The spread of carbapenem-hydrolysing β-lactamases has become a major issue in the healthcare environment, underpinning the need for rapid and accurate detection of carbapenemases amongst Gram Negative Baciilli (CRO). However, routine susceptibility methods for detecting these resistance mechanisms are not reliable, largely due to the variety of mechanisms present and the high MIC breakpoints for carbapenems, thus emphasising the need for additional testing and/or confirmation.
This laboratory receives a large number of referrals of suspected CRO’s for confirmation and uses molecular methods to confirm suspected resistance mechanisms deemed from screening through the BD Phoenix. The BD Phoenix susceptibility panels are adapted to testing low levels of antibiotics which make it ideal as an early warning system for detecting potential resistances below those defined by breakpoints for CLSI and EUCAST and to indicate the possible resistance mechanism present for more specific molecular determination of mechanisms present. The vast majority of isolates referred are from Vitek users.
1202 GNB were referred to us for antimicrobial resistance screening, of which 246 were predicted as possible CRO by the BD Phoenix, and confirmed by molecular methods. Resistance mechanisms included 155 IMP-4, 1 IMP-15, 1 IMP-18, 1 IMP-32, 10 VIM, 13 NDM, 7 KPC, 7 OXA-48, 8 OXA-23 and 43 GES-2/5. The remaining 956 isolates were not CROs, as predicted by the BD Phoenix.
It is evident that no one marker serves as a universal predictor for CROs, and that multiple low level antibiotics are needed, including cefotaxime/ceftazidime, cefoxitin and meropenem/ertapenem, amongst others. It is also evident that the breakpoints for CLSI are too high to accurately predict CRO and provide a high level of false susceptibility, necessitating additional testing. EUCAST breakpoints, while showing improvement, are not sufficiently accurate. In all cases, the BD Phoenix using the NMIC-203 panel was able to accurately and reliably predict CROs.