Intravenous anakinra for cytokine storm syndromes - Authors' reply.

We appreciate the interest in our Viewpoint on the use of intravenous anakinra in cytokine storm syndromes. Yvan Jamilloux and colleagues contribute to continuing discussions regarding the terminology of hyperinflammatory disorders; we suggest a unifying umbrella term of cytokine storm syndromes. We feel that it is the concept of hyperinflammation that is crucial; the conditions characterised by hyperinflammation overlap, and we endorse shared clinical and scientific learning.

We agree with Jamilloux and colleagues that a subgroup of patients with severe COVID-19-associated hyperinflammation might benefit from immunomodulation, but we caution against prematurely predicting optimal immunomodulatory therapies on the basis of systemic cytokine concentrations measured from peripheral blood, as local (eg, pulmonary) cytokinaemia might be more informative. The focus on inhibition of IL-6 early in the COVID-19 pandemic might be because of, at least in part, the relative ease of measurement of IL-6 and access to drug supply (tocilizumab is widely available in China), rather than a true hypothesised advantage of blocking IL-6 in preference to a different cytokine.

Nihal Martis and Alexandra Audemard-Verger discuss the timing of intervention (subclinical compared with fulminant hyperinflammatory states) and the relative merits of corticosteroids and anakinra, and suggest that anakinra should be reserved for rheumatic triggers. We have described a potential window of opportunity to reduce the risk of progression and high mortality in cytokine storm syndromes. Although steroid-free treatment regimens are difficult to achieve, we would emphasise the increased risk of infection with high-dose steroids, and the risk of masking signs of infection with IL-6 blockade (suppression of C-reactive protein or fever) as important considerations.

The use of anakinra in cytokine storm syndromes should not be confined to presentations with a rheumatic cause. Of the eight reported cases using intravenous anakinra (at the time of writing the Viewpoint), an equal number of cases had an infectious (three of eight patients) and rheumatological (three of eight patients) cause, and one patient had both an infection and a rheumatological condition (cytomegalovirus and vasculitis). As we mention in our Viewpoint, the efficacy of anakinra in non-rheumatic hyperinflammation was shown in post-hoc analyses of bacterial sepsis trials, and data suggest that IL-1 blockade might have a role in the neurotoxicity secondary to chimeric antigen receptor T-cell therapy for refractory leukaemia or lymphoma.

Additionally, cohort studies of 81 patients with severe COVID-19 suggested an efficacy signal with anakinra use.4, 5 Data from controlled trials of anakinra in patients with COVID-19 are eagerly anticipated.