Stephanie Roth

Bone marrow and tumor cell colony-forming units and human tumor xenograft efficacy of noncamptothecin and camptothecin topoisomerase I inhibitors

Topoisomerase I (TopoI), an established anticancer target, is an enzyme producing a single-strand DNA break during transcription. Several noncamptothecin TopoI inhibitors have been identified. One of these, ARC-111, was compared with two clinically used camptothecins, topotecan and irinotecan/SN-38. In mouse and human bone marrow colony formation [colony-forming units granulocyte-macrophage (CFU-GM)] assays, the IC90 values were 519 and 331 nmol/L for topotecan and SN-38 mouse CFU-GM and were 19 and 26 nmol/L for human CFU-GM, giving mouse to human differentials of 28- and 13-fold. ARC-111 produced IC90 values of 28 nmol/L in mouse and 6.2 nmol/L in human CFU-GM, thus only a 4.5-fold differential between species. Human bone marrow CFU-GM was more sensitive to topotecan than were several human cancer cell lines, but ARC-111 cytotoxicity was similar for human bone marrow CFU-GM and the seven human tumor cell lines tested. In HCT-116 xenografts, tumor growth delays (TGD) were 17 days for irinotecan and 20 days for ARC-111. In HT-29 xenografts, the TGD was 9 days for both irinotecan and ARC-111. Both ARC-111 and docetaxel had a TGD of 21 days in NCI-H460 xenografts, and both ARC-111 and gemcitabine had a TGD of 7 days in MiaPaCa2 xenograft. Current TopoI inhibitors have broad antitumor activity in human tumor xenografts that is not achieved in the clinic. This may be due to greater sensitivity of human bone marrow than mouse to the cytotoxicity of these agents. It may be possible to achieve similar levels of ARC-111 in patients as in mice allowing improved antitumor activity. [Mol Cancer Ther 2008;7(10):3212–22]

Genz-644282, a Novel Non-Camptothecin Topoisomerase I Inhibitor for Cancer Treatment

Purpose: Genz-644282 [8,9-dimethoxy-5-(2-N-methylaminoethyl)-2,3-methylenedioxy-5H-dibenzo[c,h][1,6]naphthyridin-6-one] has emerged as a promising candidate for antitumor agents. This report describes the bone marrow colony-forming unit, granulocyte macrophage (CFU-GM) and tumor cell CFU activity of topoisomerase I (Top1) inhibitors, such as Genz-644282, topotecan, irinotecan/SN-38, and ARC-111, and examines their activity in several human tumor xenograft models. Experimental Design: Colony-forming assays were conducted with mouse and human bone marrow and eight human tumor cell lines. In addition, 29 human tumor cell lines representing a range of histology and potential resistance mechanisms were assayed for sensitivity to Genz-644282 in a 72-hour exposure assay. The efficacy of Genz-644282 was compared with standard anticancer drugs (i.e., irinotecan, docetaxel, and dacarbazine) in human tumor xenografts of colon cancer, renal cell carcinoma, non–small cell lung cancer, and melanoma. Results: Human bone marrow CFU-GM was more sensitive to the Top1 inhibitors than was mouse bone marrow CFU-GM. The ratio of mouse to human IC90 values was more than 10 for the camptothecins and less than 10 for Genz-644282, which had more potency as a cytotoxic agent toward human tumor cells in culture than the camptothecins in the colony-forming and 72-hour proliferation assays. Genz-644282 has superior or equal antitumor activity in the human tumor xenografts than the standard drug comparators. Conclusions: On the basis of preclinical activity and safety, Genz-644282 was selected for development and is currently undergoing phase 1 clinical trial. Clin Cancer Res; 17(9); 2777–87. ©2011 AACR.

Endosialin is expressed in high grade and advanced sarcomas: Evidence from clinical specimens and preclinical modeling

We previously surveyed the expression of endosialin/ CD248/TEM-1 by immunohistochemistry in human clinical specimens of sarcomas and documented expression in tumor cells, stromal cells and vasculature. In the present study, we completed a retrospective analysis of the diagnostic reports available for these same samples in order to identify high-grade and metastatic disease. Our results show that endosialin can be detected in advanced disease. We screened human sarcoma cell lines in vitro for endosialin expression and developed preclinical human xenograft models of disseminated sarcoma. We found that 22 out of 42 human sarcoma cell lines were positive for endosialin with a positive correlation between mRNA and protein levels. When implanted in vivo, endosialin was expressed at all sites of dissemination. These data provide clinical and preclinical evidence that endosialin can be detected in advanced sarcoma. These results demonstrate for the first time that endosialin is a suitable therapeutic target for poor prognosis and advanced disease.

Anti-Endosialin Antibody–Drug Conjugate: Potential in Sarcoma and Other Malignancies

Endosialin/TEM1/CD248 is a cell surface protein expressed at high levels by the malignant cells of about 50% of sarcomas and neuroblastomas. The antibody–drug conjugate (ADC) anti-endosialin-MC-VC-PABC-MMAE was selectively cytotoxic to endosialin-positive cells in vitro and achieved profound and durable antitumor efficacy in preclinical human tumor xenograft models of endosialin-positive disease. MC-VC-PABC-MMAE was conjugated with anti-endosialin with 3–4 MMAE molecules per ADC. The anti-endosialin-MC-VC-PABC-MMAE conjugate was tested for activity in four human cell lines with varied endosialin levels. The HT-1080 fibrosarcoma cells do not express endosialin, A-673 Ewing sarcoma cells and SK-N-AS neuroblastoma cells are moderate expressers of endosialin, and SJSA-1 osteosarcoma cells express very high levels of endosialin. To determine whether endosialin expression was maintained in vivo, A-673 Ewing sarcoma, SK-N-AS neuroblastoma, and SJSA-1 osteosarcoma cells were grown as xenograft tumors in nude mice. The SK-N-AS neuroblastoma and the A-673 Ewing sarcoma lines were selected for in vivo efficacy testing of the anti-endosialin-MC-VC-PABC-MMAE conjugate. The treatment groups included a vehicle control, unconjugated anti-endosialin, an admix control consisting of anti-endosialin and a dose of free MMAE equivalent to the dose administered as the ADC, and the anti-endosialin-MC-VC-PABC-MMAE conjugate. The unconjugated anti-endosialin had no antitumor activity and resulted in similar tumor growth as the vehicle control. The admix control produced a modest tumor growth delay. Administration of the anti-endosialin-MC-VC-PABC-MMAE conjugate resulted in a marked prolonged tumor response of both xenograts. These proof-of-concept results break new ground and open a promising drug discovery approach to these rare and neglected tumors. Mol Cancer Ther; 14(9); 2081–9. ©2015 AACR.

Endosialin: A novel malignant cell therapeutic target for neuroblastoma

Endosialin emerged recently as a potential therapeutic target for sarcoma. Since some sarcoma subtypes, such as Ewings sarcoma, show characteristics of neuroendocrine differentiation, we wondered whether cancers with neuro-endocrine properties and/or neuroectodermal origin, such as neuroblastoma, small cell lung cancer and melanoma, may express endosialin. Endosialin protein expression was surveyed in neuroblastoma, small cell lung cancer and melanoma in human clinical specimens by immunohistochemistry (IHC) and in human cell lines by flow cytometry. Side population cells were examined to determine whether cancer stem cells can express endosialin. Endosialin-expressing neuroblastoma cell lines were implanted in immunodeficient mice and allowed to grow. The xenograft tumors were resected and tested for endosialin expression by IHC. In human clinical specimens, vascular endosialin staining was observed in neuroblastoma, small cell lung cancer and melanoma. Malignant cell staining was strongest in neuroblastoma, weak in melanoma and rare in small cell lung cancer. In human cell lines, endosialin was detected in neuroblastoma cell lines, including cancer stem cell-like side population (SP) cells, but was absent in melanoma and was both rare and weak in small cell lung cancer. Human neuroblastoma xenograft tumors were found to be positive for endosialin. Our work suggests that endosialin may be a suitable therapeutic target for neuroblastoma.

