Multiple introductions of NRCS-A Staphylococcus capitis to the neonatal intensive care unit drive neonatal bloodstream infections: a case-control and environmental genomic survey

Background: The S. capitis NRCS-A strain has emerged as a global cause of late-onset sepsis associated with outbreaks in neonatal intensive care units (NICUs) whose transmission is incompletely understood. Methods: Demographic and clinical data for 45 neonates with S. capitis and 90 with other coagulasenegative staphylococci (CoNS) isolated from sterile sites were reviewed and clinical significance determined. S. capitis isolated from 27 neonates at two hospitals between 2017-2022 underwent long-read (ONT) (n=27) and short-read (Illumina) sequencing (n=18). These sequences were compared with S. capitis sequenced from blood culture isolates from other adult and paediatric patients in the same hospitals (n=6), S. capitis isolated from surface swabs (found in 5/150 samples), rectal swabs (in 2/69 samples) in NICU patients and NICU environmental samples (in 5/114 samples). Reads from all samples were mapped to a hybrid assembly of a local sterile site strain, forming a complete UK NRCS-A reference genome, for outbreak analysis and comparison with 826 other S. capitis from the UK and Germany. Results: S. capitis bacteraemia was associated with increased length of NICU stay at sampling (median day 22 vs day 12 for other coagulase-negative staphylococci, CoNS, isolated; p=0.05). A phylogeny of sequenced S. capitis revealed a cluster comprised of 25/27 neonatal sterile site isolates, 3/5 superficial, 2/2 rectal and 1/5 environmental isolates. No isolates from other wards belonged to this cluster. Phylogenetic comparison with published sequences confirmed the cluster were NRCS-A outbreak strain, but found relatively high genomic diversity (mean pairwise distance 84.9 single nucleotide polymorphisms, SNPs) and an estimated NRCS-A S. capitis molecular clock of 5.1 SNPs/genome/year (95%CrI 4.3-5.9). Presence of S. capitis in superficial cultures did not correlate with neonatal bacteraemia, but both neonates with rectal NRCS-A S. capitis carriage identified also experienced S. capitis bacteraemia. Conclusions: S. capitis bacteraemia occurred in patients with longer NICU admission than other CoNS. Genomic analysis confirms clinically significant infections with the NRCS-A S. capitis strain, distinct from non-NICU clinical samples. Multiple introductions of S. capitis, rather than prolonged environmental persistence, was seen over 5 years of infections. Data Summary: Sequence data generated by this study has been depositied in the Short Read Archive (Bioproject PRJEB76591, Accession numbers ERS20245094-ERS20245140). All accession numbers are listed in supplementary data. Individual-level data for case-control study can not be shared without indirect identification so has not been included.