A

P(sMiAtgMRdoAwiCtt©

K

1

vitag

0d

European Journal of Radiology 65 (2008) 421–426

The contribution of diffusion-weighted MR imaging in multiplesclerosis during acute attack

Ismail Yurtsever a, Bahattin Hakyemez a,∗, Ozlem Taskapilioglu b, Cuneyt Erdogan a,Omer Faruk Turan b, Mufit Parlak a

a Uludag University Medical School, Department of Radiology, Bursa, Turkeyb Uludag University Medical School, Department of Neurology, Bursa, Turkey

Received 3 May 2006; received in revised form 27 April 2007; accepted 2 May 2007

bstract

urpose: The aims of the study are firstly, to determine the difference in diffusion-weighted imaging (DWI) in normal appearing white matterNAWM) between patients with acute multiple sclerosis (MS) and controls; secondly, to determine whether there is a correlation between EDSScores and DWI in acute plaques and also NAWM.aterials and method: Out of 50 patients with acute MS attack, 35 patients had active plaques with diffuse or ring enhancement on postcontrast

mages. Eighteen healthy volunteers constituted the control group. While 26 of 35 had relapsing-remitting, 9 had secondary progressive MS.pparent diffusion coefficients (ADC) of the active plaques, NAWM at the level of centrum semiovale and occipital horn of lateral ventricle in

he patients and NAWM in control group were measured. ADC values of active plaques were compared with WM of the patients and the controlroup. The relationship of ADC value of active plaques and WM in MS with expanded disability status scale (EDSS) was investigated by usingann–Whitney U-test.

esults: Of 63 plaques totally, 26 and 37 of the active plaques had diffuse and ring enhancement, respectively. There was no statistically significantifference between ADC value of active plaques and EDSS (p > 0.05). However, there was a statistically significant difference between ADC valuef WM occipital horn and EDSS (p < 0.05). ADC value of active plaques were higher than WM in both groups (p < 0.001). The difference betweenDC value of WM at the centrum semiovale (p < 0.05) and occipital horns (p < 0.001) in patients and controls was statistically significant. Thereas no statistically significant difference between EDSS scores, ADC value at centrum semiovale and WM around occipital horn and active plaques

n subgroups (p > 0.05).

onclusion: Apparently normal tissue in MS patients may show early abnormalities when investigated carefully enough, and there is an even

hough moderate correlation between EDSS and ADC values and early alterations of ADC value are starting in the occipital white matter alonghe ventricles. This has to be verified in larger series.

2007 Elsevier Ireland Ltd. All rights reserved.

irpae

eywords: Diffusion; Magnetic resonance; Multiple sclerosis

. Introduction

Multiple sclerosis (MS) is a chronic disease of the central ner-ous system (CNS) characterized by demyelization and axonalnjury with recurrent attacks [1]. Its pathophysiology is thought

o be related with active T cells which are sensitive to bodyntigens like myelin basic protein and myelin oligodendrocytelycoprotein [2]. Irreversible axonal injury occurs during the

∗ Corresponding author. Tel.: +90 224 4515587; fax: +90 533 4204531.E-mail address: bahattinh@hotmail.com (B. Hakyemez).

Cpsu[

t

720-048X/$ – see front matter © 2007 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.ejrad.2007.05.002

nflammatory process [3]. Four clinical types of MS, namelyelapsing-remitting (RRMS), secondary progressive (SPMS),rimary progressive (PPMS), relapsing-progressing (RPMS)re present [4]. In order to evaluate neurological deterioration,xpanded disability status scale (EDSS) is widely used [5].erebrospinal fluid (CSF) evaluation and multimodal evokedotentials are complementary methods for MS diagnosis. Theupportive CSF studies findings are the increased immunoglob-

lin G level and/or the presence of oligoclonal bands6].

Magnetic resonance imaging (MRI) is widely used inhe evaluation of MS. Plaques are usually ovoid in shape

4 rnal o

atctahamgmpegt[lmpsp[

iabN

2

2

o3Tsucacf

2

gTi(ssNawtodcw

2

gffahaoAaicalF

2

F(

22 I. Yurtsever et al. / European Jou

nd localized periventricularly perpendicular to lateral ven-ricles. Although MS plaques are typically multiple, theyan occur as a large and solitary lesion, called tumefac-ive form, which is quite rare and can be misdiagnosed as

malignant lesion [7]. On T2 images, the plaque appearsyperintense due to increased water content [8]. The lesionsre round or ovoid and iso- or hypointense compared to nor-al appearing white matter (NAWM) on T1 images. After

adolinium injection, active plaques (new developing inflam-atory lesion) may show diffuse or ring enhancement which

oints out destruction of the blood–brain barrier. However,very active plaque may not have contrast enhancement withadolinium [9]. Diffusion-weighted imaging (DWI) measureshe microscopic random translational motion of water molecules10]. It is used in differentiation of various pathologiesike ischemia, infection and tumor [11]. Diffusion of water

olecules are affected by microstructures and microdynamicrocess and apparent diffusion coefficient (ADC) can be mea-ured quantitatively [12]. The diffusion is restricted in MSlaques, and NAWM is actually affected in these patients13].

The aims of our study were firstly, to determine whether theres a difference in DWI in NAWM between patients with acute MSnd controls; secondly to determine whether there is a correlationetween EDSS scores and DWI values in acute plaques and alsoAWM.

. Materials and method

.1. Patients

Fifty patients with acute attack have been studied. A totalf 35 patients, 30 female and 5 male, aged 19–49 years (mean5 ± 2) with active plaques have been included in our study.wenty-six patients had the relapsing-remitting, nine had theecondary progressive MS. Neurologic disability was assessedsing the EDSS and ranged from 0 to 8. The control group

onsisted of 10 male and 8 female subjects, aged between 24nd 52 years (mean 34 ± 3). All controls had normal MR and nolinical evidence of MS. Approval for this study was obtainedrom our institutional review board.

tb

ig. 1. Placement of region of interest (ROI). T2-weighted image (A) shows multipB). ROIs from normal appearing white matter in occipital white matter (C) and norm

f Radiology 65 (2008) 421–426

.2. MR examination protocol

MR was performed by 1.5 T (Vision Plus, Siemens, Erlan-en). FLAIR images (TR/TE/TI 8400/114/2150) in axial plane,2W FSE (TR/TE 5400/99) and T1W SE (TR/TE 550/18)

mages in axial and sagittal planes were obtained. Field of viewFOV) 23 cm, 256 × 256 matrix, 5 mm slice thickness and 1 mmlice gap were obtained. DWI were obtained with single-shot,pin echo echo-planar imaging sequence (TR 6000, TE 139,EX 1, matrix: 96 × 200, FOV: 24 × 24, slice thickness 5 mm

nd slice gap 0 mm) in the axial plane. The diffusion gradientsere sequentially applied in the x-, y-, and z-axis directions with

wo different b values (0 and 1000 s/mm2). Trace images werebtained by simultaneous application of diffusion sensitive gra-ients on three different directions. ADC values were obtainedoncomitantly with the trace images. Postcontrast T1W imagesere obtained in axial, coronal and sagittal planes.

.3. MR analysis

The images have been evaluated at a Workstation (Sliconraphic). It has been evaluated at postcontrast T1W imagesor whether contrast enhancement where MS plaque has beenound with T2W images. Diffuse or ring enhancement has beenccepted as an active plaque. The spatial sites of these plaquesave been determined and DW trace images have been obtainedt the same level by this way. At DW trace images, the intensityf plaque is evaluated according to NAWM. Measurement ofDC values of the plaques have been performed via drawingregion of interest (ROI) with a diameter of 3–5 mm2. If there

s more than one plaque, the average of all plaques has beenalculated. At ADC maps, three different ROIs have been usedt the NAWM of centrum semiovale and the occipital horn ofateral ventricle and the average values have been determined.ig. 1 shows the placement of standard ROI.

