Efficacy and Safety of Hydroxychloroquine and Chloroquine for Treating COVID-19: A Rapid Review

Background : Since its emergence, COVID-19 has affected more than one million people with over 100,000 deaths globally as of early April 2020. In this review, we evaluated the efficacy and safety of hydroxychloroquine and chloroquine on patients with COVID-19. Methods : PubMed, CENTRAL, Google Scholar, Chinese Clinical Trial Registry and Trial Registers: ISRCTN; UMIN-CTR (Japan's Trial Register); and the WHO portal up were searched for the in vitro and in vivo studies of hydroxychloroquine and chloroquine. Results : Moderate-certainty evidence from a single published RCT to-date showed that after a five-days treatment, patients on hydroxychloroquine may have shorter symptom duration compared to the control group, including fever (MD-1.0 day (95% CI -1.48, -0.52), cough (MD -1.1 day (95% CI -1.63, -0.57) and higher likelihood of having improved chest CT appearance (RR 1.47 (95% CI 1.02, 2.11), no clear differences was observed in the risk of progressing to severe illness and risk of adverse effects. A single non-randomized study showed marked decreased in the number with positive viral load at day 6, but due to several flaws in the study, the certainty of evidence for the outcome was very low. Conclusion : Despite the evidence on possible effects of hydroxychloroquine and chloroquine, given the paucity of available evidence especially in relation to the number on-going RCTs, conclusion on the effectiveness and safety of hydroxychloroquine cannot yet be made.

of the disease (4,7,8). The virus is known to spread primarily through droplets of saliva or nasal discharge from the infected person when they cough or sneeze (9).
Researchers have been debating on the efficacy of hydroxychloroquine (HCQ) in inhibiting the SARS-CoV-2. Initial results from a placebo-controlled trial of HCQ for COVID-19 indicate that patients hospitalized with mild illness recovered more quickly with addition of the drug than with placebo at the start of a standard treatment (10). In contrast, a pilot study in China reported no difference in recovery rate among mild to moderate cases of COVID-19 when treated with HCQ (11). The efficacy of the drug in the treatment or prophylaxis of COVID-19 is still unknown. At present, there is still no treatment or vaccine for this disease (4).
HCQ is known to be a less toxic antimalarial drug compared to its analog chloroquine (CQ) because the addition of a hydroxyl group decreasing its toxicity while conserving its efficacy (12,13). It is also used as a disease-modifying anti-rheumatic drug to treat the acute and chronic rheumatoid arthritis, discoid and systemic lupus erythematosus (SLE), and juvenile idiopathic arthritis (JIA), owing to its immuno-modulatory effects (13,14). In the previous SARS outbreak in 2003, in vitro research suggested that the usage of CQ showed anti-SARS-CoV activity through virus/cell fusion interference and post-entry spread of the virus (15). Hence, the hypothesis during this outbreak suggesting the possibility of HCQ as a potential pharmacological agent for the treatment of COVID-19 infection.
Since the COVID-19 was declared as a global pandemic, scientists and pharmacological companies are desperate to find the effective treatment(s). The ill-informed and non-evident based interpretation of the efficacy of the HCQ/CQ have great potential in causing serious harm to the public. There are no clinical evidence to support the use of HCQ/CQ as a treatment or prophylaxis for SARS-CoV-2 infection. Therefore, this review aims to determine the efficacy and safety of HCQ and CQ given as a single drug or in combination with standard care with or without any other antimicrobial agent.

Methods
We searched the PubMed, CENTRAL, Google Scholar, Chinese clinical trial registry, and trial registers: ISRCTN; UMIN-CTR (Japan's Trial Register); the WHO portal up to 9th April 2020 using the search terms *chloroquine, hydroxychloroquine, coronavirus, SARS-Cov-2, 2019-NCov, and COVID-19. We screened the titles and abstracts, and included in vitro and in vivo studies of CQ/HCQ for the treatment of SARS-CoV-2. The criteria for considering studies to be included in this review is listed in Table 1.  The standard Cochrane methods were employed, as described in the Cochrane Handbook for Systematic Reviews of Interventions (16). Two review authors (NAM and CJ) independently screened for potentially eligible studies by inspecting the titles and abstracts to generate a shortlist. Two review authors (CJ and NSMD) then independently inspected the abstracts and/or full texts of these shortlisted studies further to determine final eligibility, using the predefined inclusion and exclusion criteria. We resolved any disagreement with the help of another review author (TA) as a referral. We included published studies available in full-text article.

Data extraction and management
Two review authors (NSMD and MFB) independently extracted all data from each included study using a dedicated data collection form. We collected study characteristics, including study design, setting, country, participants, interventions, comparators, outcomes and any other information considered relevant according to Cochrane Handbook for Systematic Reviews of Interventions (16). We resolved potential discrepancies through discussion and involved another review author (NML) if necessary.

