Zhang, QiQiZhangLu, JiayingJiayingLuWang, LuyaoLuyaoWangJiao, FangyangFangyangJiaoWang, MinMinWangShi, KuangyuKuangyuShi0000-0002-8714-3084Zuo, ChuantaoChuantaoZuoJiang, JiehuiJiehuiJiang2025-09-082025-09-082025-08-28https://boris-portal.unibe.ch/handle/20.500.12422/217584Introduction Tau protein aggregation is a hallmark of Alzheimer's disease (AD) pathology. Semi-quantitative analysis using regions of interest (ROIs)-based standardized uptake value ratios (SUVRs) serves as a major Tau positron emission tomography (PET) biomarker for AD diagnosis and staging. This study aims to evaluate the diagnostic performance of the second-generation tau tracer 18F-Florzolotau, including the impact of semi-quantitative reference region and ROIs methodology and partial volume error (PVE) correction. Data from the FDA-approved tracer 18F-Flortaucipir provide benchmark context, aiming to evaluate the performance.Methods A total of 842 participants from two cohorts underwent tau PET imaging with either 18F-Flortaucipir (n = 741) or 18F-Florzolotau (n = 101). The 18F-Flortaucipir cohort contains 384 normal controls, 292 patients with mild cognitive impairment, and 65 AD dementia. The 18F-Florzolotau cohort contains 27 normal controls, 26 patients with mild cognitive impairment and 48 AD dementia. SUVRs were calculated across four ROIs using six semi-quantitative methods varying by reference region and PVE-correction application. Diagnostic performance was assessed using the area under the curve (AUC) from receiver operating characteristic analysis. Partial correlations between SUVRs and clinical severity were evaluated.Results 18F-Florzolotau demonstrated high diagnostic accuracy (AUC: 0.96-0.98) for AD dementia and strong clinical correlations (|r|= 0.61-0.74). Performance varied with semi-quantitative methodology. The optimal approach used inferior cerebellar gray matter as the reference region, achieving the highest AUC and strong clinical correlations for 18F-Florzolotau. Results for 18F-Flortaucipir (AUC: 0.78-0.87; |r|= 0.29-0.45) provided consistent methodological insights supporting the choice of inferior cerebellar gray methodology.Conclusions 18F-Florzolotau shows excellent diagnostic performance for AD dementia. The semi-quantitative methodology impacts results, with inferior cerebellar gray as the recommended reference region for optimizing 18F-Florzolotau SUVR analysis in AD dementia. These findings support the clinical utility of 18F-Florzolotau tau PET in AD.en18F-Flortaucipir18F-FlorzolotauAlzheimer’s diseaseSemi-quantitative methodTau PET600 - Technology::610 - Medicine & healtharticle10.48620/911764087742310.1007/s00415-025-13342-3