Abbreviation: DGF, delayed onset of graft function.
Shortage of organs for transplantation is a major factor preventing a reduction in the growing number of patients on dialysis and on transplant waiting lists. Expansion of the donor pool by limiting the exclusion criteria for cadaver donors is one possible solution (1, 2). In 1993 we decided, at Transplant Centre, Institute of Clinical and Experimental Medicine in Prague, Czech Republic, to expand the criteria for acceptance of cadaver donors. Previously, donors older than 50 years of age and those with a history of arterial hypertension or diabetes mellitus were not accepted. As a consequence of expanded criteria, the number of renal transplants performed more than doubled from 79 kidneys in 1992 to 187 in 1994. In addition to the expansion of the criteria, improvements in the organization of the donor program, such as transplant coordination and education programs were important for this success.
There will always be a limit to the expansion of criteria for acceptance of cadaver donors. In setting up criteria, it is important to analyze risk factors associated with reduced graft survival or graft function. The factor most commonly shown to have a detrimental effect on graft survival and graft function is high donor age (1–8). With increasing age decreasing renal functional capacity can be seen as a decrease in creatinine clearance (9) associated with an increasing number of sclerotic glomeruli (10). Despite these changes renal insufficiency does not appear clinically relevant in healthy old people (11). However, a kidney from an elderly donor may not have the capacity to withstand the additional stresses of cold ischemia, nephrotoxic immunosuppression, and requirement to function in place of two native kidneys, with the reduction in functional mass leading to hyperfiltration of residual nephrons and subsequent damage (12) that in the long-term may lead to chronic allograft nephropathy and graft failure (13–16).
There are only limited data on the outcome after transplantation of kidneys with severe glomerulosclerosis. Reductions in graft survival in a small number of recipients of kidneys with more than 20% glomerulosclerosis have been reported (17). Biopsy findings in marginal donors could potentially be an instrument to exclude high-risk renal grafts from being transplanted and causing the recipient nonbeneficial trauma and complications. The objective of this study was to evaluate the procurement donor biopsy as a means to set criteria for donor acceptance in addition to medical history and donor age. The study included biopsy sampling at the donor operation in a large series of consecutive donors during a 3-year time period with evaluation of morphological findings relative to the graft functional outcome.
MATERIAL AND METHODS
This was a prospective study of renal biopsies from cadaver donor kidneys obtained from June 1993 through June 1996 in the region of the Transplant Centre at Institute of Clinical and Experimental Medicine. Inclusion criteria for the study were that a representative renal biopsy (visualizing at least six glomeruli) was available and that the kidneys were transplanted. Altogether there were 291 donors during the study period, and for logistic reasons biopsies were not taken in 56 cases. Of the 235 available donors, 12 were excluded as the kidneys were not transplanted (9 cases due to impaired donor renal function combined with biopsy finding of severe vascular changes and glomerulosclerosis from 20 to 60%, 1 donor was HbsAg+, and 2 had renal adenocarcinoma) and 23 donors were excluded because the biopsy was not representative, including only a thin subcapsular part of the renal cortex. Thus, the analysis included 200 donors. At least one kidney from each pair was transplanted and overall 387 recipients could be followed-up long-term, until November 1998 [4 kidneys were sent abroad in the highly immunized transplant project, and 9 kidneys were not transplanted due to surgical complications or donor trauma affecting one kidney]. Initially, both donor kidneys were biopsied, but as the findings were routinely very similar (left versus right kidney; number of glomeruli, r=0.74; percent sclerotic glomeruli, r=0.88; n=29;P <0.0001), the protocol was amended to include only unilateral renal biopsy. Altogether 228 patients received a kidney that had been biopsied (because 1 kidney of the 29 was not used for technical reasons).
Renal biopsy methods and classification.
Biopsy was performed by wedge excision from the renal cortex at the donor operation after in situ perfusion and retrieval of the kidneys. The biopsy sample was fixed by 10% formaldehyde and stained with hematoxylin-eosin, periodic acid-Schiff, anelinefuchsin, silver impregnation, and staining with orange G. Proportion of global sclerotic glomeruli (Fig. 1) and incidence and severity of vascular, tubular and interstitial changes were evaluated by light microscopy by one pathologist (MC) in a “blinded” fashion. The severity of changes was estimated semiquantitatively and the following criteria were used for absent, 0; mild, 1; moderate, 2; and severe changes, 3. Hyaline arteriolosclerosis (Fig. 2, A, B, and C) and fibrinoid necrosis (Fig. 3, A, B, and C): focal, small deposition, 1; focal, circulatory with narrowing of less than one-third of lumen, 2; circulatory with narrowing of more than one-third of lumen; 3. interstitial fibrosis (Fig. 4): focal, 1; less than one-third of sample, 2; more than one-third of sample, 3; tubular vacuolisation (Fig. 5) and epithelial desquamation (Fig. 6), respectively: focal, 1; less than half of tubuli, 2; more than half of tubuli affected, 3.
