J Neurol (2014) 261:1735–1744 DOI 10.1007/s00415-014-7413-9
ORIGINAL COMMUNICATION
Longitudinal MRI and neuropsychological assessment of patients with clinically isolated syndrome Tomas Uher • Jana Blahova-Dusankova • Dana Horakova • Niels Bergsland • Michaela Tyblova • Ralph H. B. Benedict • Tomas Kalincik • Deepa P. Ramasamy • Zdenek Seidl • Jesper Hagermeier Manuela Vaneckova • Jan Krasensky • Eva Havrdova • Robert Zivadinov
•
Received: 14 February 2014 / Revised: 9 June 2014 / Accepted: 10 June 2014 / Published online: 22 June 2014 Ó Springer-Verlag Berlin Heidelberg 2014
Abstract Cognitive impairment (CI) may occur in clinically isolated syndrome (CIS) patients. While the relationship between CI and magnetic resonance imaging (MRI) has been investigated extensively in multiple sclerosis (MS), MRI correlates of CI in CIS patients are unknown. To investigate the evolution of CI and to determine brain MRI structural correlates associated with CI in CIS patients. This prospective 24-month observational study examined 81 CIS patients treated with 30 lg of intramuscular interferon beta 1a once a week. MRI acquisition and neuropsychological (NP) assessment were performed at baseline, 6, 12 and 24 months. Participants were tested with Czech-validated version of Minimal Assessment of Cognitive Function in MS battery and MRI measures of lesion activity and burden, and global, tissuespecific and regional brain atrophy were performed. Over 24 months, 36 CIS patients developed clinically definite
MS (CDMS). CI was observed in 10 (12.3 %) CIS patients at baseline and at the 24 months follow-up. Eight CIS patients changed their CI status over the follow-up (four improved and four worsened). No significant difference in development of CI was detected between stable CIS patients and those who developed CDMS. In multivariate regression and mixed-effect model analyses, no significant relationship was found between NP and MRI parameters. The lack of significant relationship between MRI metrics and cognition in this group of CIS patients could be attributed to several factors including the cognitive reserve, effect of disease-modifying therapy and relatively short follow-up period.
T. Uher J. Blahova-Dusankova D. Horakova M. Tyblova E. Havrdova Department of Neurology and Center of Clinical Neuroscience, First Faculty of Medicine and General University Hospital, Charles University in Prague, Prague, Czech Republic
T. Kalincik Department of Neurology, Royal Melbourne Hospital, Melbourne, Australia
T. Uher N. Bergsland R. H. B. Benedict D. P. Ramasamy J. Hagermeier R. Zivadinov (&) Department of Neurology, Buffalo Neuroimaging Analysis Center, The Jacobs Neurological Institute, School of Medicine and Biomedical Sciences, University at Buffalo, State University of New York, 100 High St, Buffalo, NY 14203, USA e-mail: [email protected]
Keywords Multiple sclerosis Clinically isolated syndrome MRI Cognition
Z. Seidl M. Vaneckova J. Krasensky Department of Radiology, First Faculty of Medicine and General University Hospital, Charles University, Prague, Czech Republic R. Zivadinov MR Imaging Clinical Translational Research Center, School of Medicine and Biomedical Sciences, University at Buffalo, State University of New York, Buffalo, NY, USA
N. Bergsland IRCCS, Don Gnocchi Foundation, Milan, Italy T. Kalincik Department of Medicine, Melbourne Brain Centre, University of Melbourne, Melbourne, Australia
123
1736
Introduction Multiple sclerosis (MS) is an autoimmune inflammatory disease of the central nervous system leading to physical deterioration and a wide range of neuropsychiatric signs and symptoms including cognitive impairment (CI) [1]. CI is an important determinant of employment status and associated societal costs [2, 3], and adversely affects social functioning, coping, quality of life and treatment adherence of MS patients [4, 5]. It is well known that CI occurs at all stages of the disease, including patients with clinically isolated syndrome (CIS) [6–8]. Prevalence rates of CI in CIS patients range from 11 to 57 % according to different studies [7, 8] and core domains of cognition, including verbal and visual memory, information processing speed, semantic fluency, sustained attention and executive functions are affected [6, 9]. However, evidence of cognitive deterioration during the first years of disease is sparse and based predominantly on the results of cross-sectional studies [6–8, 10–15]. It is known that neuropsychological (NP) deficits are related to brain structural magnetic resonance imaging (MRI) measures in MS patients [1]. Numerous studies in early MS showed an association between CI and white matter (WM) lesions [16], whole brain [17] and cortical [18, 19] and subcortical deep gray matter (GM) atrophy, including thalamus volume loss [20–22]. Recent evidence has indicated that CI in CIS patients may be associated with an increased risk to CDMS conversion [10]. While the relationship between CI and global, tissue-specific and regional brain atrophy in MS has been investigated extensively, evidence identifying MRI metrics related to development of CI in CIS patients is lacking. For example, in two studies, the pathological changes of the WM assessed by T2 lesion volume (T2-LV) was associated with the development of CI several years later [23, 24]. The thalamic atrophy development was associated with conversion to CDMS [25], as well as CI [20–22], but the role of thalamus atrophy progression in relation to development of CI in CIS patients is unknown [26]. The aims of this study were to investigate the evolution of cognitive function in CIS patients over a period of 24 months and to determine brain MRI structural correlates associated with CI in CIS patients.
