Personalized Medicine State of the Art

1
Personalized Medicine- State of the Art

• Michael Cantor, MD
• Director, Healthcare Informatics
• Pfizer, Inc.

September 29, 2009
2
Personalized Medicine- How Far Have We Come?

• Personalized medicine, the use of
  marker-assisted diagnosis and targeted therapies
  derived from an individual's molecular profile,
  will impact the way drugs are developed and
  medicine is practiced.
• The traditional linear process of drug
  discovery and development will be replaced by an
  integrated and heuristic approach.
• Patient care will be revolutionized through
  the use of novel molecular predisposition,
  screening, diagnostic, prognostic,
  pharmacogenomic and monitoring markers.
• Knowledge of the molecular basis of disease
  will lead to novel target identification,
  toxicogenomic markers for screening and improved
  selection of clinical trial patients, which will
  fundamentally change the pharmaceutical industry.

Ginsburg and McCarthy, Trends in Biotechnology,
December 2001
3
Personalized Medicine Key Components
Driving the understanding of disease and the
discovery and development of medicines
Changes in the regulatory landscape Policy will
determine success or failure of PM Changes in
regulatory landscape pre- and post-marketing
How will PM change economics for drug discovery
and healthcare? Will the resulting benefit
actually be cheaper or more costly?
Changes in medical practice to appropriately
deliver personalized medicines to patients
4
Continuum from Empirical to Personalized Medicine
Empirical/Intuitive Medicine
Stratified/Targeted/ Personalized Medicine
Personalized/Individualized/Precision Medicine
Reduce variation in potential response to improve
prediction of whos likely to do well on a
treatment.
Right Drug Right Dose First Time
Substantial heterogeneity in treatment response
5
Personalized Medicines Progressive Staircase
A Glimpse of the Future for Patients
2020 and beyond
2012-2020

• Drugs treat underlying cause of disease
• Presymptomatic diagnosis and treatment becomes
  routine
• Therapies deliver efficacy rates to 80-90 with
  value paid for beneficial profile
• Dramatic declinein rates of medical expenditures
  due to reduction indrug induced SAEs
• Drug guarantee for medical benefits and cost
  containments

2007 2012

• 75 of oncology drugs targeted by tumor genomics
• 20 of chronic disease drugs linked to genomic
  profiling
• Risk profiles defined for CV disease
• 1000 genome sequencing
• Integration of genomic and healthcare data

2002 2006

• Delivery of data from GWAS WTCC, GAIN (2007)
• Passage of GINA (2008)
• GWAS part of drug develop.
• Risk profiles for Alz Dis
• Prospective RCT for Abacavir (2008)
• Dx tests in Onc. e.g. KRas routine
• Rise of Consumer Genetics

Late 1990s

• Affymetrix Launches 10,000 SNP chip (2002)
• HapMap (2003)
• HapMap,600K Map (2005)
• Launch of Oncotype Dx (2004)

• Candidate Gene Analysis Limited SNP Information
  (1998)
• Estimates for high density SNP map timing (2005)
  and whole genome association study (2010)

6
Personalized Medicine at Pfizer

• Maraviroc- Trofile assay for CCR5-tropic HIV
  strains
• Pan Her (anti-EGFR) Phase II trials exclude
  patients with mutated K-ras
• Biomarkers for likelihood of recurrence in Renal
  Cell Carcinoma
• Glioblastoma Multiforme cancer vaccine specific
  for tumor cells expressing variant III of EGFR
• Home monitoring technologies

7
Drug Discovery

• Tenets of PM are becoming routine for pharma
• Changing business model?
• More specialized drugs, smaller populations
• Premium due to higher likelihood of successful
  treatment
• Need therapeutic value that produces sustainable,
  profitable growth
• Smaller populations offset by lower development
  costs
• Can revenues from smaller-scale products support
  huge companies?
• Role of biomarkers yet to be completely defined
• Adaptive clinical trials based on biomarkers?
• Validation/qualification standards
• Retrospective vs. prospective data

8
Providers and Personalized Medicine

• In general, providers (JAMA 2991320-1334)
• Lack sufficient knowledge and confidence
  relating to the provision of genetic services
• Feel underprepared for assessing and managing
  genetic issues in their practice
• Lack basic genetic knowledge
• Have limited time to obtain family history
• Information overload
• Limited time with patients
• How can we overcome these barriers?
• Clinical Decision Support

9
Decision Support for Personalized Medicine

• Knowledge
• Who is eligible
• What tests are available
• Translation
• Interpreting results of tests
• Communication
• Tailored to patient educational levels?
• Available in patients language?
• Available to family members?
• Quality Control
• Is advice on interpretation of results correct?
• Are CDSS resources timely?

