PHARMACEUTICAL LICENSING STRATEGIES Best practices in deal-making, valuations and strategic management
By Steven Seget
Steven Seget Steven Seget is Principal at Delphi Pharma, and provides independent strategic consulting services to the pharmaceutical and biotechnology industries. Steven previously managed the strategic healthcare consulting function at Datamonitor and has an MBA from the London Business School.
[email protected]
Delphi Pharma provides strategic, financial and market–based solutions to clients, focusing primarily on the portfolio management, business development and licensing functions. Delphi Pharma combines an extensive research network, applied analytical expertise and an established track record to deliver high value results and measurable impact to its clients. www.delphipharma.com
Copyright © 2006 Business Insights Ltd This Management Report is published by Business Insights Ltd. All rights reserved. Reproduction or redistribution of this Management Report in any form for any purpose is expressly prohibited without the prior consent of Business Insights Ltd. The views expressed in this Management Report are those of the publisher, not of Reuters. Reuters accepts no liability for the accuracy or completeness of the information, advice or comment contained in this Management Report nor for any actions taken in reliance thereon. While information, advice or comment is believed to be correct at the time of publication, no responsibility can be accepted by Business Insights Ltd for its completeness or accuracy. REUTERS and dotted and sphere logos are the house trade marks of Reuters Limited in more than 25 countries world-wide.
Printed and bound in Great Britain by FPC Greenaway. Ormolu House, Crimscott Street, London SE1 5TE. www.greenaways.com
ii
Table of Contents Pharmaceutical Licensing Strategies
Executive Summary
10
Introducing pharmaceutical licensing
10
Licensing trends
10
Licensing process
12
Licensing valuations
13
Licensing best practices
14
Chapter 1
Introducing pharmaceutical licensing
16
Summary
16
Introduction
17
The age of the partnership
17
Definitions
19
Report outline
20
Chapter 2
Licensing trends
22
Summary
22
Introduction
23
Headline deal trends
23
Licensing deal partners
27
Licensing deal types
32
Licensing deal subjects
36
iii
Chapter 3
Licensing process
42
Summary
42
Introduction
43
A complex process
43
In-licensing versus out-licensing
45
Licensing strategy
47
Opportunity identification In-licensing Out-licensing
49 50 52
Licensing evaluations General portfolio management Applications for licensing evaluations
56 57 59
Deal-making and agreement Key elements of a pharmaceutical license agreement
60 60
Post-deal management and analysis Alliance management
61 62
Using outside agencies
64
Chapter 4
Licensing valuations
70
Summary
70
Introduction
71
Valuing deals
71
Current best practices
73
Deal-making valuation model Model inputs Evaluation modeling Model outputs Model refinements
76 77 82 88 91
Chapter 5
Licensing best practices
100
Summary
100
Introduction
101
Top licensing deals of the 21st century Genentech-Roche Idenix-Novartis Millennium-Ortho Biotech
101 101 102 102
iv
AstraZeneca-AtheroGenics
103
Preferred licensing partners Leading in-licensing companies Novartis Leading out-licensing companies Cephalon
104 104 105 107 107
Recommendations for the future Licensing trends Licensing process Licensing valuations Licensing best practices
109 109 109 110 110
Chapter 6
Appendix
112
Primary research survey
112
Sources
115
Index
116
v
List of Figures Figure 2.1: Figure 2.2: Figure 2.3: Figure 2.4: Figure 2.5: Figure 2.6: Figure 2.7: Figure 2.8: Figure 2.9: Figure 2.10: Figure 2.11: Figure 2.12: Figure 3.13: Figure 3.14: Figure 3.15: Figure 3.16: Figure 3.17: Figure 3.18: Figure 4.19: Figure 4.20: Figure 4.21: Figure 4.22: Figure 4.23: Figure 4.24: Figure 4.25: Figure 4.26: Figure 4.27: Figure 4.28: Figure 4.29: Figure 4.30: Figure 4.31: Figure 4.32: Figure 4.33: Figure 4.34: Figure 4.35: Figure 4.36: Figure 5.37: Figure 5.38: Figure 5.39: Figure 5.40: Figure 6.41:
Number and average value of top 10 pharmaceutical company licensing deals, 20012005 24 Expected change in number of licensing deals during 2006 25 Expected change in average value of licensing deals during 2006 26 Number of top 10 pharmaceutical company licensing deals by partner, 2001-2005 28 Number of top 10 biotech company licensing deals by partner, 2001-2005 30 Number of biotech out-licensing deals by partner, 2001-2005 31 Number of top 10 pharmaceutical company licensing deals by deal type, 2001-2005 33 Number of top 10 biotech company licensing deals by deal type, 2001-2005 35 Proportion of product-based licensing deals by therapy area, 2001-2005 36 Number of product-based licensing deals by therapy area, 2001-2005 37 Number of R&D licensing deals by development stage, 2001-2005 38 Number of biotech R&D licensing deals by development stage, 2001-2005 39 Expected change in number of potential partners chasing each licensing deal during 2006 44 Expected change in the length of time required to complete a licensing deal during 2006 44 The pharmaceutical licensing process 46 Parties involved in identifying potential licensing opportunities, 2006 65 Parties involved in conducting due diligence for potential licensing opportunities, 2006 66 Parties involved in the valuation and negotiation of potential licensing deals, 2006 67 Information shared between partners during licensing negotiations, 2006 74 Valuation techniques used in determining optimal licensing deal terms, 2006 75 R&D costs by phase, 2000 78 R&D lead times by phase, 2000 79 R&D success probabilities by phase, 2000 80 Drug market diffusion curve – product lifecycle 81 Likelihood of outcomes for new phase I, phase II and phase III drugs 85 Expected real values (non-discounted) for new phase I, phase II and phase III drugs 86 Discounted expected real values for new phase I, phase II and phase III drugs 87 Discounted expected real values for new phase I, phase II and phase III drugs (adjusted for lower R&D cost inflation) 88 Deal outcomes for out-licensor 89 Deal outcomes for in-licensor 89 Share of expected deal outcomes by partner 90 R&D costs by phase by therapy area, 2000 92 R&D lead times by phase by therapy area, 2000 93 R&D success probabilities by phase by therapy area, 2000 94 Peak sales and year of peak sales by therapy area, 2000 94 Discounted value of sales by therapy area, 2000 95 In-licensing partner of choice, 2006 104 Business development and licensing department, Novartis 106 Licensing process at Novartis 106 Out-licensing partner of choice, 2006 107 Licensing trends survey respondents by company focus 112
vi
Figure 6.42: Figure 6.43:
Licensing trends survey respondents by functional responsibility Licensing trends survey respondents by licensing responsibility
113 114
List of Tables Table 4.1:
R&D cost by phase and peak sales, 2006 (expressed in 2006 dollars)
vii
83
viii
Executive Summary
9
Executive Summary Introducing pharmaceutical licensing
With falling R&D productivity and continued healthcare cost containment and generic competition pushing down the returns available for successfully launched products, only those companies able to complement internal efforts with a strong partnering strategy will be able to remain competitive over the next five-to-ten years.
Since Eli Lilly and Genentech forged the first truly ‘strategic’ licensing agreement almost 30 years ago (Humulin, 1978), the licensing deal has been a fundamental part of every pharmaceutical and biotechnology company’s strategy.
The pharmaceutical industry remains one of the most risky industries in the world, with the licensing agreement providing the best safeguard for managing, or at least sharing, some of the inherent risks involved with pharmaceutical R&D.
Given its relative infancy, the strategic alliance – one that involves some level of ongoing collaboration between partners – has grown in frequency, value and complexity over the past 20 years or so.
Licensing trends
A pharmaceutical company’s limited resources to manage inter-company relationships and collaborative projects places an upper limit to the number of different agreements that can formed each year. However, there has been a consistent increase in average deal values between 2002 and 2005, likely to be the result of deal sizes increasing to include multiple development compounds. It appears that the leading pharmaceutical companies have determined that the size and quality of the deal is more important than signing a greater number of deals.
10
2001 was at the height of the ‘golden age’ for biotechnology, where company valuations were high and every pharmaceutical company wanted to access their technologies. However, a period of rationalization in the following three years saw the leading pharmaceutical companies turn away from risky biotechnology companies and back to more traditional, but relatively less risky, pharmaceutical partners. However, 2005 saw a shift in licensing activity by partner, with leading pharmaceutical companies significantly increasing licensing activity with biotech partners at the expense of intra-pharmaceutical deals.
As with deal numbers and deal values, the proportion of relationship-based alliances was also high in 2001, before falling and steadily increasing to a new peak in 2005. Again, these trends appear to confirm the leading pharmaceutical companies’ new found confidence in committing to long-term alliances.
The greatest growth in product-based pharmaceutical licensing over the past four years has been in agreements involving cancer therapies. Many biotech approaches, including monoclonal antibodies and growth factors, have their most valuable applications in the treatment of cancer.
Licensing agreements in later stages, particularly in phase II, appear to be driving growth in licensing activity over the past five years. This trend is consistent with the increased maturity of the biotech sector and the emergence of biotech companies with the resources and capabilities to develop lead drugs to later stages of clinical development before seeking a pharmaceutical partner.
11
Licensing process
Growth in the number of companies involved in chasing each licensing agreement will lead to increased pressure on licensing lead times, requiring greater levels of resources to be committed to the identification and evaluation of potential opportunities and partners.
Deal failure is often the result of one or more partners not clearly identifying their strategic aims for a licensing deal. Pharmaceutical companies must not enter into an agreement without having determined that the licensing opportunity satisfies a real and valuable objective for the company.
Having a clear understanding of what you can and cannot offer potential partners is critical in order not to over promise or waste time negotiating over the wrong deal with the wrong partner.
Many small, but ambitious biotechnology companies have managed to do a good job with the preparation of presentation materials and the identification of target licensing partners only to make a bad job of establishing a first contact with the company and failing to make any further progress.
If you are a small biotech looking to agree a blockbuster late-stage licensing agreement with a top 10 pharmaceutical company you can either spend money on a good lawyer now to negotiate and agree a favorable contract or spend double the money later down the line having to continually defend claims from your eventual licensing partner.
Biotech companies look towards specialist agencies and key investors, such as venture capitalists, in order to support the licensing process. Pharmaceutical companies appear to have greater levels of licensing resources and expertise inhouse, with the majority completing the entire licensing process without outside help.
12
Licensing valuations
According to a survey of 142 licensing executives, pharmaceutical companies are more likely than biotech companies to share financial evaluations during licensing negotiations, with more than 80% of companies sharing at least a single point financial projection and almost 50% sharing a more detailed probabilistic evaluation.
The most common evaluation technique used in determining optimal deal terms is a discounted cash flow net present value (NPV) calculation, which is used in more than 70% of companies according to surveyed licensing executives.
A deal-making valuation model is one used specifically to agree licensing terms between partners. While the approach might be similar to that used to provide recommendations for selecting the most valuable licensing opportunities to pursue the outputs of the model are quite different.
If the inputs into the valuation model cannot be agreed upon by both licensing parties then the outputs of the exercise will not provide any common ground for negotiation. Similarly, if the modeling approach and any assumption are not clear and defendable then the effective use of the evaluation model for negotiation will be limited.
The first test as to whether two companies are going to be successful in collaborating together as part of a strategic licensing agreement is whether or not they are able to negotiate an agreement based on significant common ground. If two companies are unable to agree upon the true value of the licensing asset and subsequently on the fair distribution of risks, returns and responsibilities then there is little hope that they will be able to reach a satisfactory conclusion once the real work of implementing a licensing deal begins.
13
Licensing best practices
With no two drugs or partnerships the same, there are no hard-and-fast rules for successful pharmaceutical licensing. Lessons can be learned by looking at the leading deals and deal-makers, but the application of best practices must always be directed by the specifics of the deal, rather than reverting to a list of generalized benchmarks.
The leading strategic pharmaceutical licensing deal, mentioned by the greatest number of surveyed licensing executives, is the ongoing relationship between Roche and Genentech. Over the course of the 16 year relationship Roche has acquired non-US marketing rights to a range of Genentech products, including Rituxan (rituximab) in 1995, Herceptin (trastuzumab) in 1998 and Avastin (bevacizumab) in 2003.
Novartis was considered to be the in-licensing partner of choice in 2006. The company has worked hard to position itself as a preferred licensing partner, employing a structured process providing a quick evaluation of opportunities and the early involvement of senior management to expedite decision-making. As part of this standard review process, Novartis employs a single gateway for all opportunities allowing for improved coordination and contact management.
Cephalon was considered to be the out-licensing partner of choice in 2006. Cephalon’s licensing model of acquiring rights to marketed products in order to generate revenues cash flow to help fund the development of in-house products appears to have worked well. More importantly, as an out-licensor, the company has been willing to share promotional rights for its lead products, particularly in the US and Japan, in order to maximize the returns from its internal assets.
14
CHAPTER 1
Introducing pharmaceutical licensing
15
Chapter 1
Introducing pharmaceutical licensing
Summary
With falling R&D productivity and continued healthcare cost containment and generic competition pushing down the returns available for successfully launched products, only those companies able to complement internal efforts with a strong partnering strategy will be able to remain competitive over the next five-to-ten years.
Since Eli Lilly and Genentech forged the first truly ‘strategic’ licensing agreement almost 30 years ago (Humulin, 1978), the licensing deal has been a fundamental part of every pharmaceutical and biotechnology company’s strategy.
The pharmaceutical industry remains one of the most risky industries in the world, with the licensing agreement providing the best safeguard for managing, or at least sharing, some of the inherent risks involved with pharmaceutical R&D.
Given its relative infancy, the strategic alliance – one that involves some level of ongoing collaboration between partners – has grown in frequency, value and complexity over the past 20 years or so.
16
Introduction Pharmaceutical licensing strategies: Best practices in deal-making, valuations and strategic management provides a detailed analysis of licensing strategies in the pharmaceutical and biotechnology industries. The report draws upon deal-making trend data, primary research survey results and a profile of best practices in pharmaceutical licensing in order to present a set of actionable recommendations for optimizing dealmaking. With falling R&D productivity and continued healthcare cost containment and generic competition pushing down the returns available for successfully launched products, only those companies able to complement internal efforts with a strong partnering strategy will be able to remain competitive over the next five-to-ten years.
The age of the partnership The purpose of this report is to provide a strategic perspective on the increasingly important area of pharmaceutical licensing. Questions the report attempts to answer include:
Why is licensing important in the pharmaceutical and biotechnology industries?
What are the recent trends in pharmaceutical licensing activity?
What are best practices in developing a robust licensing process?
What can be done to optimize licensing deal values through collaborative evaluations?
What does it take to become a partner of choice?
What lessons can be drawn from the key successful deals formed over the past 10 years?
