UNOS guidelines provide inadequate discriminatory criteria for kidneys that should be transplanted as single (SKT) versus dual (DKT). We evaluated the utility of the kidney donor risk index (KDRI) to define kidneys with better outcomes when transplanted as dual. Using SRTR data from 1995 to 2010 of de novo KTX recipients of adult deceased-donor kidneys, we examined outcomes of SKT and DKT stratified by KDRI group ≤1.4 (n = 49 294), 1.41–1.8 (n = 15 674), 1.81–2.2 (n = 6523) and >2.2 (n = 2791). DKT of kidneys with KDRI >2.2 was associated with significantly better overall graft survival [adjusted hazard ratio (aHR) 0.83, 95% confidence interval (CI) 0.72–0.96] compared to single kidneys with KDRI >2.2. DKT was associated with significantly decreased odds of delayed graft function (top 2 KDRI categories) and significantly decreased odds of 1-year serum creatinine level >2 mg/dL (top 3 KDRI categories). Among SKT and DKT from KDRI >2.2 there were 16.1 and 13.9 graft losses per 100 patient follow-up years, respectively. KDRI >2.2 is a useful discriminatory cut-off for the determination of graft survival benefit with the use of DKT; however, the benefit of increased graft years was less than half of single kidneys from donors in the same KDRI range.
adjusted hazard ratio
adjusted odds ratio
body mass index
cold ischemia time
donation after circulatory death
dual kidney transplantation
human leukocyte antigen
health resources and services administration
organ procurement organization
organ procurement and transplantation network
panel reactive antibody
single kidney transplantation
scientific registry of transplant recipients.
The use of marginal donor kidneys has become an increasingly accepted method to expand the donor pool; however, the quality of the donated organ in kidney transplantation is one of the most crucial factors for graft survival. Concerns over the limited life span of marginal kidneys have led some surgeons to perform dual kidney transplantation based on the concept that poor long-term outcomes may be the consequence of an imbalance between the number of viable nephrons supplied and the metabolic demand of the recipient ; therefore, concurrent transplantation of two kidneys into one recipient would increase nephron dosing  and maximize outcomes [3-14].
Various reports have demonstrated improved graft survival or function with the use of dual rather than single kidney implantation when risk factors predict poor graft function. Criteria or scoring systems for choosing to perform dual transplantation have generally included older donor age and one or more other parameter(s) such as preimplant biopsy, donor estimated creatinine clearance, kidney weight, machine perfusion characteristics and/or long cold ischemia time [2-14]. However, it is not always clear whether the criteria utilized is robust enough to select which marginal donor kidneys should be used together in order to maximize successful outcomes without reducing the organ supply—that is by dual grafting when a single graft would suffice.
Current UNOS policy states that kidneys from adult donors must be offered singly unless the offering OPO would not use the kidneys singly and the donor meets at least two of the following conditions: (1) age >60 years, (2) estimated creatinine clearance less than 65 mL/min based on admission serum creatinine, (3) terminal serum creatinine >2.5 mg/dL, (4) history of longstanding hypertension or diabetes mellitus and (5) histologic findings of glomerulosclerosis 16–49%. Unfortunately, UNOS criteria do not adequately identify ECD kidneys that would confer a graft survival advantage if transplanted as dual .
The OPTN/United Network for Organ Sharing (UNOS) Kidney Transplantation Committees' (KTC) recent proposal of a modified version of the kidney donor risk index (KDRI) originally described by Rao et al.  as a tool to improve risk assessment may also serve as useful scoring system to guide dual kidney allocation. The UNOS KDRI incorporates multiple donor parameters into a single metric quantifying risk for graft failure along a continuum.
To examine the utility of the KDRI to define kidneys with better outcomes when transplanted as DKT, we examined the outcomes of dual and single kidneys for any given level of donor risk as measured by the UNOS KDRI.
