Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print July 25, 2006, 10.1148/radiol.2403050569
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2403050569v1
240/3/811    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Vrenken, H.
Right arrow Articles by Pouwels, P. J. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Vrenken, H.
Right arrow Articles by Pouwels, P. J. W.
(Radiology 2006;240:811-820.)
© RSNA, 2006

Neuroradiology

Whole-Brain T1 Mapping in Multiple Sclerosis: Global Changes of Normal-appearing Gray and White Matter1

Hugo Vrenken, MSc, PhD, Jeroen J. G. Geurts, MSc, PhD, Dirk L. Knol, PhD, L. Noor van Dijk, MD, Vincenzo Dattola, MD, Bas Jasperse, MD, Ronald A. van Schijndel, MSc, Chris H. Polman, MD, PhD, Jonas A. Castelijns, MD, PhD, Frederik Barkhof, MD, PhD and Petra J. W. Pouwels, PhD

1 From the Departments of Radiology (H.V., J.J.G.G., L.N.v.D., J.A.C., F.B.), Clinical Epidemiology and Biostatistics (D.L.K.), Neurology (B.J., C.H.P.), and Physics and Medical Technology (R.A.v.S., P.J.W.P.), VU University Medical Center, De Boelelaan 1117, 1081 HV Amsterdam, the Netherlands; and Department of Neurosciences, Psychiatric and Anaesthesiological Sciences, University of Messina, Italy (V.D.). Received April 6, 2005; revision requested June 7; revision received September 9; accepted October 14; final version accepted November 1. Supported by the Dutch MS Research Foundation, Voorschoten, the Netherlands, through a program grant and specific project grants to H.V. (grant 98-371 MS) and J.J.G.G. (grant 00-427 MS). Address correspondence to H.V.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Purpose: To prospectively investigate whether T1 changes in normal-appearing white matter (WM) and normal-appearing gray matter (GM) in multiple sclerosis (MS) are global or regional and their relationship to disease type.

Materials and Methods: The institutional ethics review board approved study; written informed consent was obtained. Whole-brain T1 maps were obtained in 67 patients with MS and 24 healthy control subjects with three-dimensional fast low-angle shot flip angle–array method, with correction for B1 imperfections. Analysis of variance was performed on T1 histogram parameters of global normal-appearing WM and GM. Regional mean T1 values were analyzed with a multilevel approach. Multiple linear regression analysis was performed to investigate associations with clinical disability and overall atrophy. For patients, T2 lesion load was determined.

Results: T1 histograms of normal-appearing WM had significantly higher peak positions for patients with MS (792 msec ± 36 in secondary progressive [SP] MS) than for control subjects (746 msec ± 23) and were significantly broader and lower (all P < .001). Histograms for cortical normal-appearing GM were significantly shifted (peak positions, 1263 msec ± 44 in control subjects and 1355 msec ± 62 in patients with SP MS) (P < .001). Histogram peak positions were significantly higher in SP MS than in relapsing-remitting (RR) and primary progressive MS (P < .05). In SP disease, at least 31% of normal-appearing WM and 20% of cortical normal-appearing GM were affected. In MS, T1 was significantly elevated in all normal-appearing WM and cortical normal-appearing GM regions (all P < .01) but was elevated only in the thalamus in deep GM (P < .05). Cortical T1 histogram peak position was associated with clinical disability; T2 lesion load was not.

Conclusion: Results suggest that a global disease process affects large parts of both normal-appearing WM and GM in MS and effects are worse for SP MS than for RR MS.

© RSNA, 2006


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Multiple sclerosis (MS) is characterized by focal areas of demyelination (lesions), which can be visualized with magnetic resonance (MR) imaging. The volume of lesions that are visible at MR imaging, the lesion load, only correlates moderately with clinical disability measurements, possibly in part caused by disease activity outside the MR imaging–visible lesions. Histopathologic findings such as demyelination, inflammation, edema, axonal damage or loss, and glial proliferation have been reported in normal-appearing white matter (WM) (1–3). In normal-appearing gray matter (GM), small focal and diffuse noninflammatory lesions and neuronal death secondary to axonal degeneration have been suggested (4–8). In vivo, abnormalities in both normal-appearing WM and GM have been demonstrated with measurements of the magnetization transfer ratio (9,10), the diffusion tensor (11,12), and metabolite concentrations (13–16).

