White Matter Lesion Progression in LADIS
Frequency, Clinical Effects, and Sample Size Calculations
Background and Purpose—White matter lesion (WML) progression has been advocated as a surrogate marker in intervention trials on cerebral small vessel disease. We assessed the rate of visually rated WML progression, studied correlations between lesion progression and cognition, and estimated sample sizes for clinical trials with pure WML progression vs combined WML progression-cognitive outcomes.
Methods—Those 394 participants of the Leukoaraiosis and Disability Study (LADIS) study with magnetic resonance imaging scanning at baseline and 3-year follow-up were analyzed. WML progression rating relied on the modified Rotterdam Progression Scale. The Vascular Dementia Assessment Scale global score and a composite score of specific executive function tests assessed longitudinal change in cognition. Sample size calculations were based on the assumption that treatment reduces WML progression by 1 grade on the Rotterdam Progression Scale.
Results—WML progression related to deterioration in cognitive functioning. This relationship was less pronounced in subjects with early confluent and confluent lesions. Consequently, studies in which the outcome is cognitive change resulting from treatment effects on lesion progression will need between 1809 subjects per treatment arm when using executive tests and up to 18 853 subjects when using the Vascular Dementia Assessment Scale score. Studies having WML progression as the sole outcome will need only 58 or 70 individuals per treatment arm.
Conclusions—WML progression is an interesting outcome for proof-of-concept studies in cerebral small vessel disease. If cognitive outcome measures are added to protocols, then sample size estimates increase substantially. Our data support the use of an executive test battery rather than the Vascular Dementia Assessment Scale as the primary cognitive outcome measure.
- Received May 7, 2012.
- Accepted July 3, 2012.
- © 2012 American Heart Association, Inc.