key: cord-0878624-whbr5mp2
authors: Al-Abdouh, Ahmad; Bizanti, Anas; Barbarawi, Mahmoud; Jabri, Ahmad; Kumar, Ashish; Fashanu, Oluwaseun E.; Khan, Safi U.; Zhao, Di; Antar, Annukka A.R.; Michos, Erin D.
title: Remdesivir for the treatment of COVID-19: A systematic review and meta-analysis of randomized controlled trials
date: 2021-01-07
journal: Contemp Clin Trials
DOI: 10.1016/j.cct.2021.106272
sha: 07ab6710aae78029323ac397d881b0ce7f80e60a
doc_id: 878624
cord_uid: whbr5mp2

Background The nucleotide analogue prodrug remdesivir was among the first antiviral therapies to be tested in randomized controlled trials (RCTs) for COVID-19. We performed a meta-analysis to understand efficacy and safety. Methods We searched PubMed, EMBASE, Cochrane library, and ClinicalTrials.gov databases (from January 1, 2020 to November 5, 2020). We included RCTs comparing the efficacy and safety of remdesivir to control/placebo in COVID-19. Two independent investigators abstracted data, assessed the quality of evidence, and rated the certainty of evidence. Results A total of 4 RCTs with 7334 patients with COVID-19 were included. At a follow-up of 28–29 days from randomization, very low certainty evidence showed that use of remdesivir compared with control group (placebo and/or standard of care) was not associated with a significant decrease in time to clinical improvement (standardized mean difference −0.80 day; [CI, −2.12, 0.53]). However, moderate certainty of evidence showed that remdesivir was associated with higher rates of recovered patients (risk difference [RD] 0.07 [0.05, 0.08]) and discharged patients (RD 0.07 [0.03, 0.11]) and lower rates of developing serious adverse events (RD -0.05 [−0.10, −0.01]) compared with control. Moderate and very low certainty of evidence showed there was no significant difference in deaths at 28–29 days follow-up (RD -0.01 [−0.03, 0.01]) and developing any adverse events (RD 0.01 [−0.17, 0.19]) between both groups, respectively. Conclusion Patients given remdesivir are more likely to demonstrate recovery and were associated with higher rates of hospital discharge, but not with significant reduction in mean time to clinical improvement or mortality.

The devastating impact of the coronavirus disease 2019 (COVID-19) pandemic including social, economic, and health effects, caught the world unprepared. 1 As of November 2020, there have been over 1,300,000 deaths worldwide from COVID-19. While the scientific community mobilizes to identify vaccine and therapeutic candidates, public health interventions such as J o u r n a l P r e -p r o o f Journal Pre-proof contact tracing, isolation, quarantine, face mask use, and social distancing have been widely adopted to preventing transmission.

Remdesivir is an adenosine nucleotide analogue prodrug that abrogates viral replication by inducing chain termination after its incorporation into the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) RNA-dependent RNA polymerase (RdRp) enzyme. 2 At the start of the pandemic, remdesivir was one of very few nucleoside/nucleotide analogue therapeutics with human safety data that had been shown to overcome the proofreading exoribonuclease activity of the coronavirus RdRp to effectively inhibit a broad range of coronaviruses. 3, 4 With encouraging SARS-CoV-2-specific in vitro efficacy data, remdesivir became one of the earliest direct-acting antiviral therapeutics to enter randomized clinical trials (RCTs) for COVID-19. 2, 5 Its efficacy in preventing respiratory disease with early treatment was demonstrated shortly thereafter in a primate model of There are currently two RCTs that have assessed the efficacy of remdesivir vs placebo and/or standard of care in individuals with severe COVID-19, one RCT in individuals with moderate COVID-19, and one RCT that included both moderate and severe COVID-19. 5, [7] [8] [9] Patients were considered to have severe disease if their peripheral capillary oxygen saturation was equal or less than 94% on room air with evidence of pneumonia and considered to have moderate disease if their saturation was more than 94% with evidence of pneumonia.