Pharmacokinetics in Mice Implanted With Xenografted Tumors After Intravenous Administration of Tasidotin (ILX651) or Its Carboxylate Metabolite

The pharmacokinetics of tasidotin (ILX651), a depsipeptide currently in phase II for the treatment of advanced solid tumors, and tasidotin-C-carboxylate, the main metabolite, were characterized in male nude mice implanted with LOX tumors, which are sensitive to tasidotin, or H460 tumors, which are resistant to tasidotin. The pharmacokinetics of tasidotin and its metabolites were characterized after singledose administration of tasidotin (20 and 120 mg/kg), tasidotin-C-carboxylate (150 mg/kg), or tasidotin (53 mg/kg) in the presence and absence of Z-prolyl prolinal (5 mg/kg administered 1 hour prior to tasidotin administration), a competitive antagonist of prolyl oligopeptidase, the enzyme responsible for the metabolism of tasidotin to tasidotin-C-carboxylate. A secondary study was done comparing tumor growth in tasidotin-treated mice with implanted LOX tumors in the presence and absence of Z-prolyl-prolinal. After tasidotin administration, the pharmacokinetics of tasidotin and tasidotin-C-carboxylate were similar in plasma and tumors in LOX- and H460-implanted mice, indicating the resistance was not due to pharmacokinetic factors. Tumor carboxylate concentrations were much higher than in plasma after tasidotin administration. The metabolite appeared to contribute ∼17% to 33% to the total exposure in LOX tumors and 20% to 49% in H460 tumors but <5% in plasma. Less than 5% of the administered tasidotin dose was converted to tasidotin-C-carboxylate, with no apparent differences between LOX- and H460-treated animals. The presence of Z-prolyl-prolinal decreased the amount of tasidotin converted to tasidotin-C-carboxylate from 5.5% to 0.90%, a reduction of almost 80%. After tasidotin-C-carboxylate administration, the half-life was on the order of minutes compared with hours when observed after tasidotin administration. Tasidotin-C-carboxylate elimination was not dependent on tasidotin pharmacokinetics, suggesting that the rate of efflux from cells into plasma was the rate-limiting step in its elimination. Tasidotin-C-carboxylate was also further metabolized to desprolyl-tasidotin-C-carboxylate, although the metabolite ratios were <10%. Pretreatment with Z-prolyl-prolinal completely abolished the antitumor activity of tasidotin, indicating that the metabolite is the main moiety responsible for activity and that, despite tasidotin itself having activity in vitro, tasidotin is acting mainly as a prodrug.

Abstract 1388: sFLT01: An anti-angiogenic fusion protein that neutralizes placental growth factor (PlGF) and vascular endothelial growth factor (VEGF)

Proceedings: AACR 101st Annual Meeting 2010‐‐ Apr 17‐21, 2010; Washington, DC Introduction: PlGF and VEGF stimulate angiogenesis and promote the growth of tumor vasculature. PlGF is a member of the VEGF family and binds to VEGFR1. sFLT01 is a novel fusion protein comprised of the Fc portion of human IgG1 and the PlGF- and VEGF-binding domain of VEGFR1/Flt-1. The properties of sFLT01 and the potential of sFLT01 as an anti-angiogenic agent to inhibit tumor growth were investigated in several in vitro assays and in multiple xenograft tumor models. Methods: The binding kinetics of sFLT01 for both the human and murine homologues of PlGF and VEGF were assessed by Biacore. The abilities of recombinant human PlGF and VEGF to induce endothelial cell and pericyte proliferation and of sFLT01 to inhibit this stimulation were investigated in cell-based assays. The secretion of human VEGF and PlGF in culture by the HT29 colon carcinoma, H460 lung carcinoma, and A673 sarcoma human cell lines was quantified by ELISA. In efficacy studies, sFLT01 was administered by intraperitoneal injection twice per week to immunodeficient mice bearing HT29, H460, or A673 subcutaneous tumors. Antibodies specific for human IgG and VEGR2 were applied to A673 sarcoma tumor sections from mice treated with sFLT01 to visualize sFLT01 in the tumors and determine VEGFR2 expression in the cellular components. Pericytes and endothelial cells were identified with antibodies against NG2 and CD31. Results: The Biacore results indicated that sFLT01 has high affinity for human and murine PlGF and VEGF. Human recombinant PlGF and VEGF each induced the proliferation of human pericytes and endothelial cells in culture. This stimulation was inhibited by sFLT01. A673 sarcoma, HT29 colon and H460 lung carcinoma cells secreted higher levels of VEGF than PlGF in culture. In vivo, 10 mg/kg sFLT01 was effective at significantly slowing the growth of HT29 colon carcinoma and A673 sarcoma tumors compared to controls. Further analysis of the A673 sarcoma tumors in sFLT01-treated mice by immunohistochemistry revealed that sFLT01 penetrated multiple areas of the tumor. sFLT01 was detected in the vasculature, stroma, necrotic areas, and adjacent to malignant cells. sFLT01 treatment in the A673 model disrupted vessel integrity with a lack of association between endothelial cells and pericytes. A673 sarcoma cells expressed VEGFR2 in vivo. Conclusion: sFLT01 neutralizes the angiogenic activity of multiple vasculogenic VEGF family members in vitro and inhibited the proliferation of cells that form blood vessels, endothelial cells and pericytes. In vivo, sFLT01 treatment resulted in disorganized tumor vasculature thereby slowing the growth of xenografts tumors. The expression of VEGFR2 in A673 sarcoma tumors suggests that VEGF may play a role in autocrine signaling in some malignant cells. sFLT01 has antitumor and antiangiogenic activity in several human tumor xenografts and may offer therapeutic benefit. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 1388.

