Limitations
In the absence of detailed pharmacokinetic and pharmacodynamic studies in COVID-19 patients, plasma concentrations corresponding to the doses of hydroxychloroquine and chloroquine that are currently undergoing evaluation in clinical trials were best estimates based on existing literature in other patient groups. Similarly, the concentration of azithromycin used in our study, while corresponding to the reported plasma Cmax, does not take into account the possibility of tissue accumulation in patients, potentially underestimating the degree of hERG block and hence proarrhythmic risk. Since the initial completion of this work, other chemical entities have shown antiviral activity against SARS-CoV2 in vitro , including 90 positive hits from the Prestwick chemical library (50). This study is therefore not an exhaustive analysis of small molecules that could be trialled for treatment of COVID-19. However, it represents, to our knowledge, the most comprehensive in vitro dataset for how hERG-related proarrhythmia can be modified by disease associated factors and as such provides a new blueprint for high throughput in vitro analysis of disease associated proarrhythmic risk.