Treatment of Bacteremia Caused by Enterobacter spp.: Should the Potential for AmpC Induction Dictate Therapy? A Retrospective Study

Abstract

Objective:

Carbapenems are considered treatment of choice for bacteremia caused by potential AmpC-producing bacteria, including Enterobacter spp. We aimed to compare mortality following carbapenem vs. alternative antibiotics for the treatment of Enterobacter spp. bacteremia.

Patients and Methods:

We conducted a retrospective study in two centers in Israel. We included hospitalized patients with Enterobacter bacteremia treated with third-generation cephalosporins (3GC), piperacillin/tazobactam, quinolones, or carbapenem monotherapy as the main antibiotic in the first week of treatment, between 2010 and 2017. Cefepime was excluded due to nonavailability during study years. The primary outcome was 30-day all-cause mortality. Univariate and multivariate analyses were conducted, introducing the main antibiotic as an independent variable.

Results:

Two hundred seventy-seven consecutive patients were included in the analyses. Of these, 73 were treated with 3GC, 39 with piperacillin/tazobactam, 104 with quinolones, and 61 with carbapenems. All-cause 30-day mortality was 16% (45 patients). The type of antibiotics was not significantly associated with mortality on univariate or multivariate analyses. With carbapenems as reference, adjusted odds ratios (ORs) for mortality were 0.708, 95% confidence interval (CI) 0.231–2.176 with 3GC; OR 1.172, 95% CI 0.388–3.537 with piperacillin/tazobactam; and OR 0.586, 95% CI 0.229–1.4 with quinolones. The main antibiotic was not associated with repeated growth of Entrobacter spp. in blood cultures or other clinical specimens. Resistance development was observed with 3GC and piperacillin/tazobactam.

Conclusions:

Carbapenem treatment was not advantageous to alternative antibiotics, including 3GC, among patients with Enterobacter spp. bacteremia in an observational study.

Introduction

Antibiotic resistance among Gram-negative bacteria is increasing and poses a challenge for treatment. Several Gram-negative bacteria contain chromosomally encoded, inducible AmpC beta-lactamase genes. These include mainly Enterobacter spp., Serratia marcescens, Citrobacter freundii, Providencia spp., and Morganella morganii. Exposure of AmpC-positive bacteria to beta-lactams may induce hyperexpression of AmpC, conferring resistance that may develop during therapy.^1–4 This resistance has been described to develop following even a single day of beta-lactam administration.⁵ The resultant phenotype includes resistance to penicillins, cephalosporins (except cefepime), and beta-lactam/beta-lactamase inhibitors (BLBLIs).^1–4

Enterobacter cloacae and Enterobacter aerogenes have the highest induction rate among inducible AmpC-producing bacteria, ranging 3–19% in different studies. Third-generation cephalosporins (3GC) and piperacillin/tazobactam are weak inducers of AmpC, but are also susceptible to hydrolysis by the enzyme. In contrast, carbapenems are strong inducers, but are much more stable for hydrolysis.^4,6

A controversy exists regarding optimal therapy for Enterobacter bacteremia. Older studies have demonstrated high rates of resistance and increased mortality following use of 3GC in patients with Enterobacter bacteremia.^7,8 Consequently, 3GC are considered inappropriate treatment and carbapenems are considered treatment of choice for this infection.^6,9,10 Other options, including cefepime, beta-lactams beta-lactamase inhibitors, and quinolones, have been evaluated in retrospective studies with a relatively small number of patients, demonstrating noninferiority to carbapenems.^3,11

In our practice, 3GC and BLBLI are considered by some experts an acceptable definitive treatment for Enterobacter bacteremia. We aimed to explore risk factors for mortality in this infection, considering the impact of antibiotic treatment.

Methods

Data collection and patient inclusion

This was a retrospective cohort study, conducted in two tertiary, university-affiliated medical centers in Israel of 900 and 980 beds (Rabin Medical Center, Petah-Tikva, and Rambam Medical Center, Haifa, respectively). The study was approved in each center by the local ethics committees (approval number: RMC 18-2654).

