An optimized reference tissue method for quantification of tau protein depositions in diverse neurodegenerative disorders by PET with ¹⁸F-PM-PBB3 (¹⁸F-APN-1607)

ABSTRACT Positron emission tomography (PET) with 18 F-PM-PBB3 ( 18 F-APN-1607) enables high-contrast detection of tau depositions in various neurodegenerative dementias, including Alzheimer’s disease (AD) and frontotemporal lobar degeneration (FTLD). A simplified method for quantifying the radioligand binding in target regions is to employ the cerebellum as a reference (CB-ref) on the assumption that the cerebellum has minimal tau pathologies. This procedure could be valid in AD, while FTLD disorders exemplified by progressive supranuclear palsy (PSP) are characterized by occasional tau accumulations in the cerebellum, hampering the application of CB-ref. The present study aimed to establish an optimal method for defining reference tissues on 18 F-PM-PBB3-PET images of the AD and non-AD tauopathy brains. We developed a new algorithm to extract reference voxels with a low likelihood of containing tau deposits from gray matter (GM-ref) or white matter (WM-ref) by a bimodal fit to an individual, voxel-wise histogram of the radioligand retentions and applied it to 18 F-PM-PBB3-PET data obtained from age-matched 40 healthy controls (HCs) and 23 AD, 40 PSP, and five other tau-positive FTLD patients. PET images acquired at 90-110 min after injection were averaged and co-registered to corresponding magnetic resonance imaging space. Subsequently, we generated standardized uptake value ratio (SUVR) images estimated by CB-ref, GM-ref and WM-ref respectively, and then compared the diagnostic performances. GM-ref and WM-ref covered a broad area in HCs and free of voxels located in regions known to bear high tau burdens in AD and PSP patients. GM-ref allowed the most robust separation of AD and PSP patients from HCs according to the area under the curves in receiver operating characteristic curve analyses. GM-ref also provided SUVR images with higher contrast than CB-ref in FTLD patients with suspected and confirmed corticobasal degeneration. The methodology for determining reference tissues as optimized here reinforces the accuracy of 18 F-PM-PBB3-PET measurements of tau burdens in a wide range of neurodegenerative illnesses.