Risk Factors for Multidrug Resistance in Nosocomial Bacteremia Caused by Extended-Spectrum β-Lactamase-Producing Escherichia coli and Klebsiella pneumoniae

Abstract

Increasing multidrug resistance (MDR) among extended-spectrum β-lactamase-producing Escherichia coli and Klebsiella pneumoniae (ESBL-EK) is of a great concern, because the therapeutic options are severely limited. Thus, we performed a case-control study to evaluate risk factors for MDR among nosocomial bacteremia caused by ESBL-EK. All adult patients with ESBL-EK bacteremia from January 2009 through December 2010 were identified at our institution. MDR was defined as ESBL-EK that demonstrated in vitro resistance to trimethoprim-sulfamethoxazole (TMP-SMX), fluoroquinolone (FQ), and gentamicin. Case patients were those with an MDR ESBL-EK isolate, and control patients were those with a non-MDR ESBL-EK isolate. Among a total of 123 ESBL-EK isolates (74 [60.2%] E. coli and 49 [39.8%] K. pneumoniae) causing nosocomial bacteremia, 33 (26.8%) cases were due to MDR ESBL-EK. In a univariate analysis, the factors significantly associated with acquisition of MDR ESBL-EK were neutropenia, immunosuppressant use, urinary tract infection, and prior use of antibiotics, especially FQ (all p<0.05). A multivariable analysis showed that a prior receipt of FQ (odds ratio [OR]=2.93; 95% confidence interval [CI]=1.07–8.01; p=0.036), percutaneous tube insertion (OR=4.04; 95% CI=1.56–10.75; p=0.005), and neutropenia (OR=4.22; 95% CI=1.56–11.45; p=0.005) were independent risk factors for MDR among ESBL-EK bacteremia in hospitalized patients. The CTX-M-15 enzyme was predominant in both the MDR ESBL-EK and non-MDR ESBL-EK groups (55% [11/20] vs. 55.6% [15/27]). Our data suggest that strategies designed to reduce MDR in ESBL-EK bacteremia should focus on limiting the use of FQ and minimizing invasive procedures such as tube insertion.

Introduction

E scherichia coli and Klebsiella pneumoniae are major pathogens known to cause urinary tract infection, intra-abdominal infection, and primary bacteremia in both hospital and community settings. During the past three decades, antimicrobial-resistant strains that produce extended-spectrum β-lactamases (ESBLs) have emerged among Enterobacteriaceae, predominantly among E. coli and K. pneumoniae, and have become endemic in hospitals at varying levels of intensity.¹² For serious infections caused by these pathogens, carbapenems have been the most reliable treatment option, but isolation of carbapenem-resistant Enterobacteriaceae has been increasingly reported.¹²

When ESBL-producing E. coli and K. pneumoniae (ESBL-EK) isolates are susceptible to fluoroquinolone (FQ), aminoglycosides, or trimethoprim-sulfamethoxazole (TMP-SMX), therapy with these agents could be an alternative antimicrobial therapy. However, with resistance to each additional class of antibiotics, ESBL-EK infections become a greater therapeutic challenge. Furthermore, multidrug-resistant (MDR) ESBL-EK isolates (i.e., those resistant to multiple other antibiotics or antibiotic classes in addition to the oxymino β-lactam) pose significant therapeutic challenges. Despite the clinical significance of increasing MDR among ESBL-EK, little is known regarding risk factors for acquisition of MDR ESBL-EK isolates.⁹ Identification of risk factors for MDR ESBL-EK infections may help in the empirical therapeutic decision-making process and may assist in the early implementation of appropriate infection-control measures. Therefore, this study was performed to identify risk factors for MDR among ESBL-EK bacteremia in hospitalized patients.

Materials and Methods

Study design and patients

Study subjects were identified through a computerized database maintained by the hospital microbiology laboratory, which performs cultures for bacteria for all clinical specimens obtained at the Samsung Medical Center (SMC), a 1950-bed-referral tertiary care university hospital located in Seoul, South Korea. Our institute includes a 650-bed comprehensive cancer center. We reviewed the medical records of individuals found to have ESBL-EK bacteremia between January 2009 and December 2010 at SMC. Only adult patients (≥18 years) were included in the study if they had nosocomial bacteremia caused by ESBL-EK. Duplicated isolates from the same patient were excluded.

