This 2019 STAT article describes the systematic bias of the Alzheimer’s research field toward the amyloid hypothesis, a hypothesis that has failed to yield a single treatment or cure despite decades of research and the investment of billions of dollars.
Alzheimer’s disease was first discovered in 1906, when 51-year-old Auguste Deter died with what was then called “presenile dementia.” Deter’s brain was found to contain sticky protein plaques and fibrous tangles; in 1984, the former would be identified as beta-amyloid. Soon after the discovery of beta-amyloid, multiple genes were found that increased both beta-amyloid accumulation and Alzheimer’s risk, establishing the initial hypothesis that amyloid plaques were the cause of Alzheimer’s disease.
The amyloid hypothesis has been one of the most tragic stories [in] disease research.
Rachael Neve of Massachusetts General Hospital
In 1995, Athena Neurosciences created the first mouse model of Alzheimer’s disease. The mice were genetically manipulated to naturally develop amyloid plaque buildup in their brains as they aged and experience memory loss and cognitive impairment. Early tests showed clearing these amyloid buildups or preventing their accumulation reduced rates of cognitive decline. This was interpreted as further evidence that amyloid plaques played a causal role in Alzheimer’s development.
Yet even at this early stage, the research did not consistently support the amyloid hypothesis. Observational research showed little correlation between the degree of amyloid accumulation in the brains of the elderly when autopsied and their degree of cognitive impairment; individuals with large amounts of amyloid buildup also displayed no cognitive impairment, particularly within the hippocampus and frontal cortex. Conversely, the degree of synapse death correlated well with cognitive decline. In addition, the mouse models were suspected to be a poor model for the majority of Alzheimer’s cases in humans (1).
Despite the questions that arose, the amyloid hypothesis — that is, the belief that the direct cause of Alzheimer’s disease was accumulation of amyloid deposits in the brain and that clearing these amyloid deposits or preventing their initial deposition would prevent Alzheimer’s disease — quickly cemented into dogma.
The National Institutes of Health (NIH), pharmaceutical companies, and private foundations funding Alzheimer’s research focused the majority of their efforts on studies and potential treatments centered on the amyloid hypothesis. Research rooted in any alternative hypothesis struggled to find support. For example, when Ruth Itzhaki tried to publish research suggesting herpes simplex virus may play a role in Alzheimer’s development, she was rejected from 12 journals before her paper was accepted and had to campaign for years to get a single conference speaking slot. The Alzheimer’s Association’s Zenith Fellowship, designed specifically to support unconventional Alzheimer’s research, rejected her proposals, describing her hypothesis as without evidentiary basis. Subsequent research would show herpes simplex virus more than doubled risk of dementia and antiviral drugs substantially attenuate this risk increase. Other researchers studying the effects of oxidation, inflammation, blood flow, metabolism, and other factors in Alzheimer’s disease have faced similar issues, consistently finding it difficult if not impossible to secure support for their work. Many have been told explicitly that funders (private and public) are not interested in their work if it does not target amyloid.
If it weren’t for the near-total dominance of the idea that amyloid is the only appropriate drug target, we would be 10 or 15 years ahead of where we are now.
Dr. Daniel Alkon, NIH
Pharmaceutical companies exclusively funded potential Alzheimer’s treatments designed to prevent, reduce, or reverse amyloid-beta accumulation. As described in additional detail in the articles below, these trials have repeatedly and conclusively failed to deliver any benefit to patients. Trials led by Roche, Merck, Eli Lilly, and Biogen have all developed and attempted to bring to market drugs that clear or attempt to clear amyloid from the brain. Even when these drugs have successfully cleared amyloid, they have failed to slow rates of cognitive decline. In response to these failures, pharmaceutical companies have argued not that the amyloid hypothesis is incorrect but instead that we simply need to clear amyloid earlier or use stronger doses of the drugs. These failures have been so consistent and clear that pharma companies have all but given up on patients with moderate Alzheimer’s disease, arguing that once symptoms are visible, the damage is irreversible. Thirty years of Alzheimer’s research has thus yielded neither an effective treatment for the disease nor provided patients with any hope that one may be coming.
Recent research has suggested mild to moderate Alzheimer’s is reversible in at least some patients, and as described previously on CrossFit.com, treatments targeting disease factors other than amyloid show promise. Major pharma companies, however, continue to focus their efforts on amyloid, despite major failures as recent as March 2019.
