Diagnostic applications of tau PET

• Published in: Alzheimer's & dementia Vol. 17; no. S4
• Main Author: Rabinovici, Gil D.
• Format: Journal Article
• Language: English
• Published: 01.12.2021
• ISSN: 1552-5260; 1552-5279
• DOI: 10.1002/alz.049679

Background Tau PET is useful for in vivo diagnosis of Alzheimer’s disease (AD) and may be useful in detecting non‐AD tauopathies. Diagnostic applications vary by tracer and AD disease stage. Method This presentation will review existing literature on the diagnostic utility of tau PET tracers across the AD continuum and in non‐AD tauopathies. Data from in vitro (autoradiography, brain homogenate binding) and in vivo studies will be reviewed, with an emphasis on large clinical studies and PET‐to‐autopsy correlations. Result In vitro and in vivo data suggest that [18]FTP, [18F]MK6240, [18F]GTP‐1, [18F]RO948 and [18F]JNJ‐067 are AD‐selective tau tracers, while [18F]PI2620 and [18F]PM‐PBB3 (also known as [18F]APN‐1607) may bind to non‐AD tauopathies. FTP‐PET quantification shows high accuracy in differentiating AD dementia from other disorders. In a multi‐site study conducted in Lund, Seoul and UCSF (N=719), FTP SUVR in in a temporal meta‐ROI showed 89.9% sensitivity and 90.6% specificity for AD dementia versus non‐AD disorders. Sensitivity in Aβ+ MCI was lower (61.5%) with similar specificity (90.6%). RO948 PET in the Swedish BioFinder‐2 study (N=613) showed similar diagnostic performance, though sensitivity in Aβ+ MCI was lower (37.5%). Data from Mayo Clinic (N=1,281) revealed that Aβ‐PET is more sensitive than FTP‐PET for detecting the AD continuum in cognitively unimpaired (29% of subjects Aβ+ vs. 5% tau+) and MCI (55% Aβ+ vs. 30% tau+), while sensitivity was similar in clinical AD dementia (97% Aβ+ vs. 89% tau+). PET‐to‐autopsy data from the Avid A16 trial, Mayo Clinic and UCSF cohorts show high sensitivity and specificity for FTP‐PET visual reads and quantification in detecting Braak Stages V‐VI, but low sensitivity for Braak Stages I‐IV. Antemortem FTP binding in these cohorts showed poor correlation with non‐AD tau pathology. In a multi‐site German study, [18F]PI‐2620 showed 77% sensitivity and specificity in discriminating Progressive Supranuclear Palsy (N=60) from disease (N=20) and healthy (N=10) controls. Conclusion Tau PET is most useful in differentiating AD from other causes of dementia but is less sensitive in detecting preclinical and prodromal AD, which are associated with earlier Braak stages. While most tau PET tracers are AD‐specific, [18F]PI2620 and [18F]PM‐PBB3 warrant further study in non‐AD tauopathies.