J Neurol DOI 10.1007/s00415-015-7724-5
ORIGINAL COMMUNICATION
Identification of cortical lesions using DIR and FLAIR in early stages of multiple sclerosis Pierre Kolber1 • Swantje Montag1 • Vinzenz Fleischer1 • Felix Luessi1 • Janine Wilting1 • Joachim Gawehn2 • Adriane Gro¨ger1 • Frauke Zipp1
Received: 13 February 2015 / Revised: 25 March 2015 / Accepted: 26 March 2015 Ó Springer-Verlag Berlin Heidelberg 2015
Abstract The use of non-routine MRI sequences such as DIR has highlighted the role of gray matter (GM) pathology in multiple sclerosis (MS). The aim of this study was to assess the detection and relevance of cortical lesions (CLs) using MRI in early (\5 years) MS patients. 3D DIR and 3D FLAIR images at 3T from 122 patients [93 relapsing– remitting MS (RRMS), 29 clinically isolated syndrome (CIS)] were scored for CLs by two blinded readers. Patients were divided into two groups depending on the presence or absence of CLs. For FLAIR, 51 CLs were identified, of which 13 were purely intracortical and 38 mixed CLs; for DIR, this was 60 in total and 16 and 44, respectively. In both groups, there was no difference in GM fraction. Neuropsychological testing was performed for a subgroup of 66 patients. In 22.1 % of patients CLs were identified. The number of CLs revealed an association with lower working memory scores and semantical word fluency. Overall, CLs imaged with 3D FLAIR and 3D DIR sequences are found more frequently in RRMS patients than CIS and may also be a correlate for mild neuropsychological pathology.
A. Gro¨ger and F. Zipp contributed equally. & Frauke Zipp [email protected] 1
Department of Neurology, Neuroimaging Center (NIC) of the Focus Program Translational Neuroscience (FTN), University Medical Center of the Johannes GutenbergUniversity Mainz, Langenbeckstraße 1, 55131 Mainz, Germany
2
Institute of Neuroradiology, University Medical Center of the Johannes Gutenberg-University Mainz, Mainz, Germany
Keywords Multiple sclerosis Cortical lesions Magnetic resonance imaging Neuropsychological testing
Introduction Multiple sclerosis (MS), one of the most common chronic neuroinflammatory diseases and causes of disability in young adults, is traditionally considered to be caused by demyelination in white matter (WM); however, the role of gray matter (GM) is becoming increasingly prominent. Although described in anatomopathological studies over a century ago [1], cortical pathology only recently returned to the spotlight of MS research as a result of specialized MRI sequences not used in clinical routine. With conventional T2-weighted sequences, GM lesions are less visible, but by nulling the signal from the CSF and WM, double inversion recovery (DIR) allows better detection of cortical lesions (CLs). Using this sequence, CLs have even been reported to potentially precede demyelination of WM [2–4], the neuroinflammatory [5] and neurodegenerative [6] processes that characterize all MS phenotypes and occur early in the disease [7]. However, controversy remains about whether DIR should be adopted in MS clinical routine [8–10]. Other specialized sequences such as phase-sensitive inversion recovery (PSIR) have been shown to offer improvements over DIR in evaluating cortical pathology and offer complementary information [11–13]. Recent studies indicate an association of CL with clinical and cognitive impairment [14–19] in relapsing– remitting (RRMS) and progressive MS patients. The aim of this study was to evaluate the frequency and clinical impact of CLs in patients early in the disease. To ensure an accurate determination of CL load, we independently
123
J Neurol
assessed both 3D DIR and 3D fluid-attenuated inversion recovery (FLAIR) sequences, considering that they provide complementary information [13, 20, 21] and were both available as part of our standard MRI protocol for MS. This was performed for a large cohort of CIS and early RRMS patients and the results were then correlated against clinical and cognitive measures.
Materials and methods Patients This retrospective, single-center analysis was performed on 122 patients (29 CIS, 93 RRMS) diagnosed using the McDonald criteria [22] with a median age of 32 years
Table 1 Demographic data
Variable
(18–63 years), female/male ratio of 2.3, median disease duration of 1.7 years (0–5 years), median number of relapses (NOR) of 2 (1–6) and median Expanded Disability Status Scale (EDSS) [23] of 1.0 (0–4.0). At the time of MRI, 37 patients were being treated with interferon-beta (IFNb), 21 with glatiramer acetate (GA), 4 with oral immunomodulating drugs [teriflunomide (TF) or dimethyl fumarate (DF)], 15 with monthly natalizumab infusions (NA) and 45 subjects were not being treated (NT). Nine of the 29 CIS patients converted to RRMS within 2 years of diagnosis. Patients with a disease duration exceeding 5 years were not included to establish a cohort of early MS patients. Other exclusion criteria were being under 18 years of age or having a chronic progressive form of MS (SPMS, PPMS). Detailed demographic data are shown in Table 1.
Total (n = 122)
CIS (n = 29)
RRMS (n = 93)
p value
Female (n)
85
18
67
0.308c
Male (n)
37
11
26
0.308c
CIS (n)
29
–
–
–
RRMS (n)
93
–
–
–
31.75 (18–63)
32.50 (19–56)
30.50 (18–63)
0.691b
1.0 (0.0–4.0)
1.0 (0.0–3.0)
1.5 (0.0–4.0)
0.389b
2 (1–6)
1
2 (1–6)
\0.0001b,*
1.65 (0.00–5.00)
0.60 (0.00–4.00)
2.00 (0.00–5.00)
\0.0001b,*
1.82 (0.00–45.36)
1.27 (0.10–19.47)
2.05 (0.00–45.36)
0.127b
0.44 ± 0.03
0.44 ± 0.03
0.44 ± 0.03
0.837a
Interferon-beta
37
9
28
0.924c
Glatiramer acetate
21
3
18
0.262c
Teriflunomide
2
1
1
0.380c
Dimethyl fumarate Natalizumab
2 15
0 0
2 15
0.426c 0.026c,*
45
16
29
0.019c,*
Gender
Disease type
Age (years) Median (range) EDSS Median (range) NOR Median (range) DD (years) Median (range) LV (ml) Median (range) GMF Mean ± SD Therapy
Other/none
CIS clinically isolated syndrome, RRMS relapsing–remitting multiple sclerosis, GA? group of subjects treated with glatiramer acetate, GA- group of subjects not treated with glatiramer acetate, CL? group with cortical lesions, CL- group without cortical lesions, EDSS Expanded Disability Status Scale, DD disease duration, NOR number of relapses, LV lesion volume, GMF gray matter fraction * Statistically significant a
123
Derived from t test
b
Derived from Mann–Whitney U test
c
Derived from Chi-square test
J Neurol
All patients gave their written informed consent to examinations before participating in this study, which was approved by the local ethics committee and adhered to institutional guidelines in accordance with the Declaration of Helsinki. For a subgroup of 66 of the patients, we also had results from neuropsychological testing, including measurements of information processing speed [Symbol Digit Modality Test (SDMT), Trail Making Test (TMT-A)], working memory [Paced Auditory Serial Addition Test (PASAT)], visual short-term memory [block span of the Wechsler Memory Scale-Revised (WMS-R)], learning and memory [Verbal Learning and Memory Test (VLMT), with VLMT_verbal for verbal learning, VLMT_global for global learning and VLMT_longterm for longterm learning capacity], alertness [Tests of Attentional Performance (TAP)––subtest Alertness], cognitive flexibility (TMT-B) and word fluency [Regensburger Word Fluency Test (RWT), with RWT_lexical for lexical word fluency and RWT_semantical for semantical word fluency]. Neuropsychological data are given as z-scores. MR image acquisition All patients underwent an MRI examination. Measurements were performed on a 3T MR scanner (Magnetom Tim Trio, Siemens Healthcare, Erlangen, Germany) with a 32-channel head coil. The following sequences were used: 1.
2.
3.
Sagittal 3D fluid-attenuated inversion recovery (FLAIR): repetition time (TR) = 5000 ms, echo time (TE) = 388 ms, inversion time (TI) = 1800 ms, turbo factor = 141, echo train length = 848 ms, field of view (FOV) = 256 9 256 mm2, matrix size = 256 9 256, slab thickness = 192 mm, voxel size = 1 9 1 9 1 mm3, optimized variable flip angle, bandwidth = 781 Hz/Px. Sagittal 3D double inversion recovery (DIR): repetition time (TR) = 7500 ms, echo time (TE) = 307 ms, inversion time (TI) = 3000 ms, turbo factor = 216, echo train length = 657 ms, field of view (FOV) = 256 9 256 mm2, matrix size = 256 9 256, slab thickness = 192 mm, voxel size = 1 9 1 9 1 mm3, optimized variable flip angle, bandwidth = 781 Hz/ Px. Sagittal 3D T1-weighted magnetization prepared rapid gradient echo (MP-RAGE): repetition time (TR) = 1900 ms, echo time (TE) = 2.52 ms, inversion time (TI) = 900 ms, field of view (FOV) = 256 9 256 mm2, matrix size = 256 9 256, slab thickness = 192 mm, voxel size = 1 9 1 9 1 mm3, flip angle = 9°, bandwidth = 170 Hz/Px.
