INDIVIDUAL-BASED NETWORK ANALYSIS: A NOVEL APPROACH TO INVESTIGATE THE PATHOLOGICAL SPREAD OF PHF TAU USING GRAPH THEORY IN CLINICAL AND PRECLINICAL STAGES OF ALZHEIMER’S DISEASE

Hillary Protas,Dhruman D. Goradia,Valentina Ghisays,Ji Luo,Paul S. VanGilder,Pradeep Thiyyagura,Michael Malek-Ahmadi,Wendy Lee,Yinghua Chen,Vivek Devadas,Robert Bauer,Susan M. Landau,Michael W. Weiner,William J. Jagust,Kewei Chen,Yi Su,Eric M. Reiman
DOI: https://doi.org/10.1016/j.jalz.2019.06.4288
2019-01-01
Abstract:The spread of paired helical filaments(PHF) tau in the brain is currently under investigation. Different groups suggest tau could be either propagating from a seed or an imbalance in local production and clearance that leads to increased deposition. Studies have used resting state fMRI with tau PET to understand the spread of tau in humans. In this study, we introduce a new subject-based measure of tau PET connectivity that describes the spread of tau using only Flortaucipir(FTP)PET images. Here we demonstrate how FTP PET network measurements of tau spread differentiates probable Alzheimer's dementia(pAD), mild cognitive impairment(MCI), and cognitively unimpaired(CU) study participants from the Alzheimer's Disease NeuroImaging Initiative(ADNI), and suggest its promise to detect differences in network measures in CU persons at genetic risk for AD. We determined a FTP network for each subject using a method from graph theory. From the undirected weighted graph created from the inter-regional distances, the global and local (efficiency, network strength) measurements of spread of tau in the network were calculated. Next, we compared our measurements to SUVR in common regions such as prespecified ROIs (entorhinal, parahippocampal, inferior temporal, precuneus) as well as mean cortical region from Mayo and partial volume corrected(PVC)SUVR BraakI-II, BraakIII-IV, and BraakV-VI from Berkeley in their ability to distinguish 15 pAD, 63 MCI, and 81 CU research participants from one another. Finally, we compared these novel and established measurements in their ability to distinguish in CU 49 APOE4 carriers from 24 non-carriers. Like several of the established ROI flortaucipir PET measurements, each of the global and several of the local flortaucipir connectivity measurements were significantly different in the three subject groups(pAD>MCI>CU,p
What problem does this paper attempt to address?