Statistics from Altmetric.com
We have read the recent report by Konig et al1 with great interest where they described the clinical course of COVID-19 in 80 patients with systemic lupus erythematosus. The frequency of hospitalisation with COVID-19 did not differ between individuals using a synthetic antimalarial drug (AMD) and non-users. The objective of this report was to examine the association of long-term synthetic AMD exposure with hospitalisation with COVID-19 in a larger population in France.
We did a nationwide, retrospective, matched ‘exposed/unexposed’ cohort study using information from the French national health data system (online supplementary methods).2 All individuals who had six reimbursements of synthetic AMD (hydroxychloroquine (HCQ) or chloroquine) within the 12-month period preceding study entry (15 February 2020) were included. For each exposed individual, we randomly selected, from the entire French population, up to three unexposed individuals matched on age, sex, department of residence and complementary universal health insurance. The underlying diseases related to synthetic AMD exposure were recorded, as well as comorbidities and consumption of oral steroids (online supplementary tables 1 and 2). Our study was based on data available on hospital discharges as of 7 June 2020 (online supplementary methods). On this date, 83 445 patients with hospitalisation with COVID-19 have been reported, and 14 566 died in hospitals. The primary end point was the time from study entry to hospitalisation with COVID-19. The secondary end point was the time from study entry to the composite outcome intubation or death. Adjusted HR (aHR) and its 95% CI was given by a cause-specific conditional Cox proportional hazards regression model. Subgroup analyses were conducted as well as a sensitivity analysis using the Fine-Gray competing risks model, computing sub-HR (online supplementary methods). All calculations were performed using SAS V.9.4 software.
A total of 54 873 patients were long-term users of synthetic AMD (mean±SD age: 55.4±16; women: 84.3%). Exposed patients could be matched (1:3) with 155 689 unexposed individuals (50 420 matched to three unexposed individuals, 1483 with two, 1463 with one and 1507 without matching). The distribution of patients’ baseline characteristics according to synthetic AMD exposure is detailed in table 1. Most underlying diseases associated with the long-term use of synthetic AMD were lupus (n=23 120, 42.1%), inflammatory rheumatism (n=9724, 17.7%), sicca syndrome (n=3404, 6.2%), sarcoidosis (n=1073, 2%), sclerodermia (n=545, 1%) and dermatomyositis (n=389, 0.7%). Less than one-third of exposed patients had no identified underlying diseases (n=15 855, 28.9%). The primary end point of hospitalisation with COVID-19 occurred in 323 patients (128 patients in the synthetic AMD group and 195 in the unexposed group). A total of 47 patients were intubated or died (20 vs 27). In the multivariable cause-specific conditional Cox proportional hazards regression model, no difference was observed between synthetic AMD long-term exposure and unexposed group regarding the risk of hospitalisation with COVID-19 (HR, 1.15; 95% CI 0.86 to 1.55, p=0.35) (online supplementary table 3). No more difference was observed in subgroup analyses (online supplementary table 4). No difference was observed between synthetic AMD long-term exposure and unexposed group regarding the risk of intubation or death during a hospital stay for COVID-19 infection (aHR, 0.66; 95% CI 0.29 to 1.51, p=0.33).
Recently, one randomised controlled trial tested HCQ as postexposure prophylaxis.3 The incidence of COVID-19 infections did not differ between participants receiving HCQ (n=414) and those who did (n=407). However, we may question about the HCQ effective blood level achieved by the participants as more than half of randomised patients received HCQ at least 3 days after the first exposure to someone with confirmed COVID-19. In our study, individuals included in the exposed group were already exposed to the drugs at the time of the epidemic. Recently, Ferreira et al4 suggested that long-term use of HCQ confers protection against COVID-19 infection. In their study, patients who underwent PCR test for COVID-19 were suspected cases. Screening for COVID-19 infection is likely to be more frequent in patients with a chronic disease, all things otherwise being equal. Thus, we cannot exclude that the lower number of HCQ patients in the positive PCR group is not related to a larger number of patients screened in the chronic disease sample. Long-term use of synthetic AMD is associated with several illnesses, especially dysimmune conditions.5 Thus, it is important to differentiate the impact of either chronic diseases or treatment on outcomes. Even if we cannot formally exclude residual confusion, the higher risk of being hospitalised with COVID-19 did not persist when adjusting on corticosteroids use. This may be related to the immunosuppressive function of corticosteroids6 and/or to corticosteroids being a proxy of an uncontrolled chronic disease at risk of complication.7
This study does not provide evidence supporting prophylactic use of HCQ or chloroquine in the population, including the at-risk population.
Contributors ES, JB, AW, RD-S and MZ conceived and designed the experiments. LP, PH, BB, LS and JD performed the experiments. ES, LP, PH, BB, JB, AW, RD-S and MZ analysed the data. ES, LP, JB, AW, RD-S and MZ interpreted the results. ES wrote the first draft of the manuscript. All authors contributed to the writing of the manuscript. All authors agreed with the results and conclusions of the manuscript. All authors had full access to all of the data (including statistical reports and tables) in the study and can take responsibility for the integrity of the data and the accuracy of the data analysis. ES is the guarantor.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Patient consent for publication Not required.
Provenance and peer review Not commissioned; internally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.