An open-label randomized controlled trial evaluating the efficacy of chloroquine/hydroxychloroquine in severe COVID-19 patients

Despite several studies designed to evaluate the efficacy of chloroquine and hydroxychloroquine in the treatment of coronavirus disease 2019 (COVID-19), there is still doubt about the effects of these drugs, especially in patients with severe forms of the disease. This randomized, open-label, controlled, phase III trial assessed the efficacy of chloroquine or hydroxychloroquine for five days in combination with standard care compared to standard care alone in patients hospitalized with severe COVID-19. Chloroquine 450 mg BID on day 1 and 450 mg once daily from days 2 to 5 or hydroxychloroquine 400 mg BID on day 1 and 400 mg once daily from days 2 to 5 were administered in the intervention group. Patients were enrolled from April 16 to August 06, 2020, in 6 hospitals in southern Brazil. The primary outcome was the clinical status measured on day 14 after randomization with a 9-point ordinal scale. The main secondary outcomes were all-cause mortality; invasive mechanical ventilation use; the incidence of acute renal dysfunction in 28 days; and the clinical status of patients on days 5, 7, 10 and 28. All patients with a positive RT-PCR result for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) were analyzed (modified intention to treat (mITT) population). Arrythmias and cardiovascular complications were assessed as safety outcomes. A total of 105 patients were enrolled and followed for 28 days. The trial was stopped before reaching the planned sample size due to harmful effects. Patients in the intervention group had a worse clinical outcome on the 14th day (odds ratio (OR) 2.45 [1.17 to 4.93], p = 0.016) and on the 28th day (OR 2.47 [1.15 to 5.30], p = 0.020). Moreover, the intervention group had higher incidences of invasive mechanical ventilation use (risk ratio (RR) 2.15 [1.05 to 4.40], p = 0.030) and severe renal dysfunction (KDIGO stage 3) (RR 2.24 [1.01 to 4.99], p = 0.042) until the 28th day of follow-up. No significant arrythmia was noted. In patients with severe COVID-19, the use of chloroquine/hydroxychloroquine added to standard treatment resulted in a significant worsening of clinical status, an increased risk of renal dysfunction and an increased need for invasive mechanical ventilation. Trial Registration: ClinicalTrials.gov, NCT04420247. Registered 09 June 2020—Retrospectively registered, https://www.clinicaltrials.gov/ct2/show/study/NCT04420247.


Additional Statistical Analysis
Additional statistical analysis was performed to describe and compare the outcomes among chloroquine (Clq); hydroxychloroquine (HClq); and control groups, since the intervention group was composed by patients receiving chloroquine or hydroxychloroquine.
In the primary outcome analysis, we assessed the effect of the Clq group and the HClq group in relation to the control group, in addition to the Clq group comparing to the HClq group, on the 9-point ordinal scale on day 14, as an OR and confidence interval (CI) derived from an ordinal logistic regression, assuming proportional ORs adjusted for age and severity at the baseline (under invasive mechanical ventilation or not at randomization). The same regression model was used to analyze the secondaries outcomes on same scale on days 5, 7, 10 and 28 after randomization.
The effect of the Clq group and the HClq group in relation to the control group, in addition to the Clq group comparing to the HClq group on the mortality, the incidences of invasive mechanical ventilation, acute renal dysfunction by KDIGO stage 3 and coagulopathy in any moment until the 28th day after randomization are reported as proportions and differences between groups as risk ratios (RRs) with CIs, calculated by using the Wald likelihood test.
The effect of the Clq group and the HClq group in relation to the control group, in addition to the Clq group comparing to the HClq group in the number of MV-free days, the length of stay in the ICU and the length of stay in the hospital for 28 days was compared by median differences calculated with a quantile regression based on an asymmetric Laplace distribution.
A p value less than 0.05 was considered statistically significant in all analyses.
There was no missing value for the primary outcome, so imputation was not necessary.
