Reemerging kidney paired donation in Korea: simultaneous two-way swaps to overcome human leukocyte antigen and anatomic barriers
Article information
Abstract
Background
South Korea first implemented kidney paired donation (KPD) to overcome ABO-incompatible living-donor kidney transplantation. Although utilization declined as desensitization matured, persistent desensitization failure and anatomic constraints have renewed interest in KPD within a regulated framework. We report a recent single-center experience.
Methods
KPD exchanges were coordinated within an institutional registry under national supervision. We retrospectively reviewed six consecutive, simultaneous two-way exchanges (12 recipients) from 2022 to 2025. We assessed predefined desensitization failure or anticipated failure with the intended donor, compared immunologic compatibility between intended and exchange donors, and recorded perioperative desensitization, graft function, delayed graft function (DGF), and biopsy-proven acute rejection (BPAR).
Results
Five of six exchanges were undertaken for immunologic reasons. All five met criteria for desensitization failure or anticipated failure to the intended donor, with positive crossmatch and high donor-specific antibody (DSA) burden. Reassignment to the exchange donor reduced risk, lowering maximum DSA mean fluorescence intensity (MFI) from 13,228 to 1,350 and total DSA MFI from 20,542 to 2,240, although selective desensitization was still required in some recipients with residual human leukocyte antigen or ABO incompatibility. The remaining exchange addressed an anatomic constraint (adult–pediatric size mismatch). At a median follow-up of 12.2 months (range, 0.9–42.1 months), no DGF or BPAR occurred; all grafts were functioning, and the median estimated glomerular filtration rate was 66.8 mL/min/1.73 m2.
Conclusion
KPD is a safe alternative when desensitization fails or is predicted to fail, including in highly sensitized and anatomically complex candidates. These findings support renewed implementation of KPD alongside selective desensitization in Korea.
Introduction
The prevalence of end-stage kidney disease is increasing in Korea, and the gap between organ supply and demand remains substantial. Kidney transplantation (KT) provides superior patient survival, quality of life, and cost-effectiveness compared with dialysis [1]. Yet, as of 2024, the national waiting list for deceased-donor KT had expanded to 35,707 candidates, while only 644 deceased-donor transplants were performed that year. This widening disparity underscores the urgent need to expand timely access to living donor transplantation.
Owing to the persistent shortage of deceased donors, Korea has long relied on living-donor KT (LDKT), which accounts for approximately two-thirds of all KTs in recent years [1,2]. Against this backdrop, South Korea was the first country to perform kidney paired donation (KPD) in clinical practice in 1991, initially implemented to circumvent ABO blood group incompatibility [3–5]. As desensitization protocols matured, ABO-incompatible LDKT became feasible and routine, leading to a decline in KPD. More recently, several compelling factors have reemerged in support of KPD’s revival. First, improved graft and patient survival has created a growing retransplantation population, many of whom are highly sensitized and face limited opportunities for deceased-donor transplantation [6]. In Korea, retransplant candidates and highly sensitized patients—typically those with two or more positive human leukocyte antigen (HLA) crossmatches against different deceased donor offers—receive additional priority in deceased-donor kidney allocation. However, these patients with broad anti-HLA profiles often have crossmatches predicted to be positive with potential deceased donors. Consequently, organ offers cannot proceed, and their effective waiting time continues to increase despite their priority status. Second, for some candidates, desensitization may be unsafe or ineffective because of very high preexisting donor-specific antibody (DSA) levels, persistently positive crossmatch tests, unfavorable antibody kinetics, or prior failed desensitization. Third, anatomic challenges may also preclude direct living donation despite otherwise acceptable donor-recipient matching [7,8]. In these circumstances, KPD serves as a complementary strategy rather than a replacement for desensitization.
