White matter alterations in cannabis users using Diffusion MRI: a Fixel-Based Analysis study
Poster No:
1607
Submission Type:
Abstract Submission
Authors:
Arush Honnedevasthana Arun1, Suraya Dunsford2, Adam Clemente1, Thijs Dhollander3, Nadia Solowij2, Lisa Greenwood4, Valentina Lorenzetti1
Institutions:
1Australian Catholic University, Melbourne, Victoria, 2University of Wollongong, Wollongong, Wollongong, New South Wales, 3Murdoch Children's Research Institute, Melbourne, Victoria, 4The Australian National University, Canberra, ACT
First Author:
Co-Author(s):
Introduction:
Cannabis use disorders (CUD) affect > 60 million people globally and have significant adverse psychosocial outcomes (inability to quit, cravings, operating machinery/driving while intoxicated). Such outcomes have been attributed to altered brain integrity but this notion is yet to be tested in people with a DSM-5 diagnosis of CUD and using advanced measures of white-matter microstructure such as Fixel-based analysis on diffusion MRI data.
Methods:
A sample of 109 participants aged 18-32 years, comprised of 79 cannabis users stratified by CUD severity (22 mild,26 moderate, 31 severe, according to the DSM-V criteria and 30 non-using controls participated in the study. Participants underwent diffusion MRI scan on a 3T GE Architect scanner with a 48-channel head coil.
All the DWI data analysis was conducted using MRtrix3(www.mrtrix3.org) (Tournier, Smith et al. 2019). Firstly, the DWI data was denoised (Veraart, Novikov et al. 2016), Gibbs ringing artifacts were removed (Kellner, Dhital et al. 2016), motion and distortion correction was performed using the flipped DWI dataset (Andersson and Sotiropoulos 2016).
All the DWI data were upsampled to an isotopic voxel size of 1·25 mm in order to increase contrast and improve template construction (Raffelt, Tournier et al. 2017). FODs were calculated using constrained spherical deconvolution (Tournier, Calamante et al. 2007) utilising group averaged response functions for white matter, grey matter and CSF(Dhollander, Raffelt et al. 2016). FOD images were globally normalised to correct for image intensity differences between subjects.
All the subject FOD images were spatially registered and normalized to a create a symmetric study-specific template (Raffelt, Tournier et al. 2017). Fibre density (FD) was computed from the integral of each FOD lobe. Fibre cross-section (FC) for each subject is calculated by using the warp information derived from the registration process. Fibre density and cross-section (FDC) measure is a combined (product) measure of FD and FC.
General Linear Model (GLM) was utilized to compare FD, FC and FDC between cannabis and control groups, controlling for age, gender, ICV, standard drinks/year. We compared cannabis users vs control users and cannabis user individual classification (mild, moderate and severe) groups with controls.
All the DWI data analysis was conducted using MRtrix3(www.mrtrix3.org) (Tournier, Smith et al. 2019). Firstly, the DWI data was denoised (Veraart, Novikov et al. 2016), Gibbs ringing artifacts were removed (Kellner, Dhital et al. 2016), motion and distortion correction was performed using the flipped DWI dataset (Andersson and Sotiropoulos 2016).
All the DWI data were upsampled to an isotopic voxel size of 1·25 mm in order to increase contrast and improve template construction (Raffelt, Tournier et al. 2017). FODs were calculated using constrained spherical deconvolution (Tournier, Calamante et al. 2007) utilising group averaged response functions for white matter, grey matter and CSF(Dhollander, Raffelt et al. 2016). FOD images were globally normalised to correct for image intensity differences between subjects.
All the subject FOD images were spatially registered and normalized to a create a symmetric study-specific template (Raffelt, Tournier et al. 2017). Fibre density (FD) was computed from the integral of each FOD lobe. Fibre cross-section (FC) for each subject is calculated by using the warp information derived from the registration process. Fibre density and cross-section (FDC) measure is a combined (product) measure of FD and FC.
General Linear Model (GLM) was utilized to compare FD, FC and FDC between cannabis and control groups, controlling for age, gender, ICV, standard drinks/year. We compared cannabis users vs control users and cannabis user individual classification (mild, moderate and severe) groups with controls.
