More than 800 pancreas transplants have been performed in Korea since Professor Duck Jong Han conducted the first one in 1992 [1]. However, the number of pancreas transplants in Korea remains very low compared to that of kidney or liver transplants. In 2019, only seven transplant centers carried out pancreas transplants, while 80 centers performed kidney transplants [2]. The frequency of pancreas transplants in Korea has declined markedly since 2021, a trend exacerbated by the impact of coronavirus disease 2019 and the retirement of Professor Han [3].

Pancreas transplantation is the sole treatment that can maintain long-term euglycemia without requiring exogenous insulin injections [4]. Simultaneous pancreas and kidney transplantation (SPK) represents the only near-cure for patients with end-stage kidney disease who rely on insulin [5]. Pancreas after kidney transplantation (PAK) provides comparable survival benefits to recipients [6]. For certain individuals with uncontrolled diabetes, a pancreas transplant alone (PTA) offers advantages, even though it requires the use of immunosuppressants rather than insulin [7]. Pancreas transplantation is a crucial solid-organ transplantation that warrants attention. Furthermore, advancements in surgical techniques and immunosuppressive strategies have improved graft survival rates [8].

We successfully performed 100 pancreas transplants over a relatively short period of 9 years. Drawing on our expertise, we analyzed the situation of pancreas transplantation in Korea. Our goal was to share our initial experience to enhance pancreas transplant activities at other transplant centers across the country.

This study was approved by the Institutional Review Board (IRB) of the Pusan National University Yangsan Hospital (IRB No. 55-2024-086). The requirement for informed consent was waived by the IRB due to the retrospective design of the study, which did not involve any diagnostic or therapeutic intervention.

Study Design and Population

Between 2015 and July 2023, Pusan National University Yangsan Hospital performed 100 pancreas transplants. Of these, 27 patients underwent SPK or simultaneous pancreatic and living kidney transplant (SPLK), 23 patients underwent PAK, and 50 patients underwent PTA. These procedures accounted for approximately 20% of all pancreas transplants performed in Korea during this period (Fig. 1). We conducted a retrospective analysis to evaluate clinical outcomes and address several practical issues associated with solid-organ transplants.

Figure 1. The number of pancreas transplants performed in Korea between 2015 and July 2023. During the period, a total of 442 pancreas transplants were performed, of which 102 transplants were conducted at the Pusan National University Yangsan Hospital (PNUYH), Korea.

Operative Technique

Exocrine drainage

The majority of patients selected for pancreas transplants underwent duodenodenostomy for enteric drainage, totaling 86 patients, while the remaining 14 underwent duodenojejunostomy. We performed the first duodenoduodenostomy in Korea and continued with this technique until mid-2022 [9,10]. However, this operation proved more complex than duodenojejunostomy, and the advantage of performing endoscopic inspections through the graft duodenum was minimal. Therefore, we have recently transitioned to using duodenojejunostomy.

Venous outflow modification

The inferior vena cava (IVC) served as the recipient site for venous outflow anastomosis. Our technique focused on modifying the venous outflow during the bench procedure. Initially, the first six transplant cases were performed without any modifications. Subsequently, the next 15 patients received diamond-shaped patch plasty, followed by 31 patients who underwent fence angioplasty. Ultimately, 48 patients underwent aortic interpositional grafting [8,11,12]. The scheme of the pancreas transplant is illustrated in Fig. 2.

Figure 2. (A) Illustration of a duodenoduodenostomy performed for enteric drainage and an aortic interposition graft to the inferior vena cava for venous outflow. A kidney transplant was performed on the patient’s left side. (B) Duodenojejunostomy with ipsilateral kidney transplant. This is the latest version of our procedure. The kidney can be positioned on the left side if the right common iliac artery of the recipient has extensive calcification. D, duodenum; C, vena cava; A, aorta; AG, aortic graft. Illustrations created by Rune Horneland.

Immunosuppression

Antithymocyte globulin (Thymoglobulin, Sanofi; 5 mg/kg in five divided doses on postoperative days 0, 1, 2, 4, and 6) was administered to induce immunosuppression. The triple immunosuppressant regimen consisted of corticosteroids, mycophenolate mofetil (1.0–1.5 g/day), and tacrolimus, with a trough level maintained at 9–11 ng/mL during the first 3 months. Alongside routine monitoring of blood amylase and lipase levels, the steroid dosage (methylprednisolone) was tapered from 250 mg/day to 10 mg/day, and then further reduced to 4 mg/day over a 2-month period postsurgery. Initially, the target trough level for tacrolimus was set at 9 to 11 ng/mL for the first 3 months, followed by a target level of 8 ng/mL from 3 months to 1 year, and subsequently maintained within a range of 6 to 8 ng/mL.

