Long-Term Consequences of Live Kidney Donation Follow-Up in 93% of Living Kidney Donors in a Single Transplant Center

Abstract

Live kidney donation is increasing rapidly. Increases of blood pressure and proteinuria but no accelerated loss of renal function in kidney donors have been described. The credibility of this research is hampered by retrieval rates of only 50–70% of donors.

We studied renal function, blood pressure, proteinuria, parathyroid hormone, 1,25(OH)₂ cholecalciferol and calcium and phosphate excretion in a live kidney donor cohort with a 93% retrieval rate. A comprehensive physical and laboratory examination including 24-h urine collection was conducted. None of the 152 donors had renal failure. Mean time after uninephrectomy was 11 ± 7 (range: 1–28) years. GFR had declined by 25%. Blood pressure had increased from 125 ± 15/79 ± 11 to 134 ± 19/81 ± 9 mmHg (p < 0.01) but remained significantly below normal. Fifty six percent of donors developed proteinuria (>150 mg/day), but only 10% had albuminuria. Nineteen percent had increased PTH, 30% had a decreased tubular reabsorption rate of phosphate. Regarding risk factors for a higher loss of GFR, greater increases in blood pressure or proteinuria no consistent picture emerged. Because of the high incidence of proteinuria and possible changes in bone metabolism inclusion of kidney donors in registries appears worthwhile.

Introduction

Due to the continuing organ shortage the usage of kidneys from living donors has gained increasing popularity (1,2). The benefits for the recipients of these kidneys—most probably related to the superior organ quality and the short time on dialysis (3)—in terms of transplant survival and function have been described in numerous reports. The consequences of kidney donation for the donors, however, appear not so clean cut. The initial loss of renal function is in part compensated for and the mean glomerular filtration rate (GFR) has been found to be in the order of 70–75% of the value before nephrectomy (4–7). However, a reduced number of glomeruli is a risk factor for arterial hypertension (8). And indeed some researchers found an increase of blood pressure following kidney donation (4,9,10–12). Analyzing renal function or blood pressure as a function of the uninephrectomy is, however, complicated. Since both renal function and blood pressure are influenced by age it is impossible to dissociate the effects of the aging process from the influence of the time interval between nephrectomy and a later evaluation. Indeed, by using data from an age- and sex-matched normal population, some (5,13), but not all (9) authors found no difference in blood pressure between former kidney donors and the normal population. An elegant study by Williams et al. (14), using the siblings of the donors as controls, found no significant difference in blood pressure, significantly higher serum creatinine values and a significantly higher urinary protein excretion in the donors.

An increase in the amount of protein excreted by the remaining kidney, especially in male donors (14–16), has been found repeatedly (17–21). But in the absence of a correlation between protein excretion and blood pressure or renal function, the clinical significance of this finding remains unclear (12,14,22,23).

In earlier studies (24) an increase in renal phosphate excretion after unilateral nephrectomy has been described. But no further data on calcium and phosphate metabolism in kidney donors exist.

Since the follow-up of most of these studies has been in the order of only 50–70%, the reliability of these data has been questioned (25,26). It has been argued that even if only 20% of the donors are lost to follow-up one cannot assume that kidney donation is innocuous (26).

We therefore attempted to do a complete follow-up examination of renal function, blood pressure, renal protein excretion, body mass index (BMI), smoking habits and parameters of bone metabolism in all 152 living donors in our transplant center from 1973 to 2001. Seven of these donors had died from nonrenal diseases. Of the remaining 145 we collected data on 135 (93%), to our knowledge the highest number reported so far. One hundred and fifteen of the donors (79%) were evaluated in the transplant center. The remaining 20 were seen by their local physicians. A comprehensive evaluation of a number of psychosocial parameters was included and has been presented elsewhere (27).

Methods

Between 1973 and 2001, 152 living donor nephrectomies have been performed in our transplant center. Donors were contacted (if necessary repeatedly) by phone or letter and asked to come to the transplant center for a follow-up evaluation free of charge. They were instructed how to collect 24-h urine and asked to bring it to the center. In the center, blood was drawn and the donors had a complete physical examination including body weight, sonography of the remaining kidney and blood pressure measurement (mean value of three consecutive readings 2 min apart). The laboratory parameters were complete blood count, electrolytes, creatinine, urea, uric acid, cystatin C, phosphate, cholesterol, triglycerides, liver transaminases, intact parathyroid hormone, 1,25(OH)₂ cholecalciferol. The 24-h urine was evaluated for total protein, albumin, α₁ microglobulin, IgG (all measured by nephelometry), calcium and phosphate.

Data on height, blood pressure, serum creatinine and creatinine clearance before donation were taken from the records in the transplant center. The abbreviated MDRD formula (28) was used to calculate GFR. BMI was calculated as weight in kg/height in m². Tubular reabsorption of phosphate (TRP) was calculated as: 100 ×[1− (urine phosphate × serum creatinine/urine creatinine × serum phosphate)] (29).

Donors who were unable or unwilling to come to the transplant center were sent a questionnaire to be filled by their local physician asking for serum creatinine, blood pressure, urine dipstick test and body weight. In donors who refused any evaluation, we questioned the recipients if they had renal disease or renal replacement therapy.

Statistics

Data were analyzed using SPSS software (Chicago, Illinois). Values are given as mean ± SD. Comparisons between groups were made by unpaired Student's t-test or Mann-Whitney U-test as appropriate. Comparisons within groups were made by Wilcoxon test. Qualitative parameters were compared by Mc Nemar's chi square test. The significance of correlations was tested according to Spearman or Pearson. Binary logistic regression was used to evaluate the influence of different variables on renal function and proteinuria.

Results

Demographic data

The demographic data of our live kidney donor population are shown in Tables 1 and 2. Overall there were 152 kidney donors in the years from 1973 to 2001. Seventy percent of our donors were female and 30% were male. Seven donors had died in the meantime, all from nonrenal diseases. Ten of the remaining 145 donors were lost to follow-up. We collected information from the recipients on five of these donors. None of them had known renal disease. In the remaining five donors no information on their status could be obtained. One hundred thirty five donors (93% of those still alive) could be evaluated, 115 had a clinical and laboratory evaluation in the transplant center, the other 20 were seen by their local physicians (Table 1). The mean age at the time of kidney donation was 45 ± 11 years and the mean time from nephrectomy to the current evaluation was 11 ± 7 (range 1–28) years. The mean BMI at the time of evaluation was 26 ± 4 kg/m². Twenty-two percent were smokers (Table 2).

Table 1.  Live kidney donors of the Frankfurt transplant center

No. of donors (from 1973 to 2001)	152	(100%)
	107 female	(70%)
	45 male	(30%)
No. of donors who died	7	(5%)
	5 malignant diseases
	2 leukemias
	1 ovarian carcinoma
	1 breast carcinoma
	1 lung carcinoma
	2 cardiovascular deaths
No. of donors still alive	145
No. lost to follow-up	10	(7% of those alive)
	5 alive without overt renal disease
	5 unknown
No. of donors evaluated	135	(93% of those alive)
	115	(79%) evaluated in the center
	20	(14%) evaluated by local physician

Table 2.  Demographic data of the 135 kidney donors who could be evaluated

*Mean ± SD.
Female	96 (71%)
Male	39 (29%)
Age at donation (years)	45 ± 11*
Age at evaluation (years)	57 ± 11
Years since nephrectomy	11 ± 7
Body mass index	26 ± 4
Smokers, %	22

Renal function

Plasma creatinine concentrations increased from 72.5 ± 15.0 to 85.7 ± 16.8 μmol/L (p < 0.001). Cystatin C could not be measured at the time of donation in most donors. At the time of the current evaluation the mean value was 67.4 ± 13.5 nmol/L and was above the normal value (42.7–71.9 nmol/L) in 23% of donors. Endogenous creatinine clearance as calculated from the 24-h urine sample decreased from 119 ± 30 to 99 ± 30 mL/min/1.73 m². Using the MDRD 4-formula we found a decrease in GFR from 92 ± 20 to 71 ± 15 mL/min/1.73 m². The individual GFR values before donation and at the time of the current evaluation are shown in Figure 1. Considering not the absolute GFR remaining at the time of evaluation but the individual loss of GFR no significant correlation between the magnitude of the loss of GFR with time after donation was seen.

