Extending normothermic regional perfusion to pancreas recovery for clinical islet transplantation

Summary

Normothermic regional perfusion (NRP) is a strategy to improve organ quality in donation after circulatory death by restoring near-physiologic perfusion after circulatory arrest. Although its benefits are established in liver and kidney transplantation, its application to pancreas recovery and clinical islet transplantation is limited. Given the pancreas sensitivity to warm ischemia and the limited functional reserve inherent to islet transplantation, NRP may offer advantages over conventional cold recovery. We report our initial experience utilizing NRP for pancreas recovery in islet transplantation. Islet isolation yielded a high islet mass with excellent viability despite low donor body mass index, a factor traditionally associated with poor yield. Transplantation resulted in rapid improvement in glycemic control, reduced insulin requirements, resolution of hypoglycemia and favourable early graft function. This experience supports further evaluation of NRP to expand pancreas grafts utilization from donation after circulatory arrest and improve islet transplant outcomes.

Normothermic regional perfusion (NRP) is a validated strategy to refine donation after circulatory death (DCD), shifting the paradigm from passive preservation to active in situ reconditioning. Although standard cold flush techniques focus on slowing metabolic decay, NRP restores near-physiologic perfusion after circulatory arrest, allowing for both functional assessment and recovery from warm ischemic injury. Large clinical series report higher DCD organ utilization and improved outcomes when NRP is used, including a significant reduction in nonanastomotic biliary strictures in liver transplantation, and lower rates of delayed graft function and improved 12-month estimated glomerular filtration rate in kidney transplantation.¹ Although NRP has not yet definitively proven its advantages in pancreas transplantation compared with standard DCD, evidence suggests significantly higher utilization and a trend toward better 1-year graft survival when NRP is employed.² Given the sensitivity of the pancreas to hypoxic injury, the benefits established in other abdominal organs suggest that NRP may mitigate the injury to the endocrine cells. Although this advantage may not be readily apparent in whole-pancreas transplantation because of the large functional islet reserve transferred, it may become more evident in islet transplantation, where islet mass is closer to the minimum required for insulin independence and losses during procurement and isolation have greater clinical impact.

By restoring physiologically regulated circulation, NRP enables active reperfusion of the pancreas after circulatory arrest and before cold preservation. Warm ischemia is associated with rapid depletion of adenosine triphosphate, failure of membrane ion pumps, intracellular calcium accumulation, and activation of proteolytic pathways. Experimental and clinical application of NRP in abdominal organ procurement has shown restoration of aerobic metabolism during perfusion, including lactate clearance and recovery of cellular energy homeostasis.³ Given the pancreas’ susceptibility to ischemia-related microvascular injury, re-establishing oxygenated perfusion before cold preservation may attenuate microcirculatory impairment and limit the severity of ischemic injury before cold ischemia. Beyond metabolic recovery, NRP also confers a form of physiologic preconditioning³ that may be particularly relevant to islet isolation. The isolation process imposes substantial enzymatic and mechanical stress, inducing oxidative and inflammatory injury; although the cellular pathways underlying pancreatic preconditioning during NRP are not fully defined, it may trigger protective signalling pathways that enhance cellular resilience during subsequent processing. Supporting these concepts, recent European data show higher islet yields from pancreases recovered using NRP compared with conventional controlled DCD recovery, with post-transplant islet function comparable to donation after brainstem death (DBD).⁴