Materials and Methods
Study design and participants
We utilized data from the Scientific Registry of Transplant Recipients (SRTR). The SRTR data system includes data on all donor, wait-listed candidates, and transplant recipients in the United States, submitted by the members of the Organ Procurement and Transplantation Network (OPTN). The Health Resources and Services Administration (HRSA), U.S. Department of Health and Human Services provides oversight to the activities of the OPTN and SRTR contractors.
Using the SRTR Standard Analysis File, we conducted a retrospective cohort study of patients receiving a deceased-donor kidney transplants (KTX) in the United States from January 1, 1995 to December 31, 2010 with last follow-up data available from October 2011. Exclusion criteria included (1) multi-organ transplant recipients, (2) previous transplant recipients, (3) ABO incompatible transplants, (4) donor or recipient age less than 18 years, (5) recipients with human immunodeficiency virus and (6) patients with missing data to calculate the KDRI.
Exposure and outcome classification and assessment
We utilized the KDRI model adopted by the UNOS/OPTN which includes 10 deceased-donor variables (http://optn.transplant.hrsa.gov/ContentDocuments/Guide _to_Calculating_Interpreting_KDPI.pdf; February 28, 2012). The KDRI was categorized into deciles, and plotted by transplant type (single vs. dual). We examined the distribution of the KDRI score at 0.2 and 0.4 intervals and in order to achieve analyzable groups, the KDRI was categorized into four groups using 0.4 intervals: ≤1.4, 1.41–1.8, 1.81–2.2 and >2.2 which represented 66.4%, 21.1%, 8.8% and 3.8% of the transplanted kidneys, respectively.
The primary outcome was graft failure which was defined as the earliest of retransplantation, return to dialysis, or patient death. Secondary outcomes were (1) death-censored graft failure (defined as retransplantation or return to dialysis), (2) patient mortality from any cause following transplantation, (3) delayed graft function (defined as at least one dialysis session within the first week posttransplantation) and (4) creatinine level >2.0 at one year. For the DGF analysis only, recipients with unknown dialysis status at the time of transplantation, those preemptively transplanted, and cases with primary nonfunction were excluded. Data on 1-year serum creatinine were missing in 14%.
Of recovered adult kidneys, rates of discard were calculated during the same time frame across KDRI groups. The discard rate was defined as the number of kidneys not transplanted, divided by the total number of kidneys recovered for transplantation.
The following recipient factors were included in the models: age (continuous), sex, race (African-American, other), duration of maintenance dialysis prior to transplantation (none, ≤ 3 years, >3 years, missing), number of HLA-A, B and DR mismatches (≤3, >3), panel-reactive antibody (PRA) level (>30%, ≤30%, missing), body mass index (≤30 kg/m2, >30 kg/m2, missing), cold ischemia time (hours; ≤12, 12–24, >24, missing), insurance status (private, other) and year of transplantation.
Baseline characteristics by dual (DKT)/single (SKT) kidney transplant status were compared with the chi-square test for categorical variables and the t-test for continuous variables whose distribution approximated normality. Continuous variables were examined for outliers, and cases with values thought to be clinically implausible were recoded as missing. The distributions of KDRI across DKT/SKT categories were visually explored using histograms.
The Kaplan–Meier method was used to assess time to graft failure by DKT/SKT type across KDRI categories. The log-rank test was used to evaluate differences between survival curves. Cox proportional hazards models were fitted to examine the independent association between DKT/SKT status at each KDRI quartile for the time to event outcomes while adjusting for potential recipient confounders. A model was fitted for each KDRI category and stratified by transplant type (SKT vs. DKT). Time to death or failure was determined from the date of transplant until outcome, censored for loss to follow-up, end of study period (October 31, 2011). Covariates in the Cox models were assessed for adherence to the proportional hazard assumption. No important departures from proportionality were observed. Logistic regression models were fitted to examine the relationship between DKT/SKT status across each KDRI quartile and risk of DGF or 1-year creatinine >2.0 mg/dL while adjusting for potential recipient confounders. Multivariable models were fitted with the results from cases that had complete data. No data were imputed.
