Effective B Cell Depletion with Rituximab in the Treatment of Autoimmune Diseases

Immunobiol. (2002) 206, pp. 519 – 527 © 2002 Urban & Fischer Verlag http://www.urbanfischer.de/journals/immunobiol

Medizinische Poliklinik der Universität Würzburg, Germany

Effective B Cell Depletion with Rituximab in the Treatment of Autoimmune Diseases CHRISTIAN KNEITZ, MARTIN WILHELM, and HANS PETER TONY

Abstract In a pilot study four patients with systemic lupus erythematosus (SLE) and autoimmune thrombocytopenia (ITP) were treated with rituximab, a B cell depleting chimeric human/mouse anti-CD20 antibody. Treatment could be performed without serious side effects and resulted in a depletion of B cells from the peripheral blood for at least 4 months. Examination of one patient three months after treatment revealed a complete depletion of B cells in the bone marrow and in the spleen as well. The time point when peripheral B cells returned into the normal range varied between 8 months and over one year and could be observed also in the spleen. The follow up over more than 12 months revealed no significant treatment-associated side effects. Total immunoglobulin and specific antibody levels did not change except for one SLE-patient receiving additional immunosuppressive treatment including cyclophosphamide because of progressive disease. Clinical effectiveness cannot be judged by the small number of patients. However, one SLE patient with refractory severe thrombocytopenia had a very favourable response with stable platelet numbers over 100.000/ml now for more than 6 months and disappearance of anti-DNA antibodies. The treatment failure in another SLE patient could be due to the persistence of CD20-negative plasmablasts in peripheral blood which are not targeted by anti-CD20 treatment. Further studies are needed to assess the clinical benefit of B cell depletion in the treatment of autoimmune diseases.

Introduction Systemic lupus erythematosus (SLE) is a disease characterized by the production of numerous antibodies and involvement of a variety of different organ systems. SLE in human is a heterogeneous disease and associated with abnormal immunoregulation. More recent data suggest that intrinsic B cell hyperreactivity may be an essential feature (1). Several studies in animal models also point to a critical role of the B cell. MRL lpr/lpr mice spontaneously develop a disease characterized by immune complex nephritis and vasculitis similar to human SLE. B cell-deprived MRL lpr/lpr mice do not develop a disease and show a reduced number of activated autoimmune T cells (2, 3). However, genetically manipulated MRL lpr/lpr mice, which have B cells with surface Abbreviations: SLE = systemic lupus erythematosus; ITP = idiopathic thrombocytopenia 0171-2985/02/206/05-519 $ 15.00/0

520 · C. KNEITZ et al. immunoglobulin, but lack secreted immunoglobulin nevertheless develop nephritis and vasculitis (4). This implies an essential role for B cells irrespective of autoantibodies during the evolvement of disease (1). Recently a disturbed peripheral B lymphocyte homeostasis in SLE has been characterized by elevated numbers of CD27+++, CD20-, CD19+ plasmablasts the peripheral blood of patients with active disease (5). Therefore, depletion of B cells may prove to be a reasonable immunoregulatory therapy. Treatment with B cell depleting antibodies – especially rituximab – has already been established for the therapy of B cell lymphomas with few side effects in patients with low tumor burden (6). A recent study indicates that there might be a limited but suitable effect of rituximab therapy in patients with ITP (7). Rituximab is a genetically engineered chimeric human/murine anti-CD20 monoclonal antibody. The antibody is an IgG1 kappa immunoglobulin containing murine light and heavy chain variable region and human constant region sequences (8). The Fab domain (murine) of rituximab binds to the CD20 antigen on B cells and the Fc domain (human) recruits immune effector functions to mediate B cell lysis. Possible mechanisms of cell lysis include complement activation, antibody dependent cellular cytotoxicity and induction of apoptosis (7). Patients and Methods The diagnosis of ITP was established by exclusion of other disorders known to cause shortened survival of platelets. A diagnostic bone marrow aspirate has been performed in all patients at presentation. In accordance with the guidelines published by the American Society of Haematology, treatment was considered appropriate for these patients because of a platelet count less then 20 ×109/ml (9, 10). The SLE patients fulfilled the criteria of the American Congress of Rheumatology for classification of SLE. All patients gave informed consent before entering the study. Treatment regimen: Rituximab (Mabthera®, Hoffmann La-Roche, Grenzach-Whylen, Germany) at a dose of 375mg/m2 was administered intravenously once a week for a total of 4 infusions (day 1, 8, 15, and 22). The drug was reconstituted in normal saline to a concentration of 1–4 mg/ml. The initial infusion rate was 50 mg/h, with subsequent increase if no toxicity was seen. Premedication with Ranitidin (50 mg) and Clemastin (0.67 mg) was given to all patients. During the infusion blood pressure, heart rate, and vital signs were monitored. Analysis of lymphocyte subsets by immunofluorescence staining was performed by incubating PBMC with anti-CD19, anti-CD27, and anti-CD20 antibodies (PE or FITC-labeled as indicated; all antibodies from Becton-Dickinson, Heidelberg, Germany). Flow cytometric analysis was performed using a FACSCalibur (Becton Dickinson, San Jose, CA, USA). Frequencies of cell populations were calculated using CellQuest software. The absolute B cell number was calculated per milliliter of blood, based on the frequencies of these cells among lymphocytes.

