Molecular characterisation of a Klebsiella pneumoniae neonatal sepsis outbreak in a rural Gambian hospital: a retrospective genomic epidemiology investigation

This retrospective genomic epidemiology study utilized whole-genome sequencing to identify a multidrug-resistant *Klebsiella pneumoniae* ST39 strain as the cause of a fatal neonatal sepsis outbreak in a rural Gambian hospital, pinpointing contaminated multi-use intravenous fluids as the source and highlighting the global spread and high-risk nature of the SL39 sublineage.

The Gist

The Story: A "Ghost" in the Hospital Nursery

Imagine a small, rural hospital in The Gambia (a country in West Africa). This hospital has a special nursery for newborn babies. In early 2023, something terrifying happened: a "ghost" began haunting the nursery.

This ghost wasn't a spirit; it was a microscopic bacteria called Klebsiella pneumoniae. It was a super-bug, meaning it was very hard to kill with normal medicines. It started making the babies sick with blood infections (sepsis). Tragically, out of 57 babies who got sick, 34 died. That is a 60% death rate.

The doctors knew the babies were sick, but they didn't know exactly which version of the bacteria was causing the trouble, or where it was coming from. It was like trying to catch a thief in a dark room without knowing what they looked like.

The Investigation: Putting on the "Genomic Detective" Glasses

To solve the mystery, a team of scientists used a high-tech tool called Whole-Genome Sequencing. Think of this as taking a "fingerprint" or a "DNA barcode" of every single bacteria found in the hospital.

They looked at:

1. Sick babies: Blood samples from the babies who got infected.
2. The environment: Swabs from sinks, beds, and even the liquid used for IV drips.
3. Old samples: Bacteria saved from the last 10 years to see if this was a new problem or an old one.

The Big Surprise:
When they looked at the DNA, they found that the hospital had been tricked.

• The Mix-up: The lab had thought all the bacteria were the same dangerous type. But the DNA test revealed that only about 29% of the cases were actually caused by the specific "super-bug" outbreak strain. The rest were different, less dangerous bacteria.
• The Real Culprit: The true killer was a specific strain called ST39. It was a "clone" of itself, meaning it was the exact same bacteria spreading from patient to patient.

The Source: The "Poisoned Water" Bottle

The most shocking discovery was where the bacteria was hiding.

Usually, when a hospital has an outbreak, people suspect dirty hands, dirty sheets, or dirty sinks. While those were dirty, the DNA fingerprinting proved the main source was the Intravenous (IV) fluid bags.

• The Analogy: Imagine the hospital was using large jugs of water to mix medicine for the babies. Instead of using a fresh, sealed bottle for every baby, they were using one big jug, sticking a needle in it, and drawing medicine out for many different babies over several days.
• The Result: The bacteria got into that big jug. Every time they drew medicine out, they were essentially injecting the babies with the bacteria. The DNA from the bacteria in the IV bags matched the DNA from the sick babies perfectly.

The Weapon: The "Armored Tank"

Why was this bacteria so deadly?

1. The Armor: This specific strain (ST39) had a special shield called a plasmid. Think of a plasmid as a backpack the bacteria wears. Inside this backpack was a "weapon" called blaCTX-M-15.
2. The Weapon's Power: This weapon makes the bacteria immune to the most common antibiotics doctors use to treat newborns (like Cefotaxime and Gentamicin). It's like the bacteria wearing a suit of armor that bullets (antibiotics) bounce right off of.
3. The Mutation: In one case, the bacteria didn't even need the backpack. It had welded the weapon directly onto its own body (chromosome). This is even scarier because you can't just "take off" the armor; the bacteria is born with it.

The Solution: Cleaning the "Kitchen"

Once the scientists found the source (the IV bags) and the specific bacteria, the hospital team went into action. They didn't just clean; they changed the rules:

• They stopped using the big, multi-use IV bags.
• They switched to single-use, sterile water for every baby.
• They scrubbed the sinks and floors aggressively to kill the biofilm (slime) where bacteria hide.
• They trained the staff on how to prepare medicine safely.