Several countries have established nationally coordinated KPD programs to demonstrate the value of structured allocation systems. Examples include the Dutch national program with algorithm-driven allocation, the United Kingdom’s NHS Blood and Transplant Living Kidney Sharing Scheme with quarterly matching runs, and the OPTN/UNOS KPD program in the United States [9–11]. These programs have shown that systematic matching and standardized logistics can widen access to transplantation for highly sensitized patients while safeguarding donors and recipients. Building on these precedents and responding to the growing cohort of Korean candidates for whom desensitization is unsafe or ineffective, our center has operationalized a KPD pathway designed to complement desensitization.
Here, we report a single-center case series of six consecutive KPD exchanges (12 patients) performed between 2022 and 2025. We describe the clinical scenarios prompting KPD encompassing immunologic and anatomic barriers, our matching workflow, selective desensitization when indicated, and short-term graft outcomes. By outlining a pragmatic, center-specific approach aligned with national oversight by the Korea Network for Organ Sharing (KONOS), our experience aims to inform future strategies for broader implementation and scale-up of KPD within Korea’s LDKT landscape.
Methods
Design, setting, and participants
We conducted a retrospective, single-center case series of consecutive KPD LDKTs performed between January 2022 and March 2025 at Seoul National University Hospital. Eligibility required: 1) an intended living donor-recipient pair evaluated during the study period; 2) immunologic barriers precluding direct LDKT, including a positive crossmatch with the intended donor and high immunologic risk defined operationally as total DSA mean fluorescence intensity (MFI) ≥10,000 or clinician-determined high-risk incompatibility not amenable to standard desensitization; and 3) anatomic or clinical barriers making direct donation less favorable, such as complex vascular anatomy, significant donor-recipient size mismatch, or other clinical contraindications to direct LDKT. The exclusion criteria were a lack of program consent for the use of clinical data and donor age <19 years. All eligible consecutive exchanges during the study period were included. Potential participants were identified during routine LDKT evaluations in nephrology and transplant surgery clinics or during pretransplant admissions, and consecutive patients who met these criteria and consented to inclusion in the institutional KPD registry were enrolled. All living donors underwent preoperative contrast-enhanced computed tomography (CT) to evaluate renal anatomy and vascular structures, and renal volumetry was additionally performed using a commercially available semiautomated software (OncoStudio, OncoSoft Inc.) to obtain total and cortical renal volume measurements [12,13].
Kidney paired donation workflow
All candidates underwent next-generation sequencing-based HLA typing (A, B, C, DR, and DQ) and single-antigen bead assays to define DSA profiles, which were reported as the maximum and total MFI. The maximum MFI represented the highest single-antigen bead value for any donor-specific HLA antibody, and the total MFI was the sum of all donor-specific bead signals. Pairs in the registry were matched using a stepwise protocol prioritizing: 1) the absence of DSA to the exchange donor; 2) ABO compatibility or no requirement for desensitization; 3) low-level DSA considered amenable to desensitization; and 4) other anatomic or clinical considerations (including marked donor-recipient size discrepancy). The match cycles were limited to a maximum of three pairs; however, in this case series, all exchanges were two-way swaps. Partner identities were not disclosed across pairs, and all donor and recipient surgeries were performed on the same day. When required against the exchange donor, desensitization therapy was individualized according to center practice and included rituximab (single dose, 375 mg/m2 for HLA incompatibility or 200 mg for ABO incompatibility) with or without plasmapheresis and intravenous immunoglobulin (0.1–0.4 g/kg per session) as determined by the attending nephrologist. Indications, treatment regimens, and the number of sessions were recorded.
Outcomes and statistical analysis
The primary outcomes were patient and graft survival and kidney function (serum creatinine level and estimated glomerular filtration rate [eGFR]) at the last follow-up. Secondary endpoints included delayed graft function or biopsy-proven acute rejection. Given the small sample size (six exchanges, twelve recipients), analyses were descriptive: mean ± standard deviation or median (range/interquartile range [IQR]) for continuous variables and number (%) for categorical variables. No inferential statistical tests were performed.
