Alzheimer’s disease has one of the highest financial burdens. The Alzheimer’s Association estimates that the cost of the disease in the U.S. will hit $355 billion in 2021.
“Not only is Alzheimer’s one of the most expensive, it’s probably one of the most devastating diseases of humans,” said Dr. Allan Levey, professor and chairman of the department of neurology at Emory University.
However, only symptomatic treatments are available, doing nothing to slow or stop disease progression. “There’s not a single disease-modifying agent, and Alzheimer’s is now the sixth most common cause of death,” Levey said.
The last FDA-approved drug for Alzheimer’s was approved in 2003.
But there’s reason for optimism. Alzheimer’s research has ramped up steadily since the 1990s when researchers uncovered the disease’s first genetic risk factors.
NIH funding for the disease could hit $3.1 billion this year.
In the following interview, Levey shares his thoughts on historical Alzheimer’s disease drug-development efforts and provides a view of what to expect from future research into the condition and other neurodegenerative diseases.
How would you summarize the pharmaceutical industry’s progress on developing new Alzheimer’s therapies in the past two decades?
Levey: Over the last 20 years, the big pharmaceutical companies started focusing on the genetic causes of early-onset Alzheimer’s disease. These are extraordinarily rare autosomal-dominant cases that often have mutations in a gene called amyloid precursor protein that leads to overproduction and buildup of amyloid. That’s certainly one of the things that we’ve known about for a long time as being involved with the disease, but it’s never been clear it’s etiologic.
Many big pharmaceutical companies took the same strategy and invested billions in potential early-onset Alzheimer’s disease treatments. But this is a rare form of illness involving about 0.1% of cases. These are people that get the disease in their 30s and 40s for the most part. And so they’re atypical.
We’ve had a litany of clinical trials from the last 20 years worth of work by Big Pharma testing drugs in either block the production of amyloid or immunotherapies to remove amyloid, and every single one of them has failed. There’s some recent potential progress, but all shots on goal were at the same goal. All the eggs were in one basket.
What’s needed to create a more diversified approach to treat Alzheimer’s?
Levey: Look at what’s happening and other complex diseases like cancer — or take your pick of major diseases. You get to the underlying biology, and you start doing rational development of therapeutics.
The Accelerating Medicines Partnership is using this approach. The NIH said, ‘Let’s bring together a consortium of academic groups, NIH and industry to work together in an open science platform.’
What’s clear is there are now hundreds of novel molecular targets emerging from data-driven approaches that are highlighting pathways that are very important to disease. You can link them to genetic risk factors. The large genome-wide association studies now exceed a couple hundred-thousand people. There are about 40 or 50 low effect size–risk genes for Alzheimer’s disease. Studying those could help us understand what might be causal mechanisms and how the protein changes that occur in the brain relate to those genetics.
You recently co-authored a paper titled published in Nature Genetics. Can you provide a summary?
Levey: When there’s a gene identified with a disease, you’re not specifying a gene that is a risk for the disease — you’re identifying a region of a chromosome. There may be tens to hundreds of genes in that region. Then people are trying to guess which of those genes is the player so that it can be identified as a target for therapeutics.
In Alzheimer’s disease, we don’t even know exactly what gene at that locus is involved in the disease risk.
The way it has been uncovered in other diseases is to use proteins. If you can do advanced analyses, you can do Mendelian randomization if you have the data. And you ask: ‘Are there changes in proteins that we’re measuring. Are they mediating the effects of the genetic risk?’
This has been done for a few disorders in blood and autoimmune diseases. When this happens, the chances of a positive clinical trial that leads to an FDA-approved drug go up about 25 to 30 fold.
Our group has been unique in understanding the brain proteome. We have a whole genome on many of the 2,000 brains we’ve studied.
We just published in Nature Genetics the first analysis of the brain proteome and how it mediates genetic risk for Alzheimer’s disease. We’re able to identify a handful of new putative genes that might mediate risk.
We’re also involved in these other projects with NIH to turn these into novel therapeutic targets.
I lead a program known as TREAT-AD, which is a way of two centers in the country to take these new targets that are coming out of our systems biology approaches and then start taking candidates that have not been studied well — that pharma has been ignoring and start developing the reagents and the compounds hopefully to create tools that can be used to study them to further validate them.
We talked about Alzheimer’s and Parkinson’s. What are your thoughts on how the treatment of other major brain diseases could evolve in the future?
Levey: Depression is even more burdensome than Alzheimer’s. Stroke is the number one cause of disability. Parkinson’s disease is the second most common neurodegenerative disorder after Alzheimer’s.
If you ask any medical student, they would say those are four completely different diseases. But recent genetic studies suggest that genetic risk for one confers a genetic risk for all. We’ve been looking at these things in silos, but these are all related. We’ve found that depression mediates some of the effects of Alzheimer’s. So because we understand protein expression, we can now apply this to any genome-wide study of any disease and make predictions about what the protein mediators are. So we’ve already been able to see these strong links across diseases. We’ve seen this for Parkinson’s and Alzheimer’s. And it’s well known that stroke mechanisms are also clearly involved in both depression and Alzheimer’s.
So we’re trying to take a very different sort of data-driven approach using broad phenotyping. We’re setting up a precision medicine effort for brain health diseases where we can, in a very uniform way, collect information on cognition, movement, mood, depression and so forth. Our Alzheimer’s and Parkinson’s patients will have complete phenotyping.
It is mind-blowing to me that diseases like Alzheimer’s or Parkinson’s start about 20 years before symptoms begin. This is why identifying the genetic risk and biomarkers to identify these people is important. We can’t wait for them to come to a neurologist when they have symptoms. We’ve got to identify them and midlife if we’re going to start intervening and making a difference. That’s clearly going to be the future.