Vancomycin MIC as a predictor of outcome in MRSA bacteraemia in the UK context.
Hope R., Blackburn RM., Verlander NQ., Johnson AP., Kearns A., Hill R., Hopkins S., Sheridan E., Livermore DM., Vancomycin Outcome Study Group None., UK Clinical Infection Research Group None.
UNLABELLED: Received 29 November 2012; returned 20 February 2013; revised 16 May 2013; accepted 18 May 2013 OBJECTIVES: Raised vancomycin MICs have been associated with poor outcomes for methicillin-resistant Staphylococcus aureus (MRSA) bacteraemia in the USA and mainland Europe. We investigated if this also applies in the UK, where EMRSA-15 (clonal complex 22) dominates. METHODS: Isolates from UK patients receiving vancomycin therapy for MRSA bacteraemia in 2008-10 were collected, along with clinical details. Outcomes (i.e. patient survival or bacteraemia resolution) were reported 28 days after vancomycin therapy ended. The relationship between clinical outcome and MIC--as determined by CLSI and BSAC agar dilution methods--was assessed. RESULTS: Among 228 MRSA bacteraemias, 82% were caused by EMRSA-15; 65% of the patients were male and the median age was 70.5 years. MICs correlated between methods, but CLSI agar dilution testing gave a mode at 1 mg/L with only 12% of results either side, whereas the BSAC method gave a mode straddling 0.7-1 mg/L with <4% outliers. Twenty-three percent of patients died, with MRSA contributory in half; another 17% had unresolved bacteraemia at 28 days. Neither death nor unresolved bacteraemia was significantly associated with higher vancomycin MICs by either method. Rifampicin co-therapy had no quantifiable effect on outcome. The patient's age was the only significant correlate of patient outcome (P < 0.01); the underlying medical condition of the patient was important for the resolution of bacteraemia (P < 0.01), though not for overall mortality. CONCLUSIONS: Subtle vancomycin MIC differences did not correlate with worse outcomes for vancomycin monotherapy or for vancomycin/rifampicin co-therapy in MRSA bacteraemia. Regardless of the exact MIC-outcome relationship, detecting such small MIC differences seems unlikely to be reliable in routine laboratories.