Objectives: Since the development of Kidney Donor Profile Index, outcome differences based on number of donated organs per donor have not been evaluated.
Materials and Methods: We retrospectively analyzed data from the United Network for Organ Sharing national database, which identified 176311 adult renal transplant recipients from 2000 to 2019 with a deceased donor kidney from a kidney-only donor, from a donor of kidney and liver but no other organs, or from a multiorgan donor. Graft failure and transplant recipient survival were primary outcomes. A multivariate Cox proportional hazards model controlled for Kidney Donor Profile Index differences.
Results: Overall, multiorgan donors had a lower Kidney Donor Profile Index versus other donor types (odds ratio, 0.042; P < .001). Kidneys from donors with a higher Kidney Donor Profile Index were 95% less likely to be procured with other organs (P < .001). The recipient and graft survival rates for kidney transplants from kidney-only donors and from donors of kidney and liver but no other organs were 76% and 70%, respectively, whereas recipient and graft survival rates for kidney transplants from multiorgan donors were approximately 82% and 77%, respectively, at 5 years.
Conclusions: After adjustment for the Kidney Donor Profile Index, the recipients of multiorgan donor grafts demonstrated superior outcomes for graft survival and mortality compared with kidney-only donors or kidney and liver only donors. The multiorgan donor status may be an additional consideration in future renal allocation calculators.
Key words : Graft allocation, Kidney-only donor, Kidney Donor Profile Index, Kidney transplant outcomes, Multiorgan donor
Renal transplant improves survival and may reduce the financial burden for people with end-stage renal disease, and this is also true for recipients who return to dialysis after transplant.1-3 However, there remains a persistent and profound shortage of donor kidneys. In 2020, there were 22 000 renal transplants performed in the United States, and yet 90 000 candidates remained on the wait list.4,5 Despite this unmet demand, the disuse (discard) rate of recovered organs has remained at 20%.6 Some donor kidneys are discarded appropriately (eg, poor organ quality); however, some donor kidneys that would otherwise meet the criteria for transplant may be subject to discard because of ineffective allocation practices such as prolonged ischemia time or regulatory barriers. In addition, transplant centers are discouraged from accepting donor kidneys designated as extended criteria (ECD). Implant of suboptimal grafts may affect insurance reimbursement to hospitals if the result is a suboptimal transplant recipient outcome, delayed graft function, or prolonged hospital stay.7 With regard to the high demand and short supply, efficient kidney allocation continues to be a topic of intense research to optimize utilization and outcomes.
Before procurement, donor allografts are assigned a Kidney Donor Profile Index (KDPI), which represents the quality of the deceased donor kidneys relative to other types of recovered kidneys. The KDPI is a numerical estimate that combines donor clinical parameters and demographic characteristics. A lower KDPI is associated with increased graft quality and longevity and in turn influences the acceptance decision of a donor offer.8 Underutilization rates for donor kidneys are particularly high for a KDPI of 85% or greater.6 However, KDPI does not account for the number of various other organs that may have been procured from the kidney donor.
Kidney graft procurement from a multiple-organ donor (MOD) is a more complex and lengthy process compared with procurement from single-organ donors. Donor kidneys are routinely the final organ explanted in any MOD procurement. Few analyses have evaluated the outcomes of MOD transplants versus transplants from kidney-only donors (KOD). There are 2 studies that have calculated higher rates of delayed graft function in KOD transplants but no significant difference in overall recipient and graft survival compared with MOD transplants.9-11 To date there have been no evaluations of transplant outcomes of kidneys from the combined donation of kidney and liver only (KLOD). After adjustment for KDPI, we hypothesize that MOD renal grafts may have better outcomes than grafts from other donor types. Here, we compared the outcomes of KOD transplants versus MOD or KLOD transplants. Primary endpoints included recipient and graft survival among the 3 groups. Secondary endpoints included recipient, donor, and transplant characteristics among the 3 groups.