Due to the need for greater certainty in the evidence base for use of remdesivir compared to standard treatment based on disease severity and to strengthen evidence of risks and benefits, we conducted a meta-analysis to evaluate the effect of remdesivir in the treatment of moderate and severe COVID-19.

This meta-analysis was conducted following the Cochrane Collaboration guidelines and reported according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analysis). 10, 11 The protocol was registered at PROSPERO: CRD42020181509.

Data Sources and Searches. The literature search was performed without language restriction using the following electronic databases of PubMed, EMBASE, Cochrane library, and ClinicalTrials.gov from January 1, 2020 to November 5, 2020. The search strategy included broad search terms: -Remdesivir‖, -COVID-19‖, -SARS-CoV-2‖, and -Coronavirus‖ (Appendix Table 1 ).

The prespecified inclusion criteria were (1) RCTs comparing remdesivir treatment of any duration to placebo/control (2) RCTs reporting any of the following; time to clinical recovery, number of recovered patients, mortality, or adverse events. There were no limitations on the language, sample size or follow-up duration of RCTs. We excluded observational studies from this meta-analysis.

After removing the duplicates and following the selection criteria, we screened the remaining articles at the title and abstract level and then at the full text level (Figure 1 ). The process of study search and selection was performed independently by 2 investigators (A.A. and A.B.) Any conflicts were resolved by a third investigator (A.J.). Two unblinded investigators (A.A. and A.B.) independently assessed the potential risks of bias of the RCTs using the Cochrane Risk of Bias Tool at the study level as well as at outcome level (Appendix Figure 1) . 12, 13 The main outcomes of interest were mean time to clinical improvement, and the rate of recoveries, discharges from hospital, deaths, serious adverse events, and any adverse events through 28-29 days of follow up. Patients were considered to be recovered in our study if they were discharged alive from the hospital or were admitted without oxygen requirements (for Table 2 and 3). 14 We calculated the mean and standard deviation for the time to clinical improvement from the median and interquartile ranges that were provided in the selected trials as described by Wan et al 15 . Clinical improvement was defined as improvement of 2 or more points on the severity ordinal scale. In Spinner et al., 9 the 5-day and 10-day remdesivir arms were pooled together in one arm. Estimates in this meta-analysis were pooled using inverse variance random-effects model due to heterogeneity in the included population and treatment duration. 16 The Paule-Mandel method was used for estimation of 2 . We applied Hartung-Knapp/Sidik-Jonkman small-sample adjustments considering the number of studies. 17 We reported effect sizes as standardized mean difference (SMD) for the time to clinical improvement outcome and risk difference (RD) for other outcomes with 95% confidence interval (CI). The RD was reported as it gives better representation of absolute risk and doesn't overestimate the size of the effects. 18 Odds ratio was not reported in our analysis as we are evaluating data from clinical trials and using odds ratio in this kind of studies could be misleading. 19 We used I 2 statistics to measure the extent of unexplained statistical heterogeneity: I 2 greater than 50% was considered a high degree of between-study statistical heterogeneity. 20 We did not examine the publication bias as we were underpowered to detect publication bias due to small number of studies. The 95% CIs that did J o u r n a l P r e -p r o o f Journal Pre-proof not cross 0 were considered statistically significant. We used R studio for all analyses in this study.

We performed a sensitivity analysis limited to patients with moderate COVID-19 for time to clinical improvement and rates of recoveries and deaths (data were pooled from Spinner et al. 9 , and patients with moderate severity in Beigel et al. 7 and Pan et al. 21 ). We also performed another sensitivity analysis by excluding one trial at a time and

repeating the analysis (leave-one-out analysis) for all outcomes. Given the small number of studies, meta-regression analysis was not done.

The study received no funding.

We reviewed 564 articles for eligibility, 4 RCTs encompassing 7,334 patients were selected (Figure 1 ). There was some variation among RCTs with regards to designs and characteristics of participants (Table 1) . A summary of baseline characteristics and comorbidities of the included patients are available in appendix (Appendix Table 4 and showed no mortality benefit of using remdesivir. Table 4 ). Severity ordinal scale in this meta-analysis was designed to be compatible with the scales used in all included trials and has 6 categories: (1) Patients not requiring admission to the hospital (2) Admitted to hospital but not requiring supplemental oxygen (3) The GRADE certainty of evidence 14 ranged from very low to moderate as we found some issues applying indirectness (combining 5-day and 10-day treatments of remdesivir in Spinner et al. 9 ), inconsistency, and imprecision (wide confidence interval) for the studied outcomes in our analysis (Appendix Table 2 

Sensitivity analysis limited to patients with moderate COVID-19 showed the following: 

Leave-one-out analysis was done by excluding one trial a time and repeating the analysis.

For all studied outcomes, excluding any of the included trials makes the results statistically nonsignificant except for the serious adverse events where excluding Spinner et al. 9 keeps the results significant due to lower number of adverse events reported in this trial (Appendix Table 5 ).

Our meta-analysis showed that COVID-19 patients receiving remdesivir had significantly higher rates of recovery and hospital discharge with lower rates of developing serious adverse events compared to patients receiving standard of care/placebo. However, there were no significant differences in the remaining outcomes (standardized mean difference in time to clinical improvement, and rates of deaths or developing any adverse events). We did sensitivity analysis limited to patients with moderate COVID-19 from Spinner et al. 9 and a group of patients with moderate COVID-19 who were included in Beigel et al. 7 and Pan et al. 21 which showed that J o u r n a l P r e -p r o o f this category of patients are more likely to demonstrate recovery if they were given remdesivir without significant effect on mean time to clinical improvement. However, this analysis was with lower power and certainty of evidence than the main analysis due to lower number of included patients. Mortality was the only outcome that was reported in all included studies, and none of the studies showed significant decrease of mortality but were not adequately powered for mortality outcome.

In our analysis, we pooled the 5-day and 10-day treatment of remdesivir together, so our study cannot recommend certain duration of treatment. However, a head to head comparison between 5-day and 10-day remdesivir treatment was done in Goldman et al. 22 In general, remdesivir has a fair safety profile. Our analysis showed lower serious adverse events among patients in the remdesivir arm and no difference in any adverse events rates. Wang et al. 5 and Spinner et al. 9 trials showed higher rates while Beigel et al. 7 showed lower rates of any adverse events among patients in the remdesivir group. All the trials that reported adverse events showed lower rates of serious adverse events among patients in the remdesivir arm. Hepatotoxicity is usually the main concern and was identified in Phase 1 trials J o u r n a l P r e -p r o o f (unknown mechanism) and were found to be related to the dose and duration of remdesivir treatment. It will be difficult to evaluate hepatotoxicity in COVID-19 as the disease itself can cause elevation in liver enzymes and no evidence of higher rates of hepatotoxicity was found among the included trials in our analysis. Nephrotoxicity was noticed in nonclinical studies but the included trials in our analysis didn't show evidence of renal toxicity in remdesivir group. The safety profile of remdesivir in pediatric and pregnant patients is still unclear, and further trials evaluating these population will help in understanding its safety. 24 The United States Food and Drug Administration (FDA) initially approved the use of remdesivir for the treatment of COVID-19 in light of the results of Beigel et al., 7 Spinner et al., 9 and Goldman et al. 22 

Our current study has several important strengths. Our meta-analysis is the first to evaluate multiple efficacy and safety outcomes including the time to clinical improvement which is the primary outcome in most of the included trials by pooling the results of all available RCTs.

We used the Paule-Mandel method in our analysis, and we applied Hartung-Knapp/Sidik-Jonkman small-sample adjustments due to the small number of trials included in the current analysis which is consistent with general recommendations. 28 However, our findings should be considered in the context of several limitations. First, the heterogeneity of some of the outcomes we included were high which could be due to methodological heterogeneity of the included trials, the heterogeneity of the population, and the duration of remdesivir use. Second, one of the included trials was an open label trial and didn't use placebo, and one was terminated prematurely. Third, the mean time to clinical improvement was not provided in the included trials and was calculated from the median and ranges as recommended which assumes that the included population has normal distribution. 15 Fourth, this is a study level meta-analysis and a patient level meta-analysis in future would provide better evidence by better exploring the possible confounders of the results.

