You are here
Cognitive event-related potentials in multiple sclerosis: Correlation with MRI and neuropsychological findings
Multiple Sclerosis and Related Disorders, Volume 10, November 2016, Pages 192 - 197
Cognitive event-related potentials (ERPs) have been previously correlated with T2 lesion load (Τ2LL) in patients with multiple sclerosis (MS). It is currently unknown, however, whether ERPs also correlate with brain atrophy or the presence of T1 hypointense lesions (“black holes”) which reflect tissue destruction and axonal loss. The primary aim of the current study is to explore the effect of neuroradiological parameters such as brain atrophy, T1 and T2 lesion load on auditory ERPs in MS patients. In addition, we correlated cognitive impairment with neurophysiological (ERP) and neuroradiological (MRI) variables and investigated whether a combination of ERP and MRI parameters is capable of distinguishing patients suffering from secondary progressive (SP), primary progressive (PP) and relapsing-remitting (RR) MS.
Materials and methods
The study sample consisted of fifty nine MS patients (mean age±SD: 37.82±1.38 years; average disease duration: 6.76±5.3 years) and twenty six age-matched controls (mean age±SD: 41.42±15.39 years). The patients’ EDSS and NRS scores were 3.77±2.14 (range: 1–7.5) and 75.88±11.99 (range: 42–94) respectively. ERPs were recorded using the auditory “odd-ball” paradigm. T1 and T2 lesions were automatically segmented using an edge-finding tool and total lesion volumes were calculated. MRI measures of brain atrophy included third ventricle width (THIRDVW) and the ratio of mid-sagittal corpus callosum area to the mid-sagittal intracranial skull surface area (CC/MISS). Statistical analysis was performed using multiple regression, principal component (PCA) and discriminant analysis.
T1 lesion load emerged as the most significant predictor of P300 and N200 latency. The rest of the endogenous ERPs parameters (P300 amplitude, N200 amplitude) were not significantly correlated with the MRI variables. PCA of pooled neuroradiological and neurophysiological markers suggested that four components accounted for 64.6% of the total variability. Discriminant analysis based on ERP & MRI markers classified correctly 79.63% of patients in RR, PP and SP subgroups.
T1 lesion load is the most significant MRI correlate of auditory ERPs in MS patients. Importantly, ERPs in combination with MRI variables can be usefully employed for distinguishing patients with different subtypes of MS.
- The impact of T1 and T2 lesion loads as well as brain atrophy markers on auditory ERPs in MS patients was studied.
- T1 lesion load is the most significant MRI correlate of auditory ERPs in MS patients.
- ERPs in combination with MRI variables, can be usefully employed for distinguishing different MS subtypes.
Keywords: Multiple sclerosis, MRI, Event-related potentials, Brain atrophy, Cognitive impairment.
The impairment of cognitive function in multiple sclerosis (MS), particularly in the advanced stages of the disease, was first described in the 19th century by the famous neuropsychiatrist Charcot (Charcot, 1877) who noted “there is marked enfeeblement of the memory; conceptions are formed slowly”. Recent studies provided a detailed insight of cognitive dysfunction in MS and identified attentional disorders, slowing of thought processes and memory disturbances as its’ major components. These deficits occur in 30–70% of MS patients (Rao et al, 1991 and Peyser et al, 1980) and are frequently compounded by psychiatric comorbidities, most notably depression.
The pathophysiological substrate of cognitive impairment in MS has been extensively investigated with various MRI techniques as well as neurophysiological methods including cognitive event-related potentials (ERPs). A number of studies established that the latencies of N200 and P300 waves of auditory ERPs are consistently increased in MS, even in patients with a clinically isolated syndrome (Kocer et al., 2008) whereas an amplitude decrease is a less stable finding (Newton et al, 1989, Gil et al, 1993, Giesser et al, 1992, and Sundgren et al, 2015). These electrophysiological abnormalities correlate with the T2 lesion load on MRI (Sánchez et al, 2008 and Piras et al, 2003). The influence, however, of other neuroradiological parameters (i.e. brain atrophy and T1 hypointense lesions or “black holes”) on ERPs remains unexplored. This issue is of potential importance as brain atrophy and T1 lesion load, which reflects tissue destruction and axonal loss, have been shown to correlate with cognitive impairment (Honig et al., 1992).
