A multi-centre retrospective cohort study to examine the effect of high-dose steroids in COVID-19 pneumonitis admitted to Intensive Care with ARDS

ABSTRACT

Introduction: The use of systemic corticosteroids to suppress SARS-CoV-2-induced lung inflammation is advocated in the treatment of COVID-19 ARDS.1,2 Whilst the evidence for low dose early corticosteroids in COVIDARDS is well established, the effect of larger steroid doses (i.e. short-term 'pulse-dose') is yet to be investigated. Objectives: The objective of this study was to examine the effect of pulse dose steroids on ventilatory parameters such as oxygenation in COVID-19 patients with and without established fibrosis or organising pneumonia (OP). Methods: This was a multi-centre, retrospective observational study performed at four teaching hospitals, with the following inclusion criteria adult patients requiring invasive mechanical ventilation with confirmed PCR SARS-CoV-2 infection;and received high dose steroids for treatment for COVID-ARDS (defined as dose ≥ 20mg dexamethasone or an equivalent dose of methylprednisolone). This study was carried out as a service evaluation within the National Health Service (NHS) and recorded under the auspices of the clinical audit office at Imperial College HealthcareNHS Trust and Institutional Data Protection Office. Study patients were followed for 14 days or until they were discharged from the ICU and physiological or ventilatory variable data was retrospectively collected from patient records. Results: In total, 92 patients were included 14 patients 20mg/day dexamethasone;5 patients 50mg/day dexamethasone;16 patients 500mg methylprednisolone;and 57 patients 1000mg methylprednisolone. Our data demonstrate a statistically significant improvement in PaO2/FiO2 (P/F) ratio over time, from baseline to day 14, in those patients who received 1000mg Methylprednisolone (baseline PaO2 14.47 kPa, Day 3 17.51 kPa, Day 7 19.51 kPa, Day 14 22.87 kPa, p<0.001). Whilst not statistically significant, there was a trend to higher P/F ratios by day 14 in patients who received 500mg Methylprednisolone group. There was no increase in P/F ratios in those patients who received 20mg or 50mg dexamethasone. The increase in P/F ratio was most apparent in those patients who had evidence of fibrosis on CT scan, although some benefit was seen in those patients who did not fibrosis on radiological imaging. Cross sectional random effects models were used to determine the effect of 1000mg methylprednisolone on improvement in P/F ratio and demonstrate that there was an increase of P/F ratio of more than 0.38 kPa per day in those patients that received 1000mg methylprednisolone. The was no significant effect on compliance measures. There was also a trend to more ventilator free days but no difference in mortality in those patients receiving large dose methylprednisolone. Reassuringly, rates of fungal infection and pneumothorax/pneumomediastinum for patients who received steroids, including those with high dose, were equivocal. Conclusion: In this study, we present novel data suggesting that large doses of methylprednisolone may be beneficial in patients with severe COVID-19, late in the disease course when ARDS is well established. This benefit was not demonstrated in patients treated with lesser (but still high) doses of steroids (i.e. 20mg or 50mg of dexamethasone) and suggest that larger pulsed-dose steroids may induce reversibility of the disease process, particularly in those who have developed fibrosis.