17
In order to answer these questions the report brings together research and analysis from multiple sources. Leading deal-tracking databases including MedTRACK and Recombinant Capital provide licensing trend data. A survey of 142 senior licensing executives drawn from across the pharmaceutical and biotechnology industries provides a real-time assessment of current licensing processes, practices and expectations regarding future deal-making trends. Finally, company profiles and licensing case studies provide a detailed insight into successful licensing strategies and best practices.
Ever since Eli Lilly and Genentech forged the first truly ‘strategic’ licensing agreement almost 30 years ago for recombinant insulin (Humulin, 1978), the licensing deal has been a fundamental part of every pharmaceutical and biotechnology company’s strategy. Licensing has been used to provide a more flexible mechanism through which R&D, sales and marketing and, most important of all, revenue and income streams can be balanced over time. The biotechnology industry could not exist on venture capital alone, and the pharmaceutical industry would be in a sorry state without the breakthrough innovations licensed from biotech companies. The pharmaceutical industry remains one of the most risky industries in the world, with the licensing agreement providing the best safeguard for managing, or at least sharing, some of the inherent risks involved with pharmaceutical R&D.
The strategic alliance – one that involves some level of ongoing collaboration between partners – has grown in frequency, value and complexity over the past 20 years or so. Much of the growth in pharmaceutical licensing is linked to the emergence and development of the biotechnology industry, with deal-making activity broadly tracking funding levels and valuations for biotechnology companies. An illustration of the rise in value and complexity of strategic licensing is provided by the 2001 agreement for Erbitux between Bristol-Myers Squibb and ImClone Systems. Nothing illustrates the importance, value and complexity of pharmaceutical licensing better than a story that starts with a headline value of US$2 billion and concludes with an initial nonapprovable letter from the Food and Drug Administration (FDA), a rewrite of the licensing contract and, as a sub-plot, jail time for one the key protagonists. 18
Definitions The term ‘strategic pharmaceutical licensing’ presents many questions regarding scope and interpretation. In order to avoid confusion the following definitions of key terms can be applied throughout the remainder of the report:
License – an agreement between two or more parties relating to the ownership of intellectual property (IP) rights, and in particular those rights pertaining to a technology, development compound or marketed product;
Pharmaceutical license – an agreement as above that involves IP rights relating to a pharmaceutical technology, development compound or marketed product. For the purposes of this report a pharmaceutical license is used as an umbrella term to include IP relating to both pharmaceutical and biotechnology products and technologies;
Strategic license – an agreement as above that also involves an ongoing collaboration between partners, such as a co-development or co-commercialization agreement;
Licensing – the act of forming a license, relating either to process or ongoing activity in forming licenses;
Deal-making – see licensing above;
In-licensing – the act of acquiring the rights to a technology or product;
In-licensor – the company acquiring rights to a technology or product;
Out-licensing – the act of divesting the rights to a technology or product;
Out-licensor – the company divesting rights to a technology or product.
19
Report outline The report has been divided into four key sections. The licensing trends chapter presents an overview of recent trends in pharmaceutical licensing and details industry expectations for future deal-making. The licensing process chapter provides a practical guide to pharmaceutical licensing and sets out the different phases involved in optimal deal-making for both in- and out-licensors. The licensing valuations chapter outlines a simple licensing valuation model based on independent sources to support negotiation and deal-making efforts between prospective partners. Finally, the licensing best practices chapter presents a set of detailed cases studies for successful licensing deals and leading licensing partners in order to provide key lessons for optimizing strategic pharmaceutical licensing.
20
CHAPTER 2
Licensing trends
21
Chapter 2
Licensing trends
Summary
A pharmaceutical company’s limited resources to manage inter-company relationships and collaborative projects places an upper limit to the number of different agreements that can formed each year. However, there has been a consistent increase in average deal values between 2002 and 2005, likely to be the result of deal sizes increasing to include multiple development compounds. It appears that the leading pharmaceutical companies have determined that the size and quality of the deal is more important than signing a greater number of deals.
2001 was at the height of the ‘golden age’ for biotechnology, where company valuations were high and every pharmaceutical company wanted to access their technologies. However, a period of rationalization in the following three years saw the leading pharmaceutical companies turn away from risky biotechnology companies and back to more traditional, but relatively less risky, pharmaceutical partners. However, 2005 saw a shift in licensing activity by partner, with leading pharmaceutical companies significantly increasing licensing activity with biotech partners at the expense of intra-pharmaceutical deals.
As with deal numbers and deal values, the proportion of relationship-based alliances was also high in 2001, before falling and steadily increasing to a new peak in 2005. Again, these trends appear to confirm the leading pharmaceutical companies’ new found confidence in committing to long-term alliances.
The greatest growth in product-based pharmaceutical licensing over the past four years has been in agreements involving cancer therapies. Many biotech approaches, including monoclonal antibodies and growth factors, have their most valuable applications in the treatment of cancer.
Licensing agreements in later stages, particularly in phase II, appear to be driving growth in licensing activity over the past five years. This trend is consistent with the increased maturity of the biotech sector and the emergence of biotech companies with the resources and capabilities to develop lead drugs to later stages of clinical development before seeking a pharmaceutical partner.
22
Introduction Pharmaceutical licensing has undergone significant changes over the past 20-30 years. Trends in deal-making activity have shown an increase in the value and number of licensing agreements, while the types of company involved, stage of development of the deal subject and complexity of the agreement have also changed significantly. Today’s strategic licensing deals are more valuable, numerous and complex than ever before, and as a consequence companies must build competences in the licensing field in order to support these trends.
Headline deal trends The number and average value of licensing deals involving the top 10 pharmaceutical companies both increased between 2001 and 2005, as shown in Figure 2.1. The top 10 companies were considered to be those with the highest pharmaceutical product sales in 2005 (Pfizer, Sanofi-Aventis, GlaxoSmithKline, AstraZeneca, Johnson & Johnson, Roche, Merck & Co., Novartis, Wyeth and Bristol-Myers Squibb), with all licensing activity for companies acquired by the top 10 companies between 2001 and 2005 consolidated over the period.
Average deal values for the top 10 pharmaceutical companies have risen consistently between 2002 and 2005. Average deal values are based on the headline deal values released by partnering companies at the time of signing an agreement. As a result these deal values often refer to the maximum potential deal value and usually exclude any royalty payments to be paid once a drug is brought to market.
The number of licensing deals involving top 10 pharmaceutical companies has both increased and decreased at different periods between 2001 and 2005. While licensing activity in 2005 is the highest it has been over the five years, similar peaks were reached in 2001 and 2003.
23
Figure 2.1: Number and average value of top 10 pharmaceutical company licensing deals, 2001-2005
350
350 300
304
297
250 Number of deals
300
273
250
266
200
200 187
150
172
150
170 149
100
100
50
50
0
Average deal value ($ million)
285
300
0 2001
2002
Number of deals
2003
2004
2005
Average deal value ($ million)
Business Insights Ltd
Source: MedTRACK Deals & Alliances database
By looking at an analysis of the top 10 pharmaceutical companies only, we can see that the number of licensing deals that have been signed by any one company appears to have reached a plateau between 2001 and 2005. While an average of 30 deals per company is significantly more than would have been seen 20 or even 10 years ago, licensing activity in the leading companies appears to have reached a natural limit. A company’s limited resources and abilities to manage inter-company relationships and collaborative projects places an upper limit on the number of different agreements that can formed each year. However, it is noticeable that there has been a consistent increase in average deal values between 2002 and 2005. This is likely to be the result of deal sizes increasing to include multiple development compounds. It appears that the leading pharmaceutical companies have determined that the size and quality of the deal is more important than driving increases in the number of deals. As a result, the same
24
number of licensing partners are providing a greater level of licensing value over a greater number of licensed compounds.
Around two thirds of the 142 licensing executives surveyed for this report considered the number of pharmaceutical licensing deals likely to increase during 2006, as shown in Figure 2.2. This increase in licensing agreements was considered likely to be higher in the biotechnology industry than in the pharmaceutical industry. It appears that there is significant room for growth in licensing activity in the pharmaceutical and biotechnology markets, but growth will be primarily driven by smaller companies building up their licensing capabilities and increasing their deal-making activity to the levels found in established pharmaceutical and biotech companies.
Figure 2.2: Expected change in number of licensing deals during 2006
100% Proportion of survey respondents
90%
Increase significantly
80% 70%
Increase somewhat
60%
Stay roughly the same
50% 40%
Decrease somewhat
30% 20%
Decrease significantly
10% 0% Pharma
Biotech
Other
Source: Business Insights Licensing Trends Survey, 2006
25
Overall
Business Insights Ltd
The slowest level of growth in licensing activity is expected in specialty, drug delivery, generics and diagnostic companies. These companies are either well-established licensors, such as specialty and drug delivery companies, or are in industry sectors where licensing plays a less important role, such as generics and diagnostic products.
Around 60% of surveyed licensing executives expect average deal values to increase in 2006, as shown in Figure 2.3. Again, the likely increase is considered to be higher for biotech companies than for pharmaceutical companies. Only a very small percentage of survey respondents (just over 10%) expect deal values to decrease in 2006. Trends in increased average deal values found over the past four years appear set to be continued in the near future.
Figure 2.3: Expected change in average value of licensing deals during 2006
100% Proportion of survey respondents
90%
Increase significantly
80% 70%
Increase somewhat
60%
Stay roughly the same
50% 40%
Decrease somewhat
30% 20%
Decrease significantly
10% 0% Pharma
Biotech
Other
Overall
Business Insights Ltd
Source: Licensing trends survey, 2006
26
Alongside increases in industry deal numbers and headline values, it also appears that pharmaceutical licensing deals are becoming more complex. In 2004, GlaxoSmithKline and Theravance signed a multi-compound, multi-therapy area deal that involved sophisticated financial put and call options on Theravance shares in support of the biotech company’s imminent initial public offering (IPO). Other deals involving major equity stakes and multiple, cross-portfolio compounds include the 2003 agreement between Novartis and Idenix and the landmark agreement between Genentech and Roche, first signed in 1990 and revised in 1995. Sliding royalty rates, contingent equity valuations and complex territorial and market splits for co-developed and co-promoted compounds are fast becoming the norm. Those companies not able to negotiate their way through these sophisticated deal terms will quickly find themselves on the wrong end of a bad deal.
Licensing deal partners Between 2001 and 2005 it appears that the leading pharmaceutical companies have returned back to the biotech industry as a source for licensing. 2001 was at the height of the ‘golden age’ for biotechnology, where company valuations were high and every pharmaceutical company wanted to access their technologies. However, a period of rationalization in the following three years saw the leading pharmaceutical companies turn away from the more risky biotechnology companies and back to more traditional, but relatively less risky, pharmaceutical partners. However, as shown in Figure 2.4, 2005 saw a shift in licensing activity by partner, with leading pharmaceutical companies significantly increasing licensing activity with biotech partners at the expense of intra-pharmaceutical company deals.
27
Figure 2.4: Number of top 10 pharmaceutical company licensing deals by partner, 2001-2005
Number of top 10 pharma licensing deals
350 300 Pharma 90
250 135
116
128 122
200
Biotech 150 157
100
115
106
106 100
Other
50 50
63
63
51
57
2001
2002
2003
2004
2005
0
Proportion of top 10 pharma licensing deals
100% 90% 29.6%
80%
45.0%
40.7%
43.1%
Pharma
44.7%
70% 60% 50%
Biotech 51.6% 37.2%
40% 38.3%
35.7%
36.6%
30% 20% 10%
Other 16.7%
22.1%
21.2%
18.7%
18.8%
2002
2003
2004
2005
0% 2001
Business Insights Ltd
Source: MedTRACK Deals & Alliances database
28
Over the last five years the biotechnology industry has matured to become a tried and tested source for good science and collaborative development. This has been helped along by a period of consolidation in the industry to allow stronger biotech companies to emerge with full pipelines and robust development capabilities. Both pharmaceutical and biotech companies stand to benefit as a result of increased licensing activity between the two – with pharmaceutical companies gaining access to innovative and valuable compounds and biotech companies receiving improved deal values and longterm collaborative partners.
Taking a look from the other side of the industry, the trends for the top 10 biotech companies based on total 2005 revenues (Amgen, Genentech, Genzyme, Serono, CSL, Biogen Idec, Gilead, Chiron, MedImmune and Cephalon) appear to closely mimic those found in leading pharmaceutical companies. As was the case for the top 10 pharmaceutical companies, licensing activity for the top 10 biotech companies in 2005 saw an increase in the number of deals signed with other biotechnology companies.
Interestingly, it appears that the leading biotechnology companies are not behind the increase in pharmaceutical-biotech licensing found for the top 10 pharmaceutical companies, with the number of deals between leading biotech companies and pharmaceutical companies decreasing consistently between 2003 and 2005, as shown in Figure 2.5. It is clear that leading pharmaceutical and biotechnology companies are much less likely to partner amongst themselves than they are to partner with smaller companies that represent less of a strategic, competitive threat.
29
Figure 2.5: Number of top 10 biotech company licensing deals by partner, 2001-2005
Number of top 10 biotech licensing deals
100 90 80
26
Pharma
70 30
60
35
42
32
50
Biotech 48
40 30
41
34
33
32
20 Other 10
18
8
11
8
9
2001
2002
2003
2004
0 2005
Proportion of top 10 biotech licensing deals
100% 90% 80%
28.3% 38.0%
Pharma
43.2%
43.8% 51.2%
70% 60% 50%
Biotech 52.2%
40%
51.9%
42.5%
30%
44.6% 39.0%
20%
Other
10% 10.1%
13.8%
2001
2002
9.8%
12.2%
2003
2004
19.6%
0% 2005
Business Insights Ltd
Source: MedTRACK Deals & Alliances database
30
As a proportion of all biotech out-licensing deals, other biotech partners continue to provide just over 60% of all licensing partners, as shown in Figure 2.6. This proportion has been on the increase over the previous two years, and is consistent with the figures for leading biotech companies in Figure 2.5. However, with trends in licensing activity for the top 10 pharmaceutical companies showing an increase in the number of deals signed between leading pharmaceutical companies and biotech companies it appears that the growth in biotech licensing is greater for intra-biotech deals than for pharmabiotech deals. It is also evident that smaller pharmaceutical companies are not yet fully embracing the renewed interest in biotech licensing found in their larger contempories.
Figure 2.6: Number of biotech out-licensing deals by partner, 2001-2005
Proportion of biotech out-licensing partners
100% 90% 80% 70%
58.3%
58.6%
55.3%
56.9%
60.6%
Biotech
60% 50% 40% Pharma
30% 20%
41.7%
41.4%
44.7%
43.1%
39.4%
2001
2002
2003
2004
2005
10% 0%
Business Insights Ltd
Source: Recombinant Capital, Analyst’s Notebook
31
Licensing deal types In order to better understand trends in pharmaceutical licensing agreements signed over the past five years three key categorizations were used:
License/acquisition – including all agreements that involve a simple transaction involving intellectual property (IP) rights;
R&D/S&M – including all agreements involving a transactional exchange of IP rights and some exchange of expertise and know-how;
Collaboration – including all agreements involving a significant, ongoing relationship, either in R&D or marketing and sales.
