Predicting Hepatic Decompensation in Patients with Metabolic Dysfunction-Associated Steatotic Liver Disease-Related Cirrhosis: The ABID-LSM Model

Validating a Modified ABIDE Model for Predicting Hepatic Decompensation in MASLD-Related Cirrhosis

Accurate risk prediction of hepatic decompensation in cirrhosis patients with MASLD is vital for effective treatment planning.

This study validates a modified ABIDE model, which replaces esophageal varices (EV) with liver stiffness measurement (LSM) via transient elastography.

The updated model's performance was compared to other established tools to evaluate its accuracy and clinical utility.

Study Design and Methodology

A multicenter retrospective cohort study across eight Spanish tertiary centers included 388 patients with

compensated cirrhosis due to MASLD.

Exclusion criteria were prior hepatocellular carcinoma (HCC), hepatic decompensation, or liver transplantation.

Baseline risk scores—ABIDE, modified ABIDE-LSM, NFS, FIB-4, ALBI, ALBI-FIB-4, and the NAFLD decompensation risk score—were calculated for all participants.

Predictive accuracy was evaluated using competing risk regression, time-dependent area under the curve (tAUC), and Brier scores.

The study compared the performance of the ABIDE model, its modified version with LSM, and LSM alone to assess stheir predictive utility.

Evaluating the Efficacy of ABID-LSM for Predicting Hepatic Decompensation: A Longitudinal Study

During a median follow-up of 31 months (ranging from 18 to 60 months), hepatic decompensation was observed in 105 patients (27%).

In a subset of 273 patients with available liver stiffness measurements (LSM), the incidence was 20%.

The predictive accuracy of the ABID-LSM model at five years was notably higher, with a time-dependent area under the curve (tAUC) of 0.80, compared to ABIDE (tAUC 0.75, p=0.03) and LSM (tAUC 0.63, p<0.001).

Additionally, ABID-LSM demonstrated superior net reclassification compared to both ABIDE and LSM. Discriminative measures indicated that ABID-LSM outperformed ABIDE and LSM, with k-statistics of 0.73, 0.71, and 0.68, respectively, and D-statistics of 1.91, 1.77, and 1.65.

The cumulative incidence of hepatic decompensation at five years was significantly elevated using a model threshold of ≥7.1 (22% vs. 5%, sub-HR=5.3, 95% CI: 1.71-7.89; p<0.001).

The ABID-LSM model exhibited good calibration (slope 0.99) and overall performance (Integrated Brier score 0.15), outperforming other models such as NFS and ALBI (tAUC=0.72), FIB-4 (tAUC=0.74), ALBI-FIB-4 (tAUC=0.73), and NAFLD decompensation risk score (tAUC=0.65) (all p<0.001).

Key Insight

The ABIDE-LSM model demonstrates greater accuracy in predicting hepatic decompensation than existing risk prediction tools. Further studies are required to confirm its ability to reliably identify at-risk patients who may benefit from targeted pharmacological interventions.

By - Eeshan Aggarwal

Reference: Hepatology. Volume 80, Issue S1. Abstract Supplement for The Liver Meeting by the American Association for the Study of Liver Diseases (AASLD), November 15-19, 2024, San Diego, CA.