| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Neuroradiology |
1 From the Radiodiagnostic Section, Department of Clinical Physiopathology (M. Mascalchi, R.D.N., C.T., C.G., L.S.P.) and Department of Neurological Sciences (F.L., R.P., S.P.), University of Florence, Viale Morgagni 85, Florence, Italy; Unit of Neurophysiopathology, Empoli General Hospital, Italy (M. Macucci); and Neuroimaging Research Unit, Department of Neurology, University Hospital San Raffaele, Milan, Italy (M.F.). Received June 3, 2003; revision requested August 14; final revision received December 29; accepted January 29, 2004. Address correspondence to M. Mascalchi (e-mail: m.mascalchi@dfc.unifi.it).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Twenty-one carriers of the HD mutation (mean age, 58 years ± 11 [SD]) and 21 healthy control subjects (mean age, 54 years ± 13) underwent conventional, diffusion-weighted, and MT MR imaging. Volumes, mean apparent diffusion coefficients (ADCs), and MT ratios (MTRs) for left and right caudate nucleus, putamen, and cerebral periventricular white matter—as well as an index of normalized brain volume and whole-brain ADC and MT histograms—were computed. Asymmetry in volume, ADC, and MTR measurements in caudate nucleus, putamen, and periventricular white matter in control subjects and HD carriers were evaluated with Wilcoxon testing for paired samples. Differences in MR imaging variables between HD carriers and control subjects were evaluated with Mann-Whitney U testing; correlations between stages of clinical severity and MR imaging data were investigated with Spearman rank correlation testing.
RESULTS: No significant asymmetry was observed for any of the MR imaging variables. Caudate nucleus, putamen, and whole-brain volumes were smaller (P < .001 for all) in HD carriers than in control subjects. HD carriers also had increased ADC in the caudate nucleus (P = .002), putamen (P <. 001), cerebral periventricular white matter (P < .001), and whole brain (P < .001). MTR was not significantly different between HD carriers and control subjects. Correlation was observed between stages of increasing clinical disease severity and both decrease in volume of caudate nucleus (Spearman 
= –0.63), putamen ( = –0.64), and whole brain ( = –0.46) and increase in ADC in caudate nucleus ( = 0.52), periventricular white matter ( = 0.45), and whole brain ( = 0.44).
CONCLUSION: Regional and global volume loss in HD is accompanied by an increase in ADC; this correlates with disease severity.
© RSNA, 2004
Index terms: Brain, diffusion, 10.12144 • Brain, diseases, 10.884 • Brain, MR, 10.121417, 10.12144 • Huntington disease, 10.884 • Magnetic resonance (MR), diffusion tensor, 10.12144 • Magnetic resonance (MR), magnetization transfer, 10.121417
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
However, combining the results of imaging studies with the results of molecular genetics improves the genotype-phenotype correlation and has the potential to enable a betterunderstanding of the pathophysiology of the disease. In particular, results of several studies involving MR imaging demonstrated that volume loss in the neostriatum (ie, the caudate nucleus and the putamen) can be observed in presymptomatic subjects and is correlated with triplet expansion and disease progression (2–5). More recently, MR imaging techniques that exploit such new parameters as magnetization transfer and diffusion of water protons have been employed for the evaluation of degenerative diseases of the CNS such as Alzheimer disease and Parkinsonian syndromes (6–11).
Thus, the purpose of our study was to investigate whether diffusion-weighted and magnetization transfer MR imaging depict regional and/or global brain abnormalities in patients with HD.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
32) (13). Nineteen of the HD carriers were symptomatic and attended the outpatient facility for movement disorders in our hospital (Azienda Ospedaliera di Careggi, Florence) between April 2001 and April 2003. Two women of 31 and 43 years of age were asymptomatic relatives of two patients and were identified as HD carriers with molecular genetic testing during the time period of the study and were enrolled. None of the symptomatic subjects had akinetic-rigid HD. Fourteen of the symptomatic patients were undergoing chronic treatment with antidepressants (serotonin reuptake inhibitors or tricyclics) or neuroleptic drugs (eg, haloperidol).
