Gastric-juice ammonia assay for diagnosis of Helicobacter pylori infection and the relationship of ammonia concentration to gastritis severity

OBJECTIVES: To determine the test characteristics of gastric-juice ammonia concentration as measured by an ion-selective electrode and a rapid ammonia detection device for the diagnosis of Helicobacter pylori infection and to assess the relationship between gastric-juice ammonia concentration and the severity of gastritis. METHODS: Patients undergoing upper endoscopy had collection of gastric juice that was tested for ammonia using an ion-selective electrode and a rapid ammonia assay device that uses a pH-indicating membrane. A receiver operating characteristic curve was calculated for ammonia concentration. Severity of gastritis was graded using the Sydney classification (1) and correlated to gastric-juice ammonia concentration. Patients also underwent H. pylori testing by IgG serology, rapid urease testing, and histological special stain. Ammonia testing results were compared with a reference standard of two of three positive tests and with a second reference standard of a positive serology. RESULTS: 73 patients underwent endoscopy and collection of gastric juice. The receiver operating characteristic curve indicated an optimal cutoff value of 5 mM, yielding a sensitivity of 67%, specificity of 93%, positive predictive value of 67%, and negative predictive value of 93% (compared with the combined reference standard). The rapid NH3-testing device yielded a sensitivity of 83%, specificity 63%, positive predictive value 31%, and negative predictive value 95%. The severity of neutrophilic (p = 0.001) and mononuclear cell (p = 0.003) infiltration were significantly correlated with gastric-juice ammonia concentration. CONCLUSIONS: Measurement of gastric-juice ammonia concentration by ion-selective electrode or rapid detection device is a relatively insensitive and nonspecific means of H. pylori diagnosis. Gastritis severity increases with gastric-juice ammonia concentration.