If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author at: Department of Brain Sciences and UK Dementia Research Institute at Imperial College London, Burlington Danes Building, Hammersmith Hospital, DuCane Road, London, UK.
Department of Brain Sciences and UK Dementia Research Institute at Imperial College London, Burlington Danes Building, Hammersmith Hospital, DuCane Road, London, UK
Department of Brain Sciences and UK Dementia Research Institute at Imperial College London, Burlington Danes Building, Hammersmith Hospital, DuCane Road, London, UKDepartment of Computing, Imperial College London, William Penny Building, South Kensington Campus, London, UK
Patients with multiple sclerosis (MS) show different associations between brain volume loss (BVL) and physical disability.
•
Despite differences, prior studies together suggest that brain atrophy is associated with greater physical disability progression.
•
Our findings can help define future imaging biomarkers for physical disability progression and treatment monitoring in MS.
•
There is a need to engage key opinion leaders in working towards standardizing BVL and physical disability outcomes.
Abstract
Background
Multiple sclerosis (MS) is a chronic, inflammatory, demyelinating, degenerative disease of the central nervous system that affects approximately 2.8 million people worldwide. Compelling evidence from observational studies and clinical trials indicates a strong association between brain volume loss (BVL) and the accumulation of disability in MS. However, the considerable heterogeneity in study designs and methods of assessment of BVL invites questions concerning the generalizability of the reported findings. Therefore, we conducted this systematic review to characterize the relationship between BVL and physical disability in patients with MS.
Methods
A systematic literature search of MEDLINE and EMBASE databases was performed supplemented by gray literature searches. The following study designs were included: prospective/retrospective cohort, cross-sectional and case-control. Only English language articles published from 2010 onwards were eligible for final inclusion. There were no restrictions on MS subtype, age, or ethnicity. Of the 1620 citations retrieved by the structured searches, 50 publications met our screening criteria and were included in the final data set.
Results
Across all BVL measures, there was considerable heterogeneity in studies regarding the underlying study population, the definitions of BVL and image analysis methodologies, the physical disability measure used, the measures of association reported and whether the analysis conducted was univariable or multivariable. A total of 36 primary studies providing data on the association between whole BVL and physical disability in MS collectively suggest that whole brain atrophy is associated with greater physical disability progression in MS patients. Similarly, a total of 15 primary studies providing data on the association between ventricular atrophy and physical disability in MS suggest that ventricular atrophy is associated with greater physical disability progression in MS patients. Along similar lines, the existing evidence based on a total of 13 primary studies suggests that gray matter atrophy is associated with greater physical disability progression in MS patients. Four primary studies suggest that corpus callosum atrophy is associated with greater physical disability progression in MS patients. The majority of the existing evidence (6 primary studies) suggests no association between white matter atrophy and physical disability in MS. It is difficult to assign a relationship between basal ganglia volume loss and physical disability as well as medulla oblongata width and physical disability in MS due to very limited data.
Conclusion
The evidence gathered from this systematic review, although very heterogeneous, suggests that whole brain atrophy is associated with greater physical disability progression in MS patients. Our review can help define future imaging biomarkers for physical disability progression and treatment monitoring in MS.
9-HPT: Nine-Hole Peg Test; BPF: brain parenchymal fraction; BVL: brain volume loss; CDMS: clinically definite multiple sclerosis; CDW: confirmed disability worsening; CIS: clinically isolated syndrome; CSF: cerebrospinal fluid; DMT: disease modifying therapy; EDSS: Expanded Disability Status Scale; EmbaseTM: Excerpta Medica Database; HCRU: healthcare resource utilization; ICHT: Imperial College Healthcare Trust; MEDLINETM: Medical Literature Analysis and Retrieval System Online; MRI: magnetic resonance imaging; MS: multiple sclerosis; MSFC: Multiple Sclerosis Functional Composite; MSSS: Multiple Sclerosis Severity Scale; NeuroSTREAM: Neurological software tool for reliable atrophy measurement; NBV: normalized brain volume; NOS: Newcastle-Ottawa Scale; PPMS: primary progressive multiple sclerosis; PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-analyses; RRMS: relapsing-remitting multiple sclerosis; SIENA: Structural Image Evaluation, using Normalization of Atrophy; SIENAX: Structural Image Evaluation, using Normalization of Atrophy Cross-Sectional; SPMS: secondary progressive multiple sclerosis; T25FWT: Timed 25-Foot Walk Test.
is a chronic, inflammatory, demyelinating, degenerative disease of the central nervous system that affects approximately 2.8 million people worldwide (
). There are three phenotypic disease course patterns, distinguished by the occurrence and timing of relapses as well as by disability progression relative to disease onset: RRMS, which in most cases evolves into SPMS, and PPMS (
Chronic BVL provides an index of neurodegeneration. Increased rates of BVL are observed from the earliest stage of MS and proceed throughout the course of the disease (
). Among RRMS patients, the annual rate of BVL ranges from 0.5%–1.35% per year, compared to a normal age-related deterioration of 0.1%–0.3% per year in healthy individuals (
Compelling evidence from observational studies and clinical trials indicates a strong correlation between BVL and the accumulation of disability in MS. Among PPMS patients (N=487) in the placebo arm from the INFORMS trial, the baseline NBV was found to be predictive of disease progression; the relative risk of 12-week CDW was reduced by 36% in high versus low baseline NBV (
). Cohort studies have provided further supporting evidence of an association between BVL and disability accumulation in the long term. Among RRMS patients (N=480) prospectively observed for 10 years, the three groups with lowest quartiles of adjusted baseline brain volume had a significantly increased risk (odds ratios ranging from 4.13 to 6.12) of disease progression, compared to the highest quartile. Importantly, this correlation was found to be independent of the occurrence of clinical relapses (
However, the considerable heterogeneity in study designs and methods of assessment of BVL reported in the literature invites questions concerning the generalizability of these and related observations. Changes in brain volume generally are small and most typically not able to be assessed quantitatively with the use of computational tools that detect differences or changes using a range of algorithms. Cross-sectional methods (e.g., BPF and SIENAX) measure brain volume at one time point using a single MRI scan, while longitudinal methods (e.g., SIENA) measure change in brain volume over time by comparing two MRI scans acquired at different times. BPF and SIENA are the most frequently used methods (
), but even measures determined with the same algorithm can be influenced strongly by technical aspects of image acquisition that vary between sites, across MRI machines, and imaging protocols, as well as tissue artefacts arising from differences in subject position within scanners or movement (
We have undertaken a systematic review of the large volume of heterogenous literature on the association between BVL and physical disability (primarily assessed using EDSS) in MS patients. The goal of this review was to systematically characterize the relationship between BVL (measured in terms of global/total brain volume and regional brain volume) and physical disability (assessed using scales such as EDSS, MSFC, and MSSS) in patients with MS.
2. Methods
This systematic literature review was conducted according to the PRISMA guidelines (
), using a protocol that predefined study designs, populations, and outcomes of interest.
2.1 Literature search strategy
We performed a systematic literature search of MEDLINETM and EmbaseTM using embase.com. The search strategy is shown in Supplementary Table 1. Searches were run on 20th June 2021. We excluded conference abstracts.
In addition, we conducted bibliographic searches of systematic reviews and narrative reviews published in the last three years, as well as gray literature searches in Google and Google Scholar.
2.2 Eligibility criteria
We included original research articles assessing BVL in MS (using the outcomes of interest shown in Fig. 1) and MS physical disability using the measures EDSS, MSSS and MSFC (or its performance subscale scores for the 9-HPT or T25FWT). We included the following study designs: prospective/retrospective cohort, cross-sectional and case-control; and excluded randomized and other controlled trials (including single-arm trials). Only English language articles published from 2010 onwards were eligible for final inclusion. There were no restrictions on MS subtype, age, or ethnicity.
2.2.1 Eligibility assessment, quality assessment and data extraction
We screened titles and abstracts based on the eligibility criteria, and full-text copies of potentially relevant publications using the same criteria. We assessed publication quality using the NOS for observational studies (total score can range from 0 to 9, where 0 = poorest quality and 9 = highest quality;
); and extracted relevant information into a predefined Excel-based data extraction grid. Two reviewers screened articles, conducted quality assessment and extracted data, with any discrepancies resolved by a third reviewer.
3. Results
3.1 Study characteristics
Of the 1620 citations retrieved by the structured searches, 50 publications met our screening criteria and were included in the final data set. Key screening metrics are shown in Fig. 2.
Similar proportions were retrospective and prospective cohort studies (n=24 and 26, respectively). Approximately 43% of the cohort studies (n=21) had a follow-up duration of greater than or equal to five years and ∼53% (n=27) a follow-up duration of less than five years. Follow-up duration was not reported for two studies. Approximately 25% (n=13) of the studies had a sample size of <50, ∼20% (n=9) had a sample size between 51 and 100 and ∼55% (n=28) had a sample size more than 100. Geographic location varied, with most studies being conducted in Europe (70%, n=34) followed by North America (21%, n=10), Pacific (4%, n=2) and Asia and South America (2%, n=1, each). Approximately 63% (n=31) of the studies were published from 2015-2021 and the remaining studies were published from 2010-2014 (∼ 37%, n=19).
The distribution of these 50 studies by design, duration of follow up, sample size and publication date is presented in Fig. 3.
Approximately 20% (n=9) of the studies included RRMS patients only; ∼15% (n=8) CIS patients only; ∼5% (n=2) PPMS patients only and ∼60% (n=30) a combination of different MS subtypes. Average EDSS scores at baseline were not reported in ∼43% (n=21) of the studies. Average baseline EDSS scores varied considerably across the remaining studies: ≤2.0 in ∼20% (n=10), >2.0 to ≤3.0 in ∼15% (n=8), >3.0 to ≤4.0 in ∼20% (n=10), and >4.0 in one study (2%).
Correlation coefficient was the most commonly used measure of association (26%, n=13) followed by regression coefficient (22%, n=11). More than one measure was used in 14 studies (28%). Finally, ∼50% (n=24) of the studies reported multivariable analyses, ∼30% (n=16) reported univariable analyses and ∼20% (n=10) reported both types.
Table 1 presents the number of publications reporting associations between each type of BVL outcome and physical disability measure of interest.
Table 1Numbers of publications reporting associations Between Brain Volume Loss and Physical Disability Reported Across Studies.
Brain Volume Loss Measures
Numbers of publication with each Physical Disability Measure
EDSS
MSSS
MSFC (or its components)
Total Brain Volume
Whole Brain
36
3
3
Total ventricle
4
1
Lateral ventricle
10
2
1
Third ventricle
2
1
Gray Matter Volume
Total gray matter
9
1
Cortical gray matter
6
1
Deep gray matter
3
Cortical thickness
3
White Matter Volume
White matter
6
Corpus Callosum Index / Area
Corpus callosum
4
Brainstem Width
Medulla oblongata
1
Limbic Structure Volume
Amygdala, hippocampus, bilateral cingulate gyrus, and thalamus
6
1
1
Basal Ganglia Volume
Accumbens, caudate, globus pallidus, putamen, and left pallidum
3.2 Association between brain volume loss and physical disability
3.2.1 Whole brain volume
Table 2 presents an overview of the 36 primary studies reporting data on the association between whole BVL and physical disability in MS. Of these 36 studies, 19 were prospective cohorts and 17 were retrospective cohorts. The follow-up duration varied from one year to 15 years with sample sizes ranging from 20 to 1651. The average age of patients across the 36 studies ranged from 30 to 51 years. EDSS was the most commonly reported disability measure (20/36 studies, range 1.0 – 6.0 points at baseline). The studies in Table 2 have been arranged under two heads based on their follow-up duration: i) greater than or equal to 5 years (n = 15 studies) or ii) less than 5 years (n = 21 studies). Under each head, the studies are arranged in reverse chronological order, starting with the most recently published.
Table 2Association between whole brain volume loss and physical disability in MS.
Reference (Country); Quality rating
Study design / Duration
Study population
Average age (years)
BVL measurement & algorithm used
Physical disability measurement
Results
Conclusion
Association between whole BVL and physical disability in MS (greater than or equal to five years follow-up, n = 15 studies)
Total: 480 CIS: 88 RRMS: 392 Median EDSS at baseline: 1.5
Overall: 41.6
Annualized percentage change in relative brain volume (slope of follow-up year divided by the relative brain volume at baseline) over 10-years follow-up Algorithm used: SIENAX
EDSS scores at baseline, 5- and 10- years follow-up was used to classify 1) DP: An increase in the EDSS of 1.5, 1.0 and 0.5 points if the baseline EDSS score was 0.0, 1.0 to 5.0 and ≥ 5.5 points respectively assessed from baseline to year 5 (±1 year) and sustained to year 10 (±1 year) 2) Stable: Patients who do not meet the criteria for DP
BVL and physical disability (MV analysis)Without relapseLinear mixed model β = −0.00018; P = 0.0094
•
In patients with increased physical disability without relapse from baseline to 10-year follow-up there was a 0.00018 % significant annual relative BVL compared to patients with stable EDSS and without prior relapses during the 10-year follow-up
With relapseLinear mixed model β = −0.00013; P = 0.0470
•
In patients with increased physical disability with relapse from baseline to 10-year follow-up there was a 0.00013% significant annual relative BVL compared to patients with stable EDSS and without prior relapses during the 10-year follow-up
Factors adjusted in MV analysis: Age, sex, disease duration at baseline and HLA-DRB1*15:01 (an allele influencing certain clinical features of MS).
Long-term worsening (physical disability progression) was common in relapsing MS patients and was associated with accelerated brain atrophy.
Retrospective cohort 6- and 12- years follow-up groups
6 years follow-up group Total: 115 CIS: 36 RRMS: 68 PPMS: 11 12 years follow-up group Total: 79 CIS: 27 RRMS: 50 PPMS: 2 Median EDSS at baseline: 2.0
6 years follow-up group: 35.3 12 years follow-up group: 34.9
aPBV changes at 2 years follow-up Algorithm used: SIENA and VIENA
EDSS at 6-and 12- years follow-up used as a continuous variable
aPBV loss (from baseline to year 2) among6-year follow-up group (MV analysis)Linear standardized β = –0.143; P = 0.026
•
Every 1 standard deviation increase in aPBV loss over 2 years was significantly associated with a 0.143 standard deviation higher (worse) EDSS score at 6-year follow-up
aPBV loss (from baseline to year 2) among12-year follow-up group (UV analysis)Linear standardized β = – 0.037, P-value =0.743
•
Every 1 standard deviation increase in aPBV loss over 2 years was associated with a 0.037 standard deviation higher (worse) EDSS score at 12-year follow up. However, the association was not statistically significant
Factors adjusted in MV analysis: Baseline EDSS and disease-modifying therapy use during follow-up.
MS patients with early signs of neurodegeneration (or early whole brain atrophy) were more prone to develop physical disability progression.
Total: 1465 CIS: 124 RRMS: 1089 SPMS: 217 PPMS: 35 Median EDSS at baseline: NR
Overall: NR DP converters: 47.6 DP non-converters: 44.3
aPBV changes at 5 years follow-up Algorithm used: SIENA
EDSS at 5-year follow-up classified as DP converters: Increase in EDSS by ≥ 1.5, ≥ 1.0, and ≥ 0.5 from baseline <1.0, 1.0 to 5.5, and >5.5 points respectively DP non-converters: Patients not meeting the disease progression criteria
aPBV loss and DP conversion (MV analysis)Logistic OR = 0.89; P = 0.006
•
Every 1% increase in aPBV loss was significantly associated with a 12% higher risk of physical disability progression at 5 years of follow-up
Factors adjusted in MV analysis: All the significant MRI outcomes in the univariate regression analysis
PBV loss represents a viable predictive MRI marker of the development of physical disability in MS patients.
Total:192 CIS: 18 RRMS: 126 PPMS: 48 Median EDSS at baseline: NR
Overall: NR CIS: 44.8 RRMS: 43.8 SPMS/PPMS: 55.5
PBV change at 5- years follow-up Algorithm used: SIENA
EDSS change between baseline and 5 years used as a continuous variable 9-HPT and T25-fwt at 5 years used as a continuous variable
PBV loss and EDSS change (MV analysis)Hierarchical regression: R square = 0.045; P = 0.061
•
BVL (measured using PBV change) at 5-year follow-up was not associated with a change in EDSS between baseline and 5-year follow-up
PBV loss and 9-HPT (UV analysis) • BVL (measured using PBV change) at 5-year follow-up was not associated with 9-HPT at 5-yearsPBV loss and T25-fwt (UV analysis)
•
BVL (measured using PBV change) at 5-year follow-up was not associated with timed T25-fwt at 5 years
Factors adjusted in MV analysis: Age, sex, new lesion volume, and enlarging lesion volume
No significant associations were found between BVL and physical disability in MS patients.
Total: 1651 CIS: 137 RRMS: 1219 PPMS: 255 SPMS: 40 Median EDSS at baseline: 2.5 Median MSSS at baseline: 3.7
Overall: NR MS: 45.6 CIS: 39.5
aNBV and aPBV changes at 5 years follow-up Algorithm used: 1. NBV change - SIENAX 2. PBV change - SIENA
Annual change in EDSS and MSSS as a continuous variable and EDSS was further classified as DP: An absolute change in EDSS from the first to most recent follow-up MR imaging with an increase in EDSS of ≥1.5, ≥1.0 and ≥0.5 points if the baseline EDSS was 0.0, 1.0 to 5.0, ≥5.5 points respectively were confirmed after at least 24 weeks. Non-DP/ Stable- Not defined
aNBV loss and EDSS change (UV analysis)Linear mixed-effect model β = –2.187; P value = 0.0001
•
Every one-point increase in EDSS score annually was significantly associated with a 2.187 ml increased BVL (measured using aNBV loss) at 5 years follow-up
aNBV loss and MSSS change (UV analysis)Linear mixed-effect model β = –2.001; P value = 0.0001
•
Every one-point increase in MSSS score (worsening) annually was significantly associated with a 2.001 ml increased BVL (measured using aNBV loss) at 5 years follow-up
aNBV loss and DP (UV analysis)Linear mixed-effect model β = –1.281 (-12.5%); P value = 0.03
•
MS patients with DP had a 12.5% increased rate of BVL (measured using aNBV loss) compared to those without DP at 5 years follow-up
aPBV loss and EDSS change (UV analysis)Linear mixed-effect model β = –0.134; P = 0.0001
•
Every one-point increase in EDSS score annually was significantly associated with a 0.134% increased BVL (measured using aPBV loss) at 5 years follow-up
aPBV loss and MSSS change (UV analysis)Linear mixed-effect model β = –0.119; P = 0.0001
•
Every one-point increase in MSSS score (worsening) annually was significantly associated with a 0.119% increased BVL (measured using aPBV loss) at 5 years follow-up
aPBV loss and DP (UV analysis)Linear mixed-effect model β = -0.169 (–21.9%); P = 0.0001
•
MS patients with DP had a 21.9% increased rate of BVL (measured using aPBV loss) compared to those without DP at 5 years follow-up
Increased BVL was associated with physical disability progression in MS patients.
Total: 119 CIS: 3 RRMS: 92 PPMS: 4 SPMS: 20 Median EDSS at baseline: 2.5 Median MSFC at baseline: 0.03
Overall: 44.0
PBV change at 5-years follow-up Algorithm used: SIENA
EDSS and MSFC at 5 years as a continuous variable
PBV loss and EDSS (MV analysis)
•
An increase in BVL (measured using central and WBV change) over 5 years was significantly associated with a higher (worse) EDSS score at 5 years
PBV loss and MSFC (MV analysis)
•
An increase in BVL (measured using central and WBV change) over 5 years was not associated with MSFC score at 5 years
Factors adjusted in MV analysis: Age and sex
An increase in BVL over 5 years was associated with a higher (worse) EDSS score at 5 years, however, no significant associations were found between BVL and MSFC in MS patients.
