6/1/2010

1

ISSUES AND CHALLENGES IN THE DEVELOPMENT AND REIMBURSEMENT OF

PERSONALIZED MEDICINE: HOW CAN HEALTH ECONOMICS AND

OUTCOMES RESEARCH HELP?

Eric Faulkner, MPHChair, ISPOR Personalized Medicine Reimbursement & Development Working GroupSenior Director, RTI Health Solutions, Executive Director, Genomics Biotech Institute, National Association of Managed Care Physicians, Assistant Professor, Institute for Pharmacogenomics and Individualized Therapy, University of North Carolina at Chapel Hill, Research Triangle Park, NC USA

ISPOR Personalized Medicine Special Interest Group

Chair:

Eric Faulkner, MPH, Senior Director, RTI Health Solutions, and Executive Director, Genomics Biotech Institute, National Association of Managed Care Physicians, Assistant Professor, Institute for Pharmacogenomics and Individualized Therapy, University of North Carolina at Chapel Hill, Research Triangle Park, NC, USA

Leadership Committee Members:

Lieven Annemans, PhD, MSc; Finley Austin, PhD, BS; Pat Deverka, MD, MS; Lou Garrison, PhD; Mark Helfand, MD, MPH; John Hornberger, MD, MS; Dyfrig Hughes, PhD, MSc;Tracy Li, PhD; Dan Malone, RPh, PhD; Katherine Payne, PhD; Uwe Siebert, MD, MPH, MSc, ScD; Adrian Towse, MA; Dave Veenstra, PhD, PharmD; John Watkins, RPh, MPH Dyfrig Hughes,

Goal:

To develop good research practices in personalized medicine and inform appropriate health care decision and policy making using personalized medicine information

For more information, please contact: efaulkner@rti.org or 919-541-6772

Health Technology 2010: The Issues

↑ Healthcare Spending • ↑ Development Costs

↑ Cost/Quality Concerns • ↑ Value-based Purchasing

↓ P&R Opportunities • ↑ HTA & EBM Pressures

↑ Consumerism • ↑ Risk Sharing Approaches

↓ Placebo Control • ↑ Comparative Effectiveness

↓ Physician Control • ↑ Payer Control

↓ “Blockbuster” Model • ↑ Personalized Medicine

P&R = pricing and reimbursement; HTA = health technology assessment; EBM = evidence-based medicine

Payer, Provider and Manufacturer Role is a Matter of Perspective…

Trend: Struggle to Balance Innovation, Quality and Cost

Image Source: 20th Century Fox & LucasFilms, LTD.

or

Trend: Growing Pressure for More (and More) Evidence

Health decision makers are requesting more (and more) clinical

Source: Medical Guesswork. Business Week. May 29, 2006 cover story.

requesting more (and more) clinical and economic evidence on the value

of new health technologies.

Going forward, it will be important to distinguish “got to have” vs. “nice to

know” evidence…

What evidence is truly material to informed product use?

6/1/2010

2

Trend: Movement Towards Comparative Effectiveness and Value-based Purchasing

Payers and others ask:ask:

Can we get better health value for the dollars we spend?

Trend: Payers Limit or Reject Expensive New Health Technologies

Source: Times Online 2007; Telegraph 2007; WSJ 2003. Image courtesy of T. Morrow, MD. NAMCP 2006.

What is Personalized Medicine?

Personalized Medicine: Use of genetic or other molecular biomarker

information to improve the safety, effectiveness and health outcomes of patients via more

efficiently targeted risk stratification, prevention and tailored treatment management

approaches

Pharmacogenomics:Pharmacogenomics:Use of genomic tests to inform patient

treatment selection and dosing by predicting drug response

Several examples thus far…― HER2/neu: Herceptin

― KRAS/EGFR drugs: Vectibix/Erbitux

― Oncotype DX: breast cancer chemo

― CYP2C19: Plavix

Source: Scientific American 2000 and 2006

How Will Personalized Medicine Evolve?What Potential Influence on Quality & Cost?

?

Adapted from “The Evolution of Man”

Factors Driving Personalized Medicine Forward

State of the science and information systems now evolved to enable personalized medicine

Regulatory agencies have taken steps to integrate biomarkers into drug development

Payer struggles to balance quality and costs

Limit access to responders/subpopulations – ↑ effectiveness & safety p p p ↑ y

↓ wastage and unnecessary costs

Manage increasingly expensive drug categories (e.g., oncology biologics)

Aggregate affordability of health services is a significant concern in many nations

Manufacturers recognize challenging reimbursement landscape and emphasis on PM is one tool to support market access

What is Different About Personalized Medicine?

