Abstract P1-03-03: A Human in Mouse Model of Triple Negative Breast Carcinoma To Optimize Alkylating or Targeted Treatment

Abstract

Abstract Background. Breast cancer negative for ER, PR, and HER2, so called triple negative breast cancer (TNBC) occurs at a younger age and has a worse prognosis. It is considered hard to treat because it tends to rapidly develop resistance to conventional therapy and targeted therapy is not available. Therefore more efficient therapy is warranted to improve patient survival. In the development of new therapies, the availability of preclinical in vivo models is essential. Ideally, these in vivo models would recapitulate the heterogeneity in histogenetic characteristics, expression profiles, signaling pathways, and response to therapy found in human breast cancers. Xenograft models, in which human tumor tissue is engrafted directly into recipient mice, may be very useful for this purpose (Marangoni et al, Clin Can Res 2007; 13:3989). Methods. We have set up a panel of xenografts in which human TNBC can be studied in a dynamic and reproducible manner by implanting fresh breast cancer fragments in immunocompromised mice. Tumor outgrowth occurs 3 to 10 months after implantation. After outgrowth, tumors are characterized based on histology, immunohistochemistry, aCGH and gene expression profile and serially transplanted. Subsequently, responses to chemotherapeutic agents are being tested. Results. We generated five TNBC xenografts, which were all classified as basal-like based on expression profiling, and BRCA1-like based on aCGH analysis (Joosse et al., Breast Cancer Res Treat (2009) 116: 479). One model was shown to be BRCA2 mutated, whereas no BRCA1 or -2 mutations were found in the other models. We treated the TNBC xenograft models with cisplatin, since recent clinical studies have shown good response of TNBC to cisplatin (Silver et al, J Clin Oncol 2010; 28:1145, Byrski et al, J Clin Oncol 2010; 28:375). When treated with cisplatin, the BRCA2 mutated model showed only limited response, with reduced tumor growth, but no tumor eradication. In contrast, the other four TNBC xenograft models responded very well to treatment with cisplatin, resulting in complete disappearance of a palpable lesion in most cases after 21 days of treatment. However, tumors did relapse and had to be treated again. Interestingly, all tumors remained responsive to cisplatin treatments even after several relapses, but could not be completely eradicated. We have recently started treating TNBC xenograft models with the bifunctional alkylators melphalan and nimustine, because these drugs were shown to cause tumor eradication in genetically engineered mouse models of BRCA2-associated breast cancer (Evers et al, Clin Can Res 2010; 16:99). The BRCA2 mutated tumor again showed only partial response, whereas the other TNBC models showed very good initial responses, although relapses did occur. Conclusion. Triple negative breast cancer xenograft models resemble their human counterpart very well and are useful to study response to cytotoxic drugs and targeted agents in a preclinical setting, which could speed up introduction of promising treatments into clinical practice. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P1-03-03.