Wolfgang Altermann, Gerald Schlaf, Anita Rothhoff and Barbara Seliger

Transcription

1 Nephrol Dial Transplant (27) 22: doi:1.193/ndt/gfm397 Advance Access publication 5 July 27 Short Communication High variation of individual soluble serum CD3 levels of pre-transplantation patients: scd3 a feasible marker for prediction of kidney allograft rejection? Wolfgang Altermann, Gerald Schlaf, Anita Rothhoff and Barbara Seliger Martin Luther University Halle-Wittenberg, Institute of Medical Immunology, Interbranch HLA-Laboratory, Medical School, Halle, Germany Abstract Background. Previous studies have suggested that the pre-transplant levels of the soluble CD3 molecule (scd3) represent a non-invasive tool which can be used as a biomarker for the prediction of kidney allograft rejections. Methods. In order to evaluate the feasibility of scd3 for pre-transplantation monitoring the sera of potential kidney recipients (n ¼ 652) were collected four times in a 3 months interval. Serum from healthy blood donors (n ¼ 23) served as controls. The scd3 concentrations of all samples were determined using a commercially available ELISA. This strategy allowed the detection of possible variations of individual scd3 levels over time. Results. Heterogeneous scd3 concentrations were found in the samples obtained from individual putative kidney transplant recipients when quarterly measured over 1 year. Total 95% of serum samples obtained from healthy controls exhibited scd3 values <3 U/ml, whereas most recipients displayed higher serum levels (>3 U/ml). Total 524 patients (8.4%) constantly exhibited serum concentrations of <1 U/ml during the period investigated, whereas 19 patients (16.7%) showed variations by exceeding the proposed cut off of 1 U/ml for one to three times. The frequency of samples exhibiting scd3 values >1 U/ml was significantly lower than that previously reported. Conclusions. The high degree of variation does not allow the stratification of patients into high and low immunological risk groups based on a single scd3 value > 1 U/ml. Due to the heterogeneity of scd3 levels during time course and the high values of SD, its Correspondence and offprint requests to: Wolfgang W. Altermann, Gerald Schlaf, Martin Luther University Halle-Wittenberg, Institute of Medical Immunology, Magdeburger Strasse 2, 6112 Halle, Germany buero.ghatt@medizin.uni-halle.de; Gerald Schlaf, gerald.schlaf@medizin.uni-halle.de implementation as a pre-transplant marker cannot be justified to generate special provisions for the organ allocation to patients with single scd3 values >1 U/ml. Keywords: graft survival; haemodialysis; kidney transplantation; soluble CD3 Introduction The identification of pre- and post-transplant parameters of recipients bearing an increased risk of allograft rejection is an important prerequisite for the successful implementation of individually tailored immunosuppression of transplant patients. Although the treatment of transplant patients with immunosuppressive therapies has significantly increased the graft survival, a high dose immunosuppression should be avoided at least for low-risk recipients due to the massive side effects such as cancer development, infection and toxicity. The determination of panel reactive antibodies (PRA), which at present are exclusively used as indicators for an increased risk of graft rejection, is currently being critically discussed. For this reason novel markers are urgently needed for proper monitoring of pre- and post-transplant risks. The CD3 molecule, a 12 kda transmembrane molecule and member of the tumour necrosis factor receptor (TNF-R) superfamily, was originally identified as a surface antigen on the Hodgkin Reed Sternberg (HRS) cells [1 3]. Increased serum levels of the soluble 85 kda CD3 protein (scd3) in Hodgkin disease (HD) patients were further correlated with their clinical outcomes [3]. In most healthy individuals scd3 serum concentrations are only detectable at low levels. CD3, which is expressed on CD4 þ and CD8 þ T lymphocytes, B cells, natural killer (NK) cells and some cells of non-lymphoid origin, appears rather to be expressed on Th2 cytokine-producing Downloaded from at Pennsylvania State University on March 3, 214 ß The Author [27]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org

