Review
Clinical and imaging assessment of cognitive dysfunction in multiple sclerosis Maria A Rocca, Maria P Amato, Nicola De Stefano, Christian Enzinger, Jeroen J Geurts, Iris-K Penner, Alex Rovira, James F Sumowski, Paola Valsasina, Massimo Filippi, for the MAGNIMS Study Group Lancet Neurol 2015; 14: 302–17 Published Online February 4, 2015 http://dx.doi.org/10.1016/ S1474-4422(14)70250-9 Neuroimaging Research Unit (M A Rocca MD, P Valsasina MsC, Prof M Filippi MD) and Department of Neurology (M A Rocca, Prof M Filippi), Institute of Experimental Neurology, Division of Neuroscience, San Raffaele Scientific Institute, Vita-Salute San Raffaele University, Milan, Italy; Department of Neurofarba, Section of Neurosciences, University of Florence, Florence, Italy (M P Amato MD); Department of Medicine, Surgery and Neuroscience, University of Siena, Siena, Italy (N De Stefano MD); Department of Neurology, Medical University of Graz, Graz, Austria (C Enzinger MD); Department of Anatomy and Neuroscience, Section of Clinical Neuroscience, VU University Medical Centre, VUmc Multiple Sclerosis Centre Amsterdam, Amsterdam, Netherlands (J J Geurts PhD); University and University Children’s Hospital Basel, Cognitive Psychology and Methodology and Division of Paediatric Neurology and Developmental Medicine, Basel, Switzerland (I-K Penner MD); Magnetic Resonance Unit, Department of Radiology, Hospital Universitari Vall d’Hebron, Barcelona, Spain (A Rovira MD); and Neuropsychology and Neuroscience, Kessler Foundation Research Center, West Orange, NJ, USA (J F Sumowski PhD) Correspondence to: Prof Massimo Filippi, Neuroimaging Research Unit, Institute of Experimental Neurology, Division of Neuroscience, San Raffaele Scientific Institute, Vita-Salute San Raffaele University, Via Olgettina, 60, 20132, Milan, Italy fi[email protected]
302
In patients with multiple sclerosis (MS), grey matter damage is widespread and might underlie many of the clinical symptoms, especially cognitive impairment. This relation between grey matter damage and cognitive impairment has been lent support by findings from clinical and MRI studies. However, many aspects of cognitive impairment in patients with MS still need to be characterised. Standardised neuropsychological tests that are easy to administer and sensitive to disease-related abnormalities are needed to gain a better understanding of the factors affecting cognitive performance in patients with MS than exists at present. Imaging measures of the grey matter are necessary, but not sufficient to fully characterise cognitive decline in MS. Imaging measures of both lesioned and normal-appearing white matter lend support to the hypothesis of the existence of an underlying disconnection syndrome that causes clinical symptoms to trigger. Findings on cortical reorganisation support the contribution of brain plasticity and cognitive reserve in limiting cognitive deficits. The development of clinical and imaging biomarkers that can monitor disease development and treatment response is crucial to allow early identification of patients with MS who are at risk of cognitive impairment.
Introduction Multiple sclerosis (MS) has traditionally, and simplistically, been defined as a demyelinating-inflammatory white matter disorder, however, grey matter involvement is a feature of MS in most patients. The first description of cortical (including patchy demyelinating lesions) and deep grey matter lesions was presented by Charcot in 1868,1 but grey matter involvement has only been investigated in pathological and imaging studies in the past three decades.2,3 Charcot is also merited with having linked grey matter involvement to some of the clinical manifestations of MS, including cognitive impairment and convulsions. During the past decade, much effort has been put into the characterisation of the type, extent, and topography of grey matter damage in patients with MS at different stages of the disease, and to develop novel MRI sequences that can measure the grey matter involvement and track disease evolution. Some of the main drivers of this research are the need to better explain the heterogeneous and complex clinical manifestations of the disease, particularly cognitive impairment, and to develop novel therapeutic interventions. Through these studies, many advances have been made in our understanding of grey matter involvement in the disease. A few comprehensive review articles have summarised the main findings and the underlying pathological abnormalities.2–6 However, the focus on grey matter involvement in the past decade has led researchers, on some occasions, to count MS among the neurodegenerative disorders. Thus, at this stage, clarification of the distinctive features of grey matter involvement in MS is awaited. Furthermore, differentiation of how grey and white matter involvement contribute to the range of clinical manifestations of MS is also needed. These clarifications are crucial for the development of clinical and imaging biomarkers that will enable clinicians to monitor disease
evolution, predict future cognitive impairment, and define treatment response. The aim of this Review is to critically discuss the clinical scales and MRI metrics available to assess cognitive dysfunction in MS. A summary of the main advancements provided by each of these measures and practical recommendations for their use are also provided. Finally, we discuss unmet needs that should guide future research.