Kidneys were allocated to recipients on the computerized national waiting list according to complex allocation rules. In brief, blood group compatible, cross-match negative patients with a maximum compatibility index, which emphasises the absence of DR mismatch, were allocated transplants favoring cases with sensitization and clinical urgency.
Donor and recipient data.
We collected data on donor age, sex, cause of brain death, history of arterial hypertension (treatment with at least one antihypertensive drug), preharvest time in Intensive Care Unit, intraoperative last hour diuresis, intraoperative serum creatinine, type and dose of catecholamines, preservation solution, and multiple or single organ explantation. Recipient data included age, sex, number of transplants, current level of panel reactive antibodies, number of HLA antigen mismatches, duration of cold ischemia time, and patient and graft survival. Graft function was evaluated by 24-hr creatinine clearance at 1 and 3 weeks, and 3, 6, 12, 18, and 24 months after transplantation. Delayed onset of graft function was defined by one or more dialyses performed in the immediate posttransplant period. Date of graft failure was defined as date of return to chronic dialysis.
All recipients received triple immunosuppressive therapy including cyclosporine A, azathioprine, and steroids. In 21% of patients, those with sensitization (current panel reactive antibody level of more than 50%) or retransplantation, prophylaxis including OKT3 or antithymocyte globulin was given. Acute rejection was either biopsy-proven or presumed (patient received antirejection treatment). First-line therapy of acute rejection was methyl-prednisolone with OKT3 used in steroid-resistant cases.
Intraobserver variability of biopsy assessment was analyzed using weighted kappa and repeated measures analysis of variance. Multiple regression analyses were used for correlation of donor data and graft function, although simple regression analysis was used for correlation between biopsy findings and donor age, creatinine clearance and donor age, and creatinine clearance and glomerulosclerosis. Cox proportional hazards model was used for association between donor data and graft survival. Graft survival was censored for patient death with function and analyzed according to Kaplan-Meier. Unpaired t test or χ2 test was used to compare groups. Graft function was evaluated by 24-hr creatinine clearance. Data are given as mean±SD, if not otherwise indicated. P <0.05 was used in all statistical analyses. Finally, in patients treated with dialysis due to delayed onset of graft function or graft loss, creatinine clearance was arbitrarily set to the minimal value of 0.6 ml/min.
In total, 200 cadaver donors were included in the study of which 139 (69.5%) were male. Mean donor age was 39±15 years (range 7–69), 27% of donors were 50 years or older. The cause of brain death was trauma in the majority of cases (53%), 36% had spontaneous intracerebral bleeding and in 11% there were other causes. A history of arterial hypertension was noted in 29 cases (14.5%). Cardiac resuscitation had been necessary in 26 (13%) donors. Most donors (60%) were circulatory stable, whereas 25% received dopamine in doses between 6–10 μg/kg/min, and 15% in doses above 10 μg/kg/min plus noradrenaline. The donor stay in the intensive care unit ranged from 0 to 19 days (median 2.5 days). Intraoperative serum creatinine was mean 112±39 μmol/L (range 40–260) and the intraoperative last hour urine production was 290±210 ml (range 50–1080). Multiorgan explantation was performed in 48%. Euro-Collins perfusion solution was used in 63% and Viaspan solution in 37% of the cases.
There were 387 kidney recipients, 224 (57.8%) of them were transplanted in our department and the remaining in five other centers of the Czech Republic. Mean age was 47±12 years, ranging from 17 to 74 years, and 55% were male. A total of 340 (88%) patients received their first transplant, with 2 (0.5%) receiving their third renal graft and the remaining their second. Median current panel reactive antibody level was 6%, range 0–100. Most recipients (68%) had antibodies between 0 and 20% and 19 (5%) had current panel reactive antibody of more than 80%. The mean number of HLA antigen mismatches was 2.6±1.1 (range 0–6), and the mean number of DR mismatches was 0.55±0.57. Mean cold ischemia time was 20.3±4.7 hr (range 4 - 33). In 19% of transplants, the ischemia time exceeded 24 hr.
