Predicting Sex-Specific Antiarrhythmic Strategies for Atrial Fibrillation through a Regression-Guided Computational Modeling Pipeline

Atrial fibrillation (AF), the most common sustained cardiac arrhythmia, is a major contributor to stroke, heart failure, and mortality worldwide. Although AF affects both men and women at a similar rate, accumulating experimental and clinical evidence indicates that its underlying mechanisms, disease progression, and treatment responses differ by sex. However, current antiarrhythmic drug development and clinical management of AF remains largely sex neutral, likely contributing to limited efficacy and increased adverse effects. To address this gap, we developed a computational drug-screening pipeline based on experimentally constrained, sex-specific human atrial cardiomyocyte models to predict and evaluate sex-specific pharmacological strategies for AF. The pipeline integrates multivariable regression with mechanistic modeling to systematically test multi-target combinations of ion channel inhibitors and Ca²⁺ handling modulators and identify interventions that reduce arrhythmia vulnerability by restoring sex-specific electrophysiological and Ca²⁺ handling properties toward normal sinus rhythm (nSR). Application of this approach revealed a greater number of successful inhibitory drug combinations in males than in females. In males, optimal recovery to nSR primarily required inhibition of Na⁺ and K⁺ channels to prolong repolarization and refractoriness, increase Ca²⁺ transient amplitude (CaT_Amp), and reduce susceptibility to action potential duration (APD) alternans. In females, modulation of Ca²⁺-related pathways was additionally required to suppress delayed afterdepolarizations (DADs). Forward single-cell simulations confirmed the predictions of the drug-analysis pipeline, demonstrating recovery of APD, CaT_Amp, and arrhythmia vulnerability indices without introducing instabilities. Importantly, extension of these interventions to two-dimensional atrial tissue simulations demonstrated that sex-specific drug strategies reduce vulnerability to triggered activity, while suppression of reentry was most effective when combined with partial recovery of cell-cell coupling. Our results establish a multiscale computational pipeline for identifying sex-informed, multi-target antiarrhythmic therapies, amenable to experimental validation and translation to the clinic.