Ethical considerations
This study was approved by the Institutional Review Board of Seoul National University Hospital (No. H-2510-104-1684). All donors and recipients provided written informed consent for participation in the institutional KPD program and registry and for the research use of de-identified clinical data. The KPD program complies with the Declaration of Istanbul and World Health Organization Guiding Principles; all donations were voluntary, noncommercial, and operated under national oversight by KONOS. The exchange cases were approved by the Institutional Ethics Committees of Seoul National University Hospital and KONOS, and all surgeries were performed on the same day. All procedures adhered to the principles of the Declaration of Helsinki.
Results
Baseline characteristics of the study participants
Thirty-five incompatible donor-recipient pairs were enrolled in the central KPD registry. Recipient ages ranged from 12 to 76 years, and donor ages ranged from 28 to 68 years. Relationships within the registered pairs were predominantly spouses (74%), followed by parent–child (15%) and siblings (11%). Preenrollment sensitization spanned a wide range, with the calculated panel-reactive antibody class I and class II values ranging from negative to >90% for several candidates (Supplementary Table 1, available online).
During the study period, six simultaneous two-way KPD exchanges were completed, comprising 12 recipients (six males and six females) with a median age of 53 years (range, 12–67 years), including one pediatric patient (Table 1, Figure 1). The leading indication for KPD was HLA incompatibility (five/six pairs), whereas only a single exchange involved an anatomic constraint because of limited intra-abdominal space in the 12-year-old recipient. Eight recipients underwent their first KT, and four underwent second (n = 2) or third (n = 2) transplantation. Among retransplant recipients, the median dialysis interval after graft loss before KPD-LDKT was 3.8 years. Prior graft failure was attributed to chronic antibody-mediated rejection (n = 2), recurrent immunoglobulin A nephropathy (n = 1), or an unknown cause (n = 1). Dialysis exposure before KPD-LDKT ranged from preemptive to 12 years. Among non-preemptive recipients (n = 7), the median dialysis vintage was 5.5 years. Primary kidney diseases included immunoglobulin A nephropathy (n = 3), diabetic nephropathy (n = 2), primary hyperoxaluria (n = 1), autosomal dominant polycystic kidney disease (n = 1), and unknown or unspecified etiology (n = 5).
Individual characteristics and immunologic profiles of 12 kidney transplant recipients enrolled in paired exchange
Reduction in max DSA MFI after kidney paired donation.
Max DSA MFI directed against the original intended donor and the exchange donor for each transplant pair. Red bars indicate the max DSA MFI against the original donor, and orange bars indicate the max DSA MFI against the exchange donor. HLA specificities are color-coded by antibody class (Class I vs. Class II) to highlight class-specific immunologic barriers. Induction therapy used in each case (ATG or basiliximab) is indicated above the corresponding bars.
Max DSA MFI, Maximum donor-specific antibody mean fluorescence intensity; ATG, anti-thymocyte globulin; ABOi, ABO-incompatible; HLA, human leukocyte antigen; PPx, plasmapheresis.
All 12 living donors had an established relationship with their intended recipients: spouses (75%), parent–child (17%), and siblings (n = 1, 8%). The median donor age was 52 years (range, 28–67 years), and 75% of the donors were men.
Immunologic risk profile and early outcomes in the human leukocyte antigen-incompatible subgroup
Changes in immunologic risk before and after paired exchange
Within the 10 recipients in the HLA-incompatible subgroup, a complement-dependent cytotoxicity crossmatch to the intended donor was positive in three recipients. Flow cytometry crossmatching to the intended donor was positive in all 10, with an approximately even distribution between isolated B-cell positivity and combined T- and B-cell positivity. Against the intended donor, total DSA MFI ranged from 3,637 to >63,000, with eight exhibiting values ≥10,000. The maximum DSA MFI ranged from 2,563 to 23,399, with seven recipients exceeding 10,000, predominantly with Class II specificity.