Opportunity to ↑ quality via better treatment targeting and may improve cost-effectiveness (but not always)

Significant patient access & policy implications Responders vs. nonresponders

Patients in the “grey zone”

Accomplishes similar objectives as emerging risk-sharing approaches

Diagnostic evidence-assessment methods are still evolving in most major markets…lack of best practices & homogeneity Test developed separately from drug vs. co-development

Uncertainty of acceptance of adaptive HEOR approaches

Where is the evidence threshold?...moving target problem

Challenges defining value & lack of value-based pricing opportunities

Education, adoption and acceptance barriers differ from traditional pharmaceutical paradigms

6/1/2010

3

Initial Research: Setting the Stage for Future Efforts of the PM SIG

Initial PM SIG paper will focus on:

Characterization of common gaps and challenges associated with personalized medicine

Describe perspectives key stakeholders along the continuum of evidence development and technology uptake

Consider how health economics and outcomes research (HEOR) methods can help address existing gaps and decision requirements for personalized medicine

Initial paper focused only on evaluation of these issues as related to pharmacogenomic scenarios

Initial Research: Setting the Stage for Future Efforts of the PM SIG

Key Perspectives

Researcher

Clinical

HEOR

Diagnostic Test Developer

Treatment Developer

R l t Regulator

Payer

How Can HEOR Methods Help to Close Gaps and Address Decision Maker Needs?

Linking personalized medicine challenges to HEOR solutions

Identifying best practices and closing gaps

Evaluating practice and policy implications

Dan Malone, RPh, PhDProfessorUniversity of Arizona Colleges of Pharmacy and Public HealthTucson, Arizona, USAMember, ISPOR Personalized Medicine Reimbursement & Development Working Group

Role of Diagnostics in Personalizing Health Care:

1. Risk stratification Patient MORE or LESS likely to develop disease/condition?

2. Inform treatment selection Is it safe? Is it effective?

3 Inform dosage

Source: Payne K. Fish and chips all around? Regulation of DNA-based genetic diagnotics. Health Economics 2009: 18:1233-1236:Faulkner E. Evolution of personalized medicine: Role of pharmacogenomics in managed care. NAMCP webinar series 2009.

3. Inform dosage Slow metabolizers Rapid metabolizers

4. Prognostic testing How likely is the patient to respond to standard treatments?

5. Treatment monitoring Is it working? Should we switch therapies or treatment strategies?

6. Improve or optimize clinical treatment pathways

(Currently) limited information on effectively integrating biomarkers into clinical practice

The PSA debacle

Emerging complex tests like Oncotype DX

V lidit tilit d l f bi k i d f Validity, utility, and value of biomarkers are required for clinical use, but often uncharacterized at the time of discovery

Requires significant effort to clinically validate a biomarker

Researcher incentives for characterize a new marker may differ from manufacturer, regulator, and payer incentives/information needs

Trial sample size and study design challenges

Many SNPs are rare

Large study sample sizes needed to test associations

Inclusion criteria may select for certain patients and may not be generalizable to broader population

Genomic consortiums

Warfarin Pharmacogenetics Consortium

Sub issues:

Credit for contributing data

Access to combined data

Common data elements

Standards for measurement

6/1/2010

4

Effect sizes between health outcomes and genotype are weaker than expected

Relationship between genotype and health outcomes can prove challenging

Researcher Perspective

Risk ratios: 1.2 to 1.6

Multiple exposures over time may occur before adverse effects observed

Relevance to clinical practice must be considered on case-by-case basis Source: Kelly et al. BMJ 2010;340:c693

Sample collection and storage

What genetic tests should be conducted?

UK Genetic Testing Network (400 genetic tests)

EurogenTest (> 1000 tests)

Researcher Perspective

Impractical for health organizations to collect and store data in the “hope” that is useful at some later point in time.

Few organizations with active genomic programs

Downstream challenges for health outcomes and economics researchers…a short list

Insufficient coordination between basic/academic research, clinical research & downstream decision maker’s information needs

Researcher Perspective

Limited or inconsistent data to support outcomes-focused evaluation…including insufficient data systems

Evolving, inconsistent and non-transparent standards/methods for trial design & modeling for diagnostics and drug-diagnostic combo products

Lack of implementation research to understand effectiveness of integrating Dx & PM into practice

Drug & Diagnostic Manufacturer Perspective

M J Finley Austin, PhDUS Head, External Innovation EnvironmentHoffman la RocheNutley, New Jersey, USAMember, ISPOR Personalized Medicine Reimbursement & Development Working Group

Drug & Diagnostic Manufacturer Perspective

Goal: Produce a therapeutic that is safe + efficacious within the context of a defined set of patients and clinical purpose

Performance expectations: Acceptable risk increases with disease severity

IP environment: Molecule patent protected

Exact duplicates barred during patent period

Goal: Produce a measurement that provides S&E clinically meaningful information within a defined clinical context

