Kathrin Schwager

Preclinical characterization of DEKAVIL (F8-IL10), a novel clinical-stage immunocytokine which inhibits the progression of collagen-induced arthritis

IntroductionIn this article, we present a comparative immunohistochemical evaluation of four clinical-stage antibodies (L19, F16, G11 and F8) directed against splice isoforms of fibronectin and of tenascin-C for their ability to stain synovial tissue alterations in rheumatoid arthritis patients. Furthermore we have evaluated the therapeutic potential of the most promising antibody, F8, fused to the anti-inflammatory cytokine interleukin (IL) 10.MethodsF8-IL10 was produced and purified to homogeneity in CHO cells and shown to comprise biological active antibody and cytokine moieties by binding assays on recombinant antigen and by MC/9 cell proliferation assays. We have also characterized the ability of F8-IL10 to inhibit arthritis progression in the collagen-induced arthritis mouse model.ResultsThe human antibody F8, specific to the extra-domain A of fibronectin, exhibited the strongest and most homogenous staining pattern in synovial biopsies and was thus selected for the development of a fully human fusion protein with IL10 (F8-IL10, also named DEKAVIL). Following radioiodination, F8-IL10 was able to selectively target arthritic lesions and tumor neo-vascular structures in mice, as evidenced by autoradiographic analysis and quantitative biodistribution studies. The subcutaneous administration route led to equivalent targeting results when compared with intravenous administration and was thus selected for the clinical development of the product. F8-IL10 potently inhibited progression of established arthritis in the collagen-induced mouse model when tested alone and in combination with methotrexate. In preparation for clinical trials in patients with rheumatoid arthritis, F8-IL10 was studied in rodents and in cynomolgus monkeys, revealing an excellent safety profile at doses tenfold higher than the planned starting dose for clinical phase I trials.ConclusionsFollowing the encouraging preclinical results presented in this paper, clinical trials with F8-IL10 will now elucidate the therapeutic potential of this product and whether the targeted delivery of IL10 potentiates the anti-arthritic action of the cytokine in rheumatoid arthritis patients.

A Dose-Escalation and Signal-Generating Study of the Immunocytokine L19-IL2 in Combination with Dacarbazine for the Therapy of Patients with Metastatic Melanoma

Purpose: L19-IL2 is an immunocytokine composed of an antibody fragment specific to the EDB domain of fibronectin, a tumor angiogenesis marker, and of human interleukin-2 (IL2). L19-IL2 delivers IL2 to the tumor site exploiting the selective expression of EDB on newly formed blood vessels. Previously, the recommended dose of L19-IL2 monotherapy was defined as 22.5 million international units (Mio IU) IL2 equivalents. In this study, safety and clinical activity of L19-IL2 in combination with dacarbazine were assessed in patients with metastatic melanoma. Experimental Design: The first 10 studied patients received escalating doses of L19-IL2 on days 1, 3, and 5 in combination with 1 g/m2 of dacarbazine on day 1 of a 3-weekly therapy cycle. Subsequently, 22 patients received L19-IL2 at recommended dose plus dacarbazine. Up to six treatment cycles were given, followed by a maintenance regimen with biweekly L19-IL2. Results: The recommended dose of L19-IL2 in combination with dacarbazine was defined as 22.5 Mio IU. Toxicity was manageable and reversible, with no treatment-related deaths. Twenty-nine patients were evaluable for efficacy according to Response Evaluation Criteria in Solid Tumors (RECIST). In a centralized radiology analysis, eight of 29 (28%) patients achieved a RECIST-confirmed objective response, including a complete response still ongoing 21 months after treatment beginning. The 12-month survival rate and median overall survival of the recommended dose–treated patients (n = 26) were 61.5% and 14.1 months, respectively. Conclusions: The repeated administration of L19-IL2 in combination with dacarbazine is safe and shows encouraging signs of clinical activity in patients with metastatic melanoma. This combination therapy is currently evaluated in a randomized phase II trial with patients with metastatic melanoma. Clin Cancer Res; 17(24); 7732–42. ©2011 AACR.

