Spring 2005

New Frontiers in Treating Glioma

Despite the 30-plus years since President Richard M. Nixon declared the "War on Cancer," no magic bullet has been discovered to slay the disease. In fact, three decades of clinical and basic research have only demonstrated how clever a foe cancer is and how varied each cancer can be. Nonetheless, hopes are rising that a number of parallel developments are making brain tumors more vulnerable than ever before.

"Now we are starting to understand the basic science underlying tumor cell growth and proliferation, and getting the tools to monitor the behavior of cancer cells," says UCSF cancer researcher Dr. Susan Chang.

Although it has long been understood that cancers can differ enough in their natural history to make each type of cancer a separate disease, it is only recently becoming clear how varied cell activity can be in a single tumor from one part to another and as it develops over time.

"The problem is that brain tumors are very heterogeneous, with some areas dying, some very vascular, some not," Chang says. "A given drug might work in one area of the tumor and not another."

The best hope for understanding and treating individual tumors, UCSF researchers believe, will come from developments in four critical areas -- neuroimaging, molecular genetics, proteomics and local control strategies.

"The most striking change in the last few years has been the use of biologically directed therapies using novel small molecule inhibitors as agents," says Dr. Michael Prados, a neuro-oncologist. "Integrating that approach with neurosurgery and knowledge that we gain from neuroimaging, as well as knowledge about DNA, RNA and protein structure, should provide more effective therapies in a relatively short period of time."

Neurological Imaging

Currently, the role of neuroimaging in the clinical management of brain tumors is mostly limited to anatomic imaging to evaluate the extent of the tumor, to assess whether therapy is shrinking the tumor and to assist in diagnosis in some cases, such as a brain-stem glioma.

The potential for neuroimaging in diagnosing and treating glioblastoma is much broader. The rising availability of functional MRI and other imaging technologies can make imaging useful at almost every step of diagnosis and therapy. Neuroimaging can help define the extent of tumor involvement and differentiate between the tumor itself and edema or treatment effect.

Currently, oncologists take biopsies to grade tumors based on their appearance, but there is no reason that such grading could not be done as accurately or more so using imaging. "The chemical patterns in tumors can give a very different picture than X-rays or even biopsy," Chang says. "You can look at sugars, at perfusion, at water or a number of other measures." When a biopsy is needed, high-resolution neuroimaging can assist in guiding the physician to specific areas in the tumor.

High-resolution neuroimaging also may be used to characterize tumor phenotype, highlighting areas of tumor infiltration, hypoxia or angiogenesis. Together, these imaging capabilities will prove useful in triaging patients and in choosing the type of therapy to employ.

Perhaps the most powerful clinical feature of neuroimaging for brain cancer lies in the potential to evaluate the exact pathophysiology of each tumor and to assist in selecting therapeutic agents. Functional MRI (fMRI) eventually will be useful in assessing tumor response to therapeutic drugs on a physiological level. This sort of assessment is highly useful because physiochemical changes in the tumor can be spotted immediately with fMRI, so oncologists don't have to wait days or weeks to spot physical changes that indicate the tumor is responding to therapy.

"The translation of these novel imaging techniques to patient care is facilitated at UCSF because the bioengineers who develop and optimize the scanning technology work hand in hand with the physicians who are involved in the clinical management of the patient," Chang says.

Molecular Genetics

Very large-scale gene assays have introduced a revolution in oncology, one that is just beginning to be exploited. By identifying the patterns of gene activity associated with various tumors, oncologists can more accurately identify tumor subtypes and select therapies that have worked with similar tumors in the past. Further down the road, when the key oncogenic genes are identified for particular tumors, therapies will be created to attack those genes and to exploit known weak spots in the tumor's defenses.

For cancer patients and their physicians, aggressive, rapidly growing tumors are particularly frightening. Such tumors can be destructive and dangerous, offering little time to try various therapeutic options. We now know of a limited number of genes that confer aggressiveness, Chang says. If those genes or their downstream products can be blocked, it may become possible to control growth more effectively and provide time for therapeutic options to work.


Proteomics, the study of protein sequence and structure, will be an increasingly valuable tool in the future. "Proteomics refines our ability to understand tumor cell biology in a way that wasn't even possible a few years ago," Prados says. Gene-based assays are helpful, but it is also important to understand post-transcription protein modifications and how they affect cell growth. One novel approach using proteomics, for instance, would be to analyze serum from the tumor and use those tumor-specific proteins as markers of the disease and its response to therapy.

Small Molecule Control

Interfering with downstream intracellular signals to stop or slow tumor growth is another arena with great promise, Chang says, an area that UCSF is working to exploit. The success of Gleevec in treating leukemia has demonstrated that small molecules aimed precisely at blocking signaling molecules can be very useful in attacking cancer.

The challenge is in finding the right targets and exploiting them fully. There are a number of small molecules and targets currently under study, such as the epidermal growth factor receptor (EGFR), which is involved in driving cell-growing and inhibiting apoptosis.

"The margin of benefit for many new drugs as individual agents isn't that good right now," Chang says, "probably because no one knows how to select the best patients or the most appropriate targets." The problem with brain tumors is compounded because they are very heterogeneous, and drugs that may work on one part of the tumor may not work on another. The challenge for the future, Chang says, will be to intelligently attack multiple targets controlling cell growth and division, and to combine small molecule therapies with radiation and chemotherapy. These strategies are already being explored at UCSF.

Drug Delivery

When chemotherapeutic agents are developed, or when doctors try to use existing drugs, there is still the challenge of delivering those drugs to the tumor. Oncologists specializing in brain tumors face the extra hurdle of getting drugs past the blood-brain barrier.

UCSF has a number of clinical trials involving the controlled delivery of drugs to brain tumors. One of the trials, currently in phase three, involves using convection-enhanced delivery to bring a high concentration of therapeutic drugs to the tumor while providing only low levels of the drug to surrounding tissue and the rest of the body. Convection-enhanced delivery involves catheters that inject the drug under pressure into the tumor. Using this method, physicians can control the concentration of the drug much more effectively than is possible if the drug were simply allowed to diffuse into the tissue.

Dr. Susan Chang and Dr. Michael Prados may be contacted at (415) 353-7500.

Related Information

News Releases

New Strategy Used for Prion Diagnosis
A technique to detect prions in brain tissue is more sensitive than procedures currently used, according to a UCSF research team. The finding indicates that conformation-dependent immunoassay (CDI) should be the standard approach.

Brain Tumor Treatment Varies Greatly
Primary malignant brain tumors are uncommon and often associated with a poor prognosis. A new study found that treatment of these tumors varies greatly among medical centers and can conflict with accepted guidelines of care.