Quantifying tau deposition with the Parametric Estimation of Reference Signal Intensity: a 18F‐PI2620 tau‐PET and 4R‐tauopathy study

Abstract

AbstractBackgroundTau positron emission tomography (PET) can detect abnormal tau in neurodegenerative diseases. Conventional reference regions show high PET signal variability and are susceptible to contamination from partial‐volume effects of tracer uptake. The parametric estimation of reference signal intensity (PERSI) has been introduced as a new method to overcome those issues in patients with Alzheimer’s disease. The current study investigates its applicability on patients with underlying 4R‐tauopathy with the radiotracer 18F‐PI‐2620.MethodIn this preliminary, cross‐sectional analysis, 9 patients with clinically suspected 4R‐tauopathy underwent T1‐weighted MRI and 0‐60 min dynamic 18F‐PI‐2620 PET scanning. PERSI identified voxels within the white matter without specific binding for use as a reference region, which was compared with conventional anatomically defined reference regions. Time‐activity curves (TAC) and standardized uptake value ratios in the time window of 20‐40 min (SUVr20‐40) were generated for regions of interest (frontal lobe and globus pallidum). Data was analysed with FSL, PETSurfer, and R.ResultSUVr20‐40 in the globus pallidum were generally higher than in the frontal lobe for all reference regions (p < 0.001). Frontal lobe, but not globus pallidum SUVr20‐40 were lower when derived from PERSI than when using the cerebellum (p < 0.001) or pons (p < 0.001). When comparing TACs, the PERSI ROI did not show the rapid peak (perfusion) as seen in all other regions (both target and reference), resulting in an overestimation of SUVR in the early phases of the scan. However, activity was the same for all reference regions at later timepoints resulting in similar SUVr20‐40 measures in the globus pallidum.ConclusionPERSI derived reference regions do not provide an accurate estimate of early phase perfusion and as such are not appropriate for use as reference regions for reference tissue modelling of PET data. Nonetheless, PERSI provides a suitable reference region for late phase SUVR calculations which is particularly relevant for diseases and tracers were an anatomically defined reference region cannot be easily determined. Future analyses on a larger cohort will determine effect sizes and ROCs to see whether PERSI is a suitable semi‐quantitative investigation to compare between different methodologies and diagnostic groups.