Validation of noninvasive algorithms for diagnosing esophageal varices in patients with compensated advanced chronic liver disease

Introduction Non-invasive elastography-based and biochemical approaches have been evaluated for diagnosing high-risk esophageal varices (HRV). Here we aimed to validate diagnostic performance of several available non-invasive methods for HRV among the patients with compensated advanced chronic liver disease (cACLD). Material and methods Retrospective analysis of patients who underwent liver stiffness measurement (LSM) by transient elastography (TE) in a single centre over the 5-year period, with available results of esophagogastroduodenoscopy (EGD). Only patients with suspicion of cACLD as defined by LSM≥10 kPa, with no previous decompensation were included in the final analysis. Original and expanded Baveno VI criteria (B6C), Controlled attenuation parameter (CAP), Platelets count (Plt), APRI, FIB4 index, and MELD score were evaluated against the results of EGD that served as the reference method. Results Of 861 patients with available results of LSM and EGD, 73 had LSM≥10 kPa and EGD performed within 3 months (median age 62 years, 80.8% (59/73) males, 74% (54/73) alcoholic/non-alcoholic fatty liver disease, 21/73 (28.8%) with HRV). In multivariate logistic regression analysis only LSM and platelets were independently associated with HRV. The best performing tests for ruling-out HRV (% of spared EGD; % of missed HRV) were respectively: LSM<20 kPa (53.4%; 0%), B6C (38%; 0%), Expanded B6C (47.9%; 4.8%); Plt>214 (21.9%; 0%); FIB4≤1.8 (21.4%; 0%), APRI≤0.34 (12.3%; 0%). CAP, MELD=6 alone or combined with Plt>150 did not show acceptable performance in our cohort. Conclusion The best performing noninvasive algorithms for ruling-out HRV in our cohort of patients with cACLD are based on LSM, whereas biochemical tests might also be used, but with lower number of potentially spared endocopies.