Brain atrophy in relapsing-remitting multiple sclerosis: relationship with ‘black holes’, disease duration and clinical disability

Journal of the Neurological Sciences 174 (2000) 85–91 www.elsevier.com / locate / jns

Brain atrophy in relapsing-remitting multiple sclerosis: relationship with ‘black holes’, disease duration and clinical disability a a, b a a Andrea Paolillo , Carlo Pozzilli *, Claudio Gasperini , Elisabetta Giugni , Caterina Mainero , Saverio Giuliani c , Valentina Tomassini a , Enrico Millefiorini a , Stefano Bastianello c a

Department of Neurological Sciences, University of Rome ‘ La Sapienza’, Viale dell’ Universita` 30, 00185 Rome, Italy b Department of Neurology, S. Camillo Hospital, Rome, Italy c Department of Neurological Sciences, Neuroradiological Unit, University of Rome ‘ La Sapienza’, Rome, Italy Received 22 October 1999; received in revised form 16 December 1999; accepted 21 December 1999

Abstract Recent MRI studies in multiple sclerosis have highlighted the potential role of brain atrophy evaluation as a putative marker of disease progression. In the present study, we evaluated the supratentorial and infratentorial brain volume in patients with relapsing remitting multiple sclerosis (RR MS) and in healthy subjects. Moreover, we determined whether brain volumes of MS patients are associated with different aspects of brain MRI abnormalities and clinical findings. Two-dimensional acquired MRI was performed on 52 relapsingremitting multiple sclerosis and 30 healthy subjects. The volume of supratentorial and infratentorial structures was measured in selected representative slices. Gd-enhancement, T2 hyperintense, T1 hypointense (i.e. ‘black holes’) total lesion load, as well as the area of corpus callosum was calculated in the MS group and related to brain volume measures. Correlations between MRI parameters and clinical features were also considered. MS patients had significantly lower supratentorial, infratentorial brain volume and corpus callosum area than healthy subjects (P,0.01). Supratentorial brain volume was significantly related to corpus callosum area (r50.58; P,0.01) and T1 hypointense lesion load (r50.48; P,0.01), but not with T2 hyperintense lesion load. Infratentorial / supratentorial ratio was significantly associated with disease duration and EDSS score (r520.34; P50.02 and r520.49; P,0.01, respectively). This study documents that brain atrophy is an early MRI finding in RR MS and it is closely related to ‘black holes’ burden. The use of relative values (infratentorial / supratentorial ratio) may increase the conspicuity of correlation between clinical and MRI findings.  2000 Elsevier Science B.V. All rights reserved. Keywords: Multiple sclerosis; Magnetic resonance; Brain atrophy; EDSS

1. Introduction Magnetic resonance imaging (MRI) techniques have a relevant role in the diagnosis of multiple sclerosis (MS) and particularly in depicting clinically silent lesions which developed inside cerebral structures [1]. The most established MRI parameters in evaluating MS disease outcome are Gd-enhancement and T2 hyperintense lesion load,

*Corresponding author. Tel.: 139-06-4991-4716; fax: 139-06-49914705. E-mail address: [email protected] (C. Pozzilli)

which provide measures of disease activity and progression, respectively [2,3]. Several studies, however, reported that the correlation between changes of T2 lesion load and clinical disability is weak [1,4–7]. This finding has been explained by the low pathological specificity of T2 abnormalities which reflect varying degrees of edema, inflammation, demyelination and axonal loss [1]. Newer MRI techniques have been developed offering the prospect of greater pathological specificity for the more destructive pathological elements of demyelination and axonal loss. Examples include measurements of T1 hypointense lesions (black holes) [8,9], spinal cord and brain

0022-510X / 00 / $ – see front matter  2000 Elsevier Science B.V. All rights reserved. PII: S0022-510X( 00 )00259-8

