Objectives: Current guidelines recommend either the urea breath test (UBT) or the Helicobacter pylori antigen stool test (HpSA) for monitoring H. pylori infection. The aim of this study was to evaluate the agreement between the two tests in patients after treatment.
Methods: After eradication treatments, patients were tested with both UBT and HpSA. Cut-off values (delta value over baseline at 30') for UBT were positive (> or = 5 per thousand), indeterminate (3.01-4.99 per thousand), and negative (< or = 3 per thousand). Cut-off values (absorbance at 450 nm) for HpSA test were positive (> or = 0.160), indeterminate (0.159-0.140), and negative (< 0.140). Patients with either discordant or indeterminate tests underwent repeat endoscopy with multiple gastric biopsies for rapid urease test (RUT), culture, histology, and immunohistochemistry to detect H. pylori and to assess the ratio between coccoid and bacillary forms.
Results: A total of 458 patients were studied. Of these, 422 (92.2%) had concordant tests, three (0.6%) indeterminate tests (one on UBT and two on HpSA), and 33 (7.2%) discordant tests. A total of 28 patients (25 with discordant and three with indeterminate tests) underwent endoscopy. The HpSA was inaccurate in 24 cases (18 false negative, four false positive, and two indeterminate results), whereas the UBT was inaccurate in four cases (two false positive, one false negative, and one indeterminate results). Biopsy-based tests showed no bacillary or coccoid forms in all five endoscoped patients who were negative on UBT and positive on HpSA, but in one in whom the ratio between coccoid and bacillary forms was 3:1 in the antrum and corpus.
Conclusions: UBT and HpSA test give discordant or indeterminate results in nearly 8% of patients after treatment. The HpSA test is less accurate than the UBT. Coccoid forms do not cause false positive HpSA results.