|Home | About | Journals | Submit | Contact Us | Français|
Tigecycline is a glycylcycline antibiotic that inhibits protein synthesis, including that in isolates that are resistant to the tetracyclines by ribosomal protection or efflux. Infrequently, strains of enterobacteria have been reported as resistant to tigecycline due to overexpression of multidrug efflux pumps such as AcrAB (3, 4, 5). This report describes the development and mechanism of resistance to tigecycline during therapy for an Enterobacter hormaechei infection associated with lack of clinical response.
A 47-year-old man with end-stage hepatic failure was admitted to our hospital for a transplantation in February 2007. Ten days after liver transplantation, an E. cloacae strain was isolated from bile, urine, and tracheal secretions. The strain was resistant to all standard antibiotics, except amikacin, imipenem, and tigecycline (European Committee on Antimicrobial Susceptibility Testing breakpoints). The patient received imipenem and amikacin. Subsequently, various courses of antimicrobials were administered in relation to superinfections due to Stenotrophomonas maltophilia and coagulase-negative staphylococci. Further, the patient developed a pancytopenia which led to withdrawal of antimicrobials. During the following 6 weeks, the patient remained untreated. However, isolation of an E. cloacae strain, called UCN62, from blood cultures and two hepatic abscesses led to treatment with tigecycline (MICs, 1 μg/ml by Etest and 2 μg/ml by broth microdilution) at 50 mg twice daily. MICs of antibiotics for UCN62 are shown in Table Table1.1. After 3 weeks of treatment, the patient did not improve, and an isolate from bile, E. cloacae UCN63, displayed reduced susceptibility to tigecycline (MICs, 3 μg/ml by Etest and 4 μg/ml by broth microdilution; intermediate per FDA guidelines or resistant per European Committee on Antimicrobial Susceptibility Testing guidelines). In the absence of clinical improvement, the antibiotic treatment was stopped and the patient died 1 month later, against a background of sepsis and malnutrition.
SpeI restriction profiles of the DNA from the two strains were undistinguishable by pulsed-field gel electrophoresis. Ribotyping confirmed this result and identified the isolates as Enterobacter hormaechei subsp. steigerwaltii ribotype A. The difference in MICs between the pretreatment and the posttreatment isolates was confirmed by three independent determinations.
The expression of acrAB was studied by reverse transcription-PCR (6). The acrA primers are based on the sequence of the gene of strain UCN62. The following primers were used: (i) acrA gene reference, forward, 5′-GTGGGTGCCGATAACA-3′, and reverse, 5′-CCAGGACGAACTTTTTGC-3′; (ii) rrs gene reference, forward, 5′-AGTAGGGCTACACACG-3′, and reverse, 5′-CGATTACTAGCGATTCCG-3′. The rrs (16S rRNA) genes were used as a reference.
Quantitative analysis revealed that expression of acrA increased ca. 8-fold in E. hormaechei UCN63 over that in UCN62.
For this patient infected at different sites by a multiply antibiotic-resistant strain, tigecycline was used as the last resort. However, no improvement in the clinical status of the patient could be observed. In addition, a mutant with intermediate susceptibility was selected. The fatal outcome did not clearly correlate with microbiological failure since the patient was critically ill at the time of tigecycline administration.
Emergence of tigecycline resistance among Acinetobacter baumannii and Klebsiella pneumoniae strains during therapy has been recently reported (1, 2, 7, 8, 9). However, the mechanism of resistance has not been investigated in the posttreatment strains. In this study, an increase in MIC correlated with overexpression of the multidrug efflux pump AcrAB.
We thank P. H. D. Grimont for ribotyping.
Published ahead of print on 8 September 2009.