The main purpose of the present work is to study the impact of T1 and T2 lesion loads as well as brain atrophy markers on auditory ERPs in MS patients. In addition, we correlated cognitive impairment with ERP and MRI variables and investigated whether a combination of ERP and MRI parameters is capable of distinguishing patients suffering from SP, PP and RR MS.
2. Materials and methods
The study sample consisted of fifty nine MS patients (mean age±SD: 37.82±1.38 years; average disease duration: 6.76±5.3 years) and twenty six age-matched controls (mean age±SD: 41.42±15.39 years). All participants entered the study after giving informed consent for the procedures which were approved by an institutional ethics committee and were performed in accordance with the ethical standards laid down in the Declaration of Helsinki. The patients’ EDSS and NRS scores were 3.77±2.14 (range: 1–7.5) and 75.88±11.99 (range: 42–94), respectively. The group was divided into 3 subgroups: patients with secondary progressive (n=20), primary progressive (n=10) and relapsing remitting MS (n=29).
2.1. Electrophysiological study
Auditory event-related potentials were elicited using a simple discrimination task, the so-called ‘‘oddball paradigm’’. Briefly, a series of binaural tones at 70 dB sound pressure level (SPL) with a 10 ms rise/fall and a 100 ms plateau time was presented to all subjects. The auditory stimuli were presented in a random sequence with target tones of 2000 Hz occurring 20% of the time and standard tones of 1000 Hz occurring 80% of the time with an interstimulus interval of 2 s. The subject was required to distinguish between the two tones by responding to the target (i.e. mentally counting) and not responding to the standard. Patients were instructed to pay attention in distinguishing the tones, count the target tones silently and report the total number at the end of the exam.
EEG activity was recorded (filter bandpass: .1–50 Hz, 1 s epochs ranging from −100 to +900 ms relative to stimulus onset) from scalp AgCl electrodes at Cz and Pz sites according to the 10/20 system referred to linked earlobe electrodes, with a right hand ground. Artifacts caused by ocular movements±50 μV were automatically rejected. Bioelectrical activity was digitized (12 bit) in 250 Hz, 100 ms before till 900 ms after the stimulus was provided. The responses to target and non-target stimuli were averaged separately. Each patient was tested twice to ensure that waveform components are reproducible.
Latencies and amplitudes of Ν100, P200, Ν200 and P300 waves were measured after averaging of the independent waveforms. The peak of the ERP components was measured as follows: if the waveform was smooth, the maximal amplitude point was taken as a peak. Otherwise, the leading and trailing slopes of the waveform were extended, and the intersection point was determined. N100 was the waveform of maximal negativity between 80 and 160 ms and P200 wave was the waveform of maximal positivity between 150 and 250 ms. N200 wave was defined as the maximal negativity between 175 and 250 ms that appeared before the P300 wave that, in turn, was defined as the maximal positivity between 250 and 600 ms. In patients with discrete P3a and P3b waveforms, only P3b was taken into consideration8. For the needs of the present endogenous ERP study, amplitudes and latencies of N200 and P300 waves were studied.
2.2. Neuroradiological assessment
All patients were examined with a standard MS protocol using a 1.5T MR scanner including T1SE (with and without contrast administration), T2TSE, FLAIR and DWI sequences. No contrast-enhancing lesions were observed. The contouring of demyelinated plaques in the T1 and T2 sequences of MRI and brain atrophy measurements were performed using an image analysis software (ImageJ, 1.45, National Institute of Health, USA). The total volume of Τ1 (T1LL) and Τ2 lesions (T2LL) was calculated as the sum of the volumes of individual lesions. We defined ‘black holes’ as any abnormal hypointensity compared to normal-appearing white matter on a T1-weighted MRI that also appears hyperintense in the corresponding T2-weighted image (Sahraian and Radue, 2008). MRI measures of brain atrophy included the ratio of mid-sagittal corpus callosum area to the mid-sagittal intracranial skull surface area (CC/MISS) and third ventricle width (THIRDVW). The latter was measured according to the method of Benedict et al. (2004). Briefly, a line region of interest was drawn through the long axis of the ventricle, parallel to the interhemispheric fissure in the slice where the third ventricle was most visible. The width was measured by drawing a second line perpendicular to the first at its midpoint and recording its length.
