SUPPLEMENTARY MATERIALS
Supplementary Methods
Imaging acquisition
Tight but comfortable foam padding was used to minimize head motion, and earplugs were used to reduce scanner noise. T2-FLAIR images were obtained for every subject to detect clinically silent lesions. Sagittal structural images (160 sagittal slices, TR = 25 ms, TE = 4.1 ms, thickness = 1.0 mm, no gap, in-plane resolution = 231 × 232, FOV = 230 × 230 mm2, flip angle = 30°) were acquired using a fast field echo (FFE) three-dimensional T1 weighted sequence. Resting-state functional MRI (fMRI) scans were performed by echo planar imaging (EPI) sequence with scan parameters of TR = 2000 ms, TE = 50 ms, flip angle = 90°, matrix = 64 × 64, FOV = 230 × 230 mm2, slice thickness = 4.5 mm and slice gap = 0 mm. Each brain volume comprised 22 axial slices and each functional run contained 240 volumes (8 minutes). During resting state fMRI scanning, subjects were instructed to close their eyes and keep as still as possible, and not to think of anything systematically or fall asleep.
After the scan, all the participants were asked the following questions to verify the degree of their cooperation: “what were you thinking during the scan?”, “did you fall asleep just now?”, “were your eyes closed during the scan?” and “did you feel uncomfortable during the scan?” Only when the participant answered “nothing,” “no, I did not”, “yes, I kept my eyes closed” and “no, I did not feel any discomfort”; this data was included in the present study.
Corpus callosum volume calculation
Prior to processing, all scans were visually examined for motion artifacts or other distortions by a trained rater and only scans with no visible distortion were included in the sample. The automated procedures for subcortical volume measurements of different brain structures have been described previously [1; 2]. Briefly, this process includes motion correction, removal of nonbrain tissue using a hybrid watershed/surface deformation procedure [3], automated Talairach transformation, segmentation of the subcortical white matter and deep gray matter volumetric structures (including the hippocampus, amygdala, caudate, putamen, and ventricles) [1; 2], intensity normalization, tessellation of the gray-white matter boundary, automated topology correction [4] and surface deformation following intensity gradients to optimally place the gray-white matter and gray matter/CSF borders at the location where the greatest shift in intensity defines the transition to the other tissue class. This Free Surfer-based automatic procedure has been shown to be statistically comparable with manual segmentation, especially in gray matter parcellation [1].
Resting-state fMRI data preprocessing
The first 10 images were discarded to allow for signal stabilization and subject adaptation with the remaining images were corrected for slice time differences and head motion. Individual functional images were then co-registered to T1-weighted MR images which were later segmented (gray matter, white matter, and cerebrospinal fluid) and normalized to the standard structural MRI template in the Montreal Neurologic Institute space using a nonlinear transformation. The same transformation parameters were applied to the functional MRI images. To remove the sources of possible spurious variance from each voxel’s fMRI time series, we (a) removed linear trends, (b) removed nuisance signals (white matter, cerebrospinal fluid signals, six head-motion parameters), c) applied temporal bandpass filtering (0.01–0.08 Hz).
To account for differences in the geometric configuration of the cerebral hemispheres, we further transformed the preprocessed functional images to a symmetric space adhering to a previous approach [5]. Briefly, the normalized gray matter images were averaged for all participants to create a group-specific gray matter template. The group-specific gray matter template was then averaged with its left-right flipped version to generate a group-specific symmetrical gray matter template. Subject-specific gray matter images were normalized to the group-level symmetrical gray matter template before application of the resulting nonlinear transformations in order to convert subject-specific fMRI data into group-level symmetrical gray matter template space. The fMRI data was re-sampled in the symmetrical space at a resolution of 3 × 3 × 3 mm3 and spatially smoothed fMRI data with a 6-mm full-width at half-maximum isotropic Gaussian kernel.
CC midsagittal area analysis
Given the relatively frequent labelling error for the fornix and middle CC [6] and tends to overestimate white matter when compared to expert segmentation when using FreeSurfer [7], we repeated the CC midsagittal area analysis with C8 following Herron and coworkers [8]. Using geometric partitioning schemes proposed by Hofer and Frahm [9] to segment the CC into topographic compartments. The maximum extent of the CC along its anterior–posterior axis was identified, and parcellated into five or six compartments based on geometric ratios. The Hofer and Frahm parcellation incorporates a representation of five subregions of the human callosum based on diffusion imaging fiber tractography [9]. The cortical parcellation is as follows: Compartment 1 to prefrontal cortex, Compartment 2 to premotor and supplementary motor cortex, Compartment 3 to primary motor cortex, Compartment 4 to primary sensory cortex, and Compartment 5 to parietal, temporal, and occipital cortices. The CC subregion area were measured and compared between groups. As expected, we got similar results to our primary analysis with Freesurfer method. The CCS-dependent adolescents and young adults had a significantly larger midsagittal area in the mid-posterior (CC2, CC3, CC4, and CC5) subregions compared with the control group (Supplementary Table 2).