2 2796 W. Altermann et al. T lymphocytes than on Th1 cells. In addition to HD, increased serum scd3 levels have been frequently associated with diseases exhibiting Th2 dominantdriven immune responses including viral infections, lupus erythematosus and atopic dermatitis. The expressions of CD3 and CD3L as well as the potential role of CD3/CD3L interactions in humoral immunity have been extensively reviewed by Kennedy et al. [4]. Recently, there has been some evidence that high pre-transplant serum levels of scd3 indicate the risk of an impaired graft outcome of kidney transplants [5 8]. Thus, up-regulated scd3 levels were shown to be indicative for an increased risk of transplant loss emphasizing their clinical relevance and the implementation of scd3 as a predictive biomarker for allograft rejection upon transplantation of different solid organs. In all studies published so far scd3 levels as a prediction marker have only been determined once per patient prior to transplantation, whereas the pre-transplantation monitoring of scd3 levels in putative kidney recipients over time has not yet been performed. We here postulate that an additional monitoring of individual changes of scd3 levels in organ recipients over time is important for the validation and the clinical impact of this serum marker. Since individual scd3 levels showed a significant heterogeneity over time, in particular with respect to the cut off level of 1 U/ml which is currently used for distinguishing between high and low risk patients [5 8], we propose a quarterly monitoring of individual scd3 values. Patients and methods The study was initially started with 737 patients representing the complete waiting lists of the four kidney and pancreas/ kidney transplantation centres located in the Deutsche Stiftung fu r Organtransplantation (DSO) Region East and completed with 652 recipients. This reduced patient number was caused by grafting, through the deregistrations of patients for unknown reasons by the transplant centres and by their deaths during the study time. The patient samples were quarterly collected for 1 year. Serum samples from 23 healthy blood donors, kindly provided by the Blood Donation Service of the Martin Luther University in Halle (Germany), served as controls. For the analysis of scd3 levels the commercially available scd3 ELISA (Biotest AG, Dreieich, Germany) was used. Briefly, this ELISA is based on the sandwichtechnique using two different anti-scd3 monoclonal antibodies (mabs) for capturing and detecting the antigen. The kit includes the immobilized capture Ab and antigen for the standard curve pre-dropped in duplicate by the manufacturer which comprises the range between 1.6 and 1 U/ml. The day-to-day coefficient of variation was 3.9% for high (5 U/ml) and 11.4% for low (3.2 U/ml) scd3 concentrations, thus demonstrating the assay to be reliable and reproducible. To exclude an influence of the different lot numbers, standard curves for each lot were performed which showed no significant differences. Serum samples yielding optical density (OD) values higher than 2. were repeated at a higher dilution to reach OD values covered by the standard curve. All measurements were at least performed in duplicate at different days to exclude errors due to the individual handling of one experiment and to demonstrate the reproducibility of the values. The data calculations were performed using the WINSTAT program. Results Serum samples from putative recipient patients were quarterly monitored over a period of 1 year. As shown in Figure 1, most of the recipients exhibit higher scd3 serum levels (3 9 U/ml) than healthy blood donors ( 3 U/ml) which is in accordance with nearly all previous studies investigating this aspect [e.g. 9,1]. The results of the mean values (MV) at the different time points suggest no seasonal dependency of scd3 concentrations, and the high standard deviations (SD) of the MV (quarter 1: U/ml; quarter 2: U/ml; quarter 3: U/ml; quarter 4: U/ml) are in accordance with most previous studies. Although a cut off level of 1 U/ml had been proposed as a risk predictor [6 8], the frequency of scd3 serum levels >1 U/ml was considerably lower in our study when compared with the results published by Suesal and co-workers [6] (Figure 2). These lower levels were not the result of a degradation process of serum scd3 since 5-fold freezing and thawing of six serum samples with scd3 concentrations ranging from 1 to 11 U/ml and their consecutive storing for at least 2 h at room temperature or at 28C, respectively, did not lead to any alterations in the scd3 serum values (data not shown). The novel approach of the present study was the monitoring of the patient distribution with scd3 serum levels higher or lower than 1 U/ml N Blood donors Quarter 1 Quarter 2 Quarter 3 Quarter scd3 serum levels (U/ml) Fig. 1. Distinct scd3 serum levels of patients from the kidney transplantation waiting list and of healthy individuals. Soluble CD3 levels of serum samples from 652 recipient patients were quarterly obtained and analysed using a commercially available scd3 ELISA. Serum samples from 23 healthy blood donors served as controls. The diagram shows the numbers of patients exhibiting scd3 concentrations (U/ml) ranging from to higher than 5 U/ml > 5 Downloaded from at Pennsylvania State University on March 3, 214

