Radiology: MRI predicts Alzheimers progression
Image source: Chester Mathis, PhD, professor of radiology/PET facility at the University of Pittsburgh, Pennsylvania
Researchers have been well aware of the high likelihood that individuals suffering from memory loss will convert to Alzheimer’s, but  now physicians have a way of predicting an individual patient’s risk of developing the disease, according to an April 6 study published in Radiology. This closes a research gap that crippled early detection and the induction of therapeutic trials for Alzheimer’s.

Although only 1 to 2 percent of the general population converts to Alzheimer’s disease in a given year, most patients who suffer from the degenerative memory condition of mild cognitive impairment (MCI) eventually progress to develop Alzheimer’s. The ability to predict which patients would convert to the disease and which would not is critical for early detection of Alzheimer’s and for developing and eventually administering effective therapies.

Using data and images from the Alzheimer’s Disease Neuroimaging Initiative (ADNI), a private and publicly funded study consisting of more than 800 healthy MCI and Alzheimer’s patients, researchers from the University of California, San Diego (UCSD) sought to determine differences in brain atrophy via MRI among the three patient groups.

“We wanted to look at people with MCI’s MRI information to see if we could predict how likely they would be to convert to Alzheimer’s disease within the next year, to find more specific information on an individual’s risk,” said Linda K. McEvoy, PhD, from the department of radiology at UCSD and the lead author of the study.

McEvoy pointed to a large volume of literature showing the degenerative characteristics of Alzheimer’s. “The new part of this study was really utilizing the patient-specific information, which is necessary for information to be eventually translated to clinical practice.”

McEvoy and colleagues measured brain atrophy in several cortical areas as well as expansion of the ventricles using MRI, attempting to mobilize these data as risk strata for MCI patients’ conversion to Alzheimer’s. Both baseline and one-year follow-up MRIs were evaluated among a sample of 149 healthy control subjects, 104 Alzheimer’s patients and 199 patients with MCI.

Based solely on the baseline MRI scans, McEvoy said, “We found that those people with MCI not showing brain shrinkage had very low risks of conversion to Alzheimer’s, about 3 percent per year, and that for those patients with a lot of brain shrinkage, their risk was double what you’d see in normal MCI patients, so their risk of converting to Alzheimer’s was 40 percent per year.”

“We also found that if we look not just at thickness or shrinkage at a single time, but how much brain tissue they’ve lost over the period of a year, that provides even better information on how likely someone is to convert to AD. When including this information we found we could identify really high risk people, those with up to 69 percent likelihood of conversion to Alzheimer’s disease.”

Based on the volumetric scoring criteria the authors fitted to the sample, McEvoy and co-authors were able to differentiate with 91 percent sensitivity and specificity which patients were normal and which had Alzheimer’s. When using atrophy scores to predict patients’ conversion from MCI to Alzheimer’s, McEvoy and colleagues found that those MCI patients in the highest quartile had significantly greater chances of conversion to Alzheimer’s than those in the lowest quartile (odds ratio 12.0).

“The results of this study demonstrate that patient-specific estimates of the risk of conversion from MCI to AD [Alzheimer’s disease] can be derived from quantitative measures of brain atrophy obtained from both single-time-point and serial MR imaging examinations,” McEvoy and colleagues explained. “Use of these measures substantially improves risk predictions compared with risk predictions based on the clinical MCI diagnosis alone.”

McEvoy emphasized that identifying this patient-specific risk stratification might pave the way to early detection and set a foundation for therapeutic trials. Whereas previous research has chiefly highlighted the broad structural brain differences between groups of persons with and without Alzheimer’s, McEvoy and colleagues’ research facilitates the identification of early onset of the disease and could allow researchers and physicians to track progression more accurately. This more specific information on the risk and progression of the disease sets a standard against which the effects of a drug versus a placebo could be demonstrated.

McEvoy pointed out that the findings were relative to a highly selective sample of patients, which comes with both benefits and limitations. She said that validating their findings on a larger, more representative sample would be necessary, and could inform the development of an automatic, push-button method by which physicians could identify their patients’ risks of Alzheimer’s progression.

“So this information will not only be helpful for clinical trials,” McEvoy concluded, “but eventually, we hope, for patient care.”