Assessment of antral G-cell hyperplasia on routine histologic sections

Proton pump inhibitor (PPI)-associated gastric neuroendocrine tumors (NETs) are recognized as a distinct clinicopathologic category in the latest World Health Organization classification, although the relationship between chronic antisecretory therapy and gastric NET development remains incompletely understood. We evaluated antral G-cell hyperplasia as a potential tissue-level correlate of gastrin-axis activation. Archived gastric biopsies obtained over a 15-month period were reviewed, including 113 patients with gastroesophageal reflux disease/Barrett's esophagus (GERD/BE) receiving documented PPI/H2-blocker therapy and 21 controls without documented exposure. Antral G-cell hyperplasia was scored (0-2) on H&E-stained sections, and gastrin immunohistochemistry (IHC) was scored using a 3-tier scheme (0-2). Interobserver agreement was assessed using Fleiss' kappa. To address age differences, an additional 101-month search identified 9 age-matched controls. High antral H&E scores (1-2) were more frequent in the GERD/BE group than in controls (66.4% vs 28.6%; p = 0.001). Interobserver agreement improved with binary scoring (κf = 0.694; 95% CI, 0.591-0.790). Gastrin IHC scores 2 were more frequent in cases with high H&E scores (p < 0.001), and H&E and gastrin IHC scores showed a moderate positive correlation (Spearman's ρ = 0.473, p < 0.001). H&E scores 1-2 demonstrated greater discrimination for documented PPI/H2 exposure than gastrin IHC score 2 (AUC 0.689 vs 0.576). No enterochromaffin-like cell hyperplasia or NET was apparent in available body/fundus biopsies on H&E review alone. The age-matched control group showed a trend toward lower H&E scores (p = 0.070). Antral G-cell hyperplasia can be assessed using a simple binary H&E scoring scheme with acceptable interobserver reproducibility, and may represent a useful research tool for future studies.