Global Threat of Carbapenem-Resistant Gram-Negative Bacteria

1Infections caused by multidrug-resistant (MDR) and extensively drug-resistant (XDR) Gram-negative bacteria (GNB) have become major challenges for global health institutions because of the limited antibiotic options and high mortality rates. Among the clinically important XDR-GNB species that harbor plasmidic genes encoding a wide variety of carbapenemases, isolates of sequence type (ST) 258, ST11, and ST147 Enterobacterales (mainly producing Klebsiella pneumoniae carbapenemase [KPC], metallo-β-lactamase [MBL, especially New Delhi MBL {NDM}], and oxacillinase [OXA], etc.), ST111 and ST235 Pseudomonas aeruginosa (mainly producing NDM, imipenemase [IMP], Verona integron-encoded MBL [VIM], and OXA, etc.), and ST2, ST32, ST92, and ST368 Acinetobacter baumannii complex (mainly producing MBL, NDM, and OXA, etc.), they usually co-harbor other resistant β-lactamases (various extended-spectrum β-lactamases [ESBLs] and/or AmpC β-lactamases), thus constitute a worrisome global threat because of their high potential for transmission. Similar to other surveys, the study conducted by McConville et al. revealed that colonization with carbapenem-resistant (CR) Enterobacterales (CRE), which is a prerequisite for CRE infection, was an independent predictor of 90-day mortality (adjusted odds ratio [OR]: 2.3; 95% confidence interval [CI]: 1.0–5.3; P = 0.056). Furthermore, infections caused by CR- or carbapenemase-producing (CP)-GNB were shown to result in high mortality rates. This review focuses on global trends, resistance mechanisms, economic burdens, infection control policy and treatment options (including β-lactam combination agents with avibactam, zidebactam, enmetazobactam, relebactam, vaborbactam, nacubactam, durlobactam, and taniborbactam) of CR- and CP-GNB.

2The increase in carbapenem resistance in clinically important GNB gradually worsened after 2005, particularly in GNB isolates cultured from patients hospitalized in the intensive care unit (ICU). For example, from 2007 to 2008, Iran reported high imipenem resistance and MDR rates (41.8% and 56.3%, respectively) among P. aeruginosa isolates, especially those carrying class 1 integrons, or those cultured from patients undergoing surgery or hospitalized at the burn unit. Additionally, high MDR/CR rates (> 30%) among HAP-related P. aeruginosa isolates were observed in many member states of the European Union (EU) in the last decade. In contrast, the MDR rate of Taiwanese P. aeruginosa isolates was less than 18% before 2015. However, the 2016–2018 susceptibility survey of 1,127 P. aeruginosa isolates in Taiwan revealed a trend toward prominent escalation in the annual non-susceptible (NS) rates to anti-pseudomonal carbapenems (from 19.7% in 2016 to 27.5% in 2018 [ Figure 1 ]; 95% CI: 0.545–0.936; P = 0.016). When assessed using linear regression analysis, these results corresponded with the annual proportions of pneumonia-causing P. aeruginosa isolates (r = 0.980, P = 0.127). An outbreak of XDR-P. aeruginosa infections was reported in a tertiary care pediatric hospital in Italy between 2011 and 2012. The risk of colonization with CP-P. aeruginosa isolates harboring bla VIM was also reported to apparently increase with the length of hospital stay (especially > 30-day durations).

3After 2005, the isolates of MDR-A. baumannii complex harboring bla OXA-51 on the ISAba1 element and/or bla VIM/bla IMP on the class 1 integron displayed 45%–80% carbapenem resistance rates and were frequently associated with several clusters in the southern states of the EU after 2005. Moreover, the plasmids on Acinetobacter species and some clinical A. baumannii isolates that harbor genetic determinants encoding various carbapenem-hydrolyzing class D β-lactamases (bla OXA-23, bla OXA-58, bla OXA-58-like, bla OXA-72, etc.) have also been shown to confer high-level resistance to all carbapenem agents in China and Taiwan. A study conducted at a German hospital by Neidhöfer et al. observed that the bla OXA-23-encoding A. baumannii complex was more frequently introduced into the hospital by patients residing in the Arabian Peninsula than those of German ethnicity, raising the concern of ethnic factors affecting the infection due to CP-A. baumannii. Furthermore, nosocomial infections due to MDR-A. baumannii complex in pediatrics hospitalized at ICU were reported in Turkey. In contrast to a few Asian countries where an escalating antimicrobial resistance rate was observed among isolates of A. baumannii complex after 2011, isolates of A. baumannii complex accounted for solely 2.8% of the implicated organisms among episodes of hospital-acquired pneumonia (HAP) acquired in ICUs in US hospitals between 2015 and 2017.

