Nine patients with anti-neutrophil cytoplasmic antibody-positive vasculitis successfully treated with rituximab

Transcription

1 Journal of Internal Medicine 2005; 257: Nine patients with anti-neutrophil cytoplasmic antibody-positive vasculitis successfully treated with rituximab P. ERIKSSON From the Department of Rheumatology, University Hospital, Linköping, Sweden Abstract. Eriksson P (University Hospital, Linköping, Sweden). Nine patients with anti-neutrophil cytoplasmic antibody-positive vasculitis successfully treated with rituximab. J Intern Med 2005; 257: Objectives. Rituximab (RIT) is a monoclonal anti- CD20 antibody, which depletes B-lymphocytes but not plasma cells. RIT is used for treatment of B-cell lymphomas, but has also shown beneficial effects in autoimmune diseases. In this case series RIT was used in anti-neutrophil cytoplasmic antibody (ANCA)-positive vasculitis. Design. Case series with a structured follow-up of treated patients. Setting. Departments of Nephrology and Rheumatology of a university hospital. Subjects. Two women with myeloperoxidase-ancapositive microscopic polyangiitis and seven patients (five men and two women) with proteinase 3-ANCApositive Wegener s granulomatosis. All patients were resistant to conventional therapy or had relapsed repeatedly after cessation of cyclophosphamide (Cyc). Interventions. The cases were treated with intravenous infusions of RIT once a week two times (three cases) or four times (six cases). To prevent formation of antibodies to RIT, mycophenolate mofetil (five patients), azathioprine (one patient), or a short course of Cyc (two patients) were added or allowed to continue. Mainoutcome measures. Remission at 6 months assessed with Birmingham vasculitis activity score. The cases were followed 6 24 months and relapse rate was also noted. Results. Eight of nine patients responded completely and one case responded partially. Pulmonary X-ray improved (four cases), progress of lower extremity gangrene stopped (one case), remission of neuropathy was stable (one patient), renal vasculitis went into remission (two cases), and severe musculoskeletal pain improved (one case). Minor relapse in the nose occurred in two cases. No adverse events or major infections were noted. Conclusion. RIT seems promising and safe in ANCA-positive vasculitis, and controlled studies should be conducted. Keywords: anti-neutrophil cytoplasmic antibody, rituximab, therapy, vasculitis. Introduction Rituximab (RIT) is a chimeric antibody to CD20 causing lysis of B-lymphocytes and used for treatment of lymphomas. Serious side-effects have been noted more frequently in patients with large tumours [1], but no increased rate of infections has occurred. RIT has also been successfully used in patients with rheumatoid arthritis (RA) [2, 3], idiopathic thromobocytopaenic purpura [4], autoimmune haemolytic anaemia [5], cold agglutinin disease [6], systemic lupus erythematosus (SLE) [7, 8], myositis [9], and myasthenia gravis [10]. In contrast to lymphoma treatment, only few side-effects have been reported in autoimmune patients [2, 11]. Formation of antibodies to RIT has been reported in 1% of 355 lymphoma patients according to the manufacturer, and in 0 of 37 lymphoma patients in one study [12]. Wegener s granulomatosis and microscopic polyangitiis are characterized by anti-neutrophil cytoplasmic antibodies (ANCA) and small vessel vasculitis. The diseases usually respond to 540 Ó 2005 Blackwell Publishing Ltd

