Hydroxychloroquine plus standard of care compared with standard of care alone in COVID-19: a meta-analysis of randomized controlled trials

The efficacy and safety of Hydroxychloroquine (HCQ) in treating coronavirus disease (COVID-19) is disputed. This systematic review and meta-analysis aimed to examine the efficacy and safety of HCQ in addition to standard of care (SOC) in COVID-19. PubMed, the Cochrane Library, Embase, Web of sciences, and medRxiv were searched up to March 15, 2021. Clinical studies registry databases were also searched for identifying potential clinical trials. The references list of the key studies was reviewed to identify additional relevant resources. The quality of the included studies was evaluated using the Cochrane Collaboration tool and Jadad checklist. Meta-analysis was performed using RevMan software (version 5.3). Eleven randomized controlled trials with a total number of 8161 patients were identified as eligible for meta-analysis. No significant differences were observed between the two treatment groups in terms of negative rate of polymerase chain reaction (PCR) (Risk ratio [RR]: 0.99, 95% confidence interval (CI) 0.90, 1.08; P = 0.76), PCR negative conversion time (Mean difference [MD]: − 1.06, 95% CI − 3.10, 0.97; P = 0.30), all-cause mortality (RR: 1.09, 95% CI 1.00, 1.20; P = 0.06), body temperature recovery time (MD: − 0.64, 95% CI − 1.37, 0.10; P = 0.09), length of hospital stay (MD: − 0.17, 95% CI − 0.80, 0.46; P = 0.59), use of mechanical ventilation (RR: 1.12, 95% CI 0.95, 1.32; P = 0.19), and disease progression (RR = 0.82, 95% CI 0.37, 1.85; P = 0.64). However, there was a significant difference between two groups regarding adverse events (RR: 1.81, 95% CI 1.36, 2.42; P < 0.05). The findings suggest that the addition of HCQ to SOC has no benefit in the treatment of hospitalized patients with COVID-19. Additionally, it is associated with more adverse events.

The COVID-19 pandemic was identified and reported for the first time in Wuhan, China 1-3 , and has been recognized as a global public health concern by the World Health Organization (WHO) 4 . The mortality of critically ill patients with COVID-19 is considerable 5 . The initial estimations for the case fatality rate were about 2.3% in China 6 and 7.2% in Italy 7 . The 2019-nCoV infection causes clusters of severe respiratory illnesses similar to severe acute respiratory syndrome (SARS) coronavirus and, in severe cases, is associated with hospitalization, ICU admission, and frequent mortalities 8,9 . Fever, coughing, shortness of breath, muscle ache, confusion, headache, sore throat, rhinorrhea, chest pain, diarrhea, nausea, and vomiting are among the clinical manifestations of the disease 10 . Early efforts have focused on describing the clinical course, containing severe cases, and treating the sick 11 .
HCQ is used for a variety of diseases, including IgAN 13,14 , Arthritis 15,16 , Pulmonary Sarcoidosis 17 , Cutaneous lupus erythematosus [18][19][20] , Sjogren's syndrome 21 , and Type 2 diabetes mellitus 22 . HCQ prevents the activity of lysosomal enzymes. This drug can reduce the production of cytokines, including type 1 interferons, and inhibit www.nature.com/scientificreports/ T cell activation and the differentiation and expression of excitatory molecules (CD154) 23 . Currently, several observational 24,25 and clinical studies 26,27 have evaluated the efficacy of HCQ on COVID-19. Besides, several meta-analyses examined HCQ in COVID-19 regarding a few of the outcomes, such as mortality and negative rate of polymerase chain reaction (PCR). Therefore, there is an urgent need to examine more detailed outcomes based on the available evidence. The purpose of this study was to examine whether HCQ in addition to standard of care (SOC) versus SOC alone is more effective and safer in hospitalized patients with COVID-19.

Methods
The study protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) with the registration number CRD42020179425. When writing this report, we used the preferred reporting items for systematic reviews and meta-analysis (PRISMA) 28 . The PRISMA statement consists of a 27-item checklist and a 4-phase flow diagram. We used the PRISMA checklist to prepare the report and diagram for the screening process. Study selection. Two researchers independently screened the identified studies based on the inclusion criteria. Disagreements were resolved by discussion among the authors. The following inclusion criteria were used for selecting the articles: (1) hospitalized patients with suspected or confirmed COVID-19 by laboratory tests;

Search strategy.
(2) HCQ plus SOC as intervention; (3) SOC alone as control; (4) randomized controlled trial as study design; (5) published as abstract or full-text; and (6) primary and secondary outcomes of interests (negative conversion time, negative rate of PCR, mortality rate, body temperature recovery time, length of hospital stay, and any adverse events). Other studies and reports such as letters to the editor, case reports, editorial comments, observational studies, and animal models were excluded.
Data extraction and quality assessment. The Cochrane risk of bias tool was used to assess the risk of bias in five domains (selection, performance, attrition, reporting, and other) in the included studies. We used the Jadad checklist to evaluate the quality of clinical trial studies. Information on the study characteristics (place, design, and duration); patient's characteristics (age, sex, and the number of patients); intervention and control (treatment protocol); efficacy outcomes; and adverse events were extracted. Both processes were independently performed by two authors, and disagreements were resolved by discussion among the authors.
Quantitative data synthesis. A meta-analysis was conducted to compare the efficacy and safety of HCQ plus SOC versus SOC alone using the RevMan (version 5.3) software which is recommended by the Cochrane Handbook for Systematic Reviews of Interventions. The mean difference (MD) and risk ratio (RR) with a 95% confidence interval (CI) were used for continuous and dichotomous variables, respectively. Statistical heterogeneity was evaluated using I-square > 50% and Chi-square with a significance level p < 0.1. The random-effects method was used for statistical heterogeneity; otherwise, the fixed-effect method was used.

