Outbreak Caused by a Multidrug-Resistant Klebsiella pneumoniae Clone Carrying bla_VIM-12 in a University Hospital

Abstract

From November 2006 to April 2007, nine nonrepetitive isolates of Klebsiella pneumoniae with reduced susceptibility or resistance to carbapenems were recovered from clinical specimens from separate patients hospitalized in a tertiary care hospital. The imipenem-EDTA synergy test was positive for all isolates. PCR, sequencing, and transferability experiments revealed the novel bla_VIM-12 metallo-β-lactamase gene, which was plasmid mediated and located in a class 1 integron. Pulsed-field gel electrophoresis demonstrated a single macrorestriction pattern, indicating the clonal spread of VIM-12-producing K. pneumoniae.

Acquired metallo-β-lactamases (MBLs) constitute a growing class of β-lactamases that readily hydrolyze most β-lactam antibiotics, including imipenem and meropenem (20). They are usually embedded in class 1 integron structures carrying various gene cassettes and belong to the group B β-lactamases according to the classification scheme proposed by Ambler (1). Five different groups of acquired MBLs have been described (IMP, VIM, SPM, GIM, and SIM enzymes), but the first two of these groups have predominated in gram-negative organisms (8, 9, 13, 20). VIM-type MBLs are commonly encountered among nonfermenters in the Far East and southern Europe (14, 15, 18, 21, 22), but during the last few years, the respective bla genes have also spread in members of the family Enterobacteriaceae (11, 20, 23). Reports of VIM-type MBL-producing K. pneumoniae isolates from Greek hospitals are increasing (5, 7). Recently, a novel VIM-type MBL variant, VIM-12, was identified from a clinical isolate of Klebsiella pneumoniae in Greece (16). In this report, we describe an outbreak caused by K. pneumoniae harboring the bla_VIM-12 gene.

MATERIALS AND METHODS

From November 2006 to April 2007, nine nonrepetitive K. pneumoniae isolates exhibiting reduced susceptibility or resistance to imipenem were isolated from clinical specimens collected from separate patients hospitalized in different medical and surgical wards at Hippokration General Hospital, Thessaloniki, Greece. All patients had a long-term hospitalization and had been treated with multiple courses of antimicrobials, including carbapenems. Identification of the isolates to the species level and initial antibiotic susceptibility testing were performed by the Vitek-2 automated system (BioMérieux, Marcy l'Étoile, France), by which they were characterized as imipenem intermediate or resistant.

Susceptibilities to a range of antimicrobials (amikacin, amoxicillin-clavulanate, ampicillin, aztreonam, cefepime, cefotaxime, cefoxitin, ceftazidime, ciprofloxacin, cotrimoxazole, ertapenem, imipenem, meropenem, and tetracycline) were also determined by disk diffusion (3). MICs of imipenem, meropenem, ertapenem, and tigecycline were additionally determined by Etest (AB Biodisk, Solna, Sweden) and the broth microdilution method (3). Susceptibilities were interpreted according to CLSI guidelines (3), when available. For tigecycline, the U.S. FDA (Food and Drug Administration) recommendation was used (susceptible, ≤2 μg/ml; resistant, ≥8 μg/ml). Possible MBL production was tested by using the imipenem-EDTA double-disk synergy test (2, 14). Escherichia coli strains ATCC 25922 and 35218 and known VIM-12-producing E. coli strain 28 (6) were used as controls in all susceptibility assays.

The isolates were tested by PCR for the bla_VIM gene with a set of primers that amplify an internal fragment of the gene (18). For sequencing, primers amplifying the entire bla_VIM gene were also used (21). Additionally, PCRs were performed with all of the isolates for the bla_IMP MBL gene (18), as well as for the extended-spectrum β-lactamase-encoding genes bla_TEM, bla_CTX-M, and bla_SHV-5-type and the AmpC type-encoding gene bla_CMY (12, 19). Detection and mapping of a class 1 integron were carried out with primers specific for 5′ and 3′ conserved segments amplifying the variable region containing the resistance gene cassette (10). Nucleotide sequencing of both strands of the PCR products derived with primers amplifying the whole bla_VIM gene and the class 1 integron was performed with an ABI Prism 377 DNA sequencer (Applied Biosystems Division, Perkin-Elmer, Foster City, CA).

The K. pneumoniae clinical isolates studied were subjected to conjugation experiments with E. coli 26R793 (lac Rif^r) as the recipient (17). Selection of transconjugants was done on MacConkey agar plates containing rifampin at 100 μg/ml and ceftazidime at 2 μg/ml. Transconjugants were analyzed for plasmids by an alkaline lysis procedure. The plasmid DNA band was extracted from the agarose gel with the QIAquick gel extraction kit (Qiagen GmbH, Hilden, Germany) and was used as template DNA in a PCR for the detection of bla_VIM. The expression of the bla_VIM gene was tested by reverse transcriptase PCR-specific amplification with DNase-treated RNA extracts obtained with an RNeasy mini kit (Qiagen) and 16S rRNA as a control as described previously (7), with the same RNA extracts for amplification of both the 16S rRNA and bla_VIM genes.

