Diagnostic Imaging
April 2003

OVERREAD

Halted drug trial reveals MRI's role in Alzheimer's

Imaging outperforms gold standard tests in tracing disease progression

The trial of milameline as a potential treatment for Alzheimer's disease may have been a bust for the drug, but it has turned out to be promising for imaging. Researchers at the Mayo Clinic in Rochester, MN, determined that MRI was a viable biomarker of AD progression.

"The therapeutic arm of the trial was discontinued because there was no observed drug effect. So it was converted from a therapeutic trial that incorporated MRI into a natural history trial," said lead researcher, Dr. Clifford R. Jack Jr., a professor of diagnostic radiology at Mayo.

Jack and colleagues initially enrolled 362 patients with suspected AD from 38 different centers in a 52-week randomized placebo-controlled trial of milameline. They performed a baseline MRI study on all the patients and followed up on 192 patients with an additional MRI study one year later.

The group reported a 4.9% reduction in hippocampal volume in the patients, as well as a 16.1% increase in temporal horn volume.

"This is the type of anatomic change that I think can be measured accurately and that could serve as a useful biomarker of disease progression in Alzheimer's disease," Jack said.

The researchers also found that correlations between the rate of MRI volumetric change and of change in behavioral/cognitive measures were greater for the temporal horn than for the hippocampus. They compared MRI results with current psychometric tests. The study was published in the Jan. 28 issue of Neurology.

"This is the first published study done along the design of a therapeutic trial where imaging was incorporated as it would be used as a surrogate biomarker of disease progression. We achieved reliable measurements that made biological sense, and the results showed a lot less variability than current psychometric tests," Jack said.

MRI measured atrophy in almost 100% of the subjects, which is consistent with the expectations for clinical AD.

Jack cautioned that MRI cannot take the place of current gold standard measures and that more research hinges on better treatment.

"The only way to validate a true biomarker of therapeutic efficacy is if you have a drug that actually modifies progression," he said. "We can say that MRI seems to measure disease progression and ought to be a biomarker, but without a positive therapeutic trial that does modify disease progression, we can't say that MRI has been validated."

Jack advocates studying a variety of modalities to monitor Alzheimer's, including structural MR, MR spectroscopy, serum biomarkers, psychometric tests, and PET.

"A problem in the field is that many unimodality studies have been published. Serial PET data look good, so does MRI. But how does PET stack up to psychometric tests, MRI, etc.? No one really knows right now," he said.

The National Institute on Aging is designing a large multisite longitudinal prospective study of mild cognitive impairment and early AD. The study will evaluate neuroimaging and other biomarkers for disease onset and progression.

Over the course of five years, the NIA project would study both normal subjects with mild cognitive impairment and patients with AD. The study design calls for researchers to collect serial MR and PET scans, clinical and neuropsychological information, and fluid and cell samples to identify other potential biomarkers. The research team aims to include 350 patients with mild cognitive impairment, 150 normal control subjects, and 150 patients with early AD who are between the ages of 55 and 90.

"The main methods to track therapeutic efficacy at this time are primarily clinical and neuropsychological," said Neil Buckholtz, Ph.D., chief of the Dementias of Aging program at the NIA.

The issue of whether to use MRI or PET instead of these psychometric tests is still just a research question, Buckholtz said.