Carbapenemase-producing Klebsiella pneumoniae has become a worldwide recognized cause of nosocomial infections and requires urgent public health attention. The main reason of this concern is the increasing resistance to all the last-resort antibiotics, including colistin. The ideal methodology for colistin susceptibility testing still remains undefined. However, the emergence of colistin as one of the last-option treatments requires a reliable method to determine the susceptibility profile of resistant isolates. The aim of the present study was to evaluate the impact of detecting colistin resistance in Klebsiella pneumoniae isolates using MALDI-TOF MS in clinical routine practice. For this reason, 139 isolates of K. pneumoniae were collected during 2015-2017 from patients hospitalized at Pisa University Hospital. Colonies suspected to be colistin resistant were identified using MALDI-TOF MS (Bruker Daltonik GmbH, Bremen, Germany) following a protein extraction protocol. Strains were previously wholly genome sequenced. To create a customized database entry and to generate classifying algorithm models, 1.112 mass spectra were collected. In relation to their mass signals and intensities, a two dimensional peak distribution was created. The recognition capability of the algorithm based on two manually selected mass peaks was 91,8%, while cross validation was 87,6%. The proportion of correctly classified colistin-resistant K. pneumoniae was 91% and colistin-susceptible was 73%. The emergence of colistin-resistant Gram-negative organisms has a dramatic impact on patient outcomes. Our study, based on MALDI-TOF MS technology, offers rapid preliminary results on colistin resistance profile coupled with bacteria identification.
Keywords: Colistin resistance; Genetic algorithm; MALDI-TOF MS; Mass peaks; Proteome.
Copyright © 2018 Elsevier B.V. All rights reserved.