Comparison of the MICs of cefepime for extended-spectrum beta-lactamase-producing and non-extended-spectrum beta-lactamase-producing strains of Enterobacter cloacae

Sir, Extended-spectrum β-lactamase (ESBL)-producing strains of Enterobacteriaceae are a cause of increasing concern worldwide. For example, 23% of Klebsiella pneumoniae isolates recovered in European intensive care units have been reported to be ESBL producers.1 Outbreaks of infection caused by ESBL-producing strains of Enterobacter cloacae and Enterobacter aerogenes have also been described, albeit less frequently.2 Cefepime is a fourth-generation cephalosporin which exhibits poor substrate affinity for Bush group 1 chromosomally mediated β-lactamases.3 It also has excellent in vitro activity against Enterobacter spp., including constitutively derepressed strains resistant to third-generation cephalosporins such as ceftazidime and cefotaxime, and is effective in treating infections caused by these bacteria.4 The MICs of cefepime for clinical isolates of Escherichia coli harbouring a variety of ESBL TEM or SHV insertions tend to be high, but they are still within the susceptible range.4 On the basis of these MIC data and those for K. pneumoniae isolates, some authorities have suggested that fourth-generation cephalosporins are suitable for treating infections caused by ESBL-producing strains.5 However, depending on the β-lactamase, high percentages of clinical isolates of ESBL-producing E. coli and K. pneumoniae strains have been reported to be non-susceptible (intermediate or resistant) to these agents.6 There are few published data concerning the in vitro activities of fourth-generation cephalosporins against ESBL-producing strains of Enterobacter spp. At Concord Hospital we have recently experienced an upsurge of infections, including bacteraemias, caused by ESBL-producing strains of E. cloacae. In common with others,7 we observed a predictable pattern of resistance to trimethoprim and gentamicin among these isolates. We used this characteristic, together with an adaptation of the Jarlier double-disc potentiation method,8 with both cefotaxime and cefepime discs placed adjacent to a co-amoxiclav disc, to differentiate between ESBL- and non-ESBL-producing strains among 51 non-replicate clinical isolates of E. cloacae. We then evaluated the in vitro activity of cefepime against these organisms. The 51 strains were assigned to three groups on the basis of their susceptibilities to cefotaxime (susceptible, MIC ≤ 0.5 mg/L; resistant, MIC > 2 mg/L9) and probable resistance mechanisms; the MICs of cefotaxime were determined by a standard agar dilution method. The first group comprised 17 cefotaxime-susceptible strains. The second group comprised 13 cefotaxime-resistant, non-ESBLproducing strains. The MICs for 11 of the isolates in the latter group were ≥8 mg/L and it is likely that resistance to cefotaxime was mediated by stably derepressed chromosomal β-lactamases. The 21 isolates in the third group were cefotaxime-resistant ESBL producers, the MICs for 17 of which were ≥8 mg/L. The susceptibilities of the 51 strains to cefepime were determined by the Etest method according to the manufacturer's instructions. The results of the susceptibility tests are summarized in the Figure. The MICs for strains in group 1 were predictably low (MIC50 0.05 mg/L, MIC90 0.09 mg/L, range 0.023–0.125 mg/L). Those for isolates in group 2 were higher than those in group 1 (MIC50 0.5 mg/L, MIC90 1.5 mg/L, range 0.047–3 mg/L) but, according to the MIC breakpoint for susceptibility to cefepime (≤8 mg/L) recommended by the National Committee for Clinical Laboratory Standards (NCCLS),10 all of the isolates in both groups were susceptible to the drug. In contrast, the MIC50, MIC90 and range of MICs for the 21 isolates in group 3 were 16, >256 and 0.75–>256 mg/L, respectively and only nine (43%) were categorized as susceptible; four strains, the MICs for three of which were >256 mg/L, were categorized as resistant (MIC ≥ 32 mg/L10). Outbreaks caused by ESBL-producing strains of Enterobacter spp. have originated predominantly in intensive care units. In common with those caused by ESBL-producing strains of K. pneumoniae, control of these outbreaks has relied upon restricting the use of third-generation cephalosporins and the administration of targeted therapy to infected patients. Cefepime is a broad-spectrum antibiotic that has particular potential as empirical antibiotic treatment of severely ill patients and the authors of recent reviews of the management of patients with infections caused by Enterobacter spp. isolates have included it in their therapeutic armamentaria because of its clinical efficacy.4 However, these reports did not address specifically the ESBL status of the isolates and whether or not fourth-generation cephalosporins should be avoided as treatment of patients with infections caused by strains of Enterobacter spp. if they produced ESBLs. Based on the MIC data for E. coli and K. pneumoniae isolates, opinions have varied regarding the suitability of using these drugs as treatment of patients with infections caused by ESBL-producing isolates.5,6 Our results show that the MICs of cefepime for ESBL-producing strains of E. cloacae are even higher than those for ESBL-producing isolates of E. coli and K. pneumoniae. This may be conditional upon the particular ESBL being produced or it may represent a congruence of stably derepressed AmpC and extended-spectrum β-lactamases. The results of our investigation have led us to conclude that fourth-generation cephalosporins should not be used for treating serious infections caused by ESBL-producing strains of E. cloacae until the outcome of susceptibility testing is known. Distribution of MICs of cefepime for 51 clinical isolates of E. cloacae classified into three groups according to the MIC of cefotaxime and ESBL production: □, MIC ≤ 0.5 mg/L, non-ESBL-producing (n = 17); ▪, MIC > 2 mg/L, non-ESBL-producing (n = 13); , MIC > 2 mg/L, ESBL-producing (n = 21).