Image analysis and group definition Given that the sensitivity and specificity of DIR are low [11, 21, 24, 25] (signal intensity variations or vessel hyperintensities) and possible misinterpretation is common if used alone without other confirmatory sequences [21], we used a second T2-weighted MR sequence, in this case 3DFLAIR, to confirm the presence of lesions. FLAIR sequences are well established in clinical practice for the diagnosis of MS and T2-weighted sequences are included in the revised McDonald criteria [22]. DIR and FLAIR were viewed (in groups and several days apart to avoid any bias) by two raters with respect to CL load, blinded to the outcome of the other sequence and the clinical features of the patients. For scoring, both raters followed the international criteria proposed by Geurts et al. [21]. CLs were classified by their topographic location (infra- and supratentorial lesions) (Fig. 1c) and their relation to the cortical band (purely intracortical lesions or mixed white–gray matter lesions). Intracortical lesions were confined exclusively to the cortical ribbon, and did not encompass any subcortical WM (Fig. 1a, b); mixed cortical lesions were defined as lesions that overlap both the cortical ribbon and the WM (Fig. 1d). Patients were positively scored for CL (CL?) if both raters identified lesions in both scans, and otherwise negatively scored (CL-). The number of patients with CLs as well as the total number of CLs was counted. MRI post-processing For lesion and tissue segmentation, the statistical parameter mapping (SPM8) software (http://www.fil.ion.ucl.ac.uk/ spm) with lesion segmentation toolbox (LST) (http://www. applied-statistics.de/lst.html) and VBM8 toolbox (http:// dbm.neuro.uni-jena.de/vbm/) was used. First, using LST, 3D FLAIR images were co-registered to 3D MP-RAGE images and bias corrected. After partial volume estimate (PVE) label estimation, lesion segmentation was performed with an optimal threshold value (j = 0.1) for the lesion growth algorithm [26]. Subsequently, the filled MP-RAGE images were segmented into GM, WM, and cerebrospinal fluid (CSF). With regard to measuring brain atrophy [27, 28], GM results were assessed as fractions of total brain volume. Statistics Statistical analysis was performed using SPSS Statistics, Version 22.0 (IBM, Chicago, Illinois, United States of America). The Kolmogorov–Smirnov test was used to
123
J Neurol Fig. 1 Examples of cortical lesions. a Intracortical lesion (DIR left, FLAIR right). b Ushaped cortical lesion in DIR (sagittal view left, axial view right). c Infratentorial cortical lesion (DIR left, FLAIR right). d Mixed gray–white matter lesion in DIR
verify a normal distribution of data. Means with standard deviations (SD) as well as medians with ranges were calculated. Between-group differences of demographic and MRI features were evaluated with the t test, Mann–Whitney U test and Chi-square test, respectively. The Spearman rank and Pearson correlation tests were used to determine correlations between the variables. The significance level was set to p \ 0.05.
Results Using DIR images, we identified 27 patients with C1 CL, representing 22.1 % of our cohort. In total, 60 CL (16 intracortical, 44 mixed lesions) were counted (Table 2). After using FLAIR images, we confirmed the 27 subjects with CLs. However, only 51 lesions were counted, of which 13 were purely intracortical and 38 mixed CLs.
123
These 27 patients were defined as the group positively scored for CLs (CL?). We detected most CLs in supratentorial GM, with 49 CLs in DIR (35 mixed and 14 intracortical) and 41 CLs in FLAIR (32 mixed and 9 intracortical). In infratentorial GM, DIR and FLAIR images revealed a similar number of CLs (11 and 10, respectively) (Table 2). Altogether, with FLAIR images we were able to detect 85 % of the CLs seen in DIR. Clinical data (disease duration, number of relapses, EDSS, gender and comorbidities) as well as GMF and LV did not differ between CL? and CL- (Table 1). We found clear differences between the number of patients with and without CLs being treated with GA (p = 0.035) and also between those with the CIS phenotype (p = 0.006). Further, the number of CLs (counted in DIR or in FLAIR images) did not correlate with these clinical and MRI data (Table 3).
Total (n = 122)
IT
Total (n = 122)
CIS (n = 29)
DIR screening
10
IT
4
0
23
2
5
8
16
RRMS (n = 93)
50
2
8
11
29
RRMS (n = 93)
0 1
0.675c 0.426c
2
1
0.215c
0
GA? (n = 21)
0.374c
b,c
2
1
0
1
0
GA? (n = 21)
0.770c
p value
0.251
0.428
0.671
0.781
0.399
p valuea,c
25
1
6
9
19
GA(n = 101)
58
1
9
13
35
GA(n = 101)
0.126c
0.216c
0.252c
0.528c
0.031c,*
p value
0.099
0.217
0.254
0.512
0.032*
b,d
p valuea,d
3
0
0
1
2
CIS (n = 29)
48
3
6
8
30
RRMS (n = 93)
0.007*
0.258
0.428
0.356
0.053
p valuea,c
27
4
2
9
18
Total (n = 122)
1
0
0
1
1
CIS (n = 29)
26
4
2
8
17
RRMS (n = 93)
0.006c,*
0.256c
0.426c
0.354c
0.049c,*
p valueb,d
After inter-reader confirmation (FLAIR and DIR)
51
4
6
9
32
Total (n = 122)
After inter-reader confirmation (FLAIR and DIR)
1
1
0
0
0
GA? (n = 21)
1
1
0
0
0
GA? (n = 21)
26
2
3
9
18
GA(n = 101)
50
3
6
9
32
GA(n = 101)
0.035c,*
0.675c
0.516c
0.155c
0.036c,*
p valueb,d
0.033*
0.676
0.517
0.157
0.038*
p valuea,d
d
c
b
a
Statistical significance evaluation for comparison between treatment type
Statistical significance evaluation for comparison between phenotype
Derived from Chi-square test
Derived from Spearman rank test
* Statistically significant
CIS clinically isolated syndrome, RRMS relapsing–remitting multiple sclerosis, GA? group of subjects treated with glatiramer acetate, GA- group of subjects not treated with glatiramer acetate, CL cortical lesion, ST supratentorial, IT infratentorial
Total subject count
27
IT mixed CL
intracortical CL
6
2
ST intracortical CL
1
3
2
19
10
ST mixed CL
Subject count (number of subjects with C1 CL)
Classification
Total lesion count
60
IT mixed CL
intracortical CL
1
0
9
2
ST intracortical CL
3
35
14
ST mixed CL
6
CIS (n = 29)
DIR screening
Lesion count (number of CL)
Classification
Table 2 Cortical lesion scoring: frequency of cortical lesions in early phase of the disease
J Neurol
123
J Neurol Table 3 Group comparison (CL? vs CL-): clinical impact of cortical lesions in early phase of the disease Variable
CL? (n = 27)
CL- (n = 95)
p value
Correlation with number of CL in CL? group
p value
Female (n)
16
69
0.182c
Male (n)
11
26
0.182c
CIS (n)
1
28
0.006c,*
RRMS (n)
26
67
0.006c,*
32.90 (19–63)
31.10 (18–63)
0.953b
1.0 (0.0–3.5)
1.5 (0.0–4.0)
0.227b
-0.301
0.127
2 (1–6)
1 (1–5)
\0.0001b,*
-0.163
0.416
2.00 (0.00–5.00)
1.60 (0.00–5.00)
0.269b
0.312
0.114
2.05 (0.24–32.99)
1.66 (0.00–45.36)
0.246b
0.133
0.509
0.45 ± 0.04
0.44 ± 0.03
0.805a
0.038
0.850
Interferon-beta
10
27
0.390c
Glatiramer acetate
1
20
0.035c,*
Teriflunomide
0
2
0.447c
Dimethyl fumarate
1
1
0.338c
Natalizumab
4
11
0.651c
11
34
0.638c
Gender
Disease type
Age (years) Median (range) EDSS Median (range) NOR Median (range) DD (years) Median (range) LV (ml) Median (range) GMF Mean ± SD Therapy
Other/none Cognitive tests (n = 66)
CL? (n = 18)
CL- (n = 46)
p value
Correlation with number of CL
p value
0.09 ± 0.59, 18
0.00 ± 0.90, 46
0.647a
0.085
0.738
-0.61 ± 1.33, 18
-0.19 ± 1.03, 45
0.268a
0.462
0.054
-0.17 (-3.00 to 1.31), 18
-0.14 (-3.00 to 1.31), 46
0.321b
0.495
0.037*
0.15 (-1.30 to 1.50), 18
0.20 (-2.00 to 2.40), 44
0.950b
0.096
0.706
-0.37 (-2.23 to 1.99), 18
0.12 (-1.62 to 2.23), 45
0.062b
0.294
0.236
-0.21 ± 0.99, 18
0.18 ± 0.90, 45
0.140a
0.134
0.597
-0.34 (-4.57 to 0.83), 18
0.23 (-2.57 to 1.43), 45
0.088b
0.209
0.406
-0.48 (-2.50 to 2.20), 18
-0.23 (-11.11 to 1.63), 46
0.585b
0.310
0.210
-0.10 (-2.50 to 1.20), 18
0.16 (-3.00 to 2.30), 46
0.988b
0.165
0.514
TAP Mean ± SD, n SDMT Mean ± SD, n PASAT Median (range), n WMS-R Median (range), n VLMT_verbal Median (range), n VLMT_global Mean ± SD, n VLMT_longterm Median (range), n TMT_A Median (range), n TMT_B Median (range), n
123
J Neurol Table 3 continued Cognitive tests (n = 66)
CL? (n = 18)
CL- (n = 46)
p value
Correlation with number of CL
p value
-0.15 (-2.50 to 1.50), 18
0.25 (-1.60 to 1.60), 44
0.629b
0.233
0.352
-0.43 ± 1.03, 17
-0.32 ± 0.90, 41
0.678a
0.514
0.035*
RWT_lexical Median (range), n RWT_semantical Mean ± SD, n
CIS clinically isolated syndrome, RRMS relapsing–remitting multiple sclerosis, GA ? group of subjects treated with glatiramer acetate, GA group of subjects not treated with glatiramer acetate, CL ? group with cortical lesions, CL - group without cortical lesions, EDSS Expanded Disability Status Scale, DD disease duration, NOR number of relapses, LV lesion volume, GMF gray matter fraction, TAP tests of attentional performance, SDMT Symbol Digit Modality Test for information processing speed, PASAT Paced Auditory Serial Addition Test, WMSR Wechsler Memory Scale-Revised, VLMT_verbal Verbal Learning and Memory Test for verbal learning, VLMT_global Verbal Learning and Memory Test for global learning, VLMT_longterm Verbal Learning and Memory Test for longterm learning capacity, TMT_A Trail Making Test for information processing speed, TMT_B Trail Making Test for cognitive flexibility, RWT_lexical Regensburger Word Fluency Test for lexical word fluency, RWT_semantical Regensburger Word Fluency Test for semantical word fluency * Statistically significant a
Derived from t test
b
Derived from Mann–Whitney U test
c
Derived from Chi-square test
In the subgroup of 66 patients (18 CL? and 48 CL-) for whom we had neuropsychological test results, all z-scores were in the same range as those of healthy controls, indicating that there was no cognitive impairment. No differences between the CL? and CL- groups were found in the results of the cognitive tests (Table 1). In the CL? group, a significant correlation was found between the total number of CLs in FLAIR and the RWT semantic word fluency (p = 0.035), as well as with the PASAT working memory (p = 0.037) (Table 3).
