[Breath test in the diagnosis of Helicobacter pylori infection: concordance with histological methods and correlation with anatomopathological lesions of the gastric mucosa]

Abstract

Purpose: To study the concordance between 13C-urea breath test and histology in the diagnosis of Helicobacter pylori infection, and to evaluate whether there is a correlation between breath test values and histologic lesions of the gastric mucosa.

Methods: Sixty-nine patients with duodenal ulcer were prospectively studied. An endoscopy with biopsy samples (H&E stain) taken from the antrum and body was performed, and a 13C-urea breath test (measuring 13C difference: delta 13CO2) was also done. Both procedures were repeated one month after completing therapy ["classic" triple therapy (n = 28), and omeprazole+amoxycillin (n = 41)]. Eradication was defined as the absence of H. pylori both by histological and breath test methods.

Results: At the beginning of the study, 94.2% of patients (n = 65) were H. pylori positive by histological methods, and 98.6% (n = 68) were positive by the breath test (ratio of positive agreement = 0.96). Kappa for H. pylori diagnosis after therapy was 0.83 (95% CI: 0.69-0.96). A correlation between delta 13CO2 and histologic lesions was observed, both in the antrum at before treatment (Spearman coef = 0.34; p = 0.007) and after therapy in both the gastric antrum (0.63; p < 0.001) and body (0.35; p = 0.005). A significant difference was observed when comparing mean delta 13CO2 in patients with different degrees of histologic gastritis, both before treatment (antrum: W Kruskal-Wallis = 6; p < 0.05) and after therapy (antrum: W = 17; p < 0.001; body: W = 10; p < 0.05).

Conclusion: A high concordance was observed between the 13C-urea breath test and histology in the diagnosis of H. pylori infection. A correlation exists between breath test values and histologic lesions of the gastric mucosa.