Validation of the extended-spectrum b-lactamase phenotype with molecular methods
Abstract number: P880
Bruderer T., Poirel L., Adler H., Nordmann P., Frei R.
Objectives: In order to validate extended-spectrum b-lactamase (ESBL) screening and confirmation assays, we determined the presence of 6 different ESBL-encoding genes in a representative collection of Swiss isolates showing an ESBL phenotype. In the subset of isolates in which no ESBL genes were detected, the presence of SHV-1 hyperproducing penicillinase, group D oxacillinases, and plasmid-mediated AmpC b-lactamases was additionally assessed.
Methods: The clinical Enterobacteriaceae isolates expressing an ESBL phenotype were consecutively collected at the University Hospital Basel (n = 39) or obtained from several Swiss laboratories (n = 33). As phenotypic confirmatory test, three different ESBL Etest strips (AB Biodisk, Sweden) were used: ceftazidime, cefotaxime, and cefepime, each plus/minus clavulanic acid. Repetitive patient isolates and K1 hyperproducing Klebsiella oxytoca were excluded. To determine the molecular types, PCR amplifying SHV, TEM, CTX-M, VEB, PER, GES, OXA-1, OXA-10 genes as well as genes of the SHV-1 promoter region and of 6 different phylogenetic groups of AmpC was performed and amplicons were sequenced.
Results: Among 162 isolates showing an ESBL phenotype, molecular analysis demonstrated that 103 (63.6%) expressed a CTX-M type, 23 (14.2%) a SHV ESBL-type, 17 (10.5%) a TEM ESBL type, one GES-1, and one PER-1 ESBL. Among the 21 (13.0%) isolates in which no ESBL-genes could be detected, 12 Klebsiella pneumoniae and 6 Escherichia coli harboured the typical C to A mutation in the second position of the SHV-1 -10 promoter region associated with hyperproduction of SHV-1. These 18 isolates showed a characteristic susceptibility and ESBL screening/confirmation pattern that differed significantly from all molecularly confirmed ESBL isolates except for GES-1. The remaining two E. coli and one Citrobacter koseri/amalonaticus contained an OXA-1 gene explaining the false positive cefepime/cefepime + clavulanic acid Etest.
Conclusions: Almost all phenotypic ESBL isolates in which no ESBL genes could be detected were SHV-1 hyperproducers known leading to false positive results in phenotypic ESBL tests. The relatively high number of false positive results might be explained by the low incidence of ESBLs in Switzerland. Because of their unique characteristics determined in this study, SHV-1 hyperproducers can be excluded.
|Session name:||18th European Congress of Clinical Microbiology and Infectious Diseases|
|Back to top|