Brain scans prior to antidepressant treatment may predict response

Functional magnetic resonance imaging of the brain may help predict which patients will respond favorably to antidepressant treatment, according to a study published in the journal Brain.

Researchers from the University of Illinois at Chicago and the University of Michigan conducted fMRI scans of patients with major depressive disorder who were about to begin antidepressant treatment. The results showed that patients who had more active connectivity in two brain networks when they made errors on a cognitive task were less likely to respond to antidepressants.

These two networks were the error detection network, which acts when an error is noticed, and the interference processing network, which acts when deciding what information to focus on.

The increased cross-talk in these networks may reflect a tendency to ruminate on negative events, such as errors, or a deficit in emotional regulation when confronted with an error,” said Natania Crane, a graduate student in the Department of Psychiatry at the UCLA School of Medicine and lead author of the study.

Finding the right treatment can take many months. Medications used to treat major depressive disorder take eight to 12 weeks to have a noticeable effect on mood and other symptoms. In addition, patients may not respond to the first prescribed medication or suffer side effects, resulting in the need to change medications. Therefore, being able to predict responses to medications could shorten patients’ recovery time and reduce health care costs, said Scott Langenecker, assistant professor of psychology and psychiatry at UIC and lead author of the study.

Several studies that have used fMRI to identify brain regions that are hyper- or hypoactive in patients with major depressive disorder have suggested that neuroimaging may be useful for predicting patients’ response to certain medications.

In the current study, researchers examined brain activation patterns during participants’ performance of a cognitive control task and whether they predicted response to medication. The researchers used unique analysis techniques to determine which brain regions that were highly activated during errors in the cognitive task correlated with response to treatment, and how the strength of connectivity in specific brain networks predicted response to treatment.

The study involved 36 adult patients with major depressive disorder who were not receiving medication at the time of the study. The subjects underwent fMRI scans and completed a questionnaire about symptoms of depression. They then chose one of two antidepressants: escitalopram (Effexor, a selective serotonin reuptake inhibitor, 22 participants) and duloxetine (Cymbalta, a serotonin/norepinephrine reuptake inhibitor, 14 participants).

During the fMRI scan, participants were instructed to watch the letters X, Y, and Z flash on the screen. Participants were instructed to press a button each time they saw a letter, but not to press the button a second time if the same letter was repeated.

Patients were observed during and after 10 weeks of antidepressant treatment. They filled out questionnaires and were interviewed to find out if the prescribed medication reduced their symptoms.

It turned out that patients with higher brain activity in error detection and interference processing networks were less likely to experience a reduction in depressive symptoms with medication.

Using our model, we were able to predict with a very high degree of accuracy which patients would respond well to antidepressant treatment and which patients would not.”

The researchers also found that participants who made more errors during cognitive tasks were more likely to respond to antidepressant treatment.

This is an important step toward personalized medicine in the treatment of depression.” Using cognitive tests and fMRI, we can identify those who respond best to antidepressant therapy and those who need other effective treatments that work by other mechanisms, such as psychotherapy,” Langenecker says.