1 INTRODUCTION

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causative agent of coronavirus disease 2019 (COVID-19), has rapidly spread throughout the world leading to a pandemic.1-3 Up until now, some antiviral drugs have been proposed as promising therapeutic agents against SARS-CoV-2 infection including interferon,4 lopinavir/ritonavir,5 chloroquine,6 remdesivir,7 and arbidol.8

Arbidol (umifenovir) is an oral antiviral drug9 that was approved for prophylaxis in Russia and China several decades ago and used in the treatment of influenza A and B as well as other respiratory viral infections.10 In addition to Arbidol's antiviral and anti-inflammatory activities against various types of influenza viruses,11, 12 especially H1N1,13 its broad-spectrum antiviral activities against other viruses, such as Zika,14 Ebola,15 hepatitis B and C,16, 17 rhinovirus,18 respiratory syncytial virus,18, 19 coxsackie,18, 20 chikungunya,21 and adenovirus18 are shown in vitro and in vivo.

Regarding the SARS-CoV-2 infection, the antivirus effect of arbidol against SARS-CoV-2 has yet been controversial. On the one hand, the efficacy of arbidol was shown in vitro22, 23 which seems to have inhibited the infection more efficiently among other WHO-approved anti-influenza drugs including baloxavir, laninamivir, oseltamivir, peramivir, zanamivir23 by blocking the trimerization of the spike glycoprotein.22 Also, some studies suggested its beneficial effects either in monotherapy or combination therapy with other agents against COVID-19.5, 24-26 On the other hand, there exist other studies which have found no benefit of using arbidol in COVID-19 patients 27, 28 suggesting an urgent need to reach a conclusive decision on this matter. The present systematic review and meta-analysis aim to provide the latest evidence on arbidol's efficacy and safety compared with other therapeutic agents in COVID-19 treatment.

2 METHODS

We have registered the protocol of this systematic review and meta-analysis with the registry number CRD42020207821 and used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist.29

2.1 Literature search strategy

We conducted a systematic search in the leading bibliographic databases, including PubMed, the Cochrane Library, and Embase for the relevant records up to May 2021. We also searched in medRxiv, Google Scholar, and clinical registry databases, including ClinicalTrials.gov, the European Union Clinical Trials Register, and the Chinese Clinical Trial Registry for additional relevant documents. Finally, the reference lists of the included studies and review articles were screened and the search was limited to the articles the abstract or full text of which were in English. Search terms included 2019-nCoV, SARS-CoV-2, COVID-19, arbidol, and umifenovir. The following terms were used to explore PubMed: ((((((((Coronavirus[Title/Abstract]) OR (Coronavirus[MeSH Terms])) OR (COVID-19[Title/Abstract])) OR (SARS-CoV-2[Title/Abstract])) OR (COVID-19[MeSH Terms])) OR (SARS-CoV-2[MeSH Terms])) OR (2019 novel coronavirus infection[Title/Abstract])) OR (2019-nCoV infection[Title/Abstract])) AND ((Umifenovir[Title/Abstract]) OR (Arbidol[Title/Abstract])).

2.2 Study selection

Two authors (Behnam Amani and Mahsa Zareei) independently screened the identified records based on inclusion and exclusion criteria. Disagreements between the authors were resolved by discussion among authors. The studies were included based on the following criteria: (1) patients with laboratory-confirmed positive COVID-19 test; (2) arbidol as monotherapy or in combination with other therapeutic agents; (3) any therapeutic intervention as a comparison (4); efficacy and safety outcomes of interest. The primary efficacy outcomes were the negative rate of PCR (polymerase chain reaction) and PCR negative conversion time and the secondary efficacy outcomes included the rate of improvement on chest CT, rate of cough alleviation, length of hospital stay, and disease progression. The safety outcome was the incidence of adverse events reported in patients; and (5) clinical trials or observational studies. The exclusion criteria were the studies conducted on animal models, case reports, case series, letters to editors, and editorials.

2.3 Data extraction and quality assessment

We used the Cochrane collaboration tool to assess the risk of bias of randomized clinical trials.30 Quality assessment of observational studies was conducted using the Newcastle–Ottawa scale (NOS).31 We extracted data using the same data extraction form. The extracted data included (1) study characteristics (author, year, setting, and design); (2) patient's characteristics (sample size, sex, and age); (3) intervention and comparison (sample size); and (4) efficacy and safety outcomes. All steps were performed independently by two authors (Behnam Amani and Mahsa Zareei).

