We demonstrate a method for the prediction of chemotherapeutic response in patients using only before-treatment baseline tumor gene expression data. First, we fitted models for whole-genome gene expression against drug sensitivity in a large panel of cell lines, using a method that allows every gene to influence the prediction. Following data homogenization and filtering, these models were applied to baseline expression levels from primary tumor biopsies, yielding an in vivo drug sensitivity prediction. We validated this approach in three independent clinical trial datasets, and obtained predictions equally good, or better than, gene signatures derived directly from clinical data.
Bibliographical noteFunding Information:
The authors thank Keston Acquino-Michaels and Zaya Amgaabaatar for verifying the R code and reproducing all results and figures. The authors also thank Jana Heitmann, Dr Brian Stewart and Prof Cathal Seoighe for their critical review of the manuscript. The Cancer Genome Project funded by the Wellcome Trust Sanger Institute generated and made all CGP cell line drug sensitivity and baseline expression data publicly available. This study is supported by the National Institutes of Health/National Institute of General Medical Science (Pharmacogenomics of Anticancer Agents grant U01GM61393). RSH also received support from the National Institute of General Medical Science K08 (GM089941), a Circle of Service Foundation Early Career Investigator award, the National Cancer Institute R21 (CA139278), a University of Chicago Cancer Center Support Grant (#P30 CA14599), a University of Chicago Breast Cancer SPORE Career Development Award (CA125183), a Conquer Cancer Foundation of ASCO Translational Research Professorship award in memory of Merrill J Egorin, MD and the National Center for Advancing Translational Sciences of the National Institutes of Health (UL1RR024999).