ARTICLE IN BRIEF
A Medicare panel voted that there was not strong enough evidence to show that florbetapir F18 (Amyvid), a radioactive tracing agent that binds to amyloid-beta brain aggregates on PET scans, would improve health outcomes in suspected cases of early Alzheimer's disease. They were concerned about the potential for false-positives and the dearth of data in the clinical setting.
A Medicare advisory panel on Jan. 30 concluded that there is insufficient evidence that positron emission tomography (PET) using a new radioimaging contrast agent that highlights amyloid-beta deposits improves health outcomes in suspected cases of early Alzheimer disease (AD).
Members of the Medicare Evidence Development & Coverage Advisory Committee (MEDCAC) ranked the available evidence at just 2.17 out of 5 confidence points. An average of 2.5 points was needed for further consideration of other questions on the panel's agenda regarding PET and AD. [A score of 1 equals low confidence; 3, intermediate confidence; and 5, high confidence.]
MEDCAC reviews data for the Centers for Medicare & Medicaid Service (CMS), which oversees Medicare reimbursement. CMS is expected to vote on coverage in July, and while MEDCAC's reviews inform the agency's coverage decisions, CMS is not required to follow the panel's recommendations.
At issue is Eli Lilly's florbetapir F18 (Amyvid), a radioactive tracing agent that binds to amyloid-beta brain aggregates on PET scans. The US Food and Drug Administration approved florbetapir last April for differentiating other forms of mild cognitive impairment from AD, and it remains the only radioimaging agent approved for neurocognitive PET scans.
MEDCAC Chair Rita F. Redberg, MD, professor of medicine at the University of California, San Francisco School of Medicine, emphasized that the panel made no recommendation about PET reimbursement.
“MEDCAC is [only] charged with reviewing evidence in order to advise CMS on the issues posed in the voting questions,” she told Neurology Today.
“After a careful review of all of the literature and an all-day discussion of the issues and hearing from the guest speakers, the committee had low to intermediate confidence that there is adequate evidence that this scan can lead to better outcomes for patients.”
Such scans cost between $1,000 and $6,000, but average around $3,000, according to the Alzheimer's Association, and are rarely covered by insurers. The only such scans currently eligible for Medicare reimbursement are those in patients suspected of having frontotemporal dementia.
“We were not sure if there is enough, or strong enough, evidence to be able to weigh the overall health benefits — the balance of potential benefits versus harms — of this particular technology,” said panel vice chair Art Sedrakyan, MD, PhD, associate professor and director of the Patient-Centered Comparative Effectiveness Program at Weill Cornell Medical College of Cornell University.
“The value of a negative test or the ‘value of knowing’ needs more research and documentation,” he told Neurology Today. “We are also concerned about the higher chance of false-positive results particularly in the elderly.”
In a December 2012 white paper, the Institute for Clinical and Economic Review (ICER), based at the Institute for Technology Assessment at Massachusetts General Hospital in Boston, evaluated the diagnostic value and clinical benefits of testing for AD, including PET.
ICER President Steven Pearson, MD, presented an overview of the findings to the panel. He said the review found that few studies examined whether biomarker use actually improves AD diagnosis, and none explored whether or not patient outcomes improve as a result, especially long-term outcomes.
“I think the average MEDCAC vote represents a reasonable interpretation of the current state of the evidence, with the fact that studies have not yet been done to evaluate the impact of using the test results on actual clinical decision-making or patient outcomes,” he told Neurology Today.
“My opinion on needed future studies mirrors the recommendations in the white paper that we have put out. In the current era of treatments with only limited effectiveness, I think payers will require randomized clinical trials of diagnostic tests.”
He added that studies measuring the impact of PET on clinical decision-making might one day offer evidence that can be linked strongly enough to patient outcomes.
