Kidney Preservation Times, Donor Types, and Long-Term Outcomes

It was only in the 1960s with the rapid development of clinical kidney transplantation that the preservation of organs for storage transportation and sharing between centers began to be developed for clinical use. The year 1967 saw the introduction of successful perfusion preservation in an area dominated by the work of Folkert Belzer (1), followed two years later by the description of simple cold storage using a specialized solution, a concept championed by Geoffrey Collins (2). Variants on these approaches were developed over the years, including the introduction of University of Wisconsin (UW) solution (3) developed in Belzer's department in the mid 1980s. However, although periods of storage for as long as 48 hours could be achieved for even such notoriously difficult organs as the liver (4), the longer preservation times achieved under ideal laboratory conditions were achieved at a cost of ever increasing functional impairment. Despite the advance represented by UW solution in cold storage, Belzer remained an enthusiast for machine perfusion preservation and his unit produced exceptional immediate function figures for kidneys preserved using this technique (5). However, safe long-term organ preservation has not been achieved and defining, understanding, and minimizing the deleterious effects of organ storage before transplantation remains an important topic. This is of particular importance if organs are to be transported and shared in large scale regional, national, or supranational organ allocation systems to optimize tissue matching and allow highly sensitized patients to be transplanted. The relationship between immediate function and long-term graft outcome has been extensively studied. Pfaff reported a clear link in 586 renal transplant recipients between delayed renal allograft function (DGF) and graft survival; however, although DGF increased with longer cold ischemia paradoxically, this did not translate into a clear relationship between cold ischemia time and graft survival (6). The definition of delayed graft function was further refined with the introduction of the concept of slow graft function (SGF), which was shown in 896 cadaveric kidney transplants to be associated with cold ischemia time >24 hours, with a suggestion that even SGF would have an adverse effect on graft survival (7). A meeting addressing this topic in Scottsdale, Arizona in 2000 brought together 60 transplant clinicians and examined the issues raised in relation to DGF and SGF (8). They concluded that multiple factors contributed to DGF, identifying donor factors such as the process of brain death and donor age along with preservation time, as well as recipient and immune related variables with some factors having impact on long-term survival and others possibly including cold ischemia time being less important. DGF is thus seen as both an outcome measure in transplantation and also as a predictor of graft survival but with a multifactorial etiology. A systematic review of published work has subsequently suggested that pulsatile machine perfusion has a lower relative risk of DGF than cold storage, but comments on the relatively poor evidence base in terms of large good-quality studies, while suggesting that there may be a long-term graft survival advantage for machine perfusion although not demonstrating any advantage in one-year graft survival (9). These experimental and clinical studies clearly concentrated initially only on the short-term issue of successful preservation as demonstrated by a functioning graft, but did not examine long-term function as their primary endpoint. It is this area that is addressed by Opelz and Dohler in this issue of Transplantation, using the formidable volume of data available to the Collaborative Transplant Group. They were able to examine the correlation between long-term outcomes with preservation techniques and storage times over a fifteen year period in 91,674 kidney transplants. All of the usual provisos in large-scale registry studies apply in that the analyses are restricted by the data originally collected and unfortunately no data were available for analysis of delayed graft function. Nonetheless, they were able to analyze the impact of cold ischemia time and preservation technique on long-term outcome with clear advice and information for the transplant community. They show no difference in graft outcome provided ischemia times are kept below 18 hours, and advocate that kidneys should be transplanted within this time window. Beyond 18 hours, they find a progressive impact on graft survival, although the use of UW solution seems to ameliorate but not avoid this effect. Similar findings have been observed in analysis of the data held by UK Transplant, with a cutoff point of 20 hours being associated with significantly increased risk of transplant failure (graft loss or patient death) in deceased donor transplantation such that there is an 8% transplant survival difference at one year posttransplant (personal communication, C. Rudge and R. Johnson, UK Transplant, unpublished data). No obvious advantage is seen for machine perfusion but the number of cases is relatively small being only 2.6% of the data. As in other previous and smaller analyses, they emphasize the increased risk of graft failure with older donors, poor tissue-typing matches, marginal donors, and nonheartbeating donors. Recent years have seen a commendable drive towards shorter ischemia time for all categories of donor kidneys and this paper certainly supports using 18 hours as an upper limit in all cases to optimize graft survival. Properly structured studies are still needed to confirm the suggestion from many earlier studies that machine perfusion offers advantages in special circumstances such as nonheartbeating donors or extended criteria donors. Certainly there is nothing to support this contention in the present study in terms of long-term graft outcome, but the group is relatively small and did not allow statistically significant subanalyses to examine the suggested benefit from machine perfusion, although the actual effect observed was of a lower survival rate. The importance of donor age and other causes of donors being classified as “extended criteria” on long-term function is emphasized, although there did not seem to be an additional harmful effect of storage within the 18-hour limit, at least in terms of final graft outcome. The analysis is also able to compare living unrelated grafts with cadaveric grafts where ultrashort ischemia times were achieved; it also demonstrates that the living donor grafts still retain a significant advantage emphasizing the importance of additional factors in living donors and the avoidance of the harmful effects of brainstem death (10). The absence of data on delayed graft function is disappointing because this can be argued to be an additional important factor in short term management. Delayed graft function can complicate management, prolong hospital stay, increase the need for dialysis, increase the need for biopsies and other interventions and increase overall transplant costs featuring in the overall cost benefit ratio for the transplant procedure. In the absence of data from the large registries in this area, the answer will need to be drawn from properly structured clinical trials, which may yet show an important role for even shorter ischemic times and perhaps even for continuous perfusion systems be they hypothermic or normothermic. The advantages of human leukocyte antigen matching for long-term function remain and the study lends continued support to the case for inclusion of this as an important parameter in organ allocation systems, provided the 18-hour target window for cold ischemia time is achieved.