A Traceless Vascular‐Targeting Antibody–Drug Conjugate for Cancer Therapy

Monoclonal antibodies have demonstrated considerable utility in the clinical treatment of cancer, but unmodified immunoglobulins are rarely curative, especially when used as single agents. Thus, there is considerable interest in arming antibodies with bioactive payloads (e.g., drugs, radionuclides, cytokines), to improve their potency and selectivity, thus increasing activity at the tumor site while sparing normal tissues. Significant progress has beenmade in the past few years in the area of antibody–drug conjugates (ADCs) for the selective delivery of cytotoxic drugs to tumors. As a result of these investigations, new agents with pronounced clinical activities have been developed, including SGN-35 (an ADC directed against CD30-positive hematological malignancies) and trastuzumab-DM1 (which has shown activity in metastatic breast cancer). As conventional drug conjugation strategies yield heterogeneous ADC preparations, intense efforts are being devoted to the development of methods for site-selective modification of therapeutic antibodies, thus leading to products with improved performance and batch-tobatch reproducibility. Furthermore, comparative evaluations of intact immunoglobulins in IgG format and other recombinant antibody formats for ADC development have been conducted. It is generally assumed that ADCs may need to be internalized by the tumor cells for the active release of cytotoxic drugs. Once ADCs are internalized and the drug is released in the intracellular compartments, a cross-fire effect (corresponding to the migration to neighboring cells) may occur, as has been reported for the treatment of tumors consisting of a mixture of antigen-positive and antigennegative cells. However, monoclonal antibodies specific to tumor cell antigens often exhibit limited diffusion into the solid tumor mass by several mechanisms, including slow extravasation and antibody trapping by perivascular tumor cells (the so-called antigen barrier). In view of the fact that the formation of new blood vessels (angiogenesis) is a rare process in a healthy adult but a characteristic feature of virtually all types of aggressive cancers, it would be reasonable to develop vascular-targeting ADCs. Unlike the use of cell-specific markers, vascular targeting offers comprehensive tumor coverage, as the majority of cancers express splice isoforms of tenascin-C and of oncofetal fibronectin. In addition, vascular targeting helps address the issue of heterogeneity of antigen expression within the tumor mass (i.e., tumor cells which are positive or negative for the antigen). Despite the remarkable potency of cytotoxic compounds targeting the tumor vasculature and the strong dependence of growing neoplastic masses on florid angiogenesis, only limited efforts were directed in the past towards the investigation of ADCs that target tumor vascular antigens. We have recently shown that the antibody-based delivery of photosensitizers to tumor blood vessels and irradiation may induce complete and long-lasting cancer eradication, in a process that also involves the action of natural killer cells. Thus, there appears to be a strong rationale for the targeted delivery of cytotoxic agents to the tumor neovasculature for cancer therapy. Given that antibodies are large molecules compared to cytotoxic agents, potent drugs need to be used to generate ADCs that can be administered at reasonably low doses and that are compatible with industrial development activities at acceptable cost of goods. Herein, we aimed at generating a novel class of chemically defined vascular-targeting ADCs that release cytotoxic drugs with a mechanism that does not require antibody internalization. We reasoned that ADCs based on linkerless antibody modification with a potent thiolcontaining drug would allow the formation of homogeneous products by the formation of a mixed disulfide. These agents could release the cytotoxic payload in the extracellular space, when tumor cell death is initiated and releases high concentrations of reducing agents (e.g., cysteine, glutathione) to the surrounding environment. Provided that a sufficiently large amount of ADC can be delivered to the subendothelial extracellular matrix, the drug release process would be amplified as tumor cell death progresses. Dolastatins are a group of small peptides isolated from the Indian ocean hare Dolabella auricularia that bind to tubulin subunits and inhibit new microtubule assembly and depolymerize existing microtubules, thus blocking cell cycle [*] Dr. G. J. L. Bernardes, Dr. S. Tr!ssel, I. Hartmann, Dr. J. Scheuermann, Prof. D. Neri Department of Chemistry and Applied Biosciences Swiss Federal Institute of Technology (ETH Z!rich) Wolfgang-Pauli Str. 10, 8093 Z!rich (Switzerland) E-mail: dario.neri@pharma.ethz.ch Dr. G. Casi, Dr. K. Schwager Philochem AG, Libernstrasse 3, 8112 Otelfingen (Switzerland) [**] G.J.L.B. is an EMBO and Novartis Foundation Research Fellow. We thank Katrin Gutbrodt for her help in histology and immunofluorescence experiments, Dr. Kathrin Zuberb!hler and Nadine Pasche for their help during therapy experiments, and Dr. Eveline Traschel for treating tumor-bearing mice with IgG(F8). Financial contribution from ETH Z!rich, Swiss National Science Foundation, SwissBridge/Stammbach Stiftung, Kommission f!r Technologie und Innovation (KTI) and Philochem AG is gratefully acknowledged. Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/anie.201106527. Angewandte Chemie

