Kidney transplantation is considered the best therapeutic option in end-stage renal disease (ESRD) because of improved quality of life and prolongation of survival for many recipients (1). As the prevalence of ESRD increases in most countries, the demand for kidney transplantation exceeds the supply of available organs. As a result the use of donors above 50 years of age has increased in both deceased and living donor (LD) kidney transplantation. The use of grafts from older deceased donors (DD) has been evaluated in many studies and most demonstrated significantly poorer graft function and graft survival in recipients of grafts from DD above 50 years. This has been attributed to a variety of reasons, including reduced nephron mass, senescence, greater susceptibility to ischemic injury, and acute rejection episodes (2). Whether donor age above 50 years exerts a similar negative influence on the outcome of LD transplantation is poorly elucidated.
In living kidney transplantation it has been reported that kidney volume is an independent determinant of posttransplant graft outcome and that grafts from male donors have larger kidney volumes than grafts from female donors (3). Furthermore, there have been reports that when kidneys from female donors are transplanted into male recipients a significantly higher proportion of patients required antirejection treatment compared to male recipients who received a kidneys from a male donor (4, 5). Graft survival uncensored for death has been reported to be inferior in recipients of female donor kidneys as well (4). Although most reports have shown a higher serum creatinine in grafts from female donors to male recipients 1 year posttransplantation, this difference was not observed 2 years after transplantation in LD transplantation (6). Thus the effect of donor sex on outcome after transplantation is not established.
Most reports have evaluated graft survival up to 5 years after LD transplantation and early acute rejection episodes have been considered a major risk factor for graft loss (7). There are few studies that have evaluated graft survival beyond 5 years posttransplant (8, 9). The influence of early events and the impact of donor related risk factors as predictors of kidney graft outcome may change over time. Therefore, risk factors for graft failure in the early interval may not continue to be valid for graft losses 5 years after transplantation.
In Norway LD kidney transplantations have constituted approximately 40% of all renal transplantations since 1969, and older donors have been accepted after a thorough predonation screening (10, 11). The purpose of this study was to assess the influence of donor age and gender on short-term and long-term graft survival and acute rejection episodes in LD transplantation.
MATERIALS AND METHODS
All renal transplantations in Norway take place at Rikshospitalet University Hospital. Data from renal transplant recipients and donors are transferred to the Norwegian Renal Registry, a national database. A total of 1981 first renal transplants were performed between January 1, 1994 and December 31, 2004. In this prospective cohort study we included only LD recipients 18 years of age or older, leaving 739 patients eligible for this analysis. In the donor population, 346 (46.8%) were above 50 years, 118 (16.0%) were above 60 years, and 57 (7.7%) were above 65 years. During the study period, the donor age remained fairly constant, 48.1±11.4 years (1994–1996), 49.3±12.1 years (1997–2000), and 47.4±11.4 years (2001–2004). Of the total donor population 64% were women.
ABO blood group compatibility and a negative immuno-magnetic CDC B-cell and T-cell cross-matches were a prerequisite for donation. Donors had to be more than 18 years of age, but there was no upper age limit. The medical criteria were strict. Kidney function estimated by endogenous creatinine clearance more than 70 mL per min and normal urine analyses were mandatory. Blood pressure should be in the normotensive range (less than 140/90 mm Hg). Body mass index more than 30 kg per m2 and elevated fasting blood glucose were exclusion criteria. The study population consisted mainly of ethnic Norwegians and less than 1% of the donors and recipients were of non-white origin.
The database contains donor variables (age, gender, and relationship to the recipient), recipient variables (age, gender, original disease, time spent in dialysis, panel reactive antibodies), transplant factors (human leukocyte antigen [HLA] -A, -B, and -DR mismatches), and posttransplantation features including immunosuppressive regimen, rejection history, and serum creatinine values. Follow-up data were collected yearly.
In this analysis recipients were followed until graft loss, death or last recorded status up to June 30, 2005. Only early acute rejection episodes occurring within 3 months after transplantation were included in the analysis. With a rise in creatinine of 20% or more an indication biopsy was performed and evaluated by the Banff criteria. The category Banff borderline changes was included in acute rejection episodes. When indicated in the text, patient death with a functioning graft was censored as a cause of graft loss.
