Phenotypic and genotypic analysis of Helicobacter pylori resistance patterns, molecular characteristics, and CYP2C19 polymorphisms in Nanjing: a retrospective cross-sectional study

Background The continued rise in antibiotic resistance of Helicobacter pylori ( H. pylori ) poses a major challenge to clinical eradication therapy. The resistance patterns and phenotype-genotype concordance of H. pylori in the Nanjing region have not been systematically reported in recent years. Methods Data from 7227 enrolled patients at Nanjing First Hospital (2018–2024) were retrospectively analyzed. Among these, 5729 patients were culture-positive for H. pylori and underwent phenotypic antimicrobial susceptibility testing for six antibiotics. Genotypic testing for resistance-associated mutations was performed on a subset of these patients. The remaining 1498 patients, although H. pylori -positive by other diagnostic methods, were culture-negative and were excluded from resistance analyses. Results Metronidazole showed the highest resistance rate at 84.45% (4838/5729, 95% CI: 83.50–85.37), followed by clarithromycin at 50.01% (2865/5729, 95% CI: 48.71–51.31) and levofloxacin at 38.59% (2211/5729, 95% CI: 37.33–39.87). Resistance rates to amoxicillin at 0.51% (29/5729, 95% CI: 0.34–0.73), furazolidone at 0.19% (11/5729, 95% CI: 0.10–0.34), and tetracycline at 0.07% (4/5729, 95% CI: 0.02–0.18) were very low. Annual trend analysis showed that the metronidazole resistance rate remained above 92% since 2020, reaching 98.68% in 2024. Clarithromycin resistance increased from 35.80% to 60.14% and then declined slightly, while levofloxacin resistance rose from 20.99% to 44.59% before reaching a plateau. Genotypic testing for clarithromycin and levofloxacin showed excellent concordance with phenotypic susceptibility (Kappa values: 0.9046 and 0.9187, respectively; both P<0.001). Resistance mutations were highly concentrated in the 23S rRNA gene at the A2143G locus (98.93%) for clarithromycin and in the gyrA gene at the N87K locus (52.54%) for levofloxacin. Kappa values for amoxicillin, tetracycline, furazolidone, and metronidazole were all below 0.2 and not statistically significant. CYP2C19 genotyping revealed that 43.89% of patients were extensive metabolizers, 44.26% intermediate metabolizers, 11.67% poor metabolizers, and 0.18% ultrametabolizers. Conclusion Metronidazole, clarithromycin, and levofloxacin are no longer suitable as empirical first-line agents in Nanjing. Amoxicillin, furazolidone, and tetracycline remain reliable options. Genotypic testing for clarithromycin and levofloxacin resistance is highly concordant with phenotypic results and suitable for guiding individualized therapy. CYP2C19 genotyping supports personalized acid suppression strategies. Regimens containing amoxicillin, furazolidone, or tetracycline should be prioritized, with resistance gene testing recommended before using clarithromycin or levofloxacin.