Prescription Medications and Abnormal Heart Rhythms: Understanding Genetics of Drug-Induced Long QT Syndrome

Abstract

Drug-induced long QT syndrome (diLQTS) is a serious side effect of many medications, which can lead to sudden cardiac death. While some clinical risk factors for diLQTS have been identified, they do not fully explain the risk, making diLQTS largely unpredictable. Previous evidence shows that both common (minor allele frequency [MAF] >1%) and uncommon ([MAF] <1%) genetic variants are associated with diLQTS risk. However, limitations of those previous studies, such as small sample size, lack of replication, and/or assessment in real-world clinical practice, prevent the clinical utility of genomics. The objective of this dissertation research was to investigate the associations between individual uncommon candidate genetic variants (KCNE1-D85N, KCNE2-I57T, and SCN5A-G615E) and a polygenic risk score (PRS) of common variants (as published by Strauss et al.) with diLQTS risk in a large observational case-control in a real-world clinical practice setting. A retrospective case-control study was conducted utilizing patients from Michigan Genomics Initiative (MGI) treated with high-risk QT-prolonging drugs and with an electrocardiogram (ECG) measurement during drug exposure. The primary outcome was an exaggerated prolongation of the QTc interval (>60 ms change from baseline and/or >500 ms absolute value) after exposure to a high-risk QT-prolonging drug, while the secondary outcome was a composite of torsades de pointes (TdP), any ventricular arrhythmia, or sudden cardiac death. The study found that KCNE1-D85N was significantly associated with diLQTS (adjusted odds ratio=2.24 [95% CI=1.35-3.58] p=0.001) in 6,083 self-reported white patients treated with 27 different high-risk QT-prolonging medications. However, due to low minor allele frequencies, there was insufficient power to analyze the associations of KCNE2-I57T and SCN5A-G615E with diLQTS. The study also validated a PRS for diLQTS published by Strauss et al. demonstrating its association with diLQTS in self-identified white patients. We lacked the statistical power to detect the PRS as a risk factor for diLQTS in African-American and Asian patients, since the sample sizes for those race groups were smaller than the white patients. Despite the association of KCNE1-D85N and the PRS with diLQTS in white patients, there was no association found between these genetic variants and the PRS with a composite secondary outcome comprised of history of TdP, any ventricular arrhythmias, and/or sudden cardiac death documented during the drug treatment period. A recursive partitioning data mining approach, Classification and Regression Trees (CART) was employed to investigate interactions among common and uncommon genetic variants and clinical factors. Among white patients, we identified a gene-gene interaction between rs3857067 and rs12025136, whereas among African patients, none of the genetic variants were associated with the risk of diLQTS. In conclusion, our study, in conjugation with multiple previous in vitro and in vivo studies, highlights the importance of considering KCNE1-D85N as a risk factor for diLQTS in future clinical practice guidelines. Implementing the PRS in routine clinical practice could help identify patients at high risk, allowing for proactive measures to prevent this life-threatening adverse drug reaction. The pharmacogenetic interaction between rs3857067 and rs12025136 should be validated in another study with adequate statistical power, using traditional statistical methods