Clinical characteristics and temporal trends of meropenem and piperacillin/tazobactam heteroresistance in Pseudomonas aeruginosa isolates.

BACKGROUND

Heteroresistance (HR) to commonly used anti-pseudomonal agents in Pseudomonas aeruginosa has been reported in cross-sectional epidemiological surveys; however, longitudinal studies investigating the long-term epidemiological dynamics and clinical correlates of this adaptive phenotype are still lacking. This retrospective time-series study aims to systematically explore the HR phenotypic profiles and temporal dynamic of clinical P. aeruginosa isolates against piperacillin/tazobactam (TZP) and meropenem (MEM), and further clarify their potential clinical correlates.

METHODS

We comprehensively analyzed 380 non-duplicate clinical strains collected from a tertiary hospital between 2011 and 2024. The disk diffusion and Etest gradient diffusion methods were used for the preliminary determination of HR incidence rates in clinical isolates against six anti-pseudomonal antibiotics. Population analysis profiling (PAP) assays was further performed to assess the temporal dynamics and phenotypic characteristics of HR to TZP MEM. HR was defined as the presence of resistant subpopulations with an MIC at least eightfold higher than that of the dominant population at a frequency of ≥ 1 × 10⁻⁷. Subsequently, univariate and multivariate regression analyses were performed to identify clinical factors associated with HR and non-HR strains.

RESULTS

Preliminary screening indicated that 55% of the tested strains exhibited the HR phenotype, whereas confirmatory assays further verified that 49.5% of the isolates displayed HR to either TZP or MEM. The HR detection rate for MEM peaked at 35.3% during 2017-2019, followed by a gradual decline, reaching a minimum of 24.04% during the 2020-2022 pandemic. Meanwhile, the HR rate to TZP initially increased, then decreased, peaking at 43.1% during 2017-2019 and subsequently declining to 27.8%. In addition, multivariate regression analysis identified that male gender and prior antiviral exposure were independently associated with decreased risk of HR, whereas prior fluoroquinolone antibiotics exposure was independently associated with an increased risk.

CONCLUSION

Our findings highlight the widespread prevalence, distinct temporal dynamics, and diverse phenotypic features of HR, and underscore the potential clinical value of integrating HR screening into antimicrobial susceptibility testing, particularly for high-risk patient populations, to better guide therapy and mitigate the risk of treatment failure.