If everyone with AD produces the same amyloid plaque, why should APOE status matter for Leqembi? Same plaque, same antibody . . . it should be same result.
So either the plaque is somehow different (unlikely), or the stats are being botched. One way to do that is to fail to compare apples to apples, i.e., fail to compare APOE4++ treated patients to APOE4++ untreated patients, and instead compare the much more difficult APOE4++ subgroup to the entire treated universe. (Perhaps the stats had become unreliable with a too-small APOE4++ apples to apples sample.) In that situation, even a modestly improved APOE4++ patient could still look worse than the overall group.
Either way, it is problematic for blarcamesine versus APOE4++ patients, because one way or another they are qualitatively different -- qualitatively worse -- than the others. It's especially concerning because it appears that blarcamesine is operating at the genetic/epigenetic level where the potential for such a distinction would be expected to make a difference.
On a larger scale, though, it also provides hope. If APOE4++ is very bad, APOE3++ could be very good. I have already noted the apparent coincidence that the expected APOE4- patient subgroup (29%) matches 12x's prediction for cognitive improvers in the 50 mg arm over placebo (29%), if only that high dose arm is therapeutic. If not just a coincidence, this 29 percent match would help confirm an upstream MOA for blarcamesine.
Apologies for not fully playing out my reasoning here in the original post.
News 
Market Data 
Discover 