Methods Study population The participants recruited in the observational study of early interferon beta-1a treatment in high-risk subjects after CIS (SET study) were 220 CIS patients with the following
123
J Neurol (2014) 261:1735–1744
characteristics [25, 27]: (1) 18–55 years of age, (2) enrolled within 4 months from the clinical event, (3) demonstrated Expanded Disability Status Scale (EDSS) B3.5, and (4) displayed the presence of C2 T2-hyperintense lesions on diagnostic MRI and presence of C2 oligoclonal bands in cerebrospinal fluid obtained at the screening visit prior to steroid treatment. The exclusion criteria were: (1) lack of clinical and MRI follow-up data post-baseline, and (2) occurrence of second relapse before the baseline visit and (3) pregnancy. SET was an investigator-initiated, multicenter, prospective observational clinical study (clin.gov # NCT01592474), where the study design and main results were previously published [25, 27–29]. All patients started treatment at baseline with 30 lg, once-weekly, intramuscular interferon beta 1a (AvonexÒ) and received 3–5 g of methylprednisolone for the first symptom and baseline MRI was performed C30 days after steroid administration. Clinical visits occurred every 3 months for 24 months and subsequent long-term follow-up in routine clinical practice. Clinical assessments were performed using the EDSS. Of the 220 CIS patients enrolled in the SET study, 81 consecutive CIS patients entered the NP substudy. The NP substudy initiated after the validation of Czech version of Minimal Assessment of Cognitive Function in MS (MACFIMS) battery was completed [9, 30]. The study was approved by Medical Ethics Committee at the General University Hospital and First Faculty of Medicine of Charles University in Prague and the University of Buffalo, and all participating centers approved the study protocol. Written informed consent was obtained from all patients at enrolment. Neuropsychological assessment All 81 participants were tested at baseline, 6, 12 and 24 months with the Czech-validated version [31] of MACFIMS battery [9, 30]. Of those, four CIS patients at month 6 and one patient at month 12 did not undergo NP assessment. The MACFIMS battery included the following tests [9]: Controlled Oral Word Association Test (COWAT) for measuring phonemic fluency; Judgment of Line Orientation Test (JLO) for measuring visual–spatial ability; California Verbal Learning Test, second edition (CVLT-II) for measuring verbal learning and memory, with total learning (CVLT-II TL) and delayed recall CVLT-II DR subtests; Brief Visuospatial Memory Test-Revised (BVMT-R) for measuring visuospatial learning and memory, with total learning (BVMT-R TL) and delayed recall BVMT-R DR subtests; Paced Auditory Serial Addition Test 3 (PASAT-3) for measuring rapid information processing, simultaneous allocation of attention to two tasks and reasonably intact calculation; Symbol Digit Modalities Test (SDMT) for measuring rapid information processing,
J Neurol (2014) 261:1735–1744
1737
visual scanning, and to a lesser extent, working memory; and the Sorting Test from the Delis–Kaplan executive function system (D-KEFS) for measuring conceptual reasoning that assesses higher executive functions, with Sorting Test (D-KEFS ST) and Description Score (DKEFS DS) subtests. At each time point, alternate versions of MACFIMS subtests were used to minimize potential practice effect. Impairment for a single test was assessed on the level of 1.5 SD (i.e., z score \1.5 compared with a healthy population), using the regression-based norms of 134 healthy controls adjusted for age, sex, and education (Table 1)
[31]. Patients were evaluated as CI when scoring outside the normal range in two or more of the MACFIMS tests [9, 30]. For the assessment of depressive symptoms, the Beck depression inventory (BDI-II) was used. Levels of anxiety symptoms were assessed using the beck anxiety inventory (BAI). MS neuropsychological screening questionnaire (MSNQ) [9] was used to self-report CI. MRI assessment MRI acquisition and analysis were performed at baseline, 6, 12 and 24 months with a standardized protocol on the
Table 1 Baseline demographic, clinical, and neuropsychological characteristics of clinically isolated syndrome patients and healthy controls, according to the cognitive status at enrollment MS patients cognitively impaired (n = 10)
MS patients cognitively intact (n = 71)
Healthy controls (n = 134)
p value
No. of femalesa
5 (50 %)
50 (70 %)
95 (71 %)
0.195
Age at onset in years
25.7 ± 7.3; 23.5
28.6 ± 8.0; 27
33.8 ± 8.8; 33
0.216
Education in years
13.2 ± 2.4; 12
14.3 ± 2.7; 14
14.4 ± 2.5; 14
0.218
Time to baseline in days
92.3 ± 29.1; 94
84.6 ± 22.4; 83
–
0.393
EDSSb
2.5; 3.5
2.0; 2.5
–
0.533
MSFC
0.4 ± 1.0; 0.4
0.2 ± 0.6; 0.3
