Kidney Donation After Circulatory Death Is Associated With Sustained Graft Function, Comparable Graft, and Patient Survival Authors: Alan L. Xu, Feruza Kholmurodova, Georgina L. Irish, Philip A. Clayton Published: 2026 ABSTRACT Background.Kidney transplantation is the preferred treatment for kidney failure, offering various benefits, yet donor supply remains limited worldwide, including in Australia and New Zealand. We examined long-term outcomes of kidney transplant recipients from donation after circulatory determination of death (DCDD) versus donation after neurological determination of death (DNDD) donors. Although DCDD kidneys carry a higher risk of delayed graft function (GF), long-term outcomes appear equivalent, but contemporary data on GF, particularly estimated glomerular filtration rate (eGFR), remain limited.Methods.Using the Australia and New Zealand Dialysis and Transplant (ANZDATA) Registry, we included adult kidney-only transplant recipients from 2010 to 2023. Cox proportional hazards models assessed death-censored graft survival and patient survival (PS), linear mixed models evaluated GF (eGFR), and logistic regression examined acute rejection within 12 mo.Results.Among 10 144 recipients (DNDD n = 7406, DCDD n = 2738), death-censored graft survival did not differ between groups (adjusted hazard ratio [aHR] DCDD versus DNDD [aHR, 1.08; 95% confidence intervals (CI), 0.94-1.24; P = 0.27]). PS was superior in the DCDD group (aHR, 0.58; 95% CI, 0.37-0.89; P = 0.01). DCDD recipients demonstrated a steeper early increase in eGFR with comparable long-term GF. Rejection within 12 mo was more common in DCDD recipients (odds ratio 1.36; 95% CI, 1.20-1.53; P < 0.001).Conclusions.Despite a higher risk of early rejection, DCDD recipients had comparable graft and PS to DNDD recipients. Early GF improved rapidly in DCDD recipients and remained similar over time. These findings support the continued use of DCDD kidneys to expand the donor pool. INTRODUCTION The global burden of kidney failure continues to rise, with an estimated global prevalence of 0.1%.1,2 Many patients with kidney failure require kidney replacement therapy as life-saving treatment. Kidney transplantation is widely regarded as the preferred form of kidney replacement therapy and is associated with reduced mortality, improved health outcomes and quality of life.3,4 Despite this, the demand for donor kidneys continues to outpace supply worldwide.4-6 To combat this issue, implementation of DCDD (previously known as DCD) programs has resulted in an increase in DCDD donor uptake globally.7 The expanding need for kidney donors applies to Australia and New Zealand, in light of a pervasively increasing waitlist and the projected growth of the dialysis population in the coming years.8,9 DCDD kidneys carry a higher risk of delayed GF, secondary to ischemia-reperfusion injury and prolonged warm ischemic times.10-12 Historically, this risk has frequently engendered uncertainty and caution in donor allocation. However, evolving evidence and accumulated experience have prompted a gradual shift in attitudes in countries such as the United States, the United Kingdom, Australia and parts of Europe, where DCDD kidney transplantation is now routinely performed.13,14 This change is because of increasing evidence that long-term outcomes for DCDD kidneys are equivalent to DNDD (previously known as DBD).13,15 Registry data published from the United States, the United Kingdom, and Europe have often demonstrated equivalent long-term patient and graft survival, with study follow-up periods ranging between 5 and 10 y.16-21 Long-term follow-up exceeding 25 y has also been reported in some cohorts with equivalent outcomes.20,22 Despite growing data on graft and PS, long-term data on GF, including estimated glomerular filtration rate (eGFR) remain poorly characterized globally. By 1 y, eGFR has been shown to be either equivalent or slightly lower in DCDD cohorts; for example, Walls et al reported lower eGFR in the DCDD group, particularly with higher donor age.16,21,23 The limited literature examining GF beyond the first year is constrained by shorter follow-up periods or small sample sizes: Schaapherder et al16,20 found largely equivalent GF up to 5 y, while Muller et al reported lower eGFR at 20 y in a cohort of 51 patients still being followed up at the time. Rijkse et al also reported a meta-analysis of limited studies with data on eGFR up to 1 y.15 There remains a need for large-scale studies with extended follow-up of eGFR. This retrospective cohort study seeks to address these