This is sort of illustrative of a prevalent failure mode in high level R&D strategy in medicine. What they are doing is reasonable as an immediate extension of

This is sort of illustrative of a prevalent failure mode in high level R&D strategy in medicine. What they are doing is reasonable as an immediate extension of trying to map the disease process front to back, which is what pure research should be doing. The outcome of that, however, tends to be that people end up manipulating biochemistry to slow the disease process rather than fix it. You can get useful results out of this - see statins, for example. But you can't get cures. People who take statins will still end up in the same bad place in the end, just a little later. Also, it is very expensive to find safe new states for cellular operation that provide a decent benefit:harm ratio. One right thing to do in Alzheimer's is map the whole thing end to end, them make a decision on what to do. An enormous task that is basically the same thing as understanding the human brain end to end - which is a good thing in and of itself, and if Alzheimer's provokes that work (in the same way that AIDS provoked analogous major efforts to understand viral biochemistry) then good. But in the short term that doesn't produce useful therapies. The other right thing to do is to identify the fundamental differences between young, non-diseased tissue and aged, diseased tissue and work to revert them - even if you don't know how they are causing harm. Successful removal will tell you a great deal about the significance, and steer other research, and in the best case produce a useful therapy. Thus a focus on amyloid and tau, and immunotherapies to clear them, and dysfunctions in cerebrospinal fluid clearance mechanisms, with approaches ranging from simple mechanical adjustment of fluid drainage channels behind the nose to proposed stem cell therapies for the choroid plexus to addressing immunosenescence in the brain's immune system. However, Alzheimer's is such a multifaceted condition that researchers tend to focus on reverting one of these items, then don't produce wondrous benefits because the other mechanisms are still in play, and then the people who prefer to slow disease progression instead arrive to argue that this means they shouldn't look at this direct approach of reverting differences. To fix Alzheimer's will likely require clearance of both amyloid and tau, and even then it is likely that since 60% of sufferers also have vascular dementia, it will be challenging to produce good trial outcomes under the present regulatory system. If damage repair doesn't work, that means you're not repairing enough of the damage.