Note: I'm still working my way through the information on the much-hyped TauRx drug, Rember - a post on that is coming. Here's more from the same Alzheimer's meeting, though:

Elan and Wyeth unveiled the data on their widely anticipated Alzheimer’s drug bapineuzumab yesterday. This is another antibody from Elan’s shop, part of a long-running effort to induce an immune response to the amyloid protein which is thought to be a key player in the development of disease. And. . .well, this is an Alzheimer’s drug. That means it comes with all the standard baggage: it’s trying to treat an extremely difficult disease that we don’t understand very well, by a mechanism that no one can be sure will work or is even relevant. (Cue up this discussion from last week around here!)

This drug was always expected to have its best chance of working in patients without the APOE4 mutation, a lipoprotein which was identified in the 1990s as a significant risk factor for Alzheimer’s. Update: I shouldn't have used "always" there, since this was picked up during Phase II. But that shows that Wyeth and Elan did have it in mind as something to look for. The Phase III trials will, in fact, be stratified according to APOE4 status. And so it did – but not as dramatically as everyone had been hoping. About one-third of Alzheimer’s patients lack the APOE4 mutation, and this cohort showed slower decline in their brain functions with bapineuzumab treatment. But how much slower? The trial used a standard survey scale (ADAS-COG) – on that one, the existing Alzheimer’s drugs (Aricept, e.g.) show at most a 3-point effect, while bapineuzumab showed a five-point change.

That’s probably real, but I’m not sure how much that’s going to mean in the real world, and it’s certainly less than one would want. On top of that, the drug showed little or no benefit (and more side effects) in the two-thirds of the patients who have the APOE4 alleles, which meant that when all patients in the trial were taken together, improvement over placebo didn’t reach significance. And since this trial doesn’t seem to have been designed from the start to distinguish between those different patient groups, that’s the only number that you can take away with any certainty. All the other analyses are ex post facto, and thus carry less weight.

Investors, some of whom were clearly expecting a lot more than this, have not reacted well to the news: Elan’s drop has been taking the whole Irish stock exchange down along with it today. They have several other Alzheimer’s therapies in development, but the worries are starting to develop about the effectiveness of all the approaches that target amyloid. You can see some of those concerns being aired out in the latter half of the Forbes article. Some of the stronger statements are from people who are backing alternate hypotheses, which you should keep in mind, but there’s no doubt that the amyloid hypothesis for Alzheimer’s is still very much unproven. (Perhaps Lilly can shed some light today, but I doubt it, to tell you the truth). It’s going to be a long time before we can stop using that disclaimer that I had in the first paragraph.