4In 2012, a 4.6% colonization rate was reported in a rectal swab CRE survey (utilizing MacConkey agar plate for culture and Xpert MDRO cartridge testing) for patients residing in the nursing homes of Rhode Island, USA. In stark contrast, Asia and Africa are two leading continents with the highest global CR prevalence rates among Enterobacterales species. Notably, numerous outbreaks related to CRE involving a wide range of microbial species harboring diverse carbapenemase-encoding genes (bla KPC, bla NDM, bla VIM, bla OXA, etc.) have been reported in the literature. From March 2017 to September 2018, an epidemic spread of CRE, which was rapidly transmitted between hospitalized patients and associated with high mortality rates at 12 Vietnamese hospitals was reported by Tran et al.. Hospital-acquired infections and carbapenem therapy were determined to be two independent risk factors contributing to CRE colonization (ORs: 1.74 and 1.79, respectively). It is noteworthy that CRE infections were also reported among pediatrics with co-morbidities during the last decade in Los Angeles, USA. Additionally, endoscopy has surprisingly been found to be an under-recognized source of CRE transmission. Consequently, there is a strong need for the utilization of high-level disinfection for cleaning the endoscopic equipment.

5An outbreak owing to genetically related KPC-producing K. pneumoniae that persistently existed between March 2009 and February 2011 was reported at one acute care hospital in the USA. In addition, between November 2018 and October 2019, a delayed identification of a cluster involving 1,645 patients who were infected or colonized with NDM-producing CP-Enterobacterales (CPE) was reported in Tuscany, Italy. The majority (accounting for 90.9%) of the implicated CPE isolates that were mostly cultured from the intestinal tract of a total of 1,270 (77.2%) cases were ST147 and bla NDM-1-harboring CP-K. pneumoniae.

6Stenotrophomonas maltophilia, intrinsically resistant to all carbapenems, is an environmental MDR organism. Although no dominant S. maltophilia clone was identified globally, it has emerged as an important hospital-acquired pathogen globally (mainly causing primary bacteremia, pneumonia, catheter-associated infection, etc.) among hospitalized patients who receive broad-spectrum antibiotics (especially β-lactams) and/or immunosuppressive agents.

7Research has revealed that isolates of A. baumannii possess several virulence factors, including pili, outer membrane protein A, lipopolysaccharide capsule, and phospholipase. Resistance mechanisms of MDR or XDR-A. baumannii complex, listed as one of the critical priority pathogens by experts of the World Health Organization, usually include paucity of porins, constitutional expression of efflux pumps (AbeABC, AbeFGH, and AbeIJK), and expression of genes encoding various resistance β-lactamases (AmpC cephalosporinase, class B [NDM and VIM] and/or class D [OXA-23, OXA-58-like, etc.] β-lactamases). High carriage rates of diverse types of OXA enzymes (bla OXA-51 and bla OXA-23, followed by bla OXA-58, bla OXA-24/40-like, etc.)-encoding genes (77%–100%) have been reported in clinical CR-A. baumannii isolates globally. Furthermore, several acquired insertion sequences or transposons have been determined to promote the overexpression and spread of plasmid-associated bla OXA-58 genes in Acinetobacter species. In contrast to other regions, NDM-encoding genes have been frequently detected in CR-A. baumannii isolates in India. They frequently caused ventilator-associated pneumonia (VAP), catheter-associated, and bloodstream infections (BSIs) among debilitated patients hospitalized in the ICU, resulting in > 50% case-fatality rates.

8Infections caused by CR/MDR-P aeruginosa are also usually observed in immunocompromised patients (recipients of chemotherapy and/or high-dose corticosteroids), who suffer from pneumonia and require mechanical ventilation. Micek et al. observed that the prevalence rates of MDR phenotypes among P. aeruginosa isolates implicated in HAP in Europe between 2013 and 2014 ranged from 22.2% to 44.2%. Furthermore, compared to pneumonia due to non-CR/MDR P. aeruginosa isolates, patients with pneumonia caused by CR/MDR-P. aeruginosa had higher lengths of mechanical ventilation (13.1 days vs 17.0 days; P = 0.006). The case-fatality rates among patients with pneumonia caused by CR/MDR-P. aeruginosa have been reported to range from 44.7% to 64%.

9A study conducted by Kao et al., who analyzed the resistance mechanisms in 87 BSI-causing imipenem-resistant P. aeruginosa isolates collected in southern Taiwan between 2000 and 2010, revealed that carbapenemases (mainly VIM and OXA), active efflux pumps, and AmpC β-lactamase overproduction were found in 10.3%, 74.4%, and 51.3% of the P. aeruginosa isolates, respectively. The prevalence rate of metallo-β-lactamase (VIM, 6.4%) among Taiwanese imipenem-resistant P. aeruginosa isolates was similar to that in another Chinese study (8.5%). However, another molecular study that examined CR and XDR-P. aeruginosa isolates (n = 466) collected in Canada between 2007 and 2016 revealed that solely 4.3% (n = 20) harbored a carbapenemase-encoding gene, with Guiana extended-spectrum-5 producers (n = 7 [35%]) surpassing other β-lactamase producers. In stark contrast, the 2015–2017 CR-P. A study by Schäfer et al. at three medical centers in Germany revealed that 30.6% (19/62) of the samples harbored either bla VIM-1 (n = 2) or bla VIM-2 (n = 17) genes. The bla VIM carriage rate among CR-P. aeruginosa isolates was similar to that in the 2007–2009 survey conducted in Uganda (32%), and that in the 2015–2016 survey of Dubai hospitals in the United Arab Emirates (32%). Among worldwide MDR-P. aeruginosa isolates collected in the last decade, clones ST111, ST175, and ST235 have been identified to carry genomic islands. ST235 (producing KPC, and mainly identified in Europe and Asia) and ST111 (producing VIM-2, and mainly identified in all six continents, except Oceania) clones are the most worrisome class A/B carbapenemase-producers of P. aeruginosa, because they are virulent and associated with poor outcomes. IMP followed by NDM and VIM have become the two most prevalent class B carbapenemases in worldwide P. aeruginosa isolates.

