Highlights
– GAME CHANGER randomized 504 adults with hospital‑acquired or healthcare‑associated Gram‑negative bloodstream infection to cefiderocol versus clinician‑chosen standard therapy and showed non‑inferiority for 14‑day all‑cause mortality (absolute risk difference 1%, 95% CI −3 to 6).
– In the predefined carbapenem‑resistant (CR) subset (25% of participants), mortality at 14 days was 14% with cefiderocol versus 10% with standard therapy (difference 5%, 95% CI −7 to 16); the trial was underpowered to demonstrate superiority in this subgroup.
– Cefiderocol‑related serious adverse events were uncommon but observed only in the cefiderocol arm (five events: neuropsychiatric symptoms, rigors, abnormal liver chemistry, rash); overall safety was similar between groups.
Background: clinical need and therapeutic context
Bloodstream infection due to Gram‑negative bacilli is a leading cause of sepsis in hospitalized patients and is associated with substantial morbidity and mortality. The emergence and spread of carbapenem‑resistant Gram‑negative organisms (including Enterobacterales, Pseudomonas aeruginosa, and Acinetobacter spp.) have narrowed effective empirical and directed antibiotic choices and are associated with worse outcomes.
Cefiderocol is a parenteral siderophore cephalosporin with broad in vitro activity against many multidrug‑resistant Gram‑negative pathogens, including some carbapenem‑resistant strains. Its novel mechanism—iron‑chelating “trojan horse” entry through bacterial iron transporters—raises the possibility of improved penetration into resistant organisms. Before GAME CHANGER, clinical evidence for cefiderocol’s comparative efficacy in bloodstream infections caused by resistant Gram‑negatives was limited and included smaller trials and observational reports.
Study design
GAME CHANGER was an open‑label, parallel‑group, randomized trial performed in 17 tertiary hospitals across Australia, Malaysia, Singapore, Taiwan, Thailand, and Türkiye. Eligible adult patients had hospital‑acquired or healthcare‑associated bloodstream infection with Gram‑negative bacilli visible on Gram stain from an index blood culture. Randomization was 1:1 to cefiderocol (2 g IV every 8 hours) versus standard‑of‑care (SOC) antibiotics selected by the treating team. Randomization was stratified by region and comorbidity severity, using randomly permuted blocks. Patients, clinicians, and site investigators were not blinded.
The primary outcome was 14‑day all‑cause mortality after randomization. The primary analysis included participants who received at least one dose of assigned drug, had an aerobic Gram‑negative bacillus isolated from the index blood culture, and had not withdrawn consent before day 14. The trial tested non‑inferiority with a 10% absolute margin; superiority testing was planned if non‑inferiority was established. Missing data were handled by imputation. The trial was registered at ClinicalTrials.gov (NCT03869437).
Key findings
Population characteristics and enrollment
From Nov 12, 2019, to Oct 31, 2023, 513 patients were enrolled (42% female). After exclusion of nine patients who withdrew consent or whose cultures did not confirm an aerobic Gram‑negative bacillus, 504 patients constituted the main analysis population: 250 assigned to cefiderocol and 254 to SOC. A quarter (127/504) had bloodstream infection due to at least one carbapenem‑resistant organism.
Primary outcome: 14‑day mortality
At day 14, mortality in the main analysis population was 8% (20/250) in the cefiderocol arm versus 7% (17/254) in the SOC arm. The absolute risk difference was 1% (95% CI −3 to 6), which met the predefined non‑inferiority margin of 10%, indicating that cefiderocol was non‑inferior to SOC for short‑term survival. Because non‑inferiority was demonstrated, superiority testing was performed; cefiderocol was not superior to SOC.
Carbapenem‑resistant subset
In patients with at least one carbapenem‑resistant organism causing bloodstream infection (n=127), mortality at 14 days was 14% (9/64) with cefiderocol versus 10% (6/63) with SOC (absolute difference 5%, 95% CI −7 to 16), a result that did not demonstrate superiority and was imprecise. The confidence interval spans both potential harm and benefit, reflecting limited sample size in this subgroup.
Safety and adverse events
Overall adverse event profiles were similar between arms. Five treatment‑emergent serious adverse events considered probably or possibly related to the study drug occurred, all in the cefiderocol group: delirium, stupor, rigors, abnormal liver chemistry, and rash. All resolved without specific therapy except the rash, which required corticosteroid and antihistamine treatment. There were no new unexpected toxicity signals.
Interpretation and clinical implications
GAME CHANGER provides the largest randomized comparison to date of cefiderocol versus routine SOC antibiotics for hospital‑acquired and healthcare‑associated Gram‑negative bloodstream infection. The main finding—that cefiderocol is non‑inferior to SOC for 14‑day mortality—supports its role as an alternative therapeutic option in this setting, particularly when healthcare‑associated exposure or local epidemiology heightens the risk of resistant Gram‑negative pathogens.
However, the trial did not show superiority for cefiderocol overall or within the carbapenem‑resistant subset. The absence of superiority should not be interpreted as inferiority; rather, it indicates that cefiderocol performs comparably to current standard regimens in this heterogeneous, pragmatic population.
For carbapenem‑resistant infections specifically, the subgroup data are underpowered and imprecise. Although cefiderocol has attractive in vitro activity against many carbapenemase‑producing organisms, clinical outcome depends on numerous factors beyond in vitro susceptibility (e.g., site of infection, pathogen species and resistance mechanisms, host factors, timely source control, and co‑infections). Therefore, definitive conclusions about cefiderocol’s superiority or specific niche for CR bloodstream infections cannot be drawn from this dataset alone.
Strengths
Key strengths include randomized allocation, multicenter and multinational conduct, pragmatic inclusion criteria (Gram stain evidence guiding early therapy), and use of a clinically meaningful hard endpoint (all‑cause mortality at 14 days). The design reflects real‑world decision pathways where initial empiric therapy is guided by Gram stain and clinical judgment.
Limitations
Several limitations merit emphasis. First, the open‑label design can introduce bias in ancillary care and reporting of subjective outcomes, although mortality is an objective endpoint. Second, standard‑of‑care regimens were selected by treating clinicians and therefore heterogeneous; while this increases real‑world relevance, it complicates interpretation of comparative efficacy against any single SOC regimen.
Third, the trial was not powered to detect modest differences in mortality within key subgroups, notably the carbapenem‑resistant cohort or infections caused by specific species such as Acinetobacter spp. Fourth, regional differences in pathogen distribution, local resistance mechanisms, and availability of alternative agents (e.g., novel beta‑lactam/beta‑lactamase inhibitor combinations) may limit generalizability to settings with different epidemiology. Finally, although missing data were handled with imputation, imputation assumptions can influence results when loss to follow‑up is nontrivial.
How this trial informs practice
For clinicians confronting hospital‑acquired or healthcare‑associated Gram‑negative bacteraemia, GAME CHANGER supports considering cefiderocol as an alternative to existing standard therapies when resistance is a concern or when other first‑line agents are unsuitable. Antimicrobial stewardship teams should integrate local susceptibility data, species identification, and resistance mechanism testing (where available) when deciding on continuing cefiderocol after pathogen and susceptibility data are known.
Given the imprecision in the carbapenem‑resistant subgroup, cefiderocol may be particularly valuable for targeted therapy when in vitro susceptibility is confirmed and options are limited, but clinicians should remain cautious and consider combination strategies or consult infectious disease specialists for complex CR infections.
Research and policy implications
Further research is needed to clarify cefiderocol’s role in bloodstream infection caused by specific carbapenemase types and species (e.g., metallo‑beta‑lactamase producers or Acinetobacter spp.), to evaluate outcomes beyond 14 days (including relapse and long‑term mortality), and to compare cefiderocol head‑to‑head against specific newer agents in high‑risk, resistance‑enriched populations. Cost‑effectiveness and implementation studies will inform formulary decisions, particularly in resource‑limited settings where the burden of resistant Gram‑negative infection is high.
Conclusion
The GAME CHANGER trial demonstrates that cefiderocol is non‑inferior to clinician‑selected standard therapy for 14‑day mortality in adults with hospital‑acquired or healthcare‑associated Gram‑negative bloodstream infection. While adverse events attributable to cefiderocol were uncommon, the trial did not show superiority overall or in the carbapenem‑resistant subgroup. Cefiderocol therefore represents a credible therapeutic option when resistance risk is high or alternative agents are unsuitable, but additional trials or pooled analyses are required to define its optimal role against specific carbapenem‑resistant pathogens.
Funding and trial registration
GAME CHANGER was funded by the University of Queensland, Shionogi, The Henderson Foundation, and the National Medical Research Council (Singapore). The trial is registered at ClinicalTrials.gov (NCT03869437).
References
Paterson DL, Sulaiman H, Liu PY, et al; GAME CHANGER Trial Investigators. Cefiderocol versus standard therapy for hospital‑acquired and health‑care‑associated Gram‑negative bacterial bloodstream infection (the GAME CHANGER trial): an open‑label, parallel‑group, randomised trial. Lancet Infect Dis. 2025 Oct 6:S1473-3099(25)00469-4. doi: 10.1016/S1473-3099(25)00469-4. Epub ahead of print. PMID: 41067237.
Prescribing information: FETROJA (cefiderocol) injection. Shionogi Inc. (US Prescribing Information), available from FDA-approved label resources.
