Abstract
A multicenter, case-control study was conducted to assess risk factors and patient outcomes from bacteremia due to Enterobacteriaceae producing extended-spectrum β-lactamases (ESBL) and Klebsiella pneumoniae carbapenemases (KPCs). One hundred five and 20 patients with bacteremia due to ESBL and KPC-producing organisms were matched to controls that had bacteremia with non-ESBL/KPC-producing organisms, respectively. Independent risk factors for ESBL production included admission from a nursing home (odds ratio [OR], 4.64; 95% confidence interval [CI], 2.64–8.16), chronic renal failure (OR, 2.09; 95% CI, 1.11–3.92), the presence of a gastrostomy tube (OR, 3.36; 95% CI, 1.38–8.18), length of hospital stay before infection (OR, 1.02; 95% CI, 1.01–1.03), transplant recipients (OR, 2.48; 95% CI, 1.24–4.95) and receipt of antibiotics with Gram-negative activity in the preceding 30 days (OR, 1.76; 95% CI, 1.00–3.08). 28-day crude mortality rates for patients infected with ESBL or KPC-producing organisms and controls were 29.1% (34/117) and 19.5% (53/272), respectively (OR 1.70; 95% CI 1.04–2.80). On multivariate analysis, inadequate empiric therapy (OR, 2.26; 95% CI, 1.18–4.34), onset of bacteremia while in ICU (OR, 2.74; 95% CI, 1.47–5.11), Apache II score (OR, 1.17; 95% CI, 1.12–1.23), and malignancy (OR, 2.66; 95% CI, 1.31–5.41) were independent risk factors for mortality. CTX-M was the most common ESBL type in E. coli, whereas SHV predominated in Klebsiella spp. and Enterobacter spp.
Keywords: Enterobacteriaceae, bacteremia, extended-spectrum β-lactamase (ESBL), Klebsiella pneumoniae carbapenemase (KPC), carbapenem
INTRODUCTION
Extended-spectrum β-lactamases (ESBLs) are now well recognized as a major cause of cephalosporin resistance among Enterobacteriaceae [1, 2]. Two important changes in the epidemiology of ESBL have occurred in the last decade. One is the rapid global dissemination of Escherichia coli producing CTX-M-type ESBLs. CTX-M has now become the predominant ESBL type in Europe, Canada, South America and Asia [3]. Several recent studies suggest that it is finally replacing classic TEM- and SHV-type ESBLs in the United States as well [4–6]. The other change is the advent of Klebsiella pneumoniae carbapenemases, also known as KPC. Initially causing outbreaks in hospitals in the East Coast of the United States, KPC-producing K. pneumoniae have now spread worldwide [7]. The aims of the present study were to investigate the risk factors, clinical outcome and predictors of mortality in a contemporary cohort of patients with bacteremia due to ESBL- and KPC-producing Enterobacteriaceae in the United States.
PATIENTS AND METHODS
Study design and patients
This multicenter, retrospective case-control study was conducted at three medical centers in Northeastern United States (New York, Massachusetts and Pennsylvania) between 2005 and 2008. Patients with blood cultures growing E. coli, K. pneumoniae, K. oxytoca, Enterobacter cloacae and E. aerogenes were included. The two case groups consisted of those due to ESBL-producing and KPC-producing organisms. Controls were selected so that each study site and species occurred in similar proportions in both cases and controls with a ratio of three controls per case. The entire control group was used for the ESBL analysis, whereas only cases due to K. pneumoniae were included for the KPC analysis, since all KPC-producing organisms belonged to this species. Only the first episode of bacteremia was included for each patient. The study was approved by the Institutional Review Board at each participating site.
Risk factors analyses
The variables reviewed for risk factor analyses included demographic data, admission from nursing home, prior use of antibiotics with Gram-negative activity in the preceding 30 days, presence of tracheostomy tube, gastrostomy tube, intravenous line or urinary catheter at the time of infection, prior hospitalization, length of hospital stay before the onset of bacteremia, intensive care unit (ICU) admission, surgery, outpatient intravenous therapy or dialysis in the preceding one year, diabetes, the presence of chronic renal failure, liver disease, chronic obstructive pulmonary disease, cardiovascular disease, peripheral vascular disease, cerebrovascular disease, peptic ulcer disease, malignancy and history of organ transplant. Apache II score was also collected and included in the univariate analysis, but was not included in the multivariate analysis since it could not be defined for over 10% of the cases. Immunocompromised state was defined as the presence of diabetes mellitus, neutropenia, HIV infection, or receipt of steroids or other immunosuppressive agents in 30 days prior to infection.
Predictors of mortality
The primary outcome measure was death within 28 days from the onset of bacteremia. Predictors of 28-day mortality were analyzed for the whole study population, as well as within the ESBL or KPC cases and the controls separately. Patients for whom 28-day status was unknown were excluded from the mortality analyses (n = 61). The parameters reviewed in addition to those used for risk factor analyses were inadequate empiric therapy, defined as receipt of empiric antibiotics with no in vitro activity against the infecting organism, ICU stay at the time of infection, source of bacteremia and Apache Π score.
