gantenerumab is testing in MCI/prodromal. It is between asymptomatic and mild to moderate AD. I think the chance of success is reasonable in this space but higher in asymptomatic. I think roche is doing the right thing with gantenerumab. they used a pivotal 2/3 to pass bapineuzumab in the MCI hunt. I'm not sure it will work, but it was a good strategy.
Just doesn't sound good that 45% of neurons are already dead by the time they will start their testing. I think it seems more plausible that, if the amyloid hypothesis is correct, and of course that's still in play as you noted, that the best chance of success would be testing at the earlier asymptomatic stage.