Search

The charts of 54 children diagnosed with antral H. pylori were reviewed, to establish the incidence of gastroduodenal inflammation and compare therapeutic efficacies of antisecretory vs. antibacterial therapy. Histology demonstrated normal mucosa in three cases (6%) and gastric/duodenal inflammation (> or = Whitehead grade 3) in 51 biopsies (94%). 23/43 children (53%) initially responded to H2-blockers; however, by 10 mo, 13 had relapsed clinically. All of these patients subsequently responded to amoxicillin plus bismuth subsalicylate. Of the 20 children who failed to enter remission after an initial course of H2-blockers, all became symptom-free after treatment with amoxicillin/bismuth. Compared to antisecretory agents, antibacterial treatment induced clinical remission in 11/11 patients (p < 0.001), who remained symptom-free for 10 +/- 0.2 mo. Clinical remissions were maintained in significantly more patients following amoxicillin/bismuth vs. H2-blockers (44/54 vs. 10/43 courses, p < 0.001); and, the cumulative probability of remaining asymptomatic was significantly greater in the antibiotic group (p < 0.001). These data suggest that gastric colonization by H. pylori is highly predictive of mucosal pathology in children. Initial therapy should be directed toward achieving bacterial eradication, as opposed to gastric acid suppression.

To evaluate the accuracy of IgG and IgA serological tests in establishing a diagnosis of Helicobacter (Campylobacter) pylori gastric infection, 60 children presenting with chronic abdominal pain were prospectively studied. Endoscopic antral biopsies were obtained and analyzed for the presence of H. pylori using three standard methods: culture and identification of bacterial isolates, microscopic examination for morphologically characteristic bacteria, and urease production by the biopsy specimen. Concomitantly obtained serum samples were analyzed for the presence of IgG and IgA antibodies against H. pylori surface antigens using enzyme-linked immunosorbent assay (ELISA). Thirty-four of 60 (56.6%) had histological evidence of chronic active gastritis, eight of whom (13.3%) also had evidence of H. pylori infection by at least one criteria. Six of the eight infected patients had H. pylori demonstrated by all three methods. Of the eight infected patients, seven had IgG antibodies against H. pylori (sensitivity of 87%) and six had IgA antibodies (sensitivity of 75%). Among the six patients who had H. pylori infection confirmed by all three methods, all had IgG antibodies (sensitivity of 100%). In the patients without evidence of H. pylori infection, the IgG ELISA had a specificity of 96% (50/52), and the IgA ELISA had a specificity of 100% (52/52). Our data suggest that serological testing for the presence of antibodies against H. pylori may be a useful diagnostic tool in screening children with chronic abdominal pain for the presence of gastric infection with H. pylori.