Assessment of cognitive dysfunction in MS Cognitive impairment affects a large proportion of patients with MS; prevalence ranges from 40% to 70%, depending on the population studied, the tests used, and the cut-off values applied.7 Characteristically, only a few cognitive domains are affected in these patients, with prominent decrements in information processing speed and episodic memory, and less frequently in executive functions, including verbal fluency and word list generation.8,9 Core language abilities are usually spared, even in the advanced stages of the disease.7 Despite the high frequency of cognitive disturbances, these are usually underestimated during routine clinical visits, mainly because self-reports of these deficits are not a valid measure of cognitive ability,10 and the use of general cognitive assessment methods, which are not tailored for the specific deficits seen in patients with MS, is unsatisfactory. The Expanded Disability Status Scale offers only a rough estimate of cognitive function,11 and other widely used screening methods for cognitive dysfunction, such as the Mini Mental State Examination12 or the Multiple Sclerosis Neuropsychological Screening Questionnaire,10,13 are insensitive to the cognitive profile of patients with MS12 or can be affected by comorbid depression.10 The quality of a given battery should be judged on the basis of a few criteria, including standardisation; www.thelancet.com/neurology Vol 14 March 2015
Review
availability of normative data; test–retest reliability; sensitivity, specificity, and accuracy in the discrimination of patients from controls; the availability of alternative forms of the test (to avoid patients learning the answers of the test); and practicability.8 Two validated neuropsychological test batteries for the assessment of cognitive deficits in patients with MS have wide acceptance: the 45 min Brief Repeatable Battery of Neuropsychological tests (BRB-N)14 and the 90 min Minimal Assessment of Cognitive Function in Multiple Sclerosis (MACFIMS).8,9 These batteries substantially overlap with each other, but their memory tests are different.15 Advantages and disadvantages of each of these batteries have been identified.15 The BRB-N needs less time for assessment and has been translated into many European languages, whereas the MACFIMS has a stronger psychometric foundation and includes assessment of spatial processing and higher executive function abilities. Because trained personnel are usually needed to administer these batteries, the Brief International Cognitive Assessment for Multiple Sclerosis (BICAMS) initiative16 has recommended a short cognitive assessment for patients with MS, which is optimised for small centres with small teams who might not have training in neuropsychological tests, and has been constructed to maximise international use and comparability between countries. The assessment takes 15 min to complete and does not need any special equipment or intensive assessor training. The tests selected were the Symbol Digit Modalities Test,17 to assess information processing speed, which is the most common cognitive domain affected by the disease; the first five recall trials of the California Verbal Learning Test-II,18 to assess verbal short-term memory and learning; and the first three recall trials of the Brief Visuospatial Memory Test Revised,19 to assess visual short-term memory and learning. Among these tests, the Symbol Digit Modalities Test is particularly suitable for longitudinal assessment of MS-related cognitive deterioration.20 An international validation protocol for BICAMS is under development at present.21 However, BICAMS should only be used as a means for cognitive monitoring, and is not intended to replace a full neuropsychological assessment.
Comorbid fatigue and depression Fatigue7 and depression22 are highly prevalent in patients with MS. The interaction and causality between fatigue and cognitive disturbances in patients with MS is still a matter of debate (does fatigue cause cognitive deficits or vice versa?). Although some researchers noted no association between self-reported fatigue (as assessed by the Fatigue Severity Scale)23 and cognitive abilities,24 others described an association between fatigue and impairment in processing speed.25 Even more complex is the interplay among depression, fatigue, and cognitive impairment in patients with MS,26 which has led to the hypothesis that depression and fatigue might share a common www.thelancet.com/neurology Vol 14 March 2015
neurobiological mechanism. An important point that needs to be considered during assessment of depression and fatigue in patients with MS is that fatigue is one of the key symptoms in the diagnosis of major depression according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition.27 Furthermore, depression scales, such as the Montgomery-Asberg Depression Rating Scale28 and the Beck Depression Inventory,29 have a low construct validity for fatigue and depression, meaning that these are confounded variables. To minimise these effects, a reduced version of the Beck Depression Inventory, the Beck Depression Inventory-Fast Screen, which consists of only seven-items30 and excludes items that overlap with fatigue and cognitive impairment, has been proposed and validated in patients with MS.31 A novel objective instrument for differentiating cognitive and motor fatigue in MS is also being developed and validated.32 Clearly, the development of new scales with high sensitivity and specificity for these clinical manifestations of MS is needed.
White matter damage The role of lesions As summarised in the table, many studies have reported an association between the volume of brain T2 hyperintense and T1 hypointense lesions and global neuropsychological performance and performance in individual neuropsychological tests. The location of lesions in brain lobes and white matter tracts has also been shown to have a role in neuropsychological performance. From this perspective, one of the main merits of correlative clinical-lesional studies is the clarification of the importance of a disconnection mechanism (caused by the location of lesions in socalled strategic white matter tracts) to the pathogenesis of several clinical manifestations of MS (figure 1).35,63,70–72 Despite this finding, enthusiasm for assessment of white matter lesions has been lessened by at least two pieces of evidence. First, multiparametric studies,45,53,57 which have combined different MRI techniques to define the relative contributions of focal lesions or damage to the white or grey matter on cognitive performance of patients with MS, have consistently shown that white matter lesions play only a partial or complementary role compared with damage to normalappearing white matter and grey matter. This implies that white matter lesion assessment alone is not adequate to assess and monitor cognitive impairment in patients with MS. Secondly, although white matter lesions are very likely associated with cognitive performance (with discrepancies in results among studies probably related to the number and clinical characteristics of the patients enrolled, neuropsychological tests given, criteria applied for the definition of cognitive impairment, methods used for the quantification of lesions, and statistical approaches), 303
Review
Type of study
White matter Other MRI lesion measure measures included
Number of participants Neuropsychological with MS assessment
Main findings
Additional findings
Rao et al33
Crosssectional
Total lesion area Ventricular-brain ratio; size of the corpus callosum
53 (34 cognitively intact, Measures of verbal 19 cognitively impaired) intelligence, memory, abstract or conceptual reasoning, attention or concentration, language, and visuospatial skills
Total lesion area was a robust predictor of cognitive dysfunction, particularly for measures of recent memory, abstract or conceptual reasoning, language, and visuospatial problem solving
Size of the corpus callosum predicted test performance on measures on mental processing speed and rapid problem solving
SwirskySacchetti34
Crosssectional
Total lesion area Ventricular-brain ratio; size of the corpus callosum
40 (27 cognitively intact, Measures of attention or 13 cognitively impaired) concentration, language, abstract reasoning, memory, visuoperceptual skills, and information processing speed
Total lesion area correlated with most cognitive measures; left frontal lobe lesions predicted impaired abstract problem solving, memory, and word fluency; left parieto-occipital lesions predicted deficits in verbal learning and complex visual-integrative skills.
Size of the corpus callosum predicted test performance on attention, memory, word fluency, verbal learning, and visuoperceptual skills
Arnett et al35
Crosssectional
Total lesion area; frontal lobe lesions
None
43 stratified according to frontal lobe lesions
Conceptual reasoning task (WCST)
Significant association between frontal lobe lesions and WCST performance
None
Foong et al36
Crosssectional
Total T2 lesion volume; frontal lobe T2 lesion volume
None
42 (mixed clinical phenotypes)
Neuropsychological tests of executive skills
Significant correlation between frontal lesion load and impairment of executive skills
None
Miki et al37
Crosssectional
T2 lesion location in different brain lobes and U-fibres
None
53 (mixed clinical phenotypes)
11 neuropsychological tests Worse cognitive performance in of performance in executive patients with U-fibre lesions than those without control and memory
Rovaris et al38
Crosssectional
Global and frontal lobe T2 lesion volume; global and frontal lobe T1 lesion volume
MTR analysis for the whole brain and frontal lobe
30 (15 cognitively impaired and 15 unimpaired)
Measures of verbal and visuospatial memory, attention or concentration, and frontal lobe functions
High frontal lobe T2 lesion volume Low MTR of MS lesions in patients with frontal lobe deficits and whole brain tissue in patients with frontal lobe impairment
Fulton et al39
Crosssectional
T2 lesion volume
None
19 (all RRMS)
12 neuropsychological variables exploring executive functions, informationprocessing speed, verbal long-term memory, verbal short-term memory, and visuospatial ability
T2 lesion volume correlated only with the Symbol Digit Modalities Test and the Rey Auditory Verbal Learning test
None
Comi et al40
Crosssectional
Global and frontal lobe T2 lesion volume; global and frontal lobe T1 lesion volume
MTR analysis for the whole brain, frontal lobe and cerebellum
22 (11 with and 11 without frontal lobe dementia)
Neuropsychological tests of frontal lobe functions
Higher global and frontal lobe T2 and T1 lesion volume in patients with dementia vs patients without dementia; global T1 lesion volume was the best measure for discrimination of the two patient groups vs T2, global, or regional lesion volumes
Global and frontal lobe MTR measures lower in patients with dementia vs patients without dementia
Camp et al41
Crosssectional
T2 lesion volume; T1 lesion volume
Cerebral volume
63 (43 PPMS and 20 transitional progressive)
Brief Repeatable Battery of Neuropsychological tests and a reasoning test
Correlation between cognitive impairment index and T2 and T1 lesion volume
Correlation between cognitive impairment index and cerebral volume
Rovaris et al42
Crosssectional
T2 lesion volume; T1 lesion volume
Cortical or subcortical T2 lesion volume; cortical or subcortical T1 lesion volume; cortical or subcortical MTR
22 (16 cognitively impaired)
Neuropsychological tests exploring the executive functions and verbal and spatial memory
Higher T2 and T1 lesion volumes in patients with cognitive impairment vs patients without cognitive impairment
All MRI measures significantly more abnormal in patients with cognitive impairment vs patients without cognitive impairment; cortical or subcortical brain MTR only factor associated with cognitive impairment in multivariable analysis
None
(Table continues on next page)
304
www.thelancet.com/neurology Vol 14 March 2015
Review
Type of study
White matter Other MRI lesion measure measures included
Number of participants with MS
Neuropsychological assessment
Main findings
Additional findings
(Continued from previous page) Lazeron et al43
Crosssectional
FLAIR juxtacortical lesions
None
39 (mixed clinical phenotypes)
Brief Repeatable Battery of Neuropsychological tests
Correlation between cognitive impairment index and number of juxtacortical lesions
None
Foong et al44
Crosssectional
T2 lesion volume
None
25 (12 PPMS, 13 secondary progressive MS)
Measures of attention and short-term and working memory
No correlation between T2 lesion volume and neuropsychological tests
None
Filippi et al45
Crosssectional
T2 lesion volume; T1 lesion volume
Brain volume; average lesion; MTR; normal-appearing brain tissue MTR;
19 (12 cognitively impaired)
Neuropsychological tests of executive functions and verbal and spatial memory
Higher T2 and T1 lesion volumes in patients with cognitive impairment vs patients without cognitive impairment
Normal-appearing brain tissue MTR was the only factor significantly correlated with cognitive impairment in a multivariable model
Zivadinov et al46
Crosssectional
T2 lesion volume; T1 lesion volume
MTR of normalappearing brain tissue; brain parenchymal fraction
63 (all RRMS with disease duration
