Abstract :
[en] Background:
The use of temocillin (TEM) is increasing in serious infections caused by Enterobacteriaceae, including extended-spectrum β-lactamase (ESBL), as an alternative to carbapenems. Accuracy of in vitro minimal inhibitory concentration (MIC) values is of high importance in an era of antibiotic stewardship based on PK/PD. We performed a method comparison between Vitek2®, E-test® and the standard method, broth microdilution (BMD), in order to evaluate which method is reliable to determine TEM MIC.
Materials/methods:
One hundred Enterobacteriaceae were collected from respiratory samples isolated from ICU patients. MIC of temocillin was determined in parallel by 3 methods: Vitek2® (Biomérieux), E-test® (Biomérieux) and BMD, the latter following CLSI recommendations. Susceptibility to temocillin was determined according to breakpoints provided by BSAC (British Society for Antimicrobial Chemotherapy) - S: MIC ≤ 8 mg/L; R: MIC > 8 mg/L.- in order to evaluate categorical agreement (CA), essential agreement (EA), very major errors (VME) and major errors (ME), as defined in Cumitech 31A.
Results: Isolates (N=100) included: Klebsiella pneumoniae (34%), Escherichia coli (23%), Serratia sp. (18%), others (25 %). Thirty-five strains were ESBL-producers; 13% were carbapenemase-producers. 41/100 (41%) isolates were resistant to TEM according to BMD. EA of 96 % was calculated for E-test® and 95% for Vitek2® versus BMD, while CA was of 82% (E-test®) and 84% (Vitek2®). Both VME (12.2% for E-test®; 24.4% for Vitek2®) and ME (22.0% for E-test®; 10.2% for Vitek2®) exceeded the 3% threshold. As a part of errors can be explained by BSAC breakpoints (no “intermediate” category), VME and ME with MIC > ± 1 twofold dilution were calculated, for E-test® (VME:0%; ME: 6.8%) and for Vitek2® (VME: 7.3%; ME: 3.4%). Overestimation of TEM resistance (E-test®) or TEM susceptibility (Vitek2®) was observed.
Conclusions:
Compared to BMD, E-test® seems to be more reliable than Vitek2® to predict MIC of TEM. Contrary to E-test®, even when taking the adapted definition of VME and ME with MIC > ± 1 twofold dilution, VME rate for Vitek2® is still unacceptable. When the use of TEM is considered by the clinician, we would recommend to control systematically TEM MIC by E-test®, or, even better, by BMD.
