BAL HARBOUR, FLA— Neuroimaging is revealing more and more about the etiologies of depression as well as the effects of treatments such as cognitive behavior therapy and pharmacologic agents. While this relatively nascent methodology continues to be developed and refined to predict the success of specific therapies, the implications for current treatment of depression are substantial.

That is the assessment of Helen S. Mayberg, MD, Professor of Psychiatry and Behavioral Science and Neurology at Emory University School of Medicine in Atlanta. In a presentation at the 15th Annual Meeting of the American Neuropsychiatric Association, she focused on treatment as a probe of the brain’s dysfunctional circuit-network system, as illuminated by the latest advances in neuroimaging. The goals of these studies are to help to identify the clinical pathways to recovery in depression as well as brain pathways fundamental to normal and abnormal emotional processing.

IDENTIFYING WHERE

In examining abnormal brain functioning in depression, all studies hope to determine which brain regions are most critical to the clinical syndrome. One approach is to develop experimental probes whereby the neural effects of controlled mood or cognitive stress can be examined in never-depressed subjects as well as depressed patients. In Dr. Mayberg’s view, the challenge is to identify relevant behavioral stressors to the system in order to unmask points of vulnerability that may be helpful in understanding the more complex clinical syndrome.

“This approach is designed to complement our studies where we look for regional differences between patients and healthy subjects in the resting state. Unfortunately this ‘blobology’ approach has been much maligned,” Dr. Mayberg said, “although it’s an absolute requisite starting point. But once you feel as though you’ve reached a comfort zone with brain changes that are well replicated, the question becomes, how do you look at them to determine how these ‘blobs’ actually interact with each other?”she added.

TREATMENT STUDIES

Describing a series of treatment studies, Dr. Mayberg and colleagues have identified a set of brain regions modulated by different types of antidepressant treatments. Interestingly, there does not appear to be a common pathway for recovery. “You do not get the same brain changes when people recover from medication as when they recover from cognitive therapy,” Dr. Mayberg reported.

She pointed to research comparing brain images of depressed patients treated with paroxetine, which showed increases in lateral cortices and suppression of limbic circuits in both the hippocampus and the subgenual cingulate, area 25, with brain images of depressed patients who underwent cognitive therapy, in which the turning down of lateral frontal cortices as well as medial frontal and orbital frontal cortices was evident. “Because of our past findings with placebo medication response involving area 25, we were quite surprised to see the absence of changes in this region with successful CBT. Equally surprising was that there were opposite changes in the hippocampus and frontal cortex with CBT relative to medication,” Dr. Mayberg said.

The finding of unique brain changes with each treatment is consistent with preclinical and psychological theories and findings. The subgenual cingulate region is high in serotonin transporter-binding sites, with strong projections to the brain stem and hypothalamus, making the region potentially critical for modulating widespread changes in behavior mediated by the cortex, Dr. Mayberg observed. Cognitive therapy, on the other hand, appears to affect a very different set of regions, targeting predominantly cortical brain regions such as the medial frontal and dorsal cingulate—regions primarily associated with attention, self-references, and reappraisal, and specifically targeted by this form of therapy.

DEPRESSION NETWORK MODEL

These treatment findings stress the complex interactions of a distributed set of both limbic and cortical brain regions. To better communicate these complex functional interactions, Dr. Mayberg and her team have organized converging findings across experiments into a working network model of depression. The inclusion of regions within the model attempts to incorporate the key clinical symptoms of depression while also considering the way in which emotions normally affect our thinking, memory, and actions.

“A useful model must provide a framework to understand both normal and abnormal emotional regulation,” Dr. Mayberg explained, “including differences among patient subtypes and mechanisms of different types of treatment.

“If we think of depression as a neural systems disorder, we can envision that different forms of treatment … are actually targeting different components of this network in different ways, and that should resonate with everyone who takes care of patients,” Dr. Mayberg observed.

A last set of studies was presented to emphasize this point, using a Structural Equation modeling and path analysis to look for predictive markers in scans taken prior to treatment that might differentiate patients who went on to do well on cognitive therapy from those who did well on paroxetine and those who benefited from neither approach. In the first group, the state of the network reflected the cross-talk among the orbital frontal, medial frontal, and cingulate regions, Dr. Mayberg explained. “In contrast, the paroxetine group model was uninfluenced by activity in the medial frontal and orbital frontal regions,” she said. “The state of the brain in these patients was dependent on variation in the hippocampus, area 25, and lateral prefrontal regions. If we look at the patients who were medication nonresponders, the network state was distinct from the other groups, with an emphasis on inputs to area 25 without modulation by cortex.