By contrast, crossmatch to the exchange donor was negative in four recipients, B-cell-only positive in three, and both T- and B-cell positive in the remaining three. Against the exchange donor, the maximum MFI ranged from zero to 3,557, and the total MFI ranged from zero to 5,929. Two recipients had no detectable DSA against their exchange donors. Seven recipients were ABO-compatible with their exchange donors. The remaining three recipients underwent controlled ABO-incompatible transplants with initial isoagglutinin titers of 1:512, 1:256, and 1:512.
Desensitization and perioperative course
Desensitization strategies were tailored according to the incompatibility profile of each assigned exchange donor. Rituximab was administered to nine HLA-incompatible recipients, including all three ABO-incompatible recipients. Plasmapheresis was performed in seven recipients, with a median of nine sessions (range, 1–16), titrated to antibody titers. Induction therapy consisted of basiliximab or antithymocyte globulin, selected based on immunologic risk. Notably, in the first exchange pair, both recipients underwent desensitization attempts directed at their intended donors but failed. These pre-KPD treatments were not included in the plasmapheresis counts, which reflected the therapy targeting the exchange donor. This tailored desensitization allowed all recipients to undergo transplantation despite substantial preexisting immunologic barriers.
Early graft outcomes and safety
At discharge, all recipients were dialysis-independent with functioning grafts; serum creatinine ranged from 0.62 to 1.30 mg/dL and eGFR from 59.7 to 101.9 mL/min/1.73 m2. At the last follow-up, graft function remained favorable. The median serum creatinine was 1.00 mg/dL (IQR, 0.92–1.14 mg/dL; range, 0.86–1.43 mg/dL), and the eGFR had a median of 66.8 mL/min/1.73 m2 (IQR, 63.0–73.6 mL/min/1.73 m2; range, 41.5–92.3 mL/min/1.73 m2). No delayed graft function or acute rejection episodes were observed during follow-up.
Infectious events during follow-up included: 1) a 53-year-old woman with four episodes of culture-proven urinary tract infection over 16 months, predominantly Enterococcus faecium, treated with antibiotics; 2) another 53-year-old woman with BK polyomavirus viremia, managed with immunosuppression reduction and leflunomide; and 3) a 55-year-old woman with persistent cytomegalovirus viremia, managed with immunosuppression reduction, valganciclovir, and subsequently maribavir. No infection-related deaths, intensive care unit admissions, or infection-attributed graft losses occurred.
Pediatric kidney transplantation enabled by size-matched paired exchange
A 12-year-old boy with primary hyperoxaluria underwent multiple liver transplants with recurrent biliary interventions. He progressed to end-stage kidney disease; hemodialysis was initially delivered four times weekly via a tunneled catheter, and peritoneal dialysis was subsequently attempted; however, severe peritonitis led to conversion back to hemodialysis. Despite staged desensitization of the deceased donor pathway, class II HLA sensitization persisted. Ongoing dialysis became increasingly difficult to sustain, leading the family to pursue living donation from the father. However, the child had limited intra-abdominal space because of a markedly enlarged liver graft, splenomegaly, and a splenorenal shunt (Figure 2), and he harbored low-level donor-specific antibodies to his father (anti-HLA-DQ8, MFI 1,467). Therefore, KPD was used to secure a size-appropriate graft. KPD was also indicated because the adult recipient’s intended donor kidney was considered undersized, raising concerns about an inadequate nephron mass. Given the pediatric recipient’s markedly limited intra-abdominal space, size matching was a key consideration. The intended donor (the pediatric recipient’s father) had a relatively large kidney volume on CT-based volumetry (left vs. right: total renal volume, 167.8 mL vs. 140.1 mL; cortical volume, 112.0 mL vs. 99.0 mL), raising concern for size mismatch. In contrast, the exchange donor demonstrated marked renal asymmetry, with a smaller right kidney (9.0 cm), reduced right-sided split renal function on diethylenetriamine pentaacetic acid (right vs. left, 29.0% vs. 71.0%), and substantially lower right renal volume on CT-based volumetry (left vs. right: total renal volume, 189.5 mL vs. 92.9 mL; cortical volume, 146.6 mL vs. 75.1 mL). Accordingly, the smaller right kidney from the exchange donor was allocated to the pediatric recipient. In turn, the pediatric recipient’s father donated his kidney to the other adult recipient as part of the two-way paired exchange.
