Abstract
Ampicillin-sulbactam is commonly used as an empirical therapy for invasive infections where Escherichia coli is a potential pathogen. We evaluated the clinical and microbiologic characteristics of bloodstream infection due to E. coli, with focus on cases that were nonsusceptible to ampicillin-sulbactam and not producing extended-spectrum β-lactamase (ESBL). Of a total of 357 unique bacteremic cases identified between 2005 and 2008, 111 (31.1%) were intermediate or resistant to ampicillin-sulbactam by disk testing. In multivariate analysis, a history of liver disease, organ transplant, peptic ulcer disease, and prior use of ampicillin-sulbactam were independent risk factors for bloodstream infection with ampicillin-sulbactam-nonsusceptible E. coli. Among cases that received ampicillin-sulbactam as an empirical therapy, an early clinical response was observed in 65% (22/34) of susceptible cases but in only 20% (1/5) of nonsusceptible cases. Among 50 ampicillin-sulbactam-resistant isolates examined, there was no clonal relatedness and no evidence of production of inhibitor-resistant TEM (IRT). Instead, the resistance was attributed to hyperproduction of TEM-1 β-lactamase in the majority of isolates. However, promoter sequences of blaTEM-1 did not predict resistance to ampicillin-sulbactam. While the plasmid copy number did not differ between representative resistant and susceptible isolates, the relative expression of blaTEM-1 was significantly higher in two of three resistant isolates than in three susceptible isolates. These results suggest high-level blaTEM-1 expression as the predominant cause of ampicillin-sulbactam resistance and also the presence of yet-unidentified factors promoting overexpression of blaTEM-1 in these isolates.
Ampicillin-sulbactam is a β-lactam-β-lactamase inhibitor combination that has activity against a broad range of bacteria, including Gram-positive, Gram-negative, and anaerobic organisms. Available as an intravenous formulation in the United States, it is commonly used for empirical management of various conditions such as bacteremia, pyelonephritis, pneumonia, skin and soft tissue infections, and intra-abdominal infections (1). In the past, ampicillin-sulbactam has been active against Escherichia coli, as sulbactam inhibits the TEM-1 β-lactamase that is produced by over 50% of E. coli isolates in some countries (32).
Resistance to ampicillin-sulbactam has been increasing in E. coli in the last decade. A worldwide survey of E. coli strains isolated from intra-abdominal infections showed a rate of susceptibility to this combination of 46.3% in 2005, a drop from 55.5% in 2002 (3). At our institution, the susceptibility of E. coli to ampicillin-sulbactam has been maintained relatively well, but a decrease from 81% to 71% has been observed between 2004 and 2007, raising concern that the agent may no longer provide adequate coverage when used as empirical therapy for suspected E. coli infection.
Mechanisms of resistance of E. coli to β-lactam-β-lactamase combinations have been best studied for amoxicillin-clavulanic acid. These include hyperproduction of plasmid-mediated TEM-1 β-lactamase (38), production of inhibitor-resistant TEM (IRT) β-lactamases (26, 35), production of plasmid-mediated β-lactamases (including OXA-1 [39] and AmpC-type β-lactamases and extended-spectrum β-lactamases [ESBLs] [17]), and hyperproduction of chromosomal AmpC β-lactamase (36). Production of carbapenemases (for example, Klebsiella pneumoniae carbapenemase [KPC]) would also be expected to result in resistance to β-lactam-β-lactamase combinations (28). Deficiency of OmpC or OmpF porin does not affect the level of resistance per se but may contribute to resistance when combined with hyperproduction of TEM-1 (30). While there are relatively few data regarding mechanisms of resistance to ampicillin-sulbactam, it is believed that the same mechanisms affecting amoxicillin-clavulanic acid affect ampicillin-sulbactam as well (8). The bulk of the above data come from Europe, and data regarding mechanisms of resistance to β-lactam-β-lactamase combinations in isolates from the United States have been limited. In light of the continuing trend toward lower rates of susceptibility of E. coli to ampicillin-sulbactam, the goal of the present study was 2-fold: (i) to identify risk factors for and assess the clinical outcome of bacteremia due to ampicillin-sulbactam-nonsusceptible E. coli and (ii) to characterize the predominant mechanism of resistance against this agent in these isolates.
MATERIALS AND METHODS
Study design and patients.
A retrospective cohort of patients with bacteremia due to E. coli was identified at the University of Pittsburgh Medical Center, an 834-bed tertiary teaching hospital with affiliated outpatient clinics, between 2005 and 2008. Cases confirmed as caused by ESBL-producing E. coli in the clinical laboratory according to the criteria of the Clinical and Laboratory Standards Institute (CLSI) (11) were excluded from both clinical analysis and laboratory investigation. No carbapenem-resistant E. coli was present in the study sample. Clinical data were collected from electronic medical records, deidentified, and provided to the investigators. The study was approved by the Institutional Review Board of the University of Pittsburgh.
Risk factors.
Risk factors for bacteremia due to ampicillin-sulbactam-nonsusceptible E. coli were investigated by comparing patients with bacteremia caused by ampicillin-sulbactam-susceptible and -nonsusceptible isolates. The parameters reviewed included the following: age; sex; race; source of infection; immune status; presence of chronic renal failure, liver disease, chronic obstructive lung disease, cardiovascular disease, peripheral vascular disease, cerebrovascular disease, peptic ulcer disease, or malignancy; history of organ transplant, transfer from nursing home, prior use of antibiotics with activity against Gram-negative organisms in the last 30 days, hospitalization, surgery, or intensive care unit (ICU) admission in the last year; and the presence of a tracheostomy tube, gastrostomy tube, intravenous line, or urinary catheter at the time of infection. The Apache ΙΙ score (19) and modified McCabe-Jackson classification (27) were used to assess severity of illness and prognosis, respectively. An immunocompromised state was defined as the presence of diabetes mellitus, neutropenia, or HIV infection, or receipt of steroids or other immunosuppressive therapy in the last 30 days.