Takeaway: The Alzheimer’s research community provides an extreme example of the extent to which financial and intellectual biases can corrupt a field. The amyloid hypothesis, while compelling in its genesis, was maintained despite repeated demonstrations that it failed to direct us toward effective treatments in human subjects. Each failure led researchers not to question the core hypothesis but instead to move the goal posts. Bias permeated pharmaceutical companies, public funding sources, and private foundations to such an extent that those who attempted to study Alzheimer’s through any other lens found their work unsupported or even suppressed. The result has been a profound failure to develop any treatment to prevent, slow, or reverse a disease that affects millions of Americans.
Patients in Biogen’s aducanumab trials, due to disclosure laws, found out about the trial’s failure in the same public announcement that sent the stock price crashing. While patients said they had realistic expectations about the trial given the dismal history of Alzheimer’s research, many remained devastated that another potential treatment had been found to be ineffective. Patients were told by clinicians that previous tests of the amyloid hypothesis had simply treated patients too late, after the damage was done, and that aducanumab would prevent disease progression where others had failed. With the trial’s failure and a lack of viable alternatives on the market, patients and caregivers instead turned their efforts toward living with Alzheimer’s and making the best of their remaining time before memory loss became severe.
Alzheimer’s and amyloid: “It’s time to do something else” (March 2019)
Over 300 experimental Alzheimer’s drugs have failed. The majority of these drugs have, in some form, been rooted in the amyloid hypothesis that reversing or preventing the deposition of amyloid plaques will reverse or prevent Alzheimer’s. Since 2016, Roche, Merck, Eli Lilly, and Biogen have all seen major trials of amyloid-clearing drugs fail, and yet researchers continue to find reasons to keep the hypothesis alive. Multiple researchers argue that once symptoms have appeared, the disease is too late to reverse; the amyloid burden is already sufficient to cause neuronal death, and reversing it does not reverse the damage. Some, citing the well-known observation that some individuals die with brains full of amyloid but without cognitive impairment, have begun to search for mechanisms that help the brain resist amyloid-based damage. Others, however, have begun to more seriously consider, for example, that pathogens (viruses or bacteria) cause Alzheimer’s; or that impaired blood flow, inflammation, or an immune system response could trigger disease progression. Even these hypotheses, however, are often framed in terms of their impact on amyloid accumulation and toxicity.
Author Sharon Begley writes, “It’s starting to feel like a very sad real-world version of the old Monty Python sketch, where no matter how many limbs the dark knight has lopped off by his sword-wielding foe, he declares it’s only a flesh wound.”
In March 2019, Biogen and Eisai announced they were halting two Phase III trials of aducanumab after an interim analysis revealed a clear lack of benefit. The aducanumab trials were designed to specifically clear amyloid before it formed into clumps in patients who had only very early signs of the disease. This was thought to be a group considered more likely to benefit. The failure of this trial not only represents a disappointment for patients but provides the amyloid hypothesis with few potential avenues of further study.
No Alzheimer’s drug has been approved since 2003, and the consistent and complete failure of drugs targeting the amyloid hypothesis has convinced some in the research community not that we have the wrong hypothesis, but that Alzheimer’s is a disease we are powerless to cure and may at best be able to prevent or slow in very mild cases. At the time of this article’s publication, 112 therapies were being tested across 135 trials; however, 9 out of 26 agents in Phase III trials make no attempt to treat the disease or slow progression and merely aim to treat symptoms. More importantly, only 26 of the 135 trials accept patients with moderate dementia, and only 12 accept patients with severe dementia, which reflects the prevailing belief in the field (rooted in data from failed clinical trials) that amyloid-based approaches are entirely ineffective in patients with moderate or severe Alzheimer’s. This, however, means there is little hope of treatment for these more severe cases, which make up an unknown but likely meaningful share of the 5.5 million Americans with Alzheimer’s. The pharmaceutical industry has all but given up on them. One investor wrote on Twitter, “There is no proof that you can reverse the damage in later-stage patients, so I understand why no one is focused on this.” (The tweet was later deleted.)
