Background/aims: To establish rapid urease test utility for initial diagnosis of Helicobacter pylori infection in gastric ulcer patients and to determine the best site for sampling for gastric biopsies.
Methodology: Seventy consecutive gastric ulcer patients were prospectively studied. All these patients underwent three biopsies from both antrum and body (two for hematoxylin-eosin staining and one for rapid urease test -Jatrox H. p. Test-). Likewise, IgG ELISA serology and 13C-urea breath test were carried out. Gold standard for H. pylori infection was defined as two or more tests (i.e., histology, serology, breath test) with positive results.
Results: Rapid urease test yielded 96.8% sensitivity (95% CI = 89-99%) and 100% (66-100%) specificity when using biopsy specimens from the body, with identical results when biopsy specimens from both antrum and body were considered together. However, when only biopsy specimens from the antrum were used, sensitivity dropped to 72.6% (60-82%) and specificity was 100% (66-100%). As far as concordance between rapid urease test and histology is concerned, we found a "proportion of positive agreement" of 0.78 for the antrum, with 0.46 kappa statistic (P < 0.0001) and 15 McNemar statistic (P < 0.0001). For the gastric body, "proportion of positive agreement" was 0.98, with 0.94 kappa statistic (P < 0.0001) and 1 McNemar statistic (P = 0.3). Larger (P < 0.01) prevalence of both glandular atrophy (17.8%, 11-28%) and intestinal metaplasia (68.5%, 57-78%) was observed in the antrum in comparison with that in the body (4.1%, 1-11%; and 16.4%, 10-26%, respectively).
Conclusions: Biopsy specimens from the body should always be obtained when the rapid urease test is performed to diagnose H. pylori infection in gastric ulcer, since this procedure is less accurate when biopsy specimens from the antrum are used, probably due to larger prevalence of both glandular atrophy and intestinal metaplasia in the latter site. Likewise, it seems that rapid urease test from body biopsies is sufficient to reach a reliable infection diagnosis in gastric ulcer patients as this procedure performed with antrum biopsies fails to improve its overall results.