Discussion Patient-based studies with the goal of translating experimental findings to the clinic are necessary to advance the understanding and treatment of multiple sclerosis [29]. Among other challenges facing research of this disease is that the meaning and processes of neuronal compartment pathology are still not fully understood [30]. One approach to address this gap in our knowledge is to employ MRI sequences such as DIR that are not routinely used in clinical practice to unravel the overall role of GM pathology in the disease [31]. CLs are difficult to detect, but have been reported to be associated with disability and a more severe disease course [32]. Studies have shown that CLs are present in early stages of MS [2], but that their importance increases with disease progression (EDSS) [15]. However, both the frequency and specific role of these lesions are still under debate. These studies have focused on small or rather heterogeneous cohorts, including various disease phenotypes and patients across the whole course of the disease. Our study aimed to evaluate the frequency and
clinical impact of CLs in a large disease cohort consisting of only RRMS and CIS patients in very early stages of the disease, namely less than 5-year disease duration. The most common type of CL is located on the subpial layer of the cortex [33, 34] and often can only be seen postmortem or with magnetic field strengths greater than 3T [13]. Also, different MR sequences can affect the identification of lesions, with some juxtacortical lesions assigned in FLAIR images being interpreted as mixed GM/ WM lesions in DIR due to poor definition of lesion boundaries [11] (Fig. 2). Artifacts due to magnetic field inhomogeneities or vessel hyperintensities are further causes of misinterpretation. Additionally, even when following the international recommendations on lesion scoring, an agreement of only 19 % of the CLs scored was reached between experts [21], so that DIR, despite its sensitivity for CL scoring, may have a rather low specificity. To ensure reliable findings [24, 25] and to minimize any false positives resulting from artifacts in DIR images, two blinded readers viewed both 3D DIR and 3D FLAIR images to assess them for the presence of CLs; only in cases where the readers were in consensus were the patients defined as having CLs. The use of additional T2- or T1-weighted MRI sequences to classify a cortical lesion [13, 35] follows current recommendations [21] and has also been discussed recently in smaller studies [36]. In 122 patients (93 RRMS, 29 CIS) with a median disease duration of 1.5 years, we detected CLs in 22.1 % of the patients (28.0 % RRMS, 3.4 % CIS). We did not find major clinical differences between patients with and without cortical lesions, suggesting that these lesions may be less frequent in early stages of RRMS than previously thought [15, 17, 32]. In the neuropsychological analysis
123
J Neurol Fig. 2 Juxtacortical lesion scored as a mixed gray–white matter lesion. The DIR image is shown in the left panel and the FLAIR in the right
performed on a subgroup of 66 patients, we found a correlation between the number of CLs and the semantic word fluency (RWT), as well as working memory (PASAT), suggesting that GM pathology in MS starts with structural or functional alterations in the left frontal lobe of the brain. In a study on CL and WM lesions in a pediatric form of MS [37], investigators found that CL volume and GM volume were the same in both cognitively impaired and cognitively preserved children, but that WM volume was significantly decreased in patients with cognitive deficits. However, in recent studies on adults with MS [17, 18], GM pathology [38] has been associated with a bad cognitive outcome, which is in agreement with our findings. Interestingly, we note that fewer subjects treated with GA were scored positively for CL and had a lower CL count than those receiving other therapies. However, further studies with more subjects must be performed to conclude whether this finding might be of significance. In contrast to earlier studies, we observed fewer CLs than expected. Calabrese et al. [14] reported that 64 % of RRMS patients with a disease duration up to 13 years (mean disease duration of 5 years) and mean EDSS of 2 exhibited CLs as identified by 1.5T MRI, and 36 % of CIS patients with mean disease duration of 0.8 years and EDSS of 1.2 (up to 3.5). We found only 28 % of patients had CLs in the RRMS group with a median disease duration of less than 2 years and not exceeding 5 years and identified only 3 % having CLs in our group of CIS patients, whose disease duration did not exceed 4 years. These differences could be due to the shorter disease duration of our cohort as well as the different methodologies employed (2D FLAIR and 2D DIR at 1.5T versus 3D FLAIR and 3D DIR at 3T; different voxel size). Based on the results presented here, we agree with Chard et al. [9].that the identification of CLs with DIR is not yet suitable to allow clear-cut conclusions concerning disease severity or overall ‘‘neurodegeneration’’ in clinical practice. Although the use of DIR is relatively widespread, this study shows that it remains of limited value in the detection of CLs in MS. This finding is in line with other
123
studies [11, 24, 25] and the current recommendations [21], which specify that DIR should be used with other T1- or T2-weighted MRI sequences to assure reliable findings. 3D FLAIR at 3T has a similar contrast for cortical lesions and a higher signal-to-noise ratio than DIR. Furthermore, DIR is very time consuming and has a high specific absorption rate. Therefore, for high-field MRI, it seems that 3D-FLAIR might be the best sequence to study cortical pathology lesion [13, 35]. The more recently used PSIR sequence permits better detection of CLs than DIR, which may further make the utility of DIR sequences less relevant [11, 12]. Although it has been previously reported that the number of CLs increases over time and that there is an association with disability in MS, in our early cohort we found no correlation between CLs and EDSS or NOR. However, a correlation was found between the number of CLs and changes in cognitive function in word fluency and working memory, suggesting that cognitive functions are the first to be altered by GM pathology. Furthermore, a recent report on RIS patients identified 40 % as having CLs using DIR [39], which in light of the difference we observed between CIS and RRMS CL frequency must be interpreted either as a different underlying pathology or erroneous CL identification using DIR images alone. A limitation of our study is the cross-sectional design as well as detailed neuropsychological testing being available for only a subgroup of patients. Further prospective investigations focusing on early phase MS patients including a detailed analysis of the lesion distribution should increase our knowledge of the early disease. In conclusion, using 3D DIR and 3D FLAIR images in CIS and RRMS, we present evidence that CLs, although already present, are less common in early phases of the disease than previously expected. Despite a clear difference between CIS and MS, no relation with other parameters such as GMF, which characterizes the neuronal compartment, or with disease severity was detected at this early disease stage, indicating that the presence and number of CLs may not be a relevant indicator of early neurodegeneration in MS.
J Neurol Acknowledgments F.Z. is grateful for financial support from the German Multiple Sclerosis Competence Network (KKNMS, Project B7.3) funded by the Federal Ministry for Education and Research (BMBF). Conflicts of interest Pierre Kolber reports no conflicts of interest and financial disclosures. Swantje Montag reports no conflicts of interest and financial disclosures. Dr. Vinzenz Fleischer reports no conflicts of interest and financial disclosures. Dr. Felix Lu¨ssi reports no conflicts of interest and financial disclosures. Janine Wilting reports no conflicts of interest and financial disclosures. Dr. Joachim Gawehn reports no conflicts of interest and financial disclosures. Dr. Adriane Gro¨ger reports no conflicts of interest and financial disclosures. Dr. Frauke Zipp has received research grants from Teva, Merck Serono, Novartis and Bayer as well as consultation funds from Teva, Merck Serono, Novartis, Bayer Healthcare, Biogen Idec Germany, ONO, Genzyme, Sanofi-Aventis and Octapharma. Her travel compensation has been provided for by the aforementioned companies. Ethical standard All patients gave their written informed consent to examinations before participating in this study, which was approved by the local ethics committee and adhered to institutional guidelines in accordance with the Declaration of Helsinki.
13.
14.
15.
16.
17.
18.