Clinical and imaging assessment of cognitive dysfunction in multiple sclerosis Maria A Rocca, Maria P Amato, Nicola De Stefano, Christian Enzinger, Jeroen J Geurts, Iris-K Penner, Alex Rovira, James F Sumowski, Paola Valsasina, Massimo Filippi, for the MAGNIMS Study Group Lancet Neurol 2015; 14: 302–17 Published Online February 4, 2015 http://dx.doi.org/10.1016/ S1474-4422(14)70250-9 Neuroimaging Research Unit (M A Rocca MD, P Valsasina MsC, Prof M Filippi MD) and Department of Neurology (M A Rocca, Prof M Filippi), Institute of Experimental Neurology, Division of Neuroscience, San Raffaele Scientific Institute, Vita-Salute San Raffaele University, Milan, Italy; Department of Neurofarba, Section of Neurosciences, University of Florence, Florence, Italy (M P Amato MD); Department of Medicine, Surgery and Neuroscience, University of Siena, Siena, Italy (N De Stefano MD); Department of Neurology, Medical University of Graz, Graz, Austria (C Enzinger MD); Department of Anatomy and Neuroscience, Section of Clinical Neuroscience, VU University Medical Centre, VUmc Multiple Sclerosis Centre Amsterdam, Amsterdam, Netherlands (J J Geurts PhD); University and University Children’s Hospital Basel, Cognitive Psychology and Methodology and Division of Paediatric Neurology and Developmental Medicine, Basel, Switzerland (I-K Penner MD); Magnetic Resonance Unit, Department of Radiology, Hospital Universitari Vall d’Hebron, Barcelona, Spain (A Rovira MD); and Neuropsychology and Neuroscience, Kessler Foundation Research Center, West Orange, NJ, USA (J F Sumowski PhD) Correspondence to: Prof Massimo Filippi, Neuroimaging Research Unit, Institute of Experimental Neurology, Division of Neuroscience, San Raffaele Scientific Institute, Vita-Salute San Raffaele University, Via Olgettina, 60, 20132, Milan, Italy fi[email protected]
302
In patients with multiple sclerosis (MS), grey matter damage is widespread and might underlie many of the clinical symptoms, especially cognitive impairment. This relation between grey matter damage and cognitive impairment has been lent support by findings from clinical and MRI studies. However, many aspects of cognitive impairment in patients with MS still need to be characterised. Standardised neuropsychological tests that are easy to administer and sensitive to disease-related abnormalities are needed to gain a better understanding of the factors affecting cognitive performance in patients with MS than exists at present. Imaging measures of the grey matter are necessary, but not sufficient to fully characterise cognitive decline in MS. Imaging measures of both lesioned and normal-appearing white matter lend support to the hypothesis of the existence of an underlying disconnection syndrome that causes clinical symptoms to trigger. Findings on cortical reorganisation support the contribution of brain plasticity and cognitive reserve in limiting cognitive deficits. The development of clinical and imaging biomarkers that can monitor disease development and treatment response is crucial to allow early identification of patients with MS who are at risk of cognitive impairment.
Introduction Multiple sclerosis (MS) has traditionally, and simplistically, been defined as a demyelinating-inflammatory white matter disorder, however, grey matter involvement is a feature of MS in most patients. The first description of cortical (including patchy demyelinating lesions) and deep grey matter lesions was presented by Charcot in 1868,1 but grey matter involvement has only been investigated in pathological and imaging studies in the past three decades.2,3 Charcot is also merited with having linked grey matter involvement to some of the clinical manifestations of MS, including cognitive impairment and convulsions. During the past decade, much effort has been put into the characterisation of the type, extent, and topography of grey matter damage in patients with MS at different stages of the disease, and to develop novel MRI sequences that can measure the grey matter involvement and track disease evolution. Some of the main drivers of this research are the need to better explain the heterogeneous and complex clinical manifestations of the disease, particularly cognitive impairment, and to develop novel therapeutic interventions. Through these studies, many advances have been made in our understanding of grey matter involvement in the disease. A few comprehensive review articles have summarised the main findings and the underlying pathological abnormalities.2–6 However, the focus on grey matter involvement in the past decade has led researchers, on some occasions, to count MS among the neurodegenerative disorders. Thus, at this stage, clarification of the distinctive features of grey matter involvement in MS is awaited. Furthermore, differentiation of how grey and white matter involvement contribute to the range of clinical manifestations of MS is also needed. These clarifications are crucial for the development of clinical and imaging biomarkers that will enable clinicians to monitor disease
evolution, predict future cognitive impairment, and define treatment response. The aim of this Review is to critically discuss the clinical scales and MRI metrics available to assess cognitive dysfunction in MS. A summary of the main advancements provided by each of these measures and practical recommendations for their use are also provided. Finally, we discuss unmet needs that should guide future research.
