Empowering Targeted Therapy: Lessons from Rituximab

See allHide authors and affiliations

Science's STKE  13 Jul 2004:
Vol. 2004, Issue 241, pp. pe30
DOI: 10.1126/stke.2412004pe30


Rituximab, a monoclonal antibody directed against the B cell–specific protein CD20, has revolutionized lymphoma treatment by providing a highly effective form of therapy with relatively mild toxic side effects. Effective as a single agent against some forms of B cell lymphoma, rituximab also has a chemosensitizing effect, enhancing the efficacy of chemotherapy against other forms of the disease. Although the mechanisms whereby rituximab achieves its effects remain incompletely understood, these seem to involve at least three distinct phenomena: (i) antibody-dependent cell-mediated cytotoxicity, (ii) complement-mediated cell lysis, and (iii) stimulation of apoptosis in target cells. The latter occurs through interaction of complexes of rituximab and CD20 in lipid rafts, with elements of a signaling pathway involving Src kinases. Effector molecules trigger various gene expression events, leading to sensitization of malignant cells to proapoptotic stimuli. Lessons learned from the research on rituximab may be applied to the rational development of antibody-based therapies against other forms of cancer.

Monoclonal Antibodies as Therapeutic Agents in Treating Lymphoma

Three decades have elapsed since Köhler and Milstein published their seminal report describing the production of monoclonal antibodies (mAbs) (1). At that time, clinical investigators predicted the rapid transition from traditional chemotherapy to mAb-based therapy in the treatment of cancer. Nevertheless, many obstacles had to be surmounted before effective treatment using mAbs could be realized. Finally, in 1997, rituximab, a mAb to protein CD20, was approved by the U.S. Food and Drug Administration for the treatment of follicular and low-grade B cell non-Hodgkin's lymphoma. In the ensuing 7 years, hundreds of thousands of patients have benefited from this therapy.

Throughout their lifetime, lymphocytes undergo continuous stimulation toward proliferation, differentiation, or apoptosis under the influence of cytokines and cellular interactions. The numerous forms of B cell and T cell lymphoma, which originate from lymphocytes at different developmental stages, can be broadly divided into clinically indolent (for instance, follicular) and aggressive (for instance, diffuse large-cell) subtypes. In most cases, each patient’s malignant lymphocytes possess a unique antigenic signature (idiotype), so that their specific surface proteins may become therapeutic targets. This idea was first brought into clinical practice by the development of antibodies to idiotypes in the early 1980s (2). Although such antibodies were effective in treating some forms of lymphoma, the cost and time consumed in the creation of tailored antibodies to idiotypes for each patient limited the application of this therapy.

Rituximab: A mAb That Has Revolutionized Treatment of Lymphoma

With the development of rituximab, a chimeric mAb directed against the CD20 molecule, the concept of mAb therapy found its broad practical application. CD20 is a single-chain cell membrane protein restricted to benign and malignant B lymphocytes. It is present at various stages of B cell differentiation, beginning with the pre–B cell, and disappears only as the B cell completes differentiation into the plasma cell. CD20 acts during B cell receptor-mediated activation as a modulator of several signaling pathways related to tyrosine kinases of the Src family. It is an oligomeric calcium channel that by regulating Ca2+ influx (3) may participate in the regulation of lymphocyte activation and proliferation (4).

CD20 is an attractive target for the therapy of B cell neoplasms, because it is expressed on the surface of malignant cells in more than 95% of cases of B cell lymphoma and leukemia. In addition, CD20 lacks a soluble counterpart in the serum that might bind a circulating antibody, and it is not subject to shedding or internalization. A pivotal phase II study by McLaughlin demonstrated a 48% response rate of relapsed low-grade lymphomas to treatment with rituximab, with a median response duration of 13 months (5). This efficacy was comparable to results achieved with standard chemotherapy, and toxicity was milder and was largely limited to infusion-related allergic reactions. Even higher response rates(60 to 80%) and longer durations of response have been reported in previously untreated patients with low-grade lymphoma (6). Subsequently, rituximab was combined with chemotherapy, yielding a 95% response rate in patients with indolent lymphomas, with no major additive toxicity (7).

Follicular lymphomas (the most common form of indolent lymphoma) are characterized by a unique chromosomal translocation [t(14;18)] resulting in the overexpression of the bcl-2 proto-oncogene. The Bcl-2 protein acts as a potent protector against caspase-mediated apoptosis and is believed to be crucial in the neoplastic transformation of a lymphocyte into a lymphoma (8). Conventional chemotherapy is incapable of clearing cells positive for the bcl-2 rearrangement from the peripheral blood and bone marrow, as assessed by polymerase chain reaction. An unprecedented frequency of such complete molecular remissions, evidenced by bcl-2 clearance, was observed after single-agent rituximab therapy (9).