2.4 Evidence synthesis

We performed a meta-analysis using RevMan software, version 5.3. The mean difference (MD) with a 95% confidence interval (CI) was used for continuous variables and a risk ratio (RR) with 95% CI for dichotomous variables. The statistical heterogeneity was evaluated using the I2 and Chi2 tests. The random-effects model was used for studies with I2 > 50% or p < .1. Otherwise, we used the fixed-effect model.

3 RESULTS 3.1 The characteristics of studies

Figure 1 shows the literature search flow, removal of duplicates, and the screening based on title, abstract, and full text. As a result, 52 full-text articles were reviewed and sixteen studies24, 32-46 entered the final analysis. The characteristics of the studies included in the systematic review are presented in Table 1. Assessment of the risk of bias using the Cochrane collaboration tool is presented in Figure 2.

Study flow diagram

Risk of bias in the selected studies

Table 1. Characteristics of included studies Study, year Country Design Age (mean) N (M/F) Intervention (n) Comparison (n) NOS Chang Chen 2020,47 China RCT NA 236 (110/126) Arbidol (120) Favipiravir (116) RoB 2 Huang 2020,24 China R NA 27 (12/15) Arbidol (11) LPV/r (6), CQ (10) 7 Lisi Deng 2020,33 China R 44.5 33 (17/16) Arbidol + LPV/r (16) LPV/r (17) 8 Ping Xu 2020,42 China R 51.9 141 (74/67) Arbidol + IFN (71) IFN (70) 6 Qibin Liu 2020,25 China R 59.5 504 (259/245) Arbidol (257) Os (66), LPV/r (259) 6 Qiong Zhou 2020,44 China R NA 77 (31/46) Arbidol (24) Arbidol + IFN (46), IFN (7) 7 Wenyu Chen 2020,48 China RCT NA 62 (34/28) Arbidol + control (42) Control (20) RoB 2 Kaijin Xu 2020,26 China R NA 111 (47/64) Arbidol + ER (49) ER (62) 7 Xiu Lan 2020,35 China R 55.8 73 (37/36) Arbidol + LPV/r (39) LPV/r (34) 7 Jun Chen 2020,34 China R 48 134 (69/65) Arbidol (34) LPV/r (52), non-antiviral (48) 5 Xudan Chen 2020,32 China R 48 284 (131/153) Arbidol (37) Control (121), LPV/r (60), arbidol + LPV/r (16), CQ (17), Os (13), Other (16) 9 Yaya Zhou 2020,45 China R 55.5 238 (102/136) Arbidol (82) Arbidol + IFN (139) 7 Yueping Li 2020,37 China RCT 49.4 86 (40/46) Arbidol (35) LPV/r (34), control (17) RoB 2 Zhu 2020,46 China R 39.8 50 (26/24) Arbidol (16) LPV/r (34) 7 Wen 2020,41 China R 49.9 178 (81/97) Arbidol (36) LPV/r (59), control (58), arbidol + LPV/r (25) 7 Ming Li 2021,36 China R NA 62 (24/38) Arbidol (42) CQ (20) 7 Jie 2021,49 China R 65 252 (106/146) Arbidol (228) No arbidol (24) 8 Ruan 2021,50 China R 64 331 (160/171) Arbidol (273) Non-antiviral (58) 8 Ghaderkhani 2021,51 Iran RCT NA 53 (32/21) HCQ + arbidol (28) HCQ (25) RoB 2 Nojomi 2020,40 Iran RCT 56.4 100 (60/40) Arbidol (50) LPV/r (50) RoB 2 Lian 2020,38 China R 60 81 (45/36) Arbidol (45) Control (36) 8 Liu 2021,39 China R 54.8 108 (47/61) Arbidol (40) Arbidol + LHQW (68) 8 Jing Chen 2020,52 China R NA 200 (130/70) Arbidol + SFJDC (100) Arbidol (100) 8 Fang 2020,53 China R 61.5 162 (87/75) Arbidol + LHQW (113) LHQW (49) 8 Ping 2020,43 China R NA 295 (171/124) Arbidol (148) LHQW + arbidol (147) 8 Xiang-Kun 2020,54 China R NA 70 (41/29) Arbidol (30) SFJD + arbidol (40) 9 Abbreviations: CQ, chloroquine; ER, empirical regimens; F, female; HCQ, hydroxychloroquine; IFN, Interferon; LPV/r, lopinavir/ritonavir; LHQW: Lianhuaqingwen; M, male; N, number; NA, not acquired; Os, oseltamivir; R, retrospective; RCT, randomized clinical trial; RoB, risk of bias; SFJD, Shufeng Jiedu. 3.2 Comparisons 3.2.1 Arbidol versus non-antiviral treatment

