Effect of corticosteroid dosing on outcomes in high-grade immune checkpoint inhibitor hepatitis.

Abstract

Consensus guidelines recommend high-dose corticosteroids (1-2 mg/kg/day methylprednisolone equivalents) for treating grade ≥3 immune checkpoint inhibitor (ICI) hepatitis. We examined the effect of corticosteroid dosing on time to alanine aminotransferase (ALT) normalization, need for additional immunosuppression, and steroid-related complications. We conducted a retrospective cohort study of 215 ICI-treated patients from 2010 to 2020 who developed grade ≥3 (ALT>200 U/L) ICI hepatitis. Patients were grouped by initial corticosteroid dose (≥1.5 mg/kg or <1.5 mg/kg methylprednisolone equivalents). Propensity scores were calculated predicting the risk of receiving the higher steroid dose and used in inverse probability of treatment weighted (IPTW) logistic or Cox regression. The 87 patients in the ≥1.5 mg/kg group received higher initial (2.0 vs. 0.8 mg/kg/day, p<0.001) and maximum (2.0 vs. 1.0 mg/kg/day, p<0.001) steroid doses than the 128 patients in the <1.5 mg/kg group. There was no difference between the higher versus lower-dose groups in development of steroid-refractory hepatitis (OR 1.22, 95% CI 0.79-1.89, p=0.365) on IPTW-logistic regression. In patients with steroid-responsive disease, there was no difference between the two groups in time to ALT normalization using either standard Cox regression (HR 1.02, 95% CI 0.72-1.45, p=0.903) or IPTW-Cox regression (HR 1.09, 95% CI 0.78-1.51, p=0.610). The ≥1.5 mg/kg group had longer exposure to corticosteroids (median 60 vs. 44 days, p=0.005) and higher incidences of infection (18.4% vs. 7.0%, relative risk [RR] 2.6, 95% CI 1.2-5.6, p=0.011) and hyperglycemia requiring treatment (23.3% vs. 7.8%, RR 3.0, 95% CI 1.5-6.0, p=0.001). In patients with high-grade ICI hepatitis, initial treatment with 1 mg/kg/day methylprednisolone equivalents provides similar hepatitis outcomes with reduced risk of steroid-related complications when compared with higher-dose regimens.