Measuring atrophy in Alzheimer disease: a serial MRI study over 6 and 12 months.

BACKGROUND: Global brain atrophy rate calculated from serial MRI scans may be a surrogate marker of Alzheimer disease (AD) progression. Few studies have assessed atrophy in AD over short intervals. METHODS: Thirty-eight patients with AD and 19 control subjects had MRI scans at baseline, 6 months, and 1 year. Ventricular change and whole-brain volume loss were calculated directly from the regions manually outlined on registered scans and using the automated (boundary shift integral [BSI]) technique. Sample sizes required to power placebo-controlled treatment trials over 6 months and 1 year were calculated using these techniques. RESULTS: Increased rates of ventricular expansion and whole-brain atrophy were seen in AD compared with control subjects at both 6 and 12 months using manual and automated techniques (p < 0.001). Using the BSI consistently reduced measurement variability especially for whole-brain change. In clinical trials, at 6 months, significantly fewer patients would be required using the ventricular BSI (VBSI) compared with the brain BSI (BBSI) (e.g., 165 vs 410 per arm to provide 90% power to detect a 20% reduction in rate of change). At 1 year, sample size estimates were smaller than at 6 months, and the advantage of using VBSI instead of BBSI was less marked. CONCLUSIONS: In short-interval studies, using the ventricular boundary shift integral instead of the brain boundary shift integral may allow for disease-modifying effects to be demonstrated using significantly smaller sample sizes. This potential benefit must be balanced against the possibility that ventricular volumes may be more likely to be affected by factors other than neurodegeneration.