WASHINGTON (AP) - An aggressive form of brain cancer secretes a chemical that kills nearby cells and carves out a path for its own growth, say researchers studying the disease in rats.

Their findings could point to a way of slowing the growth of the disease in humans using drugs already available. The research team, led by Dr. Maiken Nedergaard of New York Medical College in Valhalla, N.Y., found that in the brain cancer known as glioma, cancer cells secrete a chemical called glutamate.

Glutamate is normally present in the brain, which uses it to transmit messages between cells. Excess amounts cause degeneration of brain cells.

"In essence, our study shows that gliomas actively secrete a compound, glutamate, that is toxic to surrounding neurons, and thereby utilize a unique way of promoting their own growth by killing their surroundings," explained Nedergaard via e-mail from Denmark.

"Our data suggest that the more malignant the glioma cells are, the more glutamate they release," Nedergaard said. "Other cancer types do not have this type of aggressive behavior. I think that the exciting part of our study is that it may give new hope to treatment since gliomas respond poorly to chemotherapy and radiation," she added.

Drugs that block the release of glutamate are already in use in other diseases, such as Parkinson's disease, and could be added to the treatment of patients with glioma, Nedergaard said.

"We found that tumor growth was reduced 30 percent to 60 percent depending on the type of gliomas," Nedergaard said of the rat studies.

"We were not able to cure with glutamate antagonist, but to increase survival. Possible development of more efficient drugs may improve the efficacy further," she said.

While the new findings come from studies of glioma in rats, Nedergaard said studies on humans are planned. She said that because the drugs that reduce the effect of glutamate are not toxic and have few side effects, they can be combined with more traditional approaches such as surgical removal, radiation and chemotherapy.

Less confident were Drs. Jeffrey D. Rothstein and Henry Brem of Johns Hopkins University, who cautioned against drawing overarching conclusions from the work. They noted that the rat studies were limited to two cancer cell lines, one of which has not proved a good model for human disease.

Nonetheless, Rothstein and Brem wrote in a commentary on the research, the study opens up a potential new treatment approach to fatal central nervous system tumors.

Nedergaard's paper and the commentary by Brem and Rothstein appear in Friday's issue of the journal Nature Medicine.

Working with Nedergaard on the project were Takahiro Takano, Jane H-C. Lin, Gregory Arcuino and Qun Gao of New York Medical College and Jay Yang of the University of Rochester Medical Center in Rochester, N.Y.

Copyright 2001 The Associated Press. All rights reserved.