Microbiologic methods
All E. coli and Klebsiella isolates that were reported as ESBL producers were subjected to PCR analyses to identify the presence of TEM, SHV and CTX-M-type β-lactamase genes [8]. Positive PCR results were followed by sequencing of the amplified products. For Enterobacter spp., all piperacillin-resistant or ceftriaxone-non-susceptible Enterobacter isolates were subjected to PCR analyses for the ESBL genes and sequencing [9]. For all species, ceftriaxone-non-susceptible isolates (corresponding to minimum inhibitory concentrations ≥16 mg/L) underwent PCR analysis for detection of the KPC gene [10].
Statistical analyses
Statistical analyses were performed using SAS version 9.2 (SAS Institute Inc., Cary, NC). For univariate analysis of risk factors and predictors of mortality, univariate logistic regression was performed. To identify independent risk factors and predictors of mortality, variables with a P value ≤ 0.15 on univariate analysis were included in a stepwise conditional multivariate logistic regression model. All P values were 2-tailed and a P value of ≤ 0.05 was considered statistically significant.
RESULTS
Demographics of the cohorts
A total of 450 patients with bacteremia due to Enterobacteriaceae (Escherichia coli, Klebsiella spp. and Enterobacter spp.) were included. One hundred twenty-five bacteremia cases were due to ESBL or KPC-producing organisms, including E. coli (n = 32), Klebsiella spp. (n = 76) and Enterobacter spp. (n = 17). Of these, 20 cases were due to KPC-producing K. pneumoniae, all of which were from the study site in New York City. The remaining 105 cases constituted the ESBL group. A total of 325 control cases were selected for analysis of the ESBL group. For analysis of the KPC group, K. pneumoniae cases from this control group were assigned (n = 176). The characteristics of these patients relative to ESBL and KPC status are shown in Tables 1 and 2, respectively. In the analysis of the ESBL and control groups, there was no significant difference in the proportion of patients > 60 years of age, sex or race between the cases and controls (Table 1). In the analysis of the KPC and control groups, there were significantly less Caucasian patients in the KPC group, reflecting the predominance of African American patients at the study site in New York City, where all the KPC cases were identified (Table 2).
Table 1.
Demographics and risk factors of ESBL-producing Enterobacteriaceae cases.
| Demographics/risk factors | Univariate analysis | Multivariate analysis | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| ESBL group | % | Control group | % | Odds Ratio | 95% Confidence Interval | p-value | Odds Ratio | 95% Confidence Interval | p-value | |
| Apache II score | 15.5a | (0–49) b | 13 a | (1–45) b | 1.04 | (1.01, 1.07) | 0.01 | |||
| Male | 54 | 51.43 | 167 | 51.38 | 1.00 | (0.65, 1.56) | 0.99 | |||
| Age >60 | 51 | 48.57 | 174 | 53.70 | 0.81 | (0.52, 1.26) | 0.36 | |||
| Immunocompromised | 63 | 60.00 | 160 | 49.23 | 1.55 | (0.99, 2.42) | 0.06 | |||
| Caucasian | 54 | 66.67 | 173 | 66.03 | 1.03 | (0.61, 1.74) | 0.92 | |||
| Hispanic | 5 | 6.17 | 27 | 10.31 | 0.57 | (0.21, 1.54) | 0.27 | |||
| Admitted from nursing home | 44 | 43.14 | 41 | 12.65 | 5.24 | (3.14, 8.73) | <.0001 | 4.64 | (2.64, 8.16) | <.0001 |
| Number of days hospitalized before infection | 5 a | (0–328) b | 1 a | (0–141) b | 1.02 | (1.01, 1.03) | <.0001 | 1.02 | (1.01, 1.03) | 0.0015 |
| Any hospitalization in past year | 84 | 80.00 | 242 | 74.46 | 1.37 | (0.80, 2.35) | 0.25 | |||
| Any ICU admission in past year | 44 | 41.90 | 95 | 29.23 | 1.75 | (1.11, 2.75) | 0.02 | |||
| Any surgery in past year | 47 | 44.76 | 113 | 34.77 | 1.52 | (0.97, 2.38) | 0.07 | |||
| Dialysis in past year | 19 | 18.10 | 29 | 8.92 | 2.26 | (1.21, 4.22) | 0.01 | |||