Results 1. Case reports

1a. ITP-patients I and II: A 39-yr (ITP-patient-I) and a 38-yr (ITP-patient-II) old male patient were referred to our department with severe thrombocytopenia (platelet count less then 20 ×109/ml). A diagnostic bone marrow aspirate was performed in both patients at presentation and showed an increased amount of megakaryocytes with normal megakaryocytic matura-

Therapy of SLE with rituximab · 521

Figure 1. IgG levels after B cell depletion with rituximab. Total IgG levels of patients investigated in the course of time

Figure 2. SLE-patient I: time course of peripheral B cells, thrombocytes and anti-DNA antibodies. Development of platelet counts (thrombocytes/ml), B cells (B cells/ml) and anti-DNA antibodies (U/ml) after monoclonal anti-CD20 antibody therapy in SLE-patient I. ↑ rituximab infusion with 375 mg/m2

tion and regular distribution of lymphoid subsets. The diagnosis of ITP was confirmed by exclusion of disorders known to cause shortened survival of platelets. Only a short and insufficient increase of platelet count has been observed during therapy with corticosteroids (1 mg/kg body weight). Whereas ITP-patient I had been treated only with steroids before, ITP-patient II had former treatment with azathioprin, intravenous gam-

522 · C. KNEITZ et al.

Figure 3. Time course of B cells in blood, spleen and bone marrow. Development of B cell counts (B cells/ml) after monoclonal anti-CD20 antibody therapy in all patients treated. FACS-analysis (FL1: CD3, FL2: CD19) of bone marrow (ITP-patient II) after two months and spleen (ITP-patient II and ITP-patient I) after 3 or 15 months, respectively, are shown. Percentages of CD3 (T cells) or CD19 (B cells) positive lymphocytes are given. ↑ rituximab infusion with 375 mg/m2

maglobulin, mycophenolat-mofetil and cyclosporin A without any persisting increase of platelet counts. Both patients were included in the study and have been treated with rituximab with no significant increase in platelet counts. Since both patients showed signs of haemorrhage, splenectomy was performed three (ITP-patient II) or 15 months (ITP-patient I) after start of rituximab treatment (Fig. 3). In ITP-patient II hyperplastic megakaryopoiesis was confirmed by an additional bone marrow biopsy shortly before splenectomy. After initial fluctuations in platelet counts ITP-patient II is in stable remission after splenectomy whereas patient I showed only a short increase in platelet counts after splenectomy. 1b. SLE-patient I (Fig. 2) A 37-year old female with a four years lasting history of SLE was initially diagnosed based on leukocytoclastic vasculitis, arthritis, positive anti-nuclear (ANA), anti-double stranded DNA (anti-dsDNA), and ribonucleoprotein antibodies. Moreover, she had severe thrombocytopenia with proof of anti-platelet autoantibodies shown to bind to the patient‘s platelets. Repeated high dose steroid therapy (dexamethasone 40 mg on four consecutive days followed by 1 mg/kg prednisolone) was performed in 1997. After an initial increase, platelet counts decreased again below 20.000/ml with signs of haemorrhage. At the end of 1997 splenectomy was performed. Platelet counts remained below 30.000/ml with 5–15 mg/day prednisolone from 1997–2000. In 2000, platelets