Most studies in which researchers reported measurements of the T1 relaxation time in normal-appearing WM have revealed an increase in T1 in patients with MS compared with control subjects (6,17–24); increased T1 in normal-appearing GM also has been reported in patients with MS (6,19,21). Direct comparisons between these studies are hampered by differences in acquisition technique, by the fact that in most studies only part of the brain was investigated because of difficulties of measuring T1 in a multisection fashion, and by the effects of B1 imperfections (25–27). The purpose of our study was to prospectively investigate whether T1 changes in normal-appearing WM and GM in patients with MS are global or regional and the relationship of these changes to disease type, clinical disability, and atrophy.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Subjects
Twenty-four healthy control subjects without any neurologic disease and without a history of neurologic symptoms or signs (13 men, 11 women; mean age, 30.8 years; range, 21.0–48.3 years), who were recruited for this study, and 67 patients with clinically definite MS (28) (25 men, 42 women; mean age, 44.0 years; range, 21.8–69.5 years), who were recruited from the outpatient clinic of the VU University Medical Center, Amsterdam, the Netherlands, on the basis of their willingness to undergo extensive MR investigations, participated in this study. The final inclusion criterion was age between 18 and 70 years; the exclusion criterion was recent disease activity (relapse within the previous 4 weeks, assessed with a history). The total group of patients with MS was classified into three groups according to clinical type: Thirteen had primary progressive (PP) MS, 36 had relapsing-remitting (RR) MS, and 18 had secondary progressive (SP) MS (29). Sixteen patients with MS (44%) were undergoing treatment with interferon beta (Rebif, Ares-Serono, Geneva, Switzerland; Avonex, Biogen, Cambridge, Mass; or Betaferon, Schering, Berlin, Germany). No patients received other disease-modifying treatments. The research protocol was approved by the institutional ethics review board, and all subjects gave informed consent. The patients with MS underwent a neurologic examination in which Expanded Disability Status Scale (EDSS) (30) and MS Functional Composite (MSFC) (31) scores were determined. These examinations were performed by medical doctors who had at least 1 year of expertise in the field of MS and who had undergone special training in the performance of examinations with the EDSS and the MSFC.

MR Imaging Protocol
All measurements were performed with a 1.5-T MR system (Magnetom Vision; Siemens, Erlangen, Germany) and the standard circularly polarized transmit-receive head coil. Oblique transverse fast spin-echo intermediate- and T2-weighted MR images were acquired with an in-plane resolution of 1 x 1 mm2 (repetition time msec/echo time msec, 2625/16 [short echo time], 98 [long echo time]; number of signals acquired, two; bandwidth, 130 Hz/pixel). Two interleaved sets of 16 sections each were acquired to obtain 32 contiguous 4-mm-thick sections.

For the T1 measurement (32), six three-dimensional fast low-angle shot image sets were acquired (20/4; number of signals acquired, one; bandwidth, 244 Hz/pixel), with nominal flip angles of 2°, 5°, 10°, 15°, 20°, and 25°. Hereafter, this data set will be referred to as the small–flip angle array. The three-dimensional slab of 128-mm thickness consisted of 32 sections of 4-mm thickness, with the same position, orientation, and resolution as the intermediate- and T2-weighted images. Acquisition time was approximately 8 minutes.

For B1 correction, three-dimensional fast low-angle shot images were acquired with nominal flip angles of 140°, 160°, 180°, 200°, and 220° (25/5; number of signals acquired, one; bandwidth, 244 Hz/pixel). Hereafter, this data set will be referred to as the large–flip angle array. These images were acquired sagitally to avoid infolding in the three-dimensional direction because a nonselective excitation pulse had to be used. The three-dimensional slab of 200-mm thickness completely enclosed the head. Section thickness was 4 mm, in-plane resolution was 2 x 2 mm2, and acquisition time was approximately 12 minutes.

Lesions in Patients with MS and Brain Volume
For the patients with MS, MR-visible abnormalities were marked and outlined by an individual (H.V., with the supervision of F.B. and J.A.C.) on the intermediate-weighted MR images by using a locally developed technique for determining the threshold, and supratentorial lesion loads were calculated. For each subject, the normalized brain volume (NBV)—normalized to skull size—was calculated from the T1-weighted three-dimensional fast low-angle shot images with 20° flip angle, and the semiautomated brain volume measurement method (Structural Image Evaluation, using Normalization, of Atrophy, SIENAX; FMRIB Analysis Group, University of Oxford, Oxford, England) was employed (33).

T1 Calculation
The images from the small–flip angle array were coregistered with trilinear interpolation (FMRIB's Linear Image Registration Tool; FMRIB Analysis Group, University of Oxford) (34,35), and the fast low-angle shot images with a 20° flip angle were used as reference. T1 was then calculated (32) for each pixel by using nonlinear least squares fitting and a hill-climbing algorithm. The values for the flip angle that were used in the calculation were corrected for the local effective B1 field strength. This local effective B1 field strength was calculated from the spatially smoothed images from the large–flip angle array by determining, for each pixel, the ratio between nominal and actual flip angles (32). Besides the T1 maps, maps of the relative spin density, which we termed M0, and the T1 fit error were generated. All B1 and T1 calculations were performed with a computer (Sparc; Sun Microsystems, Santa Clara, Calif) and software programmed by two individuals (R.A.v.S. and H.V.).

The effect of the B1 correction was assessed with performance of three complete measurements on one healthy control subject, with the head placed centrally and at two extreme positions in the coil; thus, we deliberately introduced B1 variations. The three resultant WM histograms of B1-uncorrected T1 differed greatly: peak positions varied between 701 and 749 msec, and peak widths varied between 131 and 205 msec. The B1-corrected T1 WM histograms were narrower and almost identical for the three head positions: peak positions ranged from 719 to 726 msec, and peak widths ranged from 113 to 136 msec.

Global Analysis of Normal-appearing WM and Normal-appearing GM
Automated segmentation.—To investigate the global properties of normal-appearing WM and cortical normal-appearing GM, multispectral automatic segmentation (36) was performed in each subject with template information. Inputs were the intermediate-weighted images, T2-weighted images, and fast low-angle shot images with 20° flip angle, which were coregistered by using the fast low-angle shot images obtained with a 20° flip angle as reference to achieve inherent registration of the T1 maps and segmentation masks. Infratentorial tissue was manually removed by one individual (V.D.) from both the WM and GM masks, and deep GM was removed by the same individual from the GM mask because of segmentation problems in those areas. Masks were eroded and combined with conservative nonlesioned tissue masks derived through coregistration from the outlines of the lesion on the intermediate-weighted MR images to obtain conservative normal-appearing WM and cortical normal-appearing GM masks.