Within a few months, the "ghost" was gone. No new babies got sick, and the outbreak was over.

The Bigger Picture: A Global Warning

The scientists didn't stop at just this one hospital. They compared their "fingerprint" to a global database of bacteria from all over the world.

• The Trend: They found that this specific type of bacteria (ST39) is becoming a "super-spread" clone. It is traveling across Africa, Asia, and Europe.
• The Lesson: This isn't just a local problem. It's a global warning that bacteria are evolving fast, and our current medicines are losing the battle.
• The Future: The paper suggests that in the future, we might need vaccines for babies (or even pregnant mothers) to protect them against these specific types of bacteria, just like we have vaccines for measles or polio.

Summary in One Sentence

This paper tells the story of how scientists used DNA detective work to find out that a deadly bacteria outbreak in a Gambian hospital was caused by contaminated IV fluids, leading to life-saving changes that stopped the deaths and highlighted a growing global threat of super-bugs.

Technical Summary

1. Problem Statement

• Context: Klebsiella pneumoniae is a leading cause of neonatal sepsis in sub-Saharan Africa, associated with high mortality and escalating antimicrobial resistance (AMR). First-line therapies (ampicillin, third-generation cephalosporins, gentamicin) are increasingly ineffective due to Extended-Spectrum Beta-Lactamase (ESBL) production.
• The Outbreak: In 2023, a rural hospital in The Gambia (Bansang Hospital) experienced a severe outbreak of multidrug-resistant (MDR) K. pneumoniae sepsis among neonates.
  • Scale: 57 neonatal cases with a 60% case fatality rate (CFR).
  • Initial Findings: Environmental screening suggested contamination of water, surfaces, and intravenous (IV) fluids, but the specific outbreak strain and precise transmission source remained unclear.
• Knowledge Gap: There is a lack of high-resolution genomic data on K. pneumoniae outbreaks in African neonatal units, limiting the ability to distinguish between outbreak strains, identify specific reservoirs, and understand the global epidemiology of high-risk clones.

2. Methodology

The study employed a retrospective whole-genome sequencing (WGS) approach to re-analyze clinical and environmental isolates from the outbreak and historical surveillance data.

• Study Design & Samples:
  • Clinical Isolates: 51 blood culture isolates from patients (0–59 months) during the outbreak (Jan 2023–Mar 2024), plus 56 historical isolates (2012–2022).
  • Environmental Isolates: 16 samples collected during the outbreak investigation (Oct–Dec 2023), including IV fluids, water, and surfaces.
  • Total Sequenced: 148 isolates (66 outbreak clinical, 16 environmental, 10 post-outbreak clinical, 56 historical).
• Sequencing Technology:
  • Primary: Oxford Nanopore Technologies (ONT) long-read sequencing (PromethION) to generate complete, closed genomes.
  • Validation: A subset (17 isolates) was sequenced using Illumina short-reads to validate assembly accuracy and SNP calling.
• Bioinformatics & Analysis:
  • Assembly: Hybracter for hybrid assembly; Pathogenwatch for species identification, MLST, capsular (K) and lipopolysaccharide (O) typing, and AMR detection.
  • Phylogenetics: Core-genome SNP analysis to define transmission clusters (threshold: ≤10 SNPs genetic distance and ≤28 days temporal proximity).
  • Plasmid Analysis: Complete plasmid assembly and comparative genomics (BLASTn, clinker) to track resistance gene mobility.
  • Global Context: Comparison with 684 publicly available ST39 genomes from 50 countries.

3. Key Contributions

• Refinement of Outbreak Scope: Genomic analysis corrected the species identification, revealing that ~23% of culture-confirmed "outbreak" isolates were actually K. quasipneumoniae. This refined the true outbreak size to 29 confirmed K. pneumoniae cases (ST39).
• Source Attribution: Definitively identified multi-use intravenous (IV) fluid bags as the primary transmission source by linking clinical isolates to environmental isolates from IV bags with high genetic similarity (1–3 SNPs).
• Genomic Characterization: Provided a high-resolution description of the outbreak strain (ST39-KL62), including its plasmid architecture, chromosomal resistance islands, and global lineage.
• Mechanism of Resistance Evolution: Discovered a rare instance of chromosomal integration of ESBL genes (bla_CTX-M-15), replacing a transposon via homologous recombination, offering insights into stable resistance fixation.

4. Key Results

A. Outbreak Strain Identification

• Strain: The outbreak was driven by a single clonal lineage: Sequence Type 39 (ST39) with Capsular Type KL62 and O-antigen OL2α.2.
• Transmission Cluster: 28 of the 29 outbreak isolates formed a tight cluster (0–5 SNPs difference), confirming nosocomial spread over 7 months (June 2023–Jan 2024).
• Exclusion: The strain was distinct from a 2016 outbreak in The Gambia (ST39-KL23) and was not found in historical surveillance isolates (2012–2022), indicating a new introduction.

B. Resistance Profile & Plasmid Architecture

• Phenotype: Multidrug-resistant (MDR), resistant to third-generation cephalosporins and gentamicin.
• Genotype:
  • Plasmid-borne: 25/29 isolates carried a ~187 kbp IncFIBK-IncFIIK plasmid (pNS39_A) harboring bla_CTX-M-15 (ESBL) and aac(3)-IId (gentamicin resistance).
  • Chromosomal Integration: One isolate (38277B1) lacked the plasmid but carried bla_CTX-M-15 and aac(3)-IId on an 11.2 kb chromosomal resistance island. This island replaced a ~35 kb Tn3-family transposon via ISKpn14-mediated homologous recombination.
  • Other Resistances: Chromosomal genes for chloramphenicol, tetracycline, and trimethoprim-sulfamethoxazole resistance were also present.

C. Environmental Linkage

• IV Fluids: Three IV fluid samples collected in Oct/Dec 2023 contained ST39-KL62.
  • Two matched clinical isolates (1–3 SNPs) and shared the pNS39_A plasmid.
  • One lacked the plasmid but shared the chromosomal resistance profile.
• Conclusion: The genomic evidence confirmed that contaminated multi-use IV fluids were the vector for transmission, rather than water sources or general ward surfaces (which yielded unrelated strains).

D. Global Epidemiology of ST39

• Lineage Diversification: ST39 has diversified into three main clonal groups (CG): CG39, CG10192, and CG10195.
• Outbreak Lineage: The Gambian outbreak strain belongs to CG10192, which is predominantly found in Africa and is strongly associated with the KL62 capsule and bla_CTX-M-15.
• Contrast: The earlier Gambian outbreak (2016) belonged to CG39 (KL23), which is globally distributed.
• Resistance Trends: CG10192-KL62 shows high ESBL prevalence (89%) but low carbapenemase carriage, whereas CG39 isolates show more geographic variation in resistance genes.

5. Significance and Implications

• Infection Prevention and Control (IPC): The study highlights the critical risk of multi-use IV fluids in low-resource settings. It demonstrates that even with standard environmental cleaning, contaminated reagents can sustain outbreaks.
• Genomic Surveillance Value: WGS provided actionable intelligence that culture methods could not:
  • Distinguished the outbreak strain from background noise (K. quasipneumoniae).
  • Pinpointed the exact source (IV bags) to guide targeted interventions.
  • Confirmed the outbreak was over only after genomic clearance of the specific strain.
• Public Health Policy:
  • Supply Chain: Reliance on multi-use bags due to stock-outs creates a cycle of contamination. Sustainable solutions (single-use ampoules, ring-fenced financing) are essential.
  • Vaccine Development: The dominance of KL62 (and KL2) in African neonatal sepsis supports the feasibility of developing maternal vaccines targeting these specific capsular serotypes.
  • Capacity Building: The study underscores the need for real-time genomic surveillance in low-resource hospitals to enable rapid outbreak response rather than retrospective analysis.

Conclusion: This investigation illustrates how high-resolution genomics can transform outbreak management in resource-limited settings, shifting from broad, ineffective interventions to precise, evidence-based control measures that save lives. It identifies ST39-KL62 as an emerging, high-risk, MDR clone in Africa requiring targeted surveillance and prevention strategies.