Therapy of SLE with rituximab · 523

dropped below 20.000/ml and petechial bleeding could be detected despite a prednisolone dose above 15 mg/day. Treatment with rituximab resulted in an steadily increase of platelet counts with values of more than 50.000/ml after three months. In addition, antiDNA antibodies were no more detectable after 4 months (Fig. 2). This patient initially stayed on low dose prednisolone (5 mg), which could be discontinued after normalisation of platelet counts in 2001. Polyarthralgia also improved and was not reported anymore 4 months after rituximab treatment. The cutaneous disease manifestations (anular configurated plaques, initialy confirmed as lupus manifestations by histology (lymphocytic infiltrates)) persisted. 1c. SLE-patient II A 43-year old female with a five year history of SLE was initially diagnosed based on weight loss, fever, polyarthritis, leukocytoclastic vaskulitis, lupus nephritis (WHO class IV), positive anti-nuclear (ANA), anti-double stranded DNA (anti-dsDNA) and RoSSaantibodies. In addition, she had low C3 and C4 complement levels, leukocytopenia and anaemia. Partial remission could be achieved after 13 cycles of cyclophosphamide application according to the Austin-protocol (98–5/99). Switch to a maintenance therapy with azathioprine or later mycophenolate-mofetil resulted each in a flare of disease with fatigue, increasing proteinuria, anaemia, decreased C3 and C4 levels and arthritis. Therefore, we started rituximab treatment in combination with low dose prednisolone (7.5 mg p.o.) in 08/99 (Fig. 2). After the second treatment (week 2) a vasculitic lesion of the skin (histologically confirmed leukocytoclastic vasculitis) developed on the back site of the left foot and subsequently a second one on the left knee. In the following weeks, disease activity increased with progressive proteinuria, arthritis and vasculitis of the skin. In 4/00 partial remission (normalisation of proteinuria, anti-DNA antibodies, arthritis) could be achieved by treatment with cyclophosphamide (Austin-protocol, three cycles, 15 mg/kg body weight). Maintenance therapy with cyclosporin A (3 mg/kg body weight) since June 2000 preserves partial remission. 2. Immunobiological parameters

2.1. Serum immunoglobulin and autoantibodies Serum IgG levels including IgG subclasses 1–4, IgA and IgM levels were measured before and during the study. All remained within normal limits throughout the observation period of 12 months except for SLE-patient II who required additional immunosuppressive therapy (Fig. 1). With respect to specific antibodies directed against measles, mumps, or rubeola virus and diphteria or tetanus toxoid no significant changes in the titers were observed. Interestingly, vaccination of SLE-patient II with tetanus-toxoid two months after completion of rituximab therapy caused a significant immune response (increase in tetanus titers from 0.4 to 1.5 IU/ml) although peripheral B cell counts were still significantly reduced. None of the patients developed new antibodies against other nuclear antigens. In both SLE patients the titer of anti-DNA antibodies correlated significantly with the clinical response to rituximab therapy (Fig. 2). Increased levels of antiplatelet antibodies were found only in SLE patient I and remained unchanged during therapy.

524 · C. KNEITZ et al.

Figure 4. Persistence of CD19+CD27+++CD20– B cells in SLE-patient II. CD20 and CD27 expression on peripheral B cells of SLE-patient II and I. Viable PBMC were gated for analysis. Staining with CD19/CD20 is shown for both patients at different time points as indicated. CD27 expression on CD19 positive B-cells after three months is shown as an insert. Fluorescence gate for the statistical evaluation of CD27+++ B cells is indicated, as well as the frequencies of CD20+ and CD20– B cells.

2.2. B lymphocytes Peripheral blood B cell counts, evaluated by flow cytometry as CD19+ cells, declined rapidly during treatment as indicated in figure 3. Recovery of B cells in the ITP-patients and SLE-patient I started between months three and four after end of anti-CD20 therapy (Fig. 3). The persistent low platelet counts with cutaneous bleeding required splenectomy in ITP-patient II at 3 months and ITP-patient I at 14 months after anti-CD-20 therapy. At 3 months after rituximab treatment no B cells could be detected by FACS analysis in peripheral blood, bone marrow, and spleen of ITP-patient II. In ITP-patient I at 14 months after rituximab infusion peripheral B cells as well as splenic B cells have been regenerated (Fig. 3). Similarly, in SLE-patient I B cell regeneration in the peripheral blood started 4 months after treatment, and normal B cell numbers were reached after 7 months. In SLE-patient II we could not achieve complete B cell depletion in the peripheral blood. After rituximab treatment a small but significant number of CD19+ B cells persisted. Further immunophenotyping revealed a phenotype described to plasmablasts (CD20-, CD19+, CD27+++) (Fig. 4). These cells have been seen in SLE patients with very active disease (5). After additional treatment with cyclophosphamide and cyclosporin A the CD20-, CD19+, CD27++++ subpopulation decreased in parallel to the observed clinical improvement (before therapy: 1.94%, after three months: 0.73%, after 19 months: 0,66%). This patient showed slow regeneration of peripheral B cells (total numbers of CD19+ cells) with still subnormal counts 12 months after treatment. In SLE-patient I

Therapy of SLE with rituximab · 525

(Fig. 4) and both ITP-patients (not shown) we could not detect a significant number of plasmablasts. In all patients T cells (CD3+), CD4 and CD8 positive T cells, and NKcells remained unchanged during the study period. 3. Response to therapy and side effects

