[18F]PI-2620 Tau PET signal across the aging and Alzheimer’s disease clinical spectrum

• Published in: Imaging neuroscience (Cambridge, Mass.) Mass.), 2024-10, Vol.2, p.1-16
• Main Authors: Young, Christina B., Vossler, Hillary, Romero, America, Smith, Viktorija, Park, Jennifer, Trelle, Alexandra N., Winer, Joseph R., Wilson, Edward N., Zeineh, Michael M., Sha, Sharon J., Khalighi, Mehdi, Yutsis, Maya V., Morales, Aimara P., Anders, David, Zaharchuk, Greg, Henderson, Victor W., Andreasson, Katrin I., Wagner, Anthony D., Poston, Kathleen L., Davidzon, Guido A., Mormino, Elizabeth C.
• Format: Article
• Language: English
• Subjects: [18F]PI-2620; aging; Alzheimer’s disease; tau PET
• ISSN: 2837-6056; 2837-6056
• DOI: 10.1162/imag_a_00329

[18F]PI-2620 is a second generation tracer that has shown high binding affinity for tau aggregation in Alzheimer’s disease (AD). However, [18F]PI-2620 signal in a large sample spanning the healthy aging and AD clinical spectrum as well as the stability of signal across different acquisition time windows has not yet been examined. Here, amyloid negative (Aβ-) cognitively unimpaired (CU; n = 49), amyloid positive (Aβ+) CU (n = 37), CU individuals with unknown amyloid status (n = 5), mild cognitive impairment (MCI; n = 14), dementia due to AD (n = 19), and non-AD neurodegenerative disorder (n = 54) participants were scanned with [18F]PI-2620 using a 45–75 min and/or 60–90 min acquisition time window. The impact of acquisition time on standardized uptake value ratio (SUVR) magnitude was first quantified with linear mixed models, and in participants and regions with high [18F]PI-2620 signal, SUVRs increased linearly up to 0.04 SUVR with each additional 5 min past injection time. We then accounted for differences in acquisition time using a voxel-wise correction approach and showed high correlations (all s ≥ 0.986) between SUVRs calculated from 45–75 min data and SUVRs from 60–90 min data that were interpolated to the 45–75 min scale in 15 participants who were scanned across both time windows. Using real and interpolated 45–75 min data, we next examined [18F]PI-2620 signal in Braak regions of interest and an off-target binding region (putamen) in Aβ- and Aβ+ CU, Aβ+ MCI, and Aβ+ AD dementia (n = 115) and showed that SUVRs in all Braak regions increased with greater disease severity. Within CU, higher Braak I SUVR was significantly associated with greater CSF pTau-181 (n = 35), and higher SUVRs were significantly associated with worse memory and language (n = 57). Thus, voxel-wise acquisition time corrections can be applied to combine [18F]PI-2620 datasets collected at different post-injection times, and once acquisition time is accounted for, [18F]PI-2620 signal shows the expected increases across the AD spectrum and can be used for detection of early tau elevations.