Abstract
Objective
Third-generation cephalosporins (3GCs) may be susceptible in vitro to Enterobacter spp. and Klebsiella aerogenes. However, treatment with mainly fourth-generation cephalosporins or carbapenems is currently recommended. Diversification of antimicrobial agents in therapy is required to avoid the selection pressure of resistant organisms by broad-spectrum antimicrobial agents. This study investigated the clinical efficacy of 3GC therapy for Enterobacter spp. and Klebsiella aerogenes bacteremia in a multicenter, retrospective, observational study.
Methods
Patients with Enterobacter spp. or Klebsiella aerogenes detected in blood cultures and treated with a susceptible antimicrobial agent were included in the study. Propensity score matching was performed to align patient background bases, and clinical outcomes between the 3GC and non-3GC groups were compared. Treatment success was defined as having no need for treatment escalation or the addition of other antimicrobial agents, no recurrence, or no death within 30 days.
Results
The study included 188 cases, of which 57 and 131 were included in the 3GC and non-3GC treatment groups, respectively; 53 patients in each group were matched by propensity score matching. There were no significant differences between groups in rates of switching to a susceptible antimicrobial or adding another agent, relapse within 30 days, or death within 30 days. In the 3GC group, source control was associated with favorable clinical outcomes.
Conclusion
Definitive 3GC therapy for susceptible Enterobacter spp. and Klebsiella aerogenes bacteremia is as clinically effective and valuable a targeted therapy as non-3GC therapy and can be implemented under conditions in which infection source control measures are in place.
Keywords: Enterobacter, Klebsiella, bacteremia, cephalosporins, bacterial drug resistance, beta-lactamases
Introduction
Enterobacter spp. and Klebsiella aerogenes (formerly Enterobacter aerogenes) are among the Enterobacteriaceae known to cause infections in clinical practice (1). Third-generation cephalosporins (3GCs) may be susceptible in vitro to Enterobacter spp. and K. aerogenes, but contact with β-lactam drugs promotes AmpC-type β-lactamase (AmpC) overproduction and resistance induction. 3GCs are thus not usually recommended for the treatment of invasive infections because of this potential (2-4). In contrast, fourth-generation cephalosporins (4GCs) (5) and Carbapenems (6) are stable against AmpC produced by Enterobacter spp. and K. aerogenes due to their advantageous antimicrobial structure, and thus are unlikely to be a factor in treatment failure due to AmpC enzyme induction (7,8).
However, in actual clinical practice, based on in vitro results, we have encountered cases in which 3GCs were administered as definitive treatment for infections caused by Enterobacter spp. and K. aerogenes, leading to the completion of treatment. Previous studies have also shown no significant difference in clinical outcomes of bacteremia when 3GCs were used for definitive therapy compared to non-3GC therapy (i.e. extended-spectrum penicillin antibiotics, fluoroquinolones, 4CGs, and carbapenems) (9). In addition, there is evidence that the administration of broad-spectrum antimicrobial agents may contribute to the establishment of resistance. For example, it has been reported that the increased use of imipenem is associated with the development of resistant Pseudomonas aeruginosa (10). Therefore, broad-spectrum antimicrobials, such as 4GCs and carbapenems, should not be administered more than necessary for Enterobacter spp. and K. aerogenes infections susceptible to 3GCs in vitro, and consideration should be given to implementing a switch to a more narrow-spectrum antimicrobial. In other words, it is necessary to consider a variety of antimicrobial therapies, and diversification of antimicrobial therapy is desirable.
There is currently no consensus on how treatment with 3GCs affects Enterobacter spp. and K. aerogenes bacteremia in clinical practice, and this has not yet been tested in Japan. 3GC use may reduce the need for broad-spectrum antimicrobial agents and potentially avoid the risk of resistant strains due to selection pressure. Furthermore, no studies have evaluated the impact of source control on the clinical course of treatment of Enterobacter spp. and K. aerogenes bacteremia.
We therefore compared the impact of 3GC and non-3GC treatment on the clinical course of patients with Enterobacter spp. and K. aerogenes bacteremia and retrospectively examined the impact of source control.
Materials and Methods
Research design
The present study was a retrospective, observational study conducted at two institutions in Japan. The first institution is the Tazuke Kofukai Medical Research Institute at Kitano Hospital, which functions as a general hospital, secondary medical facility, and medical research institute. The second institution is Osaka Metropolitan University Hospital, which functions as a general hospital but has areas of specialization that contribute to tertiary medical care. Kitano Hospital has 685 beds, and Osaka Metropolitan University Hospital has 965 beds.
We screened for cases of Enterobacter spp. or K. aerogenes detected in patients' blood cultures between January 2010 and December 2020. Bacteremia was defined as the detection of Enterobacter spp. or K. aerogenes in at least one set of culture samples; bacteremia caused by multiple bacteria was not excluded. In addition, based on the findings of the initially screened cases, patients treated with 3GCs, 4GCs, carbapenems, fluoroquinolones, aminoglycosides, and broad-spectrum penicillin/β-lactamase inhibitors were included in the study. Antimicrobial therapy is empirical before the drug susceptibility results are known and definitive once full knowledge of susceptibility results is attained. We targeted cases involving infections susceptible to treatment as decided by drug susceptibility testing (i.e. patients undergoing definitive treatment). Antimicrobial agents administered as definitive treatment were classified as 3GCs or non-3GCs.
