The Dutch National Living Donor Kidney Exchange Program


  • Marry De Klerk,

    Corresponding author
    1. Department of Internal Medicine – Transplantation, Erasmus MC, University Medical Center Rotterdam, Rotterdam, The Netherlands
    2. Dutch Transplantation Foundation, Leiden, The Netherlands
    Search for more papers by this author
  • Karin M. Keizer,

    1. Dutch Transplantation Foundation, Leiden, The Netherlands
    Search for more papers by this author
  • Frans H. J. Claas,

    1. National Reference Laboratory for Histocompatibility, Department of Immunohematology and Blood Transfusion, Leiden University Medical Center, Leiden, The Netherlands
    Search for more papers by this author
  • Marian Witvliet,

    1. National Reference Laboratory for Histocompatibility, Department of Immunohematology and Blood Transfusion, Leiden University Medical Center, Leiden, The Netherlands
    Search for more papers by this author
  • Bernadette J. J. M. Haase-Kromwijk,

    1. Dutch Transplantation Foundation, Leiden, The Netherlands
    Search for more papers by this author
  • Willem Weimar

    1. Department of Internal Medicine – Transplantation, Erasmus MC, University Medical Center Rotterdam, Rotterdam, The Netherlands
    2. Dutch Transplantation Foundation, Leiden, The Netherlands
    Search for more papers by this author

*Corresponding author: Marry de Klerk,


The wait time for deceased-donor kidney transplantation has increased to 4–5 years in the Netherlands. Strategies to expand the donor pool include a living donor kidney exchange program. This makes it possible that patients who cannot directly receive a kidney from their intended living donor, due to ABO blood type incompatibility or a positive cross match, exchange donors in order to receive a compatible kidney. All Dutch kidney transplantation centers agreed on a common protocol. An independent organization is responsible for the allocation, cross matches are centrally performed and exchange takes place on an anonymous basis. Donors travel to the recipient centers. Surgical procedures are scheduled simultaneously. Sixty pairs participated within 1 year. For 9 of 29 ABO blood type incompatible and 17 of 31 cross match positive combinations, a compatible pair was found. Five times a cross match positive couple was matched to a blood type incompatible one, where the recipients were of blood type O. The living donor kidney exchange program is a successful approach that does not harm any of the candidates on the deceased donor kidney waitlist. For optimal results, both ABO blood type incompatible and cross match positive pairs should participate.


In the Netherlands approximately 400 (24/million inhabitants) deceased donor kidney transplants are performed on an annual basis. This number has remained constant over the last 20 years, while an increasing influx of new patients on the waitlist resulted in wait times of 4–5 years. Strategies to expand the kidney donor pool in our country have included programs both for nonheartbeating donors and for living (non) related donors. In 2004 the proportion of deceased donor kidney transplants derived from nonheartbeating donors already exceeded 42% (171/402) and the number of living kidney donations increased to 250 providing 38% of the total number of transplanted kidneys in the Netherlands. These efforts, however, have not as yet resulted in a shorter national waitlist. Therefore, other options were explored, e.g. a program for living donor kidney exchange or crossover kidney transplantation. Such a program makes it possible for patients that cannot receive a kidney from their intended donor, due to ABO blood type incompatibility or to a positive serological cross match, to exchange donors in order to receive a kidney. The concept was described by Rapaport in 1986 (1), but despite several proposals (2,3), did not find it's way into clinical practice in Europe or in the United States until recently. In South Korea, where brain death is accepted neither socially nor legally, and donor exchange is the only alternative for living kidney donation in case of incompatibility, a crossover program is already operating for more than 10 years (4). At the moment the attitude toward living donor exchange is rapidly changing in the western world and a number of initiatives have been undertaken on a local or regional scale (5,6). We here report our 1-year experience with a national living donor kidney exchange program in which all kidney transplant centers in the Netherlands participated.

Logistics and Patients

Before embarking on a clinical program, we explored in a pilot study the acceptability of donor kidney exchange for our patients and the feasibility of a computer-based allocation program (7). It became clear that both potential donors and recipients were eager to participate, but strongly preferred anonymity. The allocation program is based on ABO blood type compatibility, while pre-definition of unacceptable HLA antigens by extensive antibody screening is used to predict negative cross matches. We also undertook a study on the ethical and psychological aspects, in which we considered a number of topics: the influence of ‘donation by strangers’ on the motivation and willingness of donor-patient couples; the issue of anonymity; the loss of the possibility of ‘medical excuses’ for unwilling donors and the view that crossover is a first step to commercial organ donation. We concluded that none of these issues seems to propose a disorderly or unethical situation (8). However, in our opinion, donor kidney exchange is barter and thus vulnerable to economic forces. Therefore, we strongly advocated the allocation of kidneys to be the responsibility of an independent organization i.e. the Dutch Transplantation Foundation. In the mean time a national committee was formed consisting of representatives of the kidney transplant centers in the Netherlands, the National Reference Laboratory for Histocompatibility and the Dutch Transplantation Foundation. Participants agreed on a protocol in which medical criteria for donor and recipient, the registration of candidates, matching, allocation, surgical and follow-up procedures were described. All patients and donors were molecularly typed for HLA-A,-B,-C, -DR and -DQ on a medium resolution level. Sera of the patients were screened for HLA alloantibodies using standard complement dependent lymphocytotoxicity and ELISA. HLA antigens toward which the patients had formed specific alloantibodies were considered not acceptable mismatches, which means that donors with these antigens were not selected for these patients. Consensus was achieved on a computerized allocation algorithm based on blood type, first identical than compatible, and match probability. The match probability is a value built up by the frequencies for a recipient compatible blood types and unacceptable HLA antigens within the actual donor pool. Other criteria are wait time counting from the first day of dialysis and donor age. Allocation procedures to match compatible combinations were scheduled every 3 months. Cross matches between the new donors and recipients were to be centrally performed in the National Reference Laboratory for Histocompatibility. Thereafter, the donors of the newly formed pairs would travel to the recipient's transplant center, where the final decision for surgery is made. A final decisive cross match was performed in the tissue typing laboratory affiliated to the transplant center of the recipient. Donation procedures in the two centers are planned within 3 months of allocation and are simultaneously performed.

