Subscribe to eTOC



Investigators report that regional temporal lobe cortical changes — detected on MRI — could serve as a biomarker for mild cognitive impairment and Alzheimer disease and could be useful as outcome measures in clinical trials.

In the hunt for biomarkers to diagnose and track progression of Alzheimer disease (AD), investigators at the University of California-San Diego (UCSD) reported that serial MRI scans of the entorhinal cortex shed the brightest light on the disease process.

The scans could measure the effectiveness of experimental treatments in blocking disease progression, said Anders M. Dale, PhD, professor of neurosciences and radiology at the UCSD School of Medicine, who led the study. That it is more accurate than clinical examination in assessing the effect of a therapy means that researchers could need fewer study subjects, a concern for many investigators and drug companies trying to identify potent disease-modifying AD medicines.

The study, reported Nov. 20 in the online version of the Proceedings for the National Academy of Sciences, was conducted as part of the Alzheimer's Disease Neuroimaging Initiative (ADNI), a consortium of public institutions and private companies working together since 2003 to test available imaging measures and other biomarkers — as well as clinical and neuropsychological testing — to track the course of mild cognitive impairment (MCI) and AD.

One ADNI goal has been to standardize and optimize image acquisitions so scientists could follow people who have been scanned in different places throughout the country and with different devices. The active ADNI cohort includes 170 AD patients, 193 healthy controls, and 364 patients with MCI. Serial MRI scans were acquired at six to 12 month intervals for participants who have been in the study for up to three years.

The UCSD scientists evaluated many brain regions, including the hippocampus, inferior temporal and middle temporal cortex, fusiform, and entorhinal cortex. They also measured the rate of volume change of whole brain and the ventricles. What had not been clear, Dr. Dale said, is the “importance of considering the normal effects of healthy aging.”

Imaging studies of whole brain or ventricles offer a picture of the shrinking AD brain but the UCSD study showed that the normal aging brain also becomes smaller over time. When scientists measured shrinkage in regions more specifically involved in AD, such as the entorhinal cortex, they were able to better distinguish the AD brain from the brain of someone following the pattern of normal aging.


TISSUE SEGMENTATION, with a six- and 12-month volume change field for a subject with mild cognitive impairment. (A) Segmentation of the baseline MRI scan, with different brain structures represented in different colors. (B) Corresponding coronal slice overlain with a heatmap representation of the voxelwise estimates of volumetric change at six months and (C) 12 months. (D) Left hemisphere cortical parcellation of the baseline MRI scan. (E) Cortical surface overlain with a heatmap representation of the estimates of cortical volumetric change at six months and (F) 12 months. The left middle-temporal gyrus has decreased in volume by 4.7 percent at six months and by 8.2 percent at 12 months; the left temporal-horn lateral ventricle has increased by 17.4 percent at six months and by 35.3 percent at 12 months. Holland D, et al. Proc Natl Acad Sci USA 2009; E-pub 2009 Nov 20.


The entorhinal cortex exhibits pathological changes years before symptoms develop, Dr. Dale said. The scientists reported that entorhinal-volume change per year in the healthy controls was 0.6 percent; in MCI, 2.5 percent, and in AD, 3.8 percent. By comparison, whole brain rate of shrinkage was identified as 0.4 percent in healthy controls, 0.9 percent in MCI, and 1.5 percent in AD.

Dr. Dale said that one of the most important observations of the new ADNI data is that “there are potential confounding effects of aging on brain change measures, and especially on ‘global’ measures such as whole brain volume or ventricle size.”

“The entorhinal cortex stands out. I am surprised it came out as well as it did. But it is certainly where I would have placed my bets,” said Dr. Dale.


The problem is that the entorhinal cortex is a small structure and it hasn't been practical to measure it accurately in large, multicenter studies. The method developed by Dominic Holland, PhD, in Dr. Dale's laboratory makes it feasible, for the first time, to obtain very precise automated measures of the structural changes in the entorhinal cortex in such studies.

“Having identified the entorhinal cortex and other regions of interest in the baseline scan, the follow-up scan is used to determine its change in shape and size,” said Dr. Dale. Scans are taken every six months and the new procedure identifies and measures spatial changes over the course of the repeated scans.

“Importantly, this procedure takes into account specific properties of the MRI scanner used, which could otherwise confound subtle anatomical changes associated with disease progression,” he added.

Today, most MRI scans on patients with memory problems are conducted to rule out infarcts or tumors. “Most radiologists will not read atrophy,” said Dr. Dale. “Now it is becoming possible to look for regional differences over time.”

Gary Small, MD, director of the University of California-Los Angeles Center on Aging, who has pioneered PET techniques for MCI and AD, agrees. “Biomarkers to track disease will allow us to develop and test medicines more efficiently.”

“We are trying to find a cholesterol test for the brain,” he said, adding that “it is a challenge. There is a an assumption that a bigger brain is a better brain and we don't know whether that is true.”

There are limitations in looking at brain structure, he said. A true surrogate biomarker, Dr. Small said, “follows the course of the disease.” But the problem is this: “Your brain may look better but you still won't be able to remember this conversation.”

Dr. Small and his colleagues are now studying PET ligands for tau, the protein that makes up the tangles in the brains of MCI and AD patients. At press time, they were preparing to describe their initial results at the American College of Neuropsychopharmacology meeting in December.

Dr. Small said that tauopathies seem to be better correlated to cognitive decline than the amyloid-filled plaques. A lot of PET imaging is now centered on the Pittsburgh Compound-B ligand that measures amyloid deposition.

“The problem is that amyloid is everywhere and we need a biomarker that tracks with the disease progression,” added Dr. Small.


• Holland D, Brewer JB, Dale AM, et al, of the Alzheimer's Disease Neuroimaging Initiative. Subregional neuroanatomical change as a biomarker for Alzheimer's disease. Proc Natl Acad Sci USA 2009; E-pub 2009 Nov 20.