The exclusion criteria were as follows: infections determined to be intermediately susceptible or resistant, incomplete antimicrobial treatment data, incomplete clinical data (e.g. blood test results and vital signs), cases when therapeutic drug choices were influenced by an infectious disease specialist, patients younger than 20 years old, and patients who did not undergo antimicrobial therapy even after the detection of bacteremia.
This study was conducted in compliance with the Ethical Guidelines for Medical Research for Humans and with the approval of the Medical Ethics Committee at Kitano Hospital and Osaka Metropolitan University Hospital (approval numbers: P210200700 and 2021-140, respectively). In addition, in conducting this study, a notice was posted on the hospital or university website stating that eligible patients could withdraw from the study at any time.
Extracted data
Extracted patient-characteristics data included the age, sex (men), presence of disseminated intravascular coagulation (DIC) syndrome, and Charlson Comorbidity Index (CCI) (11), which were used for mortality predictions based on comorbidities. Data on the use of immunosuppressants, antineoplastic agents, and steroids were also collected. The estimated glomerular filtration rate (eGFR), white blood cell (WBC) count, and C-reactive protein (CRP) levels were obtained from blood sampling data. The blood collection data were obtained within 24 h before and after the blood culture was taken. We also collected the Pitt Bacteremia Score (PBS) (12,13), which is a predictor of mortality risk in bloodstream infections.
The number of days of antimicrobial therapy required to treat bacteremia was investigated. In addition, we investigated the route of entry for suspected bacteria and the implementation of source control. Suspected sites of infection were estimated from blood data, imaging findings, culture results, medical history, and the judgment of the attending physician who was consulted during the data acquisition process. The measures taken during the implementation of source management for the causes of bacteremia were obtained from the clinician's documentation in the medical records. The source management approaches considered in this study were as follows: endoscopic retrograde cholangiopancreatography performed for cholangitis, insertion of a ureteral catheter for urinary tract infection, catheter removal for catheter-related bloodstream infection, and surgery for gastrointestinal perforation.
Finally, we evaluated the following clinical outcomes: a favorable clinical outcome was defined as a case in which there was no treatment escalation or need for additional antibiotics to which Enterobacter spp. or K. aerogenes were susceptible during the treatment period, no recurrence of bacteremia within 30 days, and a patient survival beyond 30 days (14,15). All other cases were considered to have unfavorable clinical outcomes.
Research goals
The primary endpoint was the evaluation of the clinical efficacy of 3GC treatment for Enterobacter spp. or K. aerogenes bacteremia. How the implementation of source control would affect the treatment of bacteremia was also determined.
Identification of isolates and judgment of the drug susceptibility test
Both facilities used MicroScan WalkAway (Beckman Coulter, Brea, USA) to identify bacteria and evaluate the minimum inhibitory concentration (MIC); for microbial identification, Kitano Hospital and Osaka Metropolitan University Hospital introduced the MALDI biotyper (Bruker Japan, Tokyo, Japan) in 2018 and 2019, respectively. The criteria for drug susceptibility testing were based on the Clinical Laboratory Standards Institute (CLSI) guidelines. At Kitano Hospital, treatment decisions were made using M100-S18 (16) between January 2010 and May 2013, M100-S22 (17) between June 2013 and January 2018, and M100-S26 (18) from February 2018 onwards. At Osaka Metropolitan University Hospital, treatment decisions were made using M100-S19 (19) between January 2010 and July 2013 and M100-S21 (20) from August 2013 onwards.
Statistical analyses
Categorical data are counted as the number of cases and reported as percentages of the total. Categorical data included the following: patient age (≥65 years old), sex (men), presence of DIC syndrome, use of antineoplastic agents or immunosuppressants, presumed means of bacterial invasion, CRP (≥15), PBS (≥4 points), whether or not source control was implemented, identification of isolated bacteria, use of antimicrobial agents, and clinical outcomes (favorable or unfavorable). Fisher's exact tests were used for statistical analyses. Patients with PBS ≥4 points (21) at the initial evaluation were defined as having a high mortality risk, and those with a CRP level ≥15 mg/dL (22) were defined as having a high disease severity.
Continuous variables are described as the median (interquartile range) and were analyzed using Mann-Whitney U tests. Continuous variables included the CCI, eGFR, WBC count, and duration of antimicrobial administration. Propensity score matching was performed at baseline on patients' backgrounds to reduce the effects of selection bias and confounding factors. A logistic regression analysis with 1:1 patient matching used a caliper of 0.2. Statistical results were considered significantly different when p<0.05.
All statistical analyses were performed using EZR (23) (Jichi Medical University, Saitama Medical Center, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.5.2). More precisely, it is a modified version of R commander (version 2.5-1) designed to add statistical functions frequently used in biostatistics.
