Fatores preditivos clínicos e laboratoriais de varizes esofágicas em crianças e adolescente com síndrome de hipertensão porta.

Abstract

The objective of this study is to determine the clinical and laboratory parameters that could help predict the presence of esophageal varices in children and adolescents with portal hypertension syndrome. All 111 children and adolescents with portal hypertension syndrome without previous history of upper gastrointestinal bleeds admitted to the Outpatient Pediatric Hepatology Service of the Pediatric Gastroenterology Department at Hospital das Clínicas da UFMG, between January 1998 and December 2005, were included in the study. The majority of patients were male (52.3%). Age at study entry varied from one month to 17 years, with a median of 3.7 years. Twelve patients had portal vein thrombosis (10.8%), 14 had congenital hepatic fibrosis (12.6%) and 85 (76.6%)showed cirrhosis in the liver biopsy, most frequently due to biliary atresia. Every patient underwent upper endoscopy for investigation of esophageal varices. Sixty percent of the patients showed varices at the first endoscopy: large in 15.3% and medium-sized in 13.5% of the 111 patients. All the patients with portal vein thrombosis showed varices at the first endoscopy. The prevalence of esophageal varices at first endoscopy was 78.6% amongst patients with congenital hepaticfibrosis and 52% amongst cirrhotics. The risk factors for varices were determined as: gender, age, baseline diagnosis, Child-Pugh class (for cirrhosis patients), ascitis, encephalopathy, splenomegaly, platelet count, serum aminotransferases, albumin, total bilirubin and prothrombin activity. In order to identify the independent variables associated with the presence of esophageal varices we initially used univariate analysis, followed by logistic regression. After multivariate analysis we calculated sensitivity, specificity, positive and negative predictive values, andaccuracy of the predictive variables identified amongst the cirrhotic patients, using the upper endoscopy as the reference test. Multivariate analysis of the 111 patients showed splenomegaly (OR=7,68), thrombocytopenia (OR=2,5), portal vein thrombosis, and congenital hepatic fibrosis as causes of portal hypertension (OR=6,15) to be predictive of esophageal varices. When we used the samestatistical procedures for the 85 cirrhotic patients, splenomegaly andhypoalbuminemia remained as significant predictors of the development of varices. From all the predictive factors studied only splenomegaly showed high sensibility (97.7%) and negative predictive values (91.7%), adequate for screening for esophageal varices amongst cirrhotic patients. We concluded that endoscopic screening should be performed in all portal vein thrombosis and congenital hepatic fibrosis patients due to the high prevalence of varices in these two groups. Cirrhotic patients who develop splenomegaly should also undergo upper endoscopy.