86

A. Paolillo et al. / Journal of the Neurological Sciences 174 (2000) 85 – 91

atrophy [10–13] magnetization transfer contrast and spectroscopy [14–16]. The presence of brain atrophy in MS has been well documented pathologically [17,18], and there is increasing evidence that atrophy per se is a clinically relevant entity. Earlier studies, evaluating cerebral atrophy in MS, showed some limitations. Firstly, the wide intersubject variation in normal brain size was generally not taken into account [19]. Secondly, arbitrary and indirect brain volume measurements, such as linear indexes, were used to quantify brain volume changes [20–22]. A semiautomated technique was recently applied to quantify brain volumes in both relapsing remitting (RR) and secondary progressive (SP) MS [11]. This study demonstrated that progressive cerebral atrophy can be detected in individual patients with MS, and it is correlated with worsening disability. The aims of the present study were: (1) to investigate the reproducibility of the semiautomated brain volume measurement technique in normal subjects considering not only the supratentorial structures as previously reported [11] but also the infratentorial structures; (2) to compare MRI brain volume measurements of RR MS patients with those of healthy subjects; (3) to determine whether brain volumes of MS patients are associated with disease related MRI and clinical features.

2. Materials and methods

2.1. Patients Fifty-two patients with clinically definite or laboratory supported RR MS [23] were enrolled in the study. They were 35 females and 17 males with a mean age of 32 years (range 17–44). Mean disease duration was 5.6 (range 1–10) and median Expanded Disability Status Scale (EDSS) [24] was 2 (range 1–4). Patients were excluded if they were on clinical relapse or under treatment with corticosteroids within the last month. Other neurological or psychiatric disease, a known history of alcohol intake, were also considered as exclusion criteria. Thirty healthy volunteers, matched with MS patients for age (31 years; range 20–44) and sex (21 female, nine male), served as control group. An informed consent was signed by all study participants.

2.2. MRI protocol Brain MR imaging was obtained with a superconductive 0.5 T magnet (Toshiba S 50). We performed sagittal T1 weighted images (T1WI) (TR 400, TE 18) and axial T2 spin-echo sequences (TR 2500, TE 30 / 90). The enhanced study was obtained directly in the axial plane from 5 to 10 min after injection of 0.1 mmol / kg of gadolinium diethylenetriamine penta-acetic acid (Gd-DTPA).

Eighteen slices of 5 mm thickness with 1 mm of gap were obtained for all sequences using a 25-cm field of view and 1603256 matrix with two excitations. Reproducible head positioning during the study was achieved by performing a midline sagittal scout slice at the beginning of each study. Thereby, we were able to orient axial sections on the same horizontal plane along a line that joined the most infero-anterior and infero-posterior parts of the CC. The same slice locations were used for all the sequences.

2.3. Measurement of brain volume We used the method firstly proposed by Losseff [11], based on a program which extracts cerebral volume from the skull and surrounding tissues, combining a histogrambased automatic thresholding with sequences of morphological operations. Brain extraction was obtained on the Gd-enhanced T1WI, using an algorithm integrated in a window-based image analysis package (eXKalp from D.S. Yoo, Department of Medical Physics and Bio-Engineering, University College, London, UK). Before brain extraction, the scans were examined by an experienced observer (AP), blind to the clinical details, to ensure that repositioning was adequate. To measure supratentorial brain volume (SBV), the observer selected four contiguous slices from each scan with the most caudal at the level of the velum interpositum. The same procedure was followed to obtain a volumetric measurement of infratentorial brain volume (IBV), choosing a single slice located at the level of the middle portion of IV ventricle floor. The choice of this particular slice was influenced by several reasons. Firstly, slices located below include meningeal tissue, which may interfere with the phase of brain extraction and also may be affected by sinus flow motion. Secondly, beyond this slice, supratentorial brain structures (temporal lobe) could be included in the volume analysis. Finally, this area is representative in MS for the development of atrophy as previously reported [25]. The reproducibility of either supratentorial and infratentorial brain volume measurements was evaluated in a sample of eight healthy subjects. They were submitted to monthly scan for a period of 3 months. Brain volume measure at each scan was compared with the previous one. The assumption was that changes in brain volume over a period of 1 month should be small when compared with apparent changes derived by measurement variation. We also evaluated the IBV/ SBV (I / S) ratio, both in MS patients and in control subjects.

2.4. MRI lesion assessment Gd-enhancing total lesion load (Gd-TLL), T2 total hyperintense lesion load (T2-TLL) and T1 total hypointense lesion load (T1-TLL) were calculated on the axial

A. Paolillo et al. / Journal of the Neurological Sciences 174 (2000) 85 – 91

plane. Hypointense lesion was defined as any region with low signal intensity relative to the surrounding white matter visible on enhanced T1WI and corresponding to a region of high signal intensity on T2WI. All these calculations were performed using the display program Dispunc (D.L. Plummer, University College London, UK) with a semiautomated contouring technique [26].

