Académique Documents
Professionnel Documents
Culture Documents
STRATEGIES
Best practices in deal-making,
valuations and strategic
management
By Steven Seget
Steven Seget
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
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 49
In-licensing 50
Out-licensing 52
Licensing evaluations 56
General portfolio management 57
Applications for licensing evaluations 59
Deal-making and agreement 60
Key elements of a pharmaceutical license agreement 60
Post-deal management and analysis 61
Alliance management 62
Using outside agencies 64
iv
AstraZeneca-AtheroGenics 103
Preferred licensing partners 104
Leading in-licensing companies 104
Novartis 105
Leading out-licensing companies 107
Cephalon 107
Recommendations for the future 109
Licensing trends 109
Licensing process 109
Licensing valuations 110
Licensing best practices 110
v
List of Figures
Figure 2.1: Number and average value of top 10 pharmaceutical company licensing deals, 2001-
2005 24
Figure 2.2: Expected change in number of licensing deals during 2006 25
Figure 2.3: Expected change in average value of licensing deals during 2006 26
Figure 2.4: Number of top 10 pharmaceutical company licensing deals by partner, 2001-2005 28
Figure 2.5: Number of top 10 biotech company licensing deals by partner, 2001-2005 30
Figure 2.6: Number of biotech out-licensing deals by partner, 2001-2005 31
Figure 2.7: Number of top 10 pharmaceutical company licensing deals by deal type, 2001-2005 33
Figure 2.8: Number of top 10 biotech company licensing deals by deal type, 2001-2005 35
Figure 2.9: Proportion of product-based licensing deals by therapy area, 2001-2005 36
Figure 2.10: Number of product-based licensing deals by therapy area, 2001-2005 37
Figure 2.11: Number of R&D licensing deals by development stage, 2001-2005 38
Figure 2.12: Number of biotech R&D licensing deals by development stage, 2001-2005 39
Figure 3.13: Expected change in number of potential partners chasing each licensing deal during
2006 44
Figure 3.14: Expected change in the length of time required to complete a licensing deal during
2006 44
Figure 3.15: The pharmaceutical licensing process 46
Figure 3.16: Parties involved in identifying potential licensing opportunities, 2006 65
Figure 3.17: Parties involved in conducting due diligence for potential licensing opportunities, 2006
66
Figure 3.18: Parties involved in the valuation and negotiation of potential licensing deals, 2006 67
Figure 4.19: Information shared between partners during licensing negotiations, 2006 74
Figure 4.20: Valuation techniques used in determining optimal licensing deal terms, 2006 75
Figure 4.21: R&D costs by phase, 2000 78
Figure 4.22: R&D lead times by phase, 2000 79
Figure 4.23: R&D success probabilities by phase, 2000 80
Figure 4.24: Drug market diffusion curve – product lifecycle 81
Figure 4.25: Likelihood of outcomes for new phase I, phase II and phase III drugs 85
Figure 4.26: Expected real values (non-discounted) for new phase I, phase II and phase III drugs 86
Figure 4.27: Discounted expected real values for new phase I, phase II and phase III drugs 87
Figure 4.28: Discounted expected real values for new phase I, phase II and phase III drugs
(adjusted for lower R&D cost inflation) 88
Figure 4.29: Deal outcomes for out-licensor 89
Figure 4.30: Deal outcomes for in-licensor 89
Figure 4.31: Share of expected deal outcomes by partner 90
Figure 4.32: R&D costs by phase by therapy area, 2000 92
Figure 4.33: R&D lead times by phase by therapy area, 2000 93
Figure 4.34: R&D success probabilities by phase by therapy area, 2000 94
Figure 4.35: Peak sales and year of peak sales by therapy area, 2000 94
Figure 4.36: Discounted value of sales by therapy area, 2000 95
Figure 5.37: In-licensing partner of choice, 2006 104
Figure 5.38: Business development and licensing department, Novartis 106
Figure 5.39: Licensing process at Novartis 106
Figure 5.40: Out-licensing partner of choice, 2006 107
Figure 6.41: Licensing trends survey respondents by company focus 112
vi
Figure 6.42: Licensing trends survey respondents by functional responsibility 113
Figure 6.43: Licensing trends survey respondents by licensing responsibility 114
List of Tables
Table 4.1: R&D cost by phase and peak sales, 2006 (expressed in 2006 dollars) 83
vii
viii
Executive Summary
9
Executive Summary
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.
