[Comparison of different techniques for the detection of heterogeneous resistance to methicillin in Staphylococcus aureus]

Abstract

The methods for detection of methicillin resistant S. aureus (MRSA) can fail to detect resistance because phenotypic expression is often heterogeneous (40% of strains). Seventy four strains of S. aureus [4 methicillin susceptible strains, 10 homogeneous MRSA (Ro) and 60 heterogeneous MRSA (Rh)] were isolated from different french hospitals in Paris. These strains were tested by different methods: oxacillin screen plate with 6 micrograms/ml oxacillin and 4% NaCl, agar diffusion method with 5 micrograms oxacillin disk tested either at 30 degrees C on Mueller-Hinton medium or at 37 degrees C on Mueller-Hinton plus 5% NaCl, BBL Crystal MRSA ID system tested with two inocula (0.5 and 1 McFarland equivalent bacterial suspension) at 37 degrees C for 4 h and 5 h. Dot-blot hybridization was performed under stringent condition with the mecA probe. The accuracy of the different methods for the detection of methicillin resistance is equivalent, except for the BBL crystal system with a 0.5 McFarland inoculum wich detects the resistance with an accuracy of 86% for Ro strains and 69% for Rh strains. In other respects, there was a close correlation with the detection of the phenotypic resistance and the presence of mecA gene. So this study demonstrates that these various methods can be used for the detection of methicillin resistant S. aureus. For a rapid detection (below 5 h) the BBL crystal system with a 1 McFarland inoculum can be used; the agar diffusion method remains a good method provided some conditions (inoculum, incubation temperature, addition of salt, incubation period); the oxicillin screnn plate is a very attractive method for it is easy and reliable.