Christine N. Vidal1,
Judith L. Rapoport MD2,
Kiralee M. Hayashi1,
Lauren E. McLemore1,
Jay N. Giedd MD2,
Peter Gochman MA2,
Jonathan Blumenthal MA2,
Nitin Gogtay MD2,
Arthur W. Toga PhD1,
Paul M. Thompson PhD1
1Laboratory of Neuro Imaging, Brain Mapping Division, Department of Neurology, UCLA School of Medicine
2Child Psychiatry Branch, National Institute of Mental Health, NIH, Bethesda, MD
Fig. 1: Average gray matter loss in medial cortex in COS; Hippocampal deficits in COS.
3D MRI scans of 12 COS and 12 matched healthy subjects were acquired longitudinally over a 5-year time span (age: 13.9+/-0.8SE/18.6+/-1.0SE years at initial/final scans). All subjects were matched for age, height, sex, but not IQ. Scans were normalized by affine transformation to ICBM standardized stereotaxic space. We applied the following anatomical modeling techniques:
Medial cortex mapping: 3D maps localizing brain changes were derived from MRI scans at baseline and follow-up. Gyral pattern and shape variations were encoded using high-dimensional elastic deformation mappings driving each subjectís cortical anatomy onto a group average. Changes in cortical gray matter density were mapped by computing warping fields that matched cortical patterns across hemispheres, across time and across subjects.
Hippocampal surface mapping: One rater, blind to age, gender and diagnosis, delineated the hippocampus at baseline and follow-up. All traces were uniformly redigitized to render the sampled points spatially uniform, and converted into 3D parametric surface meshes. Surface meshes were averaged across subjects to create average shape models. Distance fields relating surface boundary points to their medial curves were analyzed using surface-based nonparametric regression. Shape differences, visualized in the average anatomical maps, were assessed using permutation tests to compute a null distribution for the total surface area of suprathreshold statistics. Group differences in hippocampus volumes were assessed with multivariate regression.
We found a selective, and severe, frontal gray matter loss in both left and right medial surfaces of the brain hemispheres, in the schizophrenic subjects. Intriguingly, a sharp boundary appeared in the pattern of gray matter loss (see Figure). This separated frontal regions, which control attention and planning, and the limbic areas, which were comparatively spared. We created color-coded maps to reveal additional localized linkages between these progressive deficits and cognitive and clinical assessments. Hippocampal surface analyses revealed a significant bilateral volume decrease at time 2 in COS compared to healthy controls (average L. deficit: -15.1%+/-2.0%, p<0.0001; average R. deficit.: -14.8%+/-2.2%, p<.0.0007). No significant volume differences were detected at time 1 (when mean deficits in COS were L:-2.4%+/-3.0%, R:-4.9%+/-3.1%).
Frontal and limbic regions may have strikingly different vulnerabilities in terms of gray matter loss in COS, with frontal regions much more severely affected. Nonetheless, hippocampal deficits were detected in the average shape maps at follow-up. Such deficits are frequently reported in studies of adult onset schizophrenia . In the course of their illness, COS patients may exhibit decelerated growth in medial temporal lobe structures, regions that are among the last to mature in the teenage years.
References: References: . Rapoport JL et al., Arch. Gen. Psychiatry 56(7):649-54. . Thompson PM et al., PNAS 98(20):11650-11655, 2001. . Lawrie S, Abukmeil S, Br. J. Psych. 172:110-20, 1998.