Related Experiment Videos
A multicenter multistate modeling on the natural history of liver decompensation: Longitudinal analysis of 1777
Eunice X X Tan1,2,3, Nicole Shu Ying Tang2, Nicholas Syn2,3,4
1National University Centre for Digestive Health, National University Hospital, Singapore, Singapore.
Background:
Cirrhosis often remains silent until decompensation, after which survival deteriorates sharply. Contemporary, transition-aware estimates are limited, especially in Asian-predominant cohorts. We analyzed trajectories from compensated advanced chronic liver disease (cACLD) through sequential decompensations to death.
Method:
In this multicenter study, we analyzed retrospective data from an adult cohort with liver stiffness ≥15 kPa from vibration-controlled transient elastography across 19 institutions internationally. Decompensation included ascites, hepatic encephalopathy, spontaneous bacterial peritonitis, or variceal bleeding. Multistate models were constructed to characterize patient trajectories across 4 main clinically meaningful states: baseline cACLD, first decompensation, a subsequent second decompensation, and death. Etiology-specific subgroup analyses were conducted, and the effects of univariable predictors on transitions were also examined.
Results:
Among 1777 patients, 265 (14.9%) developed a first decompensation over a median follow-up of 4.25 years. In the main multistate analysis cohort, from baseline cACLD, first decompensation occurred in 215 patients at 3.10 per 100 person-years, while death occurred at a rate of 1.01 per 100 person-years. The rate of a second decompensation was higher following the first decompensation, at 42.73 per 100 person-years. Patients with alcohol-associated liver disease had the highest rate of progression from baseline to first decompensation, whereas patients with hepatitis B etiology had the lowest progression rate into a second decompensation following a first decompensating event.
Conclusion:
Most patients with cACLD remain compensated over 10 years of follow-up, but once decompensation occurs, trajectories accelerate, and survival worsens. Transition-aware, etiology-stratified estimates enable individualized prognostication and earlier risk-guided intervention, including timely transplant referral.
