Liu He, Ting Liu, Jie Xiang, Mingzhe Yan, Jun Wan, Xiaowei Peng, Yang Han, Congrui Xu
arr was prevalent in this cohort and was concentrated in several ST-defined genomic backgrounds. Documented RFP exposure was associated with carriage in unadjusted analysis but was not confirmed as an independent correlate after adjustment. Prospective studies with complete treatment histories and phenotypic susceptibility data are needed to define the contribution of rifamycin exposure.
BACKGROUND: Rifampicin (RFP) is central to the management of tuberculosis (TB) and nontuberculous mycobacterial (NTM) infections, and specialized infectious disease hospitals therefore use it routinely. Whether rifamycin resistance determinants are enriched among co-resident opportunistic pathogens in this setting has rarely been examined. We focused on the arr determinant in Klebsiella pneumoniae, for which prevalence, clonal context, and patient-level association with documented RFP exposure remain poorly defined.
METHODS: We retrospectively analyzed 165 patient-deduplicated clinical Klebsiella pneumoniae isolates collected during 2024 from a provincial specialized infectious disease hospital. All isolates underwent whole-genome sequencing; arr was identified using Kleborate v2 and cross-validated against RGI/CARD. Phenotypic RFP MIC data were not available for these Klebsiella pneumoniae isolates. Individual-level RFP exposure was retrieved from the hospital information system. The RFP-arr association was tested across nested logistic regression models with bootstrap-based confidence intervals, multiple-comparison correction, and exposure-misclassification sensitivity analyses.
RESULTS: arr was carried by 41 of 165 isolates (24.8%), with arr-3 accounting for 40 of 41 positive cases (97.6%). Among 164 isolates with definable exposure, arr positivity was higher in RFP-exposed than in unexposed patients (40.0% vs 20.2%; OR 2.64, 95% CI 1.18-5.89, P = 0.025). After covariate adjustment, all bootstrap 95% CIs included the null. Carriage was strongly concentrated in ST37, ST60, ST147, and ST656 (combined OR 22.67, 95% CI 7.66-67.03, P < 0.001). arr-positive isolates showed a broader genotypic resistance burden and lower predicted virulence scores than arr-negative isolates.
CONCLUSION: arr was prevalent in this cohort and was concentrated in several ST-defined genomic backgrounds. Documented RFP exposure was associated with carriage in unadjusted analysis but was not confirmed as an independent correlate after adjustment. Prospective studies with complete treatment histories and phenotypic susceptibility data are needed to define the contribution of rifamycin exposure.