Using functional brain imaging, Helen Mayberg, M.D., and colleagues at the University of Texas Health Science Center, San Antonio, have found increased activity in the cortex accompanied by decreases in limbic regions in patients who responded to either the popular antidepressant fluoxetine or to a placebo. They propose that this pattern of changes may be necessary for therapeutic response. However, patients who responded to fluoxetine also experienced unique changes in lower areas -- brainstem, striatum and hippocampus -- thought to confer additional advantage in sustaining the response long term and preventing relapse. The researchers report on their Positron Emission Tomography (PET scan) study in the May 2002 "American Journal of Psychiatry".

"Our findings do not support the notion that antidepressants work merely via a placebo effect," cautioned Mayberg, who has since moved to the Rotman Research Institute at the University of Toronto. "Patients on active medication who failed to improve did not sustain the brainstem, striatal and hippocampus changes unique to antidepressant responders."

In the randomized, double blind trial, 17 middle-aged men, hospitalized for unipolar depression, received either fluoxetine or placebo for 6 weeks. Rating scales revealed that 4 of the men responded to placebo and another 4 showed comparable improvement with the active medication. Nine patients failed to get better.

"Treatment with placebo is not absence of treatment, just absence of active medication," note the researchers, citing possible therapeutic benefits of a change in environment and the supportive therapeutic milieu of an inpatient psychiatric ward.

PET scans traced the destination of a radioactive form of glucose -- the brain's fuel -- to detect brain activity patterns. After 6 weeks, brains of men who responded to either treatment showed "remarkable concordance." Activity increased in prefrontal cortex, posterior cingulate, premotor, parietal cortex, and posterior insula. Activity decreased in parahippocampus, thalamus and hypothalamus.

Men who responded to fluoxetine, in addition, showed changes in certain lower brain areas -- brainstem, hippocampus, striatum and anterior insula. Brain areas activated in the fluoxetine responders were also somewhat larger. The brain stem and hippocampus appear to have important input in sustaining the cortical/limbic changes, suggest the researchers, who note that absence of changes in these lower brain areas in placebo responders may render them at higher risk for relapse, which several previous clinical studies have shown.

Although both placebo and antidepressant responders showed increased activity in the posterior cingulate (see graphic http://www.nih.gov/news/pr/may2002/placeboandtext.jpg)
at 6 weeks, this change had already occurred in placebo responders at 1 week. Together with other evidence, this suggests that the ability to increase activity in the posterior cingulate may be an early indicator of a brain's capacity to change and respond to treatment, says Mayberg. Medications that take a "bottom up" approach or non-drug, cognitive "top-down" interventions should work equally well. However, a need for progressively more aggressive treatments could signal "poor adaptive capacity" in the cortex/limbic network found to change in responders, say the researchers.