Overall, the mean number of glomeruli in the renal biopsies of the 200 donors (for 387 recipients) was 27.6±16.5 (range 6–101), and 25 glomeruli or more was found in 103 biopsies (for 198 recipients). The overall proportion of sclerotic glomeruli was 8.4±11.1% (range 0–47.6). In 94 (47%) biopsies, glomerular sclerosis was absent, in 71 (36%) it was present but affecting less than 20% of glomeruli and in 35 (18%) of biopsies it affected 20% or more. Incidence and severity of vascular, interstitial and tubular changes are shown in Table 1. Thrombi in the glomerular capillaries were found in 5 procurement biopsies (for 10 recipients).
Intraobserver variability in grading of morphological changes and measures of glomerulosclerosis was analyzed after a repeated assessment. Systematic error was negligible and not statistically significant. Error due to random was only 0.73 for total number of glomeruli, 0.08 for number of sclerotic glomeruli, and 0.66 for percentage of sclerotic glomeruli (n=50), all showing good agreement between assessments. Weighted kappa value for grading of hyaline arteriolosclerosis was 0.83; interstitial fibrosis 0.78; vacuolization 0.81; and for desquamation 0.83. A kappa value of more than 0.80 is regarded as very good strength of agreement (18).
The overall proportion of patients (n=387) who had one or more rejection episode was 53.3%, with 75% of episodes being biopsy-proven. A single rejection episode occurred in 37% of the recipients with 16.3% having two or more rejection episodes.
Graft and patient survival.
Overall graft survival (n=387) was 87.9% at 1 year, 84.5% at 2 years, and 81.5% at 3 years. In total 74 (19.1%) patients lost their graft during the study period. Overall patient survival was 90.5% at 1 year, 89.1% at 2 years, and 86.6% at 3 years. In total, 54 (14%) patients died during the study period, 45 of them with functioning grafts, and 9 without. The minimum follow-up was 26 (median 42) months.
Delayed onset of graft function (DGF *) was seen in 107 (27.6%) of the recipients, and 25 (6.5%) had never-functioning kidneys. The median time on dialysis in the patients with DGF was 12 days (range 1–92). In patients not receiving dialysis (i.e., no DGF in the early posttransplant period or graft loss at long-term) the level of graft function was stable or improved as measured by creatinine clearance (Table 2). In the following analyses, with the objective to include the variance of DGF and graft loss, patients on dialysis were not excluded but evaluated as having a minimum level of creatinine clearance.
Impact of biopsy findings and donor and recipient factors on graft function and survival.
Donor age and morphological changes were separately correlated with level of graft function at all time points (from 1 week to 24 months). Moderate levels of correlation were found for donor age (r ranging from 0.19–0.27), glomerulosclerosis (r range 0.14–0.22;Table 3), and for fibrosis (0.18–0.23), although the correlation coefficients were even smaller for tubular vacuolization (0.01–0.18) and tubular desquamation (0.07–0.11). These analyses were also performed in the subset of recipients with 25 glomeruli or more in the biopsy (n=198) and in those with their graft actually biopsied (not only the contralateral graft; n=228). All correlation coefficients were identical or very similar compared with those of the entire patient population.
A series of multiple regression analyses were performed with the dependent variable being graft function estimated by 24-hr creatinine clearance and independent variables including biopsy findings and most relevant donor and recipient prognostic factors such as donor age, cold ischemia time, and retransplantation (Table 4). Donor age was the single most important factor associated with both short- and long-term graft function. Similarly, but to a lesser degree, current level of PRA and number of DR mismatches were associated with both early and late function. The length of cold ischemia was prognostic for early function only, whereas retransplantation had a significant impact on long-term function and also on graft survival. The proportion of glomerulosclerosis in the biopsy did not correlate with either short- or long-term graft function or graft survival. However, tubular changes, both vacuolization and epithelial desquamation, had a significant impact on early as well as late level of graft function (up to 6 months). To solidify the findings, the outcome for the kidneys actually biopsied as well as for those with a larger biopsy (more than 25 glomeruli) was analyzed. Patterns of results were similar in these subsets and the overall analysis, although levels of significance were reduced (Table 5).
The above findings, that the addition of donor age into multivariate analyses eliminates the impact of glomerulosclerosis, was unexpected and therefore prompted further analyses. An initial finding was that of a moderate correlation between donor age and each of the morphological variables, apart from tubular desquamation changes that are probably a consequence of ischemic damage (Table 6).