All analyses were performed using SAS software, version 9.2 (SAS Institute, Inc., Cary, NC). Statistical significance was identified by a p-value of less than 0.05 and all confidence intervals also used a 95% threshold. This study was approved by the Albert Einstein College of Medicine internal review board.
A total of 74 282 kidney transplant recipients were included in the study cohort. There were 1308 in the DKT group and 72 974 in the SKT group. Of 1308 DKT, 11% (n = 141), 25% (n = 321), 30% (n = 397) and 34% (n = 449) of the kidneys were KDRI category 0–1.4, 1.41–1.8, 1.81–2.2 and >2.2, respectively. Of the 72 974 SKT 67% (n = 49153), 21% (n = 15 353), 8% (n = 6126) and 3% (n = 2342) were KDRI category 0–1.4, 1.41–1.8, 1.81–2.2 and >2.2, respectively. DKT tended to have higher KDRI and SKT tended to have lower KDRI (Figure 1). Recipient and transplant characteristics between the DKT and SKT groups across each KDRI category are shown in Tables 1a and 1b. Patients receiving dual kidneys were more likely to be older, African-American, and have less pre-transplant dialysis duration. They were less likely to be obese and sensitized. There were fewer DKT recipients with glomerular disease, and more with diabetes or hypertension, as the listed primary etiology of renal failure. DKT transplants had more HLA mismatches and longer cold ischemia times.
|Characteristic % or mean ± SD||SKT (n = 72 974)||DKT (n = 1308)||p-Value|
|Kidney donor risk index (KDRI)|
|Age, years||49.6 ± 14.8||58.9 ± 10.5||<0.0001|
|Listed primary ESRD etiology|
|Recipient BMI > 30 kg/m2||29.6||25.1||0.0011|
|Peak PRA > 80%||7.0||3.6||<0.0001|
|Recipient HLA MM >3||67.0||80.1||<0.0001|
|Characteristic % or mean ± SD||KDRI ≤ 1.4||KDRI 1.4–1.8||KDRI 1.8–2.2||KDRI ≥ 2.2|
|Age, years||47 ± 15||54 ± 12||53 ± 13||57 ± 9||57 ± 12||59 ± 10||59 ± 12||61 ± 10|
|BMI > 30 kg/m2||29||24||30||26||30||25||27||25|
|Peak PRA > 80%||8||6||6||3||5||3||4||3|
|HLA MM >3||64||74||71||81||76||80||78||82|
Kaplan–Meier curves for graft failure stratified by DKT/SKT status across KDRI categories are shown in Figure 2. The cumulative probability of graft failure among DKT recipients was similar to that of SKT recipients with KDRI ≤1.4, 1.41–1.8, and 1.81–2.2; but not >2.2 where DKT outcomes were superior. Five-year graft survival rates for SKT and DKT by KDRI were as follows: ≤1.4 (74%, 72%), 1.41–1.8 (63%, 64%), 1.81–2.2 (55%, 59%) and >2.2 (48%, 54%). On multivariate analysis, DKT was associated with significantly better graft survival with kidneys designated as KDRI >2.2 (adjusted hazard ratio (aHR) 0.83, 95% confidence interval (CI) 0.71, 0.96) relative to SKT, but no benefit of DKT was seen with kidneys from other KDRI categories (Table 2). Similar patterns were noted for the outcome of death-censored graft failure where a significant protective effect was demonstrated with the DKT use of KDRI >2.2 grafts (aHR 0.79, 95% CI 0.65, 0.97).
