Tumorgraft Models Research Articles

Combined inhibition of MEK and mTOR has a synergic effect on angiosarcoma tumorgrafts.

Angiosarcoma (AS) is a rare neoplasm of endothelial origin that has limited treatment options and poor five-year survival. Using tumorgraft models, we previously showed that AS is sensitive to small-molecule inhibitors that target mitogen-activated/extracellular-signal-regulated protein kinase kinases 1 and 2 (MEK). The objective of this study was to identify drugs that combine with MEK inhibitors to more effectively inhibit AS growth. We examined the in vitro synergy between the MEK inhibitor PD0325901 and inhibitors of eleven common cancer pathways in melanoma cell lines and canine angiosarcoma cell isolates. Combination indices were calculated using the Chou-Talalay method. Optimized combination therapies were evaluated in vivo for toxicity and efficacy using canine angiosarcoma tumorgrafts. Among the drugs we tested, rapamycin stood out because it showed strong synergy with PD0325901 at nanomolar concentrations. We observed that angiosarcomas are insensitive to mTOR inhibition. However, treatment with nanomolar levels of mTOR inhibitor renders these cells as sensitive to MEK inhibition as a melanoma cell line with mutant BRAF. Similar results were observed in B-Raf wild-type melanoma cells as well as in vivo, where treatment of canine AS tumorgrafts with MEK and mTOR inhibitors was more effective than monotherapy. Our data show that a low dose of an mTOR inhibitor can dramatically enhance angiosarcoma and melanoma response to MEK inhibition, potentially widening the field of applications for MEK-targeted therapy.

Abstract OT1-1-08: Master regulator (MR)-directed therapy in residual breast cancer patient derived xenografts (PDXs)

Abstract Background: Neoadjuvant chemotherapy (NACT) is standard of care in operable breast cancer (BC). There are no therapies that have demonstrated benefit in patients who failed to achieve a pathologic complete response (pCR). There remains an unmet need to identify appropriate targets in which to direct therapy in this high-risk population to decrease BC mortality. Computational approaches to reverse engineering of cell signaling networks from patient tumor-specific RNA expression data can identify key molecular dependencies that result in the malignant phenotype, i.e. MRs. These approaches out-perform traditional methods, such as identifying mutated or over-expressed genes, in determining true tumor dependencies. Successful targeting of MRs may be a means for rational selection of targeted drugs in the operable BC setting. Trial Design: A single center pilot study that will accrue 35 subjects with BC who will have received standard NACT and at least 1.5 cm of residual disease. At the time of definitive surgery, resected tissue will be assessed for residual disease and then allocated. After reserving required tissue for pathologic diagnosis, 0.5 cm3 will be procured for xenografting and 0.5 cm3 will be flash frozen for RNA extraction. Tissue for xenografting is immediately sent to a Champions Oncology site for implantation. We will use fresh frozen tissue to perform RNA-seq. Previously validated computational algorithms at our center, including MARINa and VIPER, will be used to interrogate the expression data and identify the top MRs in individual tumors. Upon maturation of a Champions Oncology Tumor Graft model, the second generation expansion group will be divided into four cohorts. We plan to test two drugs that target tumor-specific MRs, one vehicle control, and one negative control (paclitaxel, as these tumors have demonstrated paclitaxel resistance) in each model. Eligibility Criteria 1. Patients ≥18 years with newly diagnosed BC, deemed candidates for definitive surgery. 2. Subjects must have received standard NACT with taxane (paclitaxel or docetaxel; herceptin +/- pertuzumab if HER2+) and/or adriamycin/cytoxan. 3. Clinical or radiographic evidence of ≥1.5 cm of residual BC. 4. ECOG PS ≤2 Specific Aims: 1) To computationally infer MRs in individual patient tumors resistant to NACT 2) To determine if targeting MRs results in superior tumor growth inhibition in PDX models Statistical Methods: In the PDX model, pre- and post-treatment tumor volume (TV) is calculated: TV= width2 x length x 0.52, and is standardized as a tumor growth inhibition (TGI) percentage compared to the vehicle control. Since we anticipate the MR-directed arms will demonstrate a TGI (80%) twice that of the paclitaxel arm (40%), we have 80% power to detect a difference in 18 patients (one-sided alpha, 0.05). We assume a xenograft take-rate of 50% in the post-NACT setting and thus aim for an N of 35. Final data analysis will be adjusted using ANOVA. Present and Target Accrual: 35 patients, projected to accrue in 18-24 months. Estimate based on number of locally advanced BC patients seen, competing studies, and likelihood of participation. The study will open in Fall 2014. Citation Format: Kevin Kalinsky, Prabhjot Mundi, Dawn L Hershman, Eileen Connolly, Katherine D Crew, Hanina Hibshoosh, Andrea Calfiano, Matthew Maurer. Master regulator (MR)-directed therapy in residual breast cancer patient derived xenografts (PDXs) [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr OT1-1-08.