10In the BSI-CRE (n=83) study conducted by Tamma et al. at the Johns Hopkins Hospital, USA, between March 2013 and April 2016, 37 (45%) isolates were CP-CRE, of which 92% harbored the bla KPC-2 gene. The CP prevalence rate was highly similar to that reported in a Taiwanese 2017 study. In contrast, approximately 90% (121/135) of CRE isolates collected from cancer patients between October 2016 and September 2017 in Egypt, where NDM and OXA-48-like enzymes are prevalent, harbored one or more of the carbapenemase-encoding genes, as revealed by polymerase chain reaction (PCR). Moreover, PCR analysis of rectal swabs from 590 patients hospitalized in Kuwait between April 2017 and March 2018 revealed that 38 (65.5%) out of the 58 patients (9.8%) with rectal CRE colonization harbored the bla OXA-181-like gene (combined with bla KPC-2 [n = 5], bla VIM-1 [n = 4], and bla NDM-5 [n = 3]). Previous studies have revealed that high treatment failure rates are associated with KPC-producing Enterobacterales infections, especially among immunocompromised hosts. In similarity to invasive infections (bacteremia and pneumonia) that are caused by MDR-P. aeruginosa or MDR-A. baumannii complex resulting in poorer outcomes than those caused by susceptible isolates, the all-cause mortality rates related to diverse CRE infections have been notably reported to range from 22% to 72%. Furthermore, a study conducted by Tamma et al. revealed that 32% (12/37) of patients infected with CP-CRE BSIs died within 14 days of admission, with an adjusted OR of 4.92 as compared to non-CP-CRE BSI.

11The major mechanisms of resistance to carbapenems in S. maltophilia isolates mainly include plasmid-encoding L1/L2 β-lactamases, and chromosomally encoded MDR efflux pumps. High rates of case-fatality (42.6%) and attributable mortality (37.5%) were notably reported in patients with S. maltophilia-causing pneumonia and septicemia, respectively. Additionally, a Taiwanese study regarding S. maltophilia bacteremia indicated that pediatric patients who had malignancy or failed to remove central venous catheters were at high risk of in-hospital mortality.

12Table 1 presents the prevalence rates of genes encoding carbapenemases among isolates of CRE, CR-Pseudomonas aeruginosa, and CR-A. baumannii in different surveys conducted in different countries. Additionally, Table 2 illustrates the enzymatic and non-enzymatic mechanisms conferring carbapenem resistance among isolates of CRE, CR-P. aeruginosa, and CR-A. baumannii complex.

13The huge economic burden caused by MDR-GNB infections, most of which were limited to CRE, has been intensively analyzed. Bartsch et al. investigated the median cost of CRE infections with an incidence of 2.93 per 100,000 persons in the USA in 2015. They estimated that it cost hospitals 275 million US dollars, third-party payers 147 million US dollars, and society 553 million US dollars. In addition, costs increased proportionally with the incidence of CRE, with increases of 2.0-fold, 3.4-fold, and 5.1-fold for incidence rates of 6, 10, and 15 per 100,000 persons, respectively. In-hospital outbreaks of CR- or CP-GNB further exacerbate the already high-cost burden. An outbreak of CPE (caused by mainly NDM-producing K. pneumoniae), which occurred at a London hospital from July 2014 to October 2015, cost €1.1 million, with €54,000 spent on antibiotics for 18 patients who needed treatment, €94,000 on laboratory screening, €153,000 on Estates renovations, and €822,000 as opportunity cost (staff time, bed closure, and elective surgical missed revenue). Furthermore, an investigation conducted by Yang et al. at a tertiary teaching hospital in China from 2011 to 2016 revealed that the direct economic burden and disability-adjusted life-year loss caused by BSI due to CR-P. aeruginosa was ≥ 3-fold higher than that of carbapenem-susceptible P. aeruginosa BSI. According to Semin-Pelletier et al., frequent readmissions of patients who had previous infections due to OXA-48-producing K. pneumoniae isolates, the large number of transfers between wards, and a delay in the implementation of successive cohort units greatly contributed to the incomplete success of containing the spread of CPE. A few physicians have adopted aggressive surveys of samples from patients to interrupt the transmission of CRE/CPE. For example, a survey at a tertiary teaching hospital in Malaysia revealed that 5.74% of Enterobacterales isolates cultured from various clinical samples (of which rectal swab screening accounted for 49.3%) were CRE, most of which were K. pneumoniae isolates harboring bla NDM-1. In the multicenter survey conducted by Jimenez et al. in Miami, Florida from 2012 to 2016, active surveillance testing revealed that the overall CRE prevalence was 0.077 cases per 100 patient-admissions, while the incidence density was 1.46 cases per 10,000 patient-days. It is also noteworthy that rates of CRE (dominated by K. pneumoniae, Enterobacter cloacae, and Escherichia coli) steadily increased during the first 3 years of the study period, and declined after implementation of infection control strategies (contact precautions, environmental disinfection, etc.).