T1 histograms.—For each subject, the normal-appearing WM and cortical normal-appearing GM masks were combined with the T1 map to generate tissue-specific T1 histograms. The T1 histograms were smoothed by using a running average, and afterward, the histograms were normalized by dividing the number of pixels in each bin by the total pixel count and then characterized by three parameters: peak position, peak height, and peak width (full width at half maximum). For the patients with MS, T1 histograms also were generated for the aggregate of all lesions of each patient. Group mean histograms were constructed by averaging smoothed normalized individual histograms.

Regional Analysis of Normal-appearing WM and Normal-appearing GM
With care taken to avoid any visible focal or diffuse abnormalities, regions of interest were manually drawn by an individual (L.N.v.D., with supervision of F.B. and J.A.C.) on the intermediate- and T2-weighted MR images in 15 brain regions that represented three tissue classes (normal-appearing WM, cortical normal-appearing GM, and normal-appearing deep GM), with a sampling of each region bilaterally except the corpus callosum (one region of interest each for the genu and the splenium). The regions of interest were coregistered to the T1 maps, and inappropriate pixels were excluded by means of a pixel-by-pixel comparison with the segmentation-derived masks. Combined bilateral regions of interest that were smaller than 50 pixels (0.2 mL) were excluded. As a result, the analyzed number of measurements per region varied between 84 and 87 of 91 subjects. Mean volumes of the analyzed regions of interest ranged from 0.8 mL for the genu of the corpus callosum to 5.4 mL for the frontal cortex, and they were similar between subject groups. For each analyzed region of interest, the mean T1, standard deviation of T1, and mean T1 fit error were calculated.

Statistical Analysis
For both the global and regional measurements, the mean T1 fit error was compared among groups by using Student t tests. For both normal-appearing WM and GM, the T1 histogram peak position, height, and width were compared among the four subject groups (control subjects and patients with PP MS, RR MS, and SP MS) by using univariate analysis of variance, with subject age as a covariate and Bonferroni correction. The interaction between age and group was included if it was significant (appropriately corrected P < .05) as determined with an F test. The total group of patients with MS was compared with the group of control subjects with additional analysis of variance. NBV was compared among the four subject groups and also between the total group of patients with MS and control subjects in the same way as the histogram parameters were compared. These analyses were performed by using SPSS software (SPSS for Windows, version 11.0, 2001; SPSS, Chicago, Ill).

The regional T1 data were analyzed with software for fitting multilevel models (MLwiN; Centre for Multilevel Modelling, University of Bristol, Bristol, England) (37) by using a separate multilevel model for each of the three classes, normal-appearing WM, cortical normal-appearing GM, and normal-appearing deep GM; subjects were signified as level 2 units, and measurement occasion was signified as a level 1 unit. For each model, the explanatory variables used were group, region, age, sex, the interaction between group and region, and the other significant first-order interactions, which were identified with a model search procedure. The weight (W) was calculated for each measurement from the size (S) of the region, the mean T1 fit error (mT1fiterr), and the variance of T1 (T1var) as W = S/(mT1fiterr · T1var) and then was standardized. Sandwich estimators, which are robust to violations of normality, were used. Two sets of Wald {chi}2 tests were performed for each model: (a) For each region, an overall test was performed for assessment of differences among the four subject groups, and if the overall test results were significant, all pairwise comparisons among groups were performed, with Bonferroni correction and a comparison between control subjects and patients with MS. (b) The same tests as in a were performed but they were performed for each tissue class as a whole (ie, averaged over regions). In case of significant interactions (other than the interaction between subject group and region), the previously mentioned effects were averaged with respect to the appropriate variables (ie, evaluated at the mean value of age and by using effect coding for sex).

By using the SPSS software (SPSS for Windows, version 11.0, 2001; SPSS), correlations of normal-appearing WM and normal-appearing GM histogram peak positions and peak heights with lesion load, EDSS, MSFC, and NBV and correlations between normal-appearing WM and GM histogram parameters were investigated with the Spearman rank correlation coefficient. No Bonferroni correction was applied to the correlations.

By using the SPSS software (SPSS for Windows, version 11.0, 2001; SPSS), multiple linear regression analyses were performed in the combined group of patients with RR MS and patients with SP MS to assess which variables were associated with EDSS and MSFC scores. To achieve normal distributions of the dependent variables, natural logarithm transforms of the data were used instead of the original data. EDSS was transformed to ln(0.5 + EDSS), and MSFC was transformed to ln(1.1 – MSFC). Lesion load and the histogram peak positions and heights of normal-appearing WM and normal-appearing GM were used as explanatory variables. In separate analyses, age and ln(NBV) were added as explanatory variables. The same analyses were performed with ln(NBV) as the dependent variable and the same explanatory variables except ln(NBV).

