P30 A longitudinal perspective of antimicrobial resistance evolution within urinary tract infection-based clinical trials

Abstract

Abstract Background The mainstay treatments for urinary tract infections (UTIs) are nitrofurantoin and trimethoprim. The common route to acquire trimethoprim resistance (TriR) is via horizontal transfer of the trimethoprim-insensitive dihydrofolate reductase: dfrA. In contrast, nitrofurantoin resistance (NitR) requires the inactivation of two chromosomally encoded nitroreductases: nfsA and nfsB. The difference in antimicrobial resistance (AMR) evolution is reflected in community surveillance data: NitR incidence = ∼10% and for TriR > 30%. Longitudinal clinical UTI trials provide a unique opportunity to monitor in situ evolution of AMR and correlate its appearance to trial outcomes and the impact on uropathogens. Objectives To explore the in situ AMR evolution and its impact in Escherichia coli isolates from patients participating in the clinical trials: AnTIC and ALTAR. AnTIC was an open label randomized trial assessing the efficacy of antibiotic prophylaxis use in clean intermittent self-catheterizing patients. ALTAR compared the efficacy of the non-antibiotic alternative methenamine hippurate to antibiotic prophylaxis to treat recurrent UTIs. Methods The investigation of the evolution of nitrofurantoin and trimethoprim resistance in E. coli used general microbiology techniques, bioinformatic genome analysis and genetics to model known AMR mutations in susceptible E. coli strains. Results Trimethoprim resistance amongst E. coli trial isolates was driven by 14 out of ∼45 known dfrA allelic variants; the most common being dfrA1 and dfrA17. Growth analysis identified an allelic bias when strains were exposed to sub-MIC concentrations of trimethoprim. Growth rate analysis identified a 2%–10% slower doubling time for nitrofurantoin-resistant strains: NitS: 20.8 ± 0.7 min compared to NitR: 23 ± 0.8 min. Statistically, these data suggested no fitness advantage of evolved strains compared to the sensitive predecessor (P value = 0.13). Genetic manipulation of E. coli to mimic NitR evolution, however, supported a selective advantage. Conclusions Longitudinal sampling during antibiotic based clinical trials has provided a new perspective to the response of E. coli to UTI-related antibiotics. Correlation of nitrofurantoin response to pharmacokinetic data suggests that the low incidence of E. coli NitR is driven by selection not fitness. However, fitness correlates to dfrA allele carriage and could potentially be exploited to benefit the use of trimethoprim if other antibiotics are impeded by AMR.