Results
Flowchart of case selection
The case selection process is illustrated in Figure. We screened 307 cases of bloodstream infections caused by Enterobacter spp. or K. aerogenes over an 11-year period (i.e., January 2010 to December 2020); 119 cases were excluded because they satisfied the defined exclusion criteria. In 59 of the excluded cases, patients were treated with antimicrobials for which Enterobacter spp. or K. aerogenes were determined to have intermediate susceptibility or resistance by drug susceptibility testing. In 45 of the excluded cases, clinical data were incomplete; 7 included treatment approaches escalated by an infectious disease specialist, 5 involved patients younger than 20 years old, and 3 involved no treatment with antimicrobials. Ultimately, 188 cases were included in the study; the 3GC treatment group comprised 57 cases, and the non-3GC treatment group comprised 131 cases. Patients in the non-3GC treatment group were treated with 4GCs, carbapenems, fluoroquinolones, aminoglycosides, and piperacillin/tazobactam. Levofloxacin was administered intravenously or orally.
Figure.
Flowchart for case selection. The target period was between January 2010 and December 2020; cases in which Enterobacter spp. or K. aerogenes were detected in blood cultures were screened. In total, 119 cases were excluded from among the 307 screened, and 188 were included in the analysis. Antimicrobials administered as definitive therapy were classified as third-generation (n=57) and non-third-generation cephalosporins (n=131).
The comparison of patient data between the 3GC and non-3GC treatment groups before and after propensity score matching
Before propensity score matching, differences occurred in the source of infection, with more non-3GC cases having an origin in the lower gastrointestinal tract or intra-abdominal (5.3% vs. 32.8%, p<0.001) and more 3GC cases having an origin in the biliary system (40.4% vs. 23.7%, p=0.023) and urinary tract (26.3% vs. 11.5%, p=0.016). In contrast, there were no significant differences between the two groups in other patient characteristics before propensity score matching (Table 1).
Table 1.
Baseline Characteristics of Patients and the Clinical Background of the Third-generation and Non-third-generation Cephalosporin Treatment Groups.
| Variable | Third-generation cephalosporins (n=57) |
Non-third-generation cephalosporins (n=131) |
p value | ||
|---|---|---|---|---|---|
| Age, ≥65 years; n (%) | 44 | (77.2) | 91 | (69.5) | 0.297a |
| Sex, Male (%) | 38 | (66.7) | 84 | (61.8) | 0.868a |
| Charlson Comorbidity Index; median (IQR) | 3.0 | (2.0-4.0) | 3.0 | (2.0-5.0) | 0.440b |
| Disseminated intravascular coagulation; n (%) | 7 | (12.3) | 23 | (17.6) | 0.516a |
| Anticancer agents or immunosuppressants; n (%) | 24 | (42.1) | 57 | (43.5) | 0.874a |
| Source of Infection | |||||
| Lower gastrointestinal tract or intra-abdominal; n (%) | 3 | (5.3) | 43 | (32.8) | <0.001a |
| Biliary tract; n (%) | 23 | (40.4) | 31 | (23.7) | 0.023a |
| Urinary tract; n (%) | 15 | (26.3) | 15 | (11.5) | 0.016a |
| Respiratory system; n (%) | 3 | (5.3) | 13 | (9.9) | 0.399a |
| Intravenous catheter; n (%) | 2 | (3.5) | 7 | (5.3) | 0.725a |
| Unclear or others; n (%) | 11 | (19.3) | 22 | (16.8) | 0.681a |
| Clinical characteristics | |||||
| eGFR, mL/min/1.73 m2; median (IQR) | 55.0 | (33.7-70.0) | 58.5 | (36.3-77.4) | 0.681b |
| WBC, /μL; median (IQR) | 7,600 | (4,400-9,800) | 8,900 | (4,700-13,200) | 0.221b |
| CRP (≥15), mg/dL; n (%) | 7 | (12.3) | 26 | (19.8) | 0.297a |
| Pitt Bacteremia Score (≥4 points); n (%) | 16 | (28.1) | 44 | (33.6) | 0.500a |
| Implementation of source control; n (%) | 24 | (45.3) | 52 | (39.7) | 0.872a |
| Duration of antimicrobial administration, days; median (IQR) | 13.0 | (9.0-16.0) | 13.0 | (9.0-17.0) | 0.556b |
| Microorganisms | |||||
| Enterobacter cloacae or Enterobacter cloacae complex; n (%) | 38 | (66.7) | 92 | (70.2) | 0.731a |
| Klebsiella aerogenes; n (%) | 15 | (26.3) | 33 | (25.2) | 0.858a |
| Other Enterobacter spp.; n (%) | 4 | (7.0) | 6 | (4.6) | 0.494a |
aFisher’s exact test, bMann-Whitney U test. IQR: interquartile range, eGFR: estimated glomerular filtration rate, WBC: white blood cell, CRP: C-reactive protein
A total of 53 patients were matched by propensity score matching in the 3GC and non-3GC treatment groups. The propensity score matching results showed no significant differences in patient background, source of infection, blood collection, severity of illness, implementation of source control, duration of antimicrobial therapy, or bacteria detected between the two groups (Table 2).
