Optimization of the rapid carbapenem inactivation method for use with AmpC hyperproducers.

Abstract

Detection of carbapenemase-producing Enterobacterales (CPEs) is sometimes difficult with AmpC-hyperproducing Enterobacterales (AHEs), as they may falsely be classified as CPEs. Here, we present a rapid Carbapenem Inactivation Method (rCIM) optimized for AmpC producers (rCIM-A) that allows rapid and easy discrimination between AHEs and CPEs. Enterobacterales (n = 249), including natural AmpC producers, AHEs, CPEs and non-carbapenemase-producing carbapenem-resistant control strains were evaluated, using Carba NP, rCIM and rCIM-A. The rCIM-A differs from the rCIM by the addition of cloxacillin (400 μg/mL) to the initial antibiotic incubation step. The rCIM-A yielded a sensitivity and specificity of 84.26% (95% CI: 76.00%-90.55%) and 99.29% (95% CI: 96.11%-99.98%), respectively, while those of the rCIM were 86.11% (95% CI: 78.13%-92.01%) and 80.85% (95% CI: 73.38%-86.99%), respectively; those of Carba NP were lower at 84.04% (95% CI: 75.05%-90.78%) and 91.37% (95% CI: 85.41%-95.46%), respectively, due to indeterminate results. The rCIM-A was capable of discriminating between AHEs and true CPEs, but still failed to identify OXA-23-producing Proteus mirabilis isolates and remained only partially reliable for identifying IMI-like producers and a few MBL (2 NDM-1, 1 LMB-1, 1 TMB-1 and 1 IMP-13) producers. One chromosomally encoded AmpC variant, MIR-10, gave repeatedly positive results using all three tests and was thus considered a false positive. Specificity for AHEs greatly improved with the rCIM-A without altering the test performance for the other resistance mechanisms. It may replace the rCIM as a cheap, easy, rapid and accurate CPE detection test.