Translation Pathway for Coronary Stent Development- Clinical Endpoints

1. Translation Pathway for Coronary Stent Development-Clinical Endpoints Donald E. Cutlip, MD Beth Israel Deaconess Medical Center Baim Institute for Clinical Research Harvard Medical School

2. Disclosure Statement of Financial Interest I, Donald Cutlip, DO NOT have a financial interest/arrangement or affiliation with one or more organizations that could be perceived as a real or apparent conflict of interest in the context of the subject of this presentation.

3. Historical Perspective • Mature technology with over 20 years clinical trial experience • Evolving perspective on true benefits and risks of coronary stenting • From lesion/device focus (restenosis, PPMI) to patient oriented outcomes (Death, Any MI, Any Revascularization)

4. General Principles • Consensus of investigators, industry, and regulatory groups on determination of meaningful endpoints • Standardized definitions • Indirect comparisons between trials and devices • Facilitate regulatory review • Central independent adjudication • Allow for assessment of regulatory (safe and effective) and reimbursement (reasonable and necessary) objectives

5. Limitations of Individual Endpoints • Low frequency events and smaller differences require larger trials • Measures of device effectiveness (TLR) may not account for more serious competing risks (death, MI, stent thrombosis) • Composite endpoints such as “MACE”, “TVF”, “TLF” have generally been adopted to overcome some of these limitations.

6. Composite Endpoints – Use with Caution • Components should track in same direction • Increase event rate and magnify differences • If not, then may obscure differences and reduce power • Loss of statistical power for analysis of individual components. • Fail to account for differences in endpoint severity leading to flawed trial interpretation.

7. What about surrogate endpoints? • Useful for mechanistic implications and early feasibility but insufficient for assessment of overall safety and effectiveness • Late loss: Excellent discrimination of restenosis prevention but not stent thrombosis or disease progression. • OCT/IVUS: Correlation with healing and late safety – unlikely to replace hard clinical endpoints

8. Assessing Value and Patient-Reported Outcomes • What is the impact of stenting versus other treatments on global patient outcomes? • Previous analyses have focused on cost-effectiveness as a measure of healthcare value and been based on device oriented outcomes . • Patient oriented outcomes such as mortality, stroke, MI of clear interest, but differences between therapies unlikely or too small to measure.

9. Patient-Reported Outcome Measures (PROM) • Validated PROMs provide added assessment of outcomes that are important to patients. • May allow discrimination between therapies that appear similar based on usual clinical endpoints from reasonably sized clinical trials and usual follow-up duration. • Useful for assessment of overall healthcare value of competing therapies and incorporated into decisions of reasonable and necessary.

10. Patient-Reported Outcome Measures (PROM) • Validity – Evidence for measurement of the concept of interest. • Reliability – Results are reproducible. • Discrimination – Detects change and meaningful differences between groups. • SAQ, KCCQ and MLHFQ among PROMs approved by FDA.

11. Impact of Trial Designs on Endpoint Selection and Assessment • Comparator – Stent or alternative Therapy (medical/surgery) • Definitions consistent with both therapies • Endpoints to assess meaningful differences (eg, stroke, bleeding for anti-thrombotic comparisons) • Duration of Follow-up – Is therapy associated with late benefit. Endpoint must allow for late assessment. • Large, simple trials – Issues for endpoint assessment and adjudication. Pre-market vs. post-market.

12. Device Oriented Endpoints

13. Patient Oriented Endpoints