Towards Benchmark-level Outcomes in Donation After Circulatory Death Liver Transplantation: A Comparative Analysis of...

Donation after circulatory death (DCD) graft use has become a standard practice globally to increase access to liver transplantation (LT). Machine preservation strategies have been widely adopted to improve graft quality, mitigate ischemia-reperfusion injury, expand organ utilization and enable organ viability assessment. However, the various approaches to MP have never been directly compared, particularly in the DCD cohort which might experience the greatest benefit.

All cases of adult DCD-LT 2014-2024 preserved MP were collected from centers in Italy, The Netherlands, Switzerland, UK, and USA. DCD grafts preserved with cold storage (SCS) alone from previously established DCD-benchmarking cohorts served as comparators. Core Outcome Sets methodology was employed for standardized outcome analysis, comparing the impact of MP technologies in DCD grafts within established "benchmark" criteria to those with extended risk factors outside benchmarks.

Data were collected for 1538 perfused DCD-LT from >30 transplant centers; 995 (65%) were outside-benchmark criteria. Nearly half received NRP alone or combined with ex-situ perfusion (49%). Inside benchmark, nonanastomotic biliary strictures (NAS) were equally common in upfront NMP (14%) and back-to-base (B-t-B) NMP (14%). NRP (1%) and NRP+HOPE (2%) were protective (P<0.001). Outside benchmark, NAS was least frequent in the NRP+HOPE (2%), and NRP-alone (4%) group, and most common in the B-t-B NMP (18%, P<0.001) cohort. Upfront NMP demonstrated a lower observed NAS incidence than B-to-B NMP (9% vs. 18%). NRP, NRP+HOPE, and HOPE demonstrated a pattern of convergence in outcomes between inside- and outside-benchmark grafts in multivariable analyses. Binary logistic mixed-effects modelling demonstrated NAS-reduction with NRP+HOPE (OR=0.201, 95% CI= 0.075-0.569) and NRP alone (OR=0.281, 0.105-0.569). Other preservation modalities did not differ, including upfront NMP, which was similar to B-t-B NMP and SCS. Increasing donor age was detrimental to NAS rates (OR=1.02, 95% CI= 1.004-1.036). Donor and preservation factors demonstrated a 19% impact on NAS versus recipient factors (labMELD, age, BMI), which contributed with only 3% (Conditional Pseudo-R2=0.201, Marginal=0.187). Similar trends were observed regarding actuarial 1-year GS.

Machine preservation strategies were associated with differences in outcomes in DCD LT and with lower observed NAS incidence and favorable adjusted survival estimates in DCD LT. NMP-based preservation was not associated with statistically significant reductions in NAS compared with benchmark SCS cohorts.