Pharmacogenomic Discovery to Function and Mechanism: Breast Cancer as a Case Study
Biomedical research is in the midst of a period of rapid change, with the incorporation of a variety of omics technologies, the generation of extremely large datasets that result from the application of these analytical methods, and the need for novel computational approaches to deal with these large datasets. Pharmacogenomics represents one discipline within Clinical Pharmacology that has benefited significantly from these advances, advances that now make it possible to scan across the entire genome to identify genes associated with variation in drug response phenotypes and which will ultimately make it possible to sequence the entire genome of every patient being studied. As a result of these rapid technical developments, pharmacogenetics, a discipline that originated over half a century ago,1 and which initially focused on candidate genes that encode drug‐metabolizing enzymes, drug transporters, or drug targets,2 has evolved during the past decade into pharmacogenomics, with genome‐wide association studies (GWAS) that have identified genes that influence drug response with unfamiliar names such as ZNF423, MIR2052HG, and TCL1A.4 Those genes had not previously been associated with drug effect. However, the identification of sequence or structural variants in novel genes associated with variation in drug response is only the first step in a process that has reversed the standard pharmacogenomic approach that was applied only a few years ago. During the pre‐Genome Project era, we knew that phase I and phase II enzymes catalyzed the biotransformation of drugs, so we cloned and sequenced genes encoding drug metabolizing enzymes to determine whether variation in DNA sequence within or near those genes might be associated with variation in drug response.3 Today, we are able to use GWAS or next‐generation DNA sequencing (NGS) to discover unanticipated genes or DNA sequence variants that contribute to variation in drug response, but it is then necessary to pursue the underlying function of those variants and genes as well as mechanisms responsible for their association with drug response phenotypes. This reverse strategy, as illustrated subsequently by the results of germline pharmacogenomic studies of the endocrine therapy of breast cancer, can lead to novel insight into function and mechanisms that will facilitate the achievement of true precision medicine, either by enabling better selection of patients for a given therapy or by identifying new therapeutic targets. These principles and approaches will be illustrated by the results of a series of GWAS studies of the pharmacogenomics of the endocrine therapy of estrogen receptor positive (ER+) breast cancer. The focus in subsequent paragraphs will be on GWAS using the germline genome, although, obviously, the tumor somatic genome is also an area of intense study in breast cancer and many candidate gene studies have also been performed.12 Clearly, crosstalk between these two related genomes also contributes to variation in drug response. Therefore, the examples described here represents only one facet of genome‐wide pharmacogenomic discovery; in the case of the examples discussed subsequently, always followed by functional validation, mechanistic pursuit, and, eventually, clinical translation. Finally, we always need to bear in mind the fact that pharmacogenomic discovery, translation, and implementation are intimately interrelated processes, with each dependent on the other two (Figure1).
Breast cancer is the number one invasive cancer of women worldwide.13 In spite of major advances that have been made in the treatment of breast cancer, over 40,000 women die each year in the United States alone as a result of this disease (http://ww5.komen.org/BreastCancer/Statistics.html). A major advance in both the treatment and the chemoprevention of breast cancer was the realization that the growth and origin of many of these tumors is driven, at least in part, by estrogens.