All analyses were performed by two individuals (D.L.K. and H.V.) together. All results are expressed as the mean ± standard deviation unless indicated otherwise. Appropriately corrected P values less than .05 were considered to indicate a statistically significant difference.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Compared with control subjects, a significant reduction in NBV was observed for patients with SP MS and for the total group of patients with MS (both P < .001). NBV also was lower in patients with SP MS than in those with RR MS (P = .002). Patients with SP MS had largest lesion loads and largest disability, as reflected by highest EDSS values and lowest MSFC values (Table 1, Figs 1 and 2).


View this table:
[in this window]
[in a new window]

  		Table 1. Subject Group Descriptive Characteristics

 

Figure 1
View larger version (36K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1: A, Oblique transverse T1-weighted three-dimensional fast low-angle shot MR image (20/4; bandwidth, 244 Hz/pixel; field of view, 256 mm; in-plane resolution, 1 x 1 mm2) with 20° flip angle depicts section in patient with SP MS. B, T1 map of same section as in A. C, T1 fit error map of same section as in A. D, E, Corresponding tissue-specific T1 maps of normal-appearing WM (D) and normal-appearing GM (E) after conservative segmentation. B shows clear distinction between WM and GM.

 

Figure 2
View larger version (21K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2: Individual normalized unsmoothed histograms of cerebral normal-appearing (NA) WM (left peaks) and cerebral cortical normal-appearing GM (right peaks) of the same patient as in Figure 1 (gray lines) and a healthy control subject (black lines) for comparison. Histogram peaks have symmetric near-normal shape, with clear separation of normal-appearing WM from normal-appearing GM.

 
Mean T1 fit errors (Table 2) did not differ significantly between any pair of groups, neither in total normal-appearing WM or normal-appearing GM nor in any of the separately investigated regions. Regional mean fit errors were similar to global means.


View this table:
[in this window]
[in a new window]

  		Table 2. Global Analysis of T1

 
Global Analysis of Normal-appearing WM and Normal-appearing GM
The global analysis was based on volumes of normal-appearing WM of 141 mL ± 32 in the total group of patients with MS and 157 mL ± 25 in control subjects and on volumes of normal-appearing GM of 219 mL ± 53 in the total group of patients with MS and 308 mL ± 50 in control subjects. The interaction between group and age was not significant for any parameter, so this factor was removed from all models. In the total group of patients with MS, normal-appearing WM peak positions (Table 2) were higher than they were in control subjects (P < .001), and peaks were also lower (P = .005) and broader (P < .001). The differences with respect to control subjects were also significant for SP MS (all P < .001), whereas for RR MS, only peak position (P = .015) and height (P = .018) were significantly different from values in control subjects.

The T1 peak position for normal-appearing GM was increased significantly with respect to control subjects for the total group of patients with MS (P < .001) and for the groups with RR MS (P = .015) and SP MS (P < .001). The group with PP MS showed similar tendencies, but the differences were smaller and not statistically significant. The group with SP MS differed significantly from the group with PP MS in regard to peak positions for normal-appearing WM and normal-appearing GM (P = .019 and P = .003, respectively) and from the group with RR MS in regard to peak position for normal-appearing GM (P = .004) and in regard to peak position (P = .014), width (P = .017), and height (P = .020) for normal-appearing WM.

The group mean T1 histograms of the groups with MS (Fig 3) differed from the histograms of the group of control subjects over the entire range of T1 values, and there was a reduction of pixels with low T1 values and an increase in pixels with higher T1. From the subtraction of group mean histograms, one can calculate the minimum fractions of pixels involved in the changes in the histograms of the groups with MS compared with the group of control subjects. In normal-appearing WM, these fractions were 11% for the group with PP MS, 16% for that with RR MS, and 31% for that with SP MS. In normal-appearing GM, the fractions were 8% for the group with PP MS, 9% for that with RR MS, and 20% for that with SP MS.


Figure 3
View larger version (21K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a: Group mean T1 histograms for (a) normal-appearing WM (NAWM) and (b) normal-appearing GM (NAGM). Both horizontal and vertical scales differ between the two figures. In a, the mean T1 histogram of MS lesions, averaged over all patients, is shown for comparison. This histogram has a peak position around 1000 msec and is much wider than the histograms for normal-appearing WM. In both normal-appearing WM and normal-appearing GM, there is a shift of T1 histograms toward higher T1 values in patients with MS, which is most severe for the group with SP MS.

 

Figure 3
View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b: Group mean T1 histograms for (a) normal-appearing WM (NAWM) and (b) normal-appearing GM (NAGM). Both horizontal and vertical scales differ between the two figures. In a, the mean T1 histogram of MS lesions, averaged over all patients, is shown for comparison. This histogram has a peak position around 1000 msec and is much wider than the histograms for normal-appearing WM. In both normal-appearing WM and normal-appearing GM, there is a shift of T1 histograms toward higher T1 values in patients with MS, which is most severe for the group with SP MS.

 
Regional Analysis of Normal-appearing WM and Normal-appearing GM
The results of overall Wald {chi}2 tests indicated significant differences between subject groups in all regions (Table 3). The groups with RR MS and SP MS and the total group of patients with MS have significantly increased mean T1 in the classes of normal-appearing WM and cortical normal-appearing GM and in all individual normal-appearing WM and cortical normal-appearing GM regions, except the hippocampus in the group with RR MS. In the group with PP MS, the only significant T1 increases were found in the genu of the corpus callosum and in the parieto-occipital and striate cortex. In normal-appearing deep GM, only the thalamus had significantly increased T1 values in patients with MS compared with the group of control subjects, and this difference was driven by the group with SP MS.