The relatedness of the K. pneumoniae isolates was analyzed by pulsed-field gel electrophoresis (PFGE) with XbaI-digested genomic DNA and a CHEF-DRIII system (Bio-Rad, Hemel Hempstead, United Kingdom) as described elsewhere (18). XbaI macrorestriction patterns were compared visually.

RESULTS

The nine K. pneumoniae isolates studied were recovered from blood cultures, wounds, and bronchial secretions (two cases each) and from urine, sputum, and drainage fluid (one case each). The ages of the patients ranged from 23 to 90 years; they were admitted for several reasons and were hospitalized in six different wards of the hospital. Four of them had received carbapenems prior to the isolation of the bla_VIM-12-producing K. pneumoniae. Table 1 summarizes the characteristics of the patients and the origins of the isolates.

TABLE 1.

Origins of bla_VIM-12-containing K. pneumoniae isolates and clinical characteristics of the patients

Patient no.	Reason for admission	Warda	Source	Previous treatmentb	Collection date (mo/day/yr)	Outcome	Age (yr)/sexc
1	Polytraumatism	Orthopedic	Wound	CIP, CAZ, IPM	11/02/06	Recovery	38/M
2	Polytraumatism	Orthopedic	Blood	SAM, MET	11/15/06	Recovery	40/M
3	Mesenteric embolism	Surgery	Drainage fluid	MET, TCC	12/03/06	Death	90/M
4	Intoxication	ICU	Blood	TZP, CLI, MEM	12/12/06	Recovery	72/F
5	Polytraumatism	ICU	Wound	IPM, CIP, TCC, MET	01/02/07	Death	24/M
6	Hydrocephalus	Neurosurgery	Bronchial secretions	CAZ, IPM	02/03/07	Recovery	36/M
7	Renal transplantation	Transplant	Sputum	NOR	02/26/07	Recovery	32/M
8	Intracranial hemorrhage	ICU	Bronchial secretions		03/15/07	Death	41/M
9	Encephalitis	Neurology	Urine	AMK, SXT	04/06/07	Recovery	23/M

^aICU, intensive care unit.

^bAMK, amikacin; CAZ, ceftazidime; CLI, clindamycin; CIP, ciprofloxacin; IPM, imipenem; MEM, meropenem; MET, metronidazole; NOR, norfloxacin; SAM, ampicillin-sulbactam; SXT, cotrimoxazole; TCC, ticarcillin-clavulanate; TZP, piperacillin-tazobactam.

^cM, male; F, female.

By the automated system, all of the isolates were initially characterized as intermediately resistant or resistant to imipenem and meropenem (Table 2). By disk diffusion testing, five of the isolates were susceptible to all carbapenems, three were susceptible to imipenem and meropenem but intermediately resistant to ertapenem, and one was resistant to all three carbapenems. In accordance with disk diffusion, the Etest and broth microdilution MICs for all of the isolates but one were in the susceptible range for imipenem and meropenem (1 to 4 μg/ml for imipenem and 0.5 to 4 μg/ml for meropenem) and ranged from 0.5 to 16 μg/ml for ertapenem, while the remaining isolate had MICs of >32 μg/ml for all three carbapenems (Table 2). The isolates were also susceptible to tigecycline, whose MICs ranged from 0.032 to 2 mg/liter. All of the isolates exhibited synergy with EDTA in the imipenem-EDTA double-disk synergy test, indicating MBL production. The susceptibilities of the isolates to other antimicrobial agents by the disc diffusion method are reported in Table 3. All of the isolates were found resistant or intermediately resistant to ampicillin, amoxicillin-clavulanate, aztreonam, expanded-spectrum cephalosporins, ciprofloxacin, amikacin, cotrimoxazole, and tetracycline.

TABLE 2.

Susceptibilities of the nine K. pneumoniae isolates studied to carbapenems by different assays

K. pneumoniae isolate no.	Etest MICa			Broth microdilution MICa			Disk diffusionb			Vitek-2b
	IPM	MEM	ERT	IPM	MEM	ERT	IPM	MEM	ERT	IPM	MEM
1	3	2	4	4	2	4	S	S	S	I	I
2	4	3	12	4	4	16	S	S	S	I	I
3	4	3	6	4	2	8	S	S	S	I	I
4	3	4	3	4	4	4	S	S	S	I	I
5	>32	>32	>32	128	64	128	R	R	R	R	R
6	3	3	6	4	2	4	S	S	S	R	R
7	1	0.5	0.5	1	1	0.5	S	S	S	I	I
8	2	1.5	2	2	2	2	S	S	S	I	I
9	2	1	2	2	1	2	S	S	S	R	R

^aMICs are in μg/ml. ERT, ertapenem; IPM, imipenem; MEM, meropenem.

^bR, resistant; I, intermediately resistant; S, susceptible.

TABLE 3.