A case–control study was performed to evaluate risk factors for MDR among ESBL-EK bacteremia in hospitalized patients. All patients with a blood culture positive for an MDR ESBL-EK were designated as case patients, whereas control patients were those with an ESBL-EK isolate that did not meet the criteria for MDR. We reviewed the electronic medical records of individuals with ESBL-EK bacteremia. The data collected included age, sex, underlying disease, site of infection, severity of illness, duration of hospital stay before the onset of bacteremia, antimicrobial regimen, and any antimicrobial therapy received during the 30 days before the bacteremia. The presence of any of the following comorbid conditions was also documented: hepatic dysfunction, malignancy, renal insufficiency, neutropenia, care in an intensive care unit, recent surgical procedure, invasive procedure conducted during 72 hr before onset of bacteremia, corticosteroid use, or immunosuppressant use. Severity of infection was assessed at the time of the positive blood cultures using the Pitt bacteremia score.^3,4 Severity of comorbid conditions was calculated by the Charlson's comorbidity index score and the McCabe classification.² In addition, the presence of a central venous catheter, indwelling urinary catheter, percutaneous tube insertion, or mechanical ventilation was assessed. Since this study was retrospective, patient management and antimicrobial treatment regimens were chosen by the patients' physicians without any guidelines or intervention from the study investigators. The study was approved by the Institutional Review Board of SMC (Seoul, South Korea) as required by the local hospital policy at the time of the study.

Definitions

Bacteremia was defined as a finding of organisms in a blood culture specimen. Clinically significant ESBL-EK bacteremia was defined as the presence of ESBL-EK in the blood, documented by at least one positive blood culture, together with clinical features compatible with systemic inflammatory response syndrome. Bacteremia was categorized as polymicrobial if additional microorganisms were recovered from the blood cultures. MDR ESBL-EK was defined as an ESBL-EK that demonstrated in vitro resistance to all the following three antibiotics or antibiotic classes: TMP-SMX, FQ (i.e., ciprofloxacin and levofloxacin), and gentamicin. Non-MDR ESBL-EK was defined as ESBL-EK that did not meet the criteria for MDR. Nosocomial infection was defined as an infection that occurred >48 hr after admission to the hospital, an infection that occurred <48 hr after admission to the hospital in patients that had been hospitalized in the 2 weeks before admission, or an infection that occurred <48 hr after admission to the hospital in patients that had been transferred from another hospital or nursing home. Nosocomial infections were defined according to the criteria proposed by the Centers for Disease Control and Prevention.⁶ The site of infection was determined by physicians on the basis of the isolation of ESBL-EK from the presumed portal of entry and clinical evaluation. Primary bacteremia was defined according to the Center for Disease Control and Prevention definitions.⁶ Catheter-related bloodstream infections were judged according to the published guidelines.¹⁵ Chronic kidney disease was defined as a serum creatinine level ≥2 mg/dl, and the steroid use was defined as daily use of 20 mg prednisone for at least 2 weeks. Patients with immunosuppression included those who had immunosuppressive therapy (chemotherapeutic agents, immunosuppressive agents, or radiation therapy). Neutropenia was defined as an absolute neutrophil count <500/mm³. Severe sepsis was defined as sepsis with one or more clinical signs of organ dysfunction. Prior antibiotic therapy was defined as the receipt of any systemic antibiotics more than 48 hr in the preceding 30 days.

Microbiologic methods

Microorganisms were identified by the VITEK-GNI card (bioMérieux, Hazelwood, MO) in a microbiology laboratory. Antimicrobial susceptibility and ESBL confirmatory testing were performed by the GNI card in a VITEK II automated system using the modified broth microdilution method or the disk diffusion method, according to the recommendations of the Clinical and Laboratory Standards Institute. Strains showing intermediate antimicrobial susceptibility profiles were considered to be resistant.

For the characterization of ESBL types, polymerase chain reaction (PCR) and sequencing of PCR products were performed using available stored isolates. PCR for bla_TEM, bla_SHV, bla_CTX-M, and bla_VEB was conducted using previously described PCR primers and conditions.¹⁰ Both strands of all PCR fragments were sequenced, and the types of β-lactamase genes were identified by comparing the sequences to those in the database of G. Jacoby and K. Bush (http://lahey.org/Studies/). Available MDR ESBL-EK isolates were evaluated for clonal relatedness by pulsed-field gel electrophoresis (PFGE). For PFGE, agarose-embedded bacterial genomic DNA was digested with 20 U Xbal. The restriction fragments were separated by electrophoresis in 0.5× TBE buffer. PFGE was performed using the CHEF MAPPER XA apparatus (Bio-Rad Laboratories, Hercules, CA), as described previously.⁵ The resulting genomic DNA profiles were interpreted according to the previously established guidelines.²⁶

Statistical analysis

The Chi-square or Fisher exact test was used to compare categorical variables. Continuous variables were compared using the Student's t-test or the Mann-Whitney test. The stepwise backward logistic regression model was used to estimate the effects of multiple factors associated with MDR ESBL-EK infection. Variables with p values<0.1 in univariate analyses and risk factors reported by previous studies were candidates for inclusion in the multivariate analysis, including the infection pathogen and duration of hospitalization before bacteremia.^9,24 All variables for which p was <0.05 in the multivariate analysis were retained in the final model. Interactions between variables were not introduced into the models. Odds ratios (ORs) and their 95% confidence intervals (CIs) were calculated. All tests of significance were two tailed, with a p value<0.05 considered to be significant. IBM SPSS Statistics version 19 was used for these analyses.