The result of meta-analysis showed that there was no significant difference between arbidol and non-antiviral groups in terms of negative rate of PCR on Day 7 (RR: 0.94; 95% CI: 0.78–1.14; p = .55) and Day 14 (RR: 1.10; 95% CI: 0.96–1.25; p = .17), PCR negative conversion time (MD: 0.74; 95% CI: −0.87 to 2.34; p = .37) (Figure 3), rate of improvement on chest CT on Day 7 (RR: 1.53; 95% CI: 0.50–4.68; p = .46) and Day 14 (RR: 0.92; 95% CI: 0.56–1.54; p = .76), rate of cough alleviation on Day 7 (RR: 1.47; 95% CI: 0.64–3.39; p = .36) and Day 14 (RR: 1.19; 95% CI: 0.74–1.91; p = .47), hospital stay (MD: 3.97; 95% CI: 0.05–7.89; p = .05), and disease progression (RR: 1.88; 95% CI: 0.70–5.00; p = .21; Figure 4). Arbidol was associated with higher adverse events (RR: 2.24; 95% CI: 1.06–4.73; p = .04; Figure 4).

Forest plot of arbidol versus non-antiviral for outcomes of negative rate of PCR on Day 7 (A), negative rate of PCR on Day 14 (B), and PCR negative conversion time (C)

Forest plot of arbidol versus non-antiviral for outcomes of rate of improvement on chest CT on Day 7 (A), rate of improvement on chest CT on Day 14 (B), rate of cough alleviation on Day 7 (C), rate of cough alleviation on Day 14 (D), hospital stay (E), disease progression (H), and any adverse event (I)

3.2.2 Arbidol versus favipiravir

Only one study47 compared arbidol with favipiravir. The result showed no significant difference between arbidol and favipiravir groups in the clinical recovery rate. However, favipiravir was associated with better efficacy in relieving pyrexia and cough. The frequencies of drug-related adverse events for arbidol and favipiravir were 23.33% and 31.9%, respectively.

3.2.3 Arbidol versus chloroquine

There was no significant difference between arbidol and chloroquine in terms of negative rate of PCR on Day 14 (RR: 1.27; 95% CI: 0.64–2.51; p = .50) and PCR negative conversion time (MD: 0.69; 95% CI: −3.72 to 5.10; p = .76; Table 2). However, the length of hospital stay in patients taking chloroquine was significantly shorter than patients taking arbidol (MD: 4.59; 95% CI: 0.58–8.60; p = .02; Table 2).

Table 2. The pooled estimate of arbidol versus other therapeutic agents and sensitivity analysis Analysis No. of studies Participants Pooled estimate (95% CI) p Heterogeneity Ch2 p I2 Sensitivity analysis Arbidol versus non-antiviral Negative rate of PCR 4 405 1.21 (1.06–1.38) .005 5.79 .12 48% Arbidol versus chloroquine Negative rate of PCR on Day 14 3 137 1.27 (0.64–2.51) .50 12.54 .002 84% PCR negative conversion time 2 75 0.69 (−3.72 to 5.10) .76 14.71 .0001 93% Hospital stay 2 75 4.59 (0.58–8.60) .02 8.44 .004 88% Arbidol versus LPV/r Negative rate of PCR on Day 7 4 276 1.35 (1.03–1.76) .03 4.18 .24 28% Negative rate of PCR on Day 14 5 328 1.47 (1.06–2.04) .02 24.07 <.0001 83% PCR negative conversion time 5 328 −2.28 (−3.83 to −0.72) .004 21.91 .0002 82% Hospital stay 3 214 −1.87 (−8.01 to 4.27) .55 50.39 <.00001 96% Rate of improvement on chest CT on Day 7 2 156 1.14 (0.77–1.69) .50 0.29 0.59 0% Rate of improvement on chest CT on Day 14 2 156 0.99 (0.80–1.23) .92 0.24 0.62 0% Disease progress 2 164 1.08 (0.13–9.29) .94 5.64 0.02 82% Rate of cough alleviation on Day 7 2 141 1.61 (0.21–12.22) .64 5.48 0.02 82% Rate of cough alleviation on Day 14 2 141 0.81 (0.58–1.15) .24 0.32 0.57 0% Adverse events 5 367 0.44 (0.28–0.68) .0002 2.70 0.61 0% Arbidol + LPV/r versus LPV/r Negative rate of PCR on Day 7 2 117 2.06 (1.13–3.76) .02 0.01 0.91 0%