Pan-genome multilocus sequence typing and outbreak-specific reference-based single nucleotide polymorphism analysis to resolve two concurrent Staphylococcus aureus outbreaks in neonatal services

Abstract

We used a two-step whole genome sequencing analysis for resolving two concurrent outbreaks in two neonatal services in Belgium, caused by exfoliative toxin A-encoding-gene-positive (eta+) methicillin-susceptible Staphylococcus aureus with an otherwise sporadic spa-type t209 (ST-109). Outbreak A involved 19 neonates and one healthcare worker in a Brussels hospital from May 2011 to October 2013. After a first episode interrupted by decolonization procedures applied over 7 months, the outbreak resumed concomitantly with the onset of outbreak B in a hospital in Asse, comprising 11 neonates and one healthcare worker from mid-2012 to January 2013. Pan-genome multilocus sequence typing, defined on the basis of 42 core and accessory reference genomes, and single-nucleotide polymorphisms mapped on an outbreak-specific de novo assembly were used to compare 28 available outbreak isolates and 19 eta+/spa-type t209 isolates identified by routine or nationwide surveillance. Pan-genome multilocus sequence typing showed that the outbreaks were caused by independent clones not closely related to any of the surveillance isolates. Isolates from only ten cases with overlapping stays in outbreak A, including four pairs of twins, showed no or only a single nucleotide polymorphism variation, indicating limited sequential transmission. Detection of larger genomic variation, even from the start of the outbreak, pointed to sporadic seeding from a pre-existing exogenous source, which persisted throughout the whole course of outbreak A. Whole genome sequencing analysis can provide unique fine-tuned insights into transmission pathways of complex outbreaks even at their inception, which, with timely use, could valuably guide efforts for early source identification.

Surveillance-embedded genomic outbreak resolution of methicillin-susceptible Staphylococcus aureus in a neonatal intensive care unit

“…The admixture of isolates from multiple spa types 31 did not occur in our study. In most studies, a relatively devoid group of isolates was arbitrarily considered as separate outbreak [15][16][17][18][19] . Sometimes WGS analysis resulted in a further separation, but never in re-assortment of outbreak isolates like we observed.…”

“…In addition, WGS was demonstrated to enable differentiation between three contemporaneous MRSA outbreaks in a NICU 17 , and also pointed towards international transmission of MRSA from NICUs 18 . The sole example investigating MSSA, proposed that the application of WGS on neonatal infection isolates can also distinguish between patient-to-patient transmission and re-seeding from a separate source during an outbreak spanning multiple years 19 .…”

Surveillance-embedded genomic outbreak resolution of methicillin-susceptibleStaphylococcus aureusin a neonatal intensive care unit

“…The admixture of isolates from multiple spa types (28) did not occur in our study. In most studies, a relatively devoid group of isolates was arbitrarily considered as separate outbreak (12)(13)(14)(15)(16). Sometimes WGS analysis resulted in a further separation, but never in reassortment of outbreak isolates like we observed.…”

“…In addition, WGS was demonstrated to enable differentiation between three contemporaneous MRSA outbreaks in a NICU (14), and also pointed towards international transmission of MRSA from NICUs (15). The sole example investigating MSSA, proposed that the application of WGS on neonatal infection isolates can also distinguish between transmission and exogenous introduction during an outbreak spanning multiple years (16).…”

The Integrated Rapid Infectious Disease Analysis (IRIDA) Platform

“…In addition to SNV-based methods, IRIDA provides functionality for performing MLST. Current MLST approaches that extend classical MLST schemes to include an organism's core genome (cgMLST) or whole genome (wgMLST) have been validated for surveillance of foodborne disease, and are quickly being adopted by surveillance programs such as PulseNet International, and used for a variety of nosocomial, zoonotic, and tuberculosis outbreaks (45)(46)(47)(48)(49). IRIDA has adopted the fast, accurate, and computationally efficient MentaLiST system as its main pipeline for cgMLST-based analysis.…”