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.
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.
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 in-
house, 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.
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.
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.
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
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 non-
approvable 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:
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
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.
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.
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
300 297 304
300 285 300
200 200
187
150 172 170 150
149
100 100
50 50
0 0
2001 2002 2003 2004 2005
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.
100%
90% Increase
Proportion of survey respondents
significantly
80%
70% Increase
somewhat
60%
30% Decrease
somewhat
20%
10% Decrease
significantly
0%
Pharma Biotech Other Overall
Source: Business Insights Licensing Trends Survey, 2006 Business Insights Ltd
25
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%
90% Increase
Proportion of survey respondents
significantly
80%
70% Increase
somewhat
60%
30% Decrease
somewhat
20%
10% Decrease
significantly
0%
Pharma Biotech Other Overall
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.
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
350
Number of top 10 pharma licensing deals
300
Pharma
90
250
135 116 128
122
200
Biotech
150
157
106 106
115 100
100
50 Other
50 63 63 51 57
0
2001 2002 2003 2004 2005
100%
Proportion of top 10 pharma licensing deals
90%
29.6%
80% 40.7% Pharma
45.0% 43.1% 44.7%
70%
60%
50% Biotech
51.6%
40% 37.2% 35.7%
38.3% 36.6%
30%
20%
Other
10% 22.1% 21.2% 18.7% 18.8%
16.7%
0%
2001 2002 2003 2004 2005
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 long-
term 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
100
Number of top 10 biotech licensing deals
90
80 26 Pharma
70
30 35
60 42
32
50 Biotech
48
40
30 41 34
32 33
20
Other
10 18
8 11 8 9
0
2001 2002 2003 2004 2005
100%
Proportion of top 10 biotech licensing deals
90%
28.3%
80% 38.0% Pharma
43.8% 43.2%
51.2%
70%
60%
50% Biotech
52.2%
40%
51.9% 42.5% 44.6%
30% 39.0%
20%
Other
10% 19.6%
10.1% 13.8% 12.2%
9.8%
0%
2001 2002 2003 2004 2005
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 pharma-
biotech 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.
100%
Proportion of biotech out-licensing partners
90%
80%
58.3% 58.6% 55.3% 56.9%
70% 60.6% Biotech
60%
50%
40%
30% Pharma
10%
0%
2001 2002 2003 2004 2005
31
Licensing deal types
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
350
Number of top 10 pharma licensing deals
300 Collaboration
91 75
250 105
92 78
R&D/ S&M
200
79 95
68 86
150 87
License/
acquisition
100 74
91 83
73
88
50 Others
56 44
34 40
20
0
2001 2002 2003 2004 2005
100%
Proportion of top 10 pharma licensing deals
70%
R&D/ S&M
60% 32.0%
26.3% 23.9% 31.9%
50% 28.3%
40% License/
acquisition
30% 24.7%
31.9% 27.9%
32.2% 24.0%
20%
Others
10% 18.7% 14.8%
11.9% 13.2%
7.3%
0%
2001 2002 2003 2004 2005
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
100
Number of top 10 biotech licensing deals
90
Collaboration
80 26
20
70
26 25
R&D/ S&M
60 28
50 30
17 30
40 24
License/
20
acquisition
30
23 21
20 24 15
29 Others
10 15 15
8 11
0
2001 2002 2003 2004 2005
100%
Proportion of top 10 biotech licensing deals
70%
R&D/ S&M
60% 21.3% 36.6%
32.6%
30.4%
50% 27.0%
40% License/
28.8% acquisition
30% 22.8%
29.3% 20.3%
20% 36.7%
Others
10% 18.8% 14.9% 16.3%
9.8%
0%
2001 2002 2003 2004 2005
35
Licensing deal subjects
Cancer 20.5%
Infections 13.9%
Dermatology 6.2%
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.
24%
Proportion of product-based licensing deals
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
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%.
100%
Phase III
80%
70%
38.3% 34.5%
60% 37.2% Phase II
42.6% 39.4%
50%
40% 18.0%
19.5% Phase I
30% 20.5%
16.2% 17.8%
20%
28.3%
23.7% Pre Clinical
10% 19.4% 19.7% 20.2%
0%
2001 2002 2003 2004 2005
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.
