Integrated genomic and phenotypic profiling of clinical Klebsiella pneumoniae reveals an ESBL-dominated, lineage-diverse hospital population.

BACKGROUND

Klebsiella pneumoniae is a common cause of hospital-associated pneumonia, bloodstream infection, urinary tract infection, and drainage- or device-related infection. In clinical practice, its significance depends on resistance profile, lineage background, virulence-associated genes, specimen source, and ward context. Local data are useful because national epidemiology may not reflect the isolates recovered through routine work in a single hospital.

METHODS

We retrospectively analyzed 70 non-duplicate clinical K. pneumoniae isolates collected at a tertiary hospital from April to September 2025. The analysis combined clinical metadata, antimicrobial susceptibility testing (AST), short-read whole-genome sequencing, Antimicrobial Resistance(AMR) and vicimen-lerulence gene annotation, multilocus sequence typing (MLST), and department- and spevel summaries. The cohort came from diagnostic cultures and was not an active colonization-screening survey.

RESULTS

Respiratory specimens were the main source (41/70, 58.6%), and most isolates came from Internal Medicine (41/70, 58.6%) or intensive care units (ICUs; 21/70, 30.0%). The study-defined Extended-Spectrum Beta-Lactamase(ESBL)-associated category included 63 isolates. One isolate carried the carbapenemase gene blaKPC-3, so carbapenemase-related findings were handled as a single-isolate observation. AMR gene sets differed more between isolates than virulence gene sets (mean pairwise Jaccard distance, 0.613 versus 0.328). oqxA/oqxB and fosA were frequent, whereas beta-lactam, aminoglycoside, sulfonamide, tetracycline, and quinolone determinants varied among isolates. Virulence annotations mainly involved adhesion and siderophore-related loci. ST23 was the most common resolved sequence type (6 isolates), but its marker profile was mixed: four ST23 isolates carried aerobactin-associated genes and rmpA/rmpA2, three carried magA/wzy_K1, and two lacked the major aerobactin/rmpA-rmpA2 markers in the available annotation.

CONCLUSIONS

In this diagnostic collection, ESBL-associated and lineage-diverse K. pneumoniae isolates were more common than carbapenemase-associated isolates. The data do not indicate expansion of a single carbapenemase-producing clone. Because the study did not include active carriage screening, systematic confirmatory testing for all isolates, patient-location validation, or long-read plasmid analysis, it should not be used to estimate hospital-wide carbapenemase or hypervirulent K. pneumoniae prevalence.