We included adult patients (≥18 years) with monomicrobial E. cloacae or E. aerogenes bacteremia between January 1, 2010, and December 31, 2017. Identification of patients was performed using the microbiology laboratory computerized databases of each center. Bacteremia was defined by at least one positive blood culture bottle accompanied by signs and symptoms of infection. We included in the analysis all patients treated as their main antibiotic therapy (see definitions) with carbapenems, quinolones, 3GC, or piperacillin/tazobactam. Patients treated with antibiotic combinations (beta-lactam plus quinolone or aminoglycosides) or patients treated with aminoglycoside monotherapy were excluded. Patients were included in the study only once, for the first bacteremia fulfilling inclusion criteria. At the time the study was conducted, cefepime was not available for use in the participating centers.

The exposure variable was the main antibiotic treatment, defined as the antibiotic administered for over 96 hours in the first 7 days from culture collection. The primary outcome was 30-day all-cause mortality. Secondary outcomes included relapsed or persistent bacteremia within 90 days of culture collection, defined as positive repeated blood cultures for the same pathogen after at least 72 hours of antibiotic therapy; recurrence—growth of Enterobacter spp. from follow-up cultures of any source within 90 days of culture collection; and resistance to the main antibiotics in repeated cultures of the index pathogen.

Data on confounders and other risk factors for mortality were collected from patients' electronic charts, including patient demographics, department of hospitalization, Charlson comorbidity index, characteristics of the bacteremia, the Enterobacter sp., and the appropriateness of the empirical antibiotic treatment. Appropriate empirical antibiotic therapy was defined as administration of at least one antibiotic with in-vitro activity against the Enterobacter sp. (including aminoglycosides) within 48 hours from culture collection.

Charlson score and Pitt score were dichotomized to facilitate practical use of the data by clinicians. The cutoffs were chosen according to previous studies, demonstrating the relationship between these cutoffs and early mortality.^12,13 We chose to dichotomize infection source to pneumonia vs. other due to the higher mortality in Enterobacter pneumonia and the scarce data regarding optimal therapy.¹⁴

Laboratory methods

Bacteria were identified using the VITEK 2 system (bioMérieux) or MALDI Biotyper System (Bruker Daltonics Inc.). Antibiotic susceptibility testing was performed by the disk diffusion method, E-test, or VITEK II (bioMérieux) according to updated CLSI criteria for each year, accounting for changes in breakpoints.

Statistical analysis

Categorical variables were compared using the chi-squared or Fisher's exact tests; continuous variables were compared using the Student's t-test or Mann–Whitney U-test, as appropriate. Univariate analysis of risk factors for 30-day all-cause mortality was performed. Variables found to be statistically significant on univariate analysis (p < 0.05) and nonclinically correlated, and the exposure variable, were introduced into multivariate analysis with no selection. Goodness of fit was examined using the Hosmer and Lemeshow Test. Statistical analyses were performed using the SPSS statistics version 25 software.

Results

During the study period, we identified 311 consecutive patients with Enterobacter spp. bacteremia. Eight patients were excluded due to missing data regarding antibiotic therapy, 26 were excluded due to main antibiotic therapies different from those defined for analysis, and 277 patients were included in the analysis. The median age was 64 (interquartile range 52–76) years and 181 were of male gender (65%). Of the 277 patients, 73 were treated with 3GC as their main antibiotics, 39 with piperacillin/tazobactam, 104 with quinolones, and 61 with carbapenems (for antibiotic susceptibilities of isolates in each group, see Supplementary Table S1).

At 30 days, 45 patients died (16%), 7 in the 3GC group (7/73, 9.6%); 10 in the piperacillin/tazobactam group (10/39, 25.6%); 16 in quinolone group (16/104, 15.3%); and 12 in carbapenem group (12/61, 19.7%). Univariate and multivariate analyses of risk factors for 30-day all-cause mortality are presented in Tables 1 and 2. Variables significantly associated with mortality on univariate analysis included older age, higher Charlson comorbidity score, urinary catheter at presentation, higher PITT bacteremia score, higher creatinine, and lower albumin values. Appropriate empirical antibiotics were provided to 217/277 (78.3%) of patents and were not associated with survival. Age, comorbidity index, and low albumin remained significant in multivariate analysis. In both univariate and multivariate analyses, no significant association between type of main antibiotic therapy and mortality was demonstrated. Hosmer–Lemeshow goodness-of-fit test indicated that the model fits well (p = 0.394).

Table 1.

Univariate Analysis for Risk Factors for Mortality

Variable	Dead within 30 days (n = 45)	Alive within 30 days (n = 232)	p
Hospital—Rambam	28 (62.2%)	139 (59.9%)	0.772
Gender—male	28 (62.2)	154 (66.4)	0.591
Age	69 (60–80.5)	64 (49–75)	0.025
Charlson score ≥4	23 (51.1)	69 (29.7)	0.005
Pitt bacteremia score ≥4	14 (31.1)	31 (13.4)	0.003
Source pneumonia/unknown	21 (46.7)	87 (37.5)	0.249
Central line at presentation	18 (40.0)	118 (50.9)	0.182
Urinary catheter at presentation	22 (48.9)	76 (32.8)	0.038
Creatinine on index culture collection (mg/dL)	1.34 (1.0–3.2)	1.1 (0.8–1.9)	0.030
Albumin on index culture collection (g/dL)	2.3 (1.9–2.6) (44 patients)	2.8 (2.3–3.2) (221 patients)	<0.001
Baseline susceptible
Amikacin	43 (95.6)	218 (94)	1.0
Ceftriaxone	31 (70.5)	155 (67.7)	0.718
Ciprofloxacin	39 (86.7)	200 (86.2)	0.935
Meropenem	45 (100)	230 (99.1)	1.0
Piperacillin/tazobactam	29 (65.9)	175 (76.1)	0.156
Appropriate empirical therapy within 24 hours	36 (80)	181 (78)	0.768
Antibiotic treatment			0.139
3GC	7 (15.6)	66 (28.4)
Piperacillin/tazobactam	10 (22.2)	29 (12.5)
Quinolones	16 (35.6)	88 (37.9)
Carbapenem	12 (26.7)	49 (21.1)

Boldface indicates significant p-values.

Categorical variables are presented as number (%); continuous variables are presented as median (interquartile range).

3GC, third-generation cephalosporins.

Table 2.

Multivariate Analysis for Mortality

Variable	Odds ratio (95% confidence interval)	p
Age	1.026 (1.003–1.050)	0.029
Charlson score ≥4	2.143 (1.036–4.443)	0.040
PITT bacteremia score ≥4	1.400 (0.531–3.694)	0.496
Urinary catheter at presentation	1.044 (0.461–2.363)	0.918
Creatinine on index culture collection (mg/dL)^a	1.114 (0.853–1.455)	0.429
Albumin on index culture collection (g/dL)^b	0.271 (0.135–0.543)	<0.001
Antibiotic treatment (Carbapenem as reference)
3GC	0.708 (0.231–2.176)	0.547
Piperacillin/tazobactam	1.172 (0.388–3.537)	0.778
Quinolones	0.576 (0.229–1.452)	0.242

Boldface indicates significant p-values.

Odds ratio >1 indicates increased mortality.

Per increase of 1 mg/dL.

Per increase of 1 g/dL.

An additional analysis of risk factors for mortality among 211 patients with a non-urinary source of bacteremia is presented in Supplementary Table S2. Results of univariate and multivariate analyses were similar to the entire group, with no association between type of antibiotics and mortality (Supplementary Table S2).