References 1. Sander M (1898) Hirnrindenbefunde bei multipler Sklerose. Monatschrift Psychiatrie Neurol IV:427–436 2. Calabrese M, Gallo P (2009) Magnetic resonance evidence of cortical onset of multiple sclerosis. Mult Scler 15:933–941 3. Coebergh JA, Roosendaal SD, Polman CH, Geurts JJ, van Woerkom TC (2010) Acute severe memory impairment as a presenting symptom of multiple sclerosis: a clinical case study with 3D double inversion recovery MR imaging. Mult Scler 16:1521–1524 4. Popescu BF, Bunyan RF, Parisi JE, Ransohoff RM, Lucchinetti CF (2011) A case of multiple sclerosis presenting with inflammatory cortical demyelination. Neurology 76:1705–1710 5. Lucchinetti CF, Popescu BF, Bunyan RF, Moll NM, Roemer SF, Lassmann H, Bruck W, Parisi JE, Scheithauer BW, Giannini C, Weigand SD, Mandrekar J, Ransohoff RM (2011) Inflammatory cortical demyelination in early multiple sclerosis. N Engl J Med 365:2188–2197 6. Bo L, Vedeler CA, Nyland H, Trapp BD, Mork SJ (2003) Intracortical multiple sclerosis lesions are not associated with increased lymphocyte infiltration. Mult Scler 9:323–331 7. Kutzelnigg A, Lucchinetti CF, Stadelmann C, Bruck W, Rauschka H, Bergmann M, Schmidbauer M, Parisi JE, Lassmann H (2005) Cortical demyelination and diffuse white matter injury in multiple sclerosis. Brain 128:2705–2712 8. Calabrese M, De Stefano N (2014) Cortical lesion counts by double inversion recovery should be part of the MRI monitoring process for all MS patients: yes. Mult Scler 20:537–538 9. Chard D (2014) Cortical lesion counts by double inversion recovery should be part of the MRI monitoring process for all MS patients: no. Mult Scler 20:539–540 10. Rovira A, Auger C (2014) Cortical lesion counts by double inversion recovery should be part of the MRI monitoring process for all MS patients: commentary. Mult Scler 20:541–542 11. Sethi V, Muhlert N, Ron M, Golay X, Wheeler-Kingshott CA, Miller DH, Chard DT, Yousry TA (2013) MS cortical lesions on DIR: not quite what they seem? PLoS One 8:e78879 12. Sethi V, Yousry TA, Muhlert N, Ron M, Golay X, WheelerKingshott C, Miller DH, Chard DT (2012) Improved detection of
19.
20.
21.
22.
23.
24.
25.
26.
27.
cortical MS lesions with phase-sensitive inversion recovery MRI. J Neurol Neurosurg Psychiatry 83:877–882 Tallantyre EC, Morgan PS, Dixon JE, Al-Radaideh A, Brookes MJ, Morris PG, Evangelou N (2010) 3 Tesla and 7 Tesla MRI of multiple sclerosis cortical lesions. J Magn Reson Imaging 32:971–977 Calabrese M, Agosta F, Rinaldi F, Mattisi I, Grossi P, Favaretto A, Atzori M, Bernardi V, Barachino L, Rinaldi L, Perini P, Gallo P, Filippi M (2009) Cortical lesions and atrophy associated with cognitive impairment in relapsing–remitting multiple sclerosis. Arch Neurol 66:1144–1150 Calabrese M, Poretto V, Favaretto A, Alessio S, Bernardi V, Romualdi C, Rinaldi F, Perini P, Gallo P (2012) Cortical lesion load associates with progression of disability in multiple sclerosis. Brain 135:2952–2961 Calabrese M, Rocca MA, Atzori M, Mattisi I, Bernardi V, Favaretto A, Barachino L, Romualdi C, Rinaldi L, Perini P, Gallo P, Filippi M (2009) Cortical lesions in primary progressive multiple sclerosis: a 2-year longitudinal MR study. Neurology 72:1330–1336 Damasceno A, Damasceno BP, Cendes F (2014) The clinical impact of cerebellar grey matter pathology in multiple sclerosis. PLoS One 9:e96193 Papadopoulou A, Muller-Lenke N, Naegelin Y, Kalt G, Bendfeldt K, Kuster P, Stoecklin M, Gass A, Sprenger T, Radue EW, Kappos L, Penner IK (2013) Contribution of cortical and white matter lesions to cognitive impairment in multiple sclerosis. Mult Scler 19:1290–1296 Roosendaal SD, Moraal B, Pouwels PJ, Vrenken H, Castelijns JA, Barkhof F, Geurts JJ (2009) Accumulation of cortical lesions in MS: relation with cognitive impairment. Mult Scler 15:708–714 Geurts JJ, Pouwels PJ, Uitdehaag BM, Polman CH, Barkhof F, Castelijns JA (2005) Intracortical lesions in multiple sclerosis: improved detection with 3D double inversion-recovery MR imaging. Radiology 236:254–260 Geurts JJ, Roosendaal SD, Calabrese M, Ciccarelli O, Agosta F, Chard DT, Gass A, Huerga E, Moraal B, Pareto D, Rocca MA, Wattjes MP, Yousry TA, Uitdehaag BM, Barkhof F (2011) Consensus recommendations for MS cortical lesion scoring using double inversion recovery MRI. Neurology 76:418–424 Polman CH, Reingold SC, Banwell B, Clanet M, Cohen JA, Filippi M, Fujihara K, Havrdova E, Hutchinson M, Kappos L, Lublin FD, Montalban X, O’Connor P, Sandberg-Wollheim M, Thompson AJ, Waubant E, Weinshenker B, Wolinsky JS (2011) Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol 69:292–302 Kurtzke JF (1983) Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology 33:1444–1452 Chen JT, Pirko I (2011) MS cortical lesion or not? Double inversion recovery MRI reveals some answers and uncertainties. Neurology 76:412–413 Simon B, Schmidt S, Lukas C, Gieseke J, Traber F, Knol DL, Willinek WA, Geurts JJ, Schild HH, Barkhof F, Wattjes MP (2010) Improved in vivo detection of cortical lesions in multiple sclerosis using double inversion recovery MR imaging at 3 Tesla. Eur Radiol 20:1675–1683 Schmidt P, Gaser C, Arsic M, Buck D, Forschler A, Berthele A, Hoshi M, Ilg R, Schmid VJ, Zimmer C, Hemmer B, Muhlau M (2012) An automated tool for detection of FLAIR-hyperintense white-matter lesions in Multiple Sclerosis. Neuroimage 59:3774–3783 Chard DT, Parker GJ, Griffin CM, Thompson AJ, Miller DH (2002) The reproducibility and sensitivity of brain tissue volume measurements derived from an SPM-based segmentation methodology. J Magn Reson Imaging 15:259–267
123
J Neurol 28. Rudick RA, Fisher E, Lee JC, Simon J, Jacobs L (1999) Use of the brain parenchymal fraction to measure whole brain atrophy in relapsing–remitting MS. Multiple Sclerosis Collaborative Research Group. Neurology 53:1698–1704 29. Roep BO, Buckner J, Sawcer S, Toes R, Zipp F (2012) The problems and promises of research into human immunology and autoimmune disease. Nat Med 18:48–53 30. Siffrin V, Vogt J, Radbruch H, Nitsch R, Zipp F (2010) Multiple sclerosis—candidate mechanisms underlying CNS atrophy. Trends Neurosci 33:202–210 31. Zipp F, Gold R, Wiendl H (2013) Identification of inflammatory neuronal injury and prevention of neuronal damage in multiple sclerosis: hope for novel therapies? JAMA Neurol 70:1569–1574 32. Calabrese M, De Stefano N, Atzori M, Bernardi V, Mattisi I, Barachino L, Morra A, Rinaldi L, Romualdi C, Perini P, Battistin L, Gallo P (2007) Detection of cortical inflammatory lesions by double inversion recovery magnetic resonance imaging in patients with multiple sclerosis. Arch Neurol 64:1416–1422 33. Bo L, Vedeler CA, Nyland HI, Trapp BD, Mork SJ (2003) Subpial demyelination in the cerebral cortex of multiple sclerosis patients. J Neuropathol Exp Neurol 62:723–732 34. Kutzelnigg A, Lassmann H (2005) Cortical lesions and brain atrophy in MS. J Neurol Sci 233:55–59
123
35. Kilsdonk ID, de Graaf WL, Soriano AL, Zwanenburg JJ, Visser F, Kuijer JP, Geurts JJ, Pouwels PJ, Polman CH, Castelijns JA, Luijten PR, Barkhof F, Wattjes MP (2013) Multicontrast MR imaging at 7T in multiple sclerosis: highest lesion detection in cortical gray matter with 3D-FLAIR. Am J Neuroradiol 34:791–796 36. Vural G, Keklikoglu HD, Temel S, Deniz O, Ercan K (2013) Comparison of double inversion recovery and conventional magnetic resonance brain imaging in patients with multiple sclerosis and relations with disease disability. Neuroradiol J 26:133–142 37. Rocca MA, De Meo E, Amato MP, Copetti M, Moiola L, Ghezzi A, Veggiotti P, Capra R, Fiorino A, Pippolo L, Pera MC, Falini A, Comi G, Filippi M (2014) Cognitive impairment in paediatric multiple sclerosis patients is not related to cortical lesions. Mult Scler 38. Rovaris M, Filippi M, Minicucci L, Iannucci G, Santuccio G, Possa F, Comi G (2000) Cortical/subcortical disease burden and cognitive impairment in patients with multiple sclerosis. Am J Neuroradiol 21:402–408 39. Giorgio A, Stromillo ML, Rossi F, Battaglini M, Hakiki B, Portaccio E, Federico A, Amato MP, De Stefano N (2011) Cortical lesions in radiologically isolated syndrome. Neurology 77:1896–1899
ORIGINAL COMMUNICATION
Identification of cortical lesions using DIR and FLAIR in early stages of multiple sclerosis Pierre Kolber1 • Swantje Montag1 • Vinzenz Fleischer1 • Felix Luessi1 • Janine Wilting1 • Joachim Gawehn2 • Adriane Gro¨ger1 • Frauke Zipp1
Received: 13 February 2015 / Revised: 25 March 2015 / Accepted: 26 March 2015 Ó Springer-Verlag Berlin Heidelberg 2015
Abstract The use of non-routine MRI sequences such as DIR has highlighted the role of gray matter (GM) pathology in multiple sclerosis (MS). The aim of this study was to assess the detection and relevance of cortical lesions (CLs) using MRI in early (\5 years) MS patients. 3D DIR and 3D FLAIR images at 3T from 122 patients [93 relapsing– remitting MS (RRMS), 29 clinically isolated syndrome (CIS)] were scored for CLs by two blinded readers. Patients were divided into two groups depending on the presence or absence of CLs. For FLAIR, 51 CLs were identified, of which 13 were purely intracortical and 38 mixed CLs; for DIR, this was 60 in total and 16 and 44, respectively. In both groups, there was no difference in GM fraction. Neuropsychological testing was performed for a subgroup of 66 patients. In 22.1 % of patients CLs were identified. The number of CLs revealed an association with lower working memory scores and semantical word fluency. Overall, CLs imaged with 3D FLAIR and 3D DIR sequences are found more frequently in RRMS patients than CIS and may also be a correlate for mild neuropsychological pathology.
