Application of shear wave elastography for Baveno VI criteria and combined model using spleen stiffness measurement to rule out high-risk varices in advanced chronic liver disease.

Abstract

Purpose In 2015, Baveno VI workshop suggested consensus recommendation about non-invasive surveillance of patients with compensated advanced chronic liver disease (cACLD) that the screening endoscopy could be avoided safely in patients with platelet count over 150,000 / mm3 and liver stiffness (LS) value less than 20 kPa using transient elastography (TE). Recently, Colecchia et al. reported that by adding spleen stiffness (SS) criteria to the Baveno VI criteria sequentially, screening endoscopy can be avoided safely in a greater proportion of patients with cACLD. The purpose of this study is to compare Baveno VI criteria, SS value and a sequentially combined model of Baveno VI criteria and SS value, using 2D-SWE, to avoid screening endoscopy for high risk varices (HRV) safely in cACLD patients. Materials and Methods We enrolled patients who successfully underwent LS and SS measurements by 2D-SWE and recently underwent esophagogastroduodenoscopy(EGD) from April 2017 to July 2018. And the PLT count, LS value, SS value, presence and grade of esophageal varices (EV), child-pugh classification grade and laboratory findings were reviewed retrospectively. Patients with a decompensation symptom (GI bleeding, hepatic encephalopathy, jaundice, ascites), bleeding history or ligation history of EV, portal vein thrombosis, hepatic failure, non-cirrhotic portal hypertension, transjugular intrahapatic portosystemic shunt (TIPS), or hepatic and systemic malignancy were excluded. We evaluated the clinical significance of platelet count, LS and SS value and analyzed the patient factors for the prediction of HRV. We also compared the ratio of EGD spared patients using Baveno VI criteria, SS cut-off and a sequential combination of Baveno VI criteria with SS cut-off. Results The AUROC values of platelet count, LS and SS for estimating the HRV were 0.701, 0.757, and 0.844, respectively. Platelet count, LS and SS value were found to be independent predictors of HRV. In the ROC curve analysis of SSM, the cut-off value of SS to exclude HRV was ≤ 27.3 kPa. The ratio of spared EGD was 18.6% for Baveno VI criteria, 28.8% for SS cut-off ≤ 27.3 kPa and 36.1% for sequential combination of Baveno VI criteria and SS cut-off. The ratio of missed HRV patients at all criteria was less than 2%. Conclusions SS measurement using 2D-SWE is an excellent predictor of HRV for cACLD patients. Screening endoscopy can be avoided safely when Baveno VI criteria and SS cut-off are applied together.