Carbapenem stewardship with ertapenem and antimicrobial resistance-a scoping review

Abstract Consumption of carbapenem has increased due to extended-spectrum beta-lactamase-producing bacteria spreading. Ertapenem has been suggested as a not carbapenem-resistance inducer. We performed a scoping review of carbapenem-sparing stewardship with ertapenem and its impact on the antibiotic resistance of Gram-negative bacilli. We searched PubMed for studies that used ertapenem as a strategy to reduce resistance to carbapenems and included epidemiologic studies with this strategy to evaluate susceptibility patterns to cephalosporins, quinolones, and carbapenems in Gram-negative-bacilli. The search period included only studies in English, up to February 2018. From 1294 articles, 12 studies were included, mostly from the Americas. Enterobacteriaceae resistance to quinolones and cephalosporins was evaluated in 6 studies and carbapenem resistance in 4 studies. Group 2 carbapenem (imipenem/meropenem/doripenem) resistance on A. baumannii was evaluated in 6 studies. All studies evaluated P. aeruginosa resistance to Group 2 carbapenem. Resistance profiles of Enterobacteriaceae and P. aeruginosa to Group 2 carbapenems were not associated with ertapenem consumption. The resistance rate of A. baumannii to Group 2 carbapenems after ertapenem introduction was not clear due to a lack of studies without bias. In summary, ertapenem as a strategy to spare use of Group 2 carbapenems may be an option to stewardship programs without increasing resistance of Enterobacteriaceae and P. aeruginosa. More studies are needed to evaluate the influence of ertapenem on A. baumannii.


INTRODUCTION
Ertapenem is a carbapenem with weak activity against Pseudomonas spp. and Acinetobacter spp. 1 . In randomized controlled trials, ertapenem has been used for severe communityacquired infections and is licensed for intra-abdominal infections, community-acquired pneumonia, skin and soft tissue infections, and complicated urinary infections 2 . The importance of ertapenem increased after dissemination of extended-spectrum β-lactamases (ESBLs), which are now disseminating outside hospitals 3 .
Carbapenems from Group 1 (i.e., ertapenem) and Group 2 (i.e., meropenem) may select for resistant P. aeruginosa in vitro 4 . Nevertheless, the selection of carbapenem-resistant P. aeruginosa has been shown to be unlikely under physiological ertapenem concentrations. Considering the antimicrobial selective pressure, carbapenem-sparing stewardship strategies have increased in recent years 5 . However, some authors advocate ertapenem as a strategy to reduce resistance to meropenem and imipenem.
Considering the increasing importance of strategies to reduce antibiotic resistance, in this scoping review, we evaluated the effectiveness of an ertapenem-based stewardship strategy in reducing antibiotic resistance in Gram-negative bacilli (GNB).

Search strategy
Using PubMed, we searched for studies published in English that used ertapenem as a strategy to reduce resistance to any antibiotic. The search included studies from inception to February 2018. The keyword used was "ertapenem" in title and abstract in the advanced search option.

Data extraction and quality evaluation
Two reviewers (JT and FT) independently screened all studies based on either title or abstract for eligibility. Discrepancies were resolved through discussion. Reviewers then independently extracted the relevant data from all the publications included in the review. A third reviewer evaluated the discrepancies. The methodological quality of each publication was not analyzed using classical scores for randomized clinical trials, but basic elements for an objective evaluation were included in a table for critical analysis.

Inclusion and exclusion criteria
The inclusion criteria were as follows: i) epidemiological studies that compared different periods of ertapenem consumption (i.e., pre vs. post introduction) and ii) Evaluation of Group 2 carbapenem susceptibility pattern on Gram-negative bacilli. The exclusion criteria were: i) articles classified as case reports or individual data and/or ii) undescribed data of ertapenem consumption or susceptibility patterns.

Definitions and Gram-negative bacilli
The ertapenem consumption model was defined as DDD per patientday (i.e., DDD/100PD, DDD/1000PD). Susceptibility and resistance evaluation were described in a published original article. Susceptibility patterns were considered according to the Clinical and Laboratory Standards Institute (CLSI) or European Committee on Antimicrobial Susceptibility Testing (EUCAST). The analyzed resistances according to each GNB were: i) quinolone in E. coli and K. pneumoniae, ii) third-generation cephalosporin in E. coli and K. pneumoniae, and iii) carbapenems in E. coli, K. pneumoniae, A. baumannii, and P. aeruginosa.

Selected articles
The search criteria initially identified 1294 articles. After title and abstract reviews, only 12 articles fulfilled the inclusion criteria (Figure 1). The first study was published in 2008 and the last in 2015. The period of analysis varied between 2000 and 2011.
Of the articles, 7 were from America 6-12 , 4 from Asia 13-16 , and 1 from Europe 17 . A timeline of the ertapenem-based stewardship program of each study is presented in Figure 2.

Carbapenem consumption
Carbapenem consumption (Groups 1 and 2) was evaluated using different methods. Three studies used the slope curve and nine used comparative periods (before and after consumption). Thus, there was heterogeneity in the metrics used among authors, which complicates the establishment of a median or average value. Only 2 studies demonstrated the substitution tendency of Group 2 carbapenems to ertapenem after its introduction 10,11 .

