Evaluating the accuracy of the MBT Lipid Xtract Kit for assessing colistin resistance in comparison to broth microdilution

Colistin resistance testing methods such as broth microdilution (BMD) are time-consuming and labour intensive for clinical laboratories. MBT Lipid Xtract Kit on MALDI Biotyper Sirius System (Bruker, Billerica, MA, USA) utilizes lipidomic analysis to identify specific cell wall modifications associated with colistin resistance. We compared MBT to BMD (ComASP Colistin, Liofilchem) across 36 Gram-negative isolates (non-resistant MIC ≤2 µg ml^-1, resistant MIC ≥4 µg ml^-1). All samples were tested twice on MBT with discrepant results repeated before assessing categorical agreement between MBT and BMD. 44.4% (16/36) of isolates were colistin resistant via BMD. MBT Lipid Xtract had 80.6% agreement (29/36) with BMD, with 5/7 discrepancies corrected to match upon repeat testing. There was 100% agreement for <i>Escherichia coli</i> isolates (<i>n</i>=16). The whole-genome sequencing was completed on the two discrepant <i>Klebsiella pneumoniae</i> isolates, with variants within colistin resistance-associated loci identified (MIC 0.5 µg ml^-1: <i>arnC</i> S30T, <i>pmrB</i> T246A, <i>lapB</i> N212T, <i>lpxM</i> S253G, <i>crrB</i> Q287K and MIC >16 µg ml^-1: <i>arnC</i> S30T, <i>pmrB</i> R90insRN, <i>pmrB</i> T246A, <i>pmrA</i> E57G, <i>lpxM</i> S253G). Further evaluation, particularly for non-<i>E. coli</i>, of MBT is required prior to implementation in clinical laboratories.