Brain and spinal cord MR in benign multiple sclerosis: a follow-up study

J O U R NOFTHE A L

NEUROLOGICAL SCIENCES

ELSEVIER

Journal of the Neurological Sciences 143 ( 1996) 143-149

Brain and spinal cord MR in benign multiple sclerosis: a follow-up study Massimo Filippi “*, Adriana Campi b,Vittorio Martinelli a, Bruno Colombo ‘, Giuseppe Scotti b,Giancarlo Comi a ‘ Department of Neurolog.v,Scientific Institute O.spedaleSun R@ele, Ul?ic’er.sityofllilun, vi~zOkettino W 2Q132~ilam [t~d>’ h Department oj’Neur(]rudi(> l(]gy,Scientific Institute Ospedale San Ra~aele, unirersi~’ of ~ilan, ~il~w ItaO Received 25 March 1996; revised 3 June 1996; accepted 21 June 1996

Abstract We performed a clinical and magnetic resonance (MR) longitudinal study in 19 patients with benign multiple sclerosis (MS) to achieve a better definition of the nature of disability in MS. Patients with higher lesionvolumeson conventionalTz-weightedimagesat entrywerethosewith more frequent relapses ( p = 0.0004)and morenew MR lesions( p = 0.003)duringthe follow up. However, I/3 of these new lesions were located periventricularly and about 2/3 were small or intermediate in size. Two of the 11 patients (18Yo)with higher lesion volumes at entry developed progressive neurological deficits: in these two patients the new lesions seen on conventional Tz images had lower magnetization transfer ratios ( p = 0.005)than those presentin patientswho remainedclinicallystable and a marked increasein hypointense lesion volumeson T,-weightedimageswas alsofound.Spinalcordcross-sectionalarea at C5 and MTRvaluesfor the seeminglynormalwhitematterwere similarto thosefoundin normalcontrols.This study suggests that patients with benign MS have two different patterns of disease evolution, one characterized by very low clinical and MR activities, the other in which the lack of disabling symptomatology might be related to factors like site, size and nature of lesions. It also indicates that in patients with benign MS and high MR lesion loads the risk of developing a secondary progressive form of the disease is still present even after many years after onset. Keword.s; Benign multiple sclerosis; Brain and spinal cord magnetic resonance imaging; Magnetization transfer imaging; Disability

1. Introduction The factors underlying the development of disability in multiple sclerosis (MS) are still mostly unknown. Patients with benign MS, who are characterized by an unusually favorable outcome, have been extensively studied to better clarify the nature of disability in MS (Koopmans et al., 1989; Thompson et al., 1990; Thompson et al., 1992; Filippi et al., 1994a; Kidd et al., 1994; Filippi et al., cross-sectional study (Filippi et al., 1995b). In a previotrs 1995b), in which brain magnetic resonance imaging (MRI) and multimodal evoked potential (EP) findings were compared in two samples of patients with benign and secondary progressive MS matched for age, sex and duration of the disease, we found that patients with benign MS might be divided into two groups as regards the extent of brain MRI lesions. One group is characterized by low lesion loads and the other by lesion loads totally overlap-

‘ Corresponding author. Tel: +39 (2) 2643-2231. Fax: +39 (2) 2643-2335.

ping those of the severely disabled patients with secondary progressive MS. In the former group the lack of disabling symptomatology might be related to the mildness of the pathological process, while in the latter group other factors, such as the rate of lesion formation, their location and pathological nature might all be responsible for the absence of severe disability in these patients. Standard Tz-weighted MR sequences give information only about the site and the extent of MS lesions, not allowing any evaluation of the pathological characteristics of the process. Recent studies using magnetization decay analysis (Filippi et al., 1994a), magnetization transfer imaging (MT) (Dousset et al., 1992; Gass et al., 1994; Filippi et al., 1995a), proton spectroscopy (Davie et al., 1995; De Stefano et al., 1995), T1-weighted images (van Walderveen et al., 1995) and spinal cord MRI (Kidd et al., 1993) found putative markers of axonal loss and/or demeylination to be more strictly related to changes in disability than the volume of lesions on standard Tzweighted scans. At present, there are no long-term longitudinal MRI

