If we're going to have Medicare and private insurance companies pay $56K/year per patient for a drug that may not have positive effects that outweigh the risk of side effects, that's a major problem.
Limiting access to experimental drugs is a difficult question. I know someone who has a different poorly-understood disease (not Alzheimer's, I'm being deliberately vague to avoid debates) who managed to get enrolled in a clinical trial for an experimental treatment. She responded fantastically well to the drug, as did a small number of other patients in the study. On average, however, the response to the drug was poor.
She and several other responders have now found each other online and are trying multiple avenues to get back on the drug. They're getting desperate enough that they're pooling funds to order a custom synthesis from another country and have it analyzed for purity by a 3rd party lab. I'm terrified to think of the risks they're incurring, but they're so desperate to return to remission that they'd rather take the risks than continue to suffer. It must feel unthinkably unfair to be given a glimpse of remission, only to be forbidden to continue to buy the drug because it didn't work on a majority of patients.
In their case, it's likely that the disease has multiple causative factors and the drug in question only treats one specific cause. Without stratifying trials by these yet to be determined different causes, it's difficult for trials to show efficacy on large populations. I wonder if Alzheimer's disease could be similar, in that certain subsets of patients respond to the drug but the average patient will not.