B Cells As Therapeutic Targets In Autoimmune Neurological Disorders

Marinos C Dalakas


Nat Clin Pract Neurol. 2008;4(10):557-567. 

In This Article

Anti B-cell Agents in Neurology: The Role of Rituximab

Among all the aforementioned agents, only two, BCMA-IgG and rituximab, have been used in autoimmune neurological disorders. In mice with experimental autoimmune encephalomyelitis induced by myelin oligodendrocyte glycoprotein (MOG), BCMA-IgG prevented disease development, improved the strength of already weak animals, depleted the CD19+ B cells in the blood, spleen and lymph nodes, reduced anti-MOG-specific IgG antibody titers, and suppressed inflammation and ongoing demyelination in the brain and spinal cord.[25]

Rituximab is a chimeric mouse-human monoclonal antibody consisting of human IgG1 and kappa constant regions and a mouse variable region. It was derived from a hybridoma directed at human CD20, a 297-amino-acid transmembrane phosphoprotein that is present on all cells of the B-cell lineage except for stem cells, pro-B cells and plasma cells (Figure 2).[1,2,3,4,5,6] In contrast to BAFF and APRIL, CD20 is not secreted, and it is not shed or endocytosed when exposed to rituximab.[26] The function of CD20 is unclear—it is thought to be involved in B-cell activation and proliferation,[1,2,3,4,5,6] although CD20 knockout mice do not exhibit B-cell deficits.[1,2,3,4,5,6,26] Rituximab is approved for the treatment of rheumatoid arthritis,[27] and, as outlined in the sections that follow, its use is currently being explored in a number of autoimmune neurological disorders in which B cells have a role.

Multiple Sclerosis. In MS, B cells and antibodies are involved to varying degrees at different stages of the disease and in different subgroups of MS. In the I-IV classification of Lucchineti, for example, pattern II is characterized by prominent lymphocytic and macrophagic infiltrates, complement activation and deposits of immunoglobulins.[28] Additional evidence supporting an antibody-mediated process in patients with MS includes accumulations of clonally expanded B cells in the MS plaques; intrathecal production of IgG bands from oligoclonal populations of B cells; autoantibodies against MOG in actively demyelinating lesions; ectopic lymphoid tissue in the intermeningeal spaces; and upregulation of BAFF and APRIL.[17,21,22,28,29,30,31]

In patients with MS, 24 weeks of treatment with rituximab was shown to deplete B cells from the cerebrospinal fluid (CSF) and to suppress B-cell activation, but it did not affect the intrathecal synthesis of oligoclonal IgG bands derived from long-lived plasma cells.[32] In a phase II, controlled, multicenter clinical trial of 104 patients with relapsing-remitting MS, a 58% relative reduction in the proportion of patients who experienced a relapse was noted after 24 weeks of therapy. A significant reduction in the mean number of gadolinium-enhancing MRI lesions (the study's primary end point), was also observed in the treated patients compared with the placebo group (P <0.0001).[33] In an open-label trial of 26 patients treated with two courses of rituximab 6 months apart and followed for up to 72 weeks after commencement of treatment, reductions in relapses and MRI lesions were noted, suggesting long-lasting benefit.[34] These very encouraging results and the excellent tolerance of the drug has led to phase III trials in relapsing-remitting and primary progressive forms of MS, which are currently ongoing.

Neuromyelitis Optica. Neuromyelitis optica (NMO) is an inflammatory CNS disorder that affects the optic nerves and the spinal cord. It typically presents with myelitis and optic neuritis, and is characterized by varying degrees of sensory motor disturbances, bladder-bowel dysfunction and visual loss. In NMO, autoantibodies, collectively termed NMO-Ig, bind to cerebral microvessels.[35] The main target antigen of these autoantibodies is the aquaporin-4 water channel. NMO-Ig is derived from peripheral B cells, activates complement, and has been implicated in the induction of inflammatory demyelination and necrosis in the endothelial cells of the spinal cord.[35] Patients with NMO experience frequent relapses, and the disease is associated with high morbidity. Some acute flare-ups can respond to plasmapheresis, although the disease responds poorly to immunotherapies overall.

