Approach to Outcome Measure Development or Selection: A Regulatory Perspective

1. Approach to Outcome Measure Development or Selection: A Regulatory Perspective Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT) April 17-18, 2014 Ashley F. Slagle, MS, PhD Study Endpoints and Labeling Development (SEALD) Office of New Drugs (OND) Center for Drug Evaluation and Research (CDER)

2. Disclaimer The views expressed in this presentation are those of the speaker, and do not necessarily represent an official FDA position.

3. Treatment Benefit Treatment benefit is demonstrated by evidence that the treatment has a positive impact on a concept of interest: How long a patient lives How a patient feels or functions in daily life Treatment benefit can be demonstrated as either: A comparative advantage in how patients survive, feel or function A comparative reduction in treatment-related toxicity

4. Purpose of Outcome Assessment • To determine whether or not a drug has been demonstrated to provide treatment benefit • A conclusion of treatment benefit is described in labeling in terms of the concept of interest (COI), the thing measured by the outcome assessment

5. Types of Outcome Assessments • Survival • Biomarkers • A physiologic, pathologic, or anatomic characteristic that is objectively measured and evaluated as an indicator of some normal or abnormal biologic function, process or response to a therapeutic intervention  • Clinical outcome assessments (COAs) • Performance outcomes (PerfOs) • Clinician-reported outcomes (ClinROs) • Observer-reported outcomes (ObsROs) • Patient-reported outcomes (PROs)

6. Choice of COA Type • Determine the most appropriate reporter for the COI in the COU • If symptom intensity is the concept of interest in a patient population that can respond themselves, a PRO is most appropriate. • If clinical judgment is required to interpret an observation, a ClinRO is chosen. • If the COI can only be adequately captured by observation in daily life (outside of a healthcare setting), and the patient cannot report for him or herself, then an ObsRO is chosen. • When it would be useful to observe an actual demonstration of defined tasks demonstrating functional performance in the clinical setting, a PerfO may be appropriate.

7. Evidence of Treatment Benefit • Direct evidence of treatment benefit is derived from studies with endpoints that measure survival, or how patients feel and function in daily life. • Indirect evidence of treatment benefit is derived from studies with endpoints that measure other things that are related to how patients survive, feel or function

8. Direct Verses Indirect Evidence of Treatment Benefit Survival Pain Breathlessness Blood Pressure PSA 6MWT Indirect Evidence Direct Evidence Evidence Continuum

9. Treatment Benefit: What To Measure?All are important, but interpretation of trial results depends on knowing how treatment impacts the core disease-defining concepts first. Proximal disease Impact concepts Distal disease Impact concepts Disease impact on general life concepts Disease-defining concepts General psychologicalfunctioning Productivity Core signs,symptomsor decrements in functioning Related functioning Additional functioning Health status General physical functioning Health-related quality of life Additional S/Ss Related S/Ss Social functioning Satisfaction withhealth

10. Adequate and well-controlled efficacy (A&WC) studies • Studies that provide: • Evidence to support drug marketing authorization • Substantial evidence of effectiveness • Required by law to support a conclusion that a drug is effective • See 21 CFR 314.126 • Deemed A&WC based on multiple features of a clinical study design including: • Nature of the primary endpoint • Well-defined and reliable • Rigor of control of the Type I error rate • Prospectively planned analyses designed with rigor

11. When is a COA adequate for use in adequate and well-controlled studies? • Regulatory standard: measures are well-defined and reliable • Empiric evidence demonstrates that the score quantifies the concept of interest in the targeted context of use • What does this mean? • This means measuring the right thing (concept of interest), in the right way in a defined population (targeted context of use), and the score that quantifies that ‘thing’ does so accurately and reliably, so that the effects seen in the outcome assessment can be interpreted as a clear treatment benefit.

