An Update on Eight “New” Antibiotics against Multidrug-Resistant Gram-Negative Bacteria

Infections in the ICU are often caused by Gram-negative bacteria. When these microorganisms are resistant to third-generation cephalosporines (due to extended-spectrum (ESBL) or AmpC beta-lactamases) or to carbapenems (for example carbapenem producing Enterobacteriales (CPE)), the treatment options become limited. In the last six years, fortunately, there have been new antibiotics approved by the U.S. Food and Drug Administration (FDA) with predominant activities against Gram-negative bacteria. We aimed to review these antibiotics: plazomicin, eravacycline, temocillin, cefiderocol, ceftazidime/avibactam, ceftolozane/tazobactam, meropenem/vaborbactam, and imipenem/relebactam. Temocillin is an antibiotic that was only approved in Belgium and the UK several decades ago. We reviewed the in vitro activities of these new antibiotics, especially against ESBL and CPE microorganisms, potential side effects, and clinical studies in complicated urinary tract infections (cUTI), intra-abdominal infections (cIAI), and hospital-acquired pneumonia/ventilator-associatedpneumonia (HAP/VAP). All of these new antibiotics are active against ESBL, and almost all of them are active against CPE caused by KPC beta-lactamase, but only some of them are active against CPE due to MBL or OXA beta-lactamases. At present, all of these new antibiotics are approved by the U.S. Food and Drug Administration for cUTI (except eravacycline) and most of them for cIAI (eravacycline, ceftazidime/avibactam, ceftolozane/tazobactam, and imipenem/relebactam) and for HAP or VAP (cefiderocol, ceftazidime/avibactam, ceftolozane/tazobactam, and imipenem/relebactam).


Introduction
Two-thirds of the infections in intensive care units (ICUs) are caused by Gram-negative bacteria [1], and resistance of these bacteria to certain antibiotics poses a significant problem. Several multidrug resistant, Gram-negative microorganisms such as carbapenem-resistant Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacterales are duly included in the recent WHO list of high-priority pathogens [2].
Beta-lactam antibiotics-penicillins, cephalosporins, carbapenems, and to a lesser extent monobactams-are often the antibiotics of choice in the ICU, but they can be inactivated by beta-lactamase enzymes (the most common and important mechanism of resistance in Gram-negative bacteria). These beta-lactamases can be classified into four molecular classes: A (for example, KPCs that confer resistance to cephalosporins and to all carbapenems, and extended-spectrum beta-lactamases (ESBLs) that confer resistance to cephalosporins); B (metallo-beta-lactamases (MBLs), such as NDM, VIM, and IMP, which can lead to resistance to all carbapenems except monobactam); C (for example, AmpC (mostly chromosomal but can also be plasmidal), which confer resistance to cephalosporins); and D (for example, OXAs that confer resistance mostly to carbapenems) [3].
Among the beta-lactam antibiotics, carbapenems have the widest spectrum of activity and are often the antibiotic class of choice when Gram-negative bacteria produce ESBL or AmpC beta-lactamases. Harbouring genes encoding ESBL and AmpC betalactamases confer resistance to all members of penicillin, cephalosporines up to "third" generation, and aztreonam (a monobactam antibiotic). Resistance to carbapenems poses thus a big problem. Carbapenem nonsusceptibility can be due to carbapenemase production or other mechanisms. The distinction is important from the infection-prevention point of view because carbapenemase production refers to resistance caused by beta-lactamase production encoded by transmissible genes, and not due to, for example, efflux pumps. These transmissible genes can be possessed by Enterobacterales (denoted as carbapenemproducing Enterobacterales, CPE), or by nonfermenting Gram-negative bacteria such as Acinetobacter spp. and Pseudomonas aeruginosa [4]. There is a geographic diversity in the distribution of these CPE genes. OXA-48 is most common in Europe; KPC in the US, South America and China; NDM in South Asia; and VIM in Australia [5][6][7]. To treat carbapenem-nonsusceptible microorganisms, colistin is often used. However, these microorganisms can also develop resistance to colistin; for example, by means of the mcr-1 gene [8].
In the last five years, there have been several new antibiotics with predominant activity against Gram-negative bacteria approved by the U.S. Food and Drug Administration (FDA) and the European Medical Agency (EMA), i.e., plazomicin, eravacycline, cefiderocol; and antibiotics that combine beta-lactam with beta-lactamase inhibitors, i.e., ceftazidime/avibactam, ceftolozane/tazobactam, meropenem/vaborbactam, and imipenemcilastatin/relebactam. Next to these new antibiotics, temocillin, a beta-lactam antibiotic effective against Gram-negative bacteria that is only approved in Belgium and the United Kingdom, is also worthy of mention.
This narrative review aims to analyze the spectrum (especially against ESBL and CPE microorganisms), potential side effects, and clinical studies of these antibiotics, with a special focus on infections in critically ill patients with complicated urinary tract infections (cUTI), intra-abdominal infections (cIAI), or hospital-acquired pneumonia/ventilatorassociated pneumonia (HAP/VAP).