Materials and Methods
This retrospective observational study was performed with data from the United Network for Organ Sharing national database in which 176311 adult kidney transplant recipients were identified from January 2000 to December 2019. Of this cohort, 85% (N = 150570) of kidney transplant recipients had received their donor kidney from a procurement surgery that included other organs such as liver, lung, heart, pancreas, and/or intestine. Recipient-related data included age, sex, ethnicity, prior transplant, body mass index (BMI, calculated as weight in kilograms divided by height in meters squared), days on the wait list, glomerular filtration rate, and calculated panel reactive antibody score (cPRA) at the time of transplant. Donor-related data included age, sex, ethnicity, creatinine level, history of hypertension, KDPI, and donor status with regard to donation after cardiac death (DCD) and ECD. Transplant-related variables included human leukocyte antigen (HLA) mismatch level, cold ischemia time (CIT), and organ sharing status (local, regional, or national). The KDPI corresponds to percentiles of the Kidney Donor Risk Index for donor kidneys, which seeks to assess posttransplant outcomes according to 14 donor and transplant factors known to be independently associated with graft failure or death.12,13
The transplant recipients and donors were categorized into the following 3 groups, according to type of organ(s): (1) kidneys from donors from whom only kidneys were procured (KOD) (N = 25741; 15%), (2) kidneys procured from donors who donated a kidney and liver but no other organs (KLOD) (N = 67207; 38%), and (3) kidneys procured from multiple-organ donors (MOD) (N = 83636; 47%). Only 0.6% of KOD transplants were from single-kidney procurements, so we were unable to compare outcomes for donors of 1 kidney versus 2 kidneys. Recipient, donor, and transplant characteristics were analyzed among the aforementioned organ donor categories with analysis of variance or the Kruskal-Wallis test for continuous variables and chi-square test and Fisher exact test for categorical variables according to sample size and distribution of variables.
A Kaplan-Meier survival curve was generated using the Kaplan-Meier Product Limit method to compare graft and transplant recipient survival (in days) across all 3 donor categories. In the survival analysis, transplant recipient death and graft failure were primary endpoints. If these endpoints were not met or if vitality and graft status were unknown, then those data were censored on the last follow-up or last day of the analysis. Multivariate Cox regression analyses were performed for clinically suspected risk factors and dummy variables for KLOD and MOD transplants. The potential risk factors included recipient and donor demographic characteristics, clinical factors, and donor characteristics linked to donor quality like KDPI, as well as transplant-related variables. The dummy variable for the kidney allocation system (after 2014) was also included to control for the kidney allocation system effect. A last step was a logistic regression to investigate the determinants of MOD with donor and recipient characteristics as the explanatory variables. In all analyses, P < .5 was deemed significant. This study was exempt from Institutional Review Board approval.
Transplant recipients of MOD kidneys were significantly younger and had prior transplant history (16% MOD vs 10% KOD and 11% KLOD; P < .001) and higher cPRA (27.94 MOD vs 18.57 KOD and 20.28 KLOD; P < .001). The KOD recipients experienced fewer days on the wait list (704 days KOD vs 713 days KLOD and 715 days MOD; P < .001) (Table 1).
In terms of donor demographic characteristics, MOD donors were more likely to be African American or Hispanic compared with KOD or KLOD donors. Creatinine for KOD donors at time of transplant was significantly lower compared with other donor types. The MOD donors were less likely to be DCD (3% MOD vs 52% KOD and14% KLOD; P < .001) or ECD (6% MOD vs 20% KOD and 29% KLOD; P < .001). The MOD donors also had lower KDPI (0.33 MOD vs 0.56 KOD and 0.56 KLOD; P < .001) and lower mean CIT. The KOD donors had the longest mean CIT (over an hour) and the highest mean HLA mismatch level (Table 1). For MOD, the kidneys were most likely to be procured with the liver followed by the heart, pancreas, and intestine (Figure 1).
For transplant and recipient outcomes, MOD kidneys incurred longer graft survival times compared with KOD or KLOD kidneys. The 5-year graft survival rate for MOD was 77% compared with 70% for KOD and KLOD (Figure 2). This pattern is also evident for transplant recipient survival. The 5-year recipient survival rate for MOD was 82% versus 76% for KOD and KLOD (Figure 3). After adjustment for KDPI, graft failure rates between KOD and KLOD were not significantly different (hazard ratio [HR], 1.004; P = .765; 95% CI, 0.980-1.028). However, MOD incurred a lower risk of graft failure compared with KOD after adjustment for KDPI (HR, 0.964; P = .005; 95% CI, 0.940-0.989), as well as a lower risk of mortality (HR, 0.969; P = .036; 95% CI, 0.942-0.998). Higher KDPI scores were significantly associated with increased risk of graft failure (HR, 2.678; P < .001; 95% CI, 2.593-2.765) (Table 2), whereas KLOD recipients had an increased mortality risk compared with KOD recipients. Higher KDPI scores were associated with higher risk of transplant recipient mortality (HR, 2.897; P < .001; 95% CI, 2.793-3.005) (Table 3).
After adjustment for covariates, MOD kidneys were more likely to be donated to younger, female, and African American or Hispanic recipients who had a higher cPRA level, a higher BMI, longer wait times on the donor list, and fewer HLA matches. The MOD kidneys were more likely to be locally shared and less likely to be regionally shared. The KDPI scores for MOD were also significantly lower (odds ratio, 0.042; P < .001; 95% CI, 0.040-0.044) (Table 4).
This retrospective observational analysis inves-tigated outcomes from 2000 to 2019 of 176311 kidney transplant recipients of kidneys from deceased donors with KOD, KLOD, or MOD status. After adjustment for KDPI, the KOD and KLOD kidney transplant recipients had similar graft and recipient survival rates, whereas recipients of MOD kidneys experienced a reduction in graft failure and recipient mortality rates. The current KDPI model does not account for donor kidney procurement status. This donor status may also affect kidney transplant outcomes and may be used to more efficiently allocate donor kidneys.
Literature on survival outcomes from MOD grafts is sparse, but there are a few analyses that suggest superior rates of mortality and graft survival similar to the results of our investigation. In a European observational study, Smits and colleagues in 1996 calculated significantly better graft survival up to 5 years after transplant for MOD versus KOD transplants (58% MOD vs 46% KOD; P < .001) and a 1.28-fold higher risk of graft loss in KOD versus MOD transplants.14 However, these graft survival rates are low compared with modern rates. Today, 5-year deceased donor graft survival rates are over 70%, likely due to improved procurement and preservation techniques as well as the addition of tacrolimus to the immunosuppression regimen in the late 1990s.15,16 A more recent investigation on orthotopic heart transplant concluded that a greater number of organs recovered from a single donor was independently associated with reduced recipient mortality.17 These investigations postulated that favorable outcomes were attributed to the likelihood of MOD donors to be healthier versus KOD donors.
The procurement process could contribute to this difference, particularly extraction and overall ischemia time. Longer CIT from flush to ice and longer warm ischemia time from ice to reperfusion may cause delayed graft function and may reduce long-term survival rates for grafts and recipients.18-22 Within the CIT timeframe, the interval from aortic cross-clamp to placement of the kidneys in cold preservation solution on the back table is the organ extraction time interval. During multiple-organ procurements performed by multiple surgical teams, kidneys are frequently the last solid organ recovered, which could lead to rewarming and an additional ischemic insult. Osband and colleagues concluded that longer renal extraction time was correlated with higher incidence of delayed graft function for extraction times longer than 1 hour but with no differences in overall graft survival.22 In a more recent investigation, Goldsmith and colleagues conducted a retrospective analysis on DCD donors and calculated a 2-hour longer extraction time for MOD compared with KOD procurement, but the longer extraction time did not lead to differences in graft or recipient outcomes, including delayed graft function or 1-year graft and recipient survival rates.23 In this investigation MOD were younger donors with higher graft glomerular filtration rates at the time of transplant, which likely offsets the detrimental effects of a longer organ extraction time.23 Although MOD procurements correlate with longer renal extraction times, MOD interestingly had 1.5 hours less CIT compared with KLOD and KOD in this analysis. The longer CIT associated with KOD and KLOD may cause a cascade of reactions, magnified by reperfusion, leading to graft damage and activation of the immune response. It is unclear why KOD and KLOD incurred significantly longer CIT compared with MOD, but procedurally speaking this is unlikely to be related to longer extraction time.
In the unadjusted data, MOD transplants of ECD kidneys were uncommon (6%), whereas 20% of KOD and 29% of KLOD kidneys were ECD status. The ECD status is defined as a kidney from a deceased donor at least 60 years old or a deceased donor from 50 to 59 years old with at least 2 of the following characteristics: cerebrovascular accident as cause of death, serum creatinine >1.5 mg/dL, or history of hypertension. This criteria standard was created in 2002 to reduce discard and rejection rates. However, these donor kidneys have a higher risk of graft failure rate, so ECD kidneys are usually reserved for older recipients who would otherwise be subject to longer wait list times.24 Cholewa and colleagues have reported that KOD transplants have a higher percentage of ECD kidneys and that KOD recipients were more likely to experience delayed graft function, longer hospital stays, and worse long-term excretory function based on estimated glomerular filtration rate versus MOD recipients.10 Katsaros and colleagues analyzed the United Network for Organ Sharing database from 2000 to 2016 with propensity score matching and calculated a significantly higher rate of delayed graft function for KOD versus MOD.25 Although delayed graft function was not an outcome included in their analysis, there were significantly more ECD kidneys from KOD and KLOD donors compared with MOD donors. This could contribute to lower KDPI scores and better overall survival of grafts and recipients in the MOD setting.