2.5. Corpus callosum ( CC) area assessment For CC area evaluation, two pilot scans (transverse and coronal) were obtained to adjust for rotations of the head and to exclude artificially induced variation in CC. From these scout images, a midsagittal (double-oblique) T1WI spin-echo (SE) image was obtained (TR 400, TE 18). Quantification of the CC area was obtained from the midsagittal MR image by using the display program Dispunc [26], mentioned above, by the same observer. CC areas in both healthy controls and MS patients were obtained by the mean of two repeated measurements performed on two separate occasions.

2.6. Statistical analysis Intra-observer reproducibility was assessed using the statistical method proposed by Bland and Altman [27]. The intra-observer coefficient of variation (COV) was calculated according to the following formula: S.D. (a 1st , b 2nd ) / mean (a 1st , b 2nd ) where S.D. means standard deviation; a 1st and b 2nd are measurements obtained in twice repeated calculation on the same patient. Differences in MRI measurements between RRMS patients and control group were analysed using the Mann– Whitney U statistic. The correlations between brain volume measures (SBV, IBV, CC area) and other MRI parameters (Gd-TLL, T2-

87

TLL, T1-TLL) were calculated by Spearman’s rank correlation coefficient. The same analysis was used to investigate the relationship between brain volume measures and clinical features (age, EDSS and disease duration). A stepwise regression analysis was performed using both EDSS and MRI parameters as the dependent variables, respectively.

3. Results

3.1. Reproducibility The mean COV for SBV in normal subjects was 0.63% with a mean S.D. of measurement variation of 2.1 ml. The mean COV for IBV analysis was 0.23% with a mean S.D. of measurement variation of 0.4 ml, respectively. Hence, changes as little as 0.98% would be outside the 95% confidence limit for that occurring by change due to measurement variation.

3.2. MRI measures in MS patients and control group SBV, IBV, the I / S ratio and the CC area for both MS patients and healthy subjects are reported in Table 1. A significant reduction in SBV, IBV and CC area was found in MS patients (P,0.01) when compared with healthy subjects, while there was no difference in the I / S ratio. Males showed a significantly higher SBV than females, both in MS and healthy controls, while no significant sex differences were observed with regards to IBV and I / Sr. In the MS group, there was no significant difference in CC area between male and female, in contrast to healthy controls. Mean (range) Gd-TLL, T2-TLL and T1-TLL in MS patients, were 0.09 ml (0–2.33), 8.12 ml (1.13–42.03), and 1.04 ml (0.03–20.95), respectively. No significant

Table 1 MRI measures in MS patients and healthy subjects stratified by gender

MS patients Male (n517) Female (n535) Total (n552)

Healthy subjects Male (n511) Female (n519) Total (n530) a

SBV a (ml)

IBV a (ml)

I / S a ratio

CC a area

346.3 (32.9)*3 297.7–408.8 324.2 (21.1)* 291.1–383.6 331.4 (31.8)* 291.1–408.8

27.6 (4.2)* 21.8–35.8 27.1 (3.5)* 21.5–35.1 27.2 (3.7)* 21.5–35.8

0.08 (0.01) 0.07–0.09 0.08 (0.01) 0.06–0.09 0.08 (0.01) 0.06–0.09

509.9 (122.2)* 348.9–761.1 494.9 (106.7)* 331.2–701.4 499.9 (111.1)* 331.2–761.1

414.5 (39.8)3 376.3–499.1 394.3 (37.1) 344.4–412.1 400.5 (48.2) 344.4–499.1

30.9 (3.8) 26.1–39.1 29.8 (4.1) 24.9–36.5 30.7 (3.7) 25.7–39.1

0.07 (0.01) 0.06–0.09 0.08 (0.00) 0.08–0.1 0.07 (0.01) 0.06–0.1

685.1 (82.6)3 600–806.9 629.8 (63.2) 501.6–717.8 644.1 (87.7) 501.6–806.9

SBV, supratentorial brain volume; IBV, infratentorial brain volume; I / S ratio, IBV/ SBV ratio; CC, corpus callosum. All values are expressed as mean (S.D.) and range (below); *P,0.01 from healthy subjects; 3P,0.01 from female.