From 1994 to 1999 a triple therapy consisting of corticosteroids, azathioprine, and cyclosporine was used. Induction therapy with basiliximab and maintenance with cyclosporine and corticosteroids was used only in 2000. Since 2001 cyclosporine, mycophenolate mofetil, and corticosteroids constituted the standard immunosuppressive protocol. Since 2001 C2 monitoring has been used for the first 3 months after transplantation. Acute rejection episodes were treated according to a standard protocol consisting of a total dose of 1.25 g methylprednisolon given intravenously on four consecutive days. When the episodes were considered steroid resistant, the patients received anti-thymocyte globuline (ATG) or anti-CD3 monoclonal antibody (OKT3). The study was approved by the Regional Research Ethics Committee. The Norwegian Renal Registry has obtained concession from the National Data Inspectorate.
Comparisons between groups were done using cross tables and chi-square statistics for categorical variables and two-sided t tests for continuous variables. Continuous data are reported as mean±SD, whereas skewed data are reported as median and 25 percentiles and 75 percentiles (QI, QIII). Graft survival was estimated with the use of Kaplan–Meier life tables and all comparisons with the use of the Mantel–Haentzel log rank test. A sex difference in susceptibility to age as risk factor in donor kidneys on graft loss was explored by testing for biological interaction according to Rothman (12). The incidence of graft loss calculated per 100 patient year in recipients of grafts from male and female donors above and below 50 years was calculated. A new composite variable with four categories (SYMBOL, ābaJOURNAL/trans/04.03/00007890-200703150-00015/OV0399/v/2021-01-29T194855Z/r/image-png, and ab) was redefined where ā denotes female sex and donor age less than 50 years. Relative risk (RR) was calculated for each category with 95% confidence interval (CI). An interaction effect is defined as departure from additivity of absolute effects, and excess RR caused by interaction (RERI) was calculated: RERI=RR (ab)−RR (āb)−RR (aJOURNAL/trans/04.03/00007890-200703150-00015/OV0399/v/2021-01-29T194855Z/r/image-png)+1 where RR ab denotes RR for graft loss in recipients of kidneys from male donors above 50 years and kidneys from female donors below 50 years (SYMBOL) are used as a reference category (RR=1). Ninety-five percent confidence interval (CI) was calculated as proposed by Hosmer and Lemeshow (13). RERI of 0 means no interaction. RERI of 0.5 means that because of interaction between the two risk factors, RR is 0.5 greater than expected based on addition of the two risk factors.
Risk for graft loss was analyzed with the use of Cox regression after checking each factor for proportionality and performed in two intervals after transplantation: 3 months to 5 years and beyond 5 years. In addition graft survival was evaluated for the whole period beyond 3 months posttransplant. Risk factors for graft loss included in the analysis were the pretransplant factors: donor age, recipient age, recipient gender, donor gender, preemptive transplantation, HLA identical transplantation, and HLA DR matching and posttransplant factors: acute rejection episode and steroid resistant rejection episode. Each potential risk factor was related to the risk of outcome in a univariate manner by using a Cox proportional hazard survival model. Relative risks with 95% confidence intervals are presented. When constructing the Cox multivariate model only factors with P<0.2 from the univariate model were included. The Cox model identifying risk factors for acute rejection episode included the same pretransplant factors that were used in the model for graft survival. Statistical significance was identified by a two-sided P-value of less than 0.05. All analyses were performed with the use of the SPSS statistical software package (Version 12.0).
In the total population (n=739) the median follow-up time was 55.1 (QI28.6, QIII88.4) months and the range was 0.03 to 137.7 months. There were 71 graft losses during the study period and 74 patients died with functioning grafts. The overall death censored graft survival was 97.1% at 1 year, 92.8% at 5 years, and 81.0% at 10 years after transplantation, whereas uncensored graft survival was 94.7% at 1 year, 85.4% at 5 years, and 66.0% at 10 years after transplantation.