.4. Statistical analysis

Paired Sample Correlation test has been used to test the rela-ionship between ADC values of active plaques and NAWMetween the study and control group. Mann–Whitney U-test

le plaques. Square ROI is placed on contrast-enhanced active plaque in imageal centrum semiovale (D) are placed in all cases.

rnal o

htAowMuI

3

iausatbvto

1AimvN

pphapweit(

4

NaNma

ftfiom

Fp(

I. Yurtsever et al. / European Jou

ave been used to determine the relation of ADC values withhe EDSS scores and ages of patients. In addition, patients’DC values of NWM have been compared with ADC valuesf the control group. Additionally relationship of EDSS scoresith ADC values of RRMS and SPMS patients was evaluated.ann–Whitney U-test and Pearson Correlation test have been

sed for the relationship of the EDSS scores and age of cases.n this study, SPSS 11.0 have been used for statistical analysis.

. Results

Sixty-three active plaques were depicted on postcontrastmages, 26 and 37 out of these lesions showed diffuse (Fig. 2)nd ring-shaped enhancement (Fig. 3), respectively. EDSS val-es ranged from 0 to 6.5 (mean 3.2). There was no statisticallyignificant difference between the ADC values of active plaquesnd EDSS scores of patients according to Mann–Whitney U-est (p > 0.05). There was no statistically significant relationshipetween EDSS and ADC values at NAWM of centrum semio-ale according to Pearson Correlation test (p > 0.05). However,he relationship between the ADC values of NAWM aroundccipital horn and EDSS score was significant (p < 0.05).

The ADC value of active plaques was measured as.25–1.67 × 10−3 mm2/s (mean 1.42 ± 0.08) (Table 1). TheDC values of NAWM around centrum semiovale and occip-

tal horn, which had not contained any plaque there has beeneasured as 0.83 × 10−3 ± 0.05 and 1.01 × 10−3 ± 0.08. ADC

alues of active plaques were higher than the ADC values ofAWM (p < 0.001). When ADC values of NAWM were com-

ofen

ig. 2. A symptomatic female patient 47 in years old with RRMS and EDSS scoreriventricular white matter neighboring to the lateral ventricle. There is homogenouB). The lesion is obviously hyperintense with the diffusion-weighted trace image (C

f Radiology 65 (2008) 421–426 423

ared between the control and MS, p values were as follows;< 0.05 at centrum semiovale and p < 0.001 around occipitalorn (Table 2). When the ADC values at centrum semiovalend occipital horn both in MS and control group were com-ared to age, the increase in ADC values was found to be relatedith age (p < 0.05). There was no statistically significant differ-

nce between MS subgroups [RMMS (n = 26), SPMS (n = 9)]n terms of EDSS scores of patients, ADC values at the cen-rum semiovale WM, occipital horn WM and active plaquesp > 0.05).

. Discussion

In our study, ADC values of active plaques were higher thanAWM. But there was no statistical significance between EDSSnd change in ADC values of active plaques. ADC values ofAWM in MS patients were higher than the values of nor-al population. These abnormalities were especially prominent

round occipital white matter.MRI is the most useful technique for both diagnosis and

ollow-up and evaluation of response to treatment [13]. Althoughhe lesions are demonstrated with conventional MRI, insuf-cient information about degree of WM damage may bebtained about the lesions [14–18]. The contrast enhance-ent of plaques is due to endothelial damage and destruction

f the blood–brain barrier [19]. MS plaques show three dif-erent types of enhancement according to histopathology: nonhancement, homogenous and ring enhancement. Homoge-ously contrast-enhanced plaques have obvious inflammation

e as 3.0 (no.: 29) FLAIR (A) image shows hyperintense plaque at the rights contrast enhancement compatible with active plaque with postcontrast image). ADC value of the lesion was 1.39 × 10−3 mm2/s (D).