3 High variation of individual soluble serum CD3 levels of pre-transplantation patients U/ml, >1 ml A 5 Quarter 1, Quarter 2, Quarter 3, Quarter Per cent (%) 6 4 scd3 (U/ml) during the quarterly measurements. Based on the individual scd3 levels during the time course, 524 patients (8.4%) exhibited constant serum concentrations of <1 U/ml during the period investigated, whereas 19 of 652 recipients (2.9%) showed scd3 values higher than the proposed cut off level of 1 U/ml in all four measurements (Figure 3A). In contrast, 19 patients (16.7%) demonstrated variations by exceeding 1 U/ml scd3 one to three times. Randomly chosen representative results of patients exhibiting scd3 levels higher than 1 U/ml only once (Figure 3B) and patients exceeding this value for two or three times are shown in Figure 3C, respectively. The implementation of 3 U/ml as threshold value to distinguish between healthy persons and haemodialyis patients described previously [9,1] was confirmed by our results (Figure 1). From the 652 putative recipients only 11 patients (1.7%) showed values lower than 3 U/ml in all four serum samples, whereas 158 patients (24.2%) demonstrated scd3 serum levels lower than 3 U/ml for one to three times. It is noteworthy that only two of 652 recipients investigated (.3%) exhibited an extraordinary fluctuation of the scd3 serum levels during the study ranging from levels lower than 3 U/ml to levels significantly higher than the proposed cut off level of 1 U/ml (data not shown). Discussion Quarter 1 Quarter 2 Quarter 3 Quarter 4 Süsal (22) Fig. 2. Distribution of scd3 serum levels representing low and high risk recipients according to the proposed cut off level of 1 U/ml. The data are presented as percentage (frequency) of scd3 values 1 U/ml (white column/low risk recipients) and higher than 1 U/ml (black column/high risk recipients) during four quarterly measurements. The scd3 levels of a pre-transplant study investigating 3899 cadaver kidney recipients [Suesal et al. (6)] are shown for comparison. It has recently been suggested that the serum scd3 level may be a potential marker for the prediction of acute allograft rejection of kidneys. The determination of scd3 serum concentrations might therefore offer a promising non-invasive tool to recognize patients with an increased risk for developing an acute B scd3 (U/ml) C scd3 (U/ml) bo ps th si ki khj me mkj bdi br nn ww hh ak mn mw bs bk bg bja bt po lp sr Fig. 3. High variability of the individual scd3 levels in patients during four independent quarterly determinations. Soluble CD3 concentrations of eight randomly chosen patients (A) exhibiting scd3 levels higher than 1 U/ml during all four measurements, (B) exceeding once the proposed cut off level of 1 U/ml or (C) exceeding the proposed cut off level of 1 U/ml two to three times. graft rejection. Based on published results pre-transplant scd3 serum levels higher than 1 U/ml have been classified as a risk factor for the survival of kidney allografts [5 8,1]. In our present study we show (i) that the frequency of kidney pre-transplant patients exhibiting scd3 levels which exceed the proposed cut off of 1 U/ml is lower than that previously reported [6] and (ii) that there exists a high degree of timedependent variations in the individual scd3 levels Downloaded from at Pennsylvania State University on March 3, 214

4 2798 W. Altermann et al. temporarily exceeding the proposed upper 1 U/mlthreshold level (16.7% of the patients) as well as temporarily falling below the lower 3 U/ml-threshold level (24.2% of the patients) distinguishing between haemodialysis patients and healthy individuals. To the best of our knowledge, so far in all studies published pre-transplant scd3 levels have only been determined once per individual. Thus, there exists no information regarding possible time-dependent variations of individual scd3 levels during the course of disease. This aspect is of particular importance since in Middle Europe the median waiting time for the reception of a kidney graft is about 4 years, although many kidney recipients exhibit waiting times from 5 to 1 years. During this period the dialysis patients can develop additional diseases based on the continued intoxication due to the lack of proper kidney functions [11 14]. However, whether the high incidence of variation might be associated with clinical data of patients including intoxication, infections or other physiological factors has still to be investigated. The unexpectedly high individual time-dependent variations favour the concept that mainly individual immunological events, rather than the genetic disposition of a patient, as discussed previously, [6] may lead to the observed individual increases and decreases of serum scd3. Although a certain statistical impact for the implementation of serum scd3 as pre-transplant rejection marker had been suggested, the predictive value was highly narrowed by Rajakariar et al. [15]. These authors described an increased scd3 level only during a rejection episode of the vascular type, whereas the dominating tubulointerstitial type showed decreased scd3 levels even lower than those of healthy volunteers. Two other studies [16,17] in which no significant differences between the pretransplant serum scd3 levels of rejecting and nonrejecting patients