The characteristics of the study population grouped according to donor age above or below 50 years are summarized in Table 1. There was a higher number of female donors (P<0.0001) and graft losses (P=0.003) and higher creatinine levels (P<0.0001) in the older donor group than in the group with younger donors (Table 1). There was, however, no difference in the incidence of acute rejection episodes in patients who received a graft from a donor more than 50 years compared to a donor less than 50 years.
In female donor kidneys the incidence of acute rejection episodes was higher than in male donor kidneys (49.8% vs. 38.1%; P=0.002), but the incidence of steroid resistant rejections did not differ (18.3% vs. 14.7%; P=0.193). Recipients of female donor kidneys were older than recipients of male donor kidneys (47.2±14.0 yr vs. 45.2±14.9 yr, P<0.0001). In the multivariate model of risk factors for acute rejection episodes (Table 2) donor age above 50 years was not a significant risk factor for early acute rejection episodes, whereas donor age more than 65 years (HR=1.57; 95% CI 1.09–2.27) and number of HLA-DR mismatches were significantly associated with acute rejection episodes. Recipient age more than 50 years decreased the risk of experiencing an acute rejection episode (HR 0.69; 95% CI, 0.55–0.87).
Graft survival analysis of the whole observation period after transplantation show that recipients of kidneys from donors less than 50 years had significantly better censored graft survival (Fig. 1a). This was, however, only the case when the recipient had experienced an acute rejection episode (Fig. 1b). In the absence of acute rejection episodes there was no difference in graft survival (RR 1.55; 95% CI 0.67–3.60, P=0.31). There was no difference in graft survival in male and female donor kidneys, but female donor kidneys tended to do better beyond 5 years after transplantation (RR 1.49; 95% CI 0.93–2.37, P=0.09). This gender difference was significant in recipients who had not experienced an acute rejection episode (Fig. 1c). In recipients in whom an acute rejection episode had occurred, no difference in graft survival according to donor sex was observed (RR 1.30; 95% CI 0.74–2.30, P=0.36). Furthermore, graft survival of female donor kidneys was identical in male and female recipients (RR 0.96; 95% CI 0.45–2.05, P=0.91).
To evaluate whether there was an increased risk of graft loss due to a biological interaction between donor age and sex, an interaction analysis was performed. Female donors less than 50 years were used as a reference category as these grafts lowest incidence of graft loss during the study period (0.86/100 pt yr). The RR of graft loss in recipients of male donors less than 50 years was 2.03; 95% CI 0.91 to 4.60, in female donors more than 50 years RR=2.73; 95% CI 1.29 to 5.92 and in male donors more than 50 years RR=4.02 (1.86–8.96). No interaction between donor sex and age was found as excess RR because of biological interaction (RERI) was 0.262; 95% CI−2.10 to 2.62.
To estimate short- and long-term effects of risk factors for graft loss, the effect in recipients with graft survival 3 months to 5 years and beyond 5 years was evaluated and the overall impact of the risk factors beyond 3 months after transplantation. The study population consisted of 739 recipients, of which 710 were available for analysis beyond 3 months after transplantation and 343 were available for analysis in the late period after transplantation. There were 26 graft losses in the early period, whereas there were 28 graft losses in the late period after transplantation. Death with functioning graft occurred in 28 recipients in both the early and late period after transplantation. The univariate analysis (Table 3) shows that donor age more than 65 years was a risk factor for graft loss in all time periods after transplantation. The association between donor age as a continuous variable and death censored graft survival was RR 1.05 per year; 95% CI 1.02 to 1.07. Beyond 3 months an acute rejection episode (RR 1.88; 95% CI 1.08–3.30) and a steroid resistant rejection episode (RR 2.99; 95% CI 1.69–5.30) were additional risk factors for graft loss. During the early posttransplant period (3 months–5 years) the same risk factors apply. Beyond 5 years the risk factor pattern changed and male donor gender was the only additional risk factor for graft loss (RR 2.88; 1.30–6.37). Recipient age was not a predictor for graft loss at any time after transplantation.
