English translation of Med Sci (Paris) 2013 ; 29 : 57–63

© 2013 médecine/sciences – Inserm

Monoclonal antibodies (mAbs) have been used for cancer therapy as early as at the beginning of the 80’s, as shown by the pioneering work of Ronald Levy and his colleagues (Stanford University, USA). These authors reported in 1982 the treatment of a lymphoma patient with a mAb directed against the immunoglobulin expressed at the surface of tumor cells (« anti-idiotype » antibody) [1]. To date, sixteen mAbs have received a market approval in Europe and/or in the USA for cancer treatment (although two of them, edrecolomab and gemtuzumab, have been withdrawn from the market due to a lack of clinical efficacy and unfavorable risk/benefit ratio, respectively) (Tableau I). Other antibodies have got marketing approval in People’s Republic of China, India, and in some others Asian countries, as well as in Latin America (Cuba) and South America (Brazil) (Tableau I). The anti-tumor effects of these mAbs rely on mechanisms that differ depending on the molecules being targeted. A first category of antibodies is made of antibodies directed against tumor cells, whereas a second category includes antibodies directed against molecules that modulate tissue and cell tumor environment (Figure 1).

Figure 1. Monoclonal antibody mechanisms of action in cancer. Blockade of adhesion molecules such as EpCAM, involved in tumor cell migration and metastases, is not represented. ADCC: antibody-dependent cell cytotoxicity; C3bR: receptor for the C3b complement fragment; CDC: complement-dependent cell cytotoxicity; CTLs: cytotoxic T cells; DCs: dendritic cells; FcγR: receptors for the Fc region of IgG; Th: helper T cells.

The initial view prevailing when the antibodies belonging to the first category were developed was to use them as a « monoclonal serotherapy », to induce a passive immunity based on their ability to block the activation and/or the proliferation of tumor cells [by targeting receptors for growth factors such as the EGF receptor or the HER2/neu (erbB-2) molecule], to induce apoptosis, even slight (CD20), or to interfere with tumor cell adhesion (EpCAM). This passive immunotherapy also relies on the activation of immune effector mechanisms [antibody-dependent cell cytotoxicity (ADCC), phagocytosis and/or complement-dependent cytotoxicity (CDC) involving the activation of the classical complement pathway and the generation of a membrane attack complex (MAC) responsible for tumor cell lysis] (Figure 1).

This has led to numerous engineering efforts over the last decade, including modification of the constant region (Fc) (most of the mAbs currently on the market exhibit a human IgG1 Fc), antibody conjugation to cytotoxic/cytostatic drugs [such as monomethyl auristatin E (MMAE)], or reformatting mAbs under bi-specific forms (thus allowing the recruitment of immune cells such as T cells, as exemplified by blinatumomab, a bispecific anti-CD3 x anti-CD19 antibody).

Monoclonal antibodies in cancer treatment: from tumor cell killing to the induction of adaptive immunity

The concept of tumor immunosurveillance has been strongly supported in the recent years by studies performed on large cohorts of patients with various types of cancers among the most frequent ones [colorectal cancer (CRC) and lung cancers] [2]. In addition to their direct and rapid anti-tumor effects, mAbs could strongly impact tumor immunosurveillance by enabling anti-tumor adaptive immunity to develop and control tumor progression. Hence, one can suspect that this latter process is responsible for the long-lasting responses observed in some of the patients treated with therapeutic antibodies (Figure 2). Strikingly, the use of trastuzumab (anti-HER2/Neu) in patients with metastatic breast cancer has been correlated with an increase in the number of CD4⁺ T cells specific for HER2/neu-derived peptides [3]. Thus, it is likely that the therapeutic effect of trastuzumab not only depends on its ability to impact the intracellular HER2/neu-dependent signaling but is also due to its capacity to induce innate and adaptive anti-tumor immunity.

Figure 2. Induction of anti-tumor adaptive immunity by monoclonal antibodies. Tumor cells opsonized with antibodies are lysed through various effector mechanisms (Figure 1). Cell debris, apoptotic bodies, and antigen/antibody complexes are then captured by antigen-presenting cells (APCs) such as immature dendritic cells, leading to the activation of CD4+ T cells. Signal dangers are also involved in the sensing of tumor cells by APCs. CD4+ T cells trigger CD8+ cytotoxic T-cell and also likely B-cell activation, leading to the appearance of memory CD4+ and CD8+ T cells and likely memory B cells (not represented). This adaptive immunity opposes inhibitory mechanisms induced by tumor cells and by inflammation that parallels tumor growth. It counterbalances the role of regulatory T cells (Treg) and of immunosuppressive and/or pro-inflammatory cytokines (IL-10, IL-6…). Cross-presentation of tumor-associated antigens by APCs to CD8+ T cells as well as the presence of myeloid-derived suppressor cells (MDSC) such as M2 macrophages are not represented.

Experiments performed in tumor-bearing mice have demonstrated the critical role of CD4⁺ and CD8⁺ T cells in the control of tumor progression and in the survival induced by mAb treatment. On the one hand, the critical role of CD8⁺ T cells in the anti-tumor response induced by treating HER2/neu (ErbB-2)⁺ tumor-bearing mice with a combination of anti-ErbB-2 and anti-Death Receptor 5 (DR5 ou TRAIL-R2 standing for « Tumor necrosis factor-Related Apoptosis-Inducing Ligand-Receptor2 ») antibodies has been demonstrated by depleting CD8⁺ T cells. This depletion abolished anti-tumor protection [4]. The role of CD8⁺ T cells in the clinical response to anti-HER2/neu treatment has been also highlighted in a more recent study where anti-HER2/neu treatment of tumor-bearing Rag-1^−/- mice that lack both B and T cells did not prevent tumor progression [5]. In the same study, depletion of CD8⁺ T cells in wild-type (wt) immunocompetent mice led to the same result, thus demonstrating that the presence of CD8⁺ T cells is mandatory to achieve clinical response when using anti-HER2/neu antibodies. It was then demonstrated that anti-neu treatment induces specific CD8⁺ T cells that produce interferon-γ (IFN-γ), thus showing the essential role of this cell compartment in the induction of an anti-tumor protection. Finally, rechallenge experiments where high numbers of tumor cells were injected to mAb-treated surviving animals have shown that anti-neu treatment induces an immune memory able to protect animals [5].

On the other hand, the induction of an adaptive anti-tumor immunity by antibody treatment also requires CD4⁺ T cells (Figure 2). It has been demonstrated that CD4⁺ T cell depletion abolishes the anti-tumor protection achieved with mAbs directed against the murine Friend Leukemia Virus gp70 protein [6] or against CD20 [7]. In 2010, we could demonstrate that the long-term survival of immunocompetent mice harboring disseminated CD20⁺ tumors and treated with an anti-CD20 antibody is strictly dependent on the presence of CD4⁺ T cells [7]. Depletion experiments showed that the presence of these cells is required both at the beginning of anti-CD20 treatment and when mAb-treated long-term surviving animals are rechallenged with tumor cells in absence of any further antibody treatment [7]. This experimental setting also made it possible to show that transfer of highly purified CD4⁺ T cells isolated from anti-CD20-treated animals three weeks after tumor cell injection, i.e., one week after the end of mAb treatment, enables to protect animals injected with tumor cells but left untreated. No protection was observed when T cells are purified from either naive or untreated tumor-bearing mice [Deligne et al., submitted]. Furthermore, the transfer of lymphocytes isolated from long-term surviving mice rechallenged with tumor cells also made it possible to achieve anti-tumor protection, indicating the induction of a memory response by anti-CD20 treatment [7]. Finally, the injection of interleukin-2 (IL-2) in long-term surviving anti-CD20 treated animals at the time of tumor cell rechallenge significantly increases the survival rate. By contrast, the co-injection of IL-2 and anti-CD20 mAb at the initiation of treatment does not increase the survival rate [7]. In another report, the use of a tri-functional antibody directed against disialoganglioside (GD₂) (a molecule overexpressed by neuroblastoma and melanoma cells) and CD3 [an activating molecule expressed by T cells and associated with T cell receptors (TcR)], and binding to antigen-presenting cells (APCs) through its Fc portion has demonstrated that this antibody format allows the induction of specific anti-tumor T cells [8].

All these data suggest therefore that therapeutic mAbs that target tumor cells can induce a long-lasting protection though the induction of a specific adaptive memory anti-tumor response, a kind of “curative vaccination” in some way [9] (Figure 2). This is in full contrast with the still common view that the therapeutic use of mAbs in cancer therapy represents only passive immunotherapy having only a direct and immediate anti-tumor effect by recruiting effector mechanisms of the innate immunity. This view has largely prevented monoclonal antibody therapy from being considered as a true cancer immunotherapy, whose main goal is to induce an adaptive immunity and anti-cancer vaccination [mostly by ex-vivo or/and in-vivo manipulation of T cells, NKT cells, dendritic cells (DCs)…].

