Insurers have spoken: brain amyloid β PET imaging is not ready for clinical prime time. On July 3, the Centers for Medicare and Medicaid Services (CMS) issued a draft decision denying coverage for any clinical use of the technology. The decision met with strong criticism at the Alzheimer's Association International Conference (AAIC), which ran July 14-18 in Boston, Massachusetts. At a special public policy session held July 15, Alzheimer’s disease (AD) stakeholders lambasted the decision and strategized possible responses. Led by the association’s Maria Carrillo, this gathering saw some of the conference’s most passionate debate.
The session attracted a packed room of academic researchers, clinicians, and representatives from industry, imaging associations, and advocacy groups. Across the board, these stakeholders expressed frustration and disappointment with the CMS decision, decrying the loss of what they unanimously see as a valuable tool for patient care. The silver lining is that CMS will reimburse for scans in clinical trials that meet specific criteria; however, the overall decision will greatly limit access to amyloid scans in clinical care. An amyloid scan can cost $3,000 or more. Not only are most Alzheimer’s patients Medicare beneficiaries, but private insurers tend to follow CMS’ lead and are expected to deny reimbursement for this procedure, as well. “Having our hands cuffed behind us as we try to help patients and their families is a moral dilemma,” said Michael Devous of the University of Texas Southwestern Medical Center, Dallas.
The group also voiced their determination to take action. “This draft decision is a mistake and we should oppose it,” said Keith Johnson at Massachusetts General Hospital, Boston.
No insurance representatives spoke at the July 15 meeting. Instead, insurers aired their views at a separate July 18 AAIC session led by Steven Pearson, president of the Institute for Clinical and Economic Review (ICER). Based at MGH’s Institute for Technology Assessment, ICER, which receives unrestricted funds from several insurance companies, assesses the value of healthcare interventions across different medical fields. Pearson presented the findings from ICER’s white paper on the evidence standards needed for insurance coverage of AD diagnostic tests (see ARF related news story). Insurers will only cover tests that improve patient health outcomes, Pearson emphasized. A panel that included several insurance representatives agreed with this standard and discussed what evidence might be needed for amyloid scans. This session generated little crosstalk with AD researchers, as the audience was sparse and the panel took but a single question as time expired.
Insurers and Clinicians Clash Over Benefits of Amyloid Imaging
Amyloid β imaging became clinically available in 2012, when the Food and Drug Administration approved the first commercial Aβ PET ligand, Eli Lilly and Company’s Amyvid (see ARF related news story). To help guide its clinical use, the Alzheimer’s Association and the Society of Nuclear Medicine and Molecular Imaging convened a taskforce to develop appropriate criteria, which were published in January this year (see ARF related news story). AD stakeholders of every stripe have embraced the criteria. “I’ve never seen a coalition of common thought like we have in this setting,” noted Devous at the July 15 meeting.
In December 2012, Lilly petitioned CMS to extend coverage for amyloid β PET imaging agents (see ARF related news story). In response, CMS opened a National Coverage Analysis, which is the agency’s routine procedure for reconsidering coverage policies. In January 2013, CMS convened a meeting of the Medicare Evidence Development and Coverage Advisory Committee (MEDCAC) to examine the issue. MEDCAC voted that there was not enough evidence to determine if amyloid scans improved patient health outcomes (see ARF related news story), presaging the negative draft decision.
The sessions at AAIC highlighted the gulf between insurers’ and clinicians’ points of view. “The value of a diagnostic test depends on the availability of effective therapies,” said Robert McDonough, who heads clinical policy at Aetna, at the July 18 AAIC session. The MEDCAC panel had expressed similar views, noting that in the absence of a disease-modifying AD therapy, correct diagnosis has little effect on patient health.
Clinicians strongly disagree. “Having an accurate diagnosis is the foundation of good medical care,” said Stephen Salloway of Butler Hospital, Providence, Rhode Island. He pointed out that diagnosis affects prognosis, medications, and patient education. Ron Petersen of the Mayo Clinic in Rochester, Minnesota, said that amyloid scans would be useful to rule in AD when people have atypical symptoms. Speaking at the July 18 meeting, Ali Shamsi at Yale University, New Haven, Connecticut, stressed the value of negative amyloid PET scans for ruling out AD when people present clinically with dementia that is due to a silent stroke, other brain injury, drug abuse, or chronic mental illness. In these cases, correct diagnosis would greatly alter treatment and decisions about where best to treat and care for patients, Shamsi said. In this way, amyloid imaging also aids differential diagnosis for people who have a non-AD progressive neurodegenerative disease, such as frontotemporal dementias. Norman Foster of the University of Utah, Salt Lake City, said, “We’ve been surprised that so many people coming to memory clinics have negative scans.” Many of the speakers at the July 15 session have served as consultants for Lilly.
