A fascinating study from a European research group has unravelled the molecular epidemiology of a large European collection of carbapenem-resistant Klebsiella pneumoniae clinical isolates. Most carbapenem resistance was due to an acquired carbapenemases, transmission clusters were common within and between hospitals, carbapenemase-producing isolates are more likely to spread in hospitals, and 21 SNPs is the magic number for defining CPE person-to-person transmission using WGS.

The study was based on a collection of 1,717 Klebsiella pneumoniae clinical isolates (55% carbapenem resistant, 45% carbapenem susceptible) from 244 hospitals in 32 countries across Europe. These isolates were collected as part of the European Survey of Carbapenemase-Producing Enterobacteriaceae (EuSCAPE) in 2013-2014. The key findings were:

• Most carbapenem resistance was due to an acquired carbapenemases (72% of 944 carbapenem-resistant isolates).
• Around 2/3 of the CPE were from the four main pneumoniae lineages (ST 11, 15, 101, and 258/512). There was limited core genome diversity within the lineages, suggesting relatively recent emergence and spread.
• Rather confusingly, there were actually 4 species within the “Klebsiella pneumoniae” bucket (including the rather ridiculously named “Klebsiella quasipneumoniae”). Almost all of the carbapenem resistance isolates were from pneumoniae sensu stricto.
• The group used the idea of the “genetically nearest neighbour (gNN)” to explore possible transmission clusters. They found that the gNN was most likely to be a patient from the same hospital for carbapenemase producers, but less so for non-carbapenemase-producers. In fact, there was an inverse relationship between the degree of carbapenem resistance as measured by MIC and the proportion of gNNs from same-hospital patients. This suggests that more transmission of carbapenamase-producers happens in hospitals than non-carbapenemase producers.
• 21 SNPs is the ‘magic number’ for defining CPE person-to-person transmission using WGS.

The study makes a bold conclusion – that ‘…although there is no denial that gene flow between the one-health compartments exists [i.e. animals, sewage], this phenomenon has had no substantial impact on the course of the epidemiology of the major carbapenemase-positive clones of K. pneumoniae in Europe…’. To make this conclusion, surely you’d need to include isolates from animals, sewage, and the environment – otherwise how could you be sure that you weren’t ‘missing links’? Nonetheless, the study is clear that nosocomial transmission of K. pneumoniae CPE driven by carbapenem exposure seems to be the predominant driver of CPE transmission and expansion.