2 RITUXIMAB IN VASCULITIS 541 cyclophosphamide (Cyc) but some patients are therapy resistant. One case with Wegener s granulomatosis successfully treated with RIT [13] prompted us to use this drug in patients with ANCA-positive vasculitis who were frequently relapsing or resistant to therapy with continuous oral as well as to intermittent intravenous Cyc. Materials and methods Patients Nine patients (five men and four women) with therapy-resistant or frequently relapsing ANCA-positive vasculitis were treated with RIT after informed consent. Basic patient data are given in Table 1. The median age was 59 years (range 35 77). Seven patients had proteinase 3 (PR3)-ANCA and two patients had myeloperoxidase (MPO)-ANCA. The patients with MPO-ANCA were classified as microscopic polyangiitis and the seven patients with PR3- ANCA as Wegener s granulomatosis according to the Chappel Hill criteria with addition of nasal symptoms as a sign of Wegener s granulomatosis [14]. The vasculitis presented 1 21 years prior to treatment with RIT. Before RIT, the median number of relapses was 3.5 (range 1 5) in six patients, whereas three patients had never reached remission. The patients had experienced involvement of median 4 (range 1 5) organs (lungs, nose, kidneys, joints, skin, and peripheral nerves). All patients were resistant to conventional therapy, including Cyc, or had relapsed repeatedly after cessation of Cyc. Previous therapy included: oral Cyc (eight cases), intravenous intermittent Cyc (eight cases), corticosteroids (nine cases), mycophenolate mofetil (four cases), azathioprine (five cases), intravenous immunoglobulins (three cases), methotrexate (two cases), cyclosporin A (three cases), etoposide (one case), leflunomide (one case), anti-thymocyte globulin (one case), and plasmapheresis (two cases). Study design The patients were prospectively assessed at month 1 and every third month for at least 6 months after start of RIT. The primary endpoint was remission or not 6 months after start after RIT. Time to remission, time to relapse, side-effects, dose of prednisolone, course of C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), blood B-lymphocytes, plasma creatinine, ANCA, IgG, IgM, and IgA were also studied. Clinical disease activity was assessed with Birmingham vasculitis activity score (BVAS) [15]. Complete remission was defined as a BVAS score of zero indicating the absence of signs of new Table 1 Basic data concerning nine vasculitis patients treated with rituximab. The three cases without relapses never reached complete remission. All cases with Wegener s granulomatosis had proteinase-3 antibodies, and both cases with microscopic polyangitiis had myeloperoxidase antibodies Case no. Diagnosis/ sex Age at start of RIT/onset of vasculitis Affected organs since onset of vasculitis No. of relapses Previous medication (besides steroids) Prednisolone dose at start of RIT (mg) Total dose of RIT (mg) Follow-up (months after RIT)/ relapse 1 MP/F 60/58 S, N, K, ENT 1 c*, c /no relapse 2 WG/F 61/49 L, K, S, J, ENT 4 c*,c, m, ATG, cya, s, IVIG, /no relapse PE (92 times), etoposid 3 WG/F 49/48 L, K, S, ENT 0 c*, c, az, mtx, m, IVIG /relapse at 12 months 4 WG/M 59/56 N, K, J, ENT 4 c*, c, m, /relapse at 13 months 5 WG/M 56/49 J, K, N, S, ENT 5 c*, c, az, cya mtx, m, /no relapse IVIG, l, s, infliximab, 6 WG/M 73/72 L, N 0 c* /no relapse 7 MP/F 55/50 K 2 c*, c, az /no relapse 8 WGM 35/14 K, S, J, ENT 3 c*, c, az, m, s, cya, PE /no relapse 9 WGM 77/75 L, ENT 0 c, az /no relapse Most affected organs indicating rituximab in bold letters. WG, Wegener s granulomatosis; MP, microscopic polyangiitis; ENT, ear nose throat; J, joints; K, kidneys; L, lungs; N, peripheral nerves; S, skin; c, cyclophosphamide; ATG, anthithymocyte globulin; IVIG, intravenous immunoglobulins; s, sirolimus; az, azathioprine; l, leflunomide; cya, cyclosporin A; mtx, methotrexate; m, mycophenolate mofetil; PE, plasma exchange. *Intravenous intermittent.

