Kidney allocation: where utility and fairness meet

Transplantation is the preferred treatment option for end-stage renal disease because it offers a survival advantage over dialysis for the majority of patients [1]. Renal transplantation is saving, as well as improving, lives, but the benefits may differ among groups. Unfortunately, due to the static supply of donor organs (Figure 1, left), the disparity with the increasing numbers of patients placed on the waiting list and therewith waiting times continues to increase. The median waiting time for a deceased donor kidney within Eurotransplant (comprising 7 countries and 210 million inhabitants) is now well over 4 years, but differs significantly between countries. The allocation of the limited number of available kidneys poses a constant challenge to maintain an acceptable balance between equity of access, medical utility, efficiency in terms of logistics and finances, and lastly, balance between the participating countries. Medical utility can be measured using patient survival, years added to life compared to dialysis, kidney graft survival, quality of life or preventing the inefficient use of a scarce resource. Independent of the measure, it is often more easy to quantify utility than equity of access. Trends in kidney donors according to age (Eurotransplant 1 January 1991–1 January 2008). In the Eurotransplant kidney allocation system, all kidneys are virtually offered to the pool. Each patient receives a transparent point score on the basis of human leucocyte antigen (HLA) matching based on the ‘broad’ class I antigens and splits for class II, waiting time, mismatch probability, distance to the transplant centre, and balance for donation rates between the participating countries [2]. The paper by Glander and co-authors in this issue of Nephrol Dial Transplant challenges the inequity of access for patients with blood group O to receive a kidney transplant with a comparable median waiting time in comparison to non-O recipients [3]. In the short term, death on the waiting list is rarely the direct consequence of not receiving a transplant, but rather due to co-morbid conditions, such as ischaemic heart disease and/or diabetes. Their centre analysis shows that the inequity in access due to having blood group O with time results in significant longer median waiting times, excess mortality while waitlisted and, if they reach the front of the queue, inferior survival after transplantation. The inequity is, at least in part, an inherent consequence of the mandatory exchange rules within Eurotransplant dominated by zero antigen-mismatched kidneys allowing ABO-compatible allocation and shipment outside the region or country. A zero-mismatched kidney is fully compatible with the recipient's HLA-A, HLA-B and HLA-DR loci. The justification for this exception has been the superior kidney graft survival time for zero-mismatched kidneys compared to one to five mismatches. Another advantage that comes with preferential allocation is reduced waiting time for the fortuitous recipients of a zero-mismatched graft, including both O and non-O recipients. Basically, the current waiting times are unacceptably high, and the impact of the proposed restriction to ABO-identical allocation on waiting time, though fair, is likely to turn out small or disappointing. Even with the efficient current immunosuppressive strategies, the (slight) impact of full HLA compatibility should not be considered irrelevant for kidney graft survival, but the disadvantage for O patients with excess death while waiting is no longer justifiable. The centre analysis by Glander et al. supports the decision by the Kidney Advisory Committee and Board of Eurotransplant to implement ABO-O identical allocation of the zero-mismatched kidneys, but this change in allocation still awaits formal approval by the relevant health authorities in some countries. Another issue Glander et al. rightfully address in this context is the inherent disadvantage blood group O patients experience even if they have an offer from a live donor. The increasing awareness of unfavourable survival with prolonged time spent on the transplant waiting list [4] has, at least in some countries, resulted in an impressive increase in live donor transplants in the past decade. In the Netherlands, for instance, >50% of transplants performed nowadays are made possible only due to living donation and by active policy >50% of these patients received their transplant preemptively. The option of indirect donation via the national crossover programme, combining ABO-incompatible pairs and those with positive cross-matches has resulted in a respectable number of successful additional transplants [5]. Nevertheless, as could be expected, a substantial number of O recipients remain that may require expensive protocols to allow ABO-incompatible living donor transplants. Although the vast majority of living donors have an emotional and/or genetic relation, there is now legal opportunity for altruistic non-directed living donation in some countries. A domino-paired procedure with list exchange may offer a small additional option to O recipients, but at least facilitates additional indirect transplants as well as a transplant for a patient from the deceased donor waiting list. Despite all these efforts, however, waiting times are still between 4 and 5 years for a deceased donor kidney, somewhat lower compared to Germany, but in both countries, they remain significantly higher compared to countries with a historical ‘opting out’ legal system. It remains, however, unclear whether socioeconomical along with local cultural influences also have an impact on donation rates. The consequence of prioritization of HLA matching for all loci is that some patients have to wait longer for a suitable organ and thus accumulate on the waiting list, especially if they express