“But again, with the current therapeutic armamentarium it will be challenging to find changes in clinical decisions that are substantive enough to provide confidence that they lead to significant changes in patient outcomes. All these considerations will transform overnight if a trial using PET amyloid — or other biomarkers — to enroll patients shows a substantial clinical benefit.”
“This is disturbing for many of us involved with AD neuroimaging research,” said Michael W. Weiner, MD, professor of medicine, radiology, and psychiatry at the University of California, San Francisco, and principal investigator for the Alzheimer's Disease Neuroimaging Initiative (ADNI).
“Research shows that PET scan images closely resemble amyloid deposits observed after brain autopsies, and such information can be used to differentiate early AD from other forms of cognitive impairment,” he told Neurology Today.
This can be very beneficial for individuals and their families in terms of helping them plan for the future, he said.
Rather than basing its recommendation on whether or not PET data can lead to better patient health outcomes, the pivotal issue should be whether or not they help patients and their families get answers about neurocognitive symptoms, according to Dr. Weiner.
“Patients and their families just want to know what is going on; they want to know if they have Alzheimer's disease.”
Moreover, the benchmark rationale of basing coverage decisions on available treatments or health benefits is confusing when Medicare pays for tests for other disorders for which treatment is unavailable.
“I am not sure why Medicare would pay for one but not the other, but if this is the new [threshold] for Medicare reimbursement, why pay for anything when there is no inherent transfer to improved health outcomes?” said Dr. Weiner.
The MEDCAC vote came just days after the Society of Nuclear Medicine and Molecular Imaging and the Alzheimer's Association jointly published the first criteria for the appropriate use of the amyloid imaging in AD patients.
The guidelines were published Jan. 28 in an online advance publication in Alzheimer's & Dementia: The Journal of the Alzheimer's Association and The Journal of Nuclear Medicine.
The Society of Nuclear Medicine and Molecular Imaging said in a statement that they believe the evidence “supports the ability of beta amyloid imaging to change patient management, leading to better outcomes for patients.”
The Alzheimer's Association recommended that CMS cover PET according to the new guidelines, and said in a statement that it was “disappointed in the committee's [Jan. 30] decision,” and supports “associated insurance coverage to clarify unclear diagnoses in individuals experiencing memory and thinking problems.”
According to the guidelines, people who should be considered for a scan include those with persistent or progressive unexplained memory problems or confusion and those with problems on standard neurocognitive tests; those with possible AD but whose symptoms don't present in the usual way; and people with progressive dementia before age 65.
Testing is not appropriate for people over 65 who are already known to have AD, under the guidelines, or those without any symptoms or medical evidence of the disease. Further, a finding of brain amyloid in itself does not constitute a diagnosis of Alzheimer's dementia, the guidelines emphasize.
‘A TALL ORDER’
An entirely new coverage strategy for florbetapir amyloid PET scans may be necessary because it may be impossible to demonstrate that intervening therapeutically at the age covered by Medicare will lead to a clinical benefit, according to Samuel E. Gandy, MD, PhD, director of the Mount Sinai Center for Cognitive Health and the Mount Sinai Chair in Alzheimer's Disease Research.
AD has been redefined as a middle-age disease but one in which clinical manifestations usually do not appear before age 65, he said, but if amyloid deposits begin between age 40 and 50, those are the people that will benefit most from therapeutic intervention.
“But there is a problem here. The health care payors who take care of us in midlife are not focused on what happens after age 65, yet if we are to improve outcome, we must persuade third party payors to cover these [scans] for people who will age out of the window covered by those payors before the disease becomes clinically evident. Because of the age-specific insurer system in the US, one can envision a scenario in which neither third parties nor Medicare will pay for amyloid scans using the tracing agent,” he said.
Another issue that did not come up is whether any payor, CMS or private third party, will accept divergence of biomarker slope, or trajectory, as a therapeutic benefit of treatment following a presymptomatic scan.
“I would speculate that the CMS panel would not, and would insist on a meaningful clinical benefit — a tall order.”