10
Physicians (and other providers)

• Changes to education
• Changes to incentives
• Why order the diagnostic when you can just give
  the drug and see what happens?
• PM substance not there yet for most specialties
  outside of Oncology
• What does PM mean for procedure-based
  specialties?

11
Diagnostic Tests Using Genomic Information
12
Pfizer Experience Diagnostics

• Safety Diagnostics Irinotecan
• Approved in 1996 for metastatic colon cancer
• UGT1A128 allele (homozygotes) linked to
  increased risk of toxicity
• Label changed in 2005 noting pharmacogenomic
  information
• Questions about heterozygotes due to lack of
  prospective data
• Multiple providers of irinotecan-related
  diagnostics

• Companion Diagnostics Maraviroc
• Monogram Biosciences Trofile test approved for
  CCR5 tropism
• Pathways Sensitrop cleared in 2007
• Both cleared under CLIA
• Sensitrop had poor sensitivity (42 compared to
  Trofile results)
• Quality of diagnostics important for proper
  prescribing and maximizing effectiveness

13
Regulation

• IVDMIA update from FDA expected this year
• Complex diagnostic regulatory landscape
• FDA or CLIA
• Need consistent framework
• NY State / California regulations on personal
  genomics companies
• Proposed CA law to exempt bioinformatics firms
  from CLIA regulations
• Label for PGx is suggestive, not prescriptive,
  since regulators do not practice medicine
• Privacy regulations essential
• GINA
• Privacy in the Facebook era?
• Payers often ahead of regulators (KRAS)

14
Consumer Genomics

• This child is very thoughtful and focused, so I
  suggest she go into management
• - Scientist from Shanghai Biochip corporation,
  after examining a childs DNA test
• (CNN Asia, August 2009)

15
Consumer Genomics
16
Patients

• Consumer genomics are currently more recreation
  than health
• PM is most applicable in deadly diseases
• How to communicate risks?
• How to give informed consent?
• Privacy
• Family risk communicating to family members
• Risk of finding susceptibility to an incurable
  disease (e.g. Huntingtons)
• How to bring underserved patient populations the
  benefits of PM?

17
Comparative Effectiveness

• Biggest barrier to Genomic Medicine lack of
  adequately designed studies addressing clinical
  utility
• Once you know the genotype/drug associations,
  need studies to assess benefit of gene-guided
  management
• Goal of CER is to broaden and deepen tool for
  matching care more precisely to individuals
• Federal CER support
• More useful data collected
• Better methods developed to understand
  differences
• Standards for CER methodology can lead to clarity
  on which forms of evidence are sufficient for
  decisionmakers
• Require innovation new research or clinical
  trial models
• Molecularly Informed CER
• Stratify population, find genetic explanations
  for different responses
• NCI Comparative effectiveness Research in
  Genomic Applications in Cancer Care and
  Prevention
• Garber, AM and Tunis, SR , NEJM, 360(19).
• Khoury,MJ et al, Genetics in Medicine,
  11(10).

18
Personalized Medicine and Healthcare Reform

• True PM would be ideal
• Eliminate wasteful treatments
• Ensure better outcomes, fewer AEs
• Earlier disease detection, earlier stage
  interventions, lower costs
• Insurance issues
• What if your genotype is the non-responder?
• Provider issues
• How to keep abreast of developments in PM that
  are applicable to patients
• Privacy issues

19
Key for Personalized Medicine

• Significant therapeutic value
• Superior to existing therapies for serious
  condition(s)
• Prevent AE
• Higher likelihood of cure
• Significantly improve time to therapeutic goal
• Easy to understand risks/benefits
• Proper incentives
• Reimbursement (or no reimbursement) based on
  qualified biomarker

20
Conclusions

• Success of Personalized Medicine depends on
  significant, tangible clinical benefits
• Practicing PM is more complicated than
  initially expected
• PM currently concentrated in oncology, but
  should reach other areas in the short term
• As personalized therapies become the norm, drug
  discovery business models will have to change
• Many physicians currently feel unprepared, but
  with better decision support and changes in
  education, will have tools to practice PM
• Patient education will be a major facilitator for
  the practice of PM
• How far have we come?