Methods
Newly approved (between 2015 and 2020) antibiotics with predominant activity against Gram-negative pathogens were identified from the FDA and EMA website. A subsequent search was performed in MEDLINE up to 1 December 2020, to search for information regarding surveillance studies on the activity of these antibiotics against Enterobacterales, Pseudomonas, and Acinetobacter isolates, including those that are ESBL producers or carbapenem-nonsusceptible. The results of this search are presented in Tables S1 and S2. MEDLINE and clinical trial register (clinicaltrials.gov (accessed on 1 March 2021)) searches were performed to find clinical studies on these antibiotics. We included studies performed in adult patients only. The results of the included clinical trials are presented in Table 1 (cUTI), Table 2 (cIAI), Table 3 (HAP/VAP), and of the included observational studies in Table 4.

Plazomicin
Plazomicin is a synthetic aminoglycoside [9], and like other aminoglycosides, it inhibits bacterial protein synthesis and has dose-dependent bactericidal activity in vitro [10]. The FDA approved plazomicin (brand name Zemdri) in 2018 for use in cUTI and pyelonephritis at a dose of 15 mg/kg IV, QD. The plazomicin FDA package insert includes nephrotoxicity and ototoxicity as possible side effects. However, a pooled analysis of three studies on plazomicin showed that nephrotoxicity in patients receiving plazomicin was comparable to nephrotoxicity in comparator drugs (4.8% vs. 4.1%) [11].
Two clinical studies have been conducted in patients with complicated urinary tract infection (cUTI) ( Table 1) comparing plazomicin 15 mg/kg IV, QD with meropenem 1 g IV, TID [18] and levofloxacin 750 mg IV, QD for up to 10 days [19]. Both studies showed noninferiority of plazomicin. In these studies, only a limited number of patients were included with carbapenem resistant Enterobacterales (CRE) or P. aeruginosa. Another randomized control trial (RCT) that was stopped prematurely due to slow enrollment of patients with bloodstream infection and bacterial pneumonia due to CRE showed that death from any cause at 28 days or clinically significant disease-related complications occurred in 24% of patients who received plazomicin 15 mg/kg IV, QD, versus 50% of patients receiving colistin (5 mg colistin base/kg QD) in combination with meropenem or tigecycline for 7 to 14 days [20].

Eravacycline
Eravacycline is a fluorocycline of the tetracycline class. Like other tetracyclines, it inhibits bacterial protein synthesis. The FDA approved eravacycline (brand name Xerava) in 2018 for the treatment of cIAI at a dose of 1 mg/kg IV, BD for a total duration of 4 to 14 days. It may cause the same adverse events as other tetracyclines, such as hypersensitivity reactions and permanent tooth discoloration [29]. Further, the most common adverse events are infusion-site reactions (7.7%), nausea (6.5%), vomiting (3.7%), and diarrhea (2.3%) [29].
In two separate RCTs, eravacycline 1.5 mg/kg IV, BD was shown to be noninferior to meropenem 1 g IV, TD [34] or ertapenem 1 g IV, QD [35] (Table 2) in treating cIAI patients (>80% of the patients also underwent surgery). The majority of pathogens in these studies were Gram-negative (>70%), including ESBL-and CPE-producing isolates, but only a limited number of infections (<10%) were due to P. aeruginosa. In patients with ESBL-producing Enterobacterales, clinical cure rates were 87.5% (14/16) and 84.6% (11/13) in the eravacycline and meropenem groups, respectively [34]. However, the cure rate in those patients was also > 90%, emphasizing the importance of surgery. Eravacy-cline has been also investigated in cUTI. Two RCTs on the clinicaltrials.gov (accessed on 1 March 2021) trials register compared eravacycline with ertapenem (clinical trial register number NCT03032510) and levofloxacin (NCT01978938) in cUTI (Table 1). Both studies showed lower cure rates (84.8% vs. 94.8%) and (60.4% vs. 66.9%) than the comparators, respectively. In both studies, the eravacycline dose was 1.5 mg/kg IV, QD. No information regarding pathogens causing UTI could be found from these two studies, and no explanation can be given regarding why the latter study showed lower clinical response in both the treatment arm and control group. A small observational study including 35 ICU patients mainly with cIAI and pneumonia with a median Acute Physiology And Chronic Health Evaluation (APACHE) score of 16 showed absence of 30-day recurrence of 91%, but only 57% resolution of signs and symptoms at 30-day follow-up [36].

