Morphometric assessment of gastric antral atrophy: comparison with visual evaluation

Aims: As part of a multinational effort to reach a consensus in the definit ion and evaluation of atrophic gastritis, we applied morphometric technique s to 22 antral biopsy specimens examined visually by 12 experienced gastroi ntestinal pathologists. Methods and results: Atrophy was defined as loss of glands. Each pathologis t graded atrophy with both non-standardized and standardized approaches. Di scriminant function analyses of morphometric measurements were conducted to validate and grade atrophy. Kappa statistics were used to compare the perf ormance of each pathologist against the group mode and against the discrimi nant functions' grading of atrophy. Three morphometric indexes showed signi ficant differences among categories of atrophy utilizing non-standardized a s well as standardized visual atrophy grades: (i) the ratio of glandular le ngth to total mucosal thickness; (ii) the proportion of the secretory compa rtment area occupied by glands; and (iii) the number of glandular cross sec tions per 40x microscopic field. The discriminant function analyses verifie d all cases classified visually as either non-atrophic, or moderately/sever ely atrophic; it verified as mildly atrophic 40% of the cases classified vi sually as mildly atrophic; and classified the remaining 60% as moderately o r severely atrophic. The kappa statistics were good or excellent for the ma jority of pathologists. Conclusions: The evaluation of antral atrophy, simply defined as loss of gl ands, can be reliable and reproducible. The visual grading of atrophy as ab sent, moderate and severe is entirely consistent with objective morphometri c observations.