Begin typing your search above and press return to search.
Volume: 19 Issue: 1 January 2021


Machine Perfusion Decreases Delayed Graft Function in Donor Grafts With High Kidney Donor Profile Index

Objectives: Kidney transplant is the optimal treatment for patients with end-stage renal disease. The effects of using machine perfusion for donor kidneys with varying Kidney Donor Profile Index scores are unknown. We sought to assess the impact of machine perfusion on the incidence of delayed graft function in different score groups of kidney grafts classified with the Kidney Donor Profile Index.

Materials and Methods: We conducted a retrospective analysis from January 2008 through September 2017 of adult recipients (≥ 18 years old) undergoing kidney-only transplant from deceased donors. All transplant recipients were followed until December 2017. Reci­pients who received multiorgan transplants or kidneys from living donors were excluded from our analyses. Recipients were divided according to 5 donor categories of Kidney Donor Profile Index scores (0-20, 21-40, 41-60, 61-80, and 81-100). Logistic regression analysis was performed for each score group to determine the effects of machine perfusion on development of delayed graft function within each score group.

Results: Our study included 101 222 recipients who met the inclusion criteria. Multivariate analysis revealed that machine perfusion was associated with significantly decreased development of delayed graft function only in donors with high-risk profiles: the 61 to 80 score group (odds ratio = 0.83; confidence interval, 0.78-0.89) and the 81 to 100 score group (odds ratio = 0.72; confidence interval, 0.67-0.78).

Conclusions: Machine perfusion is beneficial in reducing delayed graft function only in donor kidneys with a higher risk profile.

Key words : Deceased-donor renal transplant, End-stage renal disease, Graft quality, Kidney transplantation


Kidney transplant is the optimal treatment for patients with end-stage renal disease.1 Ever since the development of kidney transplant, the importance of ensuring successful preservation of the organ between retrieval and implantation has been recognized. Two approaches have been developed to limit ischemia-reperfusion injury: cold storage (CS) and pulsatile machine perfusion (MP). The main rationale behind the use of MP is the belief that MP leads to a reduced rate of delayed graft function (DGF).2-4 Delayed graft function causes a need for continuing dialysis, which can result in longer hospitalization and increased costs and which is associated with poorer long-term outcomes.5 Attempts to prevent DGF through the use of MP may therefore be justified in both clinical and economic terms. It is generally accepted that only a subset of kidneys draws significant benefit from MP, but the particular subset that would yield the most cost-effective benefit has not been clearly identified. In December 2014, the Kidney Allocation System was implemented and allocation based on the Kidney Donor Profile Index (KDPI) was introduced, replacing the previous donor quality metric definitions (standard criteria donor [SCD] and extended criteria donor [ECD]).

Because the choice of CS or MP represents one of the few tangible parameters that the transplant surgeon can control to improve outcomes, we sought to assess the impact of the use of MP on the incidence of DGF in different groups of kidney grafts classified by KDPI. We hypothesized that MP would decrease the rate of DGF only when used in grafts with high KDPI scores.

Materials and Methods

We conducted a retrospective analysis from January 1, 2008 to September 30, 2017 of adult recipients (≥ 18 years old) of kidney-only transplant. All transplant recipients were followed until December 1, 2017. Recipients who received multiorgan transplants or received kidneys from living donors were excluded from our analyses.

Donor factors recorded for analyses included KDPI, preservation of the kidney via MP or CS, and cold ischemia time (CIT, measured in hours; defined as the time between flushing of donor organ with cold solution until removal from ice just before anastomosis in the recipient).

Recipient factors recorded for analyses included age, sex, race/ethnicity, body mass index, history of diabetes mellitus, previous kidney transplant, peripheral vascular disease (as recorded), calculated panel reactive antibody (CPRA), patient on dialysis before transplant, years on dialysis, human leukocyte antigen-DR (HLA-DR) mismatch with donor, and length of stay (in days) for posttransplant hospital care. Delayed graft function was defined as the recipient requiring dialysis for any reason (including hyperkalemia) within the first week after kidney transplant.

Our analysis was based on the Organ Procurement and Transplantation Network (OPTN) data released on December 1, 2017, which included data collected through September 30, 2017. The United Network for Organ Sharing, as the contractor for the OPTN, supplied these data. The interpretation and reporting of these data were the authors’ responsibility and in no way should be seen as an official policy of or interpretation by the OPTN or the US Government. The University of Washington Human Subjects Division deems the OPTN database as de-identified and publicly available and thus not human subject data. Therefore, this study was exempt from human subject review.

