Direct nucleic acid-based detection of MRSA from clinical specimens1

Abstract

Over the past few years a dramatic increase in the methicillin resistance of S. aureus isolates has been observed worldwide. Infections with methicillin-resistant Staphylococcus aureus (MRSA) are accompanied by higher rates of morbidity and mortality, extended length of stay and increased overall costs. Large-scale hygienic measures are necessary to prevent the spread of the resistant isolates to patients not yet infected or colonized by MRSA. It has been demonstrated that the sooner such measures are taken, the smaller the probability of transmission to other patients. To achieve this result requires rapid identification of MRSA carriers. Even more recent culture-based detection methods can provide results only after 24 h. The use of modern PCR-supported detection methods opens the possibility to record findings within only a few hours. While the original test concepts, which are based on the separate detection of the mecA gene and various markers for S. aureus, have proven to be a good culture verification test, they are of limited usefulness for the rapid and meaningful direct detection of MRSA in clinical specimens. Recently developed, innovative test concepts based on the targeted detection of the integration of a so-called SCCmec element (a gene cluster typically also harboring the methicillin-resistance transmitting mecA-gen) in the S. aureus genome are better suited for this purpose. At present there exist a number of commercial test systems as well as a multitude of test concepts developed in-house, all of which are based on this fundamental principle. Although these assays usually display high levels of sensitivity (90–100%), specificity (93–99%) and positive predictive values above 95%, their negative predictive values are usually distinctly lower (80–95%). Meanwhile, with the increasingly broader use of these test methods several disturbance variables have been identified that can lead to false positive or false negative results. In view of these limits a combination of PCR-supported and cultural detection methods would appear to present a sensible policy.