A. Paolillo et al. / Journal of the Neurological Sciences 174 (2000) 85 – 91

88

Table 2 Significant correlations between different MRI parameters in MS patients

b

b

SBV vs. IBV SBV vs. CC b area SBV vs. T1-TLL CC area vs. T2-TLL CC area vs. T1-TLL

ra

P

0.49 0.58 20.48 20.52 20.53

0.0005 0.0001 0.0005 0.0003 0.0003

a

By Spearman’s rank correlation coefficient. SBV, supratentorial brain volume; IBV, infratentorial brain volume; CC, corpus callosum. b

differences between males and females were found in the rate of MRI abnormalities.

3.3. Relationships between MRI measures In MS patients, significant correlations (P,0.05) between different MRI parameters are listed in Table 2. SBV Table 3 Significant correlations between MRI parameters and demographic–clinical features in MS patients

b

Age vs. SBV Age vs. IBV b Disease duration vs. CC b area Disease duration vs. I / Sr b EDSS vs. I / Sr a

ra

P

20.34 20.45 20.31 20.34 20.49

0.02 0.008 0.03 0.02 0.004

By Spearman’s rank correlation coefficient. SBV, supratentorial brain volume; IBV, infratentorial brain volume; I / Sr, IBV/ SBV ratio; CC, corpus callosum. b

correlates with IBV (r50.49; P50.0005), CC area (r5 0.58; P50.0001) and T1-TLL (r520.48; P50.0005), but not with T2-TLL. On the other hand CC area correlates with both T1-TLL (r520.52; P50.0003) and T2-TLL (r520.53; P50.0003). In healthy subjects, a significant correlation was observed between SBV and CC area (r50.39; P50.02) while only a positive trend was observed between SBV and IBV (r50.32; P50.09) (data not shown).

3.4. Relationship between MRI parameters and clinical / demographic features Significant correlations between MRI parameters and clinical features in MS patients are summarised in Table 3. Age is significantly related to SBV (r520.31; P,0.02) and IBV (r520.38; P,0.008), but not with CC area or I / S ratio. Disease duration is significantly related to CC area (r520.31; P50.03) and I / S ratio (r520.34; P, 0.02). An increase in EDSS is significantly associated with a decrease of I / S ratio (r520.49; P50.004) (Fig. 1). The results of stepwise multiple regression analysis by using EDSS as dependent variable and MRI measures as independent ones, confirm that I / S ratio was significantly associated with EDSS (F57.0, P50.01). Brain atrophy measures were also independently associated with age (SBV: F54.32, P50.04; IBV: F55.87, P50.005) and EDSS (IBV: F56.76, P50.01). In healthy controls, age is significantly related to SBV (r520.33; P,0.03). A trend was seen for association between increase in age and reduction in IBV (r520.30; P50.09) (data not shown).

Fig. 1. Scatterplot showing the relationship between EDSS score and I / S ratio (IBV/ SBV ratio); r520.49; P50.004 by Spearman’s rank correlation coefficient.

A. Paolillo et al. / Journal of the Neurological Sciences 174 (2000) 85 – 91

4. Discussion Our study aims to determine the amount of infratentorial and supratentorial atrophy in patients with RR MS by comparison with healthy controls using a two-dimensional MRI method [11,28,29]. This method is subject to certain limitations inherent in the MRI acquisition protocol which restrict to the axial plane the analysis of the brain volume. Furthermore, our evaluations of SBV and IBV, based on a restricted number of slices, are probably less valid estimates of true brain volume when compared with newer three-dimensional volume measures. Despite these limitations, we confirm that the Losseff’s technique is relatively simple to apply and highly reproducible in quantifying SBV structures. Furthermore, we extended previous findings to IBV measures and to a cohort of healthy subjects in order to avoid any possible bias related to MRI disease activity. Our data showed that brain volume was smaller in RR MS than in healthy subjects. The difference was significant corresponding to a decrease of 17%, 11%, 22%, for SBV, IBV and CC area, respectively. Several earlier studies have reported atrophy of the cerebral hemispheres and CC in MS [21,30–33]. These results are somewhat difficult to compare, however, given differences in subject samples (i.e. sample size, disease type), imaging and data acquisition protocol (i.e. CT vs. MRI) and measurement techniques. More recent studies, using a three-dimensional MRI approach, have attempted to compare brain volumes of MS patients with matched controls. Filippi et al. detected cerebral atrophy in seven out of 15 MS patients affected by either RR or SP disease [34]. Liu et al. found a significant reduction in brainstem (221%), cerebellum (219%) and white matter (212%) volumes as well as in CC area (221%) in 40 MS patients (20 RR, 20 SP), compared to controls [35]. A similar degree of brain volume reduction was detected in our patients, in spite of the fact that they had slight disability (mean EDSS score of 2) and a short disease duration (,10 years). Our findings are consistent with the recent view documenting that axonal damage occurred in acute lesions early in the disease course and that a significant percentage of axons may be lost during the earliest stages of lesion formation [36]. We found that males had greater SBV than females both in MS and in controls. On the other hand, CC area was greater in males than in females in healthy subjects, as previously reported [37,38], but not in MS patients. Whether the extent of periventricular high signal lesions is related to the development of CC atrophy and whether this effect might mask sex difference in CC size, remains difficult to establish. Nevertheless, the degree of CC atrophy is generally associated with the amount of MS lesions [39,40] as also documented in this study by the significant relationship between CC area and both T1-TLL and T2-TLL (see Table 3).

