Aim: Non-invasive methods are needed to identify esophageal varices (EV) in patients with chronic liver disease. To this end, we evaluated liver stiffness (LS)-spleen diameter-to-platelet ratio risk score (LSPS) in predicting EV among Japanese chronic hepatitis C patients.
Methods: A total of 99 patients with chronic hepatitis C who had undergone endoscopy, LS measurement and ultrasonography between 2013 and 2014 were enrolled. Clinical data were compared with those for other non-invasive markers (platelet count, aspartate aminotransferase-to-platelet ratio, FIB-4 index and platelet-to-spleen ratio), spleen size, LS and controlled attenuation parameter. Diagnostic applicability was assessed by the area under the receiver-operator curve (AUC) and predictive values along with multivariate logistic regression.
Results: LSPS was significantly correlated to the grade of EV (ρ = 0.617, P < 0.001) and was superior to the other non-invasive indices for determination of EV. LSPS was independently associated with EV by multivariate logistic regression analysis (odds ratio, 3.079; 95% confidence interval [CI], 2.137-4.438; P < 0.001). The cut-off value of LSPS for EV was 0.7, for which the AUC, sensitivity, specificity, positive predictive value, negative predictive value and accuracy were 0.928 (95% CI, 0.876-0.980), 86.3%, 89.6%, 70.4%, 95.8% and 88.9%, respectively.
Conclusion: LSPS may also identify EV in patients with chronic hepatitis C in Japan. The clinical values of LSPS for EV risk merit further validation in larger prospective studies.
Keywords: FibroScan; esophageal varices; liver stiffness; liver stiffness-spleen diameter-to-platelet ratio risk score; transient elastography.
© 2016 The Japan Society of Hepatology.