APASL ACLF Research Consortium (AARC) liver failure score defines the time frame for liver transplant in ACLF patients

Abstract

Background: ACLF is a serious ailment with rapid deterioration and >50% 3months mortality in the absence of liver transplant. An ideal prognostic model is lacking to stratify these patents.AARC score(comprised of bilirubin, INR, hepatic encephalopathy[HE], plasma lactate and serum creatinine) has been widely used to assess degree of liver failure and mortality. We evaluated whether AARC score could predict the need and appropriate time for liver transplant in ACLF patients. Methods: Prospectively collected patients from AARC registry with a complete 90 days follow‐up were analyzed. Predictors of 90‐day mortality were derived and compared with known disease severity scores. The AARC liver failure score was validated in the cohort of ACLF patients (n=40) who received LT with a propensity matched non‐transplanted cohort (n=120).The time of death, or ineligibility for LT were recorded. Results: Among 2341 patients of ACLF with complete outcome, the 90 days transplant free survival was 46.2%. In multivariate analysis the independent predictors for outcome were age [1.02(95 CI 1.01‐1.03), p=0.03],new onset sepsis in first week [3.02(95 CI 2.15‐4.02),p<0.001] and AARC score at baseline[1.46 (95 CI 1.32‐1.61), p=0.01]. Diagnosis of ACLF with age>55 years was associated with high‐risk of death in absence of LT [HR 1.83(95 CI 1.23‐2.72), p=0.003].Among the available disease severity scores, the AARC liver failure score had the best predictability with AUROC of 0.76 with HR of 1.37(95 CI 1.26‐1.47,p<0.001) with a cut‐off 10 or more with sensitivity of 75% and specificity of 67%. The mortality was 29% with score between 5‐8 and increased sharply to 62% with score of 10 and >92% with score of 13 or above. The AARC score was validated in both LT (n=40) and non‐LT cohorts(n=1279). The 40 cases, who received liver transplant, were compared with the propensity‐matched cohort(n=120). In conditional survival estimate, the prediction for need of LT by AARC score was comparable to that of transplanted cohort. The cumulative mortality was 31%, 55%, 78% at the end of first, and second and fourth week respectively. The AARC Score is dynamic; the score of 10 and above, at the end of first week predicted best need for transplant and poor transplant‐free survival(p<0.001). Conclusion: ACLF patients had high short‐term mortality. AARC score of 10 or above, with age more than 55 years reliably stratifies patients at the end of first week for early LT before sepsis develops.