Brain Structural Data Useful in Classifying Chronic Migraine

WASHINGTON — New data suggest that classifiers including cortical surface area, cortical thickness, and regional volumes are highly accurate in determining if individuals have chronic migraine.

The study was presented at the American Headache Society Annual Scientific Meeting in Washington, D.C., by Todd J. Schwedt, MD, FAHS, of the Mayo Clinic in Scottsdale, Ariz.

According to Schwedt, although diagnostic criteria in the International Classification of Headache Disorders (ICHD) represent a significant advance in the headache field, the criteria have limitations, particularly because primary headaches, like migraine, are based mostly on expert opinions.

“An accurate brain imaging classifier for migraine could potentially serve as an objective gold standard for assigning a migraine diagnosis and therefore be used for validating and optimizing the clinical diagnostic criteria,” he told Neurology Advisor.

This led Schwedt and colleagues to conduct a trial that examined the utility of magnetic resonance imaging (MRI) measures of brain structure for constructing classifiers. The goal with the classifiers was to accurately identify patients as having chronic or episodic migraine vs. being a healthy control, and to test the currently used threshold of 15 headache days per month for differentiating chronic migraine from episodic migraine.

The study included 66 patients with migraine and 54 healthy controls. The majority of patients were women (75.8%), and the mean age was 36.3 ± 11.1 years.

Participants underwent MRI for determining regional cortical thickness, cortical surface area, and volume. Data revealed the following mean classifier accuracies:

• 68% for episodic and chronic migraine vs. healthy controls
• 67.2% for episodic migraine vs. healthy controls
• 86.3% for chronic migraine vs. healthy controls
• 84.2% for chronic migraine vs. episodic migraine

According to the researchers, the classifiers were comprised of several structural measures, commonly including the temporal pole, anterior cingulate cortex, superior temporal sulcus, entorhinal cortex, medial orbital frontal gyrus, and pars triangularis.

“Using these classifiers consisting of these structural measures, we could determine if an individual brain MRI belonged to someone who has chronic migraine or to someone who was a non-migraine healthy control with 86% accuracy and we could determine if an individual brain MRI belonged to someone who has chronic migraine or to someone who has episodic migraine with 84% accuracy,” Schwedt said. “However, our classifiers for differentiating episodic migraine from healthy non-migraine controls were less accurate.”

Furthermore, Schwedt and colleagues found that the threshold of 15 headache days per month was the most accurate sub-classification of patients with migraine into lower and higher frequency subgroups.

Going forward, Schwedt said he and colleagues would like to continue their work in developing accurate classifiers for identifying migraine and migraine subgroups, as well as differentiating migraine from other headache types.

Reference

1. Schwedt TJ et al. Abstract 85751. Presented at: American Headache Society Annual Scientific Meeting 2015; June 18-21, 2015; Washington, D.C.