–
0.044
Type of onset (n)a Optic neuritis
3 (30 %)
20 (28 %)
–
0.904
Sensory/motor
4 (40 %)
35 (49 %)
–
0.582
0 (0 %)
8 (11 %)
–
0.264
3 (30 %) 50.5 ± 6.8; 50.0
8 (11 %) 61.3 ± 8.9; 61.0
– 64.9 ± 8.7; 64.0
0.105 0.001
Brainstem/cerebellar Polysymptomatic SDMT BVMT-R TL
23.2 ± 4.1; 21.5
28.7 ± 4.1; 29.0
29.3 ± 3.7; 29.0
0.001
BVMT-R DR
9.4 ± 1.8; 10
11.0 ± 1.2; 11.0
11.0 ± 1.0; 11.0
0.002
CVLT-II TL
49.3 ± 12.9; 48.5
60.4 ± 6.7; 60.0
60.1 ± 8.1; 61.0
0.006
CVLT-II DR
11.1 ± 3.5; 12.0
14.0 ± 1.9; 14.0
13.8 ± 2.1; 14.0
0.006
COWAT
30.0 ± 11.0; 30.5
39.2 ± 9.5; 39
43.0 ± 8.8; 41.0
0.021
JLO
23.7 ± 6.2; 25.5
26.3 ± 3.3; 27.0
27.1 ± 2.9; 28.0
0.199
PASAT-3
39.4 ± 12.5; 36.0
52.4 ± 8.1; 55.0
50.6 ± 7.8; 52.0
0.002
D-KEFS ST CS
8.0 ± 0.8; 8.0
9.6 ± 1.6; 10.0
11.2 ± 1.7; 11.0
0.002
D-KEFS ST DS
29.5 ± 4.9; 30.5
36.6 ± 7.2; 36.0
42.8 ± 7.6; 42.0
0.002
BDI-II
9.5 ± 7.5; 8.0
6.2 ± 5.3; 6.0
3.5 ± 3.9; 3.0
0.205
BAI
13.8 ± 11.5; 10.5
8.4 ± 6.9; 7.0
5.3 ± 4.7; 4.0
0.146
MSNQ
20.6 ± 10.2; 22
13.1 ± 6.7; 7.0
12.0 ± 6.7; 12.0
0.014
Unless otherwise indicated, all data are reported as mean ± SD, median. Differences between the cognitively impaired and cognitively intact group were tested using the v2 test (categorical) and Mann–Whitney rank-sum test (other). p values \0.01 are presented in bold EDSS Expanded Disability Status Scale, MSFC multiple sclerosis functional composite, SDMT Symbol Digit Modalities Test, BVMT-R Brief Visuospatial Memory Test-Revised, TL total learning, DR delayed recall, CVLT-II California Verbal Learning Test, second edition, COWAT Controlled Oral Word Association Test, JLO Judgment of Line Orientation Test, PASAT-3 Paced Auditory Serial Addition Test, 3 s, D-KEFS ST Sorting Test from the Delis–Kaplan executive function system, D-KEFS DS Description Score from the Delis–Kaplan executive function system, CS correct sorts, BDI-II beck depression inventory II, BAI beck anxiety inventory, MSNQ multiple sclerosis neuropsychological screening questionnaire a
Data in parentheses are percentages
b
Data are reported as median and range
123
1738
same 1.5 T imager (Gyroscan; Philips Medical Systems, Best, The Netherlands). Axial brain acquisitions included fluid attenuated inversion recovery, three-dimensional T1weighted images, and post-contrast T1 spin-echo images before and 5 min after a single injection of 0.1 mmol/kg of gadopentetate dimeglumine. The details of the MRI sequences are provided elsewhere [25]. Image analyses included a cumulative number and volume of gadolinium contrast enhanced (CE) and new and enlarging T2 lesions, and analyses of changes in whole brain and tissue-specific and regional GM and WM volumes, as previously reported [25]. Longitudinal percent changes in volumes of whole brain, GM, WM, cortical and lateral ventricles were obtained using direct methods of atrophy measurement [32, 33]. Absolute and percent volume changes for the total subcortical deep GM (SDGM) (defined as the sum of thalamus, caudate nucleus, putamen, globus pallidus, hippocampus, amygdala, and nucleus accumbens), thalamus and hippocampus at each time point were estimated, as previously reported [25]. Statistical analysis Statistical analyses were carried out with Statistica 10 (Statsoft, Tulsa, OK, USA), R version 3.0.1 (http://www.Rproject.org) and SPSS software version 16.0 (SPSS, Chicago, IL, USA). Parametric and non-parametric tests were used according to the distribution of the variables. Normality of the distribution was assessed using the Kolmogorov–Smirnov method. Non-normally distributed variables assessing T2 and CE cumulative lesion numbers and their volumes (LVs) were transformed using Cox-box transformations. Patients were divided into two groups based on their progression to CDMS, McDonald 2005 [34] and 2010 [35] criteria within the 24-month time period of the study. Patients were also classified as having CI or not having CI (NCI) at disease onset according to MACFIMS battery outcome. Group differences were examined with Student’s t test, Mann–Whitney U test, or v2 test. Multivariate regression model analyses, adjusted for age, sex, education, depression and change in treatment status over the 24 months of follow-up, were used to test which MRI outcomes were longitudinally associated with absolute changes of MACFIMS subtest outcomes. Longitudinal linear mixed-effect models with random term for intercept for each patient, adjusted for age, sex, education, depression and change in treatment status over the 24 months of follow-up were used to describe temporal associations between MRI measures and changes in the tests of the MACFIMS battery. To avoid spurious findings due to multiple comparisons, we considered the nominal p values B0.05 as a trend and as the nominal p value B0.01 as statistically significant.
