TL35: Regulatory impact and issues of joint modeling of longitudinal and time-to-event
*Mark Ernest Boye, Eli Lilly and Company  *Ping Wang, Eli Lilly and Company  

Keywords: model, survival analysis, longitudinal data analysis, surrogate endpoint

Researchers often collect event-time and repeated-measure longitudinal data (e.g., biomarkers and quality of life measurements) that may result from a common data generating process. Joint modeling (JM) of these data types can reveal the relationship between features of the trajectory and the survival outcome, and may result in less biased and more efficient estimates. Although commonly found in the statistical literature, there are few published substantive bio-medical JM applications. To address this discrepancy between this benefit and disuse, we consider the value and concerns of including JM in development programs designed to evaluate the clinical benefit of a pharmaceutical or biomedicine. Issues include: 1. Candidate applications and therapeutic areas that could most benefit. 2. Barriers to regulatory acceptance. 3. A path forward for inclusion of JM in regulatory submissions.