14Owing to the high burden of medical costs and high case-fatality rates seen for inpatients with CPE infections, attempts have been made to improve the cost-effectiveness of screening for CPE among all hospital inpatients. Among the different culture methods, the ChromID CARBA method (bioMérieux, Marcy l’Etoile, France) has been to perform rapidly and best in the detection of CPE (> 100 colony forming units/spot) from rectal swabs, in terms of sensitivity (92.4%) and specificity (96.9%). Despite a lower reagent cost, culture-based methods are, in fact, less sensitive than molecular methods for the detection of infection or colonization of CPE. Moreover, they are labor-intensive and slower to yield results. Lapointe-Shaw et al. observed that screening for CPE might be cost-effective as compared to not screening, if the prevalence of CPE was above 0.3% among the isolates of Enterobacterales under survey.

15Apart from hospital acquired CRE, the presence of CRE in the community setting is also a potentially worrisome threat to the public health of ambulatory patients. A meta-analysis conducted by Kelly et al. revealed that 5.6%–10.8% of rates of colonized CRE isolates belonged to the community-associated/community-onset category (CA/CO; defined as identification of asymptomatic CRE colonization at the time of admission) in the USA-based studies, while percentages ranging from 0.04% to 29.5% of colonized CRE categorized as CA/CO in origin were reported globally. Compared to healthcare-acquired CRE, a Taiwanese survey conducted by Tang et al. revealed that CA-CRE was more likely to occur in elderly female patients and result in urinary tract infection (UTI). In a study conducted in China, Hu et al. also found that CO-CRE isolates (n = 28, accounting for 43.8% of all infection-causing CRE isolates enrolled) were mainly cultured from urine samples (75%). Of the CO-CRE isolates, 8 (28.6%) were clonally unrelated E. coli isolates, all of which were NDM producers that were less resistant to aztreonam, ciprofloxacin, levofloxacin, and chloramphenicol. pneumoniae.

16The emergence of CRE at an LTCF (an institution that provides long-term rehabilitation and skilled nursing care) is also a major healthcare issue. The prevalence and incidence rate of colonized CRE in LTCFs in the USA ranged from 3% to 30.4% and 1.07 to 6.83 cases per 10,000 patient-days, respectively. In contrast, the CRE prevalence rates differed widely among the states of the EU. A significant association (ranging from 12% to 15.5%) between CRE colonization in nursing home residents with hospital admissions was observed in Spain (predominantly K. pneumoniae harboring bla OXA-48-like gene) and Israel (mainly KPC-producing K. pneumoniae). Accordingly, nursing home residents have become CRE reservoirs that should not be ignored. Multiple risk factors, including fecal incontinence (OR: 5.78), an immunocompromised condition (OR: 3.92), comorbidities (Charlson score > 3; OR: 4.85), use of gastrointestinal devices (OR: 19.7), intravascular indwelling devices or urinary catheters (OR: 5.21), and mechanical ventilation (OR: 3.56), sharing a room with a known CRE carrier, and prior antibiotic exposure (OR: 3.89), etc., has significantly associated with increased vulnerability to CRE colonization and/or infections in LTCF residents. It is noteworthy that patients from LTCFs who were colonized or infected with CRE had notably poor clinical outcomes, with a mortality rate of up to 75% among patients with CRE infection. In contrast, the prevalence rates of CRE colonization among residents of LTCFs ranged from 13% to 22.7% in Asia. As stated in a survey on the distribution of carbapenemase-encoding genes among LTCF residents, ST258 CR-K. pneumoniae harboring bla KPC-2 or bla KPC-3 were the predominant clones present in residents of LTCFs in the USA, whereas other carbapenemases (including NDM, VIM, IMP, and OXA-48-like) were uncommon among CRE isolates for them. In Japan, various types of IMP (predominantly IMP-42, IMP-11, followed by IMP-6) carbapenemases have been detected among overall CRE isolates in LTCF residents in Japan. Data from the CRE survey of LTCF residents in Spain revealed that OXA-48-like enzymes, followed by VIM-1, IMP, and KPC are the main carbapenemases that can be found there.

