How Old is Old for Transplantation?

Introduction

The number of elderly patients accepted in renal replacement programmes is continuously increasing. In Scotland, a country of five million people, the incidence of end stage renal disease (ESRD) has risen dramatically from 61 per million population (pmp) in 1990, to 109 pmp in 1999. The percentage of ESRD patients over 60 years has increased from 42% in 1989, to 65% in 1999. In fact, patients aged over 65 years represented more than half of the total number of new patients in 1999 (1). A similar trend was noted in the United States, where the number of patients over 65 years requiring RRT doubled in the last decade (2), Australia (3), Japan, Canada and the rest of Europe (4).

On the basis of evidence that transplantation is safe and successful (5,6) and survival with a transplant is better than that on dialysis (7,8) even in older patients, there is a general consensus that age per se should not represent a barrier to transplantation. And yet, many centers are still reluctant to accept elderly patients onto the waiting list due to their comorbid conditions (9) and shorter life expectancy. In addition, an increased age at the time of transplantation has a significant impact on long-term graft survival (10) and death with a functioning graft is a common event during the follow-up (11).

Although in the United Kingdom there is no age limit for access to transplantation and the listing and transplantation criteria are identical for all patients irrespective of their age, only 7.2% of transplant recipients are aged over 65 years (12). In the United States, where extended criteria donors (ECD) are used increasingly in recent years, only 9.9% of patients over 65 years are transplanted with non-ECD, while 22.9% receive an ECD kidney (13). In Scotland, where the UK national listing and transplantation criteria are applied, significant inequities in access to the waiting list and renal transplantation according to patients' age have been described (14). Only 26% of those aged 60–64 years and respectively 8.5% of those over 65 years are on the waiting list for transplantation within 3 years of starting RRT. After listing, only 43% of the 60–64 years old patients and respectively 29% of those over 65 years are transplanted within 3 years. This is in stark contrast to over 70% of patients aged 18–49 years listed within 3 years of starting RRT. More than 60% of those listed in this age group received a transplant within 3 years.

In the context of the current organ shortage, there is a continuous debate whether elderly patients should go onto the national waiting lists, or they should be part of specially designed schemes to which older or marginal kidneys are preferentially allocated. Currently in the United Kingdom, there are no such schemes and the selection criteria for the transplant waiting list and subsequent transplantation are identical for all patients irrespective of their age.

Therefore, the aim of this paper was to compare the results of kidney transplantation in different age groups and to investigate whether transplantation is worthwhile in the elderly or we should set an age limit for access onto the waiting list.

Methods

All adult patients who started dialysis between 1 January 1989 and 31 December 1999 and were transplanted with a cadaveric kidney (first graft) until 31 December 2000 (n = 1095) were grouped according to their age at grafting (18–49 years, 50–59 years, 60–64 years and >65 years). The kidneys were distributed according to the national criteria set by the UK Transplant. They are based on closeness of HLA with three tiers (tier 1 = no HLA mismatch, tier 2 = one mismatch for HLA-A and/or HLA-B and no mismatches for HLA-DR, tier 3 = one or two HLA-DR mismatches and/or two mismatches for HLA-A and/or HLA-B). At each level, priority is given to pediatric over adult recipients, highly sensitized over non-sensitized, local versus national recipients. If more than one recipient is identified, a point scoring mechanism is used as a discriminator based on the following criteria: recipient age, donor/recipient age difference, waiting time, matchability, sensitization and balance of exchange between centers. Although the algorithm has been revised several times throughout this study, all criteria continued to apply to all patients, irrespective of their age.

The sociodemographic and extensive comorbidity data, as well as the level of HLA matching, the length of the cold ischaemic time, patient and graft survival, the incidence of acute rejection episodes and delayed graft function and the causes of death and graft failure were obtained from the Scottish Renal Registry and UK Transplant databases and case notes, and compared between the four groups. Social deprivation was assessed with the Carstairs score, a combination of four variables (male unemployment, car ownership, social class and overcrowding), derived from the census, calculated for each postcode and classified into seven categories from 1 (least deprived) to 7 (most deprived) (15). Patient and graft half-life were calculated assuming a constant death rate beyond a year after transplantation and compared according to patients' age using a Log-rank test.

The risks of death and graft failure were calculated after adjustment for comorbidity and other sociodemographic variables for each group of patients, using a Cox proportional hazards model. Further comparisons were performed using chi-square, t-test, Mann-Whitney U test, Kruskal-Wallis and ANOVA test where appropriate.

Results

The demographic characteristics of the study population according to the age at transplantation are shown in Table 1.

More men were transplanted in the youngest age group, but there was a comparable representation of all social deprivation categories in all four groups in the study. There was a similar prevalence of diabetes as a cause of renal failure, and almost 60% of the patients in each age group started renal replacement therapy on haemodialysis. The time spent on dialysis until transplantation increased significantly with patient's age (p = 0.031, Kruskal-Wallis test), from a median of 1, 3 years for those aged 18–49 years to almost 2 years for patients aged over 65 years. On average, almost one third of all patients switched between dialysis modalities until transplantation. This occurred significantly more frequently in the younger patients (p = 0.034, chi square), while 76% of patients aged over 65 years remained on the initial RRT modality until grafting (haemodialysis or peritoneal dialysis).

