Phylogenetic context of antibiotic resistance provides insights into the dynamics of resistance emergence and spread

Abstract

Background To ameliorate the antibiotic resistance crisis, the drivers of resistance emergence and resistance spread must be better understood. Methods Whole-genome sequencing and susceptibility testing were performed on clinical carbapenem-resistant Klebsiella pneumoniae isolates collected from August 2014 to July 2015 across 12 long-term acute care hospitals. Ancestral state reconstruction partitioned patients with resistant strains into those that likely acquired resistance via de novo evolution or cross-transmission. Logistic regression was used to evaluate the associations between patient characteristics/exposures and these two pathways: resistance due to predicted within-host emergence of resistance, and resistance due to predicted cross-transmission. This framework is available in the user-friendly R package, phyloAMR (https://github.com/kylegontjes/phyloAMR). Results Phylogenetic analysis of 386 epidemic lineage carbapenem-resistant Klebsiella pneumoniae sequence type 258 isolates revealed differences in the relative contribution of de novo evolution and cross-transmission to the burden of resistance to five antibiotics. Clade-specific variations in rates of resistance emergence and their frequency and magnitude of spread were detected for each antibiotic. Phylogenetically-informed regression modeling identified distinct clinical risk factors associated with each pathway. Exposure to the cognate antibiotic was an independent risk factor for resistance emergence (trimethoprim-sulfamethoxazole, colistin, and novel beta-lactam/beta-lactamase inhibitors) and resistance spread (trimethoprim-sulfamethoxazole, amikacin, and colistin). In addition to antibiotic exposures, comorbidities (e.g., stage IV+ decubitus ulcers) and indwelling medical devices (e.g., gastrostomy tubes) were detected as unique risk factors for resistance spread. Conclusions Phylogenetic contextualization generated insights and hypotheses into how bacterial genetic background, patient characteristics, and clinical practices influence the emergence and spread of antibiotic resistance.