Transmural and rate-dependent profiling of drug-induced arrhythmogenic risks through in silico simulations of multichannel pharmacology

Background In vitro hERG blockade assays alone provide insufficient information to accurately discriminate “safe” from “dangerous” drugs. Recent studies have suggested that the integration of multiple ion channel inhibition data can improve the prediction of drug-induced arrhythmogenic risks. In this study, using a family of cardiac cell models representing electrophysiological heterogeneities across the ventricular wall, we quantitatively evaluated transmural and rate-dependent properties of drug-induced arrhythmogenicity through computer simulations of multichannel pharmacology. Methods and Results Rate-dependent drug effects of multiple ion channel inhibition on cardiac electrophysiology at their effective free therapeutic plasma concentrations (EFTPCs) were investigated using a group of in silico cell models (Purkinje (P) cells, endocardial (Endo) cells, mid-myocardial (M) cells and epicardial (Epi) cells). We found that (1) M cells are much more sensitive than the other cell types to drug-induced arrhythmias and can develop early afterdepolarization (EAD) in response to bepridil, dofetilide, sotalol, terfenadine, cisapride or ranolazine. (2) Most drug-induced adverse effects, such as pronounced action potential prolongations or EADs, occur at slower pacing rates. (3) Although most drug-induced EADs occur in M cells, the application of quinidine at its EFTPC can cause EADs in all four cell types. (4) The underlying ionic mechanism of drug-induced EADs differs across different cell types; while I NaL is the major depolarizing current during the generation of EAD in P cells, I CaL is mostly predominant in other cell types. (5) Drug-induced AP alternans with larger beat-to-beat variations occur at high pacing rates in mostly P cells, while the application of bepridil can cause alternating EAD patterns at slower pacing rates in M cells. Conclusions In silico analysis of transmural and rate-dependent properties using multichannel inhibition data can be useful to accurately predict drug-induced arrhythmogenic risks and can also provide mechanistic insights into drug-induced adverse events related to cardiac arrhythmias. Author summary In vitro hERG blockade assays alone provide insufficient information to accurately discriminate “safe” from “dangerous” drugs, and computer simulation of ventricular action potential using multichannel inhibition data could be a useful tool to evaluate drug-induced arrhythmogenic risks. Our study suggested that the profiling of drug-induced transmural heterogeneities in cellular electrophysiology at all physiological pacing frequencies can be essential for the comprehensive evaluation of drug safety, and for the quantitative investigation into ionic mechanisms underlying drug-specific arrhythmogenic events. These in silico models and approaches may contribute to the ongoing construction of a comprehensive paradigm for the evaluation of drug-induced arrhythmogenic risks, potentially increase the success rate and accelerate the process of novel drug development.