key: cord-340650-mwsa326c
authors: Elangovan, E. J.; Kumar, V. S.; Kathiravan, A.; Mallampalli, R.; Thomas, T.; Subramaniyam, G.
title: Rationale and prognosis of repurposed drugs with risk stratification of patients in oxygen support in COVID-19: A systematic review and meta-analysis
date: 2020-10-06
journal: nan
DOI: 10.1101/2020.10.04.20206516
sha: 
doc_id: 340650
cord_uid: mwsa326c

There has been rapid development of clinical trials conducted on antivirals, immunomodulators, and other therapies against COVID-19. The rising number of trials has led to duplication and a need for curation of available outcomes from treatments that have been followed across the world. The rising number of trials has led to duplication and a need for curation of available outcomes from treatments that have been followed across the world. We have conducted a systematic review and meta analysis that focus on evaluating the clinical outcomes of repurposed drugs against COVID-19 including Tocilizumab, Remdesivir, Dexamethasome, Lopinavir-ritonavir, Favipiravir, Hydroxychloroquine, and Convalescent plasma therapy. Twenty-nine articles were included in this study after thorough literature search and performed subgroup analyses based on disease severity levels. Random effects model was adopted to estimate overall treatment effect and heterogeneity. Subgroup analysis on mortality rate showed significant overall effect in the treatment group of studies having critically ill patients (p<0.01).Overall, our study confirmed that tocilizumab may probably reduce the mortality (<10%) of patients with COVID-19 with faster recovery time and reduce the risk of patients with lung disease in falling into oxygen support (P = 0.02). Patients on remedesivir showed no significant associations of comorbidities with risk of falling into oxygen supports. Hydroxychloroquine was found to be inefficacious in COVID-19 patients (OR 0.64; 95%CI [0.47-0.86]).Dexamethasone had marginal effect on overall mortality rate (OR 1.19; 95%CI [1.05-1.35]), and hence helpful for patients on mechanical ventilation or ECMO. There was also evidence suggesting that combination therapies (serpin + Favipiravir) were helpful in reducing the mortality rate in COVID-19 patients under invasive support.

adopted to estimate overall treatment effect and heterogeneity. Subgroup analysis on mortality rate showed significant overall effect in the treatment group of studies having critically ill patients (p<0.01). Meta-regression analysis was performed to study the association of drug efficacy in patients with different comorbidities and factors that influence the patient's prevalence in non-invasive and invasive ventilation support. Overall, our study confirmed that tocilizumab may probably reduce the mortality (<10%) of patients with COVID-19 with faster recovery time and reduce the risk of patients with lung disease in falling into oxygen support (P=0.02). Patients on Remdesivir showed no significant associations of comorbidities with risk of falling into oxygen supports. Hydroxychloroquine was found to be inefficacious in COVID- 19 35] ). There was also evidence suggesting that combination therapies (Serpin + Favipiravir) were helpful in reducing the mortality rate in COVID-19 patients under invasive support.

The first incidence of novel coronavirus was identified in patients with severe respiratory disease in Wuhan, China. Since then, COVID-19 outbreak has grown to 32.7 million cases resulting in 991,224 deaths as on 27th September, 2020 across the world. The effect of and response to the virus is varied based on the immune systems [1] environment risks [2] , preexisting health conditions [3, 4] , sex differences [5] and so on across different populations in and different countries. The virus has debilitated the global community and will continue to do so until an effective vaccine or antiviral is developed.

The ability of the virus to spread rapidly, causing increased risk of deaths in patients with existing health conditions has been considered to be the most alarming feature of COVID- [6, 7] At present, there are no targeted therapies or vaccines available for COVID-19. Based on previous outbreaks like Severe Acute Respiratory Syndrome Coronavirus (SARS-CoV) in 2003 [8] [9] [10] and the Middle East Respiratory Syndrome Coronavirus (MERS-CoV) in 2012 [11] [12] [13] , medical community has been working swiftly by repurposing the drugs and evaluating their efficacy in the treatment of the novel coronavirus. Current treatments are primarily based on factors such as combination of drugs, severity level of disease and respiratory support. Treatment paradigms follow WHO guidelines or those offered by health authorities in each country. Drugs such as Hydroxycholoroquine (HCQ) which were initially reported to be effective in treatment were later found to have limited positive effect on the infected [14] . Later analyses have shown other immune therapies, repurposed antivirals found to have positive effect on patients in particular phases of treatment [15, 16] .

Lack of specific treatment and drug therapies for COVID-19, has led the scientific and medical communities to run several drug trials in the past seven months. These studies generated a huge collection of data regarding the drug efficacy, adverse effects and its specificity towards certain populations. The aim of this study is to design and implement a data driven meta-analysis of existing literature and available outcomes regarding treatment for novel coronavirus. We have explored the effect of emerging treatments widely followed by present medical guidelines across countries including drugs like Tocilizumab (TCZ), Remdesivir (RM), Favipiravir (FPV), Dexamethasone (DM), LPV-r (LPV-r) and Convalescent Plasma (CP) therapy on patients at different severity levels of COVID- 19 . In addition to that, we have attempted to manifest the correlation of drug's ability to treat patients on invasive and non-invasive oxygen support with comorbid conditions like hypertension, diabetes, and cardiovascular diseases. We believe that our findings may help the medical and scientific community to better understand the association of temporal relationships of drug usage during different stages of disease on patients with preexisting health conditions.

We report a systematic review and meta-analysis, as per the recommendations of PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statements [17] . Our work has not been registered under PROSPERO.

OR Dexamethasone OR Convalescent Plasma OR Hydroxychloroquine OR Arbidol OR Corticosteroids) AND ("COVID-19" OR "SARS-CoV-2" OR "2019 nCoV") AND (Moderate OR Severe OR Critically ill OR Hospitalized OR Oxygen Therapy OR Invasive Mechanical Ventilation). The drug names in the keywords were selected from the World Health Organization (https://www.who.int/), Indian Council of Medical Research (https://www.icmr.gov.in/), and Ministry of Health and Family Welfare (https://www.mohfw.gov.in/) COVID-19 protocol guidelines. Additionally, we screened the references of the included articles to obtain more relevant papers for the study.

The following articles were excluded from the analysis: duplicates, review papers, editorials, letters, comments, other language manuscripts and studies tested on in-vitro cell culture and invivo animals. Inclusion criteria for the study were: i) randomized (RCT) or non-randomized clinical trials (nRCT), prospective or retrospective observational studies (cohort study and case series) ii) research articles, preprints and preliminary reports with comparators (Treatment Vs.

Control) or combination of treatment modalities or studies without control group iii) study population could be any age, sex and any region in the world, diagnosed with COVID-19 with either laboratory test-confirmed or Chest computer tomography (CT) iv) any one of these outcomes reported: mortality rate, recovery rate, viral clearance period, clinical improvement of patients in oxygen therapy or invasive mechanical ventilation (IMV) after drug treatment.

Data extracted from the included study articles were updated on a google spreadsheet. Any ambiguity in data extraction was clarified by discussion and consensus of the authors. Clinical insights were consistently sought. The following features were extracted: author, study type, date of publication, study period, study place, drug name, cohort size, gender, age, severity condition (mild, moderate, severe and critical); treatment combination, time from symptom onset to the treatment, dosage details, exclusion criteria of drug; precondition of patients (PaO2:FiO2, SpO2 levels, respiratory rate), comorbidity, patients requiring respiratory support such as low flow oxygen support, high flow or Non-Invasive mechanical Ventilation (NIV), IMV or extracorporeal membrane oxygenation (ECMO) (during admission and follow-up); clinical improvement length, viral clearance period; mortality rate; recovery rate and adverse effects. All the data were individually extracted for subgroups (treatment group, control or comparator) and overall outcomes of all the treatments were summarized together and plotted using ggplot2 [18] in R.

We used RoB 2.0 [19] and ROBINS-I [20] tools of Cochrane risk of bias assessment for evaluating RCTs and observational or nRCTs. Robvis package [21] from R was used for the visualization of risk bias assessment. The Newcastle-Ottawa scale (NOS) was used to assess single-arm studies [22] . The use of this scale is more sensible to control the quality level of the cohort study [23] .

To evaluate the drug treatment effects between control and test groups, an odds ratio was obtained to measure probability of events occurring between groups at disease severity. The intervention effect distribution was estimated using the Random Effects Model (REM), which is provided as an estimate of 95% confidence interval. I 2 statistic was used to measure heterogeneity within and between studies. This was performed using metafor package V2.4.0 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 6, 2020. . https://doi.org/10.1101/2020.10.04.20206516 doi: medRxiv preprint [24] in R V4.0.2. Overall proportion of all single armed studies were calculated using metaprop [25] program in R. Meta-regression was performed to assess the correlation of drug efficacy in patients at different oxygen therapy stages with comorbid conditions using lm function in R. A pvalue of ≤ 0.05 is considered as statistically significant in all our studies unless stated.

Overall, 15,831 records were identified by our searches. On removing redundant entries, 3095 numbers of papers were retained. After exclusion of review articles, 29 clinical studies (24 published and 5 pre-prints) met our eligibility criteria (PRISMA flow chart Figure 1 ). All these were available online between 11th March, 2020 to 22nd July, 2020. After definitive selection of articles, there were eight therapeutic agents in total found from 29 studies having 14,114 COVID-19 patients involved were compared in our meta-analysis. These eight agents included Table 1 .

By pooling articles based on disease severity levels, studies were subgrouped as 1) mild/moderate 2) severe ( Figure 3 ). Odds ratio of mortality rate for studies including mild/moderate patients found to have higher 95% Confidence Interval (CI), P Value of 1.00with I 2 = 0% indicating no observed heterogeneity. However, articles with patients at high risk/severe disease symptoms showed P Value < 0.001 and heterogeneity measure (I 2 ) of 87.5%. Most of the . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 6, 2020. [28] . The failure of the latter trial on Remedesivir is due to the small sample size with 2:1 randomisation. Both the studies showed an insignificant mortality rate between control and test groups [27, 28] [30, 31] having > 0.2 proportion.

. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 6, 2020. .

We performed meta-regression analysis to understand the association of factors influencing a patient's dependency on life supporting ventilation treatments. A set of variables that may likely influence the COVID-19 patients landing into non-invasive and invasive oxygen supports were identified. The variables included were time from symptom onset to treatment, treatment duration, mortality rate and comorbidities of patients (hypertension, diabetes, heart disease and lung diseases). Of all the drugs included, TCZ and RM had enough number of studies (14) with potentially relevant oxygen support information (supplemental oxygen/ NIV/ IMV) to carry out the analysis. Results showing the multivariate meta-regression for all TCZ studies are shown in There was no significant association identified on comparing the patients prevalence in invasive ventilation supports (ECMO) and its affecting factors. Similar to the non-invasive case, positive correlation was observed between the number of patients with hypertension and increased risk of falling into invasive oxygen support. On the other hand, patients with diabetes when treated on TCZ showed negative correlation with their prevalence in invasive ventilation support.

. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 6, 2020. . Furthermore, patients treated with TCZ under invasive support had no association with early treatment onset, and comorbidities conditions including pulmonary, renal and heart diseases.

Meta-regression on RM revealed no association of the patient population in oxygen therapy with any of the dependent variables except time from symptoms onset. It showed a negative correlation with the patient population under the invasive group.

The overall clinical outcomes in control and test for each treatment were compared and represented as a bar plot ( Figure 6 ). The percentage of death in the TCZ group was found to be lowest (<10%) among all treatments compared in this study. Additionally, patients in the TCZ group found to have faster recovery (clinical improvement duration 7days) than the control group and other treatments. However, there was no much difference in patients' percentage in the invasive support from the time of admission to follow-up. HCQ was the only test group where the death rate relatively scaled up than in the control group. The clinical outcomes of control and test groups from DM studies showed no significant difference (it was specific for significant reduction of death in IMV support). Although a shorter duration of clinical improvement was observed among TCZ, DM and RM treatments. A good rise in negative conversion rate of the viral RNA (during follow-up) was seen in CP and FPV therapy compared to the control groups.

The studies including RM and LPV-r observed to show lower incidence of severe adverse events, the percentage are very much comparable with that of their control group.

The risk of bias score evaluated for 29 included studies comprising 12 RCTs, 1 nRCT, and 16 observational (12 retrospective, 3 prospective and 1 case series) are presented in Figure 2 and . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 6, 2020. . Table S2 . The overall judgments for 15 studies were high risk or poor quality, 10 at moderate or fair quality and 4 at low risk of bias.

Although mortality rate is 1% of the total population of the world [32] COVID-19 has caused 991,224 deaths till now, and no therapeutic agent has received Food and Drug Administration (FDA) approval yet. There are many clinical studies reported on repurposed drugs for COVID-19 treatment. This meta-analysis indicates 29 of such studies involving 14,114 patients with COVID-19 and assess drug efficacy on the key outcomes, including mortality rate, viral clearance, clinical improvement, oxygen support, invasive ventilation, comorbid factors (diabetes, hypertension, heart disease, lung disease and renal disease) and adverse events.

According to the pooled results in subgroup analyses, assessing the severity of the disease, TCZ is found to be more effective in reducing the mortality rate (OR 27.50 [5.39-140.24 ]) of severe COVID-19 patients ( Figure 1 ). TCZ, an approved antagonist of IL-6 receptor in rheumatoid arthritis with moderate to severe condition [33] . This is also considered as a selective cytokine inhibitor in COVID-19 [34] , where multiple clinical studies have evaluated the safety and efficacy of TCZ in the severe stage of the disease.

In Capraa et al.'s study, a greater significance was seen in survival rate (p=0.004) and 92% recovery in TCZ treated patients as compared to the control [26] . But, Campochiaro's study reveals no statistical difference in mortality rate (TCZ 15% Vs. control 33%) and clinical improvement (69% Vs. 61%), due to small cohort size [35] . But still, there is a slight significance seen in mortality rate in our finding with odds of 2.70, showing TCZ is little superior over standard therapy. In fact, Nicola et al. concluded that early use and low dosage of TCZ has . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 6, 2020. . https://doi.org/10.1101/2020.10.04.20206516 doi: medRxiv preprint shown to be effective and reduces mortality rate (94%) without any adverse events, but they are limited with small study groups and no evidence exists for the same, with treatment at later stages of respiratory failure [36] . Another prospective study supports Nicola's results suggesting that a timely administration of TCZ before lung injury improves outcomes of severe COVID-19 [30] . In spite of that, three retrospective studies on TCZ [31, 37, 38] , where 2 studied with cohortcontrol and 1 with single-arm, managed to show consistent results in relation to risk of IMV or death or ICU admission.

The first study has a larger cohort of 544 serious COVID-19 patients, showing a decrease in IMV risk up to 27% in the TCZ group than standard 41.5%. But, this study limits with a residual confounding that there were 4 patients with renal insufficiency or cancer in the TCZ group whereas it was 15 in the standard [37] . Similarly, the second study even with a smaller cohort could observe that TCZ treatment strongly reduces ICU admission or death (25% Vs. 72%, p = 0.002) or IMV support (0% Vs. 32%, P = 0.006) than standard care in the critically ill COVID-19

patients [38] . The third single-arm study with 100 patients also shows clinical improvement in 77% patients having respiratory support or 74% of 43 ICU patients by day-10 [31] . Additionally, two studies shows good improvement in the SpO2 levels (89% to 97%, day 1-10) [39] and PaO2:FiO2 ratio (152±53 to 302.2 ±126, day 0-14) of TCZ treatment [40] .

A previous systematic review on TCZ reported 28 studies conducted in many different places; the results of most of these studies were favorable to TCZ therapy in severe and critically ill COVID-19 patients than standard therapy [41] . Based on this, TCZ is safe in treating both early and late stage respiratory conditions of COVID-19, but so far, no RCT has been reported in treating the severe groups, where the RCT is still on-going to measure the efficacy and safety of this treatment in case of severe COVID-19 [42] .

. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 6, 2020. . CP is another type of immune therapy that shows less or no significance (OR 2.12 [0.70-6.38];

1.70 [0. 63-4.59] ) between the control and the treatment in the subgrouping of 2 RCTs included.

Results from a RCT using CP have shown that mortality rate (p=0.95) and improvement of day-15 disease severity (p=0.58) were not significant when compared to the standard group. The study also shows that COVID-19 patients already have enough neutralizing antibody titers than donors and suggests that the use of CP at early symptom onset [43] . Another RCT also showed similar results that the mortality and clinical improvement were found to be insignificant.

Despite this, they showed a higher significance (87.2% CP Vs. 37.5% STD) in viral load negativity within 72 hrs after CP transfusion in severe COVID-19 patients [44] . Limitation in both studies was that they were soon terminated before attaining enough sample size which prevented definite conclusions of clinical benefits.

In our post-search, a large RCT done on CP management was found. This enrolled almost 464 moderately ill COVID-19 patients, and showed results in line with 2 previous RCTs, that mortality was not significantly different between CP treated and control groups 14.5% Vs. 13.5% [45] . According to these reports, CP therapy has shown to be more efficient in early negative conversion of viral RNA, but no effect in reducing the mortality rate of moderate-severe patients. Two RCTs with placebo-controlled have been reported, one [27] of which has a larger study size with improvised protocol than the other [28] . The first study favors RM over placebo with short recovery time (11 Vs. 15 days) for all clinical outcomes except patients on IMV or ECMO. They further suggest that it is supportive for hospitalized and low supplemental oxygen . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 6, 2020. . support patients [27] . A Second study trialed 2:1 randomization with limited sample shows no significance in reductions of viral load or duration of clinical improvement in severe COVID-19 patients [28] . Both RCTs describe that the mortality rates were insignificant between treatment and placebo. Also recommend that RM could be given as an antiviral agent in combination with other therapy before the condition progresses to need for IMV [27, 28] .

Subsequently, Gein et al. study on compassionate-use of RM, shows 22% death and 70% clinical improvement among patients. Also, 50% (25/53) of severe patients with age group 70 or older were discharged. However, adverse events occurred in 60% of patients (such as elevated hepatic enzymes, rash, hypotension and renal impairment) and serious adverse events in 23% (multipleorgan-dysfunction syndrome, acute kidney injury, septic shock and hypotension). This was caused by RM treatment alone [46, 47] .

Two other studies [48, 49] randomized moderate and severe patients with RM for either 5-day or 10-day courses. The clinical improvement of severe patients (without receiving mechanical ventilation) was not significant (64% Vs. 54%) between the two courses [48] . Anyhow, in moderate patients, the 5-day course had statistically significant clinical status when compared to standard care (OR 1.65 95% CI, 1.09-2.48; p = .02), whereas the 10-day course showed no effect [49] . Majority of these study outcomes suggest that RM can be used for hospitalized and lower respiratory tract infectious COVID-19 patients.

Regarding DM, Horby et al. have published a well-designed largest RCT that enrolled the highest population size (n=6425), when compared to any other studies included in this metaanalysis. The study shows a low dose of DM is most effective in patients on IMV or ECMO than without respiratory support, where the risk of death is reduced significantly in DM group than . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 6, 2020. [50] , which is why the subgroup analysis in our study shows a marginal effect in mortality rate (OR 1.19 [1.05-1.35]) . Therefore, DM could be recommended in treating long term symptom COVID-19 patients, requiring IMV than recent symptom onset. (Figure 1 ). Two RCTs from China studied HCQ with a small number of mild-moderate patients. The first is a preprint and only partially confirmed the drug effect on symptomatic outcome (like fever and cough) [51] . The other study shows no significance in viral clearance between HCQ plus standard Vs. standard alone treatment, in addition more adverse events were found high in HCQ group [52] . Recent research shows HCQ inhibits trained immunity by an epigenetic modulation. This prevents the antiviral effects of the bodily innate immune response against the SARS-CoV-2 infection [53] . Together, these results indicate that HCQ is not a promising rescue option for mild-moderate and severe COVID-19 patients.

LPV-r or Arbidol alone shows little benefit for clinical outcome of mild-moderate COVID-19 patients [54] , but Deng's study showed that combined effect is more favorable [55] . A . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 6, 2020. . https://doi.org/10.1101/2020.10.04.20206516 doi: medRxiv preprint retrospective study commends, LPV-r has a more rapid effect in viral clearance than HCQ [56] .

receiving oxygen support in case of severe COVID-19 patients. Thus no effect was seen in the mortality rate (OR 1.40 [0.71-2.76]) of the severe subgroup (Figure 1 ). Their finding says the positive viral RNA appeared till the end of trial in LPV-r group (40.7%), but it was not confirmed with the presence of viral infection. The study limits without blinding which might have influenced the outcome and premature discontinuation (14%) of the treatment [57] .

Therefore, the LPV-r treatment seems to have antiviral benefits during mild COVID-19 condition.

FPV, an inhibitor of RNA-dependent RNA polymerase of viral RNA, evaluated by in-vitro studies to be active against COVID-19. An observational study conducted in Japan, provided FPV (high dose of 1600 mg twice day 1 followed by 600 mg twice a day for 5 days) to a large population (n=2158) on compassionate basis and observed mortality rate of 1918 COVID-19 patients after a month was higher in severe group (31.7%) An observational study with 2158 COVID-19 patients, studied FPV use on compassionate basis. They show mortality was higher in the severe group (31.7%) when compared to mild (5.1%) and moderate (12.7%) by day-30. Likewise, the rate of recovery was lower in the severe group (14.7%) than the other two conditions (61.7% and 42.7%). Also, deaths were frequent in elderly and 24.65% of patients posed adverse events with FPV therapy. The study unfortunately is limited since it contains no background data of patients though they are registered from 407 hospitals [58] . A study in Thailand favors the use of FPV with 100% of clinical improvement in 27 hospitalized patients (without oxygen support) and 83.3% in 30 serious conditions (requiring supplemental oxygen or IMV) at day 28. However, the study is limited by the use of additional . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 6, 2020. . https://doi.org/10.1101/2020.10.04.20206516 doi: medRxiv preprint agents such as chloroquine and hence could not find the actual impact of FPV [59] . Another study shows, by day-7, FPV had no difference in recovery rate when compared with Arbidol therapy for moderate COVID-19 patients [60] .

Synthetic serine protease inhibitors (serpins) were earlier used to treat diseases like Acute myocardial infarction, Ischemic stroke and Pulmonary embolism [61]. More recently Nafamostat, a Serpin was found to prevent the proteolytic activity of transmembrane protease serine 2 (TMPRSS2) and thus inhibits the viral fusion with the host cells [62, 63] . The combination of two therapies (serpin and FPV), may allow the inhibition of viral entry as well as the replication. A small case series shows that FPV in combination with Nafamostat was found to have benefits on severe COVID-19. In which, 8 of 11 patients were extubated and 9 discharged from ICU. But still there is no convincing evidence supporting this study [64] . Taken together, FPV use is efficacious in mild-moderate COVID-19 patients. But for severe or critically ill conditions, the addition of Serpins with the FPV may be helpful.

In summary i) TCZ is effective in severe COVID-19, but requires a RCT to validate the results.

ii) DM works only on patients with IMV in reducing mortality, however several unresolved questions still exist. iii) RM benefits patients without having IMV support. iv) CP and LPV-r can cure only mild illness and CP was found to be efficacious in bringing down the viral load in at least hours. v) HCQ has no potency or antiviral effects. vi) FPV when used in combination with serpins found to have additive effects on severe COVID-19, but more findings are required to assess the drug tolerability.

One previous meta-analysis was reported already on repurposed drugs by June 9 (pre-print), however the current study has added more therapeutic agents with its recent trials including a . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 6, 2020. . https://doi.org/10.1101/2020.10.04.20206516 doi: medRxiv preprint huge number of patients. Our study is bound by limitations of much high risk quality evidence.

