|
Activity Number:
|
476
- SPEED: Clinical Trial Design, Longitudinal Analysis, and Other Topics in Biopharmaceutical Statistics
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2018 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biopharmaceutical Section
|
|
Abstract #329902
|
Presentation
|
|
Title:
|
Statistical Models for Longitudinal Analysis of Preclinical Efficacy Screens
|
|
Author(s):
|
William Forrest* and Bruno Alicke and Oleg Mayba and Alice Starr
|
|
Companies:
|
Genentech, Inc and Genentech and Genentech and Genentech
|
|
Keywords:
|
longitudinal;
oncology;
preclinical
|
|
Abstract:
|
Preclinical in vivo efficacy screens are commonly employed with aims such as rank ordering of candidate molecules, estimation of potency, or detection of synergy in combination therapies. A typical design is a one-way layout or dose-escalation study with multiple mice per group and repeated measures per animal. Regimen efficacy is summarized at the group level and experimental goals addressed by comparing summary statistics from relevant groups. Alternatively, efficacy may be summarized at the level of a test animal and summary statistics across groups analyzed with standard statistical techniques such as linear modeling. We develop a framework for this one-way layout study design in preclinical oncology and illustrate a statistical workflow based on generalized additive mixed models. Various summary statistics derived from fitted regression spline models are proposed and evaluated on both simulated and real data. Implementation and visualization of results are developed via extensions of open-source tools in the R language.
|
Authors who are presenting talks have a * after their name.
