BACKGROUND AND AIMS To evaluate the validity of the biopsy-based tests (histology, culture, and urease test) and serology in detecting current Helicobacter pylori infection against a background of atrophic gastritis (AG) or intestinal metaplasia (IM).

METHODS Helicobacter pylori infection was diagnosed in 651 subjects, using the predefined gold standard for H. pylori tests. The sensitivity, specificity, and positive and negative predictive values of culture, CLOtest, histology (Giemsa stain), and serology were calculated with regard to the histological grade of AG and IM. The level of serum pepsinogen (PG) I and II was also measured as a marker for the presence of AG.

RESULTS In the study population (n = 651), sensitivity and specificity, respectively, were as follows: culture, 56.2 and 100%; histology, 93.0 and 94.0%; CLOtest, 80.4 and 96.7%; serology, 96.0 and 67.5%. If the analysis is limited to those without AG or IM (n = 158) or to those younger than 40 years (n = 69), all tests, except for culture, had a sensitivity and specificity >90%. The sensitivity of CLOtest and the specificity of serology markedly decreased with progression of AG and IM, and serology was less specific in the presence of AG, as determined by a PG I/II ratio < or =4.1 (specificity, 83.7% vs 40.7% in PG I/II >4.1 and < or =4.1, respectively).

CONCLUSIONS Any one of biopsy-based tests or serology was found to be excellent for identifying current H. pylori infection among individuals without AG or IM and/or younger patients (<40 years). However, a combination of at least two tests is necessary in the clinical setting of AG or IM.