17Although P. aeruginosa is not a common colonized GNB cultured from LTCF residents, as reported by the survey conducted by O’Fallon et al. in the USA, and Italian surveys of Giufrè et al. and March et al., it has been shown to be easily transmitted through numerous routes, including patient-to-patient contact and environmental contamination. Thus, it is plausible that the CP-P. aeruginosa (harboring bla VIM) isolates were cultured from LTCF residents who have history of frequent hospitalizations. A study conducted by Raman et al. revealed that MDR/CR-P. aeruginosa can be identified from patients who were transferred from chronic care facilities and exposed to piperacillin-tazobactam (adjusted OR: 2.64) or carbapenem (adjusted OR: 4.36).

18Table 3 illustrates the summary of CRE, CR-P. aeruginosa, and CR-A. baumannii cultured from the community setting and LTCF.

19Early identification of in-hospital patients with risk factors for CRE acquisition (72.7% of which were K. pneumoniae), such as co-colonization with MDR-A. baumannii complex (adjusted OR: 15.6) or ESBL-producing GNB (adjusted OR: 4.7; might be linked to recent carbapenem use), exposure to glycopeptide antibiotics (adjusted OR: 3.6), and admissions within one year (adjusted OR: 3.9), is beneficial for avoiding potential spread. In addition, excessive consumption of carbapenems (especially group 2 agents) is considered to be an important predisposing factor that contributes to the worsening rates of infections caused by CR-P. aeruginosa (estimated OR: 2.87 - 40.96), CRE, and CR-A. baumannii complex. Consequently, appropriate antibiotic control policies (especially strict implementation of antibiotic stewardship) in hospitals, education of primary care staff to prevent the dissemination of high-risk hospital and LTCF microorganisms, and contacting isolation cohorts are of paramount importance in containing the spread of CR- or MDR/XDR-GNB strains and decreasing case-fatality rates. Similarly, there is a need to implement periodical surveillance testing to determine in-time targeted interventions (e.g., isolating or cohorting CRE carriers and nursing staff) are needed to effectively lessen the CRE outbreaks at LTCFs.

20The distinguishment of CR-GNB that are due to the non-carbapenemase-mediated mechanisms from CP-GNB is important, because CP-GNB are prone to disseminate between patients more readily than non-CP-GNB. Among the phenotypic CRE diagnostic assays commonly used in clinical microbiology laboratories, the modified Hodge test was applied earliest to detect potential KPC producers of CRE. However, it showed poor sensitivity in detection of the NDM and OXA-48-like CPE, and false-positive results are seen in CRE owing to the porin alteration combined with hyperproduction of ESBL and/or AmpC β-lactamase. The Carba NP test and its variants are suitable for detection of various carbapenemases in CRE and CR-P. aeruginosa. In spite of requiring the acquisition of dedicated reagents and being interpreted subjectively, the Carba NP test is a convenient biochemical test that provides the results within 15-30 minutes and thus has been applied in the clinical microbiological laboratories worldwide. Additionally, the CarbAcineto NP test was shown to perform well in detecting carbapenemases produced by CR-A. baumannii.

21Utilization of the modified carbapenem inactivation method (mCIM) in combination with the EDTA-modified CIM (eCIM) test could reliably differentiate MBL-producing CRE strains (those displaying a negative result on only the eCIM test) from serine-class carbapenemase producers (showing positive results for both tests). Nevertheless, it takes approximately 6-12 h to obtain the results and is interpreted subjectively as well. It is also noteworthy that the mCIM/eCIM test performed less well in detection of VIM-producing P. aeruginosa and OXA-producing A. baumannii isolates than CPE.

22The other carbapenemase detection tests include lateral flow immunoassays (antibody-based rapid diagnostic, easy-to-use methods, such as NG-Test Carba 5 [Hardy Diagnostics, CA, USA], and Resist-3 O.K.N. that detects OXA-48-like, KPC and NDM but not VIM and IMP-like MBLs [Coris BioConcept, Gembloux, Belgium]), targeted carbapenemase assays (inhibitor [phenylboronic acid for KPC and EDTA for MBL]-based, easy-to-use methods) for diverse carbapenemases, and matrix-assisted laser desorption–ionization time of flight mass spectrometry (hydrolysis approach).

