A hypo-status in drug-dependent brain revealed by multi-modal MRI

Abstract

Drug addiction is a chronic brain disorder with no proven effective cure. Assessing both structural and functional brain alterations by using multi-modal, rather than purely unimodal imaging techniques, may provide a more comprehensive understanding of the brain mechanisms underlying addiction, which in turn may facilitate future treatment strategies. However, this type of research remains scarce in the literature. We acquired multi-modal magnetic resonance imaging from 20 cocaine-addicted individuals and 19 age-matched controls. Compared with controls, cocaine addicts showed a multi-modal hypo-status with (1) decreased brain tissue volume in the medial and lateral orbitofrontal cortex (OFC); (2) hypo-perfusion in the prefrontal cortex, anterior cingulate cortex, insula, right temporal cortex and dorsolateral prefrontal cortex and (3) reduced irregularity of resting state activity in the OFC and limbic areas, as well as the cingulate, visual and parietal cortices. In the cocaine-addicted brain, larger tissue volume in the medial OFC, anterior cingulate cortex and ventral striatum and smaller insular tissue volume were associated with higher cocaine dependence levels. Decreased perfusion in the amygdala and insula was also correlated with higher cocaine dependence levels. Tissue volume, perfusion, and brain entropy in the insula and prefrontal cortex, all showed a trend of negative correlation with drug craving scores. The three modalities showed voxel-wise correlation in various brain regions, and combining them improved patient versus control brain classification accuracy. These results, for the first time, demonstrate a comprehensive cocaine-dependence and craving-related hypo-status regarding the tissue volume, perfusion and resting brain irregularity in the cocaine-addicted brain.

Keywords: arterial spin labeling; brain entropy; cerebral blood flow; resting state fMRI; voxel-based morphometry.

Fig. 1

Brain difference between cocaine patients and age-matched controls. Hot, violet, and green colors means atrophy, hypoperfusion, and hypo-BEN in cocaine patients, respectively. The left side of the image is the left side of the brain. The numbers above each slice indicates the spatial location of the slice in the MNI space.

Fig. 2

Correlations of GM tissue volume a) and CBF b) to cocaine dependence level. Violet and green mean positive and negative relation between GM volume and drug dependence level, respectively. Cool color means negative correlation between CBF and drug dependence level. The left side of the image is the left side of the brain. The numbers above each slice indicates the spatial location of the slice in the MNI space.

Fig. 3

Correlations of grey matter volume, CBF, BEN to drug craving scores. Violet, cool, and green colors mean negative correlation between craving score and grey matter volume, CBF, and BEN, respectively. The left side of the image is the left side of the brain. The numbers above each slice indicates the spatial location of the slice in the MNI space.

Fig. 4

Voxel-wise inter-imaging modality correlations. A) Tissue volume vs CBF, B) Tissue volume vs BEN, C) BEN vs CBF correlations. Hot color and green mean positive and negative correlation, respectively. Color bars indicate the display window for the correlation coefficients. The left side of the image is the left side of the brain. The numbers above each slice indicates the spatial location of the slice in the MNI space.