Results

Clinical characteristics of the study population

A total of 123 patients with ESBL-EK bacteremia, including 74 (60.2%) patients with ESBL-producing E. coli bacteremia and 49 (39.8%) patients with ESBL-producing K. pneumoniae bacteremia, were enrolled during the study period. The mean age (±standard deviation) of the patients was 57.0±15.2 years. The most common underlying diseases were hematologic malignancies (37.4% [46/123]) and solid tumors (36.6% [45/123]). The most common comorbid conditions were immunosuppressed state (54.5% [67/123]), followed by neutropenia (33.3% [41/123]). Primary bacteremia with an unknown site of infection was the most common source of bacteremia (23.6% [29/123]), followed by urinary tract infection (21.1% [26/123]). Of the 123 isolates, 33 (26.8%) were due to MDR ESBL-EK. The susceptibility rates to FQ, gentamicin, TMP-SMX, and piperacillin/tazobactam among all isolates were 29.3% (36/123), 43.9% (54/123), 27.6% (34/123), and 51.2% (63/123), respectively. The susceptibility rate to amikacin was 87.8% (108/123), and all isolates were susceptible to imipenem and meropenem. Most of the patients (89.4% [110/123]) had received antibiotics within the 30 days before bacteremia.

Risk factors for MDR among ESBL-EK bacteremia

The 33 MDR ESBL-EK isolates included 21 (63.6%) E. coli and 12 (36.4%) K. pneumoniae. Data from the 33 patients were compared to that of 90 patients with non-MDR ESBL-EK bacteremia. The baseline and demographic characteristics of the patients according to study group (MDR ESBL-EK and non-MDR ESBL-EK) are shown in Table 1. Variables such as age, sex, infection pathogen, underlying disease, and the Charlson comorbidity index were similar between the two groups. Polymicrobial bacteremia was more frequent in control patients than case patients. The presence of other comorbidities and the use of specific antibiotics in the previous month are shown in Table 2. Univariate analysis revealed that significant factors associated with the MDR ESBL-EK group were neutropenia, receipt of immunosuppressant therapy, urinary tract infection, and prior use of antibiotics, especially FQ (all p<0.05). No significant differences were found regarding the presence of severe sepsis, hospital stay before bacteremia, or ICU care.

Table 1.

Baseline and Demographic Characteristics of Multidrug Resistant Extended-Spectrum β-Lactamase-Producing Escherichia coli and Klebsiella pneumoniae (case patients) Versus Non-Multidrug Resistant ESBL-EK (Control Patients) in Nosocomial Bacteremia

Characteristic	Total (n=123)	MDR ESBL-EK (n=33)	Non-MDR ESBL-EK (n=90)	p Value
Mean age (years)±standard deviation	57.0±15.2	56.33±12.82	56.89±12.69	0.831
Male	74 (60.2)	18 (54.5)	56 (62.2)	0.441
ESBL-producing E. coli	74 (60.2)	21 (63.6)	53 (58.9)	0.682
Polymicrobial infection	23 (18.7)	1 (3.0)	22 (24.4)	0.008
Hospital stay before bacteremia, median days (IQR)	17 (7–31)	20.0 (13–45.5)	15.5 (7–29)	0.109
Comorbid-underlying disease
Diabetes mellitus	24 (19.5)	6 (18.2)	18 (20)	1.000
Liver disease	37 (30.1)	7 (21.2)	30 (33.3)	0.268
Renal disease	26 (21.1)	9 (27.3)	17 (18.9)	0.313
Solid cancer	45 (36.6)	8 (24.2)	37 (41.1)	0.095
Hematologic malignancy	46 (37.4)	15 (45.5)	31 (34.4)	0.264
Cardiovascular disease	15 (12.2)	2 (6.1)	13 (14.4)	0.350
Neurologic disease	18 (14.6)	4 (12.1)	14 (15.6)	0.778
Pulmonary disease	4 (3.3)	0 (0)	4 (4.4)	0.573
McCabe and Jackson classification				1.000
Nonfatal	17 (13.8)	4 (12.1)	13 (14.4)
Ultimately fatal	87 (70.7)	24 (72.7)	63 (70.0)
Rapidly fatal	19 (15.4)	5 (15.2)	14 (15.6)
Charlson weighted index of comorbidity, median (IQR)	3 (2–5)	3 (2–5)	3 (2–6)	0.509
Site of infection
Primary bacteremia	29 (23.6)	10 (30.3)	19 (21.1)	0.287
Urinary tract infection	26 (21.1)	11 (33.3)	15 (16.7)	0.045
Pancreatobiliary infection	23 (18.7)	5 (15.2)	18 (20.0)	0.541
Intra-abdominal infection	22 (18.0)	4 (12.1)	18 (20.0)	0.428
Catheter-related infection	11 (8.1)	1 (3)	10 (11.1)	0.285
Pneumonia	9 (7.3)	2 (6.1)	7 (7.8)	1.000

Data are presented as numbers of patients (%) unless otherwise indicated.