123
J Neurol (2014) 261:1735–1744
Results Baseline demographic, clinical, neuropsychological and MRI characteristics Table 1 shows demographic, clinical, and NP characteristics of the CIS patients, separated by cognitive status at time of enrollment. At baseline, 10 (12.3 %) of 81 CIS patients showed CI. Patients with CI had worse cognitive performance in SDMT (p = 0.001), BVMT-R TL (p = 0.001), CVLT-II TL (p = 0.006), BVMT-R DL (p = 0.002), CVLT-II DL (p = 0.006), COWAT (p = 0.021), PASAT-3 (p = 0.002), and D-KEFS ST DS (p = 0.002) compared with NCI patients. Patients with CI also had a trend for higher MSNQ score (p = 0.014) at baseline. No significant differences in demographic or clinical characteristics were found between the CI and NCI group at baseline (Table 1). No significant differences in MRI lesion and brain volumetric measures were found between the CI and NCI group at baseline (Table 2). No significant differences in demographic, clinical or MRI outcomes between 81 CIS patients enrolled and 136 not enrolled in the NP SET substudy were found. However, there was significantly lower number of T2 lesions (p B 0.001) and T2-LVs (p = 0.002) at baseline in the 81 participating in the NP substudy. Follow-up clinical and treatment characteristics During the 24 months, 36 of 81 patients (44.4 %) experienced relapses and developed CDMS. The median time from baseline to first relapse was 6.9 months (1st and 3rd quartiles, 3.8–15.3 months). The mean annual relapse rate was 0.51 ± 0.72 (SD). At the 24 months follow-up, the median of EDSS score was 2.0 (range 6.0) in the CDMS group and 1.5 (range 3.0) in the stable CIS group (p = 0.0003). At 2 years, 42 (51.9 %) CIS patients fulfilled the McDonald 2005 criteria and 57 (70.4 %) CIS patients fulfilled the McDonald 2010 criteria. Of the 81 patients who started Avonex, 70 (86.4 %) patients remained on the assigned treatment during the 24 months follow-up. The treatment changes are reported in Table 3. Changes in cognitive performance according to development of CDMS status over 24 months Table 3 shows NP characteristics of the CIS patients separated by conversion status to CDMS at the 24-month follow-up. Although CI was observed in 10 (12.3 %) CIS patients at the 24-month follow-up, 8 CIS patients changed their CI status over the follow-up (4 improved and 4 worsened). Three of 45 (7 %) stable CIS patients compared to one of
J Neurol (2014) 261:1735–1744
1739
Table 2 Baseline MRI characteristics of clinically isolated syndrome patients according to the cognitive status at enrollment
No. of CE lesions No. of CE positivity
a
No. of T2 lesions
MS patients cognitively impaired (n = 10)
MS patients cognitively intact (n = 71)
p value
1.1 ± 2.0; 0.0
0.6 ± 1.6; 0.0
0.466
3 (30 %)
16 (23 %)
0.836
6.1 ± 4.3; 5.5
9.8 ± 6.1; 8.0
0.150
No. of patients with C9 T2 lesionsa
2 (20 %)
32 (46 %)
0.133
CE lesion volume
0.08 ± 0.2; 0.0
0.08 ± 0.4; 0.0
0.545
T2 lesion volume
2.7 ± 3.5; 1.2
3.8 ± 4.0; 2.2
0.053
Normalized WB volume Normalized GM volume
1,507.4 ± 93.9; 1,485.5 790.9 ± 496.5; 775.3
1,508.4 ± 70.2; 1,518.6 795.7 ± 429.1; 807.2
0.940 0.587
Normalized WM volume
716.5 ± 57.3; 698.0
712.6 ± 38.8; 710.5
0.694
Normalized cortical volume
617.9 ± 44.1; 606.1
623.4 ± 35.0; 633.3
0.512
Normalized lateral ventricle volume
34.5 ± 7.3; 33.7
35.4 ± 10.1; 33.0
0.613
Total normalized SDGM volume
60.5 ± 42.4; 59.0
61.0 ± 38.0; 61.4
0.631
Normalized thalamus volume
20.8 ± 17.9; 20.5
20.9 ± 1.5; 21.0
0.251
Normalized hippocampus volume
9.1 ± 12.1; 9.1
9.3 ± 11.1; 9.3
0.899
Unless otherwise indicated, all data are reported as mean ± SD, median No number, CE contrast enhancing, WB whole brain, GM gray matter, WM white matter, SDGM subcortical deep gray matter a
Data in parentheses are percentages. Differences between the groups were tested using the v2 test or Mann–Whitney rank-sum test. All volumes are reported in milliliters
36 (3 %) CDMS patients improved MACFIMS outcome and one of 10 (2 %) stable CIS patients compared to 3 of 36 (8 %) CDMS patients had worsened MACFIMS outcome at the 24-month follow-up. In addition, 41 of 45 (91 %) stable CIS patients compared to 32 of 36 (89 %) CDMS patients showed stable MACFIMS outcome. CI and NCI groups did not differ significantly in terms of conversion to CDMS during the 2-year period (p = 0.706). No differences in CI status were found according to fulfillment of McDonald 2005 and 2010 criteria.
(p = 0.035) decrease of thalamus volume with a decrease of D-KEFS DS absolute change over the follow-up (Fig. 1a). There was also a trend for the associations between percent decrease of SDGM volume and decrease of COWAT absolute change (p = 0.044) (Fig. 1b). No significant associations between cumulative number of total new and enlarging T2 lesions and cumulative number of new CE lesions and cognitive outcomes were found.
Changes in MRI outcomes according to cognitive impairment and development of CDMS at 24 months
The longitudinal mixed-effect model analysis did not show any significant differences over time between the CI and NCI groups. The models with absolute changes in NP test as the outcome measures and the main predictor of interest percent changes of volumetric MRI parameters were tested. The analyses showed trends of the association between percent change of GM volume and absolute change of BVMT-R DR (p = 0.016) and between percent change of cortical volume and absolute change of BVMT-R DR (p = 0.019).
