Annals of Oncology 2: 579-583, 1991. C 1991 Klutver Academic Publishers. Printed in the Netherlands.

Original article The clinical significance of serum CD8 antigen levels in adult patients with Hodgkin's disease

Medizinische Klinik I, University of Saarland and Medizinische Klinik I, University of Cologne, Fed. Rep. Germany

Summary. Increased suppressor T-cell activity has been observed in patients with Hodgkin's disease. In order to evaluate the clinical significance of soluble CD8 antigen (sCD8), which is released from CD8 + suppressor/ cytotoxic T-lymphocytes, we determined sCD8 levels in the sera of 82 consecutive patients with newly diagnosed untreated Hodgkin's lymphoma who were entered into prospective trials of the German Hodgkin's Disease Study Group. sCD8 levels were significantly higher (p < 0.01) in stage IV (781 U/ml, n = 19) than in stages I-IIIB (443 U/ml; n - 63). Patients with B-symptoms (n - 36) had slightly higher levels (611 U/ml) than patients without (n — 46) systemic symptoms (447 U/ml; p - 0.08). In 77 patients evaluable for response, the complete remission (CR) rate of patients with sCD8 < 750 U/ml was higher (54/60 or 90%) than that of patients with sCD8 > 750 U/ml 11/17 or 65%; p - 0.01). The time to treatment failure was significantly longer in patients with sCD8 < 750 U/ml (p - 0.008), even among the group with stages 111B/TV only (p = 0.04). Our data suggest that the pretreatment levels of sCD8 in adult patients with Hodgkin's lymphoma have prognostic relevance, and that they should be determined especially in patients with advanced disease. Increased understanding of the role of sCD8 may shed light on the pathogenesis of Hodgkin's disease. Key words: CD8, Hodgkin's disease, prognostic factors Introduction

Patients and methods

Increased suppressor T-cell activity is a well known phenomenon observed in patients with Hodgkin's lymphoma [1-3]. Suppressor T cells can be readily identified by monoclonal antibodies directed against the CD8 cell surface antigen [4, 5]. Fujimoto and coworkers [6, 7] characterized a soluble form of the CD8 antigen (sCD8) which is found in the supernatants of human CD8+ T-cell leukemia cell lines and in the sera from patients with CD8+ T-cell leukemia. sCD8 was found to be a specific, physiologic product related to the membrane-associated CD8 molecule. Increased serum levels of sCD8 have been reported to be associated with a poor treatment outcome in children with acute lymphoblastic leukemia and non-Hodgkin's lymphoma [8] as well as in children with Hodgkin's disease [9]. We therefore studied the serum levels of CD8 antigen in a large cohort of adult patients with newly diagnosed Hodgkin's lymphoma who were treated with a uniform stage-directed approach. The results of our study suggest that high sCD8 levels in the pretreatment sera of adult patients have prognostic value, especially in patients with advanced stage.


Between January 1987 and March 1988, 147 untreated patients were registered for the multicenter therapeutic trials of the German Hodgkin's Study Group (Chairman: Prof. V. Diehl). Before the start of treatment, frozen serum samples were taken from 82 (56%) of these patients and used to determine sCD8 levels. Their ages ranged from 18 to 60 years (median 30). The stages according to the Ann Arbor classification were: IA: 10; IB: 0; HA: 17; IIB: 7; IIIA 15; I1IB 14; IVA 4; IVB 15. Staging was based on findings in the physical examination, diagnostic imaging (chest x-ray, abdominal sonograms and CT scans in all, and lymphangiogram in some patients), bone marrow and liver biopsy. Staging laparotomy with splenectomy was mandatory in those cases where clinical staging revealed clinical stage (CS) I or II without risk factors. Risk factors were defined as: massive mediastinal mass (> 1/3 of the maximal thoracic diameter), extranodal disease, and/or massive splenic involvement. Thirty-two patients (39%) were staged with laparotomy (PS IA: 4, PS IIA: 9, PS IIB:3, PS IHA: 10, PS HTB: 1, PS IVA:2, PS IVB: 3). Patients in PS I and II without risk factors (n = 14)

Downloaded from by university of winnipeg user on 18 January 2019

A. Gause, K. Verpoort, V. Roschansky, A. Tschiersch, D. Hasenclever, R. Schmits, V. Diehl, O. Brosteanu & M. Pfreundschuh


Spearman's rank correlation coefficient. The response to therapy was evaluated by a restaging 4-8 weeks after the end of therapy (6-12 months after the start of treatment with the respective protocol). For restaging, all primary disease manifestations were checked by appropriate diagnostic procedures. The complete remission rate was defined as the ratio of all patients in complete remission to all evaluable patients. The influence of potentially significant prognostic factors on freedom from treatment failure (FFTF) was estimated with the Cox regression model. FFTF curves were constructed by the Kaplan-Meier method. FFTF was defined as the time interval from the start of treatment to the first of the following events: death from any cause, progressive disease, failure to achieve complete remission (CR) at the end of therapy or relapse. A sCD8 level of 750 U/ml was chosen as the dividing point based on the maximum ratio of the estimated hazard for the groups.