|Parameters (reference group)||Patient mortality|
|Adjusted hazard ratio (95% CI)|
|Dual kidney transplantation (SKT)||0.83 (0.71–0.96)|
|Recipient age, continuous per year||1.02 (1.01–1.02)|
|Recipient race, African-American (other)||1.132 (1.01–1.26)|
|Recipient female (male)||0.92 (0.82–1.02)|
|Recipient body mass index >30 kg/m2 (other)||1.21 (1.07–1.38)|
|Recipient PRA >80 (PRA ≤ 80)||1.28 (0.99–1.66)|
|Recipient HLA-mismatches >3 (≤3)||1.03 (0.92–1.16)|
|Cold ischemia time, hours (<12)|
|Recipient insurance, non-private (private)||1.18 (1.04–1.33)|
|Recipient ESRD diagnosis (glomerulonephritis)|
|Dialysis duration prior to transplant (none)|
|<3 years||1.28 (0.97–1.69)|
|≥3 years||1.43 (1.08–1.90)|
|Transplant year||0.97 (0.95–0.98)|
|Machine perfusion||0.99 (0.89–1.12)|
Graft year analysis
Among KTX deriving from donors with KDRI >2.2, both kidneys were transplanted in 16.1% of 2759 cases. We compared the total number of graft losses per follow-up year between SKT and DKT transplants. There were 16.1 graft losses per 100 patient follow-up years among single kidney transplants and 13.9 graft losses per 100 patient follow-up years among dual kidney transplants. Based on these numbers, the total number of graft years per kidney was higher among solitary transplants.
Five-year patient survival rates for SKT and DKT by KDRI were as follows: ≤1.4 (85%, 80%), 1.41–1.8 (77%, 75%), 1.81–2.2 (70%, 70%) and >2.2 (65%, 67%) On multivariate analysis, there were no differences in overall patient survival with the use of DKT with kidneys from any KDRI category; however there was a trend toward a survival benefit with each increasing category of KDRI (Table 3).
|Outcome||KDRI ≤1.4||KDRI 1.41–1.8||KDRI 1.81–2.2||KDRI >2.2|
|Adjusted hazard ratio (95% CI)|
|Graft survival||0.97 (0.72, 1.30)||1.00 (0.85, 1.19)||0.90 (0.78, 1.05)||0.83 (0.71–0.96)|
|Death-censored graft survival||0.88 (0.56, 1.38)||0.97 (0.76–1.22)||0.86 (0.70, 1.06)||0.79 (0.65, 0.97)|
|Patient survival||1.06 (0.76, 1.47)||1.03 (0.84, 1.25)||1.03 (0.87, 1.22)||0.91 (0.77, 1.07)|
|Adjusted odds ratio (95% CI)|
|Delayed graft function||1.54 (1.05, 2.27)||0.91 (0.70, 1.18)||0.69 (0.54, 0.88)||0.62 (0.49, 0.79)|
|1-year creatinine >2.0||1.02 (0.53, 1.93)||0.61 (0.43, 0.86)||0.64 (0.48, 0.85)||0.49 (0.37, 0.65)|
The incidence of delayed graft function of SKT/DKT across the KDRI groups was ≤1.4 (23%, 30%), 1.41–1.8 (32%, 31%), 1.81–2.2 (35%, 29%) and >2.2 (36%, 26%), respectively. On multivariate analysis DKT was associated with significantly worse odds of DGF with KDRI ≤1.4 (aOR 1.54, 95% CI 1.05, 2.27) and a protective effect with a monotonic decrease across the higher KDRI categories 1.41–1.8 (aOR 0.91, 95% CI 0.70–1.18), 1.81–2.2 (aOR 0.69, 95% CI 0.54, 0.88) and >2.2 (aOR 0.62, 95%CI 0.49, 0.79; Table 3).
The incidence of serum creatinine >2.0 at 1 year was 22% for DKT and 17% for SKT; and across each KDRI group for SKT and DKTs respectively was ≤1.4 (10%, 9%), 1.41–1.8 (22%,15%), 1.81–2.2 (27%, 20%) and >2.2 (32%, 21%). On multivariate analysis, DKT was associated with significantly decreased odds of 1-year serum creatinine level >2 mg/dL, in the top 3 KDRI categories with 39%, 36% and 51% decreased odds, respectively (Table 3).
A total of 80 022 adult-donor kidneys were recovered for transplantation and 5740 were not transplanted. Discard rates across KDRI categories ≤1.4, 1.41–1.8, 1.81–2.2 and >2.2 were 2% (980/50 274), 8% (1357/17 031), 18% (1438/7951) and 41% (n = 1965/4756), respectively.
