Treatment of extended-spectrum β-lactamase-producing Enterobacteriaceae (ESBLs) infections: what have we learned until now?

The spread of extended-spectrum β-lactamase (ESBL)-producing Enterobacteriaceae (ESBL-PE) has dramatically increased worldwide, and this “evolving crisis” is currently regarded as one of the most important public health threats. The growing problem of ESBL-PE antimicrobial resistance seems to have a dual face between “Scylla and Charybdis”: on one hand the potential for rapid spread and dissemination of resistance mechanisms and on the other hand the injudicious overuse of antimicrobial agents and the inadequate infection control measures, especially in the health-care setting. Given the World Health Organization’s warning against a “post antibiotic era”, health-care providers are at a critical standpoint to find a “balance” between safe and effective ESBL-PE treatment and avoidance of inducing further resistance mechanisms. The aim of the review is to summarize the updated published knowledge in an attempt to answer basic everyday clinical questions on how to proceed to effective and the best ESBL-PE treatment options based on the existing published data.


Introduction
Extended-spectrum β-lactamase (ESBL) enzymes are characterized by the ability to hydrolyze third-generation cephalosporins and aztreonam but are inhibited by clavulanic acid. The spread of ESBL-producing Enterobacteriaceae (ESBL-PE) has dramatically increased worldwide, and this "evolving crisis" is currently regarded as one of the most important public health threats. The growing problem of ESBL-PE antimicrobial resistance seems to have a dual face between "Scylla and Charybdis": on one hand the potential for rapid spread and dissemination of resistance mechanisms and on the other hand the injudicious overuse of antimicrobial agents and the inadequate infection control measures, especially in the health-care setting. A multicenter study in the US reported that in 2012 the prevalence of ESBLproducing Klebsiella pneumoniae reached 16% and almost 12% for ESBL-producing Escherichia coli and that rates were much higher among intensive care unit (ICU) patients 1 . Even in the pediatric population, a meta-analysis revealed that the worldwide prevalence of ESBL producers was estimated to be 9% (11% neonates and 5% children) with an annual increase of 3.2% and a wide variability among different geographic regions (15% in Africa, 12% in South America, 11% in India, 7% in the rest of Asia, and 4% in Europe) 2 .
ESBL-PE are associated with increased morbidity and mortality rates, prolonged hospital stays, and increased costs. In a matched cohort study, the nosocomial financial burden of non-urinary tract infections (non-UTIs) caused by ESBL producers was 1.7 times higher compared with the same type of infections caused by non-ESBL producers 3 . A case control study in Canada showed that ESBL-PE infections were significantly associated with increased cost (C$10,507 versus C$7,882), hospitalization (8 versus 6 days), and mortality rates (17% versus 5%) 4 . In addition, data regarding the rates of ESBL-PE colonization (both health-care or community acquired) reveal an increasing trend over time with significant differences among several geographical regions and patient groups 5 . For high-risk patients in the ICU, the ESBL-PE colonization rates might range from 2.3% for the US to 49% for India. According to a recently published systematic review, the most frequently reported risk factor for ESBL-PE colonization and infection remains prior exposure to antimicrobials as well as recent hospitalization and recent or repeated surgery 5 . Although prior ESBL-PE colonization has been shown in a few studies to increase the risk of acquiring an ESBL-PE infection, further data are needed.
Given the World Health Organization (WHO) warning against a "post antibiotic era", health-care providers are at a critical standpoint to find a "balance" between safe and effective ESBL-PE treatment and avoidance of inducing further resistance mechanisms. The aim of this review is to summarize the updated published knowledge in an attempt to answer basic everyday clinical questions on how to proceed to effective and the best ESBL-PE treatment options based on the existing published data.

Before starting ESBL treatment
The first step before initiating ESBL-PE antimicrobial treatment is to carefully evaluate specific parameters that are directly associated with ESBL-PE therapeutic decision making. Of utmost importance is to clearly characterize (a) the isolate with the in vitro susceptibilities, (b) the location of the infection, (c) the degree of source control of the infection, and finally (d) the clinical condition of the patient (Table 1). In addition, recently published studies propose that all ESBL-PE do not belong in the same homogenous group as far as comorbidities, presentation, and outcome are concerned 6 . In particular, data have shown that bloodstream infections (BSIs) associated with ESBL-producing E. coli were more frequently of a urinary source and community onset compared with BSIs with ESBL-producing Klebsiella spp. 6 .

