|Home | About | Journals | Submit | Contact Us | Français|
Approval of an anti-CD20 chimeric monoclonal antibody, rituximab, has revolutionized cancer treatment and also validated CD20 targeting for providing benefit and improvement of overall response rate in B cell malignancies. Although many patients have benefited from the treatment of rituximab, there are still significant numbers of patients who are refractory or develop resistance to the treatment. Here we discuss pre-clinically well-defined potential mechanisms of action for rituximab and review the ways next generation anti-CD20 monoclonal antibodies can potentially exploit them to further enhance the treatment of B cell malignancies. Although the relative importance of each of these mechanism remains to be established in the clinic, well-designed clinical trials will help to define the efficacy and understanding of which effector activity of modified next generation anti-CD20 mAb will be important in the treatment of B-cell malignancies.
The treatment of B cell malignancies has undergone substantial change since initial marketing approval in 1997 of the chimeric anti-CD20 antibody rituximab for the treatment of both aggressive and indolent subtypes of Non-Hodgkin lymphoma (NHL).1 Rituximab is approved for use as monotherapy and in combination with chemotherapeutics. Treatment with rituximab has resulted in significant improvement in overall response rates and survival of patients with NHL.2–9 Despite these improvements, there are significant numbers of relapsed/refractory lymphoma patients1,10 and infusion related adverse events in the clinical setting.11
Several studies have suggested that rituximab activity is dependent on CD20 expression12 for both direct killing activity via CD20 signaling e.g., programmed cell death (PCD), sensitization of cells to chemotherapy13 and engagement of effector pathways,13 i.e., complement dependent cytotoxicity (CDC), antibody dependent cellular cytotoxicity (ADCC) and antibody dependent cellular phagocytosis (ADCP) (Fig. 1).13 Furthermore, passive immunization has been hypothesized as another potential mechanism for improving efficacy of rituximab, which supported the idea of using rituximab in a maintenance setting.14 In this study, it was shown that rituximab induced apoptosis of lymphoma cells promotes phagocytosis by dendritic cells and cross-priming of CD8 positive cytotoxic T lymphocytes. At this stage, whether this immunization effect is specific to rituximab or to chemotherapeutic regimens is still unclear in the clinical setting.
Rituximab can induce PCD as a result of CD20 signaling and this activity can be augmented when rituximab is hypercrosslinked via a secondary antibody or binding via Fc gamma receptors in vitro.15 Although how this crosslinking activity is achieved in vivo still remains to be proven, primary tumors derived from rituximab treated chronic lymphocytic leukemia (CLL) patients were shown to express activated caspase-3 and caspase-9 indicating the presence of PCD activity in vivo.16 A xenograft model has also shown that increased expression of anti-apoptotic Bcl-2 family proteins can result in rituximab insensitivity.17 Whether, a similar phenomenon applies to primary tumors remains to be determined. Recently, Lim et al.13 have summarized studies where they compared the ability of rituximab to deplete human CD20 transgenic mouse B cells in vivo in the presence or absence of a second transgene encoding high levels of Bcl-2, which blocks the intrinsic apoptosis pathway.13 They report ed that B cells expressing the Bcl-2 transgene were relatively resistant to apoptotic stimuli in vitro whereas in vivo they were just as susceptible to rituximab activity as B-cells lacking the transgene.13 The conclusion from these studies was that in a fully syngeneic system, induction of the intrinsic apoptosis pathway is not important for subsequent B cell depletion.13 While all these studies suggest that rituximab is involved in promoting cell death, whether this mechanism is critical for the depletion of CD20 positive target cells in vivo remains to be determined.
