ng th the e bes bestt amo among ng the them m is com common monly ly termed plan optimization. This term has a precise math ma thema ematic tical al mea meanin ning: g: the pr proce ocess ss of cho choosi osing ng those value valuess for the trea treatmen tment-del t-deliver ivery y varia variables bles that would result in an  extremum of the score function. However, given that usually only a subset of  the variables has been adjusted, and that one can have ha ve at bes bestt onl only y ve very ry lim limite ited d con confide fidence nce in the nume nu meri rica call me meas asur ure e of th the e pl plan an’s ’s go good odne ness ss (i (its ts score), the term optimization can be misleading.  A computer can be programme programmed d to design a treat-

simulated simula ted ann anneal ealing ing.. It is imp import ortant ant to poi point nt out that both the conduct of the search and the validity of the scoring method are susceptible to error. Some search methods are prone to find local  extrema   and get trapped in them, thus missing a possibly much better bett er solu solution tion else elsewher where e in the para paramete meterr spa space; ce; others force restrictions on the nature of the score functi fun ction on in ord order er for them to wor work. k. Sco Score res, s, whi while le object obj ectiv ive, e, can als also o be cli clinic nicall ally y mis mislea leadin ding g if th the e underlying algorithm does not adequately take into account the biology of the problem.

ment plan automatically, given the following: †

† †

† †

6.10 COMP COMPARISON ARISON OF UNIFOR UNIFORM-INTEN M-INTENSITY  SITY   VERSUS IMPT TREATMENT  VERSUS TREATMENT PLANS

  a unique   goal   ( e.g.,   D95%   for the tumor to be as high as possible); and a set of   constraints   ( e.g., V 60Gy percent, ent, and/ and/or or 60Gy   for the kidney to be   , 30 perc the th e do dose se wi with thin in th the e PR PRV V no nott to ex exce ceed ed so some me stated dose);   a  method of giving a score  to a plan; variabl iables es to adju adjust st   ( e.g.,   the the defin definition ition of   what var the th e pe penc ncil il-b -bea eam m we weig ight htss th that at de dete term rmin ine e th the e intensity profiles); and   an  initial guess  at a plan;   a method of searching   for that set of plan variables that maximizes the score.

 As an example of plan comparison, a comparison of  proto pr oton n th thera erapy py usi using ng un unifo iformrm-int intens ensity ity bea beams ms  versus uniform-intens uniform-intensity ity scanned beams versus intensity inte nsity-mod -modulat ulated ed scann scanned ed beam beamss is pres presente ented d 6 here. Side-by-side dose distributions in one trans verse plane are shown in Fig. 6.7 for three plans, namely nam ely pla plans ns usi using ng un unifo iformrm-int intens ensity ity sca scatt ttere ered d 6

This Th is se sect ctio ion n ha hass be been en pr prep epar ared ed wi with th th the e as assi sist stan ance ce of   A. Lomax, Paul Scherrer Institute, Switzerland.

115  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

formation on for a single beam (upper panels) and for a three-beam plan (lower Figure 6.7. A comparison of various techniques of beam formati panels) for a Ewing’s sarcoma (the PTV is outlined in yellow). The three techniques are as follows: (a) and (b) passively scattered beam; (c) and (d) scanned beam(s), but with a uniform dose in the target volume from each beam; (e) and (f) IMPT, in which the individual beams irradiate the target volume non-uniformly, but the combined dose is uniform within the target volume. (e) The dose distribution

beams irradiate the target volume non uniformly, but the combined dose is uniform within the target volume. (e) The dose distribution from one of the three beams. (f) The combined dose distribution from all three beams, displaying excellent conformation of the dose to the target volume (see Section 6.10). (Figure courtesy of A. Lomax, Paul Scherrer Institute, Villigen, Switzerland.)

beams; uniform-intensity scanned beams with variabl ble e mo mod dula lati tio on acr cro oss th the e fiel eld d; an and d fu full lly y intensity-modulated and optimized scanned beams.  As can be seen in Fig. 6.7a, even a single uniform-intensity scattered beam can provide excellent le nt co conf nform ormat atio ion n of do dose se to th the e di dist stal al en end d of th the e targe ta rgett an and d goo good d la late tera rall con confo form rmat atio ion. n. Ho Howe weve verr, because of the fixed modulation of Bragg peaks in depth across the whole field, it provides little highdose conformation of the dose to the proximal side of  the target volume. In practice, the dose conformation to the target volume is enhanced through the applicati ca tion on of mu multi ltipl ple, e, an angu gular larly ly se sepa para rate ted d be beam amss (Fig. 6.7b). Furt Further her impr improvem ovements ents in the deliv delivered ered dose can be achieved with beam-patching techniques. For example, the distal end of one or more beams can be di dirrec ectl tly y ab abut utte ted d ag agai ains nstt th the e la late terral ed edge ge of  another beam, such that critical structures close to the tumor can be selecti selectively vely avoi avoided. ded. In co cont ntra rast st to a sc scat atte tere red d be beam am,, th the e tr trea eate ted d  volume of a scanned beam can be matched (at relatively high dose levels) to the proximal, as well as the distal, surface surface of the target volu volume, me, resulting resulting in a bea beam m tha thatt is nea nearr-uni unifor form m wit within hin the tar target get  volume, but provides some sparing of the tissues

tissues. Figure 6.7c shows an example of the dose distribution of such a scanned beam for the identical beam geometry as shown for the uniform-i unifo rm-inten ntensity sity beam of Fig. 6.7a, and confirms that th at sc scan anni ning ng ca can n co conf nfor orm m th the e hi high gh do dose se to th the e target volume in all three dimensions, significantly reducing redu cing the doses delivered delivered to the tissues proximal to the target volume in comparison to use of a scattered beam. The dose distribution of three such scanned (but uniform within the PTV) beams, using the identical geometry as that of Fig. 6.7b, is shown in Fig. 6.7d. Although the difference between scanning and scattering in the three-beam plan is less marked mark ed tha than n in the sing single-b le-beam eam plan (co (compa mparin ring g Fig. 6.7b and d), neverthe nevertheles lesss scan scannin ning g is seen to provid pro vide e a met method hod by whi which ch the high-dose high-dose con confor for-mation can be further improved, and by which doses outside of the target can be globally reduced.  A homogenous dose distribution distribution,, or indeed any desire des ired d dos dose e dis distri tribut bution ion,, can als also o be com compu puted ted from a number of individually highly non-uniform proton beams (i.e., IMPT). By removing the restriction tio n th that at ea each ch bea beam m mu must st its itself elf be hom homoge ogenou nous, s, free rein is given to the optimization, allowing it to fully exploit all the degrees of freedom provided by

prox pr oxim imal al to it as wel elll as fu full ll sp spar arin ing g of di dissta tall

the many thousands of individually weighted and 116

 

TREATMENT TREATMENT PLANNING

Figure 6.8. Dose–volum Figure Dose–volume e his histogr tograms ams for the 3 thr three-b ee-beam eam plans shown in the low lower er panels of Fig. 6.7 for (a) the PTV PTV;; (b) the righ rightt femoral head; (c) the intestines; and (d) the volume within the body outline, but excluding the PTV. PTV. (Figure courtesy of A. Lomax, Paul Scherrerr Institute Scherre Institute,, Villig Villigen, en, Switzerlan Switzerland.) d.)

three-dimens three-di mensional ionally ly dist distribut ributed ed Bra Bragg gg peaks incidentt fr den from om two or mor more e dif differ ferent ent bea beam m dir direct ection ions. s. The re resul sultt is the pos possib sibili ility ty of con confor formin ming g a dos dose e distribution at different dose levels, using relatively few beams, whil while e sti still ll selec selective tively ly avo avoiding iding organs tha th at ar are e de deep eply ly em embe bedd dded ed wi with thin in th the e ta targ rget et  volume. Figure 6.7e shows the dose distribution for  just one of the three beams whose combined dose distribution is shown in Fig. 6.7f. When Fig. 6.7f is compared with either of Fig. 6.7d or b, the advantage of IMPT becomes clear. The DVHs for the PTV, PTV, the right femo femoral ral head, the th e in inte test stin ines es,, an and d th the e en enti tire re im imag aged ed pa pati tien entt  volume minus the PTV are shown in Fig. 6.8 for the thre three e plan plans. s. Thes These e hist histogra ograms ms quan quantita titative tively ly capture the full three-dimensional dose distribution in these VOIs, but the images shown in Fig. 6.8 are by neces necessity sity twotwo-dime dimension nsional al image imagess and cann cannot ot

well covered in all three plans. The DVH shown in Fig. 6.8 indicates that the femoral head is progressively better spared as one goes from passively scattered beams, to scanned but uniform beams, and to fully intensity-modulated beams, whereas the coverage of the intestines is worse with the passively scattered beams, but comparable for the other two beam be am-d -del eliv ive ery system ems. s. The DV DVH H sho hown wn in Fig. Fi g. 6. 6.8d 8d sh show owss ho how w mu much ch do dose se is di dist stri ribu bute ted d outs ou tsid ide e th the e PT PTV V (t (the he DV DVH H fo forr th the e RVR is no nott shown here as so much of the body is delineated). The Th e vo volu lume me ax axis is in th this is DV DVH H is sh show own n on only ly in absolu abs olute te uni units ts as a re relat lativ ive e vo volum lume e wou would ld ha have ve little littl e mean meaning ing (see Section 5.6.2.2). 5.6.2.2). Ther There e is    25 percent more dose delivered outside the PTV in the scatt sca ttere ered d bea beam m pla plan n tha than n in th the e oth other er two plans that are comparable with one another. Dose sta statis tistics tics hav have e an impo important rtant quan quantita titative tive

show howFrom the dose compare other sections. Fig. distributions 6.8a, it is clear that theinPTV is

role in comparing plans. Table a representativ sent ative e set of such stati statistic stics. s. 6.1 In shows principle, prin ciple, such 117

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

dose sta statist tistics ics (exc (except ept   Dmean) can be directly read off th the e DV DVHs. Hs. Dis Displa playin ying g the da data ta for the plans side-by-side, in tabular form, is very helpful. 6.11 6.1 1 6.11.1 6.1 1.1

SPECIA SPE CIAL L TEC TECHNIQ HNIQUES UES Intraocular trea treatments tments with protons

 As described in Section 3.2.3.1, a specialized apparatus is usually used in the treatment of intraocular tumors. The treatment-planning programs used for such treatments are also specialized. They require the th e sa same me el elem emen ents ts as ar are e re requ quir ired ed fo forr ge gene nera rall proto pr oton n bea beam m the thera rapy py,, but the these se ar are e dif differ ferent ently ly implement impl emented. ed. The majo majority rity of eye eye-tr -treatm eatment ent programs gra ms use a pr progr ogram am bas based ed on th that at rep report orted ed by Goitein Goite in and Mill Miller er (198 (1983). 3). Fur Further ther enhancements enhancements are ar e bei being ng mad made e (Do (Doble blerr and Bendl, Bendl, 200 2002) 2) usi using ng CT and MR MRI. I. The mai main n dif differ ferenc ences es bet betwe ween en the

the beam is fixed in direction (usually horizontally) and an d it is th the e pa pati tien ent’ t’ss ey eye e th that at is di disp spla lace ced d an and d rotated so as to gaze at a small-diameter fixation light lig ht wit with h pol polar ar and axial location location kno known wn to the planning plann ing prog program ram.. The secon second d diff differen erence ce is tha that, t, although the eye is rotated, the rest of the patient remains rema ins fixed fixed.. The pat patient’ ient’ss ey eyelid, elid, then then,, rema remains ins fixed in space as the eye is rotated, and this must be modeled in the planning program. 6.11.1.3 6.1 1.1.3 Output of the the planning planning proc process ess

 As in treatmen treatmentt planning for general radiation therapy,, the treatme therapy treatment-planning nt-planning program generates the th e pa para rame mete ters rs ne need eded ed to de deli live verr a tr trea eatm tmen ent. t. These include the shape and size of the beam collimator (designed in the BEV), the beam penetration and modu modulati lation, on, and the eye eye-fixa -fixation tion poin point. t. Very importantly, the planning program provides images

importantly, the planning program provides images thatt sho tha show w ho how w th the e ra radio dio-op -opaqu aque e cli clips ps sut sutur ured ed to the th e po post ster erio iorr sc scle lera rall su surf rfac ace e sh shou ould ld ap appe pear ar in orthogonal radiographs (simulating the precise geometry of the alignment x-ray systems of the treatment apparatus). These clips are used both as an aid in the tr trea eatm tment ent set setup up and to dem demar arca cate te the tumor circumference; they are an essential element

specialized specialize d eye eye-tre -treatm atmentent-plann planning ing pro program gramss and those tho se use used d in gen gener eral al ra radia diatio tion n the thera rapy py are out out-lined in Sections 6.11.1.1–6.11.1.5. 6.11.1. 6.1 1.1.1 1

Model of the the patient patient anato anatomy my

Rathe herr th than an be beiing ba base sed d on a se sett of three ee-dimensional images from which pertinent anatomic featur fea tures es ar are e ex extr trac acted ted,, eit either her man manual ually ly or aut autoomatic ma ticall ally y, the pa patie tient nt an anat atomy omy is gen genera erated ted by fitti ting ng a ge gene nerral li libr brar ary y of two wo-- an and d three ee-dimension dime nsional al str structu uctures res to a varie variety ty of par paramete ameters rs obtained in a variety of ways. Typically, the library includes the exterior of the eye, modeled as a pair of spheres or, better, ellipsoids; the optic axis; the serrata ata; th iris; iri s; th the e len lens; s; the   ora serr the e op opti ticc di disc sc an and d nerve; the macula;   etc. The tumor is modeled as a three-dimensional volume bounded by a tumor-base circumference; by a proximal-tumor high point and a di dist stal al de deep ep po poin int, t, bo both th re rela lati tive ve to th the e in inne nerr surf su rfa ace of th the e scl cler era; a; and by hyp yper erb bol olic ic or near-hyperbolic sets of curves that connect both the

of the treatment process. 6.11.1.4 6.1 1.1.4 Pres Presentat entation ion of resu results lts

 A number of specialized dose displays are provided provided.. One of these is a plot of isodose contours on the curved retina, using the same projection as is pro vided by a fundus camera. This allows the dose displa dis play y to be dir direct ectly ly com compar pared ed wit with h the fun fundus dus pictures and, in principle, overlaid upon them (see Fig. 6.10 6.10b). b). The DVH DVHss and dose-area dose-area his histogr tograms ams (the cumulative frequency of dose on the surface of  a vo volum lume) e) ar are e ro routi utinel nely y emp employ loyed. ed. Ind Indeed eed,, the their ir firstt use in tr firs trea eatme tment nt pla plann nning ing wa wass for th the e pla plann-

high point and the deep point to the circumference. Figure 6.9a shows a lateral view of such a model. 6.11.1. 6.1 1.1.2 2

ning of eye treatments (Goitein and Miller, 1983). 6.11.1.5 6.1 1.1.5 Treat reatment ment exam example ple

Beam simula simulation tion

These concepts are illustrated7 by the example of a trea tr eatm tmen entt of a uv uvea eall me mela lano noma ma of th the e le left ft eye ye.. During Duri ng a surg surgical ical int interv erventi ention, on, fou fourr rad radioio-opa opaque que tantalum clips were sutured to the outside of the eye bulb around the base of the lesion, in order to make the location of the tumor visible in orthogonal x-ray images. Postoper Postoperative ative proton-radiation therapy was planned plan ned for fou fourr dai daily ly fr fract action ionss of RB RBE-w E-weig eighte hted d absorbed doses of 15 Gy (RBE), for a total dose of  60 Gy (RBE (RBE). ).

The be The beam amss us used ed in in intr trao aocu cula larr tr trea eatm tmen ents ts ar are e uniform-i unif orm-inten ntensity sity beam beamss and, because of the simplicity of the beams and of the eye, they are usually simulate simu lated d by simp simplist listic ic mode models, ls,   i.e., by a parameterized depth–dose curve, a measure of the lateral penumbra penu mbra,, and a cros cross-sec s-section tion determined determined by the geometric geome tric proj projectio ection n of the aper aperture ture.. Dose hete heterorogeneities are generally not taken into account; the tissue tis suess ar are e tak taken en to be equ equiva ivalen lentt to wa water ter.. Th The e treatment plan has to model a beam that is, ho howe wever ver, , dif differ ferent ent in two ways wa ys..geometry First Fir st,, ra rath ther er than tha n the bea beam m bei being ng ro rota tated ted aro around und th the e pa patie tient, nt,

7

This case has been worked up by J. Verwey and A. Bolsi, Paul Scherrer Institute, Switzerlan Switzerland. d.

118  

TREATMENT TREA TMENT PLANNING Table 6.1. Dose statistics for the three plans presented presented in Section 6.10 and Figs 6.7 and 6.8.

 Volume  V olume

RBE-weighted RBE-we ighted absorbed dose [Gy (RBE)] Scattered

Target volume—PTV  Dose to 98 percent of volume,  D 98%  (near-minimum dose) Median dose,  D 50%   Dose to 2 percent of volume,  D 2%  (near-maximum dose) Relative volume receiving 95 percent of the prescribed dose,  V 95% 95% Right femoral head Relative volume receiving 20 percent of the prescribed dose,  V  Relative volume receiving 50 percent of the prescribed dose,  V 20% 50% 50% Relative volume receiving 80 percent of the prescribed Dose,  V 80% 80% Dose to 2 percent of volume,  D 2%  (near-maximum dose)

       

Uniform scanned

IMPT

48.5 54.6 55.5 93

49.7 54.7 54 55.4 93

48.9 54.6 56.2 92

54.0 49.8 34.6 102

50.3 39.7 10.3 98

27.3 9.8 2.3 85

Intestines Relative volume receiving 20 percent of the prescribed dose,  V 20% 20% Relative volume receiving 50 percent of the prescribed dose,  V 50% 50% Relative volume receiving 80 percent of the prescribed dose,  V 80% 80% Dose to 2 percent of volume,  D 2%  (near-maximum dose) Body outline minus PTV  Mean dose outside the PTV,  D mean  

     

13.8 2.2 0.2 63

10.6 1.6 0.2 57

8.9 1.4 0.1 55

7.2

5.9

5.6

 Volumes are relat  Volumes relative ive volumes except for the ‘body-outline2 PT PTV’, V’, whi which ch is an abs absolu olute te vo volum lume e (in cm3). The me mean an dose dose to th the e ‘body-outline2 PTV’ is equal to the integral dose to that volume divided by its volume (which is 10.7l).

Proto Pr otonn-bea beam m tr trea eatm tment ent wa wass pla plann nned ed usi using ng an

befor bef ore e ea each ch tr trea eatm tment ent to com compar pare e the obs observ erved ed

enhanc enha nced ed ve vers rsio ion n of th the e pr prog ogra ram m de deve velo lope ped d by Goitei Goi tein n and Mil Miller ler (19 (1983) 83).. Re Recon const struc ructio tion n of th the e GTV was based on (i) for the tumor base: the position of the clips and the distance between the clips (measured (meas ured intra-oper intra-operativ atively) ely) and the base of the tumorr, as wel tumo welll as the meas measured ured distance distance between the tumor and the optic disc and macula; and (ii) forr th fo the e tu tumo morr he heig ight ht an and d sh shap ape: e: an ul ultr tras asou ound nd stud st udy y. Th The e di dire rect ctio ion n of ga gaze ze (b (bot oth h ca caud udal al an and d medial) was chosen to avoid irradiation of the optic disc di sc,, and as much as pos osssib ible le of the an ante teri rior or segm se gmen ent. t. Th The e be beam am ap aper ertu ture re wa wass de desi sign gned ed (se (see e 8 Fig. 6.9a) around the GTV, GTV, with a lateral margin of  2 mm to account for possible microscopic extension,

radio-opaque clip positions with their desired position ti onss as ca calc lcul ulat ated ed by th the e pl plan anni ning ng sy syst stem em.. An accuracy of 0.2 mm can be achieved. Treatment was delivered as planned and was well tolerated, with mild conjunctivitis at the end of treatment.

motion uncertainty, and beam penumbra, and with a margin in depth of 3 mm to allow for range uncertainties. The dose distribution achieved is shown in sagittal section in Fig. 6.9b, as well as in a projection mimicking the fundus camera’s geometry (Fig. 6.9c) to allo allow w dir direct ect comp compariso arison n with the fund fundus us pict picture ure (Fig. 6.9d). Dose–volume histograms for a number of   volumes  volu mes of inter interest est are shown in Fig. 6.9e. The patient was treated in a seated position, the head fixed in a perforated thermoplastic mask with a bite block, and with the affected eye gazing at a ligh li ghtt to ac achi hiev eve e th the e de desi sire red d or orie ient ntat atio ion n of ga gaze ze (Fig. (Fi g. 6.9 6.9ff ). Ort Orthog hogona onall ra radio diogra graphs phs we were re tak taken en

ov over ert mu multi ple porated frac fr actio tions. ns. This Th is out increa inc reased accur ac curac acy ymust mus be ltiple incorpora incor ted through thr oughout thesed whole planplan ning nin g and del delive ivery ry pr proce ocess, ss, inc includ luding ing imm immobi obililization, planning, alignment, and radiation delivery. Specific Spec ific impl implemen ementat tations, ions, such as usin using g a gant gantryrybased delivery system with a nonisocentric robotic bed, or using a custom isocentric patient positioner with a fixed horizontal beam line, introduce their own considerations when accounting for accuracy.

6.11.2 6.1 1.2

Stereotactic Stereot actic trea treatments tments with protons

The pr The proce ocess ss of pr proto oton n rad radios iosurg urgery ery is lar largel gely y the same sam e as for fr frac actio tiona nated ted pr proto oton n ra radio dioth thera erapy py..9 What Wh at is dif differ ferent ent is the need to ac achie hieve ve a hig higher her level of conformity because of the large doses prescribe scr ibed, d, oft often en adm admini inist ster ered ed in a sin single gle ses sessio sion, n, with the cons consequen equentt lack of sta statist tistical ical ave averagin raging g

6.11.2.1 6.1 1.2.1 Immobi Immobilizatio lization n

In prin principle ciple,, the conv conventi entional onal ste stereot reotact actic ic fixa fixation tion frames used for Gamma Knife and linac-based radiosurger sur gery y can als also o be us used ed for pr proto oton n rad radios iosurg urgery ery.

8

In thi thiss spe specia cializ lized ed tr trea eatm tment ent neither neither a CTV nor a PTV is expl ex plic icit itly ly dr draw awn. n. Tre reat atme ment nt ma marg rgin inss ar are e bu buil iltt in into to th the e aperture-design aperture -design process.

9

This se This sect ctio ion n was pr prep epar ared ed wi with th th the e he help lp of M. Bu Buss ssie ie`re, Massachusetts General Hospital, Boston MA, USA.

119  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

120  

TREATMENT TREA TMENT PLANNING

with a diagnostic-quality imaging platform for the treatm tre atment ent align alignment ment.. Wi With th this appr approach oach,, minor changes in the patient’s placement within the fixation ti on fr fram ame e th that at ma may y oc occu curr be betw twee een n th the e pa pati tien entt bein be ing g CT sc scan anne ned d an and d tr trea eate ted d ar are e no nott cr crit itic ical al becaus bec ause e the pa patie tient nt (no (nott th the e fr fram ame) e) pr provi ovides des th the e

reference referenc e coord coordinat inates. es. How However ever,, it is cruci crucial al tha thatt the frame keep the patient securely immobilized.

6.11.2 6.1 1.2.2 .2 Ima Imaging ging  A requirem requirement ent for proton-t proton-thera herapy py planning planning for stereostereotactic tac tic trea treatmen tments ts is CT CT-base -based d imagi imaging. ng. In order to maxi ma ximi mize ze re reso solu luti tion on,, th the e sc scan anne ner’s r’s fie field ld of vi view ew should be minimized while, nevertheless, encompassing any material that could potentially be traversed by the proton beams. The CT-slice thickness should also als o be min minimi imized zed thr throug oughou houtt the hea head d (  1 mm), which whic h requ requires ires the comp computer uter tre treatm atment-p ent-plann lanning ing program pro gram to mana manage ge close to 200 axial axial slices. Forr ta Fo targ rget et vi visu sual aliz izat atio ion, n, a CT CT-a -ang ngio iogr grap aphy hy appro app roac ach h sho should uld be use used, d, adj adjus ustin ting g the tim time e and rate of injection of the contrast agent to optimize the th e vis visual ualiza izatio tion n of the ar area ea of int inter erest est.. The pr preesence of una unaccoun ccountedted-for for cont contras rastt mat material erial woul would d result in an incorrect estimate of the needed proton ranges. It is wise to obtain a baseline noncontrast scan sca n to ens ensur ure e pr prope operr con conver versio sion n of the CT den den-siti si ties es to pr prot oton on ran ange ges. s. Wh When en ne nece cess ssar ary y, th the e contr con tras ast-e t-enh nhanc anced ed CT st stud udy y use used d to out outlin line e th the e target(s) targe t(s) and structure( structure(s) s) of inter interest est can be fused to th the e non noncon contr tras astt CT study study use used d for dos dose e cal calcuculations. This is especially important when treating a vascular lesion such as an arteriovenous malformation mat ion (AVM) VM).. Othe Otherr imagi imaging ng moda modalitie lities, s, such as MRI, MR I, PE PET T, and pla planar nar ang angiog iogra raphy phy, can als also o be fused, to supplement the CT information.

Figure 6.10. A CT scan of a commercially commercially available stereotactic stereotactic frame fra me (lef (left) t) tha thatt pro provides vides a secu secure re fixation but woul would d pre preven ventt the use of posterior and posterior–oblique beams because of the high-density occipital pad and plastic-aluminum support. A CT scan of a frame modified for use with protons (right) that uses a thin carb carbon on suppo support rt alon along g with low low-den -density sity cush cushion, ion, enab enabling ling any port portal al dir directi ection on with only smal smalll dept depth h vari variati ations ons acr across oss individu indi vidual al tre treatm atment ent beam beams. s. (Fig (Figure ure cour courtes tesy y of M. Buss Bussie ie`re, Massachusetts General Hospital, Boston, MA, USA.)

These fr These frame amess pr prov ovide ide bot both h imm immobi obiliz lizat ation ion and a reference coordinate system. The use of the ‘built-in’ coordi coo rdina nates tes imp implie liess tha thatt th the e pr proce ocesse ssess ass associ ociat ated ed with immo immobiliz bilizati ation, on, CT scann scanning, ing, tre treatm atment ent plan plan-ning, nin g, an and d tr trea eatm tment ent can be ac accom compli plishe shed d on th the e same sa me da day y. Ho How wev ever er,, th this is ma may y no nott be pr pra act ctic ical al becau bec ause se of th the e tim time e con const stra raint intss inv involv olved ed wit with h the overall ove rall quali quality-a ty-assur ssurance ance pro program gram and the fabr fabricaication tio n of cu cust stom om dev device icess su such ch as ape apertu rture ress and compensators, or for fractionated treatments. The frames are also desi designed gned for isoce isocentri ntric-cou c-couch ch sys systems tems,, and there the refor fore e add additi itiona onall eff effort ort is re requi quire red d whe when n usi using ng no noni niso soce cent beds be ds..um If pin inva in vasi vether fixat fix ion n is inless used us ed,, cerami cer amic c ntri orriccalu alumin minum pins, s, sive rathe ra r atio than tha n sta stainl ess steel pins, will reduce artifacts that would otherwise change the CT densities necessary for range conversions.. Com sions Commer merciall cially y, ava availabl ilable e fram frames es are like likely ly to need modifications in order to control the type and amoun amo untt of ma mater terial ial tha thatt pr proto otons ns wou would ld en encou counte nterr proximal to the patient (Fig. 6.10).  An alternativ alternative e immobilization method is to use a reliab rel iable le non noninv invasi asive ve fr fram ame e wit with h an ind indepe epend ndent ent et al., coor co ordi dina nate te sy syst stem em fo forr al alig ignm nmen entt (G (Gal alll   et 1993 19 93b) b).. Us Usin ing g th thre ree e or mo more re me meta tall llic ic sp sphe here ress inserted in the outer table of the skull provides a reliable and accurate coordinate system when used

6.11.2 6.1 1.2.3 .3

Planning Plan ning

Most pl Most plan anss us use e th thre ree e or fo four ur be beam am di dire rect ctio ions ns;; however, as many as six can be routinely used for a single target. The shape and location of the target will dictate the beam directions. This is especially true for AVMs, which tend to have irregular shapes and an d ma may y be si situ tuat ated ed an anyw ywhe here re th thro roug ugho hout ut th the e brain. bra in. Stan Standard dard beam combi combinat nations ions can, how however ever,, be used for more routine targets such as pituitary adenomas aden omas and acou acoustic stic neur neuromas omas (App (Appendi endix x B.4) B.4).. Forr ex Fo examp ample, le, a st stand andard ard app approa roach ch for pit pituit uitary ary adenomas, with a typical prescription of 

Figure 6.9. (a) Beam’s-eye Beam’s-eye view of the eye, showing the design of the aperture. (b) Isodose lines in a plane through the tumor center. The portions of the eye structures behind the plane are suppressed. (c) Computer rendering of isodose lines on the curved retina. This image has the same scale as the fundus picture of the eye (d), with which it can be accurately compared. (e) Dose–volume histograms of the dose to various structure structuress with within in the eye. (f ) Pho Photogr tograph aph of a (dif (differ ferent) ent) patient patient in trea treatmen tmentt posit position. ion. (Figure (Figure cour courtesy tesy of  J. Verwey, Paul Scherrer Institute, Villigen, Switzerland.)

121  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

RBE RBE-wei -weighte ghted d absor absorbed bed dose dosess of 18– 20 Gy (RBE (RBE)) to the PTV;   , 8 Gy (RBE) to the optic chiasm and

nerves, and   , 12 Gy (RBE) to the brainstem, is to nerves, use fou fourr con conform formal al bea beams: ms: two lat later eral al bea beams, ms, one from fr om th the e le left ft,, an and d on one e fr from om th the e ri righ ght, t, pa pass ssin ing g through the temporal lobes and avoiding the brainstem ste m and opt optic ic str struct uctur ures; es; a pos poster terior– ior– sup superio erior– r– oblique obli que app approa roach ch pass passing ing thr throug ough h the brai brainste nstem m but av avoid oiding ing the tem tempor poral al lob lobes es and op optic tic st struc ruc-tures; and an anterior–superior–oblique approach, passing passi ng thr through ough the opti opticc stru structur ctures es but av avoidin oiding g the brainstem and temporal lobes. Collimation Collimat ion is determined using a BEV approa approach. ch. The exact margins will depend on the penumbra (a beamline-specific parameter) and the dose normalization. A standard practice is to normalize to the 90 percent isodose. Brass apertures provide the collimat lim ation ion,, and com compen pensa sator torss pr provi ovide de dis distal tal dos dose e shap sh apin ing. g. Mo Modu dula lati tion on ca can n be ac achi hiev eved ed us usin ing g a  variety of techniques including the use of a rotatin rotating g absorber propeller, or the superposition of discrete Bragg peaks (lamination). Lateral dose equilibrium is lost small beams, resulting a change in dept de pth– h–for dose do se profi pr ofile le (F (Fig ig.. 6. 6.1 11) 1)..in Th This is mu must st the be modeled in the planning algorithm and accounted for in the design of the SOBP. Because the goal is to have tight margins, it is import imp ortant ant to av avoid oid to th the e ext extent ent pos possib sible le ha havin ving g beams go thr through ough heterogeneou heterogeneouss regio regions, ns, as this has the effect of increasing the required modulation and broadening the penumbra. Also, unlike fractionated na ted tr trea eatm tment ents, s, pa patch tching ing and abu abutti tting ng bea beam m combinations are not used because of the resulting dose inhomogeneity that can result. 6.11.2.4 6.1 1.2.4 Trea reatment tment

Treatment sessions involve both alignment verification tio n and dos dose e del deliv ivery ery.. Be Becau cause se of the their ir we weigh ight, t,

Figure Figur e 6.1 6.11. 1. Dept Depth– h– dos dose e cu curv rves es of 170 MeV pr proto otons ns for a number of circularly shaped beams of diameters 12, 14, 16, 20, 30, and 50 mm. Belo Below w 12 mm diam diameter eter,, the peak peak-to-to-pla plateau teau rati ra tio o ra rapi pidl dly y de decr crea ease ses. s. (F (Fig igur ure e co cour urte tesy sy of M. Bu Buss ssie ie`re, Massachusetts General Hospital, Boston, MA, USA.)

large gan large gantri tries es ma may y ha have ve an iso isocen centri tricit city y tha thatt is sub-opti suboptimal mal for rad radiosur iosurgical gical toler tolerances ances.. In such cases, cas es, th the e pa patie tient nt pos positi itione onerr can be adj adjus usted ted to corr co rrec ectt fo forr th the e ga gant ntry ry-a -ass ssem embl bly y de defle flect ctio ions ns..  Alignment verification will depend on the type of  patient pat ient immo immobiliz bilizatio ation n (st (stereo ereotact tactic ic vers versus us noni noninn vasive with implanted markers) used, as well as the th e bea beam-d m-deli eliver very y met method hod (gant (gantry ry ver versus sus fixe fixed d beam be am an and d is isoc ocen entr tric ic be bed d ve vers rsus us no noni niso soce cent ntri ricc robotic robo tic posit positioner) ioner).. A pra practica cticall appr approach oach is to use diag di agno nost stic ic-q -qua uali lity ty xx-ra ray y im imag agin ing g to im imag age e th the e patient pat ient (with internal internal mark markers) ers) or the frame with inte in tegr grat ated ed ma mark rker ers. s. Al Alig ignm nmen entt ve veri rific ficat atio ion n is usually done for every beam. Typical treatment sessions last from 30 to 40 min. The proton-beam dose rate is generally from 2 to 6 Gy min 21.

122  

Journal of the ICRU Vol 7 No 2 (2007) Report 78 Oxford University Press

7

MOTI MO TION ON MA MANA NAGE GEME MENT NT

7.1

MOTION OF OF, AND WITHIN, WITHIN, THE PA PATIENT

The patient is not an inanimate object, but a living, breathing brea thing,, and movi moving ng indi individu vidual. al. The deliv delivery ery of  radiati radi ation on to the target volu volume me mus mustt invo involve lve first, an ef effo fort rt to mi mini nimi mize ze,, to th the e ex exte tent nt re reas ason onab ably ly possible, the motion and its uncertainties, and then taking tak ing the re remai mainin ning g mot motion ion an and d unc uncert ertain aintie tiess into account. Tum umor or and org organ an mot motion ion can be cla classi ssified fied int into o three categories (Langen and Jones, 2001), namely, motion of the patient as a whole (relative to some refe re fere renc nce e ob obje ject ct su such ch as th the e co couc uch h to top) p),, in inte terrfraction fra ction motion of organ organss with within in the patient, patient, and intra-fr intr a-fract action ion moti motion on of organ organss withi within n the pat patient ient during delivery of a single fraction.

doi:10.1093/jicru/ndm030

†

sated for is important. If the potential misregistra tr ati tion on is ta tak ken in into to co cons nsid ider era ati tion on du duri ring ng compensa comp ensator tor desig design n (see Sect Section ion 6.4.6 6.4.6.2), .2), then then,, the greater the possibility of misregistration, the more mor e th the e com compen pensat sator or has to be ex expan panded ded and the greater greater the vo volum lume e of dis distal tal tissue tissue tha thatt is irradiated. If the potential misregistration is not taken into account, any mismatch between compensat pen sator or and het heter eroge ogenei neitie tiess cou could ld lea lead d to an under-dose within the target volume.

7.2 SUP SUPPOR PORT T AND IMMO IMMOBILI BILIZA ZATION TION It is common to use some method of immobilization to better relate the patient to the treatment equipment me nt.. In so some me sp spec ecia iall ca case sess (s (see ee,, fo forr ex exam ampl ple, e, Sectio Sec tions ns 3.2 3.2.3. .3.1 1 and 3.2 3.2.3. .3.2), 2), the imm immobi obiliz lizat ation ion device is built into the equipment. Usually, a separate device is used and placed upon the couch top or treatment chair, often being indexed to it through the use of loc locat ating ing pin pins. s. Mec Mechan hanism ismss for pa patie tient nt support have been described in Section 3.2.2.  An immobilization device is used to hold the patient as a whole in a stable and near-motionless positi pos ition on dur during ing bot both h ima imagin ging g an and d tr trea eatme tment. nt. By doing doi ng so, th the e loc locat ation ionss of int intern ernal al org organs ans and the tumor are also constrained. Immobilization devices also al so se serv rve e to mi mini nimi mize ze th the e pr prob oble lem m th that at ar aris ises es should shou ld the pat patient ient’s ’s posit position ion durin during g trea treatmen tmentt be differ dif ferent ent fr from om the pa patie tient’ nt’ss po posit sition ion du durin ring g the planning CT. 7.2.1. Pro 7.2.1. Protonton-speci specific fic aspec aspects ts of  of  immobilization

Immobilization is particularly important in proton therapy for the following reasons: †

  The The goal of pro protonton-beam beam therapy therapy is to conf conform orm the th e tr trea eatm tmen entt vo volu lume me to th the e cl clin inic ical al ta targ rget et

 volume (CTV) as closely as possible. Protons do this th is in th the e de dept pth h di dire rect ctio ion, n, an and d th the e do dose se is rela re lati tive vely ly in inse sens nsit itiv ive e to mo moti tion on al alon ong g th the e beam direction. However, the dose conformation in th the e pl plan ane e pe perp rpen endi dicu cula larr to th the e be beam am is sens se nsit itiv ive e to la late tera rall mo moti tion ons; s; he henc nce e it ne need edss immobilization.   Heter Heterogen ogeneitie eitiess are an issu issue e for prot protons, ons, much more mo re th than an fo forr ph phot oton onss (s (see ee Se Sect ctio ion n 6. 6.4) 4).. Th The e regis reg istr trat ation ion of the com compen pensa sator tor (ei (eith ther er re real al or  virtual) with the heterogeneities to be compen-

Immobilization Immobiliza tion has been used in x-ra x-ray y ther therapy apy for man any y dec eca ade des, s, an and d man any y te tech chn niq ique uess of  mou moulag lage e d.orHow other oth er phy physic sical al con const stra raint int two have ha vespeci been bee n develope dev eloped. However ever, , pro protons tons present pre sent special al issues: †

†

 With passively spread beams, optimization of the penumbra demands that the aperture and compensat pen sator or be as clo close se to th the e pa patie tient nt as pos possib sible le (con (c onsi sisste tent nt wi with th avoi oidi ding ng ho hott sp spot otss du due e to aperture-edge scattering). This means that bulky immobilization devices can be problematic.  With photons, material on the skin surface can spoil spo il ski skin n sp spari aring. ng. Bu But, t, pr prot otons ons pr prov ovid ide e vir vir-tuall tu ally y no ski skin n sp spari aring, ng, and hen hence ce the there re is no need to avoid material on the skin surface. On the other other ha hand nd,, all ma mate teria riall in th the e bea beam m pa path th affec af fects ts the bea beam m pen penetr etrat ation ion.. Al Allow lowan ance ce for change cha nge in pen penetr etrat ation ion re requi quire ress kno knowle wledg dge e of  thickn thi ckness ess and com compo posit sition ion of su such ch ma mater terial ial.. When Wh enev ever er po poss ssib ible le,, ha havi ving ng im immo mobi bili liza zati tion on material in the path of a proton beam should be avoided.

# International Commission on Radiation Units and Measurements 2007

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

7.2. 7.2.2 2

Immobiliz Immo bilizati ation on tech techniqu niques es

 A review of immobilization methods in radiation oncolo onc ology gy is giv given en by Verh erhey ey and Be Bente ntell (19 (1999) 99).. Many types of immobilization devices are available, including bite-block/head-rest combinations for stabilizing the head, partial-body casts for stabilizing the thorax or pelvi pelvis, s, and whole-body whole-body cast casts. s. Cas Casts ts can be made  inter alia  from plaster of Paris, using conventional moulage techniques, from thermoplastic she sheets ets that ar are e dr drape aped d ov over er the pa patie tient nt whi while le warm and become firm upon cooling, and from bags of foam pellets that are made rigid by being placed under vacuum once the bag is made to conform to the patient’s surface. Immobilization techniques for special procedures are discussed in Section 3.2.3. Many Man y immob immobiliza ilization tion techn technique iquess wer were e init initially ially intr in trod oduc uced ed in pr prot oton on th ther erap apy y, an and d ha hav ve su subbsequ se quen entl tly y fo foun und d a pl plac ace e in co conv nven enti tion onal al xx-ra ray y therapy. These include the following: Stereotactic Stereota ctic head holder. holder. The fixation of the head by attaching a stereotactic frame using pins set into burr-holes made in the skull was first used in ext extern ernal al pr proto oton-b n-beam eam ra radio dioth thera erapy py for the irradiation of intracranial targets,   e.g., pituitary gland (Kjellberg  et al., 1962a). Perforated Perfora ted thermoplastic head masks. A warm, perfora fo rate ted d th ther ermo mopl plas asti ticc sh shee eett is fo form rmed ed to th the e patient’s head and allowed to set by cooling. The perforations serve to keep the head cool and to be less claustrophobic for the patient. The sheet is captured in a tennis racket-shaped frame that is attached to the couch top or chair with the use of  indexing pins (Verhey  et al., 1982). U-sha Ushaped ped ‘po ‘pod’, d’, ind index exed ed to the tr trea eatm tment ent cou couch, ch, supporting a foam-pellet bag. This technique was firstt dev firs develope eloped d for   p-mes -meson on ther therapy apy,, and then carried over to proton-beam therapy where it has been used at several proton therapy centers (von Essen et al., 1982).  Vacuum-fixed  V acuum-fixed bite block. The use of a bite block fixed to the treatment equipment ( e.g., the couch top), top ), whi while le usu usuall ally y pr prov ovidi iding ng goo good d imm immobi obiliz liz-ation, suffers from the problem that it is of questionable value in the edentulous patient and can place pla ce som some e st stra rain in on the pa patie tient. nt. Th This is pr probl oblem em hass be ha been en la larg rgel ely y ov over erco come me th thou ough gh th the e us use e of   vacuum suction on a bite block made to conform to the patient’s palate (Schulte  et al., 2000).

7.3 LOCA OCALIZA LIZATION TION Once the patient has been adequately immobilized, the th e ta targ rget et vo volu lume me in sp spac ace e is lo loca cate ted, d, re rela lati tive ve

the patient be located reproducibly relative to the trea tr eatm tmen entt eq equi uipm pmen entt an and d (i (ii) i) th that at th the e ta targ rget et  volume be in a known spatial relationsh relationship ip to the patient pat ient.. The latter is gene generally rally based on imagi imaging ng studies, as discussed in Section 6.3. There are four gene ge nera rall ap appr proa oach ches es to lo loca caliz lizat atio ion, n, wh whic ich h ar are e described below. 7.3.1 7.3. 1

Localiza Loca lization tion base based d on skin skin marks marks

In some circumstances ( e.g., cancers of the skin or lip), the target volume is best placed by the localization of overlying skin. The usual method of localization would then be to adjust the patient’s position unti un till a li ligh ghtt fie field ld co coin inci cide dent nt wi with th th the e ra radi diat atio ion n beam is aligned with the marks on the skin. 7.3.2 7.3. 2

Localiza Loca lization tion based based on on bony ana anatomy tomy

Because of the spatial accuracy Because accuracy desired in pro proton ton therapy, it is usual to relate the target volume to bony bon y lan landm dmark arkss ra rath ther er tha than n to ski skin n mar marks. ks. Fo Forr this reason, the use of laser beams in the treatment room ro om,, as is of ofte ten n do done ne in co conv nven enti tion onal al xx-ra ray y therapy, is used only as an initial step in the localization process. The localization of the target volume relative to the trea treatmen tmentt equip equipment ment based on bony anatomy anatomy proc pr ocee eeds ds in tw two o st step eps: s: (i (i)) th the e ta targ rget et vo volu lume me is locat loc ated ed re rela lativ tive e to the bon bony y ana anatom tomy y and (ii (ii)) the bony anatomy anatomy is loca located ted rela relative tive to the tre treatm atment ent equipm equ ipment ent.. The firs firstt st step ep is ac accom compli plishe shed d in the treatment-planning process, based on the planning CT study. Once the planning target volume (PTV) has bee been n del deline ineat ated, ed, the pla plann nning ing pr proce ocess ss det deterermines the beams to be used and the central axes of  these are generally aimed toward a point in space, usually usua lly the isocen isocenter ter.. The plan planning ning proc process ess then establishes the location of the aiming point relative to selected features of the bony anatomy. The most common way of locating the bony landmarks relative to the patient support system is to compar com pare e the ali alignm gnment ent ra radio diogr graph aphss tak taken en in the treatment room with digitally reconstructed radiographs (DRRs) (Goitein  et al., 1983b) computed for the same viewpoints. The DRR is a computer simulation, based on an imaging study (in this case, the planning CT study) that, by projecting and accumulating the CT voxel data onto a virtual film plane, simula sim ulates tes a rad radiog iograp raph, h, tak taken en fr from om any poi point nt of   view,   e.g., fr  view, from om th the e be beam am’s ’s-e -eye ye vi view ew or fr from om th the e  viewpoint of an x-ray tube located at a well-defined point within the trea treatmen tmentt room room.. The local localizat ization ion proce pr ocess ss gen genera erally lly con consis sists ts of mo movin ving g the pa patie tient nt until unt il an ort orthog hogona onall pai pairr of ra radio diogra graphs phs has the same spa spatial tial rela relations tionship hip to the trea treatmen tmentt equip equip--

to the treatment equipment. This requires (i) that

ment me nt as th the e pa pair ir of co corr rres espo pond ndin ing g DR DRRs Rs.. In 124

 

 MOTION MANAGEMENT 

particularr, the loca particula location tion of the bony ana anatomy tomy relative tiv e to a cro cross ss hai hairr (wh (which ich es estab tablis lishes hes the coo coordi rdi-nate system of the radiograph) is required to be the same sa me in th the e al alig ignm nmen entt ra radi diog ogra raph ph(s (s)) as in th the e DRR(s). The process of establishing this correspondence den ce can be do done ne man manua ually lly or or,, mor more e obj object ective ively ly and in principle faster, by using a computer (Sharp  et al., 2005).  Alternatively  Alternativ ely,, the radiograph radiographic ic information obta ob tain ined ed ju just st pr prio iorr to tr trea eatm tmen entt ca can n be us used ed to comp co mpar are e th the e pa pati tien ent’ t’ss po posi siti tion on re rela lati tive ve to th that at required for the plan. This information could come from fr om ra radio diogra graph phss tak taken en usi using ng a pa pair ir of (us (usual ually) ly) orthogonal x-ray tubes, or from scout views from a CT scanner. In a recent development, a set of conebeam bea m CT ima images ges is ac acqui quire red d usi using ng an x-r x-ray ay tub tube e moun mo unte ted d on th the e tr trea eatm tmen entt ga gant ntry ry (o (orr a ne near arby by cone-beam CT device) (Jaffray, 2003). These images can be compared with the CT image set used for planning, plan ning, and the geome geometric tric diff differen erences ces betw between een the bony (or other) other) ana anatom tomies ies in the two st studi udies es can be used to compute a positioning correction. 7.3.3 Loca 7.3.3 Localiza lization tion rel relati ative ve to to the immobilization device

When an im When immo mobi bili liza zati tion on de devi vice ce or lo loca cali liza zati tion on frame (see Sections 7.2.2 and 3.2.3) is used, fiducial marker mar kerss emb embedd edded ed in th the e loc locali aliza zatio tion n dev device ice or frame can serve in the same fashion as bony landmarks as described in Section 7.3.2. Because fiducial ci al ma mark rke ers can ge gen ner eral ally ly be lo loca cate ted d very accurat accu rately ely,, locali localizat zation ion in such cases can be more accurate than when using bony landmarks. 7.3.4 Loca 7.3.4 Localiza lization tion based based on on identifica identification tion of  of  target-volume markers or the tumor itself 

In som some e cir circum cumst stan ances ces,, ra radio diogra graph phica ically lly vis visibl ible e objects such as gold seeds and surgical clips can be embedd emb edded ed in or clo close se to the tumor; tumor; for example example,, gold seeds have been introduced into the prostate al.., 19 for tum tumor or loc locali aliza zatio tion n (Sh (Shipl ipley ey   et al 1979 79). ). In some som e cas cases, es, the gro gross ss tu tumor mor vo volum lume e (GT (GTV) V) its itself  elf  may be visible, for example, when using ultrasound to locate the prostate just before treatments. Such techn tec hniqu iques es pr prov ovide ide for acc accur urat ate e tar target get-v -volu olume me localization. The localization process in the case of  radi ra dioo-op opaq aque ue ma mark rker erss fo foll llo ows th tha at fo forr bo bony ny landmarks.

desirable to veri desirable verify fy,, aft after er the trea treatmen tmentt has taken plac pl ace, e, wh whet ethe herr or by ho how w mu much ch th the e pa pati tien entt ha hass moved mov ed duri during ng trea treatmen tment. t. This may pro provide vide valu valu-able information information on the efficacy of the immobilizimmobilization techniques (Verhey and Bentel, 1999; Verhey  et al., 1982). 7.4.1

Verificatio erification n using radiogr radiography aphy

There is generally no analogy to the use of portal films in pro protonton-beam beam ther therapy apy,, becau because se the pro proton ton beam usually does not penetrate the entire patient, and so it cannot be detected in an external detector. If pr proto otons ns are ene energe rgetic tic eno enough ugh to pen penetr etrat ate e th the e patient, they cannot be imaged by a simple detector such as a film as their intensity is largely uncha un change nged d by the ma mater terial ial thr throug ough h whi which ch the they y have ha ve pa pass ssed ed (s (see ee Se Sect ctio ion n 6. 6.4) 4).. Th Thus us,, th the e ra radi dioograp gr aphi hicc ve veri rific ficat atio ion n of th the e pa pati tien ent’ t’ss po posi siti tion on is usually based on an x-ray radiograph taken along the central axis of the beam (either along the beam dire di rect ctio ion n or or,, us usin ing g an xx-ra ray y tu tube be di dist stal al to th the e pati pa tien ent, t, po poin inti ting ng ba back ck to tow war ard d th the e so sour urce ce of  proto pr otons) ns).. The ra radio diogr graph aph th thus us obt obtain ained ed is com com-pared with a DRR computed from the radiographic  viewpoint by the treatm treatment-planning ent-planning program. It is also al so fe feas asib ible le to em empl ploy oy a pa pair ir of xx-ra ray y tu tube bess moun mo unte ted d in th the e tr trea eatm tmen entt ro room om an and d di dire rect cted ed tow to war ard d th the e is isoc ocen ente terr to pr prov ovid ide e flu fluor oros osco copi picc imagin ima ging g and hen hence ce re realal-tim time e loc locali aliza zatio tion n du durin ring g treatment, which can be used for adjusting the position iti on of th the e pa patie tient nt re rela lativ tive e to the beam in rea realltime. tim e. Th These ese x-r x-ray ay tub tubes es nee need d not nec necess essari arily ly be directed dir ected orth orthogona ogonally lly to one anot another her (Sch (Schweik weikard ard  et al., 2004). Proton-beam radiography has been proposed, primarily for the verification of heterogeneities within the th e pa patie tient nt whi while le in th the e tr trea eatme tment nt pos positi ition on but also, als o, pot potent ential ially ly,, to ass assis istt in pa patie tient nt loc locali aliza zatio tion n (Schn (S chneid eider er and Pe Pedr droni oni,, 199 1995). 5). Th This is tec techni hnique que requires highly specialized apparatus and has not yet found a place in clinical practice. 7.4.2 Verificatio erification n using positron emission tomography

Energeti Energ eticc pr proto otons ns an and d oth other er heavy heavy cha charge rged d par par-ticless und ticle undergo ergo collis collisions ions with atom atomic ic nucle nuclei, i, some of which result in the formation of a positron posit ron-emit -emitting ting isoto isotope. pe. Posi Positron tron emiss emission ion tomogra og raph phy y ha hass be been en us used ed to ve veri rify fy th the e lo loca cati tion on of  light ion (12C) beams (Enghardt  et al., 1999; 2004).

Experimental Experime ntal and theo theoreti retical cal evalu evaluatio ations ns indi indicate cate that the same technique works for protons as well (Hishikawa   et al., 2002; Nishio   et al., 2005; Parodi al.., 200 and Engh Enghardt ardt,, 2000; Par Parodi odi   et al 2002; 2; 200 2007a, 7a, 2007b; 2007c; Vynckier  et al., 1993). The technique

7.4 VER VERIFIC IFICA ATION Once the patient has been positioned for treatment, it is desirable to verify the alignment of the beam rela re lati tive ve to th the e ta targ rget et vol olum ume. e. It ma may y al also so be 125  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

suffers from some disadvantages, the most fundamental of which is that there is usually a physiologic log ic wa washo shout ut of the ind induce uced d ac activ tivity ity du durin ring g the time tim e per period iod nee needed ded to col collec lectt the emi emissi ssion on da data ta   et al. al . (Maccabee , 196 1969). 9). Ano Anothe therr pr probl oblem em is tha thatt

reconstructio reconstr uction n due to image blurring blurring as has been demonstrated by Chen  et al.   (2004). CT techniques have ha ve bee been n dev develo eloped ped,, usi using ng sin single gle-- or mul multiti-sli slice ce CT scanners, which have made it possible to obtain multiple sets of CT images that are correlated with

the cross-section for the activation is not constant alon al ong g th the e pa path th le leng ngth th,, an and d in fa fact ct th ther ere e is a threshol thr eshold d energ energy y such that ther there e is no acti activat vation ion near the end of range. Thus the three-dimensional distribution, even with perfect beam delivery, does not exactly match the dose distribution; this is to some som e ex exten tentt ov over ercom come e by loo lookin king g for any dis discre cre-pancy between the observed and calculated distributions of activity.

well-defined well-defi ned phas phases es of the respiratory respiratory cycle using an internal or external gating system (Ford   et al., al.., 20 al.., 200 2003; 200 3; Lo Low w   et al 2003 03;; Pan   et al 2004). 4). Th These ese so-cal socalled led fou four-d r-dime imensi nsiona onall CT sca scans ns can pr prov ovide ide basic data for planning treatments to largely overcome com e mot motion ion art artifa ifacts cts and to sel select ect phases phases of the breathing cycle when motion is at a minimum.

7.5 OR ORGAN GAN MO MOTIO TION N

Motion Moti on of th the e pa pati tien entt as a wh whol ole e wi will ll na natu tura rall lly y result in motion of internal structures. This motion is min minimi imized zed by ade adequ quat ate e imm immobi obiliz lizat ation ion of the patient, as discussed in Section 7.2, and will not be further discussed here.  A number of studies have documented the extent of motion of several organs, and a comprehensive review has been given by Langen and Jones (2001). Typically, the extent of motion can vary from a negligible amount, to excursions of several centimeters or so in tissues near to, or influenced by, diaphragmatic mo mov vem emen entt. In the ab abssen ence ce of sp spec ecia iall measures, the only way to deal with situations in which large excursions can occur is to allow generous mar margin ginss in del deline ineat ating ing bot both h the PT PTV( V(s) s) and PRVs. It is likely that if the extent of motion or the

Organs and tissues move within the body, both over the cou course rse of the ent entir ire e th thera erapy py (in (inter ter-fr -frac actio tion n motion) and during the delivery of a single fraction (intra-fraction (intra-fr action motion). Inter-fra Inter-fraction ction movemen movementt of  the th e tu tum mor and nd/o /orr org rgan anss on a da dayy-to to-d -da ay or week-to-week basis can, for example, be related to chan ch ang ges in bowel or bl blad add der fil filli lin ng, tu tum mor regr re gres essi sion on,, or ch chan ange gess in th the e pa pati tien ent’ t’ss we weig ight ht.. InterInt er-fr fract action ion mot motion ion is fr frequ equent ently ly a pr probl oblem em in trea tr eati ting ng pe pelv lvic ic tu tumo morr si site tes. s. It is a pa part rtic icul ular ar problem for proton-beam therapy because variable bowel or bladder filling can affect, by centimeters, the penet penetrat ration ion of prot protons ons pass passing ing thr through ough those organs. IntraInt ra-fr frac actio tion n mot motion ion can occ occur ur on a ra range nge of  time tim e sca scales les.. Mot Motion ion cau caused sed by the beating beating of the hear he artt is pe peri riod odic ic in na natu ture re,, wi with th a cy cycl cle e ti time me of   1 s; motion caused by respiration is periodic, with a cycle time of    4 s; motion caused by peristalsis is aperiodic and can take place over time scales of up to 1 min. Of these motions, respiration is probably of greatest importance as it can produce quite large disp di spla lace ceme ment ntss an and d ca can n af affe fect ct th the e or orga gans ns in th the e abdome abd omen n as we well ll as in the thorax thorax (Go (Goite itein, in, 2005; Langen Lan gen and Jon Jones, es, 200 2001). 1). Re Respi spira ratio tion n can som someetime ti mess re resu sult lt in ex excu curs rsio ions ns of or orga gans ns of se seve vera rall centim cen timete eters, rs, ev even en if th the e org organ an is som some e dis distan tance ce from the diaphragm ( e.g., kidney).

7.5.2 Orga 7.5.2 Organ n motion motion in the the absence absence of speci special al measures

artifacts that produces aremight underappreciated, the probabili prob ability ty of itlocal control be comp compromi romised sed (Ling et al., 2004). 7.5.3 Orga 7.5.3 Organ n motion motion under under cond condition itions s of  respiration gating 

The most obvious and simplest way to handle respirat pir atory ory mot motion ion is to tr track ack the re respi spira rator tory y cyc cycle, le, identi ide ntify fy th the e ph phase ase(s) (s) wh when en mot motion ion is lea least st,, and turn the beam off (gate the treatment) during the other phases. al..,19 Respira Res piratory tory gat gating ing (Oha (Ohara ra   et al ,1989) 89) use usess an extern ext ernal al br brea eathi thing ng mon monito itorr to ga gate te the rad radia iatio tion n beam be am on an and d of offf at a we well ll-d -defi efine ned d ph phas ase e of th the e

7.5.1 7.5. 1

The measur measureme ement nt of organ organ moti motion on

breath brea thin ing g cy cycl cle, e, th the e ex exte tern rnal al mo moni nito torr re read adin ing g having been previously correlated with tumor position. An example of a suitable monitor is a lightemitti emi tting ng dio diode de pla placed ced on the pa patie tient’ nt’ss abd abdome omen, n, the th e pos osit itio ion n of whic ich h is mon onit itor ored ed by vi vid deo camera cam erass whi while le th the e pa patie tient nt br brea eathe thess fr freel eely y. The diode movement is checked to ensure that its position is correlated with the phase of the breathing cycle cyc le an and, d, hen hence, ce, wit with h the location location of th the e tar target get

The problems caused by organ motion arise during imaging, simulation, and treatment. During simulatio la tion n wit with h pla plane ne film films, s, the images images mig might ht not be representative of the tumor position, as they are a single sin gle sho short rt ex expos posure ure taken taken at one time poi point nt in the breathing cycle. For ma many ny ot othe herr fo form rmss of im imag agin ing, g, an and d fo forr CT simu si mula lati tion on,, th the e sc scan anss ca can n pr prod oduc uce e a di dist stor orte ted d 126  

 MOTION MANAGEMENT 

 volume. The diode position can be used to gate CT or flu fluoro orosco scopic pic da data ta ac acqui quisit sition ion,, an and d can be use used d during dur ing tr trea eatme tment nt to gat gate e the ac accel celera erator tor bea beam, m, thuss re thu reduc ducing ing the eff effect ect of re respi spira rator tory y mot motion ion by

patient, while the acce patient, accelera lerator tor runs cont continuou inuously sly.. The adjustment adjustment could be achi achieved eved by movi moving ng the patient couch as a function of time, or by moving the radiation beam. In the latter case, for scattered

synchronizin synchro nizing g the dose deli delivery very with the pat patient ient’s ’s brea br eath thin ing g cy cycl cle. e. A wi wide de va vari riet ety y of po posi siti tion on-monito mon itorin ring g dev device icess ha have ve bee been n us used, ed, inc includ luding ing a strain str ain gauge or linea linearr tra transdu nsducer cer attached attached to the abdomen abdom en or thor thorax, ax, and a temp tempera erature ture-sens -sensitiv itive e device dev ice ins insert erted ed in the nos nostri trill (F (Ford ord   et al., 200 2003). 3). The use of re respi spira rator tory y gat gating ing in pa parti rticle cle th thera erapy py has been described by Minohara   et al.   (2000) and Tsunashima  et al.  (2004). Deep breath hold at inspiration has been used to redu re duce ce th the e mo moti tion on of lu lung ng tu tumo mors. rs. In th this is te tech ch-nique, the patient is verbally coached to produce a reproducible level of deep inspiration, which is then used during treatment planning and dose delivery (Hanley et al., 1999). The patient breathes through a mouthpiece connected to a spirometer, while nose breathing is restricted with a nose-clip. The coach observes obser ves the spir spiromet ometer er signa signall durin during g tre treatm atment, ent, instru ins tructs cts the pa patie tient nt to hol hold d the their ir br brea eath th at the appropriate time, and the accelerator is turned on (Mah  et al., 2000). In active-breathing control, the patient breathes through a mouthpiece connected to a pair of flow monitors and valves. The device is used to monitor the pat patient ient’s ’s brea breathin thing g pat pattern tern and is calib calibrat rated ed during simulation. By closing the valves at a preselected phase in the respiratory cycle, the patient’s

and wo and wobb bble led d pr prot oton on be beam amss th the e po posi siti tion on of th the e apertu ape rture re or set settin tings gs of the mu multi lti-le -leaf af col collim limat ator or (but (b ut pr prob obab ably ly no nott th the e co comp mpen ensa sato tor) r) wo woul uld d be modifie mod ified. d. In the case of sca scanne nned d bea beams, ms, the set set-tings of individual pencil beams would be adjusted as they are delivered. Li   et al.   (2004) (2004) hav have e simu simu-lated lat ed the time time-dep -depende endent nt effe effects cts of targe targett motio motion n for a sca scann nned ed ion bea beam. m. The They y ha have ve the theore oretic ticall ally y shown sho wn th that at onl online ine mot motion ion com compen pensat sation ion wit with h a scanne sca nned d ion bea beam m can yie yield ld a re rest stor orat ation ion of th the e dose homogeneity of   . 95 percent. However, such a tra tr ack ckin ing g sys yste tem m ha hass yet to be de demo mons nstr tra ate ted d practically.

breathing motion be halted (Wong  et al., 199 1999). 9). Th The ecan forced for cedtemporarily breat br eath h hol hold d dura du ratio tion n is 10–45 10–4 5 s, dep epen end din ing g on th the e pa pattie ien nt, an and d th the e sequence will usually need to be repeated several times duri during ng the dose deliv delivery ery,, with appr appropria opriate te resst pe re peri riod ods, s, un unti till th the e de desi sirred do dose se ha hass be been en delivered. The Th e var variou iouss met method hodss of br brea eath th con contr trol ol all ha have ve the advantage that the extent of motion of tumors and organs due to respiration can be substantially reduced.

7.6 COMP COMPENSA ENSATION TION FOR PATIENT AND ORGAN MOTION For any giv For given en set of pa patie tient nt imm immobi obiliz lizat ation ion and patie pa tient nt and org organ an loc locali aliza zatio tion n tec techni hniqu ques, es, the there re always remains some degree of residual motion and some uncertainties uncertainties about the loca location tionss of pat patient ient,, target volume(s), and organs at risk (OARs). These uncertainties must be taken into account in planning the treatment (Mageras  et al., 1996).

7.6.1 Mar 7.6.1 Margins gins at the the periphe periphery ry of the the CTV CTV or OARs: lateral margins

The first task is to add a lateral margin or margins to the beam in order to allow for uncertainties by: (i) defining a PTV (or PTVs) for the clinically determined min ed CT CTV( V(s); s); and usu usuall ally y, (ii (ii)) the PR PRVs Vs for all deli de line neat ated ed OA OARs Rs.. Th The e ma mann nner er in wh whic ich h th thes ese e margins should be established has been the matter of considerable study. When delineating delineating a PTV, the diff differen erentt type typess of 

7.5.4 7.5. 4

marginss mu margin must st be add added ed or com combin bined. ed. If mar margin ginss are added linearly, the resulting PTV can often be

Organ Orga n motion motion with with tumor tumor tracki tracking  ng 

 A problem with all the above techniques is that they th ey re redu duce ce ef effic ficie ienc ncy y, as on only ly a po port rtio ion n of th the e patient’s breathing cycle can be used for irra ir radi diat atio ion, n, or th the e ir irra radi diat atio ion n is in inte terr rrup upte ted d betwe bet ween en br brea eath th hol holds. ds. In add additi ition, on, the they y re rely ly on measurements made well in advance of treatment that are assumed to apply at the time of treatment.  A more elegant solution would be to track target moti mo tion on du duri ring ng th the e tr trea eatm tmen entt (i.e., by im imag agin ing g implan imp lanted ted see seeds ds or sur surgica gicall cli clips) ps) and adj adjus ustt the position posit ion of the beam appr appropria opriately tely relative relative to the

too la too larg rge, e, wi with th a co cons nseq eque uent nt ri risk sk of ex exce ceed edin ing g patient tolerance. A quadratic approach similar to that recommended by the Bureau International des Poids et Mesures (BIPM, 1981) can be employed. It provides a means to combine random and systematic, as wel welll as corre correlat lated ed and uncorrelate uncorrelated d unce uncerr  e et t al. al . tainties tain ties (Mij (Mijnhee nheerr , 19 1987 87a) a).. Ut Util iliz izin ing g th this is approach, in order to find the overall margin ( i.e., the in intter ern nal mar argi gin n an and d the se sett-u -up p marg rgin in toge to geth ther er), ), th the e ov over eral alll sy syst stem emat atic ic er erro rorr ca can n be deri de rive ved d by ad addi ding ng qu quad adra rati tica call lly y th the e se sepa para rate te 127

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

systematic ematic errors: syst S

2

uncertainty uncertain ty in the delin delineat eation ion shou should ld be inclu included ded in margin considerations. 2

  S ¼

set-up

organ motion

þ

1=2

2

S

S

þ

delineation



;   ð7:1Þ

and si and simi mila larl rly y, th the e over overal alll ra rand ndom om er erro rorr ca can n be deri de rive ved d by ad addi ding ng th the e se sepa para rate te ra rand ndom om er erro rors rs quadratically:

s 

¼



2 set-up

  s 

þ s 2organ motion



1=2

:

 

ð7:2Þ

Several approaches Several approaches hav have e been published published to quan quan-tiffy th ti the e CTV-to to-P -PT TV marg rgin in req equ uir irem emen entts (Austin-Seymour   et al., 199 1995; 5; Bal Balter ter   et al., 199 1996; 6; al.., 199 al.., Crook   et al 1995; 5; Goi Goitei tein, n, 198 1985; 5; Ro Roesk eske e   et al 1995; Stroom  et al., 1999; van Herk  et al., 2000). Stroom   et al.  (1999) presented a model that was test te sted ed fo forr pr pros osta tate te,, ce cerv rvix ix,, an and d lu lung ng ca canc ncer er.. A  CTV-toCTV -to-PTV PTV set-u set-up p marg margin in size, which ensu ensures res at leas le astt 95 pe perc rcen entt of th the e do dose se is de deli live vere red d to 99 percent per cent of the CTV, CTV, is give given n by 2 S þ 0.7s,   S  being the st stan andar dard d dev deviat iation ion for the sy syst stema ematic tic err error or (average set-up deviations per patient in the group of pat patient ients) s) and   s   the standard standard devia deviation tion for the rando ra ndom m err error or (da (day-t y-to-d o-day ay set set-up -up pos positi itions ons). ). Van Herk  et al.  (2000) came to a similar conclusion that the standard deviation for the systematic errors is three times larger than for the random errors. On the other hand, McKenzie (2000) pointed out that breathing-induced motion should be accounted for separat sepa rately ely,, with the brea breathin thing g margi margin n adde added d linearly to the quadrature sum of the other contributing factors. Unfortunately, this ideal approach can be applied only onl y in sit situa uatio tions ns in wh which ich one can ide identi ntify fy the causes cau ses of err errors ors and qu quant antify ify th the e unc uncert ertain aintie tiess ( e.g., by standard deviations). Currently, this is not

7.6.2 Marg 7.6.2 Margins ins at at the the peripher periphery y of of the the CTV or OARs: margin in depth

For protons, a margin in depth between the treated  volume and the CTV must be provided to allow for uncert unc ertain aintie tiess in pr proto oton-b n-beam eam pen penetr etrat ation ion.. Fo Forr reasons already given (see Section 5.1.4), the PTV  may not be a very helpful tool for designing these marg ma rgin ins. s. Th The e co corr rrec ectt ap appr proa oach ch is to as asse sess ss th the e source sou rcess of un uncer certai tainty nty in th the e pr proto oton n pen penetr etrat ation ion ( e.g., unce uncertain rtainties ties conce concerning rning the prec precise ise loca location tion of pat patient’ ient’ss skin surface and inte internal rnal organs, the effects effec ts of ove overlyin rlying g hete heterogen rogeneitie eities, s, and the dose computat comp utation ion algori algorithm) thm) and then design the compensator pensat or,, so th that at the low lower er un uncer certai tainty nty bou bound nd of  the th e tr trea eatm tmen entt vo volu lume me ju just st co cove vers rs th the e CT CTV V (s (see ee Section 8). Moti Mo tion onss of in inte tern rnal al or orga gans ns tr tran ansv sver erse se to th the e beam direction can, in addition to the possibility of  causing under-dosing the tumor periphery, have a subtle sub tle infl influen uence ce on tis tissue suess dis distal tal to the mo movin ving g organ or tumor. In the lung, for example, motion of  a tumor transverse to the beam direction can affect the proton-beam penetration distal to the region of  the tumor as cha change ngess in the shadowin shadowing g of do downwnstre st ream am tis tissue suess by th the e tum tumor or wil willl af affec fectt the penetration of protons reaching those tissues. 7.6.3 Dose vari 7.6.3 variati ation on with within in the the CTV and OARs: interplay effects

With uni unifor form m bea beams ms pr produ oduced ced by sca scatte tterin ring, g, the 1 beam is essen essentiall tially y deliv delivered ered sta statical tically ly and th the e dose internal internal to the targe targett volu volume me is, in practice, practice, largely unaffected by organ motion or patient misregistration. The situation is different if the beam

is applied dynamically, as with scanned beams. When a scanned proton beam is applied, organ moti mo tion on ca can n ca caus use e bo both th po posi siti tive ve an and d ne nega gati tive ve fluctuations in the dose distribution due to what have ha ve bee been n ter terme med d ‘in ‘inte terpl rplay ay eff effect ects’ s’ (B (Bort ortfel feld d  et al. , 2002; Phillips   et al., 1992). Interplay effects aris ar ise e be beca caus use e bo both th th the e be beam am de deli live very ry an and d th the e target of that delivery are changing with time. A  cell within an organ (Goitein, (Goitein, 2005) might move synch sy nchro rono nousl usly y wit with h th the e bea beam m wh while ile it is bei being ng scanned in some direction and, as a result, receive either eit her more dose (wh (when en th the e mo move vemen mentt is in the scan direction) or less dose (when the movement is in th the e dir direct ection ion op oppos posit ite e to th the e sca scan) n) th than an wa wass

generally possible except for a few situations ( e.g., some conformal therapy protocols). However How ever,, it shou should ld be unde understo rstood od tha thatt the delineation of the PTV is a matter of compromise and is not simply a mat mathema hematical tical concept. It requ requires ires clinical judg clinical judgemen ement, t, and thus thus,, is the resp responsib onsibility ility of the radiation oncology team. Note No te th that at th the e pe penu numb mbra ra of th the e be beam am(s (s)) is no nott considere consi dered d when deli delineat neating ing the PTV. How However ever,, when whe n sel select ecting ing th the e bea beam m siz sizes, es, the wid width th of the penum pen umbra bra has to be tak taken en int into o ac accou count nt and the beam size must be enlarged accordingly. The beam sizes are defined by the 50 percent isodose (ICRU, 1976). Stud St udie iess ha have ve sh sho own th that at in intr traa- an and d in inte terrobserver variability can be a large source of uncertainty in GTV and, especially, especially, CTV delineation. The

1

In pr prac actic tice, e, ev even en sca scatte ttere red d bea beams ms ar are e sca scanne nned d in dep depth. th. However, the period of depth scanning is so much less than that for organ motion that this motion can be ignored.

128  

 MOTION MANAGEMENT 

Figure 7.1. Schemati Schematicc illustration illustration of the interpla interplay y effect.

intended—giving rise to ‘dose mottle’. Because the total tot al ene energy rgy del deliv iver ered ed by a bea beam m is not appreci appreci-ably abl y af affec fected ted by mot motion ion wit within hin th the e pa patie tient nt,, the average dose within a volume of interest tends to

‘repainting’. ‘repaint ing’. This is furt further her discu discussed ssed in Sect Sections ions 3.2.1.2.2 and 7.6.3.2. Inte In terp rpla lay y ef effe fect ctss oc occu curr wi with th sc scan anne ned d be beam amss rega re gard rdle less ss of th the e in inte tens nsit ity y pr profi ofile le of th the e be beam am..

be unalter unaltered. ed. The Th e int interp erplay lay eff effect ect is sho shown wn sch schema ematic ticall ally y in Fig. 7.1. Pencil beams are being delivered from left to rig right ht,, alo along ng th the e dir direct ection ion of the dot dotted ted arr arrow ow.. Two situations are shown. In frames (a) and (b), a cell is moving to the right, in the same direction as the scan. At the time of frame (a), the cell is receiving dose from the third pencil beam; at the later time tim e of fr fram ame e (b) (b),, a fou fourth rth pencil pencil bea beam m has just been applied. The cell has now moved into the path of th the e fo four urth th pe penc ncil il be beam am,, an and d so re rece ceiv ives es an additional unintended dose which could as much as double dou ble the dos dose e it wou would ld ha have ve re recei ceive ved d if it we were re stat st ation ionary ary. Fr Fram ames es (c) and (d) dem demons onstra trate te the opposite case in which a cell is moving to the left— in the opposite direction to the scan. At the time of  frame fr ame (c), th the e cel celll is re recei ceivin ving g alm almos ostt no do dose; se; at the later later tim time e of fr frame ame (b), a fou fourth rth pencil pencil bea beam m has just been applied. Meanwhile, the cell has now

However,, when the inte However intensit nsity y varie variess spa spatiall tially y very rap apid idly ly wi witthi hin n a bea eam m, as ca can n be the ca case se in intensity-modulated radiation therapy (IMRT), the magnitude of the dose mottle may be greater than for a uni unifor form m int intens ensity ity beam. This eff effect ect can be red educ uced ed by li limi miti ting ng th the e ste teep epne ness ss of sp spa ati tial al changes in the beam intensity in IMRT. 7.6.3.1 7.6.3.1  effects

Experimental Experim ental obser observation vation of interplay interplay

That the interplay effect is not merely a theoretical probl pr oblem em is evi eviden denced ced by a ra radio diobio biolog logica icall ex exper per-imentt perf imen performed ormed by Gueu Gueulette lette   et al.   (2005). (2005). The They y were engaged in measuring the relative biological effective effec tiveness ness (RBE) of pro protons tons by asses assessing sing crypt cell rege regenera neration tion aft after er whol whole-pel e-pelvis vis irrad irradiati iation on of  exp xper erim imen enttal mic ice. e. The proto ton n be beam am was a scanned beam, with a pencil width at the location

moved out of the path of the fourth pencil beam, and so re recei ceive vess alm almost ost no dos dose— e—so so th that at the total dose delivered to this cell is much less than had it been stationary. Wit ith h sc scan anne ned d be beam ams, s, in inte terp rpla lay y ef effe fect ctss ar aris ise e when wh en th the e fr freq eque uenc ncie iess of th the e be beam am sc scan an an and d of  orga or gan n mo moti tion on ar are e co comp mpar arab able le.. Typ ypic ical al or orga gan n motion mot ionss ha have ve a per period iod of bet betwe ween en a fr fract action ion of a

of th the e an anim imal alss of    10 mm fu full ll wi widt dth h at ha half  lf  maximum. maxim um. They obser observed ved a much greater greater fluct fluctuuati tion on in th thei eirr res esul ults ts th than an in th the e xx-ra ray y be beam am controls cont rols or in prev previous ious exp experimen eriments ts on scat scattere tereddproton beams, and this effect was reproduced when they the y rep repeate eated d the experiment. experiment. The scat scatter ter in the results was consistent with fluctuations in dose of  approximately   + 13 per erce cen nt (1 SD), wh whil ile e the

second several Scanned-beam delivery, on the to other handseconds. hand, , has three thre e dim dimens ension ionss delivery of sca scan, n,, namely, the two lateral directions and penetration in dep depth th,, whi which ch are usu usuall ally y adm admini inist stere ered d wit with h a high, medium, medium, and low freq frequenc uency y. Sinc Since e the dose delivered by a beam should be deposited in   1 min, it is al almo most st gu guar aran ante teed ed th that at at le leas astt on one e of th the e three thr ee mot motion ionss wil willl ha have ve a per period iod on the order of  that of organ motion. The most direct way of mitigating the influence of mot motion ion on dos dose e mot mottle tle within within the CT CTV V and/or and/or OARs is by repeating the sequence of pencil beam deli de liv ver ery y sev se ver era al tim ti mes es— —a proc oces esss ter erm med

me meas asur ured ed beam be am profi pr ofile le as fla flattto wi with thin in a fe few w percent. They ascribed thewscatter movements of  the mice intestines during irradiation. Radiographs showed that the intestines of similarly-immobilized mice moved by up to   + 2 mm. Although there was wide scatter in the data, the RBE at the center of a spread-out Bragg peak, based on a fit to the data, wass en wa enti tire rely ly co cons nsis iste tent nt wi with th va valu lues es me meas asur ured ed using the same animal model in several scattered proto pr oton n bea beams ms (na (namel mely y 1.1 1.16, 6, but wit with h wid wide e con confifidence den ce lim limits its). ). Thi Thiss und underl erline iness th the e poi point nt th that at th the e mean dose might not be much altered, even though dose mottle may be quite large. 129

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

7.6.3.2 .2 7.6.3

Repainting Repain ting to reduce reduce the influence influence

of interplay effects Phillips   et al.   (1992) have quantitat quantitatively ively analyzed the th e ef effe fect ctss of or orga gan n mo moti tion on an and d th the e ne need ed fo forr repaintin repa inting g (term (termed ed ‘mul ‘multiplic tiplicity’ ity’ by the auth authors), ors), and Bortfeld   et al.   (2002) have demonstrated in a model that interplay effects can be substantial for scan sc anne ned d be beam ams, s, na name mely ly do dose se va vari riat atio ions ns of th the e order of   + 10 per percen cent. t. The They y poi point nt out tha that, t, whe when n multi mu ltiple ple fr fract action ionss ar are e giv given, en, th the e eff effect ect ten tends ds to aver av erage age out out.. Ho Howe weve verr, one doe doess not alw alway ayss giv give e many fractions, fractions, and the safest and mos mostt str straigh aighttforwa for ward rd wa way y to ov overc ercome ome int interp erpla lay y eff effect ectss is to repeat repe at the dose deliv delivery ery seve several ral time timess (obv (obviousl iously y each ea ch app applica licatio tion n ha havin ving g a pr propo oporti rtiona onatel tely y low lower er dose) dos e) dur during ing th the e app applic licat ation ion of a sin single gle sca scanne nned d beam. bea m. Pr Prov ovide ided d th the e re repai painti ntings ngs occ occur ur as async ynchr hroonous no usly ly wi with th re resp spec ectt to th the e or orga gan n mo moti tion on,, th the e redu duct ctio ion n in dos ose e fluct ctu uati tio ons sh shou ould ld be of  the order of the inverse of the square root of the numb nu mber er of re repa pain inti ting ngs. s. Th This is ar argu gume ment nt wo woul uld d suggest sugge st that somet something hing like 10 repa repaintin intings gs would be de desi sira rabl ble. e. Ho Howe wev ver er,, th ther ere e is li litt ttle le ne need ed to repaint repa int penc pencil il beams that deliv deliver er rela relative tively ly litt little le dose do se,, an and d th ther eref efor ore e on only ly a su subs bset et ne need ed to be repainted and one might conclude that only the few highest high est energy ‘layers’ need to be rep repaint ainted. ed. This would be the case when the PTV is approximately

Motion Motio n and mis misre regis gistr trat ation ion of th the e tar target get vo volum lume e with wi th re resp spec ectt to th the e ra radi diat atio ion n be beam amss is is,, at so some me level, lev el, ine inevit vitabl able. e. If the tar target get volumes volumes ar are e to be adequately irradiated, and adjacent OARs are to be

rectan rect angu gula larr in shap shape e as su sugg gges este ted d in Fi Fig. g. 7. 7.2a 2a.. Howe Ho weve verr, for a nea near-s r-sph pheri erical cal tum tumor or,, wh which ich is a

protected: it is essential that causes and possible magnitudes of motion andthe misregistration are

Figure 7.2. Sche Figure Schemat matic ic diag diagram ram of the application application of a scan scanned ned beam in several energy layers. The highest weighted spots are colored colo red red, foll follow owed ed by ora orange, nge, green, and blue in decr decreasi easing ng orderr of wei orde weight. ght. The dotte dotted d line liness are iso-range iso-range lines, with the long lo nges estt ran ange ge la labe bele led d ‘1 ‘1’. ’. (a (a)) If th the e ta targ rget et vo volu lume me is near-re near -rectan ctangula gularr, and only the dis distal tal thr three ee spots (re (red, d, ora orange, nge, and green) need to be repainted, then only three energy layers need nee d re repai painti nting. ng. (b) Fo Forr a ne near ar-sp -spher heric ical al tar target get vo volum lume, e, penci pen cil-b l-beam eam spots in ma many ny mo more re ene energy rgy la laye yers rs nee need d to be repainte repa inted d (five or six in this schematic schematic example). example). For a mor more e realist real istic ic situ situati ation, on, from 10 to 20 lay layers ers would incl include ude at leas leastt some pencil beams that would need to be repainted.

7.7 CO CONC NCLU LUSIO SION N

understood; understo od; tha thatt thei theirr poss possible ible conse consequen quences ces are understo unde rstood; od; tha thatt meas measures ures be tak taken en to mini minimize mize motion motio n and misr misregis egistra tration tion to the exte extent nt possi possible ble and clinically warranted; and that steps are taken to all allow ow for the remaini remaining ng deg degree reess of mot motion ion and misregistration.

much mo much more re re real alis isti ticc ca case se,, th the e hi high gh-d -dos ose e Br Brag agg g peaks pea ks ar are e dep deposi osited ted in ma many ny la layer yers, s, as sho shown wn in Fig. 7.2b. For example, when the spacing in depth is, say,    5 mm in a target of, say,   15 cm diameter, pencil beams spread over something like 15 layers would need to be repainted.

130  

Journal of the ICRU Vol 7 No 2 (2007) Report 78 Oxford University Press

8

doi:10.1093/jicru/ndm042

ESTIM EST IMA ATI TION ON AN AND D PR PRESE ESENT NTA ATI TION ON OF UN UNCE CERT RTAI AINT NTY  Y 

IN THE DELIVERED DOSE

8.1

THE INEVIT INEVITABILITY ABILITY OF UNCERT UNCERTAINTY  AINTY 

Radiation Radiatio n ther therapy apy is inhe inheren rently tly prob probabili abilistic stic.. One cann ca nnot ot be ce cert rtai ain n as to wh whet ethe herr a tu tumo morr wi will ll be controlled, or whether any given normal tissue will be damaged. These outcomes can be stated only in terms of prob probabili abilities. ties. In addi addition tion,, the application application of radiation involves very many factors, almost all of which have some level of associated uncertainty. For example, there are uncertainties regarding the following: †

†

  th the e ide identi ntifica ficatio tion n of the tum tumor or and the des desigignation of its histology and staging;   the spatial extent of the tumor and of organs at

many additional additional pote potentia ntiall sour sources ces of unce uncertain rtainty: ty: interp int erplay lay eff effect ectss tha that, t, wit with h bea beam m sca scanni nning, ng, can lead to th a at degree of dose inhomogeneity within thetumor tu mor that are ar e not normal nor mally ly enc encoun ounter tered ed in con con ventional radiother radiotherapy apy (see Section 7.6.3), registration tra tion unce uncertain rtainties ties ( e.g., be betw twee een n a ph phys ysic ical al or  virtual compensator and the patient) and, in charged-particle therapy, the effect of inhomogeneities within the patient that can strongly influence the th e do dose se di dist stri ribu buti tion on in th thei eirr sh shad adow ow.. In th this is environment, it can be very difficult to assess the dose do se im impl plic icat atio ions ns of th the e un unce cert rtai aint ntie iess th thro roug ugh h  visual inspection, and some form of computation computational al appro app roac ach h is re requi quire red. d. Un Unfor fortu tuna natel tely y, cur curre rentl ntly y available ava ilable radi radiothe otherapy rapy-plan -planning ning sys systems tems hav have e yet

†

†

† †

risk (O risk (OAR ARs; s; im imag ages es ma may y be in inco corr rrec ectl tly y in inte terrpreted, they may be distorted, and so forth);   for exte externalrnal-beam beam ther therapy apy,, the immo immobiliza bilization tion and localization of the patient and of the tumor within the patient, and the effects of physiologic motions on the dose delivered to any point within the patient;   the assessment of the distribution of heterogeneities, the effects of heterogeneities, and imperfections in the techniques to compensate for them;  the algorithms used to estimate dose;  the many parameters involved in the delivery of  treatments.

available radi available radiothe otherapy rapy plan planning ning sys systems tems hav have e yet to embrace uncertainty analysis.  Although, as just emphasized, there are many sources of uncertainty, some of them do not readily lend themselves to computational analysis. The following discussion is limited to the important question ti on of th the e ex exte tent nt to wh whic ich h th the e pr pres escr crib ibed ed do dose se distribu dis tribution tion is a true representa representation tion of the dis distritribution of dose the patient actually receives.

8.2

THE ESTIMA ESTIMATION TION OF UNCERT UNCERTAINTY  AINTY 

Given such uncertainties, one seeks to understand the sour sources ces of unce uncertain rtainty ty,, to red reduce uce them wheneverr pr eve prac actic ticabl able, e, and to eva evalua luate te th the e mag magnit nitud ude e and implications. The mere exercise of identifying the sources and magnitudes of uncertainty can be a valuable aspect of developing and judging a plan. In the pra practice ctice of rad radiothe iotherap rapy y, the est estimat imation ion and reporting of uncertainty has historically been at best implicit. Experienced physicians evaluating a trea treatmen tmentt plan undoubtedly undoubtedly make some mental assessment of the magnitude of the known uncertain ta inti ties es an and d wh wha at th the e co cons nseq eque uenc nces es ma may y be be.. However How ever,, the curr current ent sta state-of te-of-the -the-art -art tre treatm atmententdelivery deliv ery tech technique niquess seek grea greater ter geom geometric etric accu accu-racy rac y in dose delivery delivery, and are more complex complex than

In analyzing a radiation-treatment plan, there are at least two types of data whose uncertainties need to be estimated. The first involves the estimate of  the th e un uncer certai tainty nty in th the e dos dose e at sel select ected ed poi points nts in three-d thr ee-dimens imensions ions with within in the pat patient ient.. The secon second d type typ e inv involv olves es the es estim timat ation ion of unc uncert ertain aintie tiess in quanti qua ntitie tiess suc such h as   D98%,   D50%, EU EUD, D, TCP, an and d NTCP or in quantities used for constraints, such as the volume receiving greater than a certain dose  D , V  D, or th the e mi mini nimu mum m do dose se th that at is de deli live vere red d to a given give n volu volume me   V ,   DV . In th this is ca cate tego gory ry is al also so th the e quantification of the adequacy of dose coverage of  the planning target volume (PTV) and OARs. The terminology of uncertainty analysis has been clarifi cla rified ed in ISO (19 (1995) 95).. Ho Howe weve verr, lit little tle has bee been n repor re ported ted in th the e lit liter erat atur ure e in re rela latio tion n to mak making ing

those tho se emp employ loyed ed pr previ evious ously ly.. Su Such ch met method hodss ha have ve

estimates of uncertainty in radiotherapy. A general

# International Commission on Radiation Units and Measurements 2007

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

al..   (1991 re revi view ew wa wass pr pres esen ente ted d in Ur Urie ie   et al (1991). ). A  method has been described (Goitein, 1985) to estimate the uncertainty limits associated with a particu ti cula larr tr trea eatm tmen entt pl plan an (s (see ee Se Sect ctio ion n 8. 8.3) 3).. Do Dose se uncert unc ertain aintie tiess in pr proto oton-b n-beam eam th thera erapy py ha have ve als also o been bee n es estim timat ated ed as su sugge ggest sted ed in Goi Goitei tein n (19 (1978a 78a;; 1982a) and Lomax (2001).  At another level, a body of work has appeared in the last few years analyzing patient set-up uncertain ta inti ties es an and d mo moti tion onss fr from om th the e po poin intt of vi view ew of  determini dete rmining ng the mos mostt appr appropri opriate ate safe safety ty margi margin n arou ar ound nd a tu tumo morr vo volu lume me (v (van an He Herk, rk, 20 2004 04), ), an and d more recently a number of authors have also begun

to look into the problem of dealing with uncertainties ti es at th the e op opti timi miza zati tion on le leve vel, l, ma main inly ly fr from om th the e al.., poin po intt of vi view ew of or orga gan n mo moti tion on (B (Bec eckh kham am   et al 2002; McShan  et al., 2002).  A confidence level (CL) must be associated with any unce uncertain rtainty ty est estimat imate. e. Withou ithoutt a sta stateme tement nt of  the CL, an uncertainty estimate is meaningless. It is com common mon in re repor portin ting g rad radiot iother herapy apy res result ultss to indicate the 95 percent (2 SD) confidence intervals.