In an open-label study, six out of eight patients with NMO became relapse-free after a year of rituximab treatment, with a decline in relapse rate from 26 to zero attacks per year. In addition, seven of the patients showed a substantial improvement in their Expanded Disability Status Scale (EDSS) score.[36] In a retrospective review of 34 patients from two different centers, rituximab significantly lowered the relapse rate compared with pretreatment data, and stabilized or improved the EDSS scores in 91% of the patients.[37,38]

Paraneoplastic Neurological Disorders. Patients with paraneoplastic neurological disorders have circulating antibodies against a variety of antigens that are expressed in both brain and cancer cells. There is evidence that B cells, plasma cells and cytotoxic T cells cross the blood-brain barrier, and antibodies are synthesized intrathecally. In paraneoplastic opsoclonus-myoclonus, the number of clonally expanded B cells within the CSF correlates with clinical severity.[39] Rituximab, as an add-on therapy to IVIg or adrenocorticotropic hormone (ACTH), improved the ataxia severity scores, ameliorated myoclonus and reduced the rate of clinical relapse in 81% of 16 children with opsoclonus-myoclonus, and selectively reduced the numbers of clonally expanded B cells in the CSF.[40] In addition, the required ACTH dose was reduced by 51% after rituximab treatment.

Chronic Autoimmune Neuropathies. The chronic autoimmune neuropathies include a spectrum of predominantly demyelinating neuropathies, the most common of which are chronic inflammatory demyelinating neuropathy (CIDP), multifocal motor neuropathy (MMN) and IgM anti-myelin-associated glycoprotein (IgM-MAG) neuropathies. Evidence for a role for B cells in the pathogenesis of these conditions includes the deposition of immunoglobulins and complement on the patients' nerves, and the presence of complement-fixing antibodies against MAG and gangliosides.[41,42] In an open series of 21 patients with IgM antibodies to gangliosides, rituximab improved symptoms in 61% of the patients 6 months after therapy, and the benefits were maintained for up to 2 years with repeated infusions.[43] The IgM antibody titers dropped by 36% in the first year and by 57% in the second. Rituximab has also been reported to be effective in some patients with MMN or CIDP.[44,45] In another study, rituximab improved the symptoms in six out of nine patients with IgM-MAG neuropathy, and reduced IgM-MAG titers by a mean of 52% from baseline,[46] prompting a placebo-controlled study. In 26 randomized patients, rituximab significantly improved disability scores and reduced IgM-MAG titers after 8 months.[47] Rituximab is the first drug to demonstrate efficacy in a randomized trial in this particular neuropathy.

Stiff-person Syndrome. Stiff-person syndrome (SPS) is a rare but often misdiagnosed CNS disorder that is clinically characterized by stiffness and rigidity in the limbs and paraspinal muscles, intermittent superimposed muscle spasms, and heightened sensitivity to external stimuli. The majority of patients with SPS have antibodies against glutamic acid decarboxylase (GAD) or GABARAP, a linker protein responsible for γ-aminobutyric acid receptor clustering.[48] Anti-GAD antibodies are synthesized intrathecally, and oligoclonal bands are commonly detected in the CSF.[49] In one case report, rituximab was effective at reducing stiffness and increasing mobility 2 months after the treatment was initiated, and resulted in disappearance of GAD antibodies and normalization of the electromyogram.[50] The first double-blind controlled study using rituximab to treat SPS has now been completed at the NIH, and the results are currently being analyzed.

Inflammatory Myopathies. There are three main subsets of inflammatory myopathies: polymyositis, dermatomyositis and inclusion body myositis. In all these forms of the disease, B cells and plasma cells are present in the muscle tissues, and in dermatomyositis immunoglobulins are deposited on endomysial capillaries.[51] In 10 patients with polymyositis or dermatomyositis who had responded poorly to current therapies, rituximab increased or normalized muscle strength in 8 cases. Serum levels of creatine kinase and the required prednisone dose were concurrently reduced.[52,53] A multicenter NIH-sponsored clinical trial of rituximab is now ongoing for the treatment of polymyositis and dermatomyositis. The drug has not yet been tested in inclusion body myositis.

