Antagonistic effect of colistin on vancomycin activity against methicillin-resistant Staphylococcus aureus in in vitro and in vivo studies.

As concerns arise that the vancomycin MIC of methicillin-resistant Staphylococcus aureus (MRSA) could be increased by concurrent colistin administration, we evaluated the effect of colistin on vancomycin efficacy against MRSA via in vitro and in vivo studies. Among MRSA blood isolates collected in a tertiary-care hospital, we selected representative strains from community-associated MRSA strains (CA-MRSA; ST72-MRSA-SCCmec IV) and hospital-acquired MRSA strains (HA-MRSA; ST5-MRSA-SCCmec II). USA CA-MRSA (USA300), HA-MRSA (USA100), N315 (New York/Japan clone), and a MRSA standard strain (ATCC 43300) were used for comparison. We performed checkerboard assays to identify changes in the vancomycin MIC of MRSA following colistin exposure and evaluated the effect of a vancomycin-colistin combination using time-kill assays. We also assessed the in vivo antagonistic effect by administering vancomycin, colistin, and a combination of these two in a neutropenic murine thigh infection model. In the checkerboard assays, vancomycin MICs of all MRSA strains except N315 were increased by from 0.25 to 0.75 mu g/ml following colistin exposure. However, the time-kill assays indicated antagonism only against ST5-MRSA and USA100, when the vancomycin concentration was twice the MIC. In the murine thigh infection model with ST5-MRSA and USA100, vancomycin monotherapy reduced the number of CFU/muscle >1 log(10) compared to a combination treatment after 24 h in ST5-MRSA, indicating an antagonistic effect of colistin on vancomycin treatment. This study suggests that exposure to colistin may reduce the susceptibility to vancomycin of certain MRSA strains. Combination therapy with vancomycin and colistin for multidrug-resistant pathogens might result in treatment failure for concurrent MRSA infection.