Section 2.4: Novel Antibiotics for Carbapenem-Resistant Gram-Negative Infections: Clinical Applications and Prospects

I. Epidemiology of Carbapenem-Resistant Gram-Negative Infections

The latest trends from the National Bacterial Antimicrobial Resistance Surveillance Report indicate that the prevalence of carbapenem-resistant Gram-negative bacteria (CRGNB) in China exhibits distinct regional and lineage characteristics. The detection rate of carbapenem-resistant Klebsiella pneumoniae (CRKP) has climbed to 25%–30%, while the resistance rate of carbapenem-resistant Acinetobacter baumannii (CRAB) has remained above 70% for a long time. In the ICU, infections caused by CRGNB carry a 28-day mortality rate of 40%–60%. Timely implementation of precise antimicrobial therapy for CRGNB infections is core to improving prognosis and reducing mortality risk.

II. Complexity and Diversity of CRGNB Resistance Mechanisms

The resistance phenotype of CRGNB is not driven by a single factor but by a complex interactive network of multiple molecular mechanisms. Core mechanisms include enzymatic inactivation, permeability barrier restriction, active efflux system activation, target site modification, biofilm formation, metabolic compensation, and horizontal gene transfer mediated by mobile genetic elements. Although these mechanisms intersect, different pathogens exhibit clear preferences when establishing resistance phenotypes.

(I) Core Resistance Mechanisms of Major Gram-Negative Bacteria

• Carbapenem-resistant Enterobacterales (CRE): In China, the primary resistance determinant for CRKP is Class A enzymes [specifically Klebsiella pneumoniae carbapenemase (KPC)-2]. This enzyme possesses extremely broad substrate specificity, efficiently inactivating almost all β-lactam antibiotics. Additionally, CRE frequently co-downregulates membrane porins like OmpK35/36, restricting intracellular drug flux and producing additive resistance effects.

• Carbapenem-resistant Pseudomonas aeruginosa (CRPA): Its resistance architecture is characterized by non-enzymatic physiological barrier mechanisms. Loss of the specific membrane porin OprD is the key factor leading to decreased carbapenem permeability. Furthermore, CRPA utilizes naturally carried active efflux systems like MexAB-OprM, combined with highly developed biofilm invasive growth, effectively blocking intracellular drug penetration and accumulation. This deep coupling of multiple mechanisms constitutes the molecular pathological basis of difficult-to-treat resistant P. aeruginosa (DTR-PA).

• CRAB: Its resistance foundation lies in the high expression of Class D OXA enzymes, including activation of the endogenous blaOXA-51 gene and high expression of the acquired blaOXA-23 gene. Unlike CRE, CRAB relies more on a “high efflux-low permeability” synergistic mode, mediated by high-level expression of efflux pumps like AdeABC combined with low outer membrane permeability, jointly driving its highly resistant phenotype.

(II) Dynamic Evolution Characteristics of Resistance Mechanisms
As antimicrobial pressure evolves, CRGNB resistance mechanisms exhibit new dynamic features, increasing clinical treatment difficulty.

• Regulator Mutations Inducing Mechanism Stacking: In DTR-PA, inactivating mutations in regulator genes like mexR or nalC are frequently detected. This not only leads to sustained overexpression of the MexAB-OprM efflux pump but is often accompanied by downstream metabolic pathway alterations, enabling strains to maintain high-level resistance while enhancing survival competitiveness in the pulmonary microenvironment.

• Dynamic Succession of Enzyme Lineages and “Dual-Enzyme” Phenomenon: Epidemiological surveillance shows a shifting enzyme lineage in domestic CRKP. Although KPC-2 remains dominant, the proportion of Class B metallo-β-lactamases (MBLs) represented by NDM-1 and NDM-5 is rapidly rising. More clinically challenging is the increasing detection rate of “dual-enzyme” strains simultaneously carrying two carbapenemase genes (e.g., blaNDM-1 combined with blaOXA-48 or blaNDM-1 combined with blaKPC-2). Synergistic expression of these resistance genes significantly broadens the substrate hydrolysis spectrum, mediating high-level resistance to almost all β-lactam drugs, including novel enzyme inhibitor combinations.

