When Should We Start Taking Drugs to Prevent Alzheimer’s?

Experimental drugs for Alzheimer’s disease have a 99.6 percent failure rate. Not such a great track record. But, according to an expert in the field, that may be because we’re not giving the drugs to people who need them early enough. After all, we know that the brain starts showing the changes that lead to Alzheimer’s decades before symptoms appear.

John Hardy is a leading Alzheimer’s expert who recently won the Brain Award, the top prize in neuroscience, for his research on beta-amyloid. Hardy has long been in the business of unlocking secrets in the brain—he discovered the mutations in the genes related to early onset Alzheimer’s, kicking off a race to block the resulting beta-amyloid from building up in the brain. He’s also a father of the amyloid hypothesis, the leading theory in how changes in the brain lead to Alzheimer’s.

We spoke to Hardy about the ongoing beta-amyloid debate and what it might look like to find out your risk for Alzheimer’s via a personalized healthcare assessment called the polygenic risk score.

• Amyloid-beta is a sticky protein that builds up into plaques in Alzheimer’s disease. Scientists think it causes neuron death, which in turn destroys cognition and obliterates memory.
• The polygenic risk score is an assessment that combines dozens of genes to paint a picture of the overall Alzheimer’s risk
• Anti-amyloid drugs that aim to clear amyloid or stop it from building up would work like statins for high blood pressure, said Hardy, which are taken years before someone might face conditions that would lead to a heart attack.

Being Patient: It seems like there was a progression over the past decade to focusing on beta-amyloid as a drug target, and now there’s some doubt that amyloid might be the key to unlocking the cure for this disease. Where are we with that?

John Hardy: You know, for a very long time there was a fierce argument between what I call the amyloid believers and the non-amyloid believers. I am an amyloid believer, if you like. I think that now that argument is over, and I think the reason it’s over is because we understand more about the progression of the disease.

Amyloid in the brain starts to increase 20 or 30 years beforehand. Tau tangle pathology starts to come in later. And I think that once you start to see that this is a sequence of events, it kind of resolves the arguments.

People now almost universally accept that amyloid starts the disease, but they also accept universally that tau tangles are much closer to the clinical symptoms of the disease. That has implications from a therapy point of view. One of the things that we are discussing now is when might amyloid therapies work, and I think that everybody believes—and it’s a belief, not a fact—that anti-amyloid drugs would work much earlier.

Being Patient: So how old, typically, are people when they would need anti-amyloid drugs?

John Hardy: Well let’s say that somebody coming into the clinic at age 70 with symptoms. That would be a very typical case. But I think anti-amyloid drugs might be like statins for high blood pressure. You take statin for heart disease, but if you’re having a heart attack, you don’t pop a statin tablet. You take your statins years ahead of time. And that’s what I think people are believing with anti-amyloid drugs.

Being Patient: So to assess those people and get them the drugs at the right time, are we looking at something like a polygenic hazard score?

John Hardy: Well, that’s what I am arguing—that we should use the polygenic risk score to identify those of us in the general population who are at high risk.

Being Patient: Is doing the polygenic hazard score kind of like getting the 23 and Me test for the ApoE4 gene?

John Hardy: It’s a bit better than that. It’s a better risk assessment because that’s not what 23 and Me gives you.

Being Patient: But you’re saying that in order for amyloid therapy to work it has to be in someone’s hands at probably age 40-something.

John Hardy: In a perfect world, what you might do is have everybody do a polygenic risk score analysis by the age of about 60, and then follow up on the ones with a high risk with cerebrospinal fluid taps or amyloid scans thereafter. One could imagine doing that.

Of course, we hope that these are intermediate problems because we hope that then we will have therapies and this will not be an issue. You know, that’s the hope. This is a difficult stage just now while we can diagnose it but not treat it. But eventually we’ll get in a better position.

Being Patient: I guess on the other side of that too, is that people who seemingly have a low risk still sometimes develop dementia. If you tell someone they have a low risk they might say, “Well, I’m good. I don’t need to do the prevention techniques that are lifestyle related.”

John Hardy: Sure. But on the other hand, if you just put out generalized health care advice, it’s partially effective but it’s still generalized. Whereas if you say, “You need to watch out for Alzheimer’s disease,” that’s more effective I think. And then to someone else you say, “You need to watch out for bowel cancer.”

That focused advice is, I suspect, more effective than “Everybody needs to watch out for bowel cancer and Alzheimer’s disease.”

Being Patient: So in an ideal medical future …

John Hardy: Which we’re actually retreating from at the moment.

Being Patient: Will we have more individualized healthcare?

John Hardy: One could imagine polygenic risk scores for every disease outcome. We can do polygenic risk for, as I mentioned, bowel cancer. We can do it for rheumatoid arthritis. We could even do it for infectious diseases actually, because susceptibility is genetic. And you know, the process would just be the same, just different algorithms would fit different risks. One could imagine that future. That’s not science fiction.

Being Patient: And so the test is available now for Alzheimer’s.

John Hardy: Partly. It exists, but it’s not that you can go and get it privately. Although, I think it’s getting close to that. You can of course, as we discussed do a 23 and Me genetic assessment and that is not as good as a polygenic risk score but it gives you some of the information.

Being Patient: So, do you know how it’s being licensed? Is it relatively cheap to produce?

John Hardy: What’s happening is the company, Cytox, who I am a co-grantee with, they are marketing it to drug companies for clinical trials. Not for medical use but for clinical trial use. That’s what they’re doing at the moment.

Being Patient: What it would look like for a patient to try to get this?

John Hardy:  It’s not there yet. One could imagine a future where you got a report back, which said you have a 70 percent risk of getting Alzheimer’s disease by the age of 70. That scientifically is not a difficult future to imagine, quite soon. There are regulatory issues and all this sort of stuff, that might impinge on that.

Being Patient: Is it going to be something that’s available to the average person?

John Hardy: Not in the next three years. Scientifically we could do it now. It would be not at all difficult, for example, for me to take a blood sample from you and in a month’s time tell you what your risk is. That would be an easy thing to do.

Being Patient: Is that something people want to know?

John Hardy: In some trials they’re told. In other trials they’re not. There’s a lot of discussion amongst the trial designers about which is the most appropriate. Many of the trials, if people want to know their genetic status they can find it out.

With Huntington’s disease, [another brain disorder that causes neuron death], they thought that everyone would want to get tested. When the test became available, only something like 10 to 15 percent of people took it. When people actually started to think about what the consequences of knowing would be, they decided not to. You know, you have to respect that.

Being Patient: Yeah. I think that genetic counselors still do not recommend that people find out their ApoE4 status. Because with Huntington’s it’s like a certainty, right? And you can make choices based on it.

John Hardy: Yeah, but if you’re an E4 homozygote [with two copies of ApoE4], it’s pretty certain. Three percent of the population are E4 homozygotes. If you’re an E4 homozygote, your chance of getting disease by age 80 is 80 to 90 percent.

That’s as good as a Huntington’s gene test if you’re in the intermediate range. I think the reason the genetic counselors don’t want to do it is actually because they could not handle the number of people that they might be dealing with if they started to recommend it. I think it would just overwhelm the genetic counseling system. But, hopefully it’s a temporary situation until we find treatments.

This interview has been edited for clarity and length.