Our suggestion that a major mechanism of action of anti-tumor mAbs is the induction of a long-lasting adaptive anti-tumor response, if confirmed by other experimental and clinical data, could have major consequences on how monoclonal antibody treatments should be combined with other anti-cancer treatments, such as chemotherapy and/or radiotherapy. These latter treatments can inhibit the anti-tumor immune responses, due to the radio-sensitivity of immune cells or to the cytotoxic effect of drugs on these cells. This issue has been exemplified in experiments where HER2/neu⁺ tumor-bearing mice were treated with an anti-HER2/neu mAb either alone or in combination with doxorubicin, simultaneously, or in combination with cyclophosphamide or with paclitaxel, but only three to five days later [5]. The most important tumor regression was observed in the latter setting, while the anti-tumor effect achieved with the anti-HER2/neu mAb per se was decreased when doxorubicin was injected together with the antibody. However, the use of cyclophosphamide or of paclitaxel had a negative impact on immune memory. A strong tumor progression was observed in mice treated first with the antibody and then with one of these two drugs and further rechallenged with tumor cells. By contrast, surviving mice that had been treated only with the antibody efficiently resisted tumor rechallenge. Strikingly, the injection of paclitaxel one day before anti-HER2/neu treatment had a synergistic effect against primary tumor but did not impair anti-tumor memory response as shown by experiments where tumor cells were re-injected into surviving mice [5]. These experiments demonstrate therefore that the search for a therapeutic window allowing the optimal use of anti-tumor drugs in combination with anti-tumor antibodies should be now taken in consideration.

Monoclonal antibodies in cancer therapy: inhibiting the inhibitors

Tumor immunosurveillance mediated by cells from adaptive immunity [2] can be opposed or put on hold by suppressive mechanisms. These are exerted by regulatory T cells (Treg), myeloid-derived suppressor cells (MDSC) (such as M2 macrophages), and/or through the expression of inhibitory molecules [Cytotoxic T-Lymphocyte Antigen 4 (CTLA-4) (CD152), Programmed Death-1 (PD-1) (CD279), Lymphocyte-Activation Gene 3 (LAG-3) (CD223)…] on T cells, leading to the blockade of the anti-tumor activity (Figure 1). Several mAbs have been therefore recently developed to inhibit or deplete suppressive cells or to block the activity of inhibitory molecules, with the hope to restore or trigger an efficient anti-tumor immunity. The easiest strategy has been to use antagonist antibodies to block inhibitory molecules such as CTLA-4 and PD-1 (Figure 1). The ipilimumab antibody, a fully human mAb (raised using humanized mice), directed against CTLA-4, has been granted a market approval in 2011 for the treatment of metastatic melanoma. This antibody has led to significant clinical responses with an increased overall survival of several months [10] (Médecine/Sciences Vol. 27, N°10, 850-8, 2011). Different clinical trials are ongoing with antagonist antibodies directed against other inhibitory molecules such as PD-1 or LAG-3 [11]. Other strategies are also currently explored in preclinical models and in clinical trials. They are mostly based on the combined use of antagonist antibodies directed against inhibitory molecules such as CTLA-4 or PD-1 with either agonist antibodies capable of triggering stimulatory molecules such as CD137 (a molecule belonging to the TNF receptor superfamily) [12] or antibodies that recruit effector cytotoxic mechanisms once bound to molecules expressed by tumor cells such as CD20 or ErbB-2. It has been shown that the combined use of an anti-Erb-B2 antibody with an antagonist anti-PD-1 antibody or an agonist anti-CD137 antibody increases the therapeutic efficacy of the anti-ErbB-2 antibody [12]. This is accounted by the fact that the anti-tumor effect of the anti-ErbB-2 antibody requires the activation of CD8⁺ T cells, which leads to the production of IFN-γ that plays a critical role in the generation of a potent anti-tumor response. This T-cell activation is increased following CD137 activation or when inhibiting PD-1 signaling. In addition, using a xeno-transplantation model, Ronald Levy and his team have shown that the treatment with the anti-HER2/neu trastuzumab antibody induces CD137 expression on NK cells and that the use of an agonist anti-CD137 antibody induces a strong cytotoxicity by NK cells, even against tumor cells otherwise resistant to trastuzumab [13]. The single use of anti-CD137 antibody is currently evaluated in patients with various solid tumors (melanoma, lung cancer, ovarian cancer). Using immunocompetent mice, it has been shown that an anti-CD137 agonist antibody makes it possible to inhibit lymphoma growth in vivo and to increase survival, thanks to the generation of an adaptive long-lasting immune response. Furthermore, this effect was strengthened after depleting Treg cells with an antibody directed against folate-4 receptor (FR4) [14]. Another strategy based on the combined use of agonist antibodies with cytokines whose development had been previously discontinued due to the occurrence of severe adverse events is currently explored. It has been recently shown that treatment with interleukin-15 (IL-15), a cytokine that stimulates the anti-tumor activity of CD8⁺ T cells and of NK and NKT cells, allows an increased survival of mice bearing prostatic tumors, but only when the mice are simultaneously treated with a blocking anti-CTLA-4 antibody and with an antibody that inhibits the interaction between PD-1 ligand (PD-L1 or B7-H1 standing for « B7 Homolog 1 ») (CD274) that is expressed by tumor cells, with PD-1 [15] (médecine/sciences 2012, vol. 28, n°5, p. 481). Moreover, animals receiving the triple combination mAbs/IL-15 exhibit a decrease in the suppressive activity of CD4⁺CD25⁺ and CD8⁺CD122⁺ regulatory T cells. Thus, the combined use of cytokine and mAbs enables to restore the anti-tumor activity of Il-15. It paves the way to the use of this cytokine at lower doses, likely avoiding undesirable side effects.

Conclusion - Therapeutic monoclonal antibodies in cancer: starting a new story

After a remarkable breakthrough achieved over the last fifteen years in cancer therapy, monoclonal antibodies are now considered not only as exhibiting an immediate direct anti-tumor cytotoxicity (Figure 1) but also as drugs capable of inducing long-term modifications within the molecular and cellular networks in charge of tumor immunosurveillance (Figure 2). The induction of adaptive anti-tumor immune responses by monoclonal antibodies makes it possible to envision new therapeutic strategies aimed at generating an anti-tumor immune memory, enabling tumor immunosurveillance to play its role. The ability of anti-tumor antibodies to induce adaptive responses could be further exploited by combining their use with that of antibodies inhibiting mechanisms that suppress immunosurveillance. On-going clinical trials based on this strategy should allow a more accurate evaluation of its efficacy in a near future. Finally, this new way of thinking the anti-tumor modes of action of monoclonal antibodies should lead us to envision differently the relationship between antibody treatment and chemotherapy and/or radiotherapy, as well as to revisit strategies that have been developed to date, combining antibody treatment and the use of immunomodulators acting at different stages of adaptive immunity (hematopoietic growth factors such as GM-CSF, interleukins such as IL-2, IL-15…). Clearly, a new chapter in the history of monoclonal antibodies in cancer therapy will be scripted in the years to come …

Disclosure of potential conflicts of interest

The authors declare no competing financial interests.

Acknowledgments

We thank Dr. Riad Abes for his contribution to some of the work described in the present review.

References

All Tables

All Figures

Figure 1. Monoclonal antibody mechanisms of action in cancer. Blockade of adhesion molecules such as EpCAM, involved in tumor cell migration and metastases, is not represented. ADCC: antibody-dependent cell cytotoxicity; C3bR: receptor for the C3b complement fragment; CDC: complement-dependent cell cytotoxicity; CTLs: cytotoxic T cells; DCs: dendritic cells; FcγR: receptors for the Fc region of IgG; Th: helper T cells.

In the text

Figure 2. Induction of anti-tumor adaptive immunity by monoclonal antibodies. Tumor cells opsonized with antibodies are lysed through various effector mechanisms (Figure 1). Cell debris, apoptotic bodies, and antigen/antibody complexes are then captured by antigen-presenting cells (APCs) such as immature dendritic cells, leading to the activation of CD4+ T cells. Signal dangers are also involved in the sensing of tumor cells by APCs. CD4+ T cells trigger CD8+ cytotoxic T-cell and also likely B-cell activation, leading to the appearance of memory CD4+ and CD8+ T cells and likely memory B cells (not represented). This adaptive immunity opposes inhibitory mechanisms induced by tumor cells and by inflammation that parallels tumor growth. It counterbalances the role of regulatory T cells (Treg) and of immunosuppressive and/or pro-inflammatory cytokines (IL-10, IL-6…). Cross-presentation of tumor-associated antigens by APCs to CD8+ T cells as well as the presence of myeloid-derived suppressor cells (MDSC) such as M2 macrophages are not represented.

In the text