Abstract 16962: A Polygenic Predictor of Baseline QT is Associated With Sotalol-Induced QT Prolongation

Abstract

Background: Drug-induced QT interval prolongation raises the risk of fatal arrhythmia and is a central issue in safety pharmacology. Yet individual QT drug response is highly variable, and risk stratification remains a challenge. We hypothesized that genetic factors underlying the baseline QT influence this variation in QT drug response. Methods: A cohort of 990 healthy subjects was prospectively challenged with a single 80 mg oral sotalol dose. Exclusion criteria included abnormal ECG, cardiac disease, and QT-prolonging drug use. ECGs were obtained at baseline and 3 hours post-dose. QT response was defined as absolute change in Fridericia heart-rate corrected QT (ΔQTc). Plasma sotalol was measured 3 hours post-dose. Subjects were genotyped on the MEGA array. Principal component analysis of genotypes was used to define genetic ancestry. A polygenic score (PGS) for the baseline QTc comprised of 465,399 common variants was calculated from assayed and imputed genotypes. The difference in PGS between subjects with high QT response (ΔQTc≥60 ms) and those with ΔQTc<60ms was assessed by Mann Whitney test. A multivariable regression model was used to assess the association between PGS and ΔQTc after covariate adjustment. Results: Of the 990 subjects, 952 (96%) passed genomic quality control and were included in analysis. 62% were female (n=590), with median age 23 (interquartile range (IQR) 21-32). Self-reported ancestry was 89% European (n=850) and 11% North African (n=102). The median ΔQTc 3 hours post-sotalol was 21 ms (IQR 12-30). Ten subjects (1%) had ΔQTc≥60 ms, and their PGS was significantly higher compared to those with ΔQTc<60 ms ( P =0.0097, Figure 1A). In the regression model, PGS was positively associated with ΔQTc ( P =0.0007, Figure 1B). Conclusion: Common genetic variants associated with the baseline QT also capture part of the repolarization response to sotalol. This adds to the emerging concept that the genetic architecture of a trait can predict drug response.