0022-510X/96 \$15.00 Copyrigbt 0 1996 Elsevier Science B.V. All rights reserved PII S0022-5 10 X(96 )O0207-9

144

M. F’ilippiet al. /JourIKz1 oj”theNeurological Scieme,>143 (1996) 143–149

studies nor cross-sectional studies using MR markers of the presence of destructive pathology aimed at evaluating patients with benign MS. The first aim of this study was to achieve a better definition of the nature of disability in MS by evaluating in a prospective longitudinal study the changes which can be detected using serial conventional MRI in a long-term follow up and by studying newer MR parameters which can give in-vivo more specific information about the nature of the underlying pathological process. The second aim was to evaluate whether the extent of brain MRI lesions is predictive of a subsequent neurological evolution in patients with benign MS.

2. Patients and methods 2.1. Patients Patients with clinically definite MS (Poser et al., 1983) and a benign course were consecutively enrolled and prospectively followed by the MS Center of the Scientific Institute Ospedale San Raffaele, University of Milan. This study was approved by the local Ethical Committee and written informed consent was obtained by all the patients before inclusion in the study. Patients with benign disease were defined as those having a Kurtzke Expanded Disability Status Scale (EDSS) (Kurtzke, 1983) score of 3 or less after a disease duration of greater than 10 years. To be included in the study all patients had to have CSF oligoclonal bands and no clinical relapses and\or had not been tmatcd will] slcroids or psychoh-epic drugs in the previous three months. For all the patients, the number of relapses occurred before they entered the present study were retrospectively counted using the clinical charts. Conventional brain MRI scans were obtained twice in all the patients with an interval between the first and the second scan of at least one year. The second MRI was obtained at least three months after the end of the last relapse and of the treatment with steroids. During the follow up, the number of relapses and the entry into the secondary progressive phase of the disease were recorded. A patient was considered to have a significant progression of disability when a progressive increase of at least 1 point at the EDSS was observed during a 6 month period (confirmed by a second examination three months later) with or without superimposed relapses. At the time MRI were performed and every six months during the follow up, a full neurological examination with disability scored using the EDSS was performed by one observer (VM), unaware of MRI results. 2.2. Brain magnetic resonance imaging Tz-weighted (SE 2400/25–90, 5-mm contiguous axial, 256 X 256 image matrix) and pre- and post- (0.1 mmol\kg of gadolinium-diethy lenetriaminepentaacetic [Gd-DPTA] injected intravenously with a scanning delay of about 5–7