3 542 P. ERIKSSON or worse disease activity. Partial remission was defined by persistent disease activity for no more than one item (BVAS 1) [16]. The cases with pulmonary involvement were also evaluated with pulmonary radiographs every third month. Dose of RIT In six cases, RIT was given intravenously as 4 weekly fixed doses of 500 mg (or 375 mg m )2 body area in one case with a weight of 140 kg). In three cases, RIT was given as 2 weekly doses of 500 mg (see Table 1). Immunosuppressive therapy potentially biasing the results To prevent formation of antibodies to RIT, mycophenolate mofetil (five cases), azathioprine (one case), and a short course of Cyc during 7 weeks (two case) were added or continued. As shown in Table 2, mycophenolate was started for the first time concomitantly with RIT in three cases (cases 1, 6 and 9). Mycophenolate had previously not been successful to keep remission in two patients (cases 4 and 8) but it was continued in these cases. One case (case 3) was switched from Cyc to azathioprine 5 weeks before RIT and afterwards azathioprine was continued. Two cases (cases 5 and 7) got a 7-week course of Cyc at start of RIT, but previous courses had not resulted in longstanding remission. One case (case 2) could not tolerate any of the previous tested immunosuppressive drugs and she was treated with RIT and steroids alone. Laboratory analyses Circulating B-lymphocytes expressing CD19 were determined by FACS Calibur TM flow cytometer (Becton Dickinson) (reference: L )1, detection level: > L )1 ). ANCA analyses were performed at the department of clinical immunology (Linköping University Hospital) and included indirect immunofluorescence microscopy (IF) on ethanolfixed granulocytes (reference: not detectable at serum dilution 1 : 25; no titration was performed), conventional in-house ELISAs for MPO-ANCA (reference: <7 units) and PR3-ANCA (reference <5 units), and capture-elisa for PR3-ANCA (Wielisa, Wieslab AB, Lund, Sweden) (reference: <20 units). Table 2 Immunosuppressive medication 16 weeks before and 10 weeks after start of rituximab in the nine cases. All cases were also treated with steroids Weeks before and after start of rituximab )16 )15 )14 )13 )12 )11 )10 )9 )8 )7 )6 )5 )4 )3 )2 ) Case no. 1 c* c* c* c* c* c* c c c c c c c c c c m m m m m m m m m m 2 s s s s s s s 3 c c c c c c c c c a a a a a a a a a a a a a a a 4 g g m, g m, g m, g m, g m, g m, g m, g m m m m m m m m m m m m m m m m m 5 l, g l, g l, g l, g s, g s, g s, g s, g s, g s, g s, g s, g s s s c c c c c c c c 6 c* c* c* c* c* c* c* c* m m m m m m m m m m 7 c c c c c c c c a a a a a a c, p c c c c c c 8 c, s c, s c, s c, s c, s c, s c, s, p s s m, s m, s m, s m, s m, s m, s m, s m, s m, s m, s m, s m, s m, s m, s m, s m, s m, s 9 a a a a a a a a a a m m m m m m m m m m c, cyclophosphamide; m, mycophenolate; s, sirolimus; a, azathioprine; l, leflunomide; g, gamma globulin i.v. intermittent; p, plasmapheresis. *Intermittent intravenous (not continuous oral).

4 RITUXIMAB IN VASCULITIS 543 Other laboratory analyses included ESR (reference: 12 mm or less in men, and 20 mm or less in women), plasma CRP (reference: <10 mg L )1 ), plasma creatinine analysed with a conventional Jaffe method (reference: lmol L )1 in men and lmol L )1 in women), and plasma levels of IgG (reference: g L )1 ), IgA (reference: g L )1 ), and IgM (reference: g L )1 ) (nephelometric methods used for the Ig analyses). Ethical considerations The study protocol was approved by the local ethics committee. Results Remission As shown in Fig. 1, eight of nine cases were in complete remission (BVAS 0) and the ninth patient was in partial remission (BVAS 1) at 6 months. Median BVAS was 6 (range 2 18) before RIT, 0 (range 0 6) at month 1, 0 (range 0 2) at month 3, and 0 (range 0 1) at month 6. All three patients getting the lower dose of RIT (1000 mg) went into complete remission (cases 7, 8, 9). All patients, except case 9 who received 1000 mg of RIT, started to respond within 1 month after institution of RIT. The following effects were noted: progress of lower extremity gangrene stopped (case 1) (Fig. 2), stable remission of