852 - Genomic Clusters of Methicillin-Resistant Staphylococcus aureus (MRSA) Causing Bloodstream Infections (BSIs) in Hospitalized Adults, 2018-19

The ePoster content is available to users who are logged in.<br><br>Please log in and then come back here to use this button.

Co-Author(s)

TR

Timothy D. Read

Professor
Emory University School of Medicine
Atlanta, Georgia
Disclosure: I do not have any relevant financial / non-financial relationships with any proprietary interests.
NJ

Natasia F. Jacko

Research Coordinator
University of Pennsylvania
Philadelphia, Pennsylvania
Disclosure: I do not have any relevant financial / non-financial relationships with any proprietary interests.
RP

Robert A. Petit

Senior Data Analyst
Emory University
Atlanta, Georgia
Disclosure: I do not have any relevant financial / non-financial relationships with any proprietary interests.
DP

David A. Pegues

Professor of Medicine
Hospital of the University of Pennsylvania
Philadelphia, Pennsylvania
Disclosure: I do not have any relevant financial / non-financial relationships with any proprietary interests.

Presenting Author(s)

Michael Z. David

Assistant Professor
University of Pennsylvania Perelman School of Medicine
Philadelphia, Pennsylvania
Disclosure: GSK (Consultant)

Background: MRSA BSIs have 15-50% mortality and are commonly diagnosed in US hospitals. However, the frequency of hospital transmission of MRSA causing BSI is unknown.

Methods: We performed Illumina shotgun whole genome sequencing (WGS) of 106 sequential MRSA isolates from different adults with a BSI at two Philadelphia academic hospitals in a single health system in July 2018-June 2019. We abstracted clinical data from the electronic medical record. Genomic data were analyzed preliminarily using the Staphopia Analysis Pipeline.

Results: Among 106 subjects, 51.9% were male, 47.2% were white, 46.2% were black, 23.6% were < 40 years of age, and mean age was 53.1 years (s.d. 17 years). One isolate had WGS data that were inadequate for analysis. Of 105 genomes, 52 were clonal cluster (CC) 8, 22 were sequence type (ST) 5, and 16 were ST105; the remaining 15 strains belonged to 8 other CCs. Of CC8 strains, 44 were USA300 and 6 were USA500. There were 6 clusters (i.e., < 35 SNP differences in the core genome) among the 105 isolates. Four clusters were CC5 and two were CC8 strains. One cluster of CC5 strains involved 3 subjects, and 5 clusters involved 2 subjects. One cluster of ST8/USA300 strains were separated by only 1 SNP (Fig a). This and two other clustered pairs were from subjects who had overlapping hospital stays. Two of these paired subjects had an overlap in the same unit while the third pair was in the hospital together on a number of occasions (total of 40 days overlap) but never simultaneously in the same unit. The other three clustered pairs did not have temporally overlapping hospital stays, suggesting transmission via a hospital reservoir. One of these three pairs had hospitalizations overlapping in time, one at each study hospital, before each of them had infections with the related MRSA strains. There was not a clear-cut clustering of SNP distances among the isolate genomes into transmission and non-transmission groups, with some pairs of patient isolates separated by 40-80 SNPs (Fig. b). Figure 1.

Conclusion: We were able to discern from WGS data alone that some MRSA BSIs in 2 hospitals were likely due to strains transmitted between patients. Universal WGS of BSI strains may detect MRSA outbreaks in real time, even in the absence of overlapping hospitalizations, and is an emerging strategy to detect healthcare transmission of MRSA.