Multiple comparisons of drug efficacy between subgroups defined by genetic polymorphisms


Yi-Hsuan Tu, Department of Statistics, National Cheng Kung University, 70101 Tainan, Taiwan.



In contrast to using genome-wide association studies to discover associations between genes or single-nucleotide polymorphisms in the genome with disease status or outcome, some recent pharmacogenomic studies have focused on whether polymorphisms in genes involved in metabolizing drugs significantly impact their efficacy. Whether a drug starts as an active compound and gets metabolized and eliminated from the body or starts as an inactive compound and gets metabolized to an active form, patients in subgroups separated by polymorphism of a gene needed to metabolize the drug might derive differential benefit from that drug. With the use of the Clopidogrel in Unstable Angina to Prevent Recurrent Events trial for Plavix as an example, this article proposes Multiple Comparisons with Control Subgroup and Multiple Comparisons with the Best Subgroup as methods to infer whether some subgroups of patients derive more or less benefit than wild-type patients and which subgroup or subgroups of patients derive maximum benefit or practically maximum benefit from the drug. Copyright © 2012 John Wiley & Sons, Ltd.