Back to 2026 Abstracts

---

Ex Vivo Lung Perfusion and Allograft Failure After Lung Transplantation: A National Target Trial Emulation
*Sara Sakowitz^1,4, *Syed Shahyan Bakhtiyar^4,3, *Yas Sanaiha^3,4, *Deep Mehta⁴, Richard J. Shemin³, Peyman Benharash^3,4
1Department of Surgery, Massachusetts General Hospital, Boston, Massachusetts; ²Division of Cardiothoracic Surgery, Department of Surgery, University of Utah, Salt Lake City, Utah; ³Division of Cardiac Surgery, Department of Surgery, University of California, Los Angeles, Los Angeles, California; ⁴Cardiovascular Outcomes Research Laboratories, Department of Surgery, University of California, Los Angeles, Los Angeles, California

Objective:
Ex vivo lung perfusion (EVLP) has been championed as a disruptive technology to expand the donor pool and improve access to lung transplantation (LT). Yet contemporary evidence remains limited to single-center or retrospective studies, and the true clinical impact of EVLP at the national level is unknown. Using a real-world, national cohort, we emulated a randomized controlled trial (RCT) to determine the causal effect of EVLP on post-LT outcomes.

Methods:
All LT recipients (≥18 years) were identified in the 2017-2025 Organ Procurement and Transplantation Network. Recipients of donor allografts that underwent EVLP were categorized as EVLP (others: Non-EVLP). Using target trial emulation with clone-censor-weight methodology, we simulated an open-label RCT comparing EVLP with standard allografts. The primary outcome was allograft survival at three-years; we secondarily considered postoperative complications and patient survival.
Survival was evaluated using Kaplan-Meier and hierarchical Cox models with gamma frailty to account for clustering. Covariates for risk-adjustment were automatically selected using the least absolute shrinkage and selection operator, and included sociodemographic characteristics, transplant-specific factors, allocation score, hemodynamics, and comorbidities.

Results:
Among 18,047 LTs, 1,149 (6%) involved EVLP. EVLP utilization increased from 2% in 2017 to 6% in 2025 (P<0.001). EVLP allografts were procured from older, higher BMI donors who more commonly had a history of cigarette use, diabetes, and hypertension. EVLP organs were also more often procured from non-beating heart donors and experienced significantly longer cold ischemia time. Meanwhile, EVLP and Non-EVLP recipients were clinically similar.
Unadjusted allograft survival was lower after EVLP at one (87 vs 89%, P=0.02) and three years (68 vs 72%, P=0.007). In the emulated RCT, EVLP remained independently associated with increased hazards of allograft failure at one (HR 1.57, CI 1.23-2.01) and three-years (HR 1.26, CI 1.06-1.49). Receipt of an EVLP allograft also portended higher hazard of overall patient mortality at one (HR 1.57, CI 1.22-2.01) and three-years (HR 1.25, CI 1.05-1.48).
EVLP was further associated with higher incidence of postoperative extracorporeal membrane oxygenation (16 vs 9%, P<0.001), dialysis (11 vs 8%, P=0.006), inhaled nitric oxide within 72 hours (19 vs 12%, P<0.001), reintubation (22 vs 20%, P=0.03), and prolonged ventilation >5days (31 vs 24%, P<0.001).

Conclusion:
In this national target trial emulation, EVLP was associated with significantly increased risk of allograft failure, postoperative complications, and mortality within three years of LT. Although EVLP was initially heralded as a transformative technology to expand the donor pool, these real-world data call for caution regarding widespread adoption and underscore the need for rigorous prospective evaluation before continued expansion in lung transplantation.


TABLE 1: Recipient and Donor Characteristics, Stratified by Receipt of EVLP-Perfused Lung Allograft.
- IQR, Inter-Quartile Range; BMI, Body Mass Index; SD, Standard Deviation; LAS, Lung Allocation Score; CAS, Composite Allocation Score