The Food and Drug Administration plans to loosen the rules for approving new treatments for Alzheimer's disease.
Drugs in clinical trial would qualify for approval if people at very early stages of the disease subtly improved their performance on memory or reasoning tests, even before they developed any obvious impairments. Companies would not have to show that the drugs improved daily, real-world functioning.
For more than a decade, the only way to get Alzheimer's drugs to market was with studies showing that they improved a patient's ability not only to think and remember but also to function day to day at activities like feeding, dressing or bathing themselves.
The proposal, published online Wednesday in the New England Journal of Medicine, could help millions of people at risk of developing the disease by speeding the development and approval of drugs that might slow or prevent it.
The proposed policy could also be a boon for the pharmaceutical industry and researchers. They have often felt stymied by regulations that left them uncertain of how to get drugs tested and approved for marketing to people early in the course of Alzheimer's, when the medications are most likely to be useful.
Several studies are being planned for people at high risk of developing Alzheimer's, and the proposed regulations should lead to even more clinical trials, said Dr. P. Murali Doraiswamy, an Alzheimer's researcher and professor of psychiatry at Duke University School of Medicine.
"There is more motivation now to invest in the field," Doraiswamy said. But, he added, the proposal also comes with risks.
The FDA would require companies to study the drugs after they are on the market to show they actually benefit patients. But these studies might not be randomized clinical trials and so would not be as rigorous as the studies that led to approval.
Even with less rigorous studies, it might be hard to get the evidence, Doraiswamy said.
"A lot of companies drag their feet and never do a post-marketing study," he said. Even if they do one, it may take years to find out whether a drug works. Meanwhile, millions of people may be taking it, at great expense.
The agency is trying to resolve a quandary with its new policy: How can you show a drug works if patients are so early in the course of the disease that they have no obvious deterioration in their cognitive abilities or daily functioning?
In draft guidelines to the industry last month, the agency outlined a way to resolve the problem by changing the criteria for drug approval. Those at a very early stage of the disease, with no obvious symptoms, could be studied with cognitive tests.
Those with mild symptoms could have a combined test of function and cognition.
With the draft guidelines, the agency is soliciting comments from the public and industry and may make revisions, said a spokeswoman, Sandy Walsh. But, she added, "we are ready to move forward with the plan."
In the paper published Wednesday, the agency aimed to explain its perspective to a wider audience.
"Our goal was to provide as much regulatory clarity as possible," said Dr. Nicholas Kozauer, a clinical team leader in the agency's division of neurology products. "We would encourage companies to start thinking along these lines."
Companies and researchers were elated, saying the old regulations had hobbled the field.
"This is really a huge advance," said Dr. Eric Siemers, senior medical director for the Alzheimer's disease team at Eli Lilly & Co. "Kudos to the FDA."
Dr. Sean Bohen, senior vice president for early development at Genentech, praised the policy but said he worried about how to develop appropriate tests to identify subtle cognitive changes.
"There is a lot of uncertainty," Bohen said. "But it is an admirable effort. We have to start somewhere."
Even before the FDA issued its proposal, several companies and the National Institutes of Health have been planning studies of people who are very early in the course of the disease.
Dr. Paul S. Aisen of the University of California, San Diego, and Dr. Reisa Sperling of Harvard are leading a federally financed study that will enlist 70-year-olds who seem perfectly normal but are accumulating amyloid plaques in their brains, a sign that Alzheimer's could arise within about 15 years.
Of 1,000 people with plaques, half will get an experimental drug made by Lilly and the rest will get a placebo. The study will also include 500 70-year-olds without plaques, for comparison.
The researchers will assess the drug by giving subjects several tests, including ones that involve remembering lists of words and paragraphs and tests of an ability to follow instructions to change symbols to letters. Large studies indicate that people who are going to get Alzheimer's decline faster on these tests than those who will not get the disease. If the drug works, it may slow or stop that decline.
Later this year, Lilly will start a new test of a drug in people with mild symptoms. The drug was already tested in a mixed group of patients, some with mild symptoms and some with more advanced disease. It failed.
But the company gave its raw data to academic researchers who discovered that if they looked only at memory tests and only at patients with mild disease, it seemed to benefit them. So the company will redo the study with just those who now seem most likely to be helped.
Genentech is planning a study of its own drug in people who have no symptoms whatsoever but have a gene mutation that virtually guarantees they will get Alzheimer's eventually. The hope is that the drug will prevent or delay it.
But if there are no symptoms, what is the best way to look for subtle cognitive changes? Bohen asked. The agency will not accept changes on brain scans alone as evidence of efficacy. Yet it could take years for symptoms to develop in those with the gene.
"It feels like we are taking a risk," Bohen said.
The company, he added, will be sending a comment to the FDA.