PBV change at 5 years follow-up Algorithm used: SIENA (FSL- v 5.0.2)
EDSS and EDSS change at 5 years as a continuous variable
PBV loss and EDSS (MV analysis)Linear β = −0.11, 95% CI (−0.21 to −0.02); P = 0.018
•
Every 1 per cent increase in BVL from baseline to 5 years follow-up was significantly associated with 0.11 points higher (worse) EDSS score at 5 years
PBV loss and EDSS change (MV analysis)Linear β = -0.12, 95% CI (Bootstrap) (−0.33 to −0.005); P = 0.045
•
Every 1 per cent increase in BVL from baseline to 5 years follow-up was significantly associated with an increase (worsening) of 0.12 points in EDSS score between baseline and 5 years
Factors adjusted in MV analysis: Age, sex, disease duration and scanner upgrade (by adjusting for those subjects whose baseline and follow-up straddled the upgrade).
Increased brain atrophy was associated with higher physical disability in the first 5 years after a non-spinal CIS.
Total: 206 RRMS: 180 PPMS: 12 SPMS: 14 Median EDSS at baseline: NR
Overall: 37.0
aPBV change and aPBV change dichotomized according to the −0.4% cut-off Algorithm used: SIENA
Change in annualized EDSS used as a continuous variable over 7.5 years follow-up
aPBV loss and EDSS change (MV analysis)Pearson's partial r = −0.23, P = 0.001
•
An increase in aPBV loss was significantly associated with an increase (worsening) in annualized EDSS score over 7.5 years of follow-up
aPBV loss (according to the −0.4% cut-off) and EDSS change (MV analysis)
•
Patients with aPBV loss greater than−0.4%, cut-off had an increased (worsening) annualized EDSS score compared to patients who had aPBV loss lesser than−0.4% over 7.5 years of follow-up (P<0.001)
Factors adjusted in MV analysis: Age, disease duration, baseline EDSS and disease type.
Increased BVL was associated with physical disability progression in MS patients.
Total: 81 RRMS: 62 SPMS: 11 PPMS: 8 Median EDSS at baseline: 3.5
Overall: 42.0
WBV change and from baseline to 5-year or from baseline to 10-year follow-up Algorithm used: 1. Normalized WBV -SIENAX 2. Longitudinal WBV change- SIENA
EDSS at 5- and 10-year follow-up classified as DP: An increase of EDSS by 1.0 and 0.5 points if baseline EDSS was <6.0 and ≥6.0 respectively No DP: Patients who do not meet the criteria for physical disability progression
WBV loss and EDSS (MV analysis)At 5-year follow-up: Logistic β = −0.571; P <0.01At 10-year follow-up: Logistic β = −0.391; P <0.05
•
Every 1 per cent increase in BVL (measured using WBV change) was significantly associated with a 1.8 times higher risk of physical disability progression at 5 years and 1.5 times higher risk of physical disability progression at 10 years
Factors adjusted in MV analysis: Age, sex, scanner type, baseline volumes of specific structures, T1-lesion volumes, T2-lesion volumes, MS subtype, disease duration and use of disease-modifying treatment.
Whole-BVL was associated with significant physical disability progression at 5 and 10 years of follow-up in MS patients.
Total: 261 CIS: 18 RRMS: 97 PPMS: 69 SPMS: 77 Median EDSS at baseline: 4.0 Median MSSS at baseline: 5.4
Overall: 43.0
aPBV change at 1-2 years follow-up Algorithm used: SIENAX and SIENA
EDSS and MSSS at 10 years used as a continuous variable
aPBV loss and EDSS (MV analysis)Linear β = 0.35, 95% CI (– 0.58 to – 0.12)
•
Every 1 per cent increase in whole brain atrophy (measured using aPBV loss) over 1-2 years was significantly associated with a 0.35-points higher (worse) EDSS score at 10 years.
aPBV loss and MSS (MV analysis)Linear β = 0.39, 95% CI (– 0.62 to – 0.16) • Every 1 per cent increase in whole brain atrophy (measured using aPBV loss) over 1-2 years was significantly associated with a 0.39-points higher (worse) MSSS score at 10 years.Factors adjusted in MV analysis: Imaging protocol, disease type and EDSS at baseline
Early brain atrophy rates were related to subsequent long-term physical disability in MS patients.
aPBV change at 1-year follow-up Algorithm used: SIENA
EDSS at 9.3-years follow-up categorized as 1. EDSS=0.0 2. EDSS 1.0 to 2.5 3. EDSS ≥3.0
aPBV loss and EDSS (UV analysis)
•
aPBV loss at 1-year follow-up was significantly higher in patients who reached EDSS ≥3.0 (higher disability) than in patients that had an EDSS between 1.0 and 2.5 or 0.0 at 9.3-years follow-up
aPBV loss and EDSS (MV analysis)Ordinal logistic OR = NR; P <0.001
•
Higher aPBV loss at 1-year follow-up was significantly associated with higher physical disability measured by EDSS at 9.3-years follow-up
Factors adjusted in MV analysis: Age, sex, EDSS at onset, oligoclonal bands in cerebrospinal fluid, and T2 lesion volume at baseline and one year.
Higher brain volume atrophy during the first year of disease onset independently predicted long term physical disability in RRMS patients.
aPBV change at 1-year follow-up Algorithm used: 1. Brain atrophy analysis -SUN workstation 2. PBV change - SIENA
EDSS, MSFC (or its components-T25-fwt and 9-HPT) at 6-years follow-up used as a continuous variable.
Disability was also categorized as follows: EDSS 0.0 = None, EDSS >0.0 and <3.0 = Minimal, and EDSS ≥3.0 = Significant
Time to CDMS development
aPBV loss and EDSS at 6 years (MV analysis)Quantitative data: NR
•
aPBV change from baseline to 1 year did not predict EDSS independently at 6 years follow-up (quantitative data not reported)
aPBV loss and disability categories (UV analysis)
•
For patients with different degrees of disability at 6 years i.e. “none” (EDSS 0.0), “minimal” (EDSS >0.0 and <3.0), and “significant” (EDSS ≥3.0) the aPBV loss from baseline to 1 year remained a significant predictor of physical disability (p=0.005) in univariate regression analysis
aPBV loss and MSFC at 6 years (sub-group analysis [n=64]) (UV analysis)
•
aPBV change from baseline to 1 year was not a significant predictor of MSFC (p=0.309) or its components (timed 25-foot walk test, p = 0.340; nine-hole peg test, p = 0.567) at 6 years follow-up
aPBV loss and CDMS conversion (UV analysis)CDMS: –0.50%/ year [SD 0.63]Did not develop MS: –0.26%/year [SD 0.42]Mean difference –0.24 %/year,95% CI (0.02 to 0.46), P = 0.035
•
There was a significantly higher BVL (measured using aPBV change) over 1-year among CIS patients who developed CDMS compared to those who did not at 6-years follow-up
aPBV loss and risk of MS diagnosis (UV analysis)
•
The risk (hazard) of diagnosis with MS was reduced by 33.8% (95% CI 12.0% to 50.2%, P = 0.005) for each SD lower (i.e., less negative) atrophy rate at 6-years follow-up
Factors adjusted in MV analysis: Age and sex
Increased brain atrophy early in the course of the disease was predictive of MS diagnosis and physical disability progression at 6 years in CIS patients.