23In a study by Garonzik et al., optimum dosages of polymyxin B and colistin (polymyxin E), two old antibiotics of revival in this century, have been proposed to maximize their efficacy against infections related to MDR-GNB, with a minimum inhibitory concentration (MIC) of >1 mg/L for colistin. Clinical, pharmacokinetic (PK) and pharmacodynamic (PD) data of polymyxin B/E for several GNB species have been shown to have limited clinical efficacy, even if an intermediate result (i.e., MIC ≤ 2 mg/L) was achieved. Moreover, monotherapy with intravenous colistin against CR-K. pneumoniae or E. coli bacteremia was also not suggested, because of its association with a high (57.1%) mortality rate (hazard ratio: 5.57; 95% CI, 2.13 - 14.61; P < 0.001) as compared to other comparative antibiotics. Despite controversies, during the interval of an aerosolized colistimethate sodium (CMS) dosing (2 million units [MU]) using jet or ultrasonic nebulizer, a high pulmonary area under the concentration-time curve of colistin (ranging 18.9 – 73.1 μg•h/mL), as well as a high maximum pulmonary colistin concentration (6.00 ± 3.45 μg/mL) were achieved in humans, with no increase in nephrotoxicity. Thus, a high-dose regimen of aerosolized CMS monotherapy (4 MU administered every 8 h) was applied for the treatment of VAP caused by a few notable MDR-GNB (P. aeruginosa and A. baumannii complex predominantly). In addition, impaired uptake of fosfomycin (related to glpT, uhpT, and uhpA) and the existence of fosfomycin-modified genes (e.g., fosA3, which encodes fosfomycin-modifying enzymes) in transposon elements and conjugative plasmids confer Enterobacterales species, especially E. coli, exhibiting resistance to fosfomycin. High prevalence rates of fosfomycin resistance genes in CRE have been determined in several parts of East Asia (e.g., China and Japan), and resistance to fosfomycin has also emerged in the USA and the states of the EU. Therefore, the use of systemic colistin and fosfomycin is recommended as an adjunctive treatment for CRE or CR-P. aeruginosa infections.

24Before the launch of ceftazidime-avibactam for the clinical treatment of CPE infection in 2015, Gutiérrez-Gutiérrez et al. concluded that therapy with a combination regimen of antibiotics (including at least one in vitro active drug against the implicated BSI isolate and started in the first five days after infection) was associated with decreased 30-day all-cause mortality rates among patients infected with BSI owing to CPE (mainly KPC-producing K. pneumoniae) and high INCREMENT-CPE mortality scores (high Pitt bacteremic scores in patients who had ≥ 2 points of Charlson comorbidity score and BSI not originating from UTI or biliary tract infection). Despite its bacteriostatic nature and relatively low serum concentration under standard-dose administration (100 mg loading followed by 50 mg every 12 h), the study by Lin et al. on CRE bacteremia revealed that tigecycline was a good choice if the antibiotic had an MIC ≤ 0.5 mg/L against the isolates, based on the 90% probability of target attainment and 82% probability of cumulative fraction of response. Nevertheless, heterogeneity in adequate regimens of antibiotics against CPE has been suggested for both in vitro efficacy and clinical treatment effects. For example, Tang et al. determined that in vitro treatment with 1× MIC using combinations of amikacin or gentamicin, and tigecycline or doxycycline for 24 h resulted in bactericidal activity of 84%–100% in 13 KPC-K. pneumoniae isolates; in addition, the combination of doxycycline plus gentamicin or amikacin was synergistic for all the tested KPC-K. pneumoniae isolates. A review of CPE therapy by Tzouvelekis et al. and an investigation conducted by Daikos et al. concluded that the lowest mortality rate (18.8%) was observed in patients treated with combinations of various antibiotics. Compared to other antibiotic regimens, Tumbarello et al. further demonstrated that post-antibiogram therapy with a combination of tigecycline, colistin, and meropenem (at a dosage of 2 g every 8 h intravenously) was associated with lower mortality rates (OR, 0.11; 95% CI: 0.02 - 0.69; and P = 0.01) for patients with BSI due to KPC-producing K. pneumoniae, especially those with an MIC ≤ 16 mg/L for meropenem. In addition, Lowman et al. observed that carbapenem-based therapy improved survival in 20 critically ill patients infected with CPE harboring bla OXA-48-like gene, although the severity of the underlying illness significantly impacted their outcomes as well. The combination of doripenem (8 mg/L) and colistin (1 mg/L) was observed to have in vitro bactericidal efficacy against 75% (9/12) of KPC-producing K. pneumoniae isolates. As compared to the clinical treatment using a single antibiotic (colistin, tigecycline, or gentamicin), De Pascale et al. determined that a double carbapenem regimen (comprising meropenem at a dosage of 2 g every 8 h intravenously, and ertapenem at a dosage of either 2 g once daily or 1 g every 12 h intravenously) resulted in a significantly lower rate of 28-day mortality (47.9% vs 29.2%; P = 0.04), higher rates of clinical cure (31.3% vs 65%; P = 0.03) and microbiological eradication (25.9% vs 57.9%; P = 0.04) for patients with severe infections due to CR-K. pneumoniae (> 90% of which harbor bla KPC). A similar in vivo effect was also reported by Builk et al. in a murine thigh model with KPC-producing K. pneumoniae infection, using simulated mega-dose doripenem (2 g every 8 h) combined with ertapenem (1 g once daily).

