Prevalence of Campylobacter pylori as demonstrated by histology or CLO-test in different types of gastritis. A study in 5 clinically predefined groups of patients.

The prevalence of Campylobacter pylori infection as detected by histology was studied in 5 predefined groups of patients. The associated histologic and endoscopic findings were registered. Validity of CLO-test was tested against the histologic detection. The following groups of patients were studied: A) Non-ulcer dyspepsia (defined by one or all of three symptoms: heartburn, nausea/inappetence, halitosis/belching) B) control group (no specific symptoms, no ulcer, no history of gastric surgery) C) Duodenal ulcer D) Gastric ulcer E) Billroth I or II resection of the stomach. 200 patients were recruited for group A-C, in group D 134 patients and in group E 113 patients were studied. A mean prevalence of 60% was observed. Prevalence was highest in patients with duodenal ulcer (86%). In group D a prevalence of 65%, in A and E a prevalence of 54%, and in B of 40% were seen. The overall test sensitivity of the CLO-test compared against the histologic detection rate was 75%, the specificity 81%. Sensitivity was reduced in group A (69%) and E (53%) and in patients with inactive chronic gastritis (67%). In all groups patients with active forms of gastritis showed the highest prevalence of C. pylori infection. The specificity of the CLO-test was reduced in patients with duodenal ulcer (46%) and gastric ulcer (48%). Decreased specificity observed after therapy with histamin receptor (H2) blockers may explain this finding. The relationship of C. pylori infection with active types of gastritis or gastro-duodenal ulcer hints at a causal relation but is no definite proof of its etiologic role. The validity of the CLO-test seems questionable in patients with gastroduodenal ulcer or operated stomach.