Background: Helicobacter pylori eradication is recommended in post-gastric cancer resection, but premalignant changes may prevent the detection of H. pylori. The aim of this study was to determine appropriate biopsy site for detecting H. pylori in gastric cancer patients.
Materials and methods: Consecutive patients (194) with gastric adenocarcinoma were prospectively enrolled. Helicobacter pylori was evaluated by serology, histology and rapid urease test. Biopsy sites included antrum lesser curvature, upper body lesser curvature (UBLC) and upper body greater curvature (UBGC). Two biopsy specimens were obtained from each site for histological examination. One additional specimen was obtained from UBGC for the rapid urease test.
Results: The overall infection rate of H. pylori was 84.0% (95% CI 78.9-89.2). The sensitivity of histology for detecting H. pylori at various sites was: antrum (54.9%; 95% CI 45.7-63.9), UBLC (80.3%; 95% CI 72.2-87.0) and UBGC (95.1%; 95% CI 89.6-98.2). Specificities of all three biopsy sites were more than 95%. Sensitivity and specificity of the rapid urease test performed at UBGC were 96% and 100%, respectively. Sensitivities of histology decreased in correlation with increasing severity of atrophy and intestinal metaplasia (both P < 0.001 using the chi-square test for trend). The proportions of moderate to marked atrophy/intestinal metaplasia at UBGC (12.8%/14.7%) were significantly lower than those at antrum (50.0%/57.8%, P < 0.001 respectively) or UBLC (40.0%/48.9%, P < 0.001 respectively).
Conclusions: The UBGC side is the most sensitive and specific biopsy site to detect H. pylori in gastric cancer patients due to less frequent atrophy and intestinal metaplasia than at the antrum or UBLC side.