Myasthenia Gravis. Myasthenia gravis is a prototypic B-cell-mediated autoimmune disease caused by pathogenetic antibodies against the muscle acetylcholine receptors. Evidence from around 20 case reports suggests that rituximab is effective in most patients, but a controlled study has not yet been done.[54,55,56,57]

In general, 1 month after rituximab infusion, circulating B cells become undetectable, and their numbers remain low for at least 6 months. The cells start reappearing slowly thereafter, but even after 10 months their numbers remain below baseline.[3] The circulating memory CD20+CD27+ B cells are also depleted, and their levels remain low until month 8 (Figure 5). The B cells in the follicular splenic regions are preferentially affected, being depleted by 90%, compared with 25% depletion of marginal-zone B cells.[6] The germinal-center B cells are resistant to rituximab, even though they express CD20, possibly reflecting an inability of the antibody to access the intravascular spaces within the lymphoid tissues, or different sensitivities of B cells according to the local lymphoid microenvironment.[6] Stem cells in the bone marrow that do not express CD20 are also spared, thereby allowing the generation of new naive B cells.[58]

Figure 5.

Kinetics of CD20+CD27+ memory B cells in rituximab-treated patients. The figure highlights several B-cell molecules and their receptors, which are targeted by nine different monoclonal antibodies or fusion proteins currently in phase I-III clinical trials. Abbreviations: APRIL = a proliferation-inducing ligand; BAFF = B-cell-activating factor; BAFF-R = B-cell-activating factor receptor; BCMA = B-cell-maturation antigen; CTLA4 = cytotoxic T-lymphocyte antigen 4; LTβR = lymphotoxin-β receptor; LTβR-Ig = anti-lymphotoxin-β receptor antibody; MHC-II = major histocompatibility complex class II; TACI = transmembrane activator and calcium modulator and cyclophilin ligand interactor; TCR = T-cell receptor.

Rituximab is not expected to affect the levels of antibodies produced by plasma cells, although some reductions in these levels have been noted. In rheumatoid arthritis, for example, the titers of rheumatoid factor were shown to decrease two-to-threefold,[58,59] and in IgM-MAG neuropathy by 30-50%, after treatment with rituximab.[45,47] Such reductions can probably be attributed to depletion of CD27+ memory B cells, the precursors of short-lived plasma cells.[4] As the CD27+ memory B cells reappear, so do the short-lived plasma cells.[60] Given that the reconstituting B cells are naive cells with a new and diverse immunoglobulin rearrangement pattern,[4,58,59] it might take some time for them to be restimulated by the original antigen, hence the slow re-emergence of serum antibody titers. Antibody titers might therefore fall after rituximab treatment, and rebound slowly at a rate controlled by the replenishment of memory and short-lived plasma cells.[60] Interestingly, after several years of treatment, the antibody titers against anamnestic antigens, such as tetanus toxoid, remain stable.[59] This finding suggests that rituximab might have differential effects on 'autoreactive' B cells and their corresponding short-lived plasma cells, compared with 'non-self-reactive' B cells and their corresponding longer lived plasma cells, which are responsible for post-vaccination responses.[6] A recent study supports different roles for B cells and longer lived plasma cells in protective immunity.[61]

Rituximab can be administered intravenously at a dose of 375 mg/m2, given weekly for 4 weeks, or in two 1 g infusions, given at fortnightly intervals (total 2 g). The average half-life of the drug after completion of an infusion is 21 days. The infusions can be repeated after 6-12 months, at a point when B cells start rebounding or when the patient has relapsed. The drug is generally very well tolerated, although mild hypotension can be observed in some patients, necessitating the discontinuation of antihypertensive drugs on the day of the infusions. Anaphylactic or skin reactions can occur in rare cases, but these respond to intravenous methylprednisolone. Premedication with antihistamine is desirable to prevent the occurrence of such reactions.

