SEARCH

SEARCH BY CITATION

For proliferative lupus nephritis, intravenous cyclophosphamide and corticosteroids have been validated to be the current standard of therapeutic care (for review, see ref. 1). However, toxicities are not only common, but anticipated. These antimetabolic treatments induce the death of disease-associated lymphocytes and mononuclear cells, but other proliferating cells throughout the body are also affected.

All too often, clinicians are faced with the familiar scenario of a patient with systemic lupus erythematosus (SLE) who has recently received aggressive treatment for newly diagnosed nephritis, and now must be admitted to the hospital for pancytopenia, fever, and obvious systemic infection. Even though in most cases the patient will recover with appropriate antibiotic treatment, infection remains a leading cause of mortality, and therapeutic immunosuppressive agents are primary contributing factors (2).

Over the last 2 decades, accelerating advances in molecular and cellular immunology have brought the hope of finding new ways to interfere with the specific pathways responsible for these pathologic autoimmune responses without impeding life-preserving immune defenses. The pathophysiology of SLE is no doubt complex, and recent genetic surveys have implicated several chromosomal loci and candidate genes as contributing to disease predisposition. Clinical studies and murine models of disease have documented both interferon-α overexpression and antigen-presenting cell–T lymphocyte interactions in pathogenesis. However, in all lupus investigations, B cell hyperactivity and autoantibody production have been consistent features (3). Therefore, as an alternative to current conventional pharmaceutical agents, which unfortunately also impair primary mucosal barriers to microbial invasion and cause generalized weakening of microbial resistance, therapeutic approaches have been developed to specifically target the B cell compartment without disturbing other aspects of the immune system.

More than one therapeutic approach to impair and delete B lymphocytes is currently being explored. Although not the topic of the current commentary, there is emerging evidence that an unexpected outcome of treatments with protein A pheresis columns is the concurrent infusion of this natural bacterial toxin into the patient, which results in the targeted deletion, through antigen receptors, of certain B cells (4). There has also been encouraging progress leading to clinical trials of therapeutic blockade of ligand–receptor interactions for BAFF (also known as BLyS, TALL-1, THANK, and zTNF4), which plays an important and specific role in the survival of (autoreactive) B cells (5). Clinical benefits in SLE have also been reported from recent trials of antagonists of CD40–CD40 ligand (CD40L) interaction (6), although there have also been problems of toxicity from thrombosis in some treated patients. Antibodies to CD22, a marker on mature B lymphocytes, can also be used to induce the targeted deletion of B cells, and it is currently being evaluated in clinical trials in patients with non-Hodgkin's lymphoma (NHL) (7).

More extensive clinical experience has accumulated with a therapeutic monoclonal antibody to CD20 that was designed to affect only B-lineage cells. This pan–B cell marker is a membrane-associated glycoprotein that is highly expressed on the surface of pre-B and both resting and activated mature B lymphocytes, but not on hematopoietic stem cells, pro-B cells, or other normal tissues. CD20 surface expression diminishes, and may be undetectable on memory B cells and plasma cells, the end-differentiated antibody-producing factories of the body. Based on structural homologies, CD20 has been postulated to act as a calcium channel (for review, see ref. 8).

Despite extensive investigation, the role of CD20 in B cell physiology remains poorly understood. There is no known natural ligand, and CD20-knockout mice have no discernible immunologic phenotype (9). Providing an advantage for its use as a therapeutic target, binding of CD20 does not modulate its expression or result in substantial internalization. CD20 is also not shed, and there are no other known analogs that interfere with its use for B cell targeting (10). Most importantly, treatment with an antibody to CD20 induces the death of B lymphocytes, even without the need to conjugate the antibody to a toxin.

Best explored of the anti-CD20 antibodies developed for therapeutic applications is rituximab, a humanized chimeric monoclonal antibody specific for human CD20, with the variable regions of a murine anti-human CD20 B cell hybridoma fused to human IgG and κ constant regions (11). In 1997, rituximab became the first antibody to receive Food and Drug Administration approval for the treatment of cancer. After more than 300,000 treatments, it has become an accepted standard for therapeutic efficacy and safety in NHL.

The potential therapeutic opportunities with rituximab appear great (12), but the cumulative clinical experience outside of B cell malignancies is still limited. While much less is known about how this approach may affect an established autoimmune disease, there is increasing interest in exploring the utility of rituximab for the treatment of SLE.