Fc binding to Fc gamma receptors expressed on monocytes, macrophages, natural killer (NK) cells and neutrophils can lead not only to ADCC and ADCP activities but also direct killing via CD20 signaling due to hypercrosslinking.15–18 The early preclinical evidence for the involvement Fc-Fc gamma receptor interaction came from an in vivo study with the xenograft model, showing that rituximab activity is dependent on the gamma chain associated activating Fc receptors.19 Additional supporting evidence comes from a clinical study showing a better response with rituximab in NHL patients with higher affinity allelic variants of Fc gamma IIIa receptor.20–23 However, this correlation has not been observed in CLL patients,24 and it is hypothesized that this might be due in part to lower level of CD20 expression and the presence of higher levels of soluble CD20 in plasma.12 It also has been noted that in mice monocytes/macrophages are the main effector cells that contributes to the activity of rituximab compared to the NK cells and neutrophils in humans.25–27 Moreover, maximal monoclonal antibody (mAb) response activity has been shown to be dependent on intact compartments of the reticulo-endothelial system, as shown in experiments that surgical limitation of the hepatic blood supply correlated with lower B cell depletion.25–27 It has also been demonstrated that mouse Fc gamma receptor IV, a homolog of human Fc gamma receptor IIIa, is strongly involved in the effects of human IgG1.28 It is expressed on murine monocytes and macrophages, but not on murine NK cells, whereas human Fc gamma receptor IIIa is expressed on human NK cells, neutrophils and monocyte/macrophages.28 Due to differences of Fc gamma receptor expression profiles in human and mouse NK cells, the impact of NK cell activity in vivo preclinical models might not be relevant in, and translate to, clinical settings. Although human NK cells are shown to be mediating ADCC activity in vitro,29 questions remain as to whether NK cells are the key cells in mediating rituximab activity or whether, ADCC is the dominant mechanism in tissues.
Although Fc-Fc gamma receptor interaction is widely accepted to be critical for the activity of rituximab in vivo, CDC activities are still being debated.18,28 Supporting evidence for the role of CDC comes from studies where rituximab was shown to be capable of C1q binding and inducing CDC against malignant B cells in vitro.30–32 Moreover, when rituximab activity was tested in complement deficient or cobra venom inactivated complement studies rituximab was shown to exhibit reduced activity in vivo.33,34 Additional evidence comes from studies in patients showing that soon after rituximab infusion, complements were being consumed in vivo, and that addition of complement ex vivo was capable of restoring the activity of rituximab in CDC assays.35,36 Although earlier studies show the critical role of CDC activity of rituximab, it should be noted that this preclinical activity of rituximab was not sufficient to deplete B cells in vivo nor it did correlate with the expression of complement inhibitory receptors.25,37,38 Yet it has also been shown that deposition of C3b not only facilitated the removal of rituximab:CD20 complexes from the B lymphoma cells by Fc gamma receptor expressing macrophages through the process of trogocytosis, but it also blocked the interaction between the Fc domain of rituximab and Fc gamma receptor IIIA on NK cells thereby decreasing ADCC activity.35,39–41 Human genetic polymorphism correlation provides an additional line of evidence suggesting the involvement of complement in the mechanism of action of rituximab. A study investigating the impact of C1qA polymorphisms on the efficacy of rituximab demonstrated that follicular lymphoma patients with a low C1q expressing A allele correlated with enhanced rituximab responses compared to those patients with the high C1q expressing G allele.42 All these studies indicate that CDC plays a role in rituximab activity, but whether this activity is critical and has a positive or negative impact remains unresolved.
While successes, limitations and elucidation of the mechanism of action of rituximab have increased our understanding and helped advance the engineering of next generation anti-CD20 mAbs with the goal of improving the efficacy and decreasing associated adverse events, one still needs to better understand the potential interplay between the multiple proposed mechanisms of action-binding kinetics to different epitopes, PCD, CDC, ADCC and ADCP (Table 1). Depending on the proposed modification, next generation antibodies may be grouped in one of two categories: second or third generation anti-CD20 mAb.
