Acute-on-Chronic Liver Failure in Severe Alcohol-Associated Hepatitis: Impact on Management, Prognostication, and Urgency of Liver Transplantation

Background/Objectives: Severe alcohol-associated hepatitis (SAH) is a complication of alcoholic liver disease associated with high mortality. Effective identification of factors predicting the course, outcome, and need for early liver transplantation (eLT) in patients with SAH is crucial. Acute-on-chronic liver failure (ACLF), a syndrome characterized by an acute decompensation of cirrhosis, failure of one or more organs, and high short-term mortality, can occur in SAH, and its impact on management, prognostication, and need for eLT should be clarified. Methods: This is a single-center, retrospective study of patients with SAH admitted to the ASST GOM Niguarda in Milan between April 2016 and May 2023. Results: One hundred patients were included. Median Maddrey’s discriminant function (MDF) was 72.1 (IQR 51.6–101.3), and the median model for end-stage liver disease—sodium (MELD-Na) score was 28.5 (IQR 25.0–32.5). At presentation, 57 patients (57%) had ACLF 0–1 (54 ACLF 0—no organ failure [OF] or single non-renal OF with serum creatinine < 1.5 mg/dl and no hepatic encephalopathy; 3 ACLF 1—defined by the presence of at least renal failure or any other single OF if associated with renal dysfunction and/or grade I–II hepatic encephalopathy), and 43 patients (43%) had ACLF 2–3 (31 ACLF 2—defined by the presence of two OFs, 12 ACLF 3—defined by the presence of three or more OFs). In the latter group, circulatory failure was present in two (4.65%) and respiratory failure in three (6.98%). SAH was the only precipitating factor in 3 (100%) and 39 (90.7%) patients with ACLF 1 and ACLF 2–3, respectively, with infection being an associated factor in 5 (11.63%) and GI bleeding in 2 (4.65%) of subjects with ACLF 2–3. Non-response to medical treatment (MT) was significantly higher in patients with ACLF 2–3 (38/43, 88.4%) than in those with ACLF 0–1 (25/57, 43.8%). ACLF 2–3 was associated with a corresponding higher need for eLT (ACLF 2–3: 15/43, 34.9% vs. ACLF 0–1: 10/57, 17.54%%; p = 0.0474). ACLF status at presentation (ACLF 2–3 vs. ACLF 0–1) has a similar predictive accuracy of outcome (death or liver transplantation [LT]) to MELD-Na (as a 1-unit increase) (3.261 [1.999; 5.318], concordance index 0.702 vs. 1.154 [1.094; 1.217], concordance index 0.708). Conclusions: ACLF 2–3 is a frequent presentation of SAH, 43% of the cases. Given that most of these patients do not respond to MT (approximately 90%), fast-track evaluation for LT is advised for patients with a favorable psycho-social profile.