Single-agent activity in other lymphoproliferative malignancies, such as chronic lymphocytic leukemia (CLL) and the clinically aggressive diffuse large B cell lymphoma (DLBCL), is lower than that observed in follicular lymphoma. This lower activity may reflect lower density of CD20 expression on these tumors, as well as inherent differences in resistance to the antibody. The poor efficacy against these forms of lymphoma was overcome by combining rituximab with various chemotherapy regimens, suggesting a chemosensitizing property of the antibody. A large randomized trial established the combination of rituximab with CHOP (cyclophosphamide, vincristine, doxorubicin, and prednisone) as the new standard of treatment for DLBCL in the elderly, improving both relapse-free and overall survival over the standard use of CHOP alone (10). The ability of rituximab to purge B lymphocytes from blood led to its application in autoimmune disorders such as idiopathic thrombocytopenic purpura, vasculitis, and rheumatoid arthritis (11).

Mechanisms of Rituximab Action

Despite widespread use of this agent, relatively little is understood about the mechanism of rituximab's cytotoxicity. There are multiple potential mechanisms by which antibodies exert their therapeutic actions, short of serving as delivery systems for conjugated toxins or radioisotopes. Extensive evidence supports the involvement of three different mechanisms for rituximab's effects: (i) antibody-dependent cell-mediated cytotoxicity (ADCC), (ii) complement-mediated lysis, and (iii) a direct effect on the target cells through the disruption of signaling pathways and triggering of apoptosis (Fig. 1).

Fig. 1.

Mechanisms of action of rituximab on the B lymphocyte. Complement-mediated cytolysis involves the ability of IgG1 to engage the complement on antibody-coated cells. ADCC requires interaction between the Fc portion of rituximab and appropriate receptors on effector cells, such as natural killer cells. Apoptosis occurs upon cross-linking of rituximab/CD20 complexes in the lipid rafts [in vivo, possibly with the help of Fc receptor (FcR)–equipped cells]. This activates signaling pathways involving the Src kinases (Lyn, Fyn, and Lck) and their regulatory molecules Cbp and Csk.

Conflicting data concerning the contribution and relevance of these different mechanisms to the clinical efficacy of rituximab have emerged, and the need to analyze different studies in view of their divergent designs has only lately been recognized (12). In reviewing the available data, one must cautiously consider the inclusion of patients with dissimilar pathologies (such as follicular lymphoma, mantle cell lymphoma, and CLL) (13); the use of CD20-targeting antibodies other than rituximab (such as murine antibodies 1F5, 2H7, and B1) in vitro; and whether specific research models adequately reflect the clinical authenticity of human disease. The cytotoxic mechanism that predominates in a given clinical scenario may vary depending on the specific characteristics of different forms of lymphoma.

In CLL, which is characterized by a large number of circulating CD20+ cells, rituximab leads to a rapid complement-dependent lysis of malignant and benign B lymphocytes (14). Experimental evidence indicates that the infusion-related side effects of rituximab (15), as well as the brisk depletion of circulating B lymphocytes (16) after the treatment, are all related to the initial complement-dependent action of the antibody. However, in CLL, a population of cells persists and reexpands because of a low density of CD20 on the cell surface, along with a protective effect of CD59 against complement cytotoxicity (17). CD59 is a protein expressed by circulating leukocytes that inhibits inadvertent formation of the complement-originated membrane attack complex that would eliminate them from the bloodstream (18). The emergence of a resistant CLL population may explain the need for higher doses of rituximab to achieve clinically meaningful responses (19). In contrast, no correlation between CD59 expression and response to rituximab has been detected in follicular lymphoma, which implies that rituximab acts predominantly through a different mechanism in this disease (20).

In malignancies that are not characterized by large numbers of circulating malignant cells, such as follicular lymphomas, rituximab reaches the malignant cells outside the intravascular compartment and exerts its action through ADCC accomplished by natural killer and polymorphonuclear cells (21). The requirement for Fc receptors (and thus ADCC) for the activity of rituximab against Burkitt’s lymphoma cells implanted subcutaneously into nude mice supports this hypothesis (22), although a predominantly complement-related mechanism was again observed in a model where mice were injected with lymphoma cells intravenously (23). The role of ADCC in therapy for follicular lymphoma is further supported by studies of a single-nucleotide genetic polymorphism of the human gene encoding FcγRIIIa [CD16a, the immunoglobulin G1 (IgG1) Fc receptor], resulting in the receptor’s higher affinity for the IgG1 Fc fragment and thus a greater capacity for ADCC. This polymorphism correlates with a better response to treatment with rituximab in patients with non-Hodgkin’s lymphoma (24), but not in those with CLL (25).

