Panton-Valentine Leukocidin: not as lethal as all that?

  • Emma Nickerson, Mahidol-Oxford Tropical Medicine Research Unit and University of Oxford, Thailand
  • Miss Janjira Thaipadungpanit, Mahidol-Oxford Tropical Medicine Research Unit, Thailand
  • Dr Sarinna Tumapa, Mahidol-Oxford Tropical Medicine Research Unit, Thailand
  • Mr Gumphol Wongsuvan, Mahidol-Oxford Tropical Medicine Research Unit, Thailand
  • Miss Premjit Amornchai, Mahidol-Oxford Tropical Medicine Research Unit, Thailand
  • Mrs Vanaporn Wuthiekanun, Mahidol-Oxford Tropical Medicine Research Unit, Thailand
  • Prof Nicholas Day, Mahidol-Oxford Tropical Medicine Research Unit and University of Oxford, Thailand
  • Sharon Peacock, Mahidol-Oxford Tropical Medicine Research Unit, University of Oxford and Liverpool School of Tropical Medicine, Thailand
  • Objective: To determine associations between clinical features and outcome from serious S. aureus infection and the presence of genes encoding Panton-Valentine Leukocidin (pvl). Methods: A prospective study was conducted in a 1000-bed provincial hospital in NE Thailand over 1 year to identify patients with sterile-site cultures positive for S. aureus. Detailed clinical data and outcomes were recorded. Susceptibility testing was performed to determine methicillin susceptibility, and the presence of pvl was defined using PCR. Results: 248 patients were identified with one or more sterile-site cultures positive for S. aureus, 210 (85%) of which were caused by methicillin-susceptible S. aureus (MSSA) and 38 (15%) were caused by methicillin-resistant S. aureus (MRSA). 120/248 isolates (48%) were positive for pvl, all of which were MSSA. Genotyping of a proportion of isolates using multilocus sequence typing demonstrated a polyclonal population. pvl was positively associated with pyomyositis (p=0.02), skin and soft tissue abscesses (p<0.001) and other skin and soft tissue infections (p=0.02), but was not associated with lung infection. pvl was negatively associated with bacteraemia without a localising site of infection (p<0.001). There was no significant difference in the white cell counts of pvl-positive and pvl-negative cases. Overall mortality was 25%, with a highly significant negative association between mortality and pvl positivity: death occurred in 11% of pvl-positive cases versus 38% in pvl-negative cases (p<0.001). Conclusions: pvl is strongly associated with skin and soft tissue infection but is not associated with pneumonia, low white cell counts or a high mortality in our setting.