Significantly fewer MOD were DCD (3% MOD vs 52% KOD and 14% KLOD). Previous investigations have calculated a significant increase in the rate of delayed graft function and primary nonfunction of DCD grafts compared with donation after brain death grafts, although long-term graft function and recipient survival were equivocal.26,27 Therefore, a significant increase in DCD transplants was less likely to contribute to the lower survival rate for KOD and KLOD compared with MOD transplants.
Interestingly, in the adjusted data, MOD recipients were more likely to have a slightly higher BMI at the time of transplant (odds ratio, 1.002; P = .048; 95% CI, 1.000-1.005). The MOD recipients had a mean BMI of 27.85, which is defined as overweight, and this is surprising because obesity rates are rising among patients with end-stage renal disease.28 Segev and colleagues calculated that a significantly decreased chance of undergoing a kidney transplant is associated with increasing degrees of obesity, and obese transplant candidates were more likely to be bypassed on the wait list when a matched organ became available.29 However, morbidly obese recipients have 3-year graft and mortality rates similar to the rates for nonobese recipients, although obesity is associated with higher likelihood of a surgical complication and longer hospital stay.30 In this analysis, BMI was not a factor for differences in renal transplant outcomes.
The present study is the first analysis to compare renal transplant outcomes based on the number of donated organs after adjustment for KDPI to assess donor status as a risk factor for transplant. One of the strengths of this investigation is the use of a large national database spanning over a decade to perform calculations with appropriate power, and it includes data from the most recent update of the kidney allocation system, which was established in 2014. However, there are limitations to this analysis. Not all donor and recipient variables could be included in this analysis, such as delayed graft function, primary graft dysfunction, and warm ischemia time. Also, data were not available for donor organs that were initially procured for transplant but subsequently rejected or discarded. Therefore, the effects of these variables on overall graft or recipient survival could not be assessed.
Multiorgan donors were more likely to be younger and have fewer HLA mismatches and lower KDPI scores and were less likely to have ECD status compared with KOD and KLOD donors. These factors likely contributed to the superior 5-year survival rates for recipients and grafts in MOD transplants compared with KOD and KLOD transplants, and these rates were not affected by adjustment for KDPI. Donor organ source is not currently included in the KDPI calculation but may be an independent risk factor for renal transplant outcomes. The new renal transplant allocation system implemented in March 2021 may affect these outcomes, but the details are not yet evident.
Volume : 20
Issue : 1
Pages : 12 - 18
DOI : 10.6002/ect.2021.0371
From the 1University of Minnesota, Department of Surgery, Minneapolis, Minnesota; the 2University of Toledo, Department of Urology, Toledo, Ohio; the 3George Washington University, Department of Surgery, Washington, District of Columbia; the 4University of California Irvine School of Medicine, Irvine, California; and the 5Albany Medical Center, Department of Transplant Surgery, Albany, New York, USA
Acknowledgements: The authors have not received any funding or grants in support of the presented research or for the preparation of this work and have no declarations of potential conflicts of interest.
Corresponding author: Lauren Weaver, University of Minnesota, Department of Surgery, 420 Delaware St SE Minneapolis, MN 55455, USA
Phone: +1 216 406 8083
Table 1. Characteristics of Transplant Recipients, Donors, and Organs According to Donor Status: Kidney Only, Kidney and Liver Only, and Multiple Organs
Figure 1. Composition of Organs Recovered with Kidney(s) Based on the Total Number of Organs Procured from the Donor
Figure 2. Kaplan-Meier Graft Survival Estimates Based on Type of Organ Donor
Table 2. Graft Failure Statistics, Adjusted for Kidney Donor Profile Index (Without Other Covariates) by Cox Regression
Table 3. Transplant Recipient Mortality, Adjusted for Kidney Donor Profile Index (Without Other Covariates) by Cox Regression
Figure 3. Kaplan-Meier Overall Transplant Recipient Survival Estimates Based on Type of Organ Donor
Table 4. Logistic Regression for Association Between Donor and Recipient Characteristics of Kidneys From Multiple-Organ Donors