ESBL-EK, extended-spectrum β-lactamase-producing Escherichia coli and Klebsiella pneumoniae; MDR, multidrug resistance; IQR, interquartile range.

Table 2.

Comorbid Conditions of Patients with MDR ESBL-EK Bacteremia Compared with Non-MDR ESBL-EK

Characteristic	Total (n=123)	MDR ESBL-EK (n=33)	Non-MDR ESBL-EK (n=90)	p Value
Comorbid conditions
Neutropenia	41 (33.3)	16 (48.5)	25 (27.8)	0.031
Corticosteroid use	27 (22.0)	10 (30.3)	17 (18.9)	0.175
Receipt of immunosuppressant	67 (54.5)	23 (69.7)	44 (48.9)	0.040
Solid-organ transplantation	19 (15.4)	8 (24.2)	11 (12.2)	0.102
Hematopoietic stem cell transplantation	15 (12.2)	5 (15.2)	10 (11.1)	0.544
Recent operation	42 (34.1)	11 (33.3)	31 (34.4)	0.908
Central venous catheterization	73 (59.3)	21 (63.6)	52 (57.8)	0.558
Indwelling urinary catheter	51 (41.5)	16 (48.5)	35 (38.9)	0.339
Recent invasive procedure	25 (20.5)	4 (12.1)	21 (23.6)	0.211
Percutaneous tube	46 (37.4)	17 (51.5)	29 (32.2)	0.050
ICU care	38 (30.9)	12 (36.4)	26 (28.9)	0.427
Mechanical ventilator	23 (18.7)	7 (21.2)	16 (17.8)	0.665
Severity of illness
Severe sepsis	57 (46.3)	15 (45.5)	42 (46.7)	0.905
Pitt bacteremia score, median (IQR)	1 (0–3)	2 (0–2.5)	1 (0–3)	0.671
Prior use of any antibiotics within 1 month	110 (89.4)	33 (100.0)	77 (85.6)	0.019
Specific antibiotics
Fluoroquinolones	28 (22.8)	12 (36.4)	16 (20.5)	0.029
Cephalosporins	67 (54.5)	17 (51.5)	50 (55.6)	0.690
β-lactam or β-lactam/β-lactamase inhibitor	38 (30.9)	13 (39.4)	25 (27.8)	0.217

To identify independent risk factors associated with increasing MDR among ESBL-EK bacteremia in hospitalized patients, a logistic regression analysis was performed, including variables with an association of p<0.1 at the bivariate level and potential risk factors reported by previous studies (K. pneumoniae and duration of hospitalization before bacteremia). In the multivariate analysis, administration of FQ within the previous 30 days (adjusted odds ratio [aOR]=2.927; 95% CI=1.070–8.009; p=0.036), percutaneous tube insertion (aOR=4.035; 95% CI=1.557–10.747; p=0.005), and neutropenia (aOR=4.222; 95% CI=1.557–11.449; p=0.005) were significantly associated with MDR ESBL-EK bacteremia (Table 3).

Table 3.

Risk Factors Associated with Multidrug Resistance Among ESBL-Producing E. coli and K. pneumoniae Bacteremia in Hospitalized Patients

	Univariate analysis		Multivariate analysis
Risk factor	Unadjusted OR (95% CI)	p Value	Adjusted OR (95% CI)	p Value
Prior exposure to fluoroquinolone within 1 month	2.643 (1.084–6.446)	0.029	2.927 (1.070–8.009)	0.036
Percutaneous tube	2.353 (0.991–5.041)	0.050	4.035 (1.557–10.747)	0.005
Neutropenia	2.447 (1.074–5.578)	0.031	4.222 (1.557–11.449)	0.005

OR, odds ratio; CI, confidence interval.

Microbiologic analysis and ESBL characterization

In the results of antimicrobial susceptibility testing, strains resistant to amikacin were more frequent in the MDR ESBL-EK group (27.3% vs. 6.7%; p=0.004). However, the prevalence of resistance to piperacillin/tazobactam was similar for both groups (51.5% vs. 47.8%; p=0.839).