For the top 10 pharmaceutical companies, there has been a move away from simple licensing/acquisition deals towards more relationship-based alliances involving collaborative R&D and sales and marketing, as shown in Figure 2.7. As with deal numbers and deal values, the proportion of relationship-based alliances was also high in 2001, before falling and steadily increasing to a new peak in 2005. Again, these trends appear to confirm the leading pharmaceutical companies’ new found confidence in
committing
to
long-term
alliances.
The
smaller
number
of
simple
license/acquisitions in 2005 is also representative of a down-turn in pharmaceutical mergers and acquisitions, which are often followed by portfolio consolidation and the divestment of overlapping products/compounds.
32
Figure 2.7: Number of top 10 pharmaceutical company licensing deals by deal type, 2001-2005
Number of top 10 pharma licensing deals
350 Collaboration
300 250
75
91
105
92
78
R&D/ S&M
200 79
150 100
95 68
87
86
License/ acquisition 74 91
83 88
73
Others
50 56
34
44
2001
2002
2003
2004
30.3%
32.3%
25.3%
28.6%
20
0
40
2005
Proportion of top 10 pharma licensing deals
100% 90% 80%
Collaboration 34.5%
70% R&D/ S&M 60%
26.3%
23.9%
32.0%
31.9% 28.3%
50%
License/ acquisition
40% 30%
24.7%
31.9%
27.9% 32.2%
20% 10%
24.0%
Others 18.7%
11.9%
14.8%
2002
2003
7.3%
13.2%
0% 2001
2004
2005
Business Insights Ltd
Source: MedTRACK Deals & Alliances database
33
For the top 10 biotech companies, the five-year trend towards relationship licenses is similar to that found for the leading pharmaceutical companies. As shown in Figure 2.8, the leading biotech companies have increased the number of R&D/S&M agreements at the expense of simple license/acquisition agreements. Interestingly, the number of collaborative agreements has remained relatively unchanged between 2001 and 2005 – likely to be a function of big biotech turning to smaller biotech to drive growth in their licensing activity over the past five years. Big biotech have grown to reach a critical mass and can now avoid having to sign collaborative out-licensing agreements in order to bring their drugs to market. At the same time, the leading biotech companies have increased R&D/S&M licensing activity, particularly with other smaller biotech companies, in order to fill their pipelines and maintain a throughput in R&D.
34
Figure 2.8: Number of top 10 biotech company licensing deals by deal type, 2001-2005
Number of top 10 biotech licensing deals
100 90
Collaboration
80 70
26 20 26
25
60
30
50 40
R&D/ S&M
28
17
30
24
License/ acquisition
20
30
23
20
21 24
15
Others
29
10
15
8
11
2003
2004
15
0 2001
2002
32.9%
31.3%
2005
Proportion of top 10 biotech licensing deals
100% 90%
24.4%
28.3%
Collaboration
37.8%
80% 70%
R&D/ S&M 60% 50%
21.3%
36.6%
30.4%
32.6% 27.0%
License/ acquisition
40% 28.8%
30%
29.3%
20%
20.3%
22.8%
36.7%
10%
Others 18.8% 9.8%
14.9%
16.3%
2004
2005
0% 2001
2002
2003
Business Insights Ltd
Source: MedTRACK Deals & Alliances database
35
Licensing deal subjects More than 20% of all product-based pharmaceutical licensing agreements signed between 2001 and 2005 involved cancer indications, as shown in Figure 2.9. Agreements involving infections and central nervous system (CNS) indications accounted for approximately 14% each, followed by cardiovascular and circulatory system indications and autoimmune and inflammation indications.
Figure 2.9: Proportion of product-based licensing deals by therapy area, 2001-2005
Proportion of product-based licensing agreements, 2001-2005 20.5%
Cancer Infections
13.9%
Central Nervous System
13.8% 8.4%
Cardiovascular and Circulatory System
7.6%
Autoimmune and Inflammation
6.7%
Metabolic/ Endocrinology
6.2%
Dermatology
5.2%
Respiratory and Pulmonary System Digestive System
4.0%
Blood and Lymphatic System
3.7%
Business Insights Ltd
Source: MedTRACK Deals & Alliances database
The greatest level of growth in product-based pharmaceutical licensing over the past four years has been in agreements involving cancer therapies, as shown in Figure 2.10. As a share of all product-based licensing, cancer indications have grown to the levels found in 2001, closely matching trends found for biotech licensing in general. Many biotech approaches in monoclonal antibodies and growth factors have their most
36
valuable applications in the treatment of cancer. Other areas of licensing growth can be found in infections and autoimmune and inflammation indications, particularly in 2005. Leading therapy areas with more disappointing licensing growth rates include CNS and cardiovascular and circulatory system. With pharmaceutical licensing often involving agreements for the most successful R&D projects available, deal trends by therapy area provide a good indication of current and future R&D trends. It appears clear that we are likely to see some significant growth in the number of late stage cancer therapies coming through the pipeline, with more disappointing growth rates anticipated for CNS and cardiovascular and circulatory system therapies.
Proportion of product-based licensing deals
Figure 2.10: Number of product-based licensing deals by therapy area, 20012005
24% 22% 20% 18% 16% 14% 12% 10% 8% 6% 4% 2% 0% 2001
2002
2003
2004
2005
Cancer Infections Central Nervous System Autoimmune and Inflammation Cardiovascular and Circulatory System
Business Insights Ltd
Source: MedTRACK Deals & Alliances database
37
By stage of development, there has been some recent movement towards pre-clinical licensing in 2005, as shown in Figure 2.11. More generally, there has been trend in licensing activity from phase II to phase I licensing over the period 2001 and 2005. The general share of R&D licensing for phase III compounds has remained relatively unchanged over the five-year period at around 20%.
Figure 2.11: Number of R&D licensing deals by development stage, 2001-2005
100% Proportion of R&D licensing deals
90%
21.8%
23.2%
18.5%
22.0%
19.3%
Phase III
80% 70% 60%
38.3% 42.6%
39.4%
34.5%
Phase II
37.2%
50% 40% 30%
18.0% 19.5% 16.2%
17.8%
19.4%
19.7%
2001
2002
Phase I
20.5%
20% 10%
28.3%
23.7%
20.2%
2003
2004
Pre Clinical
0% 2005
Business Insights Ltd
Source: MedTRACK Deals & Alliances database
Looking at biotech R&D licensing deals specifically, there appears to have been a significant trend towards later stage deals between 2001 and 2005, as shown in Figure 2.12. For biotech R&D licensing deals, the proportion of agreements signed in phase II or later has grown from less than 20% in 2001 to almost 30% in 2005. Licensing agreements in later stages, particularly in phase II, appear to be driving growth in licensing activity over the past five years. This trend is consistent with the maturing of the biotech sector and the emergence of biotech companies with the resources and capabilities to develop lead drugs to later stages of clinical development before seeking
38
a pharmaceutical partner. As a consequence of this trend, the value of licensing deals will inevitably increase as the risk involved with compounds found at later stages of development is reduced. The increased number of licensing candidates available at later stages of clinical development also helps to explain the stagnant growth in the number of agreements been signed by major companies. It is now possible for pharmaceutical companies to limit risk by in-licensing a single phase II compound rather than licensing two phase I compounds or a handful of preclinical opportunities. It appears that the biotech industry is better prepared to bear the costs of unsuccessful early stage compounds alone in order to share higher rewards with its partners for successful compounds in later stages of development.
Figure 2.12: Number of biotech R&D licensing deals by development stage, 2001-2005
100% Proportion of R&D licensing deals
90% 80%
19.4%
23.5%
26.4%
30.0%
29.3%
Phase II and above
7.4% 10.5%
70%
9.9% 13.1%
60%
11.7%
Preclinical/ Phase I
50% 40%
73.2%
30%
66.0%
63.7%
56.9%
59.0%
2004
2005
Research
20% 10% 0% 2001
2002
2003
Business Insights Ltd
Source: Recombinant Capital, Analyst’s Notebook
39
40
CHAPTER 3
Licensing process
41
Chapter 3
Licensing process
Summary
Growth in the number of companies involved in chasing each licensing agreement will lead to increased pressure on licensing lead times, requiring greater levels of resources to be committed to the identification and evaluation of potential opportunities and partners.
Deal failure is often the result of one or more partners not clearly identifying their strategic aims for a licensing deal. Pharmaceutical companies must not enter into an agreement without having determined that the licensing opportunity satisfies a real and valuable objective for the company.
Having a clear understanding of what you can and cannot offer potential partners is critical in order not to over promise or waste time negotiating over the wrong deal with the wrong partner.
Many small, but ambitious biotechnology companies have managed to do a good job with the preparation of presentation materials and the identification of target licensing partners only to make a bad job of establishing a first contact with the company and failing to make any further progress.
If you are a small biotech looking to agree a blockbuster late-stage licensing agreement with a top 10 pharmaceutical company you can either spend money on a good lawyer now to negotiate and agree a favorable contract or spend double the money later down the line having to continually defend claims from your eventual licensing partner.
Biotech companies look towards specialist agencies and key investors, such as venture capitalists, in order to support the licensing process. Pharmaceutical companies appear to have greater levels of licensing resources and expertise inhouse, with the majority completing the entire licensing process without outside help.
42
Introduction The pharmaceutical licensing process is both lengthy and complex. Whether approaching licensing in order to acquire or divest intellectual property, there are five key steps that must be completed on the way to agreeing a successful licensing deal. Pharmaceutical companies must determine a clear licensing strategy, identify appropriate opportunities, evaluate those opportunities, negotiate favorable terms and then ensure agreements are successfully implemented and managed.
A complex process Pharmaceutical licensing is a complex process involving the identification and valuation of multiple potential opportunities and partners. More than 75% of surveyed licensing executives expect the number of potential partners chasing each licensing deal to increase in 2006, as shown in Figure 3.13. Growth in the number of companies involved in chasing each licensing agreement will lead to increased pressure on licensing lead times, with greater levels of resources required for the identification and evaluation of potential opportunities and partners. With more companies chasing each deal, in-licensing companies will inevitably have to expand their initial opportunity screening numbers in order to maintain the same levels of successful deals at the end of the process. As shown in Figure 3.14, more than 25% of surveyed licensing executives expect the average time taken to complete the licensing process to increase in 2006, with less than 20% expecting the time to decrease.
43
Figure 3.13: Expected change in number of potential partners chasing each licensing deal during 2006 100% Proportion of survey respondents
90%
Increase significantly
80%
Increase somewhat
70% 60%
Stay roughly the same
50% 40%
Decrease somewhat
30% 20%
Decrease significantly
10% 0% Pharma
Biotech
Other
Overall Business Insights Ltd
Source: Licensing trends survey, 2006
Figure 3.14: Expected change in the length of time required to complete a licensing deal during 2006 100% Proportion of survey respondents
90%
Increase significantly
80% 70%
Increase somewhat
60%
Stay roughly the same
50% 40%
Decrease somewhat
30% 20%
Decrease significantly
10% 0% Pharma
Biotech
Other
Overall Business Insights Ltd
Source: Licensing trends survey, 2006
44
In-licensing versus out-licensing While in-licensing or inward licensing is a very different proposition to out-licensing or outward licensing, the key stages through which companies must progress as part of the licensing process are similar. As shown in Figure 3.15, the licensing process always begins with a review of strategy. Setting out the strategic aims for licensing is essential in determining the direction and value benchmarks for subsequent phases of the process. Identifying opportunities means slightly different things to the in- and outlicensor, but involves a similar process of identifying and reviewing potential compounds/ technologies (for the in-licensor) or potential partners (for the outlicensor) in order to begin approaching and evaluating those opportunities.
Licensing valuations provide both a selection and deal structuring tool. Once a list of possible opportunities has been identified, some degree of evaluation must be undertaken in order to begin prioritizing those of greatest value in order to determine which opportunities should be pursued further or require more detailed scrutiny. Good licensing valuations also provide an essential input into the deal-making and agreement stage of the licensing process, helping to structure terms and values in order to reach a satisfactory agreement between licensing partners. Finally, it is important not to forget that the licensing process does not stop once signatures are added to a contract. The real hard work begins when agreements are being implemented and when problems arise that require solutions. In order to make sure all previous phases of the licensing process help to endorse good post-deal relations between licensing partners, it is important that deal outcomes are regularly fed-back to those charged with the responsibility for earlier phases of the licensing process. Licensing is an iterative process, whereby each deal provides lessons to inform better deal-making in the future.
45
Figure 3.15: The pharmaceutical licensing process In-licensing
Out-licensing
Where are the gaps? What can we offer?
Licensing strategy
What are our limitations? What are our needs?
What do we need? Where will we find it?
Opportunity identification
What have I got? Who will value it?
What is the value? What is the value to us?
Licensing evaluations
What is the value? What is the value to us?
What are my target/ acceptable risks, returns and responsibilities?
Deal-making and agreement
What are my target/ acceptable risks, returns and responsibilities?
What could be done better next time?
Post-deal management and analysis
What could be done better next time?
Business Insights Ltd
Source: Author’s research and analysis
46
Licensing strategy Deal failure is often the fault of one or more partners not clearly identifying their strategic aims from a licensing deal. This does not mean that all deals have to be led by top-down strategic management, nor does it mean that there is no room for opportunistic deal-making that arises out of serendipity, rather than rigorous strategic evaluations. However, companies must not enter into an agreement without having determined that the licensing opportunity satisfies a real and valuable set of objectives for the company, which from a portfolio perspective represents a sound investment when compared with the strategic alternatives.
Licensing presents an integral part of today’s corporate strategy for pharmaceutical companies. Drug development is a long-term game that involves significant rewards, but only after undertaking costly investments over a long time period with a great risk of failure. In the current pharmaceutical environment, where investors require returns in both the long and short term, but where much of the ‘low hanging fruit’ for drug discovery and development has been exploited, licensing provides companies, large and small, with ways to limit costs and risk and bring about new products or revenue streams more quickly.
Portfolio management will continue to be the catalyst for building a balanced portfolio and maintaining strategic alignment across therapy areas, geographies and time-scales. However, licensing has emerged as an effective tool for establishing alignment, particularly where changes and adjustments are required over a short time period. The unforeseen failure of a key product in phase III trials is difficult to compensate for with an injection of R&D investment into a given therapy area. However, the in-licensing of a similar late stage project or the out-licensing of the failed project for development in other indications or by a company with lower ‘success’ thresholds provide real compensatory actions that serve to balance the strategic aims of the company over a short time period.