2.3. Neuropsychometric evaluation
A number of neuropsychological tests that assess cognitive functions such as verbal memory, working memory, attention and abstract thought were applied in a subgroup of 26 patients. In particular, attention and concentration were assessed by the neuropsychometric tests, Stroop Test (Stroop, 1935) and Trail Making Test (Vlahou and Kosmidis, 2002). Abstract logic was tested by Raven Progressive Matrices Sets: A, B, C, D and E for adults (Raven, 1960) and Wisconsin Card Sorting Test (Heaton et al., 1993). The Paced Auditory Serial Addition Test (PASAT) was not included as test material because a subgroup of study participants had repeatedly performed the test in the context of previous studies and the resulting learning effects might act as a confounder in the analysis (Gronwall, 1977). Verbal memory was assessed by two subtests of the Wechsler Memory Scale Form II (Wechsler, 1945) (logical memory and subtest combinational memory) as well as the Verbal Fluency Test (Kosmidis et al., 2004). Z-scores of these tests were also calculated as well as the Z-score of the patients' total cognitive function.
2.4. Statistical analysis
Differences in ERP characteristics between MS patients and controls were tested with a t-test for independent samples and the non-parametric Mann–Whitney test, depending on normality of the variables’ distribution. Normality was tested using the Shapiro–Wilk test, as well as graphical methods (Q-Q plots). The relationship between total cognitive performance and neuroradiological parameters and ERP characteristics was investigated with stepwise multiple regression analysis, whereas the correlation between neuropsychometric tests and radiological and electrophysiological parameters was performed with Spearman's correlation coefficient. In multiple regression analysis, multicollinearity was assessed with the commonly utilized variance inflation factor (VIF) and Tolerance techniques. Variables were dropped if the VIF was greater or equal to 5 or Tolerance was less than 0.1. Principal component analysis (PCA) was applied to pooled data of neuroradiological and neurophysiological markers and age in order to reduce the dimensionality of the data and investigate the relationships between the markers. Finally, discriminant analysis was applied to establish those markers capable of distinguishing patients that were suffering from secondary progressive, primary progressive and relapsing-remitting MS. Wilk's lambda (λ) criterion was used for selecting discriminant variables. For all tests a p-value <.05 was considered statistically significant. Statistical analysis was performed using statistical packages (SPSS version 23.0, SPSS Inc., Chicago, IL, USA, STATGRAPHICS Centurion XV, Manugistics Inc., Rockville, USA).
The ERP and MRI characteristics of the participant groups are provided in Table 1 and Table 2, respectively. In line with previous studies, the ERP characteristics of the patient group were significantly different from the results in the healthy controls. For instance, P300 latency was significantly delayed in the patient group compared to controls (p=.005) whereas no significant difference was found regarding P300 amplitude (p>.05) (Table 1). On the other hand, N200 amplitude was significantly reduced in MS patients (p<.05). It should be noted that the slopes of the linear regression lines of age versus P300 latency in the control (1.592±.357, p<.001) and the patient groups (.879±.524, p>.05) were significantly different as in the latter case the relationship between the dependent variable and age was not significant (Fig. 1).