Statistical analyses
Continuous variables are presented as median and 25% interquartile range, and categorical variables are presented as percentages. Values were recorded from transplant recipient registration forms; when missing, values were from transplant candidate registration forms. Less than 1% of the CIT values were missing; the missing values were imputed by linear regression with distance and regional sharing as the variables. Diabetes mellitus, peripheral vascular disease, and years on dialysis had less than 0.5% missing data. The missing variables were imputed either by giving the majority variable or the mean value, as appropriate. Sensitivity analysis was conducted for all missing values with no change in the final results.

Analysis of variance or the Kruskal-Wallis test, as appropriate for distribution, was used to compare the demographic data between 5 KDPI groups for continuous variables. We used chi-square analysis for categorical variables. Logistic regression analysis was performed to determine the odds ratio (OR) for factors that could contribute to development of DGF after transplant. All results were considered significant at P < .05. All statistical analyses were performed using JMP-Pro version 14.0.0 (SAS Institute, Inc., Cary, NC, USA).


Demographic characteristics of donors and recipient (with N = 101 222 recipients meeting inclusion criteria) are shown in Table 1. Slightly fewer kidneys underwent MP (n = 45 875; 45.3%) versus placed in CS (n = 55 347; 54.7%) as a preservation method. Fewer recipients received kidneys from the 81 to 100 KDPI score group (14.7%) as opposed to approx­imately 20% for each of the other groups. Few kidneys (3.7%) had CIT greater than 36 hours. Of the recipients, only 5141 (5.1%) had CPRA ≥ 99, only 19 247 (19.0%) had 0 HLA-DR mismatches, and most recipients were on maintenance dialysis (82 170; 81.2%). Delayed graft function developed in 26 922 recipients (26.6%).

Recipients were divided according to 5 donor KDPI score groups (scores 0-20, 21-40, 41-60, 61-80, and 81-100) (Table 2). The percentage of recipients who received kidneys that had undergone MP and or had CIT > 36 hours increased as the KDPI levels increased. Recipient age, presence of diabetes mellitus, presence of peripheral vascular disease, 2 HLA-DR mismatches, and DGF significantly increased as KDPI levels increased. The percentage of recipients with CPRA ≥ 99, receipt of a prior kidney transplant, and 0 HLA-DR mismatches significantly decreased as KDPI levels increased. The proportion of recipients with DGF significantly increased with higher KDPI scores.

Logistic regression analysis was performed on each KDPI group to determine the effects of MP on the development of DGF for that level of KDPI (Table 3). All other donor and recipient factors in Table 2 except the categorical factor of dialysis (due to a strong association with years on dialysis) were controlled for in each logistic regression analysis. In the univariate analysis, MP significantly decreased the development of DGF in all KDPI groups except for the 0 to 20 KDPI group. In the multivariate analysis, MP was only associated with significantly decreased development of DGF in the higher KDPI score groups of 61 to 80 (OR = 0.83, confidence interval, 0.78-0.89) and 81 to 100 (OR = 0.72, confidence interval, 0.67-0.78) (Figure 1).

The length of stay for recipients who developed DGF classified by KDPI group is shown in Table 4. For all groups, recipients who developed DGF had a significantly increased length of stay of 3.2 to 3.7 days longer than recipients who did not develop DGF.


Previous clinical studies have tried to address which patient population might benefit the most from MP versus CS when transplanting ECD or donation after circulatory death (DCD) grafts.6-9 In December 2014, the new Kidney Allocation System was implemented and KDPI was introduced to replace the previous donor quality metric (SCD vs ECD). We sought to assess the impact of the use of MP on the incidence of DGF in different groups of kidney grafts classified by KDPI. The results of our study demonstrated that MP decreased DGF for kidneys from donors with KDPI > 60 compared with use of CS. A decrease in DGF can potentially lead to a shorter length of hospital stay, which may translate to lower financial burden. This is the first study to date to analyze KDPI as a dependent variable on the effects of MP on DGF.

