EXPERIMENTAL AND CLINICAL TRANSPLANTATION

Objectives: Pancreas transplant provides definitive treatment for uncontrolled insulin-dependent diabetes, yet use of donation after circulatory death pancreases remains limited due to concerns about higher complication rates and reduced graft survival versus donation after brain death. We (1) assessed circulatory death simultaneous pancreas-kidney transplants from 2015-2019 and 2020-2024, (2) reviewed donor/recipient characteristics, and (3) compared outcomes of circu-latory versus brain death donation in simultaneous pancreas-kidney transplants.
Materials and Methods: In a retrospective cohort analysis of all adult simultaneous pancreas-kidney transplants for 2015-2024 (n = 6934; United Network of Organ Sharing), we divided recipients as donation after circulatory death (n = 219) and brain death (n = 6715). We evaluated donor, recipient, and simultaneous pancreas-kidney transplant characteristics associated with graft survival.
Results: No significant difference was shown in the proportion of circulatory death transplants for 2015-2019 versus 2020-2024 (95% CI, -0.0125 to 0.004; P = .3). Circulatory death transplant recipients had a higher proportion of type 2 diabetes and were more likely to be treated at high-volume centers. Donation after circu-latory death transplants were more often from male, non-Hispanic White donors, with higher Kidney Donor Profile Indexes. Kidney delayed graft function was higher among circulatory death donors (26.50% vs 8.40%; P < .001). No significant difference in 5-year pancreas graft survival was observed between recipients of circulatory death and brain death transplants (hazard ratio = 0.88; P = .456). Higher recipient body mass index (hazard ratio = 1.04; P < .001), kidney delayed graft function (hazard ratio = 1.72; P < .001), and cytomegalovirus donor-positive/recipient-negative se-rostatus (hazard ratio = 1.29; P = .004) were indepen-dently associated with poorer graft survival.
Conclusions: Donation after circulatory death simul-taneous pancreas-kidney outcomes parallel results from brain death transplants, yet use remains low. Optimized donor management could expand the pancreas donor pool and enhance transplant access.

Key words : Donation after brain death, Graft survival, Transplantation outcomes

Pancreatic transplant is an established therapeutic option in individuals with type 1 or type 2 diabetes mellitus with uncontrolled diabetes including, but not limited to, level 3 hypoglycemia, recurrent diabetic ketoacidosis and severe glycemic lability. According to the Organ Procurement and Transplantation Network, candidates ≥18 years old can receive kidney-pancreas transplant if they meet both pancreas and kidney wait time criteria. Pancreas transplants can be performed as pancreas transplant alone (PTA), as pancreas-after-kidney transplant, or as simultaneous pancreas-kidney (SPK) transplant. The SPK transplant is the most common type of pancreas transplant performed worldwide.¹ Pancreas transplant offers insulin independence and mitigates progression of diabetic complications. Individuals who undergo SPK transplant have a 72% survival rate at 8 years, versus a 55% 8-year survival rate in patients with type 1 diabetes who undergo kidney transplant alone.²
Despite these details, the annual number of pancreatic transplants has declined during the past decade, from 954 in 2014 to 915 in 2023.^3,4 Impro-vements in alternative methods of glucose control, such as automatic insulin delivery systems and islet cell transplant, possibly contribute to this trend of fewer pancreas transplants. Another important factor to consider is a shortage of available pancreases for transplant. In 2023, there were 1876 adults on the pancreas transplant wait list and 1194 pancreas donors that same year.⁴ Mortality on the wait list in 2023 was 4.6 deaths per 100 patient-years.³
The shortage of donor pancreases has led to the consideration of expanding the criteria for acceptable donors. The utility of donation after circulatory death (DCD) pancreases has been limited due to concerns that warm ischemia time (WIT) after cardiac arrest may negatively affect graft survival. Siskind and colleagues in 2014 addressed this question by analyzing data from the United Network for Organ Sharing (UNOS) on all adult subjects who received pancreas and kidney-pancreas transplants between 1996 and 2012.⁵ They concluded that DCD and donation after brain death (DBD) have comparable outcomes and that DCD would increase the pool of pancreases available for transplant. In 2022, Gruessner and colleagues assessed the effect of DCD donors on PTA and SPK transplants for all transplants performed in the diabetic population in the United States between 2001 and 2020. They concluded that DCD pancreas transplants were not associated with an increased risk of complications or a decreased risk in organ survival.⁶
The objective of our article was to evaluate changes in DCD SPK transplants between 2015-2019 and 2020-2024 and to review characteristics and compare outcomes between DCD and DBD in SPK transplants during that same period.
Our study focuses on SPK transplants to provide a unique model by which to evaluate donor-related outcomes, because both pancreas grafts and kidney grafts are procured and transplanted concurrently under identical ischemic and immunological condi-tions. Pancreas-after-kidney transplant recipients have already undergone kidney transplant and would therefore have altered immune profiles that may affect graft viability. Candidates for PTA must demonstrate preserved renal function to tolerate the nephrotoxic effects of immunosuppression. Our focus on SPK transplant allows us to minimize confounding variables.