Identifying and Predicting the Effectiveness of Carboplatin In Vivo and In Vitro and Evaluating its Combination with Paclitaxel

Identifying and predicting the effectiveness of carboplatin and evaluating its combination with paclitaxel to avoid chemo-resistance may exhibit in some advanced malignancies. Models involving in vivo i.p. growth of human ovarian cancer cells (OCC1, CAOV3), mammary (4T1), and non-small cell lung cancer (NSCLC) tumors in athymic nude mice were used. Doses of (180, 60, 100, and 225 mg/kg) carboplatin in those xenograft growths were, respectively, applied. MTT proliferation assay was performed on samples of human ovarian cancer cell line 2774 incubated with different concentrations of carboplatin (20, 40, 60, 80, and 100 μg/mL). Predicting the response of NSCLC tumor model to carboplatin and evaluation to four cycles of carboplatin combined with paclitaxel at 50 and 15 mg/kg, respectively, in ovarian tumorgraft were performed. The energy yield by the carboplatin doses in OCC1, CAOV3, and 4T1 xenograft growths was perfectly correlated (r = 1) with the carboplatin dose logarithm. An efficient dose-energy model with perfect fit (R 2 = 1) has been established estimated energy yield by carboplatin doses. The response of NSCLC tumor model and growth inhibition in samples of cell line 2774 were predicted 100 % identical to those actually induced. Effectiveness of the paclitaxel/carboplatin combination in the ovarian tumorgraft model was ~7 times that induced by carboplatin alone. The therapeutic response to carboplatin should be predicted prior to therapy using dose-energy model to avoid non-optimal treatments. Paclitaxel/carboplatin combination is an effective way to control tumor growth in epithelial ovarian cancer patients using lower doses of carboplatin with acceptable toxicity.

Carbonic Anhydrase Expression in TNBC Breast Cancer Cells and Human Tumor Grafts

Carbonic anhydrase IX (CAIX) has emerged as a potential target for triple negative breast cancer (TNBC). Our goal is to utilize the tumor graft model to demonstrate the efficacy of carbonic anhydrase (CA) inhibitors in reducing tumor load and metastasis. Here, we compare the expression of CAIX in TNBC human breast cancer cell lines with a bank of tumor tissues generated from subject‐derived breast tumor grafts in mice that retain characteristics of the original patient samples (Dr. Alana Welm). In our sampling of TNBC cell lines, CAIX expression was observed in four of the five lines. Two showed exclusive expression of CAIX relative to other cell surface CA family members, allowing us to evaluate catalytic activity. We observed that CAIX activity, measured by 18O exchange between H216O and H13C18O3−, was related to the level of CAIX expressed and inhibited by an impermeant sulfonamide (N‐3500). We then analyzed CAIX expression in tumor‐grafted tissue. Thirteen samples were available and represented the three major breast cancer phenotypes: estrogen and progesterone receptor‐positive (ER/PR), HER2‐positive, and TNBC. Of the six TNBC tissues, four expressed CAIX. No carbonic anhydrase XII (CAXII) was observed. Cryogenically preserved TNBC tissue was used to develop orthotopic tumors in nude mice. Immunohistochemistry showed high mitotic activity with CAIX expression adjacent to necrotic tissue. The tissue was negative for ER/PR and CAXII. These observations demonstrate that the tumor graft model will be useful in evaluating the efficacy of CA inhibitors for potential clinical use. We gratefully acknowledge funding from NIH (GM25253‐DNS and CA165284‐SCF), the UF Cancer Center (SCF), and the CSCRB, Inc (SCF).

A novel patient-derived tumorgraft model with TRAF1-ALK anaplastic large-cell lymphoma translocation.

Although anaplastic large-cell lymphomas (ALCL) carrying anaplastic lymphoma kinase (ALK) have a relatively good prognosis, aggressive forms exist. We have identified a novel translocation, causing the fusion of the TRAF1 and ALK genes, in one patient who presented with a leukemic ALK+ ALCL (ALCL-11). To uncover the mechanisms leading to high-grade ALCL, we developed a human patient-derived tumorgraft (hPDT) line. Molecular characterization of primary and PDT cells demonstrated the activation of ALK and nuclear factor kB (NFkB) pathways. Genomic studies of ALCL-11 showed the TP53 loss and the in vivo subclonal expansion of lymphoma cells, lacking PRDM1/Blimp1 and carrying c-MYC gene amplification. The treatment with proteasome inhibitors of TRAF1-ALK cells led to the downregulation of p50/p52 and lymphoma growth inhibition. Moreover, a NFkB gene set classifier stratified ALCL in distinct subsets with different clinical outcome. Although a selective ALK inhibitor (CEP28122) resulted in a significant clinical response of hPDT mice, nevertheless the disease could not be eradicated. These data indicate that the activation of NFkB signaling contributes to the neoplastic phenotype of TRAF1-ALK ALCL. ALCL hPDTs are invaluable tools to validate the role of druggable molecules, predict therapeutic responses and implement patient specific therapies.