Paclitaxel Enhances Therapeutic Efficacy of the F8-IL2 Immunocytokine to EDA-Fibronectin–Positive Metastatic Human Melanoma Xenografts

The selective delivery of bioactive agents to tumors reduces toxicity and enhances the efficacy of anticancer therapies. In this study, we show that the antibody F8, which recognizes perivascular and stromal EDA-fibronectin (EDA-Fn), when conjugated to interleukin-2 (F8-IL2) can effectively inhibit the growth of EDA-Fn-expressing melanomas in combination with paclitaxel. We obtained curative effects with paclitaxel administered before the immunocytokine. Coadministration of paclitaxel increased the uptake of F8 in xenografted melanomas, enhancing tumor perfusion and permeability. Paclitaxel also boosted the recruitment of F8-IL2-induced natural killer (NK) cells to the tumor, suggesting a host response as part of the observed therapeutic benefit. In support of this likelihood, NK cell depletion impaired the antitumor effect of paclitaxel plus F8-IL2. Importantly, this combination reduced both the tumor burden and the number of pulmonary metastatic nodules. The combination did not cause cumulative toxicity. Together, our findings offer a preclinical proof that by acting on the tumor stroma paclitaxel potentiates the antitumor activity elicited by a targeted delivery of IL2, thereby supporting the use of immunochemotherapy in the treatment of metastatic melanoma.

A chemically modified antibody mediates complete eradication of tumours by selective disruption of tumour blood vessels

Background:The possibility of eradicating cancer by selective destruction of tumour blood vessels may represent an attractive therapeutic avenue, but most pharmaceutical agents investigated so far did not achieve complete cures and are not completely specific. Antibody conjugates now allow us to evaluate the impact of selective vascular shutdown on tumour viability and to study mechanisms of action.Methods:We synthesised a novel porphyrin-based photosensitiser suitable for conjugation to antibodies and assessed anticancer properties of its conjugate with L19, a clinical-stage human monoclonal antibody specific to the alternatively spliced EDB domain of fibronectin, a marker of tumour angiogenesis.Results:Here we show in two mouse model of cancer (F9 and A431) that L19 is capable of highly selective in vivo localisation around tumour blood vessels and that its conjugate with a photosensitiser allows selective disruption of tumour vasculature upon irradiation, leading to complete and long-lasting cancer eradication. Furthermore, depletion experiments revealed that natural killer cells are essential for the induction of long-lasting complete responses.Conclusions:These results reinforce the concept that vascular shutdown can induce a curative avalanche of tumour cell death. Immuno-photodynamic therapy may be particularly indicated for squamous cell carcinoma of the skin, which we show to be strongly positive for markers of angiogenesis.

