Accelerating Brain Tumor Drug Discovery Through Cellular Imaging

Andrew Kung, M.D., Ph.D.

Dana-Farber Cancer Institute, Cambridge, MA

Grant Program:

David Mahoney Neuroimaging Program

Funded in:

June 2004, for 2 years

Funding Amount:


Lay Summary

Modeling Brain Glioma to Test Potential Therapies Prior to Clinical Trials

The investigators will refine their new model for evaluating the ability of experimental treatments to curtail growth of deadly brain tumors called gliomas.  The new treatment evaluation method, if successful, will provide a faster and easier route for progressing from preclinical animal testing to clinical trials in humans, an important advance since gliomas are usually fatal within a year.

Currently, glioma treatments are tested in animals using a time-and-cost-intensive method.  The testing requires that tumor cells be injected in mouse brains.  MRI imaging is then used over time to view the extent of tumor growth during treatment with experimental therapies.  The greater the tumor cell growth, the less effective is the therapy.  The Dana-Farber investigators hypothesize that they have a faster, cheaper, and easier way to evaluate glioma treatment effectiveness in animal models.

These investigators have genetically engineered glioma cells to emit light that can be detected by a sensitive low-light camera.  Now, the researchers will refine this technique to provide an optimal method for modeling brain tumors in mice.  This method first entails using genomic analysis to determine whether growing tumor cells in the laboratory or using cells obtained directly from glioma patients produces an animal model that most closely resembles the human disease.

Next, the researchers will determine how this new imaging technique can be used to identify the molecular characteristics of tumor cells that respond to the signals that direct them to continue to grow.  To achieve this aim, investigators will create the capacity for the tumor cells to emit light in a stimulus-specific manner.  The greater the intensity of the stimulus to continue growth, the greater the amount of light that is emitted.  In this way, the researchers will be able to assess the effectiveness of various therapies that are designed to inhibit these growth-promoting signals.  Those therapies that hit growth-promoting signal targets and inhibit them will cause lower levels of light to be emitted by the tumor cells.  If this technique works, the investigators will plan to seek funding from other sources to evaluate the effectiveness of various experimental therapies in hitting a target they have identified that produces growth-promoting signals.

Significance:  If investigators can effectively use light emission by tumor cells as a reflection of the strength of signals that direct the tumor cells to grow, this new method will accelerate animal tests of new anti-tumor therapies.  Faster animal test results will facilitate earlier testing of promising therapies in humans.  Since gliomas are quickly fatal, this acceleration from animal to human testing is critical.


Accelerating Brain Tumor Drug Discovery Through Cellular Imaging

Therapies that target changes specific to tumor cells hold the promise of increased efficacy with decreased toxicity to normal tissues. Although there have been great advances in our understanding of the molecular causes of brain tumors, translation into new therapies has been slow. There are several impediments to the development of new brain tumor therapies, including complexities in modeling brain tumors in small animals and difficulties in predicting blood-tumor barrier exclusion of therapeutics. This proposal seeks to develop in vivo models that ameliorate these complexities, with the goal of accelerating the drug discovery timeline for new brain tumor therapeutics.

This pilot proposal seeks to expand the cellular imaging techniques we have previously developed in which orthotopically implanted brain tumor cells are quantified by in vivo bioluminescence imaging. We plan to establish a series of orthotopic glioma models utilizing a panel of cell lines with defined molecular changes. We will also attempt to establish orthotopic explants of primary patient tumor cells, which will be strictly propagated in vivo. Global expression profiling will be used to evaluate the fidelity of these in vivo models, along with in vitro models, in recapitulating the expression profile of primary human tumor samples. The goal here is to objectively determine the optimal system for modeling human brain tumors.

To evaluate the feasibility of this system for therapeutic evaluation, we plan, in the course of our studies, to utilize the system to evaluate a number of candidate therapeutics. We anticipate that the strongest proof of the utility of this system will be demonstrated by the establishment of new clinical trials based on in vivo efficacy studies utilizing the proposed platform.



The pace of discovery of new therapies for brain tumors can be accelerated by the development of more facile mouse brain tumor models.

The goals of this pilot award are to demonstrate the feasibility and applicability of in vivo cellular imaging for following the growth of brain tumors in mouse models, and to use such models for the rapid evaluation of candidate therapeutics.

1. Establish a panel of cell lines with representative characteristic molecular abnormalities found in human malignant gliomas, and to characterize each for in vivo modeling of human glioma.
2. Attempt to establish primary human brain tumor xenografts.
3. Evaluate the fidelity of in vivo and in vitro brain tumor models using global expression profiling.
4. Evaluate novel targeted therapeutics using in vivo cellular imaging models for establishment of human clinical trials.

The models developed through these studies have been adopted as an extant platform for the testing of potential new brain tumor therapies within our institute, and have been exported to other academic and industry institutions. This model has several advantages, including being high-throughput, high reproducibility, and orthotopic tumor location. We are now extending these approaches to the propagation of primary glioma stem cells, which hold the promise of improved fidelity in modeling human tumors.

Selected Publications

Woerner B.M., Warrington N.M., Kung A.L., Perry A., and Rubin J.B.  Widespread CXCR4 activation in astrocytomas revealed by phospho-CXCR4-specific antibodies.  Cancer Res. 2005, 65: 11392-11399 .

Szentirmai O., Baker C.H., Lin N., Szucs S., Takahashi M., Kiryu S., Kung A.L., Mulligan R.C., and Carter B.S.  Non-invasive bioluminescence imaging of luciferase expressing intracranial U87 xenografts: correlation with magnetic resonance imaging determined tumor volume and longitudinal use in assessing tumor growth and antiangiogenic treatment effect.  Neurosurg. 2006, 58: 365-372 .

Redjal N., Chan J.A., Segal R.A., and Kung A.L.  CXCR4 inhibition synergizes with cytotoxic chemotherapy in gliomas.  Clin Cancer Res. 2006, 12: 6765-6771 .