Treatment of neoplastic disease was initially based on surgery, chemotherapy and radiation. However, since the identification of tumor specific or associated antigens, immunotherapy has evolved rapidly as an attractive alternative (Van Pel and Boon 1982). Notably, immunoregulatory cytokines have been demonstrated to improve anti-tumor immune responses. Indeed, systemic administrations of IL-2, granulocyte macrophage colony stimulating factor (GM-CSF) or IL-12 increase the immunogenicity of some tumors, thereby inducing or boosting immune response to a level that there is effective eradication of the tumor. Additional mechanisms of cytokines include direct effects on tumor or tumor stroma cells; for example, tumor necrosis factors (TNF-a) directly damages the tumor-associated vasculature (Manusama et al. 1998). Physiologically, however, cytokines function as auto- or paracrine factors, with high concentrations only in close vicinity of the producing cell. Thus, systemic administration of cytokines neglects this idea, and doses which are needed in order to see a clinical effect are frequently accompanied by severe side-effects (Keilholz et al. 1997). To overcome this problem, cytokines were directly injected into solid tumors to achieve sufficient concentrations of the cytokines at the tumor site while reducing generalized toxicity (Mattijssen et al. 1994). Similarly, locally disseminated tumors can be amended by locoregional treatment, i.e., by isolated limb perfusion (Eggermont et al. 2003). However, advanced tumors in general are neither localized nor accessible, which are the prerequisites for either of these two approaches. An alternative means to achieve sufficient cytokine concentration at the tumor site which allows even micrometastases to be addressed is the genetic fusion of the cytokine to tumor specific antibodies. This strategy was published in 1991 by two independent research groups (Hoogenboom et al. 1991; Gillies et al. 1991). Importantly, both groups demonstrated that the fusion of cytokines to antibodies impaired neither the binding capacity of the mAb, nor - with the exception of a GM-CSF-antibody fusion protein - the biologic activity of the cytokine. Moreover, the clearance rate of these constructs in most cases lies between that of the mAbs and the cytokine, i.e., a marked prolongation of the serum half-life of the cytokine.
The most elaborately studied antibody-cytokine fusion proteins are those containing IL-2. IL-2 was rec ognized early on to promote T-cell proliferation and to activate the cytotoxic capacity of T and NK cells. Moreover, it is FDA approved for therapy of advanced melanoma or renal carcinoma. An antibody-IL-2 fusion protein specific for the disialoganglioside GD2, an antigen commonly overexpressed on human melanoma and neuroblastoma cells, displayed anti-tumor efficacy in preclinical models in vitro and in vivo. To this end, the ch14.18-IL-2 fusion protein led to the eradication of established experimental and spontaneous metastases in xenogeneic and syngeneic tumor models (Becker et al. 1996). The therapeutic effect depended on the fusion of antibody and cytokine and could not be ascribed to the above mentioned increased in vivo half-life of anti-body-IL-2 fusion proteins since neither the combined treatment of parental antibody and cytokine nor the application of an antibody-IL-2 fusion protein directed to an irrelevant antigen, demonstrated any therapeutic effects. Biodistribution analysis in the preclinical models demonstrated that the tumor specific fusion protein indeed accumulated within the tumor bearing organs (Becker et al. 1996). Interestingly, depending on the tumor model, the antibody-IL-2 fusion protein mediated its therapeutic effect either by innate or adaptive immunity, demonstrating IL-2's broad spectrum of activity (Fig. 15.2). Referring to this, tumor eradication in the melanoma and colon carcinoma models was dependent on T cells, whereas in the neuroblastoma model the efficacy of the antibody-IL-2 fusion protein relied on NK cells. Indeed, in the melanoma model, the
Fig. 15.2. Working mechanism of immunocytokines exemplified for tumor targeted IL-2. A mAb specific for a tumor-associated antigen allows the enrichment of cytokines in the tumor microenvironment. In the case of IL-2 it enhances antibody dependent cellular cytoxicity mediated by Fc-receptor positive effector cells like NK cells. In addition, tumor targeted IL-2 stimulates T cells to expand and attack the tumor. High concentrations of plasmin at the tumor site enable the cleavage of IL-2 from the fusion protein through the plasmin cutting site within the linker antibody-IL-2 fusion protein seemed to exert its therapeutic effect by boosting a pre-existing T-cell response and required CD4+ T-cell help. Consequently, in lymphotoxin-a (LT-a) knock out mice, which are characterized by an impaired induction of immune responses - due to the absence of secondary lymphoid tissue - antibody-IL-2 fusion protein treatment is only effective if primed tumor specific T cells are adoptively transferred prior to IL-2 fusion protein therapy (own unpublished observations).
The effect of antibody-IL-2 fusion proteins was also tested in adjuvant settings. For example, tumor targeted IL-2 clearly enhanced anti-tumor immune responses induced by DNA vaccination. Accordingly, therapeutic vaccination with tumor antigen loaded dendritic cells only induced a therapeutic effect in a murine melanoma model when vaccination was followed by IL-2 administration. This therapeutic effect could be clearly improved by the application of tumor-targeted IL-2 instead of systemic IL-2 treatment (Schra-ma et al. 2004). Interestingly, tumor targeted IL-2 therapy influenced the development of a memory immune response: accordingly treated mice were protected against tumor rechallenge to the same organ, but not directed towards other organs. These findings are substantiated by a previous report demonstrating that high doses of antibody-IL-2 fusion proteins elicited protective immunity without memory in a murine B-cell lymphoma model (Penichet et al. 1998). Intriguingly, systemic IL-2 administration increases the activity ofanti-body-IL-2 fusion protein treatment in a murine neuroblastoma model (Neal et al. 2004). Antibody-IL-2 fusion proteins, however, not only boost vaccination induced immune response, but also increase the immu-nogenicity of antigens. Targeting IL-2 to a soluble, poorly immunogenic antigen triggered an immune response which led to significant tumor growth retardation (Dela Cruz et al. 2003).
These encouraging preclinical data led to the clinical evaluation of antibody-IL-2 fusion proteins. The first phase I trials were conducted to test the efficacy of anti-body-IL-2 fusion proteins specific for GD2 or EpCAM for the treatment of metastastic melanoma (King et al. 2004) or prostate cancer, respectively (Ko et al. 2004). The antibody-IL-2 fusion proteins were generally well tolerated; in very few patients did drug related toxici-ties equal to or larger than grade 3 occur. Translational studies revealed the biological activity of immunocyto-kines, i.e., an increase in total lymphocyte and NK-cell counts as well as enhanced NK-cell and antibody-dependent cellular cytotoxic activity. Although these trials were only designed to evaluate safety and the maximum tolerated dose, the clinical outcomes of the melanoma patients were reported. To this end, 58 % of the treated patients after the first course (three doses of fusion protein) and 28 % at the end of the second course of therapy - which was completed by 52 % of the patients - presented with stable disease. Consequently, a phase II trial investigating the therapeutic activity is currently in preparation (King et al. 2004).
Other antibody-cytokine fusion proteins were generated with GM-CSF, IL-12, IFN-y, TNF-a or LT-a, but have not yet been as thoroughly investigated as the antibody-IL-2 fusion proteins. This may be at least in part due to the fact that their generation encountered some difficulties. For example, antibody-GM-CSF fusion proteins are cleared much faster from the plasma than the parental antibody (Dela Cruz et al. 2000). In addition, the biologic activity of TNF-a, LT-a and IL-12 within the fusion proteins is markedly decreased compared to the recombinant cytokines (Reisfeld et al. 1996). This is probably due to the nature of the cyto-kines: TNF- and LT- are active as trimeric structures and IL-12 as a heterodimer. Nevertheless, all these constructs possessed potent anti-tumor efficiency in pre-clinical models. Antibody-GM-CSF fusion proteins facilitate neutrophil antibody-dependent cellular cyto-toxicity in vitro and elicit a strong anti-tumor immune response eradicating solid tumor in vivo (Dela Cruz et al. 2000). Interestingly, the therapeutic effect of an anti-body-LT- fusion protein, originally designed to exert a direct apoptotic effect on tumor cells, crucially depended on the presence of immune competent cells. In a xenograft model, these were B and NK cells, whereas in a syngeneic melanoma model the anti-tumor effect was mediated by T cells (Schrama et al. 2001). The anti-tumor effect of LT-a fusion protein in the syn-geneic melanoma model was associated with an induction of tertiary lymphoid tissue next to the tumor, which actually may provide all the necessary requirements for T-cell priming. In this regard, these structures contained high endothelial venules which are essential for the emigration of naive T cells from the blood into lymphatic tissue and nai've T and antigen presenting cells (Fig. 15.3). The observations that the T-cell infiltrate within the tumor was tumor-specific and the quantity and clonality of this infiltrate increased over the course of therapy, imply that this tertiary lym-
Fig. 15.3. Induction of tertiary lymphoid tissue by anti-body-LT-a fusion proteins. Targeting LT-a to the tumor induces peritumoral tertiary lymphoid tissue. Macroscopic (A), microscopic (B) and ultrastructural (C) appearance of tertiary lymphoid tissue. Double staining of kryosections with anti-CD8 antibodies (green) and TRP-2/MHC class I tetramers (red) demonstrates the presence of tumor-specific cytotoxic T cells (yellow) in these tissues (B). Electronic microscopy reveals the induction of high endothelial venules, which allows the entry of naïve T cells (C)
phoid structure enabled priming of tumor specific T cells (Schrama et al. 2001).
In conclusion, the severe side effects of systemic administrations of high doses of cytokines can be prevented by targeting cytokines to the site of interest by an antibody and - most importantly - this approach demonstrated therapeutic efficiency in preclinical models. However, the clinical characterization of these molecules has just been started and it will be interesting to see if their preclinical therapeutic potential can be confirmed in clinical trials.
Was this article helpful?
Complete Guide to Preventing Skin Cancer. We all know enough to fear the name, just as we do the words tumor and malignant. But apart from that, most of us know very little at all about cancer, especially skin cancer in itself. If I were to ask you to tell me about skin cancer right now, what would you say? Apart from the fact that its a cancer on the skin, that is.