ABSTRACT
BACKGROUND: Immune-related hepatitis (IRH) occurs in 1 to 18% of immune checkpoint inhibitor (ICI)-treated patients. Steroids are usually recommended for grade≥3 IRH, but their impact on IRH resolution and patient survival remains unclear. METHODS: We retrospectively analyzed a prospective cohort of 339 patients treated at Saint-Louis Hospital (Paris, France) with ICIs for advanced melanoma. Cases of grade≥3 IRH were collected and analyzed. Two groups were compared for their biological features and time for IRH resolution and survival: patients who received steroids (steroids group: SG) and patients who did not (nonsteroids group: NSG). FINDINGS: Grade≥3 IRH was observed in 21 patients. Thirteen were treated with steroids (SG), and 8 were not (NSG). The median time for toxicity resolution was 49 days in SG and 24 days in NSG (P=0.62). All but one patient showed a favorable outcome. Two-year survival was 56% in SG and 54% in NSG (P=0.83). Higher transaminase (P=0.002) and bilirubin (P=0.008) and lower prothrombin (P=0.035) levels were observed in SG than in NSG. For 8 (4 SG/4 NSG) patients, ICI was resumed without any hepatitis relapse. INTERPRETATION: Favorable outcomes may be achieved spontaneously and with no steroids in patients with severe IRH. Steroid initiation should be discussed in cases of high bilirubin levels and decreased prothrombin levels. ICI could be resumed without hepatitis relapse. We propose a management algorithm for grade≥3 IRH that should be validated in larger and prospective cohorts.