evidence gaps by providing contemporary, long-term data on graft and patient outcomes, including GF, for kidney transplant recipients in Australia and New Zealand over a 14-y period. MATERIALS AND METHODS We conducted a retrospective cohort study using data from the Australia and New Zealand Dialysis and Transplant (ANZDATA) Registry. Study Population All adult patients (aged ≥18 y) who received a deceased donor kidney-only transplant in Australia or New Zealand between January 1, 2010, and December 31, 2023, were included. This aligns with the increased utilization of DCDD kidney transplants since 2009. Exclusion criteria included: re-cipients under the age of 18 at time of transplantation, recipients who received simultaneous multiorgan transplants, recipients who were transplanted outside of Australia or New Zealand, or recipients of kidneys from “pathological donors” (donor who received a nephrectomy for a renal mass or tumor). Definitions Patients were categorized into 2 groups depending on deceased donor pathway—DCDD or DNDD. DCDD donors were defined as controlled donors (modified Maastricht Category III) who experienced circulatory death following planned withdrawal of life-sustaining treatment in an intensive care unit or operating theater.24,25 DNDD donors were defined as individuals declared brain-dead based on neurological criteria. Study Outcomes The primary outcomes included patient survival (PS), death-censored graft survival (DCGS), and graft function (GF) over time. PS was defined as time from kidney transplantation to death from any cause (not censored at graft failure). DCGS was defined as the time from kidney transplantation to graft failure (defined as a permanent return to dialysis or retransplantation), censored at death with a functioning graft. GF was defined by estimated glomerular filtration rate (eGFR) using the CKD-EPI formula.26 Patients were followed up from time of transplant to December 31, 2023. The secondary outcome was rejection within 12-mo posttransplantation. Statistical Analyses Baseline characteristics were reported as mean (±standard deviation) for continuous variables that were normally distributed, or as median (+interquartile range) for skewed variables, and as counts with percentages for categorical variables. For descriptive analyses, comparisons between both groups were performed using the Chi-squared test for categorical variables, the Student’s t-test for normally distributed continuous variables, and the Mann–Whitney U test for nonnormally distributed continuous variables. The analysis of time-to-events (PS and DCGS) was visually assessed using Kaplan–Meier survival curves, and time to rejection using competing risk cumulative incidence curves. The PS analysis was restricted to primary transplants to avoid double counting the survival time of patients with subsequent transplants in the study period. Multivariable Cox proportional hazards regression models were constructed to evaluate the association between exposures and outcomes, with the donor pathway included in the model. A linear mixed-effects model was used to analyze long-term eGFR, including fixed effects for donor category and its interaction with time, as well as relevant covariates. A random intercept was included at the transplant level to account for clustering of repeated eGFR measures within each transplant episode. Additional random effects were also evaluated for transplants nested within recipients, and transplants nested within donors. As these alternative structures did not meaningfully alter the results, the final model retained a transplant-level random intercept only. eGFR data were also collected at the following timepoints: 1 mo, 2 mo, 3 mo, 6 mo, 12-mo, 2 y, 3 y, 5 y, 7 y, and 10 y. eGFR was modeled using a linear mixed-effects model, and we included all available eGFR measurements; observations with missing eGFR or missing covariates were excluded from the analysis at that time point. The secondary outcome of rejection within 12 mo posttransplant was analyzed using multivariate logistic regression to estimate adjusted odds ratios, with 95% confidence intervals (CI). Rejection was defined as the occurrence of any acute rejection episode within 365 d of transplantation. Recipients without a recorded rejection event were classified as not having experienced rejection, irrespective of available follow-up duration; therefore, recipients transplanted within 12 mo of the study end date were included in the primary analysis. A sensitivity analysis restricted to recipients transplanted on or before December 