Clinical outcome measures.
Patients with bacteremia due to E. coli were monitored after the onset of bacteremia to assess early clinical response at 96 h and 28-day mortality in response to antimicrobial therapy. Empirical antibiotic therapy was defined as administration of an antibiotic with in vitro activity against E. coli for at least 48 of the first 96 h after the first positive blood culture. Early clinical response was defined by all of the following occurring before or at 96 h from the time of first positive blood culture for a continuous 24-hour period: (i) temperature from 36.0 to 37.9°C, (ii) white blood cell count of 4,000 to 10,900/μl, (iii) no requirement for vasopressors, and (iv) systolic blood pressure of greater than 90 mm Hg. A case was documented as clinical failure when any of these criteria was not met. The primary endpoints of the study were clinical response at 96 h and mortality during the 28-day period after the first blood culture positive for E. coli.
Susceptibility testing.
For all available case-defining E. coli isolates, susceptibility testing was conducted using the disk diffusion method for ampicillin, ampicillin-sulbactam, amoxicillin-clavulanic acid, ticarcillin, ticarcillin-clavulanic acid, piperacillin, piperacillin-tazobactam, cephalothin, cefoxitin, cefuroxime, cefpodoxime, ceftazidime, cefotaxime, sulfamethoxazole-trimethoprim, gentamicin, and ciprofloxacin (BD, Sparks, MD). In addition, the MIC was determined for ampicillin-sulbactam at a fixed ratio of 2:1 using the agar dilution method. This method has been reported to better predict the in vitro antibacterial effect of this agent than the one using a fixed concentration of sulbactam in an in vitro pharmacokinetic model (15). Isolates that had high-level resistance to ampicillin-sulbactam, as implied by the absence of any inhibitory zone, were selected for further laboratory investigation, as described below.
aIEF.
Analytical isoelectric focusing (aIEF) was performed as described previously (20). Enzyme activity was detected by placing filter paper soaked in 500 μg/ml of nitrocefin over the focused gel. Isolates producing TEM-1, SHV-1, and CMY-2 were used as controls.
PCR and sequencing.
PCR analysis was performed to detect blaTEM-, blaSHV-, and blaOXA-1-type β-lactamase genes as previously described (4, 13). For isolates that had a phenotype consistent with AmpC β-lactamase production (i.e., reduced susceptibility to cefoxitin, cefuroxime, cefpodoxime, and ceftazidime), multiplex PCR to detect known groups of plasmid-mediated AmpC β-lactamases was conducted (29).
For isolates that were positive for the blaTEM gene, a second PCR to amplify the entire coding region as well as the promoter region was conducted using primers TEM prom F (5′-ATAAAATTCTTGAAGAC-3′) and TEM prom R (5′-TTACCAATGCTTAATCA-3′). Sequencing of the amplified products was conducted with the same primers as well as two internal primers (461F [5′-GGGCAAGAGCAACTCGG-3′] and 478R [5′-ACCGAGTTGCTCTTGCCC-3′]) using an ABI3730 instrument (Applied Biosystems, Foster City, CA).
Phylogenetic typing and PCR-based detection of the O25b-ST131 clone were performed as described by Clermont et al. (9, 10).
Transformation of plasmids mediating ampicillin-sulbactam resistance.
Plasmids were extracted from the study isolates using the alkaline lysis method as previously described (34) and transformed into E. coli DH10B by electroporation. Transformants were selected on LB agar containing 50 μg/ml of ampicillin (Sigma-Aldrich, St. Louis, MO). The transformants underwent susceptibility testing as well as PCR analysis to confirm transfer of the blaTEM gene.
PFGE.
The study isolates were subjected to pulsed-field gel electrophoresis (PFGE) to evaluate for clonality. Genomic DNA was prepared as described previously (31), digested with XbaI (New England Biolabs, Ipswich, MA), and subjected to electrophoresis using a CHEF DR III system (Bio-Rad, Hercules, CA) at 6 V with pulse time of 2.2 to 54.2 s and linear ramping at 14°C for 22 h. A lambda ladder (New England Biolabs, Ipswich, MA) was used as the DNA size marker. The relatedness of PFGE patterns was determined by the unweighted-pair group method using average linkages and the DICE setting clustering analysis on the Bionumerics software version 4.0 (Applied Maths, Sint-Martens-Latem, Belgium). The results were interpreted based on the criteria described by Tenover et al. (37).
Copy number assay.
To determine whether there were differences in the copy number of blaTEM-positive plasmid between isolates with high and low ampicillin-sulbactam MICs, three blaTEM-positive isolates with high ampicillin-sulbactam MICs (two with 256/128 μg/ml and one with 512/128 μg/ml) and three blaTEM-positive isolates with low MICs (two with 32/16 μg/ml and one with 16/8 μg/ml) were selected. The isolates used in this assay were clonally unrelated as determined by PFGE.
The copy number assay was conducted using primer pairs specific to blaTEM and dxs, a housekeeping gene, as described previously (22). Genomic DNA was extracted using the DNeasy blood and tissue kit (Qiagen, Valencia, CA) and assayed with a 7500 real-time system (Applied Biosystems) with the initial hot start of 95°C for 10 min and then 40 cycles of 95°C for 15 s and 58.5°C for 1 min. Relative copy number difference was determined using the 2−ΔΔCT calculation as previously described (22).