Total: 54 RRMS: 43 PPMS: 11 Median EDSS at baseline: 2.0
Overall: 35.0
aPBV change at 2- and 5.5-years follow-up Algorithm used: SIENA
Annualized EDSS change from baseline to 2 years, 5.5 years and from 2 years to 5.5 years as a continuous variable and further classified as DP: An increase in EDSS of at least 1.0 points after 5.5 years of follow-up. No DP: Patients not meeting the criteria for DP
aPBV change and EDSS change (UV analysis)Spearman's r (P-value)
•
An increase in aPBV loss from baseline to 2-years was significantly associated with an increase (worsening) in annualized EDSS change from baseline to 2-years follow-up
•
aPBV loss from baseline to 2 years was not associated with annualized EDSS change from 2-years to 5.5-years follow-up
•
aPBV loss from 2 to 5.5-years was not associated with annualized EDSS change from baseline to 2-years or from 2-years to 5.5-years follow-up
Note: Results for aPBV change from 0-2 years or 2 to 5.5 years and EDSS change from baseline to 5.5 years were not reportedBVL and disability (UV analysis)Median (IQR)
•
There were no significant differences in aPBV change from baseline to 2-years or from 2 to 5.5-years between DP and no DP groups
Early BVL was associated with physical disability worsening at 2-years follow-up however, there was no significant difference in aPBV change between DP and no DP groups.
A longitudinal observational study of brain atrophy rate reflecting four decades of multiple sclerosis: a comparison of serial 1D, 2D, and volumetric measurements from MRI images.
Total: 37 RRMS: 16 PPMS: 4 SPMS: 17 Median EDSS at baseline: NR Median MSSS at baseline: NR
Overall: 42.0
aBPV change at 9.25 years follow-up Algorithm used: Semiautomatic tool in HERMES MultiModality- Region Growing method.
EDSS and MSSS at 9.25 years follow-up (not specified how EDSS & MSSS were categorized in logistic regression analysis)
aBPV loss and EDSS (UV analysis)Exact logistic OR = 1.72, 95% CI (0.4 to 7.71); P = 0.449
•
A 0.1 cm3 decrease in aBPV (BVL) was associated with 1.72 times higher risk of physical disability progression (measured by EDSS) over 9.25 years of follow-up, however, the association was not significant
aBPV loss and MSSS (UV analysis)Exact logistic OR = 0.99, 95% CI (0.82 to 1.19); P = 0.945There was no significant association between BVL (measured using aPBV loss) and risk of physical disability progression (measured by MSSS) over 9.25 years of follow-up
Whole BVL was not associated with physical disability progression as measured by EDSS or MSSS.
Association between whole BVL and physical disability in MS (less than five years follow-up, n = 21 studies)
Total: 980 RRMS: 794 PPMS: 23 SPMS: 163 Median EDSS at baseline: 2.5
Overall: 45.7
PBV and aPBV changes at 4.8 years follow-up Algorithm used: SIENA
EDSS change at 4.8 years of follow-up classified as 1) DP: An increase in EDSS by ≥1.5, ≥1.0 and ≥ 0.5 points if baseline EDSS was ≥0.0, < 5.0 and ≥5.5 points respectively 2) Stable: Patients who do not meet the criteria of DP
PBV loss and DP (MV analysis)Linear β = 0.33, 95% CI (0.16 to 0.49); P <0.001 • Experiencing DP over one year of follow-up was significantly associated with a 0.33% increase in whole BVL (measured using PBV change) from baseline to 4.8 yearsaPBV loss and DP* (UV analysis)aPBV loss among DP patients = – 0.99%aPBV loss among stable patients = – 0.79%,Cohen's d = 0.18, P< 0.001
•
Patients with DP had a significantly greater aPBV loss from baseline to follow-up when compared to stable patients.
Factors adjusted in MV analysis: Sex, age, and changes in MRI scanner strength* Analyses based on the sub-population of subjects with T2-FLAIR and 3-D T1-weighted image MRI measures; n = 149
MS patients with physical disability progression had a significantly greater BVL than stable patients.
Retrospective cohort Overall: NR DP: 4.8 years DI: 4.7 years Stable: 4.8 years
Total: 980 RRMS: 794 PPMS: 163 SPMS: 23 Median EDSS at baseline: NR
Overall: NR DP: 47.1 DI: 44.7
Stable: 45.4
PBV and aPBV changes at 4.7 years follow-up Algorithm used: 1. Baseline-FSL SIENA or SIENAX 2. Follow-up- SIENA
EDSS change at 4.7 years of follow-up classified as 1) DP: An increase in EDSS by 1.0 or 0.5 points if baseline EDSS was >5.5 points 2) DI: A decrease in EDSS by 1.0 and 0.5 points if the baseline EDSS was 2.0 to 5.5 and >5.5 points respectively 3) Stable: Non-occurrence of disability progression or disability improvement.
PBV loss and EDSS (MV analysis)ANCOVA analysis
•
The physical disability progression group had a significantly higher total PBV loss compared with both disability improvement and stable groups between baseline and 4.7 years of follow-up
aPBV loss and EDSS (MV analysis)ANCOVA analysis
•
The physical disability progression group had a significantly higher aPBV loss compared with both disability improvement and stable groups during 4.7 years of follow-up
*Bonferroni adjusted pair-wise comparison†Factors adjusted in MV analysis: The differences between the groups were calculated usingANCOVA adjusted for age at first MR imaging, race, T2-LV, corresponding baseline structural volume, and EDSS.
MS patients in the physical disability progression group had a significantly higher BVL than both disability improvement and stable groups over 4.7 years of follow-up.
Total: 870 CIS: 56 RRMS: 634 PPMS: 31 SPMS: 149 Median EDSS at baseline: 2.5
Overall: 49.8
PBV and aPBV changes at 4.6 years follow-up Algorithm used: SIENA
EDSS change at 4.6 years of follow-up classified as 1) DP: An increase in EDSS of ≥ 1.5, ≥1.0, and ≥0.5 points if the baseline EDSS was 0.0, ≥1.0 to 5.0, and ≥5.5 points respectively 2) No DP: Patients who do not meet the criteria for DP
BVL among DP and no DP groups* (UV analysis)
•
Percentage BVL was significantly greater among patients in the disease progression group compared to the no disease progression group from baseline to 4.6 years of follow-up
•
aPBV loss was not significantly different between the disease progression and no-disease progression groups from baseline to 4.6 years of follow-up
*P-value reported after including the covariate “MRI scanner change”
MS patients with disease progression exhibited greater BVL when compared to MS patients without disease progression.
