The value of morphological analysis in duodenal ulcer therapy

Abstract

This study was designed to examine two premises: that the morphological "severity" of duodenal ulcers (DU) may influence the incidence of drug mediated healing and the morphological "quality" of healing after curative therapy may influence t he duration of remission. Biopsies taken at endoscopy from five healthy volunteers and from 84 patients suffering from DU were examined by light and electron microscopy. The endoscopic and morphological appearance of the mucosa within 8mm of the DU or scar, before and up to 1 year after therapy with either sucralfate, cimetidine, pirenzipine or misoprostol are described. Irrespective of the mode of therapy or whether the biopsies were from normal, juxta-DU or scar mucosa, specimens could be divided into 2 primary morphological classes: gastric metaplastic and non-metaplastic. Based on the degree of metaplastic differentiation and nonmetaplastic degeneration, these classes were further divided into 4 sub-classes. When correlated with the incidence of healing and duration of remission, metaplasia was generally found to be a positive and degenerative nonmetaplasia a negative prognostic criterion. Scores were awarded to primary morphological criteria and weighted to give high total s to favourable (metaplastic) and low totals to non-favourable (degenerative non-metaplastic) prognostic features. The sum of scores expressed as a percentage was termed the morphological index. This proved useful as a means of correlating mucosal morphology with DU healing and duration of remission. It also facilitated comparison of morphology within and between groups of patients before and after each drug regimen. The results showed that the morphological appearance of the ulcerative mucosa influenced healing and remission outcome. Discriminant analysis was applied to the numeric data that described the juxta-DU (group 1) and scar (group 2) morphology of patients treated with cimetidine in 2 studies. Separation between healed and not healed DU was achieved in 92% of group 1 and 100% (remission - more or less than 6 months) of group 2. When applied to the juxta-DU data from patients treated with cimetidine in a third study, the formulae predicted correctly in 88% of cases. In addition to predicting outcome, the formulae were used as standards to accommodate for natural variations in the prognosis of individual DU of patients enrolled for comparative drug studies. These data show that morphological analysis may be usefully employed in duodenal ulcer therapy.