Background: 

In the wake of the COVID-19 pandemic, various antiviral therapies have been clinically employed. Some of these are associated with significant proarrhythmic potential. To better understand the cardiotoxic side effects, we studied the electrophysiological effects of various antiviral therapies that are used during the COVID-19 pandemic using a Langendorff model of the isolated rabbit heart. 

Methods: 

52 hearts of New Zealand White rabbits were retrogradely perfused employing a Langendorff-setup. Eight catheters were placed endo- and epicardially, thereby recording monophasic action potentials. Hearts were paced at seven different cycle lengths (300-900ms), thus obtaining cycle-length dependent action potential duration at 90% of repolarization (APD₉₀), QT intervals and dispersion of repolarization. In addition, burst pacing was utilized to assess ventricular vulnerability. Thereafter, bradycardic AV-blocked hearts were perfused with a hypokalemic solution to enhance the occurrence of triggered activity (early afterdepolarizations and torsade de pointes). After generating baseline data, the hearts were assigned to four groups: In group 1 (HXC), hearts were treated with 1µM hydroxychloroquine. Thereafter, 3µM hydroxychloroquine were infused additionally. Group 2 (HXC+ AZI) was perfused with 3µM hydroxychloroquine followed by 150µM azithromycin. Group 3 (LOP) the hearts were perfused with 3µM lopinavir followed by 5µM and 10µM lopinavir. Group 4 (REM) was perfused with 1µM remdesivir followed by 5µM and 10µM remdesivir. 

Results: 

In group 1 hydroxychloroquine led to prolonged QT duration while leaving the APD unaltered. The arrhythmia susceptibility was significantly increased under 3µM hydroxychloroquine. 

The combination of hydroxychloroquine and azithromycin in group 2 let to a severe rise in QT-interval, APD and dispersion of repolarization accompanied by a further significant increase in arrhythmia susceptibility. The perfusion with lopinavir (group 3) induced a slight QT-prolongation and dispersion of repolarization while leaving the APD unaltered. Parallelly the arrhythmia susceptibility was non-significantly increased. Remdesivir (Group 4) induced a significant APD shortening while leaving the QT interval unaltered. The arrythmia incidence was tendentially but not significantly increased. In accordance with the reported bradycardia-events under remdesivir, remdesivir significantly reduced the rate of the ventricular escape rhythm. 

Conclusion:

We were able to show that mainly hydroxychloroquine- and azithromycin-based therapies have a significant proarrhythmic potential, with action potential prolongation and an increase in dispersion appearing to be causative. Lopinavir and remdesivir showed overall significantly less pronounced changes in cellular electrophysiology and concomitant less pronounced proarrhythmic effects.