No abstract

Sixty-two clinical isolates of Enterobacter aerogenes resistant to expanded-spectrum cephalosporins were collected between July 2003 and May 2005. Among these isolates, 23 (37.1%) were imipenem (IPM) susceptible, and 39 (62.9%) were IPM insusceptible, of which 89.7% (35/39) were resistant and 10.3% (4/39) were intermediate. Isolate genotypes were compared by pulsed-field gel electrophoresis. Of 62 isolates, 48 belonged to epidemic pulsotype A (77.4%). This pulsotype included 37.5% and 58.4% of ␤-lactam phenotypes b and a, respectively. Nine isolates (14.5%) belonged to pulsotype E, which included 22.3% and 77.7% of phenotypes b and a, respectively. The ␤-lactamases with pIs of 5.4, 6.5, 8.2, and 8.2 corresponded to extended-spectrum ␤-lactamases (ESBLs) TEM-20, TEM-24, SHV-5, and SHV-12, respectively. Of 39 IPM-insusceptible E. aerogenes isolates, 26 (66.6%) were determined to be metallo-␤-lactamase producers, by using a phenotypic method. Of these isolates, 24 harbored a bla IMP-1 gene encoding a protein with a pI of >9.5, and two carried the bla VIM-2 gene encoding a protein with a pI of 5.3, corresponding to ␤-lactamases IMP-1 and VIM-2, respectively. The remaining 13 (33.4%) isolates were negative for the bla IMP-1 and bla VIM-2 genes but showed an alteration of their outer membrane proteins (OMPs). Ten of these isolates produced the two possible OMPs (32 and 42 kDa), with IPM MICs between 8 and 32 g/ml, and three others produced only a 32-kDa OMP with IPM MICs >32 g/ml. This work demonstrates that, in addition to resistance to expanded-spectrum cephalosporins, IPM resistance can occur in ESBL-producing E. aerogenes isolates by carbapenemase production or by the loss of porin in the outer membrane.Enterobacter aerogenes has recently emerged as an important hospital pathogen (13). The prevalence of this bacterial species has increased considerably since the introduction of extendedspectrum cephalosporins into clinical practice (18). Various resistance mechanisms have been described in this species, such as extended-spectrum ␤-lactamases (ESBLs), plasmidmediated production (21), and hyperproduction of the Bush group 1 chromosomally mediated cephalosporinases (32). Although these enzymes are specifically characteristic of some Enterobacter spp., the appearance of similar plasmid-mediated ␤-lactamases in Klebsiella pneumoniae and Escherichia coli raises concerns about the spread of resistance (21, 32). For both ESBL-producing and AmpC-producing isolates, carbapenems are the only ␤-lactam agents active against both resistance mechanisms. Data from previous studies have shown that both ESBL and AmpC ␤-lactamase producers had reduced susceptibility to imipenem (IPM) due to either carbapenemhydrolyzing enzymes (carbapenemase) (30), decreased membrane permeability due to loss of porin in the outer membrane (14, 39), or active efflux (27).The IPM resistance of ESBL-producing E. aerogenes strains was observed for the first time at Amiens University Hospital (A.U.H.) in 2002. To determine whether the strains were epidem...

We present the first case of human spondylodiscitis due to Shewanella algae. Our patient did not have any predisposing factors. The portal of entry was probably a cutaneous lesion on the leg, exposed to seawater. Bacteria were isolated in pure culture from a needle biopsy specimen of the vertebral disk. Automated identification systems identified the organism as Shewanella putrefaciens. However, molecular biology identified it as S. algae. Treatment with ceftriaxone and amikacin, then ciprofloxacin successfully addressed the infection. We also review four published cases of human osteoarticular infections caused by Shewanella spp: two cases of arthritis and two cases of osteomyelitis. Two patients had predisposing factors, and contact with water was found in two cases. The clinical, radiological and biological characteristics of S. algae spondylodiscitis are indistinguishable from those of spondylodiscitis of other causes. A cutaneous lesion with exposure to water is a potential portal of entry. Molecular typing is necessary to obtain a precise bacteriological identification.

We conclude that (1) the interpretation of the results for LS-BMS is very easy, in contrast to LS-BS; (2) the diagnostic value of LS-BMS for detecting infected joint prostheses is good; and (3) additional data are needed to assess the accuracy of LS-BMS when the prosthesis has been removed.

No abstract