Whole-genome sequencing reveals five separate linezolid-resistant S. epidermidis clusters in Austrian hospitals

When three linezolid-resistant Staphylococcus epidermidis isolates turned up in blood cultures from patients at two neighboring Austrian hospitals between March and April 2023, infection control teams suspected a single outbreak spreading through the frequent patient transfers and shared clinical care that connected the two facilities. A genomic investigation published in Frontiers in Microbiology has now shown that the reality was more complicated: rather than one clonal outbreak, whole-genome sequencing of 53 cases identified five genetically distinct clusters belonging to three different sequence types, suggesting multiple independent lineages of the drug-resistant bacterium circulating at the same time.

The 53 cases split into 22 detected in Hospital A and 31 in Hospital B. The median age of the patients was 60 years, and 37 of the 53, or 70 percent, were male. Most patients in Hospital A had been on an oncology ward at the time of detection, accounting for 50 percent of that group, while patients in Hospital B were primarily in the surgical ICU, at 55 percent. Immunosuppression was documented in 20 of the 22 Hospital A cases, and hematological-oncological diseases were the most frequent underlying conditions in those patients. Recent linezolid exposure within 28 days was confirmed in only 7 patients, with a median treatment duration of 9 days, and information on prior linezolid therapy was unavailable for the Hospital B cases.

Genomic typing of the 52 isolates that passed quality filters revealed three sequence types: ST2 with 22 isolates, ST5 with 3, and ST22 with 27. Cluster analysis using a core genome MLST scheme with 1,828 core genes and a threshold of six or fewer allelic differences identified five clusters and eight singletons. Cluster 1, the largest, belonged to ST22 and spanned two years, comprising 18 clinical and 7 screening isolates, 24 of which were detected in Hospital B, particularly the surgical ICU, while a single case was identified in Hospital A. That Hospital A patient had not been hospitalized in Hospital B but had undergone three diagnostic procedures there between December 2022 and January 2023, before the isolate was recovered in March 2023.

The remaining clusters showed different patterns. Cluster 2, an ST2 lineage, consisted of 11 screening isolates from Hospital A, all but one from oncological patients, and spanned six months from October 2024 with isolates differing by zero to one allele. Cluster 3 grouped three ST5 isolates from Hospital A detected between March 2023 and July 2024, differing by two alleles. Cluster 4, another ST2 lineage, comprised three clinical isolates from Hospital A over a single month in summer 2023, and Cluster 5 included two ST2 isolates from Hospital B detected within a two-week window. Although three of the five clusters were ST2, their allelic distances and temporal patterns indicated independent evolutionary trajectories rather than a single spreading clone.

Resistance levels were striking. All 53 isolates were resistant to oxacillin and linezolid by disk diffusion, and 45 displayed high-level linezolid resistance, defined as an MIC of at least 48 mg/L. Kruskal-Wallis testing showed statistically significant differences in MIC across sequence types, with post-hoc comparisons indicating that ST22 isolates had significantly higher MICs than both ST2 and ST5. The resistance mechanisms differed by lineage: all ST22 Cluster 1 isolates carried the C2534U and U2504A mutations in the 23S rRNA gene, present in four of five rRNA gene copies in a representative hybrid assembly, while the ST2 clusters and singletons carried the G2576U mutation, detected in all five copies of a representative ST2 genome. Distinct amino acid substitutions were also found in the ribosomal proteins L3 and L4, with cluster-specific patterns such as G152D and D159Y in Cluster 1 and a 71-GGU-72 insertion in L4 shared by Clusters 3 and 4.

Notably, no transferable linezolid resistance genes, such as cfr, cfr(D), optrA or poxtA, were detected in any isolate, and Nanopore sequencing of a subset of 22 isolates confirmed that no plasmids carried linezolid resistance genes. This points to chromosomal mutation, rather than horizontal gene transfer, as the route to resistance in these lineages. All isolates carried the mecA cassette conferring methicillin resistance, and the qacA gene, associated with resistance to quaternary ammonium compounds, was found in 48 isolates, in all ST2 isolates and all but one ST22 isolate. Genes associated with erythromycin resistance were highly prevalent, and mutations in rpoB conferring rifampin resistance were detected in 22 isolates. The ica operon responsible for biofilm formation was present in nearly all ST2 and ST22 isolates but absent from all ST5 isolates.

Comparison with 89 publicly available LRSE genomes from earlier studies in Austria, France and Italy showed that the Austrian clusters were more closely related to isolates from the 2019 Austrian study than to those from France or Italy. The closest related ST5 isolate from the Austrian study differed by 13 alleles from Cluster 3, while the nearest French isolate differed by 16 alleles from Cluster 1 and Italian isolates were separated by 41 to 53 alleles from the ST2 clusters. The authors interpret these distances as inconsistent with recent transmission between countries but consistent with the long-term persistence and wider dissemination of hospital-adapted S. epidermidis lineages that are not restricted to a single institution.

The investigators are careful about what the data cannot show. Epidemiological information was limited to hospital, date of detection and ward at the time of identification, with no data on prior patient movements, contacts between patients, healthcare worker interactions, shared equipment or environmental reservoirs. Genomic clustering therefore indicates genetic relatedness and potential epidemiological links, not proof of direct transmission or its direction, and the six-allele cluster threshold, adopted from previous LRSE work, has not been specifically validated for these sequence types. Because linezolid administration data were incomplete, particularly for Hospital B cases, the role of antibiotic selective pressure in driving resistance could not be assessed, even though prior linezolid exposure is a known driver of resistance and published evidence indicates that up to 55 percent of LRSE cases occur in patients without such exposure.

Surveillance ended in April 2025, coinciding with the cessation of isolate submission rather than with any demonstrated resolution of transmission; clinical cases continued to occur afterward, and infection prevention and control measures, including admission screening on the most affected wards from October 2024, strict isolation, and enhanced contact precautions, remained in place. The authors caution that the apparent absence of further detections after these measures cannot be taken as evidence of their effectiveness, since it may simply reflect the end of the surveillance window. What the study does demonstrate is the utility of whole-genome sequencing for detecting and characterizing resistant clusters that would otherwise appear as a single outbreak, and the necessity of pairing genomic surveillance with comprehensive epidemiological investigation to understand and control the spread of linezolid-resistant S. epidermidis in hospitals.

Subject of Research: Genomic epidemiology of linezolid-resistant Staphylococcus epidermidis clusters in two Austrian hospitals

Article Title: Genomic epidemiology unravels simultaneous clusters of linezolid-resistant Staphylococcus epidermidis isolated in Austria, 2023–2025

Article References: Cabal Rosel, A., Kerschner, H., Weinberger, J., Stöger, A., Lippert, K., Daza Prieto, B., Mellmann, A., Apfalter, P., Ruppitsch, W., & Hartl, R. (2026). Genomic epidemiology unravels simultaneous clusters of linezolid-resistant Staphylococcus epidermidis isolated in Austria, 2023–2025. Frontiers in Microbiology, 17, Article 1934085. https://doi.org/10.3389/fmicb.2026.1934085

Image Credits: AI Generated

DOI: 10.3389/fmicb.2026.1934085

Keywords: linezolid resistance, Staphylococcus epidermidis, whole-genome sequencing, hospital outbreak, genomic epidemiology, Austria, 23S rRNA mutations, cgMLST, antimicrobial resistance, nosocomial transmission, ST22, infection control