The 1-year graft survival rates for cadaveric renal transplants have improved gradually over the last 20 years, and this success has been attributed in general to improvements in the management of acute rejection and in particular to the introduction of cyclosporine therapy (1). Despite this, analysis of the slopes of cadaveric kidney survival curves to yield graft half-lives shows that over the same period, the long-term results of renal transplantation have remained virtually unchanged (2). The data from our own unit, which show a 1-year graft survival rate of over 85% but 5- and 10-year graft survival rates of only 70% and 50% respectively, are typical (our unpublished data). This gradual attrition is largely due to the process variously named as chronic rejection, chronic allograft nephropathy, or chronic allograft damage (3). The histopathology of chronic renal transplant rejection is dominated by neointimal hyperplasia, interstitial fibrosis, and glomerulosclerosis (4), all processes that involve an excess accumulation of extracellular matrix proteins (5, 6). Although the pathogenesis is complex, multifactorial, and not completely understood, it is widely accepted that chronic allograft nephropathy represents a single pathway of tissue remodeling that converges from a variety of insults: immunological, ischemic, toxic, and others. This pathway includes cytokine-mediated infiltration of macrophages; induction of actin-positive myofibroblasts; synthesis of glycoproteins such as tenascin; and finally, the accumulation of collagen (7-9).
Although a number of drugs may be of potential benefit in ameliorating or even preventing the development of chronic rejection in renal allografts, randomized clinical trials are difficult to perform because they require large numbers of patients to be followed up over considerable periods of time. The mathematical difficulties posed by using graft survival as an end point have been well summarized by Hunsicker and Bennett (10). To take our own data as an example, to demonstrate a 10% improvement in graft survival rate at 5 years with a power of 90% and an alpha of 0.05, 473 patients would need to be entered into each limb of the study and followed for 5 years. Clearly, if we could develop ad interim surrogate end points of chronic rejection, then this would greatly facilitate human studies of therapeutic interventions. The logical place to look for surrogate end points is renal transplant biopsy material (11).
The aims of the present study were to develop and validate a technique for the measurement of immunostainable tissue components, which could be applied to routine diagnostic practice, and to assess whether such an approach could provide prognostic information on long-term graft outcome when applied to protocol renal transplant biopsy specimens.
A consecutive series of 52 patients who had received a cadaveric renal allograft during the period from 1989 to 1991 was studied. Patients' details are shown in Table 1. In all cases, a preperfusion biopsy specimen was obtained from the donor kidney, and transplant recipients subsequently had protocol needle core biopsies performed under ultrasound control at 1, 3, and 6 months after transplantation (12). In this study, only the 6-month biopsy specimens were considered further. Renal function was analyzed by measurement of glomerular filtration rate (GFR*) using a single-shot chromium 51 EDTA isotope clearance technique. These GFR measurements were made at 1, 6, 12, and 24 months after transplantation.
Patients were immunosuppressed with Sandimmune cyclosporine at 17 mg/kg/day given in twice daily divided doses, and this was reduced to a baseline level of 7 mg/kg/day by 6 weeks after transplantation. Steroids were given orally as prednisolone at a starting dose of 100 mg/day reducing to 40 mg by the first week and 10 mg on alternate days by 6 months after transplantation. Some patients received azathioprine at a dose of 1-2 mg/kg/day, adjusted according to the white cell count.
Delayed allograft function was defined as the need for dialysis in the first 7 days after transplantation except for a single dialysis in the immediate postoperative period performed specifically to treat a high serum potassium level. Acute allograft rejection was diagnosed clinically as the need for antirejection treatment with high-dose steroids, OKT3, or antithymocyte globulin but was confirmed in all cases by needle core biopsy.