100%
30.0% 29.3%
80%
7.4% Phase II
70% 10.5% and above
9.9%
13.1% 11.7%
60%
Preclinical/
50% Phase I
40%
73.2%
66.0% 63.7%
30% 56.9% 59.0% Research
20%
10%
0%
2001 2002 2003 2004 2005
39
40
CHAPTER 3
Licensing process
41
Chapter 3 Licensing process
Summary
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.
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 in-
house, with the majority completing the entire licensing process without outside
help.
42
Introduction
A complex process
43
Figure 3.13: Expected change in number of potential partners chasing each
licensing deal during 2006
100%
90% Increase
Proportion of survey respondents
significantly
80%
70% Increase
somewhat
60%
30% Decrease
somewhat
20%
10% Decrease
significantly
0%
Pharma Biotech Other Overall
100%
90% Increase
Proportion of survey respondents
significantly
80%
70% Increase
somewhat
60%
30% Decrease
somewhat
20%
10% Decrease
significantly
0%
Pharma Biotech Other Overall
44
In-licensing versus out-licensing
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
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.
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;
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;
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;
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
49
In-licensing
Do we have the required sales and marketing capabilities 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:
Geographic availability;
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.
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:
Originator/owner;
Planned indication(s);
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:
Therapeutic rationale – an outline of the mechanism of action and how and why
this differs from other products;
Intellectual process – a review of the current patent position, expiry dates and any
other intellectual property such as confidential information etc;
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.
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.
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 in-
house 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 decision-
making for internal projects and resources can easily be adapted for use in evaluating
external projects.
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.
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:
58
Monte Carlo simulations;
Real options.
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.
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.
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.
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.
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
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:
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.
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
Management consultants;
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%
No outside parties
60%
Proportion of survey respondents
Management
50%
consultancies
10%
Accounting firms
0%
Pharma Biotech Overall
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%
No outside parties
60%
Proportion of survey respondents
Management
50%
consultancies
10%
Accounting firms
0%
Pharma Biotech Overall
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%
No outside parties
60%
Proportion of survey respondents
Management
50%
consultancies
10%
Accounting firms
0%
Pharma Biotech Overall
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
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.
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.
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
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;
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.
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%
90% A decision-model
Proportion of survey respondents
simulating outcomes
80% for both partners
70%
A detailed
60% probabilistic forecast
model
50%
20%
No hard financial
10% forecasts
0%
Pharma Biotech Other Overall
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
80%
forecasts
60%
Discounted cash flow
50% NPV calculations
0%
Pharma Biotech Overall
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.
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.
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$802 million in 2000 dollars, which takes
into account the cost of all the failed projects required in earlier stages to get one
successful product. This cost estimate has become widely established as the benchmark
for R&D development costs in the pharmaceutical industry and the article’s R&D data
are therefore considered to be a good estimate of cost, lead time and risk.
77
studies are considered to occur in parallel with phase I and phase II trials, being
completed before a drug begins phase III trials. Post approval trials relate to lifecycle
drug development including the development of a launched drug in new indications or
new formulations, as well as any necessary post approval regulatory trial requirements.
160
140
Cost (US$ million, 2000)
120
100
80
140.0
60
86.3
40
20
23.5 5.2
15.2
0
Phase I Phase II Phase III Long-term Post approval
animal
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
40
35
Mean time to next phase (months)
30
25
20
33.8
15
26.0
10 18.2
12.3
5
0
Phase I Phase II Phase III Approval phase
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
Figure 4.23: R&D success probabilities by phase, 2000
80%
Probability of entering phase (from previous)
70%
60%
50%
40%
71.0% 68.5%
30%
20%
31.4%
10%
0%
Phase II Phase III Approval
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.
100%
97% 97%
100% 94% 94%
88% 88%
90%
Proportion of peak sales
70%
60%
50%
50%
38%
40%
30% 25%
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)
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)
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 decision-
tree or expected value approach it appears better that specific development phases are
actually discounted based on their technical risk rather than using a common project-
wide 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%
70%
Likelihood of outcome
60%
50%
40%
69% 69%
30%
49%
20%
29% 32%
10% 22%
15%
7% 10%
0%
Phase I only Phase 2 only Phase 3 only Launch
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$271 million for a new phase I drug) in constant 2006
US dollars.