Temocillin
Temocillin is a derivative of ticarcillin, a penicillin antibiotic targeting PBP3 primarily, that was developed and marketed in the UK in the 1980s, but was immediately abandoned because of its lack of activity against Gram-positive microorganisms, nonfermenters (including A. baumannii and P. aeruginosa), and anaerobes [42]. However, there has been a renewed interest in this antibiotic agent in the last decade as a carbapenem-sparing op-tion, given the increasing incidence of infections with Enterobacterales resistant to thirdgeneration cephalosporins.
Temocillin is unaffected by ESBL and AmpC, but is affected by OXA-48 and MBL [42][43][44][45]. There is no breakpoint for temocillin available from the two largest bodies that set up breakpoints for antibiotics, the European Committee for Antimicrobial Susceptibility Testing (EUCAST) or the Clinical and Laboratory Standards Institute (CLSI). The British Society for Antimicrobial Chemotherapy (BSAC) publishes two breakpoints, one for infections of the urinary tract (≤32 mg/L), and the other for systemic infection (≤8 mg/L) [45]. When the BSAC breakpoint for urine is used, temocillin shows activity against KPC-producing bacteria [45].
It has been suggested that temocillin may be used for the treatment of UTI, but so far, no results of RCTs are available. One observational study in UTI and BSI caused by ESBL or derepressed AmpC beta-lactamase-producing Enterobacterales showed clinical cure of 86% and microbiological cure of 84% [43]. In the clinical trial register, there are two ongoing studies (NCT03543436 and registry number NCT04478721) identified that compare temocillin to carbapenems for the treatment of treat cUTI due to Gram-negative bacteria resistant to third-generation cephalosporins.

Cefiderocol
Cefiderocol (brand name Fetroja) is a novel catechol-substituted siderophore [46]. It was approved by the FDA in 2019 at a dose of 2 g IV, TD for the treatment of cUTI, and this indication was extended to HAP/VAP in 2020. The most frequently occurring adverse reactions of cefiderocol according to the FDA package insert were diarrhea, infusionsite reactions, constipation, rash, candidiasis, cough, elevations in liver tests, headache, hypokalaemia, nausea, hypomagnesemia, and atrial fibrillation. Due to a cause that has not been established yet, cefiderocol has an FDA label warning for higher all-cause mortality versus other antibiotics in critically ill patients with multidrug-resistant Gram-negative bacterial infections, as shown in the study of Bassetti and coworkers that reported a mortality rate of 34% for cefiderocol vs. 18% in the best-available therapy group [47].
A clinical trial comparing cefiderocol to imipenem-cilastatin in patients with cUTI for a median treatment duration of 9 days showed significantly higher clinical and microbiological response than in the group on imipenem (73 vs. 55%) ( Table 1) [21]. In this study, the most common pathogens were E. coli and Klebsiella spp. (no information on their susceptibility to the third-generation cephalosporins was available), and 7% P. aeruginosa. A fresh from the press RCT comparing cefiderocol 2 g IV, TID in 145 patients with meropenem 2 g IV, TID in 146 nosocomial pneumonia patients showed comparable mortality at day 14, i.e., 12.4% vs. 11.6% [51]. A small observational study including 10 critically ill patients with bacteraemia and VAP caused by carbapenem-resistant (MIC ≥ 2 mg/L) A. baumannii, Stenotrophomonas maltophilia, or NDM-producing K. pneumoniae [52] who experienced clinical failure on previous antibiotic (including colistin) showed 70% clinical success at 30 days. Clinical success in this study was defined as survival, resolution of signs and symptoms, and no recurrent infection and no microbiological failure.

Ceftazidime/Avibactam
The FDA approved ceftazidime avibactam (brand name Avycaz) in 2015 for treating cIAI (in combination with metronidazole) and cUTI, at a dose of 2.5 g IV, TD, and extended to HAP/VAP in 2018. According to the FDA package insert, ceftazidime/avibactam had comparable possible adverse events to those of ceftazidime alone; the most commonly reported adverse reactions (in ≥5% of patients) were nausea and diarrhea, and positive direct Coombs test.
Ceftazidime/avibactam shows activity against Gram-negative bacteria producing ESBL, carbapenem nonsusceptible Enterobacterales [57], and P. aeruginosa [58] as long as they are not due to the presence of MBL genes. Data on the susceptibility of Acinetobacter to ceftazidime/avibactam is limited.

