Wednesday, July 02, 2008

Alzheimer's Disease

Some biotech blogging:
Technology Review reports that a Californian company Satoris has developed a simple blood test that can detect Alzheimer's Disease (AD) 2 years before clinicians.
Currently, AD can only be definitely diagnosed after death, and before that clinicians define Alzheimer's as dementia for which they have found no other reason. So a good diagnostic tool is certainly required.
However, two issues arise. The company says that their test is "nearly 90% accurate". That might sound good, but then again for a 70-year-old, the probability of having a AD is around 0,5%. Now, imagine that we test million 70-year-olds and 5 000 of them have AD. Our test is able to detect 4 500 of them. But also the test gives us 10% false positives among the healthy 995 000, which amounts to 99 500 cases. Alltogether we receive 104 000 positive test results but only 4 500 of them actually have the disease. So all of a sudden, our 90% accurate test can only give us a 4,3% probability of a person having AD.
The other problem is reported in the article:
However, until there are better treatments for Alzheimer's disease, [David] Michelson [vice president of neuroscience at Merck] says that diagnostic tests will be most useful as a research tool because "what's the use of confirming a diagnosis early if you can't do anything about it, if you don't have a treatment that could change what can happen?"
Alzheimer's Disease certainly remains one of the most important challenges for pharmaceutical companies, but most obviously a real breakthrough is nowhere in sight. Well, they have 39 years before i turn 65 and enter the risk group!

No comments: