Rapid, simple multi-locus variable number tandem repeat analysis: a reliable tool for Klebsiella pneumoniae outbreak
C Legouge1, P Bidet2, M Gits-Muselli2
1Service de Microbiologie, Centre National de Référence Associé Escherichia coli, Hôpital Robert-Debré, AP-HP, Paris, France.
Background:
Klebsiella pneumoniae causing nosocomial infections is increasingly multi-drug-resistant. Rapid and efficient typing tools are required for monitoring.
Aim:
To assess a simple, rapid (<5 h) multiplex polymerase chain reaction method based on multi-locus variable number tandem repeat analysis (MLVA) as a screening tool to determine whether or not K. pneumoniae strains are related.
Methods:
The global discriminatory power of the method was assessed on 72 unrelated K. pneumoniae isolates, including community carriage isolates, highly virulent strains causing liver abscess, and extended-spectrum beta-lactamase- and carbapenemase-producing strains. Suspected related strains from a suspected outbreak and a relapsed meningitis case were also studied. MLVA results were compared with whole-genome sequencing (WGS) analysis and multi-locus sequence typing (MLST).
Findings:
MLVA and MLST had similar discriminatory power, each distinguishing 54 profiles among the 72 unrelated isolates (Hunter-Gaston index 0.989). Each strain belonging to one sequence type (ST) or ST complex had its own MLVA type, with few exceptions. Two strains of ST268 and ST1119 shared the same MLVA profile, and two unrelated strains of ST307, ST86, ST45 and ST37 exhibited two different MLVA types each. Moreover, investigation of seven grouped cases of K. pneumoniae neonatal sepsis pointed to strong suspicion of a common source for five isolates, while two isolates with a different MLVA profile were excluded from this cluster.
Conclusion:
The MLVA approach is a useful, rapid and reliable tool for epidemiological investigation requiring only basic molecular biology equipment, and permits identification of sporadic isolates that are not part of an outbreak. However, analysis of strains sharing the same MLVA type by a highly discriminatory technique, such as WGS, remains necessary.
