Neuroimaging Tracers for AD Detection Not Yet Ready for Prime Time, FDA Panel Advises
ARTICLE IN BRIEF
An FDA panel met with representatives of companies developing the next generation of neuroimaging tracers to assess their viability for future clinical trials and diagnosis of Alzheimer disease.
Six years after scientists successfully used a PET neuroimaging agent — Pittsburgh Compound B, or PIB — to detect amyloid beta (Abeta) in the brain of the first human patient, an FDA advisory panel in October met to assess the viability and future use of the next generation of compounds for diagnosing Alzheimer disease (AD).
After meeting with representatives of three companies — GE Healthcare, Bayer Healthcare Pharmaceuticals, and Avid Radiopharmaceuticals — developing neuroimaging tracing agents that have a longer shelf life — two hours compared to 20 minutes for the PIB compound — the panel concluded that although the technologies do detect Abeta, they are not quite ready for prime time when it comes to AD diagnosis.
“The FDA wanted our input on the use of radioisotopes to detect the presence of beta-amyloid in Alzheimer disease,” said Mark Green, MD, director of headache medicine at Columbia University College of Physicians and Surgeons and a member of the FDA advisory panel. “What we determined was that it is a test for beta-amyloid. The question is whether it is a test for Alzheimer disease.”
The panel was called to hear from the companies developing a flourine-18 substance similar to PIB, which picks up the presence of Abeta. Its longer shelf life means that doctors can use it clinically and not just as a research tool.
The question that the advisory committee grappled with was how to test the radiolabeled substances to figure out what it tells doctors about an individual's condition. The problem with such tools is that studies have shown that even people with no signs of AD have Abeta deposition. But what the PIB studies have also helped scientists determine is that brains without Abeta deposition is a key factor in ruling out AD.
“We would have to be satisfied that they can document how useful these agents are. Sometimes it is good to have a test that rules out disease,” added Robert Temple, MD, director of the Office of Medical Policy of the FDA Center for Drug Evaluation and Research and acting director of the Office of Drug Evaluation. These companies “thought it would be valuable to say that this person does not have Alzheimer disease.”
He agreed with Columbia's Dr. Green. “How much more it can do remains to be seen,” he added.
DETERMINING A ‘STANDARD OF TRUTH’
“The purpose of the meeting was to determine if the detection of Abeta had clinical utility and decide on an acceptable standard of truth,” said William Klunk, MD, a professor of psychiatry and neurology at the University of Pittsburgh and co-developer of the PIB imaging agent along with his Pittsburgh colleague, Chet Mathis, PhD.
Although the short half-life of PIB — 20 minutes due to the carbon-11 radiolabel — makes it unsuitable as a commercial clinical tool, Drs. Klunk and Mathis have developed a derivative of PIB that has a longer half-life, about two hours. The University of Pittsburgh has licensed its compound to GE Healthcare. Dr. Klunk talked about the new f18 compounds and how to design clinical studies to test them. All three companies put forth an indication of “amyloid detection” as opposed to diagnosis of Alzheimer disease.
These Abeta imaging agents are used with PET and can pick up the presence of the sticky amyloid protein deposits in selected brain regions. Based on thousands of research PET PIB scans performed around the world (including those performed as part of the Alzheimer's Disease Neuroimaging Initiative) and several recent correlations between in vivo PIB PET data and postmortem histological analysis, it is widely accepted that PET imaging with tracers such as those proposed to the FDA could detect the presence of Abeta in the brain.
These imaging tests will not, at present, be used to diagnose AD. The reason: several investigators have shown that anywhere between 25 to 45 percent of healthy subjects have evidence of amyloid deposits on PIB scans. No one is quite sure of the implications — for example, are these preclinical signs of AD? — but the presence of amyloid in healthy subjects means that using these imaging agents to detect beta-amyloid would not be specific for AD diagnosis.
The meeting focused on how best to test the new longer-acting compounds that must undergo the standard studies to gain approval as an FDA drug. Scientists at Avid said that it would be important to test the compounds against a gold standard, and the only one for AD remains the autopsy, a snapshot of the brain that shows the presence or absence of plaques and tangles.
NEED FOR MORE POSTMORTEM STUDIES
Daniel Skovronsky, MD, PhD, chief executive officer at Avid Radiopharmaceuticals, proposed several possible strategies that may make the postmortem validation approach practical, including enrolling patients that are near the end of their life for reasons other than AD — such as hospice patients — and those with AD who have already consented to have their brains used postmortem in research studies.
But how many brain autopsy studies are necessary? And how extensive should the histological evaluations be? For example, even in typical AD cases, there are amyloid-positive and amyloid-negative brain areas. What histological stain or combination of stains will need to be employed? Dr. Klunk said that there is now evidence that there may be different forms of Abeta deposits in the brain and that these imaging agents may not detect all of them.
Despite the lingering questions, everyone agreed that this meeting had a positive outcome for the field of amyloid imaging in general. “It was a nice step forward,” said Dr. Klunk. “At the end of the day, everyone agreed that there is clinical utility in detecting amyloid in the brain.”