Abstract A41: Developing a mouse model of melanoma to identify drug resistance mechanisms

Abstract

Abstract Targeted therapies have the potential to revolutionize cancer care by providing personalized treatment strategies that are less toxic and more effective. Cutaneous melanoma is well known for its aggressive clinical behavior, propensity for lethal metastasis and therapeutic resistance to conventional therapy. Over the past decade, key mutations in the mitogen-activated protein kinase (MAPK) pathway have been identified in melanoma. The most common genetic alteration discovered in melanoma is a point mutation (T1799A) in the BRAF proto-oncogene that encodes for BRAF V600E. Mutated BRAF elicits constitutive activation of the MEK/ERK signaling pathway promoting tumor proliferation and survival and is detected in approximately 60% of patients with metastatic melanoma. Vemurafenib (PLX4032) is a selective BRAF inhibitor and the first of its kind to be tested in advanced-stage melanoma. Preliminary Phase III response rates of 48% versus 5% with standard therapy prompted expedited FDA approval of vemurafenib in August 2011. Although PLX4032 induces tumor responses in the majority of patients with mutant BRAF, the duration of response ranges from 2 to greater than 18 months until acquired resistance to therapy develops. Several acquired resistant mechanisms have been identified in human cell lines and a small fraction of patient samples. Most, if not all of these patient samples and cell lines were treated with DNA damaging alkylating agents prior to receiving vemurafenib. In addition, the mutation status of other genes frequently altered in melanoma that may impact resistance dynamics have not been assessed in most of the patient samples that develop resistance. To enhance identification and validation of vemurafenib resistance mechanisms we have expanded our somatic gene transfer mouse model of melanoma based on the RCAS/TVA retroviral delivery system to include tumors induced by mutant BRAF in the context of Ink4a/Arf and/or Pten loss. We crossed our DCT-TVA;Ink4a/Arflox/lox mice with BRAFCA mice that carry a conditional BRAFV600E allele that is Cre activated (CA), and conditional Ptenlox/lox mice. Retroviral mediated delivery of Cre to DCT-TVA;Ink4a/Arflox/lox;BRAFCA; Ptenlox/lox mice and DCT-TVA;Ink4a/Arflox/lox;BRAFCA mice resulted in tumor induction rates of 100% and 43% respectively. Median survival of 62 ± 6.7 days was observed in TVA;Ink4a/Arflox/lox;BRAFCA; Ptenlox/lox mice, all of which harbored metastatic disease (n=8). A median survival of 72 ± 6.7 days with no observed metastatic disease was observed in DCT-TVA;Ink4a/Arflox/lox;BRAFCA mice (n=6). To compare potential differences in PLX4032 responses and resistance kinetics between the cohorts (i.e., BRAFCA with Ink4a/Arf loss vs. BRAFCA with Ink4a/Arf and Pten loss), we initiated tumors in newborn mice. When median tumor size approached 200 mm3 we began oral 25mg PLX4032 treatment b.i.d. and observed antitumor responses in both groups. Mice exhibiting Pten loss were less responsive to PLX4032 and developed drug resistance more rapidly. We are currently evaluating potential mechanisms of PLX4032 resistance in both cohorts. Our melanoma mouse model provides a rapid and robust method to investigate factors associated with drug resistance in order to develop better therapeutic strategies.