Performance expectations: Risk tolerance often inverse to that of therapies

IP environment: Platform vs. Analyte

W k d t l t

Drugs Diagnostics

Exact duplicates barred during patent period, work-a-rounds require significant R&D investment, face same regulatory hurdles and vary in ultimate result for patients

Regulatory environment: Must go through FDA review of efficacy and safety before entering market

Safety, efficacy, and more so clinical utility & economic value demonstrated with data derived from prospective RCTs. Surrogate end-points can facilitate testing. Claims dictated by study design and data

Work-a-rounds to measure proven analytes are often possible with small investment and LDTs can be used instead of a kit

Regulatory environment: Can enter market without FDA review. CLIA regulates labs but does not test validity

Studies to support use range from characterizing assay performance to literature bridging to retrospective and to prospective RCTs (last, very rare). Claims dictated by study design and data if FDA reviewed. LDT claims not reviewed

Business Considerations for Drug & Diagnostic Manufacturers

Business environment: Producing a S&E drug, meeting FDA and payer demands for evidence requires very large investment. High acceptance with strong data

Access/Gate Keepers: Patient

Business environment: R&D investment dependent on type of product and route to market but considerably less than drugs. Acceptance variable

Access/Gate Keepers: Patient presents physician requests lab

Drugs Diagnostics

Access/Gate Keepers: Patient presents, physician prescribes, pharmacist fills, payer pays if meets coverage criteria (label, other studies, professional recommendations)

Flexible pricing reflecting (often) value to patients

High failure rate, but most successes recoup investment and a few surpass it considerably

presents, physician requests, lab performs and reports, payer pays if meets coverage criteria (label, other studies, professional recommendations). DTC testing

Pricing often administered with little attention to value to patients

High success rate for incremental improvements to proven measurements; novel analyte success variable

6/1/2010

5

Key Drug Developer Issues

• Is a test needed? When is it known?• Hypothesized response variance?• Response variance detected in clinical development or in use?• Will the test improve the benefit, risk of the drug and outcomes for

patients (i.e., will it add clinical value)?• Clinical evidence, utility

• Trials design or additional studies necessary to support the targeted use?g

• Is the test feasible within the clinical context where it will be applied? Turn around time, etc.?

• Labeling• What will be the intended use in the drug, diagnostic labels?

• Inclusion in medical guidelines• Will test/drug change the practice of medicine?

• Regulatory approval of the test and test parameters• FDA (or CE) required or can LDT be used?• Reliability of the test – false positives and negatives?• Availability and accessibility of the test

• Added costs and potential RoI?

Key Diagnostic Developer Issues

• At what point in the drug lifecycle is test need identified?• What if the drug fails? Any stand alone potential?• What is the evidence to support the need for a test and that the

marker will work? Drug specific? Class specific?• Clinical evidence, utility

• Will the test improve the benefit, risk of the drug and outcomes for patients (i.e., will it add value)? Screening only? Monitoring?

• How should trials be designed to support the intended use?• Is the test feasible within the clinical context where it will be applied?

• Labeling considerations• What will be the intended use in the drug, diagnostic labels?

• Inclusion in medical guidelines• Will test change the practice of medicine?

• Regulatory approval of the test• Test parameters, reliability of the test, transition from RU assay to

clinically validated assay, timing?• Platform, expected reimbursement?

• Potential for fast followers?• Costs including lost opportunity, potential RoI?

John Watkins, RPh, MPH, BCPSPharmacy Manager, Formulary Development, Premera Blue Cross, Mountlake Terrace, WA, andClinical Associate Professor, Pharmacy, University of Washington, Seattle, WAMember, ISPOR Personalized Medicine Reimbursement & Development Working Group

Targeted therapies benefit narrowly-defined populations and are priced accordingly

Efficient use requires diagnostics that identify responders and reduce NNT

Cost-benefit analysis should capture adverse

Payer Perspective: Need for Companion Diagnostics

Cost-benefit analysis should capture adverse events avoided by not treating non-responders

Companion diagnostics will help targeted drug manufacturers develop a solid value proposition

Current Procedural Terminology (CPT) code set lacks specific codes for genetic tests

Tests billed under miscellaneous codes require manual review to identify

Manual review is not cost-effective for single marker

Payer Perspective: Inadequate Procedure Coding

gtests usually priced under $250

Therefore, U.S. payers cannot efficiently preauthorize most genetic tests

Risk panels are marketed to healthy individuals

Evidence of clinical utility is rarely available

Interpretation requires sophisticated knowledge

Erroneous interpretation of results can harm

Payer Perspective: Direct to Consumer Marketing

Erroneous interpretation of results can harm

We may not cover a test but still pay unnecessary cost of poorly informed treatment decisions

If a test does not lead to actions that improve the individual’s present or future health, it is difficult to argue its cost-effectiveness