Relative quantification of blaTEM expression.
The same isolates used in the copy number assay were included in the expression study. For extraction of total RNA, 30 ml of Luria-Bertani (LB) broth was inoculated with three colonies from a blood agar plate incubated overnight for each of the six isolates. Cultures were incubated at 37°C with shaking (150 rpm) until they reached an optical density at 600 nm (OD600) of 0.5, at which point the bacteria was harvested. Total RNA was extracted using the RNeasy maxikit (Qiagen) according to the manufacturer's instructions.
After treatment with DNase I (Ambion, Austin, TX), RNAs from all six isolates were reverse transcribed into cDNAs using the High Capacity RNA to cDNA kit (Applied Biosystems) with the same primer pairs specific to blaTEM and dxs that were used in the copy number assay. The cDNA was run in an Applied Biosystems 7500 real-time system with the initial hot start of 95°C for 10 min and then 40 cycles of 95°C for 15 s and 58.5°C for 1 min. Nontemplate and non-reverse transcriptase controls were included for each run. Dissociation curves were also obtained to analyze the specificity of the PCR products.
A primer concentration of 400 nM each for blaTEM and dxs and 10 ng of template RNA were found to be optimal. Validation of the 2−ΔΔCT method was performed by measuring the efficiencies of blaTEM and dxs amplification. The amplification efficiency (E) of each target was determined by comparing amplifications using a serial dilution of RNA sample and calculating E using a previously described equation (22). The expression of dxs was used to normalize the relative expression data for blaTEM. The cycle thresholds (CT) for each gene were averaged using biological duplicates, and the relative expression was calculated using RQ = 2−ΔΔCT (25).
Statistical analysis.
A multivariable logistic regression model was used to assess the risk factors. The logistic model was developed by entering all variables that had P values of ≤0.2 in the univariate analyses into the initial model. Estimated odds ratios (ORs) and 95% confidence intervals (CIs) were estimated from this model. For the outcome analyses, the chi-square test, Fisher exact test, or median unbiased estimate was used to compare categorical variables, and the Mann-Whitney U test was used to compare continuous variables. All analyses were performed using SAS version 9.2 (SAS Institute Inc., Cary, NC), and a P value of 0.05 was considered statistically significant.
RESULTS
Study population.
A total of 370 cases of E. coli bacteremia were identified during the period between July 2005 and January 2008. Of these, 13 cases were excluded since they were caused by strains that were confirmed as producing ESBL by the phenotypic method endorsed by the CLSI (11). Of the remainder, 111 (31.1%) cases were reported as nonsusceptible to ampicillin-sulbactam using the disk diffusion testing in the clinical microbiology laboratory and constituted the case group in the clinical analysis. Sixty-three of the 111 cases were resistant to ampicillin-sulbactam, of which 50 had corresponding isolates available for laboratory studies.
Risk factors.
The demographics of the patients and risk factors associated with bacteremia due to ampicillin-sulbactam-nonsusceptible and -susceptible isolates are listed in Table 1. The ampicillin-sulbactam-nonsusceptible cases were more likely than the ampicillin-sulbactam-susceptible cases to be male. The degrees of health care exposure were similar in both groups, and accordingly, there was no significant difference in terms of the site of acquisition (community associated, health care associated, or hospital acquired) (data not shown). Severities of illness, as measured by median Apache II score and use of vasopressors, were comparable, and urine was the most common primary source of bacteremia in both groups. Risk factors that revealed significant association with nonsusceptibility in the univariate analysis included history of solid organ transplant and peptic ulcer disease. In regard to antimicrobial use in the previous 28 days, the use of ampicillin-sulbactam was a significant risk factor for ampicillin-sulbactam nonsusceptibility.
TABLE 1.
Risk factors for ampicillin-sulbactam nonsusceptibility with E. coli bacteremia
| Type of analysis and variable | Valuea for: |
P value | OR | 95% CI | |
|---|---|---|---|---|---|
| Cases (n = 111) | Controls (n = 246) | ||||
| Univariate analysis | |||||
| Demographics | |||||
| Median age, yr (range) | 60 (21-96) | 63 (20-94) | |||
| Age ≥65 yr | 41 (36.9) | 112 (45.5) | 0.13 | 0.70 | 0.44, 1.11 |
| Male gender | 58 (52.3) | 98 (39.8) | 0.03 | 1.65 | 1.05, 2.60 |