Abstract B25: KRAS mutation status is associated with enhanced dependency on purine biosynthesis and related pathways in non small cell lung cancer cells

Abstract KRAS gene mutation is linked to poor prognosis and resistance to oncology therapeutics in Non Small Cell Lung Cancer (NSCLC). We have explored the possibility of exploiting inherent differences in KRAS mutant cell metabolism to enhance the efficacy of treatment. We have identified a greater dependency on purine biosynthesis and related pathways in KRAS mutant compared to KRAS wild type NSCLC cell lines. Purine synthesis requires factors generated from other metabolic reactions including ribose-5-phosphate from the pentose phosphate pathway, THF cofactors from folate metabolism and glycine / amide nitrogen groups from glutamine and aspartate metabolism. In this study, microarray gene expression and biological pathway analysis identified higher expression of purine synthesis and accessory pathways such as folate metabolism, 5-aminoimidazole ribonucleotide biosynthesis and glycine synthesis pathways in KRAS mutant NSCLC cells compared to wildtype counterparts. KRAS knockdown and overexpression studies demonstrated the ability of KRAS to regulate expression of genes that comprise purine synthesis and folate metabolism pathways. Moreover, pathway analysis and knockdown studies suggest a role for MYC, an oncogene previously recognized to be associated with KRAS mutant tumors, in the regulation of these pathways in KRAS mutant NSCLC cells. Proliferation studies demonstrated higher responsiveness to antifolates such as methotrexate and pemetrexed in KRAS mutant NSCLC cells, both of which may interfere indirectly with purine biosynthesis. In vivo analysis of NSCLC tumorgraft models in nude mice also identified an association between KRAS mutant tumor status and response to pemetrexed. The expression of a KRAS driven folate/purine synthesis gene, Methylenetetrahydrofolate Dehydrogenase 2 (MTHFD2), was also correlated with antifolate activity suggesting its use as a possible biomarker of response to antifolates. We propose that KRAS mutation drives increased purine synthesis activity and as a result an elevated dependency on the factors needed to feed this biosynthetic pathway such as those generated by folate metabolism. Thus, antifolates can indirectly inhibit purine synthesis through the depletion of folate cofactors which may account for the stronger response to these agents in KRAS mutant cells. We are currently expanding this study to examine alternative inhibitors of purine synthesis as possible therapeutics in KRAS mutant NSCLC and other cancers. Collectively, our findings highlight that a better understanding of the molecular mechanisms underlying the dependency of cancer cells on specific metabolic pathways may result in more effective metabolic targeting and new approaches in treating specific cancers. Citation Format: Diarmuid M. Moran, Patricia B. Trusk, Karen Pry, Keren Paz, David Sidransky, Sarah S. Bacus. KRAS mutation status is associated with enhanced dependency on purine biosynthesis and related pathways in non small cell lung cancer cells. [abstract]. In: Proceedings of the AACR Special Conference on RAS Oncogenes: From Biology to Therapy; Feb 24-27, 2014; Lake Buena Vista, FL. Philadelphia (PA): AACR; Mol Cancer Res 2014;12(12 Suppl):Abstract nr B25. doi: 10.1158/1557-3125.RASONC14-B25

The Use of Patient-Derived Xenograft (PDX) Models to Predict Patient Response in Non-Small Cell Lung Cancer (NSCLC): Metastatic Non-Small Cell Lung Cancer

Advances in the treatment of NSCLC have been hampered by the lack of reproducible, biologically-relevant in vivo preclinical models. Patient-derived xenograft (PDX or TumorGraft) models have been shown to accurately reflect the characteristics of a patient’s tumor including drug response and may be useful tools for selecting personalized treatments. We sought to determine the ability of TumorGraft models from a cohort of NSCLC patients to predict a patient’s clinical response to drug treatment.

Abstract 4612: JQ1 suppresses tumor growth in tumorgraft models of pancreatic ductal adenocarcinoma

Abstract Pancreatic ductal adenocarcinoma (PDAC), the most common form of pancreatic cancer, is one of the most deadly of malignancies. Patients often present late in the course of disease limiting treatment options, with the majority of patients receiving chemotherapy. However, currently available chemotherapy has not impacted overall patient survival, and new therapies are urgently needed. Recently, our lab has documented that tumorgrafts derived specifically from primary PDAC tumors retain characteristics of the tumors of origin. In terms of drug evaluation, tumorgraft models have been shown to be predictive of clinical utility because these models retain tumor heterogeneity, recapitulate tumor architecture, and contain human stroma. To date, most efforts to use tumorgraft models to identify effective agents for the treatment of PDAC have taken an “all comers approach”; but this approach has thus far identified no effective therapies. We proposed that JQ1, a compound that targets the oncogene c-Myc, would be efficacious for the treatment of PDAC, based on molecular characteristics of these tumors. The oncogene c-Myc is amplified or overexpressed in a 30-45% of primary or metastatic PDAC tumors. Also, overexpression of this oncogene is sufficient to induce tumor formation in a genetically engineered mouse model of PDAC, suggesting a critical role in the tumorigenesis of this tumor type. JQ1 is a novel compound in that it is a relatively specific bromodomain inhibitor. Bromodomain containing proteins recognize acetylated lysine residues on histones and direct the assembly of macromolecular molecules to the chromatin for transcription. A specific subfamily of these proteins, known as BETs (bromodomain and extra-terminal), recruit c-Myc to specific sites for transcription. Exposure of PDAC cells to JQ1 decreases the expression and therefore the activity of c-Myc. To test the effectiveness of JQ1 in vivo, we administered 50 mg/kg of JQ1 i.p. once a day for 28 days and monitored tumor growth in a panel of 5 PDAC tumorgraft models. This dose and schedule of JQ1 administration was well tolerated in mice with no significant toxicity. Our data indicate that JQ1 suppressed tumor growth in all 5 models, compared to vehicle control treated mice. The data also showed a modest, JQ1-induced down regulation of c-Myc and NFkB. We conclude that JQ1 and other bromodomain inhibitors warrant further investigation as potentially effective agents for the treatment of PDAC. This work was supported by UAB/UMN SPORE in pancreatic cancer (P50 CA101955). Citation Format: Patrick L. Garcia, Tracy Gamblin, Leona N. Council, John D. Christein, J. Pablo Arnoletti, Martin J. Heslin, Joseph H. Richardson, Jun Qi, Jay E. Bradner, Karina J. Yoon. JQ1 suppresses tumor growth in tumorgraft models of pancreatic ductal adenocarcinoma. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 4612. doi:10.1158/1538-7445.AM2014-4612