Results:
Cannabis users and controls were found to differ significantly in education (p=0·009), IQ (p<0·001), and alcohol dependence (AUDIT scores; p=0·0)04, the control group had significantly higher IQ and more years of education, and lower alcohol dependence than the cannabis users. Cannabis users had significantly higher scores than controls in STAI-T (U=647 p<0·001), BDI (U= 429, p<0·001), CAPE_P (U=328.5, p<0·001), CAPE_N (U=526, p<0·0001and PSS (U= 480.5, p=<0·001).
We found no significant differences between cannabis users with controls. In summary, Figure 1 shows significant fixels (FWE-corrected p-value < 0·05) overlaid on a FOD template in mild users vs controls for the FD, FC and FDC metric in Superior Longitudinal Fasciculus (SLF) and Fornix regions. We found that parts of SLF in mild users have lower FD and FDC compared to controls. We also found lower FC in Fornix within mild cannabis users compared to controls
We found no significant differences between cannabis users with controls. In summary, Figure 1 shows significant fixels (FWE-corrected p-value < 0·05) overlaid on a FOD template in mild users vs controls for the FD, FC and FDC metric in Superior Longitudinal Fasciculus (SLF) and Fornix regions. We found that parts of SLF in mild users have lower FD and FDC compared to controls. We also found lower FC in Fornix within mild cannabis users compared to controls
·Significant fixels (p=0.05, FWE corrected) displayed for mild cannabis users > controls in Fibre density (first row), Fibre cross-section (middle row) and Fibre density & cross-section (last row) over
Conclusions:
We explored if group differences are associated with CUD severity and cannabis exposure. CUD and control groups were not significantly different. Both mild and moderate CUD groups vs controls, had lower FD in the superior longitudinal fasciculus connecting parietal and striatal-cingulate tracts implicated in addictive behavior (eg disinhibition), and this was correlated with the age of cannabis use onset. Mild CUD compared to controls had lower FC in a fornix region connecting the hippocampus to the subcortex. Different white matter integrity in mild and moderate CUD might reflect transient neuroplastic changes at the initial stages of addiction, which may normalize with the transition to severe CUD.
Emotion, Motivation and Social Neuroscience:
Social Neuroscience Other 2
Emotion and Motivation Other
Modeling and Analysis Methods:
Diffusion MRI Modeling and Analysis 1
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
White Matter Anatomy, Fiber Pathways and Connectivity
Keywords:
Addictions
MRI
White Matter
WHITE MATTER IMAGING - DTI, HARDI, DSI, ETC
1|2Indicates the priority used for review
Provide references using author date format
Dhollander, T., D. Raffelt and A. Connelly (2016). Unsupervised 3-tissue response function estimation from single-shell or multi-shell diffusion MR data without a co-registered T1 image.
Kellner, E., B. Dhital, V. G. Kiselev and M. Reisert (2016). "Gibbs-ringing artifact removal based on local subvoxel-shifts." Magnetic Resonance in Medicine 76(5): 1574-1581.
Raffelt, D. A., J. D. Tournier, R. E. Smith, D. N. Vaughan, G. Jackson, G. R. Ridgway and A. Connelly (2017). "Investigating white matter fibre density and morphology using fixel-based analysis." NeuroImage 144(Pt A): 58-73.
Tournier, J. D., F. Calamante and A. Connelly (2007). "Robust determination of the fibre orientation distribution in diffusion MRI: Non-negativity constrained super-resolved spherical deconvolution." NeuroImage 35(4): 1459-1472.
Tournier, J. D., R. Smith, D. Raffelt, R. Tabbara, T. Dhollander, M. Pietsch, D. Christiaens, B. Jeurissen, C.-H. Yeh and A. Connelly (2019). "MRtrix3: A fast, flexible and open software framework for medical image processing and visualisation." NeuroImage 202: 116137.
Veraart, J., D. S. Novikov, D. Christiaens, B. Ades-Aron, J. Sijbers and E. Fieremans (2016). "Denoising of diffusion MRI using random matrix theory." NeuroImage 142: 394-406.