Felty’s syndrome (FS) is defined by the coexistence of rheumatoid arthritis (RA), neutropenia, and splenomegaly. The mechanisms underlying the neutropenia of FS may involve both cellular and humoral immunity, with a possible role of granulocyte-colony stimulating factor (G-CSF) antibodies.¹ Various disease modifying antirheumatic drugs have been used to treat FS, but with varying success² as this syndrome may arise in response to the excessive immune reaction found in RA. Interest has focused recently on a new biological tool in the treatment of RA, rituximab, a chimeric monoclonal antibody specific for human CD20 which targets B lymphocytes.³ Accordingly, we investigated here the safety and efficacy of rituximab in two patients presenting with active RA and severe and refractory FS.

METHODS AND RESULTS

Two men, were studied, aged 67 (patient 1) and 53 (patient 2) years, with a duration of RA of 6 and 11 years, respectively. FS had been diagnosed respectively 5 and 3 years ago, and RA remained active in both patients despite corticotherapy and respectively one (sulfasalazine) and two (sulfasalazine and methotrexate) previous disease modifying antirheumatic drugs. Anti-tumour necrosis factor treatment was not used because of neutropenia and the risk of severe infection. The absolute neutrophil count was persistently less than 0.8×10⁹/l and complicated with recurrent sinopulmonary infections. There was no suggestion of congenital hypogammaglobulinaemia and, in particular, no sign of selective IgG2 immunodeficiency. Blood and bone marrow immunophenotyping did not disclose any features of myelodysplasia or lymphoproliferation, or any large granular lymphocytes. No other classical cause of neutropenia, such as toxicity, chronic infection, vitamin deficiency, or liver disease, was present. Anti-G-CSF (IgG) antibodies, which were determined by enzyme linked immunosorbent assay (ELISA),¹ were detected in one patient without previous administration of haematopoietic factor (G-CSF).

Owing to the presence of refractory RA associated with severe FS, rituximab was administered as an intravenous infusion at a dose of 375 mg/m² once weekly for 4 weeks. Concomitant treatment consisted of prednisone (15–20 mg/day) for more than 12 months in both patients and methotrexate (20 mg/week) since March 2003 in patient 2. The duration of follow up was 6 months. Rituximab was well tolerated and efficiently controlled the clinical and biological activity of RA in patient 2, who fulfilled the American College of Rheumatology 50 response criteria and showed a marked decrease in serum levels of rheumatoid factor. However, results for FS were disappointing, because no increase in neutrophil count or modification of infection rates could be detected (table 1). In patient 1, a decrease in neutrophil count was observed at week 12, but without any clinical anomaly. Biological controls showed no modification of levels of anti-G-CSF antibodies, no appearance of anti-granulocyte antibodies, and no large granular lymphocyte proliferation.⁴

DISCUSSION

Several factors might account for the lack of efficacy of rituximab in the treatment of FS. Firstly, although different autoreactive B cells may be involved in the pathology of FS, the inability of rituximab to bind to plasma cells, which are CD20 negative, might prevent it from acting on FS. Nevertheless, the efficacy of rituximab in certain conditions associated with autoantibodies is not correlated with a reduction of these antibodies, which would suggest that in addition to autoantibody production, other roles of B cells (immunoglobulins, antigen presentation, T cell cooperation) are important in the pathogenesis of such diseases.³ Secondly, a subpopulation of T lymphocytes having an antigranulocyte activity may exist independently of B cells in some forms of FS.⁵

In conclusion, the lack of efficacy of rituximab in these two patients with FS raises some important questions about the mechanisms responsible for FS and the best therapeutic strategy to adopt.