31, 2022, was performed to ensure complete 12-mo follow-up ascertainment. Rejection up to 1-y posttransplantation has been shown to be an important predictor of long-term patient and graft survival.27,28 We also split the outcome of rejection for cellular and antibody-mediated rejection separately. DGF was not included as a covariate in the rejection model, as it is considered to lie in the causal pathway between donor type and rejection, a likely mediator. For all models, all variables were examined as potential covariates and were included sequentially if they were significant (P < 0.25) on univariate analysis or were deemed clinically relevant. For all models, all available variables were examined as potential covariates and were included in the model first using forward elimination (see Tables, SDC, https://links.lww.com/TXD/A899). Then, backward elimination was used to remove variables that did not significantly improve the model or were considered clinically irrelevant. Schoenfeld’s residuals were used to test the proportional hazards assumption for the Cox proportional hazards models. Collinearity was evaluated using the variance inflation factor. Interaction terms were assessed using forward elimination, retaining only those with a P value ≤0.05. The interactions between donor category and ischemic time and era and donor category were tested and found to be statistically nonsignificant and were not included in the models. The linearity assumption was assessed for all variables. Nonlinear continuous variables were made categorical. For the DCGS models, the age at transplant was found to be nonlinear, and therefore was modeled using a fractional polynomial. KDRI was included in all our models as a covariate, and a modified KDRI was derived by removing the DCDD component before applying natural logarithmic transformation and inclusion in the models, to avoid adjusting for a component already represented by donor type. A complete case analysis was performed with very little missingness in the registry data. For categorical and continuous variables, missingness was overall <2%. For the variables included in statistical models, missingness was overall <2%. The models were assessed using the Harrell C statistic and Akaike Information Criterion. Data were analyzed using Stat/IC version 19.0 (StataCorp, College Station, TX). Ethics The research was conducted with approval from the Royal Adelaide Hospital Human Research Ethics Committee (HREC reference number, HREC/17/RAH/408) and the Central Adelaide Health Network (Reference number, R20170927). The requirement for written informed consent was waived because of the use of existing deidentified registry data and the minimal risk posed to participants. RESULTS A total of 10 144 kidney transplant recipients were included, 2738 of which were DCDD kidneys (26.9%). Descriptive characteristics are outlined in Table 1. TABLE 1.Descriptive characteristicsDNDD (n = 7406)DCDD (n = 2738) P Recipients Age, median (IQR)54.0 (43.0-62.0)55.0 (44.0-63.0) <0.001 Gender, n (%) Female2738 (37.0%)1009 (36.9%)0.91 Male4668 (63.0%)1729 (63.1%)Graft number, n (%) First6460 (87.2%)2403 (87.8%)0.31 Subsequent946 (12.8%)335 (12.2%)Mean BMI kg/m2 (SD)27.5 (6.3)27.4 (6.5)0.83Primary kidney disease, n (%) Diabetic nephropathy1340 (18.2%)466 (17.1%)0.67 Glomerular disease2791 (37.9%)1019 (37.4%) Hypertension/renovascular disease583 (7.9%)228 (8.4%) Familial/hereditary1119 (15.2%)431 (15.8%) Tubulointerstitial935 (12.7%)350 (12.8%) Other systemic diseases104 (1.4%)34 (1.2%) Miscellaneous483 (6.6%)196 (7.2%)Diabetes, n (%) Type 1193 (2.6%)67 (2.5%)0.84 Type 21756 (23.8%)659 (24.2%)History of smoking, n (%)3507 (44.0%)1187 (43.6%)0.69History of chronic lung disease, n (%)721 (9.8%)269 (9.9%)0.87History of coronary artery disease, n (%)1626 (22.0%)595 (21.8%)0.85History of peripheral vascular disease, n (%)822 (11.1%)312 (11.4%)0.65History of cerebrovascular disease, n (%)468 (6.3%)186 (6.8%)0.37Donors Age, median (IQR)49.0 (35.0-60.0)50.0 (37.0-59.0)0.35Gender, n (%) Female3466 (46.8%)939 (34.3%) <0.001 Male3940 (53.2%)1799 (65.7%)BMI kg/m2, median (IQR)26.4 (23.4-30.2)26.3 (23.4-30.8)0.30History of diabetes, n (%) No6912 (93.4%)2540 (92.8%)0.28 Yes490 (6.6%)198 (7.2%)History of hypertension, n (%)1869 (25.2%)647 (23.6%)0.17History of smoking, n (%) Current smoker2984 (40.3%)1125 (41.1%)0.14 Former smoker1613 (21.8%)638 (23.3%)Cause of death, n (%) Intracranial hemorrhage3278 (45.3%)605 (22.5%) <0.001 Traumatic brain injury1185 (16.4%)556 (20.7%) Cerebral infarction391 (5.4%)188 (7.0%) Cerebral hypoxia/ischemia2165 (29.9%)1053 (39.2%) Other neurological condition107 (1.5%)38 (1.4%) Other nonneurological condition104 (1.4%)244 (9.1%)Terminal creatinine, median (IQR)72.0 (57.0-98.0)67.0 (52.0-86.0) <0.001 Transplant profile HLA mismatch,a median (IQR)4.0 (2.0-5.0)4.0 (2.0-5.0)0.06PRA, n (%) 0-<14478 (60.5%)1791 (65.4%) <0.001 1-792139 (28.9%)722 (26.4%) 80-94352 (4.8%)95 (3.5%) 95+429 (5.8%)129 (4.7%)KDRI, median (IQR)1.3 (1.0-1.6)1.4 (1.1-1.7) <0.001 Total ischemic time,b mean (SD)11.1 (4.8)11.0 (4.7)0.55Any induction therapy6872 (92.8%)2520 (92.0%)0.20 T-cell induction670 (9.0%)303 (11.1%) 0.002 Anti-CD25 antibody6315 (85.3%)2287 (83.5%) 0.03 aHLA Mismatches included HLA-A, HLA-B, HLA-DR.bTotal ischemic time in DNDD recipients defined as the time (to the closest hour), of donor renal artery interruption or aortic clamp, until time of clamp release on the renal artery in the recipient. ANZDATA/ANZOD do not separate warm and cold ischemic time. In DCDD recipients, the time is counted from when donor systolic blood pressure is <50 mmHg, or oxygen saturation falls below 80%.29BMI, body mass index; DCDD, donation after circulatory determination of death; DNDD, donation after neurological determination of death; HLA, human leukocyte antigen; IQR, interquartile range; KDRI, unmodified kidney donor risk index; PRA, panel reactive antibodies. Descriptive characteristics aHLA Mismatches included HLA-A, HLA-B, HLA-DR. bTotal ischemic time in DNDD recipients defined as the time (to the closest hour), of donor renal artery interruption or aortic clamp, until time of clamp release on the renal artery in the recipient. ANZDATA/ANZOD do not separate warm and cold ischemic time. In DCDD recipients, the time is counted from when donor systolic blood pressure is <50 mmHg, or oxygen saturation falls below 80%.29 BMI, body mass index; DCDD, donation after circulatory determination of death; DNDD, donation after neurological determination of death; HLA, human leukocyte antigen; IQR, interquartile range; KDRI, unmodified kidney donor risk index; PRA, panel reactive antibodies. Primary Outcomes Death-censored Graft Survival and Patient Survival The Kaplan–Meier Curve analysis showed no difference in DCGS between DCDD versus DNDD over 10 y (Figure 1). Multivariate analyses indicated no statistically significant association for DCDD with DCGS (aHR, 1.08; 95% CI, 0.94-1.24; P = 0.27; Table S1, SDC, https://links.lww.com/TXD/A899). We noted superior PS in the DCDD group (aHR, 0.58; 95% CI, 0.37-0.89; P = 0.01; Figure 2; Table S2, SDC, https://links.lww.com/TXD/A899). There was no era effect. Five-year PS was 88.3% (DNDD) and 87.9% (DCDD), and 10-y PS was 70.1% (DNDD) versus 73.9% (DCDD). The median (IQR) follow-up for both DCGS and PS can be seen in Table S3 (SDC, https://links.lww.com/TXD/A899). Kaplan–Meier curve for death-censored graft survival. Kaplan–Meier curve for patient survival. Graft Function At 1-mo posttransplant, the DCDD group had a lower mean eGFR (42 mL/min/1.73m2) compared with those with DNDD kidneys (53 mL/min/1.73m2), but this became comparable by 5 y, and persisted up to 10 y (Figure 3, Tables S4 and S5, SDC, https://links.lww.com/TXD/A899). Our linear mixed-effects model demonstrated a nonlinear trajectory of eGFR over time. Compared with DNDD recipients, DCDD recipients exhibited a significantly steeper early increase in eGFR (P < 0.001 for both spline interaction terms), despite a lower baseline eGFR (Table S6, SDC, https://links.lww.com/TXD/A899). Our adjusted model confirmed that with time, the gap between DCDD and DNDD narrows, and there was no long-term difference in eGFR. Observed posttransplant eGFR (stratified by donor type). eGFR, estimated glomerular filtration rate. Secondary Outcomes Rejection Within 12 Mo Figure 4 demonstrates cumulative incidence of rejection based on donor group up to 12 mo posttransplantation. We found that DCDD kidneys showed higher odds of rejection (OR, 1.36; 95% CI, 1.20-1.53; P < 0.001; Table 2; Table S7, SDC, https://links.lww.com/TXD/A899). Results did not alter meaningfully when subtyped for rejection type (cellular rejection, OR, 1.25; 95% CI, 1.06-1.47; P = 0.007; antibody-mediated rejection, OR, 1.70; 95% CI, 1.34-2.17; P < 0.001). TABLE 2.Incidence of short-term outcomes at 12 moDNDD (n = 7406)DCDD (n = 2738) P Delayed graft function1787 (24.1%)1507 (55.0%) <0.001 Rejection at 12 mo1223 (16.5%)571 (20.9%) <0.001 DCDD, donation after circulatory determination of death; DNDD, donation after neurological determination of death. Incidence of short-term outcomes at 12 mo DCDD, donation after circulatory determination of death; DNDD, donation after neurological determination of death. Time to first rejection. DISCUSSION This study demonstrates evidence of comparable long-term DCGS, PS, and eGFR over time, despite the increased risk of rejection. Consistent with other large international registry studies, our analysis showed no significant difference in long-term DCGS among recipients of DCDD kidneys. Schaapherder et al16 reported no difference in DCGS over 10 y in a Dutch cohort of 6322 patients. A systematic review and meta-analysis of 51 studies incorporating 591 683 patients, conducted by Rijkse et al15 similarly found no significant difference in the risk of graft loss beyond 5, and up to 10 y between DCDD and DNDD transplants in 8 studies. Other studies were limited by either shorter follow-up periods or earlier cohorts; for instance, Butler et al17 reported comparable outcomes in a US cohort of 35 264 deceased kidney transplants with 4 y of follow-up. There are also limited studies from 2014 examining transplant outcomes in Australia and New Zealand. Tan et al (n = 130) reported no difference in recipient and graft survival in a single-center study, while McDonald et al examined a substantially larger cohort with similar conclusions, although with follow-up limited to 5 y.30,31 Although PS has been less frequently reported, existing studies have demonstrated equivalent survival in DCDD recipients up to 10 y across cohorts from the United Kingdom, the United States, Europe, and the aforementioned meta-analysis.15,16,19 To date, a survival benefit for DCDD kidney transplants has only been observed when compared with remaining on dialysis or receiving no transplant.32,33 In our analysis, we noted improved PS in the DCDD group, though the absolute difference was small, which may partly reflect selection bias, rather than a true biological effect (Figure 2). In this context, although the difference reached statistical significance, we do not interpret it as clinically meaningful, as we do not feel that there is a plausible mechanism by which DCDD transplantation would confer superior PS. Regardless, these results do strengthen existing evidence of at least comparable long-term PS outcomes in DCDD kidney transplant recipients. Existing data on GF, typically measured by eGFR, remain limited and are generally focused on the first year posttransplant, during which DCDD transplants have sometimes shown lower eGFR compared with DNDD transplants.23,30,34 A large-scale systematic review and meta-analysis by Rijkse et al15 showed very limited studies examining eGFR as outcomes and were mostly limited by short follow-up or older age donor groups. Schaapherder et al16,20 reported equivalent eGFR between both groups of up to 5 y, but longer-term data in contemporary cohorts remain very scarce. Notably in our study with longer-term data, although DCDD recipients had lower eGFR initially, their GF improved markedly within the first year, remained stable for longer, when compared with DNDD recipients by 10 y. The initial ischemic insult associated with the donation process, together with higher rates of DGF, likely account for lower early eGFR in DCDD organs, which improves over time as this injury resolves, matching the DNDD group. To our knowledge, this study provides the largest longer-term data on eGFR following DNDD and DCDD transplantation. These results are reassuring and provide robust evidence that DCDD kidneys can achieve sustained robust long-term GF, which is of significant clinical importance. DCDD kidney transplant recipients experienced increased rejection within 12 mo, which may be because of many reasons. The increase in rates of rejection may be secondary to increased graft immunogenicity from ischemic-reperfusion injury, and the higher incidence of delayed GF, a known independent risk factor for rejection.35,36 Although rejection data on incidence and comparisons between DCDD and DNDD in various kidney transplant cohorts have been seldom reported in the literature, Rijkse et al15 found no association between the DCDD donor type and increased risk of rejection. Conversely, the elevated rejection rates in our DCDD cohort may partly reflect increased detection of subclinical rejection because of a lower threshold to biopsy in the context of DGF, which is more commonly seen in the DCDD pathway. Rates of DNDD donation are declining, partly because of fewer patients presenting with brain death.37 In this context, DCDD kidney transplantation has become increasingly important in helping to address the ongoing