Secondary outcomes are shown in Table 3. No differences were demonstrated between the main antibiotic regimens. A trend toward lower relapsed bacteremia and recurrent growth of Enterobacter with resistance to the main antibiotics was demonstrated with carbapenems. An additional analysis among subgroups of patients with urinary vs. non-urinary source demonstrated this trend in both subgroups (Table 3).

Table 3.

Secondary Outcomes by Type of Antibiotics

Variable	3GC	Piperacillin/tazobactam	Quinolones	Carbapenems	p
Growth of Enterobacter spp. in subsequent cultures of any source within 90 days	9/73 (12.3%)	9/39 (23.1%)	19/104 (18.3%)	11/61 (18.0%)	0.520
Relapsed bacteremia	5/73 (6.8%)	8/39 (20.5%)	11/104 (10.6%)	4/61 (6.6%)	0.095
Urinary source	2/19 (10.5%)	1/8 (12.5%)	2/25 (8%)	0/14	0.644
Non-urinary	3/54 (5.6%)	7/31 (22.6%)	9/79 (11.4%)	4/47 (8.5%)	0.101
Resistance at baseline to the main antibiotics	86/273 (31.5%)	70/274 (25.5%)	38/277 (13.7%)	2/277 (0.72%)	NA
Recurrent growth (90 days) of Enterobacter with resistance to the main antibiotics	19/48 (39.6%)	17/48 (35.4%)	6/47 (12.8%)	1/47 (2.2%)	NA
Urinary source	4/11 (36/4%)	3/11 (27.3%)	1/11 (9.1%)	0/11	NA
Non-urinary	15/37 (40.5%)	14/37 (37.8%)	5/36 (13.9%)	1/36 (2.8%)	NA
p-value for difference between baseline resistance to recurrent isolate resistance	p = 0.248	p = 0.162	p = 1.000	p = 0.376	NA

NA, nonapplicable.

Discussion

In this retrospective study, including 277 patients with Enterobacter bacteremia, we found no significant association between treatment with 3GC, piperacillin/tazobactam, quinolones, or carbapenems and 30-day mortality. Independent risk factors for mortality included age, the Charlson comorbidity score, and lower albumin. The main type of antibiotic treatment was not associated with repeated growth of the same Enterobacter spp. within 90 days of index infection. Among 48 patients with repeated growth, resistance to cephalosporins and piperacillin/tazobactam increased nonsignificantly.

Thirty-day mortality in our study was 16% (45/277), similar to previous studies.³ Similar outcomes with beta-lactam beta-lactamase inhibitors vs. carbapenems for treatment of bacteremia caused by potential AmpC-producing bacteria have already been reported. In a meta-analysis of 13 observational studies, including 1,021 patients overall, no significant difference between beta-lactam beta-lactamase inhibitors vs. carbapenems for the outcome of 30-day mortality was observed.¹⁵ In another meta-analysis of observational studies, including 538 patients overall treated with quinolones vs. carbapenems, lower mortality with quinolones was observed. This finding, however, was considered to be a result of selection bias, with patients with milder infection receiving quinolones.³ Currently, quinolones are still recommended only as oral step-down or treatment of mild-moderate infections.⁶ Data regarding treatment with cephalosporins for potential AmpC-producing organisms are scarce. Since a few studies showed emergence of resistance development to cephalosporins during therapy,^7,16,17 use of these drugs for serious Enterobacter infections has been discouraged.

Our study is limited by the retrospective design and a limited sample size. Secondary outcomes were rare and thus could not be fully evaluated with adjustment. Despite the attempts to adjust the analysis of mortality through multivariate analysis, the nonsignificant association between quinolone therapy and survival is likely due to residual confounding. Similarly, we cannot rule out that the lower (although nonsignificant) odds ratio for mortality with 3GC compared with piperacillin/tazobactam was not affected by confounders/bias, although baseline differences between groups were balanced (Supplementary Table S3). The sample size did not permit further adjust through a propensity score or other method. We did not test the strains for AmpC production. Cefepime is not used in the participating centers, and thus could not be evaluated. Strengths include the focus on Enterobacter spp., which has the highest potential to induce AmpC, and the relatively large number of patients treated with quinolones and cephalosporins, not commonly evaluated in previous studies due to limited use.¹⁸