Materials and Methods

This study was exempt from institutional review board approval, because we evaluated publicly available data from the UNOS database. All adult patients (≥18 years old) who received SPK transplants from 2015 through 2024 were analyzed (n = 6934). Patients were divided into the following 2 categories: (1) DCD (n = 219) and (2) DBD (n = 6715).
Descriptive statistics (mean with SD, frequencies, and percentages for categorical values) were calcu-lated for donor, recipient, and SPK transplant charac-teristics associated with graft survival. We used the chi-square test and t test to compare the 2 categories for each of the categorical and continuous variables, respectively. We used Kaplan-Meier product limit estimates to calculate survival rates, along with 95% CI values derived from the Greenwood formula for standard error. We used a 2-proportion z test to compare the proportion of DCD transplants during 2015-2019 versus 2020-2024. A Kaplan-Meier survival curve was generated, and differences between DCD and DBD were assessed with a log-rank test. Unadjusted patient and graft survival rates were calculated at 1 year, 2 years, 3 years, 4 years, and 5 years. The results were more reliable when signifi-cance was detected, as only large effects could reach significance with such a small DCD group. However, these were less robust when results were nonsig-nificant, since modest true differences may have gone undetected.

When the periods 2015-2019 and 2020-2024 were compared, the proportion of DCD transplants increa-sed from 2.96% to 3.39% (P = .310) (Table 1). During the same periods, the average time from wait list to transplant for DBD recipients decreased significantly from 337.58 days (SD 7.28 days) to 314.38 days (SD 7.29 days) (P = .025). In contrast, the mean wait time for DCD recipients increased from 216.35 days (SD 27.04 days) to 289.55 days (SD 32.98 days) without reaching statistical significance (P = .089).
In terms of recipient characteristics, a higher pro-portion of DCD recipients had type 2 diabetes mellitus (32.88%) versus DBD recipients (22.33%;P < .001). Age distribution also differed significantly (P = .001) with fewer DCD recipients aged 18 to 39 years (32.88% vs 42.06% in DBD) and more DCD recipients ≥60 years old (6.85% vs 3.31% in DBD).The mean time to transplant was significantly shorter in DCD recipients (253.12 ± 317.28 days vs 326.74 ± 423.24 days for DBD recipients; P = .011). No significant differences were observed in sex, identity (ie, White, Black, Hispanic, Asian, other), body mass index (BMI, measured as kilograms per meter squared), dialysis status, calculated panel reactive antibody, or blood group (Table 2).
Donors after circulatory death were more likely to be male (76.71% vs 69.95%; P = .031), non-Hispanic White donors (79.45% vs 59.21%; P < .001), whereas Black or Hispanic donors were less represented in the DCD group. Cardiovascular/cerebrovascular causes of death were more common among the DCD group (58.02% vs 45.57% for DBD; P = .001). Trauma-related death was more frequent among the DBD group (54.12% vs 41.98% in DCD). Machine perfusion for the kidney was significantly lower in the DCD group (5.94% vs 39.54% in DBD; P < .001). Additionally, the DCD group had a significantly higher Kidney Donor Profile Index (KDPI) (0.28 ± 0.14 vs 0.13 ± 0.12 in DBD; P < .001), and differences were observed in donor/recipient cytomegalovirus (CMV) serostatus distribution (P < .001), with DCD donors more likely to have donor-negative/recipient-negative status (D-/R-) or donor-negative/recipient-positive status (D-/R+). There were no significant differences in donor age distribution (P = .119), donor BMI (P = .064), or human leukocyte antigen mismatch levels (P = .825) (Table 3).
We observed that DCD transplants were more frequently performed at high-volume centers (≥100 transplants per year), with 76.26% of DCD cases occurring at such centers versus 43.87% of DBD cases (P < .001). The DCD organs also had significantly longer preservation times (12.51 ± 4.58 hours vs 10.45 ± 4.60 hours for DBD organs; P < .001). Systemic venous drainage (SVD) was more common in DCD cases (98.16% vs 87.50% in DBD cases; P < .001). Portal venous drainage (PVD) was rare. Induction therapy patterns differed significantly (P < .001). The DCD recipients were less likely to receive anti-thymocyte globulin (ATG) (45.66% vs 70.17% in DBD) and more likely to receive alemtuzumab (32.42% vs 15.40% in DBD) or no induction (13.24% vs 7.07% in DBD). Steroid maintenance therapy was less common in DCD recipients (65.30% vs 72.15% in DBD; P = .026). No significant differences were observed in exocrine duct management approaches (P = .538) (Table 4).
In the multivariate Cox regression analysis for pancreas graft survival in SPK transplants, the ha-zard ratio (HR) for the DCD group was 0.88 (95% CI, 0.63-1.23; P = .456). Recipient BMI was significantly associated with graft failure (HR = 1.04 per unit increase; 95% CI, 1.02-1.05; P < .001). The donor-positive/recipient-negative (D+/R-) CMV status was significantly associated with worse pancreas graft survival versus the donor-negative/recipient-negative (D-/R-) reference standard (HR = 1.29; 95% CI, 1.08-1.53; P = .004). Kidney delayed graft function was markedly higher among DCD donors (26.50% vs 8.40% in DBD donors; P < .001) and was also associated with graft failure (HR = 1.72; 95% CI, 1.45-2.04; P < .001). The remaining variables showed no significant association with graft survival (Table 5). The Kaplan-Meier analysis also demonstrated similar mortality (log-rank test: P = .8673) between DCD and DBD subjects (Figure 1).