organ shortage. The consistent evidence of equivalent long-term outcomes is thus reassuring and underpins its sustained use. While our long-term outcomes are both reassuring and encouraging, DCDD kidney transplantation carries inherent limitations. Notably, up to one-third of potential DCDD donations do not proceed to donation, which may reduce resource efficiency and can be disappointing for donor families and intended recipients.24,38 From a resource perspective, DNDD donation may be more cost-effective overall, as it enables retrieval of additional organs, such as hearts and lungs, which are retrieved at lower rates in DCDD donation.39 Additionally, it remains important to acknowledge the additional financial, physical, and psychosocial burden that DCDD kidney transplant recipients may experience, particularly in the early posttransplant period. DGF is a well-recognized challenge associated with DCDD transplants, which contributes to increased healthcare costs, prolonged hospital admissions, and greater resource utilization.36,40-42 Accordingly, measures to minimize ischemic times and reduce the incidence of DGF remain important strategies to address these short-term challenges. Notably, the favorable outcomes observed in our cohort—despite the limited use of machine perfusion—suggest that even greater improvements could be achieved with broader adoption of normothermic or hypothermic machine perfusion to mitigate early graft dysfunction. Such technologies have demonstrated clear benefits and are already widely utilized in countries including the Netherlands, France, and the United Kingdom.43-45 A Cochrane review by Tingle et al45 demonstrated superior outcomes with reduced DGF, improved graft survival and superior cost-effectiveness with the use of hypothermic machine perfusion, when compared with static cold storage. These considerations are especially relevant in the Australian and New Zealand context, where the vast geographic landscape often results in longer organ transport times and extended cold ischemic times. The use of machine perfusion may help counteract these challenges, reducing the burden of DGF, early hospital readmissions, and the need for surveillance biopsies commonly seen in the early posttransplant period.46 At this time, there is limited use of normothermic machine perfusion, and this is not captured in the registry so cannot be reported. There is no normothermic regional perfusion (NRP) currently used in Australia or New Zealand.47 This study has several notable strengths. It represents a large-scale, comprehensive and contemporary retrospective cohort analysis spanning 14 y, with the time period coinciding with increased uptake of DCDD transplants. The robust data presented, with comparable graft survival, improved PS, and sustained long-term GF, strongly support the continued use of DCDD kidneys in transplantation. In particular, with respect to GF, this study builds on previously limited evidence by extending follow-up to 10 y in a modern cohort, providing new insights into the durability of GF among DCDD recipients. Taken together, these results offer compelling justification to promote the broader uptake of DCDD transplantation as a critical strategy to help address the persistent shortage of donor organs. There are limitations to be acknowledged. Inherent biases are associated with using registry data. Although every effort has been made to adjust for confounders, the use of DCDD kidneys may carry inherent selection bias—for example, recipients may be healthier, or conversely DCDD kidneys may be assigned to recipients with lower expected posttransplant survival, potentially biasing results. At present, there is also no available registry data on machine perfusion and therefore, this could not be adjusted for within this analysis. CONCLUSION Our study demonstrates that DCDD kidney transplants achieve comparable graft survival, improved PS and comparable long-term GF, supporting their use to increase the donor pool. Despite short-term challenges of increased DGF and rejection, future research should focus on strategies, like machine perfusion, to mitigate these risks and further enhance outcomes. ACKNOWLEDGMENTS We are grateful to the Australian and New Zealand kidney units, patients and staff for their cooperation and contributions to ANZDATA. The data reported here were supplied by the ANZDATA Registry. The interpretation and reporting of these data are the responsibility of the authors and in no way should be seen as an official policy or interpretation of the Registry. Supplementary Material