47
Formulating a good licensing strategy is really a process of determining where you are as a company today and where you would like to be in the future. The difference between the two serves as a set of objectives for future licensing activity. Key areas of analysis involve:
Therapy area and physician profiles – understanding in which areas a company has strengths and deficiencies based on the current portfolio and pipeline of new drugs;
Resources and financing – determining the excess or shortfall in the amount of money and resource capacity available compared with the amount required;
R&D pipeline – entering new indication or supporting new brands with follow-up drugs by accessing either additional marketing expertise or additional products in specific markets;
Geographical coverage – maximizing the sales of key products across geographies, particularly those in which companies have no established in-house sales force;
Drug delivery/ line extensions – partnering with drug delivery companies in order to enhance the formulation of a product in order to extend its life-cycle;
R&D technology platforms – understanding current R&D capabilities and the potential productivity benefit of novel platforms, such as proteomics or nanotechnology.
Once a strategic review has been completed, the resulting licensing strategy needs to be presented to and agreed upon by senior management. The licensing strategy serves two purposes: it provides internal clarity on what needs to be done, but also provides potential partners with the reassurance that the company is committed to a clear strategy and is unlikely to under-resource either the licensing process or post-deal implementation.
48
Having a clear understanding of what you can and cannot offer potential partners is critical in order not to over promise or waste time negotiating the wrong deal with the wrong partner.
Opportunity identification Pharmaceutical licensing is best understood using the analogy of a dating agency. There are two key functions to perform – the presentation of yourself as a potential partner and the selection of appropriate profiles of potential partners. No matter how much effort is put into these two functions, the process for identifying licensing opportunities is not an exact science and is therefore fraught with inflated expectations and considerable disappointment. As with personal dating, pharmaceutical licensing suffers from an unfortunate externality resulting from the inflation of expectations and disappointment, which is that a highly skeptical opportunity screening process has emerged.
As in the ‘market for lemons’ – a microeconomics paper written by Nobel laureate George Akerlof about the market for second hand cars – if you cannot easily determine whether an opportunity or partner is a good one you inevitably assume it is a bad one. As such companies looking to present their compounds as good licensing opportunities or themselves as good licensing partners must find new ways to credibly elevate themselves above the mass of potential opportunities available for scrutiny. Saying you are young, attractive and blonde but not attaching a photo will not work in attracting a partner today. Companies need to demonstrate their strengths rather than simply listing them as a proof of suitability.
The process of opportunity identification is somewhat different for in-licensing and out-licensing.
49
In-licensing Companies looking to identify potential compounds/technologies for in-licensing must first determine an appropriate set of search criteria with which to produce a manageable list of potential opportunities. Having determined an agreed licensing strategy, broad criteria regarding specific market sectors, geographies and development stages will already provide the focus for the opportunity search. However, refinement of this list will take into account a more targeted set of criteria including:
Do we have the required R&D capabilities for this product?
Do we have the required sales and marketing capabilities for this product?
Is the product likely to be made available for licensing?
Are we likely to be considered to be an attractive partner for this product?
A full list of search criteria – once an initial selection has been limited to the required therapy area, geographies and development stages – are used to eliminate specific opportunities. These would include the following:
Specific therapeutic classes and indications;
Geographic availability;
Development stage and expected launch dates;
Maximum and minimum projected sales performance;
Key performance objectives (once-a-day dosing, low side effects etc);
Risk profile (first in class, reformulation etc).
Identifying potential opportunities can involve significant research, but should be carried out in a logical way using the criteria identified above. Sales and marketing data, available from IMS Health or through other aggregators of product sales information provides a useful tool for identifying potential opportunities. 50
Given that a high proportion of licensing occurs for development compounds, some of the best research sources for licensing opportunities are R&D databases such as LifeScience Analytics’ MedTRACK, Informa’s PharmaProjects, ADIS R&D Insight and IMS Health’s R&D Focus. Each has the ability to refine searches by indication, technology, development stage and company. In this way, a short list of suitable targets that meet these criteria can be formed. More detailed profiles for key drugs are also found in a number of the main databases which can further refine choices around sales projections and the likely risk involved with further development.
Aside from the desk research involving sales- and R&D-based database querying, the main source for identifying potential opportunities is the old-fashioned route of networking. There are a number of well-organized licensing network events and a number of websites, such as Pharmalicensing, that are now providing networking short cuts in order to pass-on details of available opportunities to potentially interested parties. It is still the major responsibility of any licensing or business development manager to form networks through which information about potential opportunities can quickly be shared. Above and beyond the direct networks between licensing and business development managers, there are significant roles to play for R&D managers and regional marketing teams. R&D managers, in particular, are exposed to many new developments and opportunities in the various therapy area-focused conferences and industry journals. These can often provide excellent sources for identifying new opportunities or for fleshing-out an early evaluation of a potential opportunity using the most up-to-date information.
In order to reduce the broad list of potential licensing opportunities to a more manageable size, a process of prioritization must be applied. A formal review stage is undertaken using a series of licensing opportunity profiles. These profiles present the relevant product information alongside status details for each of the key selection criteria. These standardized profiles then form the basis for a formal discussion regarding potential strategic fit, value and prioritization. As mentioned earlier, any determination of opportunity value must include an introspective look at whether you as a company can offer sufficient value to the partner to make it a valuable opportunity 51
to them too. The prioritization of opportunities must include a realistic determination as to the likely success of subsequent deal negotiations. A licensing opportunity profile for a development compound should include the following:
Drug name (brand/ chemical);
Originator/owner;
Planned indication(s);
Intellectual property status;
Abstract (history, mechanism of action, potential competitive benefits etc);
Development status (global/ regional);
Market profile (size, key players etc);
Clinical profile (patient size, clinical targets etc).
Out-licensing For out-licensing, companies must look to both identify and present their own opportunities (compounds/technologies) as well as identify partnership opportunities with other companies. Before preparing presentational material for a licensing opportunity, a company looking to out-license must determine when and how to go about it. Determining when to license a compound involves the consideration of many factors such as how much risk and financial burden is the company able to bear and how well equipped is that company to continue further development of a compound alone. Deciding how to out-license a compound is often subject to the decision over when to out-license. Licensing at an early stage of clinical development may involve some element of co-development, but is unlikely to include the co-commercialization of the compound once it reaches the market. However, a late stage compound in phase III development is likely to yield significant returns for the out-licensor and include both co-development and co-commercialization rights. Other types of ‘transactional’
52
licensing with limited collaboration often occur in early stage development or involve products been divested as part of post-merger anti-trust actions etc.
Having loosely determined the optimal window for licensing and the type of deal been sought, a team of qualified staff need to be brought together to draft the presentational licensing material or licensing prospectus. Usually this involves a non-confidential dossier that can be provided to any and all interested parties, and a more detailed confidential prospectus that is only shared after initial contact has been formed and the appropriate confidentiality agreements have been signed.
The confidential prospectus will include all relevant data available to support the product’s potential. For a development project this will focus on the expected clinical profile, the associated regulatory risks and the status of patent claims. The main sections to be included in the prospectus include:
Overview/summary – a one-to-two page summary presenting the key product features and potential value;
Therapeutic rationale – an outline of the mechanism of action and how and why this differs from other products;
Chemistry and pharmacy – an outline of the product’s chemistry including likely formulations and manufacturing processes;
Intellectual process – a review of the current patent position, expiry dates and any other intellectual property such as confidential information etc;
Experimental studies – a review of relevant data and observations from experimental work, in vitro and in vivo studies and any comparative data;
Clinical development – an outline of clinical objectives, pharmacology and study results including a detailed summary of completed pivotal studies;
Therapeutic potential – a review of the product’s commercial potential and how this links with a significant market opportunity; 53
Competitive potential – a detailed profile of the product’s expected dosing, tolerance and efficacy for each indication, as well as likely launch dates.
A references and bibliography section is often included to help support some of the claims made in the experimental data and clinical development sections and to provide further references to papers not referenced within the core text of the prospectus.
The non-confidential brochure provides a more concise product summary (five-to-six pages) aimed to encourage potentially suitable partners to make contact and find out more information. A good non-confidential brochure would also provide enough information to deter unsuitable partners from making contact and wasting time in providing further information. Aside from brevity, there are two further differences between the confidential prospectus and the non-confidential brochure. The first is that some sensitive data may be removed from the latter document, replaced simply with ‘…more information is available under confidentiality’. The second difference is that the wider audience exposed to the brief, non-confidential brochure will require a short introduction to the company at large and not simply the product available for outlicensing. The brochure should include a company section that outlines scientific and commercial competences as well as key contact details.
When preparing presentational materials for marketed products the key difference is the addition of more significant therapeutic and competitive potential sections. Market data will already be available and this will be reflected in a more robust commercial evaluation for the product in current or future potential markets.
The licensing brochure and prospectus form the primary mechanism for presenting a potential out-licensing opportunity to potential partners. The second licensing opportunity process involves the active identification of potential partners.
The first filter for identifying potential partners is to look for activity in the product’s target market sectors. Who is currently marketing, developing or in-licensing products
54
for the target therapy sector or physician group? Secondary considerations involve understanding which companies are strong in targeted regional markets, such as the US, Europe and Japan. The final, but often the most important screening, is determining the closeness of fit between the licensing opportunity and the potential licensing partner’s portfolio and pipeline strategy. Would the product add something of value? Would the out-licensing partner be able to add something of value? Ideally, this process should provide around 20 companies to actively target as preferred partners.
These potential partners must be contacted very carefully, as the wrong sort of introduction with the wrong message to the wrong person may prove disastrous to future communications and potential negotiations. Identifying opportunities includes identifying the right approach to the right person within the short list of target companies. Many small, but ambitious, biotechnology companies have managed to do a good job with the preparation of presentation materials and the identification of target licensing partners only to make a bad job of establishing a first contact with the company and as a result fail to make any further progress.
55
Licensing evaluations Evaluating licensing opportunities can fast become an overly detailed and sophisticated pursuit if the right parameters are not agreed beforehand. First of all, there are two levels of valuation. The first involves a selection of the most appropriate and valuable opportunities. For an in-licensor this is the most attractive product, compound or technology, while for an out-licensor this is the most attractive partner. Once the selection process is completed a second, more detailed evaluation supports negotiations and deal-making. This deal-making evaluation is not covered in this section of the report, but is presented in detail in the licensing valuations chapter.
Licensing evaluations for selecting appropriate opportunities to advance to a more detailed stage of negotiation and deal-making are similar to those found as part of an internal portfolio management process. Valuations need to be consistent and transparent in order to provide the relative ranking and prioritization required. Like portfolio management evaluations, the key trade-offs are to be found between sophisticated financial models and more simple rating models that include strategic elements such as portfolio fit and balance. Unlike portfolio management evaluations, potential licensing agreements involve an extra level of uncertainty, which is the contractual relationship with a third party. Understanding the relative value of an underlying licensed asset is not enough to make choices among different licensing opportunities. Some expectation as to the likely deal terms available for a licensing deal is required in order to compare opportunities. However, understanding a deal’s potential terms and financial metrics at an early stage of opportunity assessment can be difficult, particularly for complex licenses which might include multiple compounds.
56
General portfolio management Licensing evaluations need to provide relative values and rankings across all external opportunities, but need not necessarily provide consistent measures across both inhouse and external opportunities. However, it is desirable in a robust portfolio management process to include both internal and external potential projects in order to arrive at optimal resource allocation decisions. Parameters set-up to aid decisionmaking for internal projects and resources can easily be adapted for use in evaluating external projects.
Portfolio evaluation approaches allow decision-makers to establish preferences across projects by measuring them against the explicit objectives identified by senior management. As a result, project evaluations must include measurable criteria to assess the extent to which key objectives are likely to be satisfied. Identifying these value criteria is a key part of the project evaluation framework. Once criteria have been established, project evaluations can begin.
Project scoring approaches accommodate the fact that multiple criteria are often required in order to select and prioritize pharmaceutical projects effectively. Different scoring systems use different criteria, and include financial measures, strategic fit, competitive advantage, market attractiveness, the degree to which projects leverage core competencies, technical feasibility and risk versus return.
After a project has been scored against relevant criteria, a weighted score is calculated. This relative superiority or attractiveness score is used as a proxy for the value of the project. Once all projects have been evaluated, minimum threshold scores can be set in order to select and prioritize the most attractive projects and eliminate those of insufficient value.
Although project scoring is a key approach to project evaluation, and enables projects to be evaluated with respect to a number of different dimensions, the project rankings
57
derived from scoring can be misleading, which as a result can lead to sub-optimal project prioritization and decision-making. Scoring systems provide a means of capturing and quantifying project information in a consistent manner for future decision-making, but as a basis for discussion rather than a decision in itself.
Pharmaceutical companies develop approaches to portfolio management in order to allocate resources optimally with respect to both the minimization of risk and the maximization of value across the portfolio. Different projects are associated with different values and risk profiles. Although additional parameters can be incorporated, the key role for project evaluations is to provide a critical assessment of value and risk across the portfolio.
A broad range of different financial tools and evaluation techniques are available for use in project valuation. Financial analysis is able to combine an evaluation of a project’s value, risks and associated costs. However, financial approaches do not represent a viable standalone approach to project evaluation. Portfolio decision-making is characterized by the range of trade-offs made across different criteria, including meeting unmet medical need while maximizing cash-flows, balancing short-term and long-term performance, and protecting existing franchises alongside investments in new technologies. It is clear that these trade-offs cannot be adequately captured by project evaluations based solely on financial analysis. Instead, financial analysis must be employed alongside a broader scoring methodology enabling non-financial dimensions to be assessed and included in decision-making.
Although it is beyond the scope of this report, a 2005 Business Insights report, Pharmaceutical Portfolio and Project Management, discusses a range of available financial analysis tools and techniques, their application in project evaluation and their relative strengths and weaknesses. Key financial approaches include:
Net present value (NPV);
Decision tree analysis and expected NPV (ENPV);
58
Monte Carlo simulations;
Real options.
Applications for licensing evaluations As mentioned previously, the key difference between an internal and external opportunity is that a licensed compound’s agreement terms will determine how risks, returns and responsibilities will be shared between parties. It is only by including an evaluation of this package of contractual terms that a potential licensing opportunity can be evaluated. Other complications for evaluating licensing opportunities include the partnering risks brought about by third-party decision-makers and the limited availability of data for generating evaluations for external opportunities.
As a check list, some of the key intellectual property (IP) licensing evaluation questions include:
Who owns the IP rights (sole versus joint ownership, originator versus sub-license, exclusive versus non-exclusive etc.)?
Is the patent in force and valid (have maintenance fees been paid, is claim enforceable, are there any blocking patents etc.)?
What is the scope and length of the patent (how useful in preventing competitors, how does regulatory progress affect legal and economic lifespan of patent etc.)?
