Imaging human reward processing with positron emission tomography and functional magnetic resonance imaging

Nina B. L. Urban, Mark Slifstein, Shashwath Meda, Xiaoyan Xu, Rawad Ayoub, Olga Medina, Godfrey D. Pearlson, John H. Krystal, Anissa Abi-Dargham

Functional neuroimaging (fMRI) studies show activation in mesolimbic circuitry in tasks involving reward processing, like the Monetary Incentive Delay Task (MIDT). In voltammetry studies in animals, mesolimbic dopamine release is associated with reward salience. This study examined the relationship between fMRI activation and magnitude of dopamine release measured with Positron emission tomography study (PET) in the same subjects using MIDT in both modalities to test if fMRI activation is related to dopamine release. Eighteen healthy subjects were scanned with [¹¹C]raclopride PET at baseline and after MIDT. Binding potential (BP(ND)) was derived by equilibrium analysis in striatal subregions and percent change across conditions (∆BP(ND)) was measured. Blood oxygen level dependence (BOLD) signal changes with MIDT were measured during fMRI using voxelwise analysis and ROI analysis and correlated with ∆BP(ND). ∆BP(ND) was not significant in the ventral striatum (VST) but reached significance in the posterior caudate. The fMRI BOLD activation was highest in VST. No significant associations between ∆BP(ND) and change in fMRI BOLD were observed with VST using ROI analysis. Voxelwise analysis showed positive correlation between BOLD activation in anticipation of the highest reward and ∆BP(ND) in VST and precommissural putamen. Our study indicates that endogenous dopamine release in VST is of small magnitude and is related to BOLD signal change during performance of the MIDT in only a few voxels when rewarding and nonrewarding conditions are interspersed. The lack of correlation at the ROI level may be due to the small magnitude of release or to the particular dependence of BOLD on glutamatergic signaling.