P A Russo1, L J Chartrand, E Seidman. 1. Department of Pathology, Gastroenterology Service, Hôpital Ste-Justine, and Université de Montréal, Montréal, Québec, Canada.
Abstract
OBJECTIVE: To prospectively evaluate and compare the sensitivity, specificity, and positive and negative predictive values of serum antigliadin (AGA) and antiendomysium antibodies (EMA) in predicting the initial diagnosis of celiac disease. DESIGN: Sera were tested prospectively for IgA and IgG AGA by enzymed-linked immunosorbent assay and IgA EMA by immunofluorescence techniques on monkey esophagus and human umbilical cord sections in 95 pediatric patients referred for duodenal biopsies. PATIENTS: Ninety-five pediatric patients were referred for duodenal biopsies, with a clinical suspicion of celiac disease; 24 of those patients had celiac disease by criteria of the European Society for Pediatric Gastroenterology and Nutrition. SETTING: A pediatric gastroenterology clinic of a tertiary care pediatric university hospital. RESULTS: EMA testing on human umbilical cords was the most specific but was also the least sensitive. All the patients with biopsy-proven celiac disease were identified by either one or both serologic tests (100% combined sensitivity). The combination of AGA and EMA on monkey esophagus resulted in a negative predictive value of 100% accuracy. CONCLUSIONS: A combination of AGA and EMA tests resulted in 100% sensitivity and 100% negative predictive value, useful in selecting patients for duodenal biopsy.
OBJECTIVE: To prospectively evaluate and compare the sensitivity, specificity, and positive and negative predictive values of serum antigliadin (AGA) and antiendomysium antibodies (EMA) in predicting the initial diagnosis of celiac disease. DESIGN: Sera were tested prospectively for IgA and IgG AGA by enzymed-linked immunosorbent assay and IgA EMA by immunofluorescence techniques on monkey esophagus and human umbilical cord sections in 95 pediatric patients referred for duodenal biopsies. PATIENTS: Ninety-five pediatric patients were referred for duodenal biopsies, with a clinical suspicion of celiac disease; 24 of those patients had celiac disease by criteria of the European Society for Pediatric Gastroenterology and Nutrition. SETTING: A pediatric gastroenterology clinic of a tertiary care pediatric university hospital. RESULTS: EMA testing on human umbilical cords was the most specific but was also the least sensitive. All the patients with biopsy-proven celiac disease were identified by either one or both serologic tests (100% combined sensitivity). The combination of AGA and EMA on monkey esophagus resulted in a negative predictive value of 100% accuracy. CONCLUSIONS: A combination of AGA and EMA tests resulted in 100% sensitivity and 100% negative predictive value, useful in selecting patients for duodenal biopsy.