Definition of Outcomes and Statistical Analysis

Pancreatic graft failure was defined as insulin dependency persisting for more than 2 months. Technical failure was defined as graft failure occurring within 3 months of transplantation, in the absence of an immunological cause. Hemorrhagic complications were defined as those requiring reoperation due to hemorrhage. Postoperative computed tomography revealed partial thrombosis, defined as thrombosis that did not obstruct the vessel.

Categorical variables were presented using absolute and relative frequencies. Quantitative variables were assessed based on the results of the Shapiro-Wilk test to determine if the data distribution was normal (mean±standard deviation [SD]) or non-normal (median, interquartile range [IQR]). The quantitative assay results were analyzed using the Kruskal-Wallis test or analysis of variance for three groups, and the Student t-test for two groups. Categorical data were examined using the Pearson chi-square test. The Kaplan-Meier method was employed to evaluate graft survival rates, and the log-rank method was used to compare survival rates across different groups. The post hoc Dunn test was conducted when significant P-values were obtained. The Fisher exact test was applied to events with low frequencies. Additionally, a Cox proportional hazards regression model was utilized to explore the relationship between various covariates and graft failure. This model included covariates such as recipient age, donor age, Pancreas Donor Risk Index (PDRI), preoperative hemoglobin A1c (HbA1c), waiting time, and transplant type (PTA and PAK compared to SPK), as well as sex (male compared to female). The Cox model was selected to handle time-to-event data and to adjust for potential confounders. The results are reported as hazard ratios (HRs) with 95% confidence intervals (CIs), accompanied by a P-value. Significant differences were identified with a P-value of less than 0.05. Graphical imaging and statistical analyses were performed using Python (Python Software Foundation).

Baseline Characteristics and Waiting Time for Grafts

Table 1 summarizes the baseline characteristics of pancreas transplant recipients across three groups. There were no significant differences in the body mass index of recipients or donors, sex distribution, donor HbA1c levels, cold ischemic time of the graft pancreas, PDRI [13], or the distribution of living donors among those who received SPK and PAK transplants.

Table 1. Baseline characteristics of pancreas transplant recipients and waiting time for transplantation

Variable	SPK group (n=27)	PAK group (n=23)	PTA group (n=50)	Total (n=100)	P-value
Recipient
Age (yr)	42.0 (35.5–48.5)	52.0 (44.8–59.3)	33.0 (23.8–42.3)	40.0 (29.0–51.0)	<0.001
Female sex	16 (59.3)	8 (34.8)	24 (48.0)	48 (48.0)	0.123
Body mass index (kg/m2)	22.1±1.9	22.6±2.8	22.5±2.7	22.4±2.5	0.497
Preoperative HbA1c (%)	7.0 (5.8–8.2)	8.1 (7.3–8.9)	9.4 (8.1–10.8)	8.2 (7.0–9.5)	<0.001
C-peptide <0.5 ng/mL	10 (37.0)	11 (47.8)	40 (80.0)	59 (59.0)	<0.001
Doses of insulin requirement (unit)	12.0 (3.2–20.8)	32.0 (18.8–45.3)	44.5 (34.9–54.1)	35.5 (21.5–49.5)	<0.001
Donor
Age (yr)	31.0±10.4	32.3±6.9	26.1±8.3	28.8±9.0	<0.001
Female sex	11 (40.7)	11 (47.8)	12 (24.0)	34 (34.0)	0.094
Body mass index (kg/m2)	22.7 (20.1–25.3)	22.5 (21.4–23.6)	22.3 (20.2–24.4)	22.6 (20.4–24.8)	0.623
HbA1c (%)	5.4±0.2	5.3±0.3	5.3±0.3	5.3±0.3	0.614
Pancreas cold ischemic time (min)	280.0 (253.5–306.5)	273.0 (247.5–298.5)	298.5 (264.0–333.0)	288.0 (259.8–316.3)	0.308
PDRI	1.23 (1.05–1.40)	1.13 (0.99–1.28)	1.08 (0.89–1.26)	1.13 (0.99–1.28)	0.116
Living kidney donor	10 (37.0)	12 (52.2)	-	22 (22.0)	0.430
Waiting time for pancreas graft (day; total number)	1,432 (1,161)	570 (641)	147 (160)	536 (819)	<0.001
Kidney from deceased donor	2,268 (642)	-	-	-	-
Kidney from living donor	178 (207)	-	-	-	-

Values are presented as median (interquartile range), number (%), or mean±standard deviation.