Abstract 5251: Endosialin: A novel cell surface therapeutic target for early-stage and late-stage neuroblastoma

Hypothesis: Endosialin emerged recently as a potential marker and therapeutic target for adult and pediatric sarcoma. Given evidence of a possible common progenitor cell for mesenchymal and neural cell lineages, we wondered whether expression of endosialin may be shared by sarcomas and neuroblastomas, which are cancers of mesenchymal origin and neural crest origin, respectively. Methods: Endosialin protein expression was studied in live human neuroblastoma cell lines by flow cytometry using a fully human monoclonal antibody against human endosialin. Endosialin-positive human neuroblastoma cells were subsequently implanted at different anatomic sites in nu/nu mice and allowed to grow. Tumors were collected, formalin fixed and subjected to immunohistochemistry using a fully human monoclonal antibody against human endosialin. Using the same IHC assay, endosialin protein expression was also studied in formalin-fixed paraffin-embedded human clinical specimens of neuroblastoma. Results: We tested 10 human neuroblastoma cell lines for endosialin protein expression by flow cytometry and found that 9/10 expressed endosialin. Several of the positive cell lines were derived from bone marrow metastases, suggesting that endosialin expression is maintained in advanced disease. We modeled endosialin-positive neuroblastoma in vivo by implanting SK-N-AS cells in mice subcutaneously and in the subrenal capsule, kidney being a site where neuroblastoma sometimes originates. Immunohistochemical analysis revealed that SK-N-AS cells, which are positive for endosialin in vitro and derived from the bone marrow metastasis of an adrenal primary tumor, formed endosialin-positive subcutaneous tumors and endosialin-positive renal tumors, demonstrating that endosialin expression is supported by different microenvironments in various anatomic locations. Immunohistochemistry of human clinical specimens of neuroblastoma showed expression of endosialin. All specimens of neuroblastoma were bone marrow metastases, demonstrating that endosialin is expressed in advanced disseminated neuroblastoma. Conclusions: Our work demonstrates that endosialin expression is shared by sarcomas and neuroblastomas. The expression of endosialin in neuroblastoma potentially opens up a new therapeutic horizon for neuroblastoma patients, including those suffering from advanced disease. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 5251.