Subsequently, we combined data on donor age with morphological data on glomerulosclerosis. The incidence of DGF, graft survival and the level of graft function, short- and long-term posttransplant, were compared between recipients of kidneys with proportion of glomerular sclerosis being 0–19% vs. 20% or more (Table 7). There were no significant differences in the incidence of DGF or graft survival, however, graft function evaluated by creatinine clearance at 3 weeks and at 1 year was significantly reduced in patients receiving a kidney with 20% or more sclerotic glomeruli. It was, however, evident that these kidneys were associated with an acceptable prognosis (with glomerular filtration rate about 40 ml/min), and, in fact recipients of kidneys with 25% or more glomerulosclerosis also proved to do well (64% immediate graft function, 24% DGF, 11% never function, and graft survival 86.7% at 1 year, 82.2% at 2 years and 74.7% at 3 years; n=45).
In consequence of the impact of donor age and its association with increasing glomerulosclerosis, it would be logical to expect increased importance of histological changes in biopsies of kidneys from elderly donors. Therefore, the analysis was repeated in recipients of grafts from donors age 50 years or older (Table 7). There were no significant associations between proportion of glomerulosclerosis and graft function, its onset or level, or graft survival, in the older donor group. As a secondary finding it was evident that all parameters of graft function were inferior in the subgroups of kidneys from elderly donors.
Finally, thrombi in the glomerular capillaries, a specific sign of increased risk, were found in five donor biopsies. Kidneys from these donors were transplanted to 10 patients, 9 of whom were followed-up and one transplanted abroad in the highly immunized transplant project. The donor age ranged from 8 to 24 years and all of the donors died from cranial trauma. Prognosis was limited with never functioning grafts in three patients (33%), delayed onset of graft function in five patients (56%), and immediate function in one recipient (11%). At 2 years, four grafts were functioning, one graft had been lost, and one recipient had died with functioning graft.
Increasing donor age is a well-recognized risk factor for renal graft survival (2, 17, 19–22) and function (6–8). This may be largely the result of an increasing proportion of glomerulosclerosis in kidneys from people of advancing years (10, 13, 23). The restriction of functional nephron mass may lead to hyperfiltration and further damage may follow (13, 15, 24). The correlation with donor age is in keeping with the findings of this study, with respect to glomerulosclerosis, and also to vascular changes and interstitial fibrosis. However, age is not always associated with a decrease in renal function; in one-third of healthy subjects in a large longitudinal study it was not, and a decrease in GFR was primarily associated with cardiovascular disease (9). In our study, 36% of donors older than 50 years had kidneys with 10% or less glomerulosclerosis. However, with increasing age the frequencies of atherosclerosis, hypertension and diabetes also increase, which may accelerate renal changes (11). Therefore, donor age alone cannot be the deciding factor for acceptance of a marginal donor. It has been been suggested by Gaber et al. (17), that a donor renal biopsy should be used to distinguish the acceptable donor in marginal cases such as elderly donors with hypertension or with spontaneous intracerebral bleeding. The authors recommend that kidneys with more than 20% glomerulosclerosis and donor age of more than 50 years should not be used for transplantation. In another recent study (25), the limit for acceptance was suggested to be 15% glomerulosclerosis. Although in the latter report, kidneys with glomerulosclerosis exceeding the limit were not transplanted; the finding by Gaber et al. (17) that kidneys with 20% or more glomerulosclerosis had delayed onset of graft function in 87% supports their recommendation. In contrast, our study clearly shows that kidneys with more than 20%, or even more than 25%, glomerulosclerosis had a similar prognosis compared with those with less extensive glomerular changes. The discrepancy in results may be explained to some extent by the fact that we had 67 kidneys with 20% or more glomerulosclerosis and the material of Gaber et al. (17) only included 8 kidneys.
An unexpected finding in this study was that the proportion of glomerulosclerosis did not vary with graft function or survival in the multivariate analysis, when donor age was taken into account. Dividing all patients into groups of more or less than 20% glomerulosclerosis, there was a significant difference in graft function at 3 weeks and at 12 months, but when taking donor age again into account, only including recipients of kidneys from donors older than 50 years, there was no difference in functional prognosis. This means that knowing the age of the donor, data on glomerulosclerosis does not add any vital information. It is also evident that although the graft function in recipients of kidneys with more than 20% glomerulosclerosis was significantly reduced, it is still at an acceptable level (about 40 ml/min), and cannot be taken as a criteria for nonacceptance of such donors.
