Looking Beyond Amyloid in Alzheimer's Diseases

Over the past 30 years, billions have been spent putting the amyloid hypothesis in Alzheimer's disease to the test.

After several failed attempts at targeting pathways involved in the accumulation of this protein in the brain, researchers are acknowledging that there's probably more to the Alzheimer's picture -- that it may be something else, or some other combination of factors, that causes the neurodegeneration that leads to the disease's characteristic cognitive impairment.

"There is no question in 2015 that amyloid is not the 'proximate' cause of cognitive impairment," David Knopman, MD, of the Mayo Clinic, told MedPage Today. "Instead, both by amyloid-dependent and amyloid-independent pathways, neurodegenerative changes such as synapse loss and neuronal death are the 'proximate' causes of cognitive impairment."

"Therefore," he continued, "exploration of therapeutic targets and therapeutic strategies directed against neurodegeneration should be an important component of the quest to treat Alzheimer's disease."

It's not that amyloid has no role in Alzheimer's. It's clearly a culprit in genetic forms of the disease, and is present in the brains of the majority of patients with non-familial, or sporadic, forms. But that doesn't mean it's causative in all cases.

Some researchers -- particularly those involved in the development of anti-amyloid drugs -- continue to support expensive tests of therapies founded on the amyloid hypothesis. They suggest that if they target amyloid earlier, before symptoms appear or when they're in extremely early stages, disease progression might be improved and prevention may even be possible.

Others are moving on to other pathways that don't involve amyloid, including tau phosphorylation, lipid metabolism, and inflammation.

Amyloid Hypothesis Debuts

Most trace the genesis of the amyloid hypothesis to a 1984 paper by George Glenner and Caine Wong that isolated beta-amyloid from the brains of Alzheimer's patients (when neuroscientists talk about "amyloid" in Alzheimer's, they're always referring to the protein beta-amyloid).

In 1992, John Hardy and Gerald Higgins published a paper in Science that concluded, on the basis of genetic evidence from Down syndrome patients -- many of whom develop Alzheimer's in their 30s or 40s -- that amyloid accumulation was the likely causative factor for their dementia. Those with Down syndrome have three copies of chromosome 21, which happens to house the gene for amyloid precursor protein (APP), instead of the normal two.

Other genetic evidence has accumulated for familial forms of the disease that point to amyloid overproduction and accumulation as the main cause. In addition to APP gene mutations, changes in the genes that encode the presenilin-1 and presenilin-2 proteins have also been found to be causative -- most notably families in Colombia that have been enrolled in epidemiological studies and clinical trials. These genes are involved in the production of secretases that cleave APP to form beta-amyloid.

Evidence was mounting that amyloid should be the right target. Two major routes of development arose: antibodies that directly targeted amyloid aggregates, and small molecule drugs that inhibited beta- and gamma-secretase, two of three enzymes involved in APP cleavage.

Amyloid Tested

Cracks started to appear in the amyloid cascade hypothesis after gamma-secretase inhibitors bombed in clinical trials. Not only did they not improve cognition; they had significant side effects -- likely because the enzyme cleaves many other important substrates besides APP.

Then, anti-amyloid antibody trials failed in patients with mild-to-moderate Alzheimer's -- most disappointingly with bapineuzumab, followed by solanezumab.

Investigators involved in those trials have conducted subanalyses that suggest the drugs may work if they're given early enough; there were signals that patients with milder disease had greater benefits.

Now, many of these agents are being studied in earlier disease -- led by the promise of early results from Biogen's aducanumab, which is being tested in patients with prodromal and mild disease who have amyloid confirmed on brain scans.

But not everyone is convinced that this strategy of the-earlier-the-better can redeem the amyloid hypothesis.

"It's a reasonable question to be answered," said John Breitner, MD, MPH, of McGill University. "Before you abandon your theory, you should turn over every stone. But you have to ask yourself, when do you say uncle?"

Some companies are also investigating the efficacy of beta-secretase inhibitors, which target a slightly different enzyme from gamma-secretase that's also involved in cleavage of the amyloid precursor protein. Early results were promising, but phase II and III trials are still ongoing.

What If Not Amyloid?

Other factors that have thrown a curveball at the amyloid hypothesis include the finding that about 20% of patients who have certain biomarkers consistent with Alzheimer's but don't have evidence of amyloid plaques on brain scans.

The converse is also true -- a quarter to a third of patients have amyloid accumulation in their brains, but no symptoms at all. This has been termed "preclinical Alzheimer's" but the course of progression, if there is one, isn't yet evident.

This latter group is being studied in the A4 prevention trial, which seeks to enroll 1,000 patients age 70 and over who meet those criteria.

Sam Gandy, MD, PhD, of Mount Sinai Hospital in New York, said the trial is "probably the last best hope for amyloid as key. Solanezumab and the like look unlikely to be beneficial enough after symptoms begin."

Researchers are starting to turn over other stones -- and there are many, at different stages of development. These include tau phosphorylation and accumulation, lipid metabolism, neuroinflammation, calcium homeostasis, autophage response, and mitochondrial cascade.

Indeed, one of the most common mutations that increases the risk of sporadic Alzheimer's is in a gene that codes for cholesterol clearance. Those with the APOE4 gene variant have higher cholesterol, but they also have higher risk of Alzheimer's. This mutation, however, doesn't definitively increase Alzheimer's risk like mutations in APP or PSEN1/2 genes do.

Knopman notes that there are "many promising non-amyloid mechanisms that could be 'druggable'; the field has to figure out which ones affect the course of the disease and the symptoms."

Some experts suspect that Alzheimer's treatment will ultimately involve a combination of therapies tailored to a patient's personalized disease profile.

Zaven Khachaturian, PhD, editor-in-chief of Alzheimer's & Dementia, said his journal has launched a workgroup of international experts to re-evaluate all of the prevailing ideas about Alzheimer's.

"There's no question that 'amyloid' is part of the Alzheimer's story," Khachaturian told MedPage Today. "What is in question, or what should be questioned, is where does it come into the picture? When and how does it affect the clinical features of the disease? Is it a necessary and sufficient factor of the disease process -- i.e. clinical expression -- or merely a co-factor among many other biological processes?"

"The simple fact is that as of AAIC15 [the Alzheimer's Association International Conference last month], there have not been any successful therapies based on the amyloid hypothesis," he added. "All the justifications offered to explain these failures by all the apologists of the amyloid hypothesis notwithstanding, we need to move on and explore alternative avenues for treatments."

• Amyloid Pioneer: It's Not Over Yet
• Two Minds on Amyloid
• Last Hopes for the Amyloid Hypothesis in Alzheimer's?

Comment