Further trials are evaluating remdesivir in inhaled form (NCT04539262 and NCT04480333), in combination with tocilizumab (NCT04409262), in combination with baricitinib (ACTT-2), in combination with intranasal interferon (ACTT-3) and in different populations such as pediatric patients (NCT04431453) and adult outpatients (NCT04501952).

All these trials will help in evaluating the efficacy and safety of remdesivir in COVID-19

patients. The overall management of COVID-19 has improved as experience with the condition has been gained. The differences in the timing of the studies and implementation of newer J o u r n a l P r e -p r o o f management methodologies can influence the differences seen within the analyzed studies, the data obtained, and the ability to translate these findings to future clinical uses.

In conclusion, remdesivir use in patients with moderate or severe COVID-19 was associated with significant increase in rates of recovery and hospital discharge and lower rates of serious adverse events. However, there was no significant difference in mean time to clinical improvement and mortality. These results suggest the need of more RCTs to evaluate the role of remdesivir in COVID-19 patients.

The authors did not receive any specific funding for this work. Dr. Michos is partially supported by the (unrestricted) Blumenthal Scholars Preventive Cardiology Fund at Johns Hopkins.

None of the authors report any disclosures.

Effects of COVID-19 pandemic in daily life

Remdesivir is a direct-acting antiviral that inhibits RNA-dependent RNA polymerase from severe acute respiratory syndrome coronavirus 2 with high potency

Controlled Trial of Ebola Virus Disease Therapeutics

Broad-spectrum antiviral GS-5734 inhibits both epidemic and zoonotic coronaviruses

Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial

Clinical benefit of remdesivir in rhesus macaques infected with SARS-CoV-2. bioRxiv Prepr Serv Biol

Remdesivir for the Treatment of Covid-19 -Final Report

Repurposed Antiviral Drugs for Covid-19 -Interim WHO Solidarity Trial Results

Effect of Remdesivir vs Standard Care on

Clinical Status at 11 Days in Patients With Moderate COVID-19: A Randomized Clinical Trial

Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement

Updated method guidelines for systematic reviews in the cochrane collaboration back review group. Spine (Phila Pa 1976)

RoB 2: a revised tool for assessing risk of bias in randomised trials

The Cochrane Collaboration's tool for assessing risk of bias in randomised trials

GRADE guidelines: 1. Introduction-GRADE evidence profiles and summary of findings tables

Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range

Statistical Primer: heterogeneity, random-or fixed-effects model analyses?

Methods to estimate the between-study variance and its uncertainty in meta-analysis

Relative risk versus absolute risk: one cannot be interpreted without the other

When can odds ratios mislead? Odds ratios should be used only in case-control studies and logistic regression analyses

Predicting the extent of heterogeneity in meta-analysis, using empirical data from the Cochrane Database of Systematic Reviews

Repurposed antiviral drugs for COVID-19 -interim WHO SOLIDARITY trial results

Remdesivir for 5 or 10 Days in Patients with Severe Covid-19

Infectious Diseases Society of America Guidelines on the Treatment and Management of Patients with COVID-19

Remdesivir (Veklury) Approval for the Treatment of COVID-19

Combined Cross-Discipline Team Leader, Division Director, and ODE Director Summary Review

Remdesivir (Veklury) Approval for the Treatment of COVID-19-The Evidence for Safety and Efficacy

Remdesivir for Severe COVID-19 versus a Cohort Receiving Standard of Care. Clin Infect Dis an Off Publ Infect Dis Soc Am

Impacts of remdesivir on dynamics and efficacy stratified by the severity of COVID-19: a simulated two-arm controlled study. medRxiv

Improved tests for a random effects meta-regression with a single covariate

Authors declare that this manuscript is original, has not been published before and is not currently being considered for publication elsewhere.We confirm that the manuscript has been read and approved by all named authors and that there are no other persons who satisfied the criteria for authorship but are not listed. We further confirm that the order of authors listed in the manuscript has been approved by all of us.We understand that the Corresponding Author is the sole contact for the Editorial process. He is responsible for communicating with the other authors about progress, submissions of revisions and final approval of proofs