Results
Study characteristics. Figure 1 shows the process of literature search, removal of duplication, and screening. Of the total 1,205 records, 1,098 were excluded based on the title and abstract. The remaining 107 records were review for full text. Finally, eleven randomized controlled trials (RCTs) 26,[29][30][31][32][33][34][35][36][37][38] with 8161 patients were included in the meta-analysis. The main characteristics of the selected studies are presented in Table 1. Assessment of the risk of bias using the Cochrane Collaboration tool is presented in Fig. 2. Outcomes. Primary outcomes. The meta-analysis demonstrated that there was no significant difference between the HCQ plus SOC group and SOC group in terms of the negative rate of PCR (RR: 0.99, 95% CI 0.90, 1.08; P = 0.76), PCR negative conversion time (MD: − 1.06, 95% CI − 3.10, 0.97; P = 0.30), and all-cause mortality (RR: 1.13, 95% CI 0.99 1.27, P = 0.06) (Fig. 3).

Sensitivity analysis and subgroup analysis.
The findings of the subgroup analysis are presented in Table 2. The findings showed that primary outcomes did not change regarding dose, sample size, treatment duration, and severity of COVID-19. We performed a sensitivity analysis based on different settings and control groups. For this reason, we included RCTs done on non-hospitalized patients and RCTs with placebo as control.
The result did not change in terms of the negative rate of PCR and mortality rate ( Table 2).

Discussion
The purpose of this study was to examine the efficacy and safety of HCQ plus SOC compared to SOC alone for COVID-19. Several RCTs have examined the efficacy and safety of HCQ in treating COVID-19. The findings of our metaanalysis showed that the addition of HCQ to SOC in patients with COVID-19 was not associated with significant improvement in any outcomes reported in patients. Previously published meta-analyses 39-42 on observational studies and RCTs also found no clinical benefits for HCQ in comparison with SOC for COVID-19 patients, which confirms our findings. Gautret et al. 43 suggested that the causes of insufficient response to treatment with HCQ in the non-respondents with COVID-19 should be examined by factors such as SARS-CoV-2 strains, genome, and other associated factors with the metabolism of HCQ in patients. A possible mechanism for HCQ inefficiency was explained by Sandeep and McGregor 44 using virtualized quantum mechanical modeling. However, Yao et al. 45 found that HCQ was more potent than Chloroquine in inhabiting SARS-CoV-2 in vitro.
It should be noted that most of the included studies in these meta-analyses were observational. There are some concerns regarding the limitations of these studies that should be considered. All kinds of biases such as confounding, reverse causation, statistical considerations, and other issues are limitations of these studies in the estimation of drug efficacy and safety 46 . The Agency for Healthcare Research and Quality (AHRQ) has provided recommendations on including observational studies in the comparative effectiveness review process for  www.nature.com/scientificreports/ comparing medical interventions 30 . Our meta-analysis showed that HCQ was not associated with a significant reduction or increase in the COVID-19 mortality rate. The results from similar meta-analyses 40,47 are in line with our findings. Moreover, we conducted a sensitivity analysis of studies done in nonhospital patients and placebo as a control, plus included RCTs regarding the negative rate of PCR and mortality rate. The result did not show any significant differences. Fiolet et al. 47 also found that HCQ was not associated with a different mortality rate in all studies and RCTs, which confirms our findings. Some studies 43,48 have supported the synergistic effect of HCQ with Azithromycin on COVID- 19. In an open-label non-randomized clinical trial, Gautret et al. 43 found that 100% of patients who received HCQ and Azithromycin as combination therapy successfully recovered from COVID-19. These authors found similar results in another study. A meta-analysis showed that HCQ alone or in combination with Azithromycin in comparison with the control group was not effective in treating COVID-19 and was associated with higher mortality rates. It seems that there is no benefit for HCQ in combination with Azithromycin, and it is associated with an increased mortality rate 47 . Contradictorily, a study on 1,438 patients hospitalized with Covid-19 in all United States Veterans Health Administration medical centers found that the rates of death in patients under treatment with HCQ alone were higher than and HCQ plus Azithromycin (27.8%, 22.1%) 25 . However, a meta-analysis of the adverse effects of long-term Azithromycin use in patients with chronic lung diseases showed that Azithromycin is associated with an increased risk of bacterial resistance (2.7-fold) 49 .
Our meta-analysis showed that as an add-on therapy to SOC was associated with more adverse events. Reported adverse events in RCTs were mainly mild. Three RCTs reported serious adverse events in patients taking HCQ and SOC. However, this difference was not statistically significant. Generally, adverse effects of antimalarial are usually rare and mild 50 . Gastric symptoms are a prevalent adverse effect of HCQ 51, 52 ; other adverse events include Cutaneous 53 , headache 54-56 , Cardiomyopathy 57-59 , and Retinopathy [60][61][62][63][64] . Regarding the short-term follow-up of the studies, it is recommended that patients who received HCQ should be monitored for possible adverse events over a longer period of time.

Limitations.
High heterogeneity between studies, the use of different treatment protocols as SOC, short follow-up periods, and lack of rigorous methodologies of the studies were among the limitations of our study. Nevertheless, the findings of this study can be beneficial for guiding clinicians in decisions regarding COVID-19 treatment.

Conclusion
According to the findings of this study, HCQ in addition to SOC appears to be not an effective treatment in primary outcomes, including negative conversion time, the negative rate of PCR, and mortality rate. In addition, HCQ did not show a significant improvement in other secondary outcomes. HCQ was also associated with higher rates of adverse events.  www.nature.com/scientificreports/