Figure 1. GFR of the individual donors before uninephrectomy and at the time of the current evaluation.

We tried to identify risk factors for a larger decrease in GFR, which are available at the time of donation by binary logistic regression using age, sex, smoking status, BMI, blood pressure and GFR at the time of donation as covariates. The outcome parameter was the decrease in GFR as a percentage of the pre-nephrectomy GFR above or below the median (25.1%). The only significant factor for a larger than average loss of GFR was a higher GFR at the time of donation (99.4 ± 22.0 vs. 84.5 ± 15.7 mL/min/1.73 m²). Univariate comparison of donors having a loss of GFR of more or less than the median by U-test showed the same results. Spearman correlation analysis showed no significant correlation between the loss of GFR and the rise in blood pressure, the amount of proteinuria or time after donation.

Blood pressure

Compared to the values before kidney donation systolic blood pressure rose significantly from 125 ± 15 to 134 ± 19 mmHg (Table 4). Diastolic blood pressure increased nonsignificantly from 79 ± 11 to 81 ± 9 mmHg. The percentage of donors with hypertension (blood pressure above 140/90 mmHg or treatment with anti-hypertensive drugs) increased from 7% to 30%. Since the rise in blood pressure in our donor population could have been caused simply by their increasing age, we also compared the mean blood pressure values with an age- and sex-matched sample from the German population (30). Both systolic and diastolic blood pressure in male and female donors were significantly lower than in the normal population at the time of donation and also at the current follow-up examination (Table 5). This is not unexpected because the donors were of course positively selected for normal blood pressure. Using the data from Table 5, we calculated the mean increase in systolic blood pressure per year for men and women as 0.8 and 1.3 mmHg, respectively. In male donors systolic blood pressure rose by 0.6 mmHg/year, in female donors the increase of systolic blood pressure per year was 0.5 mmHg.

Table 4.  Blood pressure before and after kidney donation

	Before donation	At evaluation	p-value*
*Wilcoxon test. ^†Mean ± SD. ^‡Mc Nemars chi square test.
Systolic blood pressure (mmHg)	125 ± 15^†	134 ± 19	<0.001
Diastolic blood pressure (mmHg)	79 ± 11	81 ± 9	n.s.
Hypertensive, %	7	30	<0.001^‡

Table 5.  Blood pressure in donors compared to the normal German population (n = 7101)

	Normals	Donors	p-value*
*Student's t-test. ^†Mean ± SD, mmHg.
Males aged 40–49 years	135 ± 17/88 ± 11^†	125 ± 2.1/79 ± 2.0	<0.01
Males aged 50–59 years	143 ± 19/89 ± 11	131 ± 2.8/80 ± 1.6	<0.01
Females aged 40–49 years	130 ± 18/82 ± 11	125 ± 1.8/79 ± 1.0	<0.01
Females aged 50–59 years	143 ± 21/86 ± 11	134 ± 2.1/82 ± 1.1	<0.01

Spearman correlation analysis revealed a negative correlation between blood pressure at the time of donation and the increase in blood pressure over time (R =−0.51, p < 0.001), that is, the lower the blood pressure at the time of donation, the larger the increase over time. Despite the fact that, as in the normal population, there were significant correlations of mean arterial pressure with: age, GFR and BMI; the rise in blood pressure in our donors was not significantly correlated with: GFR, age, sex, smoking status and BMI. Proteinuria also was not correlated with the rise in blood pressure.

Proteinuria

Because no patient with pathological urinary protein excretion was allowed to donate a kidney, the percentage of donors with a protein excretion of more than 150 mg/day was zero at the time of donation. Proteinuria has repeatedly been found to increase after kidney donation. But the nature of the proteins and the clinical significance of this finding remain unclear. We therefore measured—in addition to whole protein—the amount of the three differently sized marker proteins IgG, albumin and α₁-microglobulin in the 24-h urine samples of those donors who were evaluated at our center. Fifty six percent of 115 donors had a proteinuria of more than 150 mg/day, but only 10% had pathological urinary albumin concentrations (above 50 mg/L). Seven percent had increased urinary α₁-microglobulin (above 12 mg/L) pointing to tubular injury and 2% had pathological IgG excretion (above 10 mg/L) signifying glomerular disease (Table 6). In 40% of our donors urinary whole protein was increased without a concomitant rise in one of the three marker proteins making it impossible to decide if glomerular or tubular injury was present. Proteinuria was equally present in female and male donors. Using Spearman analysis, we found no correlation of proteinuria with blood pressure, GFR (at the time of evaluation or at the time of donation), BMI, age, smoking, or time since donation (Figure 2). Proteinuria also did not correlate with the magnitude of the rise in blood pressure or of the decrease in GFR. Logistic regression analysis (after dividing the donors in two groups at the median protein excretion of 162 mg/day) revealed no influence of age at donation, diastolic blood pressure at time of donation and at the current evaluation, time since donation, increase of blood pressure over time, GFR at donation and at evaluation, decrease of GFR, BMI, smoking, or sex. Systolic blood pressure at the time of donation was, however, higher in the donors with more than average protein in the urine (128 ± 14 mmHg vs. 121 ± 13 mmHg, p < 0.05 by Mann-Whitney U-test).

Table 6.  Proteinuria in kidney donors

Percentage with protein-excretion > 150 mg/day	56
Percentage with urinary albumin > 50 mg/L	10
Percentage with urinary α1-microglobulin > 12 mg/L	7
Percentage with urinary IgG > 10 mg/L	2

Figure 2. Whole urinary protein/24 h versus time after kidney donation. The correlation was not statistically significant.

Parameters of bone metabolism

The amount of 1 α hydroxylase might be diminished by removal of one kidney putting kidney donors at an increased risk for secondary hyperparathyroidism. And indeed we had found elevated PTH levels in some kidney donors. We therefore measured levels of 1,25(OH)₂ vitamin D₃, intact PTH, calcium and phosphate in the serum and the renal excretion of calcium and phosphate in our donor population. Because different PTH assays had been used, we expressed PTH in percentage of the upper limit of normal (Table 7). The mean values were all normal, except a slightly lowered tubular reabsorption rate of phosphate (TRP, 77 ± 8%). This pathological value was caused by 30% of donors with a decreased TRP. Nineteen percent of the donors had elevated PTH levels and 12% had decreased 1,25(OH)₂ vitamin D₃. There was no statistically significant correlation between PTH and 1,25 (OH)₂ vitamin D₃ or the TRP. Also there was no significant positive or negative correlation between donor age and PTH or 1,25 (OH)₂ Vitamin D₃ levels. This was true for the group as a whole as well as for both sexes analyzed separately. Also there was no significant difference in PTH or 1,25 (OH)₂ Vitamin D₃ levels between males and females.

Table 7.  Parameters of bone metabolism in kidney donors

*Mean ± SD.
Serum phosphate (mmol/L)	0.90 ± 0.19*
Serum calcium (mmol/L)	2.4 ± 0.09
Urinary calcium excretion (mmol/day)	4.2 ± 2.4
Tubular reabsorption rate of phosphate (%)	77 ± 8
Percentage with decreased tubular phosphate reabsorption	30
iPTH (Percentage of upper limit of normal)	77 ± 52
Percentage with increased iPTH	19
1,25(OH)2 Vitamin D3 (pmol/L)	77.5 ± 30.0
Percentage with decreased 1,25(OH)2 Vitamin D3	12