Assessment of risk of bias in included studies
Two review authors (NAM and NML) independently assessed one included study for risk of bias (ROB) according to the following seven criteria, in accordance with the recommendations in the Cochrane Handbook for Systematic Reviews of Interventions (16).
Random sequence generation.
Blinding of participants and personnel.
Blinding of outcome assessment.
Incomplete outcome data.
Selective outcome reporting.
Other bias.
We made a judgment of low, high, or unclear ROB, with justifications based on the information obtained from the study (10). We assessed the ROB and presented in a Risk of bias' table for one study (10). We used ROBINS-I risk of bias tool to assess the risk of bias for non-randomized comparative study (17). We included the following outcomes in the 'Summary of Findings' Whenever we identified an issue in each of the five GRADE criteria that was considered a serious risk to influence the outcome estimate, we downgraded the quality of evidence by one level, and when we considered the issue to be very serious, we downgraded the quality of evidence by two levels (18).
When we decided to downgrade the quality of evidence from the default high quality, we justified our decision and described the level of downgrading in the footnotes of the

Results
Description of studies     No clinical outcomes reported in this paper (to be reported in a separate paper).

Non-randomized trial
Those who refused HCQ were allocated to the control group: in general, patients who refused treatment tend to be prognostically worse as a group, raising concerns on ROB. The age difference between the two groups does not seem to pose a major prognostic imbalance, and other major characteristics appear comparable between the two groups. However, the authors have only taken into account of the known prognostic factors in reporting these characteristics.
The rest of control group composed of patients from other centers, for whom care regimen, in particular, and co-intervention might be different, and these were not elaborated in the paper. Viral testing, the major outcome reported, were done differently between the two groups, as in the control patients, patients were tested every alternate day's vs. HCQ patients for whom testing was done every day. In control group patients, if the positive viral results was missed on the day of reporting, results were carried forward from the previous day. This is another evidence of performance bias being present.
The intervention group enrolled 26 patients, while the control group enrolled 16. It was unclear the reason for discrepancy on the number of participants between intervention and the control group. It was unlikely to be due to a lack of relevant patients. Was there a statistical consideration to make the result appealing?
Stopping before target participants' number: the trial was stopped when 36 out of 42 target participants were reached. In general, stopping trial early for benefit tends to make the trial results more exaggerated than stopping on time. This raises the following query: since COVID patients were rapidly available consecutively, it would not have taken long to add another 6 patients to complete the trial, why stopped at 36? Could it be possible that patients 37-42 on HCQ were actually available, but adding them would have made the results less drastically favoring HCQ?
Loss of follow-up: 6 patients in HCQ group were lost to follow up, and their data were not included in the report. 4 out of the 6 were positive in their last reading. Including the results of the four from their last readings carried forward, or even taking a plausible proportion of these patients to be assumed as positive on day 6 (a convenient way is to assume 50% positive: i.e. 2 positive cases on day 6, would decrease the difference in the proportion with positive viral load in between groups).

Discussion
Through a comprehensive search strategy, we identified three in vitro studies (20)(21)(22) from the search strategy. Generally, two studies (20,21) report that the HCQ is more potent to SARS-CoV-2 than CQ.
However, the long term usage of HCQ in patients might lead to toxicity. In another study (22), it is reported that CQ and remdesivir are more effective on SARS-CoV-2. HCQ is not included in this study.
From our analysis and certainty of evidence rating, HCQ or possibly CQ may have given moderate effects on SARS-COV-2.
Besides the in vitro studies, we identified two in vivo studies that matched our selection criteria in terms of population, intervention, comparison and outcomes. We believe the study gathered in this review represented the best available evidence to answer the question that we posed in conducting this review. One in vivo study was a randomized-controlled trial which compared the use of HCQ and standard therapy in 62 patients (10). Another in vivo study was non-randomized comparative trial in which HCQ and azithromycin was compared with standard care among 36 patients (17). Overall, the use of HCQ showed reductions in the symptoms of COVID-19 and laboratory findings, measured primarily as improvement of symptoms such as fever and cough. However, both studies showed low quality of evidence. The sample size was relatively small as compared to the number of COVID-19 patients in both countries.
Despite our broad search strategies, we might have missed the relevant articles that examined HCQ due to the widely variable description of COVID-19 and corresponding interventions. There are 43 ongoing trials that are yet to be included in our analyses, and with the large number of studies and participants in most outcomes, the inclusion of these studies might change the overall findings.
There is a concern that HCQ and CQ might have adverse effects on COVID-19 patients (24). The author questioned the in vivo study by Gautret

Conclusion
There is some evidence from one RCT and one non-randomized comparative study that HCQ appears to modestly reduce SARS-CoV-2 activity, but there is so far no clear evidence that HCQ affects other outcomes as what have been mentioned by Juurlink (2020) (24) with no data on adverse effects and animal-model outcomes. However, the certainty of the estimates for all outcomes (or the quality of evidence) were very low-to-moderate, which means there is a clear possibility that the overall findings may change with further research.
In view of the very low-to-moderate certainty of evidence presented in this review, more well-

Consent for publication
Not applicable.

Availability of data and materials
Not applicable.

Competing interests
The authors declare that they have no competing interests.

Funding
This publication is funded by the National Institutes of Health, Malaysia (NMRR Research ID: 54669). Figure 1 The PRISMA flow chart for in vivo and in vitro study of hydroxychloroquine and chloroquine in COVID-19.

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download.