Annual Forum 2009 – The Biology and Treatment of Primary Brain Tumors

Tracy Batchelor, MD, MPH, Harvard Medical School, Boston, MA
Luis Parada, Ph.D., University of Texas Southwestern Medical Center, Dallas, TX

Gabriele Bergers, University of California, San Francisco, CA
Lynda Chin, MD, Dana-Farber Cancer Institute, Boston, MA
Peter Dirks, MD, University of Toronto, Toronto, Canada
Daniel Duda, Ph.D., Massachusetts General Hospital, Boston, MA
John M. L. Ebos, Ph.D., University of Toronto, Toronto, Canada
Eric Holland, MD, Ph.D., Memorial Sloan-Kettering Cancer Center, New York, NY
Mikkel Noerholm, Massachusetts General Hospital, Charlestown, MA
Greg Sorensen, Massachusetts General Hospital, Charlestown, MA
Charles D. Stiles, Ph.D., Dana-Farber Cancer Institute, Boston, MA
Roger Stupp, MD, University of Lausanne, Lausanne, Switzerland
Robert Wechsler-Reya, Ph.D., Duke University Medical Center, Durham, NC

Layout and Goals

  1. From Models of Brain Tumors to Clinical Trials
  2. Tumor Origin, Medulloblastoma and Mechanisms
  3. The Problem of Angiogenesis
  4. Platelet Derived Growth Factor (PDGF), Alternative Tumor Sources and Implications

Primary tumors of the central nervous system are the leading cause of cancer death in children, and a tumor of growing incidence in adults, in whom it is equally difficult to treat. Only recently have researchers begun to understand the basic genetic derangements that play a central role in the growth of these tumors. In children, brain tumors are found in a variety of unusual types, some with unique and characteristic appearance on pathologic examination and in their clinical behavior. In adults, primary brain tumors usually arise in astrocytes, supporting cells of the nervous system, and, in their most common, and poorly differentiated form, these tumors, called gliomas, are exceedingly difficult to treat with any of the common approaches (surgery, chemotherapy, or radiation therapy). Major efforts are now underway to elucidate the genetic changes in these tumors, the pathways and receptors activated by these changes, and the changes found in the tumor micro-environment. Animal models for some of these kinds of brain tumors have now been developed, and are providing information on tumor behavior and response to experimental treatment. New clinical trials are finding that these tumors are highly dependent on new blood vessels, and respond to treatments that destroy these vessels. Abnormal receptors and activated growth pathways are found on the cell surface or inside these tumors and these may be the subject of new treatments as well. New information indicates that multiple aberrancies may exist in the signaling PORTFOLIO of these tumors, thus requiring multiple sites of attack. In this rapidly evolving state of knowledge, the meeting, to be led by Dr. Tracy Batchelor of the Massachusetts General Hospital's Cancer Center, and Dr. Luis Parada of University of Texas Southwestern Medical Center, should provide an exciting opportunity for bringing together leaders in the genetics, biology, and treatment of this important group of tumors.

Outcome Report
The focus of the 2009 Forbeck Foundation meeting in Hilton Head was primary brain tumors. These are a leading cause of cancer death in children. Although their occurrence in adults is relatively rare, the tumors can be very aggressive with poor survival rates. In adults, primary brain tumors usually arise in cells called astrocytes, which are thought of as the supporting cells of the nervous system. These cells, in their poorly differentiated form, can produce a tumor type called gliomas, which are exceedingly difficult to treat with any of the common approaches (surgery, chemotherapy, or radiation therapy). Major efforts are now underway to elucidate the genetic changes that can occur in specific types of brain tumors. This can lead to the identification of either abnormal receptors and/or activated growth pathways that may aid the development of new, much needed, therapies. Animal models for some of these tumors have now been developed, and can also provide information on their behavior and response to experimental treatments.