There is a continuing and growing disparity between the number of kidneys available for transplant and the number of patients on wait lists. One possibility to increase the supply is to expand donor criteria.10,11 However, kidneys from ECDs tend to have higher rates of primary nonfunction, increased development of DGF, and reduced long-term survival versus kidneys from SCDs.11-14 The choice of CS or MP represents one of the few tangible parameters that the transplant surgeon can control to improve outcomes. Machine perfusion can reduce ischemia-reperfusion injury and provides an opportunity to evaluate kidney graft quality before implantation. However, a clear identification is not yet available of the subset of donor kidneys that would yield the most cost-effective benefit with the use of MP. Three meta-analyses that compared MP versus CS found that MP could reduce the rate of DGF but that the 1-year graft and patient survival rates were not different.9,15,16 None of these studies focused on good quality kidneys. A different meta-analysis of ECDs by Schold and associates17 showed that MP was associated with reduced incidence of DGF and increased 1-year graft survival. In the study, the investigators examined the Scientific Registry of Transplant Recipients database from 1994 to 2003 to compare MP with CS in ECD kidneys and found that rates of DGF were 20% with MP and 28% with CS. The study also examined paired transplanted kidneys and found that MP preservation significantly decreased the DGF rate compared with CS (19% vs. 26%; P < .001). Our group previously described the best utilization of MP for DCD kidneys.18 We identified that use of MP for DCD kidneys from donors less than 50 years of age provided no clinical benefit and in fact may have increased CIT.18

With the recent introduction of KDPI as a surrogate for graft quality, we analyzed all deceased-donor kidney transplant procedures performed in the United States between 2008 and 2017. After controlling for multiple variables, we identified that MP has a clear benefit in decreasing DGF in those kidneys with a KDPI > 60. This finding could have an important clinical and possibly financial impact for transplant centers. It is important to emphasize that the intention of our study was not to make a cost-effective analysis. A previous meta-analysis demonstrated poor evidence of economic benefits with MP.5 Economic benefits may arise from 2 sources: the short-term impact of reduction of DGF and the associated resource use and costs and the long-term impact on graft loss.

We showed that the length of posttransplant hospital stay was significantly lower in all KDPI score groups when patients did not develop DGF. Shorter hospital stay can be used as a surrogate of lower cost.

Although the present study was inherently limited by the retrospective nature of a review of a registry database, it is the largest data set available to assist in formulating guidelines. In addition, the transplant registry database does not allow us to identify those patients who received hemodialysis during the first week due to hyperkalemia; therefore, the rate of DGF might be overestimated. It was not the intention of this study to look at patient and graft survival, which ultimately is the main outcome that will lead to the implementation of a new tool. However, previous studies have demonstrated the detrimental effects of DGF on graft survival.19,20

Since the Kidney Allocation System with the KDPI was implemented, this is the first study to identify the particular donor kidneys (and ultimately, patients) that would benefit from the use of MP.

In summary, the results of the present study should encourage the use of MP versus CS on kidneys from donors with a higher risk profile (KDPI > 60). These data might also assist organ pro­curement organizations in improving resource allocation and to potentially decrease costs.


