Pauline Anderson

April 25, 2012

April 25, 2012 (New Orleans, Louisiana) — Misdiagnosis of Alzheimer's disease (AD) based on positron emission tomography (PET) scan readings appears to be a troubling problem that could get worse as more amyloid-specific tracers become available.

A new study found that almost two thirds of patients were misdiagnosed or had inconclusive data on the basis of interpretation of their PET scan in a community setting.

"We concluded that since this is such a large percentage of people who are misdiagnosed based upon their PET scans that we should caution clinicians to weigh the risks versus the benefits before ordering PET scans," said Sheena M. Shipley, BS, Departments of Neurology and Psychiatry, University of Colorado School of Medicine, Denver.

She presented the data here at the American Academy of Neurology's 64th Annual Meeting.


The US Food and Drug Administration (FDA) approved fluorodeoxyglucose (FDG)-PET in 2004 for the diagnostic evaluation of dementia. Previous studies had shown that this test can assist in the differential diagnosis of frontotemporal dementia and AD.

The research to date has shown varying diagnostic accuracy for PET, said Shipley. The current study sought to determine the accuracy of PET in the community setting, where clinicians are frequently faced with distinguishing demented from normal patients. Such clinicians will probably be tasked to do so even more in future as the bulk of the general population enters an age category where dementia is much more common.

For the study, researchers screened all patients seen at the University of Colorado Hospital Neurobehavior Clinic from September 2004 to September 2010. Diagnostic evaluations included complete medical and psychiatric history, neurologic examination, mental status exam, neuropsychological testing, and structural brain imaging.

A diagnosis was determined by a consensus of 2 fellowship-trained behavioral neurologists. This diagnosis was compared with interpretations of PET scan readings. The study included scans of 46 patients; 72% were male, and the mean age was 63.4 years.

The study showed that 30 of the 46 patients, or 65%, were misdiagnosed or had inconclusive data on the basis of the interpretation of their scan. As well, 16 patients, or 35%, had false-positive findings, in which the scans were mistakenly read as showing AD.

"What's interesting about that is that of those 16 people, 9 either had normal cognition or a reversible cause of dementia such a psychiatric cause or vitamin B12 deficiency," said Shipley.

She noted that only 11% of the scans were interpreted by neuroradiologists.

The study might reflect a lack of quality control and standardization for PET scan reading, said Helena Chui, MD, professor and chair, director, Alzheimer Disease Research Center, Department of Neurology, Keck School of Medicine, University of Southern California (USC), Los Angeles. She noted she has seen a similar lack of standardization reflected in her own practice. Community physicians, she said, send patients for an FDG-PET scan at USC, where radiologists read the scans and refer patients suspected of having AD on to her.

"Actually, when I do the full work-up, I sometimes tell these patients, the good news is that I don't think you have Alzheimer's disease," Dr. Chui noted. "Then I look at the PET scan and I also don't think the PET scan supported AD."

She advocates "joint interdisciplinary conferences," during which all those charged with reading scans, including general radiologists and neuroradiologists, can review images together.

Problem Compounded?

The problem of scan misinterpretation will soon be compounded with the arrival later this year of new amyloid tracers, said Gil D. Rabinovici, MD, assistant professor of neurology, Memory and Aging Center, University of California, San Francisco.

The concern, he said, is that clinicians, especially primary care doctors, will be tempted to opt for a scan instead of a detailed cognitive evaluation or specialist referral. "A positive amyloid scan means there is high likelihood of amyloid plaques, but that doesn't mean a patient has AD as the cause of their symptoms," he told Medscape Medical News.

He suggested that although new diagnostic technology represents a major advance in the field, increasing access to such technology could lead to further misinterpretations. "The scans must be ordered only in certain situations, as an adjunct to a clinical evaluation, cognitive testing, and structural imaging."

Even though Avid, the company now owned by Eli Lilly that is behind the launch of florbetapir, an amyloid tracer recently approved for use in living patients, has a training program in place, many — if not most — PET scans will continue to be obtained in private practice, noted Dr. Rabinovici.

Why did this current study pick up such a high percentage of misdiagnosis when past studies did not? "The populations in research studies tend to be highly selected, the study highly controlled, and therefore these studies often overestimate the value of a diagnostic, or the effect of a drug," said Dr. Rabinovici.

"The real question is how these scans will perform in real practice, in patients with comorbidities, who would have been excluded from a research study."


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: