Normothermic Regional Perfusion in Pancreas and Islet Transplantation from Donation After Circulatory Death: A Systematic Review

Technical Summary: Normothermic Regional Perfusion in Pancreas and Islet Transplantation from Donation After Circulatory Death

Problem Statement
Pancreas transplantation remains the most effective treatment for type 1 diabetes with end-stage kidney disease, yet access is shrinking due to a failure in organ utilization rather than supply. In the United States, the deceased-donor transplant rate fell significantly between 2019 and 2024, with over 25% of recovered pancreases not being transplanted. A primary barrier to expanding the donor pool is the underutilization of Donation After Circulatory Death (DCD) pancreases. These organs are highly susceptible to warm ischemia-reperfusion injury, leading to early graft loss driven by thrombosis and pancreatitis.

While Normothermic Regional Perfusion (NRP) has improved outcomes for liver and kidney transplants from DCD donors by restoring oxygenated circulation before procurement, its efficacy for the pancreas and islet cells remains uncertain. Standard DCD practice relies on Super-Rapid Recovery (SRR), which limits but cannot reverse warm ischemia. There is a critical need to determine if NRP-assisted DCD procurement yields pancreas grafts and islet isolates comparable to SRR DCD and Donation After Brain Death (DBD) standards, or if the potential extension of warm ischemia time prior to perfusion onset negates benefits.

Methodology
This systematic review adhered to PRISMA 2020 guidelines and was registered with PROSPERO (CRD420261540738). The authors searched MEDLINE, Embase, Web of Science, Cochrane CENTRAL, and ClinicalTrials.gov up to October 2026, including congress proceedings.

• Eligibility: The review included randomized trials, cohort studies, and registry analyses (published from 2016) comparing NRP-assisted DCD procurement against SRR DCD or DBD procurement.
• Outcomes: Primary outcomes included 1-year death-censored pancreas graft survival and early graft thrombosis. Secondary outcomes encompassed pancreas utilization, primary non-function (PNF), graft pancreatitis, patient survival, and islet isolation metrics (yield, viability, function).
• Analysis: Risk of bias was assessed using ROBINS-I, and certainty of evidence was rated using GRADE. Due to heterogeneity and overlapping data sources (particularly from national registries), results were synthesized narratively following the SWiM guideline rather than pooled via meta-analysis.

Key Contributions and Results
The review identified 12 reports from four independent data sources (UK, US, Netherlands, Spain), comprising 10 studies on pancreas transplantation/utilization and 2 on islet isolation.

1. Pancreas Graft Survival (NRP vs. SRR):

   • UK Registry: In 47 NRP vs. 379 non-NRP DCD simultaneous pancreas-kidney (SPK) transplants, 1-year graft survival was 93.6% vs. 89.2% (adjusted HR for graft loss 0.56; 95% CI 0.17–1.80).
   • US Registry: 1-year graft survival after NRP was 91%, described as comparable to SRR.
   • Complications: No significant differences were found in early graft thrombosis, primary non-function, or re-operation rates. Peak lipase levels were lower in NRP groups compared to SRR in single-center cohorts.

2. Pancreas Utilization:

   • NRP was associated with numerically higher utilization. In the UK, the adjusted odds ratio for transplantation was 1.6 (95% CI 0.98–2.64). In a US propensity-matched analysis, utilization was 8.8% for NRP donors versus 2.2% for non-NRP DCD donors.

3. Comparison with DBD:

   • Unadjusted registry data indicated that pancreas graft survival and donation rates after NRP DCD were comparable to DBD.
   • Early injury markers (peak amylase/lipase) suggested NRP narrows the gap between SRR DCD and DBD grafts.

4. Islet Isolation:

   • Data from a single program (Leiden, Netherlands) showed that abdominal NRP (A-NRP) yielded a median of 471,739 Islet Equivalents (IEQ), significantly higher than DCD without NRP (218,750 IEQ) and DBD (206,522 IEQ).
   • Yield per gram of pancreas was also higher after A-NRP (4,290 IEQ/g) compared to controls.

5. Kidney Outcomes in SPK:

   • Kidney delayed graft function (DGF) was markedly lower in NRP groups (6% vs. 33% in US SPK recipients), consistent with known benefits of NRP for kidneys.

Certainty of Evidence
The authors emphasize that the certainty of evidence for every outcome was rated as "very low." This limitation stems from:

• The observational nature of all included studies.
• Significant risk of bias, particularly confounding by era, center expertise, and donor selection.
• Imprecision in effect estimates (wide confidence intervals).
• Overlap of data sources (e.g., multiple reports from the same registry).
• Lack of adjusted comparisons between NRP DCD and DBD.

Significance and Claims
The paper concludes that current evidence suggests pancreas graft outcomes after NRP DCD procurement are comparable to those after SRR DCD and DBD procurement, though this conclusion is tempered by very low certainty. The data do not support the hypothesis that NRP inherently compromises pancreas grafts; conversely, NRP may increase organ utilization and reduce kidney DGF in SPK recipients.

For islet transplantation, single-program data indicate a potential doubling of islet yield with NRP, though recipient outcomes by procurement technique remain unknown.

The authors assert that before practice recommendations can be made, pooled registry analyses with standardized outcome definitions are required. They specifically call for:

• Standardized definitions of pancreas outcomes (thrombosis, pancreatitis, PNF) in NRP registries.
• Development of pancreas-specific viability criteria during NRP (e.g., perfusate biochemistry trajectories).
• Side-by-side reporting of NRP DCD, SRR DCD, and DBD outcomes, including islet yield and recipient function.

The review maintains a modest stance, stating that while NRP does not appear to be inferior, the current evidence base is insufficient to declare it superior or to confirm equivalence with DBD without further high-quality comparative data.