Role of imaging expands in diagnosing dementias

Clinicians are increasingly looking to neuroimaging to help make earlier diagnoses of dementia, differentiate the various dementias, and monitor therapeutic progress. While research is ongoing, PET and MR imaging have been shown to have value in several areas, including some that overlap.

Clinicians want to be able to diagnose the transition from normal aging to mild cognitive impairment (MCI), said Dr. Steven T. DeKosky, director of the Alzheimer's Disease Research Center at the University of Pittsburgh in the U.S. Research is just beginning to define the range of changes experienced by patients who die with MCI. Changes in presymptomatic people who will convert to MCI are even less clear, DeKosky said at the International Conference on Alzheimer's Disease and Related Disorders (ICAD) held in July.

"It is up to the imagers to tell us what the status of the brain is in patients with MCI-not just their volumes, not just their functional data, but the level of pathological changes in the brain," he said.

While data regarding the conversion from normal to MCI are sparse, researchers are increasing their knowledge of biological and physiologic differences between patients with MCI and AD. Dr. Kejal Kantarci and colleagues at the Mayo Clinic in Rochester, Minnesota, reported in a presentation at ICAD that FDG uptake levels can distinguish between patients with amnestic mild cognitive impairment (aMCI) and those with AD. They used a 3D stereotactic surface projection technique to compare FDG uptake pixel by pixel.

FDG uptake in the right temporal lobe was the most accurate marker to differentiate controls from patients with aMCI and from patients with AD. Patients with AD showed the most significant decrease in FDG uptake in the temporal and parietal lobes, with a very high level of accuracy, Kantarci said.

Dr. Maciek Bobinski, a neuroradiologist at New York University Medical Center, and colleagues reported at the American Society of Neuroradiology meeting in June that hippocampal volume measurements accurately predicted-four years before the clinical diagnosis-which MCI patients would most quickly decline to AD. Accuracy was 71%.

German researchers reported at the ICAD meeting that a fully automated MR analysis system that extracts cortical thickness, followed by linear discriminant analysis at every vertex, can reproduce a clinical diagnosis of probable AD. The top 10 performing vertices were found in the bilateral parahippocampal gyri, obtaining a high of 100% specificity and 95% sensitivity at a single vertex. A combination of the top two vertices, one in each parahippocampal gyrus, with Mini-Mental State Examination scores resulted in a perfect diagnosis, said Jason Lerch, Ph.D., a researcher at Ludwig-Maximilian University in Munich.

A new arterial spin-labeled MR perfusion technique showed significant regional hypoperfusion in AD patients similar to PET and SPECT perfusion studies, said coinvestigator Norbert Schuff, Ph.D., an associate professor of radiology at the San Francisco VA Medical Center. The MCI group did not show statistically significant regional hypoperfusion relative to the control group, but there was a trend toward hypoperfusion in the inferior left parietal lobe, the region of greatest significance in the AD group.

Researchers from multiple centers, led by Dr. Bradford C. Dickerson, an instructor in neurology at Harvard Medical School, used functional MRI to investigate memory-related medial temporal lobe activation in patients with MCI. Those who progressed to AD within 2.5 years exhibited a greater extent of medial temporal lobe activation during picture encoding, Dickerson said at ICAD.

"We hypothesize that greater medial temporal lobe activation reflects a compensatory response to accumulating AD pathology and may therefore serve as a marker for impending clinical decline," Dickerson said.