Jpn J Radiol 2025[Dec]; ? (?): ? PMID41369878show ga
PURPOSE: We aimed to compare visual and quantitative assessments between PET/CT at 90 min and sequential PET/MRI at approximately 120 min after injection of (18)F-flutemetamol, and to investigate factors affecting differences in standardized uptake value ratio (SUVr) and Centiloid scale. MATERIALS AND METHODS: Eighty-three participants underwent both PET/CT and sequential PET/MRI. Two nuclear medicine physicians performed visual interpretations. SUVr was calculated using four reference regions-pons, whole cerebellum, cerebellar gray matter (CGM), and whole cerebellum and brainstem-and compared between PET/CT and PET/MRI. Centiloid Scale was also compared between PET/CT and PET/MRI. Subgroup analyses were conducted based on Centiloid scale. Associations between scan interval, age, and amyloid burden were evaluated using stepwise regression. Cerebrospinal fluid (CSF) biomarkers were compared with imaging findings in 52 participants. Additionally, six participants underwent dynamic PET/MRI at 0, 60, 90, and 120 min post-injection. RESULTS: Visual interpretation showed high agreement between PET/CT and PET/MRI (kappa = 0.97), and 98% concordance with CSF findings. SUVr and Centiloid scale demonstrated a strong intraclass correlation coefficient of 0.96-0.98. However, SUVr and Centiloid scale were significantly higher from PET/CT than from PET/MRI, except for SUVr using CGM. Age and Centiloid scale were significant predictors of modality differences. Scan interval was also significant when CGM was used as the reference region. Dynamic PET/MRI revealed time-dependent increases in Centiloid scale among amyloid-positive participants. High concordance in visual and quantitative assessments was seen for (18)F-flutemetamol PET/CT (90 min) and subsequent PET/MRI (120 min). While diagnostic agreement was preserved, quantitative values were influenced by amyloid burden, age, and scan timing. These findings suggest a need for careful consideration when interpreting quantitative metrics across different imaging modalities and time points.