Table 2.
Baseline Characteristics of Patients and the Clinical Background of the Third-generation and Non-third-generation Cephalosporin Treatment Groups after Propensity Score Matching.
| Variable | Third-generation cephalosporins (n=53) |
Non-third-generation cephalosporins (n=53) |
p value | ||
|---|---|---|---|---|---|
| Age, ≥65 years; n (%) | 42 | (79.2) | 37 | (69.8) | 0.373a |
| Sex, Male (%) | 36 | (67.9) | 38 | (71.7) | 0.833a |
| Charlson Comorbidity Index; median (IQR) | 3.0 | (2.0-4.0) | 3.0 | (2.0-5.0) | 0.369b |
| Disseminated intravascular coagulation; n (%) | 6 | (11.3) | 7 | (13.2) | 1.000a |
| Anticancer agents or immunosuppressants; n (%) | 24 | (45.3) | 21 | (39.6) | 0.695a |
| Source of Infection | |||||
| Lower gastrointestinal tract or intra-abdominal; n (%) | 3 | (5.7) | 4 | (7.5) | 1.000a |
| Biliary tract; n (%) | 21 | (39.6) | 22 | (41.5) | 1.000a |
| Urinary tract; n (%) | 13 | (24.5) | 14 | (26.4) | 1.000a |
| Respiratory system; n (%) | 3 | (5.7) | 0 | (0) | 0.243a |
| Intravenous catheter; n (%) | 2 | (3.8) | 0 | (0) | 0.495a |
| Unclear or others; n (%) | 11 | (20.8) | 13 | (24.5) | 0.817a |
| Clinical characteristics | |||||
| eGFR, mL/min/1.73 m2; median (IQR) | 55.0 | (36.0-70.0) | 59.0 | (42.3-73.7) | 0.774b |
| WBC, /μL; median (IQR) | 7,600 | (4,300-9,800) | 8,500 | (4,700-11,500) | 0.822b |
| CRP (≥15), mg/dL; n (%) | 5 | (9.4) | 6 | (11.3) | 1.000a |
| Pitt Bacteremia Score (≥4 points); n (%) | 16 | (30.2) | 19 | (35.8) | 0.680a |
| Implementation of source control; n (%) | 21 | (39.6) | 22 | (41.5) | 1.000a |
| Duration of antimicrobial administration, days; median (IQR) | 13.0 | (9.0-16.0) | 13.0 | (9.0-15.0) | 0.869b |
| Microorganisms | |||||
| Enter obacter cloacae or Enterobacter cloacae complex; n (%) | 36 | (67.9) | 34 | (64.2) | 0.838a |
| Klebsiella aero genes; n (%) | 14 | (26.4) | 16 | (30.2) | 0.830a |
| Other Enterobacter spp.; n (%) | 3 | (5.7) | 3 | (5.7) | 1.000a |
aFisher’s exact test, bMann-Whitney U test. IQR: interquartile range, eGFR: estimated glomerular filtration rate, WBC: white blood cell, CRP: C-reactive protein
Antimicrobial agents in each treatment group
After propensity score matching, the 3GC group included 23, 5, 5, and 20 patients administered with ceftriaxone, cefotaxime, ceftazidime, and sulbactam/cefoperazone, respectively. The most common antimicrobials administered as empirical therapy in the 3GC treatment group were 3GCs (62.3%), followed by carbapenems (26.4%) and piperacillin/tazobactam (7.5%). Carbapenems (41.5%) were the most common antimicrobial administered empirical therapy in the non-3GC group, followed by piperacillin/tazobactam (34.0%) and 4GCs (11.3%). Antimicrobials administered as definitive therapy in the non-3GC treatment group were most commonly carbapenems (35.8%), followed by piperacillin/tazobactam (26.4%) and fluoroquinolone (24.5%) (Table 3).
Table 3.
Empirical and Definitive Therapies in the Third-generation and Non-third-generation Cephalosporin Treatment Groups
| antimicrobial | Third-generation cephalosporins (n=53) | Non-third-generation cephalosporins (n=53) |
p value* | |||||
|---|---|---|---|---|---|---|---|---|
| Empirical therapy | Empirical therapy | Definitive therapy | ||||||
| Third-generation cephalosporins | 33 | (62.3) | 5 | (9.4) | - | - | <0.001 | |
| Ceftriaxone or cefotaxime | 14 | (26.4) | 3 | (5.7) | - | - | ||
| Ceftazidime or cefoperazone/sulbactam | 19 | (35.8) | 2 | (3.8) | - | - | ||
| Fourth-generation cephalosporins | 1 | (1.9) | 6 | (11.3) | 7 | (13.2) | 0.113 | |
| Carbapenems | 14 | (26.4) | 22 | (41.5) | 19 | (35.8) | 0.151 | |
| Piperacillin/tazobactam | 4 | (7.5) | 18 | (34.0) | 14 | (26.4) | 0.001 | |
| Fluoroquinolones | 1 | (1.9) | 2 | (3.8) | 13 | (24.5) | 1.000 | |
Empirical therapy is antimicrobial therapy until susceptibility results are known, while definitive therapy is antimicrobial therapy after susceptibility results are known. Third-generation cephalosporins such as cefotaxime, ceftriaxone, ceftazidime, and cefoperazone/sulbactam were administered as intravenous formulations. Fourth-generation cephalosporins were administered with cefozopran and cefepime. Carbapenems were administered with meropenem and doripenem. Fluoroquinolones were administered with ciprofloxacin and levofloxacin. *Comparison of antimicrobials administered empirically in the third-generation cephalosporins treatment group and non-third-generation cephalosporins treatment group were evaluated using Fisher’s exact tests.