rein) contrast T]-weighted (SE 700/ 17, 5-mm axial contiguous slices, image matrix 192 X 256) MRI scans of the brain were performed in all the patients with a Siemens system operating at 1.5 Tesla at the beginning of the study and at the end of the follow up. Patients were repositioned in the magnet according to EC guidelines (Miller et al., 1991). Scans were examined consensually by two of us (MF and AC),unaware of patient disease patterns; hyperintense lesions on Tz-weighted scans, hypointense lesions on T, -weighted scans and enhancing lesions on post-contrast T,-weighted scans were marked on the hardcopies and their number, size (small: diameter <5 mm; intermediate: diameter between 6 and 10 mm; large: diameter> 10 mm) and location recorded. At entry, patients were divided into two groups, according to the abnormalities present on the first Tz-weighted MRI: (i) patients with low lesion load and (ii) patients with high lesion load. Patients with low lesion load were those with a lesion load lower than the lowest lesion load ( <25 points according to the above mentioned scoring system) detected in patients with secondary progressive MS matched for age, sex and disease duration (Filippi et al., 1995b). Patients with high lesion load were those with lesion loads in the range of lesion loads ( >25 points) detected in the same group of secondary progressive MS (lWppi et al., 1995b). Quantitative assessment of the Tz-weighted and of the unenhanced T,-weighted lesion load was performed by one of us (MF), using the hardcopies. The software used was the ‘‘/usr/image” library (University of North Carolina, Chapel Hill, NC) and image displuy softwwe (David Plummer, University College, London, UK) running on a Sun workstation (Sun Microsystems, Mountain View, CA). The manual tracing measurements were performed using a mouse-controlled cursor on the computer display. The cursor was moved to define the boundary of each lesion, and each outline was stored on computer disk before automatic computation of the lesion volume. The lesion volume was calculated simply as the lesion area multiplied by the slice thickness. 2.3. Spinal cord MR1 Spinal cord MRI were performed with the same machine in 15 patients at follow up. Patients were placed in a comfortable position within the coil, trying to obtain a degree of neck flexion similar in all of them. Planning scans (Tl-weighted SE) were acquired each time in the following order: (a) a coronal localiser was performed; (b) from this a sagittal localiser was planned using an oblique projection if necessary to compensate for patient misalignment. The axial 5-mm-thick slices were acquired using a gradient-echo sequence (TR = 300, TE = 15, flip angle= 15°, FOV = 20 cm, matrix = 256 X 256) and were prescribed oriented from the sagittal scan through the center of the posterior edge of C5, perpendicular to the cervical

M. Filippi et al, /.lourwl

of the Neurological Sciences [43 (1996) 143–149

cord. The resulting image shows low signal cord surrounded by bright CSF. Spinal cord cross-sectional area, transverse and antero-posterior diameters were measured by a single observer (MF) using a mouse-controlled cursor on a computer display. Ten healthy volunteers matched for age, sex, height and weight were the control group. Cord atrophy was considered to be present when the cord area was more than 2 SD below that of the mean areas obtained

for healthy controls. 2.4. Magnetization ttunsjir im.zging MTI studies were performed in 14 patients at follow up, prior to the injection of gadolinium-DTPA, by obtaining 2D GE images (TR/TE = 600/12; slice thickness = 5

mm with an interstice gap = 2 mm; matrix = 192 X 256) with and without a saturation pulse. The saturation pulse had the following parameters: off-resonance gaussian RF pulse centered 1.5 kHz below the water frequency, with a duration of 16.384 ms, a bandwidth = 250 Hz, and a power intensity of 3.4 X 10-6 Tesla. The magnetization transfer was quantified as a percentage of signal loss, according to the following equation: magnetization transfer ratio (MTR) = (So – Ss/So) X 100, in which So is the mean signal intensity for a given region without the saturation pulse and Ss is the mean signal intensity for the same region when the saturation pulse is applied. The MTR of all the new lesions appeared at follow up were measured by one of us (MF). In addition, for each patient MTR was calculated for two regions of NAWM in the frontal lobes. The white matter of patients was considered as NAWM when no visible lesions were present in at least two consecutive slices. In the same areas, MTR were also calculated in 10 age- and sex-matched healthy volunteers. All the quantitative MR measures were made by observers who were unaware of patient disease patterns.

3. Results 3.1. Clinical data Nineteen patients entered the study (15 women and 4 men; mean age i SD = 37 ~ 7 years, mean duration of the disease ~ SD = 13.6 + 2.4 years; median number of relapses = 5, range = 2– 15; median relapse rate = 0.4/year, range = O.1– 1.3/year). Eleven patients were classified as having high lesion loads and eight as having low lesion loads. At entry, the clinical variables of the two groups were not statistically different. The median duration of follow up was 20 months (SD = +-7.1 months) and did not differ for patients with high and low lesion load at entry. Median EDSS was 1.5 both at basal evaluation and follow up, the ranges were 0.0–3.0 at the beginning of the study and 0.0–4.0 at follow up. In 14 patients the same EDSS scores were found at entry and at follow up, in two there

145

was a reduction of 0.5 and in another one an increase of 0.5. One patient entered the secondary progressive phase of the disease and her EDSS increased by 1.5 points. The remaining patient presented major psychiatric disturbances and moderate deterioration of the cognitive functions (especially, long-term memory and ‘frontal’ functions) during the follow up: his EDSS moved from 1.0 to 3.0. These two latter patients had high lesion loads at entry. Six patients had a total of 18 relapses during the follow up: 14 relapses were observed in four patients with high lesion loads (seven of these relapses were detected in the two patients with increased physical or mental disability at follow up) and four in two patients with low lesion loads (chi squared= 12.7, p = 0.0004). 3.2. T2-weighted MRI ctf the brain The median lesion volume was 10840 mm~ (range = 415–65670 mm~) at basal evaluation and 11345 mm3 (range = 445-61400 mm3) at follow up (W= 36; p = 0.01). The lesion volume at basal evaluation correlated with the numbers of relapses from the onset of the disease to the beginning of the present study (SRCC = 0.5, p = 0.02), but not with the duration of the disease (SRCC = 0.1, p = n.s.). We found 455 lesions (378 in the high lesion load group and 77 in the low lesion load group) on the basal MRI scans: 214 (47’%) were located around the ventricles and 241 (53Yo) in regions separated from the ventricles. The proportions of non-periventricular lesions were 51 YO in the high lesion load group and 627o in the low lesion load group. Seventeen new lesions were found in eight patients. Six of these patients had clinical relapses during the follow up, while none of the remaining 11 patients with no new lesions had relapses (chi-squareci = 8.8, p = 0.003). Tbe numbers of relapses and new lesions during the follow up were correlated (SRCC = 0.7, p = 0.002). Six of the patients with new lesions were from the group of patients with high lesion loads at basal evaluation and 15/ 17 new lesions were found in these patients, while the other two new lesions were found in two of the patients with low lesion loads at first MRI (chi-squared = 8.7, p = 0.003). In the high lesion load group, one lesion was located in the cerebellum, one in the internal capsule, one in the occipital lobe, four in the frontal lobe, three in the parietal lobe, and five were located around lateral ventricles. Four of these 15 lesions were small, seven were intermediate and three were large). Six new lesions (one in the cerebellum, one in the occipital lobe, one in the frontal lobe, one in the parietal lobe and two periventricular) were found in the patient who developed a secondary progressive disease course during the period of follow up. The two new lesions (one small and one large) detected in the patients of the low lesion load group were both located periventricularly. In the group of patients with high lesion loads four lesions disappeared (one small in the cerebellum, one

[46

the I I patients with hi@ Icsion loads had no enhancing lcsiomson the IWO scans: the remaining six had 17 enhancing Icsions (13 small. 2 inkrrnediate and 4 large) in at least onc of the two scans. All [hc six patients who had a clinical relapse during the follow up had enhancing lesions in one of the two MRI, while only four of”the 13 with no relapses had such lesions (chi-squrtred = 5.36, p = 0.02). Considering the appearance of new lesion on Tz-weighted images obtained at follow up and the presence ot’enhancing lesions in one of the two scans as indicative of MRI activity, we found the presence of MRI activity in all the six patients who had also clinical activity (new relapses or progressive disability during the follow up) and in five of

the 13 without clinical activity (chi-squared = 4.1, p =

0.04).

