Christopher C. Gallen, M.D., Ph.D.

1
Strategic Challenges in Neuroprotective Drug
Development

• Christopher C. Gallen, M.D., Ph.D.
• Vice President, Wyeth Research

March 15, 2003Washington, D.C.
2
The Current World of Pharma

• The Big Picture Challenge for RD-Driven
  Pharmaceutical Companies
• The Challenges of CNS RD
• Meeting the Challenge
• Changing the Model
• A Strategy Going Forward

3
Health Care Costs
As a Percentage of Gross Domestic Product in
Major Industrialized Countries, 1997
U.K.
Japan
Italy
Netherlands
Canada
France
Germany
U.S.
1997 data
Source OECD-OECD Health Data, 1998.
4
Pharmaceutical Costs
U.K.
Japan
Italy
Netherlands
Canada
France
Germany
U.S.
1997 data
Source OECD-OECD Health Data, 1998.
5
What Happens When a Patent Expires?Prozac Total
Prescriptions Per Month
6
The Patent Expiration Challenge
of Total Sales to 2005 Vulnerable to Patent
Expiration
Over 100B of Products Face Generic Competition
by 2005
Source Cap Gemini Ernst and Young, 2002. Global
Market Research Analysis
7
Pharmaceutical RD Investment is High
Source PhRMA, 2001, Based on Data from PhRMA
Annual Survey and Standard Poors Compustat, a
Division of McGraw-Hill
8
Percentage of companies active in each
therapeutic area
Other companies(n24)
Major companies(n14)
Therapeutic area ordered by decreasing number
ofNASs in development on December 31st, 2001
IO0-10099 24/05/02
Source Institute for Regulatory Science
9
Nervous system NASs dominate the development
pipeline
Source Institute for Regulatory Science
10
But RD Productivity is Decreasing
Source PhRMA Annual Survey, 2001. U.S. FDA.
Global Market Research Analysis
11
Discovery and Development Costs are Increasing
Source DiMasi et al., Tufts CSDD RD Cost Study,
2002
12
Clinical Trial Number Per NDA is Increasing
Number of Trials
Source Boston Consulting Group, 1993 Peck, C.,
Drug Development Improving the Process,Food
Drug Law Journal, Vol. 52, 1997.
13
Number of Patients Per NDA is Increasing
Number of Patients
Source Boston Consulting Group, 1993 Peck, C.,
Drug Development Improving the Process,Food
Drug Law Journal, Vol. 52, 1997.
14
Number of patients per phase III study to support
first submission
Mean number of patients
Therapeutic area
Where enrolment completed 1999-01
15
RD Cycle Times are Increasing
14.8
14.2
2.6
2.8
11.6
2.1
6.1
8.1
5.5
Years
4.1
2.4
2.5
6.1
5.9
5.1
3.2
Source Joseph A. DiMasi, New Drug Development
Cost, Risk and Complexity, Drug Information
Journal, May 1995. (From RD Directions, 1996)
16
Drug Approval Times are Increasing Again
Source U.S. Food and Drug Administration
17
Time to termination by therapeutic area(for NASs
terminated 1999-2001)
Source CMR International
18
Breakdown of reasons for termination (for NASs
terminated 1999-2001)
Source CMR International
19
Attractiveness profile of industrys late stage
pipeline
High success rate, slow cycle time
High success rate, fast cycle time
39
Anti-infectives
Musculo-skeletal
42
18
Alimentary/ metabolism
25
19
CVS
Respiratory
43
Oncology
Nervous system
51
Fast cycle time, low success rate
Low success rate, slow cycle time
Bubble size current market size (IMS) number
in bubble number of NASs in phase II/III
development
20
Why are Success Rates Declining?

