Introduction: Panton-Valentine leukocidin (PVL) is a staphylococcal toxin associated with recurrent skin and soft tissue infections (SSTI) and also necrotizing pneumonia. The aim of the study was to detect PVL in S. aureus (SA) isolates from clinical samples and to assign those isolates to clonal complexes and epidemic strains.

Material and methods: The collection of clinical SA isolates and PVL detection using RT-PCR were performed at University Hospital in Mannheim. Positive isolates were then characterized at IPHT Jena using DNA-microarrays that facilitated not only the detection of virulence genes including PVL and resistance markers, but also an assignment to CCs and strains. In addition, they were tested for PVL production using an experimental lateral flow (LF) test.

Results: Out of 556 SA isolates (MSSA and MRSA), 30 yielded PVL-positive PCR results and 27 were finally available for typing. Twelve of them were mecA-positive and three harboured SCCmec-associated fusC (fusidic acid resistance). For two methicillin-susceptible isolates (MSSA), PVL could not be confirmed by array and LF.

MSSA belonged to nine different lineages, with the most common one being CC30 (n=3). MRSA were assigned to seven CCs and 10 distinct strains. Strains CC152-MRSA-IV and CC22-MRSA-IV (PVL+/tst+) were found twice each, while the 8 other strains - including "USA300" - were identified only in single cases.

A concordance of PVL detection by array and LF was observed in 26/27 isolates (96%). One CC1-MRSA isolate yielded repeatedly negative LF results for PVL despite being positive in PCR and by array. Genome sequencing did not reveal significant mutations neither in its PVL genes nor in the agr locus.

Discussion: The MRSA rate among PVL-positive isolates was higher than the MRSA rate in the national average. All PVL-positive MRSA strains identified are known to be endemic in other parts of the world, thus they might be regarded as travel-associated. Providing for its frequency and clinical relevance, the implementation of screening tools for PVL – regardless whether it is PCR-based or a lateral flow assay – in routine diagnostics should be seriously considered.