Dynamics in Liver Stiffness Measurements Predict Outcomes in Advanced Chronic Liver Disease

作者：Georg Semmler, Zhenwei Yang, L. Fritz, F Köck, Benedikt Hofer, Lorenz Balcar, Lukas Hartl, Mathias Jachs, Katharina Stopfer, Anna Schedlbauer, Daniela Neumayer, Jurij Maurer, Theresa Bucsics, Benedikt Simbrunner, Bernhard Scheiner, Michael Trauner, Mattias Mandorfer, Thomas Reiberger, David Bauer · 发表于：Gastroenterology · 年份：2023 · DOI：10.1053/j.gastro.2023.06.030 · 被引用次数：102 · 研究领域：Liver Disease Diagnosis and Treatment、Liver Diseases and Immunity、Liver Disease and Transplantation

BACKGROUND & AIMS: Liver stiffness measurements (LSMs) provide an opportunity to monitor liver disease progression and regression noninvasively. We aimed to determine the prognostic relevance of LSM dynamics over time for liver-related events and death in patients with chronic liver disease. METHODS: Patients with chronic liver disease undergoing 2 or more reliable LSMs at least 180 days apart were included in this retrospective cohort study and stratified at baseline (BL) as nonadvanced chronic liver disease (non-ACLD, BL-LSM < 10 kPa), compensated ACLD (cACLD; BL-LSM ≥ 10 kPa), and decompensated ACLD. Data on all consecutive LSMs and clinical outcomes were collected. RESULTS: There were 2508 patients with 8561 reliable LSMs (3 per patient; interquartile range, 2-4) included: 1647 (65.7%) with non-ACLD, 757 (30.2%) with cACLD, and 104 (4.1%) with decompensated ACLD. Seven non-ACLD patients (0.4%) and 83 patients with cACLD (10.9%) developed hepatic decompensation (median follow-up, 71 months). A 20% increase in LSM at any time was associated with an approximately 50% increased risk of hepatic decompensation (hazard ratio, 1.58; 95% CI, 1.41-1.79; P < .001) and liver-related death (hazard ratio, 1.45; 95% CI, 1.28-1.68; P < .001) in patients with cACLD. LSM dynamics yielded a high accuracy to predict hepatic decompensation in the following 12 months (area under the receiver operating characteristics curve = 0.933). The performance of LSM dynamics was numerically better than d...