View this table:
[in this window]
[in a new window]

  		Table 3. Regional Analysis of T1

 
Correlations with Clinical Parameters and between Histogram Parameters
The only observed correlation with a clinical score was in the total group of patients with MS; in that group, MSFC scores were correlated with peak height for normal-appearing WM ({rho} = 0.251, P = .049).

In the total group of patients with MS, NBV was correlated with peak height for normal-appearing WM ({rho} = 0.510, P < .001) (Fig 4) and with peak position for normal-appearing GM ({rho} = –0.393, P = .001). NBV also was correlated with peak height for normal-appearing WM in the group with RR MS ({rho} = 0.562, P < .001) and with peak position for normal-appearing GM in the group with PP MS ({rho} = –0.582, P = .037) and that with SP MS ({rho} = –0.485, P = .041).


Figure 4
View larger version (14K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4: Scatterplot of T1 histogram peak height for normal-appearing WM (NAWM) versus NBV in control subjects (gray squares) and patients with PP MS ({circ}), RR MS (x), and SP MS ({blacktriangleup}). With increasing heterogeneity of T1 values in normal-appearing WM, as reflected by a decreased peak height, the overall brain volume is decreased. For the total group of patients with MS, Spearman {rho} was 0.510 (P < .001). A similar relationship with NBV was observed for peak position for normal-appearing GM.

 
Lesion load was correlated with all four T1 histogram parameters for normal-appearing WM and normal-appearing GM in the total group of patients with MS, most notably with peak height for normal-appearing WM ({rho} = –0.572, P < .001) and peak position for normal-appearing GM ({rho} = 0.471, P < .001). These two correlations also were observed in the groups with PP MS and RR MS separately (P = .001–.031). In the group with SP MS, lesion load correlated with peak height for normal-appearing GM ({rho} = –0.628, P = .005).

All mutual correlations between histogram parameters were statistically significant in the group with RR MS (all |{rho}| >0.43, all P < .007) and in the total group of patients with MS (all |{rho}| >0.48, all P < .001) (Fig 5), except for the correlation between peak position for normal-appearing WM and peak height for normal-appearing GM. In the group with PP MS, the only correlation was that between peak height for normal-appearing WM and peak position for normal-appearing GM ({rho} = –0.665, P = .013). In the group with SP MS, peak position for normal-appearing WM was correlated with peak height for normal-appearing WM ({rho} = –0.624, P = .006), and further correlations were observed of peak position for normal-appearing GM with peak height for normal-appearing WM ({rho} = –0.705, P = .001), peak position for normal-appearing WM ({rho} = 0.526, P = .025), and peak height for normal-appearing GM ({rho} = –0.667, P = .002). For the group of control subjects, peak position for normal-appearing WM was significantly correlated with peak height for normal-appearing WM ({rho} = –0.458, P = .024) and peak position for normal-appearing GM ({rho} = 0.415, P = .044).


Figure 5
View larger version (13K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5: Scatterplot of T1 histogram peak positions for normal-appearing WM (NAWM) versus those for normal-appearing GM (NAGM) for control subjects and patients with PP MS, RR MS, and SP MS. Keys are the same as in Figure 4. In the total group of patients with MS, the correlation was significant (Spearman {rho} = 0.613, P < .001). Similar correlations were observed for other T1 histogram parameters.

 
In the multiple linear regression analyses, when we used the full set of explanatory variables, both ln(1.1 – MSFC score) (adjusted R2 = 0.288) and ln(0.5 + EDSS score) (adjusted R2 = 0.194) were associated with ln(NBV) alone. With only lesion load and the histogram parameters as explanatory variables, ln(1.1 – MSFC score) was associated with only peak height for normal-appearing WM (adjusted R2 = 0.086), and ln(0.5 + EDSS score) was associated with peak position for normal-appearing GM alone (adjusted R2 = 0.055). When we used the full set of explanatory variables, ln(NBV) was associated with peak height for normal-appearing WM (standardized ß = .550) and age (standardized ß = –.286) (adjusted R2 = 0.413). When we used only lesion load and the histogram parameters as explanatory variables, ln(NBV) was associated with peak height for normal-appearing WM alone (adjusted R2 = 0.343).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
 References
 
Increased T1 throughout Normal-appearing WM and Cortical Normal-appearing GM
Distinct shifts in T1 histograms were observed for normal-appearing WM in patients with MS compared with WM in control subjects. T1 values were increased in all separately investigated normal-appearing WM regions in patients with MS. Both findings agree with previous findings (17,21,23,24). The largest T1 increases occurred in the group with SP MS, a finding that agrees with observations of the magnetization transfer ratio in those patients (38,39). T1 histograms in normal-appearing WM in patients with MS were also lower and wider than those for WM in control subjects, and this finding reflects the heterogeneity of the normal-appearing WM within each patient.

Our data for histograms suggest that T1 increases also occur in a substantial part of cortical normal-appearing GM. T1 also was significantly increased in all investigated cortical normal-appearing GM regions, and this finding agrees with findings in the literature (6,19). In cortical GM, in contrast to WM, very few of the histopathologically observed focal abnormalities can be detected with conventional MR imaging (40). The changes in T1 of cortical normal-appearing GM, thus, may also reflect focal abnormalities, which may be small or large.

