Enlargement of Brain Cerebrospinal Fluid Spaces as a Predictor of Poor Clinical Outcome in Melancholia

Background: A number of recent neuroimaging findings in depression have provided new insight into the biological substratum of depressive illness. The question now is what particular relevance the structural brain alteration described may have within the clinical context of depressive patients. We investigated a possible relationship between brain cerebrospinal fluid (CSF) space changes and patient prognosis in melancholic depression.

Method: Fifty-five patients who met DSM-IV criteria for major depressive disorder with melancholic features were examined with 3-dimensional magnetic resonance imaging, and CSF volumes were measured for global brain CSF and for lateral ventricles and left and right sylvian fissure regions. Clinical outcome was prospectively assessed during a 6-month standardized antidepressive treatment period (Phase I) and in a 2-year follow-up (Phase II) of recovered patients. The outcome measurements were total days to symptom remission (Phase I) and to eventual symptom relapse or recurrence (Phase II). The study took place from July 1998 to Dec. 2001.

Results: Phase I: Enlargement of CSF spaces in the left sylvian fissure region predicted poor treatment response. Volume measurements from this region accounted for 35% of remission time variance. Median time to full clinical remission was 82 days in patients with severe changes, 51 days in the case of mild-to-moderate CSF enlargement, and 35 days in patients with no left sylvian fissure region alterations. Phase II: Severe enlargement of global cortical CSF spaces was associated with increased risk of depression relapse or recurrence. Patients with severe cortical CSF changes showed a 7.8-fold excess risk of depression relapse/recurrence compared with patients with no cortical CSF space alteration.

Conclusion: Our data suggest that MRI-detected CSF space enlargement may be an important neuroimaging marker for poor prognosis in melancholic depression.