8.3

THE PRESENT PRESENTA ATION OF UNCERT UNCERTAINTY  AINTY 

The pr prese esenta ntatio tion n of the un uncer certai tainty nty in a thr threeeedimension dime nsional al dose dis distribu tribution tion pre present sentss a chall challenenging problem because of the plethora of data. One appr ap proa oach ch is de desc scri ribe bed d by Go Goit itei ein n (1 (198 985) 5) an and d by Urie   et al.   (1991) (1991),, an ex examp ample le of wh which ich is sho shown wn al..   (1991). in Fig Fig.. 8.1 8.1,, tak taken en fr from om Uri Urie e   et al (1991). Thr Three ee dose dos e dis distri tribut bution ionss ar are e ju juxta xtapos posed: ed: th the e nom nomina inall (most (mo st lik likely ely)) dos dose e dis distri tribut bution ion,, and and,, sep separ arat ately ely, the upper upper an and d lo lowe werr bou bound ndss on the dose at each point poi nt (a (att the st stat ated ed pr proba obabil bility ity lev level). el). Th This is hig highhligh li ghts ts th the e sc scal ale e of po pote tent ntia iall pr prob oble lems ms th that at ca can n arise as a result of a beam juncture from possible trea tr eatm tmen entt un unce cert rtai aint ntie iess an and, d, in Fi Fig. g. 8. 8.1e 1e,, ho how w thes th ese e ca can n be re redu duce ced d by be beam am fe feat athe heri ring ng (s (see ee Section 6.2.4.5).  An alternativ alternative e approach is described in Lomax (2001). In this method, dose distributions are calculated la ted for a num number ber of tra transl nslat ated ed (or ro rota tated ted)) CT data da tase sets ts,, an and, d, po pote tent ntia iall lly y, fr from om da data tase sets ts wi with th altered CT numbers to simulate density uncertainties. tie s. A hyb hybrid rid dos dose e dis distri tribut bution ion,, wh which ich ind indica icates tes

Goitein (198 Goitein (1983) 3) argu argued ed tha that, t, for many purp purposes oses in radiation therapy, 1 SD is too low a CL, and 2 SDs are ar e too high, and tha thatt an 85 per percen centt con confid fidenc ence e interval, inte rval, corr correspon esponding ding to    1. 1.5 5 SDs, is a mor ore e useful interval for many applications.

the worst-case dose at any point, is then computed as follows. For points within the PTV and CTV, the dose is set to the lowest dose at that point in any of  the calcu calculate lated d dose distribution distributions. s. For those poin points ts

Figure 8.1. Disp Figure Display lay of the dose dis distrib tributio ution n in a sagit sagittal tal section of a pati patient ent whose para-aortic para-aortic nodes are bein being g trea treated ted with par paralle allell al.., 1991 opposed oppo sed x-ra x-ray y beam beams, s, usin using g beam junctionin junctioning g (Uri (Urie e   et al 1991;; rep reprod roduced uced with perm permissio ission). n). (a) Nomi Nominal nal dose dis distrib tributio ution; n; (b) absolute dose scale (color from 10 to 80 Gy; color gray   , 10 Gy); (c) the upper-bound dose at the 85 percent CL, showing the possibility of a significant region of high dose; (d) the lower-bound dose at the 85 percent CL, showing the possibility of a significant region of low dose; (e) the upper-bound dose when the junction is feathered ( 2 1, 0,   þ 1 cm). A much smaller hot spot is seen in the overlap region [compare (e) with (c)].

132  

UNCERTAINTY IN THE DELIVERED DOSE

Figure 8.2. (a) The individual beams of a three-beam IMPT plan for a thoracic chordoma, with the nominal combined dose distribution distribution at the bottom. (b) The ‘worst case’ dist distribu ribution tion resulting resulting from 5 mm shif shifts ts alon along g eac each h majo majorr axis of the patient. patient. The worst-cas worst-case e

distribution distribu tion is calc calcula ulated ted at eac each h poin pointt by taki taking ng the minimum minimum dose of these shifted shifted dose dosess with within in the CTV, and the maxi maximum mum dose outside. Note the potential cold spots (blue areas) that could occur where beams abut ( i.e., along the patch lines of the oblique beams with the posterior beams). (Figure courtesy of A. Lomax, Paul Scherrer Institute, Villigen, Switzerland.)

outside the PTV (and, hence, within normal tissue), the dose is set to the highest in any of the calculated dose dos e dis distri tribut bution ions. s. Thi Thiss the then n doc docume uments nts pot potent ential ial cold co ld sp spot otss wi with thin in th the e tu tumo morr an and, d, in th the e sa same me

Dose–volum Dose–vol ume e hi hist stog ogra rams ms (D (DVH VHs) s) (D (Drz rzym ymal ala a  et al., 1991; Shipley  et al., 1979) are also an important dose-summarizing tool. Techniques for estimating an and d dis displa playin ying g un uncer certai tainty nty ban bands ds for DVH DVHss

display, potent display potential ial hot spots within normal tissues. Such an analysis is shown in Fig. 8.2; the potential cool regions in the tumor (colored blue corresponding to a 10–20 percent dose reduction) are because of pos possib sible le jun junctio ction n pro proble blems ms with the thr three ee abut abut-ting beams. This form of data presentation has its origins orig ins in the display display of ‘re ‘region gionss of reg regret ret’, ’, whi which ch was suggested in Shalev  et al.  (1988).

et al., 1991; have be have been en re repo port rted ed (D (Drz rzym ymal ala a   et al.., 199 Niemie Nie mierko rko and Goi Goitei tein, n, 199 1994; 4; Uri Urie e   et al 1991). 1). These techniques lead to the display of a band in dose–volume space, within which a given point of  the true DVH lies (at a stated level of confidence).  An example of this is shown in Fig. 8.3, which is reproduced from Urie   et al.   (1991). In this display, as with the uncertainty bounds of Figs 8.1 and 8.2, the uncertainties in the dose at points within the patie pa tient nt ar are e gen gener erall ally y hig highly hly cor corre rela lated ted wit with h one another so that a DVH following one of the uncertainty bounds is generally not physically realizable. Unfortunately, such displays are unable to exhibit this fact. The PTV is defined largely to accommodate alignment and motion uncertainties, and the treatment plan pl an is fr freq eque uent ntly ly de desi sign gned ed su such ch th that at th the e PT PTV  V  receives a lower dose at its boundaries than in its interior inte rior.. Becau Because se the CTV mome moment-t nt-to-mo o-moment ment or day-to-day is unlikely always to be located near the edge edg e of the PT PTV V, the DVH of the PTV will tend to underestimate the doses in the possible lower-dose regions regi ons in the CTV. On the othe otherr hand and for the same sa me re reas ason ons, s, th the e DV DVH H of th the e CT CTV V wi will ll te tend nd to overes ov erestima timate te the dose dosess in the possi possible ble low lower-do er-dose se regions in the CTV. Consequently, the DVHs of the PTV and CTV probably bracket the dose the tumor

Figure Figur e 8.3 8.3.. Dose–volu Dose–volume me his histog togra ram m wi with th upp upper er and lo lowe werr bound limits (at the 85 percent CL) for the dose distributions shown in Fig. 8.1a, c, and d (Urie  et al., 1991; reproduced with permission). The potential hot and cold spots are very evident in the DV DVH, H, but the their ir spa spatia tiall loc locat ation ions, s, of cou course rse,, can cannot not be inferred from the DVH.

actually and, such, can to mate ma te anreceives unce un cert rtai aint nty y as band ba nd abou ab outt be the th eused ‘tru ‘t rue’ e’ esti(but (b ut unknowable) tumor DVH. 133

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

8.4 RECOM RECOMMENDA MENDATIONS TIONS FOR THE CONSIDERATION AND REPORTING OF UNCERTAINTY  †

†

 Those involved in designing radiation treatments shou sh ould ld an anal alyz yze e th the e un unce cert rtai aint ntie ies; s; ma mak ke an effort to minimize them to the extent practicable; ensu en sure re th that at a qu qual alit ity y as assu sura ranc nce e pr prog ogra ram m is in place to give assurance that the treatment can be gi give ven n as pr pres escr crib ibed ed;; an and d do docu cume ment nt th thei eirr assessment of the remaining uncertainties.  Treatment planning systems should provide tools for the ana analys lysis, is, qua quant ntific ificat ation ion,, and dis displa play y of  uncertainties.

†

sented, but those in summarizing quantities (see Section 5.6.2) should be estimated, together with their corresponding confidence intervals. Such an estim es timat ate e cou could ld be st stat ated ed as fol follo lows: ws: ‘Do ‘Doses ses ar are e  judged to be accura accurate te to  x  percent of the prescription ti on do dose se,, or to be wi with thin in   y   mm of th the e tr tru ue location loca tion (at the   z   percent percent CL). CL).’’ The uncer uncertaint tainty y estim es timat ate e mig might ht be bas based ed on gen generi ericc ana analys lyses es of  the particular class of treatment, in which case it should be so-noted.  For cases where unacceptably large uncertainties mi migh ghttific t ex exis ist, t, an and d forr unce fo illu il lust stra rati tive ve pu purp oses esdose in scientific scien reports reports: : the uncertain rtainties ties inrpos the distribut dist ribution(s ion(s), ), as well as thos those e in summ summarizi arizing ng

†

 For normal reporting purposes, in uncomplicated case ca ses, s, th the e un unce cert rtai aint ntie iess in th the e fu full ll th thre reeedimens dim ension ional al dos dose e dis distri tribut bution ion nee need d not be pr pree-

quantities, quantitie s, shou should ld be est estimat imated ed and pre presente sented, d, together toget her with a sta stateme tement nt of the corr correspon esponding ding confidence intervals.

134  

Journal of the ICRU Vol 7 No 2 (2007) Report 78 Oxford University Press

9

QUAL QU ALIT ITY Y AS ASSU SURA RANC NCE E

doi:10.1093/jicru/ndm031

 A comprehensiv comprehensive e quality assurance (QA) program for a proton-beam therapy facility consists of procedur ced ures es th that at ens ensur ure e a con consis sisten tentt and sa safe fe fu fulfil lfilllmentt of th men the e dos dose e pr presc escrip riptio tion n as we well ll as min minima imall radiation exposure to the personnel and the public. Practical implementation of a QA program depends on the det detail ailss of th the e pr proto oton n ac accel celer erat ator or and the selected select ed beam beam-deli -delivery very technique. technique. In many cases cases,, QA che checks cks wil willl be re relev levant ant only to th the e par partic ticula ularr equipment or to the implemented technology. Very ofte of ten n th the e in inst stit itut ution ion th that at ha hass ac acqu quir ired ed a ne new w proton-beam therapy facility plays the dual role of  both manufacturer and user, and therefore the proposed checks are a result of research and development me nt,, an and d co coul uld d be mo modi difie fied d wi with th ac accu cumu mula late ted d experience. The Th e QA mea measur sureme ements nts de descr scribe ibed d bel below ow fol follo low w the steps of the proton-therapy procedure and focus on dos ose e del eliv iver ery y an and d tr trea eatm tmen entt pla lan nnin ing g. Comm Co mmiss ission ioning ing and val valida idatio tion n of pr proto oton-b n-beam eam delivery deliv ery and trea treatmen tment-pl t-planni anning ng sy system stemss (TPS (TPSs) s) includ inc lude e ma machi chinene-sp speci ecific fic bea beamm-da data ta ac acqui quisit sition ion,, data da ta ent entry ry int into o the tr trea eatme tment nt-pl -plann anning ing sy syst stem, em,  validation of the calculations, development of operational procedures and constancy checks, as well as training of all staff concerned with the operation of  the sy syst stem. em. Ne New w typ types es of pr proto oton-b n-beam eam-d -deli elive very ry sys systems might t also require req uire biological biolog ical assessmen al.., sment al as tems a pa part rtmigh of com commis missio sionin ning g (Ka (Kagaw gawa a   etasses 2002; 200 2;t Pedroni  et al., 1995), whereas preclinical testing of  commercial comme rcially ly avai available lable faci facilitie litiess is usua usually lly limit limited ed to physics and dosimetry acceptance checks (Kooy, 2002). The data obtained during the commissioning and val valida idatio tion n pr proce ocess ss are use used d la later ter as ben benchchmark ma rkss an and d th thre resh shol olds ds fo forr pe peri riod odic ic QA ch chec ecks ks..  Acceptance, commissioning, as well as periodic engine eng ineeri ering ng and ma maint intena enance nce che checks cks of pr proto oton n accelera acce lerators tors and mult multiple iple trea treatmen tmentt room switc switchhyard ya rdss ar are e no nott di disc scus usse sed, d, as th the e ac acce cept ptan ance ce an and d qualit qua lity y con contr trol ol of par partic ticula ularr pr proto oton n ac accel celer erat ators ors and switchyards with vacuum beam lines, focusing, bending, bendin g, an and d st steer eering ing ma magne gnets, ts, po powe werr sup suppli plies, es, and coo coolin ling g equ equipm ipment ent ar are e usu usuall ally y bas based ed on th the e recommen reco mmendat dations ions of the manu manufact facturer urer and migh mightt not be relevant to another type of machine.

9.1 PRO PROTON-BE TON-BEAM AM DELIVER DELIVERY Y SYSTEM SYSTEMS S Only Onl y a fe few w pu publi blica catio tions ns and int intern ernal al rep report ortss ar are e available, which describe commissioning and periodic QA checks of different components of the beamdelivery systems (Chu   et al., 1993; JASTRO, 2004; al.., 199 Moyer Mo yers, s, 199 1999; 9; Pe Pedr droni oni   et al 1995; 5; Sch Schre reud uder er,, 2002). The nozzle of a proton-beam-delivery system using dual scattering foils consists of a fixed section that contains various components for beam shaping and bea beam m mon monito itorin ring, g, and a mo movab vable le sno snout ut th that at permits perm its posit positionin ioning g of the pat patient ient-spec -specific ific device devices, s, such su ch as ap aper ertu ture re an and d co comp mpen ensa sato tor/ r/bo bolu luss (s (see ee Sectio Sec tion n 3). Th The e sno snouts uts are des design igned ed to min minimi imize ze ra radi diat atio ion n co leak le age e isto fo the th e sed pati pa tien ent, and an part rt of  qua qu ali lity ty con nakag trol focu cus ont, th the ed aradpa iat ia tio ion n protection issues. The integrity and quality of alignment of hard hardwar ware e comp component onentss shou should ld be check checked ed thro th roug ugh h th the e an anal alys ysis is of de dept pth– h– do dose se cu curv rves es an and d late la tera rall pr profi ofile less th tha at ar are e me meas asur ured ed in a wat ater er phantom. Although measured results are needed to  verify performance specifications and provide benc be nchm hmar arks ks fo forr tr trea eatm tmen entt-pl plan anni ning ng so soft ftwa ware re (Moyers, 1999), the use of Monte Carlo simulation (Paganetti   et al., 2004) can also support dosimetry efforts in helping to define safety tolerances and to design QA procedures. The Monte Carlo simulation of an isoce isocentric ntric-gan -gantry try tre treatm atment ent nozz nozzle le descr described ibed by Kooy (2002) and by Paganetti  et al.  (2004) allows the th e sen sensit sitivi ivity ty of pr proto otonn-dos dose e dis distri tribut bution ionss in a water phantom with respect to various beam paramet am eter erss an and d ge geom omet etri rical cal mi misa sali lign gnme ment nts, s, to be studi st udied, ed, and and,, as a re resul sult, t, the tol tolera erance nce lev levels els for these parameters might be established. Figure 9.1, for example, shows the high sensitivity of the Bragg peak to beam misalignment misalignment.. Depe Dependin nding g upon the amount of material in the beam, the pristine Bragg curve is characterized by either a low-energy tail or possibly by a secondary Bragg peak at a different depth. The upper part of Fig. 9.1 shows the effect on the depth–dose distribution of a beam scraping a frame used to support the first scatterrer, whereas the lower part shows the proton energy distribution at a wate terr ph phan anto tom m su surf rfa ace ce.. On th the e ba basi siss of  measurements and results of simulation, it is possible ib le to ca calc lcul ulat ate e to tole lera ranc nces es fo forr th the e ap appr prop opri riat ate e

# International Commission on Radiation Units and Measurements 2007

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

 An important part of quality-control procedur procedures es is related to beam monitoring with multi-wire and multimul ti-seg segmen mentt ion ioniza izatio tion n cha chambe mbers rs th that at con contr trol ol devia dev iatio tions ns in bea beam m pos positi ition, on, che check ck th the e bea beam m siz size e and its unif uniformit ormity y, and control the dose delivered delivered to the patient during the treatment. Usually, ratios

of sig signa nals ls bet betwe ween en var variou iouss det detect ectors ors ar are e use used d to trigge tri ggerr sa safet fety y con contr trol ol int interl erlock ockss and ter termin minat ate e beam deliv delivery ery.. Tolera olerance nce valu values es for beam termi terminnation should be specified to ensure proper settings of range-shifter devices, selection of scattering-foil positions, and rotational velocity of modulator propell pe ller ers. s. Th The e ca cali libr bra ati tion on of th the e pr prim imar ary y do dose se monito mon itorr sho should uld be bas based ed on ion ioniza izatio tionn-cha chambe mberr measurements in a water phantom, and dose calculations using an established dosimetry code of practice ti ce (s (see ee Se Sect ctio ion n 4) 4).. Th The e mo moni nito torr ca cali libr brat atio ion n is performed on a daily basis at many proton-therapy faci fa cili liti ties es.. For da dail ily y ch chec ecks ks of sc scan anne nedd-be beam am systems, an ionization chamber is positioned in the refere ref erence nce vo volum lume e (of (often ten a 10  10  10 cm3 cube) within a water phantom, which is irradiated so as to deliver a dose of 1 Gy to the reference volume

operat oper atio iona nall pa para rame mete ters rs fo forr be beam am de deli live very ry an and d action levels that should be used in periodic checks. Charact Char acteriza erization tion and perio periodic dic sta stabilit bility y check checkss of  the mechanical and radiation isocenter for proton

(Coray et al., 2002). Figure 9.2 shows an example of  the th e re rela lati tive ve do dose sess me meas asur ured ed da dail ily y in a ho homo mo-geneous volume. The effect of the energy and the gantry angle on dosimetry should also be checked periodically. The relation between the proton-beam energy energ y and the monit monitor-c or-chamb hamber er resp response onse shou should ld be frequently checked (Coray  et al., 2002). If a rotating gantry is used, multiple irradiations from different directions are often carried out with the th e pa pati tien entt in th the e sa same me po posi siti tion on.. Th The e mo moni nito torr cali ca libr brat atio ion n va valu lue e sh shou ould ld be co corr rrec ecte ted d at ea each ch gant ga ntry ry an angl gle e if th the e va valu lue e va vari ries es as a fu func ncti tion on of gantry angle. The stability of the monitors with gantry angle and reproducibility of the beam from

gantries are important in order to facilitate better patient alignment from all beam directions. A procedu ce dure re to de dete term rmin ine e th the e sh shap ape e an and d si size ze of th the e proton gantry isocenter to within 0.2 mm is given by Moyers and Lesyna (2004). Perio Pe riodic dic QA che checks cks for wob wobbli bling ng and sca scanni nning ng delive del ivery ry tec techni hnique quess ar are e sim simila ilarr to th those ose for the pass pa ssiv ive e be beam am sc scat atte teri ring ng te tech chni niqu que. e. Ho Howe weve verr, becaus bec ause e th these ese del delive ivery ry sy syst stems ems ar are e dyn dynami amicc and accurat accu rate e deliv delivery ery of the planned dose depe depends nds on the ac accur curat ate e dep deposi ositio tion n of ind indivi ividua dually lly we weigh ighted ted pencil beams, additional QA methods are required for scann scanned-b ed-beam eam sys systems tems.. Meas Measurin uring g proc procedur edures es for the det determ ermina inatio tion n of th the e sha shape, pe, pos positi ition, on, and direction of individual pencil beams must be developed, together with methods for assessing the homogenei oge neity ty of the dose and shape shape of com comple plex x fiel fields ds,, which whi ch can be del deliv ivere ered d wit with h suc such h sy syst stems ems (Ch (Chu u  et al., 1993; JASTRO, 2004; Pedroni  et al., 2005).

Figure Figu re 9.2. Dose measured measured dail daily y in the cent center er of a ref refere erence nce  volume in a scanning beam (n   is the numb number er of dat data a poin points) ts) (Coray et al., 2002; reproduced with permission).

Figure 9.1. An example of the effec Figure effectt of beam misa misalig lignmen nmentt on the pristine Bragg peak and on the proton spectrum (Paganetti  et al., 2004; reproduced with permission).

136  

QUALITY QUALI TY ASSUR ASSURANCE ANCE

the accelerator should be verified as a part of peri-

for QA of radiotherapy TPSs. This report and the

odic QA checks (JASTRO, 2004; Moyers, 1999).

 Accurate positio  Accurate positioning ning of the pati patient ent in the proto proton n beam requires location of the patient on the support system and precise operation and correct interaction of two systems: one that performs and controls the movements required to place the patient in the prescribed position, position, and another that verifies the position iti on of th the e pa pati tien entt re rela lati tive ve to th the e be beam am ax axis is (s (see ee Section 7). The geometric/positioning measurements are ar e pe perf rfor orme med d to qu quan anti tify fy th the e acc ccur ura acy of th the e patie pa tientnt-pos positi itioni oning ng dev devices ices and the pa patien tientt alig alignnment with respect to the nozzle and gantry. The periodic od ic ch chec ecks ks of co coin inci cide denc nce e of th the e pr prot oton on-b -bea eam m isocenter isocent er and the pati patient ent setup-laser setup-laser positi positions ons can be per perform formed ed usi using ng a pho phosph sphor or ima imagin ging g pla plate te (IP (IP)) (Terunuma et al., 2003). When radiation is applied to an IP, IP, an image of the radia radiation tion field is tempo temporarily rarily

recently published report from the IAEA, TRS 430 (IAE (I AEA, A, 20 2004 04), ), de deal al wi with th an im imag agee-ba base sed d th thre reeedimens dim ension ional al TPS and can be gen gener erall ally y app applie lied d to TPSs used in planning radiotherapy with protons. Howe Ho weve verr, em emph phas asis is sh shou ould ld be gi give ven n to sp spec ecifi ificc issues not covered by TG 53 (Fraass  et al., 1998) or TRS TR S 430 (IAEA, (IAEA, 200 2004). 4). The TP TPSs Ss use used d at proton proton center cen terss wit with h a pas passiv sive e sca scatte tterin ring g bea beam-d m-deli elive very ry syst sy stem em ar are e mos mostly tly bas based ed on sca scatte tter-c r-conv onvolu olutio tion n algorithms (Moyers, 1999). This method employs as inpu in putt da data ta th the e me meas asur ured ed de dept pth– h– do dose se ta tabl bles es,, off-axis profiles, and output factors for each energy and modu modulato latorr combi combinat nation. ion. Duri During ng comm commission ission-ing,, th ing the e dos dosime imetri tricc cha chara racte cteris ristic ticss of the pr proto oton n beams, such as depth doses, off-axis profiles, fieldsize fact factors, ors, modu modulati lation on fact factors ors for diff differen erentt field sizes, penumbra sizes, and beam ranges, should be measu mea sured red and rec record orded ed for val valida idatio tion n of the TP TPS S and also stored in a database to be referenced in future periodic QA checks (JASTRO, 2004; Moyers, 1999 19 99). ). Ag Agai ain, n, fo forr sc scan anni ning ng sy syst stem ems, s, th the e be beam am

stored stor ed in th the e IP. Th The e IP is th then en ex exp pos osed ed to th the e patient position laser. As a result, the radiation field produces a ‘positive’ image while the patient setup laser creates a ‘negative’ image of its position. The advantages of this method are direct measurements in a short time, with high resolution. The periodic QA checks of stereophotogrammetric positioning positi oning sys systems tems (Jones   et al., 1995) and digital x-ray x-r ay ima imagin ging g sy syste stems ms deal wit with h the cal calibr ibrati ation on of  charge-coupl chargecoupled ed device (CCD) camer cameras as and verifi verificacation ti on of th the e al align ignme ment nt of th the e ax axial ial xx-ra ray y im imag agin ing g system sys tem (Schr (Schreuder euder,, 2002; 2004). If autom automatic atic position ti onin ing g bas based ed on com compa pari rison son of di digi gita tall lly y re reco connstru st ruct cted ed ra radi diog ogrrap aphs hs (D (DRR RR)) an and d xx-rray da data ta is

dat data-ac quisition tionth procedur pro e emen is ent somewha some t pth– differen diff erent, t, and an da-acquisi requ re quir ires es the e cedure meas me asur urem t what of de dept h– dose do se curves and lateral profiles of pencil beams, requiring sma smallll-fiel field d dos dosime imetry try equ equipm ipment ent wit with h a goo good d  et al. spatial resolution (Pedroni , 2005). TG 53 (Fraass   et al., 1998) and TRS 430 (IAEA, 2004 20 04)) do no nott co cove verr th the e ca cali libr brat atio ion n of co comp mput uted ed tomo to mogr grap aphy hy (C (CT) T) im imag ages es an and d co conv nver ersi sion on of CT Hounsf Hou nsfeld eld un units its to pr proto oton n st stopp opping ing po powe wers. rs. Th The e uncertainty in the range of protons due to inaccurate calibration of CT images and beam hardening artifa art ifacts cts is es estim timat ated ed by Sch Schaf affne fnerr and Pe Pedr droni oni (1998) to be 1.1 percent (1 SD) of the total range in soft tissue and 1.8 percent (1 SD) in cortical bone

employed, a check of the accuracy of the positioning algorithm algorit hm shoul should d be done (JASTRO, (JASTRO, 2004; Lesyna, 2004). 200 4). Pa Patie tient nt pos positi itioner onerss can also inc includ lude e high high-precision robotic systems (de Kock, 2002; Schreuder, 2004) that are capable, in combination with gantry angulations, angula tions, of locat locating ing the treatment treatment isocent isocenter er at any point within the patient so as to direct the beam from any angle through that point. The accuracy of  robotic rob otic pa patie tient nt posi positio tionin ning g dep depend endss mai mainly nly on the capabil cap ability ity of the rob robot ot con contro troll and sa safet fety y soft softwa ware re rath ra ther er th than an on th the e me mech chan anic ical al acc ccur urac acy y of th the e couch, and therefore periodic quality control involves a substantial amount of software and safety checks (de Kock, 2004).

for th for the e CT sc scan anne nerr th tha at th the ey us used ed.. Gi Giv ven th the e import imp ortanc ance e of the ac accur curac acy y of th the e CT Hou Hounsfi nsfield eld  values, it is strongly recommended that regular chec ch ecks ks of th the e co cons nsis iste tenc ncy y of th the e CT Ho Houn unsfi sfiel eld d  values be performed.  A proton-beam TPS usually has automated features to design aperture shapes from target-tissue outlin out lines es and to des design ign com compen pensat sator/ or/bol boluse usess th that at provide pro vide dis distal-e tal-edge dge targe targett cov coverag erage. e. Bolus Boluses es tha thatt contr con trol ol pen penetr etrat ation ion of the bea beam m in non non-un -unifo iform rm phant ph antoms oms re rela lativ tive e to th the e sha shape pe of th the e dis distal tal sid side e of a sp spec ecifi ified ed ta targ rget et-t -tis issu sue e re regi gion on ar are e de desi sign gned ed using usi ng pr prese esett ma margi rgins. ns. Use User-s r-spec pecifie ified d amo amoun unts ts of  bolu bo luss ex expa pans nsio ion n to ac acco coun untt fo forr la late tera rall pr prot oton on sca sc att tter er an and d pa pati tien entt-po posi siti tion on un unce cert rtai aint nty y ar are e employ emp loyed ed in bol bolus us des design ign (se (see e Sec Sectio tion n 6). Tes ests ts using usi ng ac actua tuall tis tissue sue sam sample pless to sim simula ulate te typ typica icall trea tr eatm tment entss are per perfor formed med as a fina finall tes testt to ve verif rify y the th e des design ign of a bol bolus us int intend ended ed to com compen pensat sate e for

9.2 PATIENT POSITION POSITIONING ING AND IMMOBILIZATION

9.3

TREATMENT TREA TMENT-PLANN -PLANNING ING SYSTEM SYSTEMS S

al..   (1998) The re repor portt of Fr Fraas aasss   et al (1998) cont contains ains recommendations from AAPM Task Group 53 (TG 53)

137  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

Figure 9.3. Verific Figure Verificatio ation n of pati patient ent tre treatm atment ent in a scan scanning ning beam. The calc calcula ulated ted orth orthogon ogonal al dose pro profiles files (solid curv curves) es) and the measured data points (crosses) are shown (Coray  et al., 2002; reproduced with permission).

et al., 19 the ti the tiss ssue ue ir irre regu gular larit itie iess (M (Mil ille lerr   et 1999 99;; Schaffner and Pedroni, 1998). Patien Pa tient-sp t-specific ecific periodic QA check checkss for a beamdelive del ivery ry sy syst stem em usi using ng pas passiv sive-s e-sca catte tterin ring g tec techhniques niqu es inclu include de port portal al calib calibrat rations ions and veri verificat fication ion of beam-shaping devices (apertures and compensaet al., 2002; iTL, 2001; tors/bolus tors/ boluses) es) (Hees (Heese e   et

due to nuclear interactions. The model parameterizatio za tion n da data ta ar are e st store ored d in loo lookup kup tables, tables, and the therapy planning can predict absolute doses. There are no pat patient ient-spe -specific cific devic devices; es; ther therefor efore e the indi indi- vidual information on each patient treatme treatment nt is in the control-point sequence (Chu  et al., 1993; Lomax  et al., 200 2004) 4) tha thatt is use used d to cal calcul culat ate e dos dose e dis distri tri--

JA JASTR STRO, O,calculates 2004; 200 4; Ka Kacpe cpere rek, k, per 2003; 200 3; Mo Moyer yers, s, 199 1999). 9). The TPS the dose monitor unit (MU) at the reference point in the patient but at present the th e do dose se pe perr MU is in indi divi vidu dual ally ly ve veri rifie fied d in a phantom for each patient treatment portal at many facilities (iTL, 2001; JASTRO, 2004; Moyers, 1999). Prec Pr ecis isio ion n in ma manu nufa fact ctur urin ing g th the e bo bolu luss di dire rect ctly ly affe af fect ctss th the e di dist stal al pa part rt of th the e de dept pth– h– do dose se di dist stri ri-bution, whereas the errors in the patient aperture impinge on the lateral-dose distribution. Although milling machines provide high manufacturing precision, the manufactured beam-shaping devices for each patient patient shoul should d be check checked ed again against st TPS dat data a files by comparing aperture shapes and bolus thick-

but bution ions in lude a wa water ter ph phant om. Patie Pa tientspecifi cific QA  checks che ckss inc includ e man manual ualantom. calcul cal culat ation ions snt-spe of the th e cMU MU, , range rang e check checkss and, most impo important rtantly ly,, indep independe endent nt dose dos e cal calcul culat ation ionss bas based ed on the con contr trol ol file its itself elf,, which is directly compared with that calculated by the TPS. In addition, as a weekly QA check, a field is qua quasisi-ra rando ndomly mly sel select ected ed an and d th the e dos dose e dis distri tri-bution checked in a water phantom with an array of ionization chambers. The thickness of the water colu co lum mn ca can n be se sett to th the e req equi uirred dep eptth of  measur mea sureme ement. nt. Th The e wa water ter ph phant antom om is irr irrad adia iated ted,, using usi ng th the e sam same e con contr trolol-poi point nt seq sequen uence ce as for the patie pa tient nt tr trea eatme tment nt.. Th The e mea measur sured ed dos dose e pr profil ofiles es should be then compared with the dose distribution

nesses at pre-selected points. Improve Impr ovement ment in quali quality ty cont control rol of dose delivery can be achieved through the use of proton radiography that provides verification in the relative range vivo o   in th of pro protons tons   in viv the e pa patie tient nt (S (Schn chneid eider er and Pedroni, 1995). The use of PET monitoring to check proton range, dose localization, and stability of the trea tr eatm tmen entt du duri ring ng di diff ffer eren entt fr frac acti tion onss ha hass be been en et al., 1999; 2004; investi inve stigate gated d (En (Enghard ghardtt   et Hishikawa   et al., 2002; Parodi   et al. , 2002; 2007a; 2007b; 200 7b; 200 2007c) 7c).. Re Realal-tim time e PE PET T ima imagin ging g cou could ld be cons co nsid ider ered ed a po pote tent ntia iall lly y us usef eful ul to tool ol fo forr QA in proton therapy (see Section 7.4.2). The TPS described by Lomax  et al.   (2004) calculates the dose distribution for a spot scanning-beam technique from a superposition of individual pencil beams, taking into account the density information from fr om the cor corre respo spond nding ing CT sli slices ces.. An emp empiric irical al model mod el of the pencil pencil bea beam m is use used d tha thatt tak takes es int into o accou ac count nt the at atten tenua uatio tion n of the pri primar mary y pr proto otons, ns, effects effec ts of mult multiple iple Cou Coulomb lomb scattering, scattering, and losses

recalc reca lcul ulat ated ed by th the e TP TPS S us usin ing g a ho homo moge gene neou ouss medium instead of the patient CT data. Figure 9.3 shows two orthogonal dose profiles taken during a routine patient verification (Coray  et al., 2002). The solid sol id lin line e rep repre resen sents ts th the e ca calcu lcula late ted d pr profi ofile; le; th the e cros cr osse sess sh show ow th the e me meas asu ure red d do dose ses. s. As a qu qual alit ity y chec ch eck, k, th the e rou outi tine ne do dosi sime metr try y wi with th io ioni niza zati tion on chambers should agree with the expected dose from the trea treatment tment plan withi within n the user-e user-establ stablished ished tolerances. Verifications of dose delivery for a dynamic treatment technique using a fluorescent screen and al.., 2000 CCD cam camer era a (Bo (Boon on   et al 2000)) ha have ve shown good sensiti sens itivity vity with   . 1 mm sp spat atial ial re resol solut ution ion th that at allowed the detection of deviations of a few percent from the calculated dose distribution. If resp re spir ira ato tory ry-s -syn ynch chrron oniz ized ed irrrad ir adia iati tion on is employed, CT images used for treatment planning must also be taken in respiratory-gating mode. The same specifications for irradiation gating, synchronizing timing and thresholds, should be applied for acquisitio acqu isition n of trea treatmen tment-pl t-plannin anning g CT image images. s. As 138

 

QUALITY QUALI TY ASSUR ASSURANCE ANCE Table 9.1. Quality-assur Quality-assurance ance procedures for passive beam-delivery systems. systems. Daily checks † † † †

  Apertur Aperture e alignmen alignment; t; room lasers, interlocks interlocks;; communi communication; cation; patient-positioning patient-positioning system  Depth–dose and lateral profiles (range, entrance dose, uniformity of range modulation and Bragg-peak width, flatness, symmetry)  Dose monitor calibration, check of MU value under standard condition   Individual patient treatment treatment calibration and range checks

Weekly checks † † † †

  Patien Patient-positioni t-positioning ng and imaging systems   Beam-line apparatus   Respiratory-gating equipment  Dose delivered to randomly selected patients (comparison of planned dose distributions to those measured in a water phantom)

 Annual or scheduled inspection checks † † †

 x-ray patient positioning and alignment systems   CT Hounsfield number calibration   Compr Comprehensive ehensive tests of therapy equipment equipment W

  monitor chambers, timers, beam-delivery beam-delivery termination and control interlocks, stray stray radiat radiation ion exposure to patients, gantry isocenter, depth–dose and lateral profiles, baseline data for daily QA checks

Table 9.2. Additional quality-assurance procedures procedures for scanning beam-delivery systems. systems. Daily checks † † † †

 Dose rate and monitor ratios for the pencil beam  Performance of the beam-position monitors  Depth–dose curve of a pencil beam in a water phantom  Calibration of the primary dose monitor

Weekly Wee kly checks †

 Qualitative three-dimensional check of the outline and range of the dose distribution for one patient’s irradiation field in a water phantom

Half-yearly checks

 Calibration of the primary dose monitor and the phase space of the beam tunes  Annual or scheduled inspection checks †

†

 Check of the beam characteristics W  calibration of the whole dosimetry system, performance of the scanning system in terms of dose linearity and dose-rate dependence

an example, the following issues should be investigated gate d for resp respira iratorytory-sync synchron hronized ized irra irradiat diation ion to serv se rve e as be benc nchm hmar arks ks fo forr pe peri riod odic ic QA ch chec ecks ks (JASTRO, 2004): (1) differences between observed respiration signal and actual organ movement; (2) phase uncertainty at CT scanning for treatment planning to make reference images; (3) sett setting ing of th the e th thres reshol hold d lev level el of the extent extent of  the movement; (4) move movemen mentt of org organ an du durin ring g all allow owed ed per period iod for irradiation.

9.4 EXA EXAMPL MPLES ES OF OF PERIOD PERIODIC IC CHEC CHECKS KS Each proton facility will have different QA requirements, men ts, and the items listed listed in Table abless 9.1 and 9.2 are guidelines and suggestions and do not describe alll te al test stss th tha at mu musst be ma made de at an any y pa part rtic icul ular ar facility. In Table 9.1, the procedures for passive scattering beam-delivery beam-delivery sys systems tems (iTL, 2001; JAST JASTRO, RO, 2004; 200 4; Ka Kacpe cperek rek,, 200 2003; 3; Moy Moyers, ers, 199 1999; 9; Sch Schreu reuder der,, 2002) 200 2) are listed, listed, and in Table 9.2 add additi itiona onall pr proocedur ced ures es for spo spot-s t-scan cannin ning g bea beam-d m-deliv elivery ery sy syste stems ms (Chu   et al., 1993; Coray  et al., 2002) are given. The proced pr ocedur ures es for sca scanne nned d beam beamss are giv given en mos mostly tly in term te rmss of do dosi sime metr tric ic iss issue ues, s, as th the e che check ck of dos dose e

delivery is the most important task. 139  

Journal of the ICRU Vol 7 No 2 (2007) Report 78 Oxford University Press

doi:10.1093/jicru/ndm032

PRESCR SCRIBI IBING, NG, REC RECORD ORDING ING,, AND REP REPORT ORTING ING 10 PRE TREATMENT 10.1 INTR INTRODUC ODUCTION TION Previous Previo us sect section ionss of the pr prese esent nt re repor portt ha have ve dea dealt lt with the deve developmen lopmentt of plann planning ing aims, treatment treatment plans, plan s, and tr treat eatmen mentt pr prescr escript iption ions. s. The pr presen esentt section summarizes some of this material and formulates suggestions concerning the process of prescribing,, re ing reco cord rding ing,, an and d re repo port rtin ing g of pr prot oton on-t -the hera rapy py pr proc oced edur ures. As di disc scus sed d in Se Sect ction ion 2, all doses dose do sess should bees. specified asusse RBE-weighted absorbed in units of Gy (RBE).

10.2 GENERA GENERAL L RECOM RECOMMENDA MENDATIONS TIONS FOR PRESCRIBING, RECORDING, AND REPORTING 10.2.1 The compon 10.2.1 components ents of prescr prescribin ibing, g, recording, and reporting a patient’s treatment

Once the goals of ther therapy apy have been dete determine rmined, d, the information required to execute and document a pa patie tient’ nt’ss tr trea eatme tment nt mus mustt be ac accum cumula ulated ted and reco re cord rded ed.. Th This is in info form rmat atio ion n fa fall llss in into to se seve ven n categories: 1. Initial medical note

2. Planning aims 3. Treatment planning 4. Treatment prescription

5. Technical data

History of present illness, co-morbidities, physical examination, findings on imaging and pathological studies, and general management strategy.  All the informa information tion needed to plan the treatment of the patient. The process of simulating a number of delivery strategies for a radiation treatment and choosing the best one to use for treatment Instructions for treatment delivery to achieve the planned dose distribution and authorization of the technical details to deliver the treatment plan Data required for treatment delivery according to the treatment plan (with the prescription, treatment plan, doses and technical data

The re rela latio tion n amo among ng the these se cat catego egorie riess is sho shown wn schematically in Fig. 10.1. 10.2. 10 .2.2 2

Planni Pla nning ng aim aims s

The planning aims include the desired dose levels and the acc accept eptabl able e dos dose e gr gradi adient entss to th the e tar target get  volume(s) and the organs at risk. The responsible radiation oncologist gives these aims to the planner as a bas basis is for pla planni nning ng a tr trea eatme tment nt.. Th They ey re resul sultt from a medi medical cal decis decision-m ion-makin aking g proc process, ess, based on the detailed evaluation of the patient, and diagnostic studies, consultation(s) when indicated, oncological concepts, delineation of volumes such as gross tumor volume (GTV), clinical target volume (CTV), plann pla nning ing tar target get vo volum lume e (P (PTV) TV),, and org organ an at ris risk k (OAR). Consideration is commonly given to the use of on one e or mo more re th ther erap apeu euti ticc mo moda dalit litie ies. s. In so some me cases, the initial planning aims cannot be achieved, and an ite itera rativ tive e pr proce ocess ss tak takes es pla place ce in whi which ch th the e aims ai ms ar are e pr prog ogre ress ssiv ivel ely y ad adju just sted ed un unti till a pl plan an is designed that is clinically acceptable or the patient is referred management by other modalities. To develop thefor planning aims into an approved treatment plan requires involvement by a team of physicists, dosimetrists, technologists, and physician(s). It is recommended that the planning aims be part of the arch archived ived records, with nota notation tionss descr describin ibing g any com compr promi omises ses bet betwe ween en ini initia tiall and fina finall aim aimss and the reasons for such compromises. Settin Set ting g plan plannin ning g aim aimss and cons constra traints ints and usi using ng new or developing tools for scoring the evolving treatment me nt pl plans ans du durin ring g op optim timiza izatio tion n ar are e dis discu cusse ssed d in Section 6 (planning aims and plans in Sections 6.1 and 6.6, plan assessment in Section 6.7, plan comparison and optimization in Sections 6.8 and 6.9, comparing unif uniform orm-int -intens ensity ity ver versus sus int intensi ensity-m ty-modu odulat lated ed proton therapy treatment plans in Section 6.10). 10.2.2.1 10.2. 2.1

Specifying Specify ing planning aims

6. Record of the treatment 7. Report(s) of  the treatm treatment ent

approved and fixed). Storage of all data relevant to the patient’s treatment. For example, completion note, report to the referring and other physician(s), publications.