Fig. 1. Axial unenhimced T)-weighted image of the patient who dcvclcrped psychiatric and cognitive disturbances. Hypointense lesions are visible around the lateral ventricles.

intermediate and one large in the frontal lobe and one large in the parietal lobe) and four enlarged (one in the basal ganglia and three periventricular). No such changes were found in the group of patients with low lesion loads. 3.3. T,-weighted MN of the brain The median lesion volume of hypointense lesions on unenhanced TI-weighted images was 480 mm~ (range = 0–8640 mms) at the basal evaluation and 480 mm~ (range = 0–9990 mms) at follow up (W= 6; p = n.s.). At the basal evaluation, hypointense T[-weighted lesions were found only in patients with high Tz lesion loads (Fig. 1). Increases in T]-weighted lesion volumes were found in only three patients (all of them were of the high lesion load group). These increases were due to the formation of 11 new hypointense lesions. All these lesions were present on the first TQ-weighted MRI but were isointense on the unenhanced T, -weighted scans. Four of these lesions and 20% increase in hypointense T1 lesion load were detected in the patient who developed secondary progressive MS and six with 647o increase of hypointense T, lesion load in the patient who developed severe psychiatric and cognitive disturbances and who also developed three new lesions on Tz-weighted images. All these lesions were located around the lateral ventricles. 3.4. T,-weighted A4RI after gadolinium-DTPA injection Five of the eight patients with low lesion loads had no enhancing lesions both at entry and at follow up. In the remaining four patients four enhancing lesions (three small and one large) were found in one of the two scans. Five of

Fig, 2, T2-weighted axial gradient-echo images of the cord at C5. In (A), the spinal cm-dimage of the patient who developed secondary progressive MS during the follow up is shown. Atrophy of the cord was detected. A large lesion is also visible. In (B), the spinal cord image of one of the patients who did not have neurological evaltttion during the follow-up is presented for reference.

M. Fili]y)i (,!{II./JoIim
3.5. Spinal cot-d MRI The mean (SD) spinal cord cross-sectional at C5 was 109 (10) mmz in the control group and 104 (1 I) mm2 in the patients with benign MS. This difference was not statistically significant. Spinal cord atrophy was found in 2/15 (1390) patients with benign MS. Both these patients were in the high lesion load group. One of these patients had converted to secondary progressive MS during the follow up (Fig. 2).

3.6. MTI The mean MTR of the eight new lesions present in the two patients who presented neurological evolution was 37.2 (SD = 4.8), while it was 44.6 (SD = 3.6) for the 7 new lesions present in the four patients who remained benign MS at the end of the follow up. This difference was statistically significant (t = 3.3; p = 0.005). The mean MTR of the NAWM was 48.2 (SD = 1.4) for the 13 patients studied (in one of them it was not possible to identify an area of NAWM in two consecutive slices) and 51.1 (SD = 1.5) for the normal white matter of the controls (p < 0.0001). No significant difference was found when the MTR values of the NAWM were compared between patients with high and low lesion load. In both the cases who developed progressive neurological deficits, the MTR values of the NAWM were within the 2 SD from the mean obtained for the normal white matter of controls.

4. Discussion Patients with MS may have a particularly mild disease, determining a low-grade disability even after several years from the onset (Matthews, 1991). This clinical subgroup of the disease, called benign MS, has been used in several MRI studies aimed at defining the nature of disability in MS (Koopmans et al., 1989; Thompson et al., 1990; Thompson et al., 1992; Filippi et al., 1994a; Kidd et al., 1994; Filippi et al., 1995b). However, in some of the studies (Koopmans et al., 1989; Thompson et al., 1990; Filippiet al.,1994a), patients with benign MS were found to have MRI lesion loads on conventional T2-weighted images similar to those found in patients with secondary progressive MS, whose disease course is highly disabling. This clinical\MRI discrepancy may be explained at least by three reasons. First, it is now clear that not only the extent of lesions, but also the location and the intrinsic nature are relevant in causing disability in MS (Dousset et al., 1992; Gass et al., 1994; Filippi et al., 1994a; Filippi et al., 1995a and Filippi et al., 1995b; van Walderveen et al., 1995; Davie et al., 1995; De Stefano et al., 1995). Secondly, the severity of spinal cord involvement in MS has not been studied until recently (Kidd et al., 1993) and it is surely a relevant factor for permanent disability. Thirdly,

