Abstract TP083: The Brief Cognitive Scale for Detecting Cognitive Impairment in Intracranial Atherosclerotic Disease: a Biomarker-Correlated Scale

Background: Intracranial atherosclerosis (ICAS) related cognitive impairment has been increasingly recognized, but lacks specific diagnostic tools. We sought to develop and validate a brief cognitive scale for ICAS (BCoS-ICAS). Methods: Patients were enrolled from the Peking Union Medical College Hospital (training cohort, n=609) and 17 tertiary hospitals (validation cohort, n=218) between 2021 and 2025. All underwent comprehensive neuropsychological assessments, 3D-T1 MRI, and plasma biomarker analysis (GFAP, NfL, pTau217). The BCoS-ICAS was derived using a decision-tree machine learning algorithm and partial least squares analysis in the training cohort, then validated in the independent validation cohort. Linear regression models explored associations between the BCoS-ICAS and biomarkers. Results: The BCoS-ICAS includes the Calculation subtest, Auditory Verbal Learning Test (immediate recall and short-delay recall), and Trail Making Test Part A, administered within 15 minutes. It demonstrated superior diagnostic accuracy compared to MoCA and NINDS-CSN in the training cohort (sensitivity, 0.91 [95% CI 0.87 to 0.94]; specificity, 0.82 [95% CI 0.78 to 0.86]; and accuracy, 0.86 [95% CI 0.83 to 0.89]) and external validation cohort (sensitivity, 0.92 [95% CI 0.86 to 0.96]; specificity, 0.85 [95% CI 0.76 to 0.92]; and accuracy, 0.89 [95% CI 0.84 to 0.93]). Scales correlated with total white matter volume (natural algorithm, β 1.282, 95% CI 0.136 to 2.428), plasma GFAP (β -0.003, 95% CI -0.005 to 0.000), and NfL (β -0.004, 95% CI -0.008 to -0.001). Conclusions: The BCoS-ICAS is an efficient 15-minute scale for detecting ICAS-related cognitive impairment, with biological plausibility and potential value for monitoring disease.