Clinical and biological data suggest that the occurrence of a cytokine storm may be crucial in determining the clinical features and severity of COVID-19, with interleukin-6 playing a major role; thus, inhibiting its activity by blocking its binding to the specific receptor could be useful [1].

Here we report on the use of tocilizumab (RoActemra©, Roche), a humanised anti-human interleukine-6 receptor antibody, in 20 hospitalised patients with SARS-CoV-2 infection confirmed by reverse transcriptase-polymerase chain reaction assay and severe pneumonia (PaO2/FiO2 ratio < 300) undergoing high flow oxygen or noninvasive positive-pressure ventilation with interleukin-6 serum levels higher than 20 pg/ml; mechanically ventilated patients were excluded. In these patients, tocilizumab 8 mg/kg intravenously (maximum 800 mg) were added to standard treatment (hydroxychloroquine, prophylactic enoxaparin, oxygen, and nutritional support); a further identical dose was administered 12 h later if no significant amelioration could be found. The drug was prescribed as off-label, and local Ethics Committee approval and informed consent were obtained for each patient. Patients were followed until discharge or death; 1 patient is still hospitalised.

The age range was 32 to 78 years (median 60 years) (Table I). At baseline, 18 patients were receiving noninvasive positive-pressure ventilation, and 2 patients recieved high-flow oxygen; median PaO2/FiO2 ratio was 137 (range: 101–210). The median duration of symptoms before tocilizumab therapy was 10.5 days (range: 3–21 days), and tocilizumab was administered a median of 2.5 days after admission (range: 0–13). Four patients underwent mechanical ventilation during follow-up: 1 patient died, 2 patients were extubated and discharged, and 1 patient was extubated but still in intensive care due to intervening septic complications. The time to cessation of high-flow oxygen or noninvasive positive-pressure ventilation was 5.5 days, and the median duration of hospitalisation was 15 days (range: 7–52 days). The mortality rate was 5%, and there were no clinical and laboratory differences between patients who required or did not require mechanical ventilation during hospitalisation. No adverse drug reaction was reported.

Table I

Demographic and clinical characteristics of the patients

Parametervalue
Male/female19/1
Age, median (IQR) [year]60 (55–67)
Age category, n (%):
 < 50 years2 (10)
 50 to 70 years14 (70)
 ≥ 70 years4 (20)
Oxygen-support category, n (%):
 High-flow oxygen2 (10)
 Noninvasive positive-pressure ventilation18 (90)
Time from admission to tocilizumab therapy, median (IQR) [days]2.5 (1.7–4)
Duration of symptoms before tocilizumab therapy, median (IQR) [days]10.5 (7.7–12.2)
Laboratory values, median (IQR):
 White blood cells [109/l]8.3 (6.8–9.7)
 Lymphocytes [109/l]1.2 (0.8–1.6)
 AST [IU/l]54 (38–94)
 ALT [IU/l]46 (17–86)
 Creatinine [mg/dl]0.98 (0.95–1.35)
 Ferritin [ng/ml]786 (442–1198)
 C-reactive protein [mg/l]146 (75–240)
 D-dimer [ng/ml]1574 (1145–6285)
 LDH [IU/l]429 (337–573)
 IL-6 [pg/ml]132 (90–160)
Hospitalisation, median (IQR) [days]15 (13–19.5)
Death rate, n (%)1/20 (5%)

[i] IQR – interquartile range, ALT – alanine aminotransferase, AST– aspartate aminotransferase, LDH – lactate dehydrogenase, IL-6 – interleukin-6.

It is still a matter of debate whether the clinical course of COVID-19 is different in patients with rheumatic disease or in those undergoing a disease-modifying treatment [2, 3]. A previous report from China showed that tocilizumab could be an effective treatment, able to reduce the mortality of COVID-19 patients [4]. Data from our preliminary report seems to confirm such an effect and compare well with those from other studies: the mortality rate was 14% in severe Chinese patients [5], 22% in patients treated with lopinavir-ritonavir [6], and 13% in patients treated with remdesivir [7], while the aggregated data of death and intubation was 20% (4 patients) in our series, and 32.3% in patients treated with hydroxychloroquine alone [8]. Adding Tocilizumab to standard treatment in patients with severe COVID-19 could therefore be useful.

Our study has significant shortcomings: mainly the small size of the treated cohort and the lack of a control (possibly randomised) group, which limit the interpretation of the results.

These preliminary data therefore need confirmation in larger populations by case-control studies and mainly by randomised studies.