Efficacy of Tocilizumab in Patients With COVID-19 ARDS Undergoing Noninvasive Ventilation

Francesco Menzella; Matteo Fontana; Carlo Salvarani; Marco Massari; Patrizia Ruggiero; Chiara Scelfo; Chiara Barbieri; Claudia Castagnetti; Chiara Catellani; Giorgia Gibellini; Francesco Falco; Giulia Ghidoni; Francesco Livrieri; Gloria Montanari; Eleonora Casalini; Roberto Piro; Pamela Mancuso; Luca Ghidorsi; Nicola Facciolongo


Crit Care. 2020;24(589) 

In This Article

Patients and Methods

Seventy-nine consecutive patients with severe COVID-19 pneumonia and worsening acute respiratory failure (ARF) were admitted to the Pulmonology Unit of Azienda USL-IRCCS of Reggio Emilia between March 10, 2020, and April 14, 2020, and were followed-up until May 18, 2020 (median follow-up, 60 days; interquartile range, 22–69 days). All patients (71% males) had a SARS-CoV-2 infection confirmed by a positive reverse-transcriptase polymerase chain-reaction (RT-PCR) assay for SARS-CoV-2 in a respiratory tract specimen and clinical and radiological findings compatible with COVID-19 severe pneumonia (demographic and clinical characteristics are shown in Table 1). Basic laboratory profile included IL-6, C-reactive protein (CRP), ferritin, D-dimer, lactate dehydrogenase (LDH) levels, and complete blood count (CBC) (Table 2). Serum CRP and IL-6 levels at admission were elevated in all patients with mean + SD values of 11.9 ± 7.2 mg/dl and 147.2 ± 180.4 pg/ml, respectively, indicating the hyperinflammatory state of most patients. All 79 patients needed NIV at admission according to severity of ARF based on the ratio of arterial partial pressure of oxygen to fraction of inspired oxygen (PaO2/FiO2) cut-off values recorded at ARDS onset. NIV was used preferentially for moderate-severe ARF due to availability and exercise in NIV in our unit.

All patients had a PaO2/FiO2 ratio > 100 and ≤ 200 mmHg despite oxygen delivered through a Venturi mask (FiO2 60%). Table 2 shows that the mean ± SD PaO2/FiO2 ratio in our patients at admission was 120.1 ± 41.6 mmHg. NIV was provided using Philips V680™ (Respironics INC®, Pennsylvania, USA) or Hamilton G-5 (Hamilton Medical AG, Bonaduz, Switzerland) mechanical ventilators through a full-face mask. Table 1 also shows the other characteristics of the patients at admission.

Of 79 patients included in the study, 41 were treated with TCZ. The decision to treat the patients with TCZ and the formulation of the drug we used were strictly based on drug availability. The patients were enrolled from March 11 to April 14, 2020, and were followed until May 25 (the last enrolled patients had a follow-up of 31 days). Subcutaneous (SC) TCZ was administered because intravenous (IV) TCZ was not available for a period of time; furthermore, we also did not have available both formulations at the end of March and at the beginning of April (March 20–April 5, 2020) for a period of 16 days. Therefore, all 38 patients who started NIV in this period were not treated with TCZ and they represented the control group. Patients were treated with TCZ at the beginning of NIV. The primary outcome was the in-hospital mortality rate. A secondary composite outcome of no response to TCZ was represented by the patients who were intubated or died during the hospitalization in the pulmonology unit. To obtain data on infections, one of the authors (MF) revised the medical records of all patients. Only infections occurring after administration of TCZ were included in the treatment group.

Improvement was considered when the following conditions were satisfied: increase in the PaO2/FiO2 ratio of 30% after NIV, FiO2 < 50%, respiratory rate (RR) < 30 breaths per minute, expiratory tidal volume >5 ml/kg body weight expected with a pressure support < 10 cm H2O and PEEP < 8 cm H2O. In this case, NIV was progressively suspended and a Venturi oxygen mask with variable FiO2 was started on the basis of arterial blood gas data (Table 2).

NIV failure was defined according to ERS/ATS guidelines.[14] The decision to intubate a patient was made on the basis of the following criteria: persistent or worsening of acute respiratory failure (ARF) (PaO2/FiO2 ratio < 100 mmHg, respiratory rate > 36/min) despite NIV, development of conditions requiring endotracheal intubation (EOT) in order to protect airways (coma or convulsive disorder) or to manage abundant tracheal and/or bronchial secretions, or hemodynamic or electrocardiographic instability.

This study was approved by the Local Ethics Committee of Area Vasta Emilia Nord (no. 855/2020/OSS/AUSLRE) and conducted according to the principles of the Declaration of Helsinki. Since not all patients were able to give their informed consent, the Ethics Committee waived this requirement. Informed consent was sought from all surviving patients as soon as they regained their mental competence. This study has been partially funded by the Italian Ministry of Health, grant number COVID-2020-12371808. The funder had no role in the definition of the study design neither in the interpretation and publication of the results.

Statistical Analysis

Descriptive statistics for continuous variables were presented as mean ± standard deviation (SD). The Welch's t test and the Mann-Whitney test were used for comparison of two groups, while the Brown-Forsythe and Welch ANOVA test and nonparametric Kruskal-Wallis test were used for comparison of three or more continuous variables, where appropriate. Chi-square test and Fisher exact test were used for comparison of categorical variables. The impact of TCZ on mortality and on the probability of dying or being intubated was assessed by Kaplan-Meier survival function estimates, and the differences between treatment and standard therapy were analyzed with the log-rank test. The Cox proportional hazards model was used to assess the relation between tocilizumab treatment, age, and Charlson comorbidity index at diagnosis and mortality and intubation/death. We reported age and Charlson comorbidity index-adjusted hazard ratios (HRs) and 95% confidence intervals. A p value less than 0.05 was considered significant. Statistical analysis was performed using SPSS version 22.0 (IBM Statistics, Armonk, NY: IBM Corp, USA).