Discussion

The potential long-term problems associated with unilateral nephrectomy for live kidney donation have been a matter of debate since the beginning of this surgical procedure (31–33). During the first 15–20 years most authors relied on the research on renal function and survival of persons with congenital unilateral kidneys or who had had a nephrectomy for reasons other than kidney donation to prove that live kidney donation was not associated with excess long-term medical risks (34). In the 1980s, papers by Hostetter et al. (35) and others (9) on the consequences of surgical ablation of renal mass sparked new interest in the fate of the living kidney donor. Indeed a number of reports appeared that showed the development of proteinuria in persons with a single kidney (16,20,36) and former kidney donors (10,14). Some authors also found a significant increase in blood pressure after renal donation (9,15,37). One problem with the vast majority of the papers dealing with the long-term consequences of living kidney donation was the low (50–70%) retrieval rate of the former donors. Some surgeons (25) and nephrologists therefore argued that as long as the fate of the other 30–50% of the donors is unknown the finding of normal renal function or blood pressure of those donors who were evaluated cannot serve as proof that unilateral nephrectomy is without adverse effects (26). Fehrman-Ekholm et al. in 2001 (5) were the first (and to our knowledge the only) to report a cross-sectional follow-up of more than 85% percent of 402 surviving donors from their original renal donor cohort of 451 using a questionnaire. Three of these donors had developed renal disease and one was on dialysis. The mean calculated GFR 12 years after donation was 72% of the age-adjusted normal value and the prevalence of hypertension was not higher than in the general population. Proteinuria (detected by dipstick method) was present in 12% and was correlated with hypertension and a lower GFR.

Like any one else, a kidney donor has a risk of developing renal disease at some point after donation. Indeed, using former kidney donors listed for a renal transplant as indicative of the number of donors with end-stage renal disease (ESRD) Ellison et al. (2) found a prevalence of 0.04% compared to the normal ESRD prevalence of 0.03%. Because presumably not all former donors with ESRD are being listed for a renal transplant the numbers will likely be higher. And the risk for hypertension and, probably to a lesser extent, renal disease may be increased in relatives of patients with these diseases (14). But in the absence of a comprehensive registry for all kidney donors the question if renal disease or hypertension are definitively more frequent in donors than in the normal population cannot be answered yet.

In our cohort of 152 live kidney donors 7 had died in the meantime (all from nonrenal causes). No former donor is on dialysis, two have a proteinuria of more than 1 g/day (a 67-year-old female who had donated a kidney to her son 3 years ago with a blood pressure of 130/70 mmHg, a GFR of 48 mL/min/1.73 m² and 1059 mg of urinary protein/day, and a 62-year-old male who had donated a kidney to his wife 3 years ago with a blood pressure of 140/100 mmHg, a GFR of 47 mL/min/1.73 m² and 1200 mg of urinary protein/day), and one has a proteinuria of 5.2 g/day (a 43- year-old male who had donated a kidney to his brother 16 years ago with a blood pressure of 130/90 mmHg and a GFR of 90 mL/min/1.73 m²). We were able to evaluate 79% of those still alive in our own center and had to rely on the evaluation by local physicians in 14%. In accordance with Fehrman-Ekholm et al. (5) and other studies (4,6,15,23) we found a decrease in creatinine-clearance or GFR by 20–25%. Surprisingly, but in accordance with some (14,16) but not all (5) former studies, we could not find a correlation between residual renal function and blood pressure or the amount of proteinuria in our cohort.

Blood pressure and the percentage of donors with hypertension increased significantly over time, as it does in the normal population. We therefore think that the most meaningful way to analyze the increase in blood pressure in kidney donors is to compare blood pressure itself or its increase over time to an age- and sex-matched normal population. Since kidney donors should be positively selected for normal blood pressure, one would expect their blood pressure to be lower than in the normal population at least at the time of donation. And this was indeed the case and remained true after a mean time of 11 years. The increase of blood pressure was not larger than in the normal population, although the possibility that blood pressure increased more than it would have without kidney donation is difficult to exclude. In our cohort of donors uninephrectomy, therefore, does not appear to increase the risk of developing hypertension above that of normal aging. And Williams et al. (14), as stated in the Introduction, also found similar blood pressures in donors and their siblings who had retained their two kidneys.