12. Good Measurement Principles Defines good measurement principles to consider for “well-defined and reliable” (21 CFR 314.126) PRO measures intended to provide evidence of treatment benefit All COAs can benefit from the good measurement principles described within the guidance http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM205269.pdf 12

13. Well-defined and Reliable The tool adequately measures the concept of interest in the context or clinical setting of interest To assess this, we review the tool’s measurement properties: Content validity Construct validity Reliability Ability to detect change Information to support interpretation of change 13

14. Review of ClinRO, ObsRO, PRO, and PerfO Measures: Any Differences? SAME: Instrument Targeted Claims Endpoint Model Conceptual Framework Content Validity Other Measurement Properties Interpretation of Scores Language Translation and Cultural Adaptation Data Collection Method Modifications Clinical Trial Design and Data Analysis Issues Key References DIFFERENT: Nothing?

15. Clinical Outcome Assessment Considerations • Not all patient reported, clinical-reported, observer-reported, or performance outcome assessments are appropriate Clinical Outcome Assessments • May be useful for other purposes: • Diagnostic • Prognostic • Trial eligibility and trial enrichment • Epidemiologic or population studies • Clinical practice decision-making • Measures used successfully for these other purposes will not necessarily be appropriate outcomes assessments (i.e., they may not be able to reliably detect treatment benefit in clinical trials or support labeling claims in a non-misleading way)

16. Seeking Advice from FDA • Discuss plans early! • 2 pathways: • In the context of an Investigational New Drug (IND) program • Drug Development Tool (DDT) Qualification

17. DDT Guidance (Final January 2014) Describe a process NOT evidentiary standards Qualification process described for Biomarkers, Animal Models, and Clinical Outcome Assessments (COA) http://www.fda.gov/downloads/Drugs/GuidanceComplicanceRegulatoryInformationi/Guidances/UCM230597.pdf 17

18. CDER Qualification of Clinical Outcome Assessments • DDT Qualification Guidance: COA qualification is a conclusion that within the stated context of use (COU), the results of measurement can be relied upon to represent a specific concept (COI) with a specific interpretation when used in drug development and regulatory decision-making • Plain language: Within a specific clinical context, we’re measuring the right thing, in the right way, and we can rely upon the results of the qualified assessment across clinical trials within that clinical context • CDER qualification is currently reserved for those COAs that are ultimately intended to support primary or secondary endpoints in clinical trials • Qualified instruments shall be made available publically available

19. Roadmap to Patient-Focused Outcome Measurement in Clinical Trials • Intended to illustrate how one might embark upon a sound, orderly, instrument selection or development pathway, beginning with the clinical context in which the instrument is intended to be used.

20. Roadmap to PATIENT-FOCUSED OUTCOME MEASUREMENT in Clinical Trials 1 2 3 Understanding the Disease or Condition Conceptualizing Treatment Benefit Selecting/Developing the Outcome Measure • Identify the meaningful health aspect that is the intended benefit to patients in their daily lives • Survives (e.g., length of survival) • Feels (e.g., symptom severity) • Functions (e.g., walking ability) Natural history of the disease or condition • Onset/Duration/Resolution • Diagnosis • Pathophysiology • Range of manifestations • Search for existingclinical outcome assessment measuring the concept(s) of interest in the context of use : • • Measure exists • • Measure exists but needs to be modified • • No measure exists • • Measure under development • B. Identify the measureable concept of interest that represents the meaningful health aspect, which can be: • Equivalent to the meaningful health aspect (e.g., patients’ self-reported ambulatory activities in daily life) OR • Distinct from, but related to the meaningful health aspect (e.g., 6-minute walk test) Patient subpopulations • By severity • By onset • By comorbidities • By phenotype B. Begin clinical outcome assessment development • Document content validity (qualitative or mixed methods research) • Evaluate cross-sectional measurement properties (reliability and construct validity) • Create user manual • Consider submitting to FDA for qualification for use in exploratory studies Health care environment • Treatment alternatives • Clinical care standards • Health care system perspective • C. Define context of use for clinical trials, e.g.: • Disease/Condition entry criteria • Clinical trial design • Endpoint positioning C. Complete clinical outcome assessment development: • Document longitudinal measurement properties (construct validity, ability to detect change) • Document guidelines for interpretation of treatment benefit and relationship to claim • Update user manual • Submit to FDA for qualification as effectiveness endpoint to support claims Patient/caregiver perspectives • Definition of treatment benefit • Benefit-risk tradeoffs • Impact of disease • D. Consider appropriate clinical outcome assessment type(s): • Patient-Reported Outcome (PRO) • Observer-Reported Outcome (ObsRO) • Clinician-Reported Outcome (ClinRO) • Performance Outcome (motor, sensory, cognition) U.S. Food and Drug Administration Center for Drug Evaluation and Research Office of New Drugs http://www.fda.gov/Drugs Updated on March 14, 2014

