Skip to main content

Improved Outcomes and Resource Use With Normothermic Machine Perfusion in Liver Transplantation.

JAMA surgery2025-01-29PubMed
Total: 73.0Innovation: 7Impact: 8Rigor: 7Citation: 8

Summary

In 1086 consecutive liver transplants, normothermic machine perfusion was associated with markedly better outcomes than static cold storage, particularly for DCD grafts: lower EAD (17.5% vs 50.0% in DCD), reduced AKI, shorter LOS, dramatically fewer readmissions, and substantial reductions in graft failure (HR 0.22 overall; 0.13 for DCD) and mortality (HR 0.31).

Key Findings

  • Early allograft dysfunction lowest with DCD-NMP (17.5%) vs DCD-SCS (50.0%), DBD-NMP (36.8%), DBD-SCS (27.3%) (P<.001)
  • Shorter hospital and ICU LOS with DCD-NMP (median hospital 5.0 days; ICU 1.5 days; both P=.01)
  • Lower 1-year readmissions: 86% lower for DCD-NMP vs DCD-SCS; 53% lower for DBD-NMP vs DBD-SCS
  • AKI reduced in DCD-NMP (31.1%) vs DCD-SCS (47.4%) (P=.001)
  • Graft failure reduced with NMP overall (HR 0.22) and especially in DCD (HR 0.13); mortality reduced (HR 0.31)

Clinical Implications

Anesthesia and transplant teams should consider prioritizing NMP, especially for DCD livers, to lower EAD, resource use, and graft failure. Implementation planning should address logistics, costs, and training.

Why It Matters

Provides compelling real-world evidence that NMP improves clinical and resource outcomes, supporting broader adoption and protocol changes in transplant anesthesia and surgery.

Limitations

  • Single-center retrospective design with potential selection bias and residual confounding
  • Temporal changes in practice could confound comparisons between NMP and SCS eras

Future Directions

Prospective multicenter trials and cost-effectiveness analyses to validate benefits, optimize selection criteria, and integrate NMP into standardized workflows.

Study Information

Study Type
Cohort
Research Domain
Treatment
Evidence Level
III - Retrospective observational cohort; susceptible to confounding but informative real-world effectiveness data.
Study Design
OTHER