89

In order to better understand the underlying pathological mechanisms of brain atrophy in MS we looked at the relationship between brain volumes (SBV, IBV) and the other MRI measures reflecting tissue damage. In a recent study, Simon et al. found correlation between atrophy measures (linear indexes) with themselves [41]. Our findings are consistent with this study showing significant, albeit weak, correlations between SBV and IBV as well as between SBV and CC area. It is likely that atrophy in different brain structures may occur dependently with one another, resulting from axonal degeneration processes. We found no significant relationship between SBV and MRI findings of disease activity such as Gd-TLL and T2-TLL. Previous studies also failed to demonstrate a correlation between cerebral volumes and both Gd-TLL and T2-TLL [11,34,35] suggesting that each of these measures are examining different pathological processes which may not always result in irreversible tissue destruction. In fact, Gd-enhancement indicates blood brain barrier breakdown and T2-TLL reflects varying degrees of edema, inflammation, demyelination and axonal loss. This study showed a significant correlation between SBV and T1-TLL. Recent pathological studies have demonstrated that T1 hypointense lesions ‘black holes’ are associated with tissue destruction and axonal loss [42]. This suggests that the brain volume may respond to axonal damage by shrinking and increasing neuronal packing density leading to SBV atrophy, as detected by the present technique. A previous study demonstrated no relationship between ‘black holes’ and cerebral atrophy [34] suggesting the independence of the atrophy process from other MRI measures. This discrepancy with our results might be explained by several factors. Firstly, the use of a different measurement method appears to be relevant. Our volumetric analysis is limited to only four MRI ‘slabs’ which mainly include the brain structures located around the lateral ventricles. It is not uncommon to see ‘black holes’ at the tip of the lateral ventricles and therefore, the degree of relationship between T1-TLL and SBV might be overestimated by the present method. Secondly, we use a wider sample with respect to Filippi et al. [34] and this increases the power of statistical analysis. We confirmed previous findings that showed weak or no cross-sectional correlation between SBV and EDSS [11,34,35]. In the present study the only MRI measure correlating with the EDSS is I / S ratio (r520.49; P5 0.004). There is individual variation in absolute brain volume measures related to a number of factors including ageing, sex, body height and weight [43,44]. The use of relative values (i.e. I / S ratio) seems to overcome the limit due to inter-subject brain volume variation revealing a stronger grade of association with EDSS than absolute values. The decrease of I / S ratio is associated with greater

