Clonal variation in high- and low-level phenotypic and genotypic mupirocin resistance of MRSA isolates in south-east London.

OBJECTIVES Both low-level mupirocin resistance (LMR) and high-level mupirocin resistance (HMR) have been identified. The aim of this study was to determine the epidemiology of LMR and HMR in MRSA isolates at five hospitals that have used mupirocin for targeted decolonization as part of successful institutional control programmes. METHODS All MRSA identified in three microbiology laboratories serving five central and south-east London hospitals and surrounding communities between November 2011 and February 2012 were included. HMR and LMR were determined by disc diffusion testing. WGS was used to derive multilocus sequence types (MLSTs) and the presence of HMR and LMR resistance determinants. RESULTS Prevalence of either HMR or LMR amongst first healthcare episode isolates from 795 identified patients was 9.69% (95% CI 7.72-11.96); LMR was 6.29% (95% CI 4.70-8.21) and HMR was 3.40% (95% CI 2.25-4.90). Mupirocin resistance was not significantly different in isolates identified from inpatients at each microbiology laboratory, but was more common in genotypically defined 'hospital' rather than 'community' isolates (OR 3.17, 95% CI 1.36-9.30, P = 0.002). LMR was associated with inpatient stay, previous history of MRSA and age ≥65 years; HMR was associated with age ≥65 years and residential postcode outside London. LMR and HMR varied by clone, with both being low in the dominant UK MRSA clone ST22 compared with ST8, ST36 and ST239/241 for LMR and with ST8 and ST36 for HMR. V588F mutation and mupA carriage had high specificity (>97%) and area under the curve (>83%) to discriminate phenotypic mupirocin resistance, but uncertainty around the sensitivity point estimate was large (95% CI 52.50%-94.44%). Mutations in or near the mupA gene were found in eight isolates that carried mupA but were not HMR. CONCLUSIONS Mupirocin resistance was identified in <10% of patients and varied significantly by clone, implying that changes in clonal epidemiology may have an important role in determining the prevalence of resistance in conjunction with selection due to mupirocin use.