Immunostaining. Paraffin wax sections (3 μm) cut from renal transplant biopsy specimens were dewaxed and treated with 6% H2O2 to block endogenous peroxidase. Sections were exposed to 0.1% trypsin (Difco, Detroit, MI) in 0.12% CaCl2, pH 7.8, for 25 min at 37°C. A standard indirect immunoperoxidase method was used with primary antibodies to type 3 collagen (Europath), alpha smooth muscle actin (Europath), macrophage CD68 (Dako, Glostrup, Denmark), and tenascin (Europath) applied overnight at 4°C. A counter-stain was omitted to facilitate subsequent image analysis.
Image analysis. Sections were viewed on a Zeiss photomicroscope III with a JVC TK 120E video camera attached. The camera was linked with a Y/C cable to the in-built framegrabber board of an Apple Macintosh 7100/80 AV microcomputer, and images were imported directly to the freeware image analysis program NIH-Image using the "Plug-in digitizer" software Photoshop-compatible plug-in. Sequential grey-scale images of renal cortex were grabbed using the ×10 objective, moving along the central line of each biopsy specimen from one end of the available cortex to the other, without overlapping. The presence of glomeruli in the fields was ignored.
To calculate the area fraction of a particular immunostained component, a threshold was applied to each image at a constant level that distinguished between the stained component (rendered black) and the unstained background (rendered white). The proportion of black to white pixels in the image was then calculated as a percentage. This represents the percentage volume fraction of the tissue that is occupied by the stained element (13). A set of macros was written to automate this process, such that sequential fields could be measured with just two keystrokes per field. These macros are available as a simple text file, by E-mail on request to [email protected]. All measurements were performed by the same person (E.B.), who was blinded to the graft outcome data.
The normal ranges of values for the immunostainable components being measured were determined using sections of normal renal tissue from nephrectomy specimens of patients with small renal cell carcinomas, excluding cases with histological evidence of reflux nephropathy or tumor.
Reproducibility studies. To test the reproducibility of the histomorphometric system, serial sections obtained from one biopsy specimen were immunostained for collagen 3 and analyzed in the following ways
- One section was analyzed 10 times;
- Ten sections were stained together and analyzed on the same day;
- Ten sections were stained on separate days but analyzed on the same day.
Statistical analysis. The results of quantitative immunohistochemistry were correlated with patient outcome data at 6, 12, and 24 months after transplantation. Pearson regression coefficients and their associated P values were calculated using Instat 1.12. The levels of immunostained components in biopsy specimens from different patient groups were compared using the Mann-Whitney U test.
Clinical outcome. Initial allograft function was recorded in 45 patients (87%) and the remaining 7 patients had delayed graft function. Twenty patients suffered at least one episode of acute rejection, with three of these having second rejections and one patient having three rejection episodes. Thirty-four patients were treated for hypertension after transplantation and 18 were normotensive.
Reproducibility studies. The results of analyzing serial sections from one biopsy specimen are shown in Figure 1. There was little variation in mean percentage area fraction of collagen III whether one section was analyzed 10 times, 10 sections were stained and analyzed on the same day, or 10 sections were stained on separate days but analyzed on the same day. The greatest variation seemed to arise between the three groups, which were each measured on separate days. This probably results from variation in the lighting of the microscope, which was not standardized. This problem was eliminated in subsequent studies by measuring all sections for a single comparison in one session.
Correlation with renal function. Percentage area fraction of collagen III immunostaining correlated with 6-month GFR (r=−0.42, P=0.005; Fig. 2). In addition, the area fraction of collagen III in 6-month biopsy specimens was predictive of 12-month GFR (r=−0.32, P=0.03; Fig. 3). Staining for alpha smooth muscle actin, tenascin, and macrophages did not correlate with renal functional parameters at any of the time points studied.