424 I. Yurtsever et al. / European Journal of Radiology 65 (2008) 421–426

F ). Th( ue wid 10−3

apepethtpnEbwdcivIno

scpecaa

wmrpdR

mlBia6swtfiaadhhAa

ig. 3. A symptomatic 23-year-old female with RRMS and EDSS 3.5 (no.: 24A). There is ring-shaped contrast enhancement compatible with active plaqiffusion-weighted trace image (C). Mean ADC value of the lesions was 1.45 ×

nd moderate myelin loss, whereas total demyelinization anderipheral inflammation have been shown in plaques with ring-nhancement [19,20]. In our study, 59% of contrast enhancedlaques were ring-shaped while 41% showed homogenousnhancement. The association between EDSS and hyperintensi-ies on T2W images and contrast enhancement in active plaquesas been shown to be poor [20]. However, there is an impor-ant relationship between contrast enhancement and activationeriods [21]. Scanderberg et al. have demonstrated statistical sig-ificance between hypointensity of lesion on T1W images andDSS [22]. Plaques hypointense on T1W images are thought toe result of axonal loss. The correlation of hypointense lesionsith age, the duration and natural course of the disease has beenemonstrated [23]. Khoury et al. have demonstrated significantorrelation between the number of lesions on MR and EDSSn the patients during long follow-up. However, it is contro-ersial how these findings correlate with clinical practice [24].ncrease in hypointensity of lesions on T1W images in time hadot been correlated with EDSS because DWI was our major areaf interest and long follow-up was not available in our study.

Magnetization transfer (MT), spectroscopy, DWI and diffu-ion tensor imaging (DTI) are promising techniques overcomingonventional MR limitations. MT is a technique that sup-resses protein-bound water and increases detection of contrast

nhanced plaques [25]. Filippi et al. have demonstrated statisti-al significance between EDSS and cognitive impairment, braintrophy, decrease in MT ratios [26]. N-acetylaspartate (NAA)nd NAA/creatine ratios have been shown to decrease in NAWM

ivAp

ere are multiple plaques at bilateral cerebral hemispheres with FLAIR imageth postcontrast image (B). The plaques are obviously hyperintense with themm2/s (D).

ith proton MR spectroscopy [27]. Recurrent attacks of inflam-ation and reversible axonal damage due to demyelinization

esult in irreversible neuronal loss and gliosis. At the end of thisrocess, NAA decreases in the plaques and NAWM [14]. NAAecrease in NAWM, more pronounced in SPMS and PPMS thanRMS, is correlated to EDSS [28].

DWI is sensitive to translational movements of waterolecules of tissues [19]. Destruction in myelin sheath, axonal

oss or both of them result in widening of extracellular space.ecause of increased water diffusion compared to the NAWM

n MS lesions, higher ADC values are found. ADC values ofcute MS plaques are higher than chronic ones. Approximatelyweeks prior to hyperintensity on DWI, mean ADC values

tart to increase [29]. As a result, restriction in the motion ofater molecules decreases and ADC values increases [10]. In

his study, we investigated association between EDSS and DWIndings. In order to demonstrate this, ADC measurements fromctive plaques and NAWM were performed. ADC values of thective plaques were 1.42 × 10−3 cm2/s. Roychowdhurry et al.emonstrated active plaques as hyperintense on DWI and thatigher ADC values than NAWM [19]. Annette et al. showedigher ADC values in chronic plaques than NAWM [18]. OurDC values are in accordance with other studies in the liter-

ture [30]. ADC values are higher in acute plaques because of

nflammation and edema. During chronic stage, increase in ADCalues is due to demyelinization and axonal damage. However,DC values can not discriminate between acute and chroniclaques [15,17].