Abstract 4284: Evaluation of cancer-related mutations in tumorgraft models

Abstract Personalized TumorGraft models are generated from tumor fragments resected from patients with refractory advanced cancers that are implanted and propagated in immunodeficient mice. They maintain the genetic characteristics of the original tumor and can potentially be used for individualized therapy selection. We used samples from 120 TumorGrafts to analyze genetic changes (353 mutations and 31 copy number assays) with a customized Qiagen qBiomarker Somatic mutation PCR array. We chose the alterations to be analyzed based on the recent and classic literature data on significant known somatic mutations in cancer, either with biological or therapy response relevance in different cancer types (prioritizing lung and colorectal tumors). Our models consisted of: 39 colorectal cancer samples, 30 pancreatic cancers, 25 non-small cell lung cancers (NSCLC), 11 melanomas, 7 ovarian cancers, 4 breast cancers, and 4 small cell lung cancers. Among our 120 samples, 22% (26) did not show any of the analyzed mutations or deletions, 78% (94) showed at least one mutation, and of those: 56 had 2 or more mutations (44%). In colorectal tumors, the most frequently observed mutation was Kras, followed by p53, and PIK3CA. Among NSCLC: p53, PIK3CA, deletion of p16 and mutation of Kras. In melanoma: Braf and deletion of p16. All 30 cases of pancreatic cancer showed mutations, being Kras the most frequent, followed by p16 deletion and p53 mutation. On the other tumor types, not more than three samples displayed the same mutation. A subgroup of the samples (95) was validated for mutations with other two techniques: conventional Sanger sequencing and/or IonAmpliseq Cancer panel (a panel of key cancer related genes, Life Technologies). 56/69 (81%) cases had the same mutation(s) detected by two or three techniques simultaneously. Three samples were concordantly wild type for the evaluated genes by three or two techniques. Comparison of the genetic alterations with the demographic and clinicopathological data available for these samples (including tumor stage, therapy response, patient's gender and age) is underway to evaluate possible correlations among these parameters and the molecular changes. With the recent massive data produced by next generation sequencing, and with our growing knowledge on the applicability of genetic alterations in the clinic, it is important to have cost-effective, easy straight- forward to perform and to analyze techniques in order to profile the tumors for relevant mutations on the era of personalized medicine. In the present work, we assess a quick and simple, real time PCR based method with selected targets for mutation detection and compare with two other widely used methods. Moreover we aim to profile tumors and uncover their possible genetic targets to best utilize our powerful investigational method of TumorGrafts. Further studies with larger cohorts and alternative mutation detection techniques are needed to validate and expand the present results. Citation Format: Mariana Brait, Luciane T. Kagohara, Evgeny Izumchenko, Samuel Long, Tin Khor, Elizabeth Bruckheimer, David Sidransky. Evaluation of cancer-related mutations in tumorgraft models. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 4284. doi:10.1158/1538-7445.AM2014-4284

Abstract 1674: Humanized mouse models for personalized preclinical testing of monoclonal antibodies targeting immune checkpoints

Abstract The blockade of immune checkpoints with monoclonal antibodies (mAbs) is a promising therapeutic avenue, with durable objective responses observed in patients with solid tumors, particularly melanoma, non-small cell lung cancer (NSCLC) and renal cell carcinoma. Preclinical models that recapitulate a functional human immune system will therefore be essential tools for the continued investigation of immunotherapy approaches. Champions Oncology is engaged in advanced personalized solutions and our TumorGraft models have been developed and extensively characterized as a platform for use in personalizing cancer patient care, as well as pharmaceutical development. However, because TumorGrafts are established by engrafting patient tumors into immune-deficient mice, the therapeutic efficacy of immune-modulatory drugs cannot be directly examined. To circumvent this limitation, we reconstituted the human immune system by engrafting human hematopoietic cells (HLA-A2; CD34+) into immune-compromised mice (PrkdcscidIl2rgtm1Sug) carrying the scid mutation and a targeted mutation of the Il2r-gamma gene. Ten to twelve weeks later, mature CD45+ human T cells could be detected in these mice, at which time TumorGrafts were established, followed by drug-sensitivity testing with various mAbs targeting the immune system. Fifty six well-established melanoma, colorectal, breast and lung TumorGraft models were characterized and selected with regard to their HLA type and other molecular characteristics such as BRAF mutation status (in melanoma) and KRAS mutation status (in colorectal and lung cancer) as well as expression of PD-L1. With the present study, we demonstrated the potential of combining the humanized mouse with Champions TumorGraft to generate a preclinical platform for assessing the therapeutic value of mAbs targeting immune checkpoints in various solid tumors. Citation Format: Gilson S. Baia, David Vasquez, Daniel Ciznadija, Brandy Wilkinson, David Sidransky, Amanda Katz, Keren Paz. Humanized mouse models for personalized preclinical testing of monoclonal antibodies targeting immune checkpoints. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 1674. doi:10.1158/1538-7445.AM2014-1674