6/1/2010

6

AmpliChip (Roche Diagnostics) $250 per test

Probably Cost-effective: Tamoxifen metabolism

Test cost = cost of 5-6 months of generic drug

Prodrug: poor metabolism poor response, patient not effectively treated & drug cost wasted

Payer Perspective: Case Example: CYP450 Testing

y g

Probably Not cost-effective: PPI metabolism

Rapid PPI metabolizers need high-dose PPI for effective H. pylori eradication

Incremental cost to double PPI dose: $20-40

Best to high dose everyone

Does it Improve Care + Result in Savings?Early Assessment from One BCBS Plan

Case Study: CareFirst BlueCross BlueShield internal return on investment analysis or Oncotype Dx and KRAS testing

Plan had medical & pharmacy data aggregated

Evaluated total tests completed X list cost of test vs. actual cost avoidance

Oncotype DX testing:yp g

Test cost $3,900

For every $1 spent, the plan estimated $2 in savings

KRAS testing:

Test cost $450

For every $1 spent, the plan estimated $18 in savings

When considered PTEN + BRAF + KRAS at $2,700 testing cost, ROI was still 1:5

Source: Winston Wong. CareFirst BlueCross BlueShiled. Incorporating Genomic Testing into Oncology Treatment Pathways. AMCP Annual Meeting April 2010.

How Can Health Outcomes & Economics Research Help?

David Veenstra, PhD, PharmDAssociate ProfessorUniversity of WashingtonSeattle, Washington, USAMember, ISPOR Personalized Medicine Reimbursement & Development Working Group

Research prioritization using VOI approaches

Create a robust research infrastructure to support conduct of observational studies of PM technologies

How Can HEOR Help?:Academic Research

Define an evidentiary framework to support decision-making regarding

Strategic decision making in development

Early stage modeling

Test performance thresholds

Support for value-based reimbursement

How Can HEOR Help?:Diagnostic Test Development

pp

Cost-effectiveness analyses

Pricing models

Use early stage modeling to understand potential impact of biomarkers on product efficacy and value

Facilitate participation in public-private partnerships to improve research

How Can HEOR Help?:Drug Development

infrastructure, methods, evidence framework

6/1/2010

7

Consider use of quantitative methods for assessing potential benefits and harms of genomic tests

Decision modeling

Discrete choice experiments

How Can HEOR Help?:Regulatory Decisions

Discrete choice experiments

Employ decision models of varying complexity to serve purposes ranging from providing a framework for discussions to projections of events rates to estimation of QALYs

How Can HEOR Help?:Technology Assessment & Payer Decisions

Develop evidence framework to assist decision makers in developing recommendations and decisions

Novel healthcare technologies offer great promise to improve population health, but identification of optimal strategies and efficient use is challenging

Personalized medicine is particularly

Conclusions

challenging because of the pace of technology advancements, often small effect sizes, and comparatively low regulatory hurdles/ease of patient access

Approaches used in health economics and outcomes research can facilitate prioritization of research topics, strategic development decisions, trial design, guideline development, and reimbursement decisions

However, many challenges remain, including

Transparent approaches to decision modeling acceptable to

Conclusions

Transparent approaches to decision modeling acceptable to multiple stakeholders

Assessment of patient preferences and ‘information for information’s’ sake

Potential misperceptions that there is a tension between comparative effectiveness research and personalized medicine

ISPOR members can play a role in:

Development of best practices in diagnostics and personalized medicine

Helping to streamline approaches for generating information to inform approval, reimbursement, adoption, and policymaking decisions across the continuum from early technology

Conclusions

y gydevelopment to patient access

Questions

How do we identify areas of greatest impact for PM and prioritize what to pursue?

How can we harmonize methodologies and expectations across markets for diagnostics & personalized medicine?

What different clinical & economic considerations are relevant in assessing PM technologies? What is currentlyrelevant in assessing PM technologies? What is currently missing?

Should all tests be subject to the same evidentiary expectations? If so, what is evidence is appropriate?

If some tests are subject to different evidence expectations, what approach/criteria should we use to decide/prioritize?

Can we harmonize…or will system differences prevent common approaches to assessing clinical/economic benefit?

6/1/2010

8

(Potential) Future Directions of the PM SIG

Evaluation of best practices for diagnostics and personalized medicine

Clinical & economic evidence development approaches

Technology assessment

Decision support approaches

Evaluation of case scenarios – What’s working/not working?

Consideration of and input on policy topics

Appropriateness of adaptive value characterization

Value-based pricing approaches

Harmonization of approaches

Comparative effectiveness vs. personalized medicine

Consumer-driven decision making

ISPOR Personalized Medicine Special Interest Group

Thank You!Thank You!For more information, please contact:

Eric Faulkner

Chair, ISPOR Personalized Medicine Special Interest Group

efaulkner@rti.org

919-541-6772