| White | 81 (76.4) | 176 (77.9) | 0.77 | 0.92 | 0.53, 1.59 |
| Black | 22 (20.8) | 44 (19.5) | 0.78 | 1.08 | 0.61, 1.92 |
| Health care exposures | |||||
| Admitted from nursing home | 23 (20.7) | 42 (17.1) | 0.41 | 1.27 | 0.72, 2.24 |
| Prior surgery within 30 days | 15 (13.5) | 28 (11.4) | 0.58 | 1.21 | 0.62, 2.37 |
| Any hospitalization in past yr | 88 (79.3) | 193 (78.5) | 0.86 | 1.05 | 0.61, 1.82 |
| Any ICU admission in past yr | 33 (29.7) | 63 (25.6) | 0.42 | 1.23 | 0.75, 2.02 |
| Any surgery in past yr | 32 (28.8) | 55 (22.4) | 0.19 | 1.41 | 0.85, 2.34 |
| Dialysis in past yr | 10 (9.0) | 26 (10.6) | 0.65 | 0.84 | 0.39, 1.80 |
| Outpatient intravenous therapy in past yr | 5 (4.5) | 7 (2.8) | 0.42 | 1.61 | 0.50, 5.19 |
| Devices | |||||
| Tracheostomy tube | 4 (3.6) | 14 (5.7) | 0.41 | 0.62 | 0.20, 1.93 |
| Gastrostomy tube | 2 (1.8) | 2 (0.8) | 0.42 | 2.24 | 0.31, 16.10 |
| Indwelling urinary catheter | 20 (18.0) | 28 (11.4) | 0.09 | 1.71 | 0.92, 3.19 |
| Vascular catheter | 60 (54.1) | 118 (48.0) | 0.29 | 1.28 | 0.81, 2.00 |
| Underlying conditions | |||||
| Diabetes | 27 (24.3) | 83 (33.9) | 0.07 | 0.63 | 0.38, 1.04 |
| Chronic obstructive pulmonary disease | 5 (4.5) | 26 (10.6) | 0.07 | 0.40 | 0.15, 1.06 |
| Cardiovascular disease | 27 (24.3) | 57 (23.3) | 0.83 | 1.06 | 0.63, 1.79 |
| Peripheral vascular disease | 1 (0.9) | 2 (0.8) | 0.93 | 1.11 | 0.10, 12.32 |
| Cerebrovascular disease | 4 (3.6) | 16 (6.5) | 0.27 | 0.54 | 0.17, 1.64 |
| Immunocompromised | 59 (53.2) | 122 (49.6) | 0.53 | 1.15 | 0.74, 1.81 |
| Chronic renal failure | 12 (10.8) | 22 (9.0) | 0.59 | 1.23 | 0.58, 2.58 |
| Peptic ulcer disease | 6 (5.4) | 1 (0.4) | 0.02 | 13.9 | 1.66, 117.12 |
| Liver disease | 22 (19.8) | 29 (11.8) | 0.05 | 1.84 | 1.00, 3.38 |
| Malignancy | 8 (7.2) | 23 (9.4) | 0.50 | 0.75 | 0.32, 1.73 |
| Transplant | 30 (27.0) | 39 (15.9) | 0.02 | 1.96 | 1.14, 3.36 |
| Sources of bacteremia | |||||
| Pneumonia (baseline) | 4 (3.6) | 14 (5.7) | 0.26 | ||
| UTI | 49 (44.1) | 103 (41.9) | 1.67 | 0.52, 5.32 | |
| Intra-abdominal | 6 (5.4) | 7 (2.8) | 3.00 | 0.63, 14.23 | |
| Line related | 5 (4.5) | 14 (5.7) | 1.25 | 0.28, 5.65 | |
| Other | 4 (3.6) | 1 (0.4) | 14.00 | 1.20, 163.37 | |
| Unknown | 43 (38.7) | 107 (43.5) | 1.41 | 0.44, 4.51 | |
| Severity of illness | |||||
| Mean Apache II score (range) | 11 (0-42) | 11 (0-35) | |||
| Apache II score of ≥25 | 9 (8.7) | 16 (6.9) | 0.56 | 1.28 | 0.55, 3.01 |
| Vasopressor use | 18 (16.4) | 39 (16.0) | 0.94 | 1.02 | 0.56, 1.88 |
| Antimicrobials in 28 days prior to bacteremia | |||||
| Any | 29 (26.1) | 38 (15.4) | 0.02 | 1.94 | 1.12, 3.34 |
| Oxyimino-cephalosporinsb | 1 (0.9) | 6 (2.4) | 0.35 | 0.36 | 0.04, 3.06 |
| β-Lactam-β-lactamase inhibitor combinationsc | 16 (14.4) | 17 (6.9) | 0.03 | 2.27 | 1.10, 4.68 |
| Ampicillin-sulbactam | 13 (11.7) | 7 (2.8) | 0.002 | 4.53 | 1.75, 11.69 |
| Carbapenemsd | 0 (0.0) | 1 (0.4) | 1.00 | 2.22 | 0.00, 86.43 |
| Fluoroquinolonese | 2 (1.8) | 7 (2.8) | 0.56 | 0.63 | 0.13, 3.07 |
| Multivariate analysis | |||||
| Ampicillin-sulbactam in 28 days prior to bacteremia | 0.011 | 3.651 | 1.35, 9.85 | ||
| Liver disease | 0.019 | 2.131 | 1.13, 4.00 | ||
| Peptic ulcer disease | 0.012 | 15.837 | 1.84, 136.46 | ||
| Transplant | 0.024 | 1.945 | 1.09, 3.46 | ||
Unless otherwise indicated, values are number (percentage) of cases.
Cefuroxime, cefotaxime, ceftriaxone, ceftazidime, and cefepime.
Ampicillin-sulbactam, amoxicillin-clavulanate, and piperacillin-tazobactam.
Ertapenem, imipenem, and meropenem.
Ciprofloxacin, levofloxacin, and moxifloxacin.
In the multivariate analysis, a history of liver disease, peptic ulcer disease, or organ transplant and prior use of ampicillin-sulbactam remained as independent risk factors for ampicillin-sulbactam nonsusceptibility (Table 1).
Clinical outcome.
Clinical cure was achieved within 96 h in 60 (55%) and 146 (60%) of the patients in nonsusceptible and susceptible groups, respectively. The mortalities within 28 days were 20% and 17% for these two groups, respectively (P = 0.6). Among patients who received ampicillin-sulbactam as the empirical therapy, 65% (22/34) of those with a susceptible isolate and 20% (1/5) of those with a nonsusceptible isolate had clinical cure within 96 h (P = 0.14). Among patients with ampicillin-sulbactam-nonsusceptible isolates, 73% (29/40) of those who received appropriate empirical therapy (i.e., with an agent to which the isolate was susceptible) and 30% (3/10) of those who received inappropriate empirical therapy (i.e., with an agent to which the isolate was nonsusceptible, including ampicillin-sulbactam) attained clinical cure within 96 h (P = 0.02).
