Survival Benefit From Corticosteroids in Severe Alcohol-associated Hepatitis Attributed to Clinical and Treatment Differences in a Large Multicenter Cohort.

Abstract

Introduction: Corticosteroids are recommended by multiple society guidelines for the treatment of severe alcohol-associated hepatitis (AH). However, their use remains controversial due to inconsistent studies regarding their survival benefit.

Methods: This was a retrospective cohort study of first-time hospitalizations for severe AH (Maddrey discriminant function ≥ 32) admitted to the Veterans Health Administration between January 3, 2005, and December 5, 2020, (i) evaluating the effect of corticosteroid therapy on all-cause survival, (ii) characterizing the clinical and psychosocial factors associated with corticosteroid use, and (iii) determining the effect of duration of corticosteroid therapy on all-cause survival among treatment-responsive patients (Lille score < 0.45).

Results: During the study period, 2,618 patients were admitted with severe AH, of whom 1,083 (41.37%) received corticosteroids. Although corticosteroids were significantly associated with improved all-cause survival in the unadjusted model ( P = 0.022), no survival benefit was observed in the adjusted model after accounting for baseline and admission characteristics (adjusted hazard ratio [aHR] = 1.01, P = 0.818). Psychiatry consultation was the only factor evaluated that was protective against mortality (aHR = 0.67, P < 0.001). Among the 428 patients (49.7%) responsive to corticosteroids, duration of therapy was not associated with overall survival on unadjusted ( P = 0.696) or adjusted models (aHR = 1.12, P = 0.710 for a ≥28-day course compared with a ≤7-day reference).

Discussion: Despite being recommended by clinical guidelines for severe AH, corticosteroids have low utilization with no survival benefit after accounting for differences in patient characteristics and practice patterns. Among patients with treatment response per the Lille score, no difference was observed in overall survival between shorter and longer durations of corticosteroid therapy.