Background
Traditionally, multiple sclerosis (MS) is considered as an autoimmune inflammatory disorder [
1], predominantly affecting the white matter of the central nervous system [
2]. This view is based on the high sensitivity of conventional MRI sequences in the detection of white matter abnormalities in MS patients. Nevertheless, histopathologic analyses of demyelinated foci in the cerebral cortex of patients with MS - dating back to the beginning of the 20
th century [
3] - gave first evidence for cortical involvement in MS. In the early 1960s, Brownell and Hughes described that 26 % of the MS lesions affected the gray matter [
4]. However, the classical view of MS as a pure white matter pathology has not been overcome until the beginning of the 21
st century. Due to the introduction of more sensitive imaging techniques such as double inversion recovery (DIR), gray matter involvement of the MS pathology is now well established [
5,
6]. By suppressing the signals form the cerebrospinal fluid and white matter, DIR sequences have a higher sensitivity compared to conventional MR sequences [
4] and thus, have made a major contribution in detecting focal cortical lesions in MS. Although a recent study has shown that DIR can detect only a minority of cortical lesions in MS, the same study has also shown a significant correlation between the number of DIR-hyperintense cortical lesions on the one hand and the total number of cortical lesions in post-mortem histopathologic analysis on the other hand [
7]. Moreover, DIR is currently considered the best wide-scale application pulse sequence for cortical lesion detection and offers high sensitivity, specificity and accuracy for the detection of gray matter lesions [
8].
Multiple sclerosis is associated with a variety of symptoms that are responsible for functional impairment in affected patients. Besides physical disability, cognitive deficits – which were firstly described in the second half of the 19
th century [
9] – are found in up to 70 % of patients with multiple sclerosis at both the earlier and later stages of the disease [
10]. MS-related cognitive impairment can affect various aspects of cognition. Processing speed and episodic memory seem most commonly affected [
11]. However, MS patients often exhibit significant deficits in executive functions too [
12,
13].
The extended focus in MS research including gray matter pathology had important consequences for the study of the neural correlates of MS symptoms. In recent years, the contribution of DIR-hyperintense cortical lesions to functional loss is particularly highlighted. Cross-sectional and longitudinal studies have shown correlations between the number and/or volume of cortical lesions and cognitive [
14,
15] or physical [
16] impairment. Regional associations are also described, for instance between an accumulation of cortical lesions in mesio-temporal areas and impairments in episodic memory [
12,
17].
However, these studies are not without limitations. For example, detailed cognitive investigation [
14] or DIR sequences were lacking [
15]. Here we try to overcome these shortcomings and apply a novel methodological approach to investigate the clinical relevance of cortical lesions in MS, i.e. by dichotomizing patients based on the presence or absence of cortical lesions. The resulting two subgroups were compared with regard to demography, cognition, fatigue, affective mood state, and several other established MRI markers of disease severity, for example T2 lesion load, third ventricle width, and global cortical thinning.
Discussion
The impact of gray matter pathology, in particular that of gray matter lesions, on cognitive and physical functioning in MS patients has been discussed for many years. Here we highlight the clinical relevance of DIR-hyperintense cortical lesions in patients with RRMS. The most intriguing finding of the present study was that patients with - compared to those without visible cortical lesions - differed from each other in global cortical thickness and mnestic functions, whereas no differences between these two patient groups were observed regarding EDSS, age, age at diagnosis, disease duration, or non-mnestic cognitive functions. Moreover, patients without cortical lesions showed normal cortical thickness and mnestic functions, when compared with a group of healthy controls.
That a reduction of cortical thickness can occur in MS [
36] - and that this thinning is related to global e.g. [
37] and even specific cognitive impairment [
12,
38] - has been found in previous studies. Similar to a recent finding [
39], we show an association between MS-related cortical thinning and the presence of cortical MS lesions. It is known that gray matter pathology involves both inflammatory and degenerative mechanisms, but the relationship between the two remains unclear [
6]. Gray matter atrophy might be the final step of several pathological processes, which could include cortical demyelination but also retrograde degeneration secondary to white matter lesions and, perhaps, primary neurodegeneration [
40].