Susceptibility testing.
Of the 63 ampicillin-sulbactam-resistant cases, isolates for 50 cases were available for further studies in the research laboratory. Upon MIC testing, 46 (92%) were resistant and 4 (8%) were intermediate to ampicillin-sulbactam. Susceptibility testing results for these 50 isolates with other β-lactams and penicillin-β-lactamase inhibitor combinations are shown in Table 2. For non-β-lactam agents, the rates of nonsusceptibility to gentamicin, ciprofloxacin, and sulfamethoxazole-trimethoprim were 12, 26, and 60%, respectively. None of these isolates possessed a phenotype consistent with production of known inhibitor-resistant TEM (IRT) β-lactamases, i.e., susceptibility to cephalothin (8). Thirty of the 50 isolates gave no inhibitory zone around the ampicillin-sulbactam disk, which was suggestive of high-level resistance to this agent.
TABLE 2.
β-Lactam susceptibilities of 50 ampicillin-sulbactam-resistant E. coli bacteremic isolates
| Antimicrobial | No. (%) of isolates |
||
|---|---|---|---|
| Susceptible | Intermediate | Resistant | |
| Ampicillin | 0 (0) | 0 (0) | 50 (100) |
| Amoxicillin-clavulanate | 1 (2) | 8 (16) | 41 (82) |
| Ticarcillin | 0 (0) | 0 (0) | 50 (100) |
| Ticarcillin-clavulanate | 4 (8) | 26 (52) | 20 (40) |
| Piperacillin | 0 (0) | 2 (4) | 48 (96) |
| Piperacillin-tazobactam | 36 (72) | 8 (16) | 6 (12) |
| Cephalothin | 0 (0) | 6 (12) | 44 (88) |
| Cefoxitin | 40 (80) | 2 (4) | 8 (16) |
| Cefuroxime | 28 (56) | 7 (14) | 15 (30) |
| Ceftazidime | 46 (92) | 1 (2) | 3 (6) |
PFGE, phylogenetic typing, and O25b-ST131 detection.
A dendrogram of the 50 available isolates constructed from PFGE revealed no clustering of the study isolates, indicating that they were not clonally related (data not shown). The distribution of phylogenetic types was as follows: A, 11 isolates (22%); B1, 3 isolates (6%); B2, 26 isolates (52%); and D, 10 isolates (20%). Of the 26 isolates belonging to phylogenetic type B2, 9 isolates (34.6%) were identified as O25b-ST131 using the PCR-based method.
aIEF.
Analytical IEF was performed on 37 representative isolates. It revealed a strong band at a pI of 5.4 in all except one isolate, consistent with the presence of TEM-1. Additional bands consistent with SHV-1 (pI, 7.6) and OXA-1 (pI, 7.3) were observed for two and one isolates, respectively, in a manner consistent with the PCR results described below, except for one isolate which was positive for blaOXA-1 by PCR but for which a band corresponding to OXA-1 could not be visualized by aIEF.
PCR for β-lactamase genes.
All four isolates that were nonsusceptible to ceftazidime were positive by PCR for the presence of a plasmid-mediated AmpC β-lactamase gene, including three isolates with a CMY-2-type β-lactamase and one isolate with a FOX-type β-lactamase. Forty-eight (96%), 3 (6%) and 4 (8%) isolates were positive for the blaTEM, blaSHV, and blaOXA-1 genes, respectively.
Transformants with ampicillin resistance.
Of the 30 clinical isolates that gave no inhibitory zone around the ampicillin-sulbactam disk, three carried CMY-2 or FOX-type β-lactamase genes by PCR, accounting for this phenotype. Of the remaining 27 isolates, all isolates except one were positive for the blaTEM gene by PCR, whereas two and three isolates were also positive for SHV and OXA-1, respectively. Plasmids were purified from these isolates and used to transform E. coli DH10B. As a result, ampicillin-resistant transformants were obtained from 23 of the 27 isolates. All these transformants were positive for the blaTEM gene by PCR, whereas one was positive for the OXA-1 gene as well. The MICs of ampicillin-sulbactam for the 27 clinical isolates and their transformants, when available, are listed in Fig. 1. The MICs for the clinical isolates and their transformants correlated well within the range of 32/16 to 128/64 μg/ml. However, increases in MICs over 128 μg/ml did not appear to be inherited by the transformants.
FIG. 1.
Dot plot of MICs of ampicillin-sulbactam-resistant E. coli bacteremic isolates and their blaTEM-1-positive transformants.
Sequencing of the blaTEM gene and its promoter region.
The blaTEM gene was sequenced for 22 of the 26 clinical isolates that gave no inhibitory zone around the ampicillin-sulbactam disk, were positive for the blaTEM gene, and were negative for plasmid-mediated AmpC β-lactamase genes. All sequences were consistent with TEM-1. The entire promoter region was further sequenced for 11 of these isolates. They were selected among the 26 isolates to represent various MICs of ampicillin-sulbactam. Several promoter sequences have been implicated in differential expression of the TEM-1 β-lactamase in E. coli (21). Of the 11 isolates, 6 isolates had the regular P3 promoter, whereas 5 isolates had the Pa/Pb promoter. The ampicillin-sulbactam MICs of the isolates with the P3 promoter were 32/16 (n = 2), 128/64 (n = 2), and 512/256 (n = 2) μg/ml. Those of isolates with the Pa/Pb promoter were 64/32 (n = 3) and 256/128 (n = 2) μg/ml.
Copy numbers of blaTEM-carrying plasmids.
