Biostatistic tools in pharmacogenomics--advances, challenges, potential.

Inter-individual variations in drug response are all-too common and, throughout medical history have often posed problems, many of them serious ones. The variations could stem from multiple factors, which include those of both the host (age, genetic and environmental factors) and disease (pathophysiological phenotypes, somatic mutations in case of cancers). The complex interplay of these factors can influence pharmacodynamic responses, such as adverse effects and efficacy, as well as pharmacokinetic manifestations through variability in drug absorption, distribution, metabolism and excretion. Recently, several potentially powerful tools to decipher such intricacies are emerging in various fields of science, and the translation of such knowledge to personalized medicine, called, in general, pharmacogenomics, has been promoted and has occasioned strong expectations from almost every sector of health care. However, at present, few biomarkers can predict which group of patients will respond positively, which will be non-responders and who might experience adverse reactions from the same medication and dosage. This review highlights several important aspects related to the design and statistical analysis for pharmacogenomics studies or clinical trials, which incorporate biomarkers. First, we review biomarker development: how biomarkers may be used as targets and the difference between prognostic and predictive markers. Second, in confirmatory clinical trials, we focus on issues related to study design for evaluating biomarkers and how they can be used to determine which patients might optimally benefit from a specific therapy. Finally, we review exploratory statistical screening techniques for detecting biomarkers in Phase I or pharmacokinetics studies.