  1. Wolfe RA, Ashby VB, Milford EL, et al. Comparison of mortality in all patients on dialysis, patients on dialysis awaiting transplantation, and recipients of a first cadaveric transplant. N Engl J Med. 1999;341(23):1725-1730.
    CrossRef - PubMed
  2. Ploeg RJ, van Bockel JH, Langendijk PT, et al. Effect of preservation solution on results of cadaveric kidney transplantation. The European Multicentre Study Group. Lancet. 1992;340(8812):129-137.
    CrossRef - PubMed
  3. Cho YW, Terasaki PI, Cecka JM, Gjertson DW. Transplantation of kidneys from donors whose hearts have stopped beating. N Engl J Med. 1998;338(4):221-225.
    CrossRef - PubMed
  4. Wight JP, Chilcott JB, Holmes MW, Brewer N. Pulsatile machine perfusion vs. cold storage of kidneys for transplantation: a rapid and systematic review. Clin Transplant. 2003;17(4):293-307.
    CrossRef - PubMed
  5. Wight J, Chilcott J, Holmes M, Brewer N. The clinical and cost-effectiveness of pulsatile machine perfusion versus cold storage of kidneys for transplantation retrieved from heart-beating and non-heart-beating donors. Health Technol Assess. 2003;7(25):1-94.
    CrossRef - PubMed
  6. Watson CJ, Wells AC, Roberts RJ, et al. Cold machine perfusion versus static cold storage of kidneys donated after cardiac death: a UK multicenter randomized controlled trial. Am J Transplant. 2010;10(9):1991-1999.
    CrossRef - PubMed
  7. Treckmann J, Moers C, Smits JM, et al. Machine perfusion versus cold storage for preservation of kidneys from expanded criteria donors after brain death. Transpl Int. 2011;24(6):548-554.
    CrossRef - PubMed
  8. Jiao B, Liu S, Liu H, Cheng D, Cheng Y, Liu Y. Hypothermic machine perfusion reduces delayed graft function and improves one-year graft survival of kidneys from expanded criteria donors: a meta-analysis. PLoS One. 2013;8(12):e81826.
    CrossRef - PubMed
  9. Deng R, Gu G, Wang D, et al. Machine perfusion versus cold storage of kidneys derived from donation after cardiac death: a meta-analysis. PLoS One. 2013;8(3):e56368.
    CrossRef - PubMed
  10. Port FK, Bragg-Gresham JL, Metzger RA, et al. Donor characteristics associated with reduced graft survival: an approach to expanding the pool of kidney donors. Transplantation. 2002;74(9):1281-1286.
    CrossRef - PubMed
  11. Metzger RA, Delmonico FL, Feng S, Port FK, Wynn JJ, Merion RM. Expanded criteria donors for kidney transplantation. Am J Transplant. 2003;3 Suppl 4:114-125.
    CrossRef - PubMed
  12. Pascual J, Zamora J, Pirsch JD. A systematic review of kidney transplantation from expanded criteria donors. Am J Kidney Dis. 2008;52(3):553-586.
    CrossRef - PubMed
  13. Daly PJ, Power RE, Healy DA, Hickey DP, Fitzpatrick JM, Watson RW. Delayed graft function: a dilemma in renal transplantation. BJU Int. 2005;96(4):498-501.
    CrossRef - PubMed
  14. McLaren AJ, Jassem W, Gray DW, Fuggle SV, Welsh KI, Morris PJ. Delayed graft function: risk factors and the relative effects of early function and acute rejection on long-term survival in cadaveric renal transplantation. Clin Transplant. 1999;13(3):266-272.
    CrossRef - PubMed
  15. Lam VW, Laurence JM, Richardson AJ, Pleass HC, Allen RD. Hypothermic machine perfusion in deceased donor kidney transplantation: a systematic review. J Surg Res. 2013;180(1):176-182.
    CrossRef - PubMed
  16. Bathini V, McGregor T, McAlister VC, Luke PP, Sener A. Renal perfusion pump vs cold storage for donation after cardiac death kidneys: a systematic review. J Urol. 2013;189(6):2214-2220.
    CrossRef - PubMed
  17. Schold JD, Kaplan B, Howard RJ, Reed AI, Foley DP, Meier-Kriesche HU. Are we frozen in time? Analysis of the utilization and efficacy of pulsatile perfusion in renal transplantation. Am J Transplant. 2005;5(7):1681-1688.
    CrossRef - PubMed
  18. Cantafio AW, Dick AA, Halldorson JB, Bakthavatsalam R, Reyes JD, Perkins JD. Risk stratification of kidneys from donation after cardiac death donors and the utility of machine perfusion. Clin Transplant. 2011;25(5):E530-E540.
    CrossRef - PubMed
  19. Zeraati AA, Naghibi M, Kianoush S, Ashraf H. Impact of slow and delayed graft function on kidney graft survival between various subgroups among renal transplant patients. Transplant Proc. 2009;41(7):2777-2780.
    CrossRef - PubMed
  20. Fonseca I, Teixeira L, Malheiro J, et al. The effect of delayed graft function on graft and patient survival in kidney transplantation: an approach using competing events analysis. Transpl Int. 2015;28(6):738-750.
    CrossRef - PubMed

Volume : 19
Issue : 1
Pages : 8 - 13
DOI : 10.6002/ect.2019.0139

PDF VIEW [140] KB.

From the Division of Transplantation, Department of Surgery, University of Washington, Seattle, Washington, USA
Acknowledgements: The authors have no sources of funding for this study and have no conflicts of interest to declare. This work was made possible by the Clinical and Bio-Analytics Transplant Laboratory (CBATL) in the Department of Surgery at the University of Washington Medical School.
Corresponding author: Martin I. Montenovo, Department of Surgery, Box 356410, Seattle, WA 98195, USA