147

previous studies were cross-sectional (Koopmans et al., 1989; Thompson et al., 1990; Filippi et al., 1994a; Filippi et al., 1995b) or had follow up periods lasting 6 months or less (Thompson et al., 1992; Kidd et al., 1994). In a previous cross-sectional study from our group (Filippi et al., 1995b), comparing brain MRI and multimodal EP in patients with benign and secondary progressive MS, we found that the extent of lesions together with their location and nature might all be relevant in explaining the mildness of the disease process in benign MS. In detail, we found that in some patients with benign MS the lack of disabling symptomatology might be related to the presence of only few brain MRI lesions. On the other hand, other mechanisms should be effective in the remaining patients with benign MS and MRI lesion loads completely overlapping those of patients with secondary progressive MS. This study suggests that patients with benign MS have two different patterns of disease evolution. The first is characterized by a disease activity which is very low both clinically and subclinically. Patients with low lesion loads at entry had indeed fewer relapses and new and enhancing lesions compared to those with high lesion loads at entry. Therefore, it is possible that in a subgroup of patients with benign MS the lack of disability is related to a low rate of disease activity. This finding agrees with previous studies indicating that the rate of appearance of enhancing lesions is lower in benign MS compared to early relapsing-remitting MS (Thompson et al., 1992; Kidd et al., 1994). In addition, a previous MRI study (Filippi et al., 1995c) has already suggested that in patients with relapsing-remitting MS, who are characterized by a clinical pattern of disease evolution qualitatively similar to that of benign MS, increase in disability, which is usually due to incomplete recovery from relapses, is related to new lesion formation. The second pattern of disease evolution in benign MS is characterized by higher clinical and MRI activities. Patients with this disease pattern experienced more frequent clinical relapses and had higher numbers of new and enhancing lesions. Thus, other factors have to be considered to explain the lack of disability in such patients. The size and the site of the new lesions might be two of these factors: 1/3 of the new lesions in patients with high lesion loads were located periventricularly and about 2/3 were small or intermediate. However, the intrinsic nature of the lesions seems to be even more important. Hypointense T, lesions were typically located in periventricular regions and the extent of these abnormalities has been found to correlate well with disability (van Walderveen et al., 1995), since they probably reflect tissue destruction (Loevener et al., 1995). In addition, the mean MTR value of new lesions found in patients who did not develop clinical evolution was higher than that of new lesions of the two patients with increased disability. MTR is considered a reliable and objective method to evaluate the degree of macromolecular destruction (demyelination and axonal 10SSin MS) (DOUS-

set et al., 1992) and it has been found to correlate with disability better than lesion load (Gass et al., 1994). This study also indicates that in the subgroup of patients with benign MS and high MRI lesion loads the risk of developing a secondary progressive form of the disease is still present even after many years from the onset. We found that 2/11 (18%) patients with high lesion loads developed progressive physical or mental disability during the follow up. This finding confirms previous clinical observations made on larger samples (McAlpine, 1964; Riser et al., 1971) the prognostic value of the extent of brain MRI abnormalities also for patients with benign MS, as already demonstrated for clinically isolated syndromes suggestive of MS (Filippi et al., 1994b) and, to some” extent, for patients with the established form of the disease (Filippi et al., 1995c). It is also important that other MR measures were found to be abnormal in the two patients who developed progressive neurological deficits. In these two patients we found: (a) larger increase in T, hypointense lesions. This increase was particularly high for the patient who developed psychiatric and cognitive disturbances. In this patient, the T, hypointense lesions were located periventricularly and they might have affected cortical functions through diaschisis phenomena (Comi et al., 1991); (b) reduced MTR values for the newly-formed lesions; and (c) in the patient who developed a secondary progressive MS with increased pyramidal and cerebella dysfunction,spinal cord atrophy at C5 level. Spinal cord atrophy probably results from degeneration of long tract fibers secondary to severe tissue damage and a recent MRI study found this parameter to be correlated with the degree of disability in patients with MS (Kidd et al., 1993). All these findings agree with previous studies demonstrating that several non-conventional MR parameters well correlate with the degree of disability in MS (Dousset et al., 1992; Gass et al., 1994; Filippi et al., 1994a; Filippi et al., 1995a and Filippi et al., 1995b; van Walderveen et al., 1995; Davie et al., 1995; De Stefano et al., 1995). On the contrary, it seems to be less important the role of subtle changes of the NAWM. A consequence of these considerations is that the definition of benign MS needs to be changed at least for those patients with high MRI lesion loads. In such cases, a disease duration longer than 10 years should be required together with the demonstration of the absence of clinical and MR (both conventional and non-conventional) activity.