ERP characteristics of the participant groups.
|Patient group (n=59)||Control group (n=26)||p|
|Age||36.0 (31.0, 43.0)||41.42±15.39||.655|
|N100 latency (ms)||102.97±14.14||95.08±7.80||.002|
|N100 amplitude (μV)||10.70±4.13||10.03±3.42||.494|
|N200 latency (ms)||223.0 (212.0, 245.0)||230.36±16.51||.559|
|N200 amplitude (μV)||5.45 (3.75, 8.30)||8.59 (5.56, 12.35)||.040|
|P200 latency (ms)||168.62±19.20||170.16±12.12||.665|
|P200 amplitude (μV)||10.86±5.02||13.11±3.97||.031|
|P300 latency (ms)||369.97± 40.74||337.5 (319.0, 360.0)||.005|
|P300 amplitude (μV)||13.7 (10.2, 21.2)||16.65±7.1||.587|
Data are presented as Mean ± SD or Median (1st quartile, 3rd quartile), when normality is violated. P-values derived through the independent t-test for the normally distributed variables and through the non parametric Mann-Whitney U-test for the skewed ones.
P-values <.05 were considered as significant.
MRI characteristics of the patient group.
|T1 lesion load (ml)||1.61 (.55, 4.08)|
|T2 lesion load (ml)||8.25 (3.78, 23.67)|
|Third ventricle width (mm)||5.14 (3.93, 7.14)|
aData are Means±SD or Median (1st quartile, 3rd quartile), when normality is violated.
Linear regression lines of P300 wave latency and age in controls (Α,P300 latency=280.251+1.592*age, r2=.454, p<.001) and MS patients (Β,P300 latency=338.309+.879*age, r2=.05, p>.05). The curved lines correspond to the 95% confidence interval of the regression lines.
The correlation between ERPs and neuroradiological findings was performed with multiple regression analysis, using ERP characteristics (N200, P300 latency and amplitude) as dependent variables.
Regarding P300 latency, the step-wise multiple regression model with all five predictors (neuroradiological markers T1LL, T2LL, 3RDVW, CC/MISS and age) retained T1LL as the most significant one. The optimal equation for fitting the data was: P300 latency=359.17+1.82[T1LL], R=.38, p=.004 (Fig. 2).
Linear regression analysis (with 95% confidence interval) of the correlation between P300 wave latency and lesion load in T1 sequence.
Similarly, in the multiple regression model with N200 latency as dependent variable, T1 lesion load was the most important N200 latency predictor (R=.42, p=.002). The regression equation was: N200 latency=224.19+1.46[T1LL]. The rest of the endogenous ERPs parameters (N200 & P300 amplitude) were not significantly correlated with the MRI variables.
The results of the correlations between neuropsychometric tests, neuroradiological and electrophysiological markers are summarized in Table 3. Statistically significant correlations were observed between T1 and T2 lesion load and verbal memory. Moreover, significant correlations were observed between N200 amplitude, P300 latency and verbal memory. The rest of the correlations did not reach statistical significance (Table 3).
Correlation coefficients between neuroradiological, neurophysiological parameters and Z-scores of the neuropsychological tests.
|Test||T1 LL||T2 LL||3RDVW||CC/MISS||N200 latency||N200 amplitude||P300 latency||P300 amp|
|Sum of correct answers (WCST)||−.28||−.16||.030||−.058||.33||.33||−.48||.18|
|Number of categories filled||−.21||−.03||−.18||.04||−.33||.57*||−.23||−.019|
|Number of tests for 1st category||−.27||.28||.19||−.08||.35||.37*||−.23||−.02|
|Logical memory (WMS)||−.30||−.27||−.34||.11||.07||−.13||−.62*||−.22|
|Combination learning (WMS)||−.37||−.33||−.34||−.08||.15||−.16||−.51||−.23|
|Trail part A||.17||.055||−.20||.069||−.24||.22||.32||−.052|
|Trail part B||.096||.103||.015||−.306||−.04||.09||−.43||−.017|
|Semantic information flow (word count)||−.46*||.51*||−.128||−.16||.15||−.16||−.38||.14|
|Phonological information flow (word count)||−.44*||.48*||−.136||−.20||.62||−.1||−.48*||.29|
Multiple regression analysis with total cognitive performance as dependent variable and age, ERP & MRI parameters as predictors produced an R2 of 69.67% (p=.006). The regression equation was Total cognitive=−40.1097+.234*Age+.138*N100 Latency+.292*T2 LL−1.576*T1 LL+492.146*CC/MISS. These data indicate that a model comprising ERP and MRI variables can explain about 70% of the total cognitive performance.