References:
1 Fischl B, van der Kouwe A, Destrieux C et al (2004) Automatically parcellating the human cerebral cortex. Cereb Cortex 14:11-22
2 Fischl B, Salat DH, Busa E et al (2002) Whole brain segmentation: automated labeling of neuroanatomical structures in the human brain. Neuron 33:341-355
3 Ségonne F, Dale AM, Busa E et al (2004) A hybrid approach to the skull stripping problem in MRI. Neuroimage 22:1060-1075
4 Fischl B, Liu A, Dale AM (2001) Automated manifold surgery: constructing geometrically accurate and topologically correct models of the human cerebral cortex. IEEE Trans Med Imaging 20:70-80
5 Zuo XN, Kelly C, Di Martino A et al (2010) Growing together and growing apart: regional and sex differences in the lifespan developmental trajectories of functional homotopy. J Neurosci 30:15034-15043
6 Lefebvre A, Beggiato A, Bourgeron T, Toro R (2015) Neuroanatomical diversity of corpus callosum and brain volume in autism: meta-analysis, analysis of the Autism Brain Imaging Data Exchange project, and simulation. Biol Psychiatry 78:126-134
7 Klauschen F, Goldman A, Barra V, Meyer-Lindenberg A, Lundervold A (2009) Evaluation of automated brain MR image segmentation and volumetry methods. Hum Brain Mapp 30:1310-1327
8 Herron TJ, Kang X, Woods DL (2012) Automated measurement of the human corpus callosum using MRI. Front Neuroinform 6
9 Hofer S, Frahm J (2006) Topography of the human corpus callosum revisited—Comprehensive fiber tractography using diffusion tensor magnetic resonance imaging. Neuroimage 32:989-994
Supplementary Table 1. Brain Regions with Abnormal (decrease) inter-hemispheric mirrored functional connectivity (VMHC) in CCS-dependent adolescents and young adults compared with control subjects. Results were reported at p < 0.05, corrected with a single voxel height of p < 0.01 and a cluster volume > 1836 mm3 using software [AFNIAlphaSim; http:/afni.nimh.gov/pub/dist/doc/manual/AlphaSim.pdf]) with a gray matter mask produced by the symmetrical template.
Region / MNI coordinate / BA / Peak t values / Volume (mm3)Medial OFC / 6,42,-14 / 11 / -4.9555 / 1944
Abbreviation: CCS, codeine-containing cough syrups; OFC, orbitofrontal cortex; MNI, Montreal Neurological Institute; BA, Brodmann area.
Supplementary Table 2. Midsagittal CC area measurements of all CCS-dependent adolescents and young adults (patients) compared with non-addicted control group. CCS-dependent individuals had larger midsagittal area in the middle- and posterior- CC subregions than controls.
CC / Patients$ / Non-addicted controls$ / p-valueCC1 (mm3) / 179.2 (20.1) / 169.6 (23.0) / 0.103
CC2 (mm3) / 199.9 (26.2) / 179.3 (28.3) / 0.005*
CC3 (mm3) / 75.5 (15.6) / 62.8 (12.2) / 0.001*
CC4 (mm3) / 36.6 (7.5) / 31.3 (6.3) / 0.005*
CC5 (mm3) / 240.7 (25.1) / 216.5 (25.3) / 0.001*
CC, corpus callosum; CCS, codeine-containing cough syrups.
$Values expressed as mean (standard deviation)
*Significant if p < 0.05 (Bonferroni correction)
Supplementary Figure 1. Schematic of corpus callosum segmentation showing five equal segments.
CC, corpus callosum; CC1, rostrum; CC2, genu; CC3, truncus/body; CC4, anterior splenium; CC5, roughly equivalent of the posterior splenium
Supplementary Figure 2. Both groups show adequate T2* signal in the region of the medial OFC.
The mean T2* images from the 38 controls (A) and from the 33 CCS-dependent male adolescents and young adults (C) are shown with the crosshairs on the medial OFC cluster (B) that showed reduced inter-hemispheric functional connectivity in CCS-dependent adolescents and young adults.
5