Significant differences were noted between centers, elderly patients being more likely to be listed in centers 1 and 2.

The most important differences in comorbidity conditions between transplant recipients according to their age at grafting are illustrated in Table 2. There was a significant increase in the incidence of peripheral vascular disease, arrhythmias and left ventricular hypertrophy with increased age. Ischaemic heart disease, heart failure and gastrointestinal disorders were more commonly seen in patients over 50 years compared with those aged 18–49 years.

Valvular diseases and pulmonary embolism were most common among patients over 65 years, respiratory diseases were more often seen in patients aged 60–64 years, while patients in the 18–49 year-old group had half the prevalence of CMV compared with all other groups. The overall prevalence of comorbid conditions seemed to be higher in patients aged 60–64 years rather than in those aged over 65 years. This indicates, that the eldest patients who were eventually transplanted underwent a tight assessment process and only those with fewer comorbid conditions were selected.

All transplants had comparable levels of HLA matching irrespective of the recipient's age (Table 3). Between 48% and 58% of patients were well matched (tier 1 [000 mismatches] or tier 2 [maximum 2 A and/or B mismatches but no DR mismatches]). Overall, there was a tendency to offer fewer fully matched kidneys (000 mismatches) to recipients aged >65 years, but this was compensated by a higher proportion of tier 2 (010, 010 or 110 mismatches) grafts.

As expected, the donor age increased, the older the recipient, from 38 years in the younger recipients, to 48 years in those >65 years, but with a comparable gender distribution.

Younger patients (18–49 years) were transplanted within a shorter cold ischaemic time (mean 1126 min), but had a higher incidence of acute rejection episodes, while elderly patients (>65 years) had a higher incidence of delayed graft function.

All patients had an excellent 1-year survival rate (Figure 1). There were no significant differences between the survival curves in the first year post-transplantation, but as expected, there were substantial differences in the long-term survival between those younger and those older than 60 years. It was noted that up to 5 years post-transplant, those aged >65 years had a better survival than patients aged 60–64 years. At 8 years post-transplant, half of the patients in the 60–64 years old group were alive compared with one third in the >65 group.

Similar differences were noted for the graft survival (Figure 2), 60% of the transplants performed in patients younger than 60 years functioning at 8 years, in contrast with only 30–40% in those over 60 years. However, graft survival in patients aged >65 years was comparable with that seen in the 50–59 years group for the first 5 years post-transplant. When the two groups of elderly recipients were compared, patients over 65 years had a better graft survival throughout the study period.

This advantage persisted when graft survival was censored for death with functioning graft (Figure 3). In fact, the older group had the best graft survival at all time points, but overall, the differences between the study groups were not statistically significant (p = 0.2685, Log-rank test).

An examination of the crude death rate (Table 4) showed a fourfold increase from 2.4 per 100 years of patient follow-up in the 18–49 years group to 10.35 in those aged >65 years. This was further illustrated by the differences in the proportion of patients from each group, dying within the study period. Patient half-life showed a significant reduction in the life expectancy, from 37 years in the youngest group to less than 8 years in patients >65 years (p = 0.0001, chi-square). Nevertheless, it is important to highlight that the death rate noted in patients over 60 years who were transplanted, was significantly lower than that observed in similar aged patients who were listed but remained on dialysis (16 per 100 years of patient follow-up). When the risk of death on dialysis on the waiting list was compared with the risk of death following transplantation (not adjusted for graft failure), it was observed that despite an initial (30 days) higher risk of death following transplantation (RR = 1.52, CI = 0.2–11.81), patients over 60 years have a 70% lower risk of death (RR = 0.30, CI = 0.16–0.56) beyond a year post-transplant.

When the relative risk of death, adjusted for comorbidity conditions, was determined for all transplant recipients, those over 50 years had a largely increased risk of death compared with the baseline represented by those aged 18–49 years (Table 4). A similar tendency was observed for graft failure, more than 40% of the grafts in the eldest group being lost throughout the 11 years of follow-up, a large proportion of these due to death with functioning graft. Despite a significant disproportion in the graft half-life between the four groups, a kidney allograft transplanted in patients under 60 years was likely to function for at least 9 years (one quarter of the estimated patient half-life), while patients over 65 years enjoyed an almost entire dialysis-free life following transplantation (graft half-life 6.96 and patient half-life 7.88, respectively).

A comparison of the relative risk of graft failure adjusted for the confounding comorbidity, showed no significant differences between the four groups of transplanted patients (p = 0.201, Cox regression analysis, Table 4).

The incidence of death with functioning graft was three times higher in patients aged >65 years compared with those aged 18–49 years, in contrast to immunological failures, which had an entirely opposite trend and were most frequent in the youngest age group (Table 5). However, a significant proportion of the causes of graft failure in the 18–49 years old group were unaccounted for.