Conclusions

We observed no differences in mortality or infection persistence/relapse among patients with Enterobacter sp. bacteremia treated with 3GC, quinolones, piperacillin-tazobactam, or carbapenems as the main therapy in the first week of treatment. Considering the presumed adverse implications of overuse of carbapenems, mainly carbapenem resistance development,^19,20 carbapenem-sparing treatment options for Enterobacter infections should be considered. The results of an ongoing randomized controlled trial may shed further light on this issue.²¹

Footnotes

Disclosure Statement

No competing financial interests exist.

Funding Information

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Supplementary Material

References

D'Angelo

R.G.

, Johnson

J.K

, Bork

J.T

, and Heil

E.L.

. 2016. Treatment options for extended-spectrum beta-lactamase (ESBL) and AmpC-producing bacteria. Expert Opin. Pharmacother. 17:953–967.

Harris

P.N.A.

, and Ferguson

J.K.

. 2012. Antibiotic therapy for inducible AmpC β-lactamase-producing Gram-negative bacilli: what are the alternatives to carbapenems, quinolones and aminoglycosides? Int. J. Antimicrob. Agents, 40:297–305.

Harris

P.N.A.

, Wei

J.Y

, Shen

A.W

, Abdile

A.A

, Paynter

, Huxley

R.R

, Pandeya

, Doi

, Huh

, O'Neal

C.S

, Talbot

T.R

, and Paterson

D.L.

. 2016. 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.

Jacoby

G.A.

2009. AmpC beta-lactamases. Clin. Microbiol. Rev. 22:161–182, Table of Contents.

Jacobson

K.L.

, Cohen

S.H

, Inciardi

J.F

, King

J.H

, Lippert

W.E

, Iglesias

, and VanCouwenberghe

C.J.

. 1995. The relationship between antecedent antibiotic use and resistance to extended-spectrum cephalosporins in group I beta-lactamase-producing organisms. Clin. Infect. Dis. 21:1107–1113.

Tamma

P.D.

, Doi

, Bonomo

R.A

, Johnson

J.K

, Simner

P.J

, and Antibacterial Resistance Leadership Group. 2019. A Primer on AmpC Beta-Lactamases: necessary knowledge for an increasingly multidrug-resistant world. Clin. Infect. Dis. 69:1446–1455.

Chow

J.W.

, Fine

M.J

, Shlaes

D.M

, Quinn

J.P

, Hooper

D.C

, Johnson

M.P

, Ramphal

, Wagener

M.M

, Miyashiro

D.K

, and Yu

V.L.

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

Kang

C.-I.

, Kim

S.-H

, Park

W.B

, Lee

K.-D

, Kim

H.-B

, Oh

, Kim

E.-C

, and Choe

K.-W.

. 2004. Bloodstream infections caused by Enterobacter species: predictors of 30-day mortality rate and impact of broad-spectrum cephalosporin resistance on outcome. Clin. Infect. Dis. 39:812–818.

Tamma

P.D.

, Girdwood

S.C.T

, Gopaul

, Tekle

, Roberts

A.A

, Harris

A.D

, Cosgrove

S.E

, and Carroll

K.C.

. 2013. The use of cefepime for treating AmpC β-lactamase-producing Enterobacteriaceae. Clin. Infect. Dis. 57:781–788.

10.

Harris

P.N.A.

, Alder

, and Paterson

D.L.

. 2015. Antimicrobial susceptibility reporting and treatment selection for AmpC-producing Enterobacteriaceae: what do microbiologists and infectious disease practitioners actually practice? Pathology, 47:386–388.

11.

Harris

P.N.A.

, Peri

A.M

, Pelecanos

A.M

, Hughes

C.M

, Paterson

D.L

, and Ferguson

J.K.