Forty-seven isolates of ESBL-EK were included for a further microbiological study of ESBL characterization. Twenty-nine (61.7%) of the 47 were positive for CTX-M-1-like β-lactamase, and 18 (38.3%) were positive for CTX-M-9-like β-lactamase. Seven (14.9%) isolates produced SHV-type ESBL (SHV-2, 6 isolates; SHV-12, 1 isolate). The distribution of ESBL genes is summarized in Table 4. The CTX-M-15 enzyme belonging to the CTX-M-1-like β-lactamase was dominant in both the MDR ESBL-EK group and non-MDR ESBL-EK group (55% [11/20] vs. 55.6% [15/27]). CTX-M-15 and CTX-M-14 types were the most common ESBL types among ESBL-producing E. coli isolates. The CTX-M-15 type (25% [5/20]) was the most common in MDR ESBL-producing K. pneumoniae, and the SHV-2+CTX-M-15 type (22.2% [6/27]) was the most common in non-MDR ESBL-producing K. pneumoniae (Table 4).

Table 4.

Distribution of Extended-Spectrum β-Lactamases Among ESBL-Producing E. coli and K. pneumoniae Isolates Causing Nosocomial Bacteremia

ESBL type	MDR ESBL-EK n=20	Non-MDR ESBL-EK n=27
E. coli	12 (60.0)	16 (59.3)
CTX-M-15	5 (25.0)	7 (26.0)
CTX-M-14	5 (25.0)	6 (22.2)
CTX-M-57	1 (5.0)	0
CTX-M-3	0	0
CTX-M-24	0	2 (7.4)
CTX-M-32	0	1 (3.7)
CTX-M-65	1 (5.0)	0
K. pneumoniae	8 (40.0)	11 (40.7)
CTX-M-15	5 (25.0)	2 (7.4)
CTX-M-14	1 (5.0)	3 (11.1)
CTX-M-57	0	0
CTX-M-3	1 (5.0)	0
CTX-M-24	0	0
CTX-M-32	0	0
CTX-M-65	0	0
SHV-2+CTX-M15	0	6 (22.2)
SHV-12+CTX-M-15	1 (5.0)	0

The clonality of MDR ESBL-EK isolates was characterized by PFGE. Twelve E. coli isolates and eight K. pneumoniae isolates were available for further molecular analysis. The similarity among profiles was determined by a cluster analysis in a dendrogram using a cutoff of at least 80%. PFGE patterns from the MDR ESBL-producing K. pneumoniae isolates revealed close genetic relatedness among four of the isolates tested. In contrast, there was little evidence of clonality among the 12 MDR ESBL-producing E. coli isolates tested (Fig. 1). Two small clusters of clonally related isolates occurred at different times in different places. These clusters included three CTX-M-15-producing isolates (KPN5, KPN7, and KPN8) and one CTX-M-3-producing isolate (KPN1) (Fig. 1).

FIG. 1.

Pulsed-field gel electrophoresis and dendrogram of multidrug-resistant extended-spectrum β-lactamase-producing Escherichia coli (upper) and Klebsiella pneumoniae (bottom) isolates.

Discussion

In the current study, we sought to identify hospitalized patients who were at an increased risk of bacteremia with MDR ESBL-EK. Of 123 ESBL-EK isolates causing nosocomial bacteremia in the 2-year study period, 26.8% were MDR. Compared with non-MDR ESBL-EK bacteremia, MDR ESBL-EK bacteremia was more likely to be associated with a prior exposure to FQ, percutaneous tube insertion, and neutropenia. A previous study demonstrated that the only independent risk factor for MDR ESBL-EK was the infecting organism (i.e., K. pneumoniae), and that a prior antibiotic use was not independently associated with MDR ESBL-EK.⁹ However, our study focused on ESBL-EK bacteremia in hospitalized patients and showed that K. pneumoniae was not associated with an increased risk of MDR ESBL-EK. Our study showed that prior exposure to FQ is an independent risk factor for the MDR ESBL-EK bacteremia. In previous studies, the FQ use has been associated with the acquisition of other MDR gram-negative infections, such as Pseudomonas aeruginosa and Acinetobacter baumannii.^11,17,27 In addition, Villers et al.²⁷ showed that previous use of FQ was an independent risk factor for MDR A. baumannii infection, and that the institution of a policy restricting the use of intravenous FQs could decrease the infection rate with this microorganism. Interestingly, a recent study showed correlation of FQ resistance and MDR in an ESBL-producing E. coli infection.⁷ As previously reported in an MDR gram-negative infection,¹¹ we can speculate that a prior receipt of FQ either results in selective pressure or facilitates the activation of intrinsic mechanisms that confer resistance to multiple antibiotic drug classes such as drug efflux mechanism. Our data suggest that the notion that increasing MDR among ESBL-producing organisms is more likely to involve K. pneumoniae infections may be challenged by the increase of MDR ESBL-producing E. coli, which is one of the most serious pathogens in both the community and hospitals. To our knowledge, this is the largest report to compare patients with MDR ESBL-EK bacteremia with those with non-MDR ESBL-EK. A previous study regarding risk factors for increasing MDR among ESBL-EK included a high percentage of urinary tract infections, and <10% of the patients were bacteremic.⁹ Our study focused on ESBL-EK bacteremia, which is one of the most serious types of infection, and includes comprehensive clinical and microbiological analyses regarding MDR ESBL-EK bacteremia.