The average copy numbers of blaTEM-carrying plasmids from the three isolates with high ampicillin-sulbactam MICs (256/128 to 512/256 μg/ml) and the three isolates with low ampicillin-sulbactam MICs (32/15 μg/ml) were calculated to be 4.1 and 3.5 copies per cell, respectively.
Relative quantification of blaTEM expression.
The relative expression of the blaTEM gene in the three isolates with high ampicillin-sulbactam MICs was calculated to be an average of 3.4-fold higher than that in the three isolates with low ampicillin-sulbactam MICs. As shown in Fig. 2, two of the three isolates with high ampicillin-sulbactam MICs (isolates A and C) had significantly higher expression of blaTEM-1 than the three isolates with low MICs (isolates D to F). Isolate B, which had an ampicillin-sulbactam MIC of 512/256 μg/ml, however, did not demonstrate high-level expression of blaTEM-1.
FIG. 2.
Relative expression of blaTEM-1 among isolates with high (A to C) and low (D to F) ampicillin-sulbactam MICs. The error bars indicate standard deviations. MICs represent those of ampicillin-sulbactam.
DISCUSSION
Due to its wide spectrum of activity, which includes activity against Gram-positive, Gram-negative, and anaerobic bacteria, ampicillin-sulbactam has been commonly used to treat infections such as urinary tract infections (UTI), community-acquired pneumonia, skin and soft tissue infection, and bacteremia. This wide spectrum is achieved by the addition of sulbactam, which inhibits most Ambler's class A β-lactamases that are present in numerous species, including E. coli. Ampicillin-sulbactam has therefore been considered an adequate empirical choice when E. coli is suspected to be involved in an infection. In recent years, however, resistance to this agent has become common in E. coli, reaching approximately 50% in some series from the United States (2, 14). Data from European countries suggest production of inhibitor-resistant TEM (IRT) β-lactamases as a major mechanism for this resistance, among others (8), but IRT has been reported for only a small number of isolates in the United States (6, 17). The purpose of the present study was to investigate the clinical features of bacteremia due to ampicillin-sulbactam-nonsusceptible E. coli and the mechanisms underlying the resistance.
Prior use of ampicillin-sulbactam was a strong independent risk factor for nonsusceptibility to this agent. Although we did not use the more rigorous case-case-control study design, this finding was in line with a previous study by Kaye et al., which identified prior administration of the penicillin class of antimicrobials, including ampicillin-sulbactam, as an independent risk factor for acquisition of ampicillin-sulbactam-nonsusceptible E. coli (18). This suggests that repeated courses of this agent put patients at substantial risk for a nonsusceptible organism, especially given the already high background rates of resistance to this agent. Histories of liver disease, organ transplant, and peptic ulcer disease were the other independent risk factors. The former two risk factors probably reflect the characteristics of our center, which has a robust organ transplant program. The significance of peptic ulcer disease was unclear given very small number of patients who had this condition in either group.
The 28-day mortality rates were 20 and 17% for nonsusceptible and susceptible cases, respectively, and over half of the patients in both groups had clinical cure within 96 h. However, among the five patients with a nonsusceptible organism who received empirical therapy with ampicillin-sulbactam, four experienced delayed clinical cure, whereas 22 of 34 patients with a nonsusceptible organism who received an appropriate empirical therapy had clinical cure within 96 h. This suggests that discordant therapy may indeed have some negative clinical impact. However, a larger sample size is necessary to make more rigorous comments about this observation.
The distribution of phylogenetic groups was consistent with other reports on bacteremic strains documenting the predominance of group B2, which is considered the most virulent among the four groups (5, 7, 12, 33). The finding of the presence of the O25b-ST131 clone among group B2 isolates is also in agreement with recent studies suggesting that this clone is now distributed among non-ESBL-producing as well as ESBL-producing E. coli strains (5, 16, 23).
We did not find any isolate with the phenotype commonly seen in IRT production. IRT is a relatively common mechanism of resistance to amoxicillin-clavulanate, along with hyperproduction of the chromosomal class C β-lactamase in some European countries, especially France (8, 24). Our finding is consistent with previous reports documenting the rarity of IRTs in the United States (6, 17). Instead, the majority of isolates with high-level resistance to ampicillin-sulbactam were able to reproduce a similar degree of resistance to this agent in the corresponding transformants that received plasmids encoding TEM-1, implicating hyperproduction of TEM-1 as the predominant reason for this resistance. We then investigated the mechanism underlying TEM-1 hyperproduction, with the hypothesis that it would be due to either a higher copy number of the blaTEM-1-encoding plasmids or overexpression of blaTEM-1 due to modifications of the promoter. Indeed, we found that the copy numbers of blaTEM-1-carrying plasmids were equivalent in isolates with higher and lower MICs of ampicillin-sulbactam, making this less likely to be the cause of TEM-1 hyperproduction. Instead, two of the three isolates with higher MICs displayed significantly higher levels of blaTEM-1 expression than those with lower MICs, which accounted for TEM-1 hyperproduction in these isolates. Promoter sequence variations that have been associated with overexpression of blaTEM-1 (21), however, could not be identified consistently. This suggests the presence of a yet-unidentified mechanism driving overexpression of blaTEM-1 and requires further investigation. Also, there were a number of isolates whose ampicillin-sulbactam MICs did not correlate well with those of their corresponding transformants, as seen in Fig. 1. Mechanisms other than hyperproduction of TEM-1, such as porin deficiency, may account for the resistance in these isolates.