Deep GM had increased T1 only in the thalamus. This finding seems to be consistent with previously observed T1 increase and neuronal damage or loss in the thalamus of patients with MS (5,21), as well as with observations that changes in patients with MS appear less evident in deep GM than in cortex (7,41).

From the subtraction of group mean histograms, it was calculated that a large fraction of pixels of both normal-appearing WM and cortical normal-appearing GM had increased T1 in groups with RR MS and SP MS. This finding demonstrates that a substantial fraction of both tissues must be affected by the disease. Results of the regional analyses indicate that these changes can be found throughout the brain, including areas remote from localizations that are typical for MS. It must be concluded that, in MS, disease processes—whether focal or diffuse—outside MR-visible lesions are not limited to a few sites but act throughout the brain and affect large fractions of both normal-appearing WM and cortical normal-appearing GM.

Link between T1 Changes in Normal-appearing WM and Normal-appearing GM and Relationship between Changes and Lesions, Atrophy, and Clinical Measurements
T1 histogram parameters of normal-appearing WM and normal-appearing GM are correlated with each other, with the volume of MR-visible lesions, and with atrophy. They also were correlated with clinical disability as measured by using MSFC scores, although these specific correlations had P values only slightly below the threshold for statistical significance. Given the changes in cortical T1 in patients with MS, it may be interesting to relate cortical T1 values to neuropsychologic measurements. The relationship with lesions may imply that changes of normal-appearing WM and normal-appearing GM are a secondary result of lesions or that a diffuse disease process facilitates lesion formation.

Although the adjusted R2 values were small, the results of the multiple linear regression analyses suggest that the damage to normal-appearing brain tissue has a larger contribution to the progression of atrophy and clinical disability than do T2 lesions.

Our findings in the relatively small group with PP MS suggest less pronounced T1 changes, especially in normal-appearing GM, and these findings are in agreement with those in previous reports about the relatively limited involvement of the brain in patients with PP MS (42,43).

Study Limitations
The observed differences in T1 between control subjects and patients with MS are not an artifact caused by age difference between the control group and the patients with MS because previous observations have shown that T1 remains nearly constant with age in the age range of the subjects in this study (44). Moreover, subject age and its significant interactions were included in the statistical models.

The B1 correction of this T1 measurement requires that the relationship between flip angle and B1 for the small– and large–flip angle arrays is identical. Although the small– and large–flip angle arrays were obtained with a selective and nonselective radiofrequency pulse, respectively, this condition is fulfilled because the outermost sections of the small–flip angle array, which do not contain any brain tissue, are excluded.

Monoexponential T1 relaxation was assumed within each pixel, and this assumption is justified with the conservative identification of normal-appearing WM and cortical normal-appearing GM that carefully helps prevent partial volume effects. The resultant narrow tissue-specific T1 histograms have a near-normal distribution that can be described well with a small number of parameters.

Some pixels that partially contain cerebrospinal fluid may be included. These pixels, however, may well occur more frequently in control subjects than in patients with MS, because as a result of atrophy, the sulci in patients with MS are more likely to become wide enough to be recognized as cerebrospinal fluid, in which case the erosion of the GM mask will remove the partial volume pixels from the GM mask. The much larger relative decrease of analyzed normal-appearing GM volume (29%) than of NBV (4.7%) in patients with MS compared with that in control subjects is in agreement with this interpretation.

Our results suggest that, in patients with MS, there is a global disease process that affects large parts of both normal-appearing WM and normal-appearing GM; this disease process is worse for patients with SP MS than for those with RR MS and is related to overall brain atrophy.


   ADVANCES IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
References
 


   ACKNOWLEDGMENTS
 
The authors thank the patients with MS and the volunteer control subjects who participated in this study.


   FOOTNOTES
 

Abbreviations: EDSS = Expanded Disability Status Scale • GM = gray matter • MS = multiple sclerosis • MSFC = MS Functional Composite • NBV = normalized brain volume • PP = primary progressive • RR = relapsing-remitting • SP = secondary progressive • WM = white matter

Author contributions: Guarantors of integrity of entire study, F.B., P.J.W.P.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, H.V., J.J.G.G., L.N.v.D., P.J.W.P.; clinical studies, H.V., J.J.G.G., L.N.v.D., V.D., B.J., R.A.v.S., C.H.P., J.A.C., F.B., P.J.W.P.; statistical analysis, H.V., J.J.G.G., D.L.K., J.A.C., F.B., P.J.W.P.; and manuscript editing, H.V., J.J.G.G., D.L.K., V.D., B.J., R.A.v.S., C.H.P., J.A.C., F.B., P.J.W.P.