Comparing longitudinal brain atrophy measurement techniques in a real-world multiple sclerosis clinical practice cohort: towards clinical integration?.
Total: 102 CIS: 1 RRMS: 99 SPMS: 2 Median EDSS at baseline: 2.0
Overall: 30.4
PBV and aPBV changes at 1-year follow-up Algorithm used: SIENA and icobrain
EDSS at baseline, 1 year and change from baseline to 1 year used as a continuous variable
PBV loss and physical disability – icobrain method (UV analysis)Kendall's τ = 0.148; P = 0.041
•
Increased BVL (measured by Icobrain long method) at 1-year follow-up was significantly associated with an increase (worsening) in EDSS delta from baseline to 1 year
PBV, aPBV losses and physical disability - SIENA and icobrain methods (UV analysis)Kendall's τ = NR; P = NR
•
SIENA and icobrain long measured annualized and non-annualized PBV changes measured at 1-year follow-up did not significantly correlate with EDSS at baseline or 1-year (quantitative data not reported)
Note: Results for the aPBV loss and EDSS change at one-year follow-up not reported
Increased BVL at 1 year was associated with an increase (worsening) in EDSS delta from baseline to 1 year in MS patients.
CIS: 253 RRMS: 708 SPMS: 128 PPMS: 125 Median EDSS at baseline: NR
Overall: NR CIS: 33 RRMS + SPMS: 39.7 PPMS: 48.5
Annual rate of change in the number of atrophied regions (cortex, deep gray matter, brainstem, white matter, cerebellum, and lateral ventricles) over 2.43 years Algorithm used: Desikan-Killiany-Tourville protocol
Annualized EDSS change over 2.43 years used as a continuous variable
Change in the number of atrophied regions and EDSS (MV analysis)RRMSLinear β = 0.03; P <0.0001
•
Every unit increase in the annual rate of the number of atrophied regions (measured using annualized increase in event-based model stage) was significantly associated with annual increase of 0.03 units of EDSS
SPMS and PPMSLinear β = NR; P = NR
•
There was no association between the annual rate of change in the number of atrophied regions (measured using annualized increase in event-based model stage) and annualized EDSS changes (quantitative data not reported)
Factors adjusted in MV analysis: Total intracranial volume, age at study entry, sex, scanner magnetic field, MRI protocol and disease duration.
Increased BVL was associated with physical disability progression in RRMS patients but not in SPMS and PPMS patients.
Total: 38 RRMS: 36 SPMS: 2 Median EDSS at baseline: 4.0
Overall: 36.8
PBPF change measured annually for 3 years Algorithm used: Volume measure was obtained on the MPRAGE images using the longitudinal stream included in the FreeSurfer analysis suite and the VBM (v 8) approach implemented in the SPM (v 8) software.
EDSS at the third year was classified as EDSS worsening- An increase in 1.0 and 0.5 points, when baseline EDSS was <5.5 and ≥5.5 points respectively or sustained up until the 3-year follow-up time point. EDSS stable or improved: Definition not reported Baseline EDSS used as a continuous variable
PBPF change from 2ndto 3rd-year and EDSS at 3rd-year follow-up (UV analysis)
•
Patients with EDSS worsening at 3 years follow-up had a significantly greater BVL (as measured using PBPF loss) from 2nd to 3rd
Changes in PBPF and baseline EDSS (UV analysis)
•
Baseline EDSS did not correlate with change in PBPF from baseline to 1st year or 1st to 2nd-year or 2nd to 3rd -year (quantitative data not reported)
Note: Results for the association between PBPF change from baseline to 1st year or 1st to 2nd -year and physical disability status at the third year was not reported
MS patients with EDSS worsening had a significantly greater BVL over 3 years of follow-up than patients with stable or improved EDSS.
Total: 21 RRMS: 14 SPMS: 7 Median EDSS at baseline: 4.0
Overall: 48.0
WBV change at 1-year follow-up Algorithm used: MSmetrix method
Baseline EDSS used as a continuous variable
WBV loss and EDSS (MV analysis)Hierarchical linear regression R2 = 0.48, P = 0.01
•
Baseline EDSS, age, and sex predicted around 50% of the variance in the whole brain volume change from baseline to 1-year follow-up; regression coefficient for the change in whole brain volume was not reported
Factors adjusted in MV analysis: Age and sex
Baseline EDSS, age and sex predicted around 50% of the variance in the whole brain volume change from baseline to 1-year follow-up in MS patients.
Total: 120 CIS: 63 RRMS: 53 SPMS: 4 Median EDSS at baseline: NR
Overall: NR CIS: 32.6 MS: 35.0
PBV change at 3.5 years follow-up and categorized into quartiles Algorithm used: SIENA
EDSS at baseline and 3.5 years used as a continuous variable
PBV loss and EDSS (UV analysis)EDSS at baseline:-Pearson's r = − 0.219; P = 0.016EDSS at 3.5 years follow-up:-Pearson's r = − 0.201; P = 0.028
•
BVL (measured using PBV change) from baseline to 3.5 years was weakly correlated with higher (worse) EDSS scores at baseline and 3.5 years follow-up
PBV loss in quartiles and EDSS (UV analysis)
•
Patients in different quartiles of BVL categories from baseline to 3.5 years of follow-up did not have significant differences in their median EDSS scores at baseline; P = NR
•
Patients in the quartiles with higher BVL (Q1 and Q2) had a significantly higher (worse) median EDSS score at 3.5 years of follow-up compared to patients in the quartiles with lower BVL (Q3 and Q4); P = 0.01
Higher PBV loss was generally associated with more active disease; however, it did not serve to predict the course of MS on an individual basis.
Total: 352 RRMS: 256 SPMS: 73 PPMS: 23 Median EDSS at baseline: Overall NR
Overall: 44.0
aPBV change at 2 years follow-up Algorithm used: SIENA
EDSS at 2-years follow-up classified as DP: Increase in EDSS by 1.0 and 0.5 points from baseline ≤5.5 and >5.5 points respectively No DP: Patients not meeting the disease progression criteria
aPBV loss and EDSS (MV analysis)Logistic OR = 0.59, 95% CI: (0.34 to 1.04); P = 0.07
•
Every 1% increase in aPBV loss from baseline to 2 years was associated with 1.7 times increased risk of DP at 2 years follow-up but the association was not statistically significant
Factors adjusted in MV analysis: MS center.