Acknowledgements The Dispimage package display program was written and provided by David Plummer, Department of Medical Physics, University College London (UK).

References Comi, G., Pemni, D,, Martinelli, V., Filippi, M,, Parrlesu, E., Lenzi, G., Bettinardi, V,, Vaghi, M., Pozzilli, C,, Fieschi, C., Canal, N,, Fazio, F. (1991) Cognitive function impairment in multiple sclerosis: a 18F-FDG and PET study. In: Wietholter, H,, Dichgans, J,, Mertin, J. (Eds.), Current Concepts in Multiple Sclerosis, Elsevier Science Publishers, Amsterdam, pp. 55-60, Davie, C.A., Barker, G.J., Webb, S,, Tofts, P.S., Thompson, A.J,, Hmding, A.E., McDonald, W,I., Miller, D.H. (1995) Persistent functional deficit in multiple sclerosis and artosomal dominant cerebella ataxia is associated with axonal loss. Brain, 118: 1583–1592. De Stet’ano, N., Matthews, P.M., Antcl, J.P., Preul, M., Francis, G.S,, Arnold, D.L. (1995) Chemical pathology of acute demyelinatirrg lesions and its correlation with disability. Ann Neurol 38: 901–909. Dousset, V., Grossman, R.I,, Ramer, K.N., Schnall, M.D., Young, L,H., Gonzales-Scarano, F, Lavi, E., Cohen, J.A. (1992) Experimental allergic encephalomyelitis and multiple sclerosis: lesion characterization wi~hmagrrctic transtcr imaging. Radiology, 182: 4X3–91. Filippi, M,, Barker, G.J., Horsliclcl, M.A., Sacares, P.S., MacManus, D.G., Thompson, A,J,, ‘lofts, P.S., McDonald, W,I,, Millel, D,H. ( 1994a) A quantitative brain MRI study of hmrign and sccmrdary progressive multiple sclerosis. J Neurol, 241: 246-251, Filippi, M., Campi, A., DOUSWX, V,, BwaUi, C., MartincUi, V., Canal, N., Scotti, G., Comi, G. (1995a) A magnetization transfer imaging study of normal-appearing white matter in multiple sclerosis. Neurology, 45: 478–82, Filippi, M., Campi, A., Mammi, S,, Martinelli, V,, Locatelli, T,, Scotti, G., Amadio, S., Canal, N., Comi, G, (1995b) Brain magnetic resonance imaging and multimodal evoked potentials in benign and secondary progressive multiple sclerosis, J Neurol Neurosurg Psychiatry, 58: 3 1–37. Filippi, M,, Horsfield, MA., Morrissey, S.P,, MacManus, D.G., Rudge, P., McDonald, WI., Miller, D,H. (1994b) Quantitative brain MRI lesion load predicts the course of clinically isolated syndromes suggestive of’multiple sclerosis. Neurology, 44: 635–41, Filippi, M., Paty, D.W., Kappos, L., Barkhof, F., Compston, D.A,S., Thompson, A.J., Zhao, G.J., Wiles, C.M., McDonald, W.I., Miller, D.H. ( 1995c) Correlations between changes in disability and T2weighted brain MRI activity in multiple sclerosis: a follow up study. Neurology, 45: 255–60. Gass, A., Barker, G.J., Kidd, D., Thorpe, J.W., MacManus, D.G., Brennan, A,, Tofts, P.S., Thompson, A.J., McDonald, W.I., Miller, D.H. (1994) Correlation of magnetisation transfer ratio with clinical disability in multiple sclerosis, Ann Neurol, 36: 62–67. Kidd, D., Thompson, A.J., Kendall, B.E., McDonald, WI., Miller, D.H. (1994) Benign form of multiple sclerosis: MRI evidence for less frequent and less inflammatory disease activity. J Neurol Neumsurg Psychiatry, 57: 1070-1072. Kidd, D., Thorpe, J.W., Thompson, A.J., Kendall, B.E., Moseley, I.F., MacManus, D.G., McDonald, W.I., Miller, D.H. (1993) Spinal cord MRI using multi-array coils and fast spin echo. IL Findings in multiple sclerosis. Neurology, 43: 2632–2637. Koopmans, R.A., Li, D.K.B., Grochowski, E., Cutler, P.J., Paty, D.W. (1989) Benign versus chronic progressive multiple sclerosis: magnetic resonance imaging features. Ann Neurol, 25: 74–81. Kurtzke, J.F. (1983) Rating necrologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology, 33: 1444– 1452. Loevener, L.A,, Grossman, R.I., McGowan, J.C., Ramer, K.N., Cohen, J.A. (1995) Characterization of multiple sclerosis plaques with Tl weighted MR and quantitative magnetization transfer. AJNR 16: 1473-1479. Matthews, W.B. (1991) Course and prognosis. In: Matthews, W.B. (Ed), McAlpine’s Multiple Sclerosis. Second Edition, Churchill Livingston, Edinburgh, London, Melbourne and New York, pp. 139–163.

