Background Osteosarcoma is the most common pediatric bone cancer in the United States. Advances in surgery and chemotherapy have resulted in a cure rate of 65% for children who present with localized disease. However, patients who have metastatic disease have only a 25% survival rate. Attempts to improve this outcome using various intensification regimens have resulted in only slightly better 2-year overall survival rates. Unfortunately, these results demonstrate the inability to target the tumor using intensification of chemotherapy to salvage patients with metastatic disease. Therefore, we have turned our efforts to identifying biological elements that can be modulated in this disorder. HOS is a human osteosarcoma cell line that does not form primary tumors and does not form metastatic lung nodules in immuno-deficient mice. However, treatment of HOS with the chemical N-methyl-N-nitro-N-nitrosoguanidine (MNNG) resulted in a transformed cell line, MNNG-HOS, that is able to form primary tumors and also metastatic nodules in the lung. Further characterization of the chemically transformed cell line resulted in the identification of the MET oncogene as the likely mediator of metastasis. MET is a cell surface receptor that binds a single ligand, hepatocyte growth factor, and has tyrosine kinase activity. MET is mutated or overexpressed in a variety of tumors, including many carcinomas, hematologic malignancies and soft tissue sarcomas. In addition, a study in large dogs with osteosarcoma has shown that 10% of these tumors harbor mutations in MET. These findings led us to study MET and its possible role in osteosarcoma metastasis. Research Performed in 2009 We performed immunohistochemistry utilizing a newly developed osteosarcoma tissue array that is linked to patient outcome data. Samples from eighteen patients with high-grade conventional osteosarcoma were available for analysis. Immunohistochemistry was performed using a MET specific antibody. We found high level of staining in 83% of samples tested. Staining was stratified into 0+/1+/2+ staining versus 3+ staining and correlated with outcome. A Kaplan-Meier survival analysis resulted in a p-value of 0.65 (hazard ratio 0.21 / 95% confidence interval of the hazard ratio = 0.02 to 2.26). However, the lack of statistical significance is very likely a function of the high level of MET expression observed in the majority of patient samples. We attempted to perform similar experiments using a phospho-MET antibody, but could not due to lack of an effective antibody. To overcome this limitation, we utilized a newly developed technology that utilizes electrochemiluminescence (Meso Scale Discovery) to obtain quantitative values of expression. The technology is based on a sandwich antibody technique using a MET capture antibody that is bound to a plate and a different MET detection antibody that is conjugated to a luminescent signaling molecule. The use of electricity to excite this complex results in release and activation of the luminescent molecule, allowing for quantitation of the signal. We used this approach using seven established osteosarcoma cell lines and four newly derived cell lines from patients with metastatic disease. In our analysis, we detected high level of expression (>50 ng of MET per mg of total protein) in two of the 11 cell lines tested and intermediate level of expression (5-50 ng of MET) in an additional two cell lines. In addition, we extracted protein from 23 tumors from osteosarcoma patients and found high level of MET expression (>50 ng of MET) in five tumor samples. In addition, we were able to use a phospho-(tyrosine) antibody as the detection antibody, to assess phosphorylation status. Phospho-(tyrosine)MET was detected in all but one of the cell lines and tumors tested above, suggesting that high level of expression also results in activation. We went on to test the osteosarcoma cell lines to determine if they were responsive to a highly-specific MET inhibitor, PHA-665752. Electrochemiluminescent assays showed that two cell lines tested had significantly lower levels of phospho-(tyrosine)MET levels in cells treated with PHA-665752. These levels returned close to their high baseline levels when examined 96 hours after treatment, demonstrating that the inhibitor was targeting MET. We then tested the ability of PHA-665752 to inhibit the following parameters: proliferation (MTT analysis), motility (scratch assay), migration (through an 8 micron pore) and invasion (migration through a matrigel matrix). PHA-665752 had no effect on proliferation in any of the osteosarcoma cell lines tested. However, it was effective in inhibiting one or more of the other parameters in three cell lines. These included AI, LR and MNNG-HOS. AI and LR are the two highest MET expressing cell lines and MNNG-HOS has activated MET due to a translocation. These results demonstrated that only cells that express high levels of MET have a decrease in the metastatic phenotype in vitro when MET is inhibited. To date, we have found high levels of MET expression and also of phospho-(tyrosine)MET in several cell lines derived from patients with metastatic osteosarcoma. We utilized a MET inhibitor and demonstrated inhibition of parameters of the metastatic phenotype in vitro in both of these high MET expressing cell lines. These preliminary results have allowed us to formulate the following hypothesis which we will begin to test in 2010. Hypothesis to be tested in 2010 High level expression of MET results in a metastatic phenotype. In addition, pharmaceutical inhibition of MET will attenuate metastatic disease.
Specific Aims 1. Determine the ability of a MET-specific inhibitor, currently in clinical trials, to abrogate metastatic disease in vivo using osteosarcoma cell lines that have high levels of MET expression. 2. Determine the expression pattern of MET, and whether they harbor activating mutations in MET, using 46 tumor samples from patients with osteosarcoma. Significance and Goals The successful completion of the above objectives may demonstrate that inhibition of MET, in relevant osteosarcoma cell lines that have high levels of MET expression or activation can decrease metastatic disease. These experiments will also help in determining which subset of patients has high levels of MET expression. We will also be able to determine whether mutations lead to high levels of MET expression or constitutive MET activation. Our hope is to be able to identify the subset of patients with osteosarcoma who may benefit from addition of therapy with MET inhibitors as the basis for a clinical trial. Our use of an inhibitor that is currently in early clinical trials will allow us to translate potential successes on the bench and in murine models, rapidly into the clinic. Presentations and Papers in 2009 Poster presentation at American Association for Cancer Research 2009 Annual Meeting. Oral presentation at Pediatric Academic Societies 2009 Annual Meeting. Poster presentation at Connective Tissue Oncologic Society 2009 Annual Meeting. Paper in preparation for Cancer Research.