• Promoter Evolution and Gene Copy Number Amplification: Clinical studies reveal that CRAB is undergoing promoter sequence evolution downstream of resistance genes like blaOXA-23, or significantly enhancing enzyme production levels via plasmid copy number amplification. This phenotype evolution driven by gene dosage effects allows strains to maintain extremely high minimum inhibitory concentrations (MICs) even against high-concentration enzyme inhibitor combinations, rendering conventional dosing strategies ineffective.

III. Novel Antimicrobial Treatment Strategies Based on Resistance Mechanisms

From 2024 to 2025, with the successive domestic approval of novel β-lactam antibiotics and β-lactam/β-lactamase inhibitor (BL/BLI) combinations, the treatment paradigm for CRGNB has shifted significantly from empirical combination therapy to targeted therapy centered on molecular resistance mechanisms.

(I) Treatment Strategies for KPC-Producing CRE
Accurate identification of carbapenemase typing in clinical isolates is the core prerequisite for formulating CRE treatment strategies, as the antibacterial activity of novel BL/BLIs exhibits significant enzyme specificity. Both IDSA guidelines and domestic consensus recommend nucleic acid amplification or antigen testing for enzymatic identification of CRE isolates.

For KPC-producing CRE infections, the first-line options are monotherapy with ceftazidime-avibactam (CAZ-AVI) or imipenem-cilastatin-relebactam (IMI/REL). IMI/REL consists of a carbapenem combined with a novel diazabicyclooctane (DBO) class inhibitor. Relebactam acts as a potent inhibitor, covalently and reversibly binding to and inactivating Class A and C enzymes. Additionally, relebactam can antagonize specific active efflux pumps and reverse resistance caused by chromosomal-mediated AmpC overexpression. Through this multi-target synergistic mechanism, it enhances inhibition of cell wall synthesis. Imipenem and relebactam demonstrate high pharmacokinetic synchrony in humans, with elimination half-lives of 1.2–1.5 h. This combination exhibits broad tissue distribution, with excellent penetration into pulmonary epithelial lining fluid, rapidly achieving and maintaining therapeutic levels above pathogen MICs throughout the dosing interval.

The Phase III RESTORE-IMI trial showed that in treating CRGNB infections, the 28-day clinical efficacy of the IMI/REL group was significantly superior to the colistin plus imipenem group (71% vs. 40%); simultaneously, the 28-day mortality in the IMI/REL group was lower (10% vs. 30%), and the incidence of nephrotoxicity was significantly reduced (10% vs. 56%, P=0.002). Furthermore, a 2025 Italian multicenter severe cohort study showed that in the ICU setting, the real-world clinical cure rate of IMI/REL for KPC-producing K. pneumoniaeinfections exceeded 70%.

(II) Treatment Strategies for MBL-Producing CRE
MBLs possess broad-spectrum β-lactamase activity, hydrolyzing most antibacterial drugs including carbapenems. Aztreonam is currently the only clinically available β-lactam drug with intrinsic resistance to MBL hydrolysis, but it is highly susceptible to degradation by coexisting Class A, C, or D serine enzymes. Avibactam, a DBO inhibitor, covalently blocks these serine enzymes, providing a biochemical shield for aztreonam to exert bactericidal activity against penicillin-binding protein 3 (PBP3). This combination strategy effectively dismantles the complex resistance mechanisms mediated by MBL-producing strains co-carrying multiple serine enzymes.