In July 2018, Biogen posted what seemed to be positive results for the drug BAN2401, which significantly reduced brain amyloid levels and cognitive decline compared to placebo. Upon closer inspection, however, the results were less clear. There was no dose-response effect. The amount of amyloid clearance was proportional to the dose of drug received, but only those on the highest dose saw any reduction in cognitive decline; those on the lowest doses declined more quickly than placebo. Additionally, only 30% of patients in the high-dose group had the APOE4 mutation that triples Alzheimer’s risk, compared to 70% of the placebo group, a difference that in itself could explain the observed effect. The study’s authors argue these results reflect a threshold effect and that a certain amount of amyloid clearance is required to show clinical benefit. A Phase III clinical trial of BAN2401, Clarity AD, was announced in March 2019. The article notes Lilly is continuing to fund its own multi-drug approaches to maximally suppress amyloid as a potential Alzheimer’s treatment.
What can we learn from the latest Alzheimer’s drug failure? (January 2018)
In January 2018, trial data conclusively demonstrated Eli Lilly’s Alzheimer’s drug solanezumab lacks clinical effectiveness. Solanezumab is an antibody designed to clear amyloid from the brain before plaques form, and — if plaque formation is a cause of Alzheimer’s — to prevent or reverse disease progression. Two Lilly trials testing the drug failed in 2012, with this third trial designed to test the drug specifically in patients with earlier stages of the disease in whom it was thought amyloid clearance would be more clinically meaningful; Lilly argued the damage was already done in patients with moderate Alzheimer’s. This trial, however, showed no significant effect, with 908 patients who took the drug for 80 weeks showing similar declines in cognitive performance over that period to placebo.
This is another trial demonstrating drugs that remove amyloid from the brain or prevent its accumulation fail to reverse or prevent Alzheimer’s. Nevertheless, the researchers seem unconvinced. Dr. Sam Gandy (Icahn School of Medicine, Mount Sinai) argues these early-stage patients are still being treated too late and that still-earlier treatment, prior to any symptoms, is required. However, in practice, treating patients earlier would require massive screening programs and prescription of amyloid-lowering medication to tens of millions of Americans, which would be similar to how statins are prescribed today. Dr. Lawrence Honig (Columbia), one of the trial’s authors, instead argues, “I and most Alzheimer’s researchers still believe amyloid is a valid target, but we haven’t managed to engage it adequately.” In other words, more severe forms of amyloid control are required. Lilly, consequently, has announced a trial (the Anti-Amyloid Treatment in Asymptomatic Alzheimer’s Disease, or “A4”) testing a four-times-higher dose of the drug in patients without even early-stage cognitive impairment.
Reflecting on the mid-2016 failure of yet another Alzheimer’s drug, researchers question the validity of the central hypotheses guiding Alzheimer’s research. The hypothesis that amyloid plaques form in the brain and lead to neuronal death was first proposed in the 1980s (Glenner 1984); the closely related hypothesis that “neurofibrillary tangles” of tau are similarly toxic was proposed in 1986. As early as 1991, however, clinical data began to undermine these hypotheses. Specifically, the degree of amyloid or tau buildup in the brain did not seem to correlate with disease severity — only synapse loss did. In 2008, the drug Abeta42 was found to successfully clear plaques from the brains of patients with Alzheimer’s but failed to drive cognitive improvement. The field responded not by dismissing the causal significance of amyloid beta but instead by arguing that once plaques had formed, the damage was done and earlier intervention was required. In subsequent years, support for research targeting any aspect of the disease other than amyloid clearance has been minimal. These problems are exacerbated by a reliance on animal models that fail to mimic the course of the disease in humans. Animal model studies have led to numerous drugs that resolve elements of disease pathogenesis in rats and mice (i.e., amyloid clearance) but fail to have clinical benefit in humans. The result is that despite decades of research and billions of dollars in investment, patients can turn to only four drugs, all of which do nothing to slow or reverse disease and can only temporarily improve symptoms. Any research unrelated to the amyloid hypothesis continues to be mostly driven by private foundations rather than public funding options such as the NIH.
- Alzheimer’s appears in two forms. Early Alzheimer’s, which is believed to be mostly or entirely due to genetics, involves rapid accumulation of amyloid in the brain at an early age. Auguste Deter had this form of Alzheimer’s disease. The later form of Alzheimer’s, which make up the vast majority of cases, is only partially influenced by genetics. Athena’s mouse model may be a good proxy for early Alzheimer’s, but as the linked articles below show, it and similar models have consistently failed to predict which drugs will and will not be effective at improving cognitive decline in humans with late Alzheimer’s.