Noninvasive Methods of Predicting Large Esophageal Varices in Children With Intrahepatic Portal Hypertension

Objective: Esophageal variceal bleeding is a severe complication of portal hypertension. The standard diagnostic screening test and therapeutic procedure for esophageal varices (EV) is endoscopy, which is invasive in pediatric patients. This study aimed to evaluate the role of noninvasive parameters as predictors of large varices in children with intrahepatic portal hypertension. Methods: Participants included in this cross-sectional study underwent a screening endoscopy. Variceal size, red marks, and portal gastropathy were assessed and rated. Patients were classified into two groups: Group 1 (G1) with small or no varices and Group 2 (G2) with large varices. The population consisted of 98 children with no history of gastrointestinal (GI) bleeding, with a mean age of 8.9 +/- 4.7 years. The main outcome evaluated was the presence of large varices. Results: The first endoscopy session revealed the presence of large varices in 32 children. The best noninvasive predictors for large varices were platelets (Area under the ROC Curve [AUROC] 0.67; 95% CI 0.57-0.78), the Clinical Prediction Rule (CPR; AUROC 0.65; 95% CI 0.54-0.76), and risk score (AUROC 0.66; 95% CI 0.56-0.76). The logistic regression model showed that children with a CPR value under 114 were 8.59 times more likely to have large varices. Risk scores higher than -1.2 also increased the likelihood of large varices (OR 6.09; P = 0.014), as did a platelet count/spleen size z score lower than 25 (OR 3.99; P = 0.043). The combination of these three tests showed a high negative predictive value. Conclusions: The CPR, the risk score, and the platelet count/spleen size z score could be helpful in identifying cirrhotic children who may be eligible for endoscopy.