Objectives
We reported the BPIT plot whose independent variable was time as a novel variation of the graphical analysis (NeuroImage 22: T81). In this study, we compared BPIT with a linear version of the simplified reference tissue method (SRTM; Ichise et al., 2003) and reference tissue graphical analysis (RTGA; Logan et al. 1996) in dopamine-related ligands of varying dissociation levels.
Methods
Healthy control subjects were studied with PET following a bolus injection of [11C]raclopride, a DA D2/D3 receptor ligand (n=21), [11C]methylphenidate (n=21), a DA membrane transporter ligands, or [11C]DTBZ, a VMAT-2 ligand (n=15). Radioactivity time-profiles were obtained for anterior (ant-) and posterior (post-) putamen (Pu) and caudate nucleus (CN), ventral striatum (VS), and cerebellum (Cb). BPIT utilized the plateau portion (t0-90 min) of the following plot: w1·A(T) + w2· A(t)dt where w1 and w2 are constants and A(T) represents the radioactivity in a region. The time t0 was determined on each ligand as the first frame after the slowest peak among all striatal regions across all subjects. The ligand-specific t0 values were used in variations of SRTM and RTGA where the reference region terms were excluded from respective operational equations (SRTM-R and RTGA-R). In variations including the reference region terms (SRTM+R and RTGA+R), reported circulation times were used while the brain-to-blood clearance rate constant of the reference regions (k2R) was estimated per subject (SRTM+R) or set to reported values (RTGA+R). The differences between BP estimates of two approaches were expressed by percent deviation: (BP1 − BP2)/(BP1 + BP2)/2·100, where subscripts 1 and 2 denote BP estimates of two approaches to compare.
Results
Estimates of t0 were 30, 50, and 40 min for [11C]rac, [11C]MP, and [11C]DTBZ, respectively. Using ligand-specific t0, a no slope model described the BPIT plot better than a linear model in more than 96% regions (i.e., the BPIT plot reached a plateau). Percent deviations between the approaches are listed in the Table 1. BPIT, STRM, and RTGA yielded essentially identical estimates of regional BP for [11C]rac scans. The three methods yielded identical regional BP using –R variations for [11C]MP scans, while slight deviations were observed between +R and –R variations. We found relatively large SD for [11C]DTBZ although deviations remained minimal.
Percent deviations of BP estimates between methods (mean ± standard deviation)
Conclusions
This study indicated that the BPIT plot reached a plateau at ligand-specific t0. After t0, BPIT yielded essentially identical estimates of BP as variations of SRTM and RTGA without the reference region terms. Therefore, we conclude that BPIT is an alternative approach for the ligands examined in this study. Further investigation may be due regarding whether including or excluding the reference tissue term in SRTM and RTGA for [11C]MP.