LETTER
Carbapenemase-producing Enterobacteriaceae (CPE) have been spreading worldwide and are a great concern among health care settings (1). Intriguingly, IMP-6 (encoded by the blaIMP-6 gene), one of the IMP-type metallo-carbapenemases, has been reported to confer the paradoxical imipenem-susceptible but meropenem-resistant (ISMR) phenotype to Enterobacteriaceae strains, and strains carrying blaIMP-6 have recently been isolated throughout Japan (2, 3). This phenotype might lead to misdiagnosis of CPE infections when imipenem is used for antimicrobial susceptibility testing, resulting in inappropriate antimicrobial use and failure to prevent nosocomial transmissions caused by this pathogen. The blaIMP-6 gene differs from the blaIMP-1 gene by one point mutation: adenine at nucleotide 640 in blaIMP-1 is replaced by guanine, corresponding to the amino acid substitution of glycine for serine (4). Although IMP-6 has weaker hydrolysis activity to imipenem as well as some kinds of β-lactams, such as benzylpenicillin and ceftazidime, than IMP-1 (5), CPE-carrying blaIMP-6 shows resistance to almost all the β-lactams except for imipenem, due to the concomitant production of CTX-M-2, an extended-spectrum β-lactamase (ESBL) (2, 3). Here, we investigated the antimicrobial activity of various antibacterials, including compounds recently marketed or currently under development, against the problematic ISMR strains carrying blaIMP-6.
Eighty-two ISMR phenotype Enterobacteriaceae strains (29 Escherichia coli, 22 Klebsiella pneumoniae, 20 Klebsiella oxytoca, 9 Enterobacter cloacae, and 2 Citrobacter freundii strains) isolated in a tertiary hospital in Japan from 2010 to 2014 were used. The existence of blaIMP-6 (guanine at nucleotide position 640) was confirmed by allele-specific PCR in all test strains (4), and blaCTX-M-2-group was detected by PCR in 80 of 82 strains (6). MIC was determined by the CLSI broth microdilution method (7), and the ISMR phenotype was defined as an imipenem MIC of ≤1 mg/liter and a meropenem MIC of ≥4 mg/liter (8). The MIC medium, iron-depleted cation-adjusted Mueller-Hinton broth prepared by treatment with Chelex (Bio-Rad, Hercules, CA), was used for cefiderocol (siderophore cephalosporin) (9).
The MIC90s of cefepime, ceftolozane-tazobactam, ceftazidime-avibactam, and ciprofloxacin were ≥32 mg/liter, while the MIC90s of colistin and amikacin were 0.5 and 8 mg/liter, respectively (Table 1). The two strains with resistance to colistin were K. pneumoniae and E. cloacae strains (MIC data by species are shown in the supplemental file). Recently marketed β-lactam–β-lactamase inhibitor combinations such as ceftolozane-tazobactam and ceftazidime-avibactam have been reported to be active against ESBL producers but inactive against metallo-β-lactamase producers (1, 10), and the results of the present study were consistent with those reports, although the IMP-6 enzyme has remarkable substrate specificity changes. However, the antibacterials in the compound under development, cefiderocol and aztreonam-avibactam, showed MIC90s of 1 and 0.25 mg/liter, respectively, against these ISMR strains, and there were no strains for which the MICs were ≥4 mg/liter. The reports that cefiderocol and aztreonam have quite low kcat/Km or relative Vmax/Km values for IMP-1 suggest that one reason for the good antimicrobial activities of both compounds is the low level of hydrolysis by IMP-6 as well as IMP-1 (5, 11, 12). Although colistin and amikacin showed potent activity against ISMR Enterobacteriaceae in this study, safety issues such as nephrotoxicity remain concerns in clinical settings (13, 14). Alternative treatment options for CPE infections, including those caused by IMP-6-producing strains, are needed.
Susceptibility distribution of 82 imipenem-susceptible but meropenem-resistant Enterobacteriaceae strains carrying blaIMP-6
ACKNOWLEDGMENTS
MIC data collection was supported by T. Hori and Y. Jinushi, Shionogi TechnoAdvance Research Co., Ltd.
We thank A. Naito and T. Yamaguchi for useful advice.
All authors are employees of Shionogi & Co., Ltd., and we have no conflicts of interest to declare.
FOOTNOTES
- Accepted manuscript posted online 24 April 2017.
Supplemental material for this article may be found at https://doi.org/10.1128/AAC.00576-17 .
- Copyright © 2017 American Society for Microbiology.