The results of our study demonstrate that KDRI serves as an adequate discrimination tool to determine when graft survival can be improved with the use of dual rather than single kidney transplantation. In this analysis, KDRI >2.2 was the cut-off value that conferred significantly better overall graft survival with DKT. Our results confirm existing knowledge that DKT of marginal grafts is associated with significantly decreased odds of delayed graft function (demonstrated with top 2 KDRI categories). Lastly, DKT significantly reduces the odds of 1-year serum creatinine level >2 mg/dL (top 3 KDRI categories); this finding likely points to a better functional reserve of dual transplant recipients because of an enhanced nephron mass.
Despite our finding of a significant difference in graft survival for recipients of DKT from high KDRI kidneys, the effect size is somewhat small (17%). And, although DKT kidneys from donors with a KDRI >2.2 did offer increased graft years, the protective benefit was substantially less than half as compared to single kidneys from donors in the same KDRI range. Therefore, from a resource perspective, with the strong caveat that SKT may be more selectively allocated than DKT kidneys, solitary transplants may offer substantially more total graft years to the recipient population despite an elevated risk. In other words, among those donor kidneys that can be successfully transplanted as singles, two transplants offer significantly more cumulative graft survival than dual transplantation. However, from an individualistic and programmatic perspective, a small yet significant optimization in graft survival, along with other indices of graft function, may be preferable. Also to consider is the possibility of improving organ utilization with DKT. Often the decision-making rationale is not DKT versus SKT, but rather DKT versus discard. Since 41% of kidneys with KDRI >2.2 are discarded, there appears to be room for improved utilization.
The UNOS KDRI designation is a granular risk assessment that classifies deceased-donor kidneys according to relative risk for graft loss. Although there are significant donor risk factors not captured by the index, such as biopsy characteristics, concordance statistics indicate that the tool is especially useful for distinguishing more extreme categories of graft failure risk (http://optn.transplant.hrsa.gov/ContentDocuments/Guide_to_Calculating_Interpreting_KDPI.pdf, 16). Our multivariable Cox proportional hazards model demonstrated a significant decrease in the risk of graft failure using DKT with KDRI >2.2; however there was a stepwise decrease in the point estimates for each of the top 3 KDRI categories and a narrowing of the confidence interval toward significance in KDRI category 1.81–2.2, suggesting that with a larger sample of high-KDRI kidneys the cut-off value for maximizing outcomes with DKT may be lower.
Another compelling scoring system to determine criteria for DKT versus SKT was reported by Remuzzi et al.  in 2006. The authors reported the results of a multicenter prospective analysis demonstrating excellent short- and long-term results after allocation of kidneys from donors >60 years of age based on a pre-implant biopsy 12-point histologic scoring system proposed by the Dual Kidney Transplant Group that assesses vessels, glomeruli, tubules and interstitium. Their results suggest that the histologic evaluation of donor kidneys serves to reduce variability in prognosticating outcomes. Unfortunately, the absence of UNOS data on features other than glomerulosclerosis limits the ability to validate the usefulness of histopathology in predicting graft outcome with United States registry data.
On examination of the characteristics of candidates chosen to receive DKT, it is clear that the recipients are older, thinner, more likely African-American and hypertensive or diabetic, less likely to be sensitized and have less time on dialysis. This suggests selection by transplant clinicians in the process of evaluating the most appropriate candidates for dual transplantation. Selection of optimal candidates is important in DKT because it is a more extensive procedure and is typically performed in elderly recipients which have less physiologic reserve than younger recipients. Despite the greater extent of the DKT procedure, complication rates have been shown to be comparable to those of single kidney transplants .