No significant differences were found in MRI outcomes between CIS patients who were CI at baseline vs. those who were not. No significant differences in MRI outcomes were detected between CI and NCI group at 24 months (Table 4). Significant difference in absolute change of T2-LV (p = 0.0003), percent brain volume change (PBVC) (p = 0.002), GM volume percent change (p = 0.009), cortical volume percent change (p = 0.009) and lateral ventricle volume percent change (p = 0.0001) was detected between CDMS and stable CIS group over 24 months. Association between cognitive changes and MRI outcomes over 24 months In multivariate regression analysis, a trend was shown for the association of absolute (p = 0.027) and percent
Mixed-effect model analysis over the 24 months
Discussion This study is the first prospective, longitudinal clinical investigation of the association between the global, tissuespecific and regional MRI outcomes and cognition in CIS patients on disease-modifying treatment. The results of the study indicate that CI is present in a detectable proportion
123
1740
J Neurol (2014) 261:1735–1744
Table 3 Absolute changes of clinical and neuropsychological characteristics of clinically isolated syndrome patients, according to clinically definite multiple sclerosis status at 24 months Stable CIS (n = 45)
CDMS (n = 36)
p value
EDSS
-0.3 ± 0.7; 0.0
0.5 ± 1.1; 0.5
0.0003
MSFC
0.2 ± 0.3; 0.2
0.0 ± 0.4; 0.0
0.045
Therapy status changea n (%) SDMT
0 (0 %) 3.2 ± 6.8; 3.0
14 (39 %) 1.6 ± 5.2; 1.5
ORIGINAL COMMUNICATION
Longitudinal MRI and neuropsychological assessment of patients with clinically isolated syndrome Tomas Uher • Jana Blahova-Dusankova • Dana Horakova • Niels Bergsland • Michaela Tyblova • Ralph H. B. Benedict • Tomas Kalincik • Deepa P. Ramasamy • Zdenek Seidl • Jesper Hagermeier Manuela Vaneckova • Jan Krasensky • Eva Havrdova • Robert Zivadinov
•
Received: 14 February 2014 / Revised: 9 June 2014 / Accepted: 10 June 2014 / Published online: 22 June 2014 Ó Springer-Verlag Berlin Heidelberg 2014
Abstract Cognitive impairment (CI) may occur in clinically isolated syndrome (CIS) patients. While the relationship between CI and magnetic resonance imaging (MRI) has been investigated extensively in multiple sclerosis (MS), MRI correlates of CI in CIS patients are unknown. To investigate the evolution of CI and to determine brain MRI structural correlates associated with CI in CIS patients. This prospective 24-month observational study examined 81 CIS patients treated with 30 lg of intramuscular interferon beta 1a once a week. MRI acquisition and neuropsychological (NP) assessment were performed at baseline, 6, 12 and 24 months. Participants were tested with Czech-validated version of Minimal Assessment of Cognitive Function in MS battery and MRI measures of lesion activity and burden, and global, tissuespecific and regional brain atrophy were performed. Over 24 months, 36 CIS patients developed clinically definite
MS (CDMS). CI was observed in 10 (12.3 %) CIS patients at baseline and at the 24 months follow-up. Eight CIS patients changed their CI status over the follow-up (four improved and four worsened). No significant difference in development of CI was detected between stable CIS patients and those who developed CDMS. In multivariate regression and mixed-effect model analyses, no significant relationship was found between NP and MRI parameters. The lack of significant relationship between MRI metrics and cognition in this group of CIS patients could be attributed to several factors including the cognitive reserve, effect of disease-modifying therapy and relatively short follow-up period.
T. Uher J. Blahova-Dusankova D. Horakova M. Tyblova E. Havrdova Department of Neurology and Center of Clinical Neuroscience, First Faculty of Medicine and General University Hospital, Charles University in Prague, Prague, Czech Republic
T. Kalincik Department of Neurology, Royal Melbourne Hospital, Melbourne, Australia
T. Uher N. Bergsland R. H. B. Benedict D. P. Ramasamy J. Hagermeier R. Zivadinov (&) Department of Neurology, Buffalo Neuroimaging Analysis Center, The Jacobs Neurological Institute, School of Medicine and Biomedical Sciences, University at Buffalo, State University of New York, 100 High St, Buffalo, NY 14203, USA e-mail: [email protected]
Keywords Multiple sclerosis Clinically isolated syndrome MRI Cognition
Z. Seidl M. Vaneckova J. Krasensky Department of Radiology, First Faculty of Medicine and General University Hospital, Charles University, Prague, Czech Republic R. Zivadinov MR Imaging Clinical Translational Research Center, School of Medicine and Biomedical Sciences, University at Buffalo, State University of New York, Buffalo, NY, USA
N. Bergsland IRCCS, Don Gnocchi Foundation, Milan, Italy T. Kalincik Department of Medicine, Melbourne Brain Centre, University of Melbourne, Melbourne, Australia
123
1736
Introduction Multiple sclerosis (MS) is an autoimmune inflammatory disease of the central nervous system leading to physical deterioration and a wide range of neuropsychiatric signs and symptoms including cognitive impairment (CI) [1]. CI is an important determinant of employment status and associated societal costs [2, 3], and adversely affects social functioning, coping, quality of life and treatment adherence of MS patients [4, 5]. It is well known that CI occurs at all stages of the disease, including patients with clinically isolated syndrome (CIS) [6–8]. Prevalence rates of CI in CIS patients range from 11 to 57 % according to different studies [7, 8] and core domains of cognition, including verbal and visual memory, information processing speed, semantic fluency, sustained attention and executive functions are affected [6, 9]. However, evidence of cognitive deterioration during the first years of disease is sparse and based predominantly on the results of cross-sectional studies [6–8, 10–15]. It is known that neuropsychological (NP) deficits are related to brain structural magnetic resonance imaging (MRI) measures in MS patients [1]. Numerous studies in early MS showed an association between CI and white matter (WM) lesions [16], whole brain [17] and cortical [18, 19] and subcortical deep gray matter (GM) atrophy, including thalamus volume loss [20–22]. Recent evidence has indicated that CI in CIS patients may be associated with an increased risk to CDMS conversion [10]. While the relationship between CI and global, tissue-specific and regional brain atrophy in MS has been investigated extensively, evidence identifying MRI metrics related to development of CI in CIS patients is lacking. For example, in two studies, the pathological changes of the WM assessed by T2 lesion volume (T2-LV) was associated with the development of CI several years later [23, 24]. The thalamic atrophy development was associated with conversion to CDMS [25], as well as CI [20–22], but the role of thalamus atrophy progression in relation to development of CI in CIS patients is unknown [26]. The aims of this study were to investigate the evolution of cognitive function in CIS patients over a period of 24 months and to determine brain MRI structural correlates associated with CI in CIS patients.