Four match procedures, in which a total of 60 donor–recipient combinations were enrolled, have been performed in the year 2004. There were 31 couples that participated once, 17 twice, 11 three times and 1 even all 4 rounds. We enlisted 29 ABO blood type incompatible and 31 pairs with positive lymphocyte cross matches. Table 1 show the demographic data of the donors and their recipients, stratified for ABO blood type incompatibility and cross match positivity. In the latter group long waiting sensitized female recipients with partners as intended donors predominated. In Table 2 the blood type distribution of the 29 ABO blood type incompatible couples is given. The A to 0 combination was most frequently seen (16 pairs), followed by 6 A-B or B-A combinations.

Table 1.  Donor and recipients characteristics
blood type
n = 29

Cross match
n = 31
Gender recipients (male/female)17/128/23
Gender donors (male/female)12/1718/13
Age recipients (median, range)51 (22–72)50 (16–69)
Age donors (median, range)53 (38–74)53 (29–71)
PRA % historic (median, range)2 (0–100)28 (5–100)
Wait time months (median, range)19 (0–84)26 (0–69)
Donor type:
 Partner (male/female)8/1115/5
 Sibling/other relative2/12/2
Table 2.  Blood type distribution of the 29 ABO blood type incompatible pairs


For 40 couples, 14 of 29 ABO blood type incompatible and 26 of 31 cross match positive (p = 0.008, chi-square test), the computer program predicted on the basis of blood type and unacceptable HLA antigens 47 exchange possibilities resulting in 94 new combinations. Because 21 of these 40 potentially exchangeable pairs could be matched with more than one (median: 3, range: 2–12) other pair, a further selection was made on the basis of match probability. This resulted in 26 combinations that proved to have negative cross matches thereby demonstrating the accuracy of the antibody determination in the HLA laboratories and the allocation algorithm. The drop out rate during this last procedure was comparable for both groups: 36% for the ABO blood type incompatible group and 35% for the cross match positive group, suggesting the fairness of the selection by match probability. There was no necessity for further selection on the basis of wait time or donor age. In 19 cases a new combination was already found in a first computer search, while for five patients a second and for two a third procedure was needed. The chance for a donor–recipient pair to find a matching couple was 19 of 31 (61%) during a first attempt, but thereafter chances decreased to 5 of 17 (29%) for those participating twice and 2 of 11 (18%) for the couples that tried three times.

Of the 26 successfully matched recipients, 9 belonged to an originally ABO blood type incompatible combination, while 17 had positive cross matches with their intended donors (Table 3). Five times an ABO blood type incompatible pair was combined with a cross match positive one. So, 9 of the 29 (31%) ABO blood type incompatible and 17 of the 31(55%) cross match positive combinations were matched to new donors and recipients (p = 0.11). Even some long waiting highly immunized patients could be matched to a new donor. Table 4 shows the original ABO blood type distribution of the 26 couples that were matched versus the 34 for whom no new combination was found.

Table 3.  Characteristics matched versus not-matched
n = 26
n = 34
ABO blood type incompatible920
Positive cross match1714
Gender recipients (male/female)11/1514/20
Gender donors (male/female)12/1418/16
Age recipients (median, range)51 (22–72)51 (16–59)
Age donors (median, range)54 (29–71)52 (33–74)
PRA % historic (median, range)11 (0–80)20 (0–100)
Wait time months (median, range)19 (0–69)27 (0–84)
Donor type:
 Partner (male/female)11/88/12
 Sibling/other relative23/2
Table 4.  Original blood type distribution matched (n = 26) versus not-matched (n = 34)

All original blood type B to A and O to B combinations were matched, less successful were the O to A (71%), A to B (50%), A to A (44%) and O to O (40%) blood type combinations. For all the ABO blood type incompatible pairs with a blood type O recipient it was more difficult to find a matching couple, but it was still possible in 5 of the 21 cases (24%).

After the allocation of 26 donors to their new recipient, 24 kidney transplants were performed. The median time between the moment of enrolment in the exchange program and the actual transplantation was 104 days (61–326). Two procedures were cancelled for medical reasons. In 22 instances, the original donor and recipient underwent surgery in different centers while two times both operations took place in the same center. One transplant had to be removed in the 4th postoperative week because of an irreversible rejection. The recipient was a 35 year-old male with a historical PRA of 21% and originated from the positive cross match group. All other kidneys are functioning well after a median follow-up of 8 months and none of the donors suffered from complications.


In the present report we demonstrated a successful living donor kidney exchange program in which both cross match positive and ABO blood type incompatible donor–recipient pairs participated. We were able to create new ABO blood type compatible couples with negative cross matches for 26 of the 60 (43%) candidates and this resulted in 24 kidney transplants. For the next couple of years we expect an input of at least 15 new donor–recipient pairs in each match procedure. This will presumably lead to approximately 25–30 transplantations a year. This approach of donor exchange to expand the donor pool has all the advantages of living unrelated kidney donation, which is associated with an excellent long-term outcome irrespective of HLA matching.

Alternative protocols have been developed to make direct donation possible within incompatible pairs, e.g. the use of plasma exchange to remove the isoagglutinins or anti-HLA antibodies in combination with the administration of i.v. immunoglobulins and anti-B cell antibodies. Disadvantages of these protocols are the demanding technique, the high financial costs and, more important, the high rate of rejection and graft loss (9,10).

Another alternative solution for ABO blood type or cross match incompatible pairs is the live donor-list exchange (6,11). A live donor who is incompatible with his potential recipient donates his kidney to the deceased donor waitlist with the agreement that the paired recipient would receive priority for the next compatible cadaver kidney. The recipient on the waitlist who receives a live kidney has the benefit of a better chance of graft survival. The recipient with the incompatible donor receives within a few days a cadaver kidney with a lower graft survival, but the benefit of decreasing his wait time. Although the 'nett gain' of this type of exchange in terms of survival is probably in favor of the total pool of patients on the list, it also implies that a number of individuals have to wait longer for a deceased donor kidney. Especially the blood type O recipients will suffer from the extraction of blood type O kidneys for the benefit of blood type O recipients with a living non–blood type O donor. This unfairness will not occur when the program is restricted to ABO blood type incompatible couples with non–O type recipients and ABO blood type compatible couples with a positive cross match (12). However, this limitation, while fulfilling criteria of justice and fairness, renders the program less efficient in terms of absolute numbers of transplants. On the other hand, efficiency is not always a valid argument. The highest efficiency can be reached with one large crossover pool including all couples irrespective of blood type or cross match. In our opinion it is unrealistic and even unethical to persuade a compatible donor to donate to a large anonymous pool instead of directly to a relative or friend. The efficiency argument has also been used against our approach with the strict separation of the exchange donor pool from the deceased donor pool. We are aware that our program will stagnate, when the situation arises that the exchange pool exists of non–blood type O donors and type O recipients. Participation of both ABO blood type and cross match incompatible couples, which makes it possible to combine couples from these subgroups, is therefore essential for the success of the program. We were able to find new recipients for 43% of the donors and although our program did not make optimal use of the available living donor kidneys, it did not harm any of the candidates on the waitlist for deceased donor kidneys. Without interference with other allocation systems we can not find any ethical argument against paired living donor kidney exchange, provided the participants are fully informed and unforced, and an independent organization is responsible for the matching of new combinations. However, there may be legal barriers, e.g. in Germany and the United Kingdom where law forbids living-unrelated donation in the absence of a close emotional relationship, or in France where even emotional unrelated donation is not allowed unless, in select cases, a court of law rules differently. In the United Kingdom the law will change in 2006 and UK Transplant is confident to start a crossover program that year. We encourage the various transplant organizations to embark on these programs too and, when necessary, to convince the authorities to modify transplant laws in order to make these indirect donations possible.


Without the cooperation of the seven university kidney transplantation centers this program could not have been implemented. We express our appreciation to all the surgeons, nephrologists and coordinators kidney transplantation (Academic Medical Center Amsterdam: M.M. Idu, surgeon, F. Bemelman and S. Surachno, nephrologists, S. ter Meulen, coordinator. University Medical Center Groningen: R.J. Ploeg, surgeon, J.J. Homan van der Heide, nephrologist, R. Meijer and A. Roelofs, coordinators. Leiden University Medical Center: J. Ringers, surgeon, J.W. de Fijter, nephrologist, M. van Gurp, coordinator. Academic Hospital Maastricht: L.W.E. van Heurn, surgeon, J.P. van Hooff and E.M. van Duijnhoven, nephrologists, J. Dackus, coordinator. University Medical Center St. Radboud: J.A. van der Vliet, surgeon, Ph.M.M. Dooper and A.J. Hoitsma, nephrologists, D. Pilzecker, coordinator. University Medical Center Utrecht: R.W.H. van Reedt Dortland, surgeon, R.J. Hené, nephrologist, B. Vink, coordinator. Erasmus MC University Medical Center Rotterdam: J.N.M. IJzermans, surgeon, G. Bakker, A. Luchtenburg and W. Zuidema, coordinators).

This program was supported by the Dutch Kidney Foundation.