Second generation antibodies can be tailored to be humanized or fully human with unmodified Fc domain, with the aim of reducing immunogenicity. Likewise, third generation antibodies can be modified to include engineered Fc domains with the aim of improving the therapeutic activity in all patients, particularly in genetically defined subpopulations that express a low affinity version of the Fc receptor on their immune effector cells. Second generation antibodies include ofatumumab, ocrelizumab and veltuzumab. Compared to the other CD20 mAbs, fully human IgG1 ofatumumab is at the most advanced stage: it is approved in the US and undergoing regulatory review in the EU for the treatment of CLL.43–45 Ofatumumab binds a different CD20 epitope compared to rituximab and has a slower off rate such that over 3 hours the disassociation for ofatumumab is only 20–30%, but 70–80% for rituximab.46,47 Moreover, ofatumumab exhibits not only ~10 fold higher CDC activity in rituximab sensitive tumor cell lines but also exhibits CDC activity in rituximab resistant cell lines.43,48 ADCC activities of ofatumumab have been shown to be similar to rituximab although it is a weaker PCD inducer than rituximab.43,48 Ocrelizumab is currently being developed for non-oncology indications, and has been shown to bind to the same CD20 epitope as rituximab,47,49 but the molecule has relatively higher ADCC and low CDC activities compared to rituximab.50,51 Veltuzumab is a humanized IgG1 and has similar mechanisms as rituximab, with the exception of a 2.5 fold slower off rate and higher CDC activity (mean EC50 of <100 ng/ml) compared to rituximab (EC50 of ~150 ng/ml).52,53 Phase 1/2 studies with recurrent B cell lymphomas have shown a 53% overall response rate, including 6 patients with complete response at a median follow up of 12 weeks.54 Phase 2 clinical trials using the new subcutaneous formulation of veltuzumab for NHL, CLL and ITP patients are on going.55 Clinical trials testing veltuzumab and ofatumumab may provide further insights into the importance of different epitope on PCD activity in the clinical context.
Third generation anti-CD20 mAbs in early phases of clinical development include AME133v, Pro13192 (v114), GA101 and R603/EMAB-6. TRU-015, is another third generation anti-CD20 molecule that is a small modular immunopharmaceutical drug composed of human IgG1 Fc and CH2 and CH3 hinge regions linked directly to an anti-CD20 scFv.56–58 It is slightly smaller than an IgG and has high ADCC and low complement activating ability. It is currently in Phase 2 clinical development for RA.59–62 AME-133v, an Fc protein engineered antibody, is currently being evaluated in a Phase 1/2 dose escalation study using weekly intravenous doses for four consecutive weeks in patients with relapsed/refractory follicular B cell NHL. In vitro models have shown that the Fc domain of AME-133v binds to the low-affinity variant of Fc gamma RIIIa (FF or FV) with a higher affinity (mean EC50 <10 ng/ml) thereby improving killing of B cells ~10 fold over rituximab.63,64 The improvement in efficacy of AME-133v that has been seen in preclinical settings has yet to be demonstrated in a clinical setting. Although the current clinical trial with AME-133v started in July of 2006 and had an estimated primary completion date of December 2008, clinical data remain unavailable as of October 2009.
Pro131921 (v114), is another Fc protein engineered antibody and displays 30-fold greater binding to the low-affinity variant of Fc gamma RIIIa (FF or FV) than rituximab.65 In vitro, this binding affinity exhibits improved ADCC activity up to 10 fold more than rituximab. Preclinical studies in non-human primates showed that treatment with Pro13192 (v114) resulted in a dose-dependent reversible neutropenia and thrombocytopenia.65 Phase 1/2 clinical studies to assess safety of escalating doses of Pro13192 (v114) in patients with NHL and CLL were recently terminated.
GA101 is a third-generation anti-CD20 mAb with a glyco-engineered Fc portion which exhibits improved binding affinity to FcgammaRIII by 50-fold, that results in a 10- to 100-fold increase in ADCC against CD20 positive NHL cell lines.66–70 On the other hand, CDC activity of GA101 is much lower than rituximab in vitro.66–70 Moreover, GA101 depleted normal B cells as well as B cell lymphoma, significantly more than other CD20 directed antibodies, including rituximab Fc variants with improved ADCC activity.69,70 In vivo xenograft experiments in which cobra venom factor was used to inhibit complement activity demonstrated that GA101 may be efficacious in that setting.71 Structurally, GA101 contains a modified Fc domain hinge region that results in stronger induction of apoptosis of several NHL cell lines and primary malignant B cells.66 These modifications may provide GA101 with an increased therapeutic efficacy, leading to complete responses and long-term survival in xenograft models of diffuse large B cell lymphoma and mantle cell lymphoma (MCL).71–74 In cynomolgus monkeys, GA101 compared to rituximab induced complete, rapid and long-lasting B cell depletion not only in peripheral blood but also in spleen and lymph nodes.74,75 It is still unknown whether this superior activity comes from PCD or decreased CDC activity compared to rituximab.
LFB-R603/EMAB-6 is a chimeric third generation IgG1 and is produced in rat cell line YB2/0 using EMABLING technology thus resulting in naturally low fucose content in its Fc region.76 Compared to rituximab, LFB-R603/EMAB-6 has similar CDC and PCD activities whereas FcγRIIIA binding and FcγRIIIA-dependent effector functions are higher and results in producing an ADCC plateau around 35% at 50 ng/ml, while rituximab induced less than 5% ADCC at the same concentration.76 Furthermore, LFB-R603/EMAB-6 induces higher ADCC activity against CLL cells than rituximab even when target cells express fewer CD20 molecules.76 Although the improvement in efficacy of LFB-R603/EMAB-6 that has been seen in preclinical settings, the clinical activity is yet to be demonstrated in the future.
In the face of the excitement and perhaps uncertainty generated by the next generation of anti-CD20 mAbs, their potential success in the treatment of B cell malignancies will depend on their clinically demonstrated safety and efficacy profiles. There is the potential to achieve far better therapeutic efficacy than rituximab, and, significantly, demonstrate efficacy in rituximab-resistant populations. Additionally, all third generation antibodies appear to be Fc engineered (protein or glyco-engineering) with the hopes of improving the affinity of Fc-Fc gamma receptor interaction, thereby improving the ADCC activity. Although improvement of these interactions has increased ADCC activity in vitro, the significance of this mechanism is yet to be determined in the clinic. Furthermore, one should note that modulating Fc interactions can also affect other mechanisms of cell cytotoxicity such as ADCP activity and CD20 mediated apoptosis due to hypercrosslinking via Fc gamma receptor.
Although the mechanisms of action of each CD20 mAb are well-studied in preclinical settings, the variability seen in clinical response to rituximab may also depend on level of CD20 expression, levels of circulating soluble CD20, presence and abundance of effector cells, CD20 binding epitope and kinetics, tissue distribution and tumor burden. The predictive value of preclinical models in terms of quantification of the dose-concentration-effect relationship of rituximab using pharmacokinetic-pharmacodynamic analysis, identification of the individual factors influencing the response, relative importance of each of these mechanisms of action and resistance remains to be better understood and proven in the clinic. Well-designed clinical trials will help define and refine efficacy and provide increased understanding of which activity of modified next generation anti-CD20 mAb will prevail. It is interesting to note that rituximab in combination with methotrexate is also approved for treatment of adult RA patients who have had inadequate response to TNF antagonist therapies, and the product is also currently being studied in additional indications. Evaluations of modified next generation anti-CD20 mAb as treatments for non-oncology indications can potentially add to our understanding of the critical activities required for efficacy.
We would like to thank Michele Mccolgan and Greg Roland for their help with the figure and the table, respectively.
Previously published online: www.landesbioscience.com/journals/mabs/article/10789