The comparison of clinical outcomes in the treatment groups
There were no statistically significant differences in treatment escalation or addition of other drugs during the treatment period, recurrences within 30 days, or death within 30 days between the 3GC and non-3GC treatment groups. In addition, overall clinical outcomes were similar between the two groups (Table 4). The antimicrobial agents associated with a poor clinical course in the 3GC group were ceftriaxone, cefotaxime, ceftazidime, and cefoperazone/sulbactam, administered in five, two, three, and one patient, respectively.
Table 4.
Unfavorable Clinical Outcomes between Patients in the Third-generation Cephalosporins and Non-third-generation Cephalosporins Treatment Groups
| Outcome | Third-generation cephalosporins (n=53) |
Non-third-generation cephalosporins (n=53) |
Odds ratio [95% CI] | p value | ||
|---|---|---|---|---|---|---|
| Unfavourable clinical outcome; n (%) | 11 | (20.8) | 12 | (22.6) | 0.896 [0.318-2.500] | 1.000 |
| Treatment escalation or combined use with other drugs; n (%) | 5 | (9.4) | 3 | (5.7) | 1.727 [0.316-11.73] | 0.716 |
| Number of recurrences within 30 days; n (%) | 4 | (7.5) | 3 | (5.7) | 1.357 [0.217-9.742] | 1.000 |
| Number of deaths within 30 days; n (%) | 2 | (3.8) | 6 | (11.3) | 0.310 [0.029-1.846] | 0.270 |
Fisher’s exact test. CI: confidence interval
The assessment of factors affecting clinical outcomes
Table 5 shows the factors associated with clinical outcomes in the 3GC and non-3GC treatment groups. In the 3GC treatment group, the implementation of source control had an impact on the favorable clinical outcomes, with a statistically significant difference (p=0.002). In contrast, patients with a low eGFR and high WBC count and CRP level (≥15 mg/dL) tended to have poorer clinical outcomes than others. In the non-3GC treatment group, there was a trend toward a higher incidence of DIC (p=0.039) and higher CRP (≥15) in the patients with unfavorable clinical outcomes than in those with favorable outcomes. The implementation of source control in the non-3GC treatment group may have contributed to an improved clinical course, but the difference was not statistically significant.
Table 5.
Factors Associated with Unfavorable Clinical Outcomes in the Third-generation Cephalosporins and Non-third-generation Cephalosporins Treatment Groups.
| Variable | Third-generation cephalosporin (n=53) |
p value | Non-third-generation cephalosporin (n=53) |
p value | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Unfavorable clinical outcome (n=11) |
Favorable clinical outcome (n=42) |
Unfavorable clinical outcome (n=12) |
Favorable clinical outcome (n=41) |
|||||||
| Age, ≥65 years; n (%) | 10 | (90.9) | 32 | (76.2) | 0.423a | 9 | (75.0) | 28 | (68.3) | 0.737a |
| Sex, Male; n (%) | 9 | (81.8) | 27 | (64.3) | 0.469a | 10 | (83.3) | 28 | (68.3) | 0.472a |
| Charlson Comorbidity Index; median (IQR) | 3.0 | (1.0-5.0) | 3.0 | (2.0-4.0) | 0.920b | 3.0 | (2.8-7.3) | 3.0 | (2.0-5.0) | 0.564b |
| Disseminated intravascular coagulation; n (%) | 0 | (0) | 6 | (14.3) | 0.324a | 4 | (33.3) | 3 | (7.3) | 0.039a |
| Anticancer agents or immunosuppressants; n (%) | 2 | (18.2) | 22 | (52.4) | 0.086a | 5 | (41.7) | 16 | (39.0) | 1.000a |
| Clinical characteristics | ||||||||||
| eGFR, mL/min/1.73 m2; median (IQR) | 37.0 | (27.0-53.5) | 57.3 | (44.7-78.3) | 0.089b | 48.1 | (36.5-67.7) | 59.0 | (42.6-73.7) | 0.595b |
| WBC/μL; median (IQR) | 9,200 | (5,900-13,650) | 7,200 | (3,950-9,680) | 0.174b | 9,850 | (4,180-12,450) | 8,100 | (4,700-10,500) | 0.632b |
| CRP (≥15), mg/dL; n (%) | 2 | (18.2) | 3 | (7.1) | 0.275b | 3 | (25.0) | 3 | (7.3) | 0.121a |
| Pitt Bacteremia Score (≥4 points); n (%) | 4 | (36.4) | 12 | (28.6) | 0.716a | 6 | (50.0) | 13 | (31.7) | 0.311a |
| Implementation of source control; n (%) | 0 | (0) | 21 | (50.0) | 0.002a | 3 | (25.0) | 19 | (46.3) | 0.318a |
aFisher’s exact test, bMann-Whitney U test. IQR: interquartile range, eGFR: estimated glomerular filtration rate, WBC: white blood cell, CRP: C-reactive protein
Discussion
As antimicrobial therapy against AmpC-producing bacteria such as Enterobacter spp. and K. aerogenes, treatment with 3GCs was generally not recommended due to concerns about resistance induction (24). However, it has also been reported that treatment with 3GCs does not increase the risk of poor clinical outcomes compared with treatment with antimicrobial agents other than 3GCs (9). There have also been cases of successful treatment of Enterobacter spp. and K. aerogenes bacteremia using 3GCs that were determined to be susceptible in drug susceptibility testing, and the evaluation of the therapeutic efficacy of 3GCs is an important factor in the selection of therapeutic agents. In Enterobacter spp. and K. aerogenes bacteremia, if treatment with 3GCs has clinically similar efficacy to treatment with non-3GCs, it may be possible to reduce the frequency of using broad-spectrum antimicrobial agents, such as carbapenems and 4GCs, and resolve bacteremia using different antimicrobial options, including 3GCs, thus leading to the diversification of antimicrobial therapy. The present findings suggest that definitive treatment of bacteremia caused by Enterobacter spp. and K. aerogenes with 3GCs may have non-inferior clinical efficacy to definitive treatment with agents other than 3GCs, as in previous studies (9).
