Background: Current understanding of the rates and trajectories of progression to cirrhosis complications is derived from outdated data. We examined the progression patterns of cirrhosis and their implications, focusing on the different outcomes of prolonged illness vs. early mortality using contemporary data.
Methods: We conducted a retrospective cohort study of adult patients with compensated cirrhosis (CC) from 130 Veterans Affairs hospitals diagnosed between 10/1/2010 and 08/30/2015, with follow-up through 08/31/2023. Utilizing an unadjusted semi-Markov multistate model, we examined transitions from CC to intermediate disease events (ascites, hepatic encephalopathy [HE], esophageal variceal bleeding [EVB], hepatocellular carcinoma [HCC], any two complications, or three or more complications) and, ultimately, to death. We also examined the effects of age and etiology (HCV, alcohol, MASH).
Results: Of 24,728 patients with CC (median age 62 years, 18% HCV, 25% alcohol, 26% HCV and alcohol, 30% MASH), 30% developed ascites, 10% HE, 9% HCC, and 1% developed EVB as the first decompensation-related event. Most patients with CC progressed to a single complication (50%), 3% experienced >2 complications at the onset of decompensation, and 21% died before any complications. In total, 26% remained as CC during a median follow up of 6.2 years. The risk of progressing from 1st complication to a subsequent complication was much higher than that from CC to 1st complication (Figure 1). The risk of developing a 2nd complication was higher in patients with EVB, whereas patients with ascites, HE, or HCC had a higher risk of progressing directly to death (Figure 1).
The progression intensity (instantaneous hazard) was the highest from 3 complications to death, followed by from 2 complications to death (Figure 2). Younger patients had an accelerated progression intensity across most transitions. Risk of progression to HCC was higher for patients with HCV than for alcohol related cirrhosis (HR = 2.13, 95% CI: 1.85 – 2.45). More patients with alcohol-related cirrhosis progressed to non-HCC complications and at a faster pace than HCV cirrhosis (HR from CC to ascites for HCV- vs. alcohol-related cirrhosis = 0.53, 95% CI=0.49- 0.57). Progression rates were similar for patients with alcohol and MASH across most transitions.
Conclusions: In a cohort of patients with compensated cirrhosis, half progressed while others remain compensated, and a large proportion died. The progression between states varied by age/etiology. Transitions accelerated as patients progressed in their clinical course, underscoring the need for future trials to focus on the high yield period before patients develop first complication. These data have implications for counseling patients and planning interventions.

