Five decades with grandparent donors: The Norwegian strategy and experience

Living donors (LDs) are preferred over DDs for renal transplantation in children due to superior GS. Oslo University Hospital has never restricted living donation by upper age. The aim of this study was to investigate long-term outcomes using grandparents (GPLD) compared to PLD. Retrospective nationwide review in the period 1970-2017. First renal graft recipients using a GPLD were compared to PLD kidney recipients for long-term renal function and GS. 278 children (≤18 years) received a first renal transplant: 27/251 recipients with a GPLD/PLD. GPLD (median 59 (42-74) years) were significantly older than PLD (median 41 (23-65) years, ( P < .001). Median DRAD was 52 (38-70) vs 28 (17-48) years, respectively. GS from GPLD and PLD had a 1-, 5-, and 10-year survival of 100%, 100%, and 90% vs 93%, 82%, and 72%, respectively ( P = .6). In a multivariate Cox regression analysis adjusted for gender, donor age, recipient age, and year of transplant, this finding was similar (HR 0.98; 95% CI 0.34-2.84, P = .97). Five-year eGFR was 47.3 and 59.5 mL/min/1.73

the transplantation era in 1970, grandparents have been accepted as donors on similar terms as parents.
Long-term outcome data for grandparents as donors for children with ESRD are scarce, with only a few publications. [4][5][6] The observation time varies from two to 20 years and shows satisfactory GS. These analyses are, however, limited by a relatively low number of grandparent donors.
The aim of this retrospective survey was to evaluate our longterm strategy of allowing grandparental kidney donation on equal terms with parental donation, with a focus on long-term kidney function and GS.

| ME THODS
Individual patient data were extracted from the NRR.
NRR is a national medical quality registry following all kidney transplant recipients from start of renal replacement therapy to death. Baseline data, for example, comorbidity, renal diagnosis, and drug treatment, are collected at time of transplantation. Follow-up data on graft loss and death are captured continuously, while a summary of all events during the year and status of the patient with regard to, for example, drug therapy, blood pressure, and clinical chemistry is updated by the last visit at the end of the year. The individual coverage is >99.9% since the start in 1969 and annual data coverage is >98%.
All first kidney, pediatric (≤18 years), LD transplantations performed in Norway in the period from 1970 to 2017 were retrieved.
The NNR has national coverage and a written informed consent to perform quality and research analysis on the data obtained. Follow-up was censored in December 2018. Graft loss was defined as retransplantation or start of dialysis, including recipient death with functioning graft.
Groups were compared using Kaplan-Meier analysis. The hazard ratio for graft loss for grandparent vs parent LD was calculated with Cox regression analyses adjusted for gender, donor age, recipient age, and year of transplant. In addition to this, we evaluated the risk of graft loss between pairs with a DRAD of more than 30 years vs those with an age difference of less than 30 years. The DRAD of 30 years was chosen as the average age difference between all donors and all recipients was 31 and the median was 29 years.
Kidney function at 5 (±2) years after transplantation was calculated as eGFR using the Schwartz-Lyon eGFR formula 7 for all the recipients below 18 years of age at the time of analysis. For recipients older than 18 years, the MDRD4 formula was used. 8 Recipients with graft loss before the time of creatinine measurement were given a value of zero for eGFR when comparing the two groups. Statistical analyses were performed using SPSS 25. Descriptive values are given as means (SD), medians (range), and proportions (%).

| RE SULTS
Among pediatric recipients of LD kidneys, 27 received their graft from a GPLD and 251 from a PLD. The causes of ESRD are listed in Table 1.
Demographic data of the two groups are shown in Table 2.
Data on recipients with donors with an age difference of more or less than 30 years are shown in Table 3.
The median recipient age of GPLD grafts was lower than of the PLD group, 5.7 (range 1.2-18.6) vs 13.6 (range 0.8-18.9) years. The The median follow-up time was 11.9 (range 1-46) years.
Graft survival for both groups is shown in Figure 1. Log-rank test did not show any significant difference between the groups (P = .6).
The reasons for graft loss were rejection (8/9) in the GPLD group.
In the PLD group, the reasons for graft loss were rejection (80/129), recurrence (8/129), tumor (1/129), and other causes (4/129). Death with a functioning graft as a cause of graft loss was seen in one GPLD recipients and in 36 of the PLD recipients, respectively.
Graft loss stratified by a DRAD of more vs less than 30 years is shown in Figure 2.
In univariate Cox regression analysis, the hazard ratio for graft loss between the two groups was not significantly different (HR   In comparison with these publications, our data include more grandparents and the median donor age was slightly higher. We We could also see in the multivariate analysis that tx year had a positive impact on GS (Table 5). This may be associated with improvements in treatment and better long-term prognosis, as also seen in other cohorts. 9

F I G U R E 2 Cumulative graft loss in relation to DRAD
Several issues have to be considered when evaluating a possible donor for a child, and the aim is to preserve an acceptable kidney function for a long time.
The acceptance of older LD for children has been controversial.
Especially for children, the main policy has been "young-for-young." Decreased functional reserve, a higher risk for hypertension, and vascular disease have been the reason for not accepting older donors. 10,11 Historically, a study of DDs showed that increased age difference between donor and recipient was associated with decreased GS. 12 Later, other studies have shown similar results, even though the difference in GS has decreased. 13 Due to a higher risk of vascular disease, Heidotting et al 11  More data are emerging and suggest that the use of older LD has acceptable long-term outcome. 15,16 A factor that can affect the function of a transplanted kidney is the size. Using an adult kidney to small children (<15-20 kg) will lead to a significant size mismatch. An early study from Bohlin and Berg looked at changes in GFR in both child recipients and adult donors.
The study showed an equivalent GFR in recipient and donor adjusted for body size. These results suggest that transplantation of an adult kidney to a child results in a functional adaptation to the smaller body size of the recipient within 3 months after transplantation. 17 Later, a study showed that the adult graft function did not improve with longitudinal growth as pediatric grafts did. 18 Another study from Germany confirmed this, though in this study, all adult donors were DD. 19 In our material, the eGFR at 5 years post-transplant was better for the PLD than for the GPLD groups. Unfortunately, we only had data from 17/27. The difference in eGFR should be interpreted cautiously due to the low number of measurements available in the GPLD group.
HLA compatibility is of importance for GS as more HLA mismatches increase the risk of rejections, even in the era of better immunosuppression. 20 Poor HLA matching is also associated with enhanced sensitization making it harder to find suitable donors for repeated transplantations. We strive to select well-matched donors in order to keep HLA immunization to a minimum. In our data, we observed that parental donors had a slightly better HLA class I match with the recipients as compared to the grandparental donors, although the difference was small. The marginally better-matched parental donors could be an advantage when the patient is in need of a retransplantation.
The long-term risks linked to unilateral nephrectomy in LDs have received increased attention during the last decade. There is an increased all-cause mortality risk and risk for ESRD more than a decade after kidney donation. 21 To our knowledge, this is the largest study on this topic, even though the number of recipients is limited. Another strength is the long and complete follow-up. There is a small overlap in donor age between the groups, but on a group level, the DRAD is substantial, and in spite of this, the results are promising.

| CON CLUS ION
In conclusion, our strategy to use grandparents as donors has proven to be successful. We will continue to seek grandparent donors when possible. With this strategy, we can save other potential organ donors for future transplantation and at the same time reduce the overall risk for a young parent donor. With careful selection of potential donors, chronological age alone should not exclude highly motivated grandparents. With the emerging positive data on GS and long-term results, such a policy will enable centers to expand the donor pool.