The total mortality of bacteremia caused by Enterobacter spp. within 28 days of infection has been reported to be 21% (25), indicating that it is an extremely serious infections that can affect the prognosis if not treated appropriately. Furthermore, recent reports for Enterobacter spp., Serratia spp., and Citrobacter spp. have reported 30-day mortality rates ranging from 6.2% to 8.5% (9). This improvement in mortality is presumably due to differences in causative organism species and patient background, but it may also be due to improvements in critical care because of the different periods covered. Furthermore, the treatment results in the current study were similar to those of the most recently previous studies, with 30-day mortality rates of 3.8% in the definitive 3GC group and 11.3% in the definitive non-3GC group, indicating that definitive 3GC treatment for Enterobacter spp. and K. aerogenes bacteremia does not increase mortality compared to definitive non-3GC treatment. It is noteworthy that definitive 3GC treatment for Enterobacter spp. and K. aerogenes contributes to treatment completion without increased mortality or prolonging the duration of antimicrobial therapy compared to definitive non-3GC treatment.
Although some strains were resistant to the administered antimicrobials from blood cultures, there were no deaths or relapses within 30 days of the onset of bacteremia, and there was no need for escalation or additional antimicrobial therapy. In other words, in this study, the involvement of strains other than Enterobacter detected in blood cultures did not influence a poor prognosis. Among the 3GCs, the rate of treatment failure was lowest for cefoperazone/sulbactam, but this may be due to the high rate of administration in biliary tract infections and the high rate of implementation of source control.
Focusing on the source of infection and patient characteristics prior to propensity score matching, the non-3GCs group had a higher percentage of patients reporting either the lower gastrointestinal tract or abdominal cavity as the source of infection with a statistically significant difference. This was thought to be due to the need for treatment with broad-spectrum antimicrobial agents, including coverage of anaerobic bacteria, since the choice of antimicrobial agents in cases of gastrointestinal tract perforation was made with an awareness of the involvement of Enterobacteriaceae bacteria. The most common source of infection presumed to have led to bacteremia was related to the gastrointestinal tract. After propensity score matching, Gram-negative bacilli were detected in a total of 10 patients in the 3GC group in culture specimens from the gastrointestinal tract, such as bile and ascites fluid (Enterobacter spp. in 6 samples, K. aerogenes in 2 samples, other Klebsiella spp. in 2 samples, Citrobacter spp. in 2 samples, Aeromonas spp. in 1 sample and Pseudomonas aeruginosa in 1 sample). Similarly, in the non-3GC group, Gram-negative bacilli were detected in a total of 13 patients (Enterobacter spp. in 8 samples, K. aerogenes in 6 samples, other Klebsiella spp. in 3 samples and Pseudomonas aeruginosa in 1 sample). In one case in the non-3GC group, the clinical course was not satisfactory. Enterobacter spp. were detected in bile culture and were sensitive to the administered antimicrobial agents, but recurrence of bacteremia was observed. In cases of infection with obstruction, such as cholangitis and urinary tract infections, and intra-abdominal infections with perforation of the gastrointestinal tract, in addition to conventional antimicrobial therapy, source control has been reported to affect treatment outcomes (26-28). According to our findings, when source control was performed in the 3GC group, the clinical course was good regardless of the severity of bacteremia. However, when source control was not performed in the 3GC group, the clinical course was worse than in the source control group, regardless of the severity of the bacteria. The implementation of source control also seemed to help improve the clinical course in the non-3GC group, although no statistically significant difference was observed. As in previous reports, these results suggest that source control during the period of treatment for bacteremia may lead to clinical success. For example, in infections with inadequate drainage, uncontrolled infection sites, or prolonged treatment, switching to 3GCs was not recommended because of the risk of developing AmpC resistance during treatment. These results suggest that 3GCs can be safely administered for bacteremia under conditions in which the site of infection is adequately controlled, with the number of bacteria expected to be under control. In addition, given the present findings, the systemic status of the patients since the clinical course may also be affected by the decline in the renal function (i.e. a low eGFR) and disease severity (i.e. DIC status, high CRP level, and high PBS level).
