Temporal changes in brain volume and cognition in a randomized treatment trial of vascular dementia

overtime by magnetic resonance imaging (MRI) as part of a randomized treatm ent trial of vascular dementia. Methods. Participants who met criteria for vascular dementia Underwent comprehensive neurological and neuropsychologic al testing on entrance, during, and at completion of the 1-year study. For those centers who had easily available MRI, MRI of the brain was to be perf ormed on entry and completion of the study. Image analysis was performed on : all balanced and T2-weighted MR films to determine ventricular, sulcal, i schemic, and hemispheric brain volumes. Results, Of the 105 patients who me t the criteria for vascular dementia, 40 had a baseline MR[ study that met protocol requirements and was of excellent image quality. The baseline vent ricular volume in these 40 patients with high-quality MR correlated with, m ost measures of cognitive and behavioral function; including the total Alzh eimer's Disease Assessment Score (ADAS) (r = 0.51, P = .0024), as well as a ctivities of daily living (r = 0.61, P = .0002). The baseline ischemic brai n volume correlated well only with the gait and postural stability scale tr = 0.74, P = .009). Of the 40 participants, 25 had MRL studies at baseline and at completion of the study that were comparable and of excellent image quality. For these 25 patients, the mean ventricular volumes increased by 9 % over the study year (P = .001) and the mean ischemic brain volume increas ed by 18% (P = .01). Temporal changes in the sulcal: and nonischemic brain volume did not reach significance. None of the 14 clinical score measures c hanged significantly between baseline and completion of the study in these 25 patients. Conclusion. In summary, ventricular volume correlated well wit h cognitive measures in patients with vascular dementia and was a more sens itive marker for change during the study year than the clinical scales used in this study. This study also points out the practical limitations of bra in imaging as a surrogate measure of clinical outcome in multicenter random ized treatment trials of brain disease.