RESUMO
BACKGROUND AND AIM: To determine the application range of diagnostic kits utilizing anti-Helicobacter pylori antibody, we tested a newly developed latex aggregation turbidity assay (latex) and a conventional enzyme-linked immunosorbent assay (E-plate), both containing Japanese H. pylori protein lysates as antigens, using sera from seven Asian countries. METHODS: Serum samples (1797) were obtained, and standard H. pylori infection status and atrophy status were determined by culture and histology (immunohistochemistry) using gastric biopsy samples from the same individuals. The two tests (enzyme-linked immunosorbent assay and latex) were applied, and receiver operating characteristics analysis was performed. RESULTS: Area under the curve (AUC) from the receiver operating characteristic of E-plate and latex curves were almost the same and the highest in Vietnam. The latex AUC was slightly lower than the E-plate AUC in other countries, and the difference became statistically significant in Myanmar and then Bangladesh as the lowest. To consider past infection cases, atrophy was additionally evaluated. Most of the AUCs decreased using this atrophy-evaluated status; however, the difference between the two kits was not significant in each country, but the latex AUC was better using all samples. Practical cut-off values were 3.0 U/mL in the E-test and 3.5 U/mL in the latex test, to avoid missing gastric cancer patients to the greatest extent possible. CONCLUSIONS: The kits were applicable in all countries, but new kits using regional H. pylori strains are recommended for Myanmar and Bangladesh. Use of a cut-off value lower than the best cut-off value is essential for screening gastric cancer patients.