Several limitations associated with the present study warrant mention. First, the measurement of AmpC-type β-lactamase production, which also influences the drug susceptibility testing results and the outcome of antimicrobial therapy, has not been generalized. In the present results, definitive 3GC treatment for Enterobacter spp. and K. aerogenes bacteremia was not inferior to definitive non-3GC treatment, but it is unclear how 3GC treatment affects AmpC-type β-lactamase expression. Second, there was a lack of bacteriological evaluation data in the records. Follow-up blood culture was not performed in all cases to assess the effect of antimicrobial therapy on bloodstream infections. Therefore, confirming whether or not antimicrobial administration had eliminated the isolated bacteria from the blood was impossible. Third, the study design was retrospective and observational, and the selection of antimicrobial agents and determination of treatment efficacy involved subjective judgment by monitoring clinicians. In addition, the timing of blood collection differed by case due to the study design.
Reliable treatment with 3GCs against Enterobacter spp. and K. aerogenes is not inferior to treatment with non-3GCs, but the drug sensitivity of 3GC compares unfavorably with 4GCs and carbapenems (data not shown), so 3GCs are not recommended as empiric therapy. In addition, a more detailed investigation of the factors responsible for resistance to antimicrobial therapy is needed. In this study, we focused on Enterobacter spp. and K. aerogenes, but it is worth evaluating Serratia and Citrobacter spp. in the future, as these bacteria also express AmpC-type β-lactamases and are commonly encountered in clinical settings.
Conclusion
Conclusive treatment with 3GCs for bloodstream infections caused by Enterobacter spp. and K. aerogenes was found to be as clinically effective as definitive treatment with non-3GCs. 3GCs as a treatment option are expected to broaden the range of therapeutic agents and reduce the occurrence of selection pressure by broad-spectrum antimicrobial agents for resistant organisms.
The authors state that they have no Conflict of Interest (COI).
Financial Support
This work was supported by the Research Program on Emerging and Re-emerging Infectious Diseases from the Japan Agency for Medical Research and Development, AMED under Grant (number JP21fk0108505, JP21fk0108094 and JP21fk0108133); and the Japan Society for the Promotion of Science, JSPS KAKENHI under Grant (number 18K16185, 19K08958).
References
- 1. Gupta N, Aparna, Choudhary U, et al. Enterobacter bacteremia. J Assoc Physicians India 51: 669-672, 2003. [PubMed] [Google Scholar]
- 2. Kaye KS, Cosgrove S, Harris A, et al. Risk factors for emergence of resistance to broad-spectrum cephalosporins among Enterobacter spp. Antimicrob Agents Chemother 45: 2628-2630, 2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Livermore DM, Brown DFJ, Quinn JP, et al. Should third-generation cephalosporins be avoided against AmpC-inducible Enterobacteriaceae? Clin Microbiol Infect 10: 84-85, 2004. [DOI] [PubMed] [Google Scholar]
- 4. Meini S, Tascini C, Cei M, et al. AmpC β-lactamase-producing Enterobacterales: what a clinician should know. Infection 47: 363-375, 2019. [DOI] [PubMed] [Google Scholar]
- 5. Tamma PD, Girdwood SCT, Gopaul R, et al. The use of cefepime for treating AmpC β-lactamase-producing Enterobacteriaceae. Clin Infect Dis 57: 781-788, 2013. [DOI] [PubMed] [Google Scholar]
- 6. Harris PNA, Wei JY, Shen AW, et al. Carbapenems versus alternative antibiotics for the treatment of bloodstream infections caused by Enterobacter, Citrobacter or Serratia species: a systematic review with meta-analysis. J Antimicrob Chemother 71: 296-306, 2016. [DOI] [PubMed] [Google Scholar]
- 7. Hiraoka M, Masuyoshi S, Mitsuhashi S, et al. Cephalosporinase interactions and antimicrobial activity of BMY-28142, ceftazidime and cefotaxime. J Antibiot (Tokyo) 41: 86-93, 1988. [DOI] [PubMed] [Google Scholar]
- 8. Beadle BM, Shoichet BK. Structural basis for imipenem inhibition of class C β-lactamases. Antimicrob Agents Chemother 46: 3978-3980, 2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Derrick C, Bookstaver PB, Lu ZK, et al. Multicenter, observational cohort study evaluating third-generation cephalosporin therapy for bloodstream infections secondary to Enterobacter, Serratia, and Citrobacter species. Antibiotics (Basel) 9: 254, 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Harris AD, Smith D, Johnson JA, Bradham DD, Roghmann MC. Risk factors for imipenem-resistant Pseudomonas aeruginosa among hospitalized patients. Clin Infect Dis 34: 340-345, 2002. [DOI] [PubMed] [Google Scholar]
