Utility of ancillary stains for Helicobacter pylori in near-normal gastric biopsies.

Documentation of Helicobacter pylori infection and eradication is important, prompting some clinicians and pathologists to request ancillary stains on all gastric samples that do not demonstrate H. pylori on initial histologic review. Studies evaluating the utility of ancillary stains in patients with minimal inflammation are lacking. We used Giemsa, Warthin-Starry, acridine orange, and immunohistochemical stains to search for organisms in 56 patients with biochemical evidence of H. pylori infection (positive Campylobacter-like organism test) and gastric mucosal samples interpreted to be H pylori negative by hematoxylin and eosin (H&E). We correlated the findings with severity of inflammation and patients' histories of medication use. Nineteen (34%) patients had histologically normal mucosae, 22 (39%) had chronic inflammation with or without focal activity, and 15 (27%) had chemical gastropathy. Fifty (89%) cases were negative for H. pylori with additional stains, and 6 contained bacteria that were detected with all 4 ancillary stains and on retrospective review of H&E-stained sections that also showed chronic inflammation. Eleven (20%) patients were taking proton pump inhibitors, and 4 (7%) had previously received H. pylori eradication therapy. We conclude that H&E stains demonstrate H. pylori in most infected patients, so preemptive stain requests are largely unnecessary. Failure to identify bacteria by H&E evaluation generally reflects their absence in biopsy material, even among Campylobacter-like organism test--positive patients. However, organisms may be overlooked in patients with mild inflammation and in those receiving proton pump inhibitor or antibiotic therapy, so one should consider ordering ancillary stains to enhance detection of bacteria in these settings.