| Outpatient intravenous therapy in past year | 15 | 14.29 | 24 | 7.38 | 2.09 | (1.05, 4.15) | 0.04 | |||
| Prior surgery within 30 days | 15 | 14.29 | 53 | 16.31 | 0.86 | (0.46, 1.59) | 0.62 | |||
| Antibiotics in 30 days prior to enrollment | 48 | 47.06 | 76 | 23.38 | 2.91 | (1.83, 4.64) | <.0001 | 1.76 | (1.00, 3.08) | 0.0491 |
| Carbapenem c | 8 | 7.62 | 8 | 2.46 | 3.27 | (1.20, 8.94) | 0.02 | |||
| Ampicillin/sulbactam | 6 | 5.71 | 15 | 4.62 | 1.25 | (0.47, 3.32) | 0.65 | |||
| Fluoroquinolone d | 12 | 11.43 | 28 | 8.62 | 1.37 | (0.67, 2.80) | 0.39 | |||
| Cefepime | 18 | 17.14 | 17 | 5.23 | 3.75 | (1.85, 7.58) | 0.0002 | |||
| Ceftriaxone | 4 | 3.81 | 4 | 1.23 | 3.18 | (0.78, 12.94) | 0.11 | |||
| Piperacillin/tazobactam | 8 | 7.62 | 17 | 5.23 | 1.49 | (0.63, 3.57) | 0.37 | |||
| Chronic renal failure | 28 | 26.67 | 41 | 12.62 | 2.52 | (1.46, 4.33) | 0.0008 | 2.09 | (1.11, 3.92) | 0.0223 |
| Diabetes | 40 | 38.10 | 79 | 24.31 | 1.92 | (1.20, 3.06) | 0.006 | |||
| Chronic obstructive lung disease | 11 | 10.48 | 45 | 13.85 | 0.73 | (0.36, 1.47) | 0.37 | |||
| Cardiovascular disease | 29 | 27.62 | 77 | 23.69 | 1.23 | (0.75, 2.02) | 0.42 | |||
| Peripheral vascular disease | 4 | 3.81 | 8 | 2.46 | 1.57 | (0.46, 5.32) | 0.47 | |||
| Cerebrovascular disease | 9 | 8.57 | 27 | 8.31 | 1.03 | (0.47, 2.28) | 0.93 | |||
| Peptic ulcer disease | 3 | 2.86 | 13 | 4.00 | 0.71 | (0.20, 2.53) | 0.59 | |||
| liver disease | 17 | 16.19 | 35 | 10.77 | 1.60 | (0.86, 3.00) | 0.14 | |||
| Malignancy | 14 | 13.33 | 61 | 18.77 | 0.67 | (0.36, 1.25) | 0.20 | |||
| Solid organ malignancy | 10 | 9.52 | 53 | 16.31 | 0.54 | (0.26, 1.10) | 0.09 | |||
| Transplant | 23 | 21.90 | 34 | 10.46 | 2.40 | (1.34, 4.30) | 0.003 | 2.48 | (1.24, 4.95) | 0.0103 |
| Tracheostomy tube at enrollment | 26 | 24.76 | 29 | 8.92 | 3.36 | (1.87, 6.03) | <.0001 | |||
| Gastrostomy tube at enrollment | 17 | 16.19 | 12 | 3.69 | 5.04 | (2.32, 10.95) | <.0001 | 3.36 | (1.38, 8.18) | 0.0076 |
| Indwelling urinary catheter at enrollment | 26 | 24.76 | 51 | 15.69 | 1.77 | (1.04, 3.02) | 0.04 | |||
| Vascular catheter at enrollment | 74 | 70.48 | 176 | 54.15 | 2.02 | (1.26, 3.24) | 0.004 | |||
Median.
Range.
Ertapenem, imipenem, meropenem.
Ciprofloxacin, levofloxacin, moxifloxacin.
Table 2.
Demographics and risk factors of KPC-producing K. pneumoniae cases.
| Demographics/risk factors | KPC group | % | Control group | % | Odds Ratio | 95% Confidence Interval | p-value |
|---|---|---|---|---|---|---|---|
| Apache II score | 21 a | (3–44) b | 14 a | (1–45) b | 1.11 | (1.05, 1.17) | 0.0001 |
| Male | 8 | 40.00 | 92 | 52.27 | 0.61 | (0.24, 1.56) | 0.30 |
| Age >60 | 12 | 60.00 | 95 | 54.29 | 1.26 | (0.49, 3.24) | 0.63 |
| Immunocompromised | 13 | 65.00 | 93 | 52.84 | 1.66 | (0.63, 4.35) | 0.30 |
| Caucasian | 4 | 21.05 | 96 | 67.61 | 0.13 | (0.04, 0.41) | 0.0005 |
| Hispanic | 3 | 15.79 | 13 | 9.15 | 1.86 | (0.48, 7.24) | 0.37 |
| Admitted from nursing home | 6 | 30.00 | 28 | 16.00 | 2.25 | (0.80, 6.35) | 0.13 |
| Number of days hospitalized before infection | 20 a | (0–323) b | 1 a | (0–141) b | 1.03 | (1.01, 1.05) | 0.0006 |
| Any hospitalization in past year | 16 | 80.00 | 141 | 80.11 | 0.99 | (0.31, 3.16) | 0.99 |
| Any ICU admission in past year | 9 | 45.00 | 54 | 30.68 | 1.85 | (0.72, 4.72) | 0.20 |
| Any surgery in past year | 13 | 65.00 | 64 | 36.36 | 3.25 | (1.23, 8.56) | 0.02 |
| Dialysis in past year | 1 | 5.00 | 8 | 4.55 | 1.11 | (0.13, 9.32) | 0.93 |
| Outpatient intravenous therapy in past year | 3 | 15.00 | 16 | 9.09 | 1.76 | (0.47, 6.68) | 0.40 |
| Prior surgery within 30 days | 3 | 15.00 | 26 | 14.77 | 1.02 | (0.28, 3.72) | 0.98 |
| Antibiotics in 30 days prior to enrollment | 17 | 85.00 | 40 | 22.73 | 19.3 | (5.37, 69.07) | <.0001 |
