Prior efforts to dissect etiological and pharmacological modulations in brain morphology in obsessive-compulsive disorder (OCD) are often undermined by methodological and sampling constraints, yielding conflicting conclusions and no reliable neuromarkers. Here we evaluated alteration of regional gray matter volume including effect size (Cohen's d value) in 95 drug-naïve patients (age range: 18-55) compared to 95 healthy subjects (age: 18-63), then examined pharmacological effects in 65 medicated (age: 18-57) and 73 medication-free patients (age: 18-61). Robustness of statistical outcomes and effect sizes was rigorously tested with Monte Carlo cross-validation. Relative to controls, both drug-naïve and medication-free patients exhibited comparable volumetric increases mainly in the left thalamus (d=0.90, 0.82, respectively), left ventral striatum (d=0.88, 0.67), bilateral medial orbitofrontal cortex (d=0.86, 0.71; 0.90, 0.73), and left inferior temporal gyrus (d=0.83, 0.66), and decreased volumes in left premotor/presupplementary motor areas (d=-0.83, -0.71). Interestingly, abnormalities in the thalamus and medial orbitofrontal cortex were present in medicated patients whereas entirely absent in premotor and ventral striatum. It suggests that pharmacotherapy elicited divergent responses in orbitofronto-striato-thalamic and premotor circuits, which warrants the design of longitudinal studies investigating the potential of these neuromarkers in stratified treatments of OCD.
Keywords: Cross-validation; Drug-naïve; OCD; Pharmacotherapy; Voxel-based morphometry.
Copyright © 2017 The Authors. Published by Elsevier B.V. All rights reserved.