The sera were stored at -70°C until use. They were Pretreatment sera coded and tested blind. The sera of 25 healthy persons with the same median age and age range as the patients The summary of the results is shown in Table 1 and served as controls. Fig. 1. sCD8 serum levels in 82 adult patients with newly diagnosed untreated Hodgkin's lymphoma Soluble CD8 assay ranged from 120 to 1640 U/ml. The median level in all patients with Hodgkin's disease was not significantly sCD8 antigen levels were measured with a two-site higher than that in normal controls (521 U/ml ±54 sandwich-enzyme immunoassay (CELLFREE T8 TM s.e.m. vs. 300 ±42, p - 0.08). The median sCD8 level test kit, T Cell Science, Cambridge MA) according to in patients who presented with B-symptoms was 611 the manufacturer's description (Pui et al., 1989). Brief- U/ml compared to 447 U/ml in patients without ly, sCD8 in the test samples or standards was bound to B-symptoms (p - 0.08). Patients in stage IV had signifithe wells of a polysterene 96-well microtiter plate; 100 cantly higher sCD8 levels than patients in stages I-IH ul of a 1 : 10 diluted serum sample were added into (781 U/ml vs. 443 U/ml; p - 0.005). Surgically staged each well and incubated at 37°C for 90 min. Then a (PS) patients (after splenectomy) had lower sCD8 horseradish peroxidase-conjugated anti-CD8 mono- levels than clinically staged (CS) patients (562 vs. 465 clonal antibody directed against a second epitope on the CD8 molecule was added. The reaction was developed with O-phenylenediamine, and the degree of sub- Table 1. sCD8 levels in the sera of patients with Hodgkin's lymphoma and healthy controls. strate conversion was determined at 492 nm using a Titertek ELISA reader (Flow Laboratories). sCD8 n sCD8 U/ml (± s.e jn.) concentrations were expressed in units/ml, with 1,000 300±32 25 units/ml being defined as the concentration of sCD8 Normal controls p - 0.08 present in a reference preparation provided by the sup- All patients 82 521±49 plier. With this method, 25 normal healthy adults with Stages I—III 63 the same median age and age range as the patients were 443±42 n _ f\ fiH U.I




370±38 430±70 620± 130

B Histology LP


Age (years)

Downloaded from by university of winnipeg user on 18 January 2019

were treated with extended-field radiotherapy (40 Gy). Patients in CS/PS ILIA and CS/PS I and II with one or more risk factors (n •= 35) were treated according to the HD1 protocol (Pfreundschuh et al., 1988) and received 2 double cycles of COPP + ABVD followed by extended-field radiotherapy (20 Gy vs. 40 Gy). Patients in CS/PS HEB/TV (n - 33) were treated according to the HD-3 protocol [10] and received 3 double cycles of COPP + ABVD, and in instances of complete remission (CR) were randomized into either another double cycle of COPP + ABVD or 20 Gy involved-field radiotherapy; patients with non-CR received involved-field radiotherapy for persisting nodal disease, or the CEVD-protocol [11] or autologous bone marrow transplantation for persisting systemic disease. Informed consent was obtained from all patients, and the study had been approved by the clinical trials review committee of the Medical Association of North Rhine Westphalia.




Correlation with other laboratory parameters • COS s 750

sCD8 was strongly correlated with the levels of soluble interleukin-2 receptors (sCD25; p - 0.001) and moderately with serum alkaline phosphatase (p = 0.01) and erythrocyte sedimentation rate (p - 0.02). There was no significant correlation (p > 0.1) with albumin, LDH, total leukocyte and lymphocytes counts, nor with sex, age (50 years) or histological subtype (lymphocyte predominant/nodular sclerosis vs. mixed cellularity/lymphocyte depleted).


n -

_ Nocrtpati - Ewnti - M*d tknaobt I I i i | i

17 7 16

60 8 23, no ii


• COS > 750





i i 1 < i i 1 i

' ' ' 1 ' ' ' ' 1 ' ' ' ' 1 ' ' '' 50 BO

i i i i 1 i


Fig. 2. Freedom from treatment failure (FFTF) according to sCD8 levels of all 77 patients evaluable for response. FFTF is significantly better (p - 0.008) in patients with low (750 U/ml)sCD8 levels.