Authors stated no financial relationship to disclose.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 

  1. Kornek B, Storch MK, Weissert R, et al. Multiple sclerosis and chronic autoimmune encephalomyelitis: a comparative quantitative study of axonal injury in active, inactive, and remyelinated lesions. Am J Pathol 2000;157:267–276.[Abstract/Free Full Text]
  2. Lucchinetti C, Bruck W, Noseworthy J. Multiple sclerosis: recent developments in neuropathology, pathogenesis, magnetic resonance imaging studies and treatment. Curr Opin Neurol 2001;14:259–269.[CrossRef][Medline]
  3. van Waesberghe JH, Kamphorst W, De Groot CJ, et al. Axonal loss in multiple sclerosis lesions: magnetic resonance imaging insights into substrates of disability. Ann Neurol 1999;46:747–754.[CrossRef][Medline]
  4. Bo L, Vedeler CA, Nyland H, Trapp BD, Mork SJ. Intracortical multiple sclerosis lesions are not associated with increased lymphocyte infiltration. Mult Scler 2003;9:323–331.[Abstract/Free Full Text]
  5. Cifelli A, Arridge M, Jezzard P, Esiri MM, Palace J, Matthews PM. Thalamic neurodegeneration in multiple sclerosis. Ann Neurol 2002;52:650–653.[CrossRef][Medline]
  6. Griffin CM, Chard DT, Parker GJ, Barker GJ, Thompson AJ, Miller DH. The relationship between lesion and normal appearing brain tissue abnormalities in early relapsing remitting multiple sclerosis. J Neurol 2002;249:193–199.[CrossRef][Medline]
  7. Miller DH, Thompson AJ, Filippi M. Magnetic resonance studies of abnormalities in the normal appearing white matter and grey matter in multiple sclerosis. J Neurol 2003;250:1407–1419.[CrossRef][Medline]
  8. Peterson JW, Bo L, Mork S, Chang A, Trapp BD. Transected neurites, apoptotic neurons, and reduced inflammation in cortical multiple sclerosis lesions. Ann Neurol 2001;50:389–400.[CrossRef][Medline]
  9. Filippi M, Campi A, Dousset V, et al. A magnetization transfer imaging study of normal-appearing white matter in multiple sclerosis. Neurology 1995;45:478–482.[Abstract/Free Full Text]
  10. Dehmeshki J, Chard DT, Leary SM, et al. The normal appearing grey matter in primary progressive multiple sclerosis: a magnetisation transfer imaging study. J Neurol 2003;250:67–74.[CrossRef][Medline]
  11. Cercignani M, Bozzali M, Iannucci G, Comi G, Filippi M. Magnetisation transfer ratio and mean diffusivity of normal appearing white and grey matter from patients with multiple sclerosis. J Neurol Neurosurg Psychiatry 2001;70:311–317.[Abstract/Free Full Text]
  12. Werring DJ, Clark CA, Barker GJ, Thompson AJ, Miller DH. Diffusion tensor imaging of lesions and normal-appearing white matter in multiple sclerosis. Neurology 1999;52:1626–1632.[Abstract/Free Full Text]
  13. Adalsteinsson E, Langer-Gould A, Homer RJ, et al. Gray matter N-acetyl aspartate deficits in secondary progressive but not relapsing-remitting multiple sclerosis. AJNR Am J Neuroradiol 2003;24:1941–1945.[Abstract/Free Full Text]
  14. Chard DT, Griffin CM, McLean MA, et al. Brain metabolite changes in cortical grey and normal-appearing white matter in clinically early relapsing-remitting multiple sclerosis. Brain 2002;125:2342–2352.[Abstract/Free Full Text]
  15. Sarchielli P, Presciutti O, Tarducci R, et al. Localized (1)H magnetic resonance spectroscopy in mainly cortical gray matter of patients with multiple sclerosis. J Neurol 2002;249:902–910.[CrossRef][Medline]
  16. Vrenken H, Barkhof F, Uitdehaag BM, Castelijns JA, Polman CH, Pouwels PJ. MR spectroscopic evidence for glial increase but not for neuro-axonal damage in MS normal-appearing white matter. Magn Reson Med 2005;53:256–266.[CrossRef][Medline]
  17. Griffin CM, Parker GJ, Barker GJ, Thompson AJ, Miller DH. MTR and T1 provide complementary information in MS NAWM, but not in lesions. Mult Scler 2000;6:327–331.[Abstract/Free Full Text]
  18. Lacomis D, Osbakken M, Gross G. Spin-lattice relaxation (T1) times of cerebral white matter in multiple sclerosis. Magn Reson Med 1986;3:194–202.[Medline]
  19. Larsson HB, Frederiksen J, Kjaer L, Henriksen O, Olesen J. In vivo determination of T1 and T2 in the brain of patients with severe but stable multiple sclerosis. Magn Reson Med 1988;7:43–55.[Medline]
  20. Miller DH, Johnson G, Tofts PS, MacManus D, McDonald WI. Precise relaxation time measurements of normal-appearing white matter in inflammatory central nervous system disease. Magn Reson Med 1989;11:331–336.[Medline]
  21. Parry A, Clare S, Jenkinson M, Smith S, Palace J, Matthews PM. White matter and lesion T1 relaxation times increase in parallel and correlate with disability in multiple sclerosis. J Neurol 2002;249:1279–1286.[CrossRef][Medline]
  22. Ropele S, Strasser-Fuchs S, Augustin M, et al. A comparison of magnetization transfer ratio, magnetization transfer rate, and the native relaxation time of water protons related to relapsing-remitting multiple sclerosis. AJNR Am J Neuroradiol 2000;21:1885–1891.[Abstract/Free Full Text]
  23. Vaithianathar L, Tench CR, Morgan PS, Lin X, Blumhardt LD. White matter T(1) relaxation time histograms and cerebral atrophy in multiple sclerosis. J Neurol Sci 2002;197:45–50.[CrossRef][Medline]