 As pointed out in Section 6.6.1.3, there is a wide  variety of ways to specify the dose requirements for the tumor (PTV) and the dose constraints for norma no rmall tis tissue suess (P (PR RV, RVR VR). ). In the cas case e of bot both h proton therapy and intensity-modulated radiation

# International Commission on Radiation Units and Measurements 2007

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

Figure 10.1. Block diagram, summarizing the relations among the planning aims, prescription, technical data, treatment treatment record and reports.

therapy, when specifying dose to the PTV, there is an increasing trend away from specifying doses at a single point,  viz., the ICRU Reference Point, and toward tow ard spec specifyin ifying g dose dose-sum -summari marizing zing quan quantiti tities es such su ch as th the e dos dose e del deliv iver ered ed to 50 per perce cent nt of th the e  volume of interest (VOI) ( D50%), the near-min near-minimum imum dose or the dose delivered to 98 percent of the VOI ( D98%), the near-maximum dose or the dose delivered to 2 percent of the VOI ( D2%), the least dose delivered to a specified volume,   V , of the VOI ( DV ),  etc . Additional information can be presented specifying the limits of dose inhomogeneities within the PTV and where minimum minimum doses may be located. All

1983; York, 2003) in the specification and evaluation of planning aims. When such models are used, the reference to the description of the model should be included in the record as well as the values of any selected parameters.  As stated in previous ICRU reports (ICRU, 1993b; 1999; 2004), the formulation of planning aims and hence the prescription of a treatment is the responsibility of the radiation oncologist. The goal of the present report and other ICRU reports is to suggest uniform approaches to the recording and reporting of pertinent treatment-related treatment-related information information..

of the pertinent information used in the definition of planning aims should be clearly recorded. Doses shoul sh ould d be spe specifi cified ed as RB RBEE-we weigh ighted ted abs absorb orbed ed doses [in Gy (RBE)]. In Section 6.7.3, it is pointed out that there is also growing interest in the use of a  variety of modelmodel-based based param parameters eters such as equiva equivallent uniform dose (EUD), tumo tumorr contr control ol proba probability bility (TCP), (TCP ), and normal normal-tissu -tissue e complic complication ation proba probability bility (NTCP) (NT CP) (Bortf (Bortfeld, eld, 2003; Lyman, 1985; Niemie Niemierko, rko, 1997; Niemerko and Goitein, 1993; Schultheiss  et al.,

10.2.2.2 10.2. 2.2

Normal-tissue Normaltissue const constraints raints

Normal-tissue constraints tend to be either dose or dose dos e –volu –volume me con const stra raint ints. s. In ‘se ‘seria rial-l l-lik ike’ e’ org organs ans (Withers   et al., 1988), the constraint will typically be th the e ma maxi ximu mum m do dose se to ev even en a sm smal alll vo volu lume me,, whereas for ‘parallel-like’ organs the relative size of  the th e vo volu lume me irr irrad adia iate ted d ab abov ove e th the e to toler leran ance ce le leve vell might be the most important parameter (for further discuss disc ussion ion see Sec Sectio tion n 5.6 5.6.1. .1.3). 3). In som some e ins instan tances ces,, 142

 

 PRESCRIBING, RECORDING, AND REPORTING A TREAMENT 

specific ificati ation on of nor normal mal-ti -tissu ssue e con constr straint aintss mig might ht spec be ba base sed d on pr pred edic icti tion onss of NT NTCP CP de deri rive ved d fr from om mathemati math ematical cal models models.. As noted above, all releva relevant nt parameter values of the model should be recorded.

2. the approved plan with its dose distribution(s); 3. the pr presc escrib ribed ed RB RBEE-we weigh ighted ted abs absorb orbed ed do doses ses and dose–volume constraints a the intended reference dose,   e.g.,   Dmed   or dose

Once a satisfactory been oped and accepted acce pted,, treatment the prescriptio prescriplan ption nhas and thedevelplan, in sufficient detail to allow accurate reconstruction, beco be come me pa part rt of th the e pr pres escr crip ipti tion on an and d sh shou ould ld be recorded (see Table 10.1).

at the ICRU Reference Point if specified b near near-m -min inim imum um do dose se to th the e PT PTV V ( D98%) (s (see ee Section 5.6.1.3) c ne near ar-m -max axim imum um do dose se to th the e PT PTV V ( D2%) (s (see ee Section 5.6.1.3) d   V  D  D,   i.e., the largest volume of a specified VOI that receives a dose more than or equal to the dose,  D RBE  (see Section 5.6.1.1); 4. the norm normal-t al-tissue issue cons constrai traints, nts,   e.g., th the e la larg rges estt dose ( DV ) to be received by a specified volume,  e.g., of normal tissues (see Section 5.6.1.2); 5. the the fra fraction ctionatio ation n schem scheme, e,   viz., th the e nu numb mber er of  fract fr action ions, s, int interer-fr frac actio tion n int interv erval, al, and ov over erall all treatment time; 6. the the med medica icall asp aspect ectss tha thatt af affec fectt ho how w the tr trea eattment is to be performed; 7. all technical data required to perform the treatment (see Section 10.3.4. and Table 10.2).

10.3

10.3.4 Appr 10.3.4 Approva ovall of the the prescr prescriptio iption n and technical data

10.2.2.3 Selecti 10.2.2.3 Selection on of of treatme treatment nt approa approach: ch: the trea treatment tment plan

Having established the planning aims, the oncology team tea m has two two,, ra rath ther er dif differ ferent ent,, app approa roache chess tha thatt can ca n be us used ed to de deli live verr pr prot oton on th ther erap apy y, su such ch as scattered-beam therapy and scanned-beam therapy with or without intensity modulation. If the necessary fa facil ciliti ities es ar are e av avail ailabl able, e, du durin ring g the pr proce ocess ss of  trea tr eatme tment nt pla plann nning ing,, the tea team m nee needs ds to eva evalua luate te which of these approaches and which plan would bestt ac bes achie hieve ve the pla plann nning ing aim aims. s. Th The e pr proce ocess ss and available tools are described in Sections 6.7–6.10.

10.3.1 10.3 .1

PRESCRIBING PRESCR IBING PROT PROTON-BEAM ON-BEAM THERAP THERAPY  Y  General Gene ral appro approach aches es to prescr prescribing  ibing 

 As point pointed ed out in Secti Section on 10.2.2 10.2.2.1, .1, ther there e is a variety of ways in which the oncology team can choose to expr ex press ess th the e pl plan anni ning ng ai aims ms fo forr a giv given en pa patie tient nt or group of patients, and indeed a variety of different approaches are currently in use in different centers. The Th e ge gene nera rall re recom comme mend ndat atio ions ns fo forr pr pres escri cribi bing ng,, recording, and reporting conventional external-beam radiation therapy (ICRU, 1993b; 1999; 2004) largely appl ap ply y to pr prot oton on th ther erap apy y. Th The e ma majo jorr di diff ffer eren ence ce between prescribing for the former and the latter lies in the use of dose–volume reporting and the amount of technical data required for proton therapy. 10.3.2 10.3 .2

The pre prescrip scription tion

The prescription must give sufficient technical detail

The execution of the prescribed treatment requires the specification and recording of a large volume of  technical data for each treatment field, particularly for sca scann nneded-bea beam m th thera erapy py.. Tabl able e 10. 10.2 2 lis lists ts th the e exam ex ampl ples es of da data ta th that at sh shou ould ld be re reco cord rded ed an and d repor re ported ted.. Exa Exampl mples es of suc such h tec techni hnical cal da data ta are as follows: †

†

†

 for scattered-beam therapy, all appropriate beamforming parameters such as the aperture shapes or multi-leaf collimator settings, the compensator design, and the range modulation;  for wobbled-beam therapy, the scan pattern, the penc pe ncil il be beam am si size ze an and d sp spa aci cing ng,, th the e ap aper ertu ture re shapes or multi-leaf collimator settings;   for for sca scanne nnedd-bea beam m th thera erapy py (ei (eithe therr for uni unifor formmintensity inte nsity or inte intensity nsity-modu -modulate lated d ther therapy) apy),, the

extensive files detailing the sequence of scanned pencil pen cil bea beams, ms, ea each ch wit with h th their eir ene energy rgy,, we weigh ight, t, and position.

so tha thatt the rad radiat iation ion the therap rapist ists/t s/tech echnol nologis ogists ts can fully full y und unders erstan tand d the pr prescr escript iption ion and deli deliver ver the dose acc dose accord ording ing to the pr prescr escript iption ion.. Ap Appr prova ovall of the finall pr fina presc escript ription ion mea means ns app appro roval val of all tec techni hnical cal parameters chosen to implement the treatment plan.

 A physician’ physician’ss approva approvall of a treatmen treatmentt prescription covers not only the specific treatment plan, but also the th e tec techn hnica icall asp aspect ectss re requi quire red d to imp implem lement ent th the e presc pr escrip riptio tion. n. The tec techni hnical cal fa facto ctors rs can inc includ lude e ext xten ensi sive ve da data ta fil files es,, wh whic ich h ca cann nnot ot be rea eadi dily ly inspected by eye, nor can the compliance with the quality-a qual ity-assur ssurance ance meas measures ures be asses assessed sed for each trea tr eatm tment ent pla plan. n. Th The e re reali ality ty is tha thatt the ra radia diatio tion n oncologist accepts the calculations and statements of phy physicis sicists ts and engin engineers. eers. The resp responsib onsibility ility for

10.3.3 Gene 10.3.3 General ral rec recomme ommendat ndations ions for prescribing 

 At a minimum, the prescription should include and record (see also Table 10.2): 1. the delineations of the volumes of interest (GTV, CTV, PTV, OAR, RVR,  etc.); 143  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  Table 10.1. Summ Summary ary of the data required required in each of the thr three ee area areas: s: ‘plan ‘planning ning aims’, ‘pr ‘prescr escription iption’, ’, and ‘treatment ‘treatment record’. Required items are identified by colored table cells. Items specific to, or that contain information concerning, proton-beam therapy are, in addition, crosshatched.

144  

 PRESCRIBING, RECORDING, AND REPORTING A TREAMENT  Table 10.1.  Continued

145  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  note 1  For example, patient name, identifying code, and photograph (if any). note 2  For example, age, sex, and race. note 3  For example, patient’s address, person to contact, and referring physician. note 4  For example, site, histology, staging, TNM classification, and genetic characterization (if available). note 5  For example, prior tumors, prior therapies, pertinent diagnostic-imaging results, and laboratory findings. note 6  Curative/palliative intent; intention to use concomitant or subsequent courses of therapy (surgery, chemotherapy, other systemic therapies such as biological agents); identification of treatment protocol (if any). note 7  A   A patient might receive one or more courses of treatment, in which one course might be a regime of chemotherapy, or radiation therapy for a primary tumor. Courses might be separated in time, or might be sequential or concurrent with the course of radiation therapy under consideration. note 8   A course of radiation therapy might contain one or more sequential segments in which one segment, for example, might treat the primary tumor and regional nodes to a dose of 50 Gy (RBE), followed by a second segment in which the primary tumor is given a boost dose of 20 Gy (RBE) for a total tumor dose of 70 Gy (RBE). Each segment of a radiation therapy course is represented by one and only one plan. Generally, but not necessarily, each segment will involve a single modality ( e.g., protons) that must be specified. If, however, mixed modalities are used ( e.g., protons and photons), then the planning aims and the prescription need to supply separate information for each modality used. note 9  See Section 5.6.3 for definition of the prescription dose. note 10  The goal dose will usually be defined by a dose–volume requirement, phrased in terms of the prescription dose. For example ‘the entire PTV is to receive at least 95 percent of the prescription dose’. It might also include some form of homogeneity requirement, for example, ‘the dose within the PTV to be within –5 and þ 7 percent of the prescription dose’ (see Section 6.6.1). note 11  In some proton-planning approaches, because the lateral and depth margins need to be different, the beam is designed relative to the CTV, rather than the PTV, using the appropriate margins in each direction. In this case, the margins used should be individually specified. note 12   In exte external rnal-bea -beam m radi radiothe otherap rapy y, one or mor more e tre treatm atment ent fra fractio ctions ns are deli deliver vered. ed. A singl single e fra fractio ction n inv involv olves es the deli deliver very y of  radiation by (usually) several beams in sequence, usually over a short period of time (ranging from a minute to a fraction of an hour). More than one fraction might be delivered in a day. The fractionation scheme [ e.g.,  d RBE 2 Gy (RBE) per fraction] must be specified for each segment of the treatment separately, if different. In addition, the beam sequencing must be specified ( e.g., ‘all beams delivered at each fraction’). ¼

note 13  See Section 6.6.1 for examples. note 14   Generally, but not necessarily, each segment will involve a single modality ( e.g., protons) that must be specified. If, however, mixed modalities are used ( e.g., protons and photons), then the respective modality must be specified for the individual beams used in

the segment. note 15   Each segment will have its own target-volume goal dose, which must be established such that the target-volume dose for the overall course of radiation therapy is met. note 16  As for the target volume, the dose received by each OAR (or PRV) within each segment must, when the segments are combined, meet the OARs’ dose constraint(s) for the overall course of radiation therapy. note 17  The   The type of plan (uniform-i (uniform-inten ntensity sity or inte intensit nsity-mo y-modula dulated) ted),, the type of equi equipme pment nt ( e.g., gantry or fixed beam), the type of  beam delivery ( e.g., scattered or scanned), and the beam characteristics ( e.g., pencil beam width in air for scanned beams) must all be specified. note 18  If a particular beam arrangement ( e.g., five-field prostate protocol plan) is required, it should be specified. note 19  A photograph of the immobilized patient in the treatment position is highly desirable. Otherwise, a sketch is useful. note 20  The two-dimensional dose displays and DVHs required in this category should be available either in hard-copy ( i.e., printed on paper using a non-fading ink) or, in the case of an all-digital record, available as monitor displays. note 21   At a minimum, in addition to the achieved RBE-weighted absorbed doses corresponding to the target-volume dose aims, the near-minimum near-min imum ( D98%), near-maximum ( D2%), and median ( D50%) RBE-weighted absorbed doses (see Section 5.6.1) to the PTV should be recorded. note 22   At a minimum, for each OAR (and PRV) of clinical interest, the near-maximum RBE-weighted absorbed dose ( D2%), and the achieved RBE-weighted doses corresponding to the normal-tissue dose constraints should be recorded. note 23   The two-dimensional dose displays should be shown as RBE-weighted dose displays. When relative doses are displayed, the dose to which, say, the 100 percent level correspondence should be specified. The displays should have a clearly readable representation of the RBE-weighted absorbed dose levels corresponding either to RBE-weighted absorbed dose lines or to color-wash regions. The sections should be selected to sample adequately the target volume and clinically important normal tissues. If possible, the displays should feature sagittal and/or coronal sections as well as transverse sections. note 24  Dose–volume histograms should be shown as RBE-weighted absorbed dose displays (see Section 5.6.2.2). note 25  If any difference(s) between the delivered and intended treatments might be clinically significant, the explanation for the deviation and a description of the remedial action that was required should be recorded as well.

146  

 PRESCRIBING, RECORDING, AND REPORTING A TREAMENT 

Table 10.2. Lis Listt of some technical technical data that shoul should d be recor recorded. ded. Items that are spec specific ific to or concerning, concerning, proton-beam proton-beam therapy are shaded.

Continued

147  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  Table 10.2 Continued

148  

 PRESCRIBING, RECORDING, AND REPORTING A TREAMENT 

the existence and correctness of these data lies in the hands of the responsible physicist, who should formally forma lly app approv rove e thes these e tech technical nical data tha thatt shou should ld be included in the treatment record (Table 10.2). 10.4 ADDI ADDITION TIONAL AL ASPEC ASPECTS TS OF PRES PRESCRIB CRIBING, ING, RECORDING, RECOR DING, AND REPOR REPORTING TING 10.4.1 10.4 .1

Therapy Ther apy equi equipmen pmentt

In pr proton oton-be -beam am the therap rapy y, sev severa erall elem element entss of the equi eq uipm pmen entt ne need ed to be sp spec ecifi ifica call lly y pr pres escr crib ibed ed,, recorded, and reported. These include the following: †

†

additiona additi onall fac factor torss tha thatt are dis discus cussed sed in Sec Sectio tions ns 6 and 7: (i) allowance for dose uncertainties due to heterogen ero geneiti eities es in tiss tissue ue den densiti sities, es, and (ii) pos possib sible le errors in registration of the real or virtual compensatorr wi to with th th the e pa pati tien ent’ t’ss an ana ato tomy my.. As a re resu sult lt,, th the e margins in depth and lateral must be documented.

 the type of accelerator ( e.g., cyclotron or synchrotron tr on), ), as th the ey ma may y pr prod oduc uce e di diff ffer eren entt be beam am characteristics (see Section 3);  the treatment-delivery equipment ( e.g., gantry or fixed beam);

10.5 10.5.1 10.5 .1

RECORDING RECOR DING PROT PROTON-BEA ON-BEAM M THERAP THERAPY  Y  The tre treatm atment ent rec record ord

Good recording is the basis for good reporting, and only recorded data can be analyzed. Records should be comprehensive enough to allow us to understand and an d va vali lida date te co comp mpon onen ents ts of th the e tr trea eatm tmen entt an and d trea tr eatm tmen entt ou outc tcom ome. e. In Tab able less 10 10.1 .1 an and d 10 10.2 .2 examples exa mples of dat data a req require uired d for pres prescribi cribing, ng, reco recordrding, and reporting and examples of technical data,

†

  th the e bea beam m sha shapin ping g con configu figura ratio tion(s n(s)) use used d ( e.g., scanning or scattering nozzle), including patientand field field-spec -specific ific hard hardwar ware e such as collim collimator atorss and compensators;  motion tracking, if applicable.

respectively, are given. The tr trea eatme tment nt rec record ord sho should uld inc includ lude e all inf inforormation needed to characterize the patient’s status, to doc docume ument nt ho how w th the e pa patie tient nt wa wass int intend ended ed to be trea tr eate ted, d, an and, d, if di diff ffer eren ent, t, ho how w th the e pa pati tien entt was inter ali alia a, th trea tr eated ted.. Th This is inc includ ludes, es,   inter the e pa pati tien ent’ t’ss demograp demo graphic hic dat data a and tumo tumorr sta status, tus, the pre prescrip scrip-tion and the underlying technical data, the details of ho how w th the e the thera rapy py wa wass del deliv ivere ered, d, and fol follo low-u w-up p informat info rmation. ion. If diffe differenc rence(s) e(s) betw between een the inte intended nded and delivered treatment are clinically significant, a description of the deviation and of any adjustments made mad e sho should uld be re recor corded ded as we well. ll. Th The e pa patie tient’ nt’ss radiati rad iation-t on-thera herapy py reco record rd is an arch archival ival docu document ment that should be retained for at least as long as the law prescribes.

In tr trea eatm tmen ents ts us usin ing g mo more re th than an on one e ra radi diat atio ion n modality ( e.g., prot protons ons and phot photons), ons), the rele relevant vant deta de tail ilss of th the e eq equi uipm pmen entt of ea each ch mo moda dali lity ty us used ed should be recorded. 10.4.2 10.4 .2

Beam-shap Beam -shaping ing tech techniqu niques es

The patient’s prescription and record must describe the bea beam-d m-deli elive very ry tec techni hnique que use used d ( e.g., scat scatter tered ed beam or scanned beam, in the latter case, uniform-i unif orm-inten ntensity sity or inten intensitysity-modu modulate lated d beam) (see Section 3). For scattered beams, the details of  the bea beam-s m-shap haping ing sy syst stem em mus mustt be spe specifi cified ed and documente docum ented d ( e.g., ide identi ntifica ficatio tion n of the sca scatt tteri ering ng device dev ices, s, bea beam m ape apertu rture, re, and com compen pensa sator tor). ). Fo Forr scanned beams, the details of the scanning system and pen pencil cil-be -beam am pa param ramete eters rs sho should uld be re recor corded ded ( e.g., grid pattern, spacing of pencils, pencil widths, dwell times, and repainting specifications).

10.5.2 10.5 .2

The pat patient ient’s ’s rec record ord

The patient’s record is usually a combination of a paper pap er re recor cord d (th (the e pa patie tient’ nt’ss ra radio diothe therap rapy y ‘ch ‘chart art’) ’) and an electronic record containing a large volume of information (such as three-dimensional dose distrib tr ibut utio ions ns an and d sc scan an pa patt tter erns ns of pe penc ncil il be beam ams) s).. Hard copy is likely to remain highly prevalent for some som e tim time. e. Ho Howe weve verr, th the e re recor cord d ma may y be ent entire irely ly electronic, in which case provision should be made to ensure that changing technology does not render the records inaccessible.

10.4.3 Techn 10.4.3 echnique iques s for for dealing  dealing  with hetero heterogeneities geneities

Heterogeneitie Heterogen eitiess in tissu tissue e dens densities ities pene penetra trated ted by proton beams can substantially affect proton range and dose homo homogenei geneity ty.. Thes These e hete heterogen rogeneitie eitiess can be compensated for using real (physical) or virtual (embedded (embe dded in the scann scanning ing sequ sequences ences)) comp compensaensatorss (se tor (see e Sec Sectio tion n 6.4 6.4.6. .6.2). 2). The det detail ailss of the compensation scheme should be recorded.

10.6 REPOR REPORTING TING THE TREA TREATMENT TMENT OF A SINGLE PATIENT 10.6.1 10.6 .1

Reporti Rep orting ng proton proton-bea -beam m therapy therapy

10.4.4 10. 4.4 Mar Margin gins s

In contrast to the might prescription, in which institutional freedom be desirable, thesome reporting of tr trea eatme tments nts mu must st be don done e usi using ng un unifo iform rm ter terms ms

Consideration of the plan is likely to include all the factors common to photon-beam therapy plus several 149  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

concep cepts ts for all pa patie tients nts wit within hin ea each ch dep depart art-and con ment and in all centers. If this is not followed, any useful exchange of scientific or clinical information between betw een cent centers ers becom becomes es diffi difficult cult or even impossible. Wi Withou thoutt modi modifying fying the trea treatmen tmentt tech techniqu niques es and the dose levels currently delivered in different centers, adopting the same concepts and definitions forr bo fo both th pr pres escr crib ibin ing g an and d re repo port rtin ing g re redu duce cess th the e risks of confusion. Reports normally have a focus either on a single patie pa tient, nt, gr group oupss of pa patie tients nts,, or on val valida idatio tion n and

10.6.2 10. 6.2.2 .2

comparison of technologies. The content of purpose a report depends on the intended recipient and the for which it is to be used. A completion of treat-

†

The com comple pletion tion not notee

The completion note is a summary report made by the trea treating ting physician physician and placed in the patient’s patient’s chart ch art and re reco cord rd.. A co copy py is us usua uall lly y se sent nt to th the e referr ref erring ing ph physi ysicia cian(s n(s). ). It nee needs ds to inc includ lude e tho those se aspect asp ectss of th the e pr proto oton n tr trea eatme tment nt tha thatt can bec become ome important impo rtant basebase-line line info informat rmation ion for possi possible ble consecutive therapies (Gre´goire  et al., 2004). It sho should uld,, at a min minimu imum, m, inc includ lude e the fol follow lowing ing elements:   diagnos diagnosis, is, pat patholog hological ical type and grad grade, e, stag stage e of  disease, and relevant prior history;

ment note, a report to referring physician(s) (which might be a copy of the completion note), a report sent se nt to a ph phys ysic icia ian n wh whom om th the e pa pati tien entt ha hass co connsulted sul ted,, or a rep report ort sen sentt to a ra radia diatio tion n onc oncolo ologis gistt who may wish to evaluate the patient for re-trea re-t reatmen tment, t, all requ require ire diff differen erentt leve levels ls of deta detail il and completeness. The reporting of results is of special importance in pro proton ton rad radiati iation on ther therapy apy,, as the use of prot protons ons has only rece recently ntly becom become e wide widespre spread. ad. Ther Therefor efore, e, detail det ailed ed re recor cordin ding g and rep report orting ing is par partic ticula ularly rly

†

†

†

 a summary of the overall treatment plan;   an ap appr propr opriat iate e sum summar mary y of the RB RBE-w E-weig eight hted ed absorb abs orbed ed dos doses es to the PT PTV V, CT CTV V, and GT GTV V, as well as the nearnear-minim minimum um and near near-maxi -maximum mum doses delivered to the CTV and GTV; GTV;  RBE-weighted absorbed doses to PRVs or OARs, a statement of any complications, and deviations from the planned treatment;   a st stat ateme ement nt of the pa patie tient’ nt’ss st stat atus us up upon on com com-pletion of therapy and plans for future follow-up.

10.6.2.3 10.6. 2.3 Summary repor reportt to the referrin referring g  physician

needed. Otherwise the determination of the role of  proto pr oton n th ther erapy apy as a com compon ponent ent of mod modern ern can cancer cer trea tr eatme tment nt wil willl be ver very y dif difficu ficult. lt. Pr Previ evious ous ICR ICRU U repor re ports ts ha have ve ide ident ntifie ified d thr three ee lev levels els of rep report orting ing,, 1 th thro roug ugh h 3 (I (ICR CRU, U, 19 1999 99). ). For th the e mo mosst pa part rt,, proto pr oton n ra radia diatio tion n th thera erapy py sho should uld be rep report orted ed at level 3.

The no The note te to th the e re refe ferr rrin ing g ph phys ysic icia ian( n(s) s) wo woul uld d in most cases be a copy of the completion or treatment note. In certain situations, the note might only be an abbreviation of the completion note. 10.6.2 10. 6.2.4 .4

10.6.2 10.6 .2

Patien Pa tient-spe t-specific cific rep reports orts

 A case report provides a detailed description of a patient’ss treatment suitable for case presentat patient’ presentations ions at con confer ferenc ences es or for wri writte tten n tr trans ansmis missio sion n of the deta de tails ils of an in indi divi vidu dual al pa pati tien ent’ t’ss tr trea eatm tmen entt to knowle kno wledg dgeab eable le pr profe ofessi ssiona onals. ls. Exa Examp mples les of Ca Case se

There Ther e ca can n be se seve vera rall ty type pess of pa pati tien entt-sp speci ecific fic repor re ports, ts, dep depend ending ing on the int intend ended ed aud audien ience. ce. It would not be appropriate to enumerate all requirements for these reports as they should be tailored to the particular needs of each situation. However, some general points are followed for some common types of report. 10.6.2 10. 6.2.1 .1

Casee rep Cas report ort

Reports are provided in Appendix B. 10.6.2.5 10.6. 2.5

Detailed repor reportt to a physician physician

Depending on their future role in the patient treatment, other physicians should be provided with all relevant relev ant infor informat mation ion the they y migh mightt requ require, ire,   e.g., fo forr evaluation of a complication or for consideration of  possible further therapy therapy..

The initi initial al medical medical note note

The initial medical note presents various categories of pa patie tient nt inf inform ormat ation ion.. Fo Forr ex examp ample, le, his histor tory y of  present pres ent illne illness, ss, inclu including ding age, gend gender er,, rac race, e, sym sympptoms to ms,, me medi dica call ev eval alua uati tion onss (i (inc nclu ludi ding ng im imag agin ing g studies and pathology, and treatments, if any), past medical medi cal his history tory,, espec especially ially auto autoimmu immune ne dise diseases; ases; geneti gen eticc dis diseas eases es ass associ ociat ated ed wit with h inc incre rease ased d rad radii-

10.7 REPOR REPORTING TING PROT PROTON-BEAM ON-BEAM THERAP THERAPY  Y  FOR A SERIES OF PATIENTS

ati ation on history; sensitivi sens itivity; ty; curr current ent medi medicati cations; ons; allergies; aller gies; social family history; physical examination; review revi ew of imagi imaging ng stu studies dies;; revi review ew of pat patholog hological ical findings; and general management plan.

Reporting results proton-beam therapy is very similar simi lar to the reporti rep orting ng of any other oth er rad radiat iation ion the thera rapy py modality. The reader is referred to ICRU Reports 50, 62, and 71 (ICRU, 1993b; 1999; 2004) for details. 150

 

Journal of the ICRU Vol 7 No 2 (2007) Report 78 Oxford University Press

doi:10.1093/jicru/ndm033

 APPENDIX A: IMPLEMENT IMPLEMENTA ATION OF THE TRS 398 CODE OF PRACTICE FOR IONIZATION CHAMBER DOSIMETRY 

 A.1 CALIBR CALIBRA ATION OF IONIZA IONIZATION TION CHAMBERS When an ionization chamber or dosimeter is sent to a standards laboratory for calibration, stabilitychec ch eck k me meas asur urem emen ents ts (u (usi sing ng a su suit itab able le ch chec eck k device dev ice)) sho should uld be car carrie ried d out by the user bef before ore and after the calibration. This will ensure that the cham ch ambe berr re resp spon onse se ha hass no nott be been en af affe fect cted ed by th the e trans tr anspor porta tatio tion. n. A ref refere erence nce ion ioniza izatio tion n cha chambe mberr should sho uld be cal calibr ibrat ated ed at a re refer ferenc ence e qua qualit lity y   Q0   at intervals not exceeding 2 or 3 years, or whenever the th e us user er su susp spec ects ts th that at th the e ch cham ambe berr ha hass be been en damaged. dama ged. If dir directly ectly measured measured value valuess of   kQ,Q0 (or  N  D,w,Q0) for th the e cha chambe mberr ha have ve bee been n obt obtain ained ed pr pree viously,, a recalibra  viously recalibration tion to verify the quality dependence of the chamber should be made at least every thir th ird d ti time me th that at th the e ch cham ambe berr is ca cali libr brat ated ed.. Th The e chamber cham ber shou should ld be reca recalibr librated ated at all qual qualities ities at least lea st ev every ery 6 ye years ars.. It is the respons responsibi ibilit lity y of the user to incr increase ease the freq frequenc uency y of the calibration calibrationss for cha chambe mbers rs who whose se lon long-t g-term erm st stabi abilit lity y has not been verified over a period exceeding 5 years.

to water at a PSDL (Primary Standards Dosi Do sime metr try y La Labo bora rato tory ry)) or or,, mo more re co comm mmon only ly,, aga gain insst a sec econ onda dary ry stan and dar ard d at an SSDL (Seco (S econd ndary ary Sta Standa ndards rds Dos Dosime imetry try Lab Labor orat atory ory). ). Only the latter case will be discussed here. It is assumed that the absorbed dose to water,   Dw, is known at a depth of 5 g cm22 in a water phantom phan tom for   60Co gam gamma ma ra rays ys.. Th This is is re realiz alized ed at the SSDL by means of a calibrated cavity-io cavi ty-ioniza nization tion cham chamber ber perf performin orming g measu measurerements in a water phantom. The user chamber is plac pl aced ed wi with th it itss ref efer eren ence ce po poin intt at a de dept pth h of  22

5 g cm  N  in   is a water phantom, factor obtained from and its calibration  D,w

 N  D;w  ¼

 Dw  ;  M 

 

ð A :1Þ

Calibr Cal ibrat ation ionss ma may y be car carrie ried d out eit eithe herr dir direct ectly ly

where   M   is the dos dosime imeter ter rea readin ding g cor corre recte cted d for influen infl uence ce qua quanti ntitie ties, s, so as to cor corre respo spond nd to the refer re ferenc ence e con condit dition ionss for whi which ch th the e cal calibr ibrat ation ion factor is va vali lid d. Refe ferrence co con ndi dittio ion ns recomme om mend nded ed fo forr th the e ca cali libr bra ati tion on of io ioni niza zati tion on

agai ag ains nstt a pr prim imar ary y st stan anda dard rd of ab abso sorb rbed ed do dose se

chambers in   60Co are given in Table A.1.

 A.1.1

Calibration in a   60Co beam

The te The text xt an and d ta tabl bles es in th this is Ap Appe pend ndix ix ar are e ta take ken n al almo most st exclusiv excl usively ely from secti sections ons of TRS 398 (IAE (IAEA, A, 2000 2000), ), whic which h is therefore therefor e not further referenced. # International Commission on Radiation Units and Measurements 2007

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  Ta Table ble A.1. Refe Referen rence ce condi conditions tions recommended recommended for the calibration calibration of ioniza ionization tion chambers in   60Co gamma radiation at standards laboratories or for cross calibrations in user laboratories (IAEA, 2000).

Quantity

Reference value or reference characteristic

Phantom material

Water

Phantom size Source-chamber distancea (SCD)  Air temperatureb  Air pressure Ref efer eren ence ce po poin intt of th the e io ioni niza zati tion on ch cham ambe berr

30  30  30 cm3 (approximately) 100 cm c 20 C 101.3 kPa For cy cyli lind ndri rica call ch cham ambe bers rs,, on th the e ch cham ambe berr ax axis is at th the e ce cent nter er of  the cavity volume For plane-parallel chambers on the inner surface of the entrance window, at the center of the window 5 g cm22 10  10 cm2 50 percent No reference values are recommended, but the values used should be stated in the calibration certificate. No reference values are recommended, but the dose rate used should always be stated in the calibration certificate. It should also be stated whether or not a recombination correction has been applied, and, if so, the value should be stated. 8

Depth in phantom of the reference point of the chambera Field Fiel d size at the posit position ion of the ref refere erence nce poin pointt of the cham chamber ber Relative humidity Polarizing voltage and polarity Dose rate

a

 After a wate waterr phantom with a plastic window has been filled, its dimension dimensionss may slowly change with time. When using a horizontal beam, it may therefore be necessary to check the source–surface distance and the chamber depth every few hours. bThe

temperature of the air in a chamber cavity should be taken to be that of the phantom, which should be measured; this is not necessarily the same as the temperature of the surrounding air. c In some countries the reference air temperature is 22 C. 8

152  

 APPENDIX A

 A.2

REFERENCE REFERE NCE DOSIMETRY IN THE USER

chamb cha mber er and som someti etimes mes on th the e pol polari arity ty,, mig might ht

PROTON BEAM

require several (up to 20) minutes. Indeed, failure to do so can result in errors that are larger than the effect for which one is correcting. The lea leakag kage e (an (and d ba backg ckgrou round nd)) cur curren rentt is th that at gener gen erat ated ed by th the e com comple plete te mea measur suring ing sy syst stem em in the absence of applied radiation. Leakage can also be rad radiati iation-i on-induc nduced, ed, and cham chambers bers can show no leakage leaka ge prior to irra irradiat diation ion yet have a signi significant ficant leak le akag age e af afte terr ir irra radi diat atio ion. n. Th The e le leak akag age e cu curr rren entt shou sh ould ld al alw way ayss be me meas asur ured ed be befo fore re an and d aft fter er irradiation, and should be small compared with the current obtained during the irradiation (less than  0.1 percent of the measurement current and normall ma lly y of th the e sa same me si sign gn). ). In so some me in inst stan ance ces, s, fo forr example, small-volume chambers used at low dose rates, the relative leakage current might be larger. If this is the case, the measurement current should be cor corre recte cted d for lea leakag kage, e, pa payin ying g at atten tentio tion n to the sign si gn of th the e le leak akag age e cu curr rren ent. t. Ch Cham ambe bers rs wi with th a leakage current that is larger than    1 percent of  the th e mea measur sureme ement nt cur curre rent, nt, or is var variab iable le in tim time, e,

 A.2.1 Determination of the absorbed dose to water

It is as assu sume med d th that at th the e us user er ha hass an io ioni niza zati tion on chambe cha mberr or a dos dosime imeter ter wit with h a cal calibr ibrat ation ion fa facto ctorr  N  D,w,Q0 in te term rmss of ab abso sorb rbed ed do dose se to wa wate terr at a reference-beam reference-be am quality   Q0. Fo Follo llowin wing g th the e for formal mal-ism given in Sec Sectio tion n 4.4 4.4.2, .2, the cha chambe mberr is pos posiitioned tio ned acc accord ording ing to the re refer ferenc ence e con condit dition ionss and the absorbed dose to water in the proton beam is given by  Dw;Q  ¼  M Q N  D;w;Q0 kQ;Q0 ;

 

ð A :2Þ

where  M Q  is the reading of the dosimeter incorporating at ing the pr produ oduct ct   Pki   of co corr rrec ecti tion on fa fact ctor orss fo forr influence influ ence quan quantiti tities, es, and   kQ,Q0 is the cor correc rectio tion n factor that corrects for the difference between the reference-beam quality  Q 0  and the actual quality  Q being used.

should not be used. When relative relative measu measureme rements nts are carrie carried d out in accelerator beams, it is strongly recommended that an additional monitoring dosimetry system be used during duri ng the exp experime erimental ntal procedure procedure to acco account unt for fluct flu ctua uati tion onss in th the e ra radi diat atio ion n ou outp tput ut.. Th This is is especially important when ratios of dosimeter readings are used ( e.g. for cross calibrations in different beams, for measurements using different polarities, or var varyin ying g vo volta ltages ges,,   etc.). Th The e ex exter ternal nal mon monito itorr shou sh ould ld pref efer era abl bly y be pos osiiti tion one ed wit ithi hin n th the e phant ph antom, om, alo along ng the maj major or axi axiss of th the e tr trans ansve verse rse plane, at the same depth as the chamber and at a distance of    3 or 4 cm from the centr central al axis. This This

 A.2.2 Practical considerations for fo r measurements in the proton beam

Unless the ionization chamber is designed so that it can be put directly into water, it must be used with wit h a wa water terpr proof oof sle sleev eve. e. Th The e sle sleeve eve sho should uld be made of PMMA, with a wall sufficiently thin (preferably erab ly not   . 1. 1.0 0 mm in th thick ickne ness) ss) to all allow ow the chamber to achieve thermal equilibrium with the water in   , 10 min. The sleeve should be designed so as to allow the air pressure in the chamber to reach ambient air pressure quickly; an air gap of  0.1–0.3 mm between the chamber and the sleeve is adequate. To reduce the buildup of water vapor around the chamber, the waterproof sleeve should not be lef leftt in wa water ter longer longer th than an is ne neces cessar sary y to carry out the measurements. Additional accuracy is ga gain ined ed if th the e sa same me sl slee eeve ve th that at wa wass us used ed fo forr the calib calibrat ration ion of the chamber in the standards standards lab abor ora ator ory y is al alsso used for all sub ubsseq equ uen entt measurements. Before Befo re meas measurem urements ents are made made,, the sta stabilit bility y of  the th e do dosi sime mete terr sh shou ould ld be ve veri rifie fied d us usin ing g a ch chec eck k sour so urce ce.. En Enou ough gh ti time me sh shou ould ld be al allo lowe wed d fo forr th the e dosi do sime mete terr to re reac ach h th ther erma mall eq equi uili libr briu ium. m. So Some me mains-powered electrometers are best switched on for at least 2 h before use to allow stabilization. It is alwa always ys advis advisable able to pre pre-irra -irradiat diate e an ioniz ionizati ation on cham ch ambe berr wi with th 2– 5 Gy to ac achi hiev eve e ch char arge ge eq equi uili li-brium bri um in th the e dif differ ferent ent ma mater terial ials. s. It is esp especi ecially ally important to operate the measuring system under stable conditions whenever the polarity or polarizing voltage are modified, which, depending on the

might not be possible for a small beam such as an ocular beam. In that case the use of a transmission monito mon itorr cha chambe mberr loc locat ated ed do downs wnstr tream eam fr from om th the e final collimator is recommended. If the monitor is positioned posit ioned in air air,, the possi possible ble temp tempera erature ture drifts should be taken into account (IAEA, 2000).  A.2.3

The ca The cali libr bra ati tion on co coef effic ficie ient nt fo forr an io ioni niza zati tion on chamber is valid only for the reference conditions that apply to the calibration. Any departure from the reference conditions when using the ionization chamber in the user beam should be corrected for using appropriate factors.  A.2.3.1

Temperature, pressure, and humidity

 As all chambers recommende recommended d in the present report are open to ambient air, the mass of air in the th e ca cavi vity ty vo volu lume me is su subj bjec ectt to atm tmos osph pher eric ic 153

 

Corrections for influence quantities

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

 variations. The correction factor kTP  ¼

ð273:2 þ T Þ P0 ð273:2 þ T 0 Þ P

 

taken to be the mean of the absolute values of readings taken at both polarities. For the routine use of  a giv given en ion ioniza izatio tion n cha chambe mberr, a sin single gle pol polari arizin zing g pote po tent ntia iall an and d po pola lari rity ty is no norm rmal ally ly ad adop opte ted. d. Howe Ho weve verr, th the e ef effe fect ct on th the e ch cham ambe berr re read adin ing g of  using polarizing potentials of opposite polarity for each user beam quality   Q   can be accounted for by using a correction factor

ð A :3Þ

should be applied to convert the cavity air mass to the reference conditions. In Eq. (A.3),   T   and   P   are the cavity temperature and air pressure at the time of the measurements, and   T 0   and   P0   are the reference en ce val value uess (gen (gener erall ally y 20 or 22 C and 101. 101.3 3 kPa). kPa). The Th e te temp mper erat atur ure e of th the e air in a ch cham ambe berr ca cavit vity y should be taken to be that of the phantom, which should be measured; this is not necessarily the same as th the e te temp mper erat atur ure e of th the e su surr rrou ound ndin ing g air air.. Fo Forr measure meas uremen ments ts in a wa water ter pha phanto ntom, m, the cha chambe mberr water wa terpro proof of slee sleeve ve sho should uld be ven vented ted to the at atmos mos-phere in order to obtain rapid equilibrium between the ambient air and the air in the chamber cavity. No cor correc rectio tions ns for hum humidi idity ty are nee needed ded if the calibr cal ibrat ation ion coe coeffic fficien ientt wa wass re refer ferre red d to a rel relat ativ ive e humidi hum idity ty of 50 per percen centt and is use used d in a rel relat ativ ive e 8

kpol  ¼

 

ð A :4Þ

where   M þ   and   M 2   are the elect electrome rometer ter read readings ings obtained obtai ned at posi positive tive and nega negative tive polar polarity ity,, resp respecectively, and   M  is the electrometer reading obtained with wi th th the e po pola lari rity ty us used ed rou outi tine nely ly (po (posi siti tiv ve or negative). The readings   M þ   and   M 2  should be made with care, ensuring that the chamber reading is stable follo fol lowin wing g any cha change nge in pol polari arity ty (so (some me cha chambe mbers rs can tak take e up to 20 min to st stabi abiliz lize). e). To min minimi imize ze

humidity humidi ty bet betwe ween en 20 and 80 per percen cent. t. If th the e cal caliibrat br ation ion fa facto ctorr is re refer ferre red d to dry air air,, a cor correc rectio tion n   60 fact fa ctor or sh shou ould ld be ap appl plie ied; d; fo forr Co calib calibrat rations ions,,  K hum 0.997 (ICRU, 1998). hum

th the e lerators, influ in fluen ence ceit of fluct flu ctua uati tion ons s in the th etheou outp tput utings of  accelera acce tors, is pre prefera ferable ble that tha t all readings read be no norm rmal aliz ized ed to th that at of an ex exte tern rnal al mo moni nito torr. Ideally, the external monitor should be positioned approximately at the depth of measurement, but at a di dist stan ance ce of 3– 4 cm fr from om th the e ch cham ambe berr ce cent nter er along the major axis in the transverse plane of the beam. When Wh en th the e ch cham ambe berr is se sent nt fo forr ca cali libr bra ati tion on,, a decision is normally made, either by the user or by the calibration laboratory, on the polarizing potential and polarity to be adopted for the routine use of the chamber. chamber. The calibration calibration should be carri carried ed out at this polarizing potential (and polarity, if only

¼

 A.2.3.2

j M   M þ j þ j M   M  j   ; 2 M 

Electrometer calibration

When the ionization chamber and the electrometer are calib calibrat rated ed sepa separat rately ely,, a calib calibrat ration ion coeffi coefficient cient for each is given by the calibration laboratory. The electrometer calibration coefficient   kelec   is an influence quantity and is included in the product   Pki   of  correctio corr ection n fac factors. tors. Typica ypically lly,, the calibr calibrati ation on coeffi coeffi-cient   N  D,w   for th the e ion ioniza izatio tion n cha chamb mber er is giv given en in 2

units of Gy nC 1 and that for the electrometer   kelec 21 eith ei ther er in un unit itss of nC rd rdg g (rdg electrometer reading) or, if the electrometer readout is in terms of charge, as a dimensionless factor close to unity. If the ioniz ionizati ation on chamb chamber er and the elect electrome rometer ter are ar e cal calibr ibrat ated ed tog togeth ether er,, the then n the com combin bined ed cal caliibration coefficient   N  D,w   will will typ typica ically lly be giv given en in 21 21 units of Gy rdg or Gy nC (dependi (dep ending ng on the electrometer readout) and no separate electrometer calibration factor   kelec   is required. In this case, a  value for   kelec   of unit unity y (dim (dimensio ensionless nless)) shoul should d be recorded in the worksheets.

one polarity is used for the calibration), or, if not, clearly stated. The calibration laboratory might or might not correct for the polarity effect at the calibration bra tion quali quality ty   Q0. Th This is sh shou ould ld be st stat ated ed in th the e calibration certificate. If the cal calibr ibrat ation ion lab labor orat atory ory has alr alread eady y cor cor-rected rec ted for th the e pol polari arity ty eff effect ect,, the then n th the e use userr mus mustt apply appl y the corr correctio ection n fac factor tor   kpol   derived derived using Eq. (A.4) to all measurements made using the routine cor-polarity polar ity.. If the calibr calibrati ation on labor laborato atory ry has   not cor rected  for the polarity effect, the subsequent treatment of the polarity effect depends on the facilities available to the user, and on what beam qualities must be measured.

¼

 A.2.3.3 Polarity effect The effect on a chamber reading of using polarizing pote po tent ntia ials ls of op oppo posi site te po pola lari rity ty mu must st al alwa ways ys be checke che cked d on com commis missio sionin ning. g. In cha charge rgedd-par partic ticle le beams bea ms the eff effect ect can be sig signifi nifican cantt (IA (IAEA EA,, 200 2000). 0). When a chamber is used in a beam that produces a meas me asur urab able le po pola lari rity ty ef effe fect ct,, th the e tr true ue re read adin ing g is

 A.2.3.4

Ion recombination

The incomplete collection of charge in an ionizationchamber cavity owing to the recombination of ions requires the use of a correction factor. Two separate effe ef fects cts ta take ke pl plac ace: e: (i (i)) th the e re reco comb mbin inat atio ion n of ion ionss 154

 

 APPENDIX A

formed by separate ionizing particle tracks, termed general (or volume) recombination, which is dependent on the density of ionizing particles and therefore on the dose rate; and (ii) the recombination of  ionss for ion forme med d by a sin single gle io ioni nizi zing ng pa part rtic icle le tr trac ack, k, referr ref erred ed to as ini initia tiall rec recomb ombina ination tion,, whi which ch is ind indeependent of the dose rate. Both effects depend on the chambe cha mberr geo geomet metry ry and on the app applie lied d pol polariz arizing ing  voltage.  voltag e. For beams other than heavy ions, initial recombination is generally , 0.2 percent. Forr pul Fo pulsed sed beams, beams, it is re recom commen mended ded tha thatt the correction corre ction fact factor or   ks   be de deri rive ved d us usin ing g th the e tw twoo voltage method. This method assumes a linear V  and uses the measured dependen depe ndence ce of 1/  M   on 1/ V   values of the collected charges   M 1   and   M 2   at the polari pol arizin zing g vo volta ltages ges   V 1   and   V 2, re respe specti ctivel vely y, measured using the same irradiation conditions.  V 1 is the nor normal mal ope opera ratin ting g vo volta ltage ge and   V 2   a low lower er V 2   should ideally be equal to  voltage; the ratio   V 1 / V  or lar larger ger tha than n 3. Str Strict ictly ly,, the pol polari arity ty eff effect ect wil willl 2   should  ect M 1  usi   M Eq cha change nge with hrecte the voltag tage, e,isand and each ea ch bewit corre cor cted dvol for this th effect eff using ng Equa uatio tion n (A.4). The recombination correction factor   ks  at the normal operating voltage  V 1  is obtained from

ks  ¼ a0 þ a1

2

     M 1  M 2

þ a2

 M 1  M 2

;

 

ð A :5Þ

where the cons where constant tantss   ai   are given in Table A.2 for pulsed pulse d and for puls pulsed-s ed-scanne canned d rad radiati iation. on. To minimize the influence of fluctuations in the output of  accelera acce lerators, tors, all the read readings ings shou should ld pref preferab erably ly be normal nor malize ized d to th that at of an ex exter ternal nal mon monito itorr. The external exte rnal monit monitor or shoul should d pref preferab erably ly be posi positione tioned d inside the phantom approximately at the depth of  measur mea sureme ement, nt, but at a dis distan tance ce of 3– 4 cm aw away ay from fr om the chamber chamber cen center ter along the major axis in the transverse plane of the beam. For k s , 1.03, the correction can be approximated to within 0.1 percent using the relation:  M 1 = M 2  1 ks  1 ¼  ; V 1 =V 2  1

 

ð A :6Þ

that is, the percentage correction is the percentage change in reading divided by a number that is one less than the voltage ratio. This has the advantage V 2   and of wor workin king g for non non-in -integ tegra rall val values ues of   V 1 / V  serves as a check on the evaluation using Equation (A.5). Note that the correction factor   ks  evaluated using the two-voltage method in pulsed beams corrects for both general and initial recombination.  A word of caution is required regarding the use of the two-voltage method for parallel-plane ioniz-

Table A.2. Quadratic-fit coefficients coefficients for the calculation of    by th the e tw twoo-v vol olta tage ge te tech chn niq ique ue in pu pullse sed d an and d pulsedpul sed-sca scanne nned d ra radia diation tion,, as a fun functi ction on of the vol voltage tage ratio  V 1 / V  V2    (Weinhous and Meli, 1984).

 ks

V 1 /  V 2

Pulsed a0

2.0 2.5 3.0 3.5 4.0 5.0

2.337 1.474 1.198 1.080 1.022 0.975

Pulsed scanned

 

a1

3.636 1.587 2 0.875 2 0.542 2 0.363 2 0.188

 

2

 

2

       

 

a2

2.299 1.114 0.677 0.463 0.341 0.214

 

a0

4.711 2.719 2.001 1.665 1.468 1.279

 

a1 2

 

2

     

a2

8.242 8.2 3.9 3. 977 2 2. 2.4 402 2 1. 1.6 647 2 1. 1.20 200 0 2 0. 0.75 750 0

 

 

 

4.533 4.5 2.2 2. 261 1.40 1. 404 4 0.98 0. 984 4 0.73 0. 734 4 0.47 0. 474 4

that for some plane-parallel chambers the expected V  is linear dependence of 1/  M   on 1/ V  is not satisfied in the th e vol olta tage ge in inte terv rval al us used ed fo forr th the e tw two o vol olta tage ge meth me thod od.. Th This is ef effe fect ct ca can n be co comp mpen ensa sate ted d fo forr by using the same two polarizing voltages for the dose determination in the user beam as are used for the chamber calibration at the standards laboratory, or by th the e us user er in th the e ca case se of a cr cros osss ca cali libr brat atio ion. n.  Alternatively  Alternativ ely,, the range of linearity of a chamber can be established in a pulsed beam by measuring the th e cha chambe mberr res respon ponse se ov over er a ra range nge of pol polari arizin zing g  voltages up to the manufactu manufacturer’s rer’s recommended maximum. This is a useful check on the performance of a chamber and should always be performed when whe n com commis missio sionin ning g a ne new w cha chambe mberr. If pos possib sible, le, the chamber should be used subsequentl subsequently y only at  voltages within the linear range, in which case the use of the two-voltage method is valid. In continuous radiation, the two-voltage method can also be use used d and a cor correc rectio tion n fa facto ctorr der derive ived d using the relation:

ks  ¼

  ðV 1 =V 2 Þ2  1 ðV 1 =V 2 Þ2  ð M 1 = M 2 Þ

:

 

ð A :7Þ

Note that for making recombination corrections, protonpro ton-sync synchro hrotron tron beams of longlong-puls pulse e dur duratio ation n and low low-pu -pulse lse-r -repe epetit tition ion fr frequ equenc ency y ma may y be con con-sidered as continuous. Forr re Fo relat lative ive mea measur sureme ement nts, s, for ex examp ample, le, the determination of depth–dose distributions and the measurement of output factors, the recombination corr co rrec ecti tion on sh shou ould ld be de dete term rmin ined ed in a su suffi ffici cien entt subset subs et of cond condition itionss tha thatt appr appropria opriate te corre correction ctionss can be derived. In pulsed beams, for which general recombination is dominant, the recombination correct re ctio ion n fo forr a gi give ven n ch cham ambe berr wi will ll sc scal ale e ap appr prox oxiimate ma tely ly li line near arly ly wi with th do dose se ra rate te.. In co cont ntin inuo uous us beams, beam s, the recombinatio recombination n corre correction ction is small and

ation chambers in pulsed beams. It has been shown

approximately approxima tely constant. 155

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

In scanned beams and other special beams of very high intensity, space-charge effects cannot be neglected, and the charge-collection efficiency should be assess ass essed ed by cal calibr ibrat ation ion aga agains instt a sy syst stem em th that at is independent of dose rate, such as a calorimeter. It should be noted that for the calibration of ionizatio iza tion n cha chambe mbers rs in st stand andard ardss lab labor orat atori ories es (se (see e Table A.1) A.1),, the calib calibrat ration ion certi certificat ficate e shou should ld sta state te whether or not a recombination correction has been applie app lied. d. The pr prece ecedin ding g dis discus cussio sion n and the wor workkshee sh eett is ba base sed d on th the e as assu sump mpti tion on th that at th the e ca cali li-brat br ation ion lab labora orator tory y has app applie lied d a rec recomb ombina inatio tion n correctio corr ection, n, and ther therefor efore e the pro procedu cedure re give given n for  k s  refers only to recombination the determination of  k in the user beam. If the calibration laboratory has not applied a recombination correction, the correction factor determined for the user beam quality   Q must mu st be div divide ided d by th that at app approp ropria riate te to th the e cal caliibration quality  Q 0, that is, k0s  ¼

  ks;Q : ks;Q0

 

ð A :8Þ

If   Q0  is a continuous beam,  k s,Q0 will normally be close to unity, and the effect of not applying   ks,Q0 either at calibration or using Eq. (A.8) will be negligible gib le in mos mostt cas cases. es. Ho Howe wever ver,, if   Q0   is a pu puls lsed ed beam, failure by the standards laboratory to apply ks,Q0 at the time of calibration is a potential source

of error, especially if the dose per pulse in the user beam is very different from that used at calibration. If this is the case, the user must determine   ks,Q0 in the clinic at a dose per pulse similar to that used at calibr cal ibrat ation ion (th (this is mig might ht not be th the e dos dose e per pulse normal nor mally ly use used d in th the e cli clinic nic). ). Thi Thiss det determ ermina inatio tion n does do es no nott ne need ed to be ca carr rrie ied d ou outt at   Q0; it is the matching of the calibration dose per pulse that is important impo rtant.. To avo avoid id a recu recurren rrence ce of this pro problem, blem, the user should request that a recombination correction be applied, or at least measured, at the next calibration at a standards laboratory, especially for calibration in pulsed beams.  Although  Althoug h proto proton n beams acceler accelerated ated by an isochro isochro-nouss cyc nou cyclot lotro ron n ar are e pul pulse sed d bea beams ms,, th the e pu pulse lse re reppetition rate is so high (of the order of 20 MHz) that such proton beams ought to be treated as continuous beams bea ms.. Fo Forr mo modu dula late ted d pr prot oton on be beams ams,, th the e ti time me-dependent dose can be regarded as pulsed due to the modulation modula tion process. How However ever,, even for the deepest points on the spread-out Bragg peak, the duration of  each pulse is usually long compared with the ioncoll co llec ecti tion on ti time me an and d ag agai ain n th the e be beam am sh shou ould ld be regarded as continuous in relation to recombination. It is im imp port rtan antt that the na natture (pu (puls lse ed or continu cont inuous) ous) of eac each h beam shou should ld be inv invest estiga igated ted.. The ion-recombination correction factor can be overestimated by up to 2 per cent if a cyclotron beam is assumed to be pulsed (Palmans  et al., 2006).

156  

 APPENDIX A

 A.3. DOSIMETR DOSIMETRY Y WORKSH WORKSHEET EET (IAEA, 2000).

157  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

158  

Journal of the ICRU Vol 7 No 2 (2007) Report 78 Oxford University Press

doi:10.1093/jicru/ndm039

 APPENDIX B: CLINICAL EXAMPLES DISCLAIMER

 As in ICRU Reports 50, 62, and 71 (ICRU, 1993b; 1999; 2004), clinic clinical al exam examples ples are pre presente sented d here with wit h the aim of ill illus ustr trat ating ing ho how w the re recom commen men-datio da tions ns con contai tained ned in the pr prese esent nt rep report ort ma may y be applied in practice. The six examples were obtained from different different pro proton ton ther therapy apy cent centers ers in diff differen erentt countries. count ries. They should not be cons construe trued d as ICR ICRU U recomm rec ommend endat ation ionss for cho choosi osing ng giv given en tr trea eatme tment nt techniques, volumes, or dose levels.

dRBE, respectively:  DRBE½Gy  ð RBEÞ ¼  DðGyÞ  RBE

;

dRBE ½Gy  ð RBEÞ ¼ dðGyÞ  RBE

:

 A generic RBE value of 1.1 for the therapeutic therapeutic use of protons is recommended ( i.e., for clinical protonbeam therapy, there is one value, and it is independent of total dose, dose per fraction, and tissue type). CONTRIBUTORS

DOSE PRESCRIPTION

 Absorbed doses ( D) ar are e giv iven en in Gy, wh wher erea eass RBE-wei RBE -weighte ghted d absor absorbed bed dose dosess ( DRBE) are given in Gy (RBE). Doses per fraction are stated as   d   and

The contributions of the following to the compilations of the clinical examples are gratefully acknowledged: J. Ad Adam ams, s, M. Bu Buss ssie ie`re, H. Ko Kooy oy, N. Li Lieb ebsc sch, h, J. Loeffler, T. Sakae, N. Tarbell, and T. Yock.

# International Commission on Radiation Units and Measurements 2007

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

CASE NUMBER B.1: UVEAL MELANOMA  General patient information

a. Patient Patient identification identification i. Name ii.. Age, ii Age, ge gend nder er,, an and d ra race ce iii. iv.  v..  v  vi.  vii.

Address Phone no. E-mail address Hospital ID no. Person to notify (with contact information) were a problem to arise

b. Medical Medical team team i. Ra Radia diatio tion n onc oncolo ologis gist(s) t(s) ii.. Ra ii Radi diat atio ion n ph phys ysic icis ist(s t(s)) iii. ii i. Ref efer erri ring ng ph phys ysic icia ian n

Confidential 85 ye year ar ol old d Japanese male Confidential Confidential Confidential Confident ial Confidential Confiden tial Confidential Confident ial

Confident Confid ential ial Con onfid fiden enti tial al Con onfid fiden enti tial al

Medical evaluation of presenting lesion

a. Pres Presen enti ting ng The pa The pati tien entt pr pres esen ente ted d wi with th comp co mpla lain intt myio my iode deso sops psia ia in th the e le left ft eye. b. Clinical history He devel developed oped the myiod myiodesop esopsia, sia, but decided to watch it without consulting with physicians. He  visited  visit ed a local ophth ophthalmol almologis ogistt because the symptom had progressed ove gressed overr the follo following wing few months mon ths.. He was re refer ferre red d to a Universit Univ ersity y Hosp Hospital ital with diag diag-nosis of retinal detachment.

Figure B.1.1. T2-weigh T2-weighted ted MRI image showing the tumor in the left eye.

c. Phy Physi sica call exam ex amin inat atio ion n

Proba Prob abl ble e ma mali lign gnan antt me mela lan nom oma a in th the e na nasa sall sid side e of th the e lef leftt ey eye; e; otherwise non-contributory d. Imag Imaging ing studi studies es MRI: 9.5 9.5  10.8 mm2 tumor of  left eye. Shown in Fig. B.1.1, and PET image in Fig. B.1.2 e. Tumor site Left eye f. Dia iagn gnos osis is Uvea eall mel elan anom oma a g. Grade No biopsy h. Stage Stage III, T3N0M0 i. Pr Prio iorr th ther erap apy y None No ne  j. Family history Noncontributory

Figure B.1. Figure B.1.2. 2. Meth Methion ionine–PET ine–PET imag image e demo demonst nstra rates tes an incr increaeased uptake corresponding to the choroidal tumor (arrow).

Treatment intent

a. Cura Curativ tive e by pr proto oton n th thera erapy py wit with h   DRBE,50%   of  70 Gy (RBE) b. No surgery or chemothera chemotherapy py c. Pro Probable bable outcome outcome i. Cur Cure e at 10 years: 95 percent percent ii.. Pro ii Proba babi bili lity ty of di dist stan antt me meta tast stas asis is::   , 5 percent iii. Prob Probabilit ability y of visual loss:   , 5 percent

General medical evaluation

a. History, History, physical No ev evid iden ence ce of me meta tast stat atic ic examination, tumor in chest and abdominal and imag imaging ing CT b. Co-m Co-morb orbidi iditie tiess On medica medicatio tion n for hyp hypert ertenension, gast gastritis ritis and pros prostat tatic ic hypertrophy

160  

 APPENDIX B

Treatment planning 

Treatment prescription

a. General General plan plan i. Sing Single le segment segment ii. Sing Single le anterior anterior beam b. Definitions of treatment treatment volumes volumes i. One millimeter millimeter thick CT images thr through ough-out the orbit were taken and transferred to the tre treatm atmentent-plann planning ing sys system. tem. The findings on MRI, ultrasound, and fundoscopic examinat exam ination ion wer were e utili utilized zed in the delin delineaeation ti on of tu tumo morr ma marg rgin inss fo forr th the e GT GTV V and CTV. ii. GTV GTV:: GTV was delineated as demonstr demonstrated ated iii. ii i. on CTCT V: scan. CTV was del elin inea eatted as GT GTV  V þ 0.5 mm, inc includ luding ing th the e who whole le thi thickn ckness ess of  the sclera. iv.. PTV iv PTV:: PTV was delin delineat eated ed by adding 1 mm around CTV in all directions, to allow for any variation in beam set-up and potential mov mo vem emen entt of ey eye e ba ball ll po posi siti tion on du duri ring ng treatment. c. RBE 1.10 d. Dose fractiona fractionation tion:: five fractions fractions to the PTV of  dRBE,50% 14 Gy (R (RBE BE)) or a to tota tall   DRBE,50% 70 Gy (RBE) over    7 days e. Un Uncer certai tainty nty in dos dose e and pa patie tient nt pos positi ition on not quantified ¼

¼

¼

a. Prescribed Prescribed dose dose i. Beam energy: energy: 70 MeV  MeV  ii. Dose is prescrib prescribed ed in  D RBE  [Gy (RBE)] iii. ii i. PTV: PTV: th the e do dose se at th the e ce cent nter er of th the e 22-cm cm wide wi de sp spre read ad-o -out ut Br Brag agg g pe peak ak (S (SOB OBP) P) on the th e ce cent ntra rall ax axis is wa wass 70 Gy (R (RBE BE). ). Th The e minimum dose to the PTV was    90 percent of the prescribed dose or   DRBE,98% was    63 Gy (RBE). iv.. Fractiona iv Fractionation: tion: five fractions over 7 days b. OAR: OAR: Dos Dose e con const strai raint ntss not ex excee ceeded ded by th these ese prescribed doses

f. Organs Organs at risk (OARs) and dose constraint constraintss or the th e do dose se pr prod oduc ucin ing g ma majo jorr mo morb rbid idit ity y at    5 percent Doses Dos es to opt optic ic dis discc and fovea fovea we were re bel below ow the level of concern. i. Opti Opticc disc disc   DRBE,2%   , 1 Gy (RBE) ii. Fo Fovea vea   DRBE,2%   , 1 Gy (RBE) g. The pl plan anne ned d do dose se le leve vels ls to ot othe herr st stru ruct ctur ures es were also below the constraint levels.

Patient immobilization and positioning 

a. Supine position with individualized individualized plastic cast was used for patient immobilization. There was supplemental immobilization of the head. b. Figure Figure B.1 B.1.3a .3a and b pr prese esent nt the CT sec sectio tions ns through the globe demonstrating the position of  a metal clip (a) and the optic disc (b). c. Me Metal tal clips were not nec necess essary ary to ide identi ntify fy the tumor boundary in the treatment planning, but necess nec essary ary for fiel field d pos positi itioni oning ng at eac each h tr trea eattment session. For this purpose, a few clips were sutu su ture red d on th the e sc scle lera ra re rela lati tive vely ly ne near ar to th the e tumo tu morr ma marg rgin inss be befo fore re th the e ac acqu quis isit ition ion of CT images.

Figure Figur e B.1 B.1.3. .3. Delin Delinea eatio tion n of RO ROIs Is on CT ima images ges.. Blu Blue e lin line e indicates left eye ball. (a) red line indicates metal clip. (b) Yellow line indicates optic disc.

161  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

c. Dose displa displays ys i. Axia iall pla lan ne th thrrou ough gh ce cen nter of eye in Fig. B.1.4 ii. Dose– volu volume me histogram histogramss in Figs B.1.5 B.1.5 and B.1.6

Treatment technique

The patien patientt wa wass ins instru tructe cted d to gaz gaze e at a poi point nt in space using a flickering light to achieve a fixed positio it ion n of the oc ocu ula larr gl glob obe e det eter erm min ined ed by the

Figur Fig ure e B.1 B.1.6. .6. Dose Dose-v -volu olume me his histog togra ram m (DV (DVH) H) of OA OARs. Rs. The  x-axi -axiss indi indicat cates es RBERBE-wei weighte ghted d absor absorbed bed dose ( DRBE) in Gy (RBE). (RBE ). The   y-axis indicates cumula cumulative tive volume in percenta percentage. ge. DVH of the lens is the red line and that of retina is the blue line. A part of the retina was included in the PTV.

planning planni ng pr proce ocess. ss. Th The e pos positi ition on of the flic flicker kering ing ligh li ghtt wa wass se sett in th the e tr trea eatm tmen entt pr proc oces esss to al allo low w target tar get vol volum umes es to be ade adequa quatel tely y irr irradi adiat ated ed wit with h mini mi nima mall do dose se to OA OARs Rs.. As th the e di dire rect ctio ion n of th the e gaze and hence the position of the target was satisfa is fact ctor ory y in th the e fir first st pl plan anni ning ng CT, a se seco cond nd CT

was not taken in this case. Metal clips sutured on the sclera sclera rel relat ative ively ly nea nearr to the tumor tumor we were re als also o used us ed un unde derr flu fluor oros osco copy py fo forr acc ccur ura ate pa pati tien entt positioning.

Quality assuranc assurance e Figure B.1.4. Dose distribut distribution ion in plane through the isocenter, isocenter, where wher e GTV, CTV, and PTV are indi indicat cated ed by red line lines. s. Isodo Isodose se distribution is indicated by yellow lines: from outside, 10, 30, 50, 90, and 95 percent of the prescribed dose,  D RBE 70 Gy (RBE), respectively.

a. Beam Beam pr profil ofiles es (flat (flatnes nesss and depth– depth– dos dose) e) we were re chec ch eck ked pr prio iorr to th the e sta tart rt of th the e co cour urse se of  treatment. b. Dete Determ rmin ina ati tion on of th the e do dose se at th the e re refe fere renc nce e poin po intt (c (cen ente terr of th the e be beam am an and d SO SOBP BP)) was performed. c. Tra Trans nspa pare rent nt fo foil ilss fo forr be beam am’s ’s ey eye e vi view ew an and d lateral view were overlaid on orthogonal x-ray

¼

filmss to con film confirm firm th the e pos positi itioni oning. ng. Bea Beam’ m’ss ey eye e  view with collimator (Fig. B.1.7) was printed to compar com pare e wit with h the cor corre respo spond nding ing pos positi itioni oning ng x-ray film (Fig. B.1.8). Eye position was monitored during each treatment session by a video camera with the cornea being marked directly on the video monitor. In case of detected movement of cornea of 1 mm the beam was immediately stopped manually. Equipment Figure B.1.5. Dose– volume histogram histogram (DVH) of target volumes. volumes. GTV is indi indicat cated ed in blue blue,, CTV in red, and PTV in bla black. ck. The near ne ar-mi -mini nimum mum dos dose e to the CTV wa wass   DRBE,98%,CTV  67. 67.4 4 Gy

a. Synchrotr Synchrotron-produ on-produced ced 70 MeV proton beam b. Broad-beam energy modulation modulation

¼

(RBE).  D RBE,50%,CTV 

¼

 DRBE,2%,CTV 

¼

70 Gy (RBE).

c. Imagi Imaging ng system system to position position target on beam 162

 

 APPENDIX B Continued

Lens Retina Ciliary body Iris Lid Fovea/  disc

 

15.6 ,1 ,1

 

,

1 1 ,1

 

,

 

,

70 70 70

1

70 70 ,1

 

Remaining volume at risk (RVR) i. Volume receiving V 30Gy 2.9 ml 30Gy (RBE) ii. Maxi Maximum mum dose outs outside ide PTV:   DRBE,2%   ,70 ¼

Gy (RBE) b. Total absorbed doses, D Figure B.1. Figure B.1.7. 7. Beam Beam’s ’s ey eye e vie view w with the coll collima imator tor.. Stru Structur ctural al images were printed on a transparent film. Blue circle indicates the eye ball, small white marks indicate the metal clips, red line stack indicates the tumor, blue line stack indicates the lens, and yellow line indicates the collimator. This image was used to set

 Volume/s  V olume/structur tructure e PTV

 

D50%, Gy

63.6

 

D98%, Gy

61.3

 

D2%, Gy

63.6

GTV Lens Retina Ciliary body Iris Lid Fovea/disc

up the patient by comparison with a verification x-ray film.

         

63.6 14.2 ,1 ,1 ,1 ,1 ,1

63.6  

,

1

 

63.6 63.6 63.6 63.6 63.6 63.6 ,1

Patient status at completion of treatment

a. Res Respons ponse e of tumor: none b. No interrupti interruptions ons c. No undue undue reactions reactions

 Addendum: technical information  Beam parameters and beam modification elements

a. Rotati Rotating ng range modulator: modulator: SOBP width is 20.0 mm (H2O) b. Beam course: course: C-9 (the ninth beam line) line) c. Lat Lateral eral spreading: spreading: 0.04 mm gold foil scatterer scatterer and wobbler d. Pen Penumbr umbra a (80– 20 percent) percent):: 2.5 mm e. Rang Range e comp compensa ensation tion and adju adjustm stment: ent: comp compenensating bolus made of polyethylene was used for conforming the distal edge of the beam to the shape of the PTV. The thickness of the compensating sati ng bolus at mini minimum mum thickness thickness poin pointt was 12.97 mm. f. Multi-leaf collimator: collimator: not used g. Individualized collimator: collimator: brass brass

Figure Figur e B.1 B.1.8. .8. Veri erific ficat ation ion x-r x-ray ay fil film m sho showi wing ng the ape apertu rture re outline. Metal clips are marked in blue for better visualization.

Total doses delivered

a. Total RBE-weighted RBE-weighted absorbed doses,  D RBE  Volume/   Volume/  structure

 DRBE,50%, Gy (RBE)

PTV GTV

70 70

 DRBE,98%, Gy (RBE)

67.4 70

 DRBE,2%, Gy (RBE)

70 70 Continued

163  

 PRESCRIBING RECORDING, AND REPORTING PROTON-BEAM THERAPY 

CASE NUMBER B.2: ADENOCARCINOMA OF PROSTATE General patient information

a. Patien Patientt identification identification i. Name ii.. Age, ii Age, ge gend nder er,, an and d ra race ce iii. Address iv. Phone no.  v.. E-mail address  v

Confidential 78 ye year ar ol old d Japanese male Confidential Confidential Confidential Confident ial

 vi. Hospital no. Confidential Confiden tial  vii. Person toID notify Confidential Confid ential (with contact information) were a problem to arise b. Med Medical ical team team

i. Loca Locall con contr trol ol pr proba obabil bility ity at 10 yea years: rs: 80 percent ii. GIII/IV rectal injury: 5 percent percent iii. GIII/ GIII/IV IV bladder injury: injury: 5 perc percent ent Treatment planning 

a. General General plan plan i. One treatment treatment segment segment b. Definition of treatment treatment volumes volumes i. Th The e tr trea eatm tment ent pla planni nning ng CT was ac acqui quire red d with wi th th the e pa pati tien entt in th the e sa same me po posi siti tion on,, immobi imm obiliz lizat ation ion dev device ice,, an and d con condit dition ionss as used for treatment. 5 mm CT sections were used to define GTV, CTV and organs at risk (OARs). ii. Defin Definition ition of volumes volumes  GTV: the prostate gland †

 CTV: GTV  þ 5 mm  PTV: not delineated (see Addendum) c.   DRBE,50% 76 Gy (RB (RBE) E) at 2.1 2.17 7 Gy (R (RBE BE)) per fraction in 35 fractions over 7 weeks  DRBE,98% 74 Gy (RB (RBE) E) or   DRBE,98% 2.11 Gy (RBE) per fraction d. RB RBE E 1.10 e. Organ Organss at risk with dose constraint constraintss †

i. Radia Radiatio tion n onc oncolo ologis gist(s) t(s) ii.. Ra ii Radi diat atio ion n ph phys ysic icis ist(s t(s)) iii. ii i. Ref efer erri ring ng phys physic icia ian n

†

Confident Confid ential ial Confi Co nfide dent ntia iall Con onfid fiden enti tial al

¼

¼

Medical evaluation of presenting lesion

¼

¼

a. Med edic ical al his isto torry

f. Grade g. Stage

Asymptomatic, PSA of  21 5.4 ng ml (1998), 5.8 ng ml21 (2003) Prostate normal. Otherwise noncontributory MRI: no tu tum mor see een n in pros pr osta tate te.. CT, MRI RI,, an and d bone bo ne sc scan an:: no me meta tast stat atic ic tumors noted Prostate Moderately differentiated adenocarcinoma. Ultrasound-guided Ultrasound -guided biopsy Gleason 6 T1cN0M0

h. Prior therapy i. Fam amil ily y hi hist stor ory y

None Nonc No ncon ontr trib ibut utor ory y

b. Physical Physical examination c. Im Ima agi gin ng stu tud die iess

d. Tumor site e. Diagnosis

¼

i. Ante An teri rior or rec ecta tall wal all: l:   DRBE,2% 76 Gy (RBE) ii. Blad Bladder der base: base:  D RBE,2% 76 Gy (RBE) iii. Righ Rightt and left hip: hip:  D RBE,2% 32 Gy (RBE) f. Deliv Delivery ery technique technique i. Pas Passive sive energy modu modulate lated d 250 MeV MeV pro proton ton beam ii. Lat Lateral eral opposed opposed fields iii. No allowance allowance for prostate prostate motion g. Treatmen reatment-planning t-planning system system i. Do Dose sess wer ere e es esti tima mate ted d wi with th th the e th thrree ee-dimensional dose calculation method based on CT images. ii.. Sim ii impl ple e rayy-li line ne tracin ing g met eth hod was ¼

¼

h.

General medical evaluation

used to ca used calc lcul ulat ate e ra rang nges es an and d to de desi sign gn collimators. iii. Accuracy of dose statement: not determined RVR (R RVR (Rema emaini ining ng Volu olume me at Ri Risk) sk) rec receiv eiving ing V 33 227 ml 3 3 Gy (RBE) Maxim Ma ximum um dos dose e out outsid side e PT PTV V: DRBE,2% 76 Gy (RBE) CTV and OARs are shown in Fig. B.2.1. Beam’s Beam ’s eye view of CTV is shown shown in Fig. B.2.2. ¼

a. History, History, physical examination, Noncontributory and imaging b. Co-morbidities None c. Medications Noncontributory

i.  j. k.

  ¼

Patient immobilization and positioning  Treatment intent

Supine Supi ne with head on head headres rest. t. Pa Patien tientt immo immobiliz biliz--

a. Curat Curative ive by proton radiation radiation alone b. No surgery or systemic systemic therapy therapy c. Predicted outcome

ation was achieved by means of a cradle with polystyrene beads suctioned to conform to the contours of the patient’s body. Position of the patient relative 164

 

 APPENDIX B

b. Total dose to CTV CTV::   DRBE,50% 76 Gy (RBE) at 2.17 2.1 7 Gy per fractio fraction, n, 35 fr fract action ionss in 7 we weeks eks DRBE,98% 74 Gy (RBE), DRBE,2% 78 Gy (RBE) c. RBE 1.10 d.   DRBE,2%   to OA OARs. Rs. Anterior Anterior re recta ctall wa wall: ll: 76 Gy (RBE) e. Dose distribut distributions ions are shown in Fig. B.2.3. f. Dose–vo e–vollume histograms are shown in Fig. B.2.4. ¼

¼

  ¼

Treatment technique

¼

a. Patient Patient in immobilization devices. b. The positioning positioning was confir confirmed med using orthogonal orthogonal simulator images referring to orthogonal digitally reconstru recon structed cted rad radiogram iogramss (DRR (DRRs) s) gener generate ated d in the trea treatmen tment-pla t-planning nning compu computer ter progr program. am. It wass con wa confirm firmed ed usi using ng the orth orthogo ogonal nal sim simula ulator tor images ima ges dig digita itally lly sto store red d and ort ortho hogon gonal al fluo fluoro ro-scopicc image scopi imagess durin during g ever every y trea treatmen tmentt sessio session. n. c. See Addendum Addendum for additional additional details.

Figure B.2. Figure B.2.1. 1. The CTV (re (red), d), rec rectum tum (green (green), ), urin urinary ary blad bladder der (blu (b lue e), an and d rig ight ht fem emu ur (or ora ange ge)) ar are e sh sho own in th this is three-d thr ee-dimen imension sional al rec recons onstruc tructed ted imag image. e. The left femu femurr is not shown.

Quality assurance

a. The ins instit tituti utiona onall st stand andard ard dos dose e mon monito itorr wa wass calibrated once a year at the national standards center. b. The given dose at the reference reference point (the isocenter and/or the center of SOBP) was measured under the same conditions as for treatment for each patient before starting the irradiation. c. The The do dose se-m -mon onit itor orin ing g sy syst stem em an and d th the e be beam am-delive del ivery ry sy syst stem em we were re che check cked ed ev every ery mor mornin ning g before the treatment. d. The bea beam m po posit sition ion wa wass eva evalua luated ted du durin ring g the irradiation by means of a beam-profile monitor. e. The beam current and and the integral beam charge charge were mea were measur sured ed wit with h two ind indepe epende ndent nt set setss of  monito mon itors. rs. The ratio ratio of dos doses es mea measur sured ed by th the e two ind indepe epend ndent ent mon monito itors rs wa wass con contin tinuou uously sly evaluated during the irradiation so that any significan nifi cantt cha change nge of bea beam m pa param ramete eters rs cou could ld be instantly detected during the irradiation. f. The beam flatness flatness was kept within   + 4 %, and monitore moni tored d durin during g the irra irradia diation tion by a speci special al ion chamber chamber..

Figure Figur e B.2 B.2.2. .2. The bea beam’ m’ss ey eye e vie view w of th the e CT CTV V and th the e bea beam m apertu ape rture re co confo nform rmed ed to the tu tumor mor sha shape pe wi with th the multi multi-le -leaf  af  collimator.

Equipment

to the bea beam m wa wass det determ ermine ined d by las laser er ali alignm gnment ent beams on markers on the cradle and patient skin.

a. 250 MeV proton proton beams were were supplied supplied by a slow

Treatment prescription

extr extract ionwas synchrot sync ron..21The averag av erage e absor absorbed bed doseaction rate 2 hrotron Gy min at the isocenter. b. Rotat Rotating-gantry ing-gantry system system c. Broad-beam energy modulation modulation

a. Proton-beam therapy: 250 MeV  MeV  165  

 PRESCRIBING RECORDING, AND REPORTING PROTON-BEAM THERAPY 

Figure B.2.3. Dose distribu distributions tions in the axial (a), sagittal (b), and coronal (c) planes. The white contour represents the CTV. The red contour cont our represent representss 95 per percent cent of the prescribed prescribed dose, and tha thatt in blue represen represents ts 1 per percent cent of the prescribed prescribed dose. The remaining remaining contours represent the 10–80 percent isodoses in 10 percent steps.

Total doses deliver delivered ed

a. Total RBE-weighted RBE-weighted absorbed doses,  D RBE  Volume/s  V olume/structur tructure e

 

CTV  Anterior rectal wall Bladder base Right hip/left hip

DRBE,50%, Gy (RBE)

 DRBE,98%, Gy (RBE)

 DRBE,2%, Gy (RBE)

76

74

78 76 76 32

i. RV RVR R receiving V 33 3 3 Gy (RBE)

¼

227 ml ,

ii. 77 Maxim Ma ximum um dos dose e out outsid side e CT CTV V:   DRBE,2% Gy (RBE) iii. ii i. No de devi viat atio ion n fr from om pl plan anne ned d do dose se or do dose se fractionation b. Total absorbed doses, D

Figu Figure B.2.4. Dose–volume Dose–volum e his histogr tograms ams for the CTV and OAR OARs. s. Notere that the extent of rectum evaluated is from the caudal end of the sigmoid colon to the anal verge and that of the femur is 105 mm in length from the proximal end of the right femur,  i.e., from the proximal end of the right femur to a level 10 mm below the caudal end of the CTV.

166  

 APPENDIX B

 V  Volume/st olume/structure ructure CTV  Anterior rectal wall

 

D50%, Gy

69.1

 

D98%, Gy

67.3

 

D2%, Gy

rotating rotatin g gant gantry ry.. Paramete Parameters rs of field prep prepara aration tion were we re com compu puted ted wit with h the the sim simple ple ra ray-l y-line ine-tr -traci acing ng method.

70.9 69.1

Treatment technique

Bladder base Right hip/left hip

69.1 29.1

The co The couc uch h an angl gle e wa wass 08   ( pa para rall llel el to th the e ax axis is of  gantry gan try ro rota tati tion) on).. Th The e pa patie tient nt wa wass imm immobi obiliz lized ed using usin g a crad cradle le with poly polyst styren yrene e bead beadss suct suctioned ioned to th the e con contou tours rs of th the e pa patie tient’ nt’ss bod body y. Th The e pos pos--

Patient status at completion of treatment

a. Respons Response e of tumor: none b. Ch Chan ange ge,, if an any y, of pa pati tien entt sym ympt ptom oms: s: no nott applicable c. Plan Planned ned follow-up follow-up appointment appointment and imagi imaging ng or other studies: 1 month after completion of the treatment

ition itio n was al alig ign ned wi witth la lase serr poin intter erss and mark ma rker erss on th the e cr crad adle le an and d th the e sk skin in.. Th The e po posi si-tionin tio ning g wa wass con confirm firmed ed usi using ng ort orthog hogona onall sim simuulato la torr im imag ages es re refe ferr rrin ing g to or orth thog ogon onal al DR DRRs Rs generat gene rated ed in the trea treatmen tment-pl t-planni anning ng comp computer uter program pro gram.. Beam dir directio ections ns with rota rotating ting gant gantry: ry: (1) rig right ht lat later eral, al, 2708   and and (2 (2)) le left ft la late tera ral, l, 908. The maximum depths were 207 and 212 mm in waterwa ter-equi equivalen valentt medi medium um for beam beamss (1) and (2), respect res pectively ively. The beam beamss wer were e bro broaden adened ed by the dual-ring double-scattering method. A ridge filter was se sett for 60 mm sp sprread ad--ou outt Brag agg g pea eak ks (SOBP). The beam apertures were shaped with a manual man ually ly adj adjus usted ted br brass ass mu multi lti-le -leaf af col collim limat ator or wit ith h le lea aves 50 mm th thic ick k an and d 5  50 mm2 in area. are a. Bolu Boluses ses made of wa waterter-equiv equivalent alent mat material erial wer ere e us used ed to co conf nfor orm m th the e di dist stal al ed edge gess of th the e beam be amss to th the e tu tumo morr sh shap apes es.. To acc ccou ount nt fo forr  various uncertainti uncertainties es the beam apertures were enlarg enl arged ed by 10 mm bot both h la later terall ally y and cep cephal haloocaud ca udal ally ly in th the e be beam am’s ’s ey eye e vi view ew an and d th the e be beam am rang ra nges es we were re ex exte tend nded ed 10 mm di dist stal ally ly on th the e beam axes by means of reducing the thickness of  the range shifter.

 Addendum: technical information  Beam shaping

 After delinea delineating ting the CTV, the margins were added to the CTV by means of enlarging the beam aperture tu re bot both h la later terall ally y and cep cephal halo-c o-cau audal dally ly in th the e beam’s eye view, and the distal margin of the beam was extended by reducing the thickness of the range shifter for the beam. Shape of the inner surface of  the bolus was enlarged enlarged acc accordi ording ng to the enlarged enlarged beam aper apertur ture. e. The marg margins ins cov covered ered unc uncerta ertainty inty due to set-up errors, internal variability, and distal and lateral fall-offs of the irradiation fields.  Equipment

Proton beams of 250 MeV energy were supplied supp lied by a slow slow-ext -extrac raction tion syn synchro chrotron tron with a

167  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

CASE NUMBER B.3: CARCINOMA OF LUNG

ii. GIII/ GIII/IV IV lung injury: injury:   5 percent

iii.. Other iii Other tho thora racic cic st struc ructu tures res jud judged ged to be at negligible risk

General patient information

a. Patient Patient identification identification i. Name ii. Age, gender, and race

iii. iv.  v..  v  vi.  vii.

Address Phone no. E-mail address Hospital ID no. Person to notify (with contact information) were a problem to arise

Treatment planning 

Confidential 81 year old Japanese male

a. General General plan plan i. One treatment treatment segment segment b. Definition of treatment treatment volumes volumes

Confidential Confidential Confidential Confident ial Confidential Confiden tial Confidential Confid ential

i. Trea reatmen tmentt plan planning ning CT acquired acquired with the patient pat ient in the same immo immobiliz bilizatio ation n devic device e and con condit dition ionss as use used d for tr trea eatme tment nt.. CT sections were used to define GTV, CTV, and organs at risk (OAR). ii. Defin Definition ition of volumes volumes  GTV: gross primary tumor   CTV  GTV þ 5 mm  PTV: not delineated (see Addendum)   The thr three-d ee-dimens imensional ional recon reconstru structed cted image of the right lung and the CTV  are presented in Fig. B.3.1. The beam’s eye view of the targeted structures and the be bea am ap aper erttures ar are e sh sho own in Fig. B.3.2. c.   DRBE,50% 66 Gy (R (RBE BE)) at 6.6 Gy (R (RBE BE)) pe perr frac fr actio tion, n, giv given en in 10 fr frac actio tions ns ov over er 2 we weeks eks..  DRBE,98% 63 Gy (RBE) d. RB RBE E 1.10 e. Organ Organss at risk with dose constraints constraints:: no thoracic ac ic st struc ructu ture re wa wass jud judged ged to be at sig signifi nifican cantt risk of injury † †

b. Medical Medical team team i. Ra Radi diat atio ion n on onco colo logi gist( st(s) s) ii.. Ra ii Radi diat atio ion n ph phys ysic icis ist( t(s) s) iii. ii i. Ref efe err rrin ing g physic icia ian n

Con onfid fiden enti tial al Con onfid fiden enti tial al Con onfi fiden enti tia al

†

Medical evaluation of presenting lesion

a. Medical history

d. e.

Asymptomatic. Right lung lu ng ma mass ss on ro rout utin ine e chest x-ray Physical Phy sical exam examina ination tion No abno abnormal rmal findin findings gs Imaging studies CT: 2.5 cm lesion in latera lat erall segm segment ent of righ rightt medial med ial lob lobe. e. No lym lymph ph adenopathy adenopa thy.. PET [FDG] total body scan and MRI of br bra ain in:: no evi vid den entt distant metastases Tumor site Right lung Diagnosis Carcinoma. CT guided

f. g. h. i.

Grade Stage Prior therapy Fam amil ily y hi hisstor ory y

b. c.

