Combining MRI and PET with CSF Analysis Provides a Better Prediction of Alzheimer’s Disease

The study’s findings, which were published December 11, 2012, in the journal Radiology, provide new clues into how to best detect AD before the full onset of the disease. Investigators from Duke University (Durham, NC, USA) assessed three tests--cerebrospinal fluid analysis, fluorine-18-fluorodeoxyglucose positron emission tomography (FDG-PET), and magnetic resonance imaging (MRI)--to determine whether the combination provided more accuracy than each test individually. The tests were added to routine clinical exams, including neuropsychologic testing, currently used to diagnose AD.

“This study marks the first time these diagnostic tests have been used together to help predict the progression of Alzheimer’s. If you use all three biomarkers, you get a benefit above that of the pencil-and-paper neuropsychological tests used by doctors today,” said Jeffrey Petrella, MD, an associate professor of radiology at Duke Medicine and study author. “Each of these tests adds new information by looking at Alzheimer’s from a different angle.”

AD affects more than 30 million people worldwide, and the number is expected to triple by 2050. Patients with mild cognitive impairment are at a higher risk for having Alzheimer’s, but not everyone will progress to developing the disease. Studies suggest that AD begins years to decades before it is diagnosed, with patients experiencing a phase with some memory loss, or mild cognitive impairment, before the disorder’s full onset.

Whereas there is no cure for Alzheimer’s, new treatments are being researched are likely to be most effective at the disease’s earliest stages. Researchers are working to improve early detection of AD, even before patients experience symptoms, but the disease remains difficult to diagnose and patients are often misclassified.

“Misdiagnosis in very early stages of Alzheimer’s is a significant problem, as there are more than 100 conditions that can mimic the disease. In people with mild memory complaints, our accuracy is barely better than chance. Given that the definitive gold standard for diagnosing Alzheimer’s is autopsy, we need a better way to look into the brain,” said P. Murali Doraiswamy, MBBS, professor of psychiatry and medicine at Duke and study author. “Physicians vary widely in what tests they perform for diagnosis and prognosis of mild memory problems, which in turn affects decisions about work, family, treatment, and future planning.”

The Duke investigators analyzed data from 97 older adults with mild cognitive impairment from the Alzheimer’s Disease Neuroimaging Initiative, a national study that collects data in hundreds of elderly patients with varying degrees of cognitive impairment. Participants took part in clinical cognitive testing, as well as the three diagnostic tests: MRI, FDG-PET, and cerebrospinal fluid (CSF) analysis. They then checked in with their physicians for up to four years.

The misclassification rate based only on neuropsychologic testing and other clinical data was comparatively high at 41.3%. Adding each of the diagnostic tests reduced the number of misdiagnoses so that, with all three tests combined, researchers achieved the lowest misclassification rate of 28.4%. FDG-PET, out of the three individual diagnostic tests, contributed the most information to clinical testing to identify early Alzheimer’s in patients with mild cognitive impairment.

Researchers noted that while the tests clearly provided the most diagnostic information when used in combination, additional studies are needed to better determine their role in a clinical environment. “The study used a unique data-mining algorithm to analyze MRI and PET images for ‘silent’ information that may not be apparent to the naked eye. Hence, the data should not be taken to mean that imaging should be done in every patient; rather, it sought to capture the maximum potential information available in the images,” Dr. Petrella said. “Additional studies, including those looking at the cost effectiveness of these tests, are also needed to translate the most useful biomarkers into clinical practice.”

A multicenter study led earlier in 2012 by Dr. Doraiswamy demonstrated that another type of PET scan that tags amyloid plaques in the brains of individuals with AD also helped predict the onset of the disease in those with mild cognitive impairment. Adding amyloid plaque imaging technology in additional research could complement the modalities used in this research, providing even more diagnostic data.

Related Links:
Duke University