In addition, few non-comparable single-arm observational studies and smaller cohorts are underpowered to assess the clinical outcomes that were addressed.

Based on the available evidence, this analysis has found immunotherapy (TCZ) was superior to antivirals in most of the clinical outcomes. Additionally, a low dose of DM is much helpful in patients on IMV or ECMO with significant reduction of death risk than standard care. Although it has a marginal effect on live-saving without respiratory concerns. HCQ was found to have no clinical effects for COVID-19, whereas it plays a role in inhibiting the trained immune system. LPV-r, not efficacious in severe COVID-19, yet has some antiviral effects in mild conditions. The use of CP rapidly reduces viral load within a few hours in all stages of COVID-19. The results of HCQ and LPV-r were in line with the recently reported 6-arm RECOVERY clinical trial (NCT04381936). But the trial with TCZ and other monoclonal antibody cocktails in the 6arm are still on-going, the results of which may provide an optimal dosage and safety details for an appropriate benefit of the therapy. From meta-regression analysis, a significant correlation of patients exposed to TCZ with hypertension and higher risk of falling in non-invasive oxygen support (P=0.02) was observed. On the other hand, patients on TCZ having lung diseases had lower prevalence in non-invasive support. Antivirals RM and FPV were recommended as a combinatorial therapy with other agents. A 10-day course of RM has speedy recovery in hospitalized patients (without IMV) and a 5-day course is safe in moderate COVID-19. FPV alone treats patients before the end stage of COVID-19 patients. But the combinatorial use of Serpins with FPV may be efficacious for even critical patients. Furthermore, high quality evidence is required to evaluate its usage in severe or critical COVID-19.

. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 6, 2020. 

We thank Abhimanyu Swaroop, IIT-Madras for assistance with screening of articles during the initial phase of study.

This study did not require ethical approvals as there was no direct involvement of patients.

The authors declare that they have no competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

There was no funding provided for this work . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 6, 2020. . https://doi.org/10.1101/2020.10.04.20206516 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 6, 2020. . https://doi.org/10.1101/2020.10.04.20206516 doi: medRxiv preprint Figure 2 . Risk of bias assessment for (a) Randomised controlled trails using Cochrane risk of bias (b) Non-randomized controlled trials using Newcastle-Ottawa scale . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 6, 2020. . Figure 3 . Forest plot comparing mortality of patients between control and test group and subgrouped based on severity level of disease. CI, confidence interval; df, degrees of freedom; I 2 , heterogeneity. Line of no effect is at 1. Each horizontal bar represents upper and lower 95% CI. Squares on the right and left side of the line favor the test and control group respectively. Diamond shows an overall summary of all studies is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 6, 2020. . . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 6, 2020. . is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 6, 2020. . https://doi.org/10.1101/2020.10.04.20206516 doi: medRxiv preprint

Immune response to SARS-CoV-2 and mechanisms of immunopathological changes in COVID-19

Environment and COVID-19: Pollutants, impacts, dissemination, management and recommendations for facing future epidemic threats

Can COVID 2019 induce a specific cardiovascular damage or it exacerbates pre-existing cardiovascular diseases?

Association of Blood Glucose Control and Outcomes in Patients with COVID-19 and Pre-existing Type 2 Diabetes

Sex differences in immune responses that underlie COVID-19 disease outcomes

Clinical Characteristics of Coronavirus Disease 2019 in China

The origin, transmission and clinical therapies on coronavirus disease 2019 (COVID-19) outbreak -an update on the status

A Major Outbreak of Severe Acute Respiratory Syndrome in Hong Kong

Coronavirus as a possible cause of severe acute respiratory . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) syndrome

SARS and MERS: recent insights into emerging coronaviruses

Middle East Respiratory Syndrome Coronavirus Infection in Dromedary Camels in Saudi Arabia

Emerging preclinical evidence does not support broad use of hydroxychloroquine in COVID-19 patients

Tocilizumab for severe COVID-19 related illness -A community academic medical center experience

The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration

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

ROBINS-I: a tool for assessing risk of bias in nonrandomised studies of interventions