To our knowledge, this is the most recent compa-rative analysis of DCD versus DBD SPK transplant. Donations after brain death still account for >90% of pancreas transplants globally.^6,7 Increasing the use of DCD donors could significantly expand the donor pool and help reduce wait times. The previous reviews by Siskind and colleagues⁵ (in 2014) and Gruessner and colleagues⁶ (in 2022) demonstrated comparable outcomes between DCD and DBD pancreas transplants. Our analysis builds on this foundational work by focusing on SPK transplant temporal trends and identifying evolving patterns in recipient, donor, and overall transplant charac-teristics. We explored the effects of these trends on graft and patient survival outcomes. Using a large national dataset from the UNOS registry, this review also provided key clinical insights, such as shifting wait times for DCD recipients, and factors associated with graft survival.

Utilization and waitlist trends for recipients of donation after circulatory death transplants and recipients of donation after brain death transplants Younger recipients aged 18 to 39 years were more likely to receive a DBD pancreas versus those aged 40 to 59 years and ≥60 years. This may be due to a preference for allocation of a perceived higher-quality graft for a younger recipient with a longer projected survival time and, therefore, greater potential benefit. A 2022 analysis similarly observed that pancreas transplant recipients tended to be younger (mean age of 40 years) and were more likely to undergo SPK transplant.⁶
The use of DCD transplants is often regarded as a strategy to reduce wait times.⁷ A 2022 analysis re-ported that >50% of DBD recipients experienced wait times of >180 days, whereas DCD recipients more commonly received a transplant within 30 to 180 days.⁶ In addition, a smaller census review in 2023 noted average wait times of 334 days for DCD recipients versus 431 days for recipients of DBD organs.⁷ In contrast, our data suggested that the presumed advantage of shorter wait times for DCD recipients may not be consistent across time. For the period 2015-2019, recipients of DCD did in fact experience shorter average wait times (216.4 days) than DBD recipients (337.6 days). During the 2020-2024 period, this trend reversed: the average wait time for DCD recipients increased to 289.6 days, whereas the time for DBD recipients decreased to 314.4 days. Although this difference was not statistically signi-ficant, the shift is clinically relevant given the high mortality among patients awaiting SPK transplant.³
Our analysis did not demonstrate a significant increase in DCD SPK transplants (2.96% for 2015-2019 vs 3.39% for 2020-2024; P = .31). This highlights the systemic and/or logistical barriers that may limit the potential of DCD to meaningfully reduce time to transplant. Although DCD recipients initially demonstrated shorter wait list times (Table 1), persistently low numbers suggest that transplant practices are not fully utilizing this donor source. This is particularly important in light of increasing pancreas discard rates, which rose from 27% in 2018 to 56% in 2020 among donors aged 40 to 54 years. Low DCD organ prevalence and rising discard rates are likely contributing to wait list mortality.^3,7 Further initiatives, such as the SPIRIT Compliant study, may offer insight for reassessment of potentially discarded organs for possible viability, ultimately helping to expand the donor pool and improve transplant access.⁸

