Interleukin-1 and interleukin-6 inhibition in patients with COVID-19 and hyperinflammation - Authors' reply.

We thank Juan Salvatierra and colleagues for their interest in our Article. Salvatierra and colleagues pose a critical question: how to select the right treatment and regimen for a patient with severe COVID-19. We posit that cytokine inhibition has a clear rationale in patients with severe COVID-19 with hyperinflammation, a condition caused by excess cytokine production and burdened by considerable mortality. In patients with severe COVID-19 and hyperinflammation, inhibiting excess cytokine production might reduce mortality. Hence, throughout our studies, we did not treat patients with mild or moderate disease or those without evidence of hyperinflammation, as we feel that many of these individuals have an appropriate immune response to the virus.1, 3 Conversely, we selectively and consistently evaluated anakinra in patients with severe COVID-19, respiratory insufficiency, and hyper-inflammation (defined as serum C-reactive protein ≥100 mg/L, ferritin ≥900 ng/mL, or both). In this population of patients with severe COVID-19, we initially evaluated low-dose subcutaneous anakinra (100 mg twice daily). However, we found that treatment with low-dose subcutaneously was neither associated with reductions in serum C-reactive protein nor with meaningful improvements in clinical status at day 7. Although no safety concerns emerged, the lack of marked clinical or anti-inflammatory effects led to early termination of this preliminary experience and prompted the evaluation of higher doses (5 mg/kg twice a day) and intravenous administration, which we found to be efficacious. Of note, anakinra had been previously used intravenously and at doses that far exceed those approved for licensed indications: for example, high-dose intravenous anakinra was evaluated in previous studies of septic shock and macrophage activation syndrome, both conditions sharing similarities with severe COVID-19.1, 3 Overall, anakinra has a remarkable record of safety, as a short half-life allows prompt discontinuation in the event of adverse events including bacterial infections.

It is probable that in early phases of COVID-19, viral-induced necrotic cell death of pneumocytes results in the release of IL-1α, which fosters local inflammation; eventually, this and other mediators induce the release of IL-1β by infiltrating myeloid cells, which results in runaway inflammatory responses. Thereby, it is certainly possible that lower doses of anakinra might be effective when used in earlier stages of disease development, or in patients with less severe disease, as suggested by Salvatierra and colleagues. Indeed, Huet and colleagues (among others) found lower disease regimens to be also effective in COVID-19. However, information on the best treatment regimen cannot be extrapolated from indirect comparisons between different studies. Nor should concerns about potential toxicity discourage us from titrating anakinra to the dosage that is clinically needed for optimal suppression of IL-1-mediate inflammation, since undertreating severe COVID-19 is likely to pose more threats than overdosing anakinra.