Assessment of cognitive dysfunction in MS Cognitive impairment affects a large proportion of patients with MS; prevalence ranges from 40% to 70%, depending on the population studied, the tests used, and the cut-off values applied.7 Characteristically, only a few cognitive domains are affected in these patients, with prominent decrements in information processing speed and episodic memory, and less frequently in executive functions, including verbal fluency and word list generation.8,9 Core language abilities are usually spared, even in the advanced stages of the disease.7 Despite the high frequency of cognitive disturbances, these are usually underestimated during routine clinical visits, mainly because self-reports of these deficits are not a valid measure of cognitive ability,10 and the use of general cognitive assessment methods, which are not tailored for the specific deficits seen in patients with MS, is unsatisfactory. The Expanded Disability Status Scale offers only a rough estimate of cognitive function,11 and other widely used screening methods for cognitive dysfunction, such as the Mini Mental State Examination12 or the Multiple Sclerosis Neuropsychological Screening Questionnaire,10,13 are insensitive to the cognitive profile of patients with MS12 or can be affected by comorbid depression.10 The quality of a given battery should be judged on the basis of a few criteria, including standardisation; www.thelancet.com/neurology Vol 14 March 2015
Review
availability of normative data; test–retest reliability; sensitivity, specificity, and accuracy in the discrimination of patients from controls; the availability of alternative forms of the test (to avoid patients learning the answers of the test); and practicability.8 Two validated neuropsychological test batteries for the assessment of cognitive deficits in patients with MS have wide acceptance: the 45 min Brief Repeatable Battery of Neuropsychological tests (BRB-N)14 and the 90 min Minimal Assessment of Cognitive Function in Multiple Sclerosis (MACFIMS).8,9 These batteries substantially overlap with each other, but their memory tests are different.15 Advantages and disadvantages of each of these batteries have been identified.15 The BRB-N needs less time for assessment and has been translated into many European languages, whereas the MACFIMS has a stronger psychometric foundation and includes assessment of spatial processing and higher executive function abilities. Because trained personnel are usually needed to administer these batteries, the Brief International Cognitive Assessment for Multiple Sclerosis (BICAMS) initiative16 has recommended a short cognitive assessment for patients with MS, which is optimised for small centres with small teams who might not have training in neuropsychological tests, and has been constructed to maximise international use and comparability between countries. The assessment takes 15 min to complete and does not need any special equipment or intensive assessor training. The tests selected were the Symbol Digit Modalities Test,17 to assess information processing speed, which is the most common cognitive domain affected by the disease; the first five recall trials of the California Verbal Learning Test-II,18 to assess verbal short-term memory and learning; and the first three recall trials of the Brief Visuospatial Memory Test Revised,19 to assess visual short-term memory and learning. Among these tests, the Symbol Digit Modalities Test is particularly suitable for longitudinal assessment of MS-related cognitive deterioration.20 An international validation protocol for BICAMS is under development at present.21 However, BICAMS should only be used as a means for cognitive monitoring, and is not intended to replace a full neuropsychological assessment.
