In the care of kidney transplant recipients, the renal allograft biopsy is considered the gold standard for differentiating between the various causes of renal allograft dysfunction (1, 2). However, despite its widespread use, limited information is available documenting the clinical usefulness of this procedure and rarely have prospective studies included follow-up of patients to delineate the influence of the biopsy on clinical outcome. In one prospective 44-patient study, Matas et al. found that clinical management was often changed by the biopsy results (3). However, in that study, 20% of the tissue samples were inadequate for definitive study and more than half the patients were not being treated with cyclosporine (CyA*). In addition, no long-term follow-up was provided to determine the efficacy of the postbiopsy patient management changes. Retrospective studies have similarly shown that the information provided by biopsy findings frequently results in a change in patient management (4, 5). In these studies, any addition or withdrawal of therapy was included as an alteration in patient management, but because no prebiopsy diagnosis or management plan was reported, the true clinical impact of the allograft biopsy was difficult to evaluate. Reports of the usefulness of a renal biopsy in the non-transplant setting have also been controversial; some investigators have concluded that a biopsy often results in alterations in patient management (6), whereas others suggest that the biopsy results have no clinical impact (7).
It is generally accepted that clinical signs of rejection are more easily recognized in renal allograft recipients not being treated with CyA (8). With the introduction of CyA into immunosuppressive protocols, it has become difficult to distinguish acute cellular rejection (ACR) from CyA toxicity without histopathological analysis (9). In the present study, we evaluated prospectively the clinical usefulness of the allograft biopsy in renal transplant patients receiving CyA. All patients were followed for at least 1 year to monitor the response to therapy and graft survival after a biopsy. Furthermore, in an effort to identify clinical or laboratory parameters that may have been useful in distinguishing ACR from CyA toxicity, we retrospectively analyzed the cases of biopsy-proven ACR or CyA toxicity.
During the study period between June 1994 and April 1996, a total of 82 renal allograft biopsies were performed at our institution. We prospectively analyzed the clinical usefulness of the renal allograft biopsy in 54 of these cases (47 patients, 7 patients underwent a biopsy twice). Patients received induction immunosuppression with CyA and prednisone or CyA, prednisone, and azathioprine (10). In cases of delayed graft function (DGF), patients received antithymocyte globulin and CyA was withheld until there was an improvement in renal function. The patient population consisted of 25 male and 22 female recipients, with a mean age of 37.5 years (median 38 years). At the time of the biopsy, 39 (83%) patients were receiving triple immunosuppression (CyA, azathioprine, and prednisone) and 8 (17%) patients were being maintained with CyA and prednisone. Patients had received either a cadaveric renal allograft (n=23), a living-related renal allograft (n=18), or a living-unrelated renal allograft (n=6). The original kidney diseases in these patients were diabetic nephropathy (n=10), reflux nephropathy (n=7), idiopathic focal segmental glomerular sclerosis (n=5), polycystic kidney disease (n=3), chronic glomerulonephritis of unknown origin (n=3), IgA nephropathy (n=2), membranoproliferative glomerulonephritis (n=2), Alport's syndrome (n=3), hypertensive nephrosclerosis (n=2), and others (n=10).
Before each biopsy, all clinical and laboratory data were reviewed by the medical team and an anticipated diagnosis and proposed management plan were outlined and recorded. In general, the clinical diagnosis of ACR was suspected in cases of acute allograft dysfunction with normal or sub-therapeutic trough cyclosporine levels and normal findings of renal ultrasound. Biopsies then were performed under real-time ultrasound guidance using a 15-gauge automated biopsy gun (Boston Scientific Corp., Natick, MA). Clinical indications for a biopsy included DGF (n=6), acute allograft dysfunction (n=45), and chronic allograft dysfunction (n=2). DGF was defined as oliguria (urine output <400 ml/24 hr) or requirement of hemodialysis in the first week posttransplant. Acute allograft dysfunction was defined as a rise of >0.4 mg/dl from previously stable serum creatinine levels occurring over a period of ≤1 month. Chronic allograft dysfunction was defined as a rise in serum creatinine of >0.3 mg/dl occurring over a period of ≥4 months without a subsequent return to baseline. Mean time from transplant to biopsy in the study patients was 423 days (range 3-3256, median 14). Thirty-one biopsies were performed in the 1st month posttransplant, 9 biopsies were performed in the 1st year posttransplant, and 13 biopsies were performed more than 1 year after transplantation. Renal tissue was examined using light microscopy, immunofluorescence, and electron microscopy (11, 12). Histological criteria for the diagnoses of ACR and CyA toxicity have been previously described (1, 2).
After review of the histological findings, a final diagnosis and definitive patient management plan were established. We then analyzed the incidence and significance of the changes in patient management from what had been tentatively proposed before the biopsy. To determine the impact of the postbiopsy management changes on the subsequent response to therapy and graft survival, we compared the change group (patients for whom there was a change in patient management plan after the biopsy) and the no change group (patients for whom biopsy findings confirmed original patient management plans). The time to improvement in renal function was defined as the interval from initiation of therapy (i.e., day of biopsy, in most cases) to the first evidence of a sustained decrease in serum creatinine. Allografts were considered viable until the onset of dialysis, allograft nephrectomy, or death.
To identify important clinical signs that might distinguish rejection from CyA toxicity, we analyzed the following clinical and laboratory parameters: temperature, weight gain, blood pressure (BP), CyA dose and level, interval from transplant to biopsy, serum potassium level, and rise in serum creatinine from baseline in those patients with biopsy-proven ACR (n=26) and CyA toxicity (n=12).
Proportions were compared using the Fisher's exact test for the rate of change contingency tables (13). A stepwise logistical analysis was used with forward selection and cross-validation to calculate errors in determining whether the histological diagnoses of ACR and CyA toxicity could be distinguished by the above clinical variables. A probability value of <0.05 was considered significant.
Complications. Mean drop in hematocrit observed within 24 hr after the biopsy was 0.8%±0.3%. Five patients had a drop in hematocrit >4%, and one patient (1.8%) required a transfusion in the first 24 hr after the biopsy. In an additional patient (1.8%), the allograft biopsy resulted in a large perinephric hematoma, which required surgical drainage. No macroscopic hematuria was observed in any of these patients. There were no infectious, vascular, or urological complications associated with the procedure.
Change in patient management. One biopsy (1.8%) yielded inadequate tissue (medullary tissue only) and was eliminated from the study. The biopsy findings in the remaining patient dictated a change in the previously planned patient management in 22 of 53 (41.5%) of the renal dysfunction episodes (change group). Of 22 cases who were initially anticipated to have ACR, histological findings showed 14 to have ACR, 5 to have CyA toxicity, and 3 to have other diagnoses (ACR + hemolytic uremic syndrome, ACR + CyA toxicity, hemolytic uremic syndrome). In 12 cases that were anticipated by clinical criteria to have CyA toxicity, histological findings showed 5 to have CyA toxicity and 7 to have ACR (Fig. 1). In a total of 10 cases (19%), increased or additional immunosuppression was avoided by the changes in diagnosis brought about by the biopsy. As shown in Table 1, there were no significant differences in the incidences of change in patient management among the different posttransplant period groups (38.7%, 55.6%, and 38.5% for the biopsies performed either during the 1st month, between 1 and 12 months, or beyond 1 year posttransplantation, respectively). In the acute allograft dysfunction group (n=45), there was a change in management in 44.4% of the patients. Both patients with chronic allograft dysfunction had an alteration in management. In both patients, there was a presumptive diagnosis of chronic rejection and biopsies revealed cyclosporine toxicity in one case and collapsing glomerulopathy in the other. None of the patients with DGF (n=6) had a change in management as a result of the biopsy.
Response to therapy and allograft survival. At 1 week post biopsy, 19 of 22 (86.4%) in the change group and 25 of 31 (80.6%) in the no change group demonstrated sustained improvement in renal function in response to therapy (not significant [NS]). The time to improvement in renal function was similar between the groups. Of the cases of acute allograft dysfunction (n=45), 20 were in the change group and 25 were in the no change group. Eighty-five percent (17 of 20) of the patients in the change group and 88% (22 of 25) in the no change group had a positive response to therapy in the 1st week (NS). One-year allograft survival rates were similar between the change group (79%) and the no change group (78%, NS).
As shown in Table 2, stepwise logistical analysis indicated that none of the clinical or laboratory data reviewed was correlated with the histological findings of ACR (n=26) or CyA toxicity (n=12), except for systolic BP. Systolic BP was significantly higher in the ACR group. However, if it was used for differentiating ACR from CyA toxicity, it would have an error rate of 36%, indicating that BP is not a reliable parameter to distinguish ACR from CyA toxicity. Furthermore, when patients with ACR and CyA toxicity in the first month posttransplant were compared, there was no significant difference in BP between the two groups.
Currently, the renal allograft biopsy is frequently recommended to diagnose the exact etiology of allograft dysfunction. In the past, some studies attempted to quantify the clinical usefulness of this procedure versus its associated risks and costs. The conclusions from these reports are somewhat limited in that: (1) rarely were the studies prospective; (2) rarely was there follow-up documented to determine whether the postbiopsy treatment was appropriate; (3) the studies were performed mostly before the introduction of CyA into immunosuppressive protocols; (4) many studies were done with more primitive biopsy techniques, which increased the risks of the procedure and limited the quality of tissue obtained.
In the current study, we prospectively investigated the clinical usefulness of the renal allograft biopsy in kidney transplant recipients receiving CyA-based immunosuppression. We analyzed the incidence of change in patient management plans that resulted from the histopathological findings. The patients were then followed to determine the response to therapy and to monitor allograft survival. We found that biopsy findings altered patient management in >40% of patients. It is interesting that this finding is similar to that reported in another recent prospective study in which an incorrect diagnosis was predicted by the clinicians in 26% of the patients (14). In our study, the change in patient management was bi-directional. That is, some patients anticipated to have ACR had biopsy findings consistent with CyA toxicity, whereas some patients with presumed CyA toxicity based on clinical findings, proved to have ACR by histopathological criteria. In 19% of cases, unnecessary immunosuppression was avoided by changes in diagnosis brought about by the biopsy results. In contrast to a recent report that questioned the utility of the renal allograft biopsy in the late transplant period (15), we found that patient management was changed frequently by a biopsy in all posttransplant periods. In the DGF group, no patient had a change in the anticipated patient management after a biopsy. This is a reflection of our protocol for DGF in which antithymocyte globulin is administered and CyA is withheld until adequate allograft function is achieved. Nevertheless, the renal allograft biopsy could still be useful in these circumstances to exclude causes of DGF other than ischemic injury, for example, acute humoral rejection, now a rare clinical event but one which may respond to plasma exchange therapy and tacrolimus-mycophenolate rescue (16).
Responses to therapy and allograft survival were similar between the change group and the no change group, suggesting that the changes in patient management brought about by biopsy results were appropriate. It can be hypothesized that without a biopsy, >40% of the patients initially would have received inappropriate treatment (i.e., either inadequate or excessive immunosuppression) and that the delayed initiation of treatment might have negatively affected their response to therapy and ultimate allograft survival. Admittedly, the definitive determination of the clinical usefulness of renal allograft biopsies would require a study that randomly assigned patients with allograft dysfunction to receive either biopsy-directed treatment or treatment without a biopsy. It is unlikely that this type of study could be justified, because the renal allograft biopsy is generally accepted as a clinically relevant and safe procedure.
Our analysis of the diagnostic utility of various clinical and laboratory parameters in distinguishing ACR from CyA toxicity confirmed the relative lack of specificity of all data usually available at the time of a biopsy. To further analyze the utility of these parameters, a larger study would be needed. It should be emphasized that the current use of various new immunosuppressive agents such as tacrolimus, mycophenolate mofetil, rapamycin, and antilymphocyte monoclonal antibodies, may modify the clinical and pathologic presentation of renal allograft dysfunction.
Future studies that better define the molecular mechanisms underlying acute allograft rejection may provide even more sensitive and specific diagnostic tools for clinicians. For example, the recent demonstration that intragraft expression of granzyme B and perforin genes, two molecules that participate in cytotoxic T lymphocyte mediated cell lysis, is closely associated with acute renal allograft rejection suggests that serial monitoring of such expression could be a sensitive measure of immunologic reactivity (17). Intrarenal expression of other markers such as TH1 and/or TH2 cytokine mRNA may also prove valuable in the future (18). Obviously, to offer a diagnostic advantage over conventional histology, the gene expression of rejection markers within the graft would have to correlate with their expression in peripheral blood mononuclear cells, in which case studying these molecular markers could be monitored by noninvasive means. For the foreseable future, however, the renal allograft biopsy examined by standard techniques will likely remain an essential tool in the management of renal transplant recipients.
