Public Release: 

PET scan predicts Alzheimer's more accurately

University of California - Los Angeles

Scanning a patient's brain metabolism with positron-emission tomography (PET) can improve a doctor's ability to forecast the patient's future cognitive functions by up to 30 percent, a new UCLA study discovered. Published in the November issue of the journal Molecular Genetics and Metabolism, the findings suggest that PET may offer physicians a new tool to help with earlier diagnosis and treatment of Alzheimer's disease and dementia.

The study showed that use of a PET scan sharply increased the ability of physicians to predict whether the condition of patients with early memory complaints would significantly worsen in the years after their initial examination.

"Adding PET to the diagnostic evaluation of patients with mild cognitive changes can improve our accuracy in predicting what will happen to them in the future and enhance our ability to intervene earlier," said Dr. Dan Silverman, associate director of imaging at the UCLA Alzheimer's Disease Center and assistant professor of molecular and medical pharmacology at the David Geffen School of Medicine at UCLA.

Silverman and his colleagues studied the results of PET scans performed between 1991 and 1999 on 167 UCLA patients being evaluated by their neurologists for mild cognitive complaints, such as memory loss, behavioral changes and changes in language ability. The average age of the patients was 66.

A PET scan images the brain for a metabolic pattern indicative of progressive dementia. Based on the PET scan results, the UCLA researchers predicted each of the 167 patients' future cognitive abilities. The research team then compared those predictions to the neurologists' original clinical diagnoses and to patients' cognitive conditions two to 10 years after their initial evaluation.

According to Silverman, the accuracy of the prognoses significantly increased when doctors took the PET scan results into account.

"Adding PET substantially boosted how often physicians were correct in predicting patients' future cognitive decline or stability," he noted.

For example, neurologists who diagnosed their patients as free of a progressive disease were correct 66 percent of the time. When patients' PET scans revealed a negative diagnosis, the accuracy of the doctor's predictions rose to 96 percent. However, patients in this group who had positives PET scan were 18 times more likely to develop progressive dementia in the future than those whose scans were negative.

Neurologists who diagnosed their patients with progressive dementia were correct 84 percent of the time. Adding a positive diagnosis from a PET scan boosted the accuracy of those predictions to 94 percent. Patients in this group with negative PET scans were 12 times more likely to remain cognitively stable during the follow-up period than those with positive scans.

Overall, patients with positive PET scans experienced a significant decline in general brain function years later. Those patients' performance fell significantly on the Mini-Mental State Examination (MMSE), a standard clinical test used to gauge general mental abilities.

"The predictive power of brain PET scans may prove most helpful to people with early cognitive problems who would not otherwise be suspected to have progressive dementia," Dr. Silverman said.

Patients whose doctors predicted they would maintain a stable mental course - but who received positive PET scans - experienced an average eight-point decline on the MMSE test. In comparison, patients in this group who had negative PET scans declined by an average of only two points.

This UCLA study is the first to follow a large number of clinical patients with early cognitive complaints through PET and several years of clinical monitoring, and to measure the accuracy added to clinical evaluation by visually interpreted PET scans.

Alzheimer's disease is the most common form of dementia. It often begins with mild memory lapses and gradually advances to a progressive deterioration of memory, language and mental functions. Alzheimer's affects approximately 400 million people in the United States, and the nation spends roughly $100 billion on the disease each year.


The Los Angeles Alzheimer's Association/ Turken Family Foundation and the National Institute on Aging provided funding for the study. Co-authors included Co Truong, Shanna Kim, Carol Chang, Wei Chen, Arthur Kowell, Jeffrey Cummings, Johannes Czernin, Gary Small and Michael Phelps.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.