Comorbid fatigue and depression Fatigue7 and depression22 are highly prevalent in patients with MS. The interaction and causality between fatigue and cognitive disturbances in patients with MS is still a matter of debate (does fatigue cause cognitive deficits or vice versa?). Although some researchers noted no association between self-reported fatigue (as assessed by the Fatigue Severity Scale)23 and cognitive abilities,24 others described an association between fatigue and impairment in processing speed.25 Even more complex is the interplay among depression, fatigue, and cognitive impairment in patients with MS,26 which has led to the hypothesis that depression and fatigue might share a common www.thelancet.com/neurology Vol 14 March 2015
neurobiological mechanism. An important point that needs to be considered during assessment of depression and fatigue in patients with MS is that fatigue is one of the key symptoms in the diagnosis of major depression according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition.27 Furthermore, depression scales, such as the Montgomery-Asberg Depression Rating Scale28 and the Beck Depression Inventory,29 have a low construct validity for fatigue and depression, meaning that these are confounded variables. To minimise these effects, a reduced version of the Beck Depression Inventory, the Beck Depression Inventory-Fast Screen, which consists of only seven-items30 and excludes items that overlap with fatigue and cognitive impairment, has been proposed and validated in patients with MS.31 A novel objective instrument for differentiating cognitive and motor fatigue in MS is also being developed and validated.32 Clearly, the development of new scales with high sensitivity and specificity for these clinical manifestations of MS is needed.
White matter damage The role of lesions As summarised in the table, many studies have reported an association between the volume of brain T2 hyperintense and T1 hypointense lesions and global neuropsychological performance and performance in individual neuropsychological tests. The location of lesions in brain lobes and white matter tracts has also been shown to have a role in neuropsychological performance. From this perspective, one of the main merits of correlative clinical-lesional studies is the clarification of the importance of a disconnection mechanism (caused by the location of lesions in socalled strategic white matter tracts) to the pathogenesis of several clinical manifestations of MS (figure 1).35,63,70–72 Despite this finding, enthusiasm for assessment of white matter lesions has been lessened by at least two pieces of evidence. First, multiparametric studies,45,53,57 which have combined different MRI techniques to define the relative contributions of focal lesions or damage to the white or grey matter on cognitive performance of patients with MS, have consistently shown that white matter lesions play only a partial or complementary role compared with damage to normalappearing white matter and grey matter. This implies that white matter lesion assessment alone is not adequate to assess and monitor cognitive impairment in patients with MS. Secondly, although white matter lesions are very likely associated with cognitive performance (with discrepancies in results among studies probably related to the number and clinical characteristics of the patients enrolled, neuropsychological tests given, criteria applied for the definition of cognitive impairment, methods used for the quantification of lesions, and statistical approaches), 303
Review
Type of study
White matter Other MRI lesion measure measures included
Number of participants Neuropsychological with MS assessment
Main findings
Additional findings
Rao et al33
Crosssectional
Total lesion area Ventricular-brain ratio; size of the corpus callosum
53 (34 cognitively intact, Measures of verbal 19 cognitively impaired) intelligence, memory, abstract or conceptual reasoning, attention or concentration, language, and visuospatial skills
Total lesion area was a robust predictor of cognitive dysfunction, particularly for measures of recent memory, abstract or conceptual reasoning, language, and visuospatial problem solving
Size of the corpus callosum predicted test performance on measures on mental processing speed and rapid problem solving
SwirskySacchetti34
Crosssectional
Total lesion area Ventricular-brain ratio; size of the corpus callosum
40 (27 cognitively intact, Measures of attention or 13 cognitively impaired) concentration, language, abstract reasoning, memory, visuoperceptual skills, and information processing speed
Total lesion area correlated with most cognitive measures; left frontal lobe lesions predicted impaired abstract problem solving, memory, and word fluency; left parieto-occipital lesions predicted deficits in verbal learning and complex visual-integrative skills.