Methods Study population The participants recruited in the observational study of early interferon beta-1a treatment in high-risk subjects after CIS (SET study) were 220 CIS patients with the following
123
J Neurol (2014) 261:1735–1744
characteristics [25, 27]: (1) 18–55 years of age, (2) enrolled within 4 months from the clinical event, (3) demonstrated Expanded Disability Status Scale (EDSS) B3.5, and (4) displayed the presence of C2 T2-hyperintense lesions on diagnostic MRI and presence of C2 oligoclonal bands in cerebrospinal fluid obtained at the screening visit prior to steroid treatment. The exclusion criteria were: (1) lack of clinical and MRI follow-up data post-baseline, and (2) occurrence of second relapse before the baseline visit and (3) pregnancy. SET was an investigator-initiated, multicenter, prospective observational clinical study (clin.gov # NCT01592474), where the study design and main results were previously published [25, 27–29]. All patients started treatment at baseline with 30 lg, once-weekly, intramuscular interferon beta 1a (AvonexÒ) and received 3–5 g of methylprednisolone for the first symptom and baseline MRI was performed C30 days after steroid administration. Clinical visits occurred every 3 months for 24 months and subsequent long-term follow-up in routine clinical practice. Clinical assessments were performed using the EDSS. Of the 220 CIS patients enrolled in the SET study, 81 consecutive CIS patients entered the NP substudy. The NP substudy initiated after the validation of Czech version of Minimal Assessment of Cognitive Function in MS (MACFIMS) battery was completed [9, 30]. The study was approved by Medical Ethics Committee at the General University Hospital and First Faculty of Medicine of Charles University in Prague and the University of Buffalo, and all participating centers approved the study protocol. Written informed consent was obtained from all patients at enrolment. Neuropsychological assessment All 81 participants were tested at baseline, 6, 12 and 24 months with the Czech-validated version [31] of MACFIMS battery [9, 30]. Of those, four CIS patients at month 6 and one patient at month 12 did not undergo NP assessment. The MACFIMS battery included the following tests [9]: Controlled Oral Word Association Test (COWAT) for measuring phonemic fluency; Judgment of Line Orientation Test (JLO) for measuring visual–spatial ability; California Verbal Learning Test, second edition (CVLT-II) for measuring verbal learning and memory, with total learning (CVLT-II TL) and delayed recall CVLT-II DR subtests; Brief Visuospatial Memory Test-Revised (BVMT-R) for measuring visuospatial learning and memory, with total learning (BVMT-R TL) and delayed recall BVMT-R DR subtests; Paced Auditory Serial Addition Test 3 (PASAT-3) for measuring rapid information processing, simultaneous allocation of attention to two tasks and reasonably intact calculation; Symbol Digit Modalities Test (SDMT) for measuring rapid information processing,
J Neurol (2014) 261:1735–1744
1737
visual scanning, and to a lesser extent, working memory; and the Sorting Test from the Delis–Kaplan executive function system (D-KEFS) for measuring conceptual reasoning that assesses higher executive functions, with Sorting Test (D-KEFS ST) and Description Score (DKEFS DS) subtests. At each time point, alternate versions of MACFIMS subtests were used to minimize potential practice effect. Impairment for a single test was assessed on the level of 1.5 SD (i.e., z score \1.5 compared with a healthy population), using the regression-based norms of 134 healthy controls adjusted for age, sex, and education (Table 1)
[31]. Patients were evaluated as CI when scoring outside the normal range in two or more of the MACFIMS tests [9, 30]. For the assessment of depressive symptoms, the Beck depression inventory (BDI-II) was used. Levels of anxiety symptoms were assessed using the beck anxiety inventory (BAI). MS neuropsychological screening questionnaire (MSNQ) [9] was used to self-report CI. MRI assessment MRI acquisition and analysis were performed at baseline, 6, 12 and 24 months with a standardized protocol on the
Table 1 Baseline demographic, clinical, and neuropsychological characteristics of clinically isolated syndrome patients and healthy controls, according to the cognitive status at enrollment MS patients cognitively impaired (n = 10)
MS patients cognitively intact (n = 71)
Healthy controls (n = 134)
p value
No. of femalesa
5 (50 %)
50 (70 %)
95 (71 %)
0.195
Age at onset in years
25.7 ± 7.3; 23.5
28.6 ± 8.0; 27
33.8 ± 8.8; 33
0.216
Education in years
13.2 ± 2.4; 12
14.3 ± 2.7; 14
14.4 ± 2.5; 14
0.218
Time to baseline in days
92.3 ± 29.1; 94
84.6 ± 22.4; 83
–
0.393
EDSSb
2.5; 3.5
2.0; 2.5
–
0.533
MSFC
0.4 ± 1.0; 0.4
0.2 ± 0.6; 0.3