Figure 4.26: Expected real values (non-discounted) for new phase I, phase II
and phase III drugs
3,000
Drug value (US$ million, 2006)
2,500
2,000 1,797
1,500
1,000
464
500 271
-500
Phase I only Phase 2 only Phase 3 only Launch Expected
value
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.
Figure 4.27: Discounted expected real values for new phase I, phase II and
phase III drugs
1,800
Expected drug value (US$ million, 2006)
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)
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
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
Expected drug value (US$ million, 2006)
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)
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$100
million for gaining regulatory approval. Upon the drug reaching the market, royalties
based on gross sales of 10% are payable to the out-licensor.
88
Figure 4.29: Deal outcomes for out-licensor
120
Out-licensor
100 license
payments/
Cost/ sales (US$ million, 2006)
80 royalties
Out-licensor
60 costs/ sales
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
600
In-licensor
500 costs/ sales
Cost/ sales (US$ million, 2006)
400
In-licensor
300 license
payments/
200 royalties
royalties
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
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.
1,600
1,400
Drug value (US$ million, 2006)
1,200
1,000
800
600
400
200
200 264
0
-200
-400
Phase 2 only Phase 3 only Launch Expected value
Out-licensor In-licensor
90
Having an understanding of both potential licensing cash flows and expected, risk-
adjusted 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 deal-
making negotiations around common estimates of value and not around pre-determined
target deal terms prepared in isolation by each party.
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, anti-
infective, 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.
160
Development costs (US$ million, 2000)
140
120
100
80
60
40
20
0
Phase I Phase II Phase III
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
Figure 4.33: R&D lead times by phase by therapy area, 2000
140
Mean time to complete phase (months)
120
22.1
100
80 18.2
21.0
60 12.5
15.4
92.5
40 72.1
61.0
46.4 50.5
20
0
All Analgesic/ Anti-infective Cardiovascular CNS
Anesthetic
93
Figure 4.34: R&D success probabilities by phase by therapy area, 2000
90%
Probability of entering phase (2000)
80%
70%
60%
50%
40%
30%
20%
10%
0%
Phase II Phase III Approval
Figure 4.35: Peak sales and year of peak sales by therapy area, 2000
Year 9
900
Year 12
800
Peak year sales (US$ million, 2000)
700
600
Year 10
500
Year 9
400
300
Year 6
200
100
0
All Analgesic/ Anti-infective Cardiovascular CNS
Anesthetic
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.
8,000
7,000
NPV sales (US$ million, 2000)
6,000
5,000
4,000
3,000
2,000
1,000
0
Mean 25th percentile Median 75th percentile
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
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.
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.
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
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 co-
promotion 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 co-
promote the drug.
Novartis 14
Pfizer 9
J&J 7
Amgen 6
Roche 6
GSK 5
Eli Lilly 4
AstraZeneca 3
Merck 2
Sanofi-Aventis 1
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:
Arrow Therapeutics for an early stage respiratory syncytial virus (RSV) treatment;
Arakis and Ventura for a phase II chronic obstructive pulmonary disease (COPD);
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.
Global BD&L
V. Hartmann
USA, EU Japan
Latam, Asia I. Ohhashi
Development Alliances/
Research Alliances
Out-Licensing
S. Strub
I. Csendes
Initial
Full
Preselection Technical Negotiation
Evaluation
Evaluation
S&E
Negotiation
Alliance Management
106
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 out-
licensing 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.
Cephalon 3
Ablynx 2
Amylin 2
Genentech 2
UCB 2
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 in-
licensing 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.
108
Recommendations for the future
The key recommendations from this report can be summarized under four main
categories.
Licensing trends
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
‘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;
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;
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
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%).
Drug delivery
9%
Pharma
55%
Biotech
20%
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.
Industry
analyst/
consultant
R&D 5%
6%
Corporate &
Senior
Management
13%
Licensing &
Business
Development
56%
Sales &
Marketing
20%
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 in-
licensing 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%
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) 1033-
1057
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) 211-
223
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
Amylin, 107 Novartis, 14, 23, 27, 101, 102, 104, 105, 106
Anadys Pharmaceuticals, 105 NPV, 13, 58, 74, 84, 85, 86, 87, 95
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