Histological examination of ulcer margin for diagnosing Helicobacter pylori infection in patients with gastric ulcers.

Biopsy of ulcer margin is routinely performed to exclude malignancy in patients with gastric ulcers, but its utility in diagnosing Helicobacter pylori infection has not yet been fully studied. A cohort of 50 patients with gastric ulcer was prospectively examined. Three tests including histology, rapid urease test, and urea breath test were performed in all patients for diagnosing H pylori infection. Six biopsied specimens from the margin of the gastric ulcer and 1 each specimen from antrum and body of non-ulcer part were obtained for histology using hematoxylin-eosin (H&E) stain. The criterion used for defining H pylori infection was a positive result in at least 2 of the 3 tests. H pylori infection was diagnosed in 27 (54%) of the patients. The diagnostic sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of the histological examination of the ulcer margin were 92.6%, 95.7%, 96.2%, 91.7%, and 94%, respectively. The addition of 1 specimen from the antrum or body or a combination of the 2 specimens did not increase the diagnostic yields of those for histological examination of ulcer margin alone. The diagnostic sensitivity, specificity, positive predictive value, negative predictive value, and accuracy for the rapid urease test were 96.3%, 100%, 100%, 95.8%, and 98%, respectively, and the corresponding values for the urea breath test were 88.9%, 87%, 88.9%, 87%, and 88%. We performed Giemsa stain for the 3 patients with false-negative and false-positive results of histological examination of ulcer margin using H&E stain, and all were positive for H pylori infection. In conclusion, histological examination of the ulcer margin using hematoxylin-eosin stain was quite accurate and useful for diagnosing H pylori infection in patients with gastric ulcers. A special stain is required when the diagnosis of H pylori infection is questionable on routine H&E staining.