Magesh Raghavan, T. A.

Women experience drug-induced Torsades de Pointes (TdP) at approximately twice the rate of men across more than 50 QT-prolonging drug classes, yet the quantitative ionic basis of this sex disparity remains incompletely characterised. The slow delayed rectifier current (IKs) is reduced by ~45% in female compared with male human ventricular cardiomyocytes, reducing the repolarization reserve available to compensate pharmacological IKr block. We implemented the O'Hara-Rudy (ORd) 2011 undiseased human ventricular epicardial action potential model in Python and parameterised sex variants using the most robustly established human ionic difference: GKs reduced by 45% in females (Kurokawa et al., 2016). We simulated graded IKr blockade (0-95% in steps of 5%) at three physiologically relevant pacing rates (2 Hz, 1 Hz, 0.5 Hz) after 60 beats of warm-up to approach electrophysiological steady state. Action potential duration at 90% repolarization (APD90), triangulation (APD90 - APD30), and repolarization failure (defined as APD90 > 500 ms, a conservative cellular risk marker informed by clinical QTc safety thresholds, or failure to repolarize within the cycle length) were quantified. All simulations used SciPy's Radau solver (rtol = 10-, atol = 10-) with a Numba-JIT-compiled right-hand side for computational efficiency. At baseline (0% block), the female model exhibited longer APD90 than the male at all pacing rates (+2.8 ms at 2 Hz; +4.6 ms at 1 Hz; +4.6 ms at 0.5 Hz). Under progressive IKr blockade, the absolute sex difference in APD90 amplified non-linearly: at 85% block and 1 Hz pacing the female APD90 exceeded the male by 60.4 ms (versus 4.6 ms at baseline; 13-fold amplification). At slow pacing (0.5 Hz), the sex gap was most pronounced: at 85% block, female APD90 was 1127 ms versus 939 ms for the male (+188 ms; 20% more prolonged). The critical APD threshold (>500 ms) was reached by female cells at 5 percentage points lower IKr block than male cells at 1 Hz pacing (55% vs. 60% block), both reported at the first simulated 5%-grid block level exceeding the criterion. Repolarization failure occurred 5 percentage points earlier in females at 1 Hz (90% vs. 95% block). Action potential triangulation was consistently greater in the female model at all block levels and pacing rates. A 45% reduction in IKs conductance is sufficient in this model to produce measurably greater APD90 prolongation under IKr blockade across all tested pacing rates. The non-linear amplification of the sex gap is consistent with the hypothesis that reduced IKs repolarization reserve contributes to greater female susceptibility to drug-induced QT prolongation, and supports testing sex-specific parameterizations in CiPA-style in silico cardiac safety workflows.