In evaluating a licensing opportunity it is optimal to first value the asset and then apply expected deal terms to bring about a company-specific value. As is shown in the licensing valuations chapter, a number of different parameters affect compound valuations. The costs, risks and returns associated with a compound all differ depending on the stage of development and therapy area of the specific compound. Further adjustments can be made if a specific peak sales figure can be forecast, while company specific characteristics such as size and experience can also impact on values. Consistent discount factors, applied for internal products, will allow for relative
59
rankings for licensing opportunities. Finally, expected deal terms, with respect to milestones, royalties, shared costs and responsibilities, can be applied in order to determine valuations for each licensing opportunity. These values, along with other more strategic considerations, provide the basis for selecting the opportunities of greatest potential value to the company.
Deal-making and agreement The valuations used as part of negotiations are outlined in more detail in the licensing valuations chapter. However, agreeing a successful deal involves more than a robust valuation and some persuasive negotiations. There are many pitfalls when agreeing terms for a pharmaceutical licensing agreement, a number of which are detailed below.
Key elements of a pharmaceutical license agreement The key sections of a pharmaceutical license agreement, such as preambles, recitals, definitions, grant clauses, releases, reservations and improvements, read more like a legal text book than a guide to pharmaceutical licensing. However, there are a number of key considerations expressed in these legal terms that require due consideration.
The preambles identify the respective parties, guarantors and active date of the agreement. Determining an early understanding as to which parties are signatures to the agreements, particularly where local, regional and corporate representatives are involved in early discussions, can save significant time and unnecessary confusion during negotiations. Recitals memorialize the background of a relationship and illustrate the importance of existing relationships and a reputation for successful licensing.
Grant clauses relate specifically to what is being agreed. What is the technology, product, field of use being licensed? What is the term of the agreement and over which territories? Finally, what is the specific right being granted – to sell, to use, to import?
60
Usually, grant clauses have been interrogated as part of an exchange of term sheets, but it is important that negotiations over deal terms are founded on explicit grant clauses.
Improvements in licensing contracts play an important role, particularly where ongoing relationships are anticipated. It is important that the ownership of any IP resulting from collaboration is made clear and that each party understands their obligations relating to protecting and promoting the licensed IP. It is particularly important that there is agreement over the assignment of obligations relating to the enforcement and protection of the licensed IP. If IP rights are challenged what happens to other terms of the agreement, such as royalty payments etc?
In light of all the potential pitfalls outlined above, pharmaceutical licensing continues to be one of the most complex and dynamic of legal pursuits. The good news is that there are a number of well-prepared legal firms available to help pharmaceutical companies negotiate their way through the complexities of the licensing agreement. The bad news is that, like much adversarial law, often the company with the best lawyer wins. If you are a small biotech looking to agree a blockbuster late-stage licensing agreement with a top 10 pharmaceutical company you can either spend money on a good lawyer now to negotiate and agree a favorable contract or spend double the money later down the line having to continually defend claims from your eventual licensing partner.
Post-deal management and analysis The sign of a successful strategic licensing agreement is not in the agreement and completion of the deal, but in the successful implementation of the deal terms over time. Similarly, the sign of a successful deal is not in the quality of the eventual product, over and above that which could have realistically been evaluated through due diligence, but in the quality of the relationship between licensing partners. It is clear, therefore, that much of the success in the licensing process actually occurs in the period after making an agreement. As such, pharmaceutical companies must work hard to 61
develop capabilities to help manage and support their alliances and collaborative licensing relationships.
Alliance management Alliance management is currently one of the hottest topics in pharmaceutical licensing. Companies across the pharmaceutical and biotech industries have spent the last ten years trying to determine whether it is a necessary capability and if so how it should be organized. To date, the results show very different approaches to this critical issue.
Eli Lilly were among the first to publicly promote their alliance management capabilities. Lilly established their Office of Alliance Management in 1998 as a corporate executive function looking to develop Lilly’s alliance strategy and provide a central point for alliance managers to learn and share experiences.
From an organizational perspective each Lilly alliance was assigned the following:
High-level alliance champion – vice president level or above responsible for meeting the partner counterpart to discuss strategic and governance issues;
Operational alliance leader – responsible for day-to-day management of alliance resources;
Alliance manager – responsible for providing a supportive and catalyst function, liaising between Lilly and its alliance partner.
AstraZeneca launched a collaboration management program for alliances in 2003. The collaboration management function is assigned with the responsibility for balancing AstraZeneca’s perspective with the perspective of external partners. The function provides a sense of continuity from deal-making through to implementation and is positioned within the corporate licensing department.
62
Alliance management responsibilities at Roche have been elevated to the senior management level and are integrated into every stage of decision-making. Roche’s global alliance management department combines key research, technical and commercial functions into a single team to create and manage alliances more effectively. Each partnership is assigned an alliance director who is charged with providing a consistent company perspective across the due diligence, deal signing and project execution phases.
The execution of alliances involves the management of two critical relationships. The collaborative relationship between partnering companies helps to provide effective alliance leadership, coordination and communication. However, the internal relationship between the licensing team and all other relevant functional departments is also a critical element in ensuring the successful implementation of each partner’s licensing objectives.
Pharmaceutical alliances fail when significant confusion or disconnection exists between groups within or between partnering companies. Internal groups must reach agreement on alliance objectives and on each group’s respective roles in achieving their objectives. Relevant functional teams must commit an appropriate share of their resources to the partnership and internal efforts must be coordinated and integrated both with the organization as a whole and with the efforts of the licensing partner.
The big test of alliance management capabilities is that when alliances fail for technical reasons do partners come back for more?
63
Using outside agencies The pharmaceutical licensing process requires significant resources in order to be completed successfully. Aside from the essential services provided by lawyers mentioned earlier, a number of other outside agencies provide services that can help complete the resource intensive or complex parts of the licensing process. Outside agencies offering appropriate services in support of pharmaceutical licensing include:
Management consultants;
Other specialist consultancies, including licensing or drug/marketing forecasting agencies;
Key investors, including venture capitalists and private equity funds for biotech;
Investment banks;
Accounting firms.
More than 50% of all surveyed licensing executives revealed that their companies identify potential licensing opportunities without the help of a third party agency, as shown in Figure 3.16. However, 70% of the biotech licensing executives relied on outside help to support the identification of licensing opportunities, turning to management and other consultancies, key investors and investment banks. The main recourse for pharmaceutical companies, where less than 40% of companies rely on outside help, are other specialist consultancies. These trends reveal the difference in available information for potential in-licensors and out-licensors. It appears that pharmaceutical companies, primarily comprised of potential in-licensors, are able to identify potential opportunities without significant outside help, due in large part to the promotional efforts of companies looking to out-license their compounds. For biotech companies, who are predominantly involved in out-licensing, it appears they require greater help in identifying potential partners for their technologies.
64
Figure 3.16: Parties involved in identifying potential licensing opportunities, 2006
70% Proportion of survey respondents
No outside parties 60% 50%
Management consultancies
40%
Other consultancies
30%
Key investors (venture capitalists etc)
20%
Investment banks
10% Accounting firms 0% Pharma
Biotech
Overall
Business Insights Ltd
Source: Licensing trends survey, 2006
More than 40% of the licensing executives surveyed for this report turned to third party providers when conducting due diligence exercises as part the ongoing evaluation efforts surrounding potential licensing opportunities. As shown in Figure 3.17, 60% of biotech companies used outside agencies to support licensing due diligence, primarily using specialist consultancies and accounting firms. As was the case for identifying licensing opportunities, only a small proportion of pharmaceutical companies turn to outside service providers to support their licensing due diligence.
65
Figure 3.17: Parties involved in conducting due diligence for potential licensing opportunities, 2006
70% Proportion of survey respondents
No outside parties 60% 50%
Management consultancies
40%
Other consultancies
30%
Key investors (venture capitalists etc)
20%
Investment banks
10% Accounting firms 0% Pharma
Biotech
Overall
Business Insights Ltd
Source: Licensing trends survey, 2006
More than 40% of surveyed licensing executives turned to third party service providers to support the valuation and negotiation of potential licensing deals, as shown in Figure 3.18. Outside agencies were used by more than 50% of biotech companies to support licensing valuations and negotiations. The main agencies used include specialist consultancies and key investors, such as venture capitalists. Again, around two-thirds of pharmaceutical companies were able to manage the valuation and negotiation of potential licensing agreements internally without outside help.
66
Figure 3.18: Parties involved in the valuation and negotiation of potential licensing deals, 2006
70% Proportion of survey respondents
No outside parties 60% 50%
Management consultancies
40%
Other consultancies
30%
Key investors (venture capitalists etc)
20%
Investment banks
10% Accounting firms 0% Pharma
Biotech
Overall
Business Insights Ltd
Source: Licensing trends survey, 2006
Based on the results of a survey of 142 licensing executives it appears that around 60% of biotech companies and 40% of pharmaceutical companies make use of outside agencies in supporting the licensing process. Biotech companies look towards specialist agencies and key investors, such as venture capitalists, in order to support the identification of licensing opportunities, subsequent due diligence and the valuation and negotiation of specific deals. Pharmaceutical companies appear to have greater levels of in-house licensing resource and expertise, with the majority completing the entire licensing process without outside help.
67
68
CHAPTER 4
Licensing valuations
69
Chapter 4
Licensing valuations
Summary
According to a survey of 142 licensing executives, pharmaceutical companies are more likely than biotech companies to share financial evaluations during licensing negotiations, with more than 80% of companies sharing at least a single point financial projection and almost 50% sharing a more detailed probabilistic evaluation.
The most common evaluation technique used in determining optimal deal terms is a discounted cash flow net present value (NPV) calculation, which is used in more than 70% of companies according to surveyed licensing executives.
A deal-making valuation model is one used specifically to agree licensing terms between partners. While the approach might be similar to that used to provide recommendations for selecting the most valuable licensing opportunities to pursue the outputs of the model are quite different.
If the inputs into the valuation model cannot be agreed upon by both licensing parties then the outputs of the exercise will not provide any common ground for negotiation. Similarly, if the modeling approach and any assumption are not clear and defendable then the effective use of the evaluation model for negotiation will be limited.
The first test as to whether two companies are going to be successful in collaborating together as part of a strategic licensing agreement is whether or not they are able to negotiate an agreement based on significant common ground. If two companies are unable to agree upon the true value of the licensing asset and subsequently on the fair distribution of risks, returns and responsibilities then there is little hope that they will be able to reach a satisfactory conclusion once the real work of implementing a licensing deal begins.
70
Introduction As mentioned in the licensing process chapter, there are two main reasons for evaluating licensing opportunities. One is to support the optimal selection amongst numerous opportunities and involves more simplistic and top-line valuations, often referencing a range of strategic as well as financial measures. The second reason is to support negotiations around reaching a licensing agreement and involves a more detailed financial evaluation around which specific deal terms can be agreed. The majority of this chapter will deal with the latter valuation, involving financial projections for deal-making. However, much of the approach taken can be applied to the financial element of valuations made for selecting appropriate licensing opportunities to pursue.
Valuing deals The valuation of a licensing opportunity is both an important and extremely difficult task that usually falls under the responsibility of a licensing manager more suited to making contacts and negotiations than to sitting in a dark room churning out financial projections. The problem with tasks that are both important and difficult is that they usually get handed over to specialists who neither understand the true importance of the task nor are able to fully deliver against the objectives set for the task. Those important objectives for a licensing valuation include:
Presenting an estimation of potential value for a deal;
Presenting an estimation of potential costs and risks associated with a deal;
Explaining in simple terms some of the interactions between risks and returns inherent in the deal;
71
Enabling the testing of different contract terms in order to see the impact for both partners;
Building an independent and unbiased model for representing deal values;
Producing a financial estimate that can be explained and defended by all key stakeholders.
It is the last two objectives mentioned above that represent the current challenge for licensing executives, who often have a financial model produced for them by some centralized financial or forecasting department. However, if a model cannot be shown to be both fair and clear in representing values used in negotiations with a potential licensing partner it does not provide a useful negotiation tool. It is important to make the distinction here between a valuation model used as an active negotiation tool and one that is used to set target and minimum acceptable deal terms.
Target and minimum acceptable deal terms are the established form of preparing negotiation parameters for pharmaceutical licenses. Usually, both potential partners produce their own separate valuation models, calculate target and minimum acceptable terms and then begin negotiating. Valuations are not shared explicitly, and once negotiations begin there is limited recourse to check new numbers against the valuation model. The problems with this approach are two-fold. Firstly the negotiations become needlessly adversarial. Often, there is much common ground between partners that does not require significant negotiation based on upper targets set by each party. Secondly, by taking separate approaches to valuing the deal, any discussion regarding common value becomes obsolete. The ‘win-win deal’ approach is very difficult to negotiate if one company values the deal at a distinctly different level than the other.
More and more pharmaceutical and biotech companies are beginning to move away from the “sit around a table and argue over numbers approach” to try tackling negotiations in slightly different way. The key to having more productive licensing negotiations is to try to come about an agreement over the total deal value prior to
72
negotiations. If this can be agreed upfront, then determining how that value is shared between partners, along with costs, risks and responsibilities, is a much easier exercise to embark upon.
The remainder of this chapter looks to set-out an approach to collaborative licensing valuations that provides both parties with a clear way of expressing how deal terms affect their share of deal values, costs and risks. The key to this approach is to use independent and unbiased estimates for the major valuation parameters and to keep all financial modeling at the simplest, most digestible level.
Current best practices Current practices for the sharing of licensing valuations during negotiations differ by the type of company involved, as shown in Figure 4.19. According to a survey of 142 licensing executives, pharmaceutical companies are more likely than biotech companies to share financial evaluations during licensing negotiations, with more than 80% of companies sharing at least a single point financial projection and almost 50% sharing a more detailed probabilistic evaluation. Around a third of all pharmaceutical companies share detailed decision-models during negotiations with simulated outcomes based on probabilistic parameters. For biotech companies, the picture is very different, with more than 30% of all companies failing to share any evaluation data at all during negotiations. A little over 10% of biotech companies divulge simulated decision-model outcomes during licensing discussions with a potential partner. However, only a few leading biotech companies are likely to be engaging in such sophisticated modeling techniques and therefore most do not have this information to share. As can be seen in Figure 4.20, only around 15% of biotech companies are engaged in simulation modeling in determining optimal licensing deal terms.
