BRAF-driven pediatric malignant astrocytoma: Etiology & treatment

Brain cancers are the most common type of solid tumor in children. Current therapies can result in better survival but generally leave patients with severe life-long neuro-developmental handicap. Targeted molecular therapeutics holds great promise for the development of less toxic and more effective personalized treatment strategies for cancer (1). This approach, however, requires a detailed understanding of the molecular alterations that drive tumor formation and malignant progression.

The identification of driving mutations in cancer can be used to inform the development of therapeutic agents that specifically inhibit the protein products and pathways that are deregulated due to cancer-associated gene alterations. This strategy has been used successfully in adult brain cancer, leading to clinical trials that target the epidermal growth factor receptor (EGFR), for example. Pediatric malignant astrocytomas (PMAs) in children, however, are distinct clinical entities from those seen in adults and do not harbor many of the critical genetic alterations found in their adult counterparts (2-5).

PMAs are the leading cause of cancer-related death from solid tumors in children. Because of their invasive nature, PMAs are surgically incurable, and standard chemotherapies have been only modestly, and transiently effective in their treatment. Most patients succumb to their disease within 12 months of diagnosis. In recent years, a potential link has emerged between neural progenitor (stem) cell populations and the origin of astrocytomas. This concept is central to a major objective of this proposal: to identify the type of cell that gives rise to PMAs. It is possible that further insights into the biology of neural stem cells could provide insights into the genesis as well as treatment of these tumors.

Until recently, analyses of pediatric malignant astrocytoma DNA had revealed little regarding the genetic etiology of these tumors, and consequently little was known regarding potential therapeutic targets for their treatment. However, a research effort led by UCSF investigators, including Drs. Rowitch and James, published earlier this year, provided important new insights regarding key PMA gene alterations. We have shown that an activating mutation of the BRAF oncogene is prevalent in PMAs (Schiffman et al., Cancer Research 70:512-9: reference 6) . The mutation, called BRAF:V600E, results in deregulated and heightened enzymatic activity of the BRAF protein (7). In a series of 31 PMAs we found a 23% incidence of this mutation. This observation further highlights the genetic differences between PMAs and their adult counterparts, in which BRAF mutations seldom occur.

Interestingly, in five of the seven PMAs identified as having the BRAF:V600E mutation, there were additional alterations resulting in CDKN2A gene inactivation, and associated loss of expression of a key protein that acts as a regulator of cell growth, p16. Our results suggest the possibility that the combination of BRAF activation and CDKN2A inactivation confers malignant character to astrocytes in children, and cause PMA development.

BRAF is a member of a cancer-associated signaling pathway, and there are several BRAF signaling pathway inhibitors that are currently in clinical trials for treating other types of human cancer. These include inhibitors of MAPK, a key downstream effector of BRAF, as well as BRAF itself. Consequently, the observation of BRAF mutations in PMAs provides a rationale for testing BRAF pathway inhibitors in appropriate pre-clinical models for PMA. Tumor cells lacking CDKN2A, and that do not express p16, also have an approachable therapeutic target: the cdk4 protein, whose activity is heightened in tumors lacking expression of p16. Therefore, our recently published observations highlight at least two therapeutic strategies for treating PMAs.

Summary and Significance: Until recently, the genetic underpinnings of pediatric astrocytoma have been obscure. New findings indicate that BRAF activating mutations in association with inactivating mutations of the tumor suppressor locus CDKN2A occur in as much as 20% of pediatric astrocytomas grades II-IV. This project will generate genetically faithful mouse models of this type of brain cancer. Resultant tumors will be analyzed with the help of the UCSF brain tumor research center neuropathology core. The Core is also performing detailed analysis of human pediatric astrocytomas in our cohort, with similar markers to those we propose to use in the mouse tumors. Our analysis of patient PMAs will provide information allowing validation of our mouse models. Such models, as well as BRAF:V600E human malignant astrocytoma xenografts, can immediately be used for preclinical testing of RAS/BRAF/MEK inhibitors that have been developed for blockade of this pathway in cancer treatment. Such small molecule antagonists have been used to treat other cancers, such as melanoma. Thus, there are compelling reasons to rapidly generate mouse models with these aims in mind.

The current application will utilize existing information and resources for developing mouse models to better understand the cellular origin of BRAF/CDKN2A-associated PMAs, and adapt BRAF:V600E/CDKN2A-/- tumor cells for utilization in therapeutic test models. On completion, this project will contribute new information regarding the molecular and cellular etiology of pediatric astrocytomas, and will additionally provide new cancer models, as well as related therapeutic observations, for translation to clinical practice.