|
 |
|
|||||||
|
|||||||||
Correspondence: When an article is eligible for submission of Correspondence, a link to the response form is available within the full-text article. You must be a current subscriber who has activated the online portion of your subscription in order to send a Correspondence. Any reader can read published Correspondence.
Correspondence to:
|
|
Correspondence:Correspondence published:
-
![[Read Correspondence]](/icons/misc/sectionDOWN.gif)
Reply to Kassubek
- Herminia D Rosas, H. Diana Rosas and Jill Goldstein (29 July 2003)
-
MRI-based studies on Huntington´s disease: variability of extrastriatal volume changes
- Jan Kassubek, Wilhelm Gaus and G. Bernhard Landwehrmeyer (22 July 2003)
|
|
|||
|
Herminia D Rosas, Massachusetts General Hospital 114 16th Street Charlestown, MA 02129, H. Diana Rosas and Jill Goldstein
Send Correspondence to journal:
rosas{at}helix.mgh.harvard.edu Herminia D Rosas, et al.
|
We look forward to seeing the results of Dr. Kassubek’s study, but disagree with several of his statements. Table 2 reflects absolute volume differences (unadjusted for the size of the cerebrum). While there is a great deal of variability in the size of the cerebrum in the two groups, we were interested in: differences in brain regions within the cerebrum (controlled for the overall size of the cerebrum), as well as how much atrophy occurred for particular ROIs regardless of overall size. Table 3 reflects the effect sizes for mean differences in volumes between the groups (adjusted for cerebral size); the standard deviations are not significantly different between groups on most of the ROIs. In addition, the effect sizes for the mean volume differences are quite large for extrastriatal regions. Effect size estimations are better reflections of the size of the differences between the groups, not percent differences in absolute volumes as they reflect standard deviation units from the "population mean". These are the data to which we refer when we say that there are important extrastriatal differences between groups. Even for absolute volume differences the standard deviation for the accumbens, hippocampus, amgydala, and brainstem are very similar. Since we included cerebral exterior as a covariate in the GLM, we are testing for differences between the groups on specific ROIs, unaccounted for by size of cerebrum. Our view is that the use of anatomically-based ROI segmentation provides a closer approximation to the actual anatomy of the brain. Our techniques have resulted in excellent reliabilities and have been validated across many studies of different disorders.. Previous volumetric studies that included this reliable method of segmenting the accumbens in disorders such as schizophrenia and OCD have been reported. Unreliability of segmenting the accumbens would not result in significant differences between the groups, unless there was differential misclassification of the accumbens. Unreliability would attenuate finding significant differences. This is unlikely given that raters were blind to group identification. We found reductions in the volume of the vental diencephalons in HD this area is prone to unreliability in the segmentation. Our previous published work has in fact demonstrated that the cortex also appears to undergo regionally specific and progressive thinning, although here we report global cerebral gray. The results from voxel-based morphometry are difficult to interpret, in particular wrt volume Each method, even the neuropathological “gold standard” is prone to methodological limitations. Striatal degeneration appears so severe in mid-stages of disease, that it may not be a sensitive or reliable surrogate marker of disease of disease progression. Future longitudinal studies are necessary. |
|||
|
|
|||
|
Jan Kassubek, Dept. of Neurology, University of Ulm, Ulm, Germany Dept. of Neurology, University of Ulm, Oberer Eselsberg 45, 89081 Ulm, Germany, Wilhelm Gaus and G. Bernhard Landwehrmeyer
Send Correspondence to journal:
jan.kassubek{at}medizin.uni-ulm.de Jan Kassubek, et al.
|
We read with interest the paper by Rosas et al. [1] on regional volume alterations in MR scans of 18 patients with early Huntington´s disease (HD) using a region-of-interest (ROI) based analysis. The authors stress that atrophy can be demonstrated in numerous extrastriatal areas and conclude that even in early HD “widespread degeneration” occurs. They also suggest that extrastriatal degeneration may account for some clinical symptoms traditionally attributed to striatal pathology. There is little doubt that neurodegeneration in HD extends well beyond the striatum; however, it is our impression that extrastriatal atrophy tends to vary. This impression is based on the review of neuropathological material [2] as well as on the results of a MRI study on 44 patients with early HD [3], using the extensively validated whole-brain- based technique of voxel-based morphometry for 3-D MRI analysis [4]. A closer look at the data presented by Rosas et al. shows that extrastriatal volume changes in HD are indeed not very impressive despite the emphasis the authors place on them. The volume reductions expressed as percent of normal are larger than 20% only for striatum, pallidum and amygdala; all other ROIs were 86% of control values and more. Mean absolute volumes of extrastriatal areas showed large standard deviations in the HD group resulting in considerable overlap between HD and controls. Using a general linear model for correlated data (GLM-CD), the authors concluded that extrastriatal volume reductions in HD were not due to chance. Since the standard deviations for HD-patients were substantially higher than for controls (e.g. cerebral exterior 159.79 versus 9.96) we wonder whether this inhomogeneity is compatible with the assumptions of a GLM-CD. Furthermore, it is unclear how structure volumes were ‘adjusted’ for total brain volumes (do the values given represent percentages of individual total brain volumes?) and whether the denominator ‘total brain volume’ refers to the volume of all brain tissues or to the internal volume of the skull; only the latter would not be subject to changes in the course of the disease. In addition, in our view the use of a ROI-based technique as primary method of analysis has several drawbacks. First, robust changes in areas not defined as ROIs, e.g. the hypothalamic area which exhibited significant changes in our morphometric study [3] as well as in neuropathological studies [5] were missed. Second, we are concerned that ROI definitions such as "cerebral gray" may be misleading since multiple areas with potentially variable natural history in terms of degeneration are lumped together. In addition, the surprising observation that the nucleus accumbens appears as atrophic as the caudate in the present study (which is at odds with numerous neuropathological studies 2) might reflect difficulties in interactively outlining this ROI in pathologically altered brain scans. The demonstration of quantitative extrastriatal volume loss may be of limited use in the context of clinical trials given a marked interindividual variability: disease progression may be far more robustly reflected in striatal volume changes. Longitudinal studies will show whether subregional striatal alterations can be used as progression markers in clinical trials and whether they are more reliably captured using voxel-based or ROI-based approaches. References 1). Rosas HD, Koroshetz WJ, Chen YI, et al. Evidence for more widespread cerebral pathology in early HD: An MRI-based morphometric analysis. Neurology 2003; 60 (10): 1615-1620 2). Vonsattel JPG, DiFiglia M. Huntington disease. J Neuropathol Exp Neurol 1998; 57: 369-384 3). Kassubek J, Juengling FD, Kioschies T, et al. Topography of cerebral atrophy in early Huntington´s disease: a voxel-based morphometric MRI study. J Neurol Neurosurg Psychiatry 2003; in press 4). Ashburner J, Csernansky JG, Davatzikos C, et al. Computer-assisted imaging to assess brain structure in healthy and diseased brains. Lancet Neurology 2003; 2: 79-88 5). Kremer HPH, Roos RA, Dingjan G, Marani E, Bots GT. Atrophy of the hypothalamic lateral tuberal nucleus in Huntington's disease. J Neuropathol Exp Neurol 1990; 49 (4): 371-382 |
|||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH |