Admission serum ferritin levels and effect of methylprednisolone in nonintubated patients with severe COVID-19 pneumonia

Rationale: Systemic corticosteroids control the inflammatory overresponse to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in coronavirus disease 2019 (COVID- 19). Ferritin, an acute phase reactant which can also protect the host cell from oxidative stress and has in previous studies increased the cell tolerability to infection and sepsis, could be a marker of response to corticosteroids. We sought to examine whether response to methylprednisolone differed according to admission ferritin levels in severe COVID-19. Methods: Retrospective cohort study of consecutive adults with severe COVID-19 pneumonia on high-flow oxygen (FiO2 ≥50%) admitted to an academic center from March 1 to April 15, 2020, i.e., before incorporation of corticosteroids in the guidelines for severe COVID-19. We used inverse probability of treatment weights to balance patient characteristics according to methylprednisolone use and Cox proportional hazards models to examine for significant interaction of admission serum ferritin levels with methylprednisolone for outcomes. The outcomes of interest were mortality and the composite of death or mechanical ventilation. Results: Ferritin was available in 380 of 447 (85.0%) patients (age, 60±17 years;34.2% female;13.4% Black;34.5% Hispanic;body mass index, 30.4±6.4 kg/m2;O2 saturation, 89±7%;respiratory rate, 24±8 breaths/min). Of these, 142/380 (37.4%) received methylprednisolone (median dose, 160mg/day). Ferritin did not differ between patients who received methylprednisolone vs. those who did not (median [25th-75th percentile], 992 μg/L [509, 1610] vs. 893 μg/L [474, 1467];P=0.32). Patients with elevated (>1000 μg/L) ferritin had higher procalcitonin, creatinine, and transaminase, but lower Creactive protein on admission. At 28 days, 80 patients (21.1%) had died and 102 (26.8%) were intubated. Ferritin was not associated with mortality or the composite endpoint. However, in weighted analyses, methylprednisolone use was associated with significantly lower mortality in patients with ferritin >1000 μg/L (HR 0.29;95%CI 0.12-0.67;P=0.004) and significantly higher mortality in patients with ferritin ≤1000 μg/L (HR 2.88;95%CI 1.59-5.20;P 1000 μg/L (HR 0.47;95%CI 0.30-0.74;P=0.001) but not in patients with ferritin ≤1000 μg/L (HR 1.09;95%CI 0.72-1.63;P=0.69);P=0.006 for interaction (Figure 1B). Conclusions: In nonintubated patients with severe COVID-19, methylprednisolone use was associated with reduced mortality and intubation only in patients with elevated admission ferritin levels. In contrast, there was a strong signal for higher mortality with methylprednisolone in patients with low baseline ferritin. These findings need prospective validation.