“These retrospective analyses are quite interesting in light of a new group of treatment-refractory patients where we are actually seeing overactivity of 25 in the resting state using more conventional forms of data analysis. Emerging research is beginning to suggest that the pretreatment brain state may provide important clues to understanding clinical and treatment-response variability across different depressed patient subgroups. What these more complex analyses also emphasize is that just looking at resting-state images either in an individual or in a group is interesting but doesn’t tell the whole story.”

This line of inquiry started at a point at which researchers had no idea what they would find and, in Dr. Mayberg’s words, “what we were asking the brains to show us. Once we started to see consistent patterns, we developed probes to try to see which patterns were most relevant to different patient subgroups.” That led to studies designed to understand the complicated cross-talk occurring among the affected regions. Ultimately, the end result of this research may be future classification schemes devised from a brain-based perspective, in which patients with different, individualized forms of depression will get the precise treatment they need. Dr. Mayberg added that “despite these interesting advances, we are nowhere near being able to just take a picture of the brain, look at it, and think we have something to say about how to treat people or not.”

These findings do, however, provide new incentive for future experimental studies designed to improve treatment outcome. “Deciding on a given treatment for an individual patient needs to be more evidenced based. It’s not about flipping a coin,” Dr. Mayberg concluded. “We seriously need to develop methods to identify the optimal treatment for a given individual patient, reducing the current trial and error approach and hopefully facilitating recovery more quickly.”

Negative and Positive

One observation yielded by functional MRI (fMRI) involved the manner in which depressed and nondepressed individuals process and respond to positive and negative stimuli. Dr. Mayberg cited an ongoing study that she participated in, conducted by Philippe Fossati, MD, PhD, of the Université Paris VI, in which emotional processing was analyzed not by experience but by how participants evaluated emotionally charged words and applied the terms to themselves. When Dr. Fossati and his colleagues compared brain images taken when participants evaluated the semantic meaning of words with those taken when participants applied the words to themselves, the investigators observed differences in activation of one key region of the brain—the medial frontal cortex. Surprisingly, the nondepressed controls had much higher activity in this region than did the depressed cohort when the words were positive, whereas the level of activity was comparable between groups when the words were negative. “You would think that depressed patients would have overattribution and have exaggerated responses to the negative,” Dr. Mayberg commented. “But in fact, the negative responses are really no different from those of the controls. The problem is that [depressed patients] have an attenuated response to the positive.”

—Fred Balzac

Suggested Reading
Fossati P, Hevenor SJ, Graham SJ, et al. In search of the emotional self: an fMRI study using positive and negative emotional words. Am J Psychiatry. 2003;160:1938-1945.
Goldapple K, Segal Z, Garson C, et al. Modulation of cortical-limbic pathways in major depression: treatment specific effects of cognitive behavioral therapy. Arch Gen Psych. 2004;61:34-41.
Keightley ML, Seminowicz DA, Bagby RM, et al. Personality influences limbic-cortical interactions during sad mood induction. Neuroimage. 2003;41:585-596.
Kruger S, Seminowicz D, Goldapple K, et al. Regional CBF in bipolar disorder: differences between remitted and depressed patients identified with an acute mood challenge. Biol Psychiatry. 2003;54:1274-1283.
Liotti M, Mayberg HS, McGinnis S, et al. Mood challenge in remitted unipolar depression unmasks disease-specific cerebral blood flow abnormalities. Am J Psych. 2002;159:1830-1840.
Mayberg HS. Modulating dysfunctional limbic-cortical circuits in depression: towards development of brain-based algorithms for diagnosis and optimised treatment. Br Med Bull. 2003;65:193-207.
Mayberg HS. PET imaging in depression: a neurosystems perspective. Neuroimaging Clin N Am. 2003;13:805-815.
Mayberg HS, Brannan SK, Mahurin RK, McGinnin S. Regional metabolic effects of fluoxetine in major depression: serial changes and relationship to clinical response. Biol Psychiatry. 2000;48:830-843.
Mayberg HS, Liotti M, Brannan SK, et al. Reciprocal limbic-cortical function and negative mood: converging PET findings in depression and normal sadness. Am J Psychiatry. 1999;156:675-682.
Seminowicz DA, Mayberg HS, McIntosh AR. Limbic-frontal circuitry in major depression: a path modeling meta-analysis. Neuroimage. 2004;22:409-418.

Return to table of contents