In summary, the renal allograft biopsy frequently alters patient management recommendations made on the basis of clinical and laboratory findings only. Results of approximately 40% of the biopsies in this study resulted in a change in patient management plan, and 19% allowed for an avoidance of additional immunosuppression. Whether or not patients had a change in management resulting from a biopsy did not affect their subsequent response to therapy or allograft survival.
1. Colvin RB. Pathology of renal allografts. In: Colvin RB, Bhan AK, McCluskey RT, eds. Diagnostic immunopathology. New York: Raven Press, 1995: 329.
2. Nephrology Forum. The renal allograft biopsy. Principle discussant: Robert B. Colvin. Kidney Int 1996; 50: 1069.
3. Matas AJ, Tellis A, Sablay L, et al. The value of needle renal allograft biopsy. III. A prospective study. Surgery 1985; 98: 922.
4. Kiss D, Landmann J, Mihatsch M, et al. Risks and benefits of graft biopsy in renal transplantation under cyclosporine-A. Clin Nephrol 1992; 38: 132.
5. Matas AJ, Sibley R, Mauer M, et al. The value of needle renal allograft biopsy. I. A retrospective study of biopsies performed during putative rejection episodes. Ann Surg 1983; 197: 226.
6. Richards NT, Darby S, Howie AJ, et al. Knowledge of renal histology alters patient management in over 40% of cases. Nephrol Dial Transplant 1994; 9: 1255.
7. Nephrology Forum. Renal biopsy, principle discussant: Michael P. Madaio. Kidney Int 1990; 38: 529.
8. Kao KV. Diagnosis of renal transplant rejection in the cyclosporine era. Curr Opin Nephrol Hypertens 1992; 1: 280.
9. Kolbeck PC, Scheinman JI, Sanfilippo F. Acute cellular rejection and cyclosporine nephrotoxicity monitored in biopsy in a renal allograft recipient. Arch Pathol Lab Med 1986; 110: 389.
10. Delmonico FL, Tolkoff-Rubin N, Auchincloss H Jr, et al. Management of the renal allograft recipient: immunosuppressive protocols for long-term success. Clin Transplant 1994; 8: 34.
11. Collins AB. Immunoflourescence. In: Colvin RB, Bhan AK, McCluskey RT, eds. Diagnostic immunopathology. New York: Raven Press, 1995: 699.
12. Schneeberger EE, Collins AB, Latta H, McCluskey RT. Diminished glomerular accumulation of colloidal carbon in autologous immune complex nephritis. Lab Invest 1977; 37: 9.
13. Mehta CR, Patel NR. A network algorithm for the exact treatment of Fisher's exact test in RxC contingency tables. J Am Stat Assoc 1983; 78: 427.
14. Al-Awwa IA, Hariharan S, First MR. Importance of allograft biopsy in renal transplant recipients: correlation between clinical and histological diagnosis. Am J Kidney Dis 1998; 31 (suppl 1): S15
15. Kon SP, Templar J, Dodd SM, et al. Diagnostic contribution of renal allograft biopsies at various intervals after transplantation. Transplantation 1997; 63: 547.
16. Pascual M, Saidman S, Tolkoff-Rubin N, et al. Plasma exchange and tacrolimus-mycophenolate rescue for acute humoral rejection in kidney transplantation. Transplantation 1998; 66: 1460.
17. Pavlakis M, Lipman M, Strom TB. Intragraft expression of T-cell activation genes in human renal allograft rejection. Kidney Int 1996; 49 (suppl 53): S7.
18. Suthanthiran M. Clinical application of molecular biology: a study of allograft rejection with polymerase chain reaction. Am J Med Sci 1997; 313: 264.
*Abbreviations used: ACR, acute cellular rejection; BP, blood pressure; CyA, cyclosporine; DGF, delayed graft function; NS, not significant.