Amyloid-beta. A culprit. A conspirator. A bystander.
Is it one, or is it the other? Might it wear the faces of all three?
For 20 years, the search for a cure for Alzheimer’s disease has focused on getting amyloid-beta out of the brain – preventing its formation, preventing its deposition, or clearing it away. Because amyloid-beta was thought to cause the disease, removing it should improve the disease.
It’s now apparent, after a decade of failed anti-amyloid drug trials, that this isn’t true – at least for people with the most common, late-onset Alzheimer’s. By the time symptoms appear, it’s too late to rescue much of anything. Removing amyloid doesn’t improve cognition, no matter what medication is used.
Yet several of these drugs, which had no effect in patients with symptomatic Alzheimer’s, are moving forward in large, long-term studies of groups of high-risk patients. These studies represent probably the most revolutionary turn of thought since the amyloid cascade hypothesis was proposed 20 years ago: If caught early enough, Alzheimer’s may be preventable.
How early is early enough?
In the past 5 years, researchers have learned just how much the brain can take before things go awry. Amyloid imaging studies show that the miscreant protein begins to misfold and accumulate years before symptoms arise. A very recent study of 3,000 people aged 18-100 years found that amyloid could begin its neural stranglehold 30 years before a person first forgets the way home from work (JAMA. 2015 May 19;313:1924-38).
Tau throws another wrinkle into this timeline. Not too long ago, most Alzheimer’s researchers believed tau followed amyloid-beta sequentially. But years of independent coexistence is now thought to be more likely, according to Dr. Gil Rabinovici of the University of California, San Francisco.
“We have known for decades from autopsy studies that almost all older people have tau in their mediotemporal lobes, and that it starts to show up before amyloid,” Dr. Rabinovici said. “Initially, there’s mediotemporal tau and cortical amyloid. When they intersect in the brain, that is what drives loss of neurons and, ultimately, dementia.”
The advent of tau imaging is helping to cement this new concept, he said. The very first serial imaging data, presented earlier this year, suggest a strong interaction with amyloid-beta plaque burden and rapidly declining cognition. But how long these two must tango before memory falters is still an unknown.
If preventing Alzheimer’s means preventing amyloid plaques, and amyloid begins to deposit 30 years before symptoms, is amyloid prevention even possible? Michael Wolfe, Ph.D., of Harvard University, wonders.
“There is ample evidence in support of amyloid as a key initiator of the disease – if not the key initiator – but it may be an impractical target, requiring therapeutic intervention so early that testing the idea is simply unrealistic,” said Dr. Wolfe, who has studied the molecular underpinnings of the disease for 20 years.
“There are serious concerns that these new prevention trials may fail to demonstrate efficacy of anti-amyloid approaches, as recent findings reveal that amyloid-beta deposition can begin as early as 25 years before the onset of symptoms in patients with presenilin mutations, in which the Alzheimer’s disease phenotype is certainly amyloid-driven,” he noted. “If this is the case with the much more common sporadic form of Alzheimer’s disease, then a well-designed clinical trial would be far too long and costly to ever be implemented.”
The prevention studies in the works now range from 3 to 5 years long. Dr. Reisa Sperling, principal investigator on two of the trials, hopes that’s long enough to show some signal of benefit.
“What worries me is how we select the people who are at the right stage that we will be able to track response and decline” in the span of such relatively short studies, said Dr. Sperling of Brigham and Women’s Hospital, Boston. “I don’t know whether, even at the preclinical stage, treating amyloid will be enough. If amyloid starts in the brain 10 years before dementia, and we can bend the curve of decline, slow it by 30%, I think that will cement the idea that if we treat amyloid early enough, we can change the course of Alzheimer’s.”
Because the disease flares from such a long, slow burn, the only practical way to test prevention is in cognitively normal people who carry a genetic risk or who are already amyloid-positive on imaging. All four of the amyloid prevention trials gearing up are recruiting such patients.
But none of the drugs being tested in these trials has ever actually helped a patient with Alzheimer’s. None significantly improved cognition or function in its large phase II or III studies. Their only silver linings occurred in post-hoc subgroup analyses of patients with the mildest disease. In those instances, there were hints of efficacy, showing a consistently slower rate of decline – about 35% in each analysis. This collective trend lent support for the retesting of the drugs as preventives.
Crenezumab is the active agent in the highly publicized Alzheimer’s Prevention Initiative (API) study in Medellín, Colombia, among subjects who have mutations of the presenilin genes. It didn’t succeed in either its cognitive or biomarker endpoints in a phase II trial in 2014. However, there was a 37% decrease in cognitive slowing in the mildest patients, which bolstered confidence enough to keep the study active.
The A4 Trial (Anti-Amyloid Treatment in Asymptomatic Alzheimer’s Disease) is recruiting cognitively normal but amyloid-positive subjects. It employs the anti-amyloid antibody solanezumab – a drug that failed to improve cognition in two pivotal phase III trials of patients with mild disease (EXPEDITION 1 and 2) (N Engl J Med. 2014 Jan 23;370(4):311-21).
A pooled analysis of both studies, however, found a 34% decline in cognitive slowing and a 17% decline in functional slowing, compared with placebo, in the mildest patients.
Solanezumab and another antibody, gantenerumab, are being investigated in the Dominantly Inherited Alzheimer Network (DIAN) treatment trial. The phase II/III trial is aimed at preventing Alzheimer’s in subjects with the autosomal-dominant mutation in amyloid precursor protein, presenilin-1, or presenilin-2.
But gantenerumab, too, is stumbling out of another big clinical disappointment. In 2014, a futility analysis of the phase III SCarlet RoAD trial prompted Roche to end the study, which examined gantenerumab in amyloid-positive subjects with normal cognition.