Direct Events Mediated Through Antibody Binding to CD20

The evidence that binding of CD20 by mAbs may also exert a direct apoptotic effect comes largely from investigations in vitro. These studies demonstrate growth arrest of lymphoma cells after treatment with rituximab, with no measurable involvement of complement- or cell-mediated cytotoxicity (26). In contrast, 1F5, an antibody directed at a different CD20 epitope, exhibits B cell–activating properties (27).

The difficulty in understanding how rituximab may lead to apoptosis originates from the paucity of data about the role of CD20 in vivo. To date, it appears that CD20 lacks an endogenous circulating ligand despite its homology to IgE receptor FcεRI. It forms complexes with major histocompatibility complex (MHC) class II (28) and CD40, a molecule involved in B cell activation and Ig isotype switching (29). The native activity of CD20 is related to its functional association with tyrosine kinases of the Src family, including Fyn, Lyn, and Lck. This association depends on the degree of CD20 phosphorylation, as well as on intermediary molecules, such as the Src-inactivating factor Csk-binding protein (30), which acts as an interface between CD20 and Src kinases. Recent publications suggest that CD20 is constitutively present in distinctive cell membrane structures: lipid rafts rich in cell signaling proteins, including the Src kinases (31). Initially studied in the context of endothelial potocytosis (the uptake and sequestration of molecules by caveolae) (32), lipid rafts hypothetically function as cell membrane signal-processing centers (33). Upon binding with rituximab, more CD20 complexes are rapidly translocated into these rafts, which leads to a higher concentration and cross-linking of the molecule therein (34). This translocation has been shown to down-modulate the activity of Lyn kinase in a process dependent on Csk-binding protein (35). Thus, rituximab-mediated translocation into lipid rafts may promote the engagement of cross-linked CD20 complexes in the signaling pathways, ultimately leading to susceptibility to apoptosis.

However, in contrast with previously discussed work, studies by Chan et al. demonstrate that the types of responses achieved with rituximab are very different than those observed with murine antibody to CD20 (anti-CD20) B1 (used as a backbone for tositumomab, a therapeutic 131I-radioimmunoconjugate). Rituximab stimulated CD20 translocation into the rafts and its cytotoxicity was complement-dependent, whereas B1 did not stimulate raft translocation and acted primarily by triggering apoptosis (36). The ability of rituximab to induce substantial apoptosis is observed, at least in certain cell lines, only upon further cross-linking of its complexes with CD20 using an antibody to IgG (37). Attenuation of rituximab-induced apoptosis was achieved in vitro using inhibitors of Lck and Fyn tyrosine kinases, calcium chelators, and caspase inhibitors (38, 39), indicating that all these mechanisms are critical for executing CD20-mediated cell death. The downstream effector proteins involved in mediating rituximab-dependent apoptosis after cross-linking include phospholipase C (40), caspase-3, p38 mitogen-activated protein kinase (p38 MAPK), Bax, and c-Myc (41, 42). The knowledge of their complicated networking is, however, too fragmentary to delineate a coherent mechanism operating in vivo. Conversely, B1 does not require the addition of antibody to IgG to stimulate apoptosis (43). If hyper–cross-linking also occurs by means of Fc receptors on cytotoxic cells (as suggested by in vitro incubation studies), ADCC and apoptosis may be complementary mechanisms involving cellular interactions. This model would support the capacity of rituximab to induce apoptosis in vivo with the help of ADCC-dedicated cells able to further cross-link complexes of rituximab with CD20.

One group reported caspase activation after treatment with rituximab in humans with CLL (44), supporting the assumption that apoptosis, and thus cross-linking of rituximab complexes, may take place in vivo. However, at present there are no further data confirming apoptosis in patients treated with rituximab. In fact, different cell lines have varying sensitivities to anti-CD20–mediated apoptosis, suggesting that this mechanism may occur only in selected types of lymphoma; and this sensitivity further depends on the type of antibody used, suggesting epitope specificity (45).

In both clinical and laboratory studies, rituximab exhibits chemosensitizing properties in lymphoma (46, 47). This effect is probably achieved through proapoptotic action of the mAb, leading to the modulation of intracellular proteins responsible for the regulation of cell death (48). Examples of such proteins reported in association with rituximab-triggered chemosensitization include Bcl-2 (49), Bcl-xL, X-linked inhibitor of apoptosis protein (XIAP), and Mcl-1. Researchers from the University of California in Los Angeles have described an elegant model of signaling cascade leading to sensitization to chemotherapy (50). In this model, so far confirmed in vitro in lymphoma cell lines, rituximab interacts with CD20, and through Src kinases leads to a decrease in transcription of interleukin-10 (IL-10). This disrupts an autocrine loop operating in the B lymphocyte, whereby IL-10 normally sustains the constitutive expression of the STAT3 transcription factor. STAT3 in turn is responsible for expression of bcl-2 and thus for maintaining the viability of the lymphocyte against potentially proapoptotic environmental stimuli. The secretion of IL-10 was attenuated by inhibitors of Src kinases and p38 MAPK, which implies that these molecules function in the same chain of events as rituximab. There are likely multiple undetermined intermediaries between them. If this model is operable in vivo, it could explain how rituximab-mediated signaling leads to down-regulation of Bcl-2 and thus lowers the threshold for triggering apoptosis in the lymphocyte. Metabolic damage induced by a chemotherapeutic agent would then be more likely to result in cell death.

Other Clinically Useful Antibodies

Another antibody that offers therapeutic benefit to patients, albeit less impressive than that of rituximab, is Campath-1H (alemtuzumab). This antibody targets CD52, an antigen of unknown function that is ubiquitous on lymphocytes of all lineages and exhibits a remarkable propensity to induce cell lysis when bound by an antibody. Alemtuzumab has been used primarily in the treatment of chemotherapy-resistant CLL, yielding a 33% response rate at the expense of significant infectious toxicity because of prolonged reduction in the levels of both T and B lymphocytes (51). Different heavy-chain subtypes of anti-CD52 have been investigated for possible use against CLL. A complement-specific IgM version achieved rapid purging of the peripheral blood, but not the bone marrow or spleen, whereas an IgG2-specific formulation led to ADCC-mediated long-lasting responses in all organs without affecting the serum complement levels (52). A humanized IgG1 version, which was effective in all three locations, was subsequently selected for clinical development.

When selecting a surface protein as a target for antibody-mediated therapy, it is important to consider both its biological function and how its interaction with the antibody will affect that activity. The clinical success achieved with rituximab may arise from its ability to engage diverse cytotoxic mechanisms. Development of a superior molecule may focus on enhancing its affinity for the Fc receptors in settings where ADCC is identified as the principal mode of action. This paradigm yielded interesting preclinical results in engineering more potent versions of the breast cancer drug trastuzumab (53) and a novel therapy for T cell lymphoma/leukemia, antibody to CCR4 (54). Mutated versions of rituximab with enhanced complement-binding activity also have been created (55), although their efficacy remains to be proved in the clinic.

The greatest potential for improving treatment results may reside in combinations of targeted therapies. The presence of antibodies targeting different epitopes of even a single antigen may lead to hyper—cross-linking and thus more effective ADCC and apoptosis (56). Rituximab has already been safely combined with antibody to CD52 in the clinic with encouraging results, achieving a 52% response rate in heavily pretreated B cell malignancies (57). A host of other antibodies that might augment the proapoptotic signal are being developed in early clinical trials, such as anti-CD80, anti-CD23, molecules directed against angiogenic factors (such as vascular endothelial growth factor), and HLA-DR (apolizumab) (58). One interesting pathway may involve combining rituximab with anti-CD55 or anti-CD59 to enhance complement-mediated activity in CLL, because those antibodies abolish the resistance of cells to complement in vitro (59).

In aggressive lymphomas, an additional mechanism, so far clinically unexplored, may exist: activation-induced cell death, resulting from the proapoptotic action of signals normally involved with proliferation. This phenomenon is exemplified by apoptosis of anaplastic lymphoma exposed to CD30 stimulation (60). CD30 is a molecule that is present on activated and virus-infected lymphocytes. It is involved in nuclear factor κB (NF-κB) and TRAF signaling pathways, converting a stimulatory signal into either proliferation or cell death. In CD30-positive anaplastic lymphoma, the constitutive NF-κB expression is absent and CD30 binding appears to lead to cell death rather than activation. Similarly, CD40, which is found ubiquitously on B lymphocytes, may be an attractive target with an intricate biology. Initially described as having a protective effect against apoptosis, it has the ability to associate with CD20 (61), modulate its activity (62), and stimulate both ADCC (63) and apoptosis when cross-linked by anti-CD40 (64, 65).

Ultimately, the success of novel therapies will depend on the careful choice of combination of candidate antibodies for clinical development and a more thorough understanding of cytotoxic mechanisms. Although the observation that rituximab offered high clinical response rates in B cell neoplasms was initially serendipitous, given the failure of other antibodies to CD20 to achieve similar clinical success, the subsequent exploration of its cytotoxic mechanisms has opened the door to a new generation of rationally designed antibody therapies.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
  63. 63.
  64. 64.
  65. 65.
View Abstract

Stay Connected to Science Signaling

Navigate This Article