All patients completed the planned treatment schedule. A significant clinical response was seen in SLE-patient II with a nearly complete improvement of all disease manifestations: After refractory thrombopenia for years platelet counts meanwhile are in normal range without additional steroid therapy. Polyarthralgia reported before treatment completely disappeared. Only the skin phenomena which might be manifestations of the disease, persisted. Despite modest febrile reactions none of the patients developed treatment related side effects. Most important severe infections were not observed during the study period. Discussion Successful treatment of autoimmune diseases is often limited by partial effectiveness or side effects of immunosuppressive therapy. Therefore, new alternative approaches are necessary. Enhanced B cell function is regarded as an important pathogenic event also in human SLE (1). Therefore, we started to investigate the feasibility of a treatment regime using the B cell depleting monoclonal antibody rituximab in patients with SLE and ITP. Rituximab was used according to the licensed schedule for treatment of non Hodgkin lymphoma infusing 375 mg/m2 monoclonal antibody once a week for four consecutive weeks. The treatment has been well tolerated. Over the study period of one year we did not observe any protracted toxicity attributable to B cell depletion. Specifically we did not see enhanced susceptibility to infection. Overall immunoglobulin isotype and IgG subclass levels did not change under therapy except for a small but significant decrease of total IgG in SLE-patient II (Fig. 1). However this is most likely due to the subsequent use of antiproliferative treatment with cyclophosphamide required due to progressive disease. These findings are in line with the presence of long-lived CD20-negative plasma cells, which are not targeted by anti-CD20 treatment (fig. 4). Accordingly SLEpatient II did mount a secondary antibody response after boosting with tetanus toxoid two months after the end of rituximab therapy despite the depletion of peripheral B cells. This is somewhat different to immunization studies in a primate model where interference of rituximab with the primary and memory response against a hapten (DNP) has been reported (11). Treatment with anti-CD20 monoclonal antibody resulted in all patients in a fast decrease of peripheral B cell numbers (Fig. 3). The effectiveness of B cell depletion related not only to the peripheral blood. In a single patient we could study the bone marrow two months and the spleen three months after rituximab treatment. In both lymphoid organs we were unable to demonstrate B cells by FACS analysis (Fig. 3). This suggests a quite effective depletion of the CD20 positive B cell pool in vivo. CD20 positive B cells repopulated the peripheral blood between 4 and 5 months after treatment. However, there are substantial individual differences for the time when B-cell

526 · C. KNEITZ et al. numbers return into the normal range. This varies between 8 months and over one year. B cell regeneration could also be observed in the spleen of one patient who underwent splenectomy 15 months after rituximab treatment (Fig. 3). The study cannot address the clinical effectiveness of B cell depletion in autoimmune disease. Recently a report on idiopathic thrombocytopenia (ITP) indicated an overall response rate (complete, partial or minor response) in 52% of 18 ITP-patients (7). Both ITP patients in our study did not respond to the treatment, which could be attributed to the small sample size. However, we describe one SLE-patient (patient I) who suffered from a limited stage of disease with profound thrombocytopenia as the major clinical problem. The patient required recurrent courses of high dose prednisolone for nearly three years and could rarely and only for short time periods be tapered below 10 mg/day. Rituximab treatment resulted in a very beneficial and long lasting response. Platelet counts rose slowly and stayed above 50.000/ml after three months with a strong tendency to normalisation. The more delayed platelet response was different to the study in ITP patients in which a more rapid increase of platelet counts, usually one week after the first rituximab dose, was observed. Therefore the mechanism of action is probably different and more likely due to overall decreasing disease activity in our lupus patient instead of mechanisms involving Fc receptor functions as discussed in the ITP study (7). Accordingly the patient reported a significant improvement in polyarthralgias. In parallel we also observed a fall in anti-DNA titers which turned negative 3 months after anti-CD20 treatment. This observation is in line with a recently published case report of a patient with catastrophic systemic lupus erythematosus in combination with Rosai-Dorfman sinus histiocytosis in whom rituximab treatment resulted in a rapid clinical improvement (12). However, the second SLE-patient (SLE-patient II) reported here, failed rituximab treatment. This patient with a highly active disease was heavily pretreated without inducible stable response. After treatment with rituximab she developed a flare with increasing nephrotic syndrome in the following weeks. Vasculitic skin ulcerations appeared very early under therapy. The pathomechanism of this remains unclear since increasing disease activity as well as a drug-specific side effect, as recently reported by DEREURE (13), might be accused. Upon rituximab treatment peripheral B cells rapidly declined also in this patient but a significant number of CD19+CD20-CD27+++ plasmablasts persisted in the periphery (Fig. 4). These cells have been described in SLE patients and are related to a high disease activity (5). Accordingly, their number declined in parallel to an observed clinical improvement eventually induced by cylophosphamide therapy (Fig. 4). SLE-patient I, which showed a good response to rituximab therapy, did not have elevated numbers of peripheral plasmablasts. Although we do not exactly understand the pathogenic role of this B cell subset, it could be speculated that the inability of anti-CD20 treatment do deplete these cells might be related to the treatment failure. In this context cyclophosphamide, which also targets plasma cells may be important for clinical response. In summary, B cell depletion by anti-CD20 treatment is feasible and results in complete depletion of the CD20 positive B cell pool in blood, spleen and bone marrow. A clinical efficiency can be seen in single patients and warrants further study.

