Donor Characteristics and Short-Term Kidney Allograft Outcomes

As the field of kidney transplantation searches for new ways to expand utilization of deceased donor kidneys while maintaining adequate outcomes, learning from past experience and deciding on the most promising avenues to expand practice continue. In the article by Helantera et al. (1), data from patients in the Scientific Registry of Transplant Recipients were used to study the association of older donor age and cold ischemia time—both known risk factors for delayed graft function as defined by the need for dialysis in the first post-transplant week; 90,717 transplants from a recent era (2010–2018) were analyzed for donor age and cold ischemia time as well as a number of other characteristics to build multivariable models that were used to adjust for those factors.

The prestated goal of this analysis was to determine the interaction between longer cold ischemia time, donor age, and donor quality. All three variables are known to be independently associated with higher risk of delayed graft function (2), and delayed graft function is known to associate with worse graft outcomes (3). The question was whether older or lower-quality kidneys were somehow more susceptible to damage from longer amounts of cold ischemia.

The authors report that cold ischemia time and donor age were independently associated with delayed graft function, which was not a surprising finding (1). They then looked at the interaction of older donor age (in broad categories of 50–64 and ≥65) with cold ischemia time.

Overall, the frequency of delayed graft function was 29% across the entire cohort, 45% in donors after circulatory death, and 25% in donors after brain death. There were substantial baseline differences between the recipients with and without delayed graft function as well as in donor factors, which were addressed with logistic regression and sensitivity analysis. Continuous variables were converted to categories, potentially reducing the power of the analysis. There was no attempt to adjust for substantial secular changes over the cohort period 2010–2018 with progressive adoption of deceased cardiac death donors, viral-infected kidney donors, and extremely long–cold ischemia time kidneys (>24 hours). A before and after sensitivity analysis for the new kidney allocation system effected December 2014 was performed. There was no attempt to account for organ procurement organization or transplant center differences in recipient and donor acceptance. Approaches such as propensity scores and instrumental variable analysis that might mitigate some of these potential limitations were not used. We must therefore be cautious in interpreting this rearview look at the national experience of kidney transplantation in the United States and seek not to view these results as permanent fixtures in deceased donor kidney acceptance going forward.

This study substantially affirms previous research that there is a modest association between longer cold ischemia time up to 24 hours and delayed graft function. There is also a J-shaped association between donor age and delayed graft function. It also demonstrates that Kidney Donor Profile Index, adopted December 2014, is associated with short-term outcomes such as delayed graft function, although the categories selected do not reflect the bins that are used in allocation for national, regional, and high kidney donor profile index (previously expanded criteria donors). Of note, machine perfusion not only has a tremendous effect on the odds ratio of delayed graft function but substantially attenuates the effect of longer cold ischemia time up to 24 hours. Again, these results are consistent with research in earlier cohorts, which gives greater confidence that these are important factors in early post-transplant outcomes, with cold ischemia time and machine perfusion as potentially modifiable factors. The effect of increased use of very long–cold ischemia time kidneys remains unanswered in this study.

This analysis is important and should help contribute to decision making when considering the risk of accepting a kidney offer for your patient. In addition, with the planned move to broader sharing of deceased donor kidneys to alleviate geographic disparities, the role of cold ischemia time will become a more important driver of transplant team decision making. What this analysis suggests is that, although cold ischemia time is always an important contributor to the risk of delayed graft function, this effect should be considered similarly regardless of the inherent donor quality (as driven by age and kidney donor profile index score). It is still noteworthy that cold ischemia time remains an independent risk factor for delayed graft function, with a higher odds ratio of delayed graft function for each additional hour of cold ischemia time.

So how does the kidney transplant community use this analysis to improve patient care and access to transplantation? One important outcome could be a reduction of organ discard. Although the community is highly focused on reducing organ discard, there is still significant and appropriate attention paid to transplant outcomes. Delayed graft function is an important predictor of outcomes. The authors make note that they did not analyze ultimate graft outcomes, deciding to focus instead on the short-term marker of delayed graft function. Although fair, they do note that generalizing this analysis may not lead to improved transplant outcomes because there may be other factors not captured in the outcome of delayed graft function that might affect the ultimate outcome.

As the United States transplant system moves to a broader sharing model for kidney allocation, it is an opportunity to examine best practices that can improve efficiency and promote better outcomes. Organ procurement organizations should develop policies and procedures to allow for the transport of kidneys “on pump” when shipped outside of a local procurement area. The costs of pumping and returning pumps across a distance need to be weighed against the cost of delayed graft function with longer transplant hospitalizations, increased burden on care providers, higher rates of rejection, and ultimately, reduced graft survival (4). We welcome this insightful analysis and hope for future studies that may continue to identify the boundaries of what is possible in deceased donor allocation.

Disclosures

Dr. Friedewald and Dr. Ho have nothing to disclose.

Funding

None.