Principal component analysis (PCA) was applied to pooled data of neuroradiological and neurophysiological markers and age in order to reduce the dimensionality of the data and to detect the most important causes of variability, since a significant correlation between the markers was noticed. PCA resulted in four main principal components with eigenvalues greater than 1.0, a common statistical cut-off point. The four selected components accounted for 64.6% of the total variability. In Fig. 3, neuroradiological and neurophysiological markers and age were represented as a function of both first and second principal components. The first principal component explained 25.52% of the total variability and was defined by P200 Latency, N200 Latency and N200 Amplitude. These markers were placed closed together on the positive side of PC1, indicating that they were positively correlated with each other. They were located away from the axis origin, suggesting that they were well represented from the first PC, which could be regarded as representative of the neurophysiological markers. The second principal component explained another 18.90% of the total variability and was mainly defined by T1 lesion load (T1LL), T2 lesion load (T2LL) and 3RDVW. Consequently the second principal component could be a representative of the neuroradiological markers. In Fig. 3, these markers were located closed together on the negative side of PC2, indicating a strong positive correlation. The third principal component explained another 11.04% of the total variability and was defined by CC/MISS highlighting the importance of corpus callosum integrity. This ratio was placed opposite to T1 LL, T2 LL and 3RDVW; therefore it was negatively correlated to them. The fourth principal component explained another 9.14% of the total variability and was defined by P300 latency, N100 latency and age. P300 latency and age were placed close together indicating a positive correlation (Table 4).
Principal component analysis. Plot of the two first principal components.
Classification by discriminant analysis in RR, PP & SP subgroups.
Percent of cases correctly classified: 79.63%
Α discriminant analysis was further applied in order to investigate if the patients that were suffering from primary progressive (PP), secondary progressive (SP) and relapsing-remitting (RR). MS can be distinguished by the neuroradiological and neurophysiological markers and in order to establish those markers capable of discriminating and classifying the MS patients. One discriminant function (83.77%) was statistically significant (P<.001) for distinguishing the patients. N100, N200, P300 amplitude from neurophysiological markers and T2 and T1 lesion loads from neuroradiological markers were mainly responsible for the observed discrimination (Fig. 4). The percentage of the patients that were classified into the correct group was 79.63%. Some overlapping was observed (Fig. 4), mostly for the primary progressive (PP) group. The (SP) and (RR) groups of MS patients were separated quite clearly.
Discriminant plot separating the patients (PP, RR, SP) by the neuroradiological and neurophysiological markers. Sign(+) is the centroid of each group.
ERPs, and the P300 wave in particular, were extensively used over the last decades in the context of psychophysiological studies. This wave was considered a cognitive or endogenous electrophysiological phenomenon, because in contrast to evoked potentials, that are mostly dependent on the physical characteristics of the stimuli, it requires the subject's cognitive activation. The P300 wave is elicited when the subject tries to discriminate two different stimuli (oddball paradigm). This cognitive discrimination task produces a wave of high amplitude (10–20 μv) and positive polarity, that when elicited by auditory stimuli in young adults has a latency of about 300 ms and was accordingly designated as P (positive) 300 wave (Polich, 2004).
Numerous studies in normal and abnormal populations led to the conclusion that P300 wave is a sensitive biomarker of neuronal activity that reflects working memory and concentration. In normal subjects, P300 wave latency is negatively correlated with cognitive function, especially in cognitive tests that assess the ability to rapidly allocate attentional resources. On the other hand, in patients with Alzheimer's disease and Mild cognitive impairment, P300 wave latency is increased compared to normal subjects (Howe et al, 2014 and Papaliagkas et al, 2008). Therefore, P300 latency is directly correlated with cognitive abilities in normal and abnormal conditions.
ERPs have been widely used in the assessment of cognitive functions in patients suffering from MS (Newton et al, 1989, Gil et al, 1993, Giesser et al, 1992, and Sundgren et al, 2015). According to the results of these studies, significant correlations exist between N200 and P300 wave and neuroradiological parameters such as lesion load in T2 sequence. However, the contribution of other types of lesions such as hypointense lesions in T1 MRI sequence, known as “black holes” and brain atrophy has not been studied. Moroever, P300 latency was also used to assess the efficacy of the medications used in the treatment of relapsing MS (Flechter et al., 2007).
The aim of the current study was to investigate the correlation between ERP characteristics, MRI markers and neuropsychometric tests in patients with MS. The analysis focused on the two most commonly used ERP parameters: latency and amplitude. Other possibly important parameters such as ERP morphology and their topographic distribution were not taken into account and they should be definitely studied in the future for their possible utility as diagnostic markers.
According to the results of the statistical analysis of multiple neuroradiological markers, lesion load in T1 sequence as well as corpus callosum atrophy exerted the most significant impact on ERPs, since they explained data dispersion of both N200 and P300 waves, to a greater extent compared to other parameters. Although observed for the first time, this correlation is not unexpected. The “black holes” correspond to areas of axonal damage with subsequent gliosis and therefore reflect the most destructive aspects of the demyelinating process. The precise mechanism through which demyelinating lesions cause ERP changes is not yet clear. According to some researchers, the P300 wave generator is located in the hippocampus, whereas others suggest that it is located at the level of the frontoparietal lobes (Polich, 2007). Irrespective of the initial location of the generator, the P300 wave spreads towards other cortical areas. It is therefore conceivable that the disruption of white matter connections, caused by multifocal demyelinating plaques and especially their extreme expression (i.e. the "black holes"), will affect the integrity of the neuronal network subserving ERPs and will ultimately result in prolonged ERP latencies.
The fact that no significant correlation between P300 wave latency and age was observed in the patient group, in sharp contrast to normal subjects, is a finding indicating that disease-related parameters override in significance as predictor factors the important physiological parameter of age.
The neuropsychological evaluation showed reduced performance in tests like Raven, Trail-A&B and STROOP-3 in the patient group compared to the control group (Table 5). These findings agree with previous studies of cognitive functions in patients with MS (Rovaris et al., 1998) and reveal a type of cognitive impairment that in advanced stages mimics subcortical dementia. In line with numerous previous studies, the patients’ deficits correlated with neuroradiological findings characteristic of the disease. This observation indicates that neuropsychological impairment is, to a great extent, caused by the functional disconnection of cortical areas as a consequence of the presence of demyelinating lesions. In addition, P300 latency demonstrated a high correlation with cognitive functions that are typically impaired in multiple sclerosis such as logical and verbal memory, as previously reported by other authors (Giesser et al, 1992, Kiiski et al, 2011a, and 28b).
Demographic clinical and neuropsychological features of patient and control group. Means, standard deviations and one-way ANOVA of performance on cognitive tests (Stroop Test, Trail Making Test, Raven Progressive Matrices Test, Wisconsin Card Sorting Test, Logical Memory I & II tasks of Wechsler Memory Scale-III and Verbal Fluency Test) for both groups are presented.
|Patient group (n=26) (mean±SD)||Control group (n=26) (mean±SD)||Fa|
|Age (years)||39.34 ± 11.83||41.42± 15.39|
|Disease duration (years)||6.76±5.3|
|BDI score||3.96 ±3.38|
|Raven Progressive Matrices Test- Total Correct||37.10 (11.04)||52.86 (5.24)||48.193***|
|Stroop Test Word reading-Neutral||86.45 (11.77)||105.72 (8.31)||51.920***|
|Color naming- Congruent||64.59 (13.95)||81.90 (9.53)||30.453***|
|Color naming- Incongruent||35.52 (9.42)||53.76 (11.14)||45.315***|
|Trail Making Test Part 1||44.07 (16.98)||36.96 (9.32)||3.900|
|Part 2||97.48 (47.61)||74.72 (26.85)||5.028*|
|Wisconsin Card Sorting Test- Trails to first category||16.03 (7.05)||10.62 (.62)||16.953***|
|Logical Memory I||21.24(9.03)||35.96(4.88)||59.710***|
|Logical Memory II||10.07(3.40)||13.83 (2.00)||26.310***|
|Verbal Fluency Test- Semantic Fluency||47.41(12.83)||55.34 (10.02)||6.878**|
|Phonetic fluency||29.14(12.78)||43.03 (9.51)||22.066***|
aDegrees of freedom=1, 56.