Donor characteristics
We observed that cardiovascular/cerebrovascular deaths were more prevalent among DCD donors. Trauma was more commonly reported among DBD donors as cause of death. This is in contrast to data for 2021 from a United Kingdom registry, which reported that DCD donors were more likely to have a cause of death other than cardiovascular/cerebrovascular events.⁹ These differences, coupled with comparable transplant outcomes, support the feasibility of expan-ding donor selection criteria for pancreas transplants.
The profiles of DCD and DBD donors differed across several parameters. Although previous analyses have reported that DCD donors are gene-rally younger and leaner (<50-55 years old, with BMI <30),^9,10 our data demonstrated no significant dif-ferences in donor age or BMI between DCD and DBD donors. The broader acceptance of donors with higher BMI may reflect an evolving trend toward more aggressive use of DCD organs, as transplant centers gain confidence in outcomes. The CMV seronegativity was more common among DCD donors, which reflects the more selective criteria. The DCD donors in our cohort were more frequently male, non-Hispanic White donors, with higher KDPI scores. Eligibility for DCD is a factor included in the KDPI calculation, and this factor inherently contributes to higher KDPI scores versus DBD. Kidney delayed graft function was also more prevalent in DCD organs versus DBD organs, which reflects the likely ischemic injury sustained with circulatory death.

Transplant characteristics
"There is considerable variation in transplant techniques and postoperative management across transplant centers. Our data demonstrated that DCD transplants are more frequently performed at high-volume centers, likely due to procedural complexity and substantial resources associated with these cases. Our data also aligned with the fact that individuals with type 2 diabetes mellitus are more likely to receive a DCD transplant, because the prevalence of this disease is higher in larger centers.
More than 98% of DCD pancreas transplants were performed using SVD, versus 87.5% for DBD pancreas transplants. These results are consistent with trends reported in the analysis by Gruessner and colleagues from 2022, which also demonstrated a predominant use of SVD across pancreas transplants. Although PVD more closely replicates normal physiological conditions, the use of PVD in pancreas transplants may be limited because of the increased technical complexity, including risks of anastomotic leaks, thrombosis, and severe intra-abdominal infections.¹¹ A comparative analysis published in 2001 reported superior pancreatic graft survival with PVD.¹² However, those results were later challenged by an updated UNOS analysis (for the period "1987-2016) that revealed no significant differences in patient survival or graft survival for SVD versus PVD.¹¹
Venous jump grafts are proposed to facilitate anastomosis or salvage grafts with a short portal vein. Nevertheless, an association has not been proved between the utility of venous jump grafts and improved patient survival or graft survival.¹³ Although venous jump grafts may theoretically introduce additional risks such as thrombosis, bleeding, infection, and anastomotic leak, Siskind and colleagues demonstrated no increased graft loss or mortality.¹³
Since the late 1990s, enteric drainage was the preferred method for exocrine duct management in pancreas transplants, primarily due to the high rates of urological and metabolic complications associated with bladder drainage. Furthermore, 10% to 40% of bladder-drained grafts eventually require subsequent conversion to enteric drainage.¹⁴ In our cohort, enteric drainage was used in 96.2% of DBD transplants and 97.2% of DCD transplants. Bladder drainage was employed in 3.3% of DBD cases and 2.8% of DCD cases. These trends are consistent with reports from and colleagues¹⁵ (in 2005) and Gruessner and colleagues⁶ (in 2022), who similarly demonstrated strong and sustained shifts toward enteric drainage in pancreas transplants across donor types.
Induction therapy with ATG was less commonly observed in DCD recipients versus DBD recipients, likely due to the perceived higher risk of infectious complications in the DCD population.¹⁶ In addition, ATG is associated with increased costs, the need for intensive monitoring, and a higher incidence of infusion reactions versus alternatives such as alemtuzumab, basiliximab, or no induction therapy.^17,18 These factors may contribute to reduced use of ATG in this setting.
A greater proportion of DBD recipients also received steroid maintenance therapy versus DCD recipients. There is currently no standard protocol to require different steroid management for DCD versus DBD recipients, and long-term rejection rates do not significantly differ regardless of steroids.¹⁹ Notably, a 2015 comparative analysis demonstrated that although steroid maintenance after pancreas transplant increased the risk of infections, no dif-ference or benefit was conferred in terms of graft survival or patient survival.²⁰

Graft survival outcomes and patient survival outcomes
The limited data and small sample size of DCD pancreas transplants are factors that create difficulty for comparisons with DBD pancreas transplants. In our cohort, no statistically significant difference in pancreas graft survival was observed between DCD and DBD recipients (HR = 0.98; 95% CI, 0.71-1.36; P = .910). The first large UNOS analysis also showed no significant difference in graft or recipient survival up to 15 years after transplant between DCD donors (n = 320) and DBD donors (n = 20448), although the small DCD sample did limit the statistical power of the study.⁵ An investigation from 2022 with a larger DCD cohort (n = 490) confirmed similar outcomes.⁶ Similarly, a meta-analysis of 638 DCD recipients in a study from 2017 reported no difference in overall patient survival or graft survival but reported higher risk of graft thrombosis (odds ratio = 1.67; 95% CI, 1.04-2.67; P = .006), which was not observed when donors received antemortem heparin (P = .62).²¹ A Canadian single-center analysis reported a com-parable 8-year survival rate in SPK transplant recipients, ie, recipient (87.4% for DCD vs 92.7% for DBD; P = .35), kidney graft (88.9% for DCD vs 96.9% for DBD; P = .219), and pancreas graft (77.4% for DCD vs 86.7% for DBD; P = .344).⁷ Our results complement the existing literature.
Beyond donor type, we also identified recipient characteristics that significantly influenced graft survival. Recipients with higher BMI experienced poorer pancreas graft survival. A retrospective review from 2021 by a team from Newcastle University presented an analysis of 22 years of transplant data and identified recipient BMI as an independent risk factor for posttransplant survival.²² Although no universal BMI cutoff was established, the UK National Health Service Blood and Transplant guidelines consider a BMI >30 to be a relative contraindication to transplant.²²
In addition, CMV serostatus significantly affected graft outcomes. Specifically, D+/R- recipients were independently associated with worse pancreas graft survival versus the D-/R- reference standard. This finding in our study aligned with a recent retros-pective analysis from 2024, which identified CMV D+/R- status as a significant risk factor for CMV infection and subsequent graft dysfunction.²³

Reported challenges associated with donation after circulatory death transplants and perfusion strategies to enhance donation after circulatory death transplant outcomes
Pancreas transplants from DCD grafts remain less common than pancreas transplants from DBD grafts, primarily due to concerns regarding cold ischemia time (CIT) and WIT, which leads to ATP depletion, endothelial injury, and microvascular dysfunction.²⁴ Unlike the DBD procurement process, which allows continuous organ perfusion until retrieval, DCD grafts are retrieved during a period of circulatory arrest. The UK registry data showed a median WIT of 26 minutes for DCD SPK transplant cases, with no significant difference in graft outcomes versus DBD SPK transplants.⁹ However, thresholds for acceptable WIT vary, with <60 minutes recommended for pancreas grafts by the British Transplantation Society.²⁵
In our analysis, WIT was excluded due to limited available data (n = 19). Duration of CIT also affects DCD graft viability. British guidelines recommend a CIT limit of <10 hours for DCD pancreases.²⁵ In our cohort, DCD organs had significantly longer pre-servation times than DBD organs (12.51 vs 10.45 hours; P < .001), and these preservation times did not significantly affect graft survival. The hemodynamic profile of DCD donors is another critical factor; donors treated with high-dose vasopressors may experience earlier circulatory arrest and prolonged ischemia. Froehlich and colleagues observed that vasopressors such as phenylephrine and norepinephrine were associated with better outcomes, whereas dopamine use or absence of vasopressors correlated with increased graft failure.²⁶
To address these challenges and improve DCD graft viability, various perfusion strategies are being explored. Normothermic regional perfusion (NRP) is an emerging technique that may improve DCD graft viability by restoring blood flow to abdominal organs via extracorporeal membrane oxygenation,²⁷ and NRP has reduced complications in kidney and liver transplants.^28,29 The evidence for the benefits of NRP in pancreas transplants is limited, despite its wides-pread adoption in many European countries. Bekki and colleagues discovered no significant differences in posttransplant outcomes between NRP DCD and non-NRP DCD. The NRP cohort included only 3 pancreas recipients, which limited the statistical power of that study.³⁰ Conversely, Owen and colleagues demonstrated that NRP can significantly reduce CIT in pancreas transplants; they reported a decrease from 10.2 hours in non-NRP donors to 9.7 hours in NRP donors.³¹
Machine perfusion techniques, including normot-hermic machine perfusion (NMP) and hypothermic machine perfusion (HMP), are also being inves-tigated. Unlike NRP, which involves in situ extra-corporeal membrane oxygenation, both NMP and HMP are ex situ methods. Promising results have also been demonstrated for NMP with porcine pancreases. Mazilescu and colleagues subjected pancreases to 3 hours of NMP, followed by transplant into animals. The 72-hour outcomes revealed intact islet cells and only mild signs of tissue injury.³² Success has been reported for HMP in porcine pancreases with evidence of higher oxygen pressures, as well as improved reperfusion.³³ Despite the potential benefits, these machine technologies remain underexplored in clinical practice, as shown in our analysis for which machine perfusion was strictly for the kidneys and was applied in only 5.94% of DCD transplants and 39.54% of DBD transplants.
The European Society for Organ Transplantation has emphasized the need for further research and standardization of these various preservation strategies (eg, vasopressors, NRP, NMP, HMP) to translate experimental successes into widespread clinical application.³⁴

Strengths and limitations
Our review presents a comprehensive and contem-porary analysis of pancreas transplantation using a large retrospective international cohort that includes both DCD and DBD donors. A key strength to our study is the detailed comparison of donor, recipient, and transplant characteristics. Robust statistical methods and clearly defined inclusion and exclusion criteria further strengthen the validity of our study.
The retrospective nature of our study inherently limits causal inference and is subject to potential selection bias and unmeasured confounding. The analyses herein rely on preexisting registry data; therefore, inaccuracies in data entry, such as the notably lower rates of induction therapy among DCD recipients, cannot be fully excluded and may reflect clerical or reporting inconsistencies.
The period of data collection overlapped with the COVID-19 global pandemic, which may have influenced the overall statistical patterns of DCD and DBD. Moreover, despite the analyses of national and international data, generalizability may still be affected by regional practice differences or unre-ported center-specific protocols.
Nonetheless, our study remains one of the most up-to-date and detailed evaluations of DCD pancreas transplants versus DBD pancreas transplants and provides valuable insights to inform future clinical practice and research.

Our findings showed that outcomes for DCD pancreas transplants are comparable to DBD pancreas transplants, most notably in graft survival and postoperative complications, but DCD pancreas transplants remain underused, possibly due to concerns about ischemia and logistical challenges. The frequency of use of DCD organs and the associated wait times have not significantly increased in recent years. In contrast, DBD wait times have improved, which suggests better graft allocation for DBD transplants versus DCD transplants. As organ shortages continue, the DCD pathway may serve as a solution to expand the donor pool and reduce wait list mortality. Clinicians should view the DCD pathway as a generally safe option, especially in high-volume centers, but must carefully assess risks such as high BMI and prolonged ischemia. Future efforts should focus on protocols standardization, improvement of donor and recipient selection, and advancement of preservation techniques to expand the donor pool and reduce wait list mortality.

Volume: 23 Issue: 12 December 2025

FULL TEXT