  24. van Walderveen MA, van Schijndel RA, Pouwels PJ, Polman CH, Barkhof F. Multislice T1 relaxation time measurements in the brain using IR-EPI: reproducibility, normal values, and histogram analysis in patients with multiple sclerosis. J Magn Reson Imaging 2003;18:656–664.
  25. Kingsley PB, Ogg RJ, Reddick WE, Steen RG. Correction of errors caused by imperfect inversion pulses in MR imaging measurement of T1 relaxation times. Magn Reson Imaging 1998;16:1049–1055.[CrossRef][Medline]
  26. Parker GJ, Barker GJ, Tofts PS. Accurate multislice gradient echo T(1) measurement in the presence of non-ideal RF pulse shape and RF field nonuniformity. Magn Reson Med 2001;45:838–845.[CrossRef][Medline]
  27. Clare S, Jezzard P. Rapid T(1) mapping using multislice echo planar imaging. Magn Reson Med 2001;45:630–634.[CrossRef][Medline]
  28. Poser CM, Paty DW, Scheinberg L, et al. New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann Neurol 1983;13:227–231.[CrossRef][Medline]
  29. Lublin FD, Reingold SC. Defining the clinical course of multiple sclerosis: results of an international survey. National Multiple Sclerosis Society (USA) Advisory Committee on Clinical Trials of New Agents in Multiple Sclerosis. Neurology 1996;46:907–911.
  30. Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology 1983;33:1444–1452.[Abstract/Free Full Text]
  31. Cutter GR, Baier ML, Rudick RA, et al. Development of a multiple sclerosis functional composite as a clinical trial outcome measure. Brain 1999;122:871–882.[Abstract/Free Full Text]
  32. Venkatesan R, Lin W, Haacke EM. Accurate determination of spin-density and T1 in the presence of RF-field inhomogeneities and flip-angle miscalibration. Magn Reson Med 1998;40:592–602.[Medline]
  33. Smith SM, Zhang Y, Jenkinson M, et al. Accurate, robust, and automated longitudinal and cross-sectional brain change analysis. Neuroimage 2002;17:479–489.[CrossRef][Medline]
  34. Jenkinson M, Smith S. A global optimisation method for robust affine registration of brain images. Med Image Anal 2001;5:143–156.[CrossRef][Medline]
  35. Jenkinson M, Bannister P, Brady M, Smith S. Improved optimization for the robust and accurate linear registration and motion correction of brain images. Neuroimage 2002;17:825–841.[CrossRef][Medline]
  36. Zhang Y, Brady M, Smith S. Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm. IEEE Trans Med Imaging 2001;20:45–57.[CrossRef][Medline]
  37. Rasbash J, Browne W, Goldstein H, et al. A user's guide to MLwiN. 2nd ed. London, England: Institute of Education, 2000.
  38. Filippi M, Inglese M, Rovaris M, et al. Magnetization transfer imaging to monitor the evolution of MS: a 1-year follow-up study. Neurology 2000;55:940–946.[Abstract/Free Full Text]
  39. Kalkers NF, Hintzen RQ, van Waesberghe JH, et al. Magnetization transfer histogram parameters reflect all dimensions of MS pathology, including atrophy. J Neurol Sci 2001;184:155–162.[CrossRef][Medline]
  40. Geurts JJ, Bo L, Pouwels PJ, Castelijns JA, Polman CH, Barkhof F. Cortical lesions in multiple sclerosis: combined postmortem MR imaging and histopathology. AJNR Am J Neuroradiol 2005;26:572–577.[Abstract/Free Full Text]
  41. Filippi M, Bozzali M, Comi G. Magnetization transfer and diffusion tensor MR imaging of basal ganglia from patients with multiple sclerosis. J Neurol Sci 2001;183:69–72.[CrossRef][Medline]
  42. Filippi M, Rovaris M, Rocca MA. Imaging primary progressive multiple sclerosis: the contribution of structural, metabolic, and functional MRI techniques. Mult Scler 2004;10(suppl 1):S36–S44.[Abstract/Free Full Text]
  43. Lycklama G, Thompson A, Filippi M, et al. Spinal-cord MRI in multiple sclerosis. Lancet Neurol 2003;2:555–562.[CrossRef][Medline]
  44. Steen RG, Schroeder J. Age-related changes in the pediatric brain: proton T1 in healthy children and in children with sickle cell disease. Magn Reson Imaging 2003;21:9–15.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 Arch NeurolHome page
A. Pulizzi, M. Rovaris, E. Judica, M. P. Sormani, V. Martinelli, G. Comi, and M. Filippi
Determinants of Disability in Multiple Sclerosis at Various Disease Stages: A Multiparametric Magnetic Resonance Study
Arch Neurol, August 1, 2007; 64(8): 1163 - 1168.
[Abstract] [Full Text] [PDF]

Home page
 Mult SclerHome page
H. Vrenken, P.J.W. Pouwels, S. Ropele, D.L. Knol, J.J.G. Geurts, C.H. Polman, F. Barkhof, and J.A. Castelijns
Magnetization transfer ratio measurement in multiple sclerosis normal-appearing brain tissue: limited differences with controls but relationships with clinical and MR measures of disease
Multiple Sclerosis, July 1, 2007; 13(6): 708 - 716.
[Abstract] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2403050569v1
240/3/811    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Vrenken, H.
Right arrow Articles by Pouwels, P. J. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Vrenken, H.
Right arrow Articles by Pouwels, P. J. W.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE