Anti-arrhythmic effects of Empagliflozin on human atrial myocytes of HFpEF patients

SGLT2-Inhibitors (SGLT2i) have shown a reduction in cardiovascular mortality and heart failure (HF) hospitalization in patients with and without type 2 diabetes. The mechanisms are not fully understood. Ca/Calmodulin dependent kinase II (CaMKII), which plays a key role in HF development due to stimulated late-sodium current (late I_Na), was shown to be inhibited by empagliflozin resulting in a decrease in late I_Na. Since an increase in late I_Na is considered a proarrhythmogenic factor potentially causing early and late depolarizations (EADs and DADs), its reduction by SGLT2i implies that cardiac arrhythmic activity may be reduced, which has been suggested in several clinical studies. Recently it was shown that similarly to HFrEF patients late I_Na is significantly higher in HFpEF patients. In this study we wanted to further investigate how action potential length and arrhythmic activity in HFpEF patients compares to Non-Failing (NF) patients and whether an anti-arrhythmic effect of SGLT2i could be observed in this patient group.

Human tissue samples from the right atrial appendage were obtained from 50 patients undergoing elective heart surgery and human atrial myocytes were isolated and used for patch-clamp experiments. 31 patients were classified as HFpEF patients according to ESC guidelines, 8 were considered NF. Cells were exposed to empagliflozin (10 µM), specific CaMKII inhibitor AIP, and Na-channel blocker tetrodotoxin (TTX). Current-clamp and voltage-clamp patch-clamp methods were used to measure action potentials, including EADs and DADs, and late sodium current.

Interestingly, a statistically significant increase in both APD30 (Fig. A: NF vs. HFpEF; 10.79 ± 1.603 ms vs. 16.83 ± 1.638 ms, p=0.0178, unpaired t test with Welch’s correction) and APD80 (Fig. B: NF vs. HFpEF; 55.95 ± 9.878 ms vs. 113.9 ± 14.91 ms, p=0.0036, unpaired t test with Welch’s correction) could be observed for HFpEF patients compared to NF patients. As expected, exposure to empagliflozin resulted in a decrease in pathologically enhanced late Na current in HFpEF patients (Fig. C&D: HFpEF+veh vs. HFpEF+empa; −45.69 ± 7.41 A*ms*F−1 vs. −15.67 ± 3.7 A*ms*F−1, P = 0.0187, repeated measures one-way ANOVA with Holm-Sidak’s post-hoc correction). However, this inhibition did not result in shortening of action potential duration (APD30: HFpEF+veh vs. HFpEF+empa; 16.83 ± 1.638 ms vs. 16.99 ± 1.618 ms, p=0.9435, unpaired t test with Welch’s correction; APD80:HFpEF+veh vs. HFpEF+empa; 113.9 ± 14.91 ms vs. 131.7 ± 15.38 ms, p=0.4109, unpaired t test with Welch’s correction). Still, cells that displayed arrhythmic activity in the HFpEF vehicle group showed a lower incidence of EADs and DADs when treated with SGLT2i (Fig. E: DAD+EAD: HFpEF+veh vs. HFpEF+empa, 0.08297 ± 0.04434 1/s vs. 0.009341 ± 0.006273 1/s, p=0.015, Wilcoxon matched-pairs signed rank test; Fig. F: DAD: HFpEF+veh vs. HFpEF+empa, 0.056591 ± 0.03964 1/s vs. 0.007143 ± 0.005337 1/s, p=0.0156, Wilcoxon matched-pairs signed rank test).

Discussion and Conclusion

Our data suggest that SGLT2i may reduce arrhythmic activity in human atrial cardiomyocytes of patients with HFpEF, potentially by inhibition of increased late I_Na in these cells. Noteworthy, although empagliflozin inhibits increased late I_Na, this did not suffice to reduce the observed APD prolongation in these cells, suggesting additional alterations in potassium channel or calcium channel conduction in HFpEF atria.