Recent experience with rituximab in SLE

  1. Top of page
  2. Recent experience with rituximab in SLE
  3. Novel insights from the open-label trial for lupus nephritis
  4. Emerging lessons regarding the clinical benefits of anti-CD20 therapy
  5. Can mechanisms of disease interfere with the mechanism(s) of anti-CD20 action?
  6. Is response to anti-CD20–mediated deletion influenced by prosurvival factors?
  7. REFERENCES

The literature now includes an increasing number of case reports describing efficacy of rituximab in the treatment of patients with SLE (for review, see ref. 13). There have also been 2 recent small open-label trials involving lupus patients who displayed great heterogeneity in disease severity and organ involvement. In a recent report of a phase I/II trial from the University of Rochester, Looney et al described the responses of 17 lupus patients, including 7 with proliferative glomerulonephritis (14). The study included 3 groups of patients who received different doses of rituximab (along with relatively modest doses of corticosteroids as cotreatment). Outcome analysis showed that the low-dose regimen of rituximab was suboptimal in that it resulted in only limited blood B cell depletion and lessened clinical benefits. However, the patients who received high-dose infusions, equivalent to the levels used for NHL patients, also had great variability in their measured posttreatment rituximab levels. Limited blood B cell depletion, which correlated with the inheritance of the low-affinity allele of Fcγ receptor IIIa (FcγRIIIa) (15), was more common in African American patients, who also more commonly exhibited suboptimal serum rituximab levels and poor clinical responses (16).

In Looney and colleagues' study (14), there was no general trend of induced reductions of serum titers of autoantibodies to double-stranded DNA (dsDNA). However, of the patients who exhibited clinical benefits from treatment, there were 4 with high pretreatment levels of these autoantibodies who also experienced effective B cell depletion, and treatment was associated with reductions in their levels of anti-dsDNA autoantibodies that persisted for at least 12 months (16). Human antichimeric antibodies, a now frequently observed complication that may potentially interfere with B cell depletion by rituximab and/or contribute to undesirable adverse events such as hypersensitivity reactions, were detected in 6 of these 17 SLE patients. Human antichimeric antibodies were especially frequent in those receiving the lower-dose regimen and those with poor clinical responses.

In a recent open-label British trial (17), the 6 reported lupus patients were also heterogeneous in terms of organ involvement, and only 3 had proliferative glomerulonephritis. All of these patients received rituximab at a dosage comparable with that used for NHL, but they also received cotreatment with intravenous cyclophosphamide and high-dose oral corticosteroids. Even though these additions to the treatment regimen clouded interpretation of the benefits that derived from the anti-CD20 agent, this regimen appeared to provide more uniform clinical benefits, and there was the common finding of decreases in circulating levels of autoantibodies to dsDNA.

Common to these 2 open-label studies, outcome assessments required the application of complex disease activity measures, i.e., the SLE Disease Activity Index (18) and the British Isles Lupus Assessment Group index (19), respectively, because the lupus patients displayed a diverse range of clinical manifestations. Despite this potential confounding factor, these studies have contributed to a growing optimism, since regimens that included full doses of this agent for CD20-targeted B cell depletion appeared to provide significant benefits in most of the SLE patients studied.

Novel insights from the open-label trial for lupus nephritis

  1. Top of page
  2. Recent experience with rituximab in SLE
  3. Novel insights from the open-label trial for lupus nephritis
  4. Emerging lessons regarding the clinical benefits of anti-CD20 therapy
  5. Can mechanisms of disease interfere with the mechanism(s) of anti-CD20 action?
  6. Is response to anti-CD20–mediated deletion influenced by prosurvival factors?
  7. REFERENCES

In this issue of Arthritis & Rheumatism, Sfikakis et al report on an open-label study of the clinical outcome of an anti-CD20 therapeutic regimen in 10 SLE patients with biopsy-proven class III or IV proliferative glomerulonephritis (20). Compared with the above-described investigations, the patients studied by Sfikakis and colleagues were, by design, potentially more comparable in their clinical features, which enabled the use of more clearly defined and better-validated outcome measures of disease activity.

During the 12-month followup period after treatment with standard doses of rituximab (4 weekly infusions of 375 mg/m2) combined with initial daily oral doses of 0.5 mg/kg prednisolone, 4 of 4 patients with class III nephritis and 4 of 6 with class IV nephritis achieved the primary efficacy end point of the study, partial remission defined by measures of renal function, proteinuria, and urinary sediment. Importantly, these partial remissions were achieved within 1–4 months, and in 5 of these patients complete remission was achieved by 2–8 months, which represents substantially quicker responses than are reported with the standard regimen of cyclophosphamide and corticosteroids (1). Moreover, the activity of other disease manifestations, including rash and thrombocytopenia, paralleled the clinical course of nephritis.

These treatments were also generally well tolerated: as in other reported trials in both autoimmune disease and NHL (for review, see ref. 13), most patients had no severe adverse events associated with the rituximab infusions. However, 1 of the lupus patients did develop a hypersensitivity reaction, manifested by rash and fever, which precluded completion of the rituximab regimen. Four patients also experienced infections, although the patients' complement-deficient states were deemed contributory.

In these treated patients, there was the anticipated >99% depletion of total blood B lymphocytes, which lasted a median of 5 months, while the 2 patients without clinical responses had shorter periods of depletion. Of the 5 patients with complete clinical remission, these clinical responses were sustained in 4 after B cell regeneration and lasted the full 12-month followup period. These findings therefore suggest that total blood B cell levels in treated patients do not accurately reflect the overall immunologic impact of therapy.

In the study by Sfikakis et al, also in accordance with earlier reports, the levels of total IgG and IgA were not affected by this intervention. However, total IgM levels were modestly decreased, especially in patients in whom complete remission was achieved. These findings suggest that the specific B-lineage cells responsible for the production of these different isotypes may vary in their sensitivity to rituximab-induced deletion. Speculatively, the cells that secrete IgM may be more susceptible to rituximab and/or may have more rapid turnover without subsequent replacement because their precursor cells have been deleted. Also consistent with the notion of a beneficial impact on immune complex–mediated disease, both partial and complete remissions were associated with increases in serum levels of complement factors, while decreases in complement factors correlated with treatment failure and disease recurrence after B cell regeneration.

In all patients, treatment resulted in decreases in levels of detectable serum autoantibodies to dsDNA, although these antibodies never became undetectable in any patient. However, only values for combined IgG and IgM anti-dsDNA levels were reported. While it is commonly accepted that a subset of IgG anti-dsDNA antibodies has nephritogenic potential, there is little evidence that IgM anti dsDNA autoantibodies contribute to organ damage. Moreover, not all nephritogenic autoantibodies exhibit anti-dsDNA binding activity (21), and results of mouse studies have suggested that assays that detect autoantibodies reactive with glomerular extracts may be more specific for organ-specific pathology (22). Therefore, although Sfikakis et al indicate that autoantibody levels did not correlate with clinical outcome, determinations of isotype-specific autoantibody levels might have been more informative.

While the clinical results of Sfikakis and colleagues' study are impressive, they provide little new insight into the effect of rituximab treatment on the B cell compartment, believed by many to be central to lupus pathogenesis. In fact, the effects of rituximab treatment on the lymphoid sites responsible for immunopathogenesis have not been systematically investigated. At the present time, the literature contains a single report of a case in which an SLE patient, treated with rituximab for thrombocytopenia, was shown to have both blood B cell depletion and depletion of splenic B cells at 1 month after treatment (23). However, unfortunately in this case the splenectomy was performed due to inadequate clinical response, which also suggested that the cellular impact of this therapeutic intervention is complex.

The study by Sfikakis et al did include flow cytometric surveys of blood lymphocyte subsets, aside from B cells, which proved highly informative. Concurrent with evidence of induced blood B cell depletion, there were modest but significant decreases of circulating natural killer T cells. In addition, during the months after treatment, 7 of the 10 patients exhibited overall increases in CD8+ T cells that correlated with partial remissions, and there were further increases in patients in whom complete remission was achieved. While the overall frequency of CD25+ (i.e., interleukin-2 receptor) CD4 T cells also increased with treatment, this surface phenotype alone does not accurately identify regulatory T cells, so the significance is not currently clear.

It has been reported that patients with active SLE overexpress CD40L (24), and CD40L–CD40 interactions are believed to play central roles in B–T cell collaborations that lead to germinal center reactions that promote the pathologic autoimmune responses (25). Therefore, it is notable that the laboratory studies conducted by Sfikakis et al demonstrated consistent decreases in levels of circulating CD40L-bearing CD4 T cells, which attained a mean 4-fold reduction shortly after treatment. Moreover, because CD20 is restricted to B-lineage cells, rituximab and/or corticosteroid infusions would not be anticipated to directly effect CD40L levels on T cells. Their studies also showed that levels of the activation markers CD69 and HLA–DR4 on CD4 T cells were consistently decreased in patients with partial-to-complete remission. While the authors comment that not all T cell alterations arising at the time of the B cell depletion were predictive of clinical outcome, these findings may still indicate that immunopathogenesis was affected in all treated patients, even though in some a threshold that correlated with detectable clinical benefits was not attained.

