A morphometric study of antral G-cell density in a sample of adult general population: comparison of three different methods and correlation with patient demography, helicobacter pylori infection, histomorphology and circulating gastrin levels.

Helicobacter pylori infection has been linked to hypergastrinemia and either decreased or normal G-cell content in the antral mucosa. To clarify this controversial issue, we quantitatively determined antral G-cell content on the same biopsy specimens with three different methods and examined whether these methods are intercorrelated and the relation of these methods to plasma gastrin concentrations, demography, the occurrence of H. pylori infection and chronic gastritis. Gastric antral mucosal biopsy sections from 273 adults (188 with and 85 without H pylori infection) from a general population sample were examined immunohistochemically for G-cells using cell counting, stereology (point counting) and computerized image analysis. Gastritis was scored according to the updated Sydney system. Basal plasma gastrin concentrations were measured by radioimmunoassay. The three methods for G-cell quantification were poorly correlated and the results showed no correlation with basal plasma gastrin concentrations. The antral G-cell density and scores for H. pylori colonization were positively related to age. Neither the scores for chronic inflammation, nor the scores for inflammatory activity, atrophy or intestinal metaplasia were consistently related to the antral G-cell content. In conclusion, the results of three techniques for G-cell quantification in the gastric antral mucosa were poorly intercorrelated and none of the methods correlated with plasma gastrin concentrations. Age and scores for H pylori colonization seem to be determinants of the G-cell density. That common morphometric techniques correlate poorly is of utmost importance to bear in mind when quantitative morphological studies are planned, compared or interpreted.