. 2017. Risk factors for relapse or persistence of bacteraemia caused by Enterobacter spp.: a case-control study. Antimicrob. Resist. Infect. Control. 6:14.

12.

Ternavasio-de la Vega

H.G.

, Castaño-Romero

, Ragozzino

, Sánchez González

, Vaquero-Herrero

M.P

, Siller-Ruiz

, Spalter-Glicberg

, Ramírez-Baum

, Rodríguez-Rodríguez

, García-Sánchez

J.E

, García-García

, and Marcos

. 2018. The updated Charlson comorbidity index is a useful predictor of mortality in patients with Staphylococcus aureus bacteraemia. Epidemiol. Infect. 146:2122–2130.

13.

Henderson

, Luterbach

C.L

, Cober

, Richter

S.S

, Salata

R.A

, Kalayjian

R.C

, Watkins

R.R

, Doi

, Kaye

K.S

, Evans

, Fowler

V.G

, Bonomo

R.A

, Harris

, Napravnik

, and van Duin

. 2020. The pitt bacteremia score predicts mortality in non-bacteremic infections. Clin. Infect. Dis. 70:1826–1833.

14.

Holsen

M.R.

, Wardlow

L.C

, Bazan

J.A

, Fussner

L.A

, Coe

K.E

, and Elefritz

J.L.

. 2019. Clinical outcomes following treatment of Enterobacter species pneumonia with piperacillin/tazobactam compared to cefepime or ertapenem. Int. J. Antimicrob. Agents. 54:824–828.

15.

Cheng

M.P.

, Lee

R.S

, Cheng

A.P

, De L'étoile-Morel

, Demir

, Yansouni

C.P

, Harris

, Mcdonald

E.G

, and Lee

T.C.

. 2019. Beta-lactam/beta-lactamase inhibitor therapy for potential AmpC-producing organisms: a systematic review and meta-analysis. Open Forum Infect. Dis. 6:ofz248.

16.

Choi

S.-H.

, Lee

J.E

, Park

S.J

, Choi

S.-H

, Lee

S.-O

, Jeong

J.-Y

, Kim

M.-N

, Woo

J.H

, and Kim

Y.S.

. 2008. Emergence of antibiotic resistance during therapy for infections caused by Enterobacteriaceae producing AmpC beta-lactamase: implications for antibiotic use. Antimicrob. Agents Chemother. 52:995–1000.

17.

Kaye

K.S.

, Cosgrove

, Harris

, Eliopoulos

G.M

, and Carmeli

. 2001. Risk factors for emergence of resistance to broad-spectrum cephalosporins among Enterobacter spp. Antimicrob. Agents Chemother. 45:2628–2630.

18.

Rodríguez-Baño

, Gutiérrez-Gutiérrez

, Machuca

, and Pascual

. 2018. Treatment of infections caused by extended-spectrum-beta-lactamase-, AmpC-, and Carbapenemase-producing Enterobacteriaceae. Clin. Microbiol. Rev. 31:e00079-17.

19.

Armand-Lefèvre

, Angebault

, Barbier

, Hamelet

, Defrance

, Ruppé

, Bronchard

, Lepeule

, Lucet

J.-C

, El Mniai

, Wolff

, Montravers

, Plésiat

, and Andremont

. 2013. Emergence of imipenem-resistant gram-negative bacilli in intestinal flora of intensive care patients. Antimicrob. Agents Chemother. 57:1488–1495.

20.

Shiber

, Yahav

, Avni

, Leibovici

, and Paul

. 2015. β-Lactam/β-lactamase inhibitors versus carbapenems for the treatment of sepsis: systematic review and meta-analysis of randomized controlled trials. J. Antimicrob. Chemother. 70:41–47.

21.

Trial of Meropenem Versus Piperacillin-Tazobactam on Mortality and Clinial Response—Full Text View—ClinicalTrials.gov. Available at https://clinicaltrials.gov/ct2/show/NCT02437045.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB

0.03 MB