The percentage of donors with proteinuria was much higher in our study than in the study of Fehrman-Ekholm et al. (5). The reason for this discrepancy may lie in the low sensitivity of the dipstick method in comparison to our quantitative determination from a 24-h urine. Proteinuria in about the same range that we found has also been described by other authors (12,14,19). In the report by Miller et al. (12) 39% of donors had proteinuria as tested by 24-h urine sample, all of whom were negative on standard dipstick testing. Similar to our data there was no correlation between the amount of urinary protein and time after donation. So proteinuria appeared not to be progressive. In contrast to the study published by Praga et al. (38) we could not find a significant difference in BMI between donors with and without proteinuria. Neither did we find a preponderance of males in the group of donors with proteinuria as has been suggested by the data from Zucchelli et al. (16) and Williams et al. (14).

The nature of the urinary proteins remains a matter of debate. Despite a urinary whole protein concentration above 150 mg/L in more than half of our donors, urinary albumin was increased in only 10% and pathological excretion of α₁-microglobulin or IgG, indicating tubular or glomerular injury, respectively, was found in even lower numbers. Najarian et al. (23) found an incidence of proteinuria of 22% in 57 donors but only 7% also had albuminuria. In a quantitative and qualitative comparison of urinary protein excretion in 24 healthy subjects and 21 patients with one kidney (11 with unilateral renal agenesis and 10 with uninephrectomy) Oberle et al. (20) found significantly higher whole protein excretion (206 vs. 630 mg/24 h) in the latter. Glomerular as well as tubular patterns were more frequent in the patients with a single kidney.

Although the nature of the urinary protein in persons with a single kidney has not been sufficiently clarified, most authors agree that the proteinuria is neither progressive nor associated with decreased GFR or hypertension (9,14,15,23). And this is consistent with our data.

Data on parameters of bone metabolism in kidney donors have been lacking so far. After finding an elevated PTH in some of our former kidney donors and after reviewing the early article by Pabico et al. (24) we became interested in this question. And indeed we found elevated PTH levels in 19% of our donor population. Surprisingly PTH levels were not correlated with the 1,25(OH)₂ vitamin D₃ levels. Since PTH increases the activity of the 1 α hydroxylase this may reflect a new steady state after an initial drop of the 1,25(OH)₂ vitamin D₃ levels. PTH levels are known to increase in the elderly (39). Some authors suggested this to be caused by declining active Vitamin D levels related to decreased sun exposure, diet or decreasing GFR (40). Others found a negative correlation between GFR and PTH with normal Vitamin D levels (41,42), more resembling our findings. Since this is thought to be a pathological process, no age-adjusted normal values for comparison with our data exist. Serial measurements of 1,25(OH)₂ vitamin D₃ and PTH levels shortly after donor nephrectomy could be worthwhile. In the face of a lowered TRP in 30% of our donors, an accelerated loss of bone mass might be anticipated. Unfortunately we did not collect data on bone mineral density or fracture rate, so we cannot comment on the clinical significance of this finding.

In summary we can say that our cross-sectional study with its near-complete follow-up confirms earlier research with regard to residual renal function after live kidney donation. Blood pressure in our donors remained lower than in the normal population. Like others before, we found an increase in urinary protein excretion. Pathological albuminuria or excretion of IgG was, however, rare. Nineteen percent of donors had elevated levels of intact PTH and 30% had a decreased TRP. Further studies on the nature and long-term (more than 20 years) consequences of the proteinuric state of many kidney donors, on renal function, and on bone metabolism appear both worthwhile and necessary. Given the lack of data on risk factors for the later development of proteinuria and hypertension, clinical renal disease or early osteoporosis the inclusion of all renal donors in national or international registries appears mandatory.

Acknowledgment

This study was supported by a grant from Novartis^® and Herbert-Quandt-Stiftung.