DKT is a potential strategy to maintain access without impairing outcomes of transplantation for older candidates. If the proposed UNOS policy, released for public review in early 2011, of survival matching by KDRI is accepted, the distribution of younger organs with lower risk would shift to younger recipients and organs from older donors at higher risk of graft failure would be allocated primarily to older individuals . This is projected to cut the number of deceased-donor renal transplants by 18% for candidates aged 50–64 and by 33% for those aged 65 and older over a year compared to allocation based on current rules (http://optn.transplant.hrsa.gov/kars.asp, Reports to the Board of Directors, June 17, 2009). Yet the proportion of candidates listed for kidney transplantation in these two age groups have increased faster than all younger age groups over recent years accounting for 44% and 20% respectively, of the active list by the end of 2012 (optn.hrsa.gov, accessed December 15, 2012). Unfortunately, the assertion arising from a review of SRTR data  that the 5-year probability of deceased-donor transplantation of patients aged ≥65 years is nearly equal to the probability of death on the waiting-list highlights the poor prognosis of the elderly on dialysis. Considering recent trends in transplantation of the elderly in the context of new allocation concepts, high-KDRI kidneys, in the absence of a living donor, may be the most realistic option for transplantation of seniors. DKT is an additional option for minimizing risk.
We found discard rates of donor kidneys within the top two KDRI categories were 18% and 41%, respectively. Initiatives to improve utilization of marginal kidneys are important. Before KDRI, the ECD designation was created, in part, to increase utilization. However, despite increased recovery of ECD kidneys and despite knowledge that these organs would be predicted to provide survival benefit over dialysis for appropriate recipients , transplant centers were using them with greater selectivity . Discard rates for the lowest risk ECD kidneys increased following institution of the ECD policy , indicating a reluctance to transplant kidneys with poor quality. There is great variability in organ quality that is not captured by the SCD/ECD dichotomy and it remains to be seen whether the more granular depiction of donor risk provided with the KDRI designation will result in the acceptance of kidneys that would normally be discarded under the more vague ECD designation. Given that some significant donor risk factors are not captured by the index, transplant professionals may continue to accept or decline kidneys based on additional information such as anatomy, flush characteristics, histologic features and/or machine perfusion characteristics [20, 21]. Transplant centers that request OPOs to release kidneys for dual transplantation are sometimes met with resistance from the OPO either because the OPO personnel believe the kidneys can be placed as singles or the kidneys do not meet UNOS criteria for dual allocation. Given evidence that UNOS criteria does not adequately identify ECD kidneys that would confer a graft survival advantage if transplanted as dual , more discriminating allocation criteria are needed. Allocation of kidneys with KDRI >2.2 should be considered as an option for dual placement, not only because the already poor outcomes can be improved with dual transplantation, but organ utilization may also improve particularly if more efficient placement is employed.
Our results are subject to the limitations inherent in observational data. Because kidney transplant recipients are often not randomly selected to receive dual kidneys, it is possible that they are in some unmeasured way systemically less (or more) healthy than those that received single kidneys. There is the possibility for residual confounding as a result of donor, recipient or transplant factors, not included in the analysis (not available in SRTR dataset), such as some donor biopsy characteristics, transplant technique, and degree of immunosuppression. Lastly, registry data are somewhat limited toward gaining an understanding of the causes of graft or patient loss; as such it is difficult to assess the direct association of failures that would be more reflective of donor risk factors, recipient characteristic or the interaction of these.
We demonstrate that the lower graft survival seen with the transplantation of kidneys with KDRI >2.2 can be offset by dual kidney transplantation. Given the poor prognosis of the elderly on the waiting list and the decreasing opportunities of this demographic for transplantation from deceased donors, dual kidney transplantation of high-KDRI kidneys should be considered to improve outcomes. More efficient allocation directing high-KDRI kidneys to dual transplantation should be evaluated as a way to maximize kidney utilization.
The data reported here have been supplied by the Minneapolis Medical Research Foundation as the contractor for the Scientific Registry of Transplant Recipients (SRTR). The interpretation and reporting of these data are the responsibility of the author(s) and in no way should be seen as an official policy of or interpretation by the SRTR or the US Government.
The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.