Antibody-based targeting of interferon-alpha to the tumor neovasculature: a critical evaluation

The antibody-mediated targeted delivery of cytokines, growth factors and immunomodulators offers great potential for the therapy of cancer and other serious conditions. Interferon-alpha has long been used in the clinic for the treatment of patients with certain malignancies or with viral disease. Promising anticancer activity has recently been reported for two fusion proteins consisting of immunoglobulins bearing the interferon-alpha polypeptide at the C-terminal end of the molecule. Here we describe the design, production and characterization of a novel immunocytokine, in which murine interferon-alpha2 was sequentially fused with the tumor-targeting antibody fragment scFv(F8), specific to the alternatively-spliced EDA domain of fibronectin. The resulting fusion protein (F8-IFNa) could be produced to homogeneity and was shown to retain both antigen binding activity and interferon-alpha activity. Biodistribution studies in tumor-bearing mice with radioiodinated protein preparations confirmed the ability of F8-IFNa to selectively localize at the tumor site. However, using two different murine models of cancer (F9 teratocarcinomas and Cloudman S91 melanomas in immunocompetent mice), we could not detect a striking superiority for the therapeutic performance of F8-IFNa as compared to KSF-IFNa, a fusion protein of irrelevant specificity in the mouse which was used as negative control. In the paper, we present hypotheses why the antibody-based pharmacodelivery of interferon-alpha fails to eradicate tumors, in contrast to the situation observed by our group for other immunocytokines, which benefit from a selective localization at the tumor site.

The Immunocytokine L19–IL2 Eradicates Cancer When Used in Combination with CTLA-4 Blockade or with L19-TNF

Systemic high-dose IL2 promotes long-term survival in a subset of metastatic melanoma patients, but this treatment is accompanied by severe toxicities. The immunocytokine L19-IL2, in which IL2 is fused to the human L19 antibody capable of selective accumulation on tumor neovasculature, has recently shown encouraging clinical activity in patients with metastatic melanoma. In this study, we have investigated the therapeutic performance of L19-IL2, administered systemically in combination with a murine anti-CTLA-4 antibody or with a second clinical-stage immunocytokine (L19-TNF) in two syngeneic immunocompetent mouse models of cancer. We observed complete tumor eradications when L19-IL2 was used in combination with CTLA-4 blockade. Interestingly, mice cured from F9 tumors developed new lesions when rechallenged with tumor cells after therapy, whereas mice cured from CT26 tumors were resistant to tumor rechallenge. Similarly, L19-IL2 induced complete remissions when administered in a single intratumoral injection in combination with L19-TNF, whereas the two components did not lead to cures when administered as single agents. These findings provide a rationale for combination trials in melanoma, as the individual therapeutic agents have been extensively studied in clinical trials, and the antigen recognized by the L19 antibody has an identical sequence in mouse and man.

The Immunocytokine F8-IL2 Improves the Therapeutic Performance of Sunitinib in a Mouse Model of Renal Cell Carcinoma

PURPOSE We investigated the therapeutic action of F8-IL2, a fusion protein consisting of the F8 antibody specific to the alternatively spliced extradomain-A of fibronectin, in diabody format and of human interleukin-2 in the Caki-1 (ATCC®) model of human renal cell carcinoma grafted subcutaneously in nude mice. MATERIALS AND METHODS F8-IL2 was cloned, expressed in CHO cells and purified to homogeneity. This immunocytokine was administered alone or combined with 3 standard drugs commonly used as therapy for kidney cancer, including sunitinib, sorafenib and interferon-α, in 2 sets of doses and treatment schedules. RESULTS Neither F8-IL2 nor any other therapeutic agent cured tumor bearing mice when used as a single agent. The best therapeutic results were observed for the combination of sunitinib with F8-IL2 in a continuous administration schedule, which yielded a 28% cure rate and substantial tumor growth retardation. CONCLUSIONS Considering that recombinant interleukin-2 based immunocytokines are now being investigated in several clinical trials in patients with cancer alone or combined with chemotherapy our preclinical results provide a motivation to study F8-IL2 combined with sunitinib in clinical trials in patients with kidney cancer.