–
0.044
Type of onset (n)a Optic neuritis
3 (30 %)
20 (28 %)
–
0.904
Sensory/motor
4 (40 %)
35 (49 %)
–
0.582
0 (0 %)
8 (11 %)
–
0.264
3 (30 %) 50.5 ± 6.8; 50.0
8 (11 %) 61.3 ± 8.9; 61.0
– 64.9 ± 8.7; 64.0
0.105 0.001
Brainstem/cerebellar Polysymptomatic SDMT BVMT-R TL
23.2 ± 4.1; 21.5
28.7 ± 4.1; 29.0
29.3 ± 3.7; 29.0
0.001
BVMT-R DR
9.4 ± 1.8; 10
11.0 ± 1.2; 11.0
11.0 ± 1.0; 11.0
0.002
CVLT-II TL
49.3 ± 12.9; 48.5
60.4 ± 6.7; 60.0
60.1 ± 8.1; 61.0
0.006
CVLT-II DR
11.1 ± 3.5; 12.0
14.0 ± 1.9; 14.0
13.8 ± 2.1; 14.0
0.006
COWAT
30.0 ± 11.0; 30.5
39.2 ± 9.5; 39
43.0 ± 8.8; 41.0
0.021
JLO
23.7 ± 6.2; 25.5
26.3 ± 3.3; 27.0
27.1 ± 2.9; 28.0
0.199
PASAT-3
39.4 ± 12.5; 36.0
52.4 ± 8.1; 55.0
50.6 ± 7.8; 52.0
0.002
D-KEFS ST CS
8.0 ± 0.8; 8.0
9.6 ± 1.6; 10.0
11.2 ± 1.7; 11.0
0.002
D-KEFS ST DS
29.5 ± 4.9; 30.5
36.6 ± 7.2; 36.0
42.8 ± 7.6; 42.0
0.002
BDI-II
9.5 ± 7.5; 8.0
6.2 ± 5.3; 6.0
3.5 ± 3.9; 3.0
0.205
BAI
13.8 ± 11.5; 10.5
8.4 ± 6.9; 7.0
5.3 ± 4.7; 4.0
0.146
MSNQ
20.6 ± 10.2; 22
13.1 ± 6.7; 7.0
12.0 ± 6.7; 12.0
0.014
Unless otherwise indicated, all data are reported as mean ± SD, median. Differences between the cognitively impaired and cognitively intact group were tested using the v2 test (categorical) and Mann–Whitney rank-sum test (other). p values \0.01 are presented in bold EDSS Expanded Disability Status Scale, MSFC multiple sclerosis functional composite, SDMT Symbol Digit Modalities Test, BVMT-R Brief Visuospatial Memory Test-Revised, TL total learning, DR delayed recall, CVLT-II California Verbal Learning Test, second edition, COWAT Controlled Oral Word Association Test, JLO Judgment of Line Orientation Test, PASAT-3 Paced Auditory Serial Addition Test, 3 s, D-KEFS ST Sorting Test from the Delis–Kaplan executive function system, D-KEFS DS Description Score from the Delis–Kaplan executive function system, CS correct sorts, BDI-II beck depression inventory II, BAI beck anxiety inventory, MSNQ multiple sclerosis neuropsychological screening questionnaire a
Data in parentheses are percentages
b
Data are reported as median and range
123
1738
same 1.5 T imager (Gyroscan; Philips Medical Systems, Best, The Netherlands). Axial brain acquisitions included fluid attenuated inversion recovery, three-dimensional T1weighted images, and post-contrast T1 spin-echo images before and 5 min after a single injection of 0.1 mmol/kg of gadopentetate dimeglumine. The details of the MRI sequences are provided elsewhere [25]. Image analyses included a cumulative number and volume of gadolinium contrast enhanced (CE) and new and enlarging T2 lesions, and analyses of changes in whole brain and tissue-specific and regional GM and WM volumes, as previously reported [25]. Longitudinal percent changes in volumes of whole brain, GM, WM, cortical and lateral ventricles were obtained using direct methods of atrophy measurement [32, 33]. Absolute and percent volume changes for the total subcortical deep GM (SDGM) (defined as the sum of thalamus, caudate nucleus, putamen, globus pallidus, hippocampus, amygdala, and nucleus accumbens), thalamus and hippocampus at each time point were estimated, as previously reported [25]. Statistical analysis Statistical analyses were carried out with Statistica 10 (Statsoft, Tulsa, OK, USA), R version 3.0.1 (http://www.Rproject.org) and SPSS software version 16.0 (SPSS, Chicago, IL, USA). Parametric and non-parametric tests were used according to the distribution of the variables. Normality of the distribution was assessed using the Kolmogorov–Smirnov method. Non-normally distributed variables assessing T2 and CE cumulative lesion numbers and their volumes (LVs) were transformed using Cox-box transformations. Patients were divided into two groups based on their progression to CDMS, McDonald 2005 [34] and 2010 [35] criteria within the 24-month time period of the study. Patients were also classified as having CI or not having CI (NCI) at disease onset according to MACFIMS battery outcome. Group differences were examined with Student’s t test, Mann–Whitney U test, or v2 test. Multivariate regression model analyses, adjusted for age, sex, education, depression and change in treatment status over the 24 months of follow-up, were used to test which MRI outcomes were longitudinally associated with absolute changes of MACFIMS subtest outcomes. Longitudinal linear mixed-effect models with random term for intercept for each patient, adjusted for age, sex, education, depression and change in treatment status over the 24 months of follow-up were used to describe temporal associations between MRI measures and changes in the tests of the MACFIMS battery. To avoid spurious findings due to multiple comparisons, we considered the nominal p values B0.05 as a trend and as the nominal p value B0.01 as statistically significant.