A. Gro¨ger and F. Zipp contributed equally. & Frauke Zipp [email protected] 1
Department of Neurology, Neuroimaging Center (NIC) of the Focus Program Translational Neuroscience (FTN), University Medical Center of the Johannes GutenbergUniversity Mainz, Langenbeckstraße 1, 55131 Mainz, Germany
2
Institute of Neuroradiology, University Medical Center of the Johannes Gutenberg-University Mainz, Mainz, Germany
Keywords Multiple sclerosis Cortical lesions Magnetic resonance imaging Neuropsychological testing
Introduction Multiple sclerosis (MS), one of the most common chronic neuroinflammatory diseases and causes of disability in young adults, is traditionally considered to be caused by demyelination in white matter (WM); however, the role of gray matter (GM) is becoming increasingly prominent. Although described in anatomopathological studies over a century ago [1], cortical pathology only recently returned to the spotlight of MS research as a result of specialized MRI sequences not used in clinical routine. With conventional T2-weighted sequences, GM lesions are less visible, but by nulling the signal from the CSF and WM, double inversion recovery (DIR) allows better detection of cortical lesions (CLs). Using this sequence, CLs have even been reported to potentially precede demyelination of WM [2–4], the neuroinflammatory [5] and neurodegenerative [6] processes that characterize all MS phenotypes and occur early in the disease [7]. However, controversy remains about whether DIR should be adopted in MS clinical routine [8–10]. Other specialized sequences such as phase-sensitive inversion recovery (PSIR) have been shown to offer improvements over DIR in evaluating cortical pathology and offer complementary information [11–13]. Recent studies indicate an association of CL with clinical and cognitive impairment [14–19] in relapsing– remitting (RRMS) and progressive MS patients. The aim of this study was to evaluate the frequency and clinical impact of CLs in patients early in the disease. To ensure an accurate determination of CL load, we independently
123
J Neurol
assessed both 3D DIR and 3D fluid-attenuated inversion recovery (FLAIR) sequences, considering that they provide complementary information [13, 20, 21] and were both available as part of our standard MRI protocol for MS. This was performed for a large cohort of CIS and early RRMS patients and the results were then correlated against clinical and cognitive measures.
Materials and methods Patients This retrospective, single-center analysis was performed on 122 patients (29 CIS, 93 RRMS) diagnosed using the McDonald criteria [22] with a median age of 32 years
Table 1 Demographic data
Variable
(18–63 years), female/male ratio of 2.3, median disease duration of 1.7 years (0–5 years), median number of relapses (NOR) of 2 (1–6) and median Expanded Disability Status Scale (EDSS) [23] of 1.0 (0–4.0). At the time of MRI, 37 patients were being treated with interferon-beta (IFNb), 21 with glatiramer acetate (GA), 4 with oral immunomodulating drugs [teriflunomide (TF) or dimethyl fumarate (DF)], 15 with monthly natalizumab infusions (NA) and 45 subjects were not being treated (NT). Nine of the 29 CIS patients converted to RRMS within 2 years of diagnosis. Patients with a disease duration exceeding 5 years were not included to establish a cohort of early MS patients. Other exclusion criteria were being under 18 years of age or having a chronic progressive form of MS (SPMS, PPMS). Detailed demographic data are shown in Table 1.
Total (n = 122)
CIS (n = 29)
RRMS (n = 93)
p value
Female (n)
85
18
67
0.308c
Male (n)
37
11
26
0.308c
CIS (n)
29
–
–
–
RRMS (n)
93
–
–
–
31.75 (18–63)
32.50 (19–56)
30.50 (18–63)
0.691b
1.0 (0.0–4.0)
1.0 (0.0–3.0)
1.5 (0.0–4.0)
0.389b
2 (1–6)
1
2 (1–6)
\0.0001b,*
1.65 (0.00–5.00)
0.60 (0.00–4.00)
2.00 (0.00–5.00)
\0.0001b,*
1.82 (0.00–45.36)
1.27 (0.10–19.47)
2.05 (0.00–45.36)
0.127b
0.44 ± 0.03
0.44 ± 0.03
0.44 ± 0.03
0.837a
Interferon-beta
37
9
28
0.924c
Glatiramer acetate
21
3
18
0.262c
Teriflunomide
2
1
1
0.380c
Dimethyl fumarate Natalizumab
2 15
0 0
2 15
0.426c 0.026c,*
45
16
29
0.019c,*
Gender
Disease type
Age (years) Median (range) EDSS Median (range) NOR Median (range) DD (years) Median (range) LV (ml) Median (range) GMF Mean ± SD Therapy
Other/none
CIS clinically isolated syndrome, RRMS relapsing–remitting multiple sclerosis, GA? group of subjects treated with glatiramer acetate, GA- group of subjects not treated with glatiramer acetate, CL? group with cortical lesions, CL- group without cortical lesions, EDSS Expanded Disability Status Scale, DD disease duration, NOR number of relapses, LV lesion volume, GMF gray matter fraction * Statistically significant a
123
Derived from t test
b
Derived from Mann–Whitney U test
c
Derived from Chi-square test
J Neurol
All patients gave their written informed consent to examinations before participating in this study, which was approved by the local ethics committee and adhered to institutional guidelines in accordance with the Declaration of Helsinki. For a subgroup of 66 of the patients, we also had results from neuropsychological testing, including measurements of information processing speed [Symbol Digit Modality Test (SDMT), Trail Making Test (TMT-A)], working memory [Paced Auditory Serial Addition Test (PASAT)], visual short-term memory [block span of the Wechsler Memory Scale-Revised (WMS-R)], learning and memory [Verbal Learning and Memory Test (VLMT), with VLMT_verbal for verbal learning, VLMT_global for global learning and VLMT_longterm for longterm learning capacity], alertness [Tests of Attentional Performance (TAP)––subtest Alertness], cognitive flexibility (TMT-B) and word fluency [Regensburger Word Fluency Test (RWT), with RWT_lexical for lexical word fluency and RWT_semantical for semantical word fluency]. Neuropsychological data are given as z-scores. MR image acquisition All patients underwent an MRI examination. Measurements were performed on a 3T MR scanner (Magnetom Tim Trio, Siemens Healthcare, Erlangen, Germany) with a 32-channel head coil. The following sequences were used: 1.
2.
3.
Sagittal 3D fluid-attenuated inversion recovery (FLAIR): repetition time (TR) = 5000 ms, echo time (TE) = 388 ms, inversion time (TI) = 1800 ms, turbo factor = 141, echo train length = 848 ms, field of view (FOV) = 256 9 256 mm2, matrix size = 256 9 256, slab thickness = 192 mm, voxel size = 1 9 1 9 1 mm3, optimized variable flip angle, bandwidth = 781 Hz/Px. Sagittal 3D double inversion recovery (DIR): repetition time (TR) = 7500 ms, echo time (TE) = 307 ms, inversion time (TI) = 3000 ms, turbo factor = 216, echo train length = 657 ms, field of view (FOV) = 256 9 256 mm2, matrix size = 256 9 256, slab thickness = 192 mm, voxel size = 1 9 1 9 1 mm3, optimized variable flip angle, bandwidth = 781 Hz/ Px. Sagittal 3D T1-weighted magnetization prepared rapid gradient echo (MP-RAGE): repetition time (TR) = 1900 ms, echo time (TE) = 2.52 ms, inversion time (TI) = 900 ms, field of view (FOV) = 256 9 256 mm2, matrix size = 256 9 256, slab thickness = 192 mm, voxel size = 1 9 1 9 1 mm3, flip angle = 9°, bandwidth = 170 Hz/Px.