At the time of MR imaging examination, one neurologist (R.P.), who was blinded to the MR imaging results, classified the HD carriers according to the five stages of functional capacities defined by Shoulson and Fahn (14) and Shoulson (15). These five stages take into account an individual’s engagement in occupation; ability to handle financial affairs, manage domestic responsibilities, and perform activities of daily living; and his or her need for care facilities. Stage 1 includes asymptomatic subjects and patients with very mild impairment (n = 4 in the present study); stage 2, patients with mild impairment (n = 7); stage 3, patients with moderate impairment (n = 8); stage 4, patients with severe impairment (n = 2); and stage 5, patients with extremely severe impairment (n = 0).
After the procedure was approved by our institutional review board and the aims of the study were explained, we recruited 21 volunteers whose ages matched those of the HD carriers (10 women and 11 men; mean age, 54 years ± 13; age range, 30–74 years; 19 volunteers were right handed and two were left handed) among physicians and nurses in the MR unit and their parents; these volunteers participated in the study as healthy control subjects. They had no personal or family history of neuropsychiatric disorders and gave informed consent. Mann-Whitney U testing revealed that the age difference between the HD carriers and the healthy control subjects was not statistically significant (P = .32).
MR Imaging Protocol
All subjects were examined with a clinical 1.5-T MR imaging unit (Gyroscan ACS-NT; Philips Medical Systems, Best, the Netherlands) with 23 mT/m gradient capability. In 13 HD carriers, haloperidol (Serenase; Lusofarmaco, Milan, Italy) was administered before the MR imaging examination was begun.
After scout images were obtained, 40 contiguous coronal 2-mm-thick T1-weighted inversion-recovery turbo spin-echo images (repetition time msec/echo time msec/inversion time msec, 2,964/14/400; turbo factor, 14; number of signals acquired, two) were obtained parallel to the brainstem; 20–24 transverse 5-mm-thick diffusion-weighted images and magnetization transfer–weighted images were then obtained parallel to the bicommissural line from the craniocervical junction to the vertex. The field of view was 230 mm, and the matrix was 192 x 256 for the T1-weighted and magnetization transfer images and 128 x 256 for the diffusion-weighted images. Diffusion-weighted imaging was performed with a double-shot echo-planar imaging sequence (1,600/80–102; echo-planar imaging factor, 15; number of signals acquired, two) with b values of 0 and 1,000 sec/mm2 along the three main body axes and peripheral pulse gating. For magnetization transfer imaging, a T2*-weighted gradient-recalled-echo sequence (640/12; flip angle, 20°; number of signals acquired, two) without and with an off-resonance (1.5 kHz, gaussian envelope duration, 16.4 msec; flip angle, 50°) pulse was used. The MR imaging protocol required 30 minutes.
Image Analysis
Postprocessing was performed at a dedicated workstation (Easy Vision release 4.3; Philips) by two operators (C.T. and R.D.N.) independently. They were blinded to the clinical and genetic data, and the images of the patients and control subjects were presented to the two operators randomly intermixed.
One of the operators (C.T., who had 7 years of experience in interpreting brain MR images) determined the volumes of the left and right caudate nucleus (head and body) and putamen on coronal T1-weighted MR images (Fig 1) by using a manual segmentation technique. To correct for individual size differences, for further analysis, the mean volume—calculated as (VL + VR)/2, where VL is the volume of the left and VR is the volume of the right—caudate nucleus and the mean volume of the putamen were normalized to each subject’s intracranial volume, which was measured as detailed below.
|
The ADCs and MTRs in the left and right caudate nucleus and putamen were determined by using manually drawn regions of interest (Fig 1). The size ranges of the regions of interest were 65–110 mm2 for healthy control subjects and 28–60 mm2 for HD carriers in the caudate nucleus and 120–180 mm2 for healthy control subjects and 100–170 mm2 for HD carriers in the putamen. In addition, on the ADC and MTR maps, two anterior (left and right) and two posterior (left and right) round 5-mm-diameter regions of interest were placed in the periventricular cerebral white matter at a midventricular level. The intracranial volume was manually determined with the ADC maps.
A whole-brain ADC histogram and a brain volume index—calculated as [(VIC – VCSF)/VIC]x 100, where VIC is intracranial volume and VCSF is cerebrospinal fluid volume—were computed (17). In addition, a whole-brain MTR histogram was calculated from the MTR map by applying a threshold for nonzero voxels (18). The median values of the whole-brain ADC and MTR histograms were used because, unlike other histogram metrics, median values are independent of individual head size.
So that we could assess intraoperator variability, each operator repeated, on each of 10 days, measurements of volume, ADC, and MTR of the left caudate nucleus and left putamen and measurements of intracranial volume in one healthy control subject and one HD carrier. The coefficients of variation (SDs divided by the means) ranged between 0.2% for the MTR in the left putamen in the healthy control subject and 7.0% for the volume of the left putamen in the HD carrier. The coefficient of variation for manual segmentation of the intracranial volume with ADC maps was 2%.
Statistical Analysis
We first performed a preliminary multivariate regression analysis, considering age, sex, and all the MR imaging variables.
Because the distribution of the data was not always normal (notably for the volumes of the caudate nucleus and putamen, for the brain volume index, and for the median of the whole-brain ADC in the healthy control subjects), we used nonparametric tests. Moreover, to adjust for multiple comparisons, the significance level was set at P .01 for each test.
Accordingly, after a power analysis for left-versus-right asymmetries (in which all HD carriers and healthy control subjects were considered together), we assessed the differences between the volumes, ADCs, and MTRs of the right and the left caudate nucleus, putamen, and periventricular white matter by using the Wilcoxon test for paired samples (right vs left). To control for possible effects of age on the MR imaging variables, we first assessed, by using the Spearman rank correlation in healthy control subjects and in HD carriers separately, the relationships between age and the following variables: the mean volumes of the caudate nucleus and the putamen; the brain volume index; the mean ADCs and MTRs in the caudate nucleus, putamen, and periventricular white matter; and the median whole-brain ADC and MTR values.
After a preliminary power analysis, the differences between healthy control subjects and HD carriers in terms of intracranial volume; normalized mean caudate nucleus volume; normalized mean putamen volume; mean ADCs and MTRs in the caudate nucleus, putamen, and cerebral periventricular white matter; brain volume index; and median values of the whole-brain ADC and MTR histograms were evaluated with the Mann-Whitney U test.
Last, we assessed the correlations between the stage of disease severity and the above MR imaging variables in the HD carriers by using the Spearman rank correlation coefficient.
The statistical analyses were performed by using StatView software, release 5 (SAS Institute, Cary, NC).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Multivariate regression analysis did not reveal any significant correlations between the MR imaging variables and age or sex in our sample.
Table 1 lists the volumes, ADCs, and MTRs for the right and left caudate nucleus, putamen, and periventricular white matter in healthy control subjects and HD carriers. Statistical analysis revealed an adequate power for assessing possible asymmetries of volume, whereas the power was small for assessing possible asymmetries of the ADC and MTR variables.
|
|
The analysis results indicated an adequate power for assessing possible differences between HD carriers and healthy control subjects for most of the considered MR variables but a small power for assessing possible differences in mean normalized putamen volume and brain volume index. The volumes of the caudate nucleus, putamen, and whole brain were smaller (P < .001) in HD carriers than in healthy control subjects. HD carriers also had increased ADC in the caudate nucleus (P = .002), putamen (P < .001), periventricular white matter (P < .001), and whole brain (P < .001). The MTRs in the caudate nucleus, putamen, white matter, and whole brain were not significantly different between HD carriers and control subjects.
The two asymptomatic HD carriers showed a marked (exceeding 2 SDs of the mean values in the control subjects) increase in whole-brain ADC (0.98 and 1.09 x 10–3 mm2/sec), which, in one of the carriers, was accompanied by an increase in the ADC in the putamen (1.18 x 10–3 mm2/sec) and in the caudate nucleus (0.94 x 10–3 mm2/sec) and by a decrease in the normalized volume of the putamen (0.016).
Table 3 lists the relationships between disease stage and the MR imaging variables.
|
= –0.63), putamen ( = –0.64), and whole brain ( = –0.46) and between increasing stage of clinical disease severity and increase in the ADCs in the caudate nucleus ( = 0.52), periventricular white matter ( = 0.45), and whole brain ( = 0.44) (Fig 2).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
The volumes of the caudate nucleus and putamen we observed in the healthy control subjects are similar to those previously reported (2–4). In agreement with the results of two previous studies (23,24), but at variance with the results of others (4,19,25), we observed a rightward asymmetry of the caudate nucleus in both control subjects and HD carriers; this, however, was not statistically significant.
Previous studies of MR imaging in this setting were focused on asymptomatic or paucisymptomatic subjects with HD (2–5,26). We included patients with moderate and severe disease and were successful in obtaining volume, diffusion, and magnetization transfer data in most of them.
Our results confirm that loss of bulk of the caudate nucleus and the putamen is a fundamental feature of HD (2–5,26).
Although diffusion-weighted and magnetization transfer MR techniques have been extensively used for investigating such white matter diseases as multiple sclerosis and ischemic leukoaraiosis (17,27), few researchers have explored the potential of these techniques for investigating degenerative diseases of the CNS. These are primarily gray matter diseases, being characterized by damage and loss of neurons in the cortical and subcortical gray matter (28). However, concurrent damage of the white matter secondary to neuronal dysfunction and loss has recently been reconsidered in degenerative diseases of the CNS (8–11). In fact, some studies of diffusion in patients with a probable clinical diagnosis of Alzheimer disease revealed an increase in ADC in the cortical gray matter (8) (this increase, however, was mild and nonsignificant in the hippocampus [9]) and an increase in the ADC and a decrease in the fractional anisotropy of the white matter in the temporal lobes (9), association tracts (10), and whole brain (11). To our knowledge, no data are available about diffusion changes in degenerative diseases of the CNS such as HD that predominantly affect the subcortical gray nuclei.
A significant decrease in the MTR in the hippocampus (6) and in the peak height but not in the mean MTR in the whole brain and cortical gray matter (8) have been reported in patients with a clinical diagnosis of Alzheimer disease. Moreover, a decrease in MTR was observed in the basal ganglia of patients with progressive supranuclear palsy but not in Parkinson disease (7).
We chose to examine carriers of the HD mutation with diffusion-weighted and magnetization transfer MR imaging for the following reasons: First, HD is a prototype of the degenerative diseases of the CNS that are associated with CAG triplet expansions (polyglutamine diseases) and also include spinocerebellar ataxias, dentatorubral pallidoluysian atrophy, and spinal and bulbar muscular atrophy (1,29). Second, thanks to genetic testing, we could be sure of the diagnosis and also examine HD carriers in an asymptomatic phase, which would have allowed us to achieve better coverage of the spectrum of severity of the neurodegenerative process. Finally, since deep gray nuclei sustain the major burden of the pathologic process in HD, and evaluation of these structures with diffusion-weighted and magnetization transfer techniques is relatively unaffected by contamination with cerebrospinal fluid (which hinders evaluation of neurodegenerative processes that primarily involve the cerebral cortex—such as Alzheimer disease—with these techniques), it was possible to obtain data about the potentials of these new MR imaging techniques in the evaluation of neurodegenerative processes in general.
The values of ADC in the neostriatum and whole brain in our healthy control subjects were in substantial agreement with those observed in previous studies (8,30,31), which indicated no significant asymmetry (31) and a mild increase in ADC with age that was more pronounced for white matter (20–22).
Our results demonstrate that diffusion-weighted MR imaging can depict tissue damage associated with HD as increased ADC values in the caudate nucleus and putamen. This was accompanied by an increase in whole-brain ADC. Although the precise contribution of gray and white matter changes to this finding could not be ascertained because we did not perform segmentation of gray and white matter (8), the region-of-interest analysis of the cerebral periventricular white matter revealed an increase in the ADC.
It is interesting to note that the increase in the cerebral white matter ADC is in line with the white matter loss emphasized in early HD in a report of a voxel-based volumetry study (26). Moreover, the recent evidence of cortical thinning in HD (26,32), which implies that the disease is not so regionally selective, suggests that secondary damage to the white matter, like that assumed to occur in Alzheimer disease (11), might also occur in HD. Unfortunately, we can only speculate on the basis for regional and global diffusion changes in HD. The hallmarks of brain damage in HD are neuronal loss, increased astroglia, and oligodendrocytes in the neostriatum (33). The link between these neuropathologic features and the increase in ADC in the neoastriatum of HD carriers is far from established and potentially involves changes in the diffusion characteristics of the intra- and extracellular volume fraction, changes in the permeability of cell membranes, and changes in extracellular tortuosity (34).
Two mechanisms have been hypothesized to explain the global increase in the ADC of the brain observed in other hereditary diseases of the CNS such as Leber optic neuropathy (35) and nocturnal frontal lobe epilepsy (36): (a) wallerian degeneration of white matter fibers (ie, disruption and loss of the axonal membranes or myelin secondary to gray matter disease) and (b) a less marked effect of the genetic deficit on all neuronal and axonal populations. Wallerian degeneration and diffuse microscopic damage might contribute to the whole-brain ADC histogram changes observed in HD.
The MTRs we measured in the neostriatum, periventricular white matter, and whole brain of our healthy control subjects are similar to those previously reported (7,8), taking into account the fact that we used a T2*- instead of a T1-weighted magnetization transfer sequence (7) and a different MTR histogram metric (8). The changes in MTR observed in the neostriatum, cerebral white matter, and whole brain of our HD carriers were not statistically significant. This dissociation between ADC and magnetization transfer changes in our HD carriers was unexpected, because in other degenerative diseases, an increase in ADC is combined with a reduction in MTR (6–9). One possible explanation is that while ADC is sensitive to the microscopic structure, magnetization transfer reflects the bound-to-free water ratio and could be a less sensitive marker of tissue change. Incidentally, this dissociation is an indirect confirmation that our measurements of ADC and MTR in the caudate nucleus and whole brain were not affected by partial volume effects with cerebrospinal fluid, which result in a similar degree of increase in ADC and decrease in MTR.
In recent years, advances in understanding the molecular pathophysiology of HD (1,29) have supported and helped promote new therapeutic options for HD on either experimental or clinical grounds (37–39). Because clinical features might not be sensitive enough as outcome measures, instrumental tools are desirable. Our results confirm that volume loss in the neostriatum and whole brain and ADC increase in the caudate nucleus, white matter, and whole brain are correlated with disease severity. Although volume measurements have already proved to be sensitive to disease progression and have been proposed for monitoring response to treatment (5), the potential role of ADC evaluation for these purposes deserves further investigation.
We recognize two main limitations of our study. The first was the small number of presymptomatic carriers examined, which hindered comparison of the sensitivity of volume and ADC measurements to neurodegeneration in HD. The second was the inability to assess the volume, ADC, and MTR in the cerebral cortical ribbon.
In conclusion, our results indicate that volumetric MR imaging reveals loss of volume and diffusion-weighted MR imaging reveals increased ADC in the caudate nucleus and whole brain of HD carriers and that these findings are correlated with the clinical severity of the disease.
|
FOOTNOTES |
|---|
Abbreviations: ADC = apparent diffusion coefficient, CNS = central nervous system, HD = Huntington disease, MTR = magnetization transfer ratio
Author contributions: Guarantor of integrity of entire study, M. Mascalchi; study concepts and design, M. Mascalchi; literature research, C.G.; clinical studies, M. Macucci, R.P., F.L., S.P.; data acquisition, C.T., C.G., L.S.P.; data analysis/interpretation, R.D.N., C.T.; statistical analysis, F.L.; manuscript preparation, M. Mascalchi; manuscript definition of intellectual content, F.L.; manuscript editing and revision/review, M.F.; manuscript final version approval, M. Mascalchi.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Tessa, M. Giannelli, R. Della Nave, C. Lucetti, C. Berti, A. Ginestroni, U. Bonuccelli, and M. Mascalchi A Whole-Brain Analysis in De Novo Parkinson Disease AJNR Am. J. Neuroradiol., April 1, 2008; 29(4): 674 - 680. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Kloppel, B. Draganski, C. V. Golding, C. Chu, Z. Nagy, P. A. Cook, S. L. Hicks, C. Kennard, D. C. Alexander, G. J. M. Parker, et al. White matter connections reflect changes in voluntary-guided saccades in pre-symptomatic Huntington's disease Brain, January 1, 2008; 131(1): 196 - 204. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.P. Dyke, H.U. Voss, D. Sondhi, N.R. Hackett, S. Worgall, L.A. Heier, B.E. Kosofsky, A.M. Ulug, D.C. Shungu, X. Mao, et al. Assessing Disease Severity in Late Infantile Neuronal Ceroid Lipofuscinosis Using Quantitative MR Diffusion-Weighted Imaging AJNR Am. J. Neuroradiol., August 1, 2007; 28(7): 1232 - 1236. [Abstract] [Full Text] [PDF]
|
|
Read all eLetters
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
) and putamen (*) in C, healthy control subject and, D, HD carrier with stage 3 disease.