Published in:
24-05-2022 | Hypertension | Original Article
ABC: a novel algorithm to stratify decompensation risk in patients with compensated advanced chronic liver disease (CHESS2108): an international, multicenter cohort study
Published in: Hepatology International | Issue 5/2022
Login to get accessAbstract
Background
Liver-related death is preceded by clinical decompensation; therefore, the risk stratification of decompensation in compensated advanced chronic liver disease (cACLD) is extraordinary significant.
Methods
The international, multicenter study included three cohorts from January 2009 to August 2021. In training cohort, the unfavorable Baveno VI criteria patients were used to develop the novel CHESS criteria to stratify decompensation risk. The Algorithm based on Baveno VI criteria plus CHESS criteria (ABC model) was validated in validation cohort, and used to diagnose clinically significant portal hypertension (CSPH) in hepatic venous pressure gradient (HVPG)-performed cohort.
Results
A total of 1377 cACLD patients were enrolled. In training cohort, multivariate analysis revealed that liver stiffness measurement (LSM), platelet count (PLT), albumin, alanine aminotransferase (ALT) and varices were the independent risk factors for hepatic decompensation. The novel CHESS criteria was produced (0.036 × LSM [kPa]) + (− 0.013 × PLT [109/L]) + (− 0.068 × Albumin [g/L])) + (− 0.016 × ALT [U/L]) + (0.651 × Varices [present: 1, absent: 0]), and < − 4.4, − 4.4 to − 3.1 and > − 3.1 indicated the low risk, medium risk, and high risk of decompensation, with a 3 year-time-dependent area under the curve (tAUC) of 0.851 (0.800–0.901). In validation cohort, the 3 year-tAUC of ABC model was 0.843 (0.742–0.943). Notably, in HVPG cohort, the high risk group was used to rule in CSPH with a positive predictive value of 93.0%.
Conclusions
The ABC model can stratify the risk of decompensation in cACLD. HVPG evaluation can be waived in both low risk and high risk cACLD patients as they can be managed by Baveno VI criteria and non-selective β-blockers intervention, respectively, and the remaining medium risk patients need further HVPG evaluation.