PASADENA, Calif. (AP) _ Scientists cured mice of a disease similar to multiple sclerosis and said their work could lead to successful treatment of the human disease, although a cure for people is 10 or more years away.

The new method also might help scientists develop treatments for a variety of other human autoimmune diseases, including rheumatoid arthritis, lupus and myasthenia gravis, the California Institute of Technology announced Thursday.

In autoimmune diseases, the body's immune system attacks the body's own tissues instead of simply defending against disease-causing foreign invaders.

In multiple sclerosis, certain white blood cells called helper T-cells attack the body's nerve tissue, causing paralysis, muscle cramps and weakness, slurred speech, blurred vision, fatigue, difficulty in coordination, dizziness and loss of balance.

About 250,000 Americans have multiple sclerosis.

Scientists from Caltech and Washington University in St. Louis said they designed special antibodies - called monoclonal antibodies - to destroy the white blood cells that cause a disease in mice resembling human multiple sclerosis.

''While we are quite excited about these results, I think it's important to emphasize at the outset that we have not developed a cure for human multiple sclerosis or any other human autoimmune disease,'' said Caltech biologist Leroy E. Hood.

''In my opinion, our technique is at least 10 years away from human application,'' Hood said in a prepared statement, adding that he wanted to ''avoid raising false hope in those who suffer from this debilitating and quite serious disease.''

During the past decade, other researchers showed monoclonal antibodies could to some extent reverse or prevent the mouse disease. Some even used mouse antibodies to treat human multiple sclerosis patients, who displayed only slight improvement and developed allergic reactions to the foreign antibodies.

The method used by Hood's laboratory involved designing antibodies aimed at the exact type of T-cells that cause the mouse disease, significantly boosting the proportion of mice in which the disease was prevented or reversed.

Of five sick mice given the new treatment, three showed complete reversal of paralysis within two to seven days, and a fourth's complete paralysis was reduced to slight paralysis of its tail. The fifth mouse died. Five other mice were not given the treatment, and their paralysis wasn't reversed.

Related experiments in dozens of other mice showed the disease could be prevented if they were injected with the monoclonal antibodies before scientists tried to make them sick.

The study was conducted in Hood's laboratory by research fellows Dennis Zaller and Gamal Osman, and by Osami Kanagawa at Washington University. It will be published in next month's issue of the Journal of Experimental Medicine.

Development of the new method ''certainly is significant,'' said Dr. George Ellison, a professor of neurology at the University of California, Los Angeles.

If a similar treatment can be developed for humans, ''that would be a major achievement,'' Ellison said, adding that he agreed with Hood such therapy was years in the future.

To perform the experiments, the scientists made mice sick with a disease called murine experimental allergic encephalomyelitis, or EAE, which produces symptoms similar to human multiple sclerosis.

They did this by injecting the mice with a protein found in the sheaths that coat nerve cell fibers in the brain. The rodents' immune systems then produced helper T-cells that attacked not only the injected protein but the protein in the rodents' brains.

After studying the makeup of this type of T-cell in great detail, the scientists designed monoclonal antibodies that, when injected into the sick mice, found and destroyed the disease-causing T-cells while leaving intact T- cells that protect the mice against other diseases.

Hood said use of the technique on human autoimmune diseases is years away because they are much more complex than the mouse ailment. Also, he said, understanding of human T-cells is far less advanced than the understanding of such cells in mice.