These findings may be explained by a high correlation between donor age and proportion of glomerulosclerosis. The correlation coefficient was 0.57 in our study although it was 0.90 in an autopsy study by Kaplan et al. (23). in subjects dying from causes other than renal disease. The variation in results may be due to differences in population, although the finding that the percentage of glomerulosclerosis varies depending on the cortical depth, in subcapsular, inner or outer cortex, of the biopsy or histological examination (23) provides another, possibly more plausible, explanation and suggests that comparisons of results should be limited to studies using the same sampling technique.
A further explanation for the variation in results between studies and the lack of association between glomerulosclerosis and graft survival has been suggested by Wang et al. (26) who demonstrated the necessity of obtaining 25 glomeruli in the biopsy sample to achieve a significant prediction of graft survival. Gaber et al. (17) included cases with more than 10 glomeruli and we accepted those that had more than 6 glomeruli in the study. However, we could not confirm the findings of Wang et al. (26) because in the subset of 198 patients in our study having 25 or more glomeruli in the biopsy there was no association between glomerulosclerosis and graft function or survival (Table 5) and the correlation with donor age did not increase.
Another finding in our study meriting comment is the association between tubular changes and graft function. It is logical that tubular changes, including vacuolization and epithelial desquamation, as well as prolonged cold ischemia time were significantly associated with reduction of early graft function (Table 4). An increased susceptibility due to cold ischemia damage and cyclosporine nephrotoxicity might be additive in reducing graft function. As previously demonstrated in several reports (27–29) and confirmed in this study (data not shown), delayed onset of graft function may add an impact on graft survival. In a study of postvascularization biopsies (30), no correlation between low-grade tubular lesion and DGF was seen, but in another study of procurement biopsies (31), tubular necrosis (epithelial desquamation) was found in 19 of 22 cases that developed delayed onset of graft function.
One of the most limiting factors for a study such as this one, with the objective to find the pivotal risk factors for decision making in acceptance or refusal of marginal donors, is that one can only evaluate risk factors in kidneys and recipients that were actually selected for transplantation. Although this may allow expansion of the criteria for donor acceptance, it provides no data on donors falling outside the criteria and inevitably the number of cases with unsuccessful transplants will be, and should be, too few for statistical analysis.
In summary, we did not find that the proportion of glomerulosclerosis in single procurement wedge biopsies is a major donor factor influencing graft function or survival. The study does confirm that glomerulosclerosis in this setting correlates with donor age and with the level of graft function measured by creatinine clearance in a univariate analysis. The most important factor associated with graft function in a multivariate setting was donor age. Among the morphological factors with a seemingly independent impact on onset of graft function was tubular changes. As is the case with donor age, proportion of glomerulosclerosis is a risk factor, but for neither variable there is a clear-cut level to be set as criterion of nonacceptance of a donor and the risk of excluding potentially well-functioning grafts have to be balanced. Performing a retrieval operation to evaluate potential donor kidneys and their vessels macroscopically and to obtain a biopsy, before knowing whether the organs will be acceptable or not, is a major undertaking, however, it may be called for due to the shortage of organs. From this study we conclude that in selected cases of marginal donors, in which risk factors are combined, such as old age and cardiovascular disease including hypertension, a renal procurement biopsy might provide limited additional information. The criteria for donor acceptance concerning glomerulosclerosis is difficult to determine but it should not be set too low, not below 25%, at least not in the absence of negative donor renal functional data.
1. Taylor RJ, Engelsgjerd JS. Contemporary criteria for cadaveric organ donation in renal transplantation: the need for better selection parameters. World J Urol 1996; 14:225.
2. Alexander JW, Vaughn WK. The use of “marginal” donors for organ transplantation. The influence of donor age on outcome. Transplantation 1991; 51:135.
3. Feduska NJ Jr., Cecka JM. Donor factors. Clin Transplant 1994;381.
4. Cantarovich D, Giral-Classe M, Le Sant JN, et al. Renal transplantation from cadaver donors over 60 years old. Clin Transplant 1994;237.
5. Nghiem DD, Hsia S, Carpenter BJ, et al. Renal transplantation at Allegheny General Hospital with special reference to the use of extreme-age cadaveric donors. Clin Transplant 1994;213.
6. Vianello A, Mastrosimone S, Calconi G, et al. Influence of donor age on cadaver kidney graft function and survival: univariate and multivariate analyses. Nephron 1993; 65:541.
7. Rao KV, Kasiske BL, Odlund MD, Ney AL, Andersen RC. Influence of cadaver donor age on posttransplant renal function and graft outcome. Transplantation 1990; 49:91.