Significant differences in the proportion of patients dying due to infection or vascular causes were noted between the four groups (Table 6), but it is important to note that there was no difference in cardiac death between all age groups. Nevertheless, between one in four and one in five deaths were not known in each group. If we consider these unknown cases to represent sudden deaths, and therefore have a cardiac origin, this would increase the cardiac causes of death to 50% of all kidney transplants performed in Scotland, irrespective of the recipient age.

Discussion

The demographics of ESRD demonstrate a constant increase in the age of new patients starting replacement therapy every year. Currently in Scotland, as in many other parts of the world (3,4), more than half of the new patients are aged 65 years or more, but only few of them will ever be transplanted. These patients represent a particular controversial category (5) and therefore in this analysis we decided to investigate the outcome of transplantation in those over 65 separately from patients aged 60–64 years.

Despite a general agreement that age should not represent a contraindication for transplantation, the proportion of kidney transplants performed in elderly patients in United Kingdom has not changed much in the last decade (12,16) and many units are still reluctant to accept older patients as transplant candidates. This is clearly illustrated in this analysis by 10% lower transplant rates in two of the centers, compared with the national average.

Elderly transplant recipients spend a longer time on dialysis and this may be partly due to a lengthier assessment period necessary to confirm suitability for transplantation. It is fairly clear from the present analysis that elderly transplant recipients have a higher index of comorbidity and serious conditions, which shorten the life expectancy, such as cardiovascular and respiratory conditions, are more often present in those aged over 60 years. This does indicate that the lower transplant rates may be a result of the high prevalence of comorbid conditions in this particular age group, which renders most of the patients unsuitable for transplant candidacy.

Older recipients receive kidneys from older donors, but the donor age range is comparable between the four groups, with an upper limit as high as 75 years. Donor age is a controversial point in kidney transplantation, as graft failure rates are higher with increased donor age (17,18). Nevertheless, the use of older donors is considered acceptable due to the scarcity of cadaveric kidneys and good results have been reported when such kidneys were transplanted in aged matched recipients (19). If kidneys from donors as old as 75 years can be accepted for implantation, there is no real justification why an increased number of elderly patients could not receive a kidney graft in elderly-for-elderly programmes (20,21), which would eliminate potential allocation obstacles as well as shifting younger donor kidneys to younger recipients.

All groups have excellent 1-year patient-survival rates, ranging between 87% and 98% and it is worth noting that the 91%, 1-year patient survival in the >65 years group, is comparable with that seen in patients aged 50–59 years. Beyond 3 years, there is a substantial survival benefit for younger patients, while the two groups of patients aged over 60 years have comparable, but diminished survival rates, two thirds being alive at 5 years and between 30–50% at 8 years. A similar trend is noticed for the graft survival, with comparable 1-year figures (75% to 88%) and a long-term advantage for younger recipients. The crude rate of graft loss increases with age, a significant proportion of the grafts in elderly patients being lost due to patient death. This should not necessarily be interpreted as a waste of kidneys, especially if the survival rates and the life expectancy noted in this study are set in the context of a 2.1 years median survival for all patients >65 years on renal replacement therapy (1) and a higher risk of death on dialysis on the waiting list. It is also important to note that while in the youngest group, the estimated graft half-life is nearly 10 years, against an estimated patient half-life of 37 years, a 7 years graft half-life in the eldest group ensures a dialysis-free life in the context of a half-life of 8 years, offering a substantial improvement in the quality of life.

The patient and graft survival probabilities noted here are comparable with those noted in the United States (13) and elsewhere (5,18,22–24). In addition, the comparison performed in this study has shown that results in patients over 65 years are similar to those obtained in patients aged 60–64 years and therefore these patients should be considered for transplantation. One question which still remains unanswered is whether these patients should be placed on the national waiting lists and hence compete for organs with younger candidates or whether they should be enrolled in special programmes such as ‘old-for-old’, which may allow a quicker access to a graft, but are yet to demonstrate sustainable results (25).

A critical issue for the success of transplantation in the elderly is the assessment process and there is evidence that with a strict evaluation, 80%, 5-year patient and graft survival in these patients is achievable (26). The incidence of death due to cardiovascular diseases was not significantly different in the four groups in this study, but this may be a reflection of the higher incidence of cardiovascular disease in the general population in Scotland. However, it also suggests that elderly patients, despite a higher prevalence of these conditions, when correctly assessed and selected, may not exhibit a higher risk of death as a direct result of them.

In conclusion, older patients with ESRD present health care professionals with a significant challenge and many management issues remain subject to debate. The results of this paper seem to indicate the need for a review of the assessment process in order to maximize the use of the available donor kidneys. With a prevailing shortage of organs, kidney transplantation should not be used indiscriminately in the elderly, nor should the elderly be denied access simply on the basis of age. Careful assessment of 'biological' rather than ‘chronological’ age should be used on an individual basis instead of applying rigid age limits and perhaps survival prognostic models with applicability in the assessment clinic (27) will allow us to select the most appropriate candidates in order to maximize the benefit for the elderly patients and the society.