Utility of Amyloid Imaging in Alzheimer's Still Uncertain

Pauline Anderson

January 15, 2014

The medical literature provides extremely limited data with which to evaluate the clinical utility of amyloid-beta positive-emission tomography (Aβ PET), a new summary of the evidence on imaging in the diagnostic evaluation of Alzheimer's disease (AD) concludes.

A positive Aβ PET result "is not diagnostic of AD, nor can the test be used to accurately predict the risk of the timing of progression of mild cognitive impairment," write the review authors, led by Steven D. Pearson, MD, at the Institute for Clinical and Economic Review, Boston, Massachusetts.

The review, which was part of the background material for the assessment by the Centers for Medicare and Medicare Services (CMS) of the current reimbursement status of imaging scans, was published in the January issue of JAMA Internal Medicine.

Final Decision

In April 2012, the US Food and Drug Administration (FDA) approved Aβ PET with florbetapir F-18 (Amyvid; Eli Lilly and Company) to identify Aβ plaque accumulation. A negative scan result indicates a reduced likelihood that cognitive impairment is due to AD, but a positive scan does not establish a diagnosis of AD.

In July 2013, the CMS released a draft decision memo for public comment, and the Final Decision memorandum was released on September 27, 2013. In its final decision, the CMS determined that because the evidence is insufficient to conclude that the use of PET Aβ imaging is "reasonable and necessary" for the diagnosis or treatment of illness or injury or to improve the functioning of beneficiaries with dementia or neurodegenerative disease, such imaging is not covered under the Social Security Act.

However, the CMS determined that this imaging is "promising" in 2 scenarios: first, to exclude AD in "narrowly defined and clinically difficult differential diagnoses, such as AD vs frontotemporal dementia (FTD)"; and to enrich clinical studies of treatments and preventive AD strategies. As such, CMS will cover 1 PET Aβ scan per patient in clinical studies that meet certain criteria.

For example, the objectives of the study must be to develop better treatments or prevention strategies for AD or to resolve clinically difficult differential diagnoses. The clinical studies must be approved by CMS, involve participants from appropriate populations, and be comparative and longitudinal. And where appropriate, the studies should be prospective, randomized, and use postmortem diagnosis as the endpoint. In addition, the radiotracer used for the study must be FDA approved. "All other uses are not covered," the decision document notes.

This newly published review provides a summary of findings from the relevant research supporting that decision. Evaluators looked at 14 Aβ PET studies that assessed diagnostic accuracy and 1 that assessed the impact of testing on diagnostic impression. As expected, there were no randomized, clinical trials of the use of Aβ PET imaging in clinical practice.

Of the 14 studies evaluating diagnostic accuracy, 13 compared imaging results with clinical diagnoses or with the "gold standard" of AD diagnosis, which is biopsy. A "pivotal" study that compared the sensitivity and specificity of Aβ PET with biopsy results included patients aged 51 years or older who were judged by their physician to be approaching the end of life, as well as young, healthy control individuals. In this study, 66% of all individuals had AD diagnosed by means of biopsy.

With multiple readers interpreting each scan, the analysis found that the median sensitivity of Aβ PET was 92% and the specificity was 95%. The ranges, however, were 69% to 95% and 90% to 100% for sensitivity and specificity, respectively.

One diagnostic study assessed the ability of Aβ PET to provide useful prognostic information on the likelihood of progression from mild cognitive impairment to full dementia. In this study, positive PET scan results were associated with greater decline in multiple cognitive outcome measures. Of older patients with mild cognitive impairment, 29% of those with a positive scan developed full dementia within 18 months, compared with 10% of those with a negative scan result.

However, the authors of that study noted that "the data were insufficient to predict whether, or when, cognitive deterioration would occur in individuals with positive scan results," Dr. Pearson and colleagues write.

Diagnostic Impression

The single study that looked at the impact of testing on diagnosis impression included 229 patients with suspected AD. Physicians provided a provisional diagnosis and intended management plan before and after a scan.

Scan interpretations were associated with a change in reported diagnosis in 54.6% of cases; and in at least 85% of cases, at least 1 aspect of intended management, such as the use of cholinesterase inhibitor or memantine, for example, was changed.

Aside from the small number of studies, another limitation of the available evidence is that it does not evaluate whether test results improve clinical management or patient outcome. Further, the research does not evaluate potential benefits and harms of the imaging on, for example, patients with dementia other than AD, and the ability to make life plans.

Another potential problem is that the available studies are not representative of the full spectrum of patients who would likely be considered for testing in clinical practice, because they involved highly trained physicians and patients who were very likely to have AD, said the review authors.

The clinical utility of a positive scan result "remains uncertain," Dr. Pearson and his colleagues concluded. "If tested, approximately one third of cognitively normal older adults would have a 'positive' test result for brain amyloid," they write.

They added that although the clinical utility of a negative result seems greater because the high sensitivity can essentially rule out AD, it "remains uncertain" whether "negative test results lead to important changes in subsequent clinical management or whether any such changes would produce net health benefits for patients and families."

Courageous Approach

In an accompanying "Editor's Note," Robert Steinbrook, MD, who was a member of the Medicare Evidence Development and Coverage Advisory Committee from 2010 to 2012, said that the CMS proposal is a "courageous and innovative approach," because it will allow "meaningful data" to be gathered that could potentially help to improve the lives of millions of Medicare beneficiaries.

"This approach serves the public interest," Dr. Steinbrook concludes. "It would be irresponsible for the CMS to cover Aβ PET imaging without adequate evidence about its role in AD or other neurodegenerative diseases."

Dr. Pearson reports no relevant financial relationships. Dr. Steinbrook was a member of the Medicare Evidence Development and Coverage Advisory Committee from 2010 to 2012.

JAMA Intern Med. 2014;174:133-134, 135. Abstract, Editor's Note


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: