Measurements of dopamine D1-type receptors using simultaneous PET-MRI scanner: Study on test-retest reliability

Objectives The role of dopamine D1-type receptor (D1R)-expressing neurons in the regulation of motivated behavior and future reward prediction remains elusive. In order to assess D1-mediated neuronal network regulation using truly simultaneous PET-MRI and D1R-selective [11C]SCH23390, the aim of this study was to calculate the stability of D1R-binding potential (BPnd) intra-individually in the brain between two independent PET-MRI measurements under neutral conditions.

Methods Fifteen healthy volunteers (7 females, aged 33 ± 13yrs) underwent two 90-min emission scans each after 90-sec bolus infusion of 486 ± 16MBq [11C]SCH23390 on two separate days within 2 weeks (scan1 and scan2) using mMR biograph (SIEMENS). For attenuation correction (AC), a two-point MRI Dixon VIBE sequence was obtained in parallel to the PET acquisition, resulting in segmented AC maps. After performing motion correction with SPM8, the parametric images of D1R BPnd were generated in PMOD3.2 (PMOD Technologies, Zurich, Switzerland) by multi-linear reference tissue model with two parameters with the cerebellar cortex as receptor free reference region. For VOI-analysis, BPnd data were then co-registered with the individual spatially reoriented T1 MRI data (PMOD3.2) and target VOIs were manually drawn on consecutive transversal MRI slices. SPM8 was used for voxel-wise comparison with spatially normalized BPnd images. For group comparison scan1 vs scan2 a paired t-test was applied, which was considered significant at p<0.05 (VOI-based) and p<0.005 (uncorrected with minimum cluster size of 30 voxels for SPM).

Results Neither voxel-wise nor VOI-based analyses revealed significant differences in BPnd between the two measurements with calculated BPnd of 1.47 ± 0.30 (scan1) and 1.45 ± 0.31 (scan2) for the nucleus accumbens; 1.90 ± 0.39 (scan1) and 1.91 ± 0.56 (scan2) for the head of the caudate; 1.57 ± 0.44 (scan1) and 1.58 ± 0.53 (scan2) for the putamen; 0.36 ± 0.13 (scan1) and 0.36 ± 0.13 (scan2) for the dorsolateral prefrontal cortex; and 0.29 ± 0.09 (scan1) and 0.28 ± 0.09 (scan2) for the substantia nigra/ventral tegmental area. This resulted in absolute variabilities (%) of 8.1 ± 6.7; 8.1 ± 7.2; 8.5 ± 8.0; 13.5 ± 13.3; and 14.8 ± 14.3. Mean differences according to Bland-Altman analyses were very low (0.013 ± 0.17; -0.011 ± 0.24; -0.013 ± 0.23; 0.003 ± 0.05; and 0.011 ± 0.05, respectively) and intraclass correlation (one-way, random) indicated very high agreement (0.93; 0.92; 0.95; 0.97; and 0.93).

Conclusions Test-retest reliability of D1R measurements using [11C]SCH3390 and simultaneous PET-MRI is high not only in D1R-rich brain areas but also in areas with lower D1R content. These measurements will provide the base for future multimodal joint PET-MRI data analyses in stimulation-dependent mapping of D1R-containing neurons and their effects on neuronal circuits determining behavior.