Abstract 3115: Calcitriol effects on breast cancer tumorgrafts

Abstract Antiproliferative effects of calcitriol were mainly detected in breast carcinoma lineages exposed in vitro to high hormone concentrations, which is associated with hypercalcemia in human beings. To test the hypothesis that intra-tumoral administration of calcitriol would allow higher tissue concentration of the hormone and activation of the genomic pathway, a tumorgraft model, that more closely reproduces the molecular characteristics of the primary tumor, was established. Freshly collected breast cancer samples from 14 patients were ortothopically grafted in nude mice. On the sixth week after xenografting, no palpable nodules were detected in mice implanted with 6 samples. Intra-tumoral weekly injections of calcitriol 0.06 μg (dose that allow peak serum calcitriol levels in the predicted therapeutic range) or vehicle were initiated and maintained for six weeks in mice xenografted with another 8 samples. Tumorgraft proliferation, apoptosis and target genes expression were evaluated through immnunohistochemistry reactions or RT-PCR. At the end of the experiment, tumorgrafts originated from three samples were available for analysis. VDR expression was detected in these samples as well as a trend towards higher expression of CYP24A1 mRNA (10-18 fold) in calcitriol treated samples, indicating that the genomic pathway was induced. A high proliferative index, evaluated by Ki67 expression, was observed, however, neither differences in the expression of proliferation markers (BrdU incorporation, Ki67 and CDKN1B expression) nor in apoptosis markers (cleaved caspase 3 and BCL2 expression) between vehicle and calcitriol treated tumorgrafts was detected. Additionally, no difference between groups was noticed for the expression of CDKN1A mRNA. In summary, in this small study, calcitriol antitumoral effects were not observed in tumorgrafts, established from highly proliferative breast cancer samples. This preclinical model in conjunction with pharmacokinetic calcitriol determinations might contribute to clarify the hormone effects in breast cancer. Citation Format: Maria Lucia Hirata Katayama, Victor Celso N. Fonseca-Filho, Eduardo Carneiro Lyra, Durvanei Augusto Maria, Ricardo Alves Basso, Suely Nonogaki, Juliana Mariotti Guerra, Simone Maistro, João Carlos Sampaio Goes, Maria Aparecida A. Koike Folgueira. Calcitriol effects on breast cancer tumorgrafts. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 3115. doi:10.1158/1538-7445.AM2014-3115

Abstract 1217: Development of a spontaneous in vivo cachexia model using the Champions TumorGraft™ platform

Abstract Cachexia is present in the majority of patients with advanced cancer progression and is not reversible through nutritional supplementation, leading to loss of skeletal muscle and adipose tissue. There are a lack of quality pre-clinical models for cachexia research. Previously, Champions Oncology reported a mouse cachexia model derived from a primary human renal cell carcinoma tumor implanted in immunocompromised mice. This model, CTG-0804, was developed using the innovative Champions TumorGraft™ platform, which preserves the biological properties of the original human tumor. The Champions TumorGraft platform demonstrated dramatic weight loss in these mice as the tumor grew. Plasma from these mice also showed elevated levels of human IL-6. We have now further characterized this TumorGraft model by evaluating standard of care (SOC) agents for renal cell carcinoma and anti-IL-6 compounds to determine the mechanism of cachexia. The SOC evaluation revealed that sorafenib, sunitinib, temsirolimus and bevacizumab as single agents provided a survival advantage by inhibiting tumor growth and preventing weight loss in these mice. Blockade of human IL-6 with tocilizumab or ALD518 did not inhibit tumor growth and weight loss still occurred, demonstrating the possibility of other pathways involved in the cachexia mechanism. In addition, the SOC data for sunitinib showed significant tumor growth inhibition, but no regression at the study end point. This correlated with the clinical profile, where sunitinib could not prevent metastasis and the patient eventually progressed. The TumorGraft model was molecularly characterized and mutation analysis revealed a KIT V530I mutation. This mutation has been reported in extra-abdominal aggressive fibromatosis (desmoid tumor) and acute myeloid leukemia patients that have responded to c-kit targeting therapies, such as imatanib and dasatinib. Further characterization of this TumorGraft is ongoing to include additional bioinformatics using gene expression and mutation pathway analyses. In summary, we have continued development of this cachexia model using the Champions TumorGraft platform, along with molecular evaluation, to produce a robust and well-characterized approach for cachexia investigation which can also be applied to other areas of drug development and cancer research. Citation Format: Nathan Anderson, Tin Khor, Andrew Feldhaus, Andreya Gatling, Katie Olson, John Latham, David Sidransky, Elizabeth M. Bruckheimer. Development of a spontaneous in vivo cachexia model using the Champions TumorGraft™ platform. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 1217. doi:10.1158/1538-7445.AM2014-1217

Abstract 2967: Metformin selectively targets tumor initiating cells in erbb-2 overexpressing breast cancer models