¼

†

¼

¼

  ¼

biopsy Poorly differentiated T1N0M0; Stage 1A   None Nonc nco ont ntri ribu buttor ory y

General medical evaluation

a. History, History, physical examination, and imaging b. Co-morbidities c. Medications

No important tumor findings

non-

None Noncontributory

Treatment intent

a. Curat Curative ive by proton radiation radiation alone b. No surgery or systemic systemic therapy therapy c. Predicted outcome i. Loca Locall con contr trol ol pr proba obabil bility ity at 10 ye years ars:: 80 percent

Figure B.3.1. The CTV (re Figure (red) d) and the right lung (or (orange ange)) are shown in this three-dimensional reconstructed image.

168  

 APPENDIX B

Figure B.3.2. The beam’s eye view of the CTV and the beam apertures conformed conformed with the multi-leaf multi-leaf collimators.

f. Delivery Delivery technique technique i. Pass Passiv ive e en ener ergy gy mo modu dula late ted d 15 155 5 an and d 20 200 0 MeV proton beams ii. Four fields: isocentric isocentric and co-planar iii. Respira Respiration-gated tion-gated treatment treatment delivery g. Treatme reatment-planning nt-planning system system i. Do Dose sess wer ere e estim ima ate ted d wi witth a thr hree ee-dimensional dose calculation method based on CT images. ii. Simple ray-line-tr ray-line-tracing acing method was used for calculating calcula ting ranges and designin designing g collima collimators. tors. iii. Accur Accuracy acy of dose statement: statement: not determined. determined. h. RVR RVR (R (Rema emaini ining ng Volu olume me at Ris Risk) k) rec receiv eiving ing V 30 25 ml 3 0 Gy (RBE) i. Ma Maxim ximum um dos dose e out outsid side e PT PTV V:   DRBE   ,   66 Gy

relative to the beam was determined by the use of  laser las er ali alignm gnment ent bea beams ms on mar marke kers rs on the cr cradl adle e and patient skin. Treatment prescription

a. Proton-be Proton-beam am therapy: therapy: 155 and 200 MeV  b. Total dose to CTV CTV::   DRBE,98% 63 Gy (RBE) at 6.3 Gy (RBE) per fraction, given in 10 fractions over 2 weeks   DRBE,50% 66 Gy (RBE)  DRBE,2% 67 Gy (RBE) c. Respira Respiration-gatedtion-gated-beam beam delivery d. RBE 1.10 e. OAR—righ OAR—rightt lung: D RBE,2% 62 Gy (RBE) f. Dose distributions are are presented in Fig. B.3.3 g. Dose– volu volume me histogram histogramss are shown in Fig. B.3.4 B.3.4 ¼

¼

¼

¼

¼

¼

(RBE) Patient immobilization and positioning 

Treatment technique

Supine with head on the headrest. Supine headrest. Pa Patien tientt immobilization was achieved by means of a cradle with polystyrene beads suctioned to conform to the contours of the patient’s body. Position of the patient

a. Patient Patient in the immobilization immobilization devices b. Pati tien entt po posi siti tion onin ing g was co confi nfirm rmed ed us usin ing g orth or thog ogon onal al si simu mula lato torr im imag ages es re refe ferr rrin ing g to orthogonal digitally reconstru reconstructed cted radiographs 169

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

Figure B.3.3. Dose distributions on the axial (a), sagittal (b), and coronal (c) planes. The white contour represents represents the CTV. CTV. The red contour represents 95 percent of the prescribed dose and the blue contour represents 1 percent of the prescribed dose. The remaining contours represent the 10–80 percent isodoses in 10 percent steps.

(DRRs) Rs) gen genera erated ted in the tr trea eatm tment ent pla plann nning ing (DR comp co mput uter er pr prog ogra ram. m. It wa wass co confi nfirm rmed ed us usin ing g orthog ort hogona onall sim simula ulator tor ima images ges dig digita itally lly st store ored d and ort orthog hogona onall fluo fluoro rosco scopic pic ima images ges for ev every ery treatment session. c. See Addendum Addendum for additional additional details

Quality assuranc assurance e a. The ins instit tituti utiona onall st stand andard ard dos dose e mon monito itorr wa wass calibrated once a year at the national standards center. b. The given dose at the reference reference point (the isocent ce nter er an and/ d/or or th the e ce cent nter er of th the e SO SOBP BP)) wa wass

measur meas ured ed un unde derr th the e sa same me co cond nditi ition onss as fo forr treatment before starting the irradiation.

Figure Figur e B. B.3.4 3.4.. Dos Dose– e– vo volum lume e his histog togra rams ms for the CT CTV V and the right lung.

170  

 APPENDIX B

c. Th The e do dose se-m -mon onit itor orin ing g sy syst stem em an and d th the e be beam am-delivery deliv ery sys system tem wer were e chec checked ked ever every y morn morning ing before the treatment. d. The bea beam m pos positi ition on wa wass eva evalua luated ted dur during ing th the e irradiation by means of a beam profile monitor. e. Th The e be beam am cu curr rren entt an and d th the e in inte tegr gral al be beam am charge charg e wer were e measu measured red with two indep independe endent nt sets of monitors, from which signals were sent to the accelerator system when the irradiation was completed. The ratio of doses measured by the th e two ind indepe epend ndent ent mon monito itors rs wa wass con contin tinuuously monitored during the irradiation so that a significant change of beam parameters could be instantly detected during the irradiation. f. The beam flatness flatness was kept within   + 4 percent and was monitored during the irradiation by a special ion chamber. Equipment

a. Proto Proton n bea beams ms of 155 and 200 MeV were were sup sup-plied pli ed by a slo slow w ex extr trac actio tion n sy synch nchrot rotron ron.. The average RBE-weighted absorbed dose rate was 2 Gy (RBE) min21 at the isocenter. b. Rotat Rotating ing gantry system system c. Broad-beam energy modulation modulation Total doses delivered

a. Total RBE-weighted RBE-weighted absorbed doses,  D RBE  Volume/   Volume/  structure

 DRBE,50% Gy (RBE)

 DRBE,98%, Gy (RBE)

 DRBE,2%, Gy (RBE)

CTV Right lung

66

63

67 62

i. RVR receiving receiving V 30 25 ml 30 Gy (RBE) ii. Maxim Maximum um dos dose e out outsid side e CT CTV V:   DRBE,2% , 66 Gy (RBE) iii. ii i. No dev evia iattio ion n from pla lann nned ed dos ose e or fractionation b. Total absorbed doses, D ¼

 Volume/   Volume/  structure

 D50%, Gy

CTV

60

Right lung

 

D98%, Gy

57.3

 

D2%, Gy

60.9 56.4

b. Asymp Asymptom tomat atic ic at st start art of tr trea eatm tment ent and com com-pletion of therapy c. Plan Planned ned follow-up follow-up appointment appointment and imag imaging ing or other studies: 1 month after completion of the treatment

 Addendum: technical information  Beam shaping

 After delineating the CTV, the margins were added to the CTV both laterally and cephalo-caudally by mean me anss of en enla larg rgin ing g th the e be beam am ap aper ertu ture re in th the e beam’s eye view, and the distal margin of the beam range ra nge wa wass ext extend ended ed by re redu ducin cing g the thi thickn ckness ess of  the th e ra rang nge e sh shif ifte terr fo forr th the e be beam am.. Th The e ma marg rgin inss covered uncertainties due to setup errors, internal  variability,, distal and lateral fall-offs of the  variability irradiation fields, the organ movement with breathing, and others (see Section 5.1.4.4).  Equipment

Proton beams of 250 MeV energy were supplied by a slo slow-e w-extr xtrac actio tion n sy synch nchrot rotron ron wit with h a ro rota tatin tingggant ga ntry ry sy syst stem em fo forr be beam am de deli live very ry.. Th The e av aver erag age e absorbed dose rate was 2 Gy min 21 at the isocenter. Doses Dos es we were re est estima imated ted wit with h a th three ree-d -dime imensi nsiona onall dose do se ca calc lcul ulat atio ion n me meth thod od ba base sed d on CT im imag ages es.. Para Pa ramet meters ers of fiel field d pr prepa epara ratio tion n we were re com compu puted ted with the simple ray-line-tracing method. Treatment technique

The couch angle was 08  (parallel to the axis of gantry rotation). rota tion). Beam direc directions tions with rota rotating ting gantr gantry: y: (1) right rig ht an ante teri rior or ob obli liqu que, e, 35 350 08; (2 (2)) ri rig ght an ante teri rior or oblique, 3058; (3) right posterior oblique, 2608; and (4) right posterior oblique, 2108. The maximum depths were 89, 79, 102, and 138 mm in water-equivalent medium for beams (1), (2), (3), and (4), respectively. The Th e bea beam m was br broad oadene ened d by the du dual-r al-ring ing dou double ble-scattering method. A ridge filter was set for a 40 mm spread-out Bragg peak for all beams. The beam apertures tur es wer were e shap shaped ed with a manu manually ally adjusted adjusted brass multi-leaf collimator. Each leaf was 50 mm thick and 5  50 mm2 in ar area ea.. Th The e be beam am ap aper ertu turres we werre enlarge enl arged d by 9 mm la later terall ally y, 9 mm in the cephalic cephalic dire di rect ctio ion, n, an and d 14 mm cau caudal dally ly in th the e be beam am’s ’s ey eye e  view. To tak take e acco account unt of variou variouss uncer uncertaint tainties ies the beam bea m ran ranges ges we were re ext extend ended ed 5 mm dis distal tally ly on the

beam axe beam axess by red reduci ucing ng the thicknes thicknesss of the range range shifters. Boluses were not used to conform the distal edge of the beam to the tumor shape.

 Patient status at completion of treatment

a. Res Respons ponse e of tumor: none

171  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

CASE NUMBER B.4: ACOUSTIC NEUROMA  General patient information

a. Cura Curati tive ve by si sing ngle le-d -dos ose e pr prot oton on ra radi dia ati tion on treatment. b. No surgery or systemic systemic therapy planned. planned.

a. Patien Patientt identification identification i. N ae, megend ii. Age, Ag gender er,, and and rac race e iii. iv.  v..  v  vi.  vii.

C fid enold tial 72onye year ar Caucasian female Address Confidential Phone no. Confidential E-mail address Confidential Confide ntial Hospital ID no. Confidential Confid ential Person to notify Confidential Confiden tial (with contact information) were a problem to arise

b. Med Medical ical team team i. Radiation Radiation onco oncologis logist(s) t(s) ii. R Rad adiat iation ion ph phys ysici icist(s st(s)) iii. Ref iii. efer erri ring ng ph phys ysic icia ian n

Confiden Con fidential tial Confid Co nfident ential ial Confide Confi dent ntia iall

Medical evaluation of presenting lesion

a. Med edic ical al his isto torry

b. Phys Physical ical exam examina ination tion c. Im Ima agi gin ng stu tud die iess d. Tumor site e. Diagnosis

Sev eve en year arss of dec ecrreas as-ing auditory acuity and is now nearly deaf. Increasing size of   tumor Noncontr Nonc ontributo ibutory ry 1.2 1. 25 mm sl slic ice e CT Left cerebellopontine angle Acoustic neuroma: clinical and radiolo radiological gical diagnosis. Volume 3. 3.2 2 ml No biopsy No staging None Non onco con ntr trib ibu uto tory ry ¼

f. g. h. i.

Grade Stage Prior therapy Fam amil ily y hi hisstor ory y

General medical evaluation

a. History, History, physical examination,

Treatment intent

No evidence of  other tumor detected

c. Pre Predicte dicted d outcome i. Lo Loca call co cont ntro roll pr prob obab abil ilit ity y at 7 ye year ars: s: 95 percent ii. Hearing preservation: preservation: 50 percent percent iii. Perm Permanen anentt faci facial al nerv nerve e inju injury: ry: 1 per percent; cent; temporary: 5 percent. Treatment planning 

a. General General plan plan i. Standa Standard rd ster stereota eotactic ctic rad radiosur iosurgery gery tre treatatment for this tumor. ii. Sing Single le dose dose b. Definition of treatment treatment volumes volumes i. Treat Treatment ment-pla -planning nning CT perfo performed rmed with the pa patie tient nt in th the e sam same e imm immobi obiliz lizat ation ion device and conditions as used for treatment. ii. GTV GTV:: GTV is defin defined ed as the demo demonst nstrabl rable e tumor relative to the fiducial markers for pla lan nnin ing g (F (Fig ig.. B.4 .4.1 .1a–c) a–c).. CTV was assumed to be identical to the GTV. iii. Volum olume e defi defined ned:: 3.2 ml tum tumor or visualized visualized at the th e ce cere rebe bell llop opon onti tine ne an angl gle e on im imag agin ing g studies. iv.. Fiduc iv Fiducia iall ma mark rker erss (F (Fig ig.. B. B.4. 4.2a–c). 2a–c). Th Thre ree e SS316L SS31 6LVL VL 1/16 in. diam diameter eter micr micro-sp o-spher heres es inserted insert ed in th the e out outer er tab table le of the pa patie tient’ nt’s skul sk ulll we were re us used ed fo forr pa pati tien entt al alig ignm nmen ent. t. Tar arge gett po posi siti tion on co confi nfirm rma ati tion on:: wi with th th the e patie pa tient nt in the fu fully lly imm immobi obiliz lized ed po posit sition ion,, orthogonal x-ray images were obtained for a stand st andard ard set set-up -up pos positi ition on as we well ll as oth other er coplan cop lanar ar fie fields lds pr prior ior to tr trea eatm tment ent.. Sin Single gle x-ra xray y im imag ages es we were re ob obta tain ined ed fo forr su supe peri rior or oblique obliq ue or vert vertex ex field fieldss prior to tre treatm atment. ent. Implanted micro-spheres were automatically identified on the x-ray images with custom compu com puter ter sof softw twar are e whi which ch th then en com compar pared ed them with expected projections as identified on plan planning ning DRR DRR.. Pa Patien tientt posit position ion corr correcec-

and imaging b. Coo-mo morb rbid idit itie iess

c. Present status d. Past surgery e. Prior radiation

tions we tions were re mad made. e. Th This is wa wass re repea peated ted un until til the th e int intend ended ed pos positi ition on wa wass ac achie hieve ved, d, and reconfirmed prior to treatment.  v.. PTV  v PTV:: PTV is defin defined ed as 1 mm bey beyond ond the GTV in all directions, to allow for all uncertainties included in the immobilization and setu se tup. p. Th The e PT PTV V di did d ab abut ut th the e br brai ains nste tem m surfa sur face. ce. Th The e dos dose e pr presc escrib ribed ed to the PT PTV  V  was   DRBE,98%   12.8 Gy (RBE) and thus resp re spec ecte ted d th the e do dose se co cons nstr trai aint nt of 12 Gy

Spin ina al sten enos osis is Type II diabetes Hypertension Mastoid disease as child Well controlled comorbidities C section Mastoid surgery None

 

172  

 APPENDIX B

¼

Figure Figu re B. B.4. 4.1. 1. Vie Views ws (a (a)) an and d (b (b)) ar are e la late tera rall an and d AP fie field lds, s, respectively, demonstrating the position of the fiducial markers placed in the skull. The diagnostic quality images were obtained for the positioning of the patient by the use of these markers. (c) CT slice at the level of one of the fiducial markers.

Figure B.4.2. Views (a), (b), and (c) are displays of the tumor and paths of the three beams employed.

173  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

Figure B.4.3. Three-di Three-dimensional mensional image of the tumor relative to the brainstem, chiasm, chiasm, and the 90 and 50 percent dose volumes.

(RBE) to the brainstem surface. The aperture margin was 3.5 mm beyond the PTV  to allow for the penumbra (Fig. B.4.3).  vi. Dose calculation   Three-dimensional proton pencil beam algorithm based on 1.25 mm CT slices †

 vii. The XIO syst system em was used for treatm treatment ent planning  viii. Dose Dose– – volume histogram histogramss are given in Fig. B.4.4 for the tumor and brainstem. c. Sing Single le dose: dose:  D RBE,98% 12.8 Gy (RBE) to GTV. d. RB RBE E 1.10 ¼

¼

†

of the whole head (  184 slices)   Doses are stated stated with an accuracy accuracy of  2 percent.

e. Organ Organss at risk. Dose constrain constraints ts i. Brainst Brainstem em surface: surface:  D RBE,2% 12 Gy (RBE) ii. Mid brainstem: brainstem: D RBE,2% 6 Gy (RBE) ¼

¼

Patient immobilization and positioning 

Supine Supi ne po posi siti tion on wi with th a mo modi difie fied d Gi Gill– ll– Th Thom omas– as– Cosman (GTC, Radionics Inc.) device. Treatment prescription

RBE-weighte RBE-wei ghted d dose ( DRBE) is prescribed. RBE 1.10 Single Sing le dose dose Brain-s Bra in-stem tem surfa surface: ce:   DRBE,2% 12. 12.0 0 Gy (R (RBE BE), ), viz., the near-maximum dose to the brainstem surface e.   DRBE,98%  and   DRBE,50% to the GTV are 12.8 and 13.3 Gy (RBE), respectively.

a. b. c. d.

Figure B.4.4. Dose– volu Figure volume me histograms histograms of the GTV and of the brainstem.

¼

 

¼

174  

 APPENDIX B

f. OAR OAR—br —brains ainstem tem surf surface: ace:   DRBE,2%    12 Gy (RBE)

Equipment

a. Beam Beam:: cy cycl clot otro ronn-pr prod oduc uced ed 23 230 0 Me MeV V pr prot oton on beam b. Broad-beam energy modulation modulation c. 360º isocentric isocentric gantry gantry d. Beams: protons degraded degraded to the planned range e. Port Ports: s: th thre ree e po port rtss (w (wit ith h pr proj ojec ecti tion onss cl clos ose e to orthogonal with respect to each other)

Treatment technique

a. The patient was was in the immobilization device. b. The The targ target et po posi siti tion on re rela lati tive ve to th the e fid fiduc ucia iall mark ma rker erss wa wass de dete term rmin ined ed by stu tudy dy of th the e treatme tre atment-p nt-plann lanning ing CT image images. s. For the tre treatatment, men t, ort orthog hogona onall x-r x-ray ay ima images ges of the fid fiduci ucial al mark ma rker erss we were re co comp mpar ared ed wi with th th the e di digi gita tall lly y reconstructed radiographs for each beam path. c. The GT GTV V wa wass ir irra radi diat ated ed us usin ing g a th thre reee-fie field ld appro app roac ach. h. In thi thiss cas case e of a lef left-s t-side ided d aco acous ustic tic neuroma, these comprised left anterior oblique, left le ft po post ster erio iorr ob obli liqu que, e, an and d su supe peri rior or ve vert rtex ex fields. d. Fie Field ld ra rang nge e wa wass de defin fined ed wi with th th the e di dist stal al 90 percent dose level of the SOBP at 2 mm beyond the calculated margin of the GTV. e. Fie Field ld mod modula ulatio tion n wa wass defi defined ned wit with h th the e pr proxi oxi-mal 90 percent dose level of the SOBP at 2 mm proximal to the calculated margin of the GTV. f. The 80– 20 percent percent penumbra penumbra was was    5 mm and resul re sulted ted fr from om the use of a dou double ble-sc -scat atter tering ing gantry gant ry beam beam-mod -modifica ification tion sys system tem with a SAD of 220 cm. Quality assurance

a. Individual portal hardware hardware consisted of custom brasss apert bras apertures ures and PMM PMMA A ran range ge comp compensaensators. These und underwe erwent nt a thor thorough ough QA insp inspecection, consisting of comparing vari rio ous

Total doses delivered

a. Total RBE-weighted doses,  D RBE  Volume/s  V olume/structur tructure e

GTV  Anterior surface of  brainstema   Mid-brainstem

 

DRBE,50%, Gy (RBE)

 DRBE,98%, Gy, (RBE)

 DRBE,2%, Gy (RBE)

13.3 1.1

12.8 0.09

13.5 11.9

,

0.5

a

The bra brains instem tem surf surface ace cons consider idered ed exte extends nds 4 mm bey beyond ond the GTV both superiorly and inferiorl inferiorly y.

b. Total absorbed doses,  D  Volume/s  V olume/structur tructure e

 

GTV  Anterior surface of brains brainstem tema   Mid-brainstem a

D50%, Gy   D98%, Gy   D2%, Gy

12.1 1.0 , 0.5

11.6 0.082

12.3 10.8

The bra brains instem tem surf surface ace cons consider idered ed exte extends nds 4 mm bey beyond ond the

The bra brains instem tem surf surface ace cons consider idered ed exte extends nds 4 mm bey beyond ond the GTV both superiorly and inferiorl inferiorly y.

dimensions on the devices with those required by the treatment-planning system. b. Indi Individua viduall port portals als und underwen erwentt ran range ge and modulatio la tion n mea measur sureme ements nts as we well ll as dos dose e out output put  verification,, which were compared with the  verification planned values.

 Patient status at completion of treatment

a. No change in auditory auditory acuity acuity b. General condition condition unchanged

175  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

 Addendum: technical information Quality-assurance checklist Patient   Field    Action Construction    Aperture  Apertur e geometry check (against planning printout) and thickness Range compensator shape and thicknesses Patient and field identification Calibration and dosimetry   Range (g cm22; treatment machine settings to achieve prescribed range) Modulation (g cm22; treatment machine settings to achieve prescribed modulation) Calibration point (beam coordinates) RBE-weighted absorbed dose [cGy (RBE)] ( DRBE 1200 cGy (RBE), weighted 1:1:1 for three fields) RBE  Absorbed dose (cGy) Isodose normalization (%) Output factor calcula calculation tion C [cGy MU21 at STP] Monitor units (MU at STP) Equipment settings   Treatment room Range modulator/scattering system ID Snout position (cm) ¼

Maximum field sizes LAO: L45A LPO: L55P SUP Range/SOBP LAO

3.0  3.5 cm2 3.1  3.1 cm2 3.1  3.4 cm2 10.3 cm/3.4 cm

– LP1A Verified TR 5/18/2005 Ok Ok Ok SJR SJ R 5/ 5/18 18/2 /200 005 5 8.48

– LA1A Verified TR 5/18/2005 Ok Ok Ok SJR SJ R 5/ 5/18 18/2 /200 005 5 10.27

– SU1A   Verified TR 5/18/2005 Ok Ok Ok SJR SJ R 5/1 /18/ 8/20 200 05 13.65

3.10

3.40

3.00

0,0,0 400.0

0,0,0 400.0

0,0,0 400.0

1.1 363.6 90 1.272 317.6 SJR SJ R 5/ 5/18 18/2 /200 005 5 Gantry 1 A3 18.6

1.1 363.6 90 1.336 302.4 SJR SJ R 5/ 5/18 18/2 /200 005 5 Gantry 1 A4 18.6

1.1 363.6 90 1.444 279.8 SJR SJ R 5/1 /18/ 8/20 200 05 Gantry 1 A5 20.7

LPO SUP

8.5 cm/3.1 cm 13.6 cm/3.0 cm

176  

 APPENDIX B

CASE NUMBER B.5: MEDULLOBLASTOMA (PEDIATRIC) General patient information b. Pa Patie tient nt wa wass st stand andard ard ris risk k and was to rec receiv eive e  D RBE,98%,CTV  23.4 Gy (RBE) standard doses of  D a. Patien Patientt identification identification to cer cerebr ebrosp ospina inall axi axiss and tot total al   DRBE,98%,CTV  i. Name Confidential 54 Gy (RBE) to posterior fossa. ii. Age Age,, gend gender er,, and and rac race e 10 yea yearr old old c. RBE 1.10 Caucasian male d. Prob Probabilit abilities ies of outc outcome: ome: iii. Address Confidential iv. Phone no. Confidential i. Cur Cure e at 5 years: 85 percent percent  v.. E-mail address  v Confidential Confiden tial ii. Probability of treatment-rela treatment-related ted morbidity  vi. Hospital ID no. Confidential Confide ntial  Late hearing loss: 20 percent  vii. Person to notify Confidential Confiden tial   Ne Neur uroo-co cogn gnit itiv ive e pa part rtia iall lo loss ss at 10 (with contact years: approaches 100 percent information)) were information   ¼

¼

¼

† †

a problem to arise b. Medi Medical cal team team i. Radiati Radiation on oncol oncologis ogist(s) t(s) ii. Radi Radiat ation ion ph physi ysicis cist(s) t(s) iii. ii i. Ref efer erri ring ng ph phys ysic icia ian n

†

Confidential Confiden tial Confi Co nfiden dentia tiall Con onfid fiden enti tial al

Medical evaluation of presenting lesion

a. Me Medi dica call hi hist stor ory y b. Physical exam ex amin inat atio ion n

Fou ourr wee eek k he head ada ach che e an and d emesis Normal neurological and gene ge nera rall ex exam amin inat atio ion n

  Me Meas asur urab able le100 loss lo ss of IQ at 10 ye year ars: s: approaches percent.

Treatment planning 

a. General General plan i. Th The e pa pati tien entt was tr trea eate ted d ac acco cord rdin ing g to Institutional Protocol No. 99271. ii. Patien Patientt was post near complete complete resection. iii. Pos Postope toperat rative ive prot proton on radi radiati ation on trea treatmen tment, t, in two segments. iv.. VCR chemotherapy iv chemotherapy given concurrent with,

c. Imag Imagin ing g st stud udie iess

iv.. VCR chemotherapy iv chemotherapy given concurrent with, and following, the radiation therapy. b. Definitions of treatment treatment volumes volumes i. For Segments 1 and 2, there was no GTV. ii. Fo Forr Seg Segmen mentt 1, the CTV-1 CTV-1 wa wass defi defined ned as the entirety of the CNS,   i.e., the brain and the thecal sac (to bottom of S3). iii.. Fo iii Forr Seg Segmen mentt 2, the CTV-2 CTV-2 wa wass defi defined ned as the posterior fossa. iv.. Tre iv reat atme ment nt pl plan anni ning ng CT im imag ages es wer ere e obtained with the patient in the treatment position.   For Segment 1, the patient was in the prone position and 3.75 mm CT slices were we re us used ed to sc scan an fr from om th the e su supe peri rior or aspect of cranium to the coccyx.

Pre-su Presurg rgic ical al CT and and MR MRI: I: partially calcified 5.5  3.5  3

d. e. f. g.

Tumor si sitte Diagnosis Grade Stage

h. Prior therapy i. Fami Family ly his histor tory y

3.5 cm mas masss wit within hin fou fourth rth  ventricle extending into later la teral al re reces cess; s; obs obstru tructi ctive ve hydrocephalus Poste teri rio or fos ossa sa Medulloblastoma WHO grade IV   Preoperative. CSF negative. Standard risk patient Near total resection; minimal residual tumor Pater Pa ternal nal grand grandmot mother her,, lung cancer can cer.. Mu Multi ltiple ple ma mater ternal nal breast cancers

†

†

 For Segment 2 the the supine position andpatient 2.5 mmwas CT in slices were we re us used ed to sc scan an fr from om th the e su supe peri rior or aspect of cranium to the level of C4.  v.. PTV  v PTV:: not delineated for Segments 1 or 2. Field sizes were defined to assure delivery of the planned dose to the CTV. c. Dose fractionation: fractionation: D RBE,98% 1.8 Gy (RBE) per fraction and five fractions per week. d. RB RBE E is 1.10 1.10 e. Organ Organss at risk and dose constrain constraints ts defined as the dose producing major morbidity,  viz., Grade III/IV severity of complications at   , 5 percent. This Th is pe pert rtai ains ns to rad adia iati tion on co comb mbin ined ed wi with th

General medical evaluation

a. History, History, physical examination, and imaging b. Co-morbidities c. Cur Curre rent nt me medi dica cati tion onss

No evident non-tumor disease None Anti An ti-e -eme meti ticc medication

¼

and an d

pain pa in

Treatment intent

a. Cura Curati tive ve by co comb mbin ined ed mo moda dali lity ty ther therap apie ies: s: surg su rge ery ry,, chem ch emot oth her era apy, and an d rad adia iati tion on.. Radiation to be given postoperatively.

chemotherapy. 177

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

i. Coc Cochlea hlea::  D RBE,2% 40 Gy (RBE) ii. Spin Spinal al cord: cord:  D RBE,2% 45 Gy (RBE) There was no need to reduce the dose to CTV-1 or CTV CT V-2 at an any y po poin intt to re resp spec ectt th the e de defin fined ed do dose se constraints f. Th The e XI XIO O tr trea eatm tmen entt pl plan anni ning ng sys yste tem m was employed i. Dose statement statement accuracy: accuracy: 2 percent ii. Position accuracy: accuracy: 2 mm mm ¼

¼

g. Isodose Isodose and color wa wash sh disp display layss of dos dose e dis distri tri-butions are shown in Figs B.5.1–B.5.3 and dose–  volume  volu me histograms histograms are shown shown in Fig. B.5.4. Patient immobilization and positioning 

a. For For ea each ch tr trea eatm tmen entt se sess ssio ion n of Se Segm gmen entt 1, th the e patient was anesthetized and then placed in the prone position in the immobilization device, which consisted of a base plate, rocker device and face

Figure B.5.1. Segme Figure Segment nt 1. Axia Axiall sect section ion through cerebellu cerebellum m and eyes: isodo isodose se cont contour our display (a) and color-wash color-wash display (b) of dose distributions.

178  

 APPENDIX B

Figure B.5.2. Segment 1. Spinal axis (sagittal section): isodose-contour isodose-contour display (a) and color-w color-wash ash display (b) of dose distributions. For the Segment 1 treatment the patient was prone.

mask (Me mask (Med-T d-Tec, Inc Inc.). .). Ort Orthog hogona onall x-r x-ray ay ima images ges were we re obt obtain ained ed for a sta stand ndard ard set set-up -up position position as well the treatment fields prior to treatment. These images were then compared and analyzed in the digital-ima digita l-imaging ging positi positioning oning sy system stem.. Pa Patient tient position iti on cor correc rectio tions ns wer were e mad made. e. Thi Thiss wa wass rep repea eated ted until the intended position was achieved, applied, and reconfirmed prior to each treatment. b. Fo Forr Seg Segmen mentt 2, th the e pa patie tient nt was sup supine ine in the immo im mobi bili liza zati tion on de devi vice ce.. An Anes esth thes esia ia wa wass no nott required. c. Ort Orthog hogona onall x-r x-ray ay ima images ges we were re obt obtain ained ed for a standard set-up position as well the treatment fields fiel ds pri prior or to eac each h tr trea eatme tment. nt. Th These ese ima images ges wer ere e th then en co comp mpar ared ed an and d an anal alyz yzed ed in th the e digital-imaging positioning system. Patient position corrections were made. This was repeated

Treatment prescription

a. Prescribed Prescribed dose dose i. Proton radiation radiation treatment. treatment. ii. Doses wer were e pres prescribed cribed in RBE RBE-we -weighte ighted d doses, D RBE  [Gy (RBE)]. iii.. RB iii RBE E 1.10 iv. Doses were specified as   DRBE,50%,  DRBE,98% and  D RBE,2%  v.. Fraction  v Fractionation: ation:   DRBE,98% 1. 1.8 8 Gy (R (RBE BE)) for five fractions per week for 30 fractions over a time of 6 weeks.  vi. Segment 1 To CTV CTV-1, -1,   DRBE,98% 23.4 Gy (RBE) in 13 equ equal al fr frac actio tions ns of   dRBE,98% 1. 1.8 8 Gy (RBE).  vii. Segment 2 ¼

¼

¼

¼

¼

To CTV CTV-2, -2,   DRBE,98% 30 30.6 .6 Gy (R (RBE BE)) given in 17 equal fractions of   dRBE,98% 1.8 Gy (RBE).

untill th unti the e in inte tend nded ed po posi siti tion on wa wass ac achi hiev eved ed,, applied, and reconfirmed prior to each treatment.

¼

179  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

Figure Figu re B.5.3. Total Total dose to post posterio eriorr fossa fossa:: Segm Segments ents 1 and 2. Pat Patient ient in supi supine ne position for Segment 2. Axial (a) and sagit sagittal tal (b) sections through posterior fossa with isodose contours.

180  

 APPENDIX B

  Fie ield ld mod odu ula lati tion onss fo forr the brain fieldss wer field were e defin defined ed with the prox proximal imal 98 per percen centt of th the e SO SOBP BP at the brain brain surface. c. Segm Segment ent 2. Posterior Posterior fossa †

i. Dose Dose by Segment Segment 1,   DRBE,98%   was 23.4 Gy (RBE) ii. Dos Dose e by Segment Segment 2,   DRBE,98%   was 30.6 Gy (RBE), delivered by a posterior, a right pos-

terior ter ior an and d a lef leftt pos poster terior ior obl obliqu ique e fiel fields, ds, viz., a boost treatment. Quality assurance Figure B.5.4. Dose– volu Figure volume me histograms histograms for (from left to righ right) t) brain bra in (posteri (posterior or fossa fossa)) (pink), pitu pituitar itary y glan gland d (yello (yellow), w), righ rightt cochl coc hlea ea (gre (green) en),, lef leftt coc cochl hlea ea (r (red) ed),, hyp hypoth othala alamu muss (bl (blue) ue),, and PTV (cer (cerebru ebrum) m) (re (red). d).   DRBE,PTV ,98% 55 55.5 .5 Gy (R (RBE BE)) as indicated by the vertical red line.

a. Individua Individuall porta portall hard hardwar ware e consi consiste sted d of cust custom om brasss aper bras aperture turess and PMM PMMA A ran range ge comp compensa ensa-tors. These unde underwen rwentt a thor thorough ough QA insp inspecection ti on be befo forre tr trea eatm tmen ent. t. Th This is co cons nsis iste ted d of  measuring measu ring vario various us dime dimension nsionss on the devic devices es and an d co comp mpar arin ing g th them em wi with th th the e ou outp tput ut of th the e treatment-planning system. b. Rang Range e modu modulat lation ion and absor absorbedbed-dose dose veri verificafication meas measurem urements ents wer were e made for indi individua viduall trea tr eatm tmen entt po port rtal alss an and d we were re co comp mpar ared ed wi with th

¼

b. T Tota otall dos dose e to pos poster terior ior fos fossa: sa:   DRBE,98% 54.0 Gy (RBE) in 30 fractions in 6 weeks. c. OA OAR: R: Dos Dose e con const stra raint intss we were re not ex excee ceeded ded by these prescribed doses. ¼

planned values. Equipment

Treatment technique

a. b. c. d.

Cyclotron-produced 230 MeV proton Cyclotron-produced proton beam Broad-beam passive passive energy modulation modulation 3608  rotational gantry Digita Digi tall im imag agin ing g sy syst stem em fo forr pa pati tien entt po posi siti tion on  verification e. 80– 20 pe perc rcen entt pe penu numb mbra ra wa wass    5 mm wit ith h passive double-scattering gantry beam-delivery system with SAD of 220 cm.

a. Proton Proton beam beam b. Segm Segment ent 1. 1. i. Spin Spinal al axis (the target was the thecal thecal sac)   Two matc ma tchi hing ng post po ster erio ior– r– an ante teri rior or prot pr oton on fie field ldss we were re us used ed to tr trea eatt th the e spinal axis to the inferior border of the thec th ecal al sa sac. c. Th The e SO SOBP BP fo forr th the e sp spin ine e field fie ld wa wass de desi sign gned ed fo forr th the e di dist stal al 75 percen per centt iso isodos dose e of the SOBP to be at the th e an ante teri rior or ed edge ge of th the e ver erte tebr bral al bodies, to enhance esophagus sparing. This resulted in delivery of   DRBE,98%   of   20–22 Gy (RBE) to vertebral bodies, intending inten ding to pro produce duce unif uniform orm gro growth wth modificati modifi cation. on. The prox proximal imal 90 per percent cent isodose of the SOBP was set at the posterior edge of the spinous processes.   A mo movi ving ng-j -jun unct ctio ion n te tech chni niqu que e was employed. ii. Whol Whole e brain brain   Tre reat ated ed wi with th ri righ ghtt an and d le left ft po possterio er iorr ob obli liqu que e fiel eld ds to   DRBE,98% 23.4 Gy (RBE) and with sparing of the †

Total doses delivered

a. Total RBE-weighted RBE-weighted absorbed doses,  D RBE i. Spin Spinal al axis axis   DRBE,2% 25.7 Gy (RBE) at the anterior vertebral body matchline.   DRBE,98% 22.2 Gy (RBE) at the posterior thecal sac matchline ii. Po Post steri erior or fos fossa sa by Seg Segmen ments ts 1 and 2 wa wass  DRBE,98% 54 Gy (RBE) in 30 fractions iii. ii i. The herre wer ere e no in inte terr rrup upttio ions ns in the treatment. iv.. Dose to OARs iv OARs   Do Dose se to th the e co coch chle lea a wa wass re redu duce ced d by only on ly se sev ver eral al pe perc rcen entt fo forr th the e wh whol ole e †

†

†

¼

¼

†

†

¼

†

¼

  lenses. Fiel Fi eld d ra rang nges es fo forr al alll th the e br brai ain n fie field ldss wer ere e defi efin ned wi with th the dis isttal 10 100 0 percent of the SOBP conforming to the brain surface.

brain fields, significantly more for the th e po post ster erio iorrbut foss fo ssa a bo boos ost. t. Th The e to tota tall dose to the cochlea was   DRBE 30.4– 36.7 Gy (RBE) and below the dose constraint str aint of   DRBE 40 Gy (R (RBE BE). ). Th This is ¼

¼

181  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

compares to an absorbed dose of 54–55 Gy for photon treatment in this department that results in deafness in 50 pe perc rcen entt of pa pati tien ents ts in th this is ag age e group.

b. Total absorbed doses, D Target/  structure

 D, Gy

Dose per fraction fract ion (d),

Number of 

†

†

  Lens dose:   DRBE 3–5 Gy (RBE) with oblique proton fields  Scalp dose:  D RBE     16 Gy (RBE)

Target/  structure

 DRBE, Gy (R (RB BE)  D98%   D50%   D2%

CTV1 (CSI)a CTV2 (posterior fossa) Pituitary Hypothalamus Cochlea Lens Esophagus Thyroid a

 D98%   D50%   D2%

¼

23.4 54.3

23 2 3.4 55.6 55

25.7 56.4

25.6 3.0 26.2 3.6 0.1 0

2 7. 7.1 5.3 30.4 30 5.4 2.3 0

36 3 6.5 7.8 36.7 7.6 12.4 2.6

Dos ose e per fraction (dRBE), Gy (RBE)

Number of  fractions

1..8 1 1.8 1.

13 30

CTV1 (CSI) 21.3 CTV2 (posterior 49.4 fossa) Pituitary Hypothalamus Cochlea Lens Esophagus Thyroid

23.4 51.3

23.3 2 24 4.6 2.7 4.8 23.8 27.6 3.2 4.9 0.09 2. 2 .1 0 0

33.2 7.1 33.4 6.9 11.3 2.4

fractions

1.64 1.64

13 30

Patient status at completion of treatment

a. Persistent Persistent emesis b. To continue chemotherapy chemotherapy c. Follow-u Follow-up p appointmen appointmentt in 4 weeks

Craniospinal irradiation.

182  

 APPENDIX B

 Addendum: technical information

21.3 21 50.5

Gy

Quality-assurance checklist   Patient   Field    Action   Construction  Aperture geometry check (against planning printout) and thickness Range compensator shape and thicknesses Patient and field identification

– LP2A Verified TR 5/ 5/18 18/2 /200 005 5 Ok

– RP2A Verified TR 5/ 5/18 18/2 /200 005 5 Ok

– SS2A Verified TR 5/ 5/18 18/2 /200 005 5 Ok

– IS2A Verified TR 5/ 5/18 18/2 /200 005 5 Ok

– PA1A Verified TR 5/ 5/18 18/2 /200 005 5 Ok

– RP4B Verified TR 5/ 5/18 18/2 /200 005 5 Ok

– LP4B Verified TR 5/ 5/18 18/2 /200 005 5 Ok

Ok

Ok

Ok

Ok

Ok

Ok

Ok

Ok

Ok

Ok

Ok

Ok

Ok

Ok

Calibration and dosimetry Range (g cm22) Modulation (g cm22) Calibration point (beam coordinates) RBE [cGy (RBE)] (RX dose  DRBE  D of 1200 cGy (RBE) weighted 1:1:1 for three fields) RBE factor  Absorbed dose (cGy) Isodose normalization (%) Output factor calculation C (cGy MU21 at STP) Monitor units (MU at STP)

SJR 5/18 5/18/200 /2005 5 SJR 5/18 5/18/200 /2005 5 SJR 5/18 5/18/200 /2005 5 SJR 5/18 5/18/200 /2005 5 SJR 5/18 5/18/200 /2005 5 SJR 5/18 5/18/200 /2005 5 SJR 5/18 5/18/200 /2005 5

  Equipment settings Treatment room Range modulator/scattering system ID Snout position (cm)

SJR 5/18 5/18/200 /2005 5 SJR 5/18 5/18/200 /2005 5 SJR 5/18 5/18/200 /2005 5 SJR 5/18 5/18/200 /2005 5 SJR 5/18 5/18/200 /2005 5 SJR 5/18 5/18/200 /2005 5 SJR 5/18 5/18/200 /2005 5 Gantry 1 Gantry 1 Gantry 1 Gantry 1 Gantry 1 Gantry 1 Gantry 1  A6 A6 A2 A3 A2 A5 A5

17.50 17.92 ( full ) 0,0,0

17.50 1 7.92 ( full ) 0,0,0

7.35 7.20 0,0,0

8.30 7.90 0,0,0

10.5 8.80 0,0,0

13.55 12.0 0,0,0

13.25 11.9 0,0,0

90

90

180

180

180

90

90

1.1 82 98 0.819

1.1 82 98 0.819

1.1 164 98 0.814

1.1 164 99 0.840

1.1 164 98 0.962

1.1 82 98 0.877

1.1 82 98 0.866

102.2

102.2

205.6

197.2

174

95.4

96.6

26.9

26.8

22.5

21.0

22.5

15.7

15.5

 Beams for Segment 1 (cranial spinal axis treatment) Beam number

Description Machine ID Collimator Set up/dist (cm) SSD/Wt fan SSD (cm) Field defined at Width (cm, L/R if asym) Length (cm, U/L if asym) Gantry/collimator angle (deg) Couch/pitch/roll angle (deg) Isocenter/beam entry  x /  y /  z   (cm) Tx aids Customized port ID Iso-to-port dist (cm) Compensator Iso-to-RC dist (cm) Taper angle (deg)/margin (cm) 30.0/1.00 Snout ID Degrader ID

1

2

3

4

L1 L15P WB 1G protons Sym SAD/227.0 217.5/217.5 Isocenter 27.0 27.0 285/-0 0/0/0 Iso 0.20/ 2 1  .50/1.00

L1 R15P WB 1G protons Sym SAD/227.0 217.6/217.6 Isocenter 27.0 27.0 75/-0 0/0/0 Iso 0.20/ 2   1.50/1.00

L1 SS 1G protons Sym SAD/227.0 219.0/219.0 Isocenter 26.4 26.4 0/-0 0/0/0 Iso 0.20/ 2   21 21.5 .50/ 0/1. 1.00 00

L1 IS 1G protons Sym SAD/227.0 221.2/221.2 Isocenter 26.2 26.2 0/-0 0/0/0 Iso 0.20 0. 20/  / 242.50/1.00

L1v1LWB 26.91 LWBv1RC 15.52 30.0/1.00 250

L1v1RWB 26.83 RWBv1RC 15.44 30.0/1.00 250

L1v1SS 22.46 SSv1RC 18.04 0.0/0.00 250

L1v1IS 20.95 ISv1RC 15.81 0.0/0.00 250 Continued

183  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

Continued Beam number 1

2

3

4

Compensating block ID Calculation algorithm Weight (cGy)/no. of fractions

– Pencil beam 365.0/4

– Pencil beam 365.0/4

– Broad beam 707.0/4

– Broad beam 716.0/4

 x, y, z   (cm) Target Beam spreading  Air gap (cm) Prescribed range (g cm22) Raw, w/uncert Prescribed modulation (g cm22) Raw, smr, smr w/uncert Smearing distance (cm) Uncertainty parameters Density (%) Range (cm) Dose (cGy MU21)

0.00/0.00/0.00 Brain PTV Passive 6 17.5 16.64/17.32 17.92 15.70/15.83/16.82 0.3 Nominal 3.5 0.1 0.811

0.00/0.00/0.00 Brain PTV Passive 6 17.5 16.59/17.27 17.92 15.56/15.70/16.74 0.3 Nominal 3.5 0.1 0.811

0.00/0.00/0.00 Vertebra RC Passive 10 7.2 7.17/7.52 7.2 6.08/6.21/6.69 0.3 Nominal 3.5 0.1 0.821

0.00/0.00/0.00 Vertebra RC Passive 10 8.2 8.18/8.56 7.9 6.72/6.85/7.38 0.3 Nominal 3.5 0.1 0.815

 Beams for Segment 1 ( posterior fossa treatment). Beam number 1

2

3

4

5

6

PA PF 1G protons Sym SAD/227.0 219.0/219.0 Isocenter 12.5 12.5 180/0 0/0/0 Iso 0.35/  

PA PF offcord 1G protons Sym SAD/227.0 219.0/219.0 Isocenter 12.5 12.5 180/0 0/0/0 Iso 0.35/ 

R3 R 30P PF 1G protons Sym SAD/227.0 219.3/219.3 Isocenter 12.5 12.5 240/0

L25P PF 1G protons Sym SAD/227.0 219.2/219.2 Isocenter 12.5 12.5 115/0 0/0/0 Iso 0.35/0.58/ 2 3.3 3.31 1

L15P WB 1G protons Sym SAD/227.0 218.1/218.1 Isocenter 26.7 26.7 105/0 0/0/0 Iso 0.20/2 0.2 0/2.50 .50/  / 2 1.5 1.50 0

R15P WB 1G protons Sym SAD/227.0 217.9/217.9 Isocenter 26.8 26.8 255/0

0.58/ 2 3.31

0.58/ 2 3.31

0.58/ 2 3.31

PAv1PF 15.53 PAv1PF 9.99 15.0/0.50 124

PAv2PF 15.53 PAv1PF 9.99 15.0/0.50 124

R30Pv1PF 15.69 R30Pv1PF 9.71 15.0/0.50 124

L25Pv1PF 15.48 L25Pv1PF 9.81 30.0/1.00 124

L15PsupWB 24.68 L15PsupWB 14.92 30.0/1.00 250

R15PsupWB 25.3 R15PsupWB 15.12 30.0/1.00 250

– Pencil beam 742.0/4 0.00/0.00/0.00 GTV Passive 2 10.5

– Pencil beam 370.0/2 0.00/0.00/0.00 0. Post fossa BF Passive 2 10.5

– Pencil beam 1018.0/11 0.00/0.00/0.00 Post fossa BF Passive 2 13.5

– Pencil beam 1019.0/11 0.00/0.00/0.00 0. Post fossa BF Passive 2 13.2

– Pencil beam 1236.0/13 0.00/0.00/0.00 Brain PTV Passive 6 17.7

– Pencil beam 1236.0/13 0.00/0.00/0.00 Brain PTV   Passive 6 17.9

Raw,w/uncert Prescribed modulation (g cm 22) Raw Ra w, sm smrr, sm smrr w/ w/un unce cert rt

10.03/10.48 8.8 7.38 7. 38/7 /7.7 .70/ 0/8. 8.35 35

10.03/10.48 8.8 7.38 7. 38/7 /7.7 .70/ 0/8. 8.35 35

Smearing distance (cm) Uncertainty parameters Density (%) Range (cm) Dose (cGy MU21)

0.3 Nominal 3.5 0.1 0.994

0.3 Nominal 3.5 0.1 0.994

12.88/13.43 12.57/13.11 11.98 11.98 9.87 9. 87/1 /10. 0.95 95/1 /11. 1.74 74 10 10.3 .33/ 3/10 10.5 .59/  9/  11.39 0.3 0.3 Nominal Nominal 3.5 3.5 0.1 0.1 0.940 0.929

16.85/17.54 17.92 15.69/15.91/  16.90 0.3 Nominal 3.5 0.1 0.809

17.01/17.70 17.92 15.63/16.18/  17.21 0.3 Nominal 3.5 0.1 0.809

Description Machine ID Collimator Setup/dist (cm) SSD/Wt fan SSD (cm) Field defined at Width (cm, L/R if asym) Length (cm, U/L if asym) Gantry/coll angle (deg) Couch/pitch/roll angle (deg) Isocenter/beam entry  x /  y /  z  z   (cm) Tx aids Customized port ID Iso-to-port dist (cm) Compensator Iso-to-RC dist (cm) Taper angle (deg)/margin (cm) Snout ID Degrader ID Compensating block ID Calculation algorithm Weight (cGy)/no. of fractions  x /  y /  z  z   (cm) Target Beam spreading  Air gap (cm) Prescribed range (g cm22)

Iso 0.35/ 

184

Iso 0.20/2 0.2 0/2.50 .50/  / 2 1.50

 

CASE NUMBER B.6: SKULL-BASE CHORDOMA

CASE NUMBER B.6: SKULL-BASE CHORDOMA  General patient information

a. Patient Patient identification identification i. Name ii.. Ag ii Age, e, ge gend nder er,, an and d ra race ce iii. iv.  v..  v  vi.  vii.