Risk-of-bias VISualization (robvis): An R package and Shiny web app for visualizing risk-of-bias assessments

A suggestion for quality assessment in systematic reviews of observational studies in nutritional epidemiology

Metaprop: a Stata command to perform meta-analysis of binomial data

Impact of low dose tocilizumab on mortality rate in patients with COVID-19 related pneumonia

Remdesivir for the Treatment of Covid-19 -Preliminary Report

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

Observational Study of Hydroxychloroquine in Hospitalized Patients with Covid-19

Early outcomes of tocilizumab in adults hospitalized with severe COVID19. An initial report from the Vall dHebron COVID19 prospective cohort study

Tocilizumab for the treatment of severe COVID-19 pneumonia with hyperinflammatory syndrome and acute respiratory failure: A single center study of 100 patients in

Covid-19: death rate is 0.66% and increases with age, study estimates

Tocilizumab: A Review in Rheumatoid Arthritis

The cytokine storm in COVID-19: An overview of the involvement of the chemokine/chemokine-receptor system

Efficacy and safety of tocilizumab in severe COVID-19 patients: a single-centre retrospective cohort study

Early use of low dose tocilizumab in patients with COVID-19: A retrospective cohort study with a complete follow-up

Tocilizumab in patients with severe COVID

Tocilizumab therapy reduced intensive care unit admissions and/or mortality in COVID-19 patients

Tocilizumab for patients with severe COVID-19: a retrospective, multi-center study

Pilot prospective open, single-arm multicentre study on off-label use of tocilizumab in patients with severe COVID-19

Rationale and evidence on the use of tocilizumab in COVID-19: a systematic review

A prospective, randomised, double blind placebo-controlled trial to evaluate the efficacy and safety of tocilizumab in patients with severe COVID-19 pneumonia (TOC-COVID): A structured summary of a study protocol for a randomised controlled trial

Convalescent Plasma for COVID-19. A randomized clinical trial

Effect of Convalescent Plasma Therapy on Time to Clinical Improvement in Patients With Severe and Life-threatening COVID-19

Convalescent plasma in the management of moderate COVID-19 in India: An open-label parallel-arm phase II multicentre randomized controlled trial (PLACID Trial)

Compassionate Use of Remdesivir for Patients with Severe Covid-19

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

Effect of Remdesivir vs Standard Care on Clinical Status at 11 Days in Patients With Moderate COVID-19

Dexamethasone in Hospitalized Patients with Covid-19 -Preliminary Report

Efficacy of hydroxychloroquine in patients with COVID-19: results of a randomized clinical trial

Hydroxychloroquine in patients with mainly mild to moderate coronavirus disease 2019: open label, randomised controlled trial

Hydroxychloroquine inhibits trained immunity-implications for COVID-19

Efficacy and Safety of Lopinavir/Ritonavir or Arbidol in Adult Patients with Mild/Moderate COVID-19: An Exploratory Randomized Controlled Trial

Arbidol combined with LPV/r versus LPV/r alone against

Corona Virus Disease 2019: A retrospective cohort study

Lopinavir-ritonavir versus hydroxychloroquine for viral clearance and clinical improvement in patients with mild to moderate coronavirus disease 2019

A Trial of Lopinavir-Ritonavir in Adults Hospitalized with Severe Covid-19

Preliminary Report of the Favipiravir Observational Study in Japan

Real-world Experience with Favipiravir for Treatment of COVID-19 in Thailand: Results from a Multi-center Observational Study

Favipiravir versus arbidol for COVID-19: a randomized clinical trial

Protease inhibitors and their peptidomimetic derivatives as potential drugs

The anticoagulant nafamostat potently inhibits SARS-CoV-2 infection in vitro: an existing drug with multiple possible therapeutic effects

Nafamostat Mesylate Blocks Activation of SARS-CoV-2: New Treatment Option for COVID-19

Nafamostat mesylate treatment in combination with favipiravir for patients critically ill with Covid-19: a case series

Tocilizumab for treatment of patients with severe COVID-19: A retrospective cohort study

Experimental treatment with favipiravir for COVID-19: an open-label control study