| Carbapenem c | 5 | 25.00 | 4 | 2.27 | 14.3 | (3.48, 59.10) | 0.0002 |
| Ampicillin/sulbactam | 2 | 10.00 | 11 | 6.25 | 1.67 | (0.34, 8.12) | 0.53 |
| Fluoroquinolone d | 2 | 10.00 | 13 | 7.39 | 1.39 | (0.29, 6.67) | 0.68 |
| Cefepime | 9 | 45.00 | 8 | 4.55 | 17.2 | (5.54, 53.23) | <.0001 |
| Ceftriaxone | 3 | 15.00 | 2 | 1.14 | 15.3 | (2.40, 98.27) | 0.004 |
| Piperacillin/tazobactam | 0 | 0.00 | 7 | 3.98 | 0.90 | (0.00, 6.33) | 0.93 |
| Chronic renal failure | 6 | 30.00 | 21 | 11.93 | 3.16 | (1.10, 9.13) | 0.03 |
| Diabetes | 6 | 30.00 | 41 | 23.30 | 1.41 | (0.51, 3.91) | 0.51 |
| Chronic obstructive lung disease | 4 | 20.00 | 29 | 16.48 | 1.27 | (0.40, 4.07) | 0.69 |
| Cardiovascular disease | 5 | 25.00 | 41 | 23.30 | 1.10 | (0.38, 3.20) | 0.86 |
| Peripheral vascular disease | 1 | 5.00 | 3 | 1.70 | 3.04 | (0.30, 30.65) | 0.35 |
| Cerebrovascular disease | 3 | 15.00 | 13 | 7.39 | 2.21 | (0.57, 8.55) | 0.25 |
| Peptic ulcer disease | 0 | 0.00 | 9 | 5.11 | 0.68 | (0, 4.60) | 0.74 |
| liver disease | 2 | 10.00 | 17 | 9.66 | 1.04 | (0.22, 4.87) | 0.96 |
| Malignancy | 8 | 40.00 | 41 | 23.30 | 2.20 | (0.84, 5.74) | 0.11 |
| Solid organ malignancy | 7 | 35.00 | 36 | 20.45 | 2.09 | (0.78, 5.63) | 0.14 |
| Transplant | 1 | 5.00 | 17 | 9.66 | 0.49 | (0.06, 3.91) | 0.50 |
| Tracheostomy tube at enrollment | 8 | 40.00 | 17 | 9.66 | 6.24 | (2.24, 17.38) | 0.0005 |
| Gastrostomy tube at enrollment | 9 | 45.00 | 7 | 3.98 | 19.8 | (6.19, 63.06) | <.0001 |
| Indwelling urinary catheter at enrollment | 5 | 25.00 | 25 | 14.20 | 2.01 | (0.67, 6.03) | 0.21 |
| Vascular catheter at enrollment | 15 | 75.00 | 101 | 57.39 | 2.23 | (0.78, 6.40) | 0.14 |
Median.
Range.
Ertapenem, imipenem, meropenem.
Ciprofloxacin, levofloxacin, moxifloxacin.
Risk factors for ESBL production
Independent risk factors for bacteremia caused by ESBL-producing organisms, after adjusting for potential confounding variables, included admission from a nursing home (OR, 4.64; 95% CI, 2.64–8.16, P < 0.0001), chronic renal failure (OR, 2.09; 95% CI, 1.11–3.92, P = 0.0223), the presence of a gastrostomy tube (OR, 3.36; 95% CI, 1.38–8.18, P = 0.0076), length of hospital stay before infection (OR, 1.02; 95% CI, 1.01–1.03, P = 0.0015), history of transplant (OR, 2.48; 95% CI, 1.24–4.95, P = 0.0103) and receipt of antibiotics with Gram-negative activity in the preceding 30 days (OR, 1.76; 95% CI, 1.00–3.08, P = 0.0491) (Table 1).
To characterize further the association between prior antibiotic exposure and ESBL-producing organisms, the antibiotics were divided into classes and assessed for their association. We observed that prior use of cefepime and carbapenems was significantly associated with infection with an ESBL-producing organism in the univariate analysis, but the significance did not stay in the multivariate analysis (Table 1).
Risk Factors for KPC production
Due to the relatively small number of the KPC cases, multivariate analysis was deferred for this comparison. In the univariate analysis, significant risk factors were largely consistent with those observed with the ESBL group, with the exception of surgery in the past year (Table 2). As for the individual antibiotic classes, receipt of carbapenems, cefepime, or ceftriaxone in the prior 30 days was associated with infection with a KPC-producing organism.
Empiric therapy
Of the 450 patients included in the study, 329 (73.1%) received appropriate empiric antibiotic therapy. Of the 125 cases with ESBL or KPC-producing organisms, 44 (35.2%) received appropriate empiric therapy, whereas 285 of 325 (87.7%) patients received appropriate therapy in the control group. Patients with bacteremia due to an ESBL or KPC-producing organism were significantly more likely to receive inappropriate empiric antibiotic therapy than the controls (OR, 13.12; 95% CI, 8.00–21.50; P < 0.001).