Abstract Metformin is an oral biguanide used for type II diabetes. Epidemiologic studies suggest a link between metformin use and reduced risk of breast and other types of cancers. erbB-2 expressing breast cancer is a subgroup of tumors with poor prognosis. Previous studies demonstrated that metformin is a potent inhibitor of erbB-2 overexpressing breast cancer cells; metformin treatment extends the life span and impedes mammary tumor development in ErbB-2 transgenic mice in vivo. However, the mechanisms of metformin associated anti-tumor activity, especially in prevention models, remain unclear. In this study, we investigated the mechanisms of metformin associated prevention/inhibition of ErbB-2 mediated breast cancer development by focusing on the potential effect of metformin on tumor initiating cells (TICs)/cancer stem cells (CSCs) in the MMTV-erbB-2 transgenic mouse model and its context with existing in vitro models. We report here for the first time that systemic administration of metformin (250 mg/kg/day, ip, between 8- and 18-week of age) selectively inhibits CD61high/CD49fhigh subpopulation, a group of TICs of MMTV-ErbB-2 mammary tumors, in preneoplastic mammary glands. Metformin also inhibited CD61high/CD49fhigh subpopulation in MMTV-erbB-2 tumor-derived cells, which was correlated with their compromised tumor initiation/development in a syngeneic tumor graft model. Molecular analysis indicated that metformin induced downregulation of erbB-2 and EGFR expression and inhibited the phosphorylation of erbB family members, IGF1R, Akt, mTOR and Stat3 in vivo. In vitro data indicate that low doses (1 mM) of metformin inhibited the self-renewal/proliferation of CSCs/TICs in ErbB-2-overexpressing breast cancer cells. We further demonstrated that the expression and activation of ErbB-2 were preferentially increased in CSC/TIC-enriched tumorsphere cells, which promoted their self-renewal/proliferation and rendered them more sensitive to metformin. Our results, especially the in vivo data, provide fundamental support for developing metformin-mediated preventive strategies targeting erbB-2-associated carcinogenesis. Citation Format: Pei Zhu, Meghan Davis, Amanda Blackwelder, Nora Bachman, Bolin Liu, Susan Edgerton, Leonard L. Williams, Ann D. Thor, Xiaohe Yang. Metformin selectively targets tumor initiating cells in erbb-2 overexpressing breast cancer models. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 2967. doi:10.1158/1538-7445.AM2014-2967

Abstract 3808: Preclinical efficacy of the novel, oral Selective Inhibitor of Nuclear Export (SINE) Selinexor (KPT-330) on castration resistant prostate cancer

Abstract Exportin 1/ Chromosomal Maintenance Protein 1 (XPO1/CRM1) is a key nuclear export protein whose inhibition leads to the nuclear accumulation of tumor suppressor proteins such as p53, FOXO, PTEN, pRB and I-κB. Selinexor is an orally bioavailable XPO1 inhibitor that represents a novel class of small molecule compounds with potent activity against a wide variety of cancers. Here we report the activity of Selinexor against Castration resistant prostate cancer (CRPC). CRPC progression is mediated by activation of various adaptive Androgen Receptor (AR) signaling pathways. These pathways are currently being targeted by novel anti-androgen therapies such as abiraterone acetate and enzalutamide with favorable outcomes. However, resistance to anti-androgen therapy can be developed through deregulation of tumor suppressor pathways. XPO-1 is highly expressed in prostate cancer cells and therefore we tested the effects of Selinexor on CRPC cells and tumors models. Treatment of the C4-2B prostate cancer cell line with Selinexor in vitro significantly inhibited cell proliferation and resulted in nuclear accumulation of p53 and p21. In addition, Selinexor inhibited expression of the tumor-promoting gene Ubiquitin-Conjugating Enzyme E2C (UBE2C), which is activated by AR in these cells. Selinexor was also tested in tumor graft models of two patient tumors in castrated male mice. The MDA-PCa-133 tumor is an adenocarcinoma derived from a clinical CRPC bone metastasis. Subcutaneous MDA-PCa-133 tumors in castrated male mice express full-length AR and Prostate-Specific Antigen (PSA) but lack expression of p53. The MDA-PCa-144-13 tumor is a small cell carcinoma derived from a lethal variant of prostate cancer with anaplastic clinical phenotype and does not express AR or PSA but expresses a gain-of-function p53 mutant. Vehicle or Selinexor treatment (10 mg/kg p.o. QoDX3 M/W/F) of MDA-PCa-133 tumors for 34 days resulted in almost complete inhibition of tumor growth with a 30-fold reduction of PSA (p<0.001). Treatment of MDA-PCa-144-13 tumors for 22 days also resulted in significant reduction of tumor volume (>8 fold compared to vehicle only; p<0.0047). Immunohistochemical analysis of MDA-PCa-133 tumors showed that Selinexor induced nuclear accumulation of XPO1 cargos FOXO3a and p27 with concomitant reduction in the proliferation marker KI67. In conclusion, Selinexor demonstrated robust inhibition of CRPC tumor growth and activated multiple tumor suppressors in prostate cancer cells. Selinexor is now being evaluated in Phase 1 clinical studies in patients with advanced hematological and solid tumor cancers, and preliminary results show adequate tolerability with evidence of anti cancer activity. Future clinical studies in patients with CRPC are planned and will provide more information on the use of Selinexor as monotherapy and in combination with anti-androgen therapy of CRPC. Citation Format: Sankar Narayan Maity, Guanglin Wu, Jing-Fang Lu, Anh Hoang, Yosef Landesman, Dilara McCauley, Sharon Shacham, Michael G. Kauffman, Ana M. Aparicio, Eleni Efstathiou, John C. Araujo, Christopher J. Logothetis. Preclinical efficacy of the novel, oral Selective Inhibitor of Nuclear Export (SINE) Selinexor (KPT-330) on castration resistant prostate cancer. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 3808. doi:10.1158/1538-7445.AM2014-3808