Numerous studies have demonstrated that most ESBL-producing Enterobacteriaceae are resistant to multiple antibiotic classes.^18,19,28 The resistance rates to FQ and TMP-SMX among ESBL-EK isolates in our study were 70.7% and 72.4%, respectively, which are comparable to rates found in other studies.^{14,16,23–25} Such a high prevalence of resistance to these antimicrobial agents, particularly FQ, is a well-known feature of ESBL producers and is of great concern. With regard to the treatment of ESBL-EK infections, no randomized controlled trial has been performed to date. Although carbapenems are considered the drug of choice for serious infections, alternatives such as FQ or aminoglycosides could be additional options. However, the susceptibility profile of our isolates confirms that alternatives for such infections are considerably limited. The increasing use of carbapenems has been paralleled by the rapid emergence of carbapenem resistance in nosocomial pathogens.^13,22 Our results emphasize the importance of identifying risk factors for MDR ESBL-EK to design and implement strategies that will prevent further increases in MDR ESBL-EK.⁹ The identification of risk factors for MDR in ESBL-EK infections provides vital information regarding possible means of curbing increased MDR, and thus could preserve alternative antimicrobial agents, thereby reducing the dependence on carbapenems. Prior use of antibiotics, especially FQ use, was associated with the development of MDR ESBL-EK bacteremia. Interventions to limit the emergence of ESBL-EK have traditionally focused on restricting certain antimicrobial agents associated with ESBL infections. Our data suggest that such interventions are certainly an important component in efforts to control the emergence of MDR ESBL-EK, although infection-control measures are also likely to be critical in interrupting the spread of such organisms.

Along with the prior use of FQ, percutaneous tube insertion and neutropenia were additional risk factors associated with MDR ESBL-EK bacteremia. These findings have important implications for infection control, because MDR ESBL-EK strains in a hospital environment pose a serious risk among patients with tube insertion or neutropenia. To reduce the frequency of MDR ESBL-EK bacteremia, it is important that invasive procedures, including tube insertion, should be minimized when possible, and the use of FQ should be prudently limited. Previous studies have demonstrated the emergence of FQ-resistant organisms in neutropenic cancer patients who had received prophylactic quinolone.^1,20,21 Even though oral ciprofloxacin is not routinely administered to patients with hematologic malignancies in our hospital, neutropenia was found to be associated with MDR ESBL-EK bacteremia. The close relationship between MDR ESBL-EK bacteremia and neutropenia is particularly troublesome, because the therapeutic options for febrile neutropenic patients with cancer are severely restricted.

Our study has several limitations. First, because this study was of a retrospective nature, the possibility of a limitation in precluding accurate comparisons should be borne in mind. The present study was observational, and thus unknown risk factors might have been unequally distributed between the two groups. Although selection bias is normally of concern in any case–control study, the patients of this study were tested in the same microbiology laboratory and by the same methodology, and all eligible patients were included. Thus, the potential for selection bias is likely to have been small. Second, we used a control group consisting of patients who had acquired non-MDR ESBL-EK. The selection bias introduced by the use of control patients with antimicrobial-susceptible organisms is likely to have the largest impact on antibiotics that are active against susceptible, but not against the resistant forms of the organisms.⁸ However, the research question addressed in our study was the risk factors for MDR among ESBL-EK isolated from patients with nosocomial bacteremia caused by this organism, rather than investigating the risk factors for MDR ESBL-EK among hospitalized patients in general. Third, not all ESBL-EK isolates were available for further molecular analysis. Therefore, the ESBL characterization results may not be representative of all ESBL-EK isolates. Finally, our study was conducted at a large referral center. Thus, many of our patients had serious underlying illnesses, and our findings may not be generalizable to other institutions, particularly community hospitals.

In conclusion, the emergence of MDR among ESBL-EK is particularly troublesome, because the therapeutic options for serious infections caused by ESBL-EK are severely restricted. We determined that 26.8% of ESBL-EK isolates causing nosocomial bacteremia at our hospital were MDR, and the independent risk factors for MDR ESBL-EK bacteremia were prior exposure to FQ, percutaneous tube insertion, and neutropenia. Our data suggest that strategies designed to reduce MDR in ESBL-EK should focus on limiting the use of FQ and minimizing invasive procedures such as percutaneous tube placement.