73
Figure 4.19: Information shared between partners during licensing negotiations, 2006
100% Proportion of survey respondents
90%
A decision-model simulating outcomes for both partners
80% 70%
A detailed probabilistic forecast model
60% 50%
A single point financial forecast
40% 30% 20%
No hard financial forecasts
10% 0% Pharma Biotech
Other
Overall
Business Insights Ltd
Source: Licensing trends survey, 2006
The most common evaluation technique used in determining optimal deal terms is a discounted cash flow net present value (NPV) calculation, which is used in more than 70% of companies according to surveyed licensing executives. The results shown in Figure 4.20 provide the proportions of respondents from companies utilizing the various evaluation approaches for licensing, with each separate technique able to be applied alongside another. While the use of NPV calculations are applied in similar proportions of companies in both the pharmaceutical and biotech segments, there are some small differences when it comes to the use of more sophisticated approaches. The uptake of ‘Monte Carlo’ simulations and real options is around 50% higher in pharmaceutical companies than it is in biotech companies.
74
Figure 4.20: Valuation techniques used in determining optimal licensing deal terms, 2006
Proportion of survey respondents
80% Simple product sales forecasts
70% 60%
Discounted cash flow NPV calculations
50%
Decision-tree expected NPV calculations
40% 30% 20%
`Monte carlo' simulation models
10%
Real option valuations
0% Pharma
Biotech
Overall
Business Insights Ltd
Source: Licensing trends survey, 2006
Interestingly, aside from the small difference in uptake of sophisticated modeling found between biotech and pharmaceutical companies, the utilization of modeling techniques to support licensing deal-making is similar. However, the sharing of this information with a potential partner is far greater for pharmaceutical companies than it is for biotech companies. This means that a greater proportion of biotech companies are withholding valuation data from their partners than is the case for pharmaceutical companies. The reason may be distrust between a biotech company and the often larger pharmaceutical partner. As was mentioned earlier, the company able to hire the best lawyer usually does disproportionately well during negotiations. One way to place a check on this imbalance for the biotech company is undertake extra evaluation work and keep it undisclosed in order to place a control on deal values as they evolve during negotiations. A cynic might also add that it may be possible that the quality of these biotech evaluations would not stand up to scrutiny from a more experienced licensing 75
protagonist and are therefore shielded from any unwelcome criticism. Either way, there is some work to be done for most companies, particularly those in the biotech segment, in preparing useful licensing evaluations to support the licensing negotiation process.
Deal-making valuation model A deal-making valuation model is one used specifically to agree licensing terms between partners. While the approach might be similar to that used to provide recommendations for selecting the most valuable licensing opportunities to pursue, the outputs of the model are quite different.
As mentioned earlier in this chapter, the reason to invest time in building and using a deal-making evaluation model is to help bring about a set of negotiated deal terms that maximize deal values for each partner. A ‘win-win licensing deal’ is an overused phrase in today’s pharmaceutical licensing lexicon, but the aim of every strategic licensing deal must be to find a way of maximizing deal values for both partners. A tough negotiator armed with an even tougher lawyer might be able to squeeze the last cent of value to the benefit of their company, but a reputation for leaving nothing on the table for their partners will have disastrous consequences on the company’s ability to make deals in the future.
A deal-making valuation model must demonstrate two important qualities:
Independent and unbiased inputs;
Clear and defendable outputs.
If the inputs into the model cannot be agreed upon by both licensing parties then the outputs of the exercise will not provide common ground for negotiation. Similarly, if the modeling approach and any assumptions are not clear and defendable then the use of the evaluation for negotiation will be limited. The first test as to whether two
76
companies are likely to be successful in collaborating together as part of a strategic licensing agreement is whether or not they are able to negotiate around common ground during the deal-making process. If two companies are unable to agree upon the true value of the licensing asset and on the fair distribution of risks, returns and responsibilities then there is little hope that they will be able to reach a satisfactory conclusion once the real work of implementing a deal begins.
Model inputs The best starting point for any dealing-making valuation model is to begin with a generic model based on independent industry inputs. Once this model has been completed a number of additional adjustments to the model can agreed upon by both licensing partners in order to bring about a more robust valuation of the licensed asset. However, these adjustments need to be both agreed by each party and make a relative reference to the generic inputs put together as part of the base model.
Given that most strategic pharmaceutical licensing agreements are formed during R&D, specific inputs relating to the cost, lead time and risk associated with pharmaceutical R&D must be determined. The industry numbers presented in Figure 4.21, Figure 4.22 and Figure 4.23 are taken from the Journal of Health Economics article ‘The price of innovation: new estimates of drug development costs’ written by Joseph A. DiMasi, Ronald W. Hansen and Henry G. Grabowski and published in 2003 (DiMasi, Hansen and Grabowski, 2003). The data was calculated using a random selection of 68 new drugs drawn from a survey of 10 pharmaceutical firms. The article gave a total pre-approved cost estimate of US$802 million in 2000 dollars, which takes into account the cost of all the failed projects required in earlier stages to get one successful product. This cost estimate has become widely established as the benchmark for R&D development costs in the pharmaceutical industry and the article’s R&D data are therefore considered to be a good estimate of cost, lead time and risk.
R&D out-of-pocket costs increase significantly as a product progresses through the various phases of clinical development, as shown in Figure 4.21. Long term animal
77
studies are considered to occur in parallel with phase I and phase II trials, being completed before a drug begins phase III trials. Post approval trials relate to lifecycle drug development including the development of a launched drug in new indications or new formulations, as well as any necessary post approval regulatory trial requirements.
Figure 4.21: R&D costs by phase, 2000
160
Cost (US$ million, 2000)
140 120 100 80 140.0
60 86.3
40 20 15.2
5.2
23.5
0 Phase I
Phase II
Phase III
Long-term animal
Post approval
Business Insights Ltd
Source: DiMasi, Hansen and Grabowski (2003)
As shown in Figure 4.22, the total lead time for a drug from entering human trials to being approved for launch is approximately 90 months, or 7.5 years. The greatest amount of development is spent in late stage phase III trials, but an additional 18 months are added once a marketing application is made before a drug receives final approval. These figures are based primarily on receiving approval from the Food and Drug Administration (FDA) in the US.
78
Figure 4.22: R&D lead times by phase, 2000
Mean time to next phase (months)
40 35 30 25 20 33.8
15 26.0
10 5
18.2 12.3
0 Phase I
Phase II
Phase III
Approval phase
Business Insights Ltd
Source: DiMasi, Hansen and Grabowski (2003)
Figure 4.23 shows the average attrition rates for a typical pharmaceutical product as it progresses through human testing and the approval stage. As would be expected, the greatest level of attrition occurs before a drug enters expensive phase III trials. At a late stage of clinical development only those drugs with the greatest chance of progressing through the stage and eventually being approved for marketing are entered into phase III trials.
79
Probability of entering phase (from previous)
Figure 4.23: R&D success probabilities by phase, 2000
80% 70% 60% 50% 40% 71.0%
68.5%
30% 20% 31.4%
10% 0% Phase II
Phase III
Approval
Business Insights Ltd
Source: DiMasi, Hansen and Grabowski (2003)
Alongside an analysis of R&D costs, lead times and risks, a drug valuation also requires a determination of likely sales returns for a pharmaceutical product. There are a number of different approaches to including sales forecasts for pharmaceutical drugs, but perhaps the simplest is to use a peak sales number and then apply a diffusion curve representing the likely uptake and eventual erosion of sales across the product lifecycle. The drug market diffusion curve shown in Figure 4.24 has been adapted from the Pharmacoeconomics article ‘Returns on research and development for 1990s new drug introductions’ written by Henry Grabowski, John Vernon and Joseph A. DiMasi published in 2002 (Grabowski, Vernon and DiMasi, 2002). The diffusion curve assumes a 14 year post-launch patent life and peak sales occurring in year 10. It is assumed that market sales erosion following loss of patent occurs at a rate of 50% per annum. The article estimated mean average peaks sales for a pharmaceutical drug of US$458 million in 2000 dollars. Mean average peak sales are skewed by the disproportional impact of the small number of blockbuster drugs generating significant
80
peak sales revenues, and as a result the mode (50th percentile) average peak sales are much lower than the mean. When including the direct costs associated with sales, a contribution margin of 45% is used in order to reflect both the cost of sales and direct selling costs (amounting to 55% of total sales). In order to reflect the lead time involved in marketing and sales, particularly around the launch period for a new drug, the direct costs for sales were calculated as a proportion of sales expected in the following two years. As a result, the very real scenario of investing significant marketing spend on a potential new product only to receive a final ‘non approvable’ letter from the FDA can captured by the model.
Figure 4.24: Drug market diffusion curve – product lifecycle 100% 97% 97% 94% 94% 88% 88%
100% Proportion of peak sales
90%
75%
80%
75%
70% 60%
50%
50% 38%
40% 25%
30%
19%
20%
13%
9%
10% 3% 6%
5%
2% 1%
0% 1
2
3 4
5
6
7 8
9 10 11 12 13 14 15 16 17 18 19 20
Year (post-launch)
Business Insights Ltd
Source: Author’s research and analysis
81
Evaluation modeling The first stage of evaluation modeling is to transpose all input values into current values. In order to do this we first have to arise at 2006 values and then account for any changes in the real cost and value of real inputs. This will allow evaluation results to be expressed in a common value measure (namely 2006 US dollars).
In order to account for the changes in R&D costs and likely drug values between 2000 and 2006, key trends in costs and values were extrapolated from the two key source papers – DiMasi, Hansen and Grabowski, 2003 and Grabowski, Vernon and DiMasi, 2002. Over the ten year period leading up to 2000 it was shown that clinical development costs have increased at a rate of 11.8% per annum in constant 2000 dollar values. It is clear that clinical trials have gotten larger over the past 10 years or so and this increase in costs seems broadly reasonable. Over the same ten-year period, average peak drug sales have increased at a rate of 2.9% per annum in constant 2000 dollar values. The smaller rise in sales compared with drug development costs seems reasonable given the continued rise of healthcare cost containment measures across major markets and the impact of aggressive generic competition on prices.
In order to arise at 2006 cost and revenue levels a general US dollar inflation rate must also be included alongside the price increases in constant dollar amount, and was estimated to be a further 3% per annum. For example, the 2006 cost of a phase I trial expressed in 2006 dollars will involve calculating 6 years of price increases as a result of increased trial costs (11.8%) as well as accounting for the change in dollar value between 2000 to 2006 dollars (3% per annum). The resulting sum is as follows:
2000 phase I trial cost (2000 dollars) x (1 + 11.8% + 3%) ^ 6 = 2006 phase I trial cost (2006 dollars)
82
Table 4.1: R&D cost by phase and peak sales, 2006 (expressed in 2006 dollars) 2000 (expressed in 2000 dollars)
Year-on-year increase in cost/ revenue
Year-on-year increase in dollar value
2006 (expressed in 2006 dollars)
Phase I Phase II Phase III Long term animal Post approval
15.2 23.5 86.3 5.2 140.0
11.8% 11.8% 11.8% 11.8% 11.8%
3.0% 3.0% 3.0% 3.0% 3.0%
34.8 53.8 197.5 11.9 320.5
Peak sales
458.0
2.9%
3.0%
645.1
R&D costs
All figures in US$ millions Business Insights Ltd
Source: Author’s research and analysis
Once the 2006 levels for all key costs and revenues have been established the real increase in these values over time must also be added to the model. Costs and revenues remain in 2006 dollar values but the real increases in costs and revenues (11.8% and 2.9% per annum respectively) must also be extrapolated. In this way, all costs and revenues are expressed in a common, constant value measure and reflect the real levels of increase in costs and values that will occur over time. In this way, all figures can be added together to represent true total value, without needing to discount values in order to adjust for inflation.
It is important to point out that the base case model must involve cost and revenue projections that account for real increases over time and are expressed in a common value measure (e.g. 2006 dollars) in order to allow a real valuation aggregation. Most approaches fail to separate inflation from a general level of nominal discount rate and the resulting model is needlessly complicated. By having a simple model accounting for real price increases only there is no need to discount values to begin calculating value. The baseline valuation for a licensing deal should have a real discount rate of 0%.
83
When calculating net present values, future values are discounted in order to represent the opportunity cost of investing in an alternative but similar investment. It is important to recall that net present value (NPV) measures are comparator measures and not actual values. If for a pharmaceutical investment I normal expect a rate of return equivalent to 10% per annum, then discounting a pharmaceutical investment by the 10% factor in order to arrive at an NPV only represents the value of the project over and above that expected on average from similar investments. If the resulting NPV for a project is a loss of US$10 million, this does not mean the project has no value, but rather the project is likely to return US$10 million less than the opportunity cost of capital (an average investment in a similar project). NPV is a decision-making tool rather than a valuation tool. It is better to allow a pharmaceutical company to compare different internal rates of return for projects than to simply reply on a comparison of value with a mythical alternative investment project (opportunity cost). This separation between valuing a project rather than ranking it amongst its comparators is particularly important when dealing with licensing projects, given that the discount rate used by one company can be quite different the one used by another.
It would be difficult to be more provocative than stating that net present value is the wrong tool for valuing licensing agreements, particularly given that we have established that more than 70% of companies are using NPV to support licensing evaluations. However, potential licensing partners need to relieve themselves from the need to discount value and instead look at the real value being created, adjusted for risk, and then determine how this might be shared between parties.
As well as adjusting for price inflation, the main reason for discounting values is to account for risk – the greater the risk, the greater the discount rate threshold. However, given that most approaches to pharmaceutical drug evaluations make use of a decisiontree or expected value approach it appears better that specific development phases are actually discounted based on their technical risk rather than using a common projectwide discount rate across all phases. In this way, different outcomes (e.g. failure in phase I, success in phase I but failure in phase II etc) can be weighted by their
84
likelihood and eventual outcomes can be used to calculate a weighted average of all possible outcomes based on the risk of project failure.
The likelihood of specific outcomes for a new phase I, phase II and phase III drug are shown in Figure 4.25. A new phase I drug has only a 15% chance of reaching the market, while a new phase III drug has a 69% chance of reaching the market. It is clear that an NPV projection with a common discount rate would do a bad job in capturing the differences between these two investments.
Figure 4.25: Likelihood of outcomes for new phase I, phase II and phase III drugs
80%
Likelihood of outcome
70% 60% 50% 40% 69%
69%
30% 49%
20% 32%
29%
10% 7%
10%
15%
22%
0% Phase I only
Phase 2 only New phase I
Phase 3 only
New phase II
Launch
New phase III
Business Insights Ltd
Source: Author’s research and analysis
The non-discounted real values for each potential drug outcome are shown in Figure 4.26. Obviously, only those outcomes that eventually result in a successful drug launch have a positive real value in constant 2006 US dollars. When weighted against the likelihood of different outcomes, as shown in Figure 4.25, the expected real values for new phase I, phase II and phase III drugs can be calculated. Based on these expected 85
outcomes a new phase III drug is around four times as valuable as a new phase II drug and around seven times as valuable as a new phase I drug. Given that the risks associated with drug development are accounted for in the expected value calculations and that any market risks have been averaged out into a single point peak sales forecast, the resulting valuation represents the expected additional value added by the licensing opportunity (e.g. US$271 million for a new phase I drug) in constant 2006 US dollars.