The analyses were performed through the Stata software, version 17. HFNC: high-flow nasal cannula; NIPPV: noninvasive positive-pressure ventilation; ECMO: extracorporeal membrane oxygenation; CRP: C-reactive protein. SD, standard deviation; IQR, interquartile range; Clq, chloroquine; HClq, hydroxychloroquine. a No participants had any of the following comorbidities: chronic renal failure; peripheral vascular insufficiency; or heart, liver, rheumatic, or hematological disease. b Considering asthma or chronic obstructive pulmonary disease. c Considering cancer or human immunodeficiency virus infection. d Only hospitalized patients were eligible for the trial; therefore, patients who had scores of 0, 1, 2 or 8 on a ninepoint ordinal scale were not eligible. e One missing data in the HClq group. f Two missing data in the Clq group, one missing data in the HClq group and one at the control group. g Sex missing data in the Clq group, 21 missing data in the HClq group and 25 in the control group. Data are n (%). One patient in the control group did not have ordinal scale status ascertained in 28 days for missing the follow-up; however, the clinical status was ascertained in 5, 7, 10 and 14 days. a Odds Ratio (OR) and confidence interval (CI) and p value derived from an ordinal logistic regression, assuming proportional ORs adjusted for age and severity at the baseline (under MV or not at randomization) for the modified intention to treat (mITT) population. OR >1·00 represents a clinical worsening assessed on the ordinal scale in the Clq/HClq group compared with the control group. Neutrophil/ Lymphocyte ratio D5 a 6.6 (5.1 to 8.5) 5.1 (3.3 to 9.3) 1.6 (-0.6 to 3.8) 0.389 D7 b 8.5 (5.6 to 11.1) 7.6 (5.1 to 10.8) 0.8 (-2.5 to 4.1) 0.774 D10 c 9.5 (6.5 to 14.2) 7.6 (5.2 to 11) 1.9 (-2.9 to 6.7) 0.352 Data are median (interquartile range). SOFA: Sequential Organ Failure Assessment; CRP: C-reactive protein. a On D5, 38 patients in the Clq/HClq group and 33 in the control group who were still hospitalized were considered for this mITT analysis. b On D7, 31 patients in the Clq/HClq group and 26 in the control group who were still hospitalized were considered for this mITT analysis. c On D10, 22 patients in the Clq/HClq group and 15 in the control group who were still hospitalized were considered for this mITT analysis. d On D14, 18 patients in the Clq/HClq group and 12 in the control group who were still hospitalized were considered for this mITT analysis. e On D28, 5 patients in the Clq/HClq group and 4 in the control group who were still hospitalized were considered for this mITT analysis. f Median difference with corresponding 95% CI calculated as an asymmetric Laplace distribution. g Wilcoxon rank-sum test significance. 7 (29.2) 4 (13.8) 6 (11.8) 3 2 (8.3) 0 (0) 1 (2) 4 0 (0) 0 (0) 1 (2) 5 1 (4.2) 2 (6.9) 3 (5.9) 6 0 (0) 0 (0) 0 (0) 7 1 (4.2) 0 (0) 0 (0) 8 7 (29.2) 9 (31.0) 10 (19.6) Data are n (%). One patient in the control group did not have ordinal scale status ascertained in 28 days for missing the followup; however, the clinical status was ascertained in 5, 7, 10 and 14 days. The scores on the scale were defined as follows: (0) nonhospitalized and no clinical or virological evidence of infection; (1) nonhospitalized and no limitation on activities; (2) nonhospitalized, but with limitation on activities; (3) hospitalized, but not requiring supplemental oxygen; (4) hospitalized and on oxygen via mask or nasal prongs; (5) hospitalized, on noninvasive ventilation or high-flow oxygen or pressure support ventilation in weaning mode; (6) hospitalized, intubated and on MV; (7) hospitalized on MV and additional organ support (renal replacement therapy, vasoactive drugs or extracorporeal membrane oxygenation), and (8) dead. a Odds Ratio (OR) and confidence interval (CI) and p value derived from an ordinal logistic regression, assuming proportional ORs adjusted for age and severity at the baseline (under MV or not at randomization) for the modified intention to treat (mITT) population. OR >1·00 represents a clinical worsening assessed on the ordinal scale in the Clq group compared with the control group; HClq group compared with the control group; and Clq group compared with the HClq group.