The future role of pharmacogenomics in anticancer agent-induced cardiovascular toxicity

Toxicity to the heart and vascular system is a major concern with nearly all anticancer agents. Toxicity can range from drug-induced hypertension, atherosclerosis (peripheral artery disease) and venous thromboembolism, to QT prolongation, arrhythmias, myocardial infarction/ischemia, left ventricular dysfunction and heart failure, with obvious long-term consequences [1]. Incidence of toxicity is low in the case of venous thromboembolism and myocardial infarction (<10%), rising to moderate with left ventricular dysfunction (1–28%) and common with hypertension (4–61%) [2]. Toxicity also occurs across all classes of drugs including anthracyclines, monoclonal antibodies, tyrosine kinase inhibitors, platinum derivatives, uracil derivatives and proteosome inhibitors [1]. For many of these phenotypes, the drug mechanisms responsible are unknown and even whether this is an onor off-target effect is not established. Factors that might predict a patient’s disposition to anticancer agent-induced cardiovascular cardiotoxicity are far from clear, although there is a demonstrated role for cumulative dose and infusion rate for some drugs. Many of these toxicities also occur in children, increasing the likelihood a genomic basis of predisposition by reducing the influence of factors such as BMI, diet, lifestyle, smoking, exposures, comorbidities and polypharmacy. Combined, this has piqued interest in a pharmacogenomic approach to both genetically predict which patients will experience toxicity and personalize patient care, as well as provide insights into the drug toxicity mechanisms of action by highlighting genes involved, potentially informing protective drug discovery. Pharmacogenomic research has utilized increasingly sophisticated approaches to pinpoint genetic variants associated with a particular phenotype from candidate gene association studies (CGAS) and chip-based genome-wide association studies (GWAS) using tag SNPs, to candidate gene resequencing, expression qualitative trait loci (eQTL) mapping, exome sequencing and whole-genome sequencing (WGS). The immense evolution in sequencing techniques, data storage capacity and analysis pipelines has led to a dramatic decrease in the cost of WGS per individual from about $100 million USD in 2000 to as little as $600 USD [3] in 2017. The time required to sequence one human genome is now as little as 2 days, with commercial vendors offering DNA to WGS processed data turn-around in as little as 20–30 days [4]. These advances have shifted the trend in GWAS toward adapting WGS for variant discovery, providing data on ∼3 billion base pairs, rather than genotyping 500,000–2.5 million SNPs across the genome as with chip-based GWAS. Currently, there are several ongoing large-scale human genomics projects adapting WGS being held in many countries across the world primality targeted toward cancer and rare diseases, such as the Cancer Genome Atlas, the Cancer Genome Project, and the 100,000 Genomes Project. The capacity of these projects in terms of number of individuals varies ranging between a few thousand up to two million individuals [5]. The goal of the vast majority of these projects is to identify genetic variants that are responsible for interindividual variability in drug response, and