P1‐047: REGIONAL COGNITIVE DIFFERENCES IN REFERRALS TO THE ELENBECESTAT MISSIONAD PHASE 3 PROGRAM IN EARLY ALZHEIMER'S DISEASE

developed using a similar Methodology to an established ADAScog disease progression model (Ref 1). Parameters describing placebo response were first estimated and fixed based on the published model and using the trial data from the placebo arm. Then Akaike Information Criteria (AIC) were used to select among alternative drug effect models (effects on the underlying progression rate vs an offset with no change in underlying progression rate). The best model was then tested for doseand exposure-dependency in the drug effect.Results: Models describing the drug effect as a detrimental offset occurring early in therapy without a change in underlying progression rate resulted in a large drop in AIC for ADAScog13 (-129), ADAS-cog11 (-105.7), and CDR-SB (-21.9) and a moderate drop in ADCS-ADL (-6.9) indicating a statistically favorable model (<-10.8 corresponds to p<0.001 for a nested comparison with 1df), while inclusion of a drug effect of altered progression rate resulted in little change in AIC. Addition of doseor exposure-dependency on the offset drug effect term resulted in little change in AIC for all 4 endpoints suggesting a lack of significant difference in drug effects across doses and drug exposures investigated. The magnitude of the drug effect offset at the 12 and 40 mg doses was estimated as 1.9 and 2.5 for ADAS-cog13, 1.4 and 1.9 for ADAS-cog11, 0.27 and 0.37 for CDR-SB, and -0.53 and -0.43 for ADCS-ADL, respectively.Conclusions: Overall, these Results indicate that the detrimental verubecestat drug effect in APECS occurred rapidly and did not alter the underlying disease progression rate, but rather acted as an offset. This drug effect was similar in magnitude across the doses and range of drug exposures studied. Ref 1. Clin Pharmacol Ther. 2015 Mar;97(3):210-4.