Footnotes

Acknowledgments

This study was supported by a Samsung Biomedical Research Institute grant, #SBRI GL1-B2-031-1, and by a grant from the Korea Healthcare Technology 21 R&D Project, Ministry of Health and Welfare Affairs, Republic of Korea (Grant No. A110301).

Disclosure Statement

The authors have nothing to declare.

References

Bhusal

, Mihu

C.N.

, Tarrand

J.J.

, Rolston

K.V.

2011. Incidence of fluoroquinolone-resistant and extended-spectrum beta-lactamase-producing Escherichia coli at a comprehensive cancer center in the United States. Chemotherapy, 57:335–338.

Charlson

M.E.

, Pompei

, Ales

K.L.

, MacKenzie

C.R.

1987. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J. Chronic Dis., 40:373–383.

Chow

J.W.

, Fine

M.J.

, Shlaes

D.M.

, Quinn

J.P.

, Hooper

D.C.

, Johnson

M.P.

, Ramphal

, Wagener

M.M.

, Miyashiro

D.K.

, Yu

V.L.

1991. Enterobacter bacteremia: clinical features and emergence of antibiotic resistance during therapy. Ann. Intern. Med., 115:585–590.

Chow

J.W.

, Yu

V.L.

1999. Combination antibiotic therapy versus monotherapy for gram-negative bacteraemia: a commentary. Int. J. Antimicrob. Agents, 11:7–12.

Chung

D.R.

, Lee

H.R.

, Lee

S.S.

, Kim

S.W.

, Chang

H.H.

, Jung

S.I.

, Oh

M.D.

, Ko

K.S.

, Kang

C.I.

, Peck

K.R.

et al. 2008. Evidence for clonal dissemination of the serotype K1 Klebsiella pneumoniae strain causing invasive liver abscesses in Korea. J. Clin. Microbiol., 46:4061–4063.

Garner

J.S.

, Jarvis

W.R.

, Emori

T.G.

, Horan

T.C.

, Hughes

J.M.

1988. CDC definitions for nosocomial infections, 1988. Am. J. Infect. Control, 16:128–140.

Giufre

, Graziani

, Accogli

, Luzzi

, Busani

, Cerquetti

2012. Escherichia coli of human and avian origin: detection of clonal groups associated with fluoroquinolone and multidrug resistance in Italy. J. Antimicrob. Chemother., 67:860–867.

Harris

A.D.

, Samore

M.H.

, Lipsitch

, Kaye

K.S.

, Perencevich

, Carmeli

2002. Control-group selection importance in studies of antimicrobial resistance: examples applied to Pseudomonas aeruginosa, Enterococci, and Escherichia coli. Clin. Infect. Dis., 34:1558–1563.

Hyle

E.P.

, Lipworth

A.D.

, Zaoutis

T.E.

, Nachamkin

, Fishman

N.O.

, Bilker

W.B.

, Mao

, Lautenbach

2005. Risk factors for increasing multidrug resistance among extended-spectrum beta-lactamase-producing Escherichia coli and Klebsiella species. Clin. Infect. Dis., 40:1317–1324.

10.

Kim

, Lim

Y.M.

, Rheem

, Lee

J.C.

, Seol

S.Y.

, Lee

Y.C.

, Cho

D.T.

2005. CTX-M and SHV-12 beta-lactamases are the most common extended-spectrum enzymes in clinical isolates of Escherichia coli and Klebsiella pneumoniae collected from 3 university hospitals within Korea. FEMS Microbiol. Lett., 245:93–98.

11.

Kopterides

, Koletsi

P.K.

, Michalopoulos

, Falagas

M.E.

2007. Exposure to quinolones is associated with carbapenem resistance among colistin-susceptible Acinetobacter baumannii blood isolates. Int. J. Antimicrob. Agents, 30:409–414.

12.

Kritsotakis

E.I.

, Tsioutis

, Roumbelaki

, Christidou

, Gikas

2011. Antibiotic use and the risk of carbapenem-resistant extended-spectrum-{beta}-lactamase-producing Klebsiella pneumoniae infection in hospitalized patients: results of a double case–control study. J. Antimicrob. Chemother., 66:1383–1391.

13.

Lee

H.Y.

, Chen

C.L.

, Wang

S.B.

, Su

L.H.

, Chen

S.H.

, Liu

S.Y.

, Wu

T.L.

, Lin

T.Y.

, Chiu

C.H.

2011. Imipenem heteroresistance induced by imipenem in multidrug-resistant Acinetobacter baumannii: mechanism and clinical implications. Int. J. Antimicrob. Agents, 37:302–308.

14.

Lin

C.J.

, Siu

L.K.

, Ma

, Chang

Y.T.

, Lu

P.L.

