Abstract #3227: Development of new clinically-relevant human melanoma xenograft models

Abstract

Melanomas whose incidence is still increasing, remain associated with dismal prognosis given their inherent resistance to apoptosis and their strong propensity to metastasize, notably to the lungs, liver and brain. Metastatic melanoma patients have only limited treatment options consisting of mono- or poly-therapy with dacarbazine (DTIC), cisplatin and temozolomide (TMZ). These produce only marginal clinical improvements and have no substantial impact on overall patient survival. Even biological/immuno-therapy has not contributed greater clinical benefits than DTIC alone. In order to test and develop new therapeutic options, greater attention needs to be paid to the clinical relevance of the animal model used. Numerous agents/regimens have shown really promising activity in existing pre-clinical melanoma models but efficacy could then not be demonstrated in clinical trials. Therefore our aim was to establish more reliable in vivo models through i) isolation of melanoma cells from primary and metastatic lesions, ii) in vitro characterization of their genomic (array CGH, p53 mutations) and biological properties (growth, migration, sensitivity to cytotoxic agents, expression of drug resistance and anti-apoptotic proteins) and iii) their grafting either s.c., i.v. or orthotopically into their site of origin in nude mice and when relevant follow up with histological analysis, determination of their drug sensitivity and comparison of their responses to those of established melanoma models. 35 primary melanoma cell lines have been established and to date 10 have been further characterized. In vitro these primary cultures have been found to be equally poorly sensitive to chemotherapy as established melanoma cell lines. However with primary cultures, tumor \#8220;take\#8221; and metastases were increased compared to those of established cell lines (notably spontaneous lung metastases achieved in 100% (VM-21 grafted) and 70% (VM-1) of mice). Furthermore, after the excision of primary tumors, micro-metastases in lungs were found to lead to animal death. No response to DTIC, TMZ and/or cisplatin either in terms of tumor growth or animal survival could be achieved in these new models, while the widely used B16F10 mouse model is generally drug-responsive. Furthermore, i.v. injection of these primary cells did not lead to metastases indicating that this is not an appropriate method to study melanoma metastatic processes. In contrast, we have developed two brain metastatic models through direct implantation of human brain metastases-derived melanoma cells into the brains of mice. These models are characterized by massive invasion of normal brain tissue and resistance to chemotherapy. In conclusion, in vivo models developed with primary melanoma cultures may be of greater relevance than those using established cell lines because the course of the disease and the chemotherapeutic response more closely resemble the clinical situation in humans.

Citation Information: In: Proc Am Assoc Cancer Res; 2009 Apr 18-22; Denver, CO. Philadelphia (PA): AACR; 2009. Abstract nr 3227.