HLA antigens or haplotypes not prevalent in the organ donor population. In Eurotransplant, almost all patients (97% in the period from 1996 to 2005) received a 0 or 1 HLA-DR-mismatched kidney [6]. A significant beneficial effect of HLA-A and HLA-B matching was only found in patients receiving a full HLA-DR-compatible kidney [7]. The imperative of additional HLA-B and HLA-A matching poses additional challenges to equal access that may be related to factors such as ethnicity or degree of sensitization. Compared to the general population, non-Caucasian kidney patients are not only overrepresented in the end-stage renal disease population but also underrepresented in the donor pool for deceased donor kidneys [8]. A simpler allocation system primarily based on full HLA-DR compatibility will also better serve equity of access for minorities in our society [7]. On the other hand, one should not make the mistake to change the current policy for patients in immediate need of a transplant or highly immunized patients. The acceptable mismatch programme was introduced more than 25 years ago and has significantly increased transplant rates of this cohort, defined by panel-reactive antibodies >85%, and its success depends in part on full HLA compatibility, while allowing ABO-compatible allocation [9]. This programme requires only a relative small contribution from the O donor pool, and from a utilitarian view, this is a better approach than the uncertainties related to its alternative, pre-transplant desensitization. Over the past decades, the waiting list for renal transplantation has grown significantly older, with the largest absolute and relative increase in the ≥65 years age group (Figure 1, middle). In contrast, there has been stabilization or decrease in patients aged 46–64 and <46 years, respectively [10]. There is a substantial evidence to support transplantation in the elderly but appropriately selected patients with end-stage renal disease. Transplantation is safe and provides life expectancy exceeding that on dialysis [11]. Preclusion based on age alone is considered not fair, but the difficulty of selection, taking biological age and co-morbid conditions into account, poses the greatest challenge to the principle of medical utility. Older recipients are more likely to die with a functioning transplant than younger recipients, whereas death-censored graft failure, defined as a need for re-transplantation or maintenance dialysis, is less common in older recipients [12]. Even before the now well-established Eurotransplant senior programme [13], the policy of older donors to older recipients already existed in clinical practice by local decision-making. The frequency of patients older than 55 years receiving organs from donors in the same age range was 46% higher than what would be expected by chance [12]. At present, outcome parameters in the elderly are dominated by increased death from infectious disease causes [14]. The impact of increasing age on death-censored graft outcome appears to be amplified if also the age of the kidney donor is taken into account. The therapeutic index for clinical immunosuppression appears to be even narrower in the elderly than in younger renal transplant recipients. A direct causative relation between the need to treat rejection and increased mortality may be difficult to establish, but there is no doubt that extra boluses of steroids or treatment with antibodies add significantly to post-transplant morbidity. The Eurotransplant Senior DR-compatible Program was recently launched to introduce full HLA-DR compatibility, while maintaining the ESP principle of local or regional allocation and reduced cold ischaemia times [10]. Especially in the elderly, allocation of older deceased donor kidneys based on DR-compatibility, reducing the need for rejection treatments and the additional risk of infectious complications may prove to be the preferred future approach. From a utilitarian point of view, each kidney should be transplanted into the recipient in whom it will survive the longest. Allocating young kidneys with long potential survival to elderly patients or other recipients with a much shorter life expectancy may be even considered unfair. There are, however, several issues that need to be addressed before age category-based, and in particular young-for-young, allocation can be considered. Firstly, there is an ongoing decrease in the number of available organs from younger donors, which goes up to a donor age of 46 years (Figure 1, right). Therefore, the youngest will increasingly depend on kidneys obtained from intermediate-aged donors. Secondly, even with optimal quality and matching, children are likely to need one or two re-transplant(s) and may become increasingly sensitized with each transplant performed. In the young age category, it may be the preferred approach to strive for optimal HLA matching, while allowing for ABO-compatible or ABO-A to non-A allocation for special groups, such as AB recipients and children. In summary, future allocation by age category (young, intermediate, senior) with ABO-identical and HLA-DR-compatible grafts probably best serves the delicate balance between the equity of access and medical utility for elderly and intermediate-aged transplant recipients. For the young age category, an optimal HLA matched remains preferable and thereby allowing ABO-compatible allocation. Full HLA-DR compatibility as the most important allocation criterion has the additional potential to reduce cold ischaemia time, and since the large majority of organs can be allocated locally or regionally, it will also significantly reduce transportation costs. While dealing with a life-saving but scarce medical resource, a continuous and open-minded discussion is the only way to maintain a well-balanced and socially acceptable equipoise between equity, utility and efficiency.