Discordance Between Plasma P-tau217 And PET Imaging As A Prognostic Marker In Alzheimer’s Disease:

Author Block:Hanna Cho¹, Han-Kyeol Kim², Jae Hoon Lee³, Joong-Hyun Chun⁴, You Jin Kim¹, Mina Park⁵, Rayeon Ha⁶, Daniel Connell⁷, Joel Braunstein⁷, Young Hoon Ryu³, Chul Hyoung Lyoo¹.
¹Department of Neurology, Gangnam Severance Hospital, Yonsei University College of Medicine, Seoul, Korea, Republic of, ²Department of Neurology, Wonju Severance Christian Hospital, Yonsei University Wonju College of Medicine, Wonju, Korea, Republic of, ³Department of Nuclear Medicine, Gangnam Severance Hospital, Yonsei University College of Medicine, Seoul, Korea, Republic of, ⁴Department of Nuclear Medicine, Severance Hospital, Yonsei University College of Medicine, Seoul, Korea, Republic of, ⁵Department of Radiology, Gangnam Severance Hospital, Yonsei University College of Medicine, Seoul, Korea, Republic of, ⁶Department of Psychiatry, Seoul Bukbu Hospital, Seoul, Korea, Republic of, ⁷C2N Diagnostics, LLC, St Louis, MO, USA.

Abstract:
Background: To investigate the impact of discordance between plasma biomarkers and PET imaging on longitudinal cognitive outcomes in Alzheimer’s disease.
Methods: A total of 232 participants were enrolled, all of whom underwent ¹⁸F-florbetaben (FBB), ¹⁸F-flortaucipir (FTP) PET, volumetric MRI, and neuropsychological assessments. Their plasma samples were tested for plasma p-tau217 at C2N Diagnostics. Plasma-PET discordance was defined using linear regression between plasma p-tau217 and FBB Centiloid values or FTP temporal meta-ROI SUVR. Standardized residuals were categorized into lower (L), middle (M), and upper (U) discordance groups based on ±1 SD thresholds. Longitudinal cognitive change was analyzed using linear mixed-effects models with random intercepts and slopes, adjusting for demographic factors, APOE ε4 status, and baseline PET burden.
Results: Based on linear regression residuals using FBB amyloid PET Centiloid values, participants were classified into Amyloid-L (N = 27), Amyloid-M (N = 161), and Amyloid-U (N = 44) groups. Using FTP temporal meta-ROI tau PET SUVR, participants were similarly categorized into Tau-L (N = 21), Tau-M (N = 181), and Tau-U (N = 30) groups. In linear mixed-effects models, amyloid discordance was not associated with subsequent cognitive decline. In contrast, tau discordance was significantly associated with longitudinal cognitive deterioration, with individuals showing higher tau PET burden relative to plasma p-tau217 demonstrating accelerated decline over time (MMSE: β = -2.077, p < 0.001; CDR-SB: β = 2.188, p < 0.001).
Conclusion: Individuals with disproportionately higher tau tangle burden relative to plasma p-tau217 exhibited more severe cognitive decline, indicating that plasm-PET tau discordance may serve as a prognostic biomarker for future cognitive deterioration.