neuropathy in hands (case 4), disappearance of arthritis and improvement of severe musculoskeletal pain (case 5 with partial remission), and a gradual improvement of pulmonary symptoms and improvement of pulmonary X-ray in four cases. Figure 3 shows the continuing decrease in size of the nodular lesions during a long period in case 3. The two cases with renal vasculitis also responded to RIT. The time course of plasma creatinine for case 7 is shown in Fig. 4. Renal vasculitis presented 5 years previous to RIT, and it relapsed 5 months before RIT was started. The relapse was first treated with oral Cyc, and plasma creatinine decreased from 162 lmol L )1 to approximately 130 lmol L )1. Eight weeks before RIT, azathioprine was given instead of Cyc followed by a new renal relapse with a maximum plasma creatinine of 421 lmol L )1. Cyc was reinstitued 2 weeks before start of RIT and stopped 5 weeks afterwards. Plasma creatinine decreased to 344 lmol L )1 6 months after RIT, and to 314 lmol L )1 12 months after RIT. In the other case with renal involvement (case 8), Wegener s granulomatosis presented 21 years before RIT was given. The first kidney transplantation was performed 1 year after presentation of Wegener, and the second transplantation 2.5 year before RIT. Six months before RIT, increasing haematuria and proteinuria indicated kidney biopsy showing relapse of renal vasculitis. Cyc was instituted 21 weeks before RIT and stopped 11 weeks later because of a febrile cytomegalovirus infection. As anti-rejection therapy Fig. 1 Disease activity (BVAS) in nine patients with ANCA-positive vasculitis treated with rituximab. BVAS Months after rituximab Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Case 7 Case 8 Case 9

5 544 P. ERIKSSON (a) (b) Fig. 2 The feet of case 2 with vasculitic gangrene at start of rituximab (a) and 9 months later (b). the patient had sirolimus, and mycophenolate was restarted 7 weeks before RIT. RIT was given because of continuing proteinuria, microhaematuria, and granular urine casts. Before and after Cyc, plasma creatinine was 110 and 131 lmol L )1 respectively. When RIT was initiated, plasma creatinine was 142 lmol L )1 and 12 months later it was 128 lmol L )1. Iohexol clearance was 50 ml min )1 at start of RIT, 48 ml min )1 at 6 months, and 62 ml min )1 at 12 months. Kidney biopsy 6 months after start of RIT could not detect any active vasculitic lesions, and previous granular casts had vanished although proteinuria persists. (c) Dose of prednisolone The median prednisolone dose was 17.5 mg (range 5 40) before RIT, 10 mg (range 5 20) at 1 month, Fig. 3 Pulmonary radiographs of case 2, before (a), 6 months (b), and 21 months (c) after rituximab.

6 RITUXIMAB IN VASCULITIS 545 Serum creatinine (µmol/l) weeks Fig. 4 The course of plasma creatinine in case 7 with renal vasculitis treated with rituximab. Cyclophos phamide Azathio prine Cycloph amide out Rituximab 10 mg (range 5 15) at 3 months, and 10 mg (range ) at 6 months. Other immunosuppressive drugs are shown in Table 2. Adverse reactions No serious adverse events occurred and only two ambulatory treated respiratory tract infections were registered during the observation period. One patient had cutan Herpes infection before RIT and afterwards it continued to relapse. Relapse rate In November 2004, seven patients were in remission after 6 25 months of observation, as shown in Table 1. Two patients (cases 3 and 4) experienced minor nasal relapses after 12 and 13 months respectively (BVAS 2 and 4 respectively). These relapses were easily treated with a temporary increase of prednisolone from 0 to 10 mg in case 3 and from 5 to 20 mg in case 4. At time of relapse blood B-lymphocytes were L )1 in case 3 and L )1 in case 4. CRP and ESR The median level of CRP was 28 mg L )1 (range <10 171) before RIT, 18 mg L )1 (range <10 75) at month 1, <10 mg L )1 (range <10 48) at month 3, and <10 mg L )1 (range <10 209) at 6 months. Median ESR decreased from 48 mm (range 4 90) at start to 30 mm (range 4 76) at 1 month, 20 mm (range 4 89) at 3 months, and to 17 mm (range 4 100) at 6 months (Fig. 5). Erythrocyte sedimentation rate (mm) Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Case 7 Case 8 Case 9 Fig. 5 ESR in nine patients with ANCA-positive vasculitis treated with rituximab Months after rituximab

