Genetically identical twins are highly similar in levels and spatial distribution of tau pathology: A [18F]flortaucipir PET study

Abstract

AbstractBackgroundTau accumulation starts during the preclinical phase of Alzheimer’s disease (AD) and is closely associated with cognitive decline. For preventive purposes, it is important to identify factors associated with tau accumulation and spread. Studying genetically identical twin‐pairs may give insight into genetic and environmental contributions to tau accumulation. Therefore, we assessed within‐pair similarities in 1) levels of tau and 2) spatial distribution of tau, within a sample of cognitively intact genetically identical twins.MethodFifty‐four identical twins (27 pairs) from the EMIF‐AD PreclinAD cohort underwent dynamic dual time‐point (0‐30min and 80‐100min) [18F]flortaucipir‐PET. Binding potential (BPND) (reference region: cerebellar gray) was extracted for entorhinal (Braak‐I), temporal (Braak‐III‐IV), neocortical (Braak‐V‐VI) and global (Braak‐I‐VI) regions. Additionally, for voxel‐by‐voxel analyses, PET‐images were normalized to MNI‐space. Within twin‐pair similarities in global and regional levels of [18F]flortaucipir BPND were calculated using intra‐class correlations () and compared to correlations estimated for 20 sets of 27 random (non‐twin) pairs. To examine within‐pair similarities in [18F]flortaucipir spatial distribution, each participant’s voxel‐by‐voxel [18F]flortaucipir spatial distribution was correlated to that of every other participant (). A t‐test was performed to assess whether twin‐pairs showed higher correlation coefficients (i.e. more similar spatial distribution) compared to random pairs.ResultDemographics are summarized in . Visual inspection of [18F]flortaucipir‐PET revealed both similarities and dissimilarities within pairs (). Global [18F]flortaucipir BPND was significantly correlated within pairs (r=0.46; p=0.01). Regionally, twin‐pair correlations for [18F]flortaucipir BPND was strongest in the entorhinal region (r=0.61; p<0.01), and non‐significant in the temporal region (). The correlation coefficient observed for twin‐pairs exceeded the range of correlation coefficients observed for random pairs for the entorhinal, neocortical and global region (). Examination of within‐pair similarities in voxel‐by‐voxel [18F]flortaucipir distribution showed that the correlations were higher for twin‐pairs (mean r=0.57; SD=0.06) than for random pairs (mean r=0.38; SD=0.10) (p<0.01) ().ConclusionGenetically identical twin‐pairs show substantial similarities in levels and spatial distribution of tau, suggesting genetic factors to play an important role. However, clear visual dissimilarities were also observed, indicating a role for non‐genetic factors as well. Unravelling mechanisms underlying within‐pair differences may aid in the development of preventive strategies for AD.