Dapagliflozin but not empagliflozin prevents atrial fibrillation in an experimental whole-heart model

Abstract

Abstract Background The SGLT2 inhibitors dapagliflozin and empagliflozin have an evolving role in heart failure therapy. Observational data suggest a lower risk of atrial fibrillation (AF) for patients on SGLT2 inhibitors. So far, it is unclear whether this is related to a direct antiarrhythmic potential of SGLT2 inhibitors or just a marker of improved heart function. Experimental data on the impact of SGTL2 inhibitors on atrial electrophysiology is scarce. Purpose This study aimed at evaluating the effects of the two SGLT2 inhibitors dapagliflozin and empagliflozin on atrial electrophysiology. Methods 25 rabbit hearts were explanted and retrogradely perfused employing the Langendorff-setup. Two catheters were placed on both atria to record monophasic action potentials. Hearts were paced at three different cycle lengths (350ms, 250ms, 150ms), thereby obtaining cycle-length dependent atrial action potential durations (aAPD90) and effective refractory periods (aERP). Atrial post-repolarization (aPRR) refractoriness was determined by the difference between aERP and aAPD90. AF was induced in the presence of acetylcholine (ACh, 1μM) and isoproterenol (Iso, 1μM). Thereafter, hearts were treated with dapagliflozin (n=13, 3 μM) or empagliflozin (n=12, 3 μM). Results Ach/Iso abbreviated aAPD90, aERP and prolonged aPRR in both groups (dapagliflozin group: aAPD90: −49 ms, p<0.01; aERP: −33 ms, p<0.01; aPRR: +18 ms, p=0.06; empagliflozin group: aAPD90: −40 ms, p<0.01; aERP: −25 ms, p<0.05; aPRR: +18 ms, p=0.05). Further treatment with dapagliflozin increased aAPD90 (+13 ms, p<0.01), aERP (+34 ms, p<0.05) and aPRR (+6 ms, p=ns). Additional administration of empagliflozin had no significant effect on aAPD90 (−8 ms, p=ns), aERP (−1 ms, p=ns) or aPRR (+9 ms, p=ns). Vulnerability to AF was assessed by a standardized protocol consisting of several trains of burst pacing. Ach/Iso increased the vulnerability to AF episodes in both groups (dapagliflozin group: 71 episodes vs. 1 under baseline conditions, p<0.01; empagliflozin group: 27 episodes vs. 2 under baseline conditions, p=0.08). Additional perfusion of dapagliflozin significantly reduced the occurrence of AF (to 30 episodes, p<0.05) while additional empagliflozin treatment had no significant antiarrhythmic effect (32 episodes, p=ns). Conclusion Dapagliflozin and empagliflozin had opposing effects on atrial electrophysiology. Dapagliflozin prolonged aAPD90 and aERP, thereby preventing AF. In contrast, empagliflozin did not change aAPD90 or aERP, resulting in no antiarrhythmic effect. Our findings hint at a potential antiarrhythmic benefit of dapagliflozin. Funding Acknowledgement Type of funding sources: Private grant(s) and/or Sponsorship. Main funding source(s): Hans and Gertie Fischer-Foundation (to C.E.)