M. Filippi e~al. /Juwrrol of the Neurolofiical Sciences 143 (1996) 143 149 McAlpine, D. (1964) The benign fom of multiple sclerosis: results of a long-term study. W Med J, 2; 1153-1155. Miller, D.H,, Barkhof, F., Berry, I., Kappos, L., Scntti, G., Thmnpwm, A.J. (1991) Magnetic resonance imaging in monitoring the lrcatment of multiple sclerosis: C.nnccrlcd Action Guidelines. J Nenrol Ncurosurg Psychiatry, 54: 683–8. Poser, C.M., Paty, D.W., Scheinberg, L., McDonald, W.I., Davis, F.A., Ebers, G.C., Johnson, K.P., Sibley, W.A., Silberberg, D,H., TourtelIotte, W,W. (1983) New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann Neurol, 13: 227-231. Riser, M., Gemud, J., Rascol, A., Benazet, A.M., Segria, G.M. (1971) L’6vohrtion de la scl~rose en plaques ({tude de 201 observations suivies au-del?ide 10 ans). Revue Neurologique, 124: 479–484.

149

Thompson, A.J., Kermocfe,A.G,, MacManus, D.G., Kendall, B.E., KingsIcy, D.P,E., Moseley, I.F., Mcf)onald, W.1, ( 1990) Patlcrns of disease activity in multiple sclerosis: clinical and magnetic resonance imaging study. Br Mcd J, 300: 63 1–634. Thompson, A.J., Miller, D., Ynul, B., MacManus, D., Moore, S., KingsIcy, D., Kendall, B., Feinstein, A., McDcrnald, W.1, (1992) Serial gadolinium-enhanced MRI in relapsing-remitting multiple sclerosis of varying disease duration. Neurology, 42: 60–63. van Walderveen, M.A,A., Barkhof, F., Hommes, O.R,, Polman, C.H., Tobi, H., Frequin, S.T.F.M,, Valk, J. (1995) Correlating MR imaging and clinical disease activity in multiple sclerosis: relevance of hypointense lesions on short TR/short TE (“T1-weighted”) spin-echo images. Neurology, 45: 1684–1690.