Antibiotic susceptibilities of the nine K. pneumoniae isolates studied to various antimicrobials by the disk diffusion method^a

K. pneumoniae isolate no.	AMP	AMC	ATM	CAZ	CTX	SXT	FOX	CIP	AMK	FEP	TET
1	R	R	I	R	R	R	R	R	I	I	R
2	R	R	I	R	R	R	R	R	I	I	R
3	R	R	R	R	R	R	R	R	R	I	R
4	R	R	I	R	R	R	R	R	I	I	R
5	R	R	R	R	R	R	R	R	I	I	R
6	R	R	I	R	R	R	R	I	R	I	R
7	R	R	I	R	R	R	R	I	I	R	R
8	R	R	R	R	R	R	R	R	I	I	R
9	R	R	R	R	R	R	R	R	R	I	R

^aR, resistant; I, intermediately resistant; S, susceptible. AMP, ampicillin; AMC, ampicillin-clavulanate; AMK, amikacin; ATM, aztreonam; CAZ, ceftazidime; CIP, ciprofloxacin; CTX, cefotaxime; ERT, ertapenem; FEP, cefepime; FOX, cefoxitin; IPM, imipenem; MEM, meropenem; SXT, cotrimoxazole; TET, tetracycline.

The nine K. pneumoniae isolates were PCR positive for the bla_VIM gene with both sets of consensus primers. DNA sequencing of the entire bla_VIM amplicons showed that all of the isolates carried the novel bla_VIM-12 MBL gene. The structure of the bla_VIM-12-containing integron was characterized in three representative K. pneumoniae isolates. The bla_VIM-12 gene was inserted as a gene cassette into a class 1 integron. This gene cassette, with its truncated 59-base element, was part of a 5′-to-3′ region and was flanked by two copies of an aacA7 gene cassette which confers resistance to aminoglycosides. A strong P1 promoter and an activated P2 promoter were the promoters of the 5′ conserved segment sequence. Reverse transcriptase-PCR confirmed that, irrespectively of the carbapenem MICs, the bla_VIM-12 gene was expressed in all nine isolates. PCR was negative for the bla_IMP, bla_SHV-5-type, bla_CTX-M, and bla_TEM β-lactamase genes but positive for bla_CMY in all of the isolates.

Conjugation experiments showed that the ceftazidime resistance determinant was successfully transferred to E. coli 26R793 from all nine isolates at transfer frequencies ranging from 1.7 × 10⁻² to 3.4 × 10⁻⁵ per recipient cell. Susceptibility testing confirmed the transfer of resistance to β-lactams, as well as resistance to amikacin and cotrimoxazole. The molecular sizes of the plasmids transferred from each isolate to E. coli 26R793 were approximately 70 kb. When the plasmid bands of the transconjugants were extracted from the gel and used as templates for the amplification of the bla_VIM and bla_CMY genes, the specific products were detected, suggesting that both resistance determinants resided in this transferable plasmid. Sequencing of both strands of the entire bla_VIM gene amplicons derived from the extracted plasmidic DNA confirmed the presence of bla_VIM-12 in the transconjugants. PFGE analysis showed that all of the clinical isolates belonged to a single clone producing the same macrorestriction pattern (Fig. 1).

FIG. 1.

PFGE of XbaI-digested genomic DNA from the nine nonrepetitive bla_VIM-12-positive K. pneumoniae isolates.

DISCUSSION

This report describes an outbreak caused by a single clone of K. pneumoniae carrying the novel plasmid-mediated bla_VIM-12 carbapenem-hydrolyzing β-lactamase gene. bla_VIM-12 is clustered between bla_VIM-1 and bla_VIM-2; it differs from bla_VIM-1 by eight nucleotides that are all located at the 3′ end and matches exactly the nucleotides found at the corresponding positions in the bla_VIM-2 gene (16). The spread of VIM-12-producing K. pneumoniae in our hospital might be related to the frequent transfer of patients between wards. However, other factors, such as fecal or skin carriage by staff or patients, should not be ignored. Three of the patients had received imipenem and one had received meropenem prior to the isolation of the VIM-12-producing K. pneumoniae strain. Treatment of two of the above patients with imipenem was also continued, given the subsequent laboratory reports (Etest imipenem MIC of 3 μg/ml). The third patient (patient 5) died because of severe injuries soon after the recovery of the imipenem-resistant K. pneumoniae isolate. In most cases, colistin was deemed one of the antimicrobial agents of choice according to the results of the automated system (MICs of ≤0.5 μg/liter). It should be mentioned here that we observed some discordant carbapenem susceptibility results between our automated system and the other susceptibility assays. It is of note that similar discordances have been described previously for automated systems, which were also found to overestimate resistance rates (4). Strict control measures such as isolation of infected patients, hand hygiene, and restriction of carbapenem use as monotherapy were taken, and the outbreak strain was not recovered from other patients in the hospital after the initial outbreak.

K. pneumoniae is recognized as an important pathogen mainly involved in nosocomial infections and outbreaks. The spread of MBL-encoding genes in K. pneumoniae and the emergence of new transferable MBL variants in our region may reflect a heavy usage of carbapenems for the treatment of multiresistant Acinetobacter baumannii and Pseudomonas aeruginosa infections, which are common in Greek hospitals. This evolution is worrisome and leaves limited choices for treatment regimens. The accurate and rapid detection of these MBL-producing pathogens is necessary for therapeutic considerations and infection control purposes in order to prevent further dissemination.