Figure 4.26: Expected real values (non-discounted) for new phase I, phase II and phase III drugs
Drug value (US$ million, 2006)
3,000 2,500 1,797
2,000 1,500 1,000
464
500
271
0 -500 Phase I only Phase 2 only Phase 3 only
New phase I
New phase II
Launch
Expected value
New phase III
Business Insights Ltd
Source: Author’s research and analysis
In order to illustrate the impact of different discount rates on values, the expected discounted real values for drugs at different development stages are shown in Figure 4.27. These results show that at a real discount rate of 6% the NPV of a new phase I drug falls below zero, while at a discount rate of 8% the NPV of a new phase II drug falls below zero. Given that pharmaceutical companies often base R&D investment decisions on a real NPV discount rate of around 8-10% these numbers clearly do not 86
stack up. Either the NPV model fails to accurately capture the risk associated with pharmaceutical R&D or some of the base model inputs would need adjusting to ensure the results appear reasonable.
Expected drug value (US$ million, 2006)
Figure 4.27: Discounted expected real values for new phase I, phase II and phase III drugs
1,800 1,600 1,400 1,200 1,000
phase 3
800
phase 2
600
phase 1
400 200 0 -200 0%
1%
2%
3%
4%
5%
6%
7%
8%
9% 10%
Real discount rate (annual)
Business Insights Ltd
Source: Author’s research and analysis
For example, in order to ensure the model results appear more reasonable, the real annual increase in clinical development costs could be reduced. Figure 4.28 shows the impact on the discounted expected real drug values of changing the rate at which costs escalate over time. If the rate of increase in the real cost of clinical trials is reduced from 11.8% to just 3.0%, a similar rate to the increase in likely revenues, the impact on NPV calculations is significant. Both new phase I and phase II compounds show positive NPVs with real discount rates of up to 10%. This check for reasonableness can help to inform modeling and bring about agreement between licensing partners. However, for the purpose of this illustrative example the rate of clinical cost increase has been kept at 11.8% per annum. 87
Expected drug value (US$ million, 2006)
Figure 4.28: Discounted expected real values for new phase I, phase II and phase III drugs (adjusted for lower R&D cost inflation)
2,750 2,500 2,250 2,000 1,750 phase 3
1,500
phase 2
1,250
phase 1
1,000 750 500 250 0 0%
1%
2%
3%
4%
5%
6%
7%
8%
9% 10%
Real discount rate (annual)
Business Insights Ltd
Source: Author’s research and analysis
Model outputs While an understanding of total deal values are generated by the base case model, negotiations around licensing agreements are informed by looking at how that value is shared between potential partners with respect to risks, returns and responsibilities. The first model outputs are shown in Figure 4.29 and Figure 4.30 and outline all potential cash flows for each partner over the lifecycle of the licensing agreement. As an example, a new phase II product has been used to illustrate the potential returns for each partner. The deal terms modeled involve an upfront payment to the out-licensor of US$20 million, a milestone payment of US$40 million for successfully completing phase II trials, a further US$60 million for completing phase III trials and US$100 million for gaining regulatory approval. Upon the drug reaching the market, royalties based on gross sales of 10% are payable to the out-licensor.
88
Figure 4.29: Deal outcomes for out-licensor 120 Out-licensor license payments/ royalties Out-licensor costs/ sales
Cost/ sales (US$ million, 2006)
100 80 60 40 20 0
1 2 3 4 5 6 7 8 9 1011 12 13 1415 16 17 1819 20 21 2223 24 25 2627 28 -20
Year
-40 -60 Business Insights Ltd
Source: Author’s research and analysis
Figure 4.30: Deal outcomes for in-licensor 600 In-licensor costs/ sales
Cost/ sales (US$ million, 2006)
500 400
In-licensor license payments/ royalties royalties
300 200 100 0
1 2 3 4 5 6 7 8 9 10 1112 13 1415 16 1718 1920 21 2223 2425 26 2728 -100 Year -200 -300 Business Insights Ltd
Source: Author’s research and analysis
89
Understanding potential cash flows over the lifecycle of the licensed compound does not inform each party as to its expected share of potential value. As with the total licensing opportunity evaluation, expected values are calculated using a weighted average of all potential project outcomes. Using the same licensing terms as outlined above the respective shares of expected licensing value can illustrated, as shown in Figure 4.31. These expected values allow for a negotiated agreement over the sharing of risk and expected values. In the illustrative example the maximum out-of-pocket cash flow for the in-licensor is more than double that of the out-licensor. However, in order to reward bearing a greater share of development risk the out-licensor is able to secure a greater share of the sales returns for a successful product and as a result a greater share of the expected overall licensing agreement value.
Figure 4.31: Share of expected deal outcomes by partner
Drug value (US$ million, 2006)
1,600 1,400 1,200 1,000 800 600 400 200
200 264
0 -200 -400 Phase 2 only
Phase 3 only Out-licensor
Launch
Expected value
In-licensor
Business Insights Ltd
Source: Author’s research and analysis
90
Having an understanding of both potential licensing cash flows and expected, riskadjusted licensing values provides a benchmark for constructive licensing negotiations. Agreeing on broad, independent value parameters and using a flexible but clear approach to modeling licensing opportunities helps potential partners to facilitate dealmaking negotiations around common estimates of value and not around pre-determined target deal terms prepared in isolation by each party.
As a final word on deal-making licensing valuations, it is important to generate ‘buyin’ for the collaborative valuation and negotiation process as early as possible. Ideally, an independent valuation should be provided by the out-licensor as part of any detailed licensing prospectus and any challenges to the model’s validity should be dealt with as a priority. A collaborative valuation will only prove to be a useful negotiation tool where both licensing parties have confidence in the valuation and are fully committed to the process.
Model refinements A base case valuation model is used as a starting point to arrive at mutually agreeable licensing deal valuations based on unbiased sources. Single point, largely unsubstantiated commercial claims by out-licensors are always going to be ignored, or at best challenged, by in-licensors. It is important, therefore, for companies to exhaust all industry-wide valuation parameters before beginning to add drug-specific data relating to the potential claims of the drug. As has already been shown in the base model, the likely valuation of a drug varies significantly by development stage. However, other key value parameters include therapy area and company profile.
R&D costs, lead times and risks all vary by therapy area, as shown in Figure 4.32, Figure 4.33 and Figure 4.34. Therapy area specific data for analgesic/anesthetic, antiinfective, cardiovascular and central nervous system (CNS) drugs were presented in the Drug Information Journal article ‘R&D costs and returns by therapeutic category’ written by Joseph A. DiMasi, Henry G. Grabowski and John Vernon and published in 2004 (DiMasi, Grabowski and Vernon, 2004). R&D costs vary significantly, with 91
phase III trials for anti-infective drugs (which include large scale HIV and hepatitis drug trials) more than twice as expensive as average analgesic/anesthetic and CNS drug trials.
Development costs (US$ million, 2000)
Figure 4.32: R&D costs by phase by therapy area, 2000
160 140 120 100 80 60 40 20 0 Phase I All
Phase II
Analgesic/ Anesthetic
Anti-infective
Phase III Cardiovascular
CNS
Business Insights Ltd
Source: DiMasi, Grabowski and Vernon (2004)
Drug development lead times are greatest for CNS drugs, which are almost double those found for analgesic/anesthetic and anti-infective drugs, as shown in Figure 4.33. Both CNS and cardiovascular drugs take significantly more time for regulatory review, largely as a result of the increased complexity of many of the indications involved in these therapy areas.
92
Mean time to complete phase (months)
Figure 4.33: R&D lead times by phase by therapy area, 2000
140 120 22.1
100 80
18.2 21.0
60
12.5
15.4
92.5
40
72.1
20
61.0
46.4
50.5
Analgesic/ Anesthetic
Anti-infective
0 All
Clinical phase
Cardiovascular
CNS
Approval phase
Business Insights Ltd
Source: DiMasi, Grabowski and Vernon (2004)
The probabilities of progressing through clinical phases do not vary significantly between therapy areas, as shown in Figure 4.34. However, phase II anti-infective drugs are slightly more likely to progress through into phase III trials than drugs from other therapy areas, while phase III analgesic/anesthetic drugs are more likely to receive final approval than other drugs.
93
Figure 4.34: R&D success probabilities by phase by therapy area, 2000
Probability of entering phase (2000)
90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Phase II All
Phase III
Analgesic/ Anesthetic
Approval
Anti-infective
Cardiovascular
CNS
Business Insights Ltd
Source: DiMasi, Grabowski and Vernon (2004)
Figure 4.35: Peak sales and year of peak sales by therapy area, 2000 Year 9
Peak year sales (US$ million, 2000)
900 Year 12
800 700 600 Year 10
500 Year 9
400 300 Year 6
200 100 0 All
Analgesic/ Anesthetic
Anti-infective
Cardiovascular
CNS
Business Insights Ltd
Source: DiMasi, Grabowski and Vernon (2004)
94
A range of peak sales by therapy area are also presented in DiMasi, Grabowski and Vernon, 2004. The greatest peak sales revenues are generated by CNS drugs, which generated sales almost twice that of pharmaceutical drugs on average, as shown in Figure 4.35. Cardiovascular peak sales occur slightly later in the product lifecycle to other drugs, occurring in year 12. Peak sales for cardiovascular drugs are almost twice the sales for an average anti-infective drug.
As shown in Figure 4.36, CNS drug sales are highly skewed by a small proportion of CNS drugs generating significant revenues, including multiple blockbuster products for the treatment of depression and psychosis. Cardiovascular drugs appear to promise the greatest minimum value, generating the highest NPV at the 25th percentile level.
Figure 4.36: Discounted value of sales by therapy area, 2000
NPV sales (US$ million, 2000)
8,000 7,000 6,000 5,000 4,000 3,000 2,000 1,000 0 Mean All
25th percentile
Analgesic/ Anesthetic
Median
Anti-infective
75th percentile Cardiovascular
CNS
Business Insights Ltd
Source: DiMasi, Grabowski and Vernon (2004)
95
Further differences by therapy area were presented in the Journal of Health Economics article ‘Productivity in pharmaceutical-biotechnology R&D: the role of experience and alliances’ written by Patricia M. Danzon, Sean Nicholson and Nuno Sousa Pereira and published in 2005 (Danzon, Nicholson and Pereira, 2005). Data taken from over 900 pharmaceutical firms showed that the predicted probability that an indication in phase III trials will be approved by the FDA is 22 percentage points below the sample average for CNS drugs and 21 percentage points above the average for hormonal preparations. For drugs in phase I, respiratory indications have the lowest predicted probability of being approved (30%), with hormonal preparations having the highest probability of success (78%).
Danzon, Nicholson and Pereira, 2005 also presented some key trends relating to the experience of the developing company. The likelihood of phase II success for a company that has previously participated in only one phase II trial is 69%. However, this likelihood of success increases to 81% for a firm with experience in 25 phase II compounds. The likelihood that an indication will complete phase III trials successfully for a company that has previously participated in only one phase III trial is just 51% compared with 81% for a firm that has developed 30 phase III drugs. There were no discernable trends between experience and success for phase I trials, which is perhaps consistent with the idea that experience only really impacts on large, complex trials in phase II and phase III, which require perfecting dosing and generating statistical evidence for efficacy across large patient samples.
Danzon, Nicholson and Pereira, 2005 presented further trends relating to the overall effect on development success for drugs developed within an alliance rather than by one single company. The results showed that indications developed in an alliance are no more likely to progress through phase I trials than those developed independently. However, alliance drugs are significantly more likely to complete phase II and phase III than those developed by a single company. In phase II, the likelihood of an alliance drug progressing through to phase III is 9 percentage points higher than when an
96
indication is developed independently. In phase III, co-developed indications have a 14 percentage point higher likelihood of success.
97
98
CHAPTER 5
Licensing best practices
99
Chapter 5
Licensing best practices
Summary
With no two drugs or partnerships the same, there are no hard-and-fast rules for successful pharmaceutical licensing. Lessons can be learned by looking at the leading deals and deal-makers, but the application of best practices must always be directed by the specifics of the deal, rather than reverting to a list of generalized benchmarks.
The leading strategic pharmaceutical licensing deal, mentioned by the greatest number of surveyed licensing executives, is the ongoing relationship between Roche and Genentech. Over the course of the 16 year relationship Roche has acquired non-US marketing rights to a range of Genentech products, including Rituxan (rituximab) in 1995, Herceptin (trastuzumab) in 1998 and Avastin (bevacizumab) in 2003.
Novartis was considered to be the in-licensing partner of choice in 2006. The company has worked hard to position itself as a preferred licensing partner, employing a structured process providing a quick evaluation of opportunities and the early involvement of senior management to expedite decision-making. As part of this standard review process, Novartis employs a single gateway for all opportunities allowing for improved coordination and contact management.
Cephalon was considered to be the out-licensing partner of choice in 2006. Cephalon’s licensing model of acquiring rights to marketed products in order to generate revenues cash flow to help fund the development of in-house products appears to have worked well. More importantly, as an out-licensor, the company has been willing to share promotional rights for its lead products, particularly in the US and Japan, in order to maximize the returns from its internal assets.
100
Introduction Licensing best practices provide guidance and benchmarks for successful deal-making. However, with no two drugs or partnerships the same, there can be no hard-and-fast rules for pharmaceutical licensing. Lessons can be learned by looking at the leading deals and deal-makers, but the application of best practices must always be directed by the specifics of the deal, rather than a list of generalized benchmarks. An understanding of key licensing trends, the major objectives underpinning licensing processes and the success stories to be found in recent licensing deals will help to inform and support strategic licensing. However, understanding the most important objectives for each deal-making process will always be the value added by a capable and experienced licensing manager.
Top licensing deals of the 21st century The following four deals, all mentioned as leading strategic licensing agreements by surveyed licensing executives, outline trends towards long term, multi-product alliances and a shift in power towards the biotech out-licensor. The Genentech-Roche and Idenix-Novartis agreements both illustrate the profound effect a long term agreement can have on the pipeline of the in-licensor and the financial security of the out-licensor. The Millennium-Ortho Biotech and AstraZeneca-AtheroGenics deals illustrate the new found confidence of biotech companies with strong late stage products to hold out for the best possible deal.
Genentech-Roche The leading strategic pharmaceutical licensing deal, mentioned by the greatest number of surveyed licensing executives, is the ongoing relationship between Roche and Genentech. First signed in 1990, and renewed in 1995, a licensing agreement gave Roche a 10-year option to acquire the ex-US rights to Genentech’s developmental 101
drugs once they have completed phase II trials. As a result of this relationship Roche has acquired Rituxan (rituximab) in 1995, Herceptin (trastuzumab) in 1998 and Avastin (bevacizumab) in 2003.
