But it may allow for neural regrowth, at the cost of some preformance and some memories but maybe reconnect things with time and "work". Work that would be useless since plaque forbid proper functionning anyway (IIUC).

> neural regrowth

I didn't know this was possible. Is it ?

You're talking about neurogenesis via progenitor cells and neural stem cells. Evidence was there for several decades but anyone that touched the subject was essentially blacklisted from the neuroscience community. At MIT, they would talk about how Kaplan was denied tenure and basically kicked out of his university because of his lab's controversial focus on neurogenesis. Kaplan feels similarly. Here's his personal account entitled: death of a dogma and a research career.

http://www.ncbi.nlm.nih.gov/pubmed/11576677

It was only recently discovered that healthy brains regrow neurons all the time.

Not throughout the brain. It's only known to occur in two regions, the dentate gyrus and a part of the subventricular zone.

It could be that the damage, once caused, is irreversible. I wonder if taking the drug before cognitive damage occurs will prevent it.

That's been a question for a while, and comes up every time an anti-amyloid antibody reduces plaques but doesn't affect memory impairment. A number of people believe that we need to work on early detection and to start treatment as soon as possible.

Isn't early detection and early treatment the solution to a most problems. e.g. Cancer? We still need solutions for late-stage illnesses.

It's still an open question if the plaques are a cause or symptom.

If they are a symptom then that explains why removing them doesn't change anything.

And I wonder if the drug helps the elastic nature of the brain to better reroute around the damage.