In the multivariate analysis (Table 4) donor age more than 65 years (HR 5.35; 95% CI 2.55–11.23), male donor gender (HR 2.18; 95% CI 1.25–3.81) and a steroid resistant rejection episode (HR 2.17; 95% CI 1.11–4.26) were associated with graft loss beyond 3 months after transplantation. In the early period after transplantation a steroid resistant rejection episode (HR 3.96; 95% CI 1.46–10.75) and donor age more than 65 years (HR 5.48; 95% CI 2.01–14.94) were the factors that had any positive predictive value. After 5 years donor age more than 65 years (HR 5.09; 95% CI 1.55–16.72) and male donor gender (HR 3.58; 95% CI 1.57–8.17) only achieved significance.
Of 56 donors above 65 years, 20 donated to a person above 65 years. Table 5 gives the demographics of these donors according to whether the recipient was above or below 65 years. There were fewer rejection episodes in the group where donor and recipient were of the same age and fewer graft losses. There was, however, a higher number of deaths with functioning grafts in the older group of recipients 7 (35.0%) vs. 4 (11.1%). The creatinine level 1 year after transplantation did not differ between the two groups.
In the current study we identified important factors influencing the incidence of acute rejection episodes and graft survival in LD kidney transplantation at different time periods after transplantation. We observed that the incidence of acute rejection episodes was increased when donor age was more than 65 years, but not in recipients of kidneys from donors more than 50 years. Furthermore, graft survival was not affected by donor age above 50 years as long as these recipients did not experience an acute rejection episode. In multivariate analysis of risk factors for graft loss in the early period after transplantation, 3 months to 5 years, donor age more than 65 years, and a steroid resistant rejection episode were significant predictors. However, 5 years after transplantation risk factors for graft loss changed as steroid resistant rejection episode no longer remained a predictor for graft loss, whereas male donor gender became a significant risk factor.
The incidence of early rejection episodes increased when donor age was above 65 years in LD transplantation in our study. Fijter et al. reported that the incidence of acute rejection episodes increased in deceased kidney transplantation when the donor age exceeded 50 years (14). In LD kidney transplantation the donor is screened for cardiovascular disease, the renal function is optimal, the donor is not under the stress of acute illness and cold ischemia time is minimal. These donor-related factors may contribute to the 15 year movement in donor age with respect to the risk of acute rejections that we have demonstrated.
Graft survival was not influenced by donor age above 50 years in recipients who did not experience an acute rejection episode. Conflicting results regarding the impact of donor age on graft survival in LD transplantation have been reported. In a recent paper of the entire LD transplantation experience at the University of Minnesota by Matas et al., donor age greater than 55 years and an acute rejection episode were significant risk factors in their multivariate analysis for graft loss in recipients with graft survival beyond 1 year (15). Our results indicate that graft survival in a recipient with 3 months graft survival was affected by donor age above 65 years and the occurrence of an early steroid resistant rejection episode. These results are in accordance with an observation in DD transplantation that older donor age alone was not a risk factor for worse long-term outcome as long as there was no history of hypertension and kidney function was normal (16). Normotension and normal kidney function in the LD may explain in part why donor age above 50 years was not a risk factor for graft loss at any time in our multivariate analysis. This indicates that careful selection of donors after strict medical criteria is more important than the chronological age of the donor. Studies with long term follow-up of transplantations performed with donors above 65 years are few. Thus, our result regarding donors above 65 years should be interpreted with caution, and needs to be confirmed by others. The clinical implications are yet to be decided, however, and therefore donors more than 65 years should still be considered when no other donor is available.
Early and late graft failures (death censored) reflect distinct factors in LD kidney transplantation. When patients who died with functioning grafts are excluded from the survival analysis, a more accurate association of risk factors for survival of the graft will appear. Death with preserved renal function is today increasingly because of cancer, a condition that has no direct relationship to the status of the graft (17). Risk factors for early graft losses in our study were donor age above 65 years and a steroid resistant rejection episode. De La Vega et al. who reported that although donor age above 50 years resulted in higher serum creatinine and reduced glomerular filtration rate in the recipient, no increase in graft loss during the first 3 years after transplantation was observed (18). As can be seen from the survival curves in the current study, the separation begins beyond 3 years after transplantation. The observation that only steroid resistant rejection episode was a risk factor for early graft loss is in accordance with other reports (8, 14). Furthermore, our study confirms previous studies that have shown that an early acute rejection episode is no longer a risk factor for graft failure in recipients treated with a cyclosporine based immunosuppressive regimen (19).
Five years after transplantation the risk factor pattern changed and donor age more than 65 years and male donor gender were the only remaining predictors for graft loss in both the multivariate and univariate analysis. The change in risk factor profile was not unexpected as early graft losses carry the effect of many pretransplant risks and early acute immunologic injuries. Our observation that the impact of donor age on graft loss is constant as time passes and that the rate of graft loss increases 5 years after transplantation has not been reported earlier. Toma et al. observed a poorer survival of the older (>55 years) LD grafts 5 years after the transplantation. This was, however, graft survival uncensored for death with a functioning graft (9). Our finding may reflect the fact that donor age above 65 mirrors the process of the aging graft. Halloran has proposed a role for cellular senescence in the decline of renal transplant function over time, and this may imply that these grafts are more vulnerable to further injury, as calcineurin inhibitor toxicity (20).
A rather surprising result in our study was that female donor gender presented an advantage for long-term graft survival both for male and female recipients. Earlier studies have reported poorer overall graft survival in females donating to males, and this was explained by the theory of nephron underdosing (4). Although anatomic studies have documented larger kidney weight in men, the results were inconsistent when the kidney size was corrected for the body surface area (21). In our study, overall, female donor kidneys did better than male donor kidneys, even if recipients of female donor kidneys had higher creatinine values and a higher incidence of early acute rejection episodes. Based on our observation it is tempting to speculate that there may be some survival advantage inherent in the female kidneys. This is in accordance with the reports that renal prognosis in primary chronic renal disease is considerably better in female patients than in male patients (22). Thus, our study does not lend support to gender matching.
Based on the assumption that old patients have an attenuated immune response and a reduced life expectancy, it has been proposed that kidneys from older donors should be allocated to old recipients. The subgroup of donors and recipients who were more than 65 years of age in our study showed a satisfactory 1-year creatinine and few graft losses. Indeed the “old for old” matching that has been met with varying enthusiasm in the realm of DD transplantation, may be far better suited for LD transplantation (23). For the old recipient a long waiting time for a DD graft would be highly undesirable. Donor age more than 65 years has in the whole group been identified as a risk factor for early rejection episodes and would therefore suit older recipients with an attenuated immune response better. By utilizing elderly LD in recipients above 65 years of age, kidney grafts from the DD pool can be used for younger recipients. The results in this old group of patients may be improved if they also receive an age-adapted immunosuppression in parallel with close monitoring (24).
Our study features certain strengths, being population-based from a national center with a constant LD transplantation rate, high donor age, and cyclosporine-based immunosuppressive protocol for more than two decades. Furthermore, only one patient was lost to follow-up during the study period and was therefore not included in the analysis. Registry data may be useful for identifying trends and generating hypothesis, but they are limited by an inability to determine the reasons behind those trends.
In conclusion, the result of our study demonstrated that in LD transplantation the use of donors more than 50 years did not result in an increased rejection rate or reduced graft survival when acute rejection episodes were avoided. The incidence of acute rejection episodes was only elevated when the donor was above 65 years. Our study showed that donors up to 65 years provide excellent long-term results. Although donor age above 65 years and male donor sex turned out to be significant risk factors for graft loss in our study, these results need to be confirmed. Therefore our study supports the continued use of older LD who meet carefully constructed medical criteria and who are highly motivated to donate. Furthermore, donor age should be considered more important than donor sex when a choice between potential donors has to be made.
The authors are grateful to Dr Per Fauchald for critical discussion of aspects of this manuscript. The corresponding author has been supported with grants from the Norwegian Foundation of Health and Rehabilitation.
1. Kassiske BL, Snyder J, Matas AT, et al. The impact of transplantation on survival with kidney failure. In: Terasaki PI, Cecka JM, eds. Clinical Transplant 2000. Los Angeles, UCLA Tissue Typing Laboratory 2001, pp 135–143.