7 546 P. ERIKSSON Blood B lymphocytes ( /L 1 ) Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Case 7 Case 8 Case Months after rituximab Fig. 6 Blood B-lymphocytes in nine patients with ANCA-positive vasculitis treated with rituximab. Haematology Circulating B-lymphocytes were undetectable in all patients 4 weeks after start of RIT, as shown in Fig. 6. Recovery began 6 12 months after RIT in six of nine patients but two patients have only been followed 6 and 10 months respectively. Plasma creatinine In the seven patients without active renal vasculitis plasma creatinine did not change after treatment. Median plasma creatinine was 96 lmol L )1 (range ) at start of RIT, 92 lmol L )1 (range ) at 1 month, and 97 lmol L )1 (range ) at 6 months. ANCA In the two cases with microscopic polyangiitis, the MPO-ANCA levels were below the cut-off limit before and after RIT. IF-ANCA was positive before and negative after RIT in one case and positive before as well as after RIT in the other case. In the patients with Wegener s granulomatosis, IF-ANCA did not change in seven of nine patients (remained positive 6 months after RIT in six patients and negative in one patient). In case 4, IF-ANCA was negative at month 0 but became positive at month 12 just prior to a minor nasal relapse. The median levels of capture-elisa PR3-ANCA were 74 units (range ) before RIT, 244 units (range ) after 1 month, 375 units (range ) after 3 months, and 38 units (range ) after 6 months. Plasma immunoglobulins The median plasma IgG concentration was 8.5 g L )1 (range ) before RIT, 7.2 g L )1 (range ) after 1 month, and 6.4 g L )1 (range ) after 6 months. The median plasma IgM was 0.78 g L )1 (range ) before RIT, 0.51 g L )1 (range < ) after 1 month, and 0.36 g L )1 (range < ) after 6 months. The median plasma IgA was 1.17 g L )1 (range ) before RIT, 1.07 g L )1 ( ) after 1 month, and 1.3 g L )1 (range ) after 6 months. Discussion In this case series, RIT was successfully used in nine patients with frequently relapsing or therapy-resistant ANCA-positive vasculitis. RIT seems effective, and no serious adverse advents have hitherto been noted. These findings are in agreement with 17 similar cases published as congress abstracts [17, 18]. Most of our patients had smouldering or frequently relapsing disease before the start of RIT explaining the moderate increases of initial BVAS and CRP levels in most cases. However, these patients were not in stable remission on conventional therapy, and they needed more efficient therapy.

8 RITUXIMAB IN VASCULITIS 547 The effect of RIT on disease activity was obvious within a couple of weeks in most patients. However, in one case pulmonary radiographs was unchanged at 3 months but improved at 6 months after RIT. The speed of therapeutic onset is especially important, i.e. in serious renal involvement where scarring can result in chronic renal failure and dialysis dependency. One important question in future studies will therefore be the time to onset of improvement and time to remission and not only the number of patients responding to therapy. Our patients were also given additional immunosuppressive medication, and the main reason was to prevent formation of antibodies to RIT. In the three cases who started mycophenolate concomitantly with RIT, a synergistic effect of RIT and mycophenolate cannot be excluded. However, these cases had previously been resistant to Cyc. In the other six cases (three cases who continued mycophenolate or azathioprine after start of RIT, one case with only steroids, and two cases with a short course of Cyc) RIT probably had a major impact on the induction and maintaining of disease remission. In a recent prospective randomized trial of patients with RA, RIT combined with methotrexate was more efficient than methotrexate as monotherapy, whereas RIT as monotherapy was not significantly better than methotrexate as measured with the methods proposed by the American College of Rheumatology (ACR) (ACR 50 or ACR 70) at 24 and 48 weeks respectively [2]. Two different dose regimens were tested in this study. Six patients got 4 weekly doses and three patients got 2 weekly doses of RIT. All cases experienced remission, but the data should be interpreted with caution concerning any relationship between the dose of RIT and efficacy. However, the results can hint possible doses in future randomized studies. Looney et al. [8] treated SLE-patients with a single infusion of RIT 100 or 375 mg m )2, or with 4 weekly infusions of 375 mg m )2. Profound B-cell depletion was found in 11 of 17 cases, although one of the cases getting the highest dose still had detectable B-cells in the blood. The same group reported B cell depletion to be highly correlated with both the serum RIT level at 2 months after RIT and the FccRIIIa genotype [19]. In a prospective randomized trial of 161 RA patients, Edwards et al. [2] found that two 1000 mg doses of RIT nearly completely deleted the peripheral B-cells. In a previous dose finding study, a total RIT dose below 600 mg m )2 resulted in therapeutic failure in RA patients [3]. Thus, experience in SLE and RA may also guide dosing of RIT in future randomized studies of vasculitis. The risk of infections with RIT seems not to be increased in patients with lymphoma or autoimmune diseases. Median levels of plasma IgG and IgA did not decrease below the lower reference limit in our patients. In previous studies of patients with RA, the total serum immunoglobulin levels remained within the normal range after RIT, and the levels of IgG autoantibodies decreased significantly more than did the total serum IgG. In contrast, levels of antimicrobial antibodies did not change significantly [2, 20]. However, in lymphoma patients treated with RIT an impaired immune response, especially for recall antigens, was found in another study [21]. In our cases, circulating B-lymphocyte numbers fell below the detection limit within 1 month after institution of RIT, and they started to recover 6 12 months later. This pattern is consistent with the results of other authors, i.e. in RA patients treated with RIT [2]. However, recovery of B-lymphocytes does not unavoidably lead to relapse of vasculitis, although two of our patients experienced minor nasal relapses months after the start of RIT treatment. These relapses were, however, easily treated with a minor increase of the prednisolone dose, and their previously serious vasculitic disease still remains easy to manage. Rituximab (RIT) may deplete B cells by complement-dependent cytotoxicity, antibody-dependent cell-mediated cytotoxicity and/or apoptosis [11]. The way RIT works in autoimmune diseases is not clear. Antigen presentation rather than decreased levels of autoantibodies may possibly be involved as RIT depletes B lymphocytes but not plasma cells [22]. Furthermore, B cells are essential for the differentiation of follicular dendritic cells into secondary lymphoid organs and the organization of effective lymphoid architecture. Activated B cells express co-stimulatory molecules and may produce cytokines essential for the evolution of T effector cells. Furthermore, cytokines (i.e. IL-10) produced by B cells may influence antigen-presenting dendritic cells [23]. This study was not designed to investigate the mechanisms of RIT on the disease process. However, PR3-ANCA as measured by

9 548 P. ERIKSSON capture-elisa did not disappear completely during the study period. These data, together with the rapid effect on the clinical disease activity, argue against the hypothesis that decreased antibody levels is the single or predominant mechanism of action for RIT. To conclude, in this series of nine patients with therapy resistant ANCA-positive vasculitis treatment with RIT was efficient and no serious adverse advents were noted. However, RIT should not be used as first-line therapy before larger randomized studies have been performed to compare RIT with conventional therapy. Conflict of interest statement Nothing to declare. Acknowledgements Staffan Hammerby at the Department of Radiology, University