I. Yurtsever et al. / European Journal of Radiology 65 (2008) 421–426 425

Table 1MS patients with active plaques and ADC values of normal appearing white matter and active plaques

No. MS type EDSS Age Sex ADC value (×10−3 mm2/s)

NAWM-CS NAWM-OH Active plaque

1 SPMS 4.0 24 F 0.82 1.07 1.252 RRMS 3.5 42 F 0.83 1.02 1.443 RRMS 6 47 F 0.90 1.09 1.454 RRMS 3 45 M 0.84 1.06 1.605 SPMS 2 25 F 0.78 0.96 1.486 RRMS 4 34 F 0.82 1.10 1.507 RRMS 0 33 F 0.76 0.89 1.468 RRMS 1 34 F 0.93 1.70 1.679 RRMS 3 25 M 0.77 0.98 1.44

10 SPMS 3.5 40 F 0.86 1.00 1.4811 SPMS 4.5 44 F 0.89 1.07 1.4112 RRMS 2 44 F 0.85 1.06 1.3813 RRMS 5 46 M 0.77 1.11 1.4514 RRMS 3 29 F 0.84 0.94 1.3115 RRMS 3.5 23 F 0.82 0.98 1.4016 RRMS 2.5 44 F 0.81 1.02 1.4317 RRMS 2.0 19 F 0.85 1.15 1.3318 SPMS 5.5 32 F 0.86 1.05 1.5219 RRMS 3 34 F 0.77 0.92 1.4120 SPMS 4.5 25 F 0.82 0.97 1.3521 SPMS 4 34 M 0.88 1.06 1.5322 RRMS 2 45 F 0.95 1.01 1.4223 RRMS 1 29 F 0.77 0.96 1.3924 RRMS 3.5 23 F 0.78 0.82 1.4125 RRMS 3 49 F 0.83 1.20 1.2826 RRMS 2 24 F 0.79 0.95 1.4427 RRMS 4 42 F 0.75 1.06 1.2928 RRMS 6.5 29 F 0.86 1.04 1.4229 RRMS 3 47 F 0.83 1.02 1.3930 RRMS 6.5 38 F 0.86 0.96 1.4631 RRMS 3 44 F 0.84 0.97 1.3432 SPMS 3 35 F 0.80 0.96 1.4333 RRMS 3 21 F 0.82 0.96 1.2734 RRMS 0 23 F 0.77 0.88 1.3535 RRMS 0 48 M 0.90 1.03 1.28

N cipitas

tepsaMhaAinRD

dorn

casdbtlvalue in all MS lesions and NAWM, there was no correlationbetween increased ADC values and EDSS [32]. DWI, spec-troscopy or MT tehniques to show WM anomalies in early stageof the disease may provide additional prognostic data follow-up.

Table 2ADC values of white matter (WM) in the control and patient group

ote: NAWM, normal appearing white matter; CS, centrum semiovale; OH, occale; SPMS, secondary progressive MS; RRMS, relapsing progressive MS.

There are only few studies in the literature that investigatehe association between ADC values and EDSS. While Mainerot al. [17] demonstrated poor correlation between diffusion oflaques and EDSS, Castriota Scandeberg et al. [22] showedtrong correlation. We did not find any correlation betweenctive plaques and EDSS. Controversial results may be due toS affecting not only the plaques but also NAWM. Astrocytic

yperplasia, edema, perivascular infiltration, abnormal myelinnd axonal damage in NAWM may be explanation of increasedDC values in MS patients [27]. Caramia et al. demonstrated

ncreased ADC values in NAWM [10]. In this study, there waso statistical significance between EDSS and ADC values ofRMS and SPMS patients, in accordance with the results ofroogan et al. [31].Filippi et al. found moderate correlation between EDSS and

iffusivity. In our study, ADC values at centrum semiovale andccipital horn in MS patients were 0.83 × 10−3 and 1.01 × 10−3,espectively, which were higher than the control group. Althougho statistical significance between EDSS and ADC values at

