Image analysis and statistics.
Regions of interest (ROIs) were obtained directly from the [11C]raclopride images as described previously (Volkow et al., 1994). Briefly, we identified the ROI on summed images (dynamic images taken from 10 to 54 min) that were resliced along the intercommissural plane (anterior commissure–posterior commissure line). The caudate, putamen, VS, and cerebellum were measured on four, three, two, and two planes, respectively, and right and left regions were delineated. These regions were then projected to the dynamic scans to obtain concentrations of carbon-11 versus time, which were used to calculate the K1 (transport constant from plasma to tissue) and the distribution volumes (milliliters per grams) and the blood to tissue transport constant (K1) in caudate, putamen, VS, and cerebellum using a graphical analyses technique for reversible systems (Logan plots) (Logan et al., 1990). The measure of binding potential relative to a nondisplaceable reference (BPND) was obtained as the ratio of the distribution volume in striatum to that in cerebellum minus 1 and was used to quantify DA D2/D3 receptor availability. The difference in BPND between MP and placebo (ΔBPND) was used as the measure of DA release. A two-measure within-subject repeated ANOVA to assess for drug effect (placebo vs intravenous MP) and condition (treatment-naive vs long-term treatment) was done to evaluate the effects of intravenous MP on DA and to determine whether this effect was influenced by long-term clinical treatment with oral MP. Pearson's product moment correlation analysis was done to assess whether DA changes were associated with symptom changes while on treatment [difference on the scores for inattention (CAARS A), hyperactivity (CAARS B), and total symptoms (CAARS G) obtained before treatment initiation and those after long-term clinical treatment with oral MP]. To control for potential nonspecific effects of intravenous MP, we also evaluated its effects on K1 and assessed the correlation between changes in K1 (placebo vs intravenous MP) and symptom improvement while on treatment. To control for the contribution of depression and anxiety on MP-induced DA changes, we included the Hamilton scores as covariates in the analyses.
Findings with the ROI analysis were considered significant only if corroborated with Statistical Parametric Mapping (SPM) (Friston et al., 1995). SPM was also used for exploratory analysis to assess the effects of intravenous MP on DA release in extrastriatal brain regions. The SPM analysis was performed on BPND images (obtained by computing the BPND in each pixel). A spatial normalization template matching the average BPND image contrast in the brain was developed using the BPND images from 34 healthy control subjects that participated in previous [11C]raclopride studies and collected with the same instrument and scanning sequence used for the present study. All images were carefully inspected to ensure whole-brain coverage and exclude images with potential artifacts. A 12-parameter affine transformation with 16 nonlinear iterations was used to register the images to the PD.mnc Montreal Neurological Institute (MNI) template provided with the SPM2 package. The dimensions of the bounding box of the output volumes were 90:90, 126:90, and 72:108 along x, y, and z, respectively, and the voxel size was 2 × 2 × 2 mm3. Bilinear interpolation was used for image reslicing. The average normalized and smoothed image (FWHM, 8 mm) was computed to obtain the BPND template in the MNI stereotactic space. The BPND images of the present study were then spatially normalized to the stereotactic space of MNI using the BPND template and subsequently smoothed with an 8 mm isotropic Gaussian kernel. A within-subject ANOVA model in SPM2 was used to assess for drug effect (placebo vs intravenous MP) and treatment condition (treatment-naive vs long-term treatment) to assess the effects of intravenous MP on DA and to determine whether this effect was affected by long-term clinical MP treatment. Significance was set at p < 0.001 (uncorrected, cluster size threshold >100 voxels), and the statistical maps were overlaid on an MRI structural image. Pearson's product moment correlations were used to assess the association between MP-induced DA changes (ΔBPND intravenous MP vs placebo) and the changes in the scores on the CAARS A (inattention) and CAARS B (hyperactivity) (Δtreatment-naive vs long-term treatment).