Genome sequencing traces MRSA spread in high transmission setting

Patients from two intensive care units (ICUs) in a hospital in northeast Thailand were tested over a three-month period for MRSA. During this time, 46 patients and 5 staff members tested positive at least once (16% adult and 34% pediatric patients). The genetic similarity of the MRSA isolates precluded the use of conventional low-resolution genotyping approaches for distinguishing transmission from one person to another. Therefore, whole genome sequencing was performed on a total of 76 patient isolates, including up to two repeat isolates from patients who tested positive for MRSA in the first screen. None of the patients or staff members that tested positive for MRSA had symptoms of an infection but rather were carriers.

"A striking feature of the phylogenetic tree based on S. aureus whole genome sequencing was the presence of multiple distinct clades," said senior author Sharon Peacock from the University of Cambridge and Wellcome Trust Sanger Institute. "This suggested that multiple clades of the same lineage were circulating in the hospital at the same time."

Examining single base changes in the genomes of MRSA isolates allowed researchers to infer the most likely transmission routes between infected patients. Transmission events were observed between patients within the ICU, and also between patients and staff members. These results are in contrast to a previous study performed in the UK where patient-to-patient MRSA transmission in the ICU was rare.

"Our long term goal is to use such information to inform infection control practice," said Peacock. "The degree of transmission we demonstrated in our study has directly led to the prioritization of improved hand hygiene practices at the study hospital."

Genome sequencing also revealed that MRSA clades were dynamic in the ICU over the three-month period, with some clades more prevalent early in the study and others later.

Deep sequencing of isolates taken from a single patient carrying MRSA for the longest period revealed that although all isolates were of the same clade, there were small genetic differences between them, suggesting bacterial diversity within a single carrier. This lends further support to previous work suggesting that understanding transmission networks will require measures of within-host bacterial diversity as well as traditional 'shoe-leather epidemiological' data.