Clinical outcome
For cases whose clinical outcome 28 days after enrollment was available (n = 389), 28-day crude mortality rates for patients infected with ESBL or KPC-producing organisms and controls were 29.1% (34/117) and 19.5% (53/272), respectively (OR 1.70; 95% CI 1.03–2.79; P = 0.04). When cases with KPC and ESBL-producing K. pneumoniae were compared, the 28-day crude mortality was 47.4% (9/19) for cases with KPC-producing K. pneumoniae as opposed to 27.5% (14/51) for ESBL-producing K. pneumoniae (OR, 2.38; 95% CI, 0.80–7.08; P = 0.12).
Predictors of mortality
Predictors of mortality were first assessed for the whole study population. Results from univariate and multivariate analyses are shown in Table 3. Independent predictors included inadequate empiric therapy (OR, 2.26; 95% CI, 1.18–4.34; P = 0.01), bacteremia while in ICU (OR, 2.74; 95% CI, 1.47–5.11; P = 0.002), Apache II score (OR, 1.17; 95% CI, 1.12–1.23; P < 0.001) and malignancy (OR, 2.66; 95% CI, 1.31–5.41; P = 0.007) (Table 3). Production of ESBL or KPC was a significant predictor of mortality in the univariate analysis but was not a predictor after adjusting for other confounding variables in the multivariate analysis.
Table 3.
Predictors of mortality in all patients with bacteremia due to Enterobacteriaceae.
| Predictors of mortality | Univariate analysis | Multivariate analysis | ||
|---|---|---|---|---|
| OR (95% CI) | p-value | OR (95% CI) | p-value | |
| Apache II score | 1.18 (1.14, 1.23) | <.0001 | 1.2 (1.12, 1.23) | < .0001 |
| Male | 1.0 (0.62, 1.62) | 0.99 | ||
| Caucasian | 0.85 (0.49, 1.48) | 0.57 | ||
| Hispanic | 2.0 (0.92, 4.35) | 0.08 | ||
| Age > 60 | 1.6 (0.98, 2.60) | 0.06 | ||
| Immunocompromised | 1.2 (0.74, 1.93) | 0.48 | ||
| Onset of bacteremia in ICU | 3.7 (2.24, 6.02) | <.0001 | 2.7 (1.47,5.11) | 0.002 |
| Pneumonia as a source of bacteremia | 2.1 (1.05, 4.29) | 0.04 | ||
| Length of hospital stay before bacteremia | 1.0 (1.00, 1.01) | 0.33 | ||
| Admitted from nursing home | 1.5 (0.87, 2.68) | 0.14 | ||
| Inadequate empiric therapy | 2.5 (1.52, 4.15) | 0.0003 | 2.3 (1.18,4.34) | 0.01 |
| Production of ESBL or KPC | 1.7 (1.03, 2.80) | 0.04 | ||
| Diabetes | 1.0 (0.59, 1.69) | 0.99 | ||
| Chronic renal failure | 1.1 (0.57, 2.03) | 0.82 | ||
| Dialysis | 1.0 (0.47, 2.09) | 0.98 | ||
| Chronic obstructive lung disease | 1.7 (0.90, 3.28) | 0.10 | ||
| Cardiovascular disease | 1.8 (1.08, 3.12) | 0.02 | ||
| Peripheral vascular disease | 0.8 (0.16, 3.61) | 0.74 | ||
| Cerebrovascular disease | 2.2 (1.08, 4.65) | 0.03 | ||
| Peptic ulcer disease | 1.3 (0.34, 5.06) | 0.69 | ||
| Liver disease | 2.2 (1.17, 4.15) | 0.01 | ||
| Malignancy | 1.9 (1.12, 3.37) | 0.02 | 2.7 (1.31,5.41) | 0.007 |
| Solid organ malignancy | 1.7 (0.92, 3.00) | 0.09 | ||
| Transplant | 0.4 (0.15, 0.88) | 0.02 | ||
| Hospitalization in the past year | 1.7 (0.90, 3.17) | 0.10 | ||
| ICU admission in the past year | 1.3 (0.79, 2.12) | 0.30 | ||
| Any surgery in the past year | 0.8 (0.47, 1.29) | 0.33 | ||
| Outpatient intravenous therapy in the past year | 1.3 (0.59, 2.73) | 0.54 | ||
| Any surgery in the past 30 days | 1.1 (0.58, 2.17) | 0.73 | ||
| Antibiotics with Gram-negative activity in the past 30 days | 1.1 (0.65, 1.76) | 0.80 | ||
| Presence of tracheostomy tube | 0.9 (0.48, 1.81) | 0.83 | ||
| Presence of gastrostomy tube | 1.1 (0.51, 2.49) | 0.76 | ||
| Presence of indwelling urinary catheter | 1.8 (1.03, 3.12) | 0.04 | ||
| Presence of vascular catheter | 0.9 (0.58, 1.54) | 0.83 | ||
OR, odds ratio.
95% CI, 95% confidence interval.
The predictors of mortality were then assessed for the ESBL group. Inadequate empiric therapy (OR, 6.30; 95% CI, 1.64–24.3; P = 0.007) and Apache II score (OR, 1.14; 95% CI, 1.06–1.24; P = 0.0009) were independent predictors of mortality. However, when the analysis was performed for the control group, inadequate empiric therapy, which 11% of the patients received, was no longer a predictor (P = 0.34). Apache II score (OR, 1.24; 95% CI, 1.16–1.32; P < 0.001), onset of bacteremia while in ICU (OR, 3.97; 95% CI, 1.77–8.92; P < 0.001) and malignancy (OR, 2.62; 95% CI, 1.09–6.27; P = 0.03) remained as independent predictors in the control group. Mortality analysis for the KPC group was deferred due to the small number of cases.
