DOI 10.1016/j.jarlif.2026.100081
Evaluating inclusion of continuous multivariable cognitive scores for operational enrichment of preclinical Alzheimer's disease trials: a retrospective emulation study.
Аннотация
BACKGROUND
Prevention trials for Alzheimer's Disease face significant challenges due to the slow and uncertain rate of cognitive decline in asymptomatic, amyloid-positive individuals. Biomarker positivity alone does not guarantee clinically meaningful progression, often leaving studies underpowered.
OBJECTIVES
Evaluate model-based operational trial enrichment within an amyloid-positive, cognitively normal subgroup.
METHODS
Trained on a small cohort from the National Alzheimer's Coordinating Center (N = 113), with external validation using baseline data of a similar small cohort from the Alzheimer's Disease Neuroimaging Initiative (N = 161). Gradient-boosted decision trees estimated progression risk and projected power under alternative enrollment strategies. Predictors included neuropsychological scores, demographics, medical history, and ApoE ε4 genotype. The primary outcome was clinical progression to cognitive impairment within 1- and 3-year follow-up windows.
RESULTS
In the independent testing set, the model successfully increased the effective prevalence of the outcome. The Positive Predictive Value for the 1-year window was 0.27 (a 35% relative increase over the 0.20 baseline prevalence) and 0.43 for the 3-year window (a 23% relative increase over the 0.35 baseline). Power simulations for an emulated trial (N = 500) demonstrated that this enrichment strategy consistently increased statistical power, potentially reducing the required sample size to achieve 80% power by up to 32% for small treatment effect sizes.
CONCLUSIONS
This operational methodology functions as a screening filter to optimize event rates within a highly specific, preselected trial population.
Подготовка и перевод публикации
Обработка статьи · этапы, ошибки и управление
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Тип публикации указан по данным источника. Уровень доказательности автоматически не присваивается.