Ceftolozane/Tazobactam
Ceftolozane/tazobactam is an antipseudomonal cephalosporin partnered with BLI tazobactam. This antibiotic combination was approved by the FDA (brand name Zerbaxa) in 2014 for cUTI and cIAI indications at a dose of 1.5 g IV, TD. The indication was extended to HAP/VAP in 2019. Like ceftazidime/avibactam, possible adverse events that are most commonly reported (in ≥5% of patients) were nausea and diarrhea, according to the FDA package insert.

Meropenem/Vaborbactam
The FDA approved meropenem/vaborbactam (brand name Vabomere) to treat cUTI at a dose of 4 g (meropenem 2 g and vaborbactam 2 g) IV, TD. According to the FDA package insert, the adverse reactions occurring in ≥ 3% of patients treated with this antibiotic were headache, phlebitis/infusion-site reactions, and diarrhea.
The addition of vaborbactam improves the activity of carbapenem-nonsusceptible Enterobacterales, including those that harbor KPCs. However, as mentioned above, it shows limited activities against isolates with MBL and OXA [77]. Addition of vaborbactam does not improve the activity of meropenem against A. baumannii or P. aeruginosa [56].
In treating cUTI (Table 1), meropenem/vaborbactam 2 g/2 g IV, TD showed similar clinical success when it was compared with piperacillin/tazobactam 4 g/0.5 g IV, TD in a noninferiority study: 98.4% vs. 94.0% [26]. It is worth mentioning that in this study, 5% of the isolates were resistant to piperacillin/tazobactam. An interesting comparison was made between meropenem/vaborbactam (n = 26) for the median duration of 12 days and ceftazidime/avibactam (n = 105) for a median duration of 11 days in treating mostly UTI and IAI patients due to CRE in an observational study (Table 4) [74]. In this study, around half of the patients were admitted to the ICU (median APACHE II score of 27) and the patients had other antibiotics (15.4% in meropenem/vaborbactam group vs. 51%). The clinical success (survival at 30 days, resolution of symptoms and signs, sterilization of blood cultures within 7 days of treatment, and absence of recurrent infections within 90 days) was higher in meropenem/vaborbactam than in ceftazidime/avibactam group: 69.2% vs. 61.9% (p = 0.49). In a study with various type of infections including cUTI, HAP/VAP, and bacteremia due to carbapenem-nonsusceptible pathogens, meropenem/vaborbactam showed higher clinical cure at day 28 in comparison with the best-available therapy of 65.6% vs. 33.3% [27]. It should be noted, however, that the results might have been affected by a higher proportion of patients with Charlson comorbidity index > 6 in the best-available therapy group.
Like all beta-lactam/ beta-lactamase inhibitor (BL/ BLI) presented in the present manuscript, imipenem (+cilastatin)/relebactam, further referred to as imipenem/relebactam, shows activity against Enterobacterales (including ESBL and AmpC isolates) and P. aeruginosa, but not against Acinetobacter spp. [78]. It restores activity against K. pneumoniae isolates that harbour KPCs [79], and partnering relebactam with imipenem decreased the MIC values of imipenem in P. aeruginosa isolates fourfold [78]. Despite the theoretical notion that relebactam shows activity against OXAs, a surveillance study fails to show that imipenem/relebactam has activities against OXA-positive isolates [80].