Confidential 61 ye year ar ol old d Caucasian male Address Confidential Phone no. Confidential E-mail address Confidential Confiden tial Hospital ID no. Confidential Confiden tial Person to notify Confidential Confiden tial (with contact information) were a problem to arise

b. Medi Medical cal team team i. Radia Radiatio tion n onc oncolo ologis gist(s) t(s) ii.. Ra ii Radi diat atio ion n ph phys ysic icis ist(s t(s)) iii. ii i. Ref efer erri ring ng phy physi sici cian an

Confiden Confi dentia tiall Confi Co nfide dent ntia iall Con onfid fiden enti tial al

Medical evaluation of presenting lesion

a. Medi Medical cal hist history ory Six months, months, binocular binocular oblique oblique doub do uble le vi visi sion on on do down wnga gaze ze,, drooping of left upper eyelid b. Physical Left CN III, IV palsies examination c. Imag Imaging ing stud studies ies MRI MRI:: 3 cm (max (maximum imum dime dimennsion) extradur extradural al tumor of left lateral clivus d. Tum umor or sit site e Clivu Cl ivus, s, in invad vadin ing g lef leftt ca cave vern rnou ouss sinus, encasing left cav cavernous ernous carotid car otid art artery ery, dis displa placing cing left mesial temporal lobe, effa effacing cing prepo pr eponti ntine ne cis cistern tern,, ele elevat vating ing lef eftt li lim mb of op opttic chi hias asm m. Figu Fi gure re B. B.6. 6.1 1 is th the e T2 MR MRII image of the residual lesion at start of radiation treatment. e. Di Diag agno nosi siss Chor Ch ordo doma ma (pa (path thol olog ogic ical al stu tudy dy of surgical specimen) f. Grade No grade g. Stage No staging system h. Pr Prio iorr th ther erap apy y Le Left ft pter pt erio iona nall cran cr anio ioto tomy my, trans-sylvian approach, partial tumor removal. No irradiation i. Fam Family ily histo history ry Moth Mother: er: thro throat at cance cancerr

General medical evaluation

a. History, History, physical examination, and imaging

Impairment of short-term memory since surgery, left CN III palsy, KPS 90 ¼

c. Cu Curr rren entt me medi dica cati tion onss None No ne d. Prior radi radiati ation on ther therapy apy None Treatment intent

a. Curativ Curative e b. No systemic systemic therapy therapy c. Pred Predicted icted outcome outcome at 10 years i. Tumor control probability: 60 percent percent ii. Blin Blindnes dness: s:   , 1 percent iii. Tempo Temporal ral lobe inju injury ry (sym (symptom ptomatic atic): ):   percent iv.. Br iv Brain ainst stem em inj injury ury (s (symp ymptom tomat atic) ic)::   percent  v.. Pituitary dysfunction: 100 percent  v

,

5

,

5

Treatment planning 

a. Gener General al plan plan:: pos postoper toperativ ative e combi combined ned pro proton ton and photon radiation therapy b. Definition of treatment treatment volumes volumes i. The treatment-planning treatmentplanning CT study was performed forme d with the pat patient ient in the treatment treatment positi pos ition on an and d imm immobi obiliz lized ed wit with h a cus custom tom-made thermoplastic head mask. ii. CT sec sectio tions ns of 2.5 mm we were re emp employ loyed ed in the defin definition ition of the GTV, CTV and and OAR OARss in the treatment planning system. iii.. The GTV wa iii wass th the e gro gross ss re resid sidual ual tumor tumor as seen on the imaging studies. The CTV was delineat delin eated ed to inclu include de possi possible ble micr microscop oscopic ic tumor extensions (see Fig. B.6.2). PTV was not delineated. c. Combined radiation radiation beams. beams. i. 10 MV x-ray x-ray bea beams ms we were re to be employed employed forr th fo the e in init itia iall 2 we week ekss of tr trea eatm tmen entt an and d then th en tr trea eatm tmen entt co comp mplet leted ed by 23 230 0 Me MeV  V  proton beams. d. Dose constrai constraints nts i. Bra Brainst instem em   DRBE,2% 55 55..0 Gy (R (RB BE) at center/67.0 Gy (RBE) at surface ii. Opti Opticc chia chiasm sm and nerves:   DRBE,2% 62 Gy (RBE) iii.. Fo iii Forr tr trea eatme tment nt of sku skull ll bas base e sar sarcom coma, a, no const con stra raint intss are for forma mally lly ap appli plied ed for th the e temporal lobes and the pituitary gland. e. Dose calculat calculation ion i. Thr Threeee-dim dimens ension ional al pr proto oton n pen pencil cil-be -beam am algorithm based on 2.5 mm CT slices of the whole head ii. Clarkson algorithm for the photon component iii. ii i. Do Dosse is sta tatted wi witth an acc ccu uracy of 2 percent. ¼

¼

b. Coo-mo morb rbid idit itie iess

iv.. Treatmen iv reatmentt planning used the XIO system

Obes Ob esit ity y

185  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

dose–v e–vo olume Fig. B.6.3.

histograms

are

shown

in

Treatment technique

Figure Figu re B. B.6. 6.1. 1. Ax Axia iall T2 MR MRII de demo mons nstr trat atin ing g chordoma encasing the cavernous carotid artery artery..

the th e

resi re sidu dual al

f.   DRBE,50% 76 Gy (RBE) (RBE) to th the e GT GTV V at 2 Gy (RBE) per fraction, five fractions per week ¼

Patient immobilization and positioning 

a. Supine position with a thermoplastic thermoplastic mask b. Th The e pos positi itions ons of th the e defi defined ned bon bony y lan landm dmark arkss (fiduc (fid ucia ials ls)) re rela lati tive ve to th the e be beam am we were re de dete terrmined min ed by bibi-pla planar nar rad radiog iograp raphs. hs. The pa patie tient nt pos osit itio ion n was ad adjjus uste ted d an and d the im ima agi gin ng repe re peat ated ed;; th this is co cont ntin inue ued d un unti till th the e de desi sire red d target position was achieved.

a. The CTV was irradiated irradiated using using a combination combination of  photo ph otons ns and pr proto otons. ns. The pho photon ton com compon ponent ent wass thr wa three ee fiel fields: ds: right and lef leftt la later teral al and a superior anterior oblique. b. The The pr prot oton on co comp mpon onen entt to th the e CT CTV V wa wass al also so three thr ee fiel fields: ds: rig right ht ant anterio eriorr and lef leftt ant anteri erior or oblique and a posterior. c. The The GT GTV V was ir irrrad adia iate ted d us usin ing g fiv five e pr prot oton on beam be ams, s, a ve vert rtex ex fie field ld an and d tw two o pa patc tche hedd-fie field ld combinat combi nations: ions: a left supe superior rior post posterior erior oblique paire pai red d wit with h a rig right ht la later teral al pa patch tch and a rig right ht infe in feri rior or ob obli liqu que e pa pair ired ed wi with th a le left ft po post ster erio iorr patch. d. Fie Field ld ra rang nge e wa wass de defin fined ed wi with th th the e di dist stal al 99 percent of the SOBP at the distal margin of the CTV and GTV, GTV, respectively respectivel y. e. Fie ield ld mod odu ula lattio ion n was defi efine ned d wi witth th the e proxi pr oxima mall 98 per percen centt dos dose e of the SOBP at the proximal margin of the CTV and GTV, respectively. f. Pen enum umbr bra: a: 80–20 pe perc rcen entt do dose se gr grad adie ient nt was    5 mm an and d res esu ult lte ed from use of th the e double dou ble-sc -scat atter tering ing gan gantry try bea beam m wit with h SA SAD D of  227 cm. g. See Addendum for additional technical details.

Quality assuranc assurance e

a. Individua Individuall porta portall hard hardwar ware e cons consiste isted d of cus custom tom brasss apert bras apertures ures and PMM PMMA A ran range ge comp compensaensators. These und underwe erwent nt a thor thorough ough QA insp inspecection consisting of measuring various points on the th e de devi vice cess us usin ing g re refe fere renc nce e po poin ints ts fr from om th the e computer planning system. b. Indi Individua viduall porta portals ls unde underwen rwentt ran range ge and modulatio la tion n mea measur sureme ements nts as we well ll as dos dose e out outpu putt  verification,, which were compared with  verification planned values.

Treatment prescription

a. 10 MV x-ray and 230 MeV proton proton beams b. Pre Prescrib scribed ed doses i. GTV GTV::   DRBE,50% 76. 76.0 0 Gy (RB (RBE) E) adm admini inisstered ter ed at   dRBE,50% 2 Gy (R (RBE BE)) pe perr fr frac ac-¼

¼

Equipment a. Protons: cyclotron-produced cyclotron-produced 230 MeV MeV beam i. Broad-beam energy modulation modulation

tion, five fractions per week ii. Ant Anterior erior surface of brai brainst nstem: em:   DRBE,98% 62 Gy (RBE) iii.. RB iii RBE E 1.10 c. Isod Isodos ose e co cont ntou ours rs in th the e ax axia ial, l, co coro rona nal, l, an and d sagi sa gitt ttal al pl plan anes es ar are e sh show own n in Fi Fig. g. B. B.6. 6.2 2 an and d

ii. 3608  isocentric gantry iii. Protons degraded degraded to the planned range b. Photons: 10 MV linear accelerator accelerator i. Isocentric setup c. Po Ports rts:: thr threeee-fiel field d ph photo oton n tec techni hnique que,, 4 pr proto oton n ports and 2 proton patch combinations

¼

¼

186  

CASE NUMBER B.6: SKULL-BASE CHORDOMA

Figure Figu re B. B.6. 6.3. 3. Do Dose–vo se–volu lume me hist histog ogra rams ms fo forr GT GTV V, CT CTV V, brainst brai nstem, em, chia chiasm, sm, right and left optic nerv nerve, e, and pitu pituitar itary y gland.

Total doses delivered

a. Total RBE-weighted RBE-weighted absorbed doses,  D RBE  Volume/   Volume/  structure

 Volume,  Volume, ml

 DRBE,50%, Gy (RBE)

GTV CTV Brainstem Opti ticc chi hia asm Right optic nerve Left optic nerve Pituitary

9.5 14.3 32.5 0.5 1.3

77.3 (20.4) 63.2 (20.2) 76.3 (20.4) 61 61.8 (20.2) 24.7 (13.9) 24.7 (0.6) 57. 7.8 8 (2 (20. 0.4) 4) 44 44.0 .0 (20 20.4 .4)) 7.8 (0.6) 0.5 (0.2)

82.3 (20.6) 82.3 (20.6) 82 62.8 (20.6) 62.4 62 .4 (20 20.4 .4)) 42.2 (20.2)

0.7 (0.6) 62.9 (20.4)

59.8 (20.3) 7 2. 2.8 (20.4)

1.2 0.5

 DRBE,98%, Gy (RBE)

0.2 (0.2) 5 3. 3.5 (20.4)

 DRBE,2%, Gy (RBE)

The numbers in parentheses are the doses administered by photon beams. i.   DRBE,50% 77.3 Gy (RB (RBE) E) was delivered delivered to the GTV in 38 fr frac actio tions, ns, one fr fract action ion per day, five fractions per week. The treatment wass not int wa interr errup upted ted du durin ring g the 54 da days ys.. That the delivered RBE-weighted absorbed dose exceeded the prescribed dose by 1.3 Gy (RBE) was the consequence of the need to employ patch fields. ii. Dose constraint constraint to the anterior anterior surface of the brainstem was exceeded by ,1.5 percent. ¼

b. Total absorbed dose, D

Figure B.6.2. Isodo Figure Isodose se contours contours in the axial (a), coronal coronal (b), and sagittal (c) planes. GTV and CTV are represented by solid black and dotted black lines, respectively.

 Volume/s  V olume/structur tructure e

Volume, ml   D50%, Gy   D98%, Gy   D2%, Gy

GTV CTV Brainstem Optic chiasm Right optic nerve Left optic nerve Pituitary

9.5 14.3 32.5 0.5 1.3 1.2 0.5

72.1 71.2 23.7 54.4 7.2 – 62.9

59.2 58.0 22.5 41.9 – – 53.5

76.7 76.7 59.0 58.6 40.2 56.2 72.8

187  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY 

follow foll ow-u -up p ev eval alua uati tion onss ar are e to be co cond nduc ucte ted d and appr appropria opriate te imag imaging ing stud studies ies unde undertak rtaken en in 6 months, 12 months and annually thereafter.

Patient status at completion of treatment

a. No acute acute reactions reactions b. General condition condition unchanged. unchanged. c. The patie pa tient nt lives liv some som e 12re 000 miles distan dis tant. t. In par partne tnersh rship ip es with wit h the refer ferrin ring g ph phys ysicia ician, n,

 Addendum: technical information Quality-assurance checklist

Patient Field  Action Cons Co nstr truc ucti tion on  Aperture geometry and thickness

– RA1A Verified TR 6/ 6/18 18/2 /200 005 5 Ok

– PA1A Verified TR 6/ 6/18 18/2 /200 005 5 Ok

LA1A Verified TR 6/ 6/18 18/2 /200 005 5 Ok

SU1A Verified TR 6/ 6/18 18/2 /200 005 5 Ok

LS1A Verified TR 6/ 6/18 18/2 /200 005 5 Ok

RI1A Verified TR 6/ 6/18 18/2 /200 005 5 Ok

LP1A Verified TR 6/ 6/18 18/2 /200 005 5 Ok

– RL1A   Verified TR 6/ 6/18 18/2 /200 005 5 Ok

Range compensator shape and thicknesses Patient and field identification Calibration and dosimetry Range Rang e ( g cm22) Modulation (g cm22) Calibration point (beam coordinates) Fraction RBE-weighted dose [cGy (RBE)] RBE factor Fraction absorbed dose (cGy) Isodose normalization (%) Output factor (cGy MU21 at STP)

Ok

Ok

Ok

Ok

Ok

Ok

Ok

Ok

Ok

Ok

Ok

Ok

Ok

Ok

Ok

Ok

HK 6/1 6/18/2 8/2005 005

HK 6/1 6/18/2 8/2005 005

HK 6/1 6/18/ 8/200 2005 5

HK 6/1 6/18/2 8/200 005 5

HK 6/1 6/18/2 8/2005 005

HK 6/1 6/18/2 8/2005 005

HK 6/1 6/18/2 8/2005 005

HK 6/1 6/18/ 8/200 2005 5

13.4 6.7 0,0,0

15.7 5.4 0,0,0

11.5 5.4 0,0,0

14.5 5.4 0,0,0

12.7 4.8 0,0,0

9.9 3.4 0,0,0

15.6 8.3 0,0,0

11.0 5.4 0,0,0

100

100

100

100

200

200

200

200

1.1 90.9

1.1 90.9

1.1 90.9

1.1 90.9

1.1 181.8

1.1 181.8

1.1 181.8

1.1 181.8

100 1.053

100 1.227

100 1.184

100 1.181

100 1.340

100 1.134

100 1.064

100 1.207

Monitor units (MU at STP) Equipm Equ ipment ent set settin tings gs Treatment room Range modulator/  scattering system ID Snout position (cm)

86.3

74.1

76.8

77.0

139.8

160.3

171.0

150.7

HK 6/1 6/18/2 8/2005 005 Gantry 1  A5

HK 6/1 6/18/2 8/2005 005 Gantry 1 A6

HK 6/1 6/18/2 8/2005 005 Gantry 1 A4

HK 6/1 6/18/2 8/200 005 5 Gantry 1 A5

HK 6/1 6/18/2 8/2005 005 Gantry 1 A5

HK 6/1 6/18/2 8/2005 005 Gantry 1 A4

HK 6/1 6/18/2 8/2005 005 Gantry 1 A6

HK 6/1 6/18/2 8/2005 005 Gantry 1 A5

16.3

17.5

13.7

15.4

13.7

15.2

16.6

20.4

188  

Journal of the ICRU Vol 7 No 2 (2007) Report 78 Oxford University Press

doi:10.1093/jicru/ndm034

REFERENCES

 AAPM (1980) American Associatio Association n of Physicis Physicists ts in Protocol for Neutr Neutron on Beam Medicine: Task Group 18.   Protocol  Dosimetry, AAPM Report No. 7 (American Association of Physicists in Medicine: New York).  AAPM (1983) American Associatio Association n of Physicis Physicists ts in Medicine. ‘A protocol for the determination of absorbed dose dos e fr from om hig high-e h-ener nergy gy pho photo ton n and el elect ectro ron n bea beams. ms.’’

 Radiation Dosimetry. IAE IAEA-C A-CN-9 N-96 6 (In (Inte terna rnatio tional nal  Atomic Energy Agency Agency,, Vienna).  Andreo, P., Nahum, A. and Brahme, A. (1986) ‘Chamber-dependent ‘Chamber-depen dent wall correction factors in dosimetry.’  Phys. Med. Biol .   31, 11 1189– 89– 11 1199. 99.  Archambeau,  Archambea u, J. O., Bennett, G. W., Levine, G. S., Cow Cowen, en, R. and Akanuma, A. (1974) ‘Proton radiation therapy.’

 Med. Phys.  10 , 741– 771 771..  AAPM (1986) American Associatio Association n of Physicis Physicists ts in Protoc ocol ol fo forr He Heav avy y Medi Me dici cine ne:: Tas ask k Gr Grou oup p 20 20..   Prot Charge Cha rged-P d-Part articl iclee The Thera rapy py Bea Beam m Dos Dosime imetry try, AA AAPM PM Report Rep ort No. 16 (Ame (American rican Associatio Association n of Phys Physicis icists ts in Medicine: New York).  Agosteo, S., Biratt Birattari, ari, C., Caravagg Caravaggio, io, M., Silari, M. and Tosi, G. (1998) ‘Secondary neutron and photon dose in proton therapy.’  Radiother. Oncol.  48 , 293– 305 305..  AJCC (2002) American Joint Committee on Cancer Cancer.. In Cancer Staging Manual, 6th edn, Greene, F. L., Page, D. L., Fleming, I. D., Fritz, A. G., Balch, C. M., Haller, D. G. and Morrow, M. Eds. (Springer: New York).  Almond, P. R. and Smathers Smathers,, J. B. (1977) ‘Physics inter int ercom compar pariso isons ns for neu neutr tron on ra radia diatio tion n the thera rapy py.’ .’  Int. J. Radiat. Oncol. Biol. Phys.  3 , 169– 176 176..  Amaldi,  Amald i, U., Ardui Arduini, ni, G., Cambr Cambrai, ai, R., Camp Campi, i, D., Ger Gerardi, ardi, F., Gra Gramat matica, ica, R., Leo Leone, ne, R., Manfr Manfredi, edi, G., Ninis Ninis,, M., Petrucci, G., Rossi, S., Sangaletti, L., Silari Sil ari,, M., Tos osi, i, G. G.,, Vec ecchi chi,, L. and Weis eiss, s, M. (19 (1994) 94) ‘A hosp hospital-b ital-based ased hadr hadronth ontherap erapy y faci facility lity for Ital Italy y,’ pp. 329–339 in   Hadrontherapy in Oncology, Amaldi, U. and Larsson, B. Eds (Elsevier Science B.V.: Amsterdam).

 Radiology  110 , 445– 457 457.. Introduction to Radiol Radiological ogical Physics  Attix, F. H. (1986)   Introduction and Radiation Dosimetry  (John Wiley: New York).  Austin-Seymour  Austin-Se ymour,, M., Kalet, I., McDonald, J., Krom Kr omho hout ut-S -Sch chir iro, o, S. S.,, Ja Jack cky y, J. J.,, Hu Humm mmel el,, S. an and d Ungerr, J. (1995 Unge (1995)) ‘Thr ‘Three ee dimen dimension sional al plan planning ning targ target et  volumes: A model and a softwar software e tool.’   Int. J. Radiat. Oncol. Biol. Phys.,  33 , 1073–1080. Bacher Bac her,, R., Bla Blattman ttmann, n, H., Boeh Boehringe ringerr, T., Cor Coray ay,, A., Egger Egg er,, E., Pe Pedr droni oni,, E., Phi Philli llips, ps, M. and Sch Scheib eib,, S. (198 (1 989) 9) ‘D ‘Dev evel elop opme ment nt an and d fir first st re resu sult ltss of di disc scre rete te dyna dy nami micc sp spot ot scann scannin ing g wi with th pr prot oton ons, s,’’ pp pp.. 9– 12 in  Proc. Int. Heavy Particle Therapy Workshop (PTCOG/  EORTC/ECNEU), PSI Report Report No No.. 69 (P (Paul aul Scherre Scherrerr Institute, Villigen, Switzerlan Switzerland. d. Balter, J. M., Ten Haken, R. K. and Lamm, K. L. (1996) ‘Tre ‘T rea atm tmen entt se setu tup p ver erifi ifica cati tion on,’ ,’ pp pp.. 47 471–493 1–493 in Teleth eletherap erapy: y: Pres Present ent and Futur Futuree, Ma Mack ckie ie,, T. R. an and d Palt Pa lta, a, J. R. Ed Eds. s. (A (Adv dvan ance ced d Me Medi dica call Pu Publ blis ishi hing ng:: Madison, WI). Bartolo Bar tolotta tta,, A., Indo Indovin vina, a, P. L., Ono Onori, ri, S. and Ros Rosat ati, i, A. (1984 (19 84)) ‘D ‘Dos osim imet etry ry fo forr Co Coba balt lt-60 -60 ga gamm mma a ra rays ys wi with th alanine.’  Radiat. Prot. Dosim.  9 , 277–281.

 Amerio, S., Bori  Amerio, Boriano, ano, A., Bour Bourhale haleb, b, F., Cirio Cirio,, R., Don Donetti etti,, M., Fidanzio, Fidanzio, A., Garelli, Garelli, E., Gior Giordanen danengo, go, S., Mado Madon, n, E., Marchetto, F., Nastasi, U., Peroni, C., Piermattei, A., Sanz San z Fr Frei eire re,, C. J., Sar Sardo, do, A. and Tre revis visio iol, l, E. (20 (2004) 04) ‘Dosim ‘Do simet etric ric cha chara racte cteriz rizat ation ion of a lar large ge ar area ea pix pixelelMed. Ph Phy ys.   31, segme se gmente nted d io ioniz nizat ation ion cha chambe mberr.’   Med 414–420.  Ando, K., Koike, S., Kaw Kawachi, achi, K., Hiraoka, T., Ohara, H., Yokota, M., Inada, T., Hirokawa, Y., Sato, S. and Eguchi, K. (1985) ‘Relative biological effectiveness of the therapeutic proton beams at NIRS and Tsukuba Universi Univ ersity ty.’ .’ Nippo Nippon n Igaku Hosh Hoshasen asen Gakk Gakkai ai Zass Zasshi, hi, ( Nippon Acta Radiologica)  45 , 531– 535 535..  Andreo, P. P. (1992) ‘Absorbed dose beam quality factors for the dos dosime imetry try of hig high-e h-ene nergy rgy pho photon ton bea beams. ms.’’   Phys.  Med. Biol.  37 , 2189–2211.  Andreo, P. (2002) ‘Role of the IAEA codes of practice in the radiation dosimetry chain,’ pp. 21–33 in  Proc. Int.  Symp. on Standards and Codes of Practice in Medical

Bartol Bart olot otta ta,, A. A.,, On Onor ori, i, S. an and d Pa Pant ntal alon oni, i, M. (1 (199 990) 0) ‘Alani ‘Al anine ne EPR dos dosime imetry try as a tr trav aveli eling ng dos dosime imetri tricc Radiat. Phys Phys.. system sys tem for inte intercom rcomparis parison on purpo purposes. ses.’’   Radiat. Chem.  35 , 708– 712 712.. Bartolotta, A., Fattibene, P., Onori, S., Pantaloni, M. and Pete tettti ti,, E. (1 (199 993) 3) ‘So Sour urcces of un unce cert rtai ain nty in Appl. Rad Radiat iat.. Iso Isott. therapy ther apy leve levell alan alanine ine dosi dosimetr metry y.’   Appl. 44, 13– 17 17.. Baumert, Baum ert, B. G., Lomax, A. J., Miltchev Miltchev, V. and Davis, J. B. (20 (2001) 01) ‘A com compar pariso ison n of dos dose e dis distri tribut bution ionss of  proton pro ton and pho photon ton bea beams ms in st ster ereo eota tacti cticc con confor formal mal Radiat iat.. Onc Oncol. ol. radiothe radi otherapy rapy of bra brain in lesi lesions. ons.’’   Int. J. Rad  Biol. Phys.  29 , 1439–1449. Baumert, B. G., Norton, I. A., Lomax, A. J. and Davis, J. B. (200 (2004) 4) ‘Dos ‘Dose e confo conforma rmation tion of inte intensit nsity-mod y-modula ulated ted ste terreo eota tact ctic ic ph phot oton on be beam ams, s, pr prot oton on be beam ams, s, an and d intensity inte nsity-mod -modulat ulated ed prot proton on beam beamss for intr intracr acrania aniall Int. J. Ra Radi dia at. On Onco col. l. Bio iol. l. Ph Phy ys.   60, lesions.’   Int. 1314–1324.

# International Commission on Radiation Units and Measurements 2007

 

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  Beck Beckham, ham, W. W. A., Keall, P. J. and Sieb Siebers, ers, J. V. (2002) (2002) ‘A  fluence flue nce-co -conv nvolu olutio tion n met method hod to cal calcul culat ate e ra radia diatio tion n

Blosser Bloss er,, H. G. G.,, Ge Gelbk lbke, e, G. K. K.,, La Lawt wton, on, D. D.,, Ma Marti rti,, F.,  Vincent,, J., York, R. C. and Zeller  Vincent Zeller,, A. (1993)  Proposal

therapy dose dist therapy distribut ributions ions that inco incorpor rporate ate rand random om set-up error.’  Phys. Med. Biol.  47 , 3465–3473. Beeckm Bee ckman, an, W., Jon Jongen gen,, Y., Lai Laisne sne,, A. and Lan Lannoy noye, e, G. (1991) (199 1) ‘Pr ‘Preli elimina minary ry des design ign of a re reduce duced-c d-cos ostt pro proton ton therapy facility using a compact high field isochronous Instrum. Methods Phys. Res.   B56/57, cyclotron.’   Nucl. Instrum. 1201–1204. Bell Be lli, i, M. M.,, Ce Cera ra,, F., Ch Cher erub ubin ini, i, R. R.,, Ha Haqu que, e, A. M. I. I.,, Ianzin Ian zini, i, F., Mo Mosch schini ini,, G. G.,, Sap Sapor ora, a, O. O.,, Sim Simone one,, G., Taboc abocchini chini,, M. A. and Tiv Tivero ero,, P. (1993 (1993)) ‘Ina ‘Inactiva ctivation tion and mut mutat ation ion ind induct uction ion in V79 ce cells lls by lo low w ene energy rgy protons: re-evaluation of the results at the LNL facility.’  Int. J. Radiat. Biol.  63 , 331– 337 337.. Bett Be tteg ega, a, D. D.,, Ca Calz lzol olar ari, i, P., Ch Chau auve vel, l, P., Co Cour urdi di,, A. A.,, Herault, Her ault, J., Iborr Iborra, a, N., Marc Marchesin hesini, i, R., Massa Massariell riello, o, P.,

to Cons Construct truct a 250 MeV Super Supercondu conducting cting Isoch Isochronou ronouss Cyclotron to Serve as an Advanced Cancer Treatment  Facility and as a Manufacturing Prototype for Commerc Comm ercial ial Prod Production uctionss of such Cyclot Cyclotrons rons, Re Repor portt MSUCL MS UCL-87 -874 4 (N (Nat ation ional al Sup Super erco condu nducti cting ng Cyc Cyclot lotro ron n Laboratory Labora tory,, Michigan State University, University, East Lansing, MI). Boag, Boa g, J. W. (19 (1966) 66) ‘Ioniza ‘Ionizatio tion n Cha Chambe mbers, rs,’’ pp. 1 –72 in  Radiation Dosimetry, Vol. II, Attix, F. H. and Roesch, W. C. Eds. (Academic Press: New York). Boag, J. W. and Wilson, T. (1952) ‘The saturation curve at high ionization intensity.’  Br. J. Appl. Phys.  3 , 222–229. Bolsi, A., Fogliata, A. and Cozzi, L. (2003) ‘Radiotherapy of small intracranial tumours with different advanced technique tech niquess usin using g phot photon on and pro proton ton beam beams: s: a tre treatat-

Poli,, G. L. an Poli and d Tal allo lone ne,, L. (2 (200 000) 0) ‘R ‘Radi adiob obio iolo logic gical al studies on the 65 MeV therapeutic proton beam at Nice Radia iat. t. Bi Biol. ol.   76, using usi ng huma human n tum tumour our cel cells.’ ls.’   Int. J. Rad 1297–1303. Bichsel, H. (1995) ‘Calculated Bragg curves for ionization Med. Ph Phys ys.   22, cham ch ambe bers rs of di diff ffer eren entt sh shap apes es.’ .’   Med. 1721–1726. Bilski, P., Budzanow Budzanowski, ski, M., Hoffmann, W., Molokano Molokanov v,  A., Olko, P. and Waligo´ rski, M. P. R. (1997) ‘Inve ‘In vest stiga igatio tion n of effi efficie ciency ncy of the thermo rmolum lumine inesce scence nce Radiat. Prot. detector dete ctorss for part particle icle ther therapy apy beam beams.’ s.’   Radiat.  Dosim.  70 , 501– 504 504.. Binns, Bin ns, P. J. and Ho Hough ugh,, J. H. (19 (1997) 97) ‘Se ‘Secon condar dary y dos dose e exposur exp osures es durin during g 200 MeV prot proton on ther therapy apy.’ .’   Radiat.  Prot. Dosim .  70 , 441– 444 444..

ment planning study study.’ .’  Radiother. Oncol.  68 , 1–14. Bonnett, D. E. (1993 Bonnett, (1993)) ‘Curr ‘Current ent dev develop elopment mentss in pro proton ton therapy: A review.’  Phys. Med. Biol.  38 , 1371–1392. Boon, S. N. (1998)  Dosimetry and quality control of scanning proton beams.  Doctoral Thesis, Mathematics and Natura Nat urall Scie Sciences nces (Rijk (Rijksuniv sunivers ersteit teit Gro Groninge ningen, n, The Netherlands). Boon, S. N., van Luijk Luijk,, P., Schippers, J. M., Meer Meertens, tens, H., Denis, J. M., Vynckier, S., Medin, J. and Grusell, E. (1998) ‘Fast 2D phantom dosimetry for scanning proton beams.’  Med.. Phys. 25, 464–475.  Med Boon Bo on,, S. N. N.,, va van n Lu Luij ijk, k, P., Bo¨ hrin hringer ger,, T., Cor Coray ay,, A., Lomax, Loma x, A., Ped Pedron roni, i, E., Scha Schaffner ffner,, B. and Schi Schippers ppers,, J. M. (2000) ‘Performance of a fluorescent screen and CCD came camera ra as a twotwo-dime dimensio nsional nal dosi dosimetr metry y sy syste stem m

Binns, P. P. J., Hough, J. H., Jones, D. T. L. and Schreuder Schreuder,,  A. N. (1993) ‘Microdo ‘Microdose se Spectra for Protons of  Annual ual Re Repor port, t, NA NAC/A C/AR/ R/ Differe Diff erent nt LET LET.’ .’ In   NAC Ann 93–01   (Nation (National al Acce Acceler lerato atorr Cent Centre: re: Fau Faure, re, Sout South h  Africa 88).

Phys. s.   27, for dynam dynamic ic tre treatme atment nt tech technique niques.’ s.’   Med. Phy 2198–2208. Bort Bo rtfe feld ld,, T. (200 (2 003) 3) ‘P ‘Phy hysi sica call opt ptim imiiza zati tio on. n.’’ In  Intensity-Modulated Radiation Therapy: The State of  the Art, Palta, J. R. and Mackie, T. R. Eds. (Medical

BIPM (1981) Bureau International des Poids et Mesures.  Recommendation R(I)-1, page R(I) 15 in BIPM Com. ´ ta Con ons. s. E talon lonss Me Mes. s. Ra Ray y. Io Ioni nisa sants nts   (Sec (Secti tion on I) 6, (Offilib, Paris, France). BIPM (2006) Bureau International des Poids et Mesures. Thee In Th Inte tern rnat ation ional al Sy Syst stem em of Un Units its (SI (SI)), 8th edn dn.. (BIPM, Pavillon de Breteuil, Se` vres Cedex, France). Bjo¨ rk, P., Kno¨o¨ s, T. and Nilsson, P. P. (2000) ‘Compar ‘Comparative ative dosimetry of diode and diamond detectors in electron beams for intraoperative radiation therapy.’  Med. Phys.

Physics Publishing: Madison, WI). Bortfeld, Bortfe ld, T., Jok Jokiva ivarsi rsi,, K. K.,, Go Goite itein, in, M., Ku Kung, ng, J. and Jiang, S. B. (2002) ‘Effects of intra-fraction motion on IMRT IMR T dos dose e del deliv ivery ery:: Sta Statis tistic tical al ana analy lysis sis and sim simuulation.’  Phys. Med. Biol .  47 , 2203–2220. Bortfeld, T., Trofimov, A., Shipley, W. U., Chan, W. F. A.,  Adams, J., DeLane DeLaney y, T. F., Kooy Kooy,, H., Ku ¨ fer fer,, K.-H. and Monz, Mo nz, M. (20 (2004) 04) ‘Ad ‘Advan vance cess in com compar parat ativ ive e pr proto oton n therapy treatment planning at NPTC in Boston.’  Proc.  ESTRO  23 , S122.

27, 2580–2588. ¨ w, B., Montelius, Blomquist, Blomquis t, E., Russ Russell, ell, K. R., Sten Stenerlo erlo  A., Grusell, E. and Carlsson, J. (1993) ‘Relat ‘Relative ive biological eff effect ectiv ivene eness ss of int interm ermedi ediat ate e ene energy rgy pr proto otons. ns.   60 Comparisons Compariso ns with Co gamm gamma-ra a-radiat diation ion usin using g two cell lines.’  Radiother. Oncol.  28 , 44–51. Blosse Blo sserr, H., Joh Johnso nson, n, D., La Lawto wton, n, D., Ro Ronni nninge ngen, n, R., Burlei Bur leigh, gh, R. and Go Gotts ttscha chalk, lk, B. (19 (1991) 91)   Preliminary  Design Study Exploring Building Features Required  for a Proton Therapy Facility for Ontario Cancer  Institute, Report MSUCL 76a (National Superconducting Supercon ducting Cyclotron Laboratory, Laboratory, Michigan State University Univers ity,, East Lansing, MI).

Value uess of g for ph phot oton on en ener ergi gies es. Boutillon, Boutillo n, M. (1987 (1987))   Val CCEMRI, Offilib Report (I) 85–18 (Comite´   Consultatif  ´ tal pour po ur le less E talons ons de Me Meas asur ures es des Ra Rayo yonne nneme ments nts Ionisants, Offilib, Paris, France). Boutillon, M. (1989) ‘Gap correction for the calorimetric measurement of absorbed dose in graphite with a   60Co beam.’ Phys. Med. Biol .  34 , 1809–1821. Bout Bo util illo lon, n, M. an and d Per errroc oche he-R -Rou oux, x, A. M. (1 (198 987) 7) ‘Re-evaluation of the   W  value for electrons in dry air.’  Phys. Med. Biol.  32 , 213– 219 219.. Boyd, T. J., Crandall, K. R., Hamm, R. W., Hansborough, L. D., Hoeberling, R. F., Jameson, R. A., Knapp, E. A., Muelle Mue llerr, D. W., Po Potte tterr, J. M., Sto Stoke kes, s, R. H., Sto Stovall vall,,

190  

 REFERENCES J. E., Sturgess, R. G., Swenson, Swenson, D. A., Taller allerico, ico, P. P. J.,

‘Strip ‘St rip ion ioniza izatio tion n cha chambe mbers rs as 3-D det detect ector or for had hadro ron n

Wangler, T. P. and Wilkerson, L. C. (1982) ‘The PIGMI Pion and linear line ar acceler accelerat ator or technology technology,’ pp. 3– 22 in   Pion  Heavy Ion Radiother Radiotherapy: apy: Pre-Cli Pre-Clinical nical and Clinical  Studies, Sk Skar arsg sgar ard, d, L. D. Ed Ed.. (E (Els lsev evie ierr Sc Scie ienc nce e Publishing Co., Inc.: New York). Bradshaw, W. W., Cadena, D. G., Jr., Crawford, G. W. and Spetzler, H. A.W. (1962) ‘The use of alanine as a solid dosimeter.’  Radiat. Res.  17 , 11–21. Brahme, A. (1984) ‘Dosimetric precision requirements in radiation therapy.’  Acta Radiol. Oncol.  23 , 379– 391 391.. ¨ ll Brah Br ahme me,, A. A.,, Ka llma man, n, P. an and d Lind nd,, B. K. (1 (19 989 89)) ‘Optimiza ‘Opt imization tion of pro proton ton and hea heavy vy ion therapy using an adaptive inversion algorithm.’  Radiother. Oncol.  15 , 189–197. Bred Br ede, e, H. J. J.,, He Heck cker er,, O. an and d Se Selb lbac ach, h, H. J. (1 (199 994) 4)

Nucl.. In Inst stru rum. m. Me Metho thods ds Phy Phys. s. Re Res. s.   A389, therapy.’   Nucl 499–512. Brusas Bru sasco, co, C., Voss oss,, B., Sch Schard ardt, t, D., Kr Kra a¨me merr, M. an and d Kraft, G. (2000) ‘A dosimetry system for fast measurementt of 3D dep men depthth-dos dose e pr profil ofiles es in cha charge rged-p d-part articl icle e tumor tum or the thera rapy py wi with th sca scanni nning ng te techn chniqu iques. es.’’   Nucl.  Instrum. Methods Phys. Res.   B168, 578– 592 592.. Bucc Bu ccio ioli lini ni,, M. M.,, Cu Cutt tton one, e, G. G.,, Eg Egge gerr, E. E.,, Gu Guas asti ti,, A. A.,, Mazzocchi, S., Raffaele, L. and Sabini, M. G. (1999) ‘A  comparison of the TLD-100 response to   60Co beam and 62 MeV protons.’  Phys. Med.  XV , 71– 77 77.. Buccio Buc ciolin lini, i, M. M.,, Cut Cutto tone, ne, G., Gua Guast sti, i, A., Lo Nig Nigro ro,, S., Mazzocchi, S., Raffaele, L., Sabini, M. G., Kacperek, A. and Egger, E. (2000) ‘Check on the use of thermoluminesc ne scen ence ce de dete tect ctor orss fo forr pr prot oton on do dose se di disstr trib ibut utio ion n

‘Measurement ‘Measure ment princ principles iples and exp experim erimenta entall setu setup p for determin dete rmining ing the heat defe defect ct of wa water ter for highhigh-ener energy gy protons, deuterons and alpha particles.’ In   Proc. NPL Calorimetry Workshop   (National Physical Labora Laboratory tory,, Teddington, UK). Bred Br ede, e, H. J. J.,, He Heck cker er,, O. an and d Ho Holl llna nage gel, l, R. (1 (199 997) 7) ‘Measure ‘Mea surement ment of the heat defe defect ct in wa water ter and A-150 plastic plas tic for highhigh-ener energy gy prot protons, ons, deute deuteron rons, s, and alpha particles.’ Radiat. Prot. Dosim.  70 , 505– 508 508.. Brede, H. J., Heinemann, E., Binns, P. P. J., Langen, K. M., Jones, Jon es, D. T. L. and Sch Schre reude uderr, A. N. (19 (1999) 99) ‘W ‘Wat ater er calorime calo rimetry try and ioni ionizat zation ion chamb chamber er dosi dosimetr metry y in the prot pr oton on thera therapy py be beam am,’ ,’ pp pp.. 12 123–124 3–124 in   National  Accelerator Annual Report, March 1999   (National  Accelerator  Accelera tor Centre, Faure Faure,, South Africa).

measurements.’ Phys. Med.  XVI , 117– 120 120.. Bucciolini, Bucciolin i, M., Banc Bancii Buon Buonamic amici, i, F., Mazz Mazzocch occhi, i, S., De  Angelis, C., Onori, S. and Cirrone, G. A. P. (2003) ‘Dia ‘D iamo mond nd de dete tect ctor or ve vers rsus us si sili lico con n di diod ode e an and d io ion n chamber in photon beams of different energy and field size.’ Med. Phys.  30 , 2149–2154. Bush, D. A., Slater, J. D., Shin, B. B., Cheek, G., Miller, D. W. and Sla Slater ter,, J. M. (20 (2004a 04a)) ‘Hy ‘Hypof pofra racti ctiona onated ted proton pro ton bea beam m ra radio diothe thera rapy py for st stage age I lun lung g can cancer cer.’ .’ Chest  126 , 1198–1203. Bush, D. A., Hillebrand, D. J., Slater, J. M. and Slater, J. D. (20 (2004b 04b)) ‘Hi ‘Highgh-dos dose e pr proto oton n bea beam m ra radio diothe thera rapy py of  hepat hep atoce ocellu llular lar car carcin cinom oma: a: pr preli elimin minary ary re resul sults ts of a phas ph ase e II tr tria iall.’   Gastroenterol.   127   (Sup (Suppl pl.. 1) 1),, S189–S193.