90

A. Paolillo et al. / Journal of the Neurological Sciences 174 (2000) 85 – 91

disability indicating that the major contribution to this relationship is represented by IBV value as supported by the significant association between IBV values with EDSS resulting from the stepwise multiple regression analysis. Infratentorial lesion volume on T2 correlated with EDSS better than supratentorial lesion volume [45]. In one recent report [46], the estimated volume of infratentorial structures from three dimensional-acquired MRI, was found to be significantly correlated with EDSS despite a considerable variance in volume estimates at each disability level. Studies of cerebellar atrophy have also shown significant correlation between the degree of atrophy and cerebellar deficit [25]. Taken together, these findings indicate that demyelination and axonal loss in the infratentorial structures, rather than in the supratentorial ones, are more likely to cause disability in MS. Clinical disability as measured by the EDSS scores is strongly weighted towards locomotion and hence more likely to correlate with destructive changes in the posterior fossa than changes in the brain. However, our results open some questions. Firstly, this cohort of RR MS patients is characterised by a short disease duration (mean 5.6 years) and low EDSS score (range 1–4). At this level of the scale the EDSS score is more weighted to impairment of functional systems than to motor disability [47]. Secondly, we observed a significant association of I / S ratio with disease duration suggesting that a greater decrease in IBV more than SBV may also be a feature of disease duration per se. It is plausible that the percentage of tissue destruction relative to the volume of IBV may be greater than that for SBV implying an early detection of atrophy in the infratentorial structures. Future longitudinal evaluations are required in early MS to evaluate the predictive value of I / S ratio for future disability and to determine whether changes in this measure correlate with EDSS changes over time.

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14] [15]

Acknowledgements We wish to thank Prof. Cesare Fieschi and Prof. Luigi Bozzao for their helpful suggestions and Mr Yoo for giving us the brain volume measurement program.

[16]

[17]

References [18] [1] Miller DH, Grossmann RI, Reingold SC, Mc Farland HF. The role of magnetic resonance techniques in understanding and managing multiple sclerosis. Brain 1998;121:3–24. [2] Harris JO, Frank JA, Patrons N, McFarlin DE, McFarland HF. Serial gadolinium-enhanced magnetic resonance imaging scans in patients with early, relapsing-remitting multiple sclerosis: implications for clinical trials and natural history. Ann Neurol 1991;29:548–55. [3] Miller DH, Albert PS, Barkhof F, Francis G, Frank JA, Hodgkinson S. Guidelines for the use of magnetic resonance technique in

[19]

[20]

monitoring the treatment of multiple sclerosis. Ann Neurol 1996;39:6–16. Gass A, Barker GJ, Kidd D, Thorpe JW, MacManus D, Brennan A, Tofts PS, Thompson AJ, McDonald WI, Miller DH. Correlation of magnetization transfer ratio with clinical disability in multiple sclerosis. Ann Neurol 1994;36:62–7. Filippi M, Paty DW, Kappos L, Barkhof F, Compston DA, Thompson AJ, Zhao GJ, Wiles CM, McDonald WI, Miller DH. Correlation between changes in disability and T2-weighted brain MRI activity in multiple sclerosis: a follow-up study. Neurology 1995;45:255–60. Gasperini C, Horsfield M, Thorpe JW, Kidd D, Barker GJ, Tofts PS, MacManus DG, Thompson AJ, Miller DH, McDonald WI. Macroscopic and microscopic assessments of disease burden by MRI in multiple sclerosis: relationship to clinical parameters. J Magn Reson Imag 1996;6:580–4. Lycklama a Nijeholt GJ, van Walderveen MAA, van Waesberghe JH, Polman C, Scheltens P, Rosier PF, Jongen PJ, Barkhof F. Brain and spinal cord abnormalities in multiple sclerosis. Correlation between MRI parameters, clinical subtypes and symptoms. Brain 1998;121:687–97. van Walderveen MAA, Barkhof F, Hommes OR, Polman CH, Tobi H, Frequin ST, Valk J. Correlating MRI and clinical disease activity in multiple sclerosis: relevance of hypointense lesions on short-TR / short-TE (T1-weighted spin-echo images). Neurology 1995;45:1684–90. Truyen L, van Waesberghe JH, van Walderveen MAA, van Oosten BW, Polman CH, Hommes OR, Ader HJ, Barkhof F. Accumulation of hypointense lesions (‘black holes’) on T1 spin-echo MRI correlates with disease progression in multiple sclerosis. Neurology 1996;47:1469–76. Kidd D, Thorpe JW, Thompson AJ, Kendall BE, Moseley IF, MacManus DG, McDonald WI, Miller DH. Spinal cord MRI using multi-array coils and fast spin-echo. Neurology 1993;43:2632–7. Losseff NA, Wang L, Lai HM, Yoo DS, Gawne-Cain ML, McDonald WI, Miller DH, Thompson AJ. Progressive cerebral atrophy in multiple sclerosis. A serial MRI study. Brain 1996;19:2009–19. Lycklama a Nijeholt GJ, Barkhof F, Scheltens P, Castelijns JA, Ader H, van Waesberghe JH, Polman C, Jongen SJ, Valk J. MRI of the spinal cord in multiple sclerosis. A serial MRI study. Am J Neuroradiol 1997;18:1041–8. Filippi M, Campi A, Colombo B, Pereira C, Martinelli V, Baratti C, Comi G. A spinal cord MRI study of benign and secondary progressive multiple sclerosis. J Neurol 1996;243:502–5. Grossman RI, McGowan JC. Perspectives on multiple sclerosis. Am J Neuroradiol 1998;19:1251–65, Review. van Waesberghe JH, van Walderveen MA, Castelijns JA, Scheltens P, Lycklama a Nijeholt GJ, Polman CH, Barkhof F. Patterns of lesion development in multiple sclerosis: longitudinal observations with T1-weighted spin-echo and magnetization transfer MR. Am J Neuroradiol 1998;19(4):675–83. Arnold DL, Wolinsky JS, Matthews PM, Falini A. The use of magnetic resonance spectroscopy in the evaluation of the natural history of multiple sclerosis. J Neurol Neurosurg Psychiatry 1998;1(Suppl 1):94–101. McDonald WI, Miller DHM, Barnes D. The pathological evolution of multiple sclerosis. Neuropathol Appl Neurobiol 1992;18:319–34, Review. Mews I, Bergmann M, Bunkoski S, Gullotta F, Bruck W. Oligodendrocyte and axon pathology in clinically silent multiple sclerosis lesions. Multiple Sclerosis 1998;4:55–62. Blatter DD, Bigler ED, Gale SD, Johnson SC, Anderson CV, Burnett BM, Parker N, Kurth S, Horn SD. Quantitative volumetric analysis of brain MR: normative database spanning 5 decades of life. Am J Neuroradiol 1995;16:241–51. Rao SM, Bernardin L, Leo GJ, Ellington L, Ryan SB, Burg LS. Cerebral disconnection in multiple sclerosis: relationship to atrophy of the corpus callosum. Arch Neurol 1989;46:918–20.

