Introduction

The shapes of plasma concentration versus time (Cp-time) profiles from large clinical trials are often highly variable, even in well-controlled trials involving homogeneous cohorts. Sources of variability include genetic, developmental, demographic, dietary, and lifestyle differences. Pharmacokinetic scientists are tasked with assessing the variabilities in these data sets and identifying mechanistic sources of those variabilities. Rapidly processing data to identify and graphically depict trends is a standard practice in pharmacokinetic modeling which, when properly implemented, can effectively guide modeling efforts. Kohonen self-organizing maps are an unsupervised clustering tool for organizing heterogeneous data, and this poster demonstrates that they can be a tool that pharmacokinetic scientists may find useful in understanding complex, highly variable clinical trial results.

17th North American Regional ISSX Meeting, October 16 – 20, 2011, Atlanta GA

By Jason L. Boyd, Robert Fraczkiewicz, Grazyna Fraczkiewicz, Robert D. Clark, and Walter S. Woltosz