Image analysis and group definition Given that the sensitivity and specificity of DIR are low [11, 21, 24, 25] (signal intensity variations or vessel hyperintensities) and possible misinterpretation is common if used alone without other confirmatory sequences [21], we used a second T2-weighted MR sequence, in this case 3DFLAIR, to confirm the presence of lesions. FLAIR sequences are well established in clinical practice for the diagnosis of MS and T2-weighted sequences are included in the revised McDonald criteria [22]. DIR and FLAIR were viewed (in groups and several days apart to avoid any bias) by two raters with respect to CL load, blinded to the outcome of the other sequence and the clinical features of the patients. For scoring, both raters followed the international criteria proposed by Geurts et al. [21]. CLs were classified by their topographic location (infra- and supratentorial lesions) (Fig. 1c) and their relation to the cortical band (purely intracortical lesions or mixed white–gray matter lesions). Intracortical lesions were confined exclusively to the cortical ribbon, and did not encompass any subcortical WM (Fig. 1a, b); mixed cortical lesions were defined as lesions that overlap both the cortical ribbon and the WM (Fig. 1d). Patients were positively scored for CL (CL?) if both raters identified lesions in both scans, and otherwise negatively scored (CL-). The number of patients with CLs as well as the total number of CLs was counted. MRI post-processing For lesion and tissue segmentation, the statistical parameter mapping (SPM8) software (http://www.fil.ion.ucl.ac.uk/ spm) with lesion segmentation toolbox (LST) (http://www. applied-statistics.de/lst.html) and VBM8 toolbox (http:// dbm.neuro.uni-jena.de/vbm/) was used. First, using LST, 3D FLAIR images were co-registered to 3D MP-RAGE images and bias corrected. After partial volume estimate (PVE) label estimation, lesion segmentation was performed with an optimal threshold value (j = 0.1) for the lesion growth algorithm [26]. Subsequently, the filled MP-RAGE images were segmented into GM, WM, and cerebrospinal fluid (CSF). With regard to measuring brain atrophy [27, 28], GM results were assessed as fractions of total brain volume. Statistics Statistical analysis was performed using SPSS Statistics, Version 22.0 (IBM, Chicago, Illinois, United States of America). The Kolmogorov–Smirnov test was used to
123
J Neurol Fig. 1 Examples of cortical lesions. a Intracortical lesion (DIR left, FLAIR right). b Ushaped cortical lesion in DIR (sagittal view left, axial view right). c Infratentorial cortical lesion (DIR left, FLAIR right). d Mixed gray–white matter lesion in DIR
verify a normal distribution of data. Means with standard deviations (SD) as well as medians with ranges were calculated. Between-group differences of demographic and MRI features were evaluated with the t test, Mann–Whitney U test and Chi-square test, respectively. The Spearman rank and Pearson correlation tests were used to determine correlations between the variables. The significance level was set to p \ 0.05.
Results Using DIR images, we identified 27 patients with C1 CL, representing 22.1 % of our cohort. In total, 60 CL (16 intracortical, 44 mixed lesions) were counted (Table 2). After using FLAIR images, we confirmed the 27 subjects with CLs. However, only 51 lesions were counted, of which 13 were purely intracortical and 38 mixed CLs.
123
These 27 patients were defined as the group positively scored for CLs (CL?). We detected most CLs in supratentorial GM, with 49 CLs in DIR (35 mixed and 14 intracortical) and 41 CLs in FLAIR (32 mixed and 9 intracortical). In infratentorial GM, DIR and FLAIR images revealed a similar number of CLs (11 and 10, respectively) (Table 2). Altogether, with FLAIR images we were able to detect 85 % of the CLs seen in DIR. Clinical data (disease duration, number of relapses, EDSS, gender and comorbidities) as well as GMF and LV did not differ between CL? and CL- (Table 1). We found clear differences between the number of patients with and without CLs being treated with GA (p = 0.035) and also between those with the CIS phenotype (p = 0.006). Further, the number of CLs (counted in DIR or in FLAIR images) did not correlate with these clinical and MRI data (Table 3).
Total (n = 122)
IT
Total (n = 122)
CIS (n = 29)
DIR screening
10
IT
4
0
23
2
5
8
16
RRMS (n = 93)
50
2
8
11
29
RRMS (n = 93)
0 1
0.675c 0.426c
2
1
0.215c
0
GA? (n = 21)
0.374c
b,c
2
1
0
1
0
GA? (n = 21)
0.770c
p value
0.251
0.428
0.671
0.781
0.399
p valuea,c
25
1
6
9
19
GA(n = 101)
58
1
9
13
35
GA(n = 101)
0.126c
0.216c
0.252c
0.528c
0.031c,*
p value
0.099
0.217
0.254
0.512
0.032*
b,d
p valuea,d
3
0
0
1
2
CIS (n = 29)
48
3
6
8
30
RRMS (n = 93)
0.007*
0.258
0.428
0.356
0.053
p valuea,c
27
4
2
9
18
Total (n = 122)
1
0
0
1
1
CIS (n = 29)
26
4
2
8
17
RRMS (n = 93)
0.006c,*
0.256c
0.426c
0.354c
0.049c,*
p valueb,d
After inter-reader confirmation (FLAIR and DIR)
51
4
6
9
32
Total (n = 122)
After inter-reader confirmation (FLAIR and DIR)
1
1
0
0
0
GA? (n = 21)
1
1
0
0
0
GA? (n = 21)
26
2
3
9
18
GA(n = 101)
50
3
6
9
32
GA(n = 101)
0.035c,*
0.675c
0.516c
0.155c
0.036c,*
p valueb,d
0.033*
0.676
0.517
0.157
0.038*
p valuea,d
d
c
b
a
Statistical significance evaluation for comparison between treatment type
Statistical significance evaluation for comparison between phenotype
Derived from Chi-square test
Derived from Spearman rank test
* Statistically significant
CIS clinically isolated syndrome, RRMS relapsing–remitting multiple sclerosis, GA? group of subjects treated with glatiramer acetate, GA- group of subjects not treated with glatiramer acetate, CL cortical lesion, ST supratentorial, IT infratentorial
Total subject count
27
IT mixed CL
intracortical CL
6
2
ST intracortical CL
1
3
2
19
10
ST mixed CL
Subject count (number of subjects with C1 CL)
Classification
Total lesion count
60
IT mixed CL
intracortical CL
1
0
9
2
ST intracortical CL
3
35
14
ST mixed CL
6
CIS (n = 29)
DIR screening
Lesion count (number of CL)
Classification
Table 2 Cortical lesion scoring: frequency of cortical lesions in early phase of the disease
J Neurol
123
J Neurol Table 3 Group comparison (CL? vs CL-): clinical impact of cortical lesions in early phase of the disease Variable
CL? (n = 27)
CL- (n = 95)
p value
Correlation with number of CL in CL? group
p value
Female (n)
16
69
0.182c
Male (n)
11
26
0.182c
CIS (n)
1
28
0.006c,*
RRMS (n)
26
67
0.006c,*
32.90 (19–63)
31.10 (18–63)
0.953b
1.0 (0.0–3.5)
1.5 (0.0–4.0)
0.227b
-0.301
0.127
2 (1–6)
1 (1–5)
\0.0001b,*
-0.163
0.416
2.00 (0.00–5.00)
1.60 (0.00–5.00)
0.269b
0.312
0.114
2.05 (0.24–32.99)
1.66 (0.00–45.36)
0.246b
0.133
0.509
0.45 ± 0.04
0.44 ± 0.03
0.805a
0.038
0.850
Interferon-beta
10
27
0.390c
Glatiramer acetate
1
20
0.035c,*
Teriflunomide
0
2
0.447c
Dimethyl fumarate
1
1
0.338c
Natalizumab
4
11
0.651c
11
34
0.638c
Gender
Disease type
Age (years) Median (range) EDSS Median (range) NOR Median (range) DD (years) Median (range) LV (ml) Median (range) GMF Mean ± SD Therapy
Other/none Cognitive tests (n = 66)
CL? (n = 18)
CL- (n = 46)
p value
Correlation with number of CL
p value
0.09 ± 0.59, 18
0.00 ± 0.90, 46
0.647a
0.085
0.738
-0.61 ± 1.33, 18
-0.19 ± 1.03, 45
0.268a
0.462
0.054
-0.17 (-3.00 to 1.31), 18
-0.14 (-3.00 to 1.31), 46
0.321b
0.495
0.037*
0.15 (-1.30 to 1.50), 18
0.20 (-2.00 to 2.40), 44
0.950b
0.096
0.706
-0.37 (-2.23 to 1.99), 18
0.12 (-1.62 to 2.23), 45
0.062b
0.294
0.236
-0.21 ± 0.99, 18
0.18 ± 0.90, 45
0.140a
0.134
0.597
-0.34 (-4.57 to 0.83), 18
0.23 (-2.57 to 1.43), 45
0.088b
0.209
0.406
-0.48 (-2.50 to 2.20), 18
-0.23 (-11.11 to 1.63), 46
0.585b
0.310
0.210
-0.10 (-2.50 to 1.20), 18
0.16 (-3.00 to 2.30), 46
0.988b
0.165
0.514
TAP Mean ± SD, n SDMT Mean ± SD, n PASAT Median (range), n WMS-R Median (range), n VLMT_verbal Median (range), n VLMT_global Mean ± SD, n VLMT_longterm Median (range), n TMT_A Median (range), n TMT_B Median (range), n
123
J Neurol Table 3 continued Cognitive tests (n = 66)
CL? (n = 18)
CL- (n = 46)
p value
Correlation with number of CL
p value
-0.15 (-2.50 to 1.50), 18
0.25 (-1.60 to 1.60), 44
0.629b
0.233
0.352
-0.43 ± 1.03, 17
-0.32 ± 0.90, 41
0.678a
0.514
0.035*
RWT_lexical Median (range), n RWT_semantical Mean ± SD, n
CIS clinically isolated syndrome, RRMS relapsing–remitting multiple sclerosis, GA ? group of subjects treated with glatiramer acetate, GA group of subjects not treated with glatiramer acetate, CL ? group with cortical lesions, CL - group without cortical lesions, EDSS Expanded Disability Status Scale, DD disease duration, NOR number of relapses, LV lesion volume, GMF gray matter fraction, TAP tests of attentional performance, SDMT Symbol Digit Modality Test for information processing speed, PASAT Paced Auditory Serial Addition Test, WMSR Wechsler Memory Scale-Revised, VLMT_verbal Verbal Learning and Memory Test for verbal learning, VLMT_global Verbal Learning and Memory Test for global learning, VLMT_longterm Verbal Learning and Memory Test for longterm learning capacity, TMT_A Trail Making Test for information processing speed, TMT_B Trail Making Test for cognitive flexibility, RWT_lexical Regensburger Word Fluency Test for lexical word fluency, RWT_semantical Regensburger Word Fluency Test for semantical word fluency * Statistically significant a
Derived from t test
b
Derived from Mann–Whitney U test
c
Derived from Chi-square test
In the subgroup of 66 patients (18 CL? and 48 CL-) for whom we had neuropsychological test results, all z-scores were in the same range as those of healthy controls, indicating that there was no cognitive impairment. No differences between the CL? and CL- groups were found in the results of the cognitive tests (Table 1). In the CL? group, a significant correlation was found between the total number of CLs in FLAIR and the RWT semantic word fluency (p = 0.035), as well as with the PASAT working memory (p = 0.037) (Table 3).