A. Paolillo et al. / Journal of the Neurological Sciences 174 (2000) 85 – 91 [21] Pozzilli C, Bastianello S, Padovani A, Passafiume D, Millefiorini E, Bozzao L, Fieschi C. Anterior corpus callosum atrophy and verbal fluency in multiple sclerosis. Cortex 1991;27:441–5. [22] Swirsky-Sacchetti T, Mitchell DR, Seward J, Gonzales C, Lublin F, Knobler R, Field HL. Neuropsychological and structural brain lesions in multiple sclerosis: a regional analysis. Neurology 1992;42:1291–5. [23] Poser CM, Paty DW, Scheimberg L et al. New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann Neurol 1983;13:227–31. [24] Kurtzke JF. Rating neurological impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology 1983;33:1444– 52. [25] Davie CA, Barker GJ, Webb S, Tofts PS, Thompson AJ, Harding AE, McDonald WI, Miller DH. Persistent deficit in multiple sclerosis and autosomal dominant cerebellar ataxia is associated with axon loss. Brain 1995;118:1583–92. [26] Grimaud J, Lai M, Thorpe J, Adeleine P, Wang L, Barker GJ, Plummer DL, Tofts PS, McDonald WI, Miller DH. Quantification of MRI lesion load in multiple sclerosis: a comparison of three computed-assisted techniques. Magn Reson Imaging 1996;14:495– 505. [27] Bland JM, Altman DG. Statistical methods of assessing agreement between two methods of clinical measurements. Lancet 1986;1:307– 10. [28] Paolillo A, Coles AJ, Molineux PD, GawneCain M, MacManus D, Barker GJ, Compston DAS, Miller DH. Quantitative MRI in patients with secondary progressive MRI treated with monoclonal antibody Campath 1H. Neurology 1999;53:751–7. [29] Coles AJ, Wing MG, Molyneux PD, Paolillo A, Davie CM, Geoff H, Miller D, Waldmann H, Compston A. Monoclonal antibody treatment exposes three mechanisms underlying the clinical course of multiple sclerosis. Ann Neurol 1999;46:296–304. [30] Loizou LA, Rolfe EB, Hewazy H. Cranial computed tomography in the diagnosis of multiple sclerosis. J Neurol Neurosurg Psychiatry 1982;45:905–12. [31] Noseworthy JH, Paty DW, Ebers GC. Neuroimaging in multiple sclerosis. Neurol Clin 1984;2:759–77. [32] Rao SM, Glatt S, Hammeke TA, McQuillen MP, Khatri BO, Rhodes AM, Pollard S. Chronic progressive multiple sclerosis. Relationship between cerebral ventricular size and neuropsychological impairment. Arch Neurol 1985;42:678–82. [33] Huber SJ, Paulson GW, Shuttleworth EC, Chakeres D, Clapp LE, Pakalnis A, Weiss K, Rammohan K. Magnetic resonance imaging correlates of dementia in multiple sclerosis. Arch Neurol 1987;44(7):732–6. [34] Filippi M, Mastronardo G, Rocca MA, Pereira C, Comi G. Quantitative volumetric analysis of brain magnetic resonance imaging from patients with multiple sclerosis. J Neurol Sci 1998;158:148–53.