When neuroradiological and electrophysiological parameters were investigated with regard to their ability to classify patients into disease subtypes, it turned out that a combination of ERP & MRI markers classified correctly 79.63% of patients in RR, PP and SP subgroups. It should be noted that the discriminatory power of this approach was inferior regarding PP patients. These preliminary, cross-sectional data suggest that ERPs, in combination with MRI, can be used to differentiate between the RR and SP stages of MS. Accordingly, they may constitute a biomarker that could be applied in the research or clinical domain for monitoring the transition from the RR to the SP phase of MS. This possibility should be explored in future, longitudinal studies of MS cohorts with combined electrophysiological, neuroradiological and neuropsychological monitoring.
In conclusion, T1 lesion load is the most significant MRI correlate of auditory ERPs in MS patients. In addition, ERPs in combination with MRI variables, can be usefully employed for distinguishing patients with different subtypes of MS.
- Charcot, Charcot J., 1877. Lectures on the disease of the nervous system delivered at La Salpetriere. In: New Sydenham Society. London , pp. 194–195.
- Rao et al., 1991 S.M. Rao, G.J. Leo, L. Bernandin, et al. Cognitive dysfunction in multiple sclerosis 1 frequency, patterns and prediction. Neurology. 1991;41:685-691
- Peyser et al., 1980 J.M. Peyser, K.R. Edwards, C.M. Poser, et al. Cognitive function in patients with multiple sclerosis. Arch. Neurol.. 1980;37:577-591
- Kocer et al., 2008 B. Kocer, T. Unal, B. Nazliel, et al. Evaluating sub-clinical cognitive dysfunction and event-related potentials (P300) in clinically isolated syndrome. Neurol. Sci.. 2008;29:435-444
- Newton et al., 1989 M.R. Newton, G. Barrett, M.M. Callanan, et al. Cognitive event-related potentials in multiple sclerosis. Brain. 1989;112:1637-1660
- Gil et al., 1993 R. Gil, L. Zai, J.P. Neau, et al. Event-related auditory evoked and multiple sclerosis. Electroencephalogr. Clin. Neurophysiol.. 1993;:182-187
- Giesser et al., 1992 B.S. Giesser, M.M. Schroeder, Laroccang, et al. Endogenous event-related potentials as indices of dementia in multiple sclerosis patients. Electroencephalogr. Clin. Neurophysiol.. 1992;82:320-329
- Sundgren et al., 2015 M. Sundgren, V.V. Nikulin, L. Maurex, et al. P300 amplitude and response speed relate to preserved cognitive function in relapsing-remitting multiple sclerosis. Clin. Neurophysiol.. 2015;126:689-697
- Sánchez et al., 2008 M.P. Sánchez, A. Nieto, J. Barroso, et al. Brain atrophy as a marker of cognitive impairment in mildly disabling relapsing-remitting multiple sclerosis. Eur. J. Neurol.. 2008;15:1091-1099
- Piras et al., 2003 M.R. Piras, I. Magnano, E.D. Canu, et al. Longitudinal study of cognitive dysfunction in multiple sclerosis: neuropsychological, neuroradiological, and neurophysiological findings. J. Neurol. Neurosurg. Psychiatry. 2003;74:878-885
- Honig et al., 1992 L.S. Honig, R.E. Ramsay, W.A. Sheremata. Event-related potential P300 in multiple sclerosis. relation to magnetic resonance imaging and cognitive impairment. Arch. Neurol.. 1992;49:44-50
- Sahraian and Radue, 2008 M.A. Sahraian, E.W. Radue. T1 hypotenuse lesions (black holes). M.A. Sahraian, E.W. Radue (Eds.) MRI Atlas of MS Lesions (Springer-Verlag, Heidelberg, 2008) 51-55
