* Test was 94 percent accurate in detecting Alzheimer's
* Test was 84 percent accurate at ruling out disease
By Julie Steenhuysen
CHICAGO, Sept 13 (Reuters) - A simple blood test could diagnose Alzheimer's disease, U.S. researchers said on Monday, a finding that could give more people a chance to be tested.
Other teams have shown spinal fluid tests, which require a spinal tap, can detect early changes that signal the onset of Alzheimer's, which affects at least 26 million people globally.
And imaging companies such as privately held Avid Radiopharmaceuticals, General Electric's <GE.N> GE Healthcare and Germany's Bayer <BAYGn.DE> are racing to finish clinical trials on new agents that can make brain lesions called plaques visible on positron emission tomography or PET scanners [ID:nN11108128].
A blood test would make Alzheimer's diagnosis much simpler, said Sid O'Bryant of Texas Tech University Health Sciences Center in Lubbock.
"A blood test opens access to anyone. Any clinic can do it. Even home healthcare nurses could do it," said O'Bryant, whose findings appear in the Archives of Neurology.
Alzheimer's, an incurable, progressive brain disease, is the most common cause of dementia. It is currently diagnosed by symptoms, and only confirmed by brain examination after death.
To get brain images, patients need to have access to a center that can do the test. And taking spinal fluid requires a lumbar puncture, an often painful procedure many people would rather avoid, O'Bryant said in a telephone interview.
The problem with blood tests, he said, has been accuracy.
O'Bryant's blood test looks at more than 100 proteins, and combines that with information about patients, including whether they carry a key Alzheimer's risk gene known as APOE4.
Then, a computer analysis arrives at a risk score.
They tested this score in people with Alzheimer's and without. "Our overall success rate of detecting those with Alzheimer's disease is 94 percent. Our overall correctness of classifying those without Alzheimer's disease is 84 percent," O'Bryant said.
The next step will be to see if the test can predict who will develop Alzheimer's. The spinal tap test appears to be able to do so.
A separate study in the same journal looked to see if the diabetes drug pioglitazone -- sold as Actos by Takeda Pharmaceutical Co <4502.T> -- cuts inflammation that can kill brain cells in people with Alzheimer's.
A pilot study by Dr. David Geldmacher of the University of Virginia Health System tested the drug in 25 people. Roughly half had Alzheimer's and the other half had normal brain function. None were diabetic.
Geldmacher said the study proved a large test was feasible, but noted the study was done before there was any sign that drugs in this class can cause heart problems.
And he said recent results of a large study that looked at GlaxoSmithKline's <GSK.L> Avandia or rosiglitazone -- a drug in the same class -- showed no benefit as an Alzheimer's therapy.
"I think in the end the current diabetes drugs are not going to play out to be effective Alzheimer's therapies," Geldmacher said in a telephone interview.
Researchers increasingly think the best time to treat Alzheimer's is in the earliest stages, before the toxic proteins that form in the brain kill too many brain cells. (Editing by Todd Eastham)
Our Standards: The Thomson Reuters Trust Principles.