G

CM

l horn; ADC, apparent diffusion coefficient; EDSS, expanded disability status

entrum semiovale was found, there was moderate correlationt the level of occipital horn (p < 0.05). There is more diffu-ion anomaly at the occipital WM than centrum semiovale. Theifference of ADC of NAWM/WM at the occipital horn maye linked to the so called “periventricular leucoencephalopa-hy” a somewhat physiologic finding. This must be verified witharge series. Although Filippi et al. found an increased ADC

roup Centrum semiovale WM(ADC × 10−3 mm2/s)

Occipital horn WM(ADC × 10−3 mm2/s)

ontrol (n = 18) 0.79 ± 0.04 0.89 ± 0.04S (n = 35) 0.83 ± 0.05 1.01 ± 0.07

4 rnal o

Ua

aivoi

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

26 I. Yurtsever et al. / European Jou

nfortunately, our patients were evaluated during acute attacknd no follow-up was present.

Apparently normal tissue in MS patients may show earlybnormalities when investigated carefully enough, and that theres an even though moderate correlation between EDSS and ADCalues and early alterations of ADC value are starting in theccipital white matter along the ventricles. This has to be verifiedn a larger series.

eferences

[1] Hauser SL, Goodkin DE. Demyelinating disease. In: Wilson JD, editor.Harrison principles of internal medicine. New York: Mc Graw-Hill; 2001.p. 2452–60.

[2] Bar-Or A, Oliveria EM, Anderson DE, et al. Moleculer pathogenesis ofmultiple sclerosis. J Neuroimmunol 1999;100:252–9.

[3] Trapp BD, Bo L, Mork S, et al. Pathogenesis of tissue injury in MS lesions.J Neuroimmunol 1999;98:49–56.

[4] Lublin FD, Reingold SC. Defining the clinical course of multiple sclero-sis: results of an international survey. National Multiple Sclerosis Society(USA) Advisory Committee on Clinical Trials of New Agents in MultipleSclerosis. Neurology 1996;46:907–11.

[5] Kurtzke JF. Rating neurological impairment in multiple sclerosis: anexpanded disability status scale (EDSS). Neurology 1983;33:1444–52.

[6] Poser CM, Paty DW, Scheinberg L. New diagnostic criteria for multiplesclerosis: guidelines for research protocols. Ann Neurol 1983;13:227–31.

[7] Osborn AG. Acquired metabolic, white matter, and degenerative diseasesof the brain. In: Osborn AG, editor. Diagnostic neuroradiology. Missouri:WB Saunders; 1994. p. 749–52.

[8] McDonald WI, Miller DH, Barnes D. The pathological evolution of mul-tiple sclerosis. Neuropathol Appl Neurobiol 1992;18:319–34.

[9] Miller DH, Barkhof F, Nauta JJ. Gadolinium enhancement increases thesensitivity of MRI in detecting disease activity in multiple sclerosis. Brain1993;116:1077–94.

10] Caramia F, Pantano P, Di Legge S, et al. A longitudinal study of MR dif-fusion changes in normal appearing white matter of patients with earlymultiple sclerosis. Magn Reson Imaging 2002;20:383–8.

11] Castillo M, Mukherji SK. Diffusion-weighted imaging in the evaluation ofintracranial lesions. Semin Ultrasound CT MRI 2000;21:405–15.

12] Hajnal JV, Doran M, Hall AS, et al. MR imaging of anisotropicallyrestricted diffusion of water in the nervous system: technical, anatomic,and pathologic considerations. J Comput Assist Tomogr 1991;15:1–18.

13] Guo AC, MacFall JR, Provenzale JM. Multiple Sclerosis: diffusion tensorMR imaging for evaluation of normal-appearing white matter. Radiology2002;222:729–36.

14] Inglese M, Li BS, Rusinek H, et al. Diffusely elevated cerebral cholineand creatine in relapsing-remitting multiple sclerosis. Magn Reson Med2003;50:190–5.