were observable demonstrated the lack of the predictive pre-transplant value of this molecule. These two studies together with others [16 19] rather point to an importance of scd3 posttransplant monitoring for the prediction of allograft rejection/survival. Indeed, a substantial difference in the post-transplant scd3 levels between rejecting and non-rejecting patients has been described in the early post-transplantation period [16 19]. Patients with rejections had significantly higher scd3 values at post-transplant days 5 [17,19] to days 14 or 15 [1,16,18] but with different threshold values proposed by the authors which allow the distinction between rejecting and non-rejecting kidney recipients. Thus, an integration of the individualized evaluation of posttransplant scd3 serum level as one biomarker, together with accompanying diseases which affect the immunological reactivity post-transplantation, may be a feasible approach for the non-invasive post-transplant prediction of acute kidney allograft rejection. In conclusion, our data do not support the transplantation of only HLA well-matched organs to recipients displaying pre-transplant scd3 levels higher than 1 U/ml only once as proposed previously [6 8], but rather argue for a quarterly monitoring of scd3 concentration prior to transplantation. Acknowledgement. We would like to thank L. Pellenen for excellent secretary help and the Blood Donation Service of the Martin-Luther- University Halle for providing us with serum samples from healthy volunteers. Conflict of interest statement. None declared. References 1. Schwab U, Stein H, Gerdes J et al. Production of monoclonal antibody specific for Hodgkin and Sternberg-Reed cells of Hodgkin s disease and a subset of normal lymphoid cells. Nature 1982; 299: Pizzolo G, Vinante F, Chilosi M et al. Serum levels of soluble CD3 molecule (Ki-1 antigen) in Hodgkin s disease. Relationship with disease activity and clinical stage. Br J Haemtol 199; 75: Nadali G, Vinante F, Ambrosetti A et al. Serum levels of soluble CD3 are elevated in the majority of untreated patients with Hodgkin s disease and correlate with clinical features and prognosis. J Clin Oncol 1994; 12: Kennedy MK, Willis CR, Armitage RJ. Deciphering CD3 ligand biology and its role in humoral immunity. Immunology 26; 118: Pelzl S, Opelz G, Wiesel M et al. Soluble CD3 as a predictor of kidney graft outcome. Transplantation 22; 73: Suesal C, Pelzl S, Doehler B, Opelz G. Identification of highly responsive kidney transplant recipients using pretransplant soluble CD3. J Am Soc Nephrol 22; 13: Suesal C, Pelzl S, Opelz G. Strong human leukocyte antigen matching effect in nonsensitized kidney recipients with high pretransplant soluble CD3. Transplantation 23; 76: Cinti P, Pretagostini R, Arpino A et al. Evaluation of pretransplant immunologic status in kidney-transplant recipients by panel reactive antibody and soluble CD3 determinations. Transplantation 25; 79: Barbano G, Cappa F, Prigione I et al. Plasma levels of soluble CD3 are increased in children with chronic renal failure and with primary growth deficiency and decrease during treatment with recombinant human growth hormone. Nephrol Dial Transplant 21; 16: Sengul S, Keven K, Gormez U, Kutlay S, Erturk S, Erbay B. Identification of patients at risk of acute rejection by pretransplantation and posttransplantation monitoring of soluble CD3 levels in kidney transplantation. Transplantation 26; 81: Nakao K, Nagake Y, Okamoto A, Ichikawa H, Yamamura M, Makino H. Serum levels of soluble CD26 and CD3 in patients on hemodialysis. Nephron 22; 91: Descamps-Latscha B, Herbelin A. Long-term dialysis and cellular immunity: a critical survey. Kidney Int 1993; 43: Descamps-Latscha B, Chatenoud L. T cells and B cells in chronic renal failure. Semin Nephrol 1996; 16: Taskapan H, Utas C, Oymak FS, Guelmez I, Oezesmi M. The outcome of tuberculosis in patients on chronic hemodialysis. Clin Nephrol 2; 54: Rajakariar R, Jivanji N, Varagunam L et al. High pre-transplant soluble CD3 levels are predictive of the grade of rejection. Am J Transpl 25; 5: Downloaded from at Pennsylvania State University on March 3, 214

5 High variation of individual soluble serum CD3 levels of pre-transplantation patients Slavcev A, Lacha J, Honsova E et al. Soluble CD3 and HLA antibodies as potential risk factors for kidney transplant rejection. Transpl Immunol 25; 14: Dong W, Shunliang Y, Weizhen W et al. Prediction of acute renal allograft rejection in early post-transplantation period by soluble CD3. Transpl Immunol 26; 16: Roy R, Dugre FJ, Berube L, Langlois S, Lachance JG. CD3 and TNF-alpha serum level as indicators of ongoing acute kidney graft rejection. Hum Immunol 22; 63: Pelzl S, Opelz G, Daniel V, Wiesel M, Suesal C. Evaluation of posttransplantation soluble CD3 for diagnosis of acute renal allograft rejection. Transplantation 23; 75: Received for publication: Accepted in revised form: Downloaded from at Pennsylvania State University on March 3, 214