A Novel Non-invasive Score Precisely Predicts Development of Esophageal Varices in Patients with Chronic Viral Hepatitis C
Abstract
Background:
Portal hypertension is a major complication of cirrhosis, leading to the development of gastroesophageal varices (GEVs). All patients with cirrhosis should be screened by endoscopy for esophageal varices (EVs) at the time of diagnosis. In recent years, several non-invasive methods for detecting EV have been evaluated. Our aim was to combine the imaging data, FibroScan, fibrosis markers, and blood parameters to propose a new score for the prediction of EV.
Materials and Methods:
180 HCV (Hepatitis C virus) positive patients with cirrhosis were enrolled in this cross-sectional study. APRI score, AAR score, FIB4 score, King score, and (PC\SD) (Platelet count/Spleen Diameter) were calculated. Abdominal ultrasonography and FibroScan were done for all patients. Data were collected to develop a scoring system as a non-invasive index combined from eight parameters; platelet count, serum albumin, spleen diameter, portal vein diameter, PC/SD ratio, FIB4, APRI, and FibroScan.
Results:
ROC curve analysis of different variables revealed that serum albumin had the highest AUC (0.83, cut off point 2.55) followed by PC/SD ratio (0.82, cut off 681), then came platelet count (0.77, cut off 91.0×103), spleen diameter (0.70, cut off 15.35), APRI (0.73, cut off 1.31), and both FIB4 (0.70, cut off 4.33) and PV diameter (0.67 cut off 13.1) and lastly FibroScan (0.69, cut off 44.6). Using ROC curve analysis to study the discrimination ability of this combination, the AUC was 0.89 (95% CI 0.81-0.98, p <0.001), which gives the score a good discrimination ability at the cut off point of 4.5 (or 56.25 %), with 85% sensitivity and 95% specificity.
Conclusion:
Our novel score could be a reliable tool to predict EV instead of a single parameter.
Keywords
Full Text:
PDFCopyright (c) 2020 GOVARESH
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.