When Senator Edward M. Kennedy of Massachusetts was diagnosed with a malignant brain tumor in May, Americans were reminded of the gloomy prognosis faced by many brain tumor patients. According to the National Cancer Institute, of the 20,000 people diagnosed with malignant brain tumors in the United States every year, two-thirds will die in five years or less, only a modest improvement in survival rates from those seen 30 years ago.

Glioblastoma multiforme, the most malignant type of brain tumor, is especially hard to treat because it spreads quickly throughout the brain. Isolated tumor cells invade the surrounding area, migrating deep into normal tissue and making complete tumor removal almost impossible with conventional methods such as chemotherapy, radiation and surgery. In addition, there is a risk of the brain being damaged in the treatment process, causing some functional loss.

To avoid these problems, for over a decade scientists have been exploring whether viruses could be made to infect and destroy tumor cells, leaving normal cells intact. While some viruses have a natural affinity for cancer cells, others have to be genetically engineered to increase their tumor-destroying potential. But despite some promising leads, no virus has yet been found that can be used to successfully treat brain tumors in people. Recent work by School of Medicine researchers, however, has unveiled a new virus candidate that might have the potential to completely destroy brain tumors.

The engineered virus, called VSVrp30, was first described in 2005 by a team of researchers led by Anthony N. van den Pol, Ph.D., professor of neurosurgery, in the Journal of Virology. The group tested nine viruses against brain tumor cells and found that a virus called vesicular stomatitis virus, or VSV (which causes a mild disease in cattle), worked best. The virus was grown for many generations, alternating between cultures of human glioblastoma cells and normal human cells. Viruses that grew on normal cells were discarded until the researchers arrived at a virus population that could completely destroy a tumor with minimal infection of normal cells. “We made the virus do what we wanted it to do by putting evolutionary pressure on it,” says Guido Wollmann, M.D., associate research scientist in the Department of Neurosurgery and lead author of the 2005 study. The resultant virus, with its high selectivity for tumor cells, was named VSVrp30.

Wollman, postdoctoral fellow Koray Özduman, M.D., van den Pol and Joseph M. Piepmeier, M.D., the Nixdorff-German Professor of Neurosurgery, recently tested VSVrp30’s efficacy against brain tumors in live animals and published their results in the February issue of the Journal of Neuroscience.

The group injected human brain tumor cells, with an inserted gene from coral that would cause them to glow red under the microscope, into the brains of mice. Solid brain tumors, similar to those seen in humans, soon formed in the mice’s brains. VSVrp30, modified with an inserted jellyfish gene that causes green fluorescence, was then injected into the mice’s tail veins. Within 72 hours, the researchers saw that the virus, glowing green, had selectively infected and destroyed the red brain tumor cells while sparing normal cells, even when two or three tumors were present in different parts of the brain.

With VSVrp30’s high selectivity for tumor cells the researchers hope they can use the virus to locate and infect cancer cells not only in the main body of the tumor but also cells that have dispersed to other parts of the body.

As a control, normal brain cells were injected into mice and, as predicted, VSVrp30 did not infect these cells. One of the greatest difficulties in getting drugs to reach the brain is a protective blood-brain barrier. “We found that the virus managed to cross the blood-brain barrier and infect tumors,” says van den Pol, “but it did not cross the blood-brain barrier when there was no tumor” because the presence of a tumor impairs the barrier, allowing the virus to pass through.

The biggest concern about using viral therapies is the possibility that the viruses might infect normal cells. So far, however, the virus seems to have a preference for tumor cells. Normal cells, when they sense a virus, release interferon that serves as a signal for other cells to up-regulate their antiviral defense. But tumor cells have poor antiviral defense and don’t release interferon when they sense a virus. This, combined with the high sensitivity of VSVrp30 to interferon, reduces the affinity of the virus for normal cells, the scientists say.

The group also found that the virus could travel along a nerve fiber and enter the brain. When they injected VSVrp30 near the olfactory nerve, the virus entered the brain through the olfactory bulb, a technique that allowed the scientists to avoid placing the virus into the bloodstream. “This is another route we can explore to enhance targeting the problematic part of the brain,” says van den Pol.