Types of ESBLs and KPC
The types of ESBLs, including KPC-type β-lactamases, are summarized in Table 4. Of 29 E. coli cases for which isolates were available, 21 (72.4%) had CTX-M-type ESBLs, of which 11 (52%) were CTX-M-15. Only 7 (24.1%) had SHV or TEM-type ESBLs. For 73 Klesbiella spp. cases for which isolates were available, 20 (27.4%) were KPC producers (all K. pneumoniae). They were found to produce either KPC-2 (n = 8) or KPC-3 (n = 12). Of the remaining 53 ESBL-, non-KPC-producing Klesbiella spp., 47 (88.7%) and 6 (11.3%) had SHV- and CTX-M-type ESBLs, respectively. For Enterobacter spp., 14 of 17 (82.4%) ESBL cases produced SHV-type ESBLs, whereas 3 (17.6%) produced CTX-M-type ESBLs.
Table 4.
Types of ESBLs and KPC produced by the study isolates.
| ESBL | Species (number of isolates) | ||
|---|---|---|---|
| E. coli | Klebsiella spp. | Enterobacter spp. | |
| CTX-M-1 group | 13 | 2 | - |
| CTX-M-2 group | 2 | 3 | - |
| CTX-M-9 group | 6 | 1 | 3 |
| SHV | 6 | 47 | 14 |
| TEM | 1 | - | - |
| KPC | - | 20 | - |
| Others | 1a | 1b | - |
| Unknown | 3 | 2 | - |
| Total | 32 | 76 | 17 |
CMY-33 β-lactamase.
OXY-2 β-lactamase of K. oxytoca.
DISCUSSION
The emergence and spread of CTX-M-type ESBLs and KPC-type carbapenemases are the two major developments in β-lactam resistance that have taken place in the last decade. CTX-M-type ESBLs became the predominant ESBLs in E. coli in many parts of the world by early 2000s [3]. While initially felt rare, recent reports indicate that CTX-M-type ESBLs are becoming more common in the United States as well, especially in E. coli [4]. KPC-type carbapenemases rapidly spread among K. pneumoniae, which have caused nosocomial outbreaks in hospitals throughout the East Coast of the United States [7]. Studies from the United States that address clinical characteristics of bacteremia due to ESBL-producing Enterobacteriaceae were conducted either before the advent of these newer β-lactamases or without molecular characterization [2, 11]. The present study was thus undertaken to identify the clinical features of bacteremia due to ESBL and KPC-producing Enterobacteriaceae in a contemporary cohort of patients from tertiary hospitals in the Northeastern United States.
Our results indicate that CTX-M-type ESBLs are now common among bacteremic isolates of Enterobactericeae in this geographic area. In E. coli, the majority had this group of ESBLs. About half of them had CTX-M-15, which is increasingly recognized as the most common CTX-M-type ESBL in many parts of the world [12], but CTX-M-2 and 9 groups were also present, which are known to predominate in South America and Spain, respectively [3]. SHV and TEM-type ESBLs comprised the rest, though much less common than CTX-M-type ESBLs. This trend was also reported in a recent study of ESBL-producing E. coli conducted in Philadelphia [13]. Taken together, we may assume that CTX-M-type ESBLs are now prevalent in this species in the Northeastern United States. In Klebsiella spp. and Enterobacter spp., SHV was still the most common ESBL type, but CTX-M-type ESBLs were also present in both organisms. All 20 KPC-producing organisms were from the study site in New York City, where they are known to be endemic [14].
The 28-day mortality of patients with bacteremia due to organisms producing ESBL or KPC was significantly higher at 29.1% compared with 19.5% for those in the control group. This is in line with the results of a recent meta-analysis that pooled data from 16 studies and showed significantly increased mortality in ESBL-asscociated bacteremia [1]. The subset of patients with KPC-producing K. pneumoniae in our study had even higher mortality at 47.4% as opposed to 27.5% for those with ESBL-producing K. pneumoniae. While not statistically significant likely due to the relatively small number of patients, our data strengthen the previously reported association between KPC-producing K. pneumoniae bacteremia and increased mortality [14].
The independent risk factors for ESBL-associated bacteremia included recent antibiotic use, admission from nursing home, length of hospital stay prior to infection, chronic renal failure, presence of a gastrostomy tube and history of transplant. Recent antibiotic use has been pointed out as an independent risk factor in multiple studies that investigated ESBL-associated bacteremia [11, 15–18]. Admission from nursing home has also been identified as a risk factor for ESBL production in several studies investigating bacteremia [19–21]. The presence of a gastrostomy tube has been associated with acquisition of ESBL-producing organisms among nursing home residents [22, 23], whereas chronic renal failure has not been associated with acquisition of ESBL-producing organisms. Independent risk factors for mortality in the ESBL group included higher Apache II score at the time of positive blood culture and inappropriate empiric therapy. Association of higher Apache II score with increased mortality is not surprising and this was an independent risk factor for mortality in control cases as well. Inappropriate empiric therapy, defined by use of antibiotic(s) that lack in vitro activity against the causative organism, has been identified as independent risk factor for death in ESBL-associated bacteremia in several studies [24, 25]. We may therefore speculate that the significantly increased mortality with inappropriate empiric therapy, which occurred more frequently in the ESBL and KPC groups than the control group, contributed to the excess mortality. Our study has several limitations. First, we analyzed ESBL-producing Enterobacteriaceae together, not by species. The selection of controls to be in the same proportions by species as the cases helped address some of the potential bias arising from this. Second, our analysis of the KPC cases was limited due to the relatively small number of cases. Finally, we were not able to associate types of ESBLs with clinical attributes because of their heterogeneity.
In conclusion, mortality associated with bacteremia due to ESBL or KPC-producing Enterobactereciae continues to be high despite improved detection in the clinical laboratory. CTX-M-type ESBLs are now common in E. coli and present in other species as well in Northeastern United States. The frequent occurrence of delay in appropriate therapy likely contributes to the excess mortality. Use of antimicrobials with activity against ESBL-producing organisms such as carbapenems is generally accepted as appropriate in unwell patients with risk factors. However, the advent of K. pneumoniae producing KPC-type carbapenemase further complicates this issue.
Acknowledgments
The authors thank L. G. Clarke for his assistance in data collection; we also thank the members of the clinical microbiology laboratory staff at the participating sites for identification and collection of study isolates. Y. Doi was supported by a career development award from the National Institute of Allergy and Infectious Diseases (K22AI080584).
Funding: AstraZeneca provided research support for this study, but was not involved with the study design, completion, data analysis, or writing of the manuscript.
Footnotes
Transparency declarations: D. L. Paterson has served on advisory boards for Novartis, AstraZeneca, Merck, Johnson & Johnson, Cubist, Leo Pharmaceuticals and Pfizer. A. Y. Peleg has acted as an advisor to Abbott Molecular and Ortho-McNeil-Janssen. B. Polsky is a member of the speakers’ bureau for Ortho-McNeil. Y. Doi has served on advisory board for Pfizer and has received research support from Merck. Other authors report no potential conflict of interest.
References
- 1.Schwaber MJ, Carmeli Y. Mortality and delay in effective therapy associated with extended-spectrum β-lactamase production in Enterobacteriaceae bacteraemia: a systematic review and meta-analysis. J Antimicrob Chemother. 2007;60:913–920. doi: 10.1093/jac/dkm318. [DOI] [PubMed] [Google Scholar]
- 2.Zaoutis TE, Goyal M, Chu JH, et al. Risk factors for and outcomes of bloodstream infection caused by extended-spectrum β-lactamase-producing Escherichia coli and Klebsiella species in children. Pediatrics. 2005;115:942–949. doi: 10.1542/peds.2004-1289. [DOI] [PubMed] [Google Scholar]
- 3.Rossolini GM, D’Andrea MM, Mugnaioli C. The spread of CTX-M-type extended-spectrum β-lactamases. Clin Microbiol Infect. 2008;14 (Suppl 1):33–41. doi: 10.1111/j.1469-0691.2007.01867.x. [DOI] [PubMed] [Google Scholar]
- 4.Castanheira M, Mendes RE, Rhomberg PR, et al. Rapid emergence of blaCTX-M among Enterobacteriaceae in U.S. Medical Centers: molecular evaluation from the MYSTIC Program (2007) Microb Drug Resist. 2008;14:211–216. doi: 10.1089/mdr.2008.0827. [DOI] [PubMed] [Google Scholar]
- 5.Lewis JS, 2nd, Herrera M, Wickes B, et al. First report of the emergence of CTX-M-type extended-spectrum β-lactamases (ESBLs) as the predominant ESBL isolated in a U.S. health care system. Antimicrob Agents Chemother. 2007;51:4015–4021. doi: 10.1128/AAC.00576-07. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Moland ES, Black JA, Hossain A, et al. Discovery of CTX-M-like extended-spectrum β-lactamases in Escherichia coli isolates from five US States. Antimicrob Agents Chemother. 2003;47:2382–2383. doi: 10.1128/AAC.47.7.2382-2383.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Nordmann P, Cuzon G, Naas T. The real threat of Klebsiella pneumoniae carbapenemase-producing bacteria. Lancet Infect Dis. 2009;9:228–236. doi: 10.1016/S1473-3099(09)70054-4. [DOI] [PubMed] [Google Scholar]
- 8.de Oliveira Garcia D, Doi Y, Szabo D, et al. Multiclonal outbreak of Klebsiella pneumoniae producing extended-spectrum β-lactamase CTX-M-2 and novel variant CTX-M-59 in a neonatal intensive care unit in Brazil. Antimicrob Agents Chemother. 2008;52:1790–1793. doi: 10.1128/AAC.01440-07. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Qureshi ZA, Paterson DL, Pakstis DL, et al. Risk factors and outcome of extended-spectrum β-lactamase-producing Enterobacter cloacae bloodstream infections. Int J Antimicrob Agents. 37:26–32. doi: 10.1016/j.ijantimicag.2010.09.009. [DOI] [PubMed] [Google Scholar]
- 10.Doi Y, Potoski BA, Adams-Haduch JM, et al. Simple disk-based method for detection of Klebsiella pneumoniae carbapenemase-type β-lactamase by use of a boronic acid compound. J Clin Microbiol. 2008;46:4083–4086. doi: 10.1128/JCM.01408-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Paterson DL, Ko WC, Von Gottberg A, et al. International prospective study of Klebsiella pneumoniae bacteremia: implications of extended-spectrum β-lactamase production in nosocomial Infections. Ann Intern Med. 2004;140:26–32. doi: 10.7326/0003-4819-140-1-200401060-00008. [DOI] [PubMed] [Google Scholar]
- 12.Nicolas-Chanoine MH, Blanco J, Leflon-Guibout V, et al. Intercontinental emergence of Escherichia coli clone O25:H4-ST131 producing CTX-M-15. J Antimicrob Chemother. 2008;61:273–281. doi: 10.1093/jac/dkm464. [DOI] [PubMed] [Google Scholar]
- 13.McGettigan SE, Hu B, Andreacchio K, et al. Prevalence of CTX-M β-lactamases in Philadelphia. J Clin Microbiol. 2009;47:2970–2974. doi: 10.1128/JCM.00319-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Bratu S, Landman D, Haag R, et al. Rapid spread of carbapenem-resistant Klebsiella pneumoniae in New York City: a new threat to our antibiotic armamentarium. Arch Intern Med. 2005;165:1430–1435. doi: 10.1001/archinte.165.12.1430. [DOI] [PubMed] [Google Scholar]
- 15.Qavi A, Segal-Maurer S, Mariano N, et al. Increased mortality associated with a clonal outbreak of ceftazidime-resistant Klebsiella pneumoniae: a case-control study. Infect Control Hosp Epidemiol. 2005;26:63–68. doi: 10.1086/502488. [DOI] [PubMed] [Google Scholar]
- 16.Du B, Long Y, Liu H, et al. Extended-spectrum β-lactamase-producing Escherichia coli and Klebsiella pneumoniae bloodstream infection: risk factors and clinical outcome. Intensive Care Med. 2002;28:1718–1723. doi: 10.1007/s00134-002-1521-1. [DOI] [PubMed] [Google Scholar]
- 17.Tumbarello M, Spanu T, Sanguinetti M, et al. Bloodstream infections caused by extended-spectrum-β-lactamase-producing Klebsiella pneumoniae: risk factors, molecular epidemiology, and clinical outcome. Antimicrob Agents Chemother. 2006;50:498–504. doi: 10.1128/AAC.50.2.498-504.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Rodriguez-Bano J, Navarro MD, Romero L, et al. Risk-factors for emerging bloodstream infections caused by extended-spectrum β-lactamase-producing Escherichia coli. Clin Microbiol Infect. 2008;14:180–183. doi: 10.1111/j.1469-0691.2007.01884.x. [DOI] [PubMed] [Google Scholar]
- 19.Endimiani A, Luzzaro F, Brigante G, et al. Proteus mirabilis bloodstream infections: risk factors and treatment outcome related to the expression of extended-spectrum β-lactamases. Antimicrob Agents Chemother. 2005;49:2598–2605. doi: 10.1128/AAC.49.7.2598-2605.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Schiappa DA, Hayden MK, Matushek MG, et al. Ceftazidime-resistant Klebsiella pneumoniae and Escherichia coli bloodstream infection: a case-control and molecular epidemiologic investigation. J Infect Dis. 1996;174:529–536. doi: 10.1093/infdis/174.3.529. [DOI] [PubMed] [Google Scholar]
- 21.Rodriguez-Bano J, Picon E, Gijon P, et al. Community-onset bacteremia due to extended-spectrum β-lactamase-producing Escherichia coli: risk factors and prognosis. Clin Infect Dis. 2010;50:40–48. doi: 10.1086/649537. [DOI] [PubMed] [Google Scholar]
- 22.Wiener J, Quinn JP, Bradford PA, et al. Multiple antibiotic-resistant Klebsiella and Escherichia coli in nursing homes. Jama. 1999;281:517–523. doi: 10.1001/jama.281.6.517. [DOI] [PubMed] [Google Scholar]
- 23.Mendelson G, Hait V, Ben-Israel J, et al. Prevalence and risk factors of extended-spectrum β-lactamase-producing Escherichia coli and Klebsiella pneumoniae in an Israeli long-term care facility. Eur J Clin Microbiol Infect Dis. 2005;24:17–22. doi: 10.1007/s10096-004-1264-8. [DOI] [PubMed] [Google Scholar]
- 24.Anderson DJ, Engemann JJ, Harrell LJ, et al. Predictors of mortality in patients with bloodstream infection due to ceftazidime-resistant Klebsiella pneumoniae. Antimicrob Agents Chemother. 2006;50:1715–1720. doi: 10.1128/AAC.50.5.1715-1720.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Tumbarello M, Sanguinetti M, Montuori E, et al. Predictors of mortality in patients with bloodstream infections caused by extended-spectrum-β-lactamase-producing Enterobacteriaceae: importance of inadequate initial antimicrobial treatment. Antimicrob Agents Chemother. 2007;51:1987–1994. doi: 10.1128/AAC.01509-06. [DOI] [PMC free article] [PubMed] [Google Scholar]