Pharmacokinetically, aztreonam and avibactam demonstrate excellent penetration into alveolar epithelial lining fluid and peritoneal tissues. Both components are primarily renally excreted, with highly synchronized elimination half-lives (~2 h). In clinical practice, the empirical combination of ceftazidime-avibactam (CAZ-AVI) with aztreonam requires synchronous administration and extended infusion to ensure synergistic effects. In contrast, the fixed-dose combination of aztreonam-avibactam (ATM-AVI) ensures both components are delivered in a standardized ratio, streamlining administration while maintaining optimal pharmacokinetic/pharmacodynamic (PK/PD) synergy.

Population pharmacokinetic simulations demonstrate that under standard dosing regimens, aztreonam-avibactam (ATM-AVI) achieves an excellent probability of target attainment (PTA) in patients with complicated intra-abdominal infections and severe pneumonia. Even against metallo-β-lactamase (MBL)-producing pathogens with high MIC values, adequate in vivo drug exposure is consistently maintained. This synchronized elimination profile and reliable tissue penetration enable ATM-AVI to sustain therapeutic concentrations at the site of infection, maximizing bactericidal activity against resistant Gram-negative pathogens.

The clinical value of ATM-AVI has been validated across multiple Phase III trials. The pivotal REVISIT trial evaluated its efficacy and safety in complicated intra-abdominal infections, hospital-acquired pneumonia (HAP), and ventilator-associated pneumonia (VAP). Results showed a clinical cure rate of 68.4% (193/282) in the ATM-AVI group, which was non-inferior to the meropenem-colistin group (65.7%, 92/140), with significantly lower nephrotoxicity compared to colistin-containing regimens [4]. The ASSEMBLE trial for MBL-producing strain infections demonstrated that the clinical success rate of ATM-AVI was significantly higher than that of best available therapy (42% vs. 15.0%) [5].

(III) Treatment of Carbapenem-Resistant Acinetobacter baumannii (CRAB)

Both IDSA and ESCMID recommend combination therapy with at least two agents possessing in vitro activity for CRAB infections. The current first-line regimen is sulbactam-durlobactam (SUL-DUR) combined with a carbapenem. If SUL-DUR is unavailable, alternatives include high-dose sulbactam combined with at least one other active agent (e.g., polymyxin B, minocycline, tigecycline, or cefiderocol).

Durlobactam, a novel β-lactamase inhibitor (BLI), exhibits broad-spectrum, irreversible inhibitory activity against Class A, C, and D β-lactamases. Its core pharmacological advantage lies in its ability to efficiently block Class D enzymes, particularly oxacillin-hydrolyzing OXA-23, via covalent binding. This protects sulbactam from hydrolysis, thereby restoring and enhancing its bactericidal activity against Acinetobacter baumannii penicillin-binding proteins (PBPs). Additionally, the combination synergistically inhibits bacterial cell wall synthesis through a multi-target mechanism.

Pharmacokinetically, sulbactam and durlobactam exhibit high temporal synchrony in humans, with elimination half-lives of approximately 2 hours. Studies confirm the combination achieves excellent permeability and concentration stability in epithelial lining fluid (ELF). Population PK simulations in Chinese populations verify that under standard dosing (1 g/1 g every 6 hours with 3-hour extended infusion), PTA approaches 100% for CRAB isolates with MIC ≤ 4 μg/mL.

The Phase III ATTACK trial (including a Chinese subgroup) demonstrated that for CRAB-caused HAP/VAP and bloodstream infections, the 28-day all-cause mortality in the SUL-DUR group was 19.0%, significantly lower than the polymyxin group (32.3%). Nephrotoxicity was also significantly lower (13.2% vs. 37.6%, P<0.001) [6]. Based on this efficacy and safety profile, IDSA has listed SUL-DUR as a first-line option for CRAB infections [7].

(IV) Treatment of Difficult-to-Treat Pseudomonas aeruginosa (DTR-PA)

For DTR-PA infections, first-line therapy involves novel β-lactam/β-lactamase inhibitor (BL/BLI) combinations, including ceftolozane-tazobactam (C/T), ceftazidime-avibactam (CAZ-AVI), and imipenem-cilastatin-relebactam (IMI-REL). Cefiderocol serves as an alternative. If DTR-PA is confirmed to produce MBLs, cefiderocol is the drug of choice.