PBV loss was not a predictor of physical disability progression at 2 years in MS patients.
aPBV (cut-off: –0.86%) and PBPV changes at 1-year follow-up Algorithm used: SIENA
EDSS at 4 years follow-up was classified as Confirmed worsening: An increase in 1.0 and 0.5 points if prior EDSS was ≤5.0 and ≥5.5 points respectively confirmed at a further visit with a time gap of at least 6 months Progression-free patients: Patients who do not meet the criteria for confirmed worsening at 4 years
aPBV loss and disability (MV analysis)Cox HR = 4.647, 95% CI (1.479 to 14.603); P = 0.009
•
Patients losing more than 0.86% of brain volume between baseline and 1 year had 4.7 times higher risk of developing confirmed EDSS worsening at 4-years follow-up
PBPV loss and disability (UV analysis)
•
PBPV loss at 1-year follow-up was not significantly associated with confirmed EDSS worsening at 4-years follow-up
Factors adjusted in MV analysis: Age, sex, time from the first relapse to treatment, annualized relapse rate before treatment, presence of further MS attacks within the first year of follow-up, number of new or enlarging lesions (NEL) after 12 months of therapy, presence of contrast-enhancing lesions (CEL) in either baseline or follow-up MRI scan, and baseline volume values.
Increased BVL in the first year of interferon-beta therapy was predictive of subsequent physical disability progression in RRMS patients.
Total: 62 RRMS: 59 SPMS: 3 Median EDSS at baseline: 4.0
Overall: 34.7
aPBV change from baseline to 1-, 2- and 3-years follow-up Algorithm used: SIENA
EDSS measured at 1-, 2- and 3-years follow-up was classified as EDSS worsening: An increase in EDSS of 1.0 and 0.5 points if baseline EDSS was <5.5 and ≥5.5 points respectively. Stabilization / improvement: Any other EDSS changes.
aPBV loss at 1-year and EDSS at 1-year follow-up (MV analysis)Binomial logistic OR = 1.7, 95% CI (0.91 to 3.47); P = 0.094
•
Every 1 per cent increase in BVL from baseline to 1 year was associated with a 1.7 times higher risk of EDSS worsening at 1 year; however, this association was not statistically significant
aPBV loss at 1-year and EDSS at 2-, 3-years follow-up (MV analysis)Quantitative data NR • No significant association was found between percentage brain volume changes from baseline to 1-year follow-up and disability status at 2 years (P = 0.242) or 3 years (P = 0.1)aPBV loss at 2-years and EDSS at 2-, 3-years follow-up (MV analysis)Quantitative data NR
•
No significant association was found between percentage brain volume changes between baseline to 2 years follow-up and disability status at 2 years (P = 0.366) or 3 years (P = 0.217)
aPBV loss at 3-years and EDSS at 3-years follow-up (MV analysis)Binomial logistic OR = 1.55, 95% CI (0.94 to 2.54); P = 0.084
•
Every 1 per cent increase in BVL from baseline to 3 years was associated with a 1.55 times higher risk of EDSS worsening at 3 years; however, this association was not statistically significant
Factors adjusted in MV analysis: Baseline EDSS, age, disease duration, and the number of Gd-enhancing lesions at baseline.
No significant association was observed between BVL occurring during natalizumab therapy and physical disability progression in MS patients.
Visual evoked potential and magnetic resonance imaging are more effective markers of multiple sclerosis progression than laser polarimetry with variable corneal compensation.
Total: 41 CIS: 2 RRMS: 37 SPMS: 2 Median EDSS at baseline: NR
Overall: NR MS patients with optic neuritis: 34.0 MS patients without optic neuritis: 37.6
WBV loss at 1-year follow-up Algorithm used: Semi-automated radiological protocol based on the manual segmentation technique
Baseline EDSS used as a continuous variable
WBV loss and EDSS (UV analysis)MS patients with optic neuritis-Spearman's r = 0.42, P = 0.006MS patients without optic neuritis-Spearman's r = 0.5, P = 0.01
•
An increase in BVL from baseline to 1 year was significantly associated with a higher (worse) baseline EDSS score in MS patients with or without optic neuritis
An increase in BVL from baseline to 1 year was associated with a higher (worse) baseline EDSS score in MS patients.
PWBV change measured at baseline, 6 months and yearly thereafter until 4 years follow-up. Algorithm used: SIENA
EDSS at 4 years follow-up was classified as SDP: An increase of EDSS 1.0 and 1.5 if the baseline EDSS was > 0.0 and 0.0 sustained over at least 24 weeks after the end of the study respectively. SDI: At least a 1.0-point decrease in EDSS. Conversion from CIS to CDMS used as a binary variable
BVL and EDSS (MV analysis)*P<0.05 for differences between SDP and stable plus SDI group in the confirmatory analysis that excluded MRI data between baseline and 6 months
•
There was a significant decrease in the PWBV (P<0.05) among CIS patients who developed SDP compared to patients who remained stable or improved in their disability status over 4-years follow-up
Conversion from CIS to CDMS (MV analysis)
•
There was a significant decrease in the PWBV (P = 0.007) among converters to CDMS compared to non-CDMS converters over 4-years follow-up
Factors adjusted in MV analysis: Age, sex, time from the first event to baseline assessment, and treatment status
Whole-brain atrophy was associated with SDP and conversion to CDMS in patients with CIS over 4 years.
PBV and BPF changes at 1-year follow-up Algorithm used: 1. PBV change- SIENA 2. BPF- Semi-automated algorithm implemented on SPM (v5) and MRIcro software
EDSS at 1 and 2 years and EDSS change from 0-1 and 0-2 years used as a continuous variable
PBVloss and EDSS (UV analysis)Spearman's r = – 0.198; P = 0.009 • Increased BVL (measured using PBV change) from baseline to 1-year follow-up was associated with higher (worse) EDSS scores at 1 yearPBVloss and EDSS change (UV analysis)Spearman's r = –0.195; P = 0.011
•
Increased BVL (measured using PBV change) from baseline to 1-year follow-up was associated with an increase (worsening) in EDSS delta from baseline to 1-year follow-up
BPF loss and EDSS change (UV analysis)Spearman's r = –0.188; P = 0.017
•
Increased BVL (measured using BPF change) from baseline to 1-year follow-up was associated with an increase (worsening) in EDSS delta from baseline to 2 years follow-up
Note: Correlation between 1) BPF loss and EDSS at 1 or 2 years or EDSS change from 0-1 year, 2) PBV loss and EDSS at 2 years or EDSS change from 0 - 2 years were not reported
Very early brain atrophy development was associated with an increased physical disability in the first and second years in CIS patients.
Total: 39 Subtypes not mentioned Median EDSS at baseline: 4.0
Overall: 35.0
PBV change for the 1st and 2nd year follow-up Algorithm used: SIENA
EDSS change from 0-1 and 1-2 years used as a continuous variable
PBV loss and EDSS (UV analysis)Spearman's r = NR
•
BVL (measured using PBV change) occurring during the first or second years of follow-up did not correlate with EDSS changes during the 1st or 2nd -years of follow-up, respectively (quantitative data not reported)
BVL occurring during the first or second years of follow-up was not associated with physical disability progression in MS patients.
