The numbers alone are striking. Twenty-eight of the 30 isolates—93 percent—were classified as multidrug-resistant, meaning they were resistant to three or more antibiotic classes. Seven of those, or a quarter of the MDR isolates, were extensively drug-resistant, showing non-susceptibility across eight antibiotic classes including the carbapenems, drugs typically reserved as last-line therapies for serious Gram-negative infections. Only amikacin, an aminoglycoside, retained activity against the XDR strains. Six of the seven XDR isolates came from a single institution, UTH, which serves as Zambia’s national referral hospital, while one originated from NUTH. Non-susceptibility was pervasive across the broader panel: 90 percent of isolates resisted ceftriaxone and cefepime, 87 percent resisted ceftazidime, and 90 percent resisted trimethoprim-sulfamethoxazole.

To understand what was driving this resistance, the team turned to whole-genome sequencing on an Illumina NextSeq 2000 platform, generating 150-base-pair paired-end reads that were quality-trimmed, de novo assembled, and mined with tools including the Comprehensive Antibiotic Resistance Database and PlasmidFinder. Multilocus sequence typing based on seven housekeeping genes assigned the isolates to sequence types, the standard shorthand for tracking bacterial lineages. The result was remarkable diversity: 18 distinct sequence types among 29 confirmed K. pneumoniae isolates, including two novel types never before designated. One isolate initially presumed to be K. pneumoniae turned out to be K. oxytoca, a related species, underscoring the value of genomic species confirmation in clinical microbiology.

Within that diversity, the researchers identified a roster of internationally notorious high-risk clones: ST11, ST15, ST101, ST307, ST377, and ST607. These lineages have been implicated in hospital outbreaks on multiple continents and are strongly associated with extended-spectrum beta-lactamases and carbapenemases. UTH hosted all five of the dominant hospital-associated types detected in the study, and several types—ST15, ST307, ST377, and ST607—appeared at more than one hospital, suggesting circulation through patient referral networks or regional dissemination rather than independent emergence. Notably, ST607 has only been reported twice before in Africa, in Malawi and Kenya, where it emerged as a major ESBL-producing clone carrying IncFIB plasmids and the blaCTX-M-15 gene—precisely the profile of the Zambian isolates.

Among the seven XDR isolates, five belonged to clinically dominant clones: two ST377 isolates plus single representatives of ST101, ST307, and ST607. The remaining two belonged to ST13 and ST831, lineages that had not previously been reported with XDR profiles, hinting that resistance is spreading into clones not traditionally considered high-risk. Hierarchical clustering of susceptibility and resistance gene data grouped the isolates into five clusters, Kp1 through Kp5, with the most resistant cluster, Kp5, concentrated at UTH. The authors suggest that the national referral hospital may act as a central hub that accumulates diverse MDR lineages and seeds regional facilities through inter-facility patient transfers.

The resistance determinants themselves were dominated by plasmid-borne beta-lactamase genes. blaCTX-M-15, the quintessential ESBL gene of the modern era, was present in 93 percent of isolates, while blaTEM variants appeared in 70 percent. The carbapenemase gene blaNDM-5, which encodes an enzyme capable of destroying even last-resort carbapenem antibiotics, was detected in 13 percent of isolates. Aminoglycoside-modifying enzymes were widespread, led by aac(6′)-Ib-cr5 at 77 percent and aac(3)-IIe at 64 percent, and every single isolate carried emrD, a multidrug efflux pump gene. Genes conferring resistance to sulfonamides, fosfomycin, and quinolones, including oqxA at 83 percent, rounded out a resistance repertoire of 40 distinct genes detected across the collection.

Plasmid analysis revealed the likely engine of this dissemination. Nine plasmid replicon types were identified, but one stood out: IncFIB(K)_1_Kpn3, found in 26 of 30 isolates, or roughly 86 percent. This replicon is known to be associated with dozens of resistance profiles, including ESBLs and carbapenemases. Its presence across phylogenetically distinct lineages—dominant high-risk clones and minor types alike—points to horizontal gene transfer as a critical force sustaining multidrug resistance in these hospitals. Several sequence types, including ST39, ST86, ST101, ST5133, and ST831, co-harboured both IncFIB(K)_1_Kpn3 and IncHI1B_1_pNDM-MAR, the latter a replicon strongly linked to NDM-mediated carbapenem resistance. Such plasmid promiscuity means that even if a dominant resistant clone is eliminated, the resistance genes themselves can persist and jump into new bacterial backgrounds.

Virulence profiling added a further dimension of concern. All isolates carried genes for enterobactin iron acquisition, adherence, and immune evasion, and complete yersiniabactin loci were found in 18 isolates spanning multiple sequence types, including high-risk MDR lineages. One ST86 isolate, ZLU0042372, carried complete aerobactin and salmochelin loci—a siderophore combination characteristic of hypervirulent K. pneumoniae—making it the leading candidate hypervirulent isolate in the study, although the authors caution that definitive classification would require functional validation. Encouragingly, classic hypervirulence markers such as rmpA and peg-344 were absent from all isolates. Phylogenetically, the Zambian isolates did not form hospital-specific lineages; instead, they dispersed across the tree and several clustered with non-Zambian reference genomes, situating local strains within global genetic backgrounds.

The study has limitations that the authors acknowledge. Short-read sequencing cannot reliably resolve whether resistance genes sit on plasmids or the chromosome, and the relatively small sample—just three isolates from Livingstone and six from Ndola—constrains conclusions about diversity at those sites. The restriction to bloodstream isolates may also bias findings toward strains adapted to invasive disease. Even so, as the first genomic characterisation of K. pneumoniae in Zambia, the work establishes a baseline that was previously missing from sub-Saharan Africa, a region where genomic surveillance of this pathogen has lagged despite its recognised priority in Zambia’s National Action Plan on antimicrobial resistance.

The implications are clear. With XDR strains already circulating in referral hospitals and mobile plasmids shuttling resistance genes across clonal boundaries, the window for containment is narrowing. The authors call for strengthened laboratory surveillance across all tertiary hospitals in Zambia, coordinated genomic monitoring between facilities, and reinforced infection prevention and control alongside antimicrobial stewardship programmes. Without such interventions, they warn, these clones could disseminate from tertiary centres into secondary hospitals and the wider community, turning a hospital problem into a national public health crisis.

Subject of Research: Genomic surveillance of multidrug-resistant and extensively drug-resistant Klebsiella pneumoniae bloodstream infections in Zambian tertiary hospitals

Article Title: Genomic insights into multidrug and extensively drug-resistant Klebsiella pneumoniae from bloodstream infections at three Zambian tertiary hospitals

Article References: Genomic insights into multidrug and extensively drug-resistant Klebsiella pneumoniae from bloodstream infections at three Zambian tertiary hospitals. (n.d.). https://doi.org/10.3389/frabi.2026.1933211

Image Credits: AI Generated

DOI: 10.3389/frabi.2026.1933211

Keywords: Klebsiella pneumoniae, antimicrobial resistance, genomic surveillance, Zambia, XDR, high-risk clones, carbapenemase, NDM-5, ESBL, plasmids, bloodstream infections, whole-genome sequencing