In this meta-analysis of multicentre RCTs, we found that IL-6 inhibitors administration in patients with COVID-19 pneumonia is associated with a significant reduction in longest follow-up mortality. Furthermore, it is associated with reduction in 28/30-day mortality, need for intubation, and clinical worsening. The beneficial effect on the need for intubation and clinical worsening was confirmed also when analysing studies with a low risk of bias. Results on mortality are largely driven by the RECOVERY trial. .
Relationship with previous studies
Several meta-analyses and systematic review on this topic have been recently published [18, 50,51,52,53,54,55,56,57,58,59,60]. However, most of these meta-analyses did not include the most recently published trials such as the RECOVERY trial .
The largest meta-analysis published so far by the World Health Organization (WHO) Rapid Evidence Appraisal for COVID-19 Therapies (REACT) Working Group included a total 27 randomized studies, 18 of which were not published at time of the meta-analysis publication. They found that IL-6 inhibition was associated with improved 28-day survival and reduction in the composite endpoint of need for invasive mechanical ventilation, extracorporeal membrane oxygenation, or death. Our study also found a reduction in 28/30-day mortality, together with a reduction in need for invasive mechanical ventilation and in a composite endpoint of clinical worsening. Compared with the WHO REACT meta-analysis, our study focused on a different primary endpoint (longest follow-up mortality) and on studies with published results (even if not peer-reviewed). Furthermore, due to different search strategy, inclusion criteria and last update, we included four trials not included in the WHO REACT meta-analysis [35, 39, 42, 44]. There are also some methodological difference in the statistical analysis. For example, we used a Mantel–Haenszel weighted risk ratio versus an inverse variance-weighted odds ratio. However, analysis of our data using a similar approach did not lead to change in magnitude and direction of results. We also analysed risk of bias using somewhat more restrictive criteria for unpublished, open-label and single-centre trials, and therefore our risk of bias analysis results in a lower proportion of low risk of bias studies [22,23,24,25]. Finally, we also performed a TSA. Overall, despite these methodological differences, we believe that our work is complementary to the WHO REACT study and that it confirms the potential beneficial effect of IL-6 inhibitors in patients with COVID-19 also found by the WHO REACT colleagues. In particular, our study suggest that IL-6 inhibitors benefit may extend beyond 28/30 days, and our TSA showed that the required information size was reached, suggesting firm evidence on beneficial effect of IL-6 inhibitors.
Significance of study findings and what this study adds to our knowledge
Our meta-analysis provides some evidence that tocilizumab administration may be beneficial in patients with COVID-19 pneumonia, by reducing the risk of death and the risk of intubation without increasing risk of secondary infections and adverse events.
Most of the studies enrolled patients about 10 days after symptoms onset, with a moderate–severe disease. Only one RCT was specifically performed in ICU patients , and several studies excluded patients requiring mechanical ventilation at baseline. Therefore, we are unable to comment on efficacy and safety of IL-6 inhibitors in critically ill patients and particularly on those receiving mechanical ventilation, which may have the highest risk of secondary infections [61,62,63].
Nevertheless, the REMAP-CAP study (the only study entirely performed in an ICU setting) is one of the two studies showing a significant improvement in patients survival without higher rate of severe infections . However, less than 30% of patients enrolled in the REMAP-CAP were receiving invasive mechanical ventilation at baseline.
Another interesting finding of our systematic review is that approximately half of included studies did not screen patients on the basis of baseline elevated inflammatory markers [37, 38, 40, 41, 48], including the REMAP-CAP trial . However, the RECOVERY trial enrolled patients with elevated baseline C-reactive protein, somehow supporting the rationale of administering tocilizumab to patients with documented inflammation . Moreover, in accordance with the results of the RECOVERY trial and WHO REACT meta-analysis , we also observed that the concomitant use of steroids and IL-6 inhibitors was associated with a significant reduction in clinical worsening.
In addition, we also explored the risk of development of secondary infections as well as serious adverse events and found no differences between patients receiving IL-6 inhibitors and controls, suggesting safety of this therapy.
Notably, only five trials were double-blinded [36, 38, 40, 45, 47] and only three were judged to carry a low risk of bias [36, 38, 40]. All others RCTs were open-label. Therefore, we believe that the overall treatment effect of IL-6 inhibitors may be overestimated, as it has been showed that blinded trials generally have a 40% higher NNT when compared to unblinded trials . Nevertheless, we do acknowledge that the ongoing pandemic and the pressure on most healthcare systems prompted the need to test therapies in a short timeframe, thereby not allowing in most cases to organize double-blind placebo-controlled trials.
Notably, the reduction in need for invasive mechanical ventilation as well as in clinical worsening were confirmed also when analysing double-blind, low risk of bias studies.
Collectively, results of our meta-analysis suggest that administration of IL-6 inhibitors may be beneficial in patients in a relatively early stage of the disease not undergoing invasive mechanical ventilation and treated together with systemic steroids. Although further high-quality evidence is required, use of IL-6 inhibitors may be justified in a context of pandemic and high pressure on healthcare systems and intensive care units [64, 65].
Strengths and limitations of the study
Our meta-analysis includes only multicentre RCTs, thereby with highest internal and external validity, and therefore carry the highest level of evidence [23, 24, 66, 67]. Furthermore, we explored both safety and efficacy outcomes, and investigated possible subgroup interactions including concomitant use of steroids.
However, our results are limited by the overall high risk of bias of available studies and heterogeneity among inclusion criteria and concomitant treatments. We also included trials that have not yet undergone peer-review. Nevertheless, this has been a common practice in the ongoing COVID-19 pandemic .
We were unable to obtain additional data from investigators, and therefore we could not perform detailed analysis on interaction between corticosteroids and IL-6 inhibitors.
Statistical significance is lost when removing the RECOVERY trial  from analysis. However, beneficial effect of IL-6 inhibitors on clinically relevant secondary outcomes is confirmed in several sensitivity analyses, including low risk of bias trials.
The COVID-19 pandemic is still ongoing and several treatments are under investigation. We cannot comment on possible interaction between other promising treatment strategies such as anti-virals or monoclonal antibodies and IL-6 inhibitors.
Future studies and prospects
Our study highlighted current gaps of knowledge in IL-6 inhibitors use among patients with COVID-19. In particular, there is still insufficient evidence on efficacy and safety of IL-6 inhibitors among patients with critical disease requiring invasive mechanical ventilation. Furthermore, despite several mRCTs have been published, the quality of evidence remains low since several published trials lacked blinding .
Future studies should be designed as double-blinded and placebo-controlled in order to provide the greatest level of scientific rigour. Our study can provide baseline information to calculate sample size for such a future trial.
Furthermore, future studies should assess the possible interactions between available treatments for COVID-19, especially the combination of both systemic steroids and tocilizumab in specific subgroup of patients with a more pronounced inflammatory phenotype.