Statistical Evaluation of Novel Composite Endpoints in Early Alzheimer’s Disease
Peter Castellucio, Eli Lilly and Company   Veronika Logovinsky, Eisai Inc.   Peter Quarg, Novartis Pharmaceuticals  Nandini Raghavan, Janssen R&D  Angshuman Sarkar, Novartis Pharmaceuticals  David Shera, Merck & Co.  Tobias Verbeke, OpenAnalytics  *Jinping Wang, Eisai  Peng Yu, Eli Lilly  Xin Zhao, J&J Janssen 

Keywords: Novel Composite Endpoints, Early Alzheimer’s Disease, Clinical Endpoints Working Group

BACKGROUND No validated endpoints have been shown to be sensitive to disease progression and therapeutic effect in early stages of AD. Consequently, it is challenging to quantify clinically meaningful progression and treatment effects in such prodromal AD population. Recently, a few novel composite endpoints have been developed independently by a number of companies and are under evaluation by the Clinical Endpoints Working Group (CEWG) of the ADNI Private Partner Scientific Board (PPSB) through collaborative efforts. This collaboration is developing an R-package that can be utilized by participating companies to conduct standardized analyses on their data sets and will share the results without completely releasing proprietary data, thus allowing for companies to compare results across different internal data sets in a uniform manner.

NOVEL ENDPOINTS The novel composite endpoints along with methods used for developing these endpoints will be presented.

STATISTICAL EVALUATION The sensitivity of novel composite endpoints to measure disease progression and the sample sizes required for clinical trials using these endpoints were evaluated using a number of statistical methods. Statistical analysis methods used for evaluation and method for handling missing data will be discussed and preliminary results will be presented.