Background: Current guidelines recommend that cirrhotic patients should undergo upper gastrointestinal endoscopy for screening of esophageal varices (EV). Several studies evaluated the ability of non-invasive markers to predict the presence of EV, including ultrasound parameters.
Aims: We aimed to identify non-invasive models to predict the presence of EV.
Methods: 36 patients with compensated cirrhosis were studied, clinical, biological, ultrasonographic, with Doppler study, and endoscopic data were collected and used to calculate APRI score (AST to platelets ratio index) and congestion index (CI) of the portal vein (dividing portal vein cross-sectional area by portal blood velocity). The performance of the scores was evaluated by the area under the receiver operating characteristic curve (AUROC).
Results: 36 patients were included, 64% males, with mean age 60±8,6 years. On endoscopy, 12 patients (33%) had no EV, 14 patients (39%) had small EV and 10 patients (28%) had large EV. Portal vein diameter was higher in patients with EV than those without EV, while platelet count and portal blood velocity were lower, however with no statistical difference, (p>0,05). APRI score and CI were good predictors for the presence of EV (AUROC 0.74 and 0,635 respectively). The combination of the APRI score for liver
fibrosis and CI as a surrogate of portal hypertension resulted in an even more powerful predictor of the presence of EV (AUROC 0,797) and the best cutoff value was 0,075 (sensitivity 79,2%, specificity 75%, positive predictive value 86,4% and negative predictive value 64,3).
Conclusions: CI of the portal vein and APRI score are good noninvasive tests to predict esophageal varices. The combination
of both models is even more powerful to be used as a method to select cirrhotic patients who need to undergo upper GI endoscopy.