We detected DIR-hyperintense cortical lesions in 76 % of our RRMS patients, supporting the notion of a high prevalence of these lesions [
16]. In further agreement with previous findings [
41], we observed an uneven spatial distribution of cortical lesions over the cerebral cortex, with a prominent accumulation in memory-relevant mesiotemporal regions, particularly in the bilateral parahippocampal gyrus. While the importance of the hippocampus for memory function is known since the classical description of the patient H.M. in 1957 [
42], parahippocampal involvement in memory functions was recognized only two decades ago [
43]. Squire and Zola-Morgen identified the anatomical components of what is termed the medial temporal memory system [
44]. By now it is well known that bilateral damage to the medial temporal lobe causes severe learning and memory impairments. This relationship has also been shown in MS patients. Learning and memory is the most frequently disrupted cognitive domain in MS, reported in 40–60 % of patients [
45]. Moreover, Coebergh et al. [
46] described a patient with acute memory impairment, associated with hippocampal and cortical lesions. Cortical lesions were also associated with cognitive decline in a group of 13 MS patients [
17]. In this study, a significant correlation between hippocampal lesion load and visuospatial memory was observed. Based on the present and previous findings, we thus conclude that mesiotemporal cortical lesions are highly prevalent in RRMS patients and play a crucial role in the development of mnestic dysfunction.
The association we found between memory impairment and both cortical thinning as well as cortical lesions seems particularly intriguing: One might speculate that mnestic dysfunction in MS patients could indicate cortical involvement of the MS pathology in general. Related to this assumption, a rarely occurring variant of so-called 'cortical MS' has been described in previous studies [
47,
48]. The condition was predominantly characterized by the presence of neurobehavioral symptoms (e.g. depression, apathy) and neuropsychological deficits (e.g. agraphia, anomia) suggesting cortical dysfunction. However, detailed and explicit imaging data for detecting cortical involvement (e.g. DIR) is missing in these studies. By approaching from the imaging side, our patient group with cortical involvement may reflect a different and somehow incomplete variant of cortical MS, as they showed distinct mnestic deficits in association with cortical involvement, but no other cortical symptoms such for example depression. Moreover, we found no patient with cortical lesions in the absence of subcortical lesions. In all patients, including those of the CL group, the majority of MS lesions was located subcortically. Taking this into account, we here propose that our CL group may represent a 'cortically dominant' subtype of MS. In these patients, pathophysiological processes might be different from those of patients without cortical involvement.
Beside the difference between the two patient groups in mnestic functions, both patient group showed executive deficits compared to the healthy control group. This finding is in line with previous studies that have shown significant executive dysfunction in MS patients [
12,
13,
49].
Much research and clinical development in MS has focused on the inflammatory mechanisms of the disease. Meanwhile, multiple disease-modifying drugs (DMD) are available that target the inflammatory pathology of MS, in particular the development of new white matter lesions [
50]. A recent study demonstrated that DMD - in particular IFN β-1a and glatiramer acetate - can reduce the accumulation of cortical lesions too [
51]. In addition, Filippi and colleagues [
52] showed that the presence of at least one cortical lesion is associated with a high risk of conversion from clinically isolated syndrome (CIS) to definite MS within a short period. Together with our results, these findings highlight the relevance of cortical lesions as a “target” in the development of new DMD, and to include cortical lesions as a primary outcome variable in disease and treatment monitoring.
This study is not without limitations. As already mentioned earlier, DIR represents only a limited snapshot of the real cortical pathology that is present in MS patients. The combination of DIR and a T1-weighted phase-sensitive inversion recovery (PSIR) sequence would substantially improve the sensitivity of detecting lesions present in the cortex of MS patients [
53]. Moreover, the present analyses are limited due to the low MRI field (1.5 T) applied. It has been shown that high field (3 T) and ultra-high field (7 T) MRI systems deliver a higher detection rate of cortical lesions in vivo [
54]. Finally, a further limitation of the present study is the relatively small and unbalanced sample size.
Acknowledgments
Not applicable.