In brains of mice, the key to Alzheimer's?

In the long, painstaking story of scientific research into Alzheimer's disease, this could turn out to be the chapter when the focus swings definitively from mice to men.

After years of silence surrounding the progressive disorder that decimates the human brain, Alzheimer's has gone public in a bold way. Probably not since the devastating infectious epidemics of the 19th century - like tuberculosis or Spanish flu - has any one illness inspired such a widespread and desperate desire for a cure.

But for all the discussion about new insights, and hopes raised by sophisticated advances in neuroimaging, there is still no definitive consensus on what drives Alzheimer's. The search for a cause and a cure could take years and even decades, researchers say, because they don't yet have the tools they need to understand how the disease works.

One problem in making this clear, they point out, is public confusion about what they call "mouse-heimer's" - observations based on experiments with genetically manipulated mouse models that exhibit some traits, but not all, of dementia's relentless march through the human brain.

"Green tea extract and exercise hinder progress of Alzheimer's disease in mice," reads a typical recent release from a university press office. The trouble is, the mice used in such experiments most closely resemble only those humans with three rare genetic mutations that can lead to early-onset Alzheimer's - about 2 percent of all people who get the disease.

"Mouse models have their role," says Michael Greicius, medical director of the Stanford Center for Memory Disorders and associate professor of neurology at the Stanford School of Medicine. "They've been designed - and this is what they do - to mimic the postmortem appearance of the human brain with Alzheimer's disease. But they don't really get the disease like humans do; they have some memory impairments, but they're kind of subtle."

This spring, Greicius urged health journalists writing about each new discovery based on this imperfect model "to take it with a massive grain of salt, and try not to agitate my patients and their families."

The laboratory mice that mimic some aspects of Alzheimer's disease are a useful first step in exploring new therapies, scientists say, but drawing conclusions from this basic research is misleading.

One major distinction: The mice can be programmed to develop the telltale buildup of amyloid plaque in the brain - sticky accretions of beta amyloid proteins that are early markers of the disease. But they do not go on to show the tau protein pathology that accompanies later stages of the disease in human brains - toxic "tangles" of tau that are always present in full-blown Alzheimer's.

It's an important distinction, agrees Michael Mullan, a consultant and former CEO at The Roskamp Institute in Sarasota, which has sponsored Alzheimer's research since 2003. Mullan and current CEO Fiona Crawford were part of a team of scientists who discovered in the early 1990s that the onset of Alzheimer's was related to the brain's accumulation of beta amyloid.

"We're partially responsible" for the mouse-heimer's confusion, Mullan says, "because we found the genetic errors."

These engineered lab mice can be made to contract what might be considered a mild case of Alzheimer's. "You don't have to do too much to them to stop them getting it," Mullan says. "It's not a robust model of late-onset Alzheimer's. You're basically inducing Alzheimer's in an otherwise healthy brain, and a young brain. We have about 300 drugs back there," he adds, gesturing to the Roskamp labs, "that stop mice from getting Alzheimer's. The real trick is translating that to human studies."

Mullan believes this is already starting to happen. In March the biotech firm Biogen Idec announced early success in a clinical safety trial with 166 patients of an antibody that both removed amyloid plaque from brains and clearly slowed the progress of cognitive decline.

Even if Biogen Idec's drug performs well in Phase 3 trials, researchers say, our understanding of Alzheimer's is still at the toddler stage. Greicius emphasizes that while brain scans have helped fill in some blanks, there is still a great deal to be learned from studying donated brains of people who die of the disease.

And Mullan predicts that the path ahead will remain one of trial and error. "The way these things usually play out is that when you have a new treatment for something, it's a big step forward," he says. "Then what you see over a period of three or four years are the people who don't do well. And then those people become a target for new approaches. And then we'll find out: Does the antibody itself stop working over a period of time? But that's how science works - those small steps."