91

[35] Liu C, Edwards SGM, Gong Q, Roberts N, Blumhardt LD. Three dimensional MRI estimates of brain and spinal cord atrophy in multiple sclerosis. J Neurol Neurosurg Psychiatry 1999;66:323–30. [36] Trapp BD, Peterson J, Ransohoff RM, Rudick R, Mork S, Bo L. Axonal transection in the lesions of multiple sclerosis. N Engl J Med 1998;338:278–85. [37] Kertesz A, Polk M, Howell J, Black SE. Cerebral dominance, sex and callosal size in MRI. Neurology 1987;37:1385–8. [38] Witelson SF. Hand and sex differences in the isthmus and genu of the human corpus callosum; a postmortem morphological study. Brain 1989;112:799–835. [39] Pelletrier J, Habib M, Lyon-Caen O, Salamon G, Poncet M, Khalil R. Functional and magnetic resonance imaging correlates of callosal involvement in multiple sclerosis. Arch Neurol 1993;50:1077–82. [40] Barkhof F, Elton M, Lindeboom J, Tas MW, Schmidt WF, Hommes OR, Polman CH, Kok A, Valk J. Functional correlates of callosal atrophy in relapsing-remitting multiple sclerosis patients. A preliminary MRI study. J Neurol 1998;245:153–8. [41] The Multiple Sclerosis Collaborative Research Group, Simon JH, Jacobs LD, Campion MK, Rudick RA, Cookfair DL, Herndon RM, Richert JR, Salazar AM, Fisher JS, Goodkin DE, Simonian N, Lajaunie M, Miller DE, Wende K, Martens-Davidson A, Kinkel RP, Munschauer FE, Brownscheidle CM. A longitudinal study of brain atrophy in relapsing multiple sclerosis. Neurology 1999;53:139–48. [42] van Walderveen MA, Kamphorst W, Scheltens P, van Waesberghe JH, Ravid R, Valk J, Polman CH, Barkhof F. Histopathologic correlate of hypointense lesions on T1-weighted spin-echo MRI in multiple sclerosis. Neurology 1998;50(5):1282–8. [43] Pfefferbaum A, Mathalon DH, Sullivan EV, Rawles JM, Zipursky RB, Lim KO. A quantitative magnetic resonance imaging study of changes in brain morphology from infancy to late adulthood. Arch Neurol 1994;51:874–87. [44] Blatter DD, Bigler ED, Gur RC, Johnson SC, Anderson CV, Burnett BM, Parker N, Kurth S, Horn SD. Quantitative volumetric analysis of brain MR: normative database spanning 5 decades of life. Am J Neuroradiol 1995;16:241–51. [45] Giugni E, Pozzilli C, Bastianello S, Gasperini C, Paolillo A, Koudriavtseva T, Frontoni M, Farina D, Bozzao L. MRI measures and their relations with clinical disability in relapsing-remitting and secondary progressive multiple sclerosis. Multiple Sclerosis 1997;3:221–5. [46] Edwards SGM, Gong QY, Liu C, Zvartau ME, Jaspan T, Roberts N, Blumhardt LD. Infratentorial atrophy on magnetic resonance imaging and disability in multiple sclerosis. Brain 1999;122:291–301. [47] Willoughby EW, Paty DW. Scales for rating impairment in multiple sclerosis: a critique. Neurology 1988;38:1793–8.