Introduction
Since the discovery of the AIDS virus and its in vitro culture, clinical observations have been supplemented or even replaced by monitoring the blood or plasma viral load of patients with chronic infections such as HIV and viral hepatitis. It is commonly accepted that a decrease in the viral load attests to an improvement in the infection or even to a cure (1), making it possible to use it to judge the therapeutic effectiveness of new antiviral drugs. Indeed, this method is widely used as a gold standard marker in randomised clinical trials (2). In addition, viral load monitoring has been applied to monitoring the effectiveness of treatments for acute viral infections such as cytomegalovirus (CMV) (3) and Epstein-Barr virus (EBV) (4). Monitoring viral load through quantitative PCR has been recommended as a way of monitoring therapeutic efficacy (5). For example, the effectiveness of EDP-938, a non-fusion replication inhibitor of respiratory syncytial virus (RSV), was evaluated in a randomised controlled trial involving volunteers who were intra-nasally inoculated with the RSV-A Memphis 37b strain. This trial concluded that all EDP-938 regimens were better than placebo in terms of lowering of the viral load (6). More recently, several randomised trials on the treatment of COVID-19 have used viral load as the primary outcome, demonstrating that ivermectin (7) reduced SARS-CoV-2 viral load in comparison with convalescent plasma (8). Metformin glycinate has been reported to reduce SARS-CoV-2 viral load in a double-blind Phase IIb clinical trial (9). In a very preliminary paper, we reported that 26 patients with COVID-19 treated with HCQ with or without azithromycin had a significant shorter virus shedding period compared to 16 untreated patients with COVID-19 (10). This paper was severely criticised, and we published an additional paper responding to these criticisms, in which we confirmed the reduction of the viral load in patients treated with HCQ (11). Subsequently, in another observational study, we reported that the persistence of viral shedding over ten days was more frequent in patients who were not treated with HCQ-AZ (12). A number of confounding factors may affect the outcome. The viral clearance of SARS-CoV-2 has been reported to depend upon age, given that the duration of shedding is shorter in younger patients (13–15). Other confounding factors include the timescale between the onset of symptoms and admission (16), being immunocompromised, and the initial viral load (17). Armed with this new knowledge, we aim here to re-analyse our database, investigating the role of HQC on the viral shedding of patients with COVID-19.