Response to therapy The complete response rate was 84% for all patients; 13% suffered from progressive disease (Tab. 2). Of 60 patients with sCD8 < 750 U/ml, 54 (90%) achieved complete remission (CR). This is significantly higher i 0.04

Table 2. Results of therapy in 77 patients treated for Hodgkin's lymphoma according to sCD8 levels. All

sCD8 750 U/m]

60 54 (90%)' 1 5(8%)

17 11 (65%)" 1 5(29%)

No of p«tt


0. 10

Evaluable for response Complete remissions Partial remissions Progressive disease • p 750 11 7




Fig. 3. Freedom from treatment failure (FFTF) in 32 patients with stages IIIB/IV according to sCD8 levels. FFTF is significantly better (p-0.04) in patients with low ( 750 U/ml) sCD8 levels.

Downloaded from by university of winnipeg user on 18 January 2019

(p - 0.01) than for patients with sCD8 > 750 U/ml (11/17 or 65%). When looking at patients in the advanced stages IIIB/rV only (all of whom were treated according to the same protocol), the difference in CR rate remained significant (p - 0.05), despite the low number of patients (750 U/ml: 5/11 or 45%. After a median observation time of 23 and 16 months, respectively, the FFTF curve of patients with sCD8 < 750 U/ml was significantly better (p = 0.008) than the curve of patients with sCD8 > 750 U/ml (Fig. 2). The difference in FFTF for the 32 patients in stages UIB/TV was also significant (p = 0.04; Fig. 3). I* tU MB MIA III! IVA m In a multivariate analysis without an interaction term n • 10 n - 17 n - 7 n - II n • 14 n - 4 n - 13 (stage x sCD8), only stage YEB/TV retained its sigFig. 1. sCD8 levels (mean±s.ejn) in the pretreatment sera of 82 nificance (p = 0.01) as an independent parameter for newly diagnosed adult patients with Hodgkin's lymphoma. The the prediction of treatment failure. sCD8 was not siglevels of surgically staged patients (PS) did not differ from the levels nificant (p > 0.1) nor were any of the other factors conof clinically staged patients (CS) in the respective stage. sidered for this model such as age (50 years), sex, B symptoms, histology (lymphocyte U/ml); however, this was due to the low proportion of predominant and nodular sclerosis vs. mixed cellularity spelenectomized patients in stage IV; when analyzed according to stage, there was no difference between CS and PS patients. Likewise, the higher sCD8 values in 1 0older patients were due to the higher incidence of advanced stages in these patients. In the healthy controls, 6 there was no difference in sCD8 levels between the p • 0001 three age groups (< 30 vs. 30-50 vs. > 50 years). 5


and lymphocyte depleted), soluble interleukin-2 receptor levels (1000 U/ml), and erythrocyte sedimentation rate (80 mm/h). A Cox regression model with an interaction term or restricted to IIIB/rV did not converge due to the low number of patients in this subgroup. Folluw-up sera


While soluble tumor-associated antigens are widely used as serum markers for solid tumors, serum tumor markers play only a limited role in hematological malignancies. To date, the erythrocyte sedimentation rate (ESR), which is rather nonspecific, is the only widely used marker for prognosis and disease activity in Hodgkin's lymphoma [12-14]. Soluble CD30 antigen [15,16] seems to be more specific, but is detectable in only about one-third of patients with Hodgkin's disease. In this study, higher sCD8 levels were associated with advanced stage and related to lower CR rates and shorter times to treatment failure. The latter was also true for the subgroup of patients in stages IIIB/IV who were all treated with the same protocol. However, in a multivariate analysis, sCD8 did not retain its impact as an independent prognostic factor due to the low number of patients in this subgroup. Nevertheless, as patients in advanced stages IIIB/IV who had sCD8 levels >750 U/ml had a significantly lower CR rate and a shorter time to treatment failure, the determination of Acknowledgements sCD8 may help in identifying patients in stages 1LLB/TV who are at a particularly high risk of treatment failure, Supported by a BMFT grant 01 GA 8815/7. We thank and for whom intensified treatment programs might be the participants of the German Hodgkin Study Group considered. for supplying us with sera from patients with Hodgkin's In a similar study of children with Hodgkin's lym- disease. phoma [9], high sCD8 levels evolved as an independent prognostic factor for a poor treatment outcome. This may be due to different treatment strategies in the two References patient populations studied or because suppressor cell 1. Engleman EG, Benike C, Hoppe RT et al. Suppressor cells of activity contributes more to the clinical outcome in mixed lymphocyte reaction in patients with Hodgkin's disease. children than in adult patients in whom the response to Transplantation Proceedings 1979; 11: 1827-9. cytotoxic therapy seems to be mainly determined by 2. Hillinger SM, Herzig GP. Impaired cell-mediated immunity in Hodgkin's disease mediated by suppressor lymphocytes and the tumor mass and only to a lesser degree by the level monocytes. Journal of Clinical Investigation 1978; 61:1620-7. of immunosuppression. 3. Twomey JJ, Laughter AH, Rice et al. Spectrum of immunoWhile being predicitive to treatment outcome, sCD8 deficiencies with Hodgkin's disease. Journal of Clinical Investilevels did not correlate well with disease activity: sCD8 gation 1980; 66:629-37.

Downloaded from by university of winnipeg user on 18 January 2019

For the evaluation of sCD8 as a marker for disease activity we studied the available sera of 52 patients taken after the end of therapy. In nearly all patients, sCD8 levels decreased after therapy; however, this was the case both in patients who achieved complete remission and in those who suffered from progressive disease. So far, a complete set of follow-up sera until relapse is available from only four patients. Therefore, the utility of sCD8 as a marker for disease activity can not yet be defined.

levels decreased after therapy even in patients who suffered from progressive disease, and a few patients in CR had higher sCD8 levels than before therapy. This may be due to the fact that sCD8 is released by reactive lymphocytes and not by the neoplastic Hodgkin's and Reed-Sternberg cells. Therefore, sCD8 levels are influenced by many non-specific or intercurrent events that are not related to the activity of the malignant disease. However, the real value of sCD8 as a marker for disease activity or an early indicator of relapsing disease can only be defined by follow-up of sCD8 levels during long-term remission and/or relapse in a larger number of patients. The functional role of sCD8 patients with Hodgkin's lymphoma is still unknown. As sCD8 levels have been shown to parallel the number of activated CD8+ cells in the peripheral blood (CD8+/HLA-DR+ [17], the high sCD8 levels in patients with advanced Hodgkin's lymphoma could reflect the presence of increased suppressor/cytotoxic activity. The strong correlation with soluble interleukin-2 receptor (sCD25 antigen) levels suggests that both molecules might be released from the same cell population. In this respect the data presented here correspond well to our recent analysis of soluble interleukin-2 receptors (sCD25) in the same patient population [18]: low sCD25 levels correlated with a good prognosis and FFTF was 100% in patients with sCD25 < 1000 U/ml. Thus, by assaying for pretreatment sCD8 and sCD25, two extremely different populations of patients with Hodgkin's disease can be identified: one with an excellent (low sCD25) and one with a very poor prognosis (high sCD8 levels in stages HIB/ IV). If our results, obtained under the specific conditions given (i.e., the treatment strategy of the German Hodgkin's Disease Study Group), can be confirmed in an independent set of patients, possibly treated with another therapeutic approach, a less intensive treatment regimen for patients with low sCD25 and a more aggressive approach for patients with high sCD8 should be considered.

583 13. Loffler M, Pfreundschuh M, Hasenclever D et al. Prognostic risk factors in advanced Hodgkin's lymphoma. Blut 1988; 56: 949-55. 14. Tubiana M, Henry-Amar M, van der Werf-Messing et al. A multivariate analysis of prognostic factors in early stage Hodgkin's disease. International Journal of Radiation Oncology Biology and Physics 1985; 11: 23-9. 15. Pfreundschuh M, Pohl C, Berenbeck C et al. Detection of a soluble form of the CD30 antigen in sera of patients with lymphoma, adult T-cell leukemia and infectious mononucleosis. Int J Cancer 1990; 45:869-74. 16. Pizzolo G, Vinante F, Chilosi M et al. Serum levels of soluble CD30 molecule (Ki-1 antigen) in Hodgkin's disease: relationship with disease activity and clinical stage. Brit J Haematol 1990; 75: 282-4. 17. Tomkinson B, Brown MC, Ip SH et al. Soluble CD8 during T cell activation. J Immunol 1989; 142: 2230-5. 18. Gause A, Roschansky V, Tschiersch A et al. Low serum interleukin-2 receptor levels correlate with a good prognosis in patients with Hodgkin's lymphoma. Ann Oncol 2 (suppl. 2) 1991:43-7. Received 6 February 1991; accepted 22 May 1991.