E. coli susceptibility
Three studies analyzed ertapenem consumption and E. coli carbapenem resistance rate 6,12,13 and one did not specify resistance among Enterobacteriaceae isolates 9 ( Tables 1 and Supplementary Data  -Table 2). Increased ertapenem consumption did not increase E. coli resistance to carbapenems. Quinolones were analyzed by 4 studies and third-generation cephalosporins by 6, and presented bias on results 6,11-16 (Supplementary Data - Table 2). Only 1 publication found a significant increase in quinolone resistance, although higher ciprofloxacin consumption was observed as well 15 . An increased resistance rate to third-generation cephalosporin was observed in 4 studies, but ceftriaxone, ceftazidime, and beta-lactamase inhibitor consumption rates were also higher in 3 studies 6,13,15,16 .

K. pneumoniae susceptibility
Three studies analyzed ertapenem consumption and K. pneumoniae carbapenem resistance rate 6,12,13 , and one did not specify resistance among Enterobacteriaceae isolates 9 ( Table 1). Increased consumption of ertapenem changed the susceptibility patterns of carbapenems in some studies. One study showed a slight improvement in carbapenem susceptibility 13 . Another study found a higher incidence of resistance to Group 2 carbapenems on univariate analysis; however, higher consumption of meropenem/imipenem was observed 6 . Quinolones and third-generation cephalosporin susceptibility were analyzed in 4 and 6 studies respectively 6,12-16 (Supplementary Data - Table 2). Increased third-generation cephalosporin resistance was observed in 4 studies 6,12,13,16 .

DISCUSSION
We conducted a scoping review to better understand Gramnegative bacilli antibiotic resistance and ertapenem consumption. Twelve studies evaluated ertapenem consumption as an intervention to change Group 2 carbapenem resistance. After this strategy, the Group 2 carbapenem was reduced in 3 studies. Carbapenem resistance in Enterobacteriaceae did not increase after ertapenem consumption. However, non-fermenting Gram-negative bacilli demonstrated changes in susceptibility patterns. Carbapenemresistant in A. baumannii increased in 2 of 6 studies, while 4 observed no difference. P. aeruginosa improved carbapenem susceptibility in 3 of the 12 studies, while 7 observed no differences and 2 increased carbapenem resistance.
The hypothesis that ertapenem has the potential to select P. aeruginosa and A. baumannii resistant to Group 2 carbapenems is due to its limited action on non-fermenting Gram-negative bacilli (NF-GNB). Previous reviews did not observe higher rates of carbapenem resistance in NF-GNB despite an increase in ertapenem consumption 18,19 . K. pneumoniae carbapenem resistance rate did not increase overall and it was positively affected by routine utilization of ertapenem in one study. Lee et al. (2010) observed an improvement in susceptibility to carbapenems, ceftazidime, and levofloxacin after ertapenem introduction 13 . Changes in the resistance rate of K. pneumoniae to cephalosporin and quinolones were observed. Hsu et al. (2010) demonstrated lower resistance to ceftriaxone and ciprofloxacin but this was not correlated with antibiotic consumption 15 . Goff and Mangino (2008) observed higher resistance to cephalosporins in the latter period and inferred it was due to multiple hospitalizations 12 . Overall, Enterobacteriaceae carbapenem resistance was not affected by ertapenem consumption. These results are in accordance with stable CRE colonization rates after patients using ertapenem as surgical prophylaxis 20 .
A. baumannii demonstrated predominantly no difference in the results and worst susceptibility patterns in 2 studies 13,15 . However, there was a significant increase in consumption in Group 2 carbapenems and other broad-spectrum antibiotics.
Yoon et al converged with these results when they concluded that carbapenem resistance rate is correlated with Group 2 carbapenem consumption 16 .
Carbapenem-resistant P. aeruginosa was not increased by ertapenem use in the majority of studies. Increased resistance rates were demonstrated in a study with higher Group 2 carbapenem consumption 13 . Nevertheless, Lim et al. (2013) observed a negative impact on carbapenem susceptibility even with no difference in Group 2 carbapenem consumption in both periods 14 . Similar to A. baumannii, other studies found that P. aeruginosa resistance was affected by Group 2 carbapenem consumption but not by ertapenem 21,22 . These studies converged with two positive results in the present review 11,17 , in which lower resistance was correlated with less usage of imipenem. Only one study directly associated ertapenem consumption with better carbapenem susceptibility 9 .
The present study has several limitations. Methods heterogeneity may make certain conclusions difficult when studies were not comparable between each other. Other factors may have influenced the carbapenem resistance rate of Group 2, such as higher meropenem/imipenem consumption, without multivariate analysis evaluation. However, this article presents a relevant issue in infectious disease practice and may help stewardship programs to adequately choose carbapenem therapeutic regimens without affecting the bacterial resistance rate.

CONCLUSION
The majority of studies did not demonstrate a rising Group 2 carbapenem resistance rate in Enterobacteriaceae and P. aeruginosa after ertapenem introduction. The rate of resistance to Group 2 carbapenems on A. baumannii is not clear. However, studies did demonstrate that worsening carbapenem resistance was associated with Group 2. If a carbapenem group is needed in an antimicrobial stewardship program, ertapenem may be an option to spare Group 2 carbapenem usage without increasing resistance in Enterobacteriaceae and P. aeruginosa.

AUTHORS' CONTRIBUTION
TZ: Wrote manuscripts, articles review and selection, analysis and interpretation of data. JPT: Conception and design of the study, articles reviewer and selection, manuscript review. JG: Final manuscript review. FFT: Conception and design of the study, articles reviewer and selection, final approval of the version to be submitted.

CONFLICTS OF INTEREST
Felipe Tuon conducts research for CNPQ.

FINANCIAL SUPPORT
The authors have no funding to report. SUPPLEMENTARY DATA -