SPK, simultaneous pancreas and kidney; PAK, pancreas after kidney; PTA, pancreas transplant alone; HbA1c, hemoglobin A1c; PDRI, Pancreas Donor Risk Index.

The PTA group exhibited distinct characteristics when compared with the SPK and PAK groups. The preoperative HbA1c level in the PTA group was higher, and the C-peptide level was lower. Additionally, recipients in this group were younger and had a shorter waiting time. The donors for the PTA group were also younger. Insulin requirements were higher in the PTA group. Superior kidney function observed in the PTA group may have facilitated the normalization of insulin metabolism, potentially explaining the increased insulin doses required by this group.

The waiting time for the graft pancreas revealed several critical issues. The waiting time for isolated pancreas transplants, such as PAK and PTA, was considerably shorter compared to the waiting time for SPK from a brain-dead donor, which exceeded 6 years. This suggests that recipients in need of only a pancreas, and not a kidney, have a greater likelihood of obtaining an ideal pancreas transplant. Furthermore, this indicates that a substantial number of pancreases are discarded and go unused in Korea.

Clinical Results

The analysis of surgical complications, rejection episodes, and graft failure incidents among pancreas transplant recipients in the SPK, PAK, and PTA groups revealed significant findings (Table 2). The groups experienced various postoperative complications that required surgical or radiological intervention. Although there was no statistically significant difference (P=0.310), hemorrhagic complications were noted in 7.4%, 21.7%, and 12.0% of the patients in the SPK, PAK, and PTA groups, respectively. Other surgical complications, such as adhesions, wound dehiscence/incisional hernia, technical failure, and duodenal graft leakage, were rare and showed no significant differences between groups.

Table 2. Summary of surgical complications, rejections, and graft failure episodes

Variable	SPK group (n=27)	PAK group (n=23)	PTA group (n=50)	Total (n=100)	P-value
Postoperative complication necessitating radiologic or surgical intervention
Bleeding	2 (7.4)	5 (21.7)	6 (12.0)	13 (13.0)	0.310
Graft duodenal leakage	0	0	0	0	-
Adhesion	0	0	1 (2.0)	1 (1.0)	>0.999
Wound dehiscence/incisional hernia	1 (3.7)	0	1 (2.0)	2 (2.0)	>0.999
Technical failure	1 (3.7)	0	1 (2.0)	2 (2.0)	>0.999
Thrombotic graft failure	0	0	1 (2.0)	1 (1.0)	0.279
Partial thrombosis	3 (11.1)	6 (26.1)	13 (26.0)	22 (22.0)	-
Biopsy-proven acute rejection (pancreas)	2 (7.4)	2 (8.7)	9 (18.0)	13 (13.0)	0.328
TCMR	2 (7.4)	2 (8.7)	8 (16.0)	12 (12.0)	0.464
AMR	0	0	1 (2.0)	1 (1.0)	-
Biopsy-proven acute rejection (kidney)	0	2 (8.7)	-	2 (2.0)	-
TCMR	0	1 (4.3)	-	1 (1.0)	-
AMR	0	1 (4.3)	-	1 (1.0)	-
Pancreas failure	3 (11.1)	3 (13.0)	15 (30.0)	21 (21.0)	0.086
Retransplant	1 (3.7)	0	4 (8.0)	5 (5.0)	-
Graftectomy	1 (3.7)	2 (8.7)	1 (2.0)	4 (4.0)	-
Death with functioning graft	1 (3.7)	1 (4.3)	1 (2.0)	3 (3.0)	-
Kidney failure	2 (7.4)	2 (8.7)	-	4 (4.0)	>0.999
BK virus nephropathy	1 (3.7)	1 (4.3)	-	2 (2.0)	-
Death with functioning graft	1 (3.7)	1 (4.3)	-	2 (2.0)	-

Values are presented as number (%).

SPK, simultaneous pancreas and kidney; PAK, pancreas after kidney; PTA, pancreas transplant alone; TCMR, T cell-mediated rejection; AMR, antibody-mediated rejection.

The rates of partial thrombosis in the SPK, PAK, and PTA groups were 11.1%, 26.1%, and 26.0%, respectively. However, partial thrombosis did not affect the technical success of the graft. While no cases of thrombotic graft failure were reported in the SPK and PAK groups, only one case was observed in the PTA group.

Variations in acute pancreatic rejection were noted across the groups, confirmed via biopsy. In the PTA, PAK, and SPK groups, 18.0%, 8.7%, and 7.4% of the patients experienced acute rejection, respectively (P=0.328). Antibody-mediated rejection and T cell-mediated rejection of the grafted kidney were only observed in the PAK group, with no cases reported in the SPK cohort.

