Acute allograft rejection is a cause of early transplant dysfunction and is associated with late kidney graft failure (1). To prevent irreversible graft damage, it is important to diagnose and treat acute rejection in its earliest phase. In current clinical practice, serum creatinine measurement, urine volume assessment, and needle biopsy of the graft are standard techniques for the diagnosis of acute rejection (2,3). The recognition of rejection at an early stage continues to pose a problem. Needle biopsy, the most reliable method, is an invasive technique and cannot be performed daily. However, noninvasive methods have not been able to replace it because they are less reliable. Posttransplantation monitoring of neopterin and other immunologic mediators, such as soluble interleukin 2 receptor and soluble tumor necrosis factor receptor, are limited in their use because urinary retention or dialysis significantly affects serum levels of these proteins (4).
CD30, a member of the tumor necrosis factor-nerve growth factor receptor family, is a large 120-kDa transmembrane glycoprotein that was originally identified on the surface of Hodgkin’s and Reed Sternberg cells (5). After activation of CD30+ T cells, a soluble form of CD30 (sCD30) is released into the bloodstream (6). It has been reported that ligation of CD30 by the natural ligand (CD30L) can induce apoptotic and proliferative signals in CD30+ lymphocytes (7). Martinez et al. (8) found that primary alloantigenic stimulation results in the generation of CD30+ CD4+ and CD30+ CD8+ T-cell subsets. We reported recently that sCD30 measured before transplantation could be used for the prediction of kidney graft outcome: increased pretransplantation sCD30 was significantly associated with graft loss (9,10). The aim of the present study was to determine whether monitoring of plasma sCD30 during the early posttransplantation period would differentiate patients with an impending acute allograft rejection from patients with an uncomplicated course or with acute tubular necrosis (ATN).
Between January 1995 and June 2001, more than 350 patients received a kidney graft at the Heidelberg transplant center. For the present study, we selected all recipients with available plasma samples on day 0 before transplantation and on days 3 to 5, 7 to 9, 12 to 14, and 17 to 19 after transplantation, who received cyclosporine-based immunosuppression without the use of antithymocyte globulin or OKT3, and who did not experience viral, bacterial, or fungal infections until posttransplantation day 20. Fifty-six recipients who fulfilled these criteria were studied. The patients were separated into three groups. Twenty patients with primary graft function and an uncomplicated course without acute rejection were categorized as group UC. Eleven patients with primary nonfunction showing a prolonged phase of ATN without evidence of acute rejection were categorized as group ATN. Twenty-five patients who experienced an episode of biopsy-proven acute rejection during the first 20 days were categorized as group R.
In all three groups, glomerulonephritis and cystic kidney disease were the most frequent causes of end-stage renal failure. The mean number of HLA-A+B+DR mismatches was 2±0 in groups UC and R and 2±1 in group ATN. The median day for diagnosis of acute rejection in group R patients was day 9. Interstitial rejections were diagnosed in 13 patients, vascular rejections in 3 patients, and signs of vascular and interstitial rejection were observed in the remaining 9 patients. Banff ’97 classification of biopsy specimens was available for 14 patients: 5 were graded as borderline, 2 as grade IA, 6 as grade IIA, and 1 as grade IIB. Antirejection therapy with methylprednisolone pulses was initiated on day 9 (median) and terminated on day 11 (median). None of the patients lost the graft. Group ATN patients received intermittent hemodialysis treatment until the onset of graft function (median day: 15). Hemodialysis treatment was performed using polysulfone membranes (Fresenius, Bad Homburg, Germany). Further demographic details are summarized in Table 1.
Plasma was snap-frozen within 2 hr after the blood was drawn and stored at −30°C until testing. sCD30 plasma levels were measured using a commercial ELISA kit obtained from BenderMedSystems (Vienna, Austria). Wilcoxon test, Mann-Whitney U test, Fisher exact test, and receiver operating characteristic (ROC) curves (11) were used for statistical analysis. Sequential changes in plasma sCD30 levels are expressed as relative values measured on days 3 to 5 (median day for all groups: 4), 7 to 9 (median day for all groups: 8), 12 to 14 (median day for all groups: 13), and 17 to 19 (median day for groups UC and R: 18; for group ATN: 19), compared with the baseline value on day 0, which was set at 100%.
Except for the time period between day 0 before transplantation and days 3 to 5 after transplantation, sCD30 plasma levels decreased continuously in all three patient groups. Importantly, a significant decrease in sCD30 from day 0 to days 3 to 5 was observed only in group UC and group ATN patients (group UC: from 100% to 54±4%, P <0.0001; group ATN: from 100% to 76±3%, P =0.003). In contrast, plasma sCD30 levels remained high in the 25 group R patients during the first 3 to 5 days after transplantation despite immunosuppression (100% vs. 109±8%, P =NS) (Table 2). The increased sCD30 values in group R patients might have been caused by the presence of 7 regraft recipients and 10 recipients with primary nonfunction and the necessity of repeated hemodialysis until graft function. However, when these 17 patients were excluded and the analysis of group R patients was restricted to first graft recipients with primary graft function, nearly identical sCD30 values were obtained at the respective sampling time points (compared with all group R patients, days 3 to 5: 114±16% vs. 109±8%; days 7 to 9: 89±25% vs. 76±13%; days 12 to 14: 36±7% vs. 49±12%; and days 17 to 19: 37±9% vs. 37±5%;P =NS for all comparisons). At the sampling time points of days 3 to 5, 7 to 9, and 12 to 14, group R and group ATN patients exhibited a significantly higher sCD30 than patients in group UC (R vs. UC:P <0.0001, P <0.0001, and P =0.001, respectively; ATN vs. UC:P =0.001, P =0.025, and P =0.009, respectively). The difference between group R and group ATN patients was statistically significant on days 3 to 5 and 7 to 9 (P =0.001 and P =0.041, respectively).
