As Alzheimer’s drugs keep failing clinically, the consensus among researchers in the field is that for the current approach – which is focused on amyloid beta – to work, treatments will have to start earlier.

That comes with an unwelcome corollary, though: the trials to test such drugs are long and getting longer.

The treatment duration in the trial that led to the approval of Aricept (donepezil) was six months. Now, the standard trial testing drugs even in mild cognitive impairment is 12-18 months.

And the prevention trials that currently have the highest hopes riding on them are even longer still. The largest of those studies, the anti-amyloid in asymptomatic Alzheimer’s disease (A4) study, will treat patients with solanezumab (Eli Lilly & Co.) for three years.

Solanezumab missed both the cognitive and functional endpoints in two double-blind, placebo-controlled trials in patients with mild-to-moderate disease in 2012. (See BioWorld Today, Aug. 27, 2012.)

But posthoc analysis of those trials suggested that the drug was having an effect in mild AD, leading to Lilly’s decision to test the antibody both preventively and at earlier stages of AD.

SHORTER TRIALS POSSIBLE

Not every anti-Alzheimer’s drug would necessarily need to be tested for years. Approaches targeting other facets of the disease, such as tau tangles or brain inflammation, might be able to prove their worth in shorter trials.

But given that more than 10 phase III trials have failed to show an effect of anti-amyloid agents in advanced stages of AD, it is now pretty clear that if this approach can be made to work at all, it will work early in the disease.

Two other prevention trials are testing the amyloid hypothesis in patients with early onset forms of AD. The Dominantly Inherited Alzheimer Network (DIAN) trial will treat patients with solanezumab or gantenerumab (Genentech Inc./Roche AG) for two years. And the Alzheimer’s Prevention Initiative trial is testing crenezumab (Genentech Inc./Roche AG) for up to five years.

Even those trial durations are not really particularly long compared to the likely true duration of the physiological changes that ultimately lead to Alzheimer’s disease. The earliest changes in the brain start a decade or more before memory loss becomes apparent.

Still, long trials that are getting longer are not anyone’s first choice for how to run clinical trials to say the least.

In the case of AD, the saving grace for such trials is the sheer size of the problem, in terms of the patient population that could benefit from a disease-modifying drug, versus the cost of doing nothing.

DAUNTING ECONOMICS

The economics of such trials, Eric Siemers acknowledged, “are a bit daunting at times.” But “if it turns out that treatment delays dementia, that becomes a public health issue.”

Siemers is the medical director of Lilly’s Alzheimer’s disease team, which is part of the public-private partnerships testing solanezumab in the A4 and DIAN prevention trials.

Because AD is a looming public health crisis, the costs of the A4 trial are being borne in part by the National Institutes on Aging, by philanthropical organizations, and by Eli Lilly, which is supplying the antibody as well as some grant support and the time of those who work on the project.

Gad Marshall, who is the associate medical director of clinical trials at the Brigham and Women’s Hospital Center for Alzheimer’s Research and Treatment, noted that such trials are expensive, but if they work, they are also cost-effective: if it becomes possible to treat Alzheimer’s disease, he told BioWorld Insight, “we’re talking about saving billions at the expense of millions” in trial costs.

Still, first it’s necessary to come up with those millions. And longer trials also mean larger trials, further adding to their cost. “The longer the trial, the more likely you are to lose people,” Marshall said, especially in a disease like Alzheimer’s, where some number of individuals will develop age-related issues, both related and unrelated to AD. Altogether, long prevention trials could end up totaling $100 million dollars.

On the other hand, the size of trials could be reduced through enriching patient populations and using adaptive trial design.

Reducing the number of patients necessary to see a signal will depend on being able to select patients at high risk of progressing to cognitive impairment. This is the principle behind the prevention trials being conducted in high-risk populations with early onset mutations.

In trials that are not limited to those with specific mutations, biomarkers can identify individuals at high risk. The A4 trial includes imaging to make sure that trial participants have amyloid pathology in their brain – though this, too, adds to the initial cost of the trial.

BIOMARKERS PREDICT 90%

A paper published earlier this month reported that a panel of 10 blood biomarkers could predict with 90 percent accuracy which cognitively normal individuals would develop memory problems within three years. (See BioWorld Today, March 9, 2014.)

Both the accuracy and the ease of such a test could help enrich clinical trial populations and improve the chance of detecting any therapeutic efficacy, similarly to selecting patients with specific mutations in clinical trials of targeted cancer agents.

Judith Quinlan, senior vice president of adaptive trial design and implementation, at contract research organization Aptiv Solutions Inc., told BioWorld Insight that such biomarkers can also be used to run adaptive trials. Interim analyses of biomarkers could be used, for example, to “turn off certain doses within a trial” if those doses are not showing an effect on biomarkers.

Such an approach can allow investigators to combine trial phases that are traditionally separate, which speeds up overall clinical development because it eliminates the time gaps between the end of a phase I and the beginning of a phase II trial, and so by using adaptive design, “I can perhaps cross the finish line earlier” overall even if the early phases of the trial are larger.

Good predictive biomarkers could in principle also be used as surrogate endpoints. But so far, such biomarkers are lacking. While the FDA is open to adaptive trial design, the agency wrote in draft guidance documents released in 2013 that “until there is widespread evidence-based agreement in the research community that an effect on a particular biomarker is reasonably likely to predict clinical benefit, we will not be in a position to consider an approval based on the use of a biomarker as a surrogate outcome measure in AD (at any stage of the illness).”