Abstract A14: sFLT01: An antiangiogenic protein that neutralizes placental growth factor

Introduction: Anti‐angiogenic agents have shown clinical value in combination with chemotherapy by targeting the VEGF pathways. The development of tumor vasculature in human renal cell carcinomas (RCC) is stimulated not only by VEGF but by other angiogenic growth factors such as placental growth factor (PlGF) [1,2]. In the current study, we used a novel fusion protein, sFLT01, that includes human VEGFR1 domain 2 and an IgG1 Fc, which binds to both human VEGF and PlGF. The anti‐angiogenic activity of sFLT01 was evaluated in several preclinical models of RCC. Methods: The binding affinity of sFLT01 for PlGF was evaluated in vitro in binding assays quantified by ELISA and Octet and in proliferation assays where endothelial cells are stimulated by PlGF. The ability of sFLT01 to slow tumor growth was evaluated in the 786‐O human RCC xenograft model as well as in the mouse RENCA RCC orthotopic and subcutaneous tumor models. sFLT01 was delivered by intraperitoneal injection twice per week at doses ranging from 5–25 mg/kg. Microvessel density (MVD) and other vascular parameters were analyzed in the syngeneic and xenograft tumors by immunohistochemical methods with an antibody against mouse CD31 to identify endothelial cells. Results: sFLT01 bound to human PlGF with great affinity and inhibited HUVEC proliferation with an IC90 of approximately 2.3 nM. Subcutaneous 786‐O RCC tumors were inhibited at a dose of 25 mg/kg and median survival time was increased by 33% in the orthotopic RENCA model. sFLT01 reduced MVD in subcutaneous RENCA tumors by approximately 50% and lumen area by approximately 66%. In the 786‐O xenograft model, 5 or 25 mg/kg sFLT01 reduced the blood vessel area, lumen size, and vessel perimeter compared to control. Conclusions: sFLT01 is effective at inhibiting blood vessel growth in tumors by binding VEGF and PlGF thereby preventing the development of new vasculature and decreasing existing vasculature. sFLT01 reduces intratumoral blood vessel count, prolongs survival, and delays tumor progression in the RENCA and 786‐O renal cell carcinoma models. As an anti‐angiogenic agent, sFLT01 may be useful in treating human renal cell carcinomas. Citation Information: Mol Cancer Ther 2009;8(12 Suppl):A14.

Abstract C220: Genz‐644282, a non‐camptothecin topoisomerase one inhibitor, demonstrates a wide spectrum of in vitro and in vivo antitumor activity

Genz‐644282 (GZ) is a novel non‐camptothecin topoisomerase I (Top1) inhibitor. The in vitro and in vivo activity of GZ and its M1 and M2 metabolites were explored and compared with the activity of camptothecin Top1 inhibitors. In vitro in mouse, rat, dog, and human GZ exhibited high metabolic stability, plasma binding of 88–93% and exhibits concentration dependent partitioning into red blood cells. In vivo, GZ has a large volume of distribution and low‐to‐moderate clearance in mouse, rat and dog. In nude mice, the t 1/2 for GZ is 3.6 h (po), 10.4 h (ip) and 5.1h (iv) and longer in tumor‐bearing mice. In human HCT‐116 colon ca, HT‐29 colon ca and NCI‐H460 NSCLC cells the concentration response for Genz‐6244282, M1 and M2 are the same. Upon 72h exposure of the cells to GZ, M1 or M2 the IC 50 concentrations were 0.5‐0.65 nM and the IC 90 concentrations were 1.8–2 nM. In order to evaluate the antitumor activity of GZ as compared to several approved anticancer agents, the compound was tested in seven xenograft models: LOX‐IMVI melanoma, DLD‐1 and HCT‐15 colon, MDA‐MB‐231 breast, NCI‐H292 and NCI‐H1299 lung ca. GZ was compared against two of its metabolites, Genz‐649974 (GZ‐74) and Genz‐649975 in the HCT‐116 colon ca resulting in comparable activity with GZ‐74. GZ was administered intravenously on a QODx3 schedule for 2 cycles. The tumor growth delay, TGD, (T‐C) and increase in lifespan, ILS, (T/C) for each study are listed in the table below. All of the GZ dosages were well tolerated resulting in a maximum body weight loss of ≤20%, except for the high dosages in the HCT‐15 and NCI‐H292 in which there was a maximum body weight loss of 25.7 and 20.9%, respectively. Based on these findings and other data, GZ was selected to be a development candidate. Citation Information: Mol Cancer Ther 2009;8(12 Suppl):C220.