Abstract 4410: KRAS mutation status is associated with enhanced dependency on purine biosynthesis and related pathways in non small cell lung cancer cells

Abstract KRAS gene mutation is linked to poor prognosis and resistance to oncology therapeutics in Non Small Cell Lung Cancer (NSCLC). We have explored the possibility of exploiting inherent differences in KRAS mutant cell metabolism to enhance the efficacy of treatment. We have identified a greater dependency on purine biosynthesis and related pathways in KRAS mutant compared to KRAS wild type NSCLC cell lines. Purine synthesis requires factors generated from other metabolic reactions including ribose-5-phosphate from the pentose phosphate pathway, THF cofactors from folate metabolism and glycine / amide nitrogen groups from glutamine and aspartate metabolism. In this study, microarray gene expression and biological pathway analysis identified higher expression of purine synthesis and accessory pathways such as folate metabolism, 5-aminoimidazole ribonucleotide biosynthesis and glycine synthesis pathways in KRAS mutant NSCLC cells compared to wildtype counterparts. KRAS knockdown and overexpression studies demonstrated the ability of KRAS to regulate expression of genes that comprise purine synthesis and folate metabolism pathways. Moreover, pathway analysis and knockdown studies suggest a role for MYC, an oncogene previously associated with KRAS mutant tumors, in the regulation of these pathways in KRAS mutant NSCLC cells. Proliferation studies demonstrated higher responsiveness to antifolates such as methotrexate and pemetrexed in KRAS mutant NSCLC cells, both of which may interfere indirectly with purine biosynthesis. In vivo analysis of NSCLC tumorgraft models in nude mice also identified an association between KRAS mutant tumor status and response to pemetrexed. The expression of a KRAS driven folate/purine synthesis gene, Methylenetetrahydrofolate Dehydrogenase 2 (MTHFD2), was also correlated with antifolate activity suggesting its use as a possible biomarker of response to antifolates. We propose that KRAS mutation drives increased purine synthesis activity and as a result an elevated dependency on the factors needed to feed this biosynthetic pathway such as those generated by folate metabolism. Thus, antifolates can indirectly inhibit purine synthesis through the depletion of folate cofactors which may account for the stronger response to these agents in KRAS mutant cells. We are currently expanding this study to examine alternative inhibitors of purine synthesis as possible therapeutics in KRAS mutant NSCLC and other cancers. Collectively, our findings highlight that a better understanding of the molecular mechanisms underlying the dependency of cancer cells on specific metabolic pathways may result in more effective metabolic targeting and new approaches in treating specific cancers. Citation Format: Diarmuid M. Moran, Patricia B. Trusk, Karen Pry, David Sidransky, Keren Paz, Sarah S. Bacus. KRAS mutation status is associated with enhanced dependency on purine biosynthesis and related pathways in non small cell lung cancer cells. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 4410. doi:10.1158/1538-7445.AM2014-4410

Abstract 1187: The Champions TumorGraft Bank: A demographically-rich repository of preclinical TumorGraft models

Abstract Many oncology pharmaceuticals fail during phased clinical trials, having been advanced based on research using flawed preclinical models that do not accurately replicate human tumor biology. For example, numerous models are based on genetically-engineered mice, which are often generated from limited numbers of genetic aberrations. As such, they may not accurately represent either the chaotic heterogeneity that exists within intact human tumors or throughout the clinical population. TumorGrafts, in which patient-derived tumor tissue is directly engrafted into immunodeficient mice, offer a step forward in this unresolved issue. These models maintain the complex intra-tumoral heterogeneity and biology of an intact malignancy, as well its 3-dimensional interplay with stromal components and other cells fluxing into the immediate environment. The intrinsic cross-talk between the different compartments of the tumor is also retained. We wanted to generate a repository of these tumor models and make them available to the research community for use in pharmaceutical development and basic research processes. Here we describe our extensive bank of live TumorGrafts comprising a range of different tumor types, from more common cancers, including lung, colon, and breast, to a number of rare subtypes such as adenoid cystic carcinoma. These models were originally developed for personalizing cancer treatments and are derived from patient populations across all ages and ethnicities, encompassing both treatment-naïve and heavily-treated individuals. Furthermore, because of our distinct patient-focus, we maintain detailed clinical histories, as well as molecular data where available. Hence, the bank is a reasonable surrogate of the population from which treatment groups are typically drawn for clinical trials and could potentially be of value for predetermining target populations for therapeutic intervention. The Champions TumorGraft Bank provides superior pre-clinical models that faithfully capture all the biological and molecular features of cancer and can serve basic and clinical research groups as an increasingly valuable source for drug development and biomarker discovery. Citation Format: Tin Khor, David Vasquez, Amanda Katz, Gilson Baia, Daniel Ciznadija, David Sidransky, Keren Paz. The Champions TumorGraft Bank: A demographically-rich repository of preclinical TumorGraft models. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 1187. doi:10.1158/1538-7445.AM2014-1187