In conclusion, nonsusceptibility to ampicillin-sulbactam is common among bacteremic E. coli isolates. Prior use of this agent is a strong risk factor for this resistance. While the 28-day mortalities are comparable for bacteremia due to these isolates and susceptible isolates, patients with nonsusceptible isolates may experience a delayed clinical response when empirical therapy is discordant. The predominant mechanism of ampicillin-sulbactam resistance in E. coli is hyperproduction of TEM-1, which likely represents an evolutionary response to overcome selective pressure from β-lactam-β-lactamase inhibitor combinations, especially ampicillin-sulbactam.
Acknowledgments
We thank Lloyd G. Clarke for his assistance with clinical data collection.
R.I.W.-T. was a recipient of a Medical Scholars Program scholarship from the Infectious Diseases Society of America for this study. Y.D. was supported by the National Institutes of Health (grants 1R03AI079296 and 1K22AI080584) and the Pennsylvania Department of Health (grant 4100047864). Database management was supported in part by AstraZeneca.
AstraZeneca was not involved with the study design or completion, data analysis, or writing of the manuscript.
Footnotes
Published ahead of print on 6 December 2010.
REFERENCES
- 1.Akova, M. 2008. Sulbactam-containing β-lactamase inhibitor combinations. Clin. Microbiol. Infect. 14(Suppl. 1):185-188. [DOI] [PubMed] [Google Scholar]
- 2.Al-Hasan, M. N., B. D. Lahr, J. E. Eckel-Passow, and L. M. Baddour. 2009. Antimicrobial resistance trends of Escherichia coli bloodstream isolates: a population-based study, 1998-2007. J. Antimicrob. Chemother. 64:169-174. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Baquero, F., et al. 2009. In vitro susceptibilities of aerobic and facultatively anaerobic gram-negative bacilli isolated from patients with intra-abdominal infections worldwide: 2005 results from Study for Monitoring Antimicrobial Resistance Trends (SMART). Surg. Infect. (Larchmt.) 10:99-104. [DOI] [PubMed] [Google Scholar]
- 4.Bert, F., C. Branger, and N. Lambert-Zechovsky. 2002. Identification of PSE and OXA β-lactamase genes in Pseudomonas aeruginosa using PCR-restriction fragment length polymorphism. J. Antimicrob. Chemother. 50:11-18. [DOI] [PubMed] [Google Scholar]
- 5.Bert, F., et al. 2010. Genetic diversity and virulence profiles of Escherichia coli isolates causing spontaneous bacterial peritonitis and bacteremia in patients with cirrhosis. J. Clin. Microbiol. 48:2709-2714. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Bradford, P. A., et al. 2004. Emergence of carbapenem-resistant Klebsiella species possessing the class A carbapenem-hydrolyzing KPC-2 and inhibitor-resistant TEM-30 β-lactamases in New York City. Clin. Infect. Dis. 39:55-60. [DOI] [PubMed] [Google Scholar]
- 7.Bukh, A. S., et al. 2009. Escherichia coli phylogenetic groups are associated with site of infection and level of antibiotic resistance in community-acquired bacteraemia: a 10 year population-based study in Denmark. J. Antimicrob. Chemother. 64:163-168. [DOI] [PubMed] [Google Scholar]
- 8.Canton, R., M. I. Morosini, O. M. de la Maza, and E. G. de la Pedrosa. 2008. IRT and CMT β-lactamases and inhibitor resistance. Clin. Microbiol. Infect. 14(Suppl. 1):53-62. [DOI] [PubMed] [Google Scholar]
- 9.Clermont, O., S. Bonacorsi, and E. Bingen. 2000. Rapid and simple determination of the Escherichia coli phylogenetic group. Appl. Environ. Microbiol. 66:4555-4558. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Clermont, O., et al. 2009. Rapid detection of the O25b-ST131 clone of Escherichia coli encompassing the CTX-M-15-producing strains. J. Antimicrob. Chemother. 64:274-277. [DOI] [PubMed] [Google Scholar]
- 11.Clinical and Laboratory Standards Institute. 2008. Performance standards for antimicrobial susceptibility testing; 18th informational supplement M100-S18. Clinical and Laboratory Standards Institute, Wayne, PA.
