Introduction
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has to date infected ~110 million people worldwide (December 2020), resulting in ~2.5 million deaths ( COVID-19 Dashboard, Johns Hopkins University). There are currently no small molecule medications approved to prevent or treat COVID-19. Throughout 2020, there was considerable interest in the use of hydroxychloroquine and chloroquine, two drugs used for treatment of lupus erythematosus, other inflammatory disorders and malaria, which are also effective in inhibiting replication of SARS-CoV-2 in vitro (1). Open-label observational studies purported to show therapeutic benefit for these drugs either when used as a monotherapy, or in combination with the macrolide antibiotic azithromycin (2). Consequently, the Food and Drug Administration (FDA) gave limited authorisation for emergency use of hydroxychloroquine in COVID-19 patients. However, this was subsequently been withdrawn with interim results from major clinical trials reporting negative results for efficacy against COVID-19 (3). Despite this, more than 150 clinical trials testing hydroxychloroquine and chloroquine, alone or in combination with azithromycin, for treatment of COVID-19 are still recruiting (www.clinicaltrials.gov), with many using doses several times higher than those recommended for their approved indications.
Significant concern exists around the cardiac safety profiles of hydroxychloroquine, chloroquine and azithromycin (4-6) as they are all known to prolong the QT interval on the electrocardiogram (ECG), primarily via block of the hERG/Kv11.1 potassium channel, and thus increase the risk of arrhythmias and sudden cardiac death (7). There are hundreds of cases of drug-induced arrhythmias, and cardiac arrest, associated with these drugs in the FDA’s adverse events reporting system (FAERS; www.fda.gov). In the context of COVID-19 patients, this concern was most recently highlighted in a multinational registry analysis that showed increased risk of arrhythmias in patients treated with hydroxychloroquine and chloroquine, either with or without azithromycin (8). However, this study was subsequently retracted due to concerns over the veracity of the data (9). Consequently, there is scant and inconsistent data regarding the cardiac safety of these drugs at doses relevant to COVID-19.
In addition to the intrinsic risk associated with these drugs, many factors associated with the pathophysiology of COVID-19 may increase the risk of proarrhythmic activity. For example, fever, electrolyte changes including hypokalaemia and hypermagnesaemia, and systemic acidosis associated with respiratory failure, are common in COVID-19 patients (8, 10, 11) and can all modify potency of hERG block and by extension proarrhythmic risk (12-15). Indeed, emerging evidence suggests that the QT prolongation observed for these drugs appears more profound in COVID-19 patients than in healthy individuals (5) suggesting a role for patient attributes in modifying proarrhythmic risk. Thus, there is a need for a thorough assessment of the cardiac safety profile of chloroquine, hydroxychloroquine and azithromycin in the context of changes in pathophysiological state associated with COVID-19, to guide clinical decision making and management of these patients.
Here, we have undertaken a multiplatform, preclinical cardiac safety evaluation combining high-throughput screening of hERG potency, in silico prediction of population risk, and analysis of repolarization changes in human induced pluripotent stem cell derived cardiomyocytes (hiPSC-CM) from multiple individuals to assess both acute and long term effects. Our data demonstrate a significant proarrhythmic risk for hydroxychloroquine and chloroquine at doses being proposed to treat COVID-19 and suggest that clinicians should implement long term QT interval monitoring in trials, particularly in patients with electrolyte imbalances (Graphical Abstract). Aside from specific relevance of our results to use of these drugs in treatment of COVID-19, this study also acts as a blueprint for how high-throughput in vitro screening, combined with in silico simulations, can be used to build reference datasets for how proarrhythmic risk of QT prolonging drugs can be modified by metabolic changes associated with disease in order to help guide clinicians in management of patients.