hospital, Linköping, Sweden, provided the radiographic pictures. Thanks to Dr Charlotte Dahle at the Department of Clinical Immunology, Linköping University Hospital, for valuable help with the immunological methods. The valuable comments offered by Prof. Thomas Skogh, Department of Rheumatology, Örebro University Hospital, are appreciated. References 1 Hagberg H, Holmbom E. Risk factors for side effects during first infusion of rituximab definition of a low risk group. Med Oncol 2000; 17: Edwards JC, Szczepanski L, Szechinski J et al. Efficacy of B-celltargeted therapy with rituximab in patients with rheumatoid arthritis. N Engl J Med 2004; 350: Leandro MJ, Edwards JC, Cambridge G. Clinical outcome in 22 patients with rheumatoid arthritis treated with B lymphocyte depletion. Ann Rheum Dis 2002; 61: Saleh MN, Gutheil J, Moore M et al. A pilot study of the anti- CD20 monoclonal antibody rituximab in patients with refractory immune thrombocytopenia. Semin Oncol 2000; 27: Zecca M, De Stefano P, Nobili B, Locatelli F. Anti-CD20 monoclonal antibody for the treatment of severe, immunemediated, pure red cell aplasia and hemolytic anemia. Blood 2001; 97: Berentsen S, Tjonnfjord GE, Brudevold R et al. Favorable response to therapy with the anti-cd20 monoclonal antibody rituximab in primary chronic cold agglutinin disease. Br J Haematol 2001; 115: Leandro M, Edwards J, Cambridge G, Ehrenstein M, Isenberg D. An open study of B lymphocyte depletion in systemic lupus erythematosus. Arthritis Rheum 2002; 46: Looney RJ, Anolik JH, Campbell D et al. B-cell depletion as a novel treatment for systemic lupus erythematosus: a phase I/ II dose-escalating trial of rituximab. Arthritis Rheum 2004; 50: Levine T. A pilot study of rituximab therapy for refractoy dermatomyositis. 66th meeting of the American college of Rheumatology Arthritis Rheum 2002; 46(Suppl.): 9 (abstract 1299). 10 Zaja F, Russo D, Fuga G, Perella G, Baccarani M. Rituximab for myasthenia gravis developing after bone marrow transplant. Neurology 2000; 55: Edwards JCW, Leandro MJ, Cambridge G. B-lymphocyte depletion therapy in rheumatoid arthritis and other autoimmune disorders. Biochem Soc Trans 2002; 30: Piro LD, White CA, Grillo-Lopez AJ et al. Extended rituximab (anti-cd20 monoclonal antibody) therapy for relapsed or refractory low-grade or follicular non-hodgkin s lymphoma. Ann Oncol 1999; 10: Specks U, Fervenza FC, McDonald TJ, Hogan MC. Response of Wegener s granulomatosis to anti-cd20 chimeric monoclonal antibody therapy. Arthritis Rheum 2001; 44: Jenette JC, Falk R, Andrassy K et al. Nomenclature of systemic vasculitides. Proposal of an international consensus conference. Arthritis Rheum 1994; 37: Luqmani R, Exley A, Kitas G, Bacon P. Disease assessment and management of the vasculitides. Bailliere s Clin Rheumatol 1997; 11: Jayne D, Rasmussen N, Andrassy K et al. A randomized trial of maintenance therapy for vasculitis associated with antineutrophil cytoplasmic autoantibodies. N Engl J Med 2003; 349: Keogh KA, Wilam ME, Specks U. Rituximab a potential mechanistic-based therapy for treatment of refractory ANCAassociated vasculitis. Kidney Blood Press Res 2003; 26: Jayne DRW, Burns S, Smith K. B-cell depletion with rituximab for refractory vasculitis. Kidney Blood Press Res 2003; 26: Anolik JH, Campbell D, Felgar RE et al. The relationship of FcgammaRIIIa genotype to degree of B cell depletion by rituximab in the treatment of systemic lupus erythematosus. Arthritis Rheum 2003; 48: Cambridge G, Leandro MJ, Edwards JC et al. Serologic changes following B lymphocyte depletion therapy for rheumatoid arthritis. Arthritis Rheum 2003; 48: van der Kolk LE, Baars JW, Prins MH, van Oers MH. Rituximab treatment results in impaired secondary humoral immune responsiveness. Blood 2002; 100: Looney RJ. Treating human autoimmune disease by depleting B cells. Ann Rheum Dis 2002; 61: Dörner T, Burmester G. The role of B cells in rheumatoid arthritis: mechanisms and therapeutic targets. Curr Opin Rheumatol 2003; 15: Correspondence: Per Eriksson, Department of Rheumatology, University Hospital, Linköping, Sweden. (fax: ; per.eriksson@lio.se).