Discussion Patient-based studies with the goal of translating experimental findings to the clinic are necessary to advance the understanding and treatment of multiple sclerosis [29]. Among other challenges facing research of this disease is that the meaning and processes of neuronal compartment pathology are still not fully understood [30]. One approach to address this gap in our knowledge is to employ MRI sequences such as DIR that are not routinely used in clinical practice to unravel the overall role of GM pathology in the disease [31]. CLs are difficult to detect, but have been reported to be associated with disability and a more severe disease course [32]. Studies have shown that CLs are present in early stages of MS [2], but that their importance increases with disease progression (EDSS) [15]. However, both the frequency and specific role of these lesions are still under debate. These studies have focused on small or rather heterogeneous cohorts, including various disease phenotypes and patients across the whole course of the disease. Our study aimed to evaluate the frequency and
clinical impact of CLs in a large disease cohort consisting of only RRMS and CIS patients in very early stages of the disease, namely less than 5-year disease duration. The most common type of CL is located on the subpial layer of the cortex [33, 34] and often can only be seen postmortem or with magnetic field strengths greater than 3T [13]. Also, different MR sequences can affect the identification of lesions, with some juxtacortical lesions assigned in FLAIR images being interpreted as mixed GM/ WM lesions in DIR due to poor definition of lesion boundaries [11] (Fig. 2). Artifacts due to magnetic field inhomogeneities or vessel hyperintensities are further causes of misinterpretation. Additionally, even when following the international recommendations on lesion scoring, an agreement of only 19 % of the CLs scored was reached between experts [21], so that DIR, despite its sensitivity for CL scoring, may have a rather low specificity. To ensure reliable findings [24, 25] and to minimize any false positives resulting from artifacts in DIR images, two blinded readers viewed both 3D DIR and 3D FLAIR images to assess them for the presence of CLs; only in cases where the readers were in consensus were the patients defined as having CLs. The use of additional T2- or T1-weighted MRI sequences to classify a cortical lesion [13, 35] follows current recommendations [21] and has also been discussed recently in smaller studies [36]. In 122 patients (93 RRMS, 29 CIS) with a median disease duration of 1.5 years, we detected CLs in 22.1 % of the patients (28.0 % RRMS, 3.4 % CIS). We did not find major clinical differences between patients with and without cortical lesions, suggesting that these lesions may be less frequent in early stages of RRMS than previously thought [15, 17, 32]. In the neuropsychological analysis
123
J Neurol Fig. 2 Juxtacortical lesion scored as a mixed gray–white matter lesion. The DIR image is shown in the left panel and the FLAIR in the right
performed on a subgroup of 66 patients, we found a correlation between the number of CLs and the semantic word fluency (RWT), as well as working memory (PASAT), suggesting that GM pathology in MS starts with structural or functional alterations in the left frontal lobe of the brain. In a study on CL and WM lesions in a pediatric form of MS [37], investigators found that CL volume and GM volume were the same in both cognitively impaired and cognitively preserved children, but that WM volume was significantly decreased in patients with cognitive deficits. However, in recent studies on adults with MS [17, 18], GM pathology [38] has been associated with a bad cognitive outcome, which is in agreement with our findings. Interestingly, we note that fewer subjects treated with GA were scored positively for CL and had a lower CL count than those receiving other therapies. However, further studies with more subjects must be performed to conclude whether this finding might be of significance. In contrast to earlier studies, we observed fewer CLs than expected. Calabrese et al. [14] reported that 64 % of RRMS patients with a disease duration up to 13 years (mean disease duration of 5 years) and mean EDSS of 2 exhibited CLs as identified by 1.5T MRI, and 36 % of CIS patients with mean disease duration of 0.8 years and EDSS of 1.2 (up to 3.5). We found only 28 % of patients had CLs in the RRMS group with a median disease duration of less than 2 years and not exceeding 5 years and identified only 3 % having CLs in our group of CIS patients, whose disease duration did not exceed 4 years. These differences could be due to the shorter disease duration of our cohort as well as the different methodologies employed (2D FLAIR and 2D DIR at 1.5T versus 3D FLAIR and 3D DIR at 3T; different voxel size). Based on the results presented here, we agree with Chard et al. [9].that the identification of CLs with DIR is not yet suitable to allow clear-cut conclusions concerning disease severity or overall ‘‘neurodegeneration’’ in clinical practice. Although the use of DIR is relatively widespread, this study shows that it remains of limited value in the detection of CLs in MS. This finding is in line with other
123
studies [11, 24, 25] and the current recommendations [21], which specify that DIR should be used with other T1- or T2-weighted MRI sequences to assure reliable findings. 3D FLAIR at 3T has a similar contrast for cortical lesions and a higher signal-to-noise ratio than DIR. Furthermore, DIR is very time consuming and has a high specific absorption rate. Therefore, for high-field MRI, it seems that 3D-FLAIR might be the best sequence to study cortical pathology lesion [13, 35]. The more recently used PSIR sequence permits better detection of CLs than DIR, which may further make the utility of DIR sequences less relevant [11, 12]. Although it has been previously reported that the number of CLs increases over time and that there is an association with disability in MS, in our early cohort we found no correlation between CLs and EDSS or NOR. However, a correlation was found between the number of CLs and changes in cognitive function in word fluency and working memory, suggesting that cognitive functions are the first to be altered by GM pathology. Furthermore, a recent report on RIS patients identified 40 % as having CLs using DIR [39], which in light of the difference we observed between CIS and RRMS CL frequency must be interpreted either as a different underlying pathology or erroneous CL identification using DIR images alone. A limitation of our study is the cross-sectional design as well as detailed neuropsychological testing being available for only a subgroup of patients. Further prospective investigations focusing on early phase MS patients including a detailed analysis of the lesion distribution should increase our knowledge of the early disease. In conclusion, using 3D DIR and 3D FLAIR images in CIS and RRMS, we present evidence that CLs, although already present, are less common in early phases of the disease than previously expected. Despite a clear difference between CIS and MS, no relation with other parameters such as GMF, which characterizes the neuronal compartment, or with disease severity was detected at this early disease stage, indicating that the presence and number of CLs may not be a relevant indicator of early neurodegeneration in MS.
J Neurol Acknowledgments F.Z. is grateful for financial support from the German Multiple Sclerosis Competence Network (KKNMS, Project B7.3) funded by the Federal Ministry for Education and Research (BMBF). Conflicts of interest Pierre Kolber reports no conflicts of interest and financial disclosures. Swantje Montag reports no conflicts of interest and financial disclosures. Dr. Vinzenz Fleischer reports no conflicts of interest and financial disclosures. Dr. Felix Lu¨ssi reports no conflicts of interest and financial disclosures. Janine Wilting reports no conflicts of interest and financial disclosures. Dr. Joachim Gawehn reports no conflicts of interest and financial disclosures. Dr. Adriane Gro¨ger reports no conflicts of interest and financial disclosures. Dr. Frauke Zipp has received research grants from Teva, Merck Serono, Novartis and Bayer as well as consultation funds from Teva, Merck Serono, Novartis, Bayer Healthcare, Biogen Idec Germany, ONO, Genzyme, Sanofi-Aventis and Octapharma. Her travel compensation has been provided for by the aforementioned companies. Ethical standard All patients gave their written informed consent to examinations before participating in this study, which was approved by the local ethics committee and adhered to institutional guidelines in accordance with the Declaration of Helsinki.
13.
14.
15.
16.
17.
18.
References 1. Sander M (1898) Hirnrindenbefunde bei multipler Sklerose. Monatschrift Psychiatrie Neurol IV:427–436 2. Calabrese M, Gallo P (2009) Magnetic resonance evidence of cortical onset of multiple sclerosis. Mult Scler 15:933–941 3. Coebergh JA, Roosendaal SD, Polman CH, Geurts JJ, van Woerkom TC (2010) Acute severe memory impairment as a presenting symptom of multiple sclerosis: a clinical case study with 3D double inversion recovery MR imaging. Mult Scler 16:1521–1524 4. Popescu BF, Bunyan RF, Parisi JE, Ransohoff RM, Lucchinetti CF (2011) A case of multiple sclerosis presenting with inflammatory cortical demyelination. Neurology 76:1705–1710 5. Lucchinetti CF, Popescu BF, Bunyan RF, Moll NM, Roemer SF, Lassmann H, Bruck W, Parisi JE, Scheithauer BW, Giannini C, Weigand SD, Mandrekar J, Ransohoff RM (2011) Inflammatory cortical demyelination in early multiple sclerosis. N Engl J Med 365:2188–2197 6. Bo L, Vedeler CA, Nyland H, Trapp BD, Mork SJ (2003) Intracortical multiple sclerosis lesions are not associated with increased lymphocyte infiltration. Mult Scler 9:323–331 7. Kutzelnigg A, Lucchinetti CF, Stadelmann C, Bruck W, Rauschka H, Bergmann M, Schmidbauer M, Parisi JE, Lassmann H (2005) Cortical demyelination and diffuse white matter injury in multiple sclerosis. Brain 128:2705–2712 8. Calabrese M, De Stefano N (2014) Cortical lesion counts by double inversion recovery should be part of the MRI monitoring process for all MS patients: yes. Mult Scler 20:537–538 9. Chard D (2014) Cortical lesion counts by double inversion recovery should be part of the MRI monitoring process for all MS patients: no. Mult Scler 20:539–540 10. Rovira A, Auger C (2014) Cortical lesion counts by double inversion recovery should be part of the MRI monitoring process for all MS patients: commentary. Mult Scler 20:541–542 11. Sethi V, Muhlert N, Ron M, Golay X, Wheeler-Kingshott CA, Miller DH, Chard DT, Yousry TA (2013) MS cortical lesions on DIR: not quite what they seem? PLoS One 8:e78879 12. Sethi V, Yousry TA, Muhlert N, Ron M, Golay X, WheelerKingshott C, Miller DH, Chard DT (2012) Improved detection of
19.
20.
21.
22.
23.
24.
25.
26.
27.
cortical MS lesions with phase-sensitive inversion recovery MRI. J Neurol Neurosurg Psychiatry 83:877–882 Tallantyre EC, Morgan PS, Dixon JE, Al-Radaideh A, Brookes MJ, Morris PG, Evangelou N (2010) 3 Tesla and 7 Tesla MRI of multiple sclerosis cortical lesions. J Magn Reson Imaging 32:971–977 Calabrese M, Agosta F, Rinaldi F, Mattisi I, Grossi P, Favaretto A, Atzori M, Bernardi V, Barachino L, Rinaldi L, Perini P, Gallo P, Filippi M (2009) Cortical lesions and atrophy associated with cognitive impairment in relapsing–remitting multiple sclerosis. Arch Neurol 66:1144–1150 Calabrese M, Poretto V, Favaretto A, Alessio S, Bernardi V, Romualdi C, Rinaldi F, Perini P, Gallo P (2012) Cortical lesion load associates with progression of disability in multiple sclerosis. Brain 135:2952–2961 Calabrese M, Rocca MA, Atzori M, Mattisi I, Bernardi V, Favaretto A, Barachino L, Romualdi C, Rinaldi L, Perini P, Gallo P, Filippi M (2009) Cortical lesions in primary progressive multiple sclerosis: a 2-year longitudinal MR study. Neurology 72:1330–1336 Damasceno A, Damasceno BP, Cendes F (2014) The clinical impact of cerebellar grey matter pathology in multiple sclerosis. PLoS One 9:e96193 Papadopoulou A, Muller-Lenke N, Naegelin Y, Kalt G, Bendfeldt K, Kuster P, Stoecklin M, Gass A, Sprenger T, Radue EW, Kappos L, Penner IK (2013) Contribution of cortical and white matter lesions to cognitive impairment in multiple sclerosis. Mult Scler 19:1290–1296 Roosendaal SD, Moraal B, Pouwels PJ, Vrenken H, Castelijns JA, Barkhof F, Geurts JJ (2009) Accumulation of cortical lesions in MS: relation with cognitive impairment. Mult Scler 15:708–714 Geurts JJ, Pouwels PJ, Uitdehaag BM, Polman CH, Barkhof F, Castelijns JA (2005) Intracortical lesions in multiple sclerosis: improved detection with 3D double inversion-recovery MR imaging. Radiology 236:254–260 Geurts JJ, Roosendaal SD, Calabrese M, Ciccarelli O, Agosta F, Chard DT, Gass A, Huerga E, Moraal B, Pareto D, Rocca MA, Wattjes MP, Yousry TA, Uitdehaag BM, Barkhof F (2011) Consensus recommendations for MS cortical lesion scoring using double inversion recovery MRI. Neurology 76:418–424 Polman CH, Reingold SC, Banwell B, Clanet M, Cohen JA, Filippi M, Fujihara K, Havrdova E, Hutchinson M, Kappos L, Lublin FD, Montalban X, O’Connor P, Sandberg-Wollheim M, Thompson AJ, Waubant E, Weinshenker B, Wolinsky JS (2011) Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol 69:292–302 Kurtzke JF (1983) Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology 33:1444–1452 Chen JT, Pirko I (2011) MS cortical lesion or not? Double inversion recovery MRI reveals some answers and uncertainties. Neurology 76:412–413 Simon B, Schmidt S, Lukas C, Gieseke J, Traber F, Knol DL, Willinek WA, Geurts JJ, Schild HH, Barkhof F, Wattjes MP (2010) Improved in vivo detection of cortical lesions in multiple sclerosis using double inversion recovery MR imaging at 3 Tesla. Eur Radiol 20:1675–1683 Schmidt P, Gaser C, Arsic M, Buck D, Forschler A, Berthele A, Hoshi M, Ilg R, Schmid VJ, Zimmer C, Hemmer B, Muhlau M (2012) An automated tool for detection of FLAIR-hyperintense white-matter lesions in Multiple Sclerosis. Neuroimage 59:3774–3783 Chard DT, Parker GJ, Griffin CM, Thompson AJ, Miller DH (2002) The reproducibility and sensitivity of brain tissue volume measurements derived from an SPM-based segmentation methodology. J Magn Reson Imaging 15:259–267
123
J Neurol 28. Rudick RA, Fisher E, Lee JC, Simon J, Jacobs L (1999) Use of the brain parenchymal fraction to measure whole brain atrophy in relapsing–remitting MS. Multiple Sclerosis Collaborative Research Group. Neurology 53:1698–1704 29. Roep BO, Buckner J, Sawcer S, Toes R, Zipp F (2012) The problems and promises of research into human immunology and autoimmune disease. Nat Med 18:48–53 30. Siffrin V, Vogt J, Radbruch H, Nitsch R, Zipp F (2010) Multiple sclerosis—candidate mechanisms underlying CNS atrophy. Trends Neurosci 33:202–210 31. Zipp F, Gold R, Wiendl H (2013) Identification of inflammatory neuronal injury and prevention of neuronal damage in multiple sclerosis: hope for novel therapies? JAMA Neurol 70:1569–1574 32. Calabrese M, De Stefano N, Atzori M, Bernardi V, Mattisi I, Barachino L, Morra A, Rinaldi L, Romualdi C, Perini P, Battistin L, Gallo P (2007) Detection of cortical inflammatory lesions by double inversion recovery magnetic resonance imaging in patients with multiple sclerosis. Arch Neurol 64:1416–1422 33. Bo L, Vedeler CA, Nyland HI, Trapp BD, Mork SJ (2003) Subpial demyelination in the cerebral cortex of multiple sclerosis patients. J Neuropathol Exp Neurol 62:723–732 34. Kutzelnigg A, Lassmann H (2005) Cortical lesions and brain atrophy in MS. J Neurol Sci 233:55–59
123
35. Kilsdonk ID, de Graaf WL, Soriano AL, Zwanenburg JJ, Visser F, Kuijer JP, Geurts JJ, Pouwels PJ, Polman CH, Castelijns JA, Luijten PR, Barkhof F, Wattjes MP (2013) Multicontrast MR imaging at 7T in multiple sclerosis: highest lesion detection in cortical gray matter with 3D-FLAIR. Am J Neuroradiol 34:791–796 36. Vural G, Keklikoglu HD, Temel S, Deniz O, Ercan K (2013) Comparison of double inversion recovery and conventional magnetic resonance brain imaging in patients with multiple sclerosis and relations with disease disability. Neuroradiol J 26:133–142 37. Rocca MA, De Meo E, Amato MP, Copetti M, Moiola L, Ghezzi A, Veggiotti P, Capra R, Fiorino A, Pippolo L, Pera MC, Falini A, Comi G, Filippi M (2014) Cognitive impairment in paediatric multiple sclerosis patients is not related to cortical lesions. Mult Scler 38. Rovaris M, Filippi M, Minicucci L, Iannucci G, Santuccio G, Possa F, Comi G (2000) Cortical/subcortical disease burden and cognitive impairment in patients with multiple sclerosis. Am J Neuroradiol 21:402–408 39. Giorgio A, Stromillo ML, Rossi F, Battaglini M, Hakiki B, Portaccio E, Federico A, Amato MP, De Stefano N (2011) Cortical lesions in radiologically isolated syndrome. Neurology 77:1896–1899