Therapy of SLE with rituximab · 527 Acknowledgments

We acknowledge Hoffmann La-Roche for supplying rituximab monoclonal antibody used in this study and R. JAHNS for critical comments.

References 1. LIPSKY, P. E. 2001. Systemic lupus erythematosus: an autoimmune disease of B cell hyperactivity. Nat. Immunol. 9: 764. 2. CHAN, O. T., and M. J. SHLOMCHIK. 1998. A new role for B cells in systemic autoimmunity: B cells promote spontaneous T cell activation in MRL-lpr/lpr mice. J. Immunol. 160: 51. 3. CHAN, O. T., M. P. MADAIO, and M. J. SHLOMCHIK. 1999. B cells are required for lupus nephritis in the polygenic, Fas-intact MRL model of systemic autoimmunity. J. Immunol. 163: 3592 4. CHAN, O., G. L. HANNUM, M. HABERMANN, P. MADAIO, and M. J. SHLOMCHIK. 1999. A novel mouse with B cells but lacking serum antibody reveals an antibody-independent role of B cells in lupus. J. Exp. Med. 10: 1639 5. ODENDAHL, M., A. JAKOBI, A. HARNSEN, E. FEIST, F. HIEPE, G. R. BURMESTER G. R., P. E. LIPSKY, A. RADBRUCH, and T. DÖRNER. 2000. Disturbed peripheral B lmyphocyte homeostais in systemic lupus erythematosus. J. Immunol. 165: 5970. 6. SACCHI, S., M. FEDERICO, G. DASTOLI, C. FIORANI, G. VINCI, V. CLO, and B. CASOLARI. 2001. Treatment of B-cell non-Hodgkin’s lymphoma with anti CD 20 monoclonal antibody Rituximab. Crit. Rev. Oncol. Hematol. 37: 13. 7. STASI, R., A. PAGANO, E. STIPA, and S. AMADORI. 2001. Rituximab chimeric anti-CD20 monoclonal antibody treatment for adults with chronic idiopathic thrombocytopenic purpura. Blood 98: 952. 8. REFF, M. E., K. CARNER, K. S. CHAMBERS et al. 1994. Depletion of B cells in vivo by a chimeric mouse human monoclonal antibody to CD20. Blood 83: 443 9. GEORGE, J. N., S. H. WOOLF, and G. E. RASKOB. 1996. Idiopathic thrombocytopenic purpura.: a practice guideline developed by explicit methods for the American Society of Hematology. Blood 88: 3. 10. Diagnosis and Treatment of Idiopathic Thrombocytopenic Purpura. 1997. Recommendations of the American Society of Hematology. The American Society of Hematology ITP practice guideline panel. Ann. Intern. Med. 126: 319. 11. GONZALEZ-STAWINSKI, G. V., P. B. YU, S. D. LOVE, W. PARKER, and R. D. DAVIS, Jr. 2001. Hapten-induced primary and memory humoral responses are inhibited by the infusion of antiCD20 monoclonal antibody (IDEC-C2B8, Rituximab). Clin. Immunol. 98: 175. 12. PETSCHNER, F., U. A. WALKER, A. SCHMITT-GRAFF, M. UHL, and H. H. PETER. 2001. Catastrophic systemic lupus erythematosus with Rosai-Dorfman sinus histiocytosis. Successful treatment with anti-CD20/rituximab]. Dtsch. Med. Wochenschr. 126: 998. 13. DEREURE, O., R. NAVARRO, J. F. ROSSI, and J. J. GUILHOU. 2001. Rituximab-induced vasculitis. Dermatology 203: 83. PD Dr. CHRISTIAN KNEITZ, Medizinische Poliklinik der Universität Würzburg, Klinikstr. 6–8, D-97070 Würzburg, [email protected]