Decision making on ESBL-PE antimicrobial treatment
Clinical question 1: Carbapenems or β-lactam/β-lactamase inhibitor combinations in ESBL-PE infections? Carbapenems possess the broadest spectrum of β-lactam antibiotics with greatest potency against Gram-negative bacteria and are characterized by stability to hydrolysis by the majority of β-lactamases 7 . Several studies have shown that carbapenem treatment is associated with improved outcomes in patients with severe ESBL-PE infections and remains the "gold standard" treatment for serious and invasive ESBL-PE infections 8,9 . Specific considerations among carbapenems are the induction of carbapenem resistance and their side effects, especially as far as their epileptogenic effect is concerned 10 . Most studies have evaluated either meropenem or imipenem for the treatment of ESBL-PE, although a recently published multinational retrospective study compared the clinical efficacy of ertapenem with that of other carbapenems in ESBL-PE BSIs 11 . Cure rates were similar (90.6% with ertapenem and 75.5% with other carbapenems in empiric and 89.8% and 82.6% in targeted treatment), and no differences have been observed for mortality among the two groups, but for patients with severe sepsis a non-significant trend favoring other carbapenems was observed 11 .
Among β-lactam/β-lactamase inhibitor (BLBLI) combinations, the combination of piperacillin-tazobactam (PTZ) has been extensively studied as an alternative carbapenem-sparing option against ESBL-PE infections 12 . In 2012, a meta-analysis compared the mortality rates among carbapenems and alternative regimens, including BLBLIs, for the treatment of ESBL-PE BSIs 13 . According to their results, differences were noticed in mortality rates when administered as either definitive or empirical therapy, although they mentioned one study's significant heterogeneity 13 . Since the question about the role of BLBLIs remained, six subsequent studies from 2012 to 2017 tried to elucidate the role of BLBLIs against ESBL-PE with rather conflicting results 14-18 . These controversies were interpreted by substantial differences among the studies' design.  27 showed significantly lower mortality rates at 30 and 14 days, respectively (17% versus 59% and 41% versus 20%, respectively). In particular, in the study by Lee et al., a significant association was observed between the mortality rates of the patients receiving cefepime and the MIC of the drug. In particular, for cefepime MIC of not more than 1 μg/mL, the mortality rate was 16.7%; for MIC of 2-8 μg/mL, the rates reached 45.5%; while for MIC of at least 16 μg/mL, the rates were 100% (p = 0.035) 31 . Even after propensity score adjustment, cefepime remained inferior compared with carbapenem (adjusted odds ratio 6.8, 95% confidence interval (CI) 1.5-31.2, p = 0.01). A subsequent randomized controlled trial was conducted comparing PTZ, cefepime, and ertapenem for the treatment of ESBL-PE UTIs caused by E. coli showing inferiority of cefepime compared with the other treatment options. The efficacy of cefepime was 33.3% compared with 94% efficacy of PTZ treatment 20 .

Authors' recommendations.
For serious invasive ESBL-PE infections with high inoculum and lack of source control, cefepime seems to be inferior compared with carbapenems because of two significant issues: increased MICs of the drug because of high inoculum effect and failure to achieve its pharmacodynamic targets in severe ESBL-PE infections. Cefepime could be administered only in non-severe ESBL-PE UTIs, where high drug concentrations could be achieved and when simultaneously low MIC of the drug is reported (MICs ≤2 μg/mL).

Clinical question 3: What is the role for fosfomycin, aminoglycosides, or temocillin in ESBL-PE infections?
Fosfomycin is an old bactericidal antibiotic agent (phosphonic compound) with a unique mode of action of inhibiting bacterial cell wall biosynthesis 32,33 . A recently published literature review concerning the susceptibility of contemporary Gram-negative bacteria revealed that, for ESBL-producing E. coli, susceptibilities ranged from 81% to 100% and, for ESBL-producing K. pneumoniae, from 15% to 100% 34 . Owing to its low molecular weight, its hydrophilicity, and its negligible serum protein binding, the drug achieves good tissue penetration and high concentrations in the serum, soft tissue, lungs, bone, cerebrospinal fluid, and heart valves 35,36 . Especially for the urinary tract, the drug achieves high concentrations for a prolonged period of time. Finally, in critically ill patients, a significant increase of fosfomycin volume of distribution is observed; therefore, the current paucity of data on fosfomycin in critically ill patients prevents accurate dosing guidance 36 . Clinical data concerning the efficacy of intravenous fosfomycin against ESBL-PE invasive infections are very limited and focus mainly on UTI treatment 32,37-42 . A randomized clinical trial ("FOREST") comparing the safety and efficacy of fosfomycin versus meropenem in bacteremic UTIs caused by ESBL-producing E. coli is ongoing 38 . Fosfomycin as monotherapy for the treatment of multidrugresistant organism (MDRO)-associated invasive infections is limited by the emergence of drug resistance to fosfomycin during treatment 39 . A more recently published meta-analysis conducted by Grabein et al. tried to summarize the current clinical evidence of intravenous fosfomycin in 128 studies 43 . According to their results, the drug showed comparable clinical or microbiological efficacy compared with other antibiotics when used for sepsis/bacteremia, urinary tract, respiratory tract, bone and joint, and central nervous system infections 43 . The pooled estimate for resistance development during fosfomycin monotherapy was 3.4% (95% CI 1.8%-5.1%).
Up-to-date data concerning the role of aminoglycosides in combating MDRO infections showed that for ESBL infections they can be used as part of a combination regimen, especially for UTIs and intra-abdominal infections (IAIs), as a carbapenemsparing option 8 . An in vitro synergistic effect has been confirmed for the concomitant administration of aminoglycosides plus β-lactams, while the monotherapy is not generally recommended for ESBL-PE infections, except for ESBL-PE non-bacteremic UTIs, mainly owing to the high risk of resistance development [44][45][46][47] . Among newer aminoglycosides, plazomicin (formerly ACHN-490), a next-generation aminoglycoside synthetically derived from sisomicin, is recently gaining more attention against MDRO infections 47,48 . The unique characteristic of plazomicin is its resistance to inactivation by aminoglycoside-modifying enzymes compared with other agents of the same group 47,48 . However, plazomicin is not active against bacterial isolates expressing ribosomal methyltransferases 47,48 . In two studies, plazomicin has been shown to be more potent than other aminoglycosides in treating Enterobacteriaceae 49,50 .
Temocillin is a β-lactamase-resistant carboxypenicillin active against both ESBL-PE and AmpC-producing Enterobacteriaceae and with limited activity against Pseudomonas, Acinetobacter spp., and anaerobic bacteria. Although this carbapenem-sparing drug option is licensed in only a few European countries (UK and Belgium), data from a multicenter study in the UK among 92 infection episodes (42 BSIs) treated with temocillin showed promising results 51 . In particular, both clinical and microbiological cure rates were reported to be 86% and 84% 51 . In addition, a prospective randomized controlled trial conducted in Belgium in 2014 showed that for critically ill patients the optimal dose regimen for temocillin in order to achieve its pharmacological targets with longer free-serum concentrations is 2 g three times a day administered by continuous infusion 52 .

