JAMA: Major PET trial of Alzheimers criticized

Florbetapir-PET scan of high beta-amyloid burden (score 4 of 4), correlated with Alzheimer’s. Image Source: JAMA, 2011; 305: 275-283.

Consumer advocacy organization Public Citizen has submitted a letter to the editor to the Journal of the American Medical Association, criticizing the Clark et al study, published Jan. 19, which reported a strong quantitative correlation between the biomarker florbetapir F18's signal on PET scans and the presence of beta-amyloid in the brain at death, suggesting that imaging biomarkers may be applied for diagnosing and determining early risks for Alzheimer's disease.

The letter, penned by Michael Carome, MD, and Sidney Wolfe, MD, from the health research group of Public Citizen in Washington, D.C., suggested that Clark and colleagues submitted median values for PET scan reader scores to JAMA, but withheld “critical individual reader score data” that Avid Radiopharmaceuticals submitted to the FDA on Sept. 17, 2010.

“FDA analyses of these data show substantial interreader variability among independent, extensively trained readers of the florbetapir-PET scans for individuals in the autopsy cohort, emphasizing that florbetapir-PET imaging fails to provide an accurate and reliable assessment of amyloid burden,” the letter stated.

Also, Carome and Wolfe pointed out that the published measurement of the amyloid burden for this correlation was the median rating of three expert independent readers who underwent extensive training, a process that Clark and his co-authors acknowledged “is not likely to be replicated in clinical settings.”

Therefore, the data on the interreader variability for the three independent readers are “crucial to evaluating the potential clinical utility of florbetapir-PET,” the letter indicated.

The letter concluded that the study implemented florbetapir-PET in “a rigorously controlled setting with well-trained readers, using patient populations at two extremes of the spectra for both age and health. However, despite these optimal conditions, the test yielded disparate results when looking at the analysis of the individual readers.

“If widely deployed in the real-world setting,” Carome and Wolfe continued, “with more variability in reader training and skill and in the patient population for whom florbetapir-PET presumably is intended, the performance of the test will most likely be worse. For these reasons, in our opinion the FDA should not approve florbetapir for diagnosis of Alzheimer's disease.”