Rapid quantitative assessment of gastric corpus atrophy in tissue sections

N. C.T. van Grieken, M. M. Weiss, G. A. Meijer, E. Bloemena, J. Lindeman, G. J.A. Offerhaus, S. G.M. Meuwissen, J. P.A. Baak, E. J. Kuipers*

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

Background/Aims - Grading of Helicobacter pylori induced atrophic gastritis using the updated Sydney system is severely limited by high interobserver variability. The aim of this study was to set up a quantitative test of gastric corpus mucosal atrophy in tissue sections and test its reproducibility and correlation with the Sydney scores of atrophy. Method - Mucosal atrophy was assessed in 124 haematoxylin and eosin stained corpus biopsy specimens by two experienced gastrointestinal pathologists (EB, JL) according to the updated Sydney system as none (n = 33), mild (n = 33), moderate (n = 33), or pronounced (n = 25). In each specimen, the proportions of glands, stroma, infiltrate, and intestinal metaplasia in the glandular zone were measured as volume percentages using a point counting method. The optimal point sample size, intra-observer and interobserver reproducibility, discriminative power for degrees of atrophy, and correlations with H pylori status were evaluated. Results - Counting 400 points in 200 fields of vision provided the smallest sample size that still had excellent intra-observer and interobserver reproducibility (r ≥ 0.96). Overall, the volume percentage of glands (VPGL), infiltrate (VPI), and stroma (VPS) correlated well with the Sydney scores for atrophy (p ≤ 0.003). However, no differences were found between non-atrophic mucosa and mild atrophy. No correlation was found between age and either the Sydney grade of atrophy or the VPGL or VPS. In non-atrophic mucosa and mild atrophy, H pylori positive cases showed a significantly higher VPI than did H pylori negative cases. A lower VPGL was seen in H pylori positive cases than in H pylori negative cases in the mild atrophy group. VPS did not correlate with H pylori status within each grade of atrophy. Conclusion - Point counting is a powerful and reproducible tool for the quantitative analysis of mucosal corpus atrophy in tissue sections. These data favour the combination of "none" and "mild" atrophy into one category, resulting in a three class grading system for corpus atrophy, when using the updated Sydney system.

Original languageEnglish
Pages (from-to)63-69
Number of pages7
JournalJournal of Clinical Pathology
Volume54
Issue number1
DOIs
Publication statusPublished - 2001

Cite this