Space-limiting hepatosplenomegaly in the pediatric recipient prior to kidney transplantation.
Space optimization included preoperative splenic artery embolization and intraoperative splenectomy before pediatric transplantation. Postoperatively, the splenorenal shunt was coil-embolized (day 2), ureteroneocystostomy was reperformed with stenting (day 12), and the stent was removed at 1 month. Later, the child underwent endoscopic retrograde biliary drainage revision, developed carbapenem-resistant Pseudomonas aeruginosa bacteremia, and experienced one episode of urosepsis, with no infection-related readmissions after 1 year. At the last follow-up, the eGFRs were 48.0 and 52.0 mL/min/1.73 m2 in the child and adult recipients, respectively.
Post-donation kidney function and follow-up outcomes in living donors
All 12 living donors showed preserved renal function after nephrectomy. Median follow-up from donation to the most recent assessment was 13.7 months (IQR, 4.0–21.0 months). The median serum creatinine at discharge was 1.27 mg/dL (range, 0.83–1.60 mg/dL), and the median eGFR at discharge was 62.3 mL/min/1.73 m2. At the most recent follow-up, the median serum creatinine was 1.20 mg/dL (range, 0.85–1.64 mg/dL), and the median eGFR was 64.9 mL/min/1.73 m2.
Discussion
This single-center case series shows that KPD can be delivered safely under national oversight using simultaneous two-way swaps and uniform perioperative protocols. The predominance of HLA-incompatible pairs illustrates how KPD can downstage immunologic risk without extreme desensitization. In addition, the pediatric size-mismatch case highlights that KPD extends beyond immunologic indications, providing a practical solution to anatomical and technical constraints while preserving safety. Unlike many reports focusing primarily on immunologic incompatibility, our series also included an anatomical-only indication—a size-matched adult–pediatric swap—illustrating deliberate use of KPD in Korea to address technical constraints when direct related donation is unsafe.
Despite favorable clinical outcomes and high participant satisfaction with KPD programs, their wider adoption in Korea has been tempered by advances in desensitization protocols and the increasing analytical sensitivity of DSA assays, which together have reduced the practical barriers posed by ABO/HLA incompatibility. In addition, strict anonymity and limited bidirectional disclosure, which are required in the Korean KPD process, can make some candidates uneasy about receiving organs from a non-family donor and can complicate informed decision-making. Consequently, although Korea pioneered KPD early, national uptake was not consolidated and program activity remained limited for years.
Korea remains heavily reliant on LDKT, as the deceased-donor kidney supply has not kept pace with demand. National registry data indicate that living donation accounts for roughly two-thirds of total KTs with excellent contemporary outcomes [2]. This success has created a growing cohort returning for retransplantation, for whom sensitization is a central barrier contributing to the median waiting time for deceased-donor kidneys exceeding a decade and causing further delays for highly sensitized candidates [14]. As this cohort expands, the number of patients requiring desensitization has increased, renewing interest in KPD. Within this context, KPD is reemerging as a complementary pathway that can convert positive to negative crossmatches, reduce DSA burden, and de-risk transplantation without supplanting desensitization. We acknowledge that certain candidates—particularly those with blood type O and/or extreme sensitization—may have persistently low matching probability in KPD pools. In our experience, KPD has therefore been considered a pragmatic salvage option for selected candidates who had no feasible route to direct living donation despite active desensitization efforts.
