Extensively drug-resistant Pseudomonas aeruginosa (XDR-PA) has evolved into a major threat to global healthcare systems over the past decades, particularly for patients with advanced age, immune deficiency, and serious conditions. From a global epidemiological perspective, XDR-PA accounts for 15%-30% of clinical Pseudomonas aeruginosa (P. aeruginosa) isolates from intensive care units, which is mainly driven by the overuse of broad-spectrum antibiotics and cross-transmission (Mirzaei et al., 2020; Montero et al., 2020). In China, the detection rate of XDR-PA has risen from less than 4% to approximately 9% over the past decade (Wang et al., 2021; Yang et al., 2025). More importantly, recent studies reported that the XDR-PA with hypervirulent clone has been sporadically identified around the world (de Paula-Petroli et al., 2018; Zhang et al., 2022). These isolates carry multiple drug resistance genes (e.g., blaOXA-902, blaOXA-486, and blaNDM-1), secrete various virulence factors (e.g., exoU, exoS, and elastases), and possess strong biofilm formation capabilities, making them difficult to eradicate and contributing to elevated mortality rates, particularly in elderly patients with underlying diseases (Li et al., 2024; Kang et al., 2026).

In general, the combination of ceftazidime-avibactam (CZA) and aztreonam (ATM) medication is recommended as a first-line option against metallo-β-lactamase (MBL) producing Enterobacteriaceae by current guidelines (Mantzarlis et al., 2025). However, for XDR-PA, things might not happen so straightforwardly. Besides carbapenemase production, other drug-resistant mechanisms, such as the loss of porin and the overexpression of drug efflux pumps, might be commonly included (Meletis and Karastergiou, 2025). These multifaceted mechanisms not only pose a challenge in therapeutic options but underscore the necessity of antibiotic stewardship and infection control measures to limit the spread of multidrug-resistant strains. On this occasion, the evaluation of the efficacy and optimal dosing regimens against XDR-PA in real−world clinical settings becomes particularly crucial and urgent.

In this study, we report the first case worldwide of successfully treating severe pneumonia caused by New Delhi metallo-β-lactamase (NDM) positive XDR-PA with a combination of CZA and high-dose ATM. This updated therapeutic regimen balances efficacy and safety, offering an individualized and cost-effective option for refractory pulmonary infections among elderly and critically ill populations in future clinical practice.

This case represents the first confirmed instance demonstrating that high-dose aztreonam enhanced the combined antimicrobial activity of CZA against NDM positive XDR-PA in severe pneumonia therapy. The patient in this study, presenting with advanced age, multiple underlying comorbidities, and prolonged bedrest, constituted a high-risk population for infection with hypervirulent XDR-PA strains (Li et al., 2024). After reviewing the patient’s culture records with positive results during his current hospital stay, we discovered that since the first positive culture isolated from the patient’s sputum, the XDR-PA was repeatedly detected in his peripheral blood and BALF several times, indicating the condition was growing worse. This strain is highly invasive and hard to eradicate once it infects an elderly and immunocompromised patient.

According to the latest report of antimicrobial resistance threats released by Centers for Disease Control and Prevention of the United States in July 2024, XDR-PA is regarded as a pathogen with “serious threat” level for several consecutive years, leading to over 700 million dollars attributable healthcare costs in 2017. Compared with 2019, the number of hospital-acquired XDR-PA infection cases increased by 32% in 2020, which may be associated with the prolonged length of hospital stay or secondary infections related to COVID-19 sequelae (Schwartz et al., 2024). In China, P. aeruginosa remains the second leading cause of hospital-acquired pneumonia (Society PIAoCT, 2022). Notably, among patients with ventilator-associated pneumonia, XDR-PA is noteworthy for its high prevalence, with isolation rates reaching nearly 30% (Yin et al., 2021). More seriously, the latest data released by China Antimicrobial Surveillance Network indicate that some XDR-PA strains have developed resistance even to novel high-level antibiotics, such as ceftolozane-tazobactam (MIC range = 0.25->128 μg/mL, MIC₅₀ = 1 μg/mL, MIC₉₀ = 128 μg/mL, strain number = 246, resistant rate = 13.8%) and meropenem-vaborbactam (MIC range = 0.03->64 μg/mL, MIC₅₀ = 8 μg/mL, MIC₉₀ = 32 μg/mL, strain number = 302, resistant rate = 28.8%) (Guo et al., 2025). This highlights the urgent need to study the responsible mechanisms and develop effective response strategies.

Findings from several recent multicenter epidemiological studies in China demonstrate that MBL production is present in over 50% of XDR-PA strains, largely comprising IMP and VIM types, with a consistently upward trend in prevalence (Wang and Wang, 2020; Xu et al., 2020). Furthermore, as a growing number of newfound MBL subtypes are identified worldwide, the situation regarding the prevention and control of XDR-PA becomes increasingly challenging (Zhai et al., 2025).

In our work, accurate carbapenemase typing was established through concordant results from the inhibitor enhancement method and lateral flow immunochromatographic assay, enabling us to select an appropriate combined AST schedule at a very early stage. The detection of NDM, a less common but significant MBL type among XDR-PA strains in China, signals its ongoing spread and emphasizes the necessity of continuous monitoring to improve patient management and infection control.

At present, cefiderocol and aztreonam-avibactam (AZA) are generally recommended as the first-line and second-line agents, respectively, against MBL producing XDR-PA on the international stage (Hidalgo-Tenorio et al., 2024). However, these two agents have only recently been approved for marketing in China. The application is limited due to the high cost and the lack of large-scale domestic data supporting their clinical efficacy. Therefore, the CZA + ATM combination regimen is currently widely adopted as an alternative in clinical practice in China. Meanwhile, clinical guidelines also specify that avibactam does not significantly enhance the bactericidal activity of ATM against CRPA (Hidalgo-Tenorio et al., 2024) — a conclusion consistent with our observations from the broth disk elution test in Figure 1B. Besides, a recent study demonstrated that, compared with monotherapy, the combination of CZA and ATM therapy exhibited a stronger antimicrobial effect on MBL producing CRPA in vitro, yielding favorable outcomes even when the strains simultaneously harbor AmpC over-expression or OprD deficiency (Montero et al., 2025).

Niklas and his colleagues recently reported that the inhibitory effect on XDR growth reaches its maximum when the concentration ratio of avibactam (AVI) to ATM is optimized at 1:4 in vitro. Further increasing the concentration of ATM appears to offer no additional benefit due to a saturation effect (Klein et al., 2025). In our study, we evaluated the synergistic effect of CZA (15/10 μg/mL) and ATM (30 μg/mL) using the broth disk elution method. The concentration ratio of AVI to ATM was set at 1:3 (Figure 1B), which was near the saturation threshold and was subsequently demonstrated to achieve favorable anti-infective outcomes in the patient.

According to the pharmacokinetic/pharmacodynamic principles, both CZA and ATM are time-dependent antibacterial agents with a short post-antibiotic effect. Therefore, to improve the probability of target attainment of the combined regimen, we extended the infusion duration of ATM from 1 h to 3 h and shortened the dosing interval from q8h to q6h. CZA and ATM were delivered via a Y-shaped infusion tube simultaneously during intravenous administration. Additionally, given the patient’s hypoalbuminemia (Supplementary Table 2), the concentration of free drug in the bloodstream may have been further elevated. Regarding safety, there is no incompatibility between CZA and ATM. Both agents are primarily eliminated via the kidneys as unchanged parent compounds. Thus, renal function and plasma drug concentrations were closely monitored during treatment.

Once the infection was effectively controlled, sputum or BLAF from the patient should be collected again for etiological testing, and a stepwise de-escalation therapy should be considered based on the AST results. For example, reducing the dosage of ATM for the sake of renal function protection and administering aminoglycosides or colistin via aerosol inhalation in addition to ongoing intravenous agents. Meanwhile, attention should be paid to preventing infections caused by carbapenem-resistant Acinetobacter baumannii complex due to prolonged ICU stay. Empirical strategies for such coverage could include, for example, high-dose ampicillin-sulbactam combined with tigecycline, minocycline, or colistin.

To summarize, we detailed the clinical features and diagnostic thinking of an elderly patient with severe pneumonia. More importantly, we also introduced an effective antimicrobial therapy against NDM positive XDR-PA, characterized by cost-effective and dose-flexible, offering additional options for personalized treatment of pulmonary infection in the future.

We attribute the successful treatment of this case to our adherence to “three actions of boldness with three principles of prudence” at several critical junctures that determined the clinical outcome. First was bold hypothesizing coupled with careful verification. The enlarged diameter of the inhibition zone between adjacent ATM disks was observed incidentally in the combined AST, which inspired us that ATM dosage might influence the synergistic effect. The conjecture was ultimately validated by the broth disk elution assay (Figure 1). Second was bold recommendations paired with rigorous oversight. We strictly followed the treatment indications by fully integrating the combined AST results, drug instructions, and expert consensus statements. After a comprehensive assessment of the patient’s condition in collaboration with clinical pharmacists (Supplementary Table 2), we proposed a novel therapeutic regimen, namely CZA + ATM_hi, and recommended its clinical use for the patient’s anti-infective treatment. Third was bold implementation paired with meticulous assessment. Given the recurrent infection and critical condition of the patient, a tripartite agreement was reached among clinicians, clinical pharmacists, and microbiologists in our team. We ultimately decided to proceed with CZA + ATM_hi combination therapy under the close surveillance of the patient’s liver and renal function (Supplementary Figure 3), while timely evaluating the anti-infective effect (Figure 2). Through these collaborative efforts, CZA and ATM worked like two sharp swords, striking together with a remarkable synergistic antibacterial effect against the NDM positive XDR-PA. Finally, the patient exhibited a favorable clinical response and was successfully rescued from the critical state.