Off-target binding of 18F-MK-6240 tau PET imaging demonstrated by high spatial resolution dedicated brain PET scanner

OBJECTIVE: To verify the regional distribution, especially off-target binding, of the tracer 18F-MK-6240, on tau PET images obtained with a high-resolution dedicated brain PET scanner. METHODS: We retrospectively analyzed the 18F-MK-6240 PET data from subjects with cognitive decline who were enrolled in the BATON study. All scans were acquired using the same high-resolution brain-dedicated PET system. We calculated the standardized uptake value ratios (SUVRs) by using manually defined restricted cerebellar volumes of interest (VOIs) excluding meningeal spill-in, and we compared these SUVRs with those obtained using the conventional CenTauR cerebellar gray matter reference. We also quantified the off-target uptake in regions including the meninges, retina, and ethmoid sinus. Group comparisons and interaction effects were analyzed by t-tests and an ANOVA, and ROC analyses were performed to assess the diagnostic accuracy of each method. RESULTS: Off-target accumulations were clearly observed in the substantia nigra, globus pallidum, and pineal gland. Higher extra-cerebral cortical uptake was observed in the meninges and centrum semiovale in tau-positive individuals, although these differences did not remain significant after a multiple comparison correction. The SUVRs derived with the use of restricted cerebellar VOIs were significantly higher than those from the CenTauR cerebellar VOIs in both the tau-positive and tau-negative groups (p < 0.01). A significant interaction between the diagnostic group and the reference VOI choice was detected (p = 0.019). The ROC analysis showed excellent diagnostic performance, with AUCs of 0.930 for CenTauR and 0.984 for the restricted reference VOIs (a nonsignificant difference). CONCLUSION: 18F-MK-6240 PET imaging using a high-resolution brain-dedicated PET scanner revealed notable off-target binding and meningeal spill-in, which significantly affect quantification depending on the reference VOI selection. Optimizing reference regions by excluding extracerebral contamination enhances the reliability of tau PET analyses and may be crucial for longitudinal studies and multicenter trials.