A fourth study is in the works as well. The API APOE4 trial will test an anti-amyloid antibody and a beta secretase inhibitor in 1,300 cognitively healthy people who are homozygous for apolipoprotein E epsilon-4. The active immunotherapy in the APOE4 trial is Novartis’ CAD106. The name of the BACE inhibitor has not yet been announced, said Novartis spokesperson Katrina Lucking.
Target – or targets?
The consistently poor performance of anti-amyloid antibodies in symptomatic patients has some wondering just why they are still advancing. After all of the letdowns, it should be obvious that targeting amyloid alone won’t change Alzheimer’s outcomes, said Peter Davies, Ph.D., of the Feinstein Institute for Medical Research in Manhasset, N.Y.
“The [amyloid-only] horse is dead,” said Dr. Davies, who shared the American Academy of Neurology’s 2015 Potamkin Prize for Alzheimer’s Research with Dr. Sperling. “Should we just keep flogging it? I don’t see the virtue in just continuing to pound away at something that’s not likely to yield anything. Why would you use a drug to prevent Alzheimer’s that has no activity in patients who have Alzheimer’s? That’s the kind of question you just ask from a common-sense point of view.”
Dr. Davies is one of an increasing number of researchers who think any effective treatment will be similar to diabetes management: a multipronged regimen that starts early, perhaps with lifestyle modification, develops in intensity, and targets several disease pathways. These might include drugs that focus on tau, inflammation, and immune response, for example.
“I think that is an accurate way to think of it,” said Dr. Ralph Nixon of New York University Langone Medical Center. “Lifestyle and dietary changes could be implemented at a very early stage. Perhaps there are nutritional factors that could alter these pathogenic mechanisms in a low-level way and delay onset.
“Obviously, that is not going to be enough for many people, though,” Dr. Nixon cautioned. “At some point, we have to up the level of treatment and either be more aggressive or add additional modalities. This is when you get into the idea of multidimensional treatment.”
This “cocktail” approach will be the focus of the next generation of clinical trials, Dr. Sperling said.
“We need to be preparing ourselves for combination therapy trials,” he said. “While I’m hopeful that we will see benefits to starting anti-amyloid treatment really early, it’s possible that in prodromal Alzheimer’s, there is already a lot of tau that we will also need to address. We need to start talking about how to go after different mechanisms, incorporating tau therapies, and maybe anti-inflammatory and neuroprotective agents.
“If we are going to have an effective therapy by 2025, as we’ve committed to, then we need to get these combination studies off the ground in the next 5 years,” Dr. Nixon noted.
She herself feels confident that research is on the cusp of something big.
“I believe that we will have an effective disease-modifying treatment 5-10 years from now. I think it is likely to be an anti-amyloid treatment within 5 years, and my hope is that by 10 years, we will have combination therapies, both anti-amyloid and anti-tau.”
How effective is ‘effective’?
The new prevention studies are founded on these best-case-scenario cognitive benefits of around 35% over placebo. This doesn’t mean, however, that patients are experiencing a 35% cognitive boost over baseline. The difference is described as delay in progression, not a halt in progression – and certainly not recovering. The results generally amount to a 1- to 2-point difference on the Alzheimer’s Disease Assessment Scale cognitive subscale (ADAS-cog).
“This is what we’re going for?” asked Dr. Lon Schneider, director of the California Alzheimers Disease Centers. “A 37% difference comes out to a decline of 4.5 points rather than 6 points. Maybe an effect size of 4 or 5 points would be worthwhile. But 1.5 points?
“In my opinion, this is nothing clinically – but now people think it is meaningful because it’s a ‘delay in progression.’ Isn’t that great? But this is what we work with,” Dr. Schneider explained. “If we stand back and take a really honest look at this, we can say there is not much faith in treating amyloid beyond making a very small impact on Alzheimer’s.”
Dr. Richard Caselli, a neurologist at the Mayo Clinic Arizona in Scottsdale, agreed.
“A 30% ‘benefit’ is certainly better than nothing, but slowing disease progression is a far cry from a cure,” Dr. Caselli said.
What would such a difference look like clinically? “Probably nothing, unless there was some serious adverse event,” Dr. Caselli noted. “Slowing progression means the disease has not stopped progressing, so everything we would have seen we would still see.” There is already a tremendous variation in the pace of cognitive decline, he added, so while some families might notice something, others would not.
He sometimes wonders if a really effective symptomatic treatment wouldn’t offer patients a more practical benefit.
“We hoped we would see this with the cholinesterase inhibitors and memantine, but we really don’t,” Dr. Caselli said. “We assume it’s helping, and occasionally families say patients seem brighter; but the differences are subtle. On the other hand, if a new anti-agitation med were effective, that would be obvious and helpful to everyone.”
Perhaps the only thing everyone in Alzheimer’s research seems to agree upon is that the more we learn, the more we know is left to learn.
“This is a complex disease, and we are just learning how complex it is,” Dr. Schneider said. “It has multiple targets and a complex metabolism. It’s not just amyloid – it’s not a single destination.”
He likened Alzheimer’s to diabetes – another complex disease with a number of pathologic pathways.
“Diabetes used to be considered an insulin deficiency disease – the illness where you gave insulin and everything got better,” Dr. Schneider observed. “Insulin is hugely helpful, but it’s by no means a cure, which is what we’re looking for in Alzheimer’s.
“Will one target – amyloid – really make a difference in something this complex? I guess it’s possible,” he said. “But we already know it won’t be as effective as insulin is in diabetes. We need to look into other pathways as well.”