PII-172 - IN SILICO PREDICTION OF ION CHANNEL EFFECTS OF METHADONE LEADING TO QT INTERVAL PROLONGATION.

J. Lindsey¹, M. Yue², S. Quinney³, J. Tisdale⁴, B. Overholser⁴; ¹Purdue University, ²Purdue University, West Lafayette, IN, United State, ³School of Medicine, Indiana University, Indianapolis, IN, USA, ⁴College of Pharmacy, Purdue University, Indianapolis, IN, USA.

Background: R,S methadone prolongs the QT interval, a risk factor for arrhythmias. Methadone’s QT prolonging ability has been attributed to I(Kr) inhibition, but recent studies show inhibition of other ion channels. Given the difficulty in delineating specific ion channels contributing to QT prolongation for drugs with multiple effects, the goal was to use mathematical modeling and simulation of cardiac electrophysiology to predict specific ion channel effects for methadone induced QT prolongation.

Methods: A cardiac action potential (AP) model was developed and used for pseudo-ECG simulations using the Cardiac Safety Simulator version 2.1 (Certara, Sheffield, UK). Single cell APs were simulated using published in vitro ion channel effects (e.g. I(Kr), I(K1), and I(Ca,L)) with R,S methadone (EC50 and h). A stepwise process of adding and omitting ion channel effects was used to assess individual ion channels. The Ten Tusscher model was used and APD90 calculated for comparison to isolated cardiomyocyte studies for model refinement. The developed AP model simulated pseudo-ECGs and heart-rate correct QT (QTc) were compared to published clinical trial data. For clinical trial simulations, the String Length Sjorgen model was used for differences in ventricular wall thickness among sexes and different ages. The simulation period (RR interval) varied between 800 ms to 1000 ms with 1.5% circadian rhythm variability.

Results: The final AP model included inhibitory effects on I(Kr), I(K1), and I(Ca,L). Clinical trials of 20 patients were simulated to generate pseudo-ECGs. Each patient’s QTc was calculated at 26 concentrations. The mean (SD) for observed QTc were 411 (22), 414 (25), 421 (22), and 438 (26) ms at 0-0.6, 0.6-1.0, 1.0-2.0, and 2.0-3.0 mcM of R,S methadone, respectively. The model (Figure) adequately characterized QTc intervals within 2.7% of the observed data at all concentration ranges. The model was 1.5 and 2-fold more accurate at predicting QTc intervals than lone I(Kr) / I(K1) inhibition at 0.6-1.0 and 1.0-2.0 mcM, respectively.

Conclusion: Through modeling and simulation, R,S methadone best mimicked (1) single cell APD90 and (2) clinical trial QTc intervals using reported in vitro effects on I(Kr), I(K1), and I(Ca,L). Further exploration of the role of stereoselective methadone block on the interplay of these ion channels is warranted.