Emerging lessons regarding the clinical benefits of anti-CD20 therapy

  1. Top of page
  2. Recent experience with rituximab in SLE
  3. Novel insights from the open-label trial for lupus nephritis
  4. Emerging lessons regarding the clinical benefits of anti-CD20 therapy
  5. Can mechanisms of disease interfere with the mechanism(s) of anti-CD20 action?
  6. Is response to anti-CD20–mediated deletion influenced by prosurvival factors?
  7. REFERENCES

These clinical findings support the relevance of observations from murine models, which have suggested that autoreactive B cells contribute to lupus pathogenesis by mechanisms beyond the production of pathogenic autoantibodies (for review, see ref. 26). In studies designed to investigate the mechanisms responsible for the therapeutic benefits, rituximab treatment of immunodeficient mice implanted with rheumatoid synovial explants was shown to greatly reduce T cell production of proinflammatory cytokines (27), presumably because there was a deletion of the B cells that acted as antigen-presenting cells and costimulators for autoreactive T cells involved in pathogenesis. Such benefits of rituximab were also implicated in the significant clinical responses documented in a recent multi-arm controlled study of its use in seropositive rheumatoid arthritis (RA) (28).

In recent years there has been increasing interest in the development of biomarkers, involving cellular or serologic assays, to complement available laboratory and clinical measures of disease activity. For obvious reasons, there has been interest in blood profiles of remaining B cell subsets, particularly plasma cell and memory cells, with the hope of better understanding the variations in response to treatment of the underlying immunopathobiology. By providing the first evidence that T cell activation markers, like the levels of CD40L on CD4 T cells, can be used as biomarkers for such B cell–targeted interventions, the study by Sfikakis et al may have identified a better means to assay disease activity. However, additional studies will be needed to evaluate and validate the utility of such tests as surrogate end points (29).

Can mechanisms of disease interfere with the mechanism(s) of anti-CD20 action?

  1. Top of page
  2. Recent experience with rituximab in SLE
  3. Novel insights from the open-label trial for lupus nephritis
  4. Emerging lessons regarding the clinical benefits of anti-CD20 therapy
  5. Can mechanisms of disease interfere with the mechanism(s) of anti-CD20 action?
  6. Is response to anti-CD20–mediated deletion influenced by prosurvival factors?
  7. REFERENCES

In different systems, rituximab has been shown to delete B cells by reliance on at least 3 different cellular mechanisms: B cell apoptosis can be induced by hyper-crosslinking of membrane-associated CD20 molecules (8); there is also evidence that rituximab can induce complement-dependent cytotoxicity (8); and B cells coated with anti-CD20 antibodies may be cleared by antibody-dependent cell-mediated cytotoxicity via interactions with cellular receptors for the IgG1 constant regions (i.e., FcγR), especially FcγRIIIa, which is expressed on a variety of cells, including phagocytic cells (30). However, most reported mechanistic studies have used simplified in vitro systems with cell lines that may not be representative of in vivo responses. Moreover, there is also evidence that clinical responses of different B cell malignancies to rituximab vary in their dependence on specific antibody-mediated deletional pathways. These insights appear relevant when considering how rituximab may affect patients with autoimmune disease who have abnormalities in the same pathways.

Inherited complement deficiencies are strong risk factors for human and murine SLE. In fact, human C1q deficiency is reported to have near 100% penetrance for lupus, and acquired complement deficiency states may be even more common in SLE (31). Moreover, complement has also been implicated in rituximab-induced B cell deletion (for review, see ref. 32), and C1q-deficient mice have displayed impairment of rituximab-induced depletion of human CD20-transduced syngeneic lymphoma cells (33). Hence, the evidence suggests that inherited or acquired deficiencies in complement, which contribute to autoimmunity and dysregulated autoantibody production, can also interfere with the efficiency of rituximab-mediated B cell deletion.

FcγR have also been shown to provide a central interface between the adaptive and innate immune systems, which allows specific IgG responses to recruit proinflammatory mononuclear cells into immune responses. Defects in the regulation of these responses can also be factors predisposing to autoimmune disease. In fact, the inheritance of the above-mentioned FcγRIIIa (F176) low-affinity allele is also reported to be a strong risk factor for the development of SLE and nephritis across many ethnic populations (34, 35). It is therefore relevant that, as discussed above, inefficient B cell depletion seen in a subset of rituximab-treated lupus patients correlated with the inheritance of the FcγRIIIa genotype associated with lower-affinity binding interactions (15).