We applied NRP in a recent case at our centre involving a 38-year-old donor with irreversible hypoxic–ischemic brain injury. For donors whose families consent to the NRP retrieval technique, withdrawal of care occurs in the operating room, ensuring no contact between the family and the procurement team. During the agonal phase (defined as the time when systolic blood pressure drops below 50 mm Hg), 30 000 units of intravenous heparin was administered. The donor subsequently reached asystole and, after a standard 5-minute no-touch period, abdominal NRP was established via a surgical approach. We performed intra-abdominal cannulation of the aorta and the inferior vena cava through a rapid thoracolaparotomy; the descending thoracic aorta was cross-clamped below the left subclavian artery and, to ensure exclusion of cerebral circulation, we placed a venting cannula in the aortic arch as an additional safeguard against cerebral perfusion. The donor also underwent neuromonitoring as part of a research protocol. The cannulas were connected to a Spectrum Medical Quantum Perfusion module (Spectrum Medical Inc), and the circuit was primed with a crystalloid solution containing 30 000 units of heparin, supplemented by effluent blood from the donor. Functional warm ischemic time accounted for 33 minutes, and NRP was consequently initiated and maintained for 124 minutes. We regulated systemic oxygenation via the incorporated dual membrane oxygenator, with sweep gas flows adjusted to maintain physiologic PaO₂ and PaCO₂ levels. During this phase, we visually and manually assessed the pancreas, which appeared well perfused with no evidence of notable fibrosis, fat infiltration, or pancreatitis. Perfusate monitoring showed favourable metabolic trends, characterized by a progressive decline in lactate and stable glucose levels, indicating active hepatic glycolysis and subsequent splanchnic metabolic stabilization. Both kidneys and the liver were also retrieved and successfully transplanted at our centre.

Islet isolation from this pancreas yielded 400 000 islet equivalents (IEQ), corresponding to 7312 IEQ/kg of recipient body weight, with 92% viability. These parameters exceed current clinical recommendations for a single infusion, which suggest a minimum of 3500 IEQ/kg and viability greater than 80%.⁵ Compared with a previous case series of conventional DCD isolations in age-matched cohorts,⁶ our yield of 400 000 IEQ was numerically comparable but was achieved in a donor with a substantially lower body mass index (BMI; 16). For reference, the same series reported a mean yield of 480 017 IEQ from DBD donors with an average BMI of 26.⁶ In standard isolation practice, a low BMI is a recognized negative predictor for success as islet mass generally correlates with donor weight;⁷ indeed, pancreases from such donors would traditionally be declined for islet transplantation on this basis. The achievement of more than 7000 IEQ/kg suggests that NRP may have contributed to a better preservation of islet mass, potentially mitigating the apoptotic attrition that can occur during the agonal phase in donors with limited metabolic reserves.

This isolation was used for a 69-year-old recipient with type 1 diabetes and hypoglycemia unawareness (Clarke score of 6). Basiliximab was administered for induction, and maintenance immunosuppression was continued with tacrolimus and mycophenolate mofetil. Etanercept was administered before transplant and on post-transplant days 3, 7, and 10, as per our centre’s standard of practice. The recipient’s periprocedural course was uneventful. She reported no abdominal pain, resumed oral intake on the day of the procedure and was discharged on postprocedure day 2. Within 4 weeks, her total daily dose of insulin decreased from approximately 21 units to 6 to 7 units, and her C-peptide level increased from 184 pmol/L to 832 pmol/L. Continuous glucose monitoring demonstrated near resolution of hypoglycemia, which was her main pretransplant concern, with time in range (3.9 mmol/L to 10.0 mmol/L) increasing from 74% to 96%, and time below range (< 3.9 mmol/L) decreasing from 24% to 3% (Figure 1). Four weeks after her transplant, her BETA-2 score was 15.0, consistent with favourable early graft function.⁸ This represents one of the first applications of NRP to islet transplantation in North America.

Fig. 1.

Twenty-four–hour glucose trend in the 2 weeks (A) before and (B) after islet transplantation.

The recovery of highly viable islets from a DCD donor after a considerable agonal period provides further support for broader adoption of NRP in DCD pancreas procurement for islet transplantation. Although the pancreas currently lacks validated, organ-specific biomarkers to predict islet potency during the perfusion phase, the potential regenerative benefits of NRP on pancreatic tissue seem non-negligible, and islet transplantation may be the occasion to demonstrate its utility compared with whole-organ transplantation. As is now increasingly clear in liver transplantation, NRP may provide an opportunity to standardize the quality of DCD grafts, narrowing the functional gap between DCD and DBD donors. Future work should focus on prospective, standardized data collection across centres to further define the role of NRP in routine clinical practice. This case supports the surgical feasibility of the technique and provides a rational framework toward expanded pancreas utilization in DCD donors. Given the growing proportion of DCD donors in Canada, with medical assistance in dying contributing to this increase and representing a valuable source of grafts for islet transplantation,⁹ NRP may represent a practical and valuable strategy to expand the pool of pancreases available for islet isolation and to enhance graft quality and transplant outcomes compared with conventional DCD.