Digital Neurodegenerative Pathology After Repetitive Head Impacts

Boston University Medical Campus, Boston MA

Investigators

Abstract

Summary Millions of individuals who play contact sports or serve in the military are exposed to repetitive head impacts (RHI) and are at increased risk for Alzheimer’s disease (AD) and AD-related dementias, including chronic traumatic encephalopathy (CTE). CTE is a neurodegenerative tauopathy caused in part by RHI. Our studies show that the neuropathological substrate of RHI is not limited to tau, but is heterogeneous with beta-amyloid, pTDP-43, alpha-synuclein deposition, vascular changes, axonal and myelin loss, gliosis, and neuroinflammation also present. CTE has not been assessed in most brain banks and the number of identified cases is small for well-powered clinicopathologic and ‘omic studies in diverse populations. Further, the mechanisms by which RHI triggers neurodegeneration are unknown and discrepancies exist between symptoms in life and neuropathology at death, making diagnosis in life challenging. Recent advances in digital pathology have revolutionized our ability to comprehensively examine neurodegenerative pathology. Using novel computational algorithms, we can efficiently examine thousands of glass slides to identify previously unseen, novel neuropathologic changes. Currently, our team leads an NINDS funded effort to harmonize neuropathological and clinical data across multiple brain banks (UNITE, Framingham Heart Study, Boston University Alzheimer’s Disease Research Center). Across all brain banks, each brain is carefully characterized by expert neuropathologists, who diagnose and stage all forms of neurodegenerative disease, and clinicians who collect from family members and medical records detailed information about RHI exposure and clinical outcomes. With over 150,000 glass slides already digitized to generate whole slide images (WSI), we propose to leverage one of the largest neuropathology digital libraries in the world. Our overarching hypothesis is that computational algorithms can accurately identify CTE pathology, and can spatially pinpoint heterogeneous neuropathological features on WSIs associated with RHI exposure and with clinical symptoms commonly ascribed to patients exposed to RHI. Specifically, we propose to develop multimodal computational algorithms to identify the pathognomonic CTE lesion (Aim 1), to examine how RHI (including sport, military combat, duration of football play, cumulative head impact index, age of first exposure to RHI) and non-RHI (including genetic, substance, sleep, vascular) risk factors alter the regional distribution of heterogeneous neuropathologies. (Aim 2) and characterize how the regional distribution of heterogeneous neuropathologies contribute to clinical symptoms (including cognitive, functional, mood, behavior and motor symptoms) among donors with RHI exposure (Aim 3). We will utilize 19 tissue regions per case coupled with comprehensive staining, and gold standard neuropathological diagnoses. From this project, we will develop robust computational algorithms to find CTE in diverse brain banks, to discover RHI-related, neurodegenerative mechanisms and to identify clinicopathologic correlates to inform diagnosis in life.