Premium
P3‐252: Optimizing Quantitative Strategies for TAU PET Imaging Across Different Radiotracers
Author(s) -
Seibyl John P.,
Barret Olivier,
Jennings Danna,
Marek Kenneth
Publication year - 2016
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1016/j.jalz.2016.06.1915
Subject(s) - voxel , nuclear medicine , tau pathology , partial volume , medicine , alzheimer's disease , radiology , disease
Background:Scintigraphic markers of tau deposition challenge simple quantitative analysis based on the heterogeneity of uptake patterns. Standard volume of interest strategies potentially underestimate brain tau by sampling in both normal and abnormal voxels within involved regions. Other challenges include identification of an optimized reference tissue, the need for partial volumeerror correction, off target binding affecting quantitation, and the Identification and correction of biases associated with SUVr quantitation methods for different tau radiotracers. Regarding this critical bias issue, the present investigation aims to identify practical non-invasive scan analyses resulting in more reliable quantitative assessment of brain tau burden. Methods:AV1451 or THK5351 PET imaging data were acquired in 4 young healthy controls (25-40 y; MMSE 30), 10 older healthy controls (50-75 y; MMSE 29-30), and13 Alzheimer’s patients (55-85 y; MMSE 14-29) with serial delayed imaging to determine the stability and signal:noise properties of dynamic SUVR measures using different sampling thresholds of SUVr voxel clusters within VOIs. Results:Dynamic SUVr time-activity data for key cortical regions demonstrate cut-off threshold dependent effects on time to plateau, absolute SUVr, and degree of separation between healthy volunteers and AD participants. Selection of different thresholds adjusts the amount of bias in SUVrs when compared to a gold standard quantitative estimate of regional tau deposition using full pharmacokinetic modeling with metabolite corrected arterial input function obtained for both tau radiotracers by our group in separate studies. Conclusions:Semi-quantitative outcome measures like SUVr for estimating brain tau with both AV 1451 and THK 5351 may be optimized by adjusting SUVr cut-off thresholds based onminimizing bias estimates referenced to tau measures based on fully modeled data sets. This is despite differences in pharmacokinetics, off target binding, and signal:noise characteristics of these radiotracers.