Several countries operate centralized KPD frameworks with scheduled national match runs and standardized logistics (Table 2), including the UK NHS Blood and Transplant Living Kidney Sharing Scheme, the OPTN/UNOS program in the United States, the Dutch algorithmic program with virtual crossmatch [15], the Canadian interprovincial service [16,17], and ANZKX (Australia and New Zealand Kidney Exchange) in Australia/New Zealand [18]. Observational data indicate that, with appropriate selection and standardized processes, outcomes after KPD are comparable to those after standard LDKT, even when cold ischemia time is longer because kidneys are shipped [19–23]. By contrast, Korea’s recent activity has been largely single center rather than nationally coordinated; our series reflects this setting, using simultaneous two-way exchanges with local surgery and minimal ischemia. Although small and noncomparative, the absence of delayed graft function or biopsy-proven rejection and the preservation of early eGFR in our cohort align with international benchmarks, supporting clinical feasibility in the Korean context. Although this report cannot address nationwide implementation, it indicates that simultaneous two-way KPD is practically achievable under existing oversight and may be worth considering in selected cases. Larger, multicenter studies with long-term follow-up will be required to determine whether outcomes of KPD are comparable to those of conventional living donor transplantation or desensitization-based approaches.
Selected international KPD frameworks and common design choices
Without administering validated patient- or donor-reported instruments, we noted several recurring operational points from coordinator and physician documentation: counseling needs exceeded those for routine LDKT because of unfamiliarity; decision-making was smoother when KPD and desensitization were presented side-by-side with timelines and crossmatch/DSA scenarios; explaining that multiple post-virtual crossmatch visits might reduce cancellations; and same-day operations across pairs were valued for perceived fairness and safety. In our experience, alleviating anxiety often required repeated, in-depth counseling over multiple encounters with both coordinators and clinicians, particularly as many candidates were considering KPD in a practical salvage context with no feasible alternative. These observations are descriptive and constitute a limitation; prior qualitative work has reported high satisfaction and good psychosocial adjustment among KPD participants and increased willingness when processes are clear [24–27], whereas preferences regarding anonymity and optional, program-mediated contact vary across settings [28,29]. Korea’s gradual uptake likely reflects the established effectiveness of intrafamilial ABO-incompatible and HLA-incompatible LDKT and the prevailing anonymity framework [30,31], but our experience demonstrates that simultaneous two-way swaps under KONOS oversight are feasible and safe within these safeguards.
This study had several limitations. It was a retrospective, single-center series with a small sample size, short follow-up, and no contemporaneous comparator group such as incompatible LDKT managed solely with desensitization. The exchanges we report were all simultaneous two-way swaps performed locally; therefore, the findings may not be generalizable to multicenter chains or shipped-graft models. We did not systematically collect standardized patient- or donor-reported outcome measures, and several posttransplant events (including infectious complications) were captured descriptively rather than through prespecified adjudication.
Within these limitations, this series documents the reemergence of KPD in Korea as a practical clinical tool at the center level, reflecting renewed feasibility under existing national oversight rather than a formal nationwide implementation. Under national authorization and same-day surgery, KPD enabled transplantation across substantial HLA barriers and, in one case, an anatomic barrier (adult–pediatric size mismatch) that could not have been addressed by desensitization alone. In the context of Korea’s persistent organ shortage, prolonged waiting time for deceased-donor kidneys, and a growing pool of highly sensitized and retransplant candidates, these findings indicate that KPD can be delivered safely and can expand access to LDKT in contemporary Korean practice.
Supplementary Materials
Supplementary data are available at Kidney Research and Clinical Practice online (https://doi.org/10.23876/j.krcp.25.391).
Notes
Conflicts of interest
All authors have no conflicts of interest to declare.
Funding
This work was supported by a grant from the National R&D Program for Cancer Control, Ministry of Health and Welfare, Republic of Korea (RS-2023-CC137810).
Data sharing statement
The data presented in this study are available from the corresponding author upon reasonable request.
Authors’ contributions
Conceptualization: EK, HL
Data curation: EK
Formal analysis: EK
Investigation: EK, KOM, HL
Methodology: KOM, HGK, EYS, AC, SM, JH, SP, YCK, YSK, HL
Resources: KOM, HGK, YA, AC, AH, SM, JH, SP, YCK, YSK, HL
Supervision: HL
Writing–original draft: EK, HL
Writing–review & editing: EK, HL
All authors read and approved the final manuscript.