Validated Tumor Graft Model Reveals Delayed Local Recurrence of Renal Cell Tumors After High-Dose Single Fraction Ablative Radiation Therapy

Validated Tumor Graft Model Reveals Delayed Local Recurrence of Renal Cell Tumors After High-Dose Single Fraction Ablative Radiation Therapy

KRAS Mutation Status Is Associated with Enhanced Dependency on Folate Metabolism Pathways in Non–Small Cell Lung Cancer Cells

KRAS gene mutation is linked to poor prognosis and resistance to therapeutics in non-small cell lung cancer (NSCLC). In this study, we have explored the possibility of exploiting inherent differences in KRAS-mutant cell metabolism for treatment. This study identified a greater dependency on folate metabolism pathways in KRAS mutant compared with KRAS wild-type NSCLC cell lines. Microarray gene expression and biologic pathway analysis identified higher expression of folate metabolism- and purine synthesis-related pathways in KRAS-mutant NSCLC cells compared with wild-type counterparts. Moreover, pathway analysis and knockdown studies suggest a role for MYC transcriptional activity in the expression of these pathways in KRAS-mutant NSCLC cells. Furthermore, KRAS knockdown and overexpression studies demonstrated the ability of KRAS to regulate expression of genes that comprise folate metabolism pathways. Proliferation studies demonstrated higher responsiveness to methotrexate, pemetrexed, and other antifolates in KRAS-mutant NSCLC cells. Surprisingly, KRAS gene expression is downregulated in KRAS wild-type and KRAS-mutant cells by antifolates, which may also contribute to higher efficacy of antifolates in KRAS-mutant NSCLC cells. In vivo analysis of multiple tumorgraft models in nude mice identified a KRAS-mutant tumor among the pemetrexed-responsive tumors and also demonstrated an association between expression of the folate pathway gene, methylenetetrahydrofolate dehydrogenase 2 (MTHFD2), and antifolate activity. Collectively, we identify altered regulation of folate metabolism in KRAS-mutant NSCLC cells that may account for higher antifolate activity in this subtype of NSCLC.

Patient‐derived xenografts for individualized care in advanced sarcoma

BACKGROUNDPatients with advanced, metastatic sarcoma have a poor prognosis, and the overall benefit from the few standard-of-care therapeutics available is small. The rarity of this tumor, combined with the wide range of subtypes, leads to difficulties in conducting clinical trials. The authors previously reported the outcome of patients with a variety of common solid tumors who received treatment with drug regimens that were first tested in patient-derived xenografts using a proprietary method (“TumorGrafts”).METHODSTumors resected from 29 patients with sarcoma were implanted into immunodeficient mice to identify drug targets and drugs for clinical use. The results of drug sensitivity testing in the TumorGrafts were used to personalize cancer treatment.RESULTSOf 29 implanted tumors, 22 (76%) successfully engrafted, permitting the identification of treatment regimens for these patients. Although 6 patients died before the completion of TumorGraft testing, a correlation between TumorGraft results and clinical outcome was observed in 13 of 16 (81%) of the remaining individuals. No patients progressed during the TumorGraft-predicted therapy.CONCLUSIONSThe current data support the use of the personalized TumorGraft model as an investigational platform for therapeutic decision-making that can guide treatment for rare tumors such as sarcomas. A randomized phase 3 trial versus physician's choice is warranted.

Histological and biochemical analysis of DNA damage after BNCT in rat model

To understand the mechanism of tumor cell death induced by boron neutron capture therapy (BNCT) and to optimize BNCT condition, we used rat tumor graft models and histological and biochemical analyses were carried out focusing on DNA damage response. Rat lymphosarcoma cells were grafted subcutaneously into male Wister rats. The rats with developed tumors were then treated with neutron beam irradiation 45min after injection of 330mg/kg bodyweight boronophenylalanine (10BPA) (+BPA) or saline control (–BPA). BNCT was carried out in the National Nuclear Center of the Republic of Kazakhstan (neutron flux: 1×109nvt/s, fluence: 6×1011nvt) with the presence of background γ-irradiation of 33Gy. 6 and 20h after BNCT treatment, tumors were resected, fixed and subjected to immunohistochemistry and biochemical analyses. Immunostaining of nuclei showed that double strand break (DSB) marker gamma H2AX staining was high in 20h/+BPA sample but not in 20h/–BPA samples. Poly(ADP-ribose), DSB and single strand break markers of DNA, also demonstrated this tendency. These two markers were observed at low levels in unirradiated tissues or 6h after BNCT either under −BPA and +BPA conditions. HMGB1 level increased in 6h/+BPA but not in 6h/−BPA or 20h/+BPA samples. The persistent staining of γH2AX and poly(ADP-ribose) in +BPA group suggests accumulated DSB damage after BNCT. The early HMGB1 upregulation and γH2AX and poly(ADP-ribose) observed later might be the markers for monitoring the DNA damage induced by BNCT.