Derivation of a tau‐PET based pathology accumulation index to estimate global tau‐load: Neuroimaging / New imaging methods. (7th December 2020)

• Record Type: Journal Article
• Authors: Leuzy, Antoine; Lilja, Johan; Ossenkoppele, Rik; Palmqvist, Sebastian; Janelidze, Shorena; Stomrud, Erik; Strandberg, Olof; Smith, Ruben; Hansson, Oskar
• Is Part Of: Alzheimer's & dementia. Volume 16(2020)Supplement 4
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-12-07
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.045966
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: As cortical tau pathology in Alzheimer's disease (AD) has been shown to correlate closely to neurodegeneration and onset of cognitive impairment, there is a growing interest in the further refinement of tau‐PET methods for clinical use. Similar to amyloid‐PET, tau‐imaging will likely be approved for visual‐read supported by quantification (i.e. standardized uptake value ratio [SUVR]). Recently, we showed that a spatial normalization‐based weighting‐parameter (w, termed Aβ‐index) derived from amyloid‐PET was highly correlated to SUVR. As this measure carries several advantages over SUVR (e.g. no need to define ROIs and no need for an MRI), we here aimed to extend this method to tau‐PET. Method: Preliminary findings are based on 245 subjects with and without cognitive impairment scanned with [ 18 F]flortaucipir as part of the Swedish BioFINDER study. The relationship between w (termed Tau‐index) and SUVR (temporo‐parietal target‐ROI (Vogels et al., 2018) and inferior cerebellar cortex as reference) was assessed via linear regression and coefficient‐of‐determination (R 2 ). Additional analyses are ongoing with a larger cohort (AVID AO5 and Expedition‐3), including cases with antemortem [ 18 F]flortaucipir and post‐mortem tau immunohistochemistry for neurofibrillary tangle pathology. Findings will also be validated in BioFINDER‐2 (n=800, scanned with the second‐generation tau tracer [ 18 F]RO948). Both cohorts will also include longitudinal data. Diagnostic performance of Tau‐index and SUVR will be assessed against visual‐read and post‐mortem tau‐histopathology using area‐under‐the‐curve (AUC) values from receiver‐operating characteristic analyses. These results will be presented at AAIC2020. Result: Strong associations were observed between [ 18 F]flortaucipir Tau‐index and SUVR (R 2 =0.933, P<0.001;Figure 1). No significant difference was seen between AUC‐values from Tau‐index and SUVR for the separation of AD from CU controls (Tau‐index=0.913, SUVR=0.920;Figure 2). Conclusion: The proposed Tau‐index is simpler to implement clinically and may also facilitate the comparison of tau‐PET across different tracers/processing approaches. The proposed Tau‐index showed a clear association to SUVR and carries an advantage over the latter in that it does not require ROI definition nor the use of an MRI, which is important considering that many elderly subjects have contraindications for MRI. The findings require replication in the larger data sets as well as against neuropathology, which will be presented at the conference.