CMS’ Answer: Coverage with Evidence Development
Now about that silver lining. In the draft decision, CMS agreed that PET Aβ imaging could be promising in specific scenarios, such as excluding AD in clinically difficult diagnoses, or to enrich clinical trials. To prove that scans improve patient outcomes in these scenarios, CMS proposed a “coverage with evidence development” (CED) process. The agency will cover a single PET Aβ scan for cognitively impaired Medicare beneficiaries participating in randomized clinical trials. These trials must address the question of whether amyloid imaging leads to improved health outcomes. They must also satisfy various other strictures, including “postmortem diagnosis as the endpoint where appropriate.”
Speakers at the July 15 session roundly condemned this last requirement as absurd, noting that an autopsy endpoint could require trials to last 15 to 20 years. Even without that criterion, AD stakeholders criticized the CED process as unnecessarily slow and burdensome. For example, one upcoming trial that is generating considerable excitement in the field, the Anti-Amyloid Treatment in Asymptomatic AD (A4) Trial, will use amyloid scans to screen presymptomatic patients for inclusion; these people are expected to live for decades.
Moreover, AD researchers wondered how such trials could be run in practice. “I’m concerned about how we would actually generate evidence of benefit,” said Reisa Sperling of Brigham and Women’s Hospital, Boston. She wondered what outcome metrics would be acceptable. Diagnostic tests by themselves do not improve symptoms or survival, Sperling noted. Even if paired with an experimental AD drug, it might be hard to see clinical change in a prodromal AD population. Would CMS consider improvement on cognitive tests to be a meaningful clinical outcome? Sperling asked. Researchers note that none of that is clear.
Previously, researchers in the field of oncology used the CED mechanism to gain coverage for PET ligands through the National Oncologic PET Registry (NOPR). This process took seven years. However, the bar has been set higher for Aβ ligands, said Gail Rodriguez, president of the Medical Imaging and Technology Alliance (MITA), a Washington, D.C.-based industry association. Under NOPR, PET scans were reimbursed for all Medicare beneficiaries as long as their physician submitted data regarding the treatment decision to the registry. Clinical trials were not required.
Numerous scientists perceived a double standard between coverage decisions on fluorodeoxyglucose (FDG) PET and amyloid PET. CMS covers the use of FDG PET to distinguish between frontotemporal dementia (FTD) and AD, but not amyloid PET, despite evidence that the latter is more effective for this purpose.
Some believe the real issues are financial. Rodriguez suggested that CMS wants to throttle access to the scans because it fears a deluge of demand for which it cannot afford to pay. Howard Fillit, director of the Alzheimer’s Drug Discovery Foundation, New York City, struck a moderate tone, noting that in the current climate of health care reform and cost-cutting, CMS’ focus on improved outcomes is reasonable. Insurers fear that everyone over 50 will want this test, Fillit said. He suggested reassuring CMS that primary care physicians will not be able to order the scans. In a plenary session on amyloid imaging, Stephen DeKosky at the University of Virginia, Charlottesville, noted that insurers are increasingly denying even FDG PET scans.
AD Stakeholders Weigh Options
At the July 15 meeting, participants strategized about ways to persuade CMS to change its ruling, and do so during the ongoing 30 day period in which CMS will accept public comments. Many advocated for flexing political muscle, pointing out that the CMS verdict clashes with the National Alzheimer’s Project Act’s (NAPA) directive to “optimize care quality” for AD patients (see ARF related news story; ARF related news story). Others suggested mobilizing a grassroots effort of Alzheimer’s Association members, Medicare recipients, general care practitioners, and neurologists. Researchers fretted over the limited time available before the draft decision is made final on August 2. Carrillo said the association will request that CMS extend the comment period to 90 days.
Many agreed that the best strategy is to press for coverage for certain limited indications, particularly where there is already a precedent for CMS coverage of FDG PET scans. In addition to the differential diagnosis of FTD versus AD, a precedent exists for coverage of early onset AD, said Jason Karlawish at the University of Pennsylvania, Philadelphia. Researchers agreed this might serve as a foot in the door toward eventual broader coverage.
Meanwhile, Lilly continues to gather evidence to try to support the clinical utility of amyloid β PET scans. At AAIC, Noam Kirson of The Analysis Group Inc., Boston, presented cost data from a study funded by Lilly. He reported that Medicare beneficiaries with Parkinson’s disease or vascular dementia who were incorrectly diagnosed with AD accumulated about $12,000 more per year in reimbursed Medicare expenses compared to peers who were correctly diagnosed. Once the misdiagnosis was corrected, their costs dropped back to those of peers. This would suggest that ruling out AD early could save money, Kirson concluded.
Another talk at AAIC correlated early diagnosis with better patient outcomes. Ilona Hallikainen from the University of Eastern Finland, Kuopio, reported that people with AD managed their daily activities better and had less psychological and behavioral symptoms when they were diagnosed and started treatment early in the disease rather than later. After three years of follow-up, people who were diagnosed when they had a clinical dementia rating (CDR) of 0.5 were coping better than peers diagnosed at CDR 1. The results imply that early diagnosis might help AD patients live at home longer, Hallikainen said.—Madolyn Bowman Rogers.