Post-treatment LSM rather than change during treatment predicts decompensation in patients with cACLD after HCV cure

Abstract

Baveno VII has defined a clinically significant (i.e., prognostically meaningful) liver stiffness measurement (LSM)-decrease in cACLD by ≥20% associated with a final LSM<20kPa, or any decrease to <10kPa. However, these rules have not yet been validated against direct clinical endpoints. We retrospectively analysed cACLD patients (LSM≥10kPa) with paired liver stiffness measurement (LSM) before (BL) and after (FU) HCV-cure by interferon-free therapies from 15 European centers. The cumulative incidence of hepatic decompensation was compared according to these criteria, considering hepatocellular carcinoma and non-liver-related death as competing risks. 2335 patients followed for a median of 6 years were analysed. Median BL-LSM was 16.6kPa with 37.1% having ≥20kPa. After HCV-cure, FU-LSM decreased to a median of 10.9kPa (<10kPa: 1002 [42.9%], ≥20kPa: 465 [19.9%]) translating into a median LSM-change of -5.3 (-8.8-[-2.4])kPa corresponding to -33.9 (-48.0-[-15.9])%. Patients achieving a clinically significant decrease (65.4%) had a significantly lower risk of hepatic decompensation (subdistribution hazard ratio [SHR]: 0.12 [95%CI: 0.04-0.35], p<0.001). However, these risk differences were primarily driven by a negligible risk in patients with FU-LSM <10kPa (5y-cumulative incidence: 0.3%) compared to a high risk in patients with FU-LSM ≥20kPa (16.6%). Patients with FU-LSM 10-19.9kPa (37.4%) also had a low risk of hepatic decompensation (5y-cumulative incidence: 1.7%), and importantly, the risk of hepatic decompensation did not differ between those with/without an LSM-decrease ≥20% (p=0.550). FU-LSM is key for risk stratification after HCV-cure and should guide clinical decision-making. LSM dynamics do not hold significant prognostic information in patients with FU-LSM 10-19.9kPa, and thus, their consideration is not of sufficient incremental value in the specific context of HCV-cure.