Size of the corpus callosum predicted test performance on attention, memory, word fluency, verbal learning, and visuoperceptual skills
Arnett et al35
Crosssectional
Total lesion area; frontal lobe lesions
None
43 stratified according to frontal lobe lesions
Conceptual reasoning task (WCST)
Significant association between frontal lobe lesions and WCST performance
None
Foong et al36
Crosssectional
Total T2 lesion volume; frontal lobe T2 lesion volume
None
42 (mixed clinical phenotypes)
Neuropsychological tests of executive skills
Significant correlation between frontal lesion load and impairment of executive skills
None
Miki et al37
Crosssectional
T2 lesion location in different brain lobes and U-fibres
None
53 (mixed clinical phenotypes)
11 neuropsychological tests Worse cognitive performance in of performance in executive patients with U-fibre lesions than those without control and memory
Rovaris et al38
Crosssectional
Global and frontal lobe T2 lesion volume; global and frontal lobe T1 lesion volume
MTR analysis for the whole brain and frontal lobe
30 (15 cognitively impaired and 15 unimpaired)
Measures of verbal and visuospatial memory, attention or concentration, and frontal lobe functions
High frontal lobe T2 lesion volume Low MTR of MS lesions in patients with frontal lobe deficits and whole brain tissue in patients with frontal lobe impairment
Fulton et al39
Crosssectional
T2 lesion volume
None
19 (all RRMS)
12 neuropsychological variables exploring executive functions, informationprocessing speed, verbal long-term memory, verbal short-term memory, and visuospatial ability
T2 lesion volume correlated only with the Symbol Digit Modalities Test and the Rey Auditory Verbal Learning test
None
Comi et al40
Crosssectional
Global and frontal lobe T2 lesion volume; global and frontal lobe T1 lesion volume
MTR analysis for the whole brain, frontal lobe and cerebellum
22 (11 with and 11 without frontal lobe dementia)
Neuropsychological tests of frontal lobe functions
Higher global and frontal lobe T2 and T1 lesion volume in patients with dementia vs patients without dementia; global T1 lesion volume was the best measure for discrimination of the two patient groups vs T2, global, or regional lesion volumes
Global and frontal lobe MTR measures lower in patients with dementia vs patients without dementia
Camp et al41
Crosssectional
T2 lesion volume; T1 lesion volume
Cerebral volume
63 (43 PPMS and 20 transitional progressive)
Brief Repeatable Battery of Neuropsychological tests and a reasoning test
Correlation between cognitive impairment index and T2 and T1 lesion volume
Correlation between cognitive impairment index and cerebral volume
Rovaris et al42
Crosssectional
T2 lesion volume; T1 lesion volume
Cortical or subcortical T2 lesion volume; cortical or subcortical T1 lesion volume; cortical or subcortical MTR
22 (16 cognitively impaired)
Neuropsychological tests exploring the executive functions and verbal and spatial memory
Higher T2 and T1 lesion volumes in patients with cognitive impairment vs patients without cognitive impairment
All MRI measures significantly more abnormal in patients with cognitive impairment vs patients without cognitive impairment; cortical or subcortical brain MTR only factor associated with cognitive impairment in multivariable analysis
None
(Table continues on next page)
304
www.thelancet.com/neurology Vol 14 March 2015
Review
Type of study
White matter Other MRI lesion measure measures included
Number of participants with MS
Neuropsychological assessment
Main findings
Additional findings
(Continued from previous page) Lazeron et al43
Crosssectional
FLAIR juxtacortical lesions
None
39 (mixed clinical phenotypes)
Brief Repeatable Battery of Neuropsychological tests
Correlation between cognitive impairment index and number of juxtacortical lesions
None
Foong et al44
Crosssectional
T2 lesion volume
None
25 (12 PPMS, 13 secondary progressive MS)
Measures of attention and short-term and working memory
No correlation between T2 lesion volume and neuropsychological tests
None
Filippi et al45
Crosssectional
T2 lesion volume; T1 lesion volume
Brain volume; average lesion; MTR; normal-appearing brain tissue MTR;
19 (12 cognitively impaired)
Neuropsychological tests of executive functions and verbal and spatial memory
Higher T2 and T1 lesion volumes in patients with cognitive impairment vs patients without cognitive impairment
Normal-appearing brain tissue MTR was the only factor significantly correlated with cognitive impairment in a multivariable model
Zivadinov et al46
Crosssectional
T2 lesion volume; T1 lesion volume
MTR of normalappearing brain tissue; brain parenchymal fraction
63 (all RRMS with disease duration
