Background and Aims-It is unknown whether Helicobacter pyloni infection activates complement in vivo. Mucosal deposition of various activation products of the complement system may contribute to the pathogenesis of chronic gastritis and was therefore studied by immunohistochemistry. Patients and Methods-Ethanol fixed antrum or body gastric tissue sections from 24 patients infected with H pylori (determined by bacterial immunohistochemistry) and 22 uninfected patients were examined by immunofluorescence with monoclonal antibodies to activation neoepitopes in C3b and in the terminal complex (TCC). As a control group, biopsy samples from the gastric stump of 23 Billroth II operated patients were studied.Results-Patchy, bright staining for TCC occurred below the surface epithelium and around the glands in Hpylori positive and negative gastritis as weli as in stump gastritis but seldom in normal mucosa. Activated C3 was present at the apical face of the surface epithelium, significantly more often in the antrum and body from patients with than without H pyloni infection (p=0*05 and p=003 respectively), and particularly in samples with granulocyte infiltration (p=0 04). Many bacteria were coated with activated C3 towards the pit openings but seldom within the foveolae. Conclusions-Local complement activation was shown to take place in simple chronic gastritis, associated as weli as unassociated with H pylori infection, and also in stump gastritis. The fact that activated C3 was seldom seen on H pyloni within the foveolae, suggested that the bacterium evades complement attack in this location.

SUMMARY Immunoglobulin (Ig)-producing immunocytes were quantified by paired immunofluorescence staining in specimens of gastric antral (n=52) and body (n=117) mucosa obtained from 45 patients with various gastrointestinal disorders. Enumerations were carried out in a 500 gm wide zone from the muscularis mucosae to the lumen ('tissue unit'). The specimens were divided into three categories according to the degree of inflammation, and each specimen received a grade for atrophy (0-2). The total number of IgA-, IgM-and IgG-producing cells per tissue unit increased strikingly with increasing degree of inflammation, both in antral and body mucosa. IgA immunocytes predominated (61-91%) in all specimens, but the IgG isotype showed the largest relative increase (four to 17-fold), particularly in the basal part of the mucosa. In this layer of the gastric body the proportion of IgG cells was also significantly raised in association with atrophy, irrespective of degree of inflammation. Locally produced IgG may be of protective significance in terms of internal (or 'second line') defence but may at the same time maintain immunopathological mechanisms contributing to the chronicity of gastritis. It has recently been shown that epithelial expression of secretory cornonent and uptake of IgA are enhanced in gastritis. 5 The associated cellular IgA response in the lamina propria, however, related to the degree of inflammation has not been previously quantified. Also, the relative participation of immunocytes of other isotypes (IgM, IgG, IgD, and IgE) in gastritis has not been conclusively

The mucosal distribution of G cells was quantitatively mapped in resected stomachs from 42 patients (12 with gastric ulcer, 11 with duodenal ulcer, 14 with duodenal ulcer and uremia, and 5 with gastric cancer). Along the histological border of the proximal part of the pyloric antrum there was in all patient categories a transitional zone of varying extent, with a low G-cell density before the cells disappeared in the body of the stomach. The proximal end of the duodenum contained considerably fewer G cells than in the antrum, and the number was virtually equal in all groups. Within the antrum there was in the material as a whole a gradual increase in G-cell density from the proximal to the distal part, but this difference was not apparent for the gastric ulcer patients. When corresponding antral segments were compared between the various patient groups, the G-cell density was found to be significantly decreased in the distal antrum of the gastric ulcer patients. In all patient categories, except the duodenal ulcer group with uremia, the circumferential distribution of G cells showed reduced density along the curvatura minor. For the material as a whole there were great individual variations in the overall antral G-cell density, in the antral area corresponding to the distribution of G cells and in the total G-cell mass; these three variables were not significantly related to diagnosis, age or sex.

One hundred and seven consecutive patients with hematemesis and/or melena and a diagnosis of duodenal, gastric, or esophageal ulcers were interviewed immediately before or after endoscopy about the use of non-steroid anti-inflammatory drugs (NSAIDs) and symptoms before the hemorrhage. If the patients admitted no symptoms of abdominal pain or discomfort, nausea, vomiting, or heartburn, they were classified as having no ulcer symptoms before the hemorrhage. Patients who had not taken NSAIDs during the last 48 h before the hemorrhage were classified as not having taken NSAIDs. Significantly fewer patients had ulcer symptoms in the group that had used NSAIDs than in the other group (p less than 0.01). This may be interpreted as a possible masking effect by NSAIDs on ulcer symptoms. Physicians and patients should be aware of this possible effect of NSAIDs.