Discussion
The eight antibiotics reviewed here add to the armamentarium of antibiotics against carbapenem-resistant Gram-negative microorganisms. Almost all of these new antibiotics are approved for cUTI (all except eravacycline) and most of them for cIAI (eravacycline, ceftazidime/avibactam, ceftolozane/tazobactam, and imipenem/relebactam). Ceftazidime/avibactam was the first antibiotic approved for HAP/VAP by the FDA in 2018, and over the years, ceftolozane/tazobactam (2019), imipenem/relebactam (2020), and cefiderocol (2020) were also approved by the FDA for this indication.
Local epidemiology of multidrug-resistant microorganisms plays a role in choosing these new antibiotics. In our setting, where the prevalence of Enterobacterales resistant to third-generation cephalosporins of less than 10%, and to carbapenem of less than 1% (EARS-Net Surveillance Atlas of Infectious Diseases 2019), these new antibiotics are clearly not suitable as an empirical treatment. Targeted treatment can be performed only after susceptibility tests (ASTs) and eventual molecular testing have been performed. Table 5 can assist in making a choice of antibiotics in treating multidrug-resistant microorganisms. There are enough options for treating cUTI, cIAI, or HAP/VAP due to KPC CPE (all except ceftolozane/tazobactam can be used). The options become more limited to treat HAP/VAP due to OXA-48 CPE (only cefiderocol and ceftazidime/avibactam), and even more limited to treat HAP/VAP due to MBL CPE (only cefiderocol), but the combination cef-tazidime/avibactam with aztreonam (a monobactam antibiotic that is active against NDM) can perhaps be used, since aztreonam is active against MBL). For cIAI or HAP/VAP due to carbapenem-nonsusceptible P. aeruginosa caused by elevated efflux, derepressed AmpC, or porine loss, ceftazidime/avibactam or ceftolozane/tazobactam may be used. In case of HAP/VAP due to MDR A. baumannii, only cefiderocol is available, and perhaps eravacycline in cIAI.
In order to use these antibiotics appropriately, ASTs should be available in clinical microbiology laboratories. EUCAST has published a breakpoint for all new antibiotics except for plazomicin (and for temocillin). These breakpoints may be different than those from CLSI; for example, for plazomicin. The ASTs protocols may be also different than routine ASTs. ASTs for cefiderocol, for example, need an addition of iron in the media [84].
Surveillance studies show that many of these antibiotics are more active than other older antibiotics. For example, eravacycline is 2 to 4 times more potent (i.e., 2 to 4 times lower MIC) than tigecycline against Enterobacterales [30,31], and ceftazidime/avibactam is more active than ceftazidime alone against Enterobacterales [55]. However, in our opinion, this should not lead to the routine use of these antibiotics for the treatment of infections with Gram-negative microorganisms. We believe that these antibiotics should only be used when other options are exhausted. Inappropriate use of these new antibiotics may also lead to development of resistance.
More studies are needed that investigate the use of these antibiotics to treat infections that are caused by ESBL or carbapenem-nonsusceptible microorganisms. Many RCTs that lead to approval of these antibiotics as reviewed here often did not include these microorganisms. Exactly these microorganisms are an important reason why these new antibiotics were developed. It is also important to consider whether these antibiotics should be used alone or in combination with other antibiotics for treating an MDR bacterial infections. Combination therapy should also be considered to prevent resistance development, and for synergy when MICs are high and options are limited. The combination of these and older antibiotics should also be investigated to add to the possible therapeutic armamentarium. For example, the combination of ceftazidime/avibactam and aztreonam may be a therapeutic option against CPE due to NDM, as mentioned above. The possible combination of these new antibiotics and colistin is also interesting, since according to many surveillance studies (Tables S1 and S2), colistin is still susceptible to the carbapenem-nonsusceptible Enterobacterales, Acinetobacter spp., and P. aeruginosa.
Developing antibiotics active against Gram-negative microorganisms that harbor MBL remains a challenge. While several inhibitors for serine beta-lactamases (i.e., group A, C, and D) are available as reviewed above, not many MBL inhibitors are included. We can only speculate that this is perhaps caused by the structure of the MBL [85]. In vitro, MBL can be inhibited, for example, by aztreonam (a monobactam antibiotic) or boronic acid. There are several BL/BLI antibiotics with potential activity against MBL in clinical trials at this moment, such as aztreonam/avibactam (clinicaltrials.gov (accessed on 1 March 2021) identifier NCT03329092, in patients with cIAI and HAP/VAP; and cefepime/taniborbactam, NCT03840148, in patients with cUTI). Avibactam is partnered with aztreonam to prevent aztreonam from being hydrolyzed by non-MBL beta-lactamases. While this antibiotic seemed to be active against Enterobacterales with MBL genes (NDM, VIM, IMP), it was less active against P. aeruginosa with MBL genes [86]. The reason is perhaps the presence of other mechanisms than beta-lactamases.
In conclusion, we have reviewed eight novel antibiotics that are predominantly active against Gram-negative bacteria, and all have potential in the treatment of infections due to carbapenem-nonsusceptible microorganisms.
Author Contributions: Conceptualization, E.Y. and M.v.W.; methodology, all authors; formal analysis, all authors; investigation, all authors; resources, all authors; data curation, E.Y.; writing-original draft preparation, E.Y.; writing-review and editing, all authors. All authors have read and agreed to the published version of the manuscript.
Funding: There is no funding received in writing this systematic review.
Institutional Review Board Statement: This is a systematic review, no institutional review board approval was needed.
Informed Consent Statement: This is a systematic review, no patients are involved.
Data Availability Statement: This is a systematic review, all data are presented in (Supplementary) tables.

Conflicts of Interest:
The authors declare no conflict of interest.