• Discovery issues
• Conceptual issues re disease models
• Clinical Trial issues

21
Genomic Targets Promise and Concerns

• The Promise - improved diagnostics, fundamentally
  targeted treatments
• Reality Proliferation of targets - but targets
  with a limit
• Within target heterogeneity
• Challenging targets - known models of molecular
  dysfunction
• Most targets likely loss of function
• Large market diseases polygenic
• Twin concordance rates disturbing

22
Technological Challenges

• Structure-based Drug Design
• Match molecules to targets different from in-situ
  conformation
• Fit for in vitro viral proteins likely CNS
  proteins
• Combinatorial Chemistry
• Existing libraries limited by origins - monoamine
  GPCRs, steroid receptors and serine-aspartyl
  proteases

23
Why is CNS Particularly Challenging?

• Normal Functioning
• Intimate connections, fine timing and pattern
  code
• Parallel paths, multiple systems/step
• Instantaneous mutual regulation
• Self regulation of the system over time
• Antagonists versus agonists
• Single target bullets best for probes
• Therapies generally multi-target

24
CNS Disease Animal Models can be Misleading

• Model congruity with disease
• Understand the animal model
• Understand the human disease
• Show them to be congruent in all important
  respects
• Cell Culture
• Cell-cell interactions, relation to nutritional
  systems, exogenous environment, phospholipid
  composition all differ
• Mouse Models
• Major failures of single genes
• Strain differences suggest a cause for concern

25
Meeting the Challenge Clinical Rigor

• Success rates are too low to tolerate avoidable
  flaws
• Animal testing under one set of conditions, human
  trials under another
• Ignoring the does it make scientific sense?
  test
• Animal models measuring very different dependent
  variables
• Inadequate determination of dose and duration

26
Using Technology to do Better Trials

• Key Near-time trial conduct and analysis
• Scrutinize blinded data to detect poor sites
• Exploratory development - double-blind but not
  triple blind
• Exploratory data analysis oriented database and
  approach for better programs and submissions
• Modeling and simulation for better trials
• Adaptive trial designs to optimize dose-ranging

27
Experimental Medicine - Part of the Solution

• Is the compound absorbed?
• Does the compound penetrate to the desired site
  of action? For appropriate period of time?
• Mechanism consistent with hypothesis?
• Biological effect?
• Free of class-associated limiting toxicities?

28
Disease Models

• Reality is a complex set of interactions
• Each step can be modeled as differential
  equations
• Myriad publications describe individual pieces
• Supplemented with research to test the model
• Technology allows generation of increasingly
  sophisticated disease models
• Stronger model will produce the insights on
  target selection and effective therapies
• Core Intellectual Property

29
Electronic Technologies can Improve Chemistry

• NIH Protein Structure Initiative
• Increased supercomputer modeling of protein
  folding and interactions
• Virtual screening
• Virtual combinatorial chemistry
• Moving past target to cross-assessing potential
  toxic interactions and metabolism

30
Biological Technologies Have Great Promise

• 35 of the 37 NAS launched in 2001
• Biologics have important attractions
• Typically less toxic, more predictable
• Increasingly human derived
• Easier to predict distribution, metabolism and
  elimination
• Faster development
• Higher success rates
• Huge ability to match potential targets

31
Changing the Business Model

• Historical
• Platform oriented
• First line treatments, one size fits all, mass
  population, easy (oral) treatment, ameliorating
  chronic disease
• One treatment per disease
• Next Generation
• Disease focus
• Defined populations
• Administered by specialists
• Targeted treatments
• Expand treatments to capture therapeutic
  subpopulations
• Polypharmacy in cases (similar to oncology
  development)

32
Pharma and Academicians

• Partnership
• Intellectual challenge of deciphering targets
• Building disease models
• Closer ongoing collaborative contact
• Remote presence technologies
• Secure e-data sharing

33
Pharma and Regulators

• Shifting to a model of early POC studies in man
  for both target and molecule validation calls for
  earlier consultations
• Partnership
• Closer ongoing collaborative contact
• Rolling dossiers
• Marketing rights will change from being one-off
  to continuous evaluation

34
A Strategy Going Forward

• Focus on intellect and collaboration
• Pharma focus on disease model
• Experimental medicine model
• Tap the power of the information revolution
• Tap the power of biologic-based technologies
• Adapt the Business Model