A STUDY TO MODEL RATES OF CHANGE ON NEUROPSYCHOLOGICAL TEST MEASURES IN SUBJECTS DIAGNOSED WITH BEHAVIORAL VARIANT FRONTOTEMPORAL DEMENTIA AND HEALTHY SUBJECTS

Abstract

Information about the sensitivity of cognitive tests to frontotemporal dementia (FTD) natural progression is limited and highly variable since neuropsychological assessments in FTD natural history studies are typically performed yearly, using tests optimized for clinical description and diagnosis (Knopman DS, et al. Brain 2008;131(Pt 11):2957–68; Ranasinghe KG, et al. Neurology 2016;86(7):600–10). This study goal is to identify cognitive measures predictive of behavioral variant FTD (bvFTD) progression over 12 months or less, to estimate this effect size and its clinical meaningfulness by examining their relationship with changes in behavioral, functional, quality of life (QOL), and caregiver burden measures. 70 subjects (40 with early-stages, symptomatic bvFTD with their caregivers and 30 healthy volunteers, matched by age and education) will be enrolled in this observational study conducted at 10 FTD centers within United States. Subjects will undergo cognitive tests during 13 monthly visits (quarterly at site, otherwise at home). Tests were selected for their previously demonstrated sensitivity to change over short intervals, relative lack of learning and cultural effect, breadth of cognitive domains relevant to bvFTD. Frequent assessments should increase sensitivity to decline (relative to annual assessments) by improving slope estimates. BvFTD subjects and caregivers will also complete behavioral, functional, QOL and caregiver burden assessments, where relationships with cognitive measures will be investigated. Measures requiring expert clinicians (Modified Clinical Dementia Rating, Neuropsychiatric Inventory, Clinical Global Impression of Change) will only be conducted at quarterly site visits. Contributing blood samples for DNA/RNA and biomarker explorations is voluntary. The minimal “standardized effect size” that can be detected in this study (considering 20% dropout) is 0.76 standard deviation units with 80% power and 0.05 significance level. For each cognitive outcome measure and composite measure, mixed-effects regression will be used to model both subject-specific and group-level slopes and estimate their within- and between-subject variance components. The study is currently in the start-up phase. Results will be presented at study completion. Identification of cognitive measures that are clinically meaningful, brief, sensitive to change, and predictive of bvFTD progression over short time would greatly benefit bvFTD drug development.