Collagen III staining in renal transplant biopsy specimens taken 6 months after transplantation predicted long-term graft function (Fig. 4). Patients with a collagen III of ≤40% had relatively stable GFR over a period of 24 months of follow-up. In contrast, patients with a collagen III of >40% had a significant decrement in renal function at 6 months after transplantation, which did not improve over the next 18 months. At the 24-month follow-up period, a percentage area of collagen III of >40% in the 6-month biopsy specimens was associated with a significantly lower GFR, compared with a percentage are of ≤40% (30.8±3.6 vs. 44.8±3.6 ml/min/1.73 m2, P=0.01). In normal kidneys, the percentage area fraction of collagen III immunostaining was 15±3%.
Effect of donor age and other factors. Although there was no relationship between donor age and collagen III levels in preperfusion biopsy specimens, there was a positive correlation between collagen III immunostaining in the 6-month protocol biopsy specimens and donor age (r=0.33, P=0.03; Fig. 5). Patients who suffered at least one acute rejection had a higher level of collagen III in their 6-month biopsy specimen than those in the no rejection group (median [95% confidence interval] =42 [35-44] vs. 33 [31-37], U=149, P=0.054; Fig. 6). Delayed graft function and HLA matching had no effect on the 6-month graft levels of collagen III. There was no correlation between collagen III immunostaining and cold ischemic time, anastomotic time, recipient weight, posttransplantation hypertension, or cyclosporine dosage.
This study suggests that the histomorphometric measurement of immunostained collagen III in renal biopsy specimens taken 6 months after transplantation is a useful surrogate ad interim marker of chronic allograft damage. The demonstration of an inverse correlation between collagen III measured at 6 months after transplantation and renal function measured in the longer term is likely to be particularly useful because this time point may be early enough to allow a change in therapy in an attempt to halt or even reverse the fibrotic process. Hayry's group (14) have shown that a histological chronic allograft damage index correlates with long-term graft function, but their method has shortcomings. It involves a rather complex analysis of graft histology, taking account of interstitial inflammation and fibrosis, glomerular mesangial matrix increase and sclerosis, vascular intimal thickening, and tubular atrophy. Moreover, the changes of chronic allograft damage were assessed in 2-year biopsy specimens, and this is probably too late to allow therapeutic interventions to be effective in preventing subsequent deterioration of allograft function. A correlation between histopathological determinants in biopsy specimens taken 6 months after transplantation and diminished renal function and proteinuria at 2 years was demonstrated by Nickerson et al. (15).
It might have been expected that early events in chronic allograft nephropathy, such as macrophage infiltration, might provide the earliest measure of the level of chronic damage. Indeed, a correlation has been reported between inflammatory cell infiltration in early protocol biopsy specimens and function at 1 year (16). In native kidneys, the number of actin-positive myofibroblasts have been reported to correlate with prognosis in IgA nephropathy (17), diabetes (18), and membranous nephropathy (19). The lack of any correlation between graft function and macrophage infiltration, actin, or tenascin expression in the present study was therefore disappointing. Nonetheless, a correlation with the last stage of chronic damage, collagen deposition, was demonstrated. This could be explained if the level of immunological, ischemic, and toxic insults varies considerably in the early months, resulting in random variation in "early" markers, which becomes smoothed in the later expression of collagen. If so, it may well be that repeated biopsies to measure changes in expression, rather than absolute levels, would be more informative. This would also reduce random variation due to sampling error, because the development of chronic allograft nephropathy is not entirely uniform throughout the kidney.
A number of recent studies suggest that some of the newer immunosuppressive agents, such as FK506 (20), mycophenolate mofetil, and rapamycin (21, 22), may have the potential to prevent or at least ameliorate the changes of chronic allograft damage. These studies have largely been based on retrospective data or have involved animal models, and this serves to highlight the difficulty of studying therapeutic strategies in human renal transplants. Prospective randomized trials need large numbers of patients and can only really be performed on a multicenter basis, making them difficult to organize and expensive to run. Measurement of immunostainable tissue components such as collagen III can be used as a surrogate end point in clinical studies and may go some way to obviate the need for the study of large numbers of patients. As an example, a controlled trial of the effects of two different immunosuppressive regimens powered to demonstrate a 10% difference in renal allograft fibrosis at 6 months after transplantation would only require 23 patients to be randomized to each treatment arm. Clearly, most renal transplantation units would have the patient numbers to achieve a result from such a study in a relatively short period of time, and this kind of analysis should prove to be a useful tool in the rapid assessment of new treatments.
