From Our 2008 Archives

Finding Points to Possible Blood Test for Brain Tumors

By Jeffrey Perkel
HealthDay Reporter

MONDAY, Nov. 17 (HealthDay News) -- Researchers have stumbled across a novel mechanism by which brain tumors model their environment to nurture their own growth.

The findings expose a fundamental biological process of which oncologists were previously unaware, and which may one day be exploited to combat glioblastoma, experts said. But just as significantly, they provide a potential means to monitor cancer progression and treatment by a simple blood test, rather than having to rely on either brain imaging or biopsies -- something that currently is not possible.

"Wouldn't it be awesome to know that a tumor is not fully treated or recurring before you ever see it on an X-ray or an MRI?" mused Dr. Paul Graham Fisher, a neuro-oncologist at Stanford University. "Or even beyond that, I mean, this gets way down the road, could you do a blood test to detect a brain tumor? That would be just awesome."

In findings reported in the Nov. 16 issue of Nature Cell Biology, Xandra Breakefield of Massachusetts General Hospital, and her team, led by Johan Skog, report that glioblastoma cells secrete small membrane-enclosed sacs called microvesicles. Filled with proteins and genetic material, these vesicles are picked up by neighboring cells in the brain, where they apparently induce them to alter their gene expression program to suit the needs of the tumor. Specifically, these cells can be coaxed into forming new blood vessels to supply nutrients to the growing cancer mass.

"We think the tumor cells bud off these vesicles filled with information, genetic and protein information, to actually take over their environment," said Breakefield. "I mean, they are doing it for a purpose, and they're doing it with a vengeance."

In fact, the tumors secrete so many of these vesicles that they can be found in the patient's bloodstream, outside the brain. When Skog analyzed the protein and RNA content of these microvesicles (also called exosomes) in the blood samples of glioblastoma patients, he found that he could obtain a sort of molecular snapshot of the tumor.

"By analyzing the RNA in these serum exosomes, we could determine the mutational profile in the tumor without doing a biopsy," explained Skog. "So, this is an enabling technology to look at what's going on inside the tumor at a given moment."

Like all cancers, glioblastoma cells often contain dozens of mutations. One of these, called EGFRvIII, is a mutation in the gene for a growth factor receptor called EGFR. Skog first screened 30 glioblastoma biopsies for the presence of EGFRvIII; he found it in 14 samples. When he examined blood samples from the same patients, he was able to detect the mutation in 7 samples, including two corresponding to biopsy samples that had come back negative for EGFRvIII.

That particular observation made quite an impression on Fisher. "They were really good at showing that a lot of the tumors that expressed [EGFRvIII], they were able to find it in these microvesicles in the blood, and you go, 'wow'," he said. "And then, even on top of that, they found a few tumors where they weren't able to find it, but we know that EGFRvIII is very specific to glioblastoma in a patient like this, and they were able to find it in the blood but not in the tumor, and you go, 'Hot darn, that's just incredible'."

It may therefore one day be possible to test blood samples from glioblastoma patients to monitor patient progress, check for recurrence, and make individualized treatment decisions. For instance, said Breakefield, one current chemotherapeutic is a monoclonal antibody that specifically targets the EGFRvIII mutation. "Obviously, those people would respond better to this if they had that mutation," she said.

More remotely, it may one day be possible to run early screens for the occurrence of cancer long before symptoms arise, when the cancer may be, if not curable, at least more amenable to treatment.

At the moment, the only way to assess the genetic profile of glioblastoma is via brain biopsy. And the only way to detect a brain tumor noninvasively is using brain imaging, such as by MRI or PET scans. Often, by the time patients develop symptoms of glioblastoma, the disease is so advanced as to be incurable, and often, inoperable.

Other cancer types also apparently extrude microvesicles, including breast cancer, suggesting these findings may be generalizable to other solid tumors. But, "we were the first to show for any cancer that you could pick up a mutated RNA in the exosomes of the cancer," Breakefield said.

According to statistics from the National Cancer Institute, about 22,000 Americans will be diagnosed this year with some form of brain or other nervous system tumor; 13,000 brain cancer patients will die this year. Earlier this year, U.S. Sen. Ted Kennedy (D-Mass.) was diagnosed with malignant glioma.

"I think [this study] represents a novel finding and potentially a major breakthrough in the way we measure gene expression in tumors and the ease by which it can be done by simply a blood test," said Dr. John Yu, Director of Surgical Neuro-Oncology at Cedars-Sinai Medical Center, Los Angeles.

First, however, these findings will have to be validated, and their clinical relevance established, cautioned Dr. Jeremy Rich of the Cleveland Clinic, in Ohio. "There's a lot of work that needs to be done in the interval, but to say that little parts of the tumor basically are being given off, and that we can assess those, is a pretty exciting concept," Rich said.

Concluded Fisher, "You have to mitigate the enthusiasm, because it is the first time through, and there are a lot of things that are first-time-through that don't pan out. But as I was reading this, I thought, this is awesome, it really is capitalizing on a new avenue of research in oncology."

SOURCES: Xandra Breakefield, Ph.D., professor, neurology, Harvard Medical School, and geneticist, Massachusetts General Hospital, Boston; Johan Skog, Ph.D., instructor, neurology, Harvard Medical School, and assistant in neurology, Massachusetts General Hospital, Boston; Jeremy N. Rich, M.D., chairman, Department of Stem Cell Biology and Regenerative Medicine, Cleveland Clinic, Ohio; John S. Yu, M.D., director, surgical neuro-oncology, Cedars-Sinai Medical Center, Los Angeles; Paul Graham Fisher, M.D., associate professor, neurology and pediatrics, chief, division of child neurology, and The Beirne Family Director of Neuro-Oncology, Stanford University, Palo Alto, Calif.; Nov. 16, 2008, Nature Cell Biology, online

Copyright © 2008 ScoutNews, LLC. All rights reserved.