Letters to the Editor Specific Detection of Methicillin-Resistant Staphylococcus Species by Multiplex PCR

Vannuffel and coworkers (1) offer an interesting approach to the problem of rapid and specific detection of methicillin-resistant Staphylococcus aureus in clinical specimens. This ap-proach is based on the premise that methicillin-resistant S. aureus strains harbor both mecA and femA genes, while methi-cillin-resistant coagulase-negative staphylococci (CNS) carry the mecA gene only. The authors detect both genes directly in clinical samples (e.g., blood samples) by multiplex PCR and claim that this simultaneous detection identifies both the species and its phenotypic resistance. We see a major problem with this approach: the species determination fails as soon as the clinical sample does not contain a pure culture of either S. aureus or CNS. Indeed, the femA gene is present in all S. aureus strains, irrespective of the presence of the mecA gene. This means that a mixed popula-tion of methicillin-susceptible S. aureus (e.g., in blood) and methicillin-resistant CNS (e.g., from skin contamination) will also generate two DNA fragments in the multiplex PCR and result in the erroneous identification of methicillin-resistant S. aureus. We agree with Vannuffel et al. that a rapid and reliable test for identification of methicillin-resistant S. aureus might help the clinicians in reducing the use of glycopeptide antibiotics in the hospital. Unfortunately, simultaneous detection of femA and mecA genes directly in clinical samples is not the answer