Diagnostic yield of gastric biopsy specimens when screening for preneoplastic lesions.

ABSTRACT

The Sydney system recommends sites and numbers of stomach biopsies (mapping) for evaluation of Helicobacter pylori-associated lesions. The diagnostic yield of the recommended mapping technique in populations at high risk for gastric preneoplastic lesions has not been established. We evaluated pathology data from 733 endoscopies performed as part of an intervention study that assessed the effects of H. pylori treatment on preneoplastic conditions. Two pathologists assessed whether the mapping sequence of the 7 biopsy specimens obtained during each endoscopy was correctly followed and graded the specimens using the Sydney classification for gastritis. If the mapping sequence was followed, then we evaluated whether the amount of information obtained from 3 biopsy samples approximated that obtained from 5 and 7 biopsy samples. The mapping sequence was followed in only 239 (33%) endoscopies, indicating that experienced endoscopists can inadvertently misidentify sites in the stomach when obtaining specimens. When data from 7 specimens were used, H. pylori was found in 205 endoscopies, atrophy in 152, metaplasia in 135, and dysplasia in 22. When data from 3 specimens were used, the sensitivity was 99% for presence of H. pylori, 82% for atrophy and metaplasia, and 81% for dysplasia. When data from 5 specimens were used, the sensitivity was 100% for H. pylori, 96% for atrophy, and 95% for metaplasia and dysplasia. Although site-specific biopsy mapping is difficult in practice, the recommendations of the Sydney system as to the location and number of gastric biopsy specimens can adequately identify significant gastric histopathology.