BACKGROUND: Late-life depressive disorders have been linked to cerebrovascular disease (the vascular depression hypothesis). Treatment resistance may be associated with vascular-based lesions in the white matter and basal ganglia. Virchow-Robin spaces (VRS) are cerebrospinal fluid spaces associated with microangiopathy of small cerebral vessels. This study tested the hypothesis that dilation of Virchow-Robin spaces seen on Magnetic Resonance Imaging (MRI) is associated with treatment resistance in elderly depressed individuals. METHODS: 50 patients with late-onset (age over 60 years) major depressive disorder (29 responders to monotherapy, 21 non-responders to monotherapy) and 35 normal volunteers were recruited. Assessment of deep white matter lesions [WML] and periventricular hyperintensities [PVH] (both with the Scheltens rating scale score, [Scheltens, P., Barkhof, F., Leys, D., et al. (1993) A semiquantative rating scale for the assessment of signal hyperintensities on magnetic resonance imaging. J Neurol Sci;114(1):7-12.]) and the severity of VRS dilatation (using a new scale) were scored from MRI images. Statistical group comparisons and multiple regression analyses were performed to quantify the relationship between imaging features and clinical outcome. RESULTS: There was a trend for greater WML Scheltens scores in the monotherapy resistant group compared to responders and control subjects, but only using the basal ganglia VRS score was there a statistically significant difference. A score of 2 or greater on the VRS score was 80% sensitive and 62% specific in predicting non-response to antidepressant monotherapy. The VRS score accounted for 38% of the variance in the multiple regression model and the PVH score, which was an independent predictor of outcome, contributed another 6%. LIMITATIONS: Numbers are small and type II errors possible, especially for the Scheltens ratings. Treatment response was limited to response or non-response to monotherapy and was retrospectively derived. The VRS scale was originally designed for use in patients with vascular dementia and has not been used before in affective disorders. Although all depressed subjects were late-onset, it is possible that depression led to vascular disease rather than vice versa. CONCLUSION: VRS dilatation is common in diseases associated with microvascular abnormality, which is the presumed basis of vascular depression in the elderly. VRS score may be useful in determining which patients are less likely to respond to antidepressant monotherapy. Prospective studies of patients with a wider range of treatment responses are indicated. ; Journal Article Research Support, Non-U.S. Gov't Netherlands