- 12.Cooke, N. M., S. G. Smith, M. Kelleher, and T. R. Rogers. 2010. Major differences exist in frequencies of virulence factors and multidrug resistance between community and nosocomial Escherichia coli bloodstream isolates. J. Clin. Microbiol. 48:1099-1104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.de Oliveira Garcia, D., et al. 2008. Multiclonal outbreak of Klebsiella pneumoniae producing extended-spectrum beta-lactamase CTX-M-2 and novel variant CTX-M-59 in a neonatal intensive care unit in Brazil. Antimicrob. Agents Chemother. 52:1790-1793. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Eagye, K. J., J. L. Kuti, C. A. Sutherland, H. Christensen, and D. P. Nicolau. 2009. In vitro activity and pharmacodynamics of commonly used antibiotics against adult systemic isolates of Escherichia coli and Pseudomonas aeruginosa at forty US hospitals. Clin. Ther. 31:2678-2688. [DOI] [PubMed] [Google Scholar]
- 15.Firsov, A. A., et al. 1996. Predictors of effect of ampicillin-sulbactam against TEM-1 β-lactamase-producing Escherichia coli in an in vitro dynamic model: enzyme activity versus MIC. Antimicrob. Agents Chemother. 40:734-738. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Johnson, J. R., et al. 2009. Epidemic clonal groups of Escherichia coli as a cause of antimicrobial-resistant urinary tract infections in Canada, 2002 to 2004. Antimicrob. Agents Chemother. 53:2733-2739. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Kaye, K. S., et al. 2004. Variety of β-lactamases produced by amoxicillin-clavulanate-resistant Escherichia coli isolated in the northeastern United States. Antimicrob. Agents Chemother. 48:1520-1525. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kaye, K. S., A. D. Harris, H. Gold, and Y. Carmeli. 2000. Risk factors for recovery of ampicillin-sulbactam-resistant Escherichia coli in hospitalized patients. Antimicrob. Agents Chemother. 44:1004-1009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Knaus, W. A., E. A. Draper, D. P. Wagner, and J. E. Zimmerman. 1985. APACHE II: a severity of disease classification system. Crit. Care Med. 13:818-829. [PubMed] [Google Scholar]
- 20.Kruger, T., et al. 2004. Infections with nontyphoidal Salmonella species producing TEM-63 or a novel TEM enzyme, TEM-131, in South Africa. Antimicrob. Agents Chemother. 48:4263-4270. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Lartigue, M. F., V. Leflon-Guibout, L. Poirel, P. Nordmann, and M. H. Nicolas-Chanoine. 2002. Promoters P3, Pa/Pb, P4, and P5 upstream from blaTEM genes and their relationship to β-lactam resistance. Antimicrob. Agents Chemother. 46:4035-4037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Lee, C., J. Kim, S. G. Shin, and S. Hwang. 2006. Absolute and relative QPCR quantification of plasmid copy number in Escherichia coli. J. Biotechnol. 123:273-280. [DOI] [PubMed] [Google Scholar]
- 23.Leflon-Guibout, V., et al. 2008. Absence of CTX-M enzymes but high prevalence of clones, including clone ST131, among fecal Escherichia coli isolates from healthy subjects living in the area of Paris, France. J. Clin. Microbiol. 46:3900-3905. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Leflon-Guibout, V., V. Speldooren, B. Heym, and M. Nicolas-Chanoine. 2000. Epidemiological survey of amoxicillin-clavulanate resistance and corresponding molecular mechanisms in Escherichia coli isolates in France: new genetic features of blaTEM genes. Antimicrob. Agents Chemother. 44:2709-2714. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Livak, K. J., and T. D. Schmittgen. 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2−ΔΔCT method. Methods 25:402-408. [DOI] [PubMed] [Google Scholar]
- 26.Martin, O., et al. 2010. Population analysis and epidemiological features of inhibitor-resistant-TEM-β-lactamase-producing Escherichia coli isolates from both community and hospital settings in Madrid, Spain. J. Clin. Microbiol. 48:2368-2372. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Meagher, A. K., A. Forrest, C. R. Rayner, M. C. Birmingham, and J. J. Schentag. 2003. Population pharmacokinetics of linezolid in patients treated in a compassionate-use program. Antimicrob. Agents Chemother. 47:548-553. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Papp-Wallace, K. M., et al. 2010. Inhibitor resistance in the KPC-2 β-lactamase, a preeminent property of this class A β-lactamase. Antimicrob. Agents Chemother. 54:890-897. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Perez-Perez, F. J., and N. D. Hanson. 2002. Detection of plasmid-mediated AmpC β-lactamase genes in clinical isolates by using multiplex PCR. J. Clin. Microbiol. 40:2153-2162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Reguera, J. A., F. Baquero, J. C. Perez-Diaz, and J. L. Martinez. 1991. Factors determining resistance to β-lactam combined with β-lactamase inhibitors in Escherichia coli. J. Antimicrob. Chemother. 27:569-575. [DOI] [PubMed] [Google Scholar]
- 31.Ribot, E. M., et al. 2006. Standardization of pulsed-field gel electrophoresis protocols for the subtyping of Escherichia coli O157:H7, Salmonella, and Shigella for PulseNet. Foodborne Pathog. Dis. 3:59-67. [DOI] [PubMed] [Google Scholar]
- 32.Sanders, C. C., and W. E. Sanders, Jr. 1992. β-Lactam resistance in gram-negative bacteria: global trends and clinical impact. Clin. Infect. Dis. 15:824-839. [DOI] [PubMed] [Google Scholar]
- 33.Sannes, M. R., M. A. Kuskowski, K. Owens, A. Gajewski, and J. R. Johnson. 2004. Virulence factor profiles and phylogenetic background of Escherichia coli isolates from veterans with bacteremia and uninfected control subjects. J. Infect. Dis. 190:2121-2128. [DOI] [PubMed] [Google Scholar]
- 34.Sidjabat, H. E., et al. 2009. Clinical features and molecular epidemiology of CMY-type β-lactamase-producing Escherichia coli. Clin. Infect. Dis. 48:739-744. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Sirot, D., et al. 1994. Clinical isolates of Escherichia coli producing multiple TEM mutants resistant to β-lactamase inhibitors. J. Antimicrob. Chemother. 33:1117-1126. [DOI] [PubMed] [Google Scholar]
- 36.Stapleton, P., et al. 1995. Incidence and mechanisms of resistance to the combination of amoxicillin and clavulanic acid in Escherichia coli. Antimicrob. Agents Chemother. 39:2478-2483. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Tenover, F. C., et al. 1995. Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing. J. Clin. Microbiol. 33:2233-2239. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Wu, P. J., K. Shannon, and I. Phillips. 1994. Effect of hyperproduction of TEM-1 β-lactamase on in vitro susceptibility of Escherichia coli to β-lactam antibiotics. Antimicrob. Agents Chemother. 38:494-498. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Zhou, X. Y., F. Bordon, D. Sirot, M. D. Kitzis, and L. Gutmann. 1994. Emergence of clinical isolates of Escherichia coli producing TEM-1 derivatives or an OXA-1 β-lactamase conferring resistance to β-lactamase inhibitors. Antimicrob. Agents Chemother. 38:1085-1089. [DOI] [PMC free article] [PubMed] [Google Scholar]