Importantly, on days 3 to 5, an increased sCD30 was observed only in group R patients (14 of 25), whereas without exception all group ATN and UC patients showed a decrease of sCD30 (Fisher exact test, R vs. ATN:P =0.002; R vs. UC:P <0.0001). This indicated that a nondecreasing sCD30 level during the first 3 to 5 posttransplantation days is a predictor of impending graft rejection. The individual relative sCD30 values obtained on posttransplantation days 3 to 5 in groups UC, ATN, and R are shown in Figure 1. Antirejection therapy with methylprednisolone caused a 29% decrease of plasma sCD30 in group R (P =0.004). No rebound effect was observed in subsequent samples.
For monitoring of sCD30 to be clinical meaningful, it is important to define operational cut-off (ROC) levels. ROC curves are useful for demonstrating the reciprocal association of sensitivity and specificity at all possible cut-off levels for a given test. The sampling time point that allows a differentiation between two patient groups with the highest specificity and sensitivity was determined by calculating the area under the ROC curve (AUC). In line with the results shown in Table 2, the sampling time point at days 3 to 5 was found to be the earliest and also the optimal time point, with the highest area under the ROC curve for differentiating group R from group UC (specificity: 100%; sensitivity: 88%;P <0.0001) and from group ATN (specificity: 91%; sensitivity: 72%;P =0.001) and for differentiating group ATN from group UC (specificity: 75%; sensitivity: 73%;P =0.001) (Table 3).
We reported recently that increased sCD30 serum levels before transplantation are associated with graft loss (9,10). In the current study, we demonstrate that measurement of plasma sCD30 during the early posttransplantation period allows the identification of recipients with impending acute rejection. Conventional rejection diagnosis (e.g., monitoring of serum creatinine or biopsy of the graft) does not recognize the immune process at an early cellular response stage but rather assesses progressive inflammatory destruction of the graft with first signs of functional failure. sCD30, in contrast, allowed the identification of group R patients as early as 3 to 5 days after transplantation, 5 days (median) before an acute rejection was diagnosed by conventional methods. Importantly, at this early time point, sCD30 allowed a differentiation of group R from group ATN patients. Conventional noninvasive rejection parameters, such as serum creatinine or neopterin, are not able to differentiate between these two patient groups, and it is necessary to perform fine-needle or core biopsies (4). Needle biopsies cannot be performed daily and, despite technical refinements, biopsy-associated complications, such as sampling errors, hematuria, bleeding, arteriovenous fistulas, and even graft loss can occur (12,13).
However, before it can be claimed that sCD30 measurement is able to replace invasive biopsy techniques, further investigations are needed involving larger numbers of patients, preferably in a multicenter study. The present single-center study provides an indication that sCD30 is a useful immunologic marker of increased alloreactivity, not only before (9,10) but also after transplantation.
The authors thank Martina Stolp, Nicole Schach, and Roland Seidel for excellent technical assistance, and Bernd Döhler and Gunter Laux for assistance with the computer analysis.
1. Opelz G. Collaborative Transplant Study: 10-year report. Transplant Proc 1992; 24: 2342.
2. 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.
3. Pascual M, Vallhonrat H, Cosimi AB, et al. The clinical usefulness of the renal allograft biopsy in the cyclosporine era: a prospective study. Transplantation 1999; 67: 737.
4. Rebhandl W, Felberbauer FX, Resch S, et al. Value of circulating immune parameters in renal transplantation. Transplant Proc 1997; 29: 1881.
5. Durkop, Latza U, Hummel M, et al. Molecular cloning and expression of a new member of the nerve growth factor receptor family that is characteristic for Hodgkin’s disease. Cell 1992; 68: 421.
6. Romagnani S, Del Prete G, Maggi E, et al. CD30 and type 2 T helper (Th2) responses. J Leukoc Biol 1995; 57: 726.
7. Muta H, Boise LH, Fang L, et al. CD30 signals integrate expression of cytotoxic effector molecules, lymphocyte trafficking signals, and signals for proliferation and apoptosis. J Immunol 2000; 165: 5105.
8. Martinez OM, Villanueva J, Abtahi S, et al. CD30 expression identifies a functional alloreactive human T-lymphocyte subset. Transplantation 1998; 65: 1240.
9. Pelzl S, Opelz G, Wiesel M, et al. Soluble CD30 as a predictor of kidney graft outcome. Transplantation 2002; 73: 3.
10. Süsal C, Pelzl S, Döhler B, et al. Identification of highly responsive kidney transplant recipients using pretransplant soluble CD30. J Am Soc Nephrol 2002; 13: 1650.
11. Zweig MH, Campbell G. Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clin Chem 1993; 39: 561.
12. Huraib S, Goldberg H, Katz A, et al. Percutaneous needle biopsy of the transplanted kidney: technique and complications. Am J Kidney Dis 1989; 14: 13.
13. Beckingham IJ, Nicholson ML, Bell PR. Analysis of factors associated with complications following renal transplant needle core biopsy. Br J Urol 1994; 73: 13.