C/T is specifically designed to address DTR-PA resistance mechanisms. Ceftolozane’s molecular modification of the ceftazidime side chain significantly enhances affinity for P. aeruginosa PBPs. It also exhibits excellent AmpC stability and is less affected by efflux pump upregulation or OprD porin loss. Tazobactam further broadens the enzyme inhibition spectrum by suppressing certain extended-spectrum β-lactamases (ESBLs).

C/T demonstrates excellent tissue penetration. Clinical studies confirm ceftolozane achieves significantly higher ELF penetration than traditional cephalosporins, rapidly reaching therapeutic concentrations in pulmonary lesions. Primarily renally excreted unchanged, its half-life is 2.3–3.0 hours. Extended infusion optimizes time above MIC, ensuring high PTA in complex infection sites. The Phase III ASPECT-NP trial in mechanically ventilated severe pneumonia patients showed C/T was non-inferior to meropenem for 28-day all-cause mortality (24.0% vs. 25.3%) [8]. Post-hoc analysis revealed superior clinical cure rates for P. aeruginosa pneumonia compared to traditional BL + aminoglycoside regimens [9]. The 2025 CACTUS study (MDR-PA pneumonia & bacteremia) showed C/T clinical success significantly outperformed CAZ-AVI (aOR 2.07, 95% CI 1.16–3.70) [10].

(V) Salvage Therapy for Extensively Drug-Resistant Gram-Negative Bacilli (XDR-GNB)

Cefiderocol exhibits excellent tissue penetration and a wide volume of distribution. Primarily renally excreted unchanged with a half-life of 2–3 hours, it is a time-dependent antibiotic. Prolonged infusion (e.g., 3-hour regimen) significantly optimizes PTA, particularly in critically ill patients with complex pathophysiology. Studies confirm it achieves concentrations well above pathogen MICs in the lungs, abdomen, and urinary tract.

Cefiderocol is a strategic antimicrobial for XDR-GNB, regarded as a “last-resort” agent for CRGNB infections. Its core innovation is the siderophore mechanism: a catechol moiety chelates environmental Fe³⁺, mimicking iron substrates to hijack bacterial iron transport systems for active uptake into the periplasmic space. This bypasses porin loss and efflux-mediated resistance. It also exhibits exceptional enzymatic stability, being the only monotherapy globally resistant to hydrolysis by all four classes (A, B, C, D) of β-lactamases, with potent in vitro activity against CRAB, Stenotrophomonas maltophilia, and MBL-producing CRE.

Recent large-scale real-world studies establish cefiderocol’s role in salvage therapy. The PROVE study (multicenter ICU cohorts in Europe/US) and data from ESCMID Global 2025 & IDWeek 2025 show an overall clinical cure rate of ~70% in critically ill patients. Early empirical treatment yielded a significantly higher cure rate (73.7%) vs. late salvage (54.3%) [11]. The CEFL-ID study in immunocompromised patients (hematologic malignancies, transplants) confirmed a 28-day clinical success rate of 53.3% even in highly complex infections [12]. These data prove cefiderocol can provide >50% survival expectancy even against super-resistant strains failing novel BL-BLIs. It was officially approved for clinical use in China on January 8, 2026.

Summary: 2025 CRGNB treatment strategies have shifted to precision coverage based on rapid enzymatic detection (Table 2-4-1): KPC producers → IMI-REL or CAZ-AVI; MBL producers → ATM-AVI or cefiderocol; CRAB → SUL-DUR; DTR-PA → C/T or IMI-REL.

(VI) Future Prospects of Antimicrobial Application in the Precision Medicine Era

Despite significant advances in treating CRGNB with novel antimicrobials, clinical application faces significant challenges.