The positive correlation between donor age and collagen III staining in 6-month biopsy specimens raises the possibility that the predictive changes at this time point reflect donor changes rather than those arising de novo. However, the lack of a correlation between donor age and collagen III in preperfusion biopsy specimens shows that the effect is not present at the time of transplantation but develops later, suggesting that older kidneys are more susceptible to fibrogenic influences. Even if donor changes are important, the usefulness of the method described is not diminished.
Although the development of chronic allograft rejection is a complex multifactorial process, acute rejection is thought to be the leading risk factor (23). In support of this, patients with at least one rejection episode were, in general, found to have increased levels of collagen III in their 6-month biopsy specimens. There was, however, a considerable degree of overlap between the rejection and nonrejection groups. Other factors are likely to be important despite the lack of any correlation between collagen III and potentially deleterious factors such as delayed graft function, prolonged ischemic time, and poor DR matching in this analysis.
The level of macrophage infiltration has been found to be a prognostic marker in various human and experimental models of disease (24, 25), but we found no relationship between leukocyte infiltration and outcome in the renal transplant patients analyzed here. In the 6-month biopsy tissue studied, levels of leukocyte infiltration were quite low and showed little variation between kidneys with high and low levels of collagen III deposition. Tenascin immunoreactivity was also studied because this glycoprotein seems to be produced when matrix remodeling is active and is known to increase smooth muscle cell proliferation (26). Although staining for tenascin showed considerable variation in the material under study, it did not correlate with renal function at any time and does not seem to be a useful marker of chronic allograft rejection.
Immunohistochemistry has proven to be a popular method in the investigation of the molecular cascades involved in the development of chronic transplant rejection. The vast majority of studies have used semiquantitative scoring systems by grading the level of immunostaining on an arbitrary scale from 0 to 3. Our own previous study of the relationship between renal allograft interstitial fibrosis and long-term renal function involved the use of a laborious point-counting technique (27). Although this latter technique is equally predictive of renal function, the use of a semiautomated image analysis method has a number of clear advantages-being rapid, easy to perform, inexpensive, and reproducible.
This study involved a relatively small number of patients, and it was not possible to perform a meaningful analysis of the relationship between immunostained renal function tissue components and allograft survival. Nonetheless, the results demonstrate that poor long-term graft function can be predicted using relatively early (6 month) protocol biopsy specimens immunostained for collagen III.
1. Thorogood J, Van Houwelingen HC, Van Rood JJ, Zantvoort FA, Schreuder GMTh, Persijn GG. Long term results of kidney transplantation in Eurotransplant. In: Paul LC, Solez K, eds. Organ transplantation: long term results. New York: Dekker, 1992: 33.
2. Opelz G, Mickey MR, Terasaki PI. Calculations of long term graft and patient survival in human kidney transplantation. Transplant Proc 1997; 9: 27.
3. Paul LC, et al. Diagnostic criteria for chronic rejection/accelerated graft atherosclerosis in heart and kidney transplants: proposals from the fourth Alexis Carrel Conference on Chronic Rejection and Accelerated Arteriosclerosis in Transplanted Organs. Transplant Proc 1993; 25: 2020.
4. Kasiske BL, Kalil RSN, Lee HS, Rao KV. Histopathological findings associated with chronic, progressive decline in renal allograft function. Kidney Int 1991; 80: 514.
