Anakinra for patients with COVID-19

The Article published by Giulio Cavalli and colleagues in The Lancet Rheumatology,¹ in combination with our case series of nine patients,² supports the targeting of the IL-1 pathway as a safe and promising approach in the various types of lung involvement in patients with COVID-19. These complementary studies that cover the entire spectrum of severity of lung involvement in hospitalised patients with COVID-19, provide the first efficacy and safety data for anakinra, depending on the doses used and the stage of lung injury.

The study by Cavalli and colleagues focused on moderate-to-severe acute respiratory distress syndrome (ARDS) associated with COVID-19, for which low-dose anakinra was ineffective, prompting a significant increase in anakinra dosage that yielded better outcomes.¹ However, in our study,² a low-dose regimen, similar to that proposed by Cavalli and colleagues,¹ was associated with good outcomes in patients with moderate-to-severe COVID-19 and oxygen-dependent pneumonia with a high risk of worsening (due to comorbidities, intense inflammatory syndrome, or both).

In a longitudinal transcriptomic cytokine analysis, Ong and colleagues³ showed that peak expression of IL-1α and IL-β preceded the nadir of respiratory function in patients with severe COVID-19 pneumonia. Therefore, the early use of anakinra during COVID-associated lung injury might prevent the progression to ARDS and mechanic ventilation.

Because of safety concerns, Cavalli and colleagues chose to stop anakinra as soon as liver transaminases exceeded the upper normal limit by more than three times.¹ However, such an increase in liver transaminases is common in patients COVID-19.⁴ We noticed in patients treated with low-dose anakinra, as well as in untreated patients (data not shown), that concentrations of liver transaminases and triglycerides increase, even when C-reactive protein concentrations decreased. Therefore, the increase in triglyceride levels, which is not a classic anakinra side-effect, and in liver transaminases could correspond to haemophagocytic lymphohistiocytosis-like manifestations of COVID-19. Of note, increased liver transaminases might also have been a side-effect of lopinavir–ritonavir treatment in the study by Cavalli and colleagues.1, 4 Moreover, liver injury and increased liver transaminases are associated with COVID-19 severity.4, 5 Discontinuation of anti-IL-1 therapy due to increased liver transaminases could result in a lost opportunity for patients to receive anakinra who need it most, although the withdrawal of anakinra did not shorten the median duration of treatment in the study by Cavalli and colleagues.¹ Thus, we believe that only stronger and continuous increases in liver transaminases, mostly those that appeared after the onset of anakinra, should be considered to avoid premature discontinuation of therapy.

We hope that the ongoing randomised controlled trials will confirm these promising results and provide answers to the outstanding questions.