March 10th, 2014 Efficacy Endpoints for Anticonvulsant Therapies Interview 1

Agenda Guidance from Project Advisory Board Overview of Concept Map (CMAP) objectives CMAP review Questions related to review concepts Next steps

The Project Advisory Board has emphasized modeling efficacy-specific concepts The focus of the CMAP should be on concepts directly involved in efficacy determination E.g. proportion of subjects achieving spasm cessation, % change in 28-day total partial seizure frequency, etc. In general, derived data will reside in ADaM, while this effort will focus on standardizing collected CRF data Only secondary endpoints which are important for labeling should be considered for inclusion on the CMAP Covariates which are (1) unique to DNP, (2) used in determination or validation of efficacy, should be included on the CMAP

Superiority studies determine efficacy by comparing event frequencies between study arms Trials generally compare changes in event (e.g. spasm, seizure) frequency or occurrence between placebo and dosage arms E.g. % change in seizure frequency, spasm count reduction Changes are generally assessed within each study arm comparing to a baseline timepoint E.g. % change in total partial seizure frequency from baseline to double-blind phase Frequency or occurrence of events are determined over temporal phases E.g. % change in total partial seizure frequency per 28 days Statistical test determines significance of difference in event between groups E.g. ANCOVA, Chi-square test

CMAP review

Open questions based on the CMAP… Are timepoints and phases associated with endpoints protocol specific? E.g. % change in total partial seizure frequency per 28 days from baseline to double-blind phase What is the temporal distinction between a timepoint and a phase? E.g. is the number of days in the baseline phase important? Does 28 days correspond to the length of the double-blind phase? For clinical assessments, are these compared to an earlier timepoint? E.g. Physician global assessment – is this assessed over a period of time between arms, or within an arm between timepoints? What are the numeric scores which correspond to clinical assessments? Are these important to capture? E.g. Physician global assessment, Clinical global impression, Behavior, Well being

Open questions based on the CMAP (continued)… Is it important to capture the source of event counts or occurrence? E.g. Spasms being recorded by Caregiver vs. Patient, observation via Diary vs. CCTV EEG, etc. Can Partial Seizures and Infantile Spasms be considered subtypes of general seizures? Can endpoints be generalized for adult and pediatric populations? E.g. can % change in IS frequency be considered equivalent to % change in partial seizure frequency, since % change in an event is being calculated?

Subsequent CMAP iterations and modifications… Based on feedback today, this CMAP will be updated to reflect proposed changes A definition file, where each concept on the map not currently in another data standard (e.g. SDTM) will be created Concepts specific to DNP will be defined here and flagged to be added to a data standard Subsequent interviews intend to involve reviewing CMAP modifications and verifying concept definitions