Corticosteroids in severe COVID-19: a critical view of the evidence

 

Corticosteroids in severe COVID-19: a critical view of the evidence

 

by Daniel De Backer, Elie Azoulay and Jean-Louis Vincent 

 

Critical Care volume 24, Article number: 627 (2020)

 

Since December 2019, SARS-CoV-2 has infected millions of people worldwide, causing excess deaths and a surge in demand for ICU beds. With no effective therapies against SARS-CoV-2, randomized trials of several potential therapeutic agents, including steroids, have been conducted. Although use of steroids in patients with ARDS [1] and severe viral pneumonia [2, 3] has been challenged, several arguments support the biological plausibility of steroid use in patients with severe COVID-19. First, autopsy studies in COVID-19 patients showed lymphocyte alveolitis, acute fibrinous injury and organizing pneumonia [4], which are all probably steroid-sensitive. Second, COVID-19 leads to activation of endothelial cells causing not only systemic inflammation but also microvascular thrombosis, pulmonary infarcts and venous thromboembolism [4, 5]. Admittedly, there are also arguments against steroid use. First, viral particles are often found at autopsy [4], and steroids may decrease viral clearance. Second, steroids only influence the inflammatory component of the inflammation–thrombosis–hypoxia interaction [6], suggesting that steroids may be less effective once thrombi have developed.

The RECOVERY trial compared administration of 6 mg/day dexamethasone for 10 days to usual care in 6425 hospitalized patients with SARS-CoV-2 infection. Survival was significantly higher in the dexamethasone-treated patients, especially in the subgroup of 1007 patients receiving invasive mechanical ventilation [7]. As a result of the RECOVERY findings, three further steroid trials, focusing on ICU patients, were stopped prematurely after inclusion of 384 [8], 299 [9] and 149 [10] patients, respectively. A meta-analysis of the available data concluded that administration of systemic steroids was associated with a decrease in 28-day mortality [11]. Nevertheless, although administration of steroids appears promising, several limitations must be considered when interpreting the results