2. Moreso F, Seron D, Gil-Verent S, et al. Donor age and delayed graft function as predictor of renal allograft survival in rejection free patients. Nephrol Dial Transplant
1999; 4: 930.
3. Poggio ED, Hila S, Stephany B, et al. Donor kidney volume and outcomes following live donor kidney transplantation. Am J Transplant
2006; 6: 616.
4. Zeier M, Döhler B, Opelz G, et al. The effect of donor gender on graft survival. J Am Soc Nephrol
2002; 13: 2570.
5. Vereerstraeten P, Wissing M, de Pauw L, et al. Male recipients of kidneys from female donors are at increased risk of graft loss from both rejection and technical failure. Clin Transplant
1999; 13: 181.
6. Øien CM, Varberg Reisæter A, Leivestad T, et al. Gender imbalance among donors in living kidney transplantation: the Norwegian experience. Nephrol Dial Transplant
2005; 20: 783.
7. Kerr SR, Gillingham KJ, Johnson EM, et al. Living donors >55 years: to use or not to use. Transplantation
1999; 67: 999.
8. Prommool S, Jhangri GS, Cockfield S, et al. Time dependency of factors affecting renal allograft survival. J Am Soc Nephrol
2000; 11: 565.
9. Toma H, Tanabe K, Tokumoto T, et al. Time dependent risk factors influencing the long-term outcome in living donor grafts: donor age is a crucial risk factor for long term graft survival more than 5 years after transplantation. Transplantation
2001; 72: 941.
10. Westlie L, Leivestad T, Holdaas H, et al. Report from the Norwegian national hospitals living donor registry: one year data, January 1, 2002. Transplant Proc
2003; 35: 777.
11. Fauchald P, Sødal G, Albrechtsen D, et al. The use of elderly living donors in renal transplantation. Transpl Int
1991; 4: 51.
12. Rothman KJ. Measuring interactions. In: Rothman KJ, ed. Epidemiology: An Introduction. New York, NY, Oxford University Press 2002, p 168.
13. Hosmer DW, Lemeshow S. Confidence interval estimation of interaction. Epidemiology
1992; 3: 452.
14. De Fijter JW, Mallat MJK, Doxiadis II, et al. Increased immunogenicity and cause of graft loss of old donor kidneys. J Am Soc Nephrol
2001; 12: 1538.
15. Matas AJ, William PD, Sutherland DE, et al. 2,500 living donor kidney transplants: a single-center experience. Ann Surg
2001; 234: 149.
16. Pessione F, Choen S, Durand D, et al. Multivariate analysis of donor risk factors for graft survival in kidney transplantation. Transplantation
2003; 75: 361.
17. Holdaas H, Fellström B, Jardine AG, et al. Effect of fluvastatin on cardiac outcomes in renal transplant recipients: a multicentre, randomized, placebo-controlled trial. Lancet
2003; 361: 2024.
18. De La Vega LS, Torres A, Bohorquez HE, et al. Patient and graft outcomes from older living kidney donors are similar to those from younger donors despite lower GFR. Kidney Int
2004; 66: 1654.
19. Sijpkens YWJ, Doxiadis IIN, Mallat MJK, et al. Early versus late acute rejection episodes in renal transplantation. Ttransplantation
2003; 75: 204.
20. Halloran PF, Melk A, Barth C. Rethinking chronic allograft nephropathy: the concept of accelerated senescence. J Am Soc Nephrol
1999; 10: 167.
21. Neugarten J, Kasiske B, Silbiger SR, et al. Effects of sex on renal structure. Nephron
2002; 90: 139.
22. Neugarten J. Gender and progression of renal disease. J Am Soc Nephrol
2002; 13: 2807.
23. Kasiske BL, Snyder J. Matching older kidneys with older patients does not improve allograft survival. J Am Soc Nephrol
2002; 13: 1067.
24. Martins PN, Pratschke J, Pascher A, et al. Age and immune response in organ transplantation. Transplantation
2005; 79: 127.