In 1999, Roche exercised its option to buy Genentech outright. However, this position was soon reversed and by the beginning of 2000 Roche had sold 40% of Genentech’s shares back to the public. Aside from turning a profit of around US$2 billion through the acquisition and divestment, the decision to keep Genentech an independent company has remained the critical success factor for the 16-year agreement. Independence in a biotech company, both in management and the freedom to form partnerships with other companies, is critical for maintaining a creative and innovative partner for new drug development.
Idenix-Novartis Another licensing deal of note, mentioned by several surveyed licensing executives, is the 2003 agreement between Novartis and Idenix. Primarily formed around two clinical stage hepatitis B products, the deal also included the right for Novartis to access all suitable Idenix drugs reaching phase II trials, similar to the Roche-Genentech deal. Significantly, Novartis was able to use this deal to rapidly accelerate the development of its antiviral therapeutic franchise. Successful products arising out of the deal will be co-promoted by both partners in the US and Europe. As well as providing Idenix’s private shareholders with significant funds (51% of the company was acquired for US$255 million), the deal also promised significant milestone payments to Idenix in order to fund the continued development of the two hepatitis B drugs along with further development activity in other areas.
Millennium-Ortho Biotech A recent deal that truly illustrates the new-found bargaining power of biotech companies with late stage compounds is the 2003 deal between Millennium and Ortho Biotech (Johnson & Johnson) for cancer therapy Velcade (bortezomib). Millennium managed to retain all US commercialization rights, as well as double digit royalties for
102
European sales and up to US$500 million in upfront and milestone payments. Millennium maximized its returns from Velcade by advancing development and bringing multiple potential partners to the table. Once detailed negotiations with Ortho Biotech were underway, Millennium was then able to find a sophisticated deal structure to mirror the specific business objectives of each partner.
In late 2002, Millennium was speaking to more than twenty different interested parties about Velcade, conducting CEO-level talks with more than ten different companies. Having narrowed down the list of potential partners in early 2003, Millennium collected five term sheets and three deal structures. Following Velcade’s approval in May 2003, another two parties re-established their interest. It was against this significant level of competition that Ortho Biotech was able to secure the deal.
Millennium considered three different deal structures, the first of which was a copromotion agreement in the US alongside a profit-sharing arrangement on Velcade and one of the partner’s products. The second deal involved co-promotion in the US with Millennium retaining over 50% of profits. A third deal structure, centering on partnering for the ex-US rights only, was finally selected in an attempt to retain medium to long term upside as well as balance the strategic goals of both partners. Not only did the deal result in a clear split between marketing rights, but the agreement also made a clear distinction between future development responsibilities which were divided by tumor rather than by geography.
AstraZeneca-AtheroGenics Another recent example of a biotech company holding out for a good deal is AtheroGenics’ recent 2005 agreement with AstraZeneca for atherosclerosis treatment AGI-1067. The anti-inflammatory cardiovascular drug was already in phase III trials when the agreement was made, which could be worth up to US$1 billion plus royalties. The deal comes after AtheroGenics reacquired full rights for the drug back in 2001, following an earlier agreement for AGI-1067 with Schering-Plough in 1999. The agreement includes potential royalty rates in the mid-30s depending on cumulative
103
sales levels as well as a 125-person dedicated sales force funded by AstraZeneca to copromote the drug.
Preferred licensing partners Leading in-licensing companies A survey of 142 licensing executives revealed Novartis to be the in-licensing partner of choice in 2006, as shown in Figure 5.37. Other leading in-licensing partners include Pfizer, Johnson & Johnson (J&J) and Roche. Interestingly, Amgen was the fourth ranked in-licensing company, illustrating the biotech company’s affinity with smaller biotech companies as a preferred alternative partner to big pharma.
Figure 5.37: In-licensing partner of choice, 2006 Number of survey respondents Novartis
14
Pfizer
9 7
J&J Amgen
6
Roche
6
GSK
5
Eli Lilly
4 3
AstraZeneca 2
Merck Sanofi-Aventis
1
Business Insights Ltd
Source: Licensing trends survey, 2006
104
Novartis Aside from the Idenix agreement detailed above, Novartis has stepped up its licensing activity over the past five years. Recent deals signed in 2006 include:
An agreement with Human Genome Sciences for hepatitis C interferon drug Albuferon, set to enter phase III trials in late 2006;
An agreement with Servier for a novel phase III antidepressant.
In 2005, Novartis signed deals with:
Senju Pharmaceutical for an early stage glaucoma treatment;
SeBo for an early stage renal disease treatment;
Arrow Therapeutics for an early stage respiratory syncytial virus (RSV) treatment;
Anadys Pharmaceuticals for an early stage hepatitis C treatment;
Arakis and Ventura for a phase II chronic obstructive pulmonary disease (COPD);
Otsuka for a phase III treatment for dry eye.
Novartis has worked hard to position itself as a licensing partner of choice. Its website includes a downloadable presentation explaining the company’s key qualities as a licensing partner. Figure 5.38 shows the structure of the business development and licensing team at Novartis and Figure 5.39 shows the licensing process as described in the company’s ‘Novartis: Your partner of choice’ presentation1. The website also includes a clear explanation of the process for licensing employed at Novartis, highlighting a structured process providing a quick evaluation of opportunities and an
1
‘Novartis: Your partner of choice’ presentation,
http://www.novartis.com/downloads_new/bizdevelopment/Novartis_Presentation.ppt
105
early involvement of senior management to expedite decision-making. As part of this standard review process, Novartis employs a single gateway for all opportunities allowing for improved coordination and contact management.
Figure 5.38: Business development and licensing department, Novartis Global BD&L V. Hartmann USA, EU Latam, Asia
Japan I. Ohhashi
Finance M. Ceulemans
Project Management J.M. Séquier
Research Alliances S. Strub
Development Alliances/ Out-Licensing I. Csendes
Search/Evaluation G. Cupit
Mature Business A. Hörning
Primary Care S. Saxena
Partnering H. Girsault
Ophthalmology S. De Vries
Alliance Management D. Weston
Transplantation M. Grannatt
Oncology G. Golumbeski
Anti-Infectous M. Grannatt
Business Insights Ltd
Source: ‘Novartis: Your partner of choice’ presentation
Figure 5.39: Licensing process at Novartis
Preselection
Initial Technical Evaluation
Full Evaluation
Negotiation
S&E Negotiation Alliance Management Source: Novartis: Your partner of choice’ presentation
106
Business Insights Ltd
Leading out-licensing companies A survey of licensing executives revealed that the leading out-licensing partner in 2006 is Cephalon, as shown in Figure 5.40. However, the survey selections of leading outlicensing partner were dispersed across a wide range of companies, with no single company receiving more than 3 votes. Other leading out-licensing partners include Ablynx, Amylin, Genentech and UCB.
Figure 5.40: Out-licensing partner of choice, 2006 Number of survey respondents 3
Cephalon Ablynx
2
Amylin
2
Genentech
2
UCB
2
Business Insights Ltd
Source: Licensing trends survey, 2006
Cephalon Cephalon has an interesting business model, based on limiting risk through licensing agreements and acquisitions. The company has both in-licensed and out-licensed the rights to pharmaceutical products in order to facilitate its business model. From an inlicensing perspective, Cephalon has acquired rights to Trisonex (arsenic trioxide) from Cell Therapeutics in 2005, the rights to Gabitril (tiagabine) from Sanofi-Synthelabo in 2002 and the rights to Actiq (fentanyl) from Elan in 2002 after acquiring Anesta in 2000.
Experience from being on the in-licensing side of multiple licensing agreements has led to a number of recent examples of successful out-licensing deals with development and marketing partners. In 2006, Cephalon signed an agreement with Takeda
107
Pharmaceuticals North America in order to add 500 Takeda sales reps to co-promote narcolepsy treatment Provigil (modafinil) in the US. The agreement included an option to extend the relationship to include new treatment Nuvigil (armodafinil). A similar agreement signed with McNeil Consumer and Specialty Pharmaceuticals in 2005 provides co-promotion support for Attenace (modafanil) in preparation for the approval of a pediatric label for attention deficit/hyperactivity disorder (ADHD). A 2003 agreement with Tanabe Seiyaku provided the company with a license to promote Actiq (fentanyl) in Japan.
Cephalon’s licensing model of acquiring rights to marketed products in order to generate revenues to fund the internal development of in-house products appears to work well. More importantly, as an out-licensor, the company has been willing to share promotional rights for its products, particularly in the US and Japan, in order to maximize the returns from its internal assets. By out-licensing the US promotional rights for key products without having to share any co-development costs and risks, the resulting licensing agreements have been kept simple and focused. Moreover, the agreements with Takeda and McNeil illustrate Cephalon’s willingness to partner with market leaders (and potential competitors) in order to maximize total licensing revenues.
108
Recommendations for the future The key recommendations from this report can be summarized under four main categories.
Licensing trends
As licensing deals continue to increase in value and complexity, pharmaceutical and biotech companies must look to extend their deal-making to form long lasting partnerships involving multiple products;
As more and more pharmaceutical licensing moves towards long term relationship building, companies must work even harder to find the right collaborator with the right deal in order to protect a reputation for being a reliable and productive licensing partner.
Licensing process
Pharmaceutical companies must align licensing activities around a determined strategy that is consistent with broader, corporate-level objectives;
When considering licensing opportunities, companies must also be introspective and evaluate what they can add to the deal, as a developer, marketer or collaborative partner;
A distinction between a licensing valuation used to select appropriate opportunities and a valuation used to inform negotiations and deal-making must be made in order to ensure the different objectives for each exercise are satisfied;
‘The proof of the pudding is in the eating’ and all licensing efforts should be focused on creating deals that can ultimately be successfully implemented, not just successfully agreed and signed.
109
Licensing valuations
Licensing valuations should ideally be shared and form the basis for open discussions over total project values and appropriate deal terms for sharing risks, returns and responsibilities;
Licensing valuations for deal-making should be based largely on independent and unbiased model inputs in order to limit any areas of subjective disagreement;
Valuations should be simple and clear to understand in order that all interactions between costs, risks and revenues are fully understood;
Discount rates are not necessary to understand value and should not be used to create a base case deal-making valuation;
By creating licensing valuations based on a common valuation measure (e.g. 2006 US dollars) a common language for discussing deal terms between licensing partners is provided.
Licensing best practices
Best practices are specific to a particular deal and should not be applied as a more general set of licensing rules;
Lessons can be learned from the leading deals and deal-makers – which tend to involve clear licensing objectives, long lasting agreements and creative deal terms – in order to capture the unique characteristics of each deal and deal-making partner.
110
CHAPTER 6
Appendix
111
Chapter 6
Appendix
Primary research survey A survey of 142 licensing executives was completed in early 2006 in order to identify trends and best practices for this report. As shown in Figure 6.41, survey respondents were drawn from across all major healthcare industry segments, including pharmaceuticals, biotechnology, drug delivery, generics and medical equipment and diagnostics. The survey was dominated by executives from pharmaceutical companies (55%) and biotech companies (20%).
Figure 6.41: Licensing trends survey respondents by company focus Medical equip/ diagnostics 3%
Other 9%
Generics 4% Drug delivery 9% Pharma 55%
Biotech 20%
Business Insights Ltd
Source: Licensing trends survey, 2006
As shown in Figure 6.42, survey respondents were drawn from a range of different functional responsibilities, though each respondent had some level of decision-making responsibilities for pharmaceutical licensing. Executives from the licensing and business development function made up the majority of respondents (56%), with a
112
further 20% involved in sales and marketing and 13% involved in corporate functions or senior management.
Figure 6.42: Licensing trends survey respondents by functional responsibility
R&D 6%
Industry analyst/ consultant 5%
Corporate & Senior Management 13% Licensing & Business Development 56% Sales & Marketing 20%
Business Insights Ltd
Source: Licensing trends survey, 2006
As shown in Figure 6.43, survey respondents were comprised of 63% that had responsibilities relating to both in- and out-licensing, 22% with responsibilities for inlicensing alone and 15% with responsibilities for out-licensing alone.
113
Figure 6.43: Licensing trends survey respondents by licensing responsibility Out-licensing 15%
In-licensing 22% Both 63%
Business Insights Ltd
Source: Licensing trends survey, 2006
114
Sources Cockburn and Henderson, 2001
Scale and scope in drug development: unpacking the advantages of size in pharmaceutical research. Iain M. Cockburn and Rebecca M. Henderson. Journal of Health Economics 20 (2001) 10331057
Danzon, Nicholson and Pereira, 2005
Productivity
in
pharmaceutical-biotechnology
R&D: the role of experience and alliances. Patricia M. Danzon, Sean Nicholson, Nuno Sousa Pereira. Journal of Health Economics 24 (2005) 317-339
DiMasi, Grabowski and Vernon, 2004
R&D costs and returns by therapeutic category. Joseph A. DiMasi, Henry G. Grabowski and John Vernon. Drug Information Journal 38 (2004) 211223
DiMasi, Hansen and Grabowski, 2003
The price of innovation: new estimates of drug development costs. Joseph A. DiMasi, Ronald H. Hansen and Henry G. Grabowski. Journal of Health Economics 22 (2003) 151-185
Grabowski, Vernon and DiMasi, 2002
Returns on research and development for 1990s new drug introductions. Henry Grabowski, John Vernon
and
Joseph
A.
DiMasi.
Pharmacoeconomics 20 Suppl 3 (2002) 11-29
115
Index Ablynx, 107
McNeil, 108
Alliance management, 62, 63
Merck & Co, 23
Amgen, 104
Millennium, 101, 102, 103
Amylin, 107
Novartis, 14, 23, 27, 101, 102, 104, 105, 106
Anadys Pharmaceuticals, 105
NPV, 13, 58, 74, 84, 85, 86, 87, 95
Arakis, 105
Ortho Biotech, 101, 102, 103
Arrow Therapeutics, 105
Otsuka, 105
AstraZeneca, 23, 62, 101, 103
Pfizer, 23, 104
AtheroGenics, 101, 103
Pharmaceutical, 1, ii, iii, 12, 17, 19, 23, 43, 49, 58, 63, 67, 109
Biotechnology, 12, 67, 103 Portfolio management, 47 Cardiovascular, 95 Roche, 14, 23, 27, 63, 101, 102, 104 Cephalon, 14, 107, 108 Sanofi-Aventis, 23 CNS, 36, 37, 91, 92, 95, 96 SeBo, 105 Deal making, 19, 60, 76 Senju Pharmaceutical, 105 Drug delivery, 48 Servier, 105 Eli Lilly, 10, 18, 62 Takeda, 107, 108 Genentech, 10, 14, 18, 27, 101, 102, 107 Tanabe Seiyaku, 108 GlaxoSmithKline, 23, 27 UCB, 107 Human Genome Sciences, 105 Valuation, 75 Idenix, 27, 101, 102, 105 Ventura, 105 Johnson & Johnson, 102, 104
116