- 11. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 40: 373-383, 1987. [DOI] [PubMed] [Google Scholar]
- 12. 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 140: 26-32, 2004. [DOI] [PubMed] [Google Scholar]
- 13. Rhee JY, Kwon KT, Ki HK, et al. Scoring systems for prediction of mortality in patients with intensive care unit-acquired sepsis: a comparison of the Pitt bacteremia score and the Acute Physiology and Chronic Health evaluation ii scoring systems. Shock 31: 146-150, 2009. [DOI] [PubMed] [Google Scholar]
- 14. Ng TM, Khong WX, Harris PNA, et al. Empiric piperacillin-tazobactam versus carbapenems in the treatment of bacteraemia due to extended-spectrum beta-lactamase-producing Enterobacteriaceae. PLoS One 11: e0153696, 2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Cheng L, Nelson BC, Mehta M, et al. Piperacillin-tazobactam versus other antibacterial agents for treatment of bloodstream infections due to AmpC β-lactamase-producing Enterobacteriaceae. Antimicrob Agents Chemother 61: 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Clinical and Laboratory Standards Institute. Performance Standards for Antimicrobial Susceptibility Testing: 18th informational supplement. M100-S18. Clinical and Laboratory Standards Institute, Wayne, PA, 2008. [Google Scholar]
- 17. Clinical and Laboratory Standards Institute. Performance Standards for Antimicrobial Susceptibility Testing: 22nd informational supplement. M100-S22. Clinical and Laboratory Standards Institute, Wayne, PA, 2012. [Google Scholar]
- 18. Clinical and Laboratory Standards Institute. Performance Standards for Antimicrobial Susceptibility Testing: 26th informational supplement. M100-S26. Clinical and Laboratory Standards Institute, Wayne, PA, 2016. [Google Scholar]
- 19. Clinical and Laboratory Standards Institute. Performance Standards for Antimicrobial Susceptibility Testing: 19th informational supplement. M100-S19. Clinical and Laboratory Standards Institute, Wayne, PA, 2009. [Google Scholar]
- 20. Clinical and Laboratory Standards Institute. Performance Standards for Antimicrobial Susceptibility Testing: 21st informational supplement. M100-S21. Clinical and Laboratory Standards Institute, Wayne, PA, 2011. [Google Scholar]
- 21. Chow JW, Yu VL. Combination antibiotic therapy versus monotherapy for gram-negative bacteraemia: a commentary. Int J Antimicrob Agents 11: 7-12, 1999. [DOI] [PubMed] [Google Scholar]
- 22. Kohno S, Seki M, Watanabe A; CAP Study Group. Evaluation of an assessment system for the JRS 2005: A-DROP for the management of CAP in adults. Intern Med 50: 1183-1191, 2011. [DOI] [PubMed] [Google Scholar]
- 23. Kanda Y. Investigation of the freely available easy-to-use software ‘EZR’ for medical statistics. Bone Marrow Transplant 48: 452-458, 2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Chow JW, Fine MJ, Shlaes DM, et al. Enterobacter bacteremia: clinical features and emergence of antibiotic resistance during therapy. Ann Intern Med 115: 585-590, 1991. [DOI] [PubMed] [Google Scholar]
- 25. Al-Hasan MN, Lahr BD, Eckel-Passow JE, Baddour LM. Temporal trends in Enterobacter species bloodstream infection: a population-based study from 1998-2007. Clin Microbiol Infect 17: 539-545, 2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Azuhata T, Kinoshita K, Kawano D, et al. Time from admission to initiation of surgery for source control is a critical determinant of survival in patients with gastrointestinal perforation with associated septic shock. Crit Care 18: R87, 2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Solomkin JS, Mazuski JE, Bradley JS, et al. Diagnosis and management of complicated intra-abdominal infection in adults and children: guidelines by the Surgical Infection Society and the Infectious Diseases Society of America. Clin Infect Dis 50: 133-164, 2010. [DOI] [PubMed] [Google Scholar]
- 28. Miura F, Okamoto K, Takada T, et al. Tokyo Guidelines 2018: initial management of acute biliary infection and flowchart for acute cholangitis. J Hepatobiliary Pancreat Sci 25: 31-40, 2018. [DOI] [PubMed] [Google Scholar]