2012. Molecular epidemiology of ciprofloxacin-resistant extended-spectrum beta-lactamase-producing Klebsiella pneumoniae in Taiwan. Microb. Drug Resist., 18:52–58.

15.

Mermel

L.A.

, Allon

, Bouza

, Craven

D.E.

, Flynn

, O'Grady

N.P.

, Raad

I.I.

, Rijnders

B.J.

, Sherertz

R.J.

, Warren

D.K.

2009. Clinical practice guidelines for the diagnosis and management of intravascular catheter-related infection: 2009 Update by the Infectious Diseases Society of America. Clin. Infect. Dis., 49:1–45.

16.

Mosqueda-Gomez

J.L.

, Montano-Loza

, Rolon

A.L.

, Cervantes

, Bobadilla-del-Valle

J.M.

, Silva-Sanchez

, Garza-Ramos

, Villasis-Keever

, Galindo-Fraga

, Palacios

G.M.

et al. 2008. Molecular epidemiology and risk factors of bloodstream infections caused by extended-spectrum beta-lactamase-producing Klebsiella pneumoniae A case–control study. Int. J. Infect. Dis., 2008; 12:653–659.

17.

Paramythiotou

, Lucet

J.C.

, Timsit

J.F.

, Vanjak

, Paugam-Burtz

, Trouillet

J.L.

, Belloc

, Kassis

, Karabinis

, Andremont

2004. Acquisition of multidrug-resistant Pseudomonas aeruginosa in patients in intensive care units: role of antibiotics with antipseudomonal activity. Clin. Infect. Dis., 38:670–677.

18.

Paterson

D.L.

, Ko

W.C.

, Von Gottberg

, Mohapatra

, Casellas

J.M.

, Goossens

, Mulazimoglu

, Trenholme

, Klugman

K.P.

, Bonomo

R.A.

2004. Antibiotic therapy for Klebsiella pneumoniae bacteremia: implications of production of extended-spectrum beta-lactamases. Clin. Infect. Dis., 39:31–37.

19.

Pitout

J.D.

2008. Multiresistant Enterobacteriaceae: new threat of an old problem. Exp. Rev. Anti. Infect. Ther., 6:657–669.

20.

Rangaraj

, Granwehr

B.P.

, Jiang

, Hachem

, Raad

. 2010. Perils of quinolone exposure in cancer patients: breakthrough bacteremia with multidrug-resistant organisms. Cancer, 116:967–973.

21.

Reuter

, Kern

W.V.

, Sigge

, Dohner

, Marre

, Kern

, von Baum

2005. Impact of fluoroquinolone prophylaxis on reduced infection-related mortality among patients with neutropenia and hematologic malignancies. Clin. Infect. Dis., 40:1087–1093.

22.

Rodriguez-Avial

, Rodriguez-Avial

, Merino

, Picazo

J.J.

2012. Klebsiella pneumoniae: development of a mixed population of carbapenem and tigecycline resistance during antimicrobial therapy in a kidney transplant patient. Clin. Microbiol. Infect., 18:61–66.

23.

Schwaber

M.J.

, Navon-Venezia

, Schwartz

, Carmeli

2005. High levels of antimicrobial coresistance among extended-spectrum-beta-lactamase-producing Enterobacteriaceae. Antimicrob. Agents Chemother., 49:2137–2139.

24.

Serefhanoglu

, Turan

, Timurkaynak

F.E.

, Arslan

2009. Bloodstream infections caused by ESBL-producing E. coli and K. pneumoniae: risk factors for multidrug-resistance. Braz. J. Infect. Dis., 13:403–407.

25.

Simner

P.J.

, Zhanel

G.G.

, Pitout

, Tailor

, McCracken

, Mulvey

M.R.

, Lagace-Wiens

P.R.

, Adam

H.J.

, Hoban

D.J.

2011. Prevalence and characterization of extended-spectrum beta-lactamase- and AmpC beta-lactamase-producing Escherichia coli: results of the CANWARD 2007–2009 study. Diagn. Microbiol. Infect Dis., 69:326–334.

26.

Tenover

F.C.

, Arbeit

R.D.

, Goering

R.V.

, Mickelsen

P.A.

, Murray

B.E.

, Persing

D.H.

, Swaminathan

1995. Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing. J. Clin. Microbiol., 33:2233–2239.

27.

Villers

, Espaze

, Coste-Burel

, Giauffret

, Ninin

, Nicolas

, Richet

1998. Nosocomial Acinetobacter baumannii infections: microbiological and clinical epidemiology. Ann. Intern. Med., 129:182–189.

28.

Wiener

, Quinn

J.P.

, Bradford

P.A.

, Goering

R.V.

, Nathan

, Bush

, Weinstein

R.A.

1999. Multiple antibiotic-resistant Klebsiella and Escherichia coli in nursing homes. JAMA, 281:517–523.