Q&A: Alzheimer's trial disconnect

While preclinical studies identify ways to prevent Alzheimer's disease in animals, human trials test these same therapies in symptomatic patients -- long after they are most likely to be effective

Jan 26, 2011
Jef Akst
Alzheimer's disease (AD) is a growing threat that currently afflicts some 35 million people worldwide. Without the advent of preventive therapies, the neurodegenerative disease will strike as many as 100 million people by 2050. And while laboratory studies in animal models of AD continue to uncover promising avenues for disease prevention, clinical trials in humans target patients who are already showing signs of neural degeneration.
Image: Wikimedia commons, Alzheimer Forschung Initiative e.V.
Disease biologist linkurl:Todd Golde;http://www.neuroscience.ufl.edu/faculty+research/golde.html of University of Florida College of Medicine talked to The Scientist about this disconnect, its consequences, and possible solutions to the problem -- the topic of an opinion piece he co-authored, published online today (January 26) in Neuron.The Scientist: What are the current treatments for Alzheimer's disease? Todd Golde: The two [classes of drugs] that have been approved are designed to enhance cognitive function. One is the cholinesterase inhibitors, and the other one is memantine. [Both yield] a transient increase in cognitive function, but it doesn't last. And there's no evidence that the course of the disease is changed or altered. It's really a fairly modest effect. So there's a real unmet medical need.TS: What is the problem with how these treatments are being tested in humans?TG: Over the last two decades, we've gained a fairly good understanding of certain steps that probably play an initiating role in Alzheimer's disease pathogenesis. I think the data is quite compelling that the accumulation of [amyloid-beta] peptides and aggregates drives this disease. But there's a misalignment between what we do in the laboratory, where we're largely doing a prevention study against that pathological feature, and in the clinic, where we're actually treating a symptomatic individual who has longstanding pathology and massive signs of neural degeneration. Why would removing the trigger of the disease at that point really have a major benefit? TS: What are the possible consequences of this mistake?TG: The biggest one is that it makes it much less likely that the therapies are going to show any signs of efficacy. Basically, they won't get approved. And if it costs $300-$600 million to conduct two pivotal phase III trials for an Alzheimer's disease drug, it's not going to be too long [before] people aren't going to invest. And now you want a prevention trial that's going to cost significantly more and is also going to take longer...so long that the patent is essentially expired by the time the drug gets approved.Let's say you do have a disease modifying drug that slows the course of cognitive decline, but only does so in patients who have fairly advanced Alzheimer's disease. Does an individual really want that if it doesn't improve their function -- if it just slows the course and they're pretty far gone? That's a tough ethical question. If [a drug] is only partially effective, are you just prolonging suffering rather than actually improving it?TS: What are some things that need to be addressed to solve this problem?TG: First and foremost, when we consider prevention, a drug or a therapy has to be "safe enough." And when you're talking about treating people who are asymptomatic or "well," that bar is pretty high. I think that's the first issue from the medical and scientific side.The second one is if you want to embark on realistic trials where you could actually execute them in a reasonable amount of time, it will be accepting biomarkers as surrogate endpoints for the disease. And that is not without risks. At this point I think we've make incredible progress [identifying biomarkers for] Alzheimer's disease pathology, but we don't have enough longitudinal data to say how long it will take any individual to progress on to additional neurodegeneration and cognitive impairment. Then linked tightly to [these medical or scientific issues] are the financial obstacles of running prevention trials. I'm a scientist, I'm not a politician, I'm not a policy maker, I'm not a drug manufacturer. But I think these kinds of issues need to be spoken about in those types of settings if we really want to move toward a country that is serious about preventing illness.TS: Do you think this prevention-treatment dilemma exists in drug research for other diseases? Is this a common misalignment between preclinical studies and human trials?TG: For degenerative processes, I think for sure. We largely use genetics to help us identify the root cause, we find some therapies to target that, usually most effective in our animals models in a preventive setting. Then the trials are always done on people [with] pretty advanced disease, and they again fail to translate. When we think about this idea of personalized genomic prediction of disease, if we do this testing and we want to realize the promise of it, then we have to think about therapies in the setting of prevention, not in terms of treatment.T.E. Golde, et al, "Anti-Ab therapeutics in Alzheimer's disease: the need for a paradigm shift," Neuron, 69:203-13, 2011.
**__Related stories:__***linkurl:Alzheimer's drugs hurt brain?;http://www.the-scientist.com/blog/display/57251/
[22nd March 2010]*linkurl:Alzheimer's clue found;http://www.the-scientist.com/blog/display/55868/
[29th July 2009]*linkurl:New gene in Alzheimer's disease;http://www.the-scientist.com/news/display/41520/
[15th January 2007]