Efficacy and secondary infection risk of tocilizumab, sarilumab and anakinra in COVID‐19 patients: A systematic review and meta‐analysis

Summary As the pandemic progresses, the pathophysiology of coronavirus disease 2019 (COVID‐19) is becoming clearer and the potential for immunotherapy is increasing. However, clinical efficacy and safety of immunosuppressants (including tocilizumab, sarilumab and anakinra) treatment in COVID‐19 patients are not yet known. We searched PubMed, Embase Medline, Web of Science and MedRxiv using specific search terms in studies published from 1 January 2020 to 20 December 2020. In total, 33 studies, including 3073 cases and 6502 controls, were selected for meta‐analysis. We found that immunosuppressant therapy significantly decreased mortality in COVID‐19 patients on overall analysis (odds ratio = 0.71, 95% confidence interval = 0.57–0.89, p = 0.004). We also found that tocilizumab and anakinra significantly decreased mortality in patients without any increased risk of secondary infection. In addition, we found similar results in several subgroups. However, we found that tocilizumab therapy significantly increased the risk of fungal co‐infections in COVID‐19 patients. This represents the only systematic review and meta‐analysis to investigate the efficacy and secondary infection risk of immunosuppressant treatment in COVID‐19 patients. Overall, immunosuppressants significantly decreased mortality but had no effect on increased risk of secondary infections. Our analysis of tocilizumab therapy showed a significantly increased risk of fungal co‐infections in these patients.


| Study selection
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Checklist was used to improve the reporting of our meta-analysis. We searched PubMed, EMBASE, MEDLINE, Web of Science and MedRxiv using the search terms immunosuppressants, anakinra, sarilumab, siltuximab, tocilizumab, bacterial/fungal coinfection, coronavirus, severe acute respiratory syndrome coronavirus 2, SARS-CoV-2, 2019-nCoV and COVID-19 for studies published from 1 January 2020 to 20 December 2020, and we manually searched the references of select articles for additional relevant articles ( Figure 1).

| Data extraction and verification
The inclusion criteria for the meta-analysis were as follows: (1) research focus on immunosuppressants (such as tocilizumab, anakinra, sarilumab, siltuximab, sirukumab, etc.) and COVID-19; (2) the number of cases and controls; (3) randomised controlled trial (RCT) or retrospective study, including case-control study and cohort study; (4)  the citations that met our inclusion criteria and extracted all data. If at least two of the three researchers agreed, the study was included in the meta-analysis. Next, everyone extracted while the other crosschecked the data. Disagreements were resolved by reviewing and discussing.

| Statistical analysis
The statistical significance of the pooled odds ratio (OR) was determined with the Z-test, considering the values of p < 0.05 to be statistically significant. Data were pooled from the meta-analysis with the random-effects model using the DerSimonian and Laird method and the fixed-effects model using the Mantel-Haenszel method. In cases where I 2 < 50% and the p-value for heterogeneity was >0.10, thus indicating an absence of heterogeneity between studies being compared, the fixed-effects model was used to evaluate the summary ORs. Conversely, if I 2 ≥ 50% or the p-value for heterogeneity was ≤0.10, thus indicating a higher degree of heterogeneity between studies but still met our inclusion criteria for meta-analysis, we used the random-effects model to evaluate the summary ORs. To evaluate the influence of individual data sets on overall pooled ORs, we conducted forest plot analysis to determine the stability of our results. We also carried out sensitivity analysis in which a single study within the overall metaanalysis was deleted one at a time. We applied Funnel plots and Egger's linear regression test to assess publication bias. All statistical analyses were carried out using STATA version 11.0 (Stata Corporation College Station).

| Study selection and characteristics
The combined search terms yielded all related articles, and the primary review of titles and abstracts identified 157 articles that   Table 1.

| Mortality
We Specific data are summarised in Figure 2a and Table 2.

| Fungal co-infection risk
We found tocilizumab therapy significantly increased the risk of

| Publication bias and sensitivity analysis
Begg's funnel plot and Egger's test were performed to assess publication bias. We additionally conducted sensitivity analyses by omitting one study at a time in the calculation of a summary outcome.
Although the sample sizes for cases in all eligible studies varied, corresponding pooled proportions and 95% CIs were not qualitatively altered between studies with small and large sample sizes. No other single study influenced pooled proportion and 95% CI qualitatively.

| DISCUSSION
The World Health Organization has declared that COVID-19 may progress to a pandemic associated with substantial morbidity and mortality and is a public health emergency of international concern as of 1 February 2020. 11 In a recent study, anakinra significantly reduced both the need for invasive mechanical ventilation in the ICU and mortality among patients with severe COVID-19, with no serious side-effects. 6,15,16 Kooistra et al. 16 and Cauchois et al. 6 also investigated whether anakinra was effective at reducing clinical signs of hyperinflammation in   To the best of our knowledge, this is the only systematic review and meta-analysis conducted to investigate the efficacy and secondary infection risk of immunosuppressants treatment in COVID-19

F I G U R E 2 Forest plot of the associations between immunosuppressants and mortality and secondary infection risk in COVID-19 patients. Forest plot of association between (a) immunosuppressants and mortality and (b) immunosuppressants and secondary infection risk
patients. In our meta-analysis, we also found that tocilizumab significantly decreased mortality in COVID-19 patients without any increased risk of secondary infection. In addition, we found similar results in several subgroups. However, we found that tocilizumab therapy significantly increased the risk of fungal co-infection in COVID-19 patients. Therefore, our data suggest that clinicians should be aware of antifungal therapy when COVID-19 patients are receiving tocilizumab therapy.
There are several limitations to our current study, which needs to be addressed. Firstly, only 33 studies were included, and the relatively small total sample size had limited power for the exploration of real associations. Secondly, subgroup analyses involved relatively small groups, which may not impart sufficient statistical power to explore real associations and are more likely to reveal greater beneficial effects than large-scale trials. Thirdly, every doctor has a different treatment for clinical diagnostic and treatment algorithms, which would allow for adjustment by other factors. 3,[25][26][27] In addition, the inclusion of zero-event trials can sometimes decrease the effect size estimate and narrow the CIs.

| CONCLUSION
Overall, immunosuppressants significantly decreased mortality in COVID-19 patients without any increased risk of secondary infection. Our analyses of tocilizumab therapy showed that there was a significantly increased risk of fungal co-infections.