These perspectives are important when considering why the response profile in the trial described by Sfikakis and colleagues may appear better than those described in earlier reports. Because the patients studied by Sfikakis et al were all Caucasian ethnic Greeks, it is possible that they were more homogeneous in their genetic inheritance of factors that may affect clinical response patterns, a notion also suggested by the findings, by investigators at the same institution, in recent comparative studies of patients with Sjögren's syndrome (36). Hence, the impressive rates of response in these Greek patients may reflect their good fortune to have not inherited immunogenetic profiles that interfered with the mechanism(s) of action of this B cell–deleting agent.

Is response to anti-CD20–mediated deletion influenced by prosurvival factors?

  1. Top of page
  2. Recent experience with rituximab in SLE
  3. Novel insights from the open-label trial for lupus nephritis
  4. Emerging lessons regarding the clinical benefits of anti-CD20 therapy
  5. Can mechanisms of disease interfere with the mechanism(s) of anti-CD20 action?
  6. Is response to anti-CD20–mediated deletion influenced by prosurvival factors?
  7. REFERENCES

The susceptibility of CD20-bearing B-lineage cells to deletion by rituximab may also be affected by local prosurvival influences. BAFF, a member of the tumor necrosis factor α family, is a major determinant of the survival of all B-lineage cells after exit from the bone marrow, and it has also been shown to affect the outcome of antigen receptor–mediated encounters. Moreover, overexpression of BAFF has also been implicated in the pathogenesis of RA, SLE, and other autoimmune diseases (37). For the longevity of plasma cells, which do not express CD20 and hence are not directly affected by rituximab, survival signals are reported to include interleukin-5, interleukin-6, tumor necrosis factor α, ligands for CD44, and stromal cell–derived factor 1 (CXCL12) (38). In addition, recent findings in a well-characterized murine lupus system suggest that a defect in chemotaxis to stromal cell–derived factor 1 contributed to the pathologic accumulation of plasma cells into the spleen (39).

Autoimmune pathogenesis is also affected by the availability of these prosurvival and chemotactic factors, and expression may be altered especially in ectopic lymphoid tissues that are directly responsible for pathology. While it is currently unknown how any of these factors affect B cell responsiveness to anti-CD20 treatment, these influences may explain the improvement in the anti-dsDNA response found with the rituximab and cyclophosphamide cotreatment regimen in the British lupus trial (17), while such effects were not seen in all patients after the monotherapy regimen in the trial reported by Looney et al (14). Improved outcome with a rituximab cotreatment regimen was also demonstrated in the recent RA trial (28), which showed significant improvement with a cyclophosphamide/rituximab cotreatment regimen, compared with rituximab alone. Importantly, in this RA study there was also a better rate of clinical response in the group that received cotreatment with methotrexate, which was not used at cytotoxic doses, but which clearly can have subtle and varied effects on cellular metabolism.

While it is currently unknown how these agents act/interact with rituximab, the benefits of cotreatment are not due to simple effects on the in vivo pharmacokinetics of rituximab distribution and excretion (40). It is therefore intriguing to speculate that such second agents may instead provide benefits by limiting the availability of local B cell survival and chemotactic factors, concurrent with or independent from reducing local disease-associated inflammation. Notably, even though the regimen in the Greek SLE patients included only rituximab and modest doses of corticosteroids most had previously received cyclophosphamide or mycophenolate mofetil. One may therefore wonder whether benefits can also derive from sequential treatments, perhaps due to residual effects on non–B cell local mesenchymal cells and their precursors that affect clinical outcome.

In conclusion, findings of the reported clinical trial in SLE patients (20) suggest that rituximab may have an attractive safety and efficacy profile. Although the data are still limited, rituximab may be associated with less risk than current regimens with cyclophosphamide, which, in addition to impairing defenses from infection, is responsible for an unacceptable risk of induced ovarian failure (41). In contrast, serious infections were not common after rituximab treatment in this open-label study. A possible explanation may derive from evidence that protective antibody levels induced by prior vaccination are not affected by rituximab, presumably due to a lack of effects on long-lived plasma cells, although this clearly requires further investigation. Although there is still much to learn about the effects of anti-CD20 on disease pathogenesis, the current report provides exciting data and brings a promising treatment one step closer to the clinic.

REFERENCES

  1. Top of page
  2. Recent experience with rituximab in SLE
  3. Novel insights from the open-label trial for lupus nephritis
  4. Emerging lessons regarding the clinical benefits of anti-CD20 therapy
  5. Can mechanisms of disease interfere with the mechanism(s) of anti-CD20 action?
  6. Is response to anti-CD20–mediated deletion influenced by prosurvival factors?
  7. REFERENCES