5. Furness PN. Extracellular matrix and the kidney. J Clin Pathol 1996; 49: 355.
6. El Nahas AM. Pathways to renal fibrosis. Exp Nephrol 1995; 3: 71.
7. Paul LC. Chronic renal transplant loss. Kidney Int 1995; 47: 1491.
8. Lemstrom K, Koskinen P, Hayry P. Molecular mechanisms of chronic renal allograft rejection. Kidney Int 1995; 48: 2.
9. Hayry P, Yilmaz S. Chronic allograft rejection: an update. Transplant Proc 1994; 26: 3159.
10. Hunsicker LG, Bennett LE. Design of trials of methods to reduce late allograft loss: the price of success. Kidney Int 1995; 48 (suppl 52): S120.
11. Dimeny E, Wahlberg J, Larsson E, Fellstrom B. Can histopathological findings in early renal allograft biopsies identify patients at risk for chronic vascular rejection? Clin Transplant 1995; 9: 79.
12. Nicholson ML, Attard AR, Bell A, Donnelly PK, Veitch PS, Bell PRF. Renal transplant biopsy using real time ultrasound guidance. Br J Urol 1990; 65: 564.
13. Furness PN. The use of digital images in pathology. J Pathol 1997; 183: 252.
14. Isoniemi H, Taskinen E, Hayry P. Histological chronic allograft damage index accurately predicts chronic renal allograft rejection. Transplantation 1994; 58: 1195.
15. Nickerson P, Jeffrey J, McKenna R, Gough J, Rush D. Do renal allograft function and histology at 6 months post-transplant predict graft function at 2 years? Transplant Proc 1997; 29: 2589.
16. Rush DN, Jeffrey JR, Gough J. Sequential protocol biopsies in renal transplant patients: clinicopathological correlates using the Banff schema. Transplantation 1995; 59: 511.
17. Goumenos DS, Brown CB, Shortland J, el Nahas AM. Myofibroblasts, predictors of progression of mesangial IgA nephropathy? Nephrol Dial Transplant 1994; 9: 1418.
18. Pedagogos E, Hewitson T, Fraser I, Nicholls K, Becker G. Myofibroblasts and arteriolar sclerosis in human diabetic nephropathy. Am J Kidney Dis 1997; 29: 912.
19. Roberts I, Burrows C, Shanks J, Venning M, McWilliam L. Interstitial myofibroblasts: predictors of progression in membranous nephropathy. J Clin Pathol 1997; 50: 123.
20. Gjertson DW, Cecka MJ, Terasaki PI. The relative effects of FK506 and cyclosporine on short and long term kidney graft survival. Transplantation 1995; 60: 1384.
21. Gregory CR, Huang X, Pratt RE, et al. Treatment with rapamycin and mycophenolic acid reduces arterial intimal thickening produced by mechanical injury and allows endothelial replacement. Transplantation 1995; 59: 655.
22. Azuma H, Binder J, Heemann U, Schmid C, Tullius SG, Tilney NL. Effects of RS61443 on functional and morphological changes in chronically rejecting rat kidney allografts. Transplantation 1995; 59: 460.
23. Basadonna GP, Matas AJ, Gillingham KJ, et al. Early versus late acute renal allograft rejection: impact on chronic rejection. Transplantation 1993; 55: 993.
24. Eddy AA. Interstitial macrophages as mediators of renal fibrosis. Exp Nephrol 1995; 3: 76.
25. Furness PN, Rogers-Wheatley L, Harris KPG. Semiautomatic quantitation of macrophages in human renal biopsy specimens in proteinuric states. J Clin Pathol 1997; 50: 118.
26. Wallace WA, Howie SE, Lamb D, Salter DM. Tenascin immunoreactivity in cryptogenic fibrosing alveolitis. J Pathol 1995; 175: 415.
27. Nicholson ML, McCulloch TA, Harper S, et al. Early measurement of interstitial fibrosis predicts long-term renal function and graft survival in renal transplantation. Br J Surg 1996; 83: 1082.
* Abbreviations: GFR, glomerular filtration rate;