Rituximab has been used in combination with other immunosuppressants, such as corticosteroids, mycophenolate, cyclophosphamide, azathioprine or methotrexate,[6,62] for the treatment of vasculitis and rheumatoid arthritis, without additional complications. This experience differs from that with natalizumab, which requires discontinuation of the other immunosuppressants for 2-3 months before initiating therapy.[63] It remains to be determined whether combination therapy will be more effective than monotherapy in difficult neurological cases. Rituximab has been also used effectively in some cases of pediatric SLE in two infusions of 750 mg/m2 administered 2 weeks apart, either alone or in combination with corticosteroids and cyclophosphamide,[1] suggesting that it can be used in children with difficult autoimmune neurological disorders.

The resistance of long-lived plasma cells to rituximab probably explains its excellent safety profile, the absence of infections, and the patients' retained ability to produce immunoglobulins and mount an antibody response against anamnestic antigens. In spite of the apparent plasticity of the immune system, which enables it to compensate for the peripherally depleted B cells, vigilance is still required to guard against the possibility of infections in patients receiving repeated doses or concurrent immunosuppressants.[3] Such infections might not be limited to common bacterial or viral agents, but might also extend to agents that cause latent infections, such as JC virus or herpesviruses, as has been experienced with natalizumab.[63] Rare cases of progressive multifocal leukoencephalopathy have been reported in patients with SLE receiving rituximab, although a cause-and-effect relationship has not been established.[64]

A consistent observation in many series is the elevation of BAFF levels after rituximab treatment, probably as an inherent compensatory mechanism to drive B-cell production.[1,2,3,4,5,6] Theoretically, combining rituximab with one of the agents against BCMA or TACI-IgG, which reduce the survival of BAFF-dependent, immunoglobulin-producing, long-lived plasma cells, might have a prolonged effect on B cells and autoantibody levels. Such a combination therapy might be attractive in the future, in view of increased levels of BAFF in autoimmune neurological disorders.

Rituximab depletes B cells through three mechanisms (Figure 6): antibody-dependent cellular cytotoxicity, whereby antibody-coated cells bind to the Fc receptors of macrophages or natural killer cells; activating the membrane attack complex on B cells (complement-dependent cytotoxicity); and inducing apoptosis by changing the lipid raft environment on the CD20+ B-cell membrane.[1,2,3,4,5,6]

Figure 6.

Mechanisms of action of rituximab. Rituximab induces cell death through three mechanisms. (A) Antibody-dependent cell-mediated-cytotoxicity. Rituximab recruits macrophages and natural killer cells by binding to their Fcγ receptors. (B) Complement-mediated cytotoxicity. Rituximab activates complement and generates membrane attack complexes. (C) Induction of apoptosis.

The most impressive observation in all published series, and from our own experience, is the long-lasting benefit of rituximab, sometimes exceeding 6-8 months after therapy. The degree of clinical response, however, varies from patient to patient, probably reflecting the varying degree of contribution of B cells to the autoimmune process, as discussed earlier. It is difficult to ascertain which of the B-cell functions depicted in Figure 1 is primarily influenced by the drug and is responsible for the noted benefit. Diminished production of pathogenetic autoantibodies might be a contributing factor, but this decrease in production might be insufficient to be clinically meaningful.[27,59] The effects of rituximab on other B-cell functions—effects that include blockade of costimulatory molecules required for clonal expansion of T cells, inhibition of the antigen-presenting role of B cells, suppression of the cytokine network, inhibition of immune complexes, and induction of immunoregulatory T cells—might be more important in explaining the noted clinical benefit. Accordingly, rituximab might be beneficial not only in antibody-mediated disorders of the CNS and PNS, but also in other autoimmune diseases where both B cells and T cells contribute to disease pathogenesis.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.