Benefits of Corticosteroids in Alcohol-Associated Hepatitis Complicated by Acute-on-Chronic Liver Failure

Introduction: Alcohol-associated hepatitis (AH) is typically characterized by systemic inflammation and is often accompanied by multiorgan failure, which can progress to acute-on-chronic liver failure (ACLF). While steroids are the mainstay of treatment for AH, their efficacy in patients presenting with concurrent ACLF is understudied. We evaluated the characteristics of those treated with steroids and their impact on the clinical course of AH-induced ACLF. Methods: Single center retrospective cohort of all patients presenting with AH (per NIAAA criteria) from 2016-2021. ACLF was defined using the EASL-CLIF criteria: 1) single renal failure, 2) single non-renal failure with renal dysfunction (Cr 1.5-1.9) and/or hepatic encephalopathy, or 3) 2 or more organ failures. The primary end point was all cause 30-day mortality. Severity of AH was measured by mDF, MELD-Na, CLIF-C OF and CLIF-C ACLF scores. CLIF-C OF evaluates the severity of each organ failure (hepatic, renal, cerebral, circulatory, respiratory, and coagulation) with a cumulative score, while CLIF-C ACLF expands on CLIF-C OF by taking age and WBC count into consideration. Results: 109 patients presented with concurrent AH and ACLF (median age 43.4 years, 83.5% male, 81.3% Hispanic). Median (IQR) mDF was 66.9 (50.5-94.5), MELD-Na was 32 (27-36), CLIF-C OF was 10 (9-11), and CLIF-C ACLF was 47 (42-54) at presentation. Sixty-eight (62.4%) patients were treated with steroids. Patients treated with steroids had a significantly higher mDF (78.4 vs 63.7, P=0.02), but similar MELD-Na, CLIF-C OF, and CLIF-C ACLF (P >0.05) (Table 1). In patients treated with steroids, the median (IQR) Lille score after 4 or 7 days was 0.215 (0.076-0.417), and 36 (52.9%) were considered steroid responders. Overall, 23 (21.1%) patients died within 30 days of hospital admission and 33 (30.3%) died within 90 days of admission (Figure 1). Using models adjusted for mDF, patients treated with steroids (vs not treated) were less likely to die within 30 days (OR 0.35, 95% CI 0.14-0.93), but more likely to be re-hospitalized within 90 days (OR 4.69, 95% CI 1.54-14.32). There was no effect of steroids on 90-day mortality (P=0.30). Conclusion: Amongst patients with AH and ACLF, those receiving corticosteroids had higher mDF scores and were less likely in ICU. Treatment with steroids reduced all cause 30-day mortality but did not influence 90-day mortality.Figure 1.: 30-day and 90-day mortality of patients who did and did not receive steroid treatment. Table 1. - Characteristics of AH Patients and ACLF With and Without Steroid Treatment TotalN=109 SteroidsN=68 No steroidsN=41 P-value Age 43.4 (35.4-52.5) 41.5 (34.7-50.5) 47.9 (39.1-52.8) 0.18 Male (%) 83.5 85.3 80.5 0.51 Hispanic (%) 81.3 79.4 80.5 0.86 mDF 66.9 (50.5-94.5) 78.4 (55.2-100.6) 63.7 (43.7-84.5) 0.02 MELD-Na 32 (27-36) 32 (28-35) 32 (27-37) 0.92 CLIF-C OF 10 (9-11) 10 (9-11) 10 (9-13) 0.23 CLIF-C ACLF 47 (42-54) 47 (42-51) 48 (43-58) 0.11 Required ICU level of care (%) 44.4 35.3 58.5 0.02 Length of stay 7 (5-14) 7 (4-17) 7 (5-12) 0.87 90-day rehospitalization (%) 41.9 45.6 12.2 0.01 30-day mortality (%) 21.1 14.7 31.7 0.04 90-day mortality (%) 30.3 26.5 36.6 0.27 Categorical variables are presented as % and continuous variables as median (IQR).