In the PTA group, the rate of pancreatic failure was significantly higher (30.0%) compared to the SPK (11.1%) and PAK (13.0%) groups, with the P-value approaching significance at 0.086 (log-rank test). There were no retransplantations in the PAK group, whereas one patient in the SPK group and four in the PTA group underwent this procedure. Graftectomies were performed once in the SPK group, twice in the PAK group, and once in the PTA group. Each cohort reported a single death related to functional grafting: an anesthetic accident in the SPK group, septic shock in the PAK group, and malignancy in the PTA group. The PTA group also experienced one technical failure where, despite intraoperative Doppler ultrasonography, venous flow did not resume after reperfusion. Graftectomies were performed on the day of transplantation, suggesting that the problem lay with the graft itself, particularly an obstruction in the microcirculation.

Kidney failure was observed in 7.4% and 8.7% of the patients in the SPK and PAK groups, respectively (P>0.999). BK nephropathy and death with a functioning kidney graft occurred only once in the SPK and PAK groups.

Graft Survival Rates

Graft survival rates are depicted in Fig. 3. The pancreatic graft survival rates were 91.0% at 1 year, 78.5% at 5 years, and 75.4% at 9 years. The survival rate following PTA was numerically lower than those observed after SPK or PAK. However, the difference in survival rates between SPK/PAK and PTA was not statistically significant (P=0.059) (Fig. 3C).

Figure 3. Kaplan-Meier survival curves. (A) The graft survival rates of pancreas transplants performed at Pusan National University Yangsan Hospital. The 1-, 5-, and 9-year graft survival rates were 91.0%, 78.5%, and 75.4%, respectively. Graft survival was defined as insulin independence. (B) The graft survival rates of SPK, PAK, and PTA. The 1-, 5-, and 9-year graft survival rates were 92.3%, 87.9%, and 87.9% for SPK, 87.0%, 87.0%, and 87.0% for PAK, and 91.7%, 69.2, and 64.6% for PTA, respectively. (C) The graft survival rate between SPK/PAK versus PTA. With borderline statistical significance (P=0.059, log-rank test), PTA had a lower graft survival rate. SPK, simultaneous pancreas and kidney; PAK, pancreas after kidney; PTA, pancreas transplant alone.

We conducted a Cox proportional hazards analysis to assess the influence of various factors on graft failure. The covariates we considered were age, donor age, PDRI, HbA1c level, waiting time, type of transplant (PTA and PAK compared with SPK), and sex (male vs. female). Our findings indicated that none of these variables significantly predicted graft failure, as all had P-values greater than 0.05 (Table 3).

Table 3. Cox proportional hazards model for graft failure

Covariate	Coefficient	Hazard ratio (95% CI)	P-value
Age	0.00	1.00 (0.97–1.04)	0.78
Donor age	0.01	1.01 (0.94–1.08)	0.83
PDRI	–0.03	0.97 (0.14–6.60)	0.98
Preoperative HbA1c (%)	–0.01	0.99 (0.80–1.23)	0.92
Waiting time	0.00	1.00 (1.00–1.00)	0.97
PTA (vs. SPK)	0.77	2.16 (0.54–8.57)	0.27
PAK (vs. SPK)	–0.72	0.49 (0.08–3.03)	0.44
Male (vs. female)	0.11	1.12 (0.38–3.27)	0.84

CI, confidence interval; PDRI, Pancreas Donor Risk Index; HbA1c, hemoglobin A1c; PTA, pancreas transplant alone; SPK, simultaneous pancreas and kidney; PAK, pancreas after kidney.

The number of organ donations following brain death in Korea has significantly increased, yet it still falls short of meeting the demand for organ transplantation [14]. For example, in 2022, the waiting period to receive a kidney from a brain-dead donor surpassed 6 years [15]. Despite these challenges, our experience suggests that the issue of organ scarcity has not been adequately addressed in the context of pancreas transplantation. The prolonged wait for SPK transplants was more due to the limited availability of kidneys from brain-dead donors than to the availability of pancreas grafts. Recipients with living kidney donors typically underwent successful pancreas transplants within 6 months. Furthermore, the waiting time for PTA was under 5 months. In 2022, there were 169 brain-dead donors aged between 19 and 49 years who were potential candidates for pancreas donation. However, despite this pool of potential donors, only 31 pancreas transplants were carried out that year. By comparison, 677 kidneys and 342 livers were transplanted in the same timeframe [15]. These statistics indicate that, unlike kidneys or livers, there was no shortage of pancreas grafts in Korea for recipients requiring only a pancreas transplant.

