Scientists say the study released on Tuesday is too small to show whether the drug will ultimately prove effective, but any sign of progress in Alzheimer's research is welcome. Scientists expect Baxter's results next year.

"Over the last few years there have been a lot of conspicuous failures with Alzheimer's drugs," said Dr. Norman Relkin of Weill Cornell Medical College in New York. Relkin presented three-year results on Baxter's study at the Alzheimer's Association International Conference in Vancouver.

Most Alzheimer's patients typically decline over three to six months, so to have the disease stabilize in four patients -all treated for three years with the same dose of the same drug - is "a very unexpected and very positive finding," he said.

There are no licensed drugs that can slow the progression of Alzheimer's, and currently approved medications only treat symptoms. The disease is the most common form of dementia, affecting 5.1 million Americans and 35 million people globally.

Relkin presented three-year data on 16 patients in Baxter's Phase 2 study of Gammagard, which originally enrolled 24 participants.

Overall, all 11 study participants who got Gammagard for the full three years showed improvements in thinking abilities, behavior and daily functioning.

Of these, the four patients who received the most effective dose (0.4g/kg) for the full 36 months fared the best, showing no decline on several standard measures of cognition, memory, daily functioning and mood.

Many drugs fail despite very encouraging Phase 2 results, Relkin said. The real proof of Gammagard's effectiveness will come early next year with the release of results of the large-scale, Phase 3 study, which he is also leading.

Experts predict that without an effective treatment, the number of Americans with Alzheimer's will double by 2050 to more than 10 million, and related healthcare costs could soar above $1 trillion a year.

Other Alzheimer's treatments in late-stage clinical trials - Johnson & Johnson and Pfizer Inc's bapineuzumab and Eli Lilly's solanezumab - use powerful, manmade antibodies to target a protein called beta amyloid that accumulates in the brains of patients with Alzheimer's disease.

Late-stage studies of bapineuzumab and solanezumab are expected within the next few months.

Gammagard is different from them. The drug is an intravenous immune system treatment made from natural antibodies taken from young, healthy blood donors. Known generically as intravenous immunoglobulin, or IVIG, the therapy is typically used to fight infections in patients with weakened immune systems.

Compared with solanezumab and bapineuzumab, which Relkin describes as "industrial-strength vacuum cleaners" used to remove beta amyloid from the brain, he says IVIG's natural antibodies are more like "gentle, whisk brooms."

"It may be that a gentler approach is a more realistic approach in the long run," he said.

RELATIVE INDEPENDENCE

Dr. Sterling Williams, an obstetrician/gynecologist from Silver Spring, Maryland, has watched the drug at work. His wife Joice, 71, is one of the four patients who has benefited from Gammagard and is completing her sixth year on the treatment.

"There has been some decline, but it has been very slow, and she is still relatively independent in functioning at home and in her activities of daily living," said Williams, who is also 71. "Every morning, she gets up, fixes breakfast, washes clothes, irons -- all of that is fine."

She becomes a bit confused in a new environment, such as a hotel, Williams said.

Some of the other study patients also still function independently. "They travel by themselves," Relkin said. "They take part in community activities. They volunteer."

Over time there have been declines, he said, but these have been consistent with normal aging.

Dr. Ralph Nixon, an Alzheimer's researcher at New York University Langone Medical Center, said the finding was encouraging because it is taking place in a population of patients who were already affected by the disease.

"It's one of the first trials to really show some signs of therapeutic benefit in an affected population," something he described as a "very meaningful and hopeful sign."

Dr. Laurie Ryan, program director for Alzheimer's disease clinical trials at the National Institute of Aging, one of the National Institutes of Health, said the study's findings suggested the treatment should be tested further.

If Gammagard does prove effective in the larger trial, supply and cost may limit its use.

Each dose contains pooled antibodies from the plasma of 1,000 blood donors. Typically, each of the two treatments a month that patients need costs between $1,500 and $3,000, depending on a person's weight.

If the drug works, the challenge will be determining which of the antibodies are actually helping patients and developing a product that is easier and cheaper to make, Relkin said.

(Editing by Lisa Von Ahn)

• A Permanent Economic Emergency