Authors' recommendations.
Fosfomycin is strongly suggested for ESBL-PE UTIs and as a step-down therapy in sourcecontrolled ESBL-PE infections. A randomized clinical trial ("FOREST") comparing the safety and efficacy of fosfomycin versus meropenem in bacteremic UTIs caused by ESBLproducing E. coli is ongoing 41 . Other options of source-controlled ESBL-PE UTIs are aminoglycosides, especially for cystitis infections. In addition, they can be used as part of a combination regimen, especially for UTIs and IAIs, as a carbapenemsparing option 8 . For temocillin, larger clinical studies among different patient groups are needed in order to establish their role as a valuable carbapenem-sparing option against ESBL-PE BSIs.
Clinical question 4: What is the role of the newly approved drugs against ESBL-PE infections?
In Table 2, some of the new drugs active against multidrugresistant bacteria, including ESBL-producing ones, are reported. Among newer BLBLIs developed, two of them-ceftazidimeavibactam and ceftolozane-tazobactam-have already received US Food and Drug Administration (FDA) approval and therefore will be discussed further.
Ceftazidime-avibactam is a combination of cephalosporin with a new non-BLBLI that is generally active against Enterobacteriaceae and P. aeruginosa producing class A β-lactamases (ESBLs and KPCs) and class C β-lactamases (AmpCs) and some Enterobacteriaceae producing class D β-lactamases (OXAs) but lacks activity against class B carbapenemases and is less active against anaerobes compared with other BLBLIs. A phase 3 trial (RECLAIM 1 and RECLAIM 2) conducted by Mazuski et al.  Ceftolozane-tazobactam is a co-formulation of a novel cephalosporin with an old β-lactamase inhibitor. Ceftolozane is a new cephalosporin based on the scaffold of ceftazidime-with only one modification of the side chain at the 3-position of the cephem nucleus-with improved activity against multidrug-resistant Pseudomonas spp. Ceftolozane, like other oxyimino-cephalosporins such as ceftazidime and ceftriaxone, is not stable against class A, B, or D β-lactamases (mainly ESBLs or carbapenemases). The combination with tazobactam significantly broadens its spectrum against ESBL-PE and against few anaerobes 56,57 . In 2014, the FDA approved the administration of the combination drug for the treatment of complicated UTIs and cIAIs based on previously published clinical trials (ASPECT trials) 58-60 . In particular, the drug was evaluated in phase 3 non-inferiority clinical trials versus levofloxacin 750 mg daily in complicated UTI or meropenem 1 g q8h in cIAI. The UTI trial compared ceftolozane 1,000 mg q8h versus ceftazidime 1,000 mg q8h, including pyelonephritis, and demonstrated similar microbiologic and clinical outcomes, as well as a similar incidence of adverse effects after 7 to 10 days of treatment, respectively. The second cIAI trial has been conducted comparing ceftolozanetazobactam 1,000/500 mg and metronidazole 500 mg q8h versus meropenem 1,000 mg q8h in the treatment of cIAI. The recommended FDA dosage is 1 g/0.5 g 8 hourly for 7 days in complicated UTIs and 4 to 14 days in cIAIs, respectively 61 .

Grant information
The author(s) declared that no grants were involved in supporting this work.