By 2022, only seven transplant centers in Korea had performed at least one pancreas transplant. Although the number of brain-dead donors remained stable, the number of pancreas transplants declined [15]. This indicates that many viable pancreases were discarded. Transplant surgeons strive to maximize the use of potentially transplantable organs. However, due to low activity in pancreas transplantation, the field is experiencing a severe crisis. Consequently, a more feasible and simpler surgical technique is needed to address this issue and increase the utilization rate of pancreatic grafts.

The first duodenoduodenostomy for enteric drainage in Korea was performed to facilitate the monitoring of immunity by allowing endoscopic observation of the duodenal mucosa of the graft [10]. Although a previous study indicated that signs of pancreatic graft rejection are not detectable in the duodenal mucosa [16], specific signs of other pathologies were observed in the graft's duodenal mucosa. Furthermore, there were no instances of immediate postoperative leakage or leakage associated with graft failure. Positioned retrocolically, duodenoduodenostomy reduces the risk of postoperative adhesive complications and has been established as a safe procedure [9,17].

Although duodenoduodenostomy offers several advantages, it also has limitations. This procedure necessitates a larger incision in the upper abdomen for transplantation and poses a risk of leakage if graftectomy becomes necessary, although the need for graftectomy is rare. We encountered one case of leakage following graftectomy for posttransplant lymphoproliferative disease. Owing to these issues, we have recently shifted to using duodenojejunostomy for enteric drainage.

Modifying the venous outflow is a critical aspect of our procedure. Thrombotic graft failure is the primary cause of early graft failure [18]. As previously reported, the transplanted pancreas is intrinsically thrombogenic [8–12]. Partial thrombosis often occurs at the ends of splenic vessels or mesenteric roots. However, a partial thrombus that did not impede blood flow did not affect the graft survival rate [19]. Over time, the approach to outflow modification has evolved at our center. The final version of the procedure involved an aortic interposition graft, which proved to be simple, technically feasible, and easily applicable [11]. The venous anastomotic location was consistently the IVC, regardless of the technique used. The IVC is an ideal site for preventing thrombotic graft failure [20]. We strongly recommend using an aortic interposition graft and the IVC as the venous outflow site for a pancreas transplant, particularly for centers starting to perform this procedure. We encountered only one instance of thrombotic graft failure, which was related to the condition of the graft.

Previously, the graft survival rate for isolated pancreas transplants such as PAK or PTA was significantly lower than that for SPK [21]. However, recent reports indicate that the survival rates for patients with PAK are now comparable to those for patients with SPK [3,6,22]. In this study, the outcomes for both SPK and PAK were similar, and both groups achieved excellent results. PAK may be the most suitable option for recipients who either have a living kidney donor or have already undergone a kidney transplant, especially since the waiting period for SPK is excessively long in Korea. SPLK offers an additional alternative for these patients; however, the transplant team often faces significant challenges due to the complexity of the procedure, which requires the coordination of three teams for pancreatic retrieval, kidney procurement from a living donor, and recipient surgery.

One of the most notable findings of this study was the noninferiority of the PTA outcomes, despite PTA traditionally being considered to have a lower graft survival rate compared to SPK [4,5,7]. Although acute rejections were more common in the PTA group, the difference was not statistically significant. Moreover, the graft survival rate following PTA was comparable to that observed after SPK or PAK. In the initial phase of the study (until 2019), the PTA cohort experienced a high number of acute rejections. In response, we revised our immunosuppressive protocols and introduced an early detection method for acute rejection in PTA patients [23], which led to a reduction in the incidence of acute rejections. Globally, the frequency of PTA is declining due to various factors, including advancements in insulin delivery systems, continuous glucose monitoring devices, and successful outcomes of islet transplants [3]. However, in Korea, the range of alternative treatment options remains limited, with pancreas transplantation being a notable exception. Therefore, PTA could represent a highly significant treatment option, provided that recipients are carefully selected.

This study has several limitations, including its retrospective nature and the fact that it is based on the experience of a single center, which may introduce certain biases. Despite these limitations, we hope that this study will contribute to the further development of pancreas transplant programs in Korea.

Conflict of Interest

No potential conflict of interest relevant to this article was reported.

Funding/Support

This study was supported by the Research Institute for Convergence of Biomedical Science and Technology, Pusan National University Yangsan Hospital and research grant from the Korean Society for Transplantation (2024-00-01002-004).

Author Contributions

Conceptualization: JHR. Formal analysis: KHY. Funding acquisition: BHC. Investigation: KHY, SHJ. Visualization: RH. Writing–original draft: JHR. Writing–review & editing: all authors. All authors read and approved the final manuscript.