123
J Neurol (2014) 261:1735–1744
Results Baseline demographic, clinical, neuropsychological and MRI characteristics Table 1 shows demographic, clinical, and NP characteristics of the CIS patients, separated by cognitive status at time of enrollment. At baseline, 10 (12.3 %) of 81 CIS patients showed CI. Patients with CI had worse cognitive performance in SDMT (p = 0.001), BVMT-R TL (p = 0.001), CVLT-II TL (p = 0.006), BVMT-R DL (p = 0.002), CVLT-II DL (p = 0.006), COWAT (p = 0.021), PASAT-3 (p = 0.002), and D-KEFS ST DS (p = 0.002) compared with NCI patients. Patients with CI also had a trend for higher MSNQ score (p = 0.014) at baseline. No significant differences in demographic or clinical characteristics were found between the CI and NCI group at baseline (Table 1). No significant differences in MRI lesion and brain volumetric measures were found between the CI and NCI group at baseline (Table 2). No significant differences in demographic, clinical or MRI outcomes between 81 CIS patients enrolled and 136 not enrolled in the NP SET substudy were found. However, there was significantly lower number of T2 lesions (p B 0.001) and T2-LVs (p = 0.002) at baseline in the 81 participating in the NP substudy. Follow-up clinical and treatment characteristics During the 24 months, 36 of 81 patients (44.4 %) experienced relapses and developed CDMS. The median time from baseline to first relapse was 6.9 months (1st and 3rd quartiles, 3.8–15.3 months). The mean annual relapse rate was 0.51 ± 0.72 (SD). At the 24 months follow-up, the median of EDSS score was 2.0 (range 6.0) in the CDMS group and 1.5 (range 3.0) in the stable CIS group (p = 0.0003). At 2 years, 42 (51.9 %) CIS patients fulfilled the McDonald 2005 criteria and 57 (70.4 %) CIS patients fulfilled the McDonald 2010 criteria. Of the 81 patients who started Avonex, 70 (86.4 %) patients remained on the assigned treatment during the 24 months follow-up. The treatment changes are reported in Table 3. Changes in cognitive performance according to development of CDMS status over 24 months Table 3 shows NP characteristics of the CIS patients separated by conversion status to CDMS at the 24-month follow-up. Although CI was observed in 10 (12.3 %) CIS patients at the 24-month follow-up, 8 CIS patients changed their CI status over the follow-up (4 improved and 4 worsened). Three of 45 (7 %) stable CIS patients compared to one of
J Neurol (2014) 261:1735–1744
1739
Table 2 Baseline MRI characteristics of clinically isolated syndrome patients according to the cognitive status at enrollment
No. of CE lesions No. of CE positivity
a
No. of T2 lesions
MS patients cognitively impaired (n = 10)
MS patients cognitively intact (n = 71)
p value
1.1 ± 2.0; 0.0
0.6 ± 1.6; 0.0
0.466
3 (30 %)
16 (23 %)
0.836
6.1 ± 4.3; 5.5
9.8 ± 6.1; 8.0
0.150
No. of patients with C9 T2 lesionsa
2 (20 %)
32 (46 %)
0.133
CE lesion volume
0.08 ± 0.2; 0.0
0.08 ± 0.4; 0.0
0.545
T2 lesion volume
2.7 ± 3.5; 1.2
3.8 ± 4.0; 2.2
0.053
Normalized WB volume Normalized GM volume
1,507.4 ± 93.9; 1,485.5 790.9 ± 496.5; 775.3
1,508.4 ± 70.2; 1,518.6 795.7 ± 429.1; 807.2
0.940 0.587
Normalized WM volume
716.5 ± 57.3; 698.0
712.6 ± 38.8; 710.5
0.694
Normalized cortical volume
617.9 ± 44.1; 606.1
623.4 ± 35.0; 633.3
0.512
Normalized lateral ventricle volume
34.5 ± 7.3; 33.7
35.4 ± 10.1; 33.0
0.613
Total normalized SDGM volume
60.5 ± 42.4; 59.0
61.0 ± 38.0; 61.4
0.631
Normalized thalamus volume
20.8 ± 17.9; 20.5
20.9 ± 1.5; 21.0
0.251
Normalized hippocampus volume
9.1 ± 12.1; 9.1
9.3 ± 11.1; 9.3
0.899
Unless otherwise indicated, all data are reported as mean ± SD, median No number, CE contrast enhancing, WB whole brain, GM gray matter, WM white matter, SDGM subcortical deep gray matter a
Data in parentheses are percentages. Differences between the groups were tested using the v2 test or Mann–Whitney rank-sum test. All volumes are reported in milliliters
36 (3 %) CDMS patients improved MACFIMS outcome and one of 10 (2 %) stable CIS patients compared to 3 of 36 (8 %) CDMS patients had worsened MACFIMS outcome at the 24-month follow-up. In addition, 41 of 45 (91 %) stable CIS patients compared to 32 of 36 (89 %) CDMS patients showed stable MACFIMS outcome. CI and NCI groups did not differ significantly in terms of conversion to CDMS during the 2-year period (p = 0.706). No differences in CI status were found according to fulfillment of McDonald 2005 and 2010 criteria.
(p = 0.035) decrease of thalamus volume with a decrease of D-KEFS DS absolute change over the follow-up (Fig. 1a). There was also a trend for the associations between percent decrease of SDGM volume and decrease of COWAT absolute change (p = 0.044) (Fig. 1b). No significant associations between cumulative number of total new and enlarging T2 lesions and cumulative number of new CE lesions and cognitive outcomes were found.
Changes in MRI outcomes according to cognitive impairment and development of CDMS at 24 months
The longitudinal mixed-effect model analysis did not show any significant differences over time between the CI and NCI groups. The models with absolute changes in NP test as the outcome measures and the main predictor of interest percent changes of volumetric MRI parameters were tested. The analyses showed trends of the association between percent change of GM volume and absolute change of BVMT-R DR (p = 0.016) and between percent change of cortical volume and absolute change of BVMT-R DR (p = 0.019).
No significant differences were found in MRI outcomes between CIS patients who were CI at baseline vs. those who were not. No significant differences in MRI outcomes were detected between CI and NCI group at 24 months (Table 4). Significant difference in absolute change of T2-LV (p = 0.0003), percent brain volume change (PBVC) (p = 0.002), GM volume percent change (p = 0.009), cortical volume percent change (p = 0.009) and lateral ventricle volume percent change (p = 0.0001) was detected between CDMS and stable CIS group over 24 months. Association between cognitive changes and MRI outcomes over 24 months In multivariate regression analysis, a trend was shown for the association of absolute (p = 0.027) and percent
Mixed-effect model analysis over the 24 months
Discussion This study is the first prospective, longitudinal clinical investigation of the association between the global, tissuespecific and regional MRI outcomes and cognition in CIS patients on disease-modifying treatment. The results of the study indicate that CI is present in a detectable proportion
123
1740
J Neurol (2014) 261:1735–1744
Table 3 Absolute changes of clinical and neuropsychological characteristics of clinically isolated syndrome patients, according to clinically definite multiple sclerosis status at 24 months Stable CIS (n = 45)
CDMS (n = 36)
p value
EDSS
-0.3 ± 0.7; 0.0
0.5 ± 1.1; 0.5
0.0003
MSFC
0.2 ± 0.3; 0.2
0.0 ± 0.4; 0.0
0.045
Therapy status changea n (%) SDMT
0 (0 %) 3.2 ± 6.8; 3.0
14 (39 %) 1.6 ± 5.2; 1.5
