Heterogeneous distribution of tau pathology in the behavioural variant of Alzheimer’s disease
Ellen H. Singleton, Oskar Hansson, Yolande A.L. Pijnenburg, Renaud La Joie, William G. Mantyh, Pontus Tideman, Erik Stomrud, Antoine Leuzy, Maurits Johansson, Olof Strandberg, Ruben Smith, Evi Berendrecht, Bruce L. Miller, Leonardo Iaccarino, Lauren Edwards, Amelia Strom, Emma E. Wolters, Emma M. Coomans, Denise Visser, Sandeep S.V. Golla, Hayel Tuncel, Femke H. Bouwman, John C. van Swieten, Janne M. Papma, Bart van Berckel, Philip Scheltens, Anke A. Dijkstra, Gil D. Rabinovici, Rik Ossenkoppele
Abstract
OBJECTIVE: The clinical phenotype of the rare behavioural variant of Alzheimer's disease (bvAD) is insufficiently understood. Given the strong clinico-anatomical correlations of tau pathology in AD, we investigated the distribution of tau deposits in bvAD, in-vivo and ex-vivo, using positron emission tomography (PET) and postmortem examination. METHODS: F]RO948 PET. We converted tau PET uptake values into standardised (W-)scores, adjusting for age, sex and mini mental state examination in a 'typical' memory-predominant AD (n=205) group. W-scores were computed within entorhinal, temporoparietal, medial and lateral prefrontal, insular and whole-brain regions-of-interest, frontal-to-entorhinal and frontal-to-parietal ratios and within intrinsic functional connectivity network templates. For the postmortem study, the percentage of AT8 (tau)-positive area in hippocampus CA1, temporal, parietal, frontal and insular cortices were compared between autopsy-confirmed patients with bvAD (n=8) and typical AD (tAD;n=7). RESULTS: Individual regional W-scores ≥1.96 (corresponding to p<0.05) were observed in three cases, that is, case #5: medial prefrontal cortex (W=2.13) and anterior default mode network (W=3.79), case #2: lateral prefrontal cortex (W=2.79) and salience network (W=2.77), and case #7: frontal-to-entorhinal ratio (W=2.04). The remaining four cases fell within the normal distributions of the tAD group. Postmortem AT8 staining indicated no group-level regional differences in phosphorylated tau levels between bvAD and tAD (all p>0.05). CONCLUSIONS: Both in-vivo and ex-vivo, patients with bvAD showed heterogeneous distributions of tau pathology. Since key regions involved in behavioural regulation were not consistently disproportionally affected by tau pathology, other factors are more likely driving the clinical phenotype in bvAD.