22. Understanding the Disease or Condition

23. Understanding the Disease or Condition

24. Understanding the Disease or Condition

25. Understanding the Disease or Condition

26. FDA’s Patient Focused Drug Development Initiative • Systematically gather patients’ perspectives on their condition and available therapies to treat their condition • 20 public meetings over the course of PDUFA V, each focused on a specific disease area • http://www.fda.gov/ForIndustry/UserFees/PrescriptionDrugUserFee/ucm368342.htm

28. Conceptualizing Treatment Benefit

29. Conceptualizing Treatment Benefit

30. Defining Context of Use Each of the following variables can impact the adequacy of a COA to support a claim: Disease definition including, if appropriate Disease subtype Disease severity History of previous treatment Patient subpopulations Patient demographics Reporting ability Culture and language Clinical trial design and objectives Endpoint positioning Endpoint definitions Analysis plan Methods for interpretation of study results Targeted labeling claim Clinical practice and study setting Inpatient vs. outpatient Geographic location Clinical practice variation 30

31. Endpoint Definition and Positioning • Create study objectives based on the COI in the COU • Position the outcomes as trial endpoints that will be interpretable in comparison with a control group • Define endpoints using COA scores • Plan analysis • Measurement of change over time in individual patients that are combined for a means of assessing a group score • Analysis of means • Analysis of proportions • Hierarchy for testing multiple assessments

32. Conceptualizing Treatment Benefit

34. Selecting/Developing the Measure

35. Selecting/Developing the Measure

36. Selecting/Developing the Measure

37. COA Wheel and Spokes • This diagram identifies the key components of the documentation submitted to CDER to support COA qualification • The Wheel and Spokes diagram also represents the general iterative process used in developing a COA for qualification.

39. Spoke I

40. Spoke II

41. What Is Content Validity? • Content validity is the extent to which the content of an instrument represents important aspects of a given concept for an intended use and for a defined target population • Establishing content for a new instrument may involve both qualitative and quantitative research methods. Qualitative data are essential for establishing content validity of a COA • Input from the target population is essential

42. Conceptual Framework • An explicit description or diagram of the relationships between the questionnaire or items in an assessment and the concepts measured • Describes how the individual items contribute to the total score that will be analyzed and ultimately described in labeling

43. Conceptual Framework Score of Domain A Item 1 Item 2 Item 3 Item 4 Item 5 Item 6 Domain Concept A Total Score Overall Concept Score of Domain B Domain Concept B

44. Spoke III

45. Qualification for Use in Exploratory Studies / as Exploratory Endpoints At this point in time, submitters may consider the option of submitting evidence for COA qualification. Qualification at this point in development will be for use in exploratory analyses for purposes of testing other measurement properties.

46. Spoke IV

47. COA Qualification for Use as Primary or Secondary Endpoint When all measurement properties are tested, evidence will be reviewed to support COA qualification for use in adequate and well-controlled studies as a primary or secondary endpoint measure of effectiveness.

48. Spoke V

49. Conclusion The roadmap to a well-defined and reliable outcome assessment begins with a full understanding of the disease or condition to be tested An assessment cannot be chosen or developed without a well-defined context of use, understanding of the meaningful health aspect, and targeted concept of interest to be assessed The science of measurement continues to evolve with new tools and methods for efficient development and modification of assessments There is no one size fits all approach to measure development, we all must endeavor to be flexible, while applying good measurement principles as appropriate for each unique situation