Correspondence to: Michael Pfreundschuh, M.D. Medizinische Klinik I Universitat des Saarlandes D-6650 Homburg Fed. Rep. Germany

Downloaded from by university of winnipeg user on 18 January 2019

4. Reinherz EL, Kung PC, Goldstein G et al. A monoclonal antibody reactive with the human cytotoxic/suppressor T-cell subset previously defined by a heteroantiserum termed TH2. Journal of Immunology 1980; 124: 1301-7. 5. Schlossman SF, Reinherz EL. Human T-cell subsets in health and disease. Springer Seminars in Inununopathology 1980; 7: 9-18. 6. Fujimoto J, Levy S, Levy R. Spontaneous release of the Leu-2 (T8) molecule from human T cells. J Exp Med 1983; 159: 752-66. 7. Fujimoto J, Stewart SJ, Levy R. Immunochemical analysis of the released Leu-2 (T8) molecules. J Exp Med 1984; 160: 116-24. 8. Pui C-H, Ip SH, Dodge RK et al. Serum levels of CD8 antigen in childhood lymphoid malignancies: a possible indicator of increased suppressor cell acitivy in poor-risk patients. Blood 1988; 72: 1015-21. 9. Pui C-H, Ip SH et al. Increased serum CD8 antigen levels in childhood Hodgkin's disease relate to advanced stage and poor treatment outcome. Blood 1989; 73: 209-13. 10. Pfreundschuh M, Loffler M, Ruhl U et al. Therapy of Hodgkin's lymphomas. Results of the German Hodgkin's Study Group. Onkologie 1988; 11: 48-52. 11. Pfreundschuh M, Schoppe WD, Fuchs R et al. Lomustine, etoposide, vindesine, and dexamethasone (CEVD) for Hodgkin's lymphoma refractory to COPP and ABVD. A multicenter trial of the German Hodgkin Study Group. Cancer Treatment Reports 1987; 71: 1203-7. 12. Friedman S, Henry-Amar M, Cosset JM et al. Evolution of erythrocyte sedimentation rate as predictor of early relapse in posttherapy early-stage Hodgkin's disease. Journal Clinical Oncology 1988; 6: 596-601.

Annals of Oncology 2: 584, 1991.

Classified advertisements from ESMO-members

Please contact: ANNALS OF ONCOLOGY Kluwer Academic Publishers P.O. Box 17 3300 AA Dordrecht The Netherlands

Tel.: 31-{0)78-334238 a * : 3 1 -{0)78-334254


Fourth International ARTAC Congress on Therapeutic Trials in Cancer and AIDS Scientific concepts, Methods and New Drugs September 25-27 1991, Paris, France Organizing committee: J.M. Andrieu, D. Betpomme, P. Fumoleau, S. Kaye, H. Maass, M. Marty, M. Plccart, T. Philip, J. Robert, A. Santoro, G. Schwartsmann; With the participants: J.M. Alexandra, Q. Atassl, M. Bdron, P. Conte, U. Creutzlg, H. Reblg, C. Franks, P. Hilgard, L Israel, C. Jasmin, F. Lavelle, U H. U, H.M. Pinedo, Y. Pommier, M. Rozencwelg, M. Schneider, L Schwarzenberg, H. Tagnon, M. Tonato, B. Winograd, P. Workman Scops of the congress: Special workshop on ovarian carcinoma; statistical methods; organization ol dlnlcal research In Europe; arrBcancer membrane attacking molecules; topoisomerases l-ll Inhibiting agents; taxol and other new drugs; new drug acceptation procedures In Europe; therapeutic intensification; chemoresistance; adjuvant and neoadjuvant chemotherapy; other treatments: differentiation inductors, InterteuWn II; therapeutic trends in AIDS, quality of life In cancer patients Deadline for submission o l abstracts: September 1,1991 Deadline for registration: September 15,19S1 For further Information and registration contact. D. Belpomme, Chairman of the "ARTAC" _ , ,„ t „ „ „ 36 Ruede Silly Tei.: 33.1 92100 Boutogne-BUlancourt Fax: (33.1)


Downloaded from by university of winnipeg user on 18 January 2019

These spaces are available for personal advertisements of ESMO members (free of charge), and for meeting announcements (Dfl. 200,- per space). Maximum approximately 130 words per space.

The clinical significance of serum CD8 antigen levels in adult patients with Hodgkin's disease.

Increased suppressor T-cell activity has been observed in patients with Hodgkin's disease. In order to evaluate the clinical significance of soluble C...
475KB Sizes 0 Downloads 0 Views