25Ceftazidime combined with avibactam, a diazabicyclooctane (DBO) β-lactamase inhibitor, is effective against most Ambler class A/C/D enzymes, providing a new option for the treatment of CRE/CPE infections; however, mutations in Asp179Tyr, Val240Gly, Ala240Val, Ala177Glu, Thr243Met substitutions, and 165-166Glu-Leu insertion, have been shown to compromise the in vitro efficacy of ceftazidime-avibactam while sparing several other novel antibiotics against CRE isolates. In addition, recent investigations have revealed that a significant proportion of CR-E. cloacae complex exhibited significantly higher NS rates to ceftazidime-avibactam. In the small case series covering diverse CRE infections (n = 37, with five-sixths being CR-K. pneumoniae), Shields et al. observed that ceftazidime-avibactam therapy achieved a success rate of 70% and 50%, respectively, for bacteremia and pneumonia, and 100% for acute pyelonephritis. Furthermore, resistance to ceftazidime-avibactam among CRE developed notably following its therapy for 10–19 days (median, 15 days). The overall 30-day survival rate of the series by Shields et al. was 76%, similar to that of another survey on CR-K. pneumoniae therapy using antibiotic combination regimens (72.5%). In a small cohort (n = 38) survey conducted by Alraddadi et al. to explore the outcomes of patients with various infections caused by CRE (largely OXA-48-producing K. pneumoniae isolates), ceftazidime-avibactam therapy provided a better clinical remission rate than comparative antibiotics, including colistin, tigecycline, and meropenem (80% vs 53.6%; P = 0.14). Nevertheless, no difference was observed in the 30-day all-cause mortality rates between the two groups (50% vs 57.1%; P = 0.71). Furthermore, in a head-to-head study conducted by Ackley et al. who compared the clinical efficacy and development of resistance between ceftazidime-avibactam and meropenem-vaborbactam (a β-lactamase inhibitor that comprises structurally boronic acid [BA]) against CRE, a 2.9% rate of resistance developed after ceftazidime-avibactam monotherapy, while similar rates of clinical success and 90-day mortality were observed between the two groups.

26Among other potential novel antibiotics, ceftolozane-tazobactam was inactive against isolates of CPE isolates, several AmpC-producing Enterobacterales, and most ESBL-producing K. pneumoniae. Cefepime in combination with enmetazobactam (formerly AAI101, a β-lactamase inhibitor that structurally comprises penicillanic acid sulfone), exhibited lower MICs against isolates of ESBL-producing Enterobacterales and KPC-producing E. coli than CP-K. pneumoniae. In contrast, cefepime combined with zidebactam (formerly WCK 5222, a structural DBO β-lactamase inhibitor) displayed superior in vitro activity against the Ambler class B enzyme producers of Enterobacterales than ceftazidime-avibactam and imipenem in combination with relebactam (also a structural DBO β-lactamase inhibitor), meropenem in combination with vaborbactam, cefepime-enmetazobactam, durlobactam (formerly ETX 2514, a structural DBO β-lactamase inhibitor, penicillin-binding protein [PBP] 2 inhibitor, and in vitro activity against P. aeruginosa AmpC β-lactamase), and plazomycin. Furthermore, the anti-CPE spectra of aztreonam-avibactam and cefepime combined with taniborbactam [also a structural BA β-lactamase inhibitor against all metallo-β-lactamases (MBL), except for the IMP types], are considerably similar to those of cefepime-zidebactam and cefiderocol. Nacubactam (formerly OP0595 or RG6080, a structural DBO β-lactamase and PBP2 inhibitor) was demonstrated to be active in vitro against diverse β-lactamase producers, including a few NDM or VIM-producing Enterobacterales species (mainly E. coli and Enterobacter spp.). When nacubactam was combined with meropenem, it was shown to be active in vitro against the CRE isolates of any β-lactamase producer. Phase 3 clinical investigation is currently being conducted to test the efficacy of nacubactam in combination with meropenem against CRE isolates harboring the bla KPC gene.

27As stated in several studies, septicemia or pneumonia caused by MDR-P. aeruginosa isolates resulted in poor patient outcomes. Recommendations for the treatment of MDR/XDR-GNB infections have been published based on the susceptibility and PK/PD profiles of conventional and novel antibiotics. Antibiotic combination therapy is likely to select mutants displaying a broader resistance phenotype (e.g., mutational inactivation of the repressor gene mexR that regulates the multidrug efflux operon mexAB–oprM for P. aeruginosa) than before. Nevertheless, a few combination regimens of dual antibiotics exhibited synergistic or additive effects in vitro (determined using fractional inhibitory concentration index) against CR- or MDR/XDR-P. aeruginosa isolates. In addition, the combination of gentamicin and ciprofloxacin has good potential for inhibiting biofilm formation synthesized from P. aeruginosa (71.4%). De-escalation of antibiotics into a single agent is strongly recommended when susceptibility of the implicated GNB (including MDR/XDR-P. aeruginosa) isolate is known, and there is a significant improvement in the patient’s condition. Among the novel antibiotics that have been launched to combat MDR-GNB isolates, ceftolozane-tazobactam has excellent in vitro activity against global CR- and XDR-P. aeruginosa strains, including those with overexpression of efflux pumps but no carbapenemase production. Despite not being validated by several randomized clinical studies, ceftazidime-avibactam, cefepime-zidebactam, cefiderocol, imipenem-relebactam (showing excellent in vitro activity relative to imipenem solely against OprD-losing P. aeruginosa isolates with Pseudomonas-derived cephalosporinase hyper-production), meropenem-vaborbactam, meropenem-nacubactam, and cefepime-taniborbactam are considered as promising alternatives against infections caused by CR- or MDR-P. aeruginosa isolates. Although aztreonam-avibactam has excellent in vitro potential to inhibit CPE (including producers of MBL; MIC90 ≤ 8 mg/L), relatively high MIC90 levels of this novel antibiotic were observed against the global overall and MBL-positive P. aeruginosa isolates tested (32 mg/L and 32 mg/L, respectively).

28A 3-h intravenous infusion of 2 g meropenem every 8 h produced a high percentage (72.89 ± 22.40%) of time above the serum concentration of 8 mg/L after the third dose. This regimen provides therapeutic benefits for the treatment of VAP caused by CR-A. baumannii isolates. Unfortunately, a high in vitro resistance to meropenem (MIC > 64 mg/L) was exhibited by most XDR/CR-A. baumannii and CRE isolates, whose susceptibility was not restored by sulbactam addition. Because of the extremely high likelihood of resistance to the majority of antibiotics, various regimens of antibiotic combinations have been proposed for the treatment of XDR-A. baumannii in the PubMed database. For instance, Jean et al. demonstrated that a prolonged intravenous infusion (> 3 h) of imipenem combined with tigecycline provided a significantly better survival rate for patients with XDR-A. baumannii bacteremic VAP than imipenem plus sulbactam (64.3% vs 14.3%). Time-kill kinetic analysis proved that the former regimen inhibited the in vitro growth of XDR-A. baumannii. In addition, treatment with a colistin-carbapenem (doripenem) regimen also significantly improved the 28-day survival rates among solid-organ transplant recipients with various XDR-A. baumannii infections (OR: 7.88; 95% CI: 1.60 – 38.76; P = 0.01). In partial similarity to the meta-analysis conducted by Kengkla et al., who substituted tigecycline for carbapenem, Pongpech et al. suggested that the triple combination of meropenem, colistin, and sulbactam was a good regimen in vitro against carbapenem-NS A. baumannii complex that did not harbor genes encoding class B carbapenemase. Furthermore, to effectively treat severe VAP or BSI caused by MDR or XDR-A. baumannii complex, Piperaki et al. suggested a combination regimen comprising two active in vitro agents. Antibiotic options include high-dose ampicillin-sulbactam, high-dose tigecycline (or minocycline), and polymyxin or aminoglycoside. If two in vitro active agents are not available, in vitro synergy studies are valuable in choosing the most appropriate targeted combination scheme. Novel antibiotics, including ceftolozane-tazobactam, imipenem-relebactam, meropenem-vaborbactam, ceftazidime-avibactam, aztreonam-avibactam, cefepime-enmetazobactam, and cefepime-zidebactam, have significantly poor in vitro activity against XDR-A. baumannii complex, while exhibiting excellent in vitro efficacy against CRE or CR-P. aeruginosa. In contrast, cefiderocol, a novel siderophore modified from ceftazidime, exhibited excellent in vitro activity against several carbapenem-NS GNB species, including A. baumannii complex; however, compared to the best available therapy, higher all-cause mortality rates were observed during the early hospitalization course when cefiderocol was administered for the treatment of patients with nosocomial pneumonia and BSI caused by CR-Acinetobacter species. The combination of sulbactam with durlobactam has been shown to effectively inhibit CR-Acinetobacter species. In addition to cefiderocol and eravacycline (a novel fluorocycline agent of the tetracycline family), several novel drugs and new combination regimens, including TP-6076 (a fully synthetic fluorocycline antibiotic under development), WCK 4234 (a structural DBO β-lactamase inhibitor) with meropenem, LN-1-255 (a β-lactamase inhibitor modified from the penicillanic acid sulfone), taniborbactam, SPR741 (a cationic peptide derived from polymyxin B as an antibiotic adjuvant), and phage therapy (mostly applied in animal models for the treatment of infections caused by A. baumannii complex and P. aeruginosa), etc., have been investigated to evaluate their feasibility for clinical use and potential efficacy against troublesome CR/XDR-A. baumannii isolates.

29Table 4 compares the spectra of novel antibiotics against carbapenemase-producing GNB. Additionally, Table 5 presents the spectra of important carbapenemase inhibitors against various plasmid-mediated carbapenemases in GNB.

30In the present MDR/XDR-GNB era, we have encountered an antibiotic pipeline scenario. Although a few novel antibiotics have been effective in vitro against several CR-GNB, their clinical efficacy requires further validation. The judicious prescription of these valuable antibiotics, strict implementation of antibiotic stewardship policy, adequate disinfection of equipment and environment of hospital and LTCF settings, in combination with in-time screening to initiate necessary cohort isolation, are a few measures that must be vigorously undertaken to lessen the rapidly worsening trends of resistance in clinically important GNBs.

31S-SJ and P-RH collected and analyzed the data. S-SJ and DH participated in writing the manuscript. S-SJ, DH, and P-RH read and approved the final version of the manuscript. All authors contributed to the manuscript and approved the submitted version.

32The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

33All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.