Source of variation in the diagnosis of Helicobacter pylori-associated multifocal atrophic gastritis.

Abstract

Introduction: Multiple sampling from different sites of the stomach as well as the number of fragments of gastric mucosa available for histopathologic evaluation are important sources of variation when classifying and grading chronic gastritis.
Objective: To estimate the sensitivity of the number of fragments of gastric mucosa necessary to establish the diagnosis of atrophic gastritis with intestinal metaplasia, gastric dysplasia and H. pylori infection. In addition, this study will attempt to assess the intra-observer variability in the classification of these premalignant gastric lesions.
Methods: This is a 6 year-cohort study, wherein 1958 gastric endoscopic procedures performed by two gastroenterologists were reviewed. Five gastric biopsy samples were obtained from the antrum, body and lesser curvature during each procedure. One pathologist was in charge of reviewing the five histopathology samples for each subject and providing a definitive diagnosis which was used as the gold standard. Each gastric mucosa sample reviewed led to an individual diagnosis for that sample which was compared with the gold standard. Intra-observer variability was assessed in 127 individuals who correspond to a random sample of 20% of the total endoscopic procedures performed during the 72 month-follow-up.
Results: The sensitivity of the diagnosis of intestinal metaplasia (IM) and gastric dysplasia increased proportionally with the number of gastric mucosa samples reviewed. The lesser curvature of the stomach had the highest sensitivity for the diagnosis of IM and dysplasia, among all the stomach regions studied. Just one sample of gastric mucosa attained a sensitivity of 95.9% for the detection of H. pylori infection. The intra-observer agreement for the diagnosis of multifocal atrophic gastritis was 86.1% and the kappa value was 0.79 (95% CI 0.76-0.85). Alcohol-fixed biopsy specimens were inadequate to diagnose H. pylori infection and to assess dysplasia.
Conclusion: The number of mucosa gastric fragments reviewed, the fixation method used, and the biopsy site are all important factors in order to ensure a correct classification of chronic gastritis.