International scientists gathering in Washington for a conference this week expressed cautious optimism that they may finally be on the right track to fight Alzheimer's, a disease that already affects millions around the world and is expected to more than double by 2050 as the population ages.

On Saturday, researchers for the first time brought together dozens of Alzheimer's afflicted families - patients, patients-to-be and their healthy loved ones - from as far as Australia and Britain to meet face to face.

Alzheimer's usually strikes older adults, affecting about 1 in 9 people age 65 or over. Less than 1 percent of cases worldwide are the autosomal dominant form, caused by inheriting a gene with a particular mutation that triggers the disease well before the senior years. Children of an affected parent have a 50 percent chance of inheriting their family's bad gene. But if they do, they almost always get sick about the same time their parent did.

"Finally, I got to talk to other people who are going through the same thing," said Dean DeMoe, who with four other siblings inherited the family's bad gene. One sister was spared.

His wife, Deb, said he experiences early memory changes known as mild cognitive impairment, but DeMoe still holds a job with an oil company and said, "I don't dwell on it."

Families' first question: Why not try to fix the gene defect that causes this form of Alzheimer's instead of targeting its downstream effects? Why, asked others, can't desperate families get faster access to experimental drugs, as AIDS patients once did?

Many of these families are part of the Dominantly Inherited Alzheimer Network (DIAN) study that monitors the health of family gene carriers and their healthy relatives in several countries. Recently, it showed that silent changes in the brain can precede the first memory problems by 20 years.

Now scientists think the best hope against Alzheimer's is to treat high-risk people long before symptoms appear, aiming to at least stall the disease if not prevent it.

On Sunday, researchers at the Alzheimer's Association International Conference reported possible new ways to predict who will get sick with the more common late-onset Alzheimer's - vital to testing such treatments - and a single test probably won't be enough, said Johns Hopkins University neuroscientist Marilyn Albert.

-Tracking about 350 people starting in middle age, Albert's team found a combination of tests predicted development of mild cognitive impairment within five years. They include a spinal tap to measure toxic levels of Alzheimer's hallmark amyloid and tau proteins; MRI scans to detect shrinking brain regions; and two standard memory assessments. The combination isn't ready for doctors' offices, but should help drug companies tell who to enroll in early-stage treatment studies, she said.

-Scientists at VU University Medical Center in Amsterdam found another protein, named neurogranin, in spinal fluid. It may signal that connections called synapses are dying, making it harder for brain cells to communicate.

-Researchers at the University of Alberta, Canada, are beginning to hunt a saliva test for earlier markers of cognitive decline.

A second DIAN study now is testing whether either of two experimental drugs might give those gene carriers more symptom-free years by fighting buildup of sticky amyloid in the brain. That study soon will expand to test additional drugs.

"The goal here really is to get drugs approved to help everyone," said Dr. Randall Bateman of Washington University in St. Louis, who oversees the DIAN drug study.