New clinical trials are finding that gliomas are highly dependent on new blood vessel development. Our understanding of how the tumor cell communicates with host blood vessels to increase their rate of division and thus their number offers a new approach to destroying the cancer cells if therapies can be developed to specifically destroy these vessels. In this field there is a desperate need to define new approaches to treatment. Dr. Tracy Batchelor of Massachusetts General Hospital Cancer Center, and Dr. Luis Parada of the University of Texas Southwestern Medical Center, brought together a group of clinicians and basic scientists to review new information on the development and treatment of brain tumors. Dr. Peter Dirks reviewed his work on the identification of cell populations with stem cell like properties in mouse models of medulloblastoma (a pediatric tumor) and glioma. He reviewed the importance of culture conditions in the enumeration of these cells and the evidence that these cells maintain their tumorigenic properties in vivo. The identification and characterization of these cell populations could lead to the development of drug screening platforms to identify agents that target cancer stem cells that are present in these tumor types. Dr. Luis Parada reviewed data provided by Lynda Chin from The Cancer Genome Atlas (TCGA) study. Glioblastoma was the first tumor selected for genomic screening in this large-scale national project. Investigators probed glioblastoma for the expression of 600 well-characterized, cancer-associated genes. There were six major genetic techniques may tell us more about the biology of malignant brain tumors and may offer a non-invasive means of monitoring physiological changes in these tumors during therapy. Dr. Greg Sorensen reviewed the use of these MRI techniques in a study of 31 glioblastoma patients in which a number of biological effects of anti-VEGF therapy were detectable. MRI techniques are being developed to predict patients who will or will not respond to anti-angiogenic therapy. Dr. John Ebos reviewed the potential adverse effects of anti-angiogenic agents including provocative data generated in mouse models of cancer that anti-angiogenic therapies, in the form of receptor tyrosine kinase inhibitors, may increase the metastatic potential of certain tumor types. These observations raise potentially important questions regarding the timing of anti-angiogenic therapy and the possible benefit of drug combinations to reduce the possibility of this effect.

The importance of the brain tumor microenvironment was highlighted and reviewed by Dr. Gabrielle Bergers. The beneficial effect of bevacizumab on brain tumors is transient and the tumor eventually becomes more invasive through a number of potential pathways. She reviewed preclinical glioma models that implicate a number of cytokines, c-MET, SDF1, in this switch to a more infiltrative tumor growth pattern. Possible molecular targets to prevent or reverse this invasive tumor growth were reviewed.

A session was chaired by Dr. Chuck Styles and emphasized the use of mouse models to study glioma. Dr. Kwon presented his very preliminary studies in which he is using material derived from NF1; p53;Pten based tumor models to identify genes that are transcriptionally modified in a significant way compared to the expression in neural progenitors which are the source for these tumors. He has used these data to compare with the TGCA data and find commonalities. As a first example of his proposed method, Dr. Kwon has identified a gene that is highly expressed both in human and mouse tumors and that is implicated in controlling reactive oxygen species, a potentially damaging agent in tumor development. He discussed the methods of validation of this gene and moreover the approaches he plans to take to identify additional genes in a high throughput manner. This was a good platform for discussion of the many technical obstacles the field is faced with. Dr. Holland discussed his PDGF inducible glioma mouse models in the context of Ink4a/Arf null background. Injection of a retrovirus with either a nestin or GFAP promoter can induce tumor development and he proposes this as a powerful model to study tumor development of PDGFR sub-classes. Dr. Holland presented some mathematical models that would support the stochastic capacity of tumor formation as a function of the rate of mutation and proliferation in the pool of potentially tumor forming cells. In this scenario, in principle, any cell could undergo the appropriate number of mutations to give rise to glioma. In another set of experiments in which his group uses side population FACS sorting to enrich for tumor cells, Dr. Holland demonstrated that agents such as temozolimide enrich this population resulting in more aggressive tumorigenic cells. Dr. Becher studies diffuse infiltrating pontine gliomas, a very rare but terrible form of glioma in children. These tumors arise in the Pons and are very aggressive. In a novel set of experiments Dr. Becher provided compelling evidence that he can successfully model this rare tumor in mice. This is a huge advance since due to their rarity, material to study these tumors is very difficult to obtain. Dr. Becher plans to use this model to identify therapeutic targets that may enhance the treatment of this disease. In the final session, Dr. Parada expanded on the discussion of physiologically relevant mouse models and their potential to provide more informative and a quantitative evaluation of therapeutic modalities. The cancer stem cell hypothesis was further discussed in a continuation of discussion broached by Drs, Holland and Wechsler-Reya, among others. The value of continued development of models not only for validation and biological studies but also as improved pre-clinical models of glioma was discussed.

In summary, the meeting embodied many of the current and timely topics that surround continued efforts by both clinicians and basic scientists to tackle this devastating disease that is not curable. It can be said that there was a dynamic, candid, and open-minded discussion that served to both educate both sides of this front and to identify possibilities for collaboration.