Amarantus Bioscience Holdings Inc (AMBS)

[joboggi]

That is a very good question, and there are several answers.

1. Amyvid is the criterion standard test for studying AD compounds. If you do the searches you will find that in trained hands the SEN and Spec is well over 90%. It is early, and the Open is on, but you can be sure that the numbers are in the 90s. I do not want to get it wrong, but it is a little higher than that.
2. Amyvid can show progression of the plaque. Lympro cannot do that.
3. Currently, after being followed for SEVEN years, numbers that were 92% and 91% dropped down to 73% sen, 81% spec, and 78% accuracy. see posters two and three. Keep in mind that they were BOTH tested with version TWO of the Lympro. Unlike what we have been led to believe of late, the Version two dates to SEVEN years ago, when Provista had control of the test. So, there is NO utility in using Lympro for differenting between AD and whatever dementia they meany by Other dementia in this study.
4. The KEY to poster three is the LENGTH of the study. Diagnoses change over time. Most commonly they get diagnoses added, or at other times changed.
5. There were not nearly enough people in the study or in any of the planned studies to figure out if there is a reason to use Lympro.
6. The preliminary evidence is that there is no reason to use Lympro, and especially not if the 18 subjects all were PD dementia subjects.

7.What is a good Specificity? Well, if you are designing a screening test then you would not want to miss any patients with the disease and could tolerate false positives as long as there were other means of defining your patients. SO, that speaks to a VERY high specificity.

8. If you are looking for a test to be used widely to discern AD v non AD then the current numbers of 88 % sen and 82% spec are horrible and easily eclipsed by Amyvid.

9. If like the gene Apo e, you can get a test that PREDICTS some people will be MORE LIKELY to get AD that is valuable because then you can put say 1000 people with apo 2 on one side and 1000 people with apo e on the other side, give half of each the treatment and half of each the placebo and see if that changes the rate of AD that you expect from historical controls that would be statistically significant and from the control (placebo ) group that would be statistically significant as they are doing in a study right now.

10. Over time, as you study a test of this sort, the numbers invariably go DOWN. That is because of the changing diagnoses, and because the test is used in a larger number of people by TRULY independent bodies who find the ACTUAL results. All too often contract researchers give the desired results, with a wink. They are the compounds that wind up being pulled from the market years later.

11. IT IS COMPLICATED. I could go on, but well, it is, and if was not complicated you would not need a very high degree in order to work in this field.

12. If you had a number in the 90s you could work with it. If you had a number in the say high 70s and it showed something UNIQUE about your patients it would get used in research with an asterick, but not in larger populations.