- Benedict et al., 2004 R.H. Benedict, B.W. Guttman, I. Fishman, et al. Prediction of neuropsychological impairment in multiple sclerosis-comparison of conventional MRI measures of atrophy and lesion burden. Arch. Neurol.. 2004;61:226-230
- Stroop, 1935 J.R. Stroop. Studies of interference in serial verbal reactions. J. Exp. Psychol.. 1935;18:643-662
- Vlahou and Kosmidis, 2002 K. Vlahou, M. Kosmidis. Trail making test in the Greek population. Normative data for clinical and scientific applications. Psychology. 2002;9:336-352
- Raven, 1960 J.C. Raven. Guide to the Standard Progressive Matrices. (Lewis, London, HK, 1960)
- Heaton et al., 1993 R.K. Heaton, G. Chelune, J.L. Talley, et al. Wisconsin Card Sorting Test manual: Revised and Expanded. (Psychological Assessment Resources, Odessa, FL, 1993)
- Gronwall, 1977 D.W.A. Gronwall. Paced auditory serial-addition task: a measure of recovery from concussion. Percept. Mot. Skills. 1977;44:367-373
- Wechsler, 1945 D. Wechsler. A standardized memory scale for clinical use. J. Psychol.. 1945;19:87-95
- Kosmidis et al., 2004 Μ Kosmidis, K. Vlahou, P. Panagiotaki, et al. The verbal fluency task in the Greek population: normative data, and clustering and switching strategies. J. Int. Neuropsychol. Soc.. 2004;10:164-172
- Polich, 2004 J. Polich. Clinical application of the P300 event-related brain potential. Phys. Med. Rehabil. Clin. N. Am.. 2004;18:133-136
- Howe et al., 2014 A.S. Howe, A. Bani-Fatemi, V. De Luca. The clinical utility of the auditory P300 latency subcomponent event-related potential in preclinical diagnosis of patients with mild cognitive impairment and Alzheimer's disease. Brain Cognit.. 2014;86:64-74
- Papaliagkas et al., 2008 V. Papaliagkas, V. Kimiskidis, M. Tsolaki, et al. Usefulness of event-related potentials in the assessment of mild cognitive impairment. BMC Neurosci.. 2008;9:107
- Flechter et al., 2007 S. Flechter, J. Vardi, Y. Finkelstein, L. Pollak. Cognitive dysfunction evaluation in multiple sclerosis patients treated with interferon beta-1b: an open-label prospective 1 year study. Isr. Med. Assoc. J.. 2007;9:457-459
- Polich, 2007 J. Polich. Updating P300: an integrative theory of P3a and P3b. Clin. Neurophysiol.. 2007;118:2128-2148
- Rovaris et al., 1998 M. Rovaris, M. Filippi, L. Minicucci, et al. Relation between MRI abnormalities and patterns of cognitive impairment in multiple sclerosis. Neurology. 1998;50:1601-1608
- Kiiski et al., 2011a H. Kiiski, R.B. Reilly, R. Lonergan, et al. Change in PASAT performance correlates with Change in P3 ERP amplitude over a 12-month period in multiple sclerosis patients. J. Neurol. Sci.. 2011;305:45-52
- b H. Kiiski, R. Whelan, R. Lonergan, et al. Preliminary evidence for correlation between PASAT performance and P3a and P3b amplitudes in progressive multiple sclerosis. Eur. J. Neurol.. 2011;18:792-795
a Laboratory of Clinical Neurophysiology, AHEPA Hospital, Aristotle University of Thessaloniki, Greece
b Laboratory of Mathematics and Statistics, Agricultural University of Athens, Greece
c Department of Psychology, Aristotle University of Thessaloniki, Greece
d 3rd Department of Neurology, Aristotle University of Thessaloniki, “G Papanikolaou” General Hospital, Exohi, Greece
e Department of Radiology, Faculty of Medicine, University of Crete, University Hospital, Heraklion 71110 Stavrakia, Greece
f Department of Radiology, “G Papanikolaou” General Hospital, Exohi, Greece
⁎ Corresponding author.
© 2016 Elsevier B.V., All rights reserved.