(1) Dynamic Evolution of Resistance Mechanisms
Clinical monitoring has identified KPC-2 gene mutations during CAZ-AVI therapy, leading to altered enzymatic activity and treatment failure. Additionally, some strains co-carry multiple resistance enzymes, and SUL-DUR-resistant CRAB has emerged, underscoring the critical need for dynamic resistance monitoring when deploying novel agents.

(2) Pathophysiological Variations in Critically Ill Patients
Sepsis and septic shock often involve hyperdynamic circulation, capillary leak, or acute kidney injury, significantly altering antimicrobial volume of distribution and clearance. For critically ill patients, 2024 clinical guidelines strongly emphasize optimizing dosing via “extended infusion” [13]. For novel BL/BLIs, a 3–4 hour continuous infusion is recommended to maximize PTA, which is decisive when treating strains near susceptibility breakpoints.

With multiple novel antimicrobials entering clinical practice, CRGNB management has undergone a paradigm shift, marking the transition from “empirical combination therapy” to “precision treatment based on molecular resistance mechanisms.” Carbapenemase-typing-guided therapy, combined with individualized therapeutic drug monitoring (TDM) and dynamic resistance assessment, forms a closed-loop management system that significantly optimizes antimicrobial strategies in critically ill patients.
(Qu Hongping, Dai Yunqi, Ruijin Hospital, School of Medicine, Shanghai Jiao Tong University)

References

1. Zeng M, Xia J, Zong ZY, et al. Guidelines for diagnosis, treatment, and prevention of carbapenem-resistant Gram-negative bacterial infections. Chin J Infect Chemother. 2024;24(2):135-151.

2. Motsch J, Murtade Oliveira C, Stus Y, et al. RESTORE-IMI 1: imipenem/relebactam vs colistin plus imipenem. Clin Infect Dis. 2020;70(9):1799-1808.

3. Marino A, Pipitone G, Venanzi Rullo E, et al. Restoring control: real-world success with imipenem-relebactam in critical MDR infections. Pathogens. 2025;14(7):685.

4. Carmeli Y, Cisneros JM, Paul M, et al. Aztreonam-avibactam versus meropenem for serious Gram-negative infections (REVISIT). Lancet Infect Dis. 2025;25(2):218-230.

5. Daikos GL, Cisneros JM, Carmeli Y, et al. Aztreonam-avibactam for MBL-producing Gram-negative pathogens (ASSEMBLE). JAC Antimicrob Resist. 2025;7(4):dlaf131.

6. Kaye KS, Shorr AF, Wunderink RG, et al. Sulbactam-durlobactam vs colistin for CRAB infections (ATTACK). Lancet Infect Dis. 2023;23(9):1072-1084.

7. Tamma PD, Heil EL, Justo JA, et al. IDSA 2024 Guidance on antimicrobial-resistant Gram-negative infections. Clin Infect Dis. 2024;ciae403.

8. Kollef MH, Novacek M, Kivistik C, et al. Ceftolozane-tazobactam vs meropenem for nosocomial pneumonia (ASPECT-NP). Lancet Infect Dis. 2019;19(12):1299-1311.

9. Timsit JF, Huntington JA, Wunderink RG, et al. Ceftolozane/tazobactam vs meropenem in ventilated HAP/VAP (ASPECT-NP subset). Crit Care. 2021;25(1):290.

10. Shields RK, Abbo LM, Ackley R, et al. Ceftazidime-avibactam vs ceftolozane-tazobactam for MDR-PA (CACTUS). Lancet Infect Dis. 2025;25(5):574-584.

11. Clancy CJ, Cornely OA, Marcellas W, et al. Cefiderocol effectiveness in Gram-negative infections (PROVE US interim). Infect Drug Resist. 2024;17:4427-4443.

12. Souges S, Faure E, Parize P, et al. Cefiderocol in immunocompromised patients with MDR-GNB (CEFL-ID). J Infect. 2025;90(1):106376.

13. Abdul-Aziz MH, Hammond NE, Bretts J, et al. Prolonged vs intermittent β-lactam infusions in sepsis. JAMA. 2024;332(8):638-648.