Overall: NR Stable CIS: 29 Progressing to CDMS: 28
PWBV change at 6 months follow-up Algorithm used: SIENAX
Conversion from CIS to CDMS used as a binary variable
PWBV loss in CIS and CDMS groups (MV analysis)
•
Relative change in PWBV over 6 months showed a consistent trend to decrease faster in the CDMS group, this trend did not reach the level of statistical significance (P>0.1, mixed models)
Factors adjusted in MV analysis: Age, sex, and treatment changes
BVL over 6 months showed a consistent trend to decrease faster in the CDMS group, however, this was not statistically significant.
PBV change from baseline at 6-monthly intervals until 1.5 years Algorithm used: SIENA
Baseline EDSS as a continuous variable
PBV loss and EDSS (UV analysis)P-value: not significant
•
There were no significant correlations between BVL (measured using PBV change) during baseline and 1.5 years; baseline and 0.5 years; 0.5 years and 1-year; and 1-year and 1.5- years and EDSS at the baseline
No correlation was found between brain atrophy over 1.5 years and EDSS at baseline in RRMS patients.
Total: 84 RRMS: 57 PPMS: 20 SPMS: 7 Median EDSS at baseline: 3.2
Overall: 42.6
BPF and standardized BPF changes over 3.2-years follow-up Algorithm used: BPF -automated template-driven segmentation (TDS+) from the dual echo images
Probability of developing SDP over the 3.2 -years follow-up SDP: EDSS worsening (an increase of 1.0 or 0.5 points in EDSS if the baseline score was <6.0 and ≥6.0 points respectively) sustained for at least three months.
BPF change and SDP in the whole cohort (MV analysis)Quantitative data not reported.
•
BVL (measured using BPF change) was not found to be a significant predictor for developing SDP during the 3.2-years follow-up (quantitative data not reported)
BPF change and SDP in RRMS and SPMS patients (MV analysis)Quantitative data not reported.
•
BVL (measured using BPF change) was not found to be a significant predictor for developing SDP during the 3.2-years follow-up in the RRMS and SPMS patients (quantitative data not reported)
Standardized BPF change and SDP in the whole cohort (MV analysis)Logistic OR = 0.46, 95% CI (0.08 to 2.73); P = 0.39
•
BVL (measured using standardized BPF change) was not found to be a significant predictor for developing SDP during the 3.2-years follow-up.
Standardized BPF change and SDP in RRMS and SPMS patients (MV analysis)Logistic OR = NR; P>0.1
•
BVL (measured using standardized BPF change) was not found to be a significant predictor for developing SDP during the 3.2 years in the RRMS and SPMS patients.
Factors adjusted in MV analysis: Length of follow-up, age, and baseline EDSS.
BVL was not found to be a significant predictor for developing SDP during 3.2 -years of follow-up in MS patients.
Total: 390 The sample size of subtypes not reported Median EDSS at baseline: NR
Overall: NR
aPBV change at 14-months follow-up Algorithm used: SIENA
EDSS at baseline and annualized change in EDSS used as a continuous variable at 14-months follow-up
aPBV loss and EDSS at baseline (UV analysis)Spearman's r = –0.12; P <0.005
•
An increase in annualized BVL (measured using aPBV change) was associated with a significantly higher (worse) EDSS score at the baseline
aPBV loss and physical disability by sub-types
•
The correlation between annualized BVL (measured using aPBV change) and baseline EDSS for subgroups of patients (CIS/RRMS/PPMS/SPMS) was weak or absent (quantitative data not reported)
aPBV loss and annualized EDSS change (overall and by sub-types) (UV analysis)Quantitative data not reported
•
The correlation between annualized BVL (measured using aPBV change) and annualized EDSS change in overall MS patients and subgroups (CIS/RRMS/PPMS/SPMS) was weak or absent over 14-months follow-up
Increased annualized BVL was weakly correlated with higher baseline physical disability in MS patients. Correlation between annualized BVL and disability by sub-types of MS was weak or absent.
Brain volume loss and physical disability (greater than or equal to five years of follow-up; n =15)
•
Of the 15 studies investigating the association between whole brain volume change and EDSS, 13 found a significant relationship such that an increase in the whole BVL over the period of observation was associated with physical disability progression (
Of the three studies investigating the association between whole brain volume change and MSSS, two found that an increase in the BVL was significantly associated with a higher (worse) MSSS score (
Of the three studies investigating the association between whole brain volume change and MSFC or component performance measures, none reported a significant association between them (
One study that reported on the association between whole brain volume change and conversion from CIS to CDMS showed a significant association such that an increased BVL was associated with a greater risk of conversion from CIS to CDMS (
Brain volume loss and physical disability (less than five years of follow-up; n=21)
•
Of the 20 studies reporting on the association between whole brain volume change and EDSS, 13 found a significant relationship such that an increase in the whole BVL was associated with a worsening of EDSS (
Comparing longitudinal brain atrophy measurement techniques in a real-world multiple sclerosis clinical practice cohort: towards clinical integration?.
Visual evoked potential and magnetic resonance imaging are more effective markers of multiple sclerosis progression than laser polarimetry with variable corneal compensation.
Of the two studies reporting on the association between whole brain volume change and conversion from CIS to CDMS, one showed a significant association such that an increased BVL was associated with a greater risk of conversion from CIS to CDMS (
Enlargement of the ventricular CSF space provides an alternative measure to brain parenchymal atrophy for assessment of neurodegeneration.
Table 3 presents an overview of the 15 primary studies reporting data on the association between ventricular enlargement and physical disability in MS. Of these 15 studies, six were prospective and nine were retrospective. The follow-up duration varied from two years to 30 years with a sample size ranging from 37 to 1651. The average age of patients across the 15 studies ranged from 32 to 50 years. The baseline median: i) EDSS reported in nine studies ranged from 1.5 – 4.0 points; ii) MSSS reported in two studies was 3.7 and 5.4. Fourteen out of 15 studies reported ventricular “volume” change whereas one study (
) reported change in the widths (i.e., a “linear” measure) of the ventricle. The studies in Table 3 have been arranged under three heads based on the type of ventricular atrophy, which was measured using the atrophy of total ventricle (four studies), lateral ventricle (ten studies), and third ventricle (two studies). One study (
A longitudinal observational study of brain atrophy rate reflecting four decades of multiple sclerosis: a comparison of serial 1D, 2D, and volumetric measurements from MRI images.
) provided data on both lateral and third ventricular atrophy and is therefore repeated under both heads. Under each head, the studies are arranged in reverse chronological order, starting with the most recently published.
Table 3Association between ventricular (total, lateral, and third) atrophy and physical disability in MS.
Reference (Country); Quality rating
Study design / Duration
Study population
Average age (years)
BVL measurement
Physical disability measurement
Results
Conclusion
Association between total ventricular atrophy and physical disability in MS (n