15] Poser CM, Brinar VV. Diagnostic criteria for multiple sclerosis. Clin NeurolNeurosurg 2001;103(1):1–11.

[

[

f Radiology 65 (2008) 421–426

16] Werring DJ, Brassat D, Droogan AG, et al. The pathogenesis of lesionsand normal-appearing white matter changes in multiple sclerosis: a serialdiffusion MRI study. Brain 2000;123:1667–76.

17] Mainero C, De Stefano N, Iannucci G, et al. Correlates of MS dis-ability assessed in vivo using aggregates of MR quantities. Neurology2001;56:1331–4.

18] Nusbaum AO, Lu D, Tang CY, et al. Quantitative diffusion measurementsin focal multiple sclerosis lesions: correlations with appearance on T1-weighted MR images. AJR Am J Roentgenol 2000;175:821–5.

19] Roychowdhury S, Maldjian JA, Grossman RI. Multiple Sclerosis: compar-ison of trace apparent diffusion coefficients with MR enhancement patternof lesions. AJNR Am J Neuroradiol 2000;21:869–74.

20] Rovaris M, Comi G, Filippi M. The role of non-conventional MR tech-niques to study multiple sclerosis patients. J Neurol Sci 2001;186(Suppl. 1):S3–9.

21] Paolillo A, Pozzilli C, Gasperini C, et al. Brain atrophy in relapsing-remitting multiple sclerosis: relationship with ‘black holes’, diseaseduration and clinical disability. J Neurol Sci 2000;174:85–91.

22] Castriota Scanderberg A, Tomaiuolo F, Sabatini U, et al. Demyelinatingplaques in relapsing-remitting and secondary-progressive multiple scle-rosis: assessment with diffusion MR imaging. AJNR Am J Neuroradiol2000;21:862–8.

23] Tourbah A, Stievenart JL, Abanou A, et al. Correlating multiple MRIparameters with clinical features: an attempt to define a new strategy inmultiple sclerosis. Neuroradiology 2001;43:712–20.

24] Khoury SJ, Guttmann CR, Orav EJ, et al. Longitudinal MRI in multiplesclerosis: correlations between disability and lesion burden. Neurology1994;44:2120–4.

25] van Waesberghe JH, Castelijns JA, Lazeron RH, et al. Magnetization trans-fer contrast (MTC) and long repetition time spin-echo MR imaging inmultiple sclerosis. Magn Reson Imaging 1998;16:351–8.

26] Filippi M, Bozzali M, Comi G. Magnetization transfer and diffusion tensorMR imaging of basal ganglia from patients with multiple sclerosis. J NeurolSci 2001;183:69–72.

27] Guo AC, Jewells VL, Provenzale JM. Analysis of normal-appearingwhite matter in multiple sclerosis: comparison of diffusion tensor MRimaging and magnetization transfer imaging. AJNR Am J Neuroradiol2001;22:1893–900.

28] Miller DH, Thomson AJ, Filippi M. Magnetic resonance studies of abnor-malities in the normal appearing white matter and grey matter in multiplesclerosis. J Neurol 2003;250:1407–19.

29] Rocca MA, Cercignani M, Iannucci G, et al. Weekly diffusion-weighted imaging of normal-appearing white matter in MS. Neurology2000;55:882–4.

30] Christiansen P, Gideon P, Thomsen C, et al. Increased water self-diffusionin chronic plaques and in apparently normal white matter in patients withmultiple sclerosis. Acta Neurol Scand 1993;87:195–9.

31] Droogan AG, Clark CA, Werring DJ, et al. Comparison of multiple sclerosisclinical subgroups using navigated spin echo diffusion-weighted imaging.Magn Reson Imaging 1999;17:653–61.

32] Filippi M, Cercignani M, Inglese M, et al. Diffusion tensor magnetic res-onance imaging in multiple sclerosis. Neurology 2001;56:304–11.