The antibody-mediated targeted delivery of interleukin-10 inhibits endometriosis in a syngeneic mouse model

BACKGROUND Endometriosis is still a highly underdiagnosed disease, and the current medical and surgical treatment of endometriosis is associated with a high recurrence rate. This study investigates the use of derivatives of the human antibody F8, specific to the alternatively spliced extra-domain A of fibronectin (Fn), for the imaging and treatment of endometriosis. METHODS Immunohistochemistry and immunofluorescence was used to evaluate antigen expression in endometriotic tissue of human endometriosis and of a syngeneic mouse model of the disease. The in vivo targeting performance of a fluorescent derivative of the F8 antibody was assessed by imaging mice with endometriosis using a near-infrared fluorescence imager, 24 h following i.v. injection of the antibody conjugate. Furthermore, the mouse model was used for therapy experiments using two recombinant F8-based immunocytokines [F8-interleukin-10 (IL10) and F8-IL2] or saline for the treatment groups. RESULTS A very strong vascular expression of splice isoforms of Fn and of tenascin-C was observed in human endometriotic lesions by immunohistochemistry and immunofluorescence techniques. After i.v. administration, a selective accumulation of the F8 antibody in endometriotic lesions could be observed in a syngeneic mouse model. These targeting data were used as a basis for therapy experiments with a pro-inflammatory (F8-IL2) and an anti-inflammatory (F8-IL10) cytokine fusion protein of the F8 antibody. The average lesion size in the F8-IL10 treatment group was clearly reduced compared with the saline control group and with the F8-IL2 group, for which no therapeutic effects were observed. CONCLUSIONS The F8 antibody targets endometriotic lesions in vivo in a mouse model of endometriosis and may be used for the non-invasive imaging of the disease and for the pharmacodelivery of anti-inflammatory cytokines, such as IL10.

Murine analogues of etanercept and of F8-IL10 inhibit the progression of collagen-induced arthritis in the mouse

IntroductionEtanercept is a fusion protein consisting of the soluble portion of the p75-tumor necrosis factor receptor (TNFR) and the Fc fragment of human IgG1, which is often used for the treatment of patients with rheumatoid arthritis. F8-IL10 is a human immunocytokine based on the F8 antibody and interleukin-10, which is currently being investigated in rheumatoid arthritis with promising clinical results. We have aimed at expressing murine versions of these two fusion proteins, in order to assess their pharmaceutical performance in the collagen-induced model of rheumatoid arthritis in the mouse.MethodsTwo fusion proteins (termed muTNFR-Fc and F8-muIL10) were cloned, expressed in chinese hamster ovary (CHO) cells, purified and characterized. Biological activity of muTNFR-Fc was assessed by its ability to inhibit TNF-induced killing of mouse fibroblasts, while F8-muIL10 was characterized in terms of muIL10 activity, of binding affinity to the cognate antigen of F8, the alternatively-spliced EDA domain of fibronectin, by quantitative biodistribution analysis and in vivo imaging. The therapeutic activity of both fusion proteins was investigated in a collagen-induced mouse model of arthritis. Mouse plasma was analyzed for anti-drug antibody formation and cytokine levels were determined by bead-based multiplex technology. The association of F8-IL10 proteins with blood cells was studied in a centrifugation assay with radiolabeled protein.ResultsBoth fusion proteins exhibited excellent purity and full biological activity in vitro. In addition, F8-muIL10 was able to localize on newly-formed blood vessels in vivo. When used in a murine model of arthritis, the two proteins inhibited arthritis progression. The activity of muTNFR-Fc was tested alone and in combination with F8-huIL10. The chimeric version of F8-IL10 was not better then the fully human fusion protein and showed similar generation of mouse anti-fusion protein antibodies. Incubation studies of F8-muIL10 and F8-huIL10 with blood revealed that only the fully human fusion protein is not associated with cellular components at concentrations as low as 0.2 μg/ml, thus facilitating its extravasation from blood vessels.ConclusionsThe described products may represent useful research tools for the study of the anti-arthritic properties of TNF blockade and of IL10-based immunocytokines in syngeneic immunocompetent models of arthritis.