Mice that get weekly injections of an experimental treatment have remained healthy for more than 500 days so far after getting a dose of the rogue proteins that cause the disease. Untreated mice generally died within about 200 days.
The treatment sharply reduced the buildup of the dangerous proteins in the spleen, and the researchers hope that with further work the treatment will have the same effect in the brain.
The results are far from providing a useful treatment for the human version of mad cow disease, called Creutzfeldt-Jakob disease, or CJD. For one thing, the experimental treatment failed when begun in mice that had already developed symptoms.
Still, the work indicates that the approach is worth pursuing, researchers from Imperial College in London and University College London conclude in Thursday's issue of the journal Nature.
CJD is a rare fatal disease that can be caused by infection, inherited, or produced without any known cause. Symptoms include progressive dementia and irregular jerking movements.
The disease has made headlines in recent years because about 130 cases in Britain have been linked to eating contaminated beef products.
CJD is caused by misshapen versions of so-called "prion" proteins. They act like germs, spreading in the body and converting normal prion proteins to the misshapen version.
In the new work, the scientists injected the mice in their bellies with disease-causing prions. Then, a week or 30 days later, they began regular injections of other proteins, called antibodies, designed to latch on to either the normal or the misshapen prions. The antibodies appear to hamper the conversion of normal proteins into hazardous ones, although it's not clear how.
It is also unclear whether symptoms would appear if the treatments were stopped.
The antibodies didn't work if the disease-causing prions were injected directly into the brain. The researchers said the antibodies probably can't easily get into the brain, which would also explain why they don't work in mice that already show symptoms.
Dr. Paul Brown of the National Institute of Neurological Disorders and Stroke in Bethesda, Md., said the failure of the antibodies to work in those two cases poses practical barriers to using them in human therapy. Still, he said they might prove useful someday for staving off CJD in some people, including surgeons or patients exposed to contaminated surgical instruments.