8. Kumar MS, Panigrahi D, Dezii CM, et al. Long-term function and survival of elderly donor kidneys transplanted into young adults. Transplantation 1998; 65:282.
9. Rowe JW, Andres R, Tobin JD, Norris AH, Shock NW. The effect of age on creatinine clearance in men: a cross-sectional and longitudinal study. J Gerontol 1976; 31:155.
10. Nyengaard JR, Bendtsen TF. Glomerular number and size in relation to age, kidney weight, and body surface in normal man. Anat Rec 1992; 232:194.
11. Fliser D, Franek E, Ritz E. Renal function in the elderly—is the dogma of an inexorable decline of renal function correct? Nephrol Dial Transplant 1997; 12:1553.
12. Johnson LB, Kuo PC, Dafoe DC, et al. The use of bilateral adult renal allografts—a method to optimize function from donor kidneys with suboptimal nephron mass. Transplantation 1996; 61:1261.
13. Brenner BM, Milford EL. Nephron underdosing: a programmed cause of chronic renal allograft failure. Am J Kidney Dis 1993; 21:66.
14. Chertow GM, Brenner BM, Mackenzie HS, Milford EL. Non-immunologic predictors of chronic renal allograft failure: data from the United Network of Organ Sharing. Kidney Int Suppl 1995; 52:S48.
15. Isoniemi H, Nurminen M, Tikkanen MJ, et al. Risk factors predicting chronic rejection of renal allografts. Transplantation 1994; 57:68.
16. Paul LC. Chronic renal transplant loss. Kidney Int 1995; 47:1491.
17. Gaber LW, Moore LW, Alloway RR, Amiri MH, Vera SR, Gaber AO. Glomerulosclerosis as a determinant of posttransplant function of older donor renal allografts. Transplantation 1995; 60:334.
18. Brennan P, Silman A. Statistical methods for assessing observer variability in clinical measures. Br Med J 1992; 304:1491.
19. Braun WE, Popowniak KL, Nakamoto S, Gifford RW, Jr., Straffon RA. The fate of renal allografts functioning for a minimum of 20 years (level 5A)—indefinite success or beginning of the end? A proposed classification of long-term allograft survivals. Transplantation 1995; 60:784.
20. Cecka JM, Terasaki PI. Optimal use for older donor kidneys: older recipients. Transplant Proc 1995; 27:801.
21. Jacobbi LM, McBride VA, Etheredge EE, et al. The risks, benefits, and costs of expanding donor criteria. A collaborative prospective three-year study. Transplantation 1995; 60:1491.
22. Lloveras J, Arias M, Andres A, et al. Five-year follow-up of 250 recipients of cadaveric kidney allografts from donors older than 55 years of age. Transplant Proc 1995; 27:981.
23. Kaplan C, Pasternack B, Shah H, Gallo G. Age-related incidence of sclerotic glomeruli in human kidneys. Am J Pathol 1975; 80:227.
24. Terasaki PI, Koyama H, Cecka JM, Gjertson DW. The hyperfiltration hypothesis in human renal transplantation. Transplantation 1994; 57:1450.
25. Ratner LE, Kraus E, Magnuson T, Bender JS. Transplantation of kidneys from expanded criteria donors. Surgery 1996; 119:372.
26. Wang HJ, Kjellstrand CM, Cockfield SM, Solez K. On the influence of sample size on the prognostic accuracy and reproducibility of renal transplant biopsy. Nephrol Dial Transplant 1998; 13:165.
27. Troppmann C, Gillingham KJ, Gruessner RW, et al. Delayed graft function in the absence of rejection has no long-term impact. A study of cadaver kidney recipients with good graft function at 1 year after transplantation. Transplantation 1996; 61:1331.
28. Troppmann C, Gillingham KJ, Benedetti E, et al. Delayed graft function, acute rejection, and outcome after cadaver renal transplantation. A multivariate analysis. Transplantation 1995; 59:962.
29. Pfaff WW, Howard RJ, Patton PR, Adams VR, Rosen CB, Reed AI. Delayed graft function after renal transplantation. Transplantation 1998; 65:219.
30. Gaber LW, Gaber AO, Tolley EA, Hathaway DK. Prediction by postrevascularization biopsies of cadaveric kidney allografts of rejection, graft loss, and preservation nephropathy. Transplantation 1992; 53:1219.
31. Rohr MS. Renal allograft acute tubular necrosis. II. A light and electron microscopic study of biopsies taken at procurement and after revascularization. Ann Surg 1983; 197;663.