Brede Br ede,, H. J., Hec Hecke kerr, O. and Ho Holln llnage agel, l, R. (20 (2000) 00) ‘An

Buts Bu tson on,, M. J. J.,, Ch Cheu eung ng,, T. an and d Yu, P. K. N. (2 (200 001) 1)

absorbed dose to water calorimeter for collimated radiation fields.’  Nucl. Instrum. Methods Phys. Res .   A455, 721–732. Brede, H. J., Greif, K-D., Hecker, O., Heeg, P., Heese, J., Jone Jo nes, s, D. T. L. L.,, Kl Klug uge, e, H. an and d Sc Scha hard rdt, t, D. (2 (200 006) 6) ‘Absorbed dose to water determination with ionization chambe cha mberr dos dosime imetry try and cal calori orimet metry ry in re rest stric ricte ted d neutr neu tron, on, pho photo ton, n, pr prot oton on and hea heavyvy-ion ion ra radia diatio tion n fields.’ Phys. Med. Biol .  51 , 3667–3682. Therapy and Breue Br euerr, H. and Smit, B. J. (20 (2000) 00)   Proton Therapy  Radiosurgery   (Springer-Verlag: Berlin). Broe Br oerse rse,, J. J., Co Coppo ppola, la, M. and Bur Burger ger,, G. (19 (1978) 78)   A  European Neutron Dosimetry Intercomparison Project (ENDIP (EN DIP): ): Re Resul sults ts and Eva Evalua luation tion. EUR 6004 EN

‘Radio ‘Ra diochr chromi omicc film dos dosime imetry try in wa water ter pha phanto ntoms. ms.’’  Phys. Med. Biol.  46 , N27–N31. Cambri Cam bria, a, R., He´ ra rault ult,, J., Br Brass assart art,, N. N.,, Sil Silari ari,, M. and Chauvel, P. (1997) ‘Proton beam dosimetry: A comparison so n be betw twee een n th the e Fa Fara rada day y cu cup p an and d an io ioni niza zati tion on chamber.’  Phys. Med. Biol.  42 , 11 1185– 85– 11 1196. 96. Carlsson, C. A. and Carlsson, G. A. (1970) ‘Proton dosimetry et ry:: me meas asur urem emen entt of de dept pth h do dose sess fr from om 18 1855-Me MeV  V  protons by means of thermolumines thermoluminescent cent LiF.’ LiF.’   Radiat.  Res.  42 , 207– 219 219.. Case, J. A., von Goeler, F., Rabin, M. S. Z. and Wagner, M. (1994) ‘Effects of detector geometry and sensitivity Therapy apy on th the e sh shap ape e of th the e Br Brag agg g pe peak ak.’ .’   Proton Ther Co-Operative Group Meeting,  Cambridge, MA .

(Comm (Co mmiiss ssio ion n of the th e Eurrope Eu pean an Commu mun nit itiies es:: Luxembourg). Broerse, J. J., Mijnheer, B. J., Eenma, J. and Wootton, P. (1979)) ‘Dos (1979 ‘Dosimet imetry ry inte interco rcompar mparison isonss and prot protocol ocolss for therapeuticc applications of fast therapeuti fast neutron beams,’ beams,’ in pp. High LET Radia iati tio ons in Clinical 117–123   Hi  Radiotherapy, Bar Barend endsen sen,, G. W., Br Broe oerse rse,, J. J. and Breur, K. Eds. (Pergamon Press: Oxford). Broerse, J. J., Mijnheer, B. J. and Williams, J. R. (1981) ‘European protocol for neutron dosimetry for external beam ther therapy apy.. Eur Europea opean n Neut Neutron ron Dos Dosimet imetry ry Gro Group up (ECNEU).’ Br. J. Radiol.  54 , 882– 898 898.. Brusasco, C., Cattai, A., Cirio, R., Dellacasa, G., Donetti, M., Isoardi, P., Marchet Marchetto, to, F., Per Peroni, oni, C, Ro Rolan lando, do, V., V., Rusp Ru spa, a, M. M.,, So Sola lano no,, A. an and d Za Zamb mber erna nard rdi, i, C. (1 (199 997) 7)

Castro, J. R., Quivey, J. M., Lyman, J. T., Chen, G. T. Y., Philli Phi llips, ps, T. T. L., Tobia Tobias, s, C. A. and Alpen, Alpen, E. L. (19 (1980) 80) ‘Curr ‘Cu rrent ent st stat atus us of cli clinic nical al par partic ticle le ra radio diothe thera rapy py at Lawrence Berkeley Laboratory.’  Cancer  46 , 633– 641 641.. Caumes, Caum es, J., Os Ostro trowsk wski, i, A., Stei Steinsch nschaden aden,, K., Manc Mancaux, aux, M.,, Ca M. Canc nce, e, M. M.,, Si Sime meon on,, J. P., Sa Saba batt ttie ierr, R. an and d Brete Br eteau, au, N. (19 (1984) 84) ‘D ‘Dire irect ct cal calibr ibrat ation ion of ion ioniza izatio tion n chambers with a TE calorimeter at the Orleans cyclotron neutron facility.’  Strahlenther.  160 , 127– 128 128.. ´ talons de CCEMRI (1977) Comite´   Consultatif pour les E Mesur Me sures es des Ra Rayo yonne nnemen ments ts Ion Ionisa isants nts (Se (Secti ction on I). ´ . CC Correction Corr ection d’Hum d’Humidite idite´  CCEM EMRI RI R( R(I) I)-3 -30 0 (B (Bur urea eau u ` International des Poids et Mesures, Se vres, France). ´ talons de CCEMRI (1985) Comite´   Consultatif pour les E Mesur Me sures es des Ra Rayo yonne nnemen ments ts Ion Ionisa isants nts (Se (Secti ction on I).

191  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY   Effect of a change of stopping-power values on the W  values recommended by ICRU for electrons in dry air. CCEMRI (I)/85–8 (Bureau International des Poids et Mesures, Se` vres, France). Cella, Cel la, L., Loma Lomax, x, A. and Mir Miralbe albell, ll, R. (2001 (2001)) ‘Po ‘Potent tential ial role rol e of inte intensity nsity modulated modulated pro proton ton beam beamss in pro prosta state te cancer radiotherapy.’  Int. J. Radiat. Oncol. Biol. Phys. 49, 217– 223 223.. Censor Cen sor,, Y. (2003 (2003)) ‘Ma ‘Mathem thematic atical al optim optimizat ization ion for the invers inv erse e pr probl oblem em of int intens ensity ity-mo -modul dulat ated ed ra radia diatio tion n Intensity Modul Modulated ated Radia Radiation tion Ther Therapy: apy: therapy.’ In   Intensity Thee St Th Stat atee of th thee Ar Artt, Pa Palta lta,, J. R. and Mackie, Mackie, T. T. R. Eds. (Medical Physics Publishing: Madison, WI). Chadwick, M. B., Jones, D. T. L., Arendse, G. J., Cowley,  A. A., Richter Richter,, W. A., Lawrie, J. J., Newman, R. T.,

Chu, W. W. T. (199 (1995a) 5a) ‘Ion beam beamss for cancer cancer tre treatm atment ent – a perspective.’   Nucl. Instrum. Methods Phys. Res.   B99, 835–838. Chu, W. T (1995b) ‘Radiation detectors’, pp. 234–245 in  Ion Beams in Tumor Therapy, Linz, U. Ed. (Chapman and Hall: Weinheim, Germany). Chu, W., McEvoy, M., Nyman, M., Renner, T., Gonzales, B., Sin Singh, gh, R. P. and Str Stradn adner er,, R. (19 (1985) 85) ‘W ‘Wobb obbler ler dosime dos imetry try for the bio biome medic dical al pr progr ogram am at the LBL BEVALAC.’  IEEE Trans. Nucl. Sci.   NS-32, 3324–3326. Chu,, W. T., Re Chu Renne nnerr, T. R. and Lud Ludew ewigt igt,, B. A. (19 (1989) 89) ‘Dynamic beam delivery for three-dimensional conforWorkshop on mal ther therapy apy,’ pp. 295– 295– 328 in   EULIMA Workshop the Potential Value Value of Light Ion Beam Therapy Therapy, Nice,  France, Cha Chauve uvel, l, P. and Wamb ambers ersie, ie, A. Eds Eds.. EU EUR R

Pilcher, J. V., Smit, F. D., Steyn, G. F., Koen, J. W. and Stander Stan der,, J. A. (1999 (1999)) ‘Nuc ‘Nuclear lear interactio interaction n cro cross ss sec  Nuc Nucl. l. Phys Phy s   tionss for prot tion proton on radi radiothe otherapy rapy.’ . A654, 1051c–1057c. Chandna, Chan dna, S., Dw Dwarak arakana anath, th, B. S., Khai Khaitan, tan, D., Lazar Mathew Ma thew, T. and Jain, V. (2002 (2002)) ‘Lo ‘Low-do w-dose se radi radiatio ation n hypersensitivity in human tumor cell lines: Effects of  cell-cel cell -celll cont contact act and nutr nutritio itional nal depri deprivati vation.’ on.’   Radiat.  Res.  157 , 516– 525 525.. Chapman, P. H., Ogilvy, C. S. and Butler, W. E. (1993) ‘A  new stereotactic alignment system for charged particle radiosur radi osurgery gery at the Harv Harvard ard Cyclo Cyclotro tron n Labo Laborat ratory ory,, Stereotactictic Radiosurgery, Bosto Bos ton,’ n,’ pp. 105–108 in   Stereotactictic

12165 EN (Commission of the European Communities: Luxembourg). Chu,, W. T., Lud Chu Ludew ewigt igt,, B. A. and Re Renne nnerr, T. R. (19 (1993) 93) ‘Instrumentation for treatment of cancer using proton Rev.. Sc Sci. i. In Insstr trum um.   64, and an d li ligh ghtt-io ion n be beam ams. s.’’   Rev 2055–2122. Ciesiels Cies ielski, ki, B. and Wi Wielop elopolsk olski, i, L. (1994 (1994)) ‘The effects of  dose and radiation quality on the shape and power satRadiat. Res. urat ur atio ion n of the EPR sig signal nal in ala alanin nine.’ e.’   Radiat. 140, 105–111. Cirio, R., Garelli, E., Schulte, R., Amerio, S., Boriano, A., Bour Bo urha halleb, F., Cout utrrakon, G., Donet etti ti,, M., Giordanen Gior danengo, go, S., Kos Koss, s, P., Mado Madon, n, E., Mar Marchet chetto, to, F.,

 Alexander, E., III, Loeffler  Alexander, Loeffler,, J. S. and Lunsford, L. D. Eds. (McGraw-Hill: New York).

Nastas Nast asi, i, U. U.,, Pero Peroni ni,, C., C., Sa Sant ntua uari ri,, D. D.,, Sard Sardo, o, A. A.,, Scie Sc ielz lzo, o, G. G.,, St Stas asi, i, M. and and Trev evis isio iol, l, E. (200 (2004) 4)

Char, D. H., Quivey, J. M., Castro, J. R., Kroll, S. and Philli Phi llips, ps, T. (19 (1993) 93) ‘He ‘Heliu lium m ion ionss ve versu rsuss iod iodine ine 125 brachytherapy in the management of uveal melanoma.  A prospect prospective, ive, randomize randomized, d, dynamically balanced trial.’ Ophthalmology  100 , 1547–1554. Charpa Cha rpak, k, G. (19 (1970) 70) ‘Ev ‘Evolu olutio tion n of the aut automa omatic tic spa spark rk chambers.’ Ann. Rev. Nucl. Sci .  20 , 195– 254 254.. Chauv Cha uvene enet, t, B., Bal Balte te`s, D. an and d De Dela laun unay ay, F. (1 (199 997) 7) ‘Comp ‘C ompari ariso son n of gr graph aphite ite-to -to-w -wat ater er abs absorb orbeded-dos dose e   60 transfers for Co photon beams using ionometry and Fricke dosimetry.’  Phys. Med. Biol.  42 , 2053–2063. Chen, G. T. (1988) ‘Dose volume histograms in treatment Int. J. Ra Radi dia at. On Onco col. l. Bi Biol ol.. Ph Phys ys.   14, planning.’   Int. 1319–1320. Chen, G. T., Singh, R. P., P., Castro, J. R., Lyman, J. T. and

‘Two-dimensional ‘Two-dimension al and quasi-thre quasi-three-dimensional e-dimensional dosimetry of hadron and photon beams with the Magic Cube Phys. Med. Biol. and the Pixe Pixell Ioniz Ionizati ation on Chamb Chamber er.’ .’   Phys. 49, 3713–3724. Cole, Co le, F., Liv Livdah dahl, l, P. V., Mil Mills, ls, F. and Teng eng,, L. (19 (1987) 87) ‘Design ‘Des ign and appli applicat cation ion of a pro proton ton ther therapy apy acc acceler eleraaProc.. 19 1987 87 IE IEEE EE Pa Part rtic icle le torr,’ pp to pp.. 19 1985–198 85–1987 7 in   Proc  Accelerator Conference, Lindstrom, E. R. and Taylor, L. S. Eds. (IEEE Press, Piscataway, NJ). Constable, I. J., Goitein, M., Koehler, A. M. and Schmidt, R. A. (1976 (1976)) ‘Smal ‘Small-fiel l-field d irra irradiat diation ion of monk monkey ey eyes with protons and photons.’  Radiat. Res.  65 , 304– 314 314.. Const Con stant antino inou, u, C., Har Harrin ringto gton, n, J. C. and DeWerd erd,, L. A. (199 (1 992) 2) ‘An el elec ectr tron on de dens nsit ity y ca cali libr brat atio ion n pha phant ntom om fo forr CT-based treatment planning computers.’  Med. Phys.,  19,

Quivey, J. M. (1979) ‘Treatment planning for heavy ion Int. J. Ra Radi diat at.. On Onco col. l. Bi Biol ol.. Ph Phys ys.   5, radiotherapy.’   Int. 1809–1819. Chen,, G. T. Y., Kung Chen Kung,, J. H. and Beaudette, Beaudette, K. P. (2004 (2004)) ‘Artifacts in computed tomography scanning of moving objects.’  Semin. Radiat. Oncol .,  14 , 19–26. Cheu Ch eung ng,, T., Bu Buts tson on,, M. J. an and d Yu, P. K. N. (2 (200 005) 5) ‘Post-irradiation colouration of GafChromic EBT radiochromic film.’  Phys. Med. Biol.  50 , N281–N285. Chiba, Chib a, T., Tokuu okuuye, ye, K., Mat Matsuzak suzaki, i, Y., Sugah Sugahara ara,, S., Chuganji, Y., Kagei, K., Shoda, J., Hata, M., Abei, M., Igaki, Iga ki, H., Tana anaka, ka, N. and Aki Akine ne,, Y. (20 (2005) 05) ‘Pr ‘Proto oton n beam ther therapy apy for hepa hepatoce tocellula llularr car carcinom cinoma: a: A ret retrorospective review of 162 patients.’   Clin. Cancer. Res.   11, 3799–3805.

325–327. Coray, A., Pedroni, E., Boehringer, T., Lin, S., Lomax, A. and Goit Goitein, ein, G. (200 (2002) 2) ‘Do ‘Dosime simetry try with the scan scanned ned proton pro ton beam on the Paul Scherrer Scherrer Ins Institut titute e gant gantry ry,’ ,’ Proc. Int Int.. Sym Symp. p. on Stan Standar dards ds and pp.. 295–30 pp 295–302 2 in   Proc. Codes Cod es of Pra Practi ctice ce in Med Medica icall Rad Radiat iation ion Dos Dosime imetry try. IAEA-CN-9 IAEACN-96 6 (Inte (Internat rnationa ionall Atom Atomic ic Ener Energy gy Agen Agency cy,,  Vienna.).  Vienna. ). Cour Co urant ant,, E. D. D.,, Liv Living ingst ston on,, M. S. and Sny Snyder der,, H. S. (1952)) ‘The strong-focu (1952 strong-focusing sing sync synchro hrotro tron– n– a new high energy accelerator.’  Phys. Rev.  88 , 11 1190– 90– 11 1196. 96. Courd Co urdi, i, A., Br Brass assart art,, N. N.,, He Hera rault ult,, J. and Cha Chauv uvel, el, P. (1994) (19 94) ‘Th ‘The e dep depthth-dep depen enden dentt ra radia diatio tion n re respo sponse nse of  human mela melanoma noma cells expo exposed sed to 65 MeV prot protons. ons.’’  Br.. J. Radiol.  67 , 800– 804  Br 804..

192  

 REFERENCES Coutrakon, G., Miller, D., Wong, J., Gere, L. and Binns, R. (1990) ‘Characterization of the Loma Linda proton beam bea m usi using ng a pla plast stic ic sci scinti ntilla llator tor and CCD cam camer era a  Med. Phys readout.’ .  17 , 543 (Abstract). Coutrakon, G., Miller, D., Kross, B. J., Anderson, D. F., DeLuc De Luca, a, P. M. Jr Jr.. and Siebers, Siebers, J. (19 (1991a 91a)) ‘A bea beam m intensity monitor for the Loma Linda cancer therapy proton accelerator.’  Med. Phys.  18 , 817–820. Coutr Co utrak akon on,, G., Bau Bauman man,, M. M.,, Les Lesyna yna,, D. D.,, Mi Mille llerr, D., Nusbaum, Nusb aum, J., Slat Slater er,, J., Joha Johannin nning, g, J., Mir Miranda, anda, J., DeLuca, P. M. Jr., Siebers, J. and Ludewigt, B. (1991b) ‘A pr prot otot otyp ype e be beam am de deli live very ry sy syst stem em fo forr th the e pr prot oton on medical medi cal acce acceler lerato atorr at Loma Linda Linda.’ .’   Med. Phys.   18, 1093–1099.

Particles es 32, Abs Abstra tracts cts of the 37t 37th h Pro Proton ton LABS.’ In   Particl LABS.’ Thera The rapy py CoCo-Ope Opera rativ tivee Gr Group oup Mee Meetin ting, g, Cap Capee Town own,,  South Africa. Deas De asy y, J. O. O.,, Cha Chao, o, K. S. C. and Markman Markman,, J. (20 (2001) 01) ‘Uncertainties in model-based outcome predictions for treatment planning.’  Int. J. Radiat. Oncol. Biol. Phys . 51, 1389–1399. Dela De lacr croi oix, x, S. S.,, Br Brid idie ierr, A. A.,, Ma Maza zal, l, A. A.,, Dau aurres es,, J. J.,, Ostrowsky Ostro wsky,, A., Nauray Nauraye, e, C., Kacper Kacperek, ek, A., Vynckier ynckier,, S., Brassard, N. and Habrand, J. L. (1997) ‘Proton dosimetry comparison involving ionometry and calorimetry.’  Int. J. Radiat. Oncol. Biol. Phys.  37 , 711–718. Dendale, Den dale, R., Lumb Lumbros roso-Le o-Le Rouic, L., Noe Noel, l, G., Feu Feuvre vret, t, L., Lev Levy y, C., De Dela lacr croix oix,, S., Me Meye yerr, A., Na Naur uray aye, e, C.,

Coutrakon, G., Hubbard, J., Johanning, J., Maudsley, G., Slaton, T. T. and Morton, P. (1994) ‘A performance study of the Loma Linda prot proton on medi medical cal acce acceler lerato atorr.’   Med.  Phys.  21 , 1691–1701. Coutrakon, G., Cortese, J., Ghebremedhin, A., Hubbard, J., Johanning, J., Koss, P., Maudsley, G., Slater, C. R. and Zucc Zuccare arelli, lli, C. (1997 (1997)) ‘Mic ‘Microd rodosim osimetry etry spec spectra tra of  the Loma Linda prot proton on beam and rel relativ ative e biol biologic ogical al effectiveness effectivene ss comparisons.’ Med. Phys.  24 , 1499–1506.

Mazal, A., Mamm Mazal, Mammar ar,, H., Gar Garcia, cia, P., D’He D’Hermie rmies, s, F., Frau, Fra u, E., Plan Plancher cher,, C., Asse Asselain lain,, B., Schli Schlienge engerr, P., Mazeron, J. J. and Desjardins, L. (2006) ‘Proton beam radio ra diothe thera rapy py for uv uveal eal mel melano anoma: ma: Re Resul sults ts of Cur Curie ie Inst In stit itut ut-O -Ors rsay ay Pr Prot oton on Th Ther erap apy y Ce Cent nter er (I (ICP CPO) O).’ .’  Int. J. Radiat. Oncol. Biol. Phys.  65 , 780– 787 787.. Dieckman Die ckmann, n, K., Geo Georg, rg, D., Bogn Bogner er,, J., Zehe Zehetma tmayer yer,, M., Petersch, B., Chorvat, M., Weitmann, H. and Potter, R. (200 (2 006) 6) ‘Opt ptiimi miz zin ing g lina li nacc-ba base sed d ste terreotacti ticc

Cox, R., Thacker, J., Goodhead, D. T. and Munson, R. J. (1977) ‘Mutation and inactivation of mammalian cells by various ionising radiations.’  Nature  267 , 425– 427 427.. Crook, Cro ok, J. M., Raymond, Raymond, Y., Salh Salhani, ani, D., Yang, H. and Esche Esc he,, B. (19 (1995) 95) ‘Pr ‘Pros osta tate te mot motion ion dur during ing st stand andard ard rad adio ioth ther erap apy y as as asse sess ssed ed by fid fiduc ucia iall ma mark rker ers. s.’’

radiothe radio thera rapy py of uv uveal eal mel melano anomas mas:: 7 ye years ars cli clinic nical al Int. J. Ra Radi diat at.. On Onco col. l. Bi Biol. ol. Ph Phys ys..   66, experience.’   Int. S47–S52. Dobler, B. and Bendl, R. (2002) ‘Precise modelling of the eye for proton therapy of intra-ocular tumours.’  Phys.  Med. Biol.  47 , 593– 613 613..

 Radiother.. Oncol.  37 , 35–42.  Radiother Cumming, J. B. (1963) ‘Monitor reactions for high energy proton beams.’  Ann. Rev. Nucl. Sci .  13 , 261– 286 286.. Cuttto Cu tone ne,, G., Raff ffa aele le,, L., Bar aro one Tong ngh hi, L., Rove Ro velli lli,, A., Sab Sabini ini,, M. G., Egg Egger er,, E., Ka Kacpe cpere rek, k, A., Brai Br ai,, M. M.,, Ba Bart rtol olot otta ta,, A. A.,, Ter eri, i, G. G.,, On Onor ori, i, S. an and d Fattibene, Fattib ene, P. (1999) ‘First dosimetry intercomp intercomparison arison Phys. Med.   XV , resul re sults ts for the CA CAT TANA Pr Proj oject ect.’ .’   Phys. 121–130. Daftari, Daf tari, I., Cas Castena tenadas, das, C., Pet Petti, ti, P. L., Singh Singh,, R. P. and  Verhey  V erhey, L. J. (1999) ‘An application of GafChr GafChromic omic MD-55 film for 67.5 MeV clinical proton beam dosimetry.’  Phys. Med. Biol.  44 , 2735–2745. Linearr Pr Prog ogra ramm mmin ing g an and d Dan anttzi zig g, G. B. (1 (196 963 3)   Linea  Extensions   (Princet (Princeton on Univ Universi ersity ty Pre Press, ss, Prin Princeto ceton, n,

Domen, S. R. (1980 Domen, (1980)) ‘Abso ‘Absorbed rbed dose water water calo calorime rimeter ter.’ .’  Med. Phys.  7 , 157– 159 159.. Domen, Dom en, S. R. (198 (1987) 7) ‘Adva ‘Advances nces in calo calorime rimetry try for radi radi-Dosime imetry try of  ation at ion dosim dosimet etry ry,’ ,’ pp. 245–320 in   The Dos  Ionizing Radiation, Vol. II , Kase, K. R., Bjarngard, B. E. and Att Attix, ix, F. H. Eds Eds.. (Ac (Acade ademic mic Pr Press ess:: Or Orlan lando, do, FL). Domen Do men,, S. R. (19 (1994) 94) ‘A se seale aled d wa water ter calorime calorimeter ter for Res. s. Na Natl. tl. Ins Inst. t. Stan Stand. d. measuring meas uring abso absorbed rbed dose dose.’ .’   J. Re Technol .  99 , 121– 141 141.. Domen Do men,, S. R. and Lam Lamper perti, ti, P. J. (19 (1974) 74) ‘A hea heat-l t-loss oss compensated calorimeter: Theory, design, and performance.’  J. Res. Natl. Bur. Stand.  78A , 595– 610 610.. Drzy Dr zyma mala la,, R. E. E.,, Mo Moha han, n, R. R.,, Br Brew ewsste terr, L. L.,, Ch Chu, u, J. J.,, Goit itei ein, n, M., Har arms ms,, W. an and d Urie ie,, M. (1 (199 991) 1)

NJ). Daures Daur es,, J., Cha Chauv uvene enet, t, B. and Os Ostr trow owski ski,, A. (19 (1994) 94) ‘State-of the-art of calorimetry at LPRI.’ In  Proc. NPL Calorimetry Workshop   (National Physical Labora Laboratory tory,, Teddington, UK). De Angelis, C., Ono Onori, ri, S., Pacilio, Pacilio, M., Cirrone, Cirrone, G. A. P., Cutt Cu tto one ne,, G., Ra Raff ffae aele le,, L., Buc ucccio ioli lini ni,, M. and Mazzocchi, S. (2002) ‘An investigation of the operating char ch arac acte teri rist stic icss of tw two o PT PTW W di diam amon ond d de dete tect ctor orss in photon and electron beams.’  Med. Phys.  29 , 248– 254 254.. Debus, J., Schulz-Ertner, D., Schad, L., Essig, M., Rhein, B., Thil Thillmann lmann,, C. O. and Wanne annenma nmacher cher,, M. (2000 (2000)) ‘Stereotactic ‘Stereo tactic fractionated radiotherapy for chordomas and chondrosarcomas of the skull base.’  Int. J. Radiat. Oncol. Biol. Phys.,  47 , 591– 596 596..

‘Dose-volume ‘Dose-vo lume histogr histograms.’ ams.’   Int. J. Radiat. Oncol. Biol.  Phys.  21 , 71– 78 78.. Dudley, R. A. (1966) ‘Dosimetry with photographic emulRadiation Dosime Dosimetry try,, Vol. II , sions, sio ns,’’ pp. 325– 387 387.. in   Radiation  Attix, F. H. and Roesch, W. C. Eds. (Academic Press: New York). DuSautoy, A. R. (1996) ‘The UK primary standard calorimeterr for photo mete photon-be n-beam am abso absorbed rbed dose meas measure urement ment.’ .’  Phys. Med. Biol.  41 , pp. 137–151. Ebert, Ebe rt, P. P. J., Hardy, K. A. and Cad Cadena ena,, J. D. G. (19 (1965) 65) ‘Ene ‘E nerg rgy y de depe pend nden ence ce of fr free ee ra radi dica call pr prod oduc ucti tion on in alanine.’ Radiat. Res.  26 , 178– 197 197.. Eggerr, E., Scha Egge Schalenbo lenbourg, urg, A., Zogr Zografos afos,, L., Ber Bercher cher,, L., Boehri Boe hringe ngerr, T., Cha Chamot mot,, L. and Go Goite itein, in, G. (20 (2001) 01) ‘Maxim ‘Ma ximizi izing ng loc local al tum tumor or con contr trol ol and sur surviv vival al aft after er

de Kock, E. A. (2002) ‘The robot-based patient positioning syst sy stem em for hig high h pr preci ecisio sion n ra radio diothe thera rapy py at iTh iThemb emba a

prot pr oton on be beam am ra radi diot othe hera rapy py of uv uvea eall me mela lano noma ma.’ .’  Int. J. Radiat. Oncol. Biol. Phys.  51 , 138– 147 147..

193  

 PRESCRIBING, RECORDING, AND REPORTING PROTON-BEAM THERAPY  Enghardt, W., Debus, J., Haberer, T., Hasch, B. G., Hinz, R., Ja¨kel, O., Kra¨mer mer,, M., Lauc Lauckner kner,, K., Pa Pawe welke lke,, J. and Po¨nisch, F. F. (1999) ‘Positr ‘Positron on emission tomography for quality assurance of cancer therapy with light ion beams.’ Nucl. Phys.  A654, 1047c–1050c. Enghardt, W., Crespo, P., Fiedler, F., Hinz, R., Parodi, K., Pawe Pa welke lke,, J. and Po¨ nisc nisch, h, F. (2004 (2004)) ‘Char ‘Charged ged hadr hadron on tumour tumo ur ther therapy apy moni monitori toring ng by mean meanss of PET PET.’ .’   Nucl.  Instrum. Methods. Phys. Res .  A525, 284– 288 288.. Falkmer, S., Fors, B., Larsson, B., Lindell, A., Naeslund, J. an and d St Sten enso son, n, S. (1 (196 962) 2) ‘P ‘Pil ilot ot st stud udy y on pr prot oton on Radiol iol.   58, irradia irr adiation tion of huma human n car carcino cinoma.’ ma.’   Acta Rad 33–51.

Fuk ukum umot oto, o, S. S.,, En Endo do,, K. K.,, Mu Muto to,, K. K.,, Ak Akis isad ada, a, M. M.,, Kitaga Kit agawa wa,, T., Ina Inada, da, T., Tsuj sujii, ii, H., Ma Maruh ruhas ashi, hi, A., Hayakawa, Y., Takada, Y. and Tada, J. (1989) ‘Tsukuba Particle cle Medical Medi cal Proton Synchrotro Synchrotron,’ n,’ p. 70 in   Heavy Parti Therapy Ther apy Worksh orkshop op (PTC (PTCOG, OG, EORT EORTC, C, ECNE ECNEU) U), PS PSII Repor Re portt No No.. 69 (P (Paul aul Sch Scherr errer er Ins Instit titute ute,, Vi Villi lligen gen,, Switzerland). Fukum Fu kumur ura, a, A., Ka Kanai nai,, T., Kan Kanema ematsu tsu,, N. N.,, Yusa usa,, K. K.,, Maruhash Maru hashi, i, A., Noht Nohtomi, omi, A., Nish Nishio, io, T., Shim Shimbo, bo, M.,  Akagi, T., Yanou, T., Fukuda, S., Hasegaw Hasegawa, a, T., Kusano, Y. and Masuda, Y. (2002) ‘Proton beam dosimetry et ry:: pr prot otoc ocol ol an and d in inte terrco comp mpar aris ison on in Ja Japa pan, n,’’ Proc. Int Int.. Sym Symp. p. on Sta Standa ndards rds and pp.. 32 pp 321– 1– 32 326 6 in   Proc.

Fattibene, Fattib ene, P., Calic Calicchia chia,, A., d’Err d’Errico, ico, F., De Ange Angelis, lis, C., Egger, E. and Onori, S. (1996) ‘Preliminary assessment of Li LiF F an and d al alan anin ine e de dete tect ctor orss fo forr th the e do dosi sime metr try y of    Rad Radia iat. t. Prot Pr ot. . Dosi Do sim m prot pr oton on the thera rapy py be beams ams.’ .’ .   66, 305–309. Fidanz Fid anzio io,, A., Aza Azario rio,, L., De Ang Angeli elis, s, C., Pa Pacil cilio, io, M.,

Codes of Pra Codes Practi ctice ce in Med Medica icall Rad Radiat iation ion Dos Dosime imetry try, IAEA-CN-9 IAEACN-96 6 (Inte (Internat rnationa ionall Atom Atomic ic Ener Energy gy Agen Agency cy,,  Vienna. Gall,, K. P., Verhe Gall erhey y, L., Alons Alonso, o, J., Cast Castro, ro, J., Col Collier lier,, J. M.,, Ch M. Chu, u, W., Da Daft ftar ari, i, I. I.,, Go Goit itei ein, n, M. M.,, Ku Kubo bo,, H. H.,, Ludewigt Lude wigt,, B., Munz Munzenrid enrider er,, J., Pe Petti, tti, P., Ren Renner ner,, T.,

Onori, S. and Piermattei, A. (2002) ‘Radiochromic film dosimetr dosi metry y for a 62 MeV clini clinical cal pro proton ton beam beam.’ .’   Phys.  Med.  XV111, 15– 23 23.. Field, S. B. (1977) ‘Early and late normal tissue damage after fast neutrons.’   Int. J. Radiat. Oncol. Biol. Phys . 3, 203– 210 210.. Fitzek, M. M., Thornton, A. F., Rabinov, J. D., Lev, M. H., Pardo, F. S., Munzenrider, J. E., Okunieff, P., Bussiere, M., Bra Braun, un, I., Hoch Hochberg berg,, F. H., Hedl Hedley-W ey-Whyte, hyte, E. T.,

Rosenth Rose nthal, al, S., Smi Smith, th, A., Sta Staple ples, s, J., Sui Suit, t, H. and Thornt Tho rnton, on, A. (19 (1993a 93a)) ‘St ‘Stat ate e of the art art?? Ne New w pr proto oton n medica med icall fac facili ilitie tiess for the Mas Massa sachu chuse setts tts Ge Gener neral al Hosp Ho spit ital al an and d th the e Un Univ iver ersi sity ty of Ca Cali lifo forn rnia ia Da Davi viss Instru trum. m. Met Method hodss Phy Phys. s. Res Res.. Medical Center Center.’ .’   Nucl. Ins B79, 881– 884 884.. Gall Ga ll,, K. P., Ver erhe hey y, L. J. an and d Wag agne nerr, M. (1 (199 993b 3b)) ‘Com ‘C ompu pute terr-as assi sist sted ed posi po siti tion onin ing g of radi ra diot othe hera rapy py patients pati ents usin using g impla implanted nted radi radiopaq opaque ue fiduci fiducials.’ als.’   Med.

Liebsch, Liebsc h, N. J. and Har Harsh, sh, G. R. (19 (1999) 99) ‘Ac ‘Accel celer erat ated ed fractio fra ctionate nated d pro proton/ ton/photo photon n irra irradiat diation ion to 90 coba cobalt lt gray equivalent for glioblastoma multiforme: results of  Neuros rosurg urg.   91, a ph phas ase e II pr pros ospe pect ctiv ive e tr tria ial. l.’’   J. Neu 251–260. Fleming, D. M. and Glass, W. A. (1969) ‘Endothermic proRadiat. Res.   37, cesses cess es in tiss tissue-e ue-equiva quivalent lent plas plastic.’ tic.’   Radiat. 316–322. Fogl Fo glia iata ta,, A. A.,, Bo Bols lsi, i, A. an and d Co Cozz zzi, i, L. (2 (200 002) 2) ‘C ‘Cri riti tica call appraisa appr aisall of tre treatme atment nt tech technique niquess base based d on conv convenentional tio nal pho photon ton bea beams, ms, int inten ensit sity y mod modula ulate ted d pho photon ton beams and proton beams for therapy of intact breast.’  Radiother.. Oncol.  62 , 137– 145  Radiother 145.. Fong, Fo ng, P. P. M. M.,, Ke Keil, il, D. C., Does, Does, M. D. and Gore, Gore, J. C. (2001)) ‘Po (2001 ‘Polyme lymerr gels for magne magnetic tic res resonanc onance e imagi imaging ng

 Phys.  20 , 11 1153– 53– 11 1159. 59. Gall, K., Rosenthal, S. K. and Smith, A. R. (1994) ‘Proton dosimetry protocol comparisons,’ pp. 231–233 in  Proc.  NIRS International Seminar on the Application of   Heavy Ion Accelerators to Radiation Therapy of Cancer in Co Conne nnecti ction on wit with h XXI PTC PTCOG OG Mee Meetin ting g, Kan Kanai, ai, T. and an d Tak akad ada, a, E. Ed Eds. s. NI NIRS RS-M -M-1 -103 03/H /HIM IMA ACC-00 008 8 (Nation (Na tional al Inst Institute itute of Radio Radiologi logical cal Scie Sciences nces,, Chiba Chiba,, Japan. 231–233). Gasto Gas torf, rf, R., Hum Humphr phries ies,, L. and Ro Rozen zenfel feld, d, M. (19 (1986) 86) ‘Cylindric ‘Cyli ndrical al cham chamber ber dimen dimension sionss and the corr correspo espondnd N   A   Med. Phys ing values of   Awall   and   N gas  /( ).’ .   13, gas x ion 751–754. Geso, M., Ackerly, T. and Patterson, W. (2004) ‘Improving radiochromic film’s sensitivity by wrapping it with UV 

of radiation dose distributions at normal room atmosphere.’  Phys. Med. Biol .  46 , 3105–3113. Fonteno Fon tenot, t, J. D., Ne Newhau whauser ser,, W. D. and Titt Titt,, U. (2005 (2005)) ‘Design tools for proton therapy nozzles based on the double-scattering double-sca ttering foil technique.’   Radiat. Prot. Dosim. 116, 211–215. Ford, Fo rd, E. C., Magera Mageras, s, G. S., Yor Yorke ke,, E. and Ling, C. C. (2003)) ‘Res (2003 ‘Respira piration tion-cor -correl related ated spir spiral al CT CT:: A meth method od of  measurin meas uring g res respir pirato atory-in ry-induced duced anat anatomic omic moti motion on for radiation radiatio n treatme treatment nt planning.’ Med. Phys.  30 , 88– 97 97.. Nuclea earr Pa Part rtic icle less in Ca Canc ncer er Fowl Fo wler er,, J. F. (1 (198 981) 1)   Nucl Treatment (Adam Hilger Ltd.: Bristol, UK). Fraa Fr aass ss,, B. B.,, Do Dopp ppk ke, K. K.,, Hu Hunt nt,, M. M.,, Kut utch cher er,, G. G.,, Starks Sta rkscha chall, ll, G., Ste Stern, rn, R. and Van Dyk Dyke, e, J. (19 (1998) 98) ‘Ame ‘A meri rica can n As Asso soci ciat atio ion n of Ph Phys ysic icis ists ts in Me Medi dici cine ne

phosphor.’  Med. Phys.  31 , 1014–1016. Glimelius, Glimeliu s, B., Ask, A., Bjel Bjelken kengre gren, n, G., Bjor Bjork-Er k-Erikss iksson, on, T., Blo Blomqu mquis ist, t, E., Joh Johans ansson son,, B., Kar Karlss lsson on,, M. and Zackriss Zack risson, on, B. (200 (2005) 5) ‘Num ‘Number ber of pati patients ents potentially potentially eligible for proton therapy.’ therapy.’  Acta Oncol.  44 , 836– 849 849.. Goitein, Goit ein, M. (1977 (1977)) ‘The measuremen measurementt of tiss tissue ue hete heterorodens de nsit ity y to gu guid ide e ch char arge ged d pa part rtic icle le ra radi diot othe hera rapy py.’ .’  Int. J. Radiat. Oncol. Biol. Phys.  3 , 27–33. Goitein, Goit ein, M. (1978 (1978a) a) ‘Co ‘Compen mpensat sation ion for inhom inhomogen ogeneiti eities es in cha charge rged d par partic ticle le ra radio diothe thera rapy py usi using ng com comput puted ed Int. J. Ra Radi diat at.. On Onco col. l. Bi Biol ol.. Ph Phys ys.   4, tomography.’   Int. 499–508. Goitein, M. (1978b) ‘A technique for calculating the influence en ce of th thin in in inho homo moge gene neit itie iess on ch char arge ged d pa part rtic icle le beams.’ Med. Phys.  5 , 258– 264 264..

Radiation Therapy Committee, Task Group 53: Quality assuran assu rance ce for clin clinical ical radi radiothe otherapy rapy tre treatme atment nt planning.’  Med. Phys.  25 , 1773–1829.

Goitein, M. (1980) ‘Calculation of the uncertainty in the Am.. Me Med. d. dose deli deliver vered ed in radi radiatio ation n ther therapy apy.’ .’   J. Am  Assoc.  244 , 1347–1350.

194  

 REFERENCES Go Goite itein, in, M. (19 (1982a 82a)) ‘Li ‘Limit mitat ation ionss of tw two-d o-dime imensi nsiona onall treatment treat ment planning programs. programs.’’  Med. Phys.  9 , 580– 586 586.. Goite Go itein, in, M. (19 (1982b 82b)) ‘Ap ‘Appli plica catio tions ns of co compu mpute ted d tom tomoography in radiotherapy treatment planning,’ pp. 195– Progress in Medic Medical al Radia Radiation tion Physi Physics, cs, Vol. I, 293 in   Progress Orton, C. G. Ed. (Plenum Press, New York). Goitein, M. (1983) ‘Non standard deviations.’   Med. Phys. 10, 709–711. Goitein, M. (1985) ‘Calculation of the uncertainty in the dose deli deliver vered ed in radi radiati ation on ther therapy apy.’ .’   Med. Phys.   12, 608–612. Goitein, M. (1992) ‘The comparison of treatment plans.’

‘Modulation’’. http:/ ‘Modulation’’. http://huhepl.harvard. /huhepl.harvard.edu/~gottscha edu/~gottschalk/  lk/  BGdocs.zip/SOBP.pdf. Gotts Go ttscha chalk, lk, B., Pla Platai tais, s, R. and Pa Pagan ganett etti, i, H. (19 (1999) 99) ‘Nuclear ‘Nuc lear inte intera ractio ctions ns of 160 MeV prot protons ons sto stopping pping in copper: coppe r: A tes testt of Monte Carlo nucle nuclear ar mode models.’ ls.’   Med.  Phys.  26 , 2597–2601. Graf Gr affm fman an,, S. (1 (197 975) 5)   On the Evaluation of New  Irradiation Modalities in Tumor Therapy: An  Experimental and Clinical Study with Special ¨   University  Reference to High Energy Protons, Um Umea ea Medi Me dica call Di Diss sser erta tati tion ons, s, Ne New w Se Seri ries es No No.. 1 (U (Ume mea a¨ University, Sweden).

 Semin. Radiat. Oncol. ,   2, 246– 256 256.. Goit ite ein, M. (1 (199 995) 5) ‘L ‘Le ess sso ons to be

Graffman, S., Brahme, A. and Larsson, B. (1985) ‘Proton radiothe radi otherapy rapy with the Upps Uppsala ala cycl cyclotr otron. on. Expe Experien rience ce

lear arn ned

fro fr om