Objectives: Because of the emergence of plasmid-mediated (mcr-1 and mcr-2 genes) and chromosomally-encoded colistin resistance, reliable methods for detecting colistin resistance/susceptibility in routine laboratories are required. We evaluated the respective performances of the BD Phoenix automated system, the newly-developed Rapid Polymyxin NP test and the broth microdilution (BMD) reference method to detect colistin resistance in Enterobacteriaceae, and particularly those producing MCR-1 and MCR-2.Methods: Colistin susceptibility of 123 enterobacterial clinical isolates (40 colistinsusceptible and 83 colistin-resistant isolates) was tested with the Phoenix automated system, the Rapid Polymyxin NP test and the BMD method. Molecular mechanisms responsible for plasmid-mediated and chromosomally-encoded colistin resistance mechanisms were investigated by PCR and sequencing.Results: Considering BMD as a reference method, the Phoenix system failed to detect ten colistin-resistant isolates (one Escherichia coli, one Klebsiella pneumoniae, seven Enterobacter spp., and one Salmonella enterica). The Rapid Polymyxin NP test failed to detect the same single E. coli isolate. Those two latter methods detected the sixteen E. coli, K. pneumoniae and S. enterica isolates producing the plasmid-encoded MCR-1 and MCR-2.
Conclusion:The Phoenix system and the Rapid Polymyxin NP test are reliable techniques for detecting plasmid-mediated MCR-1 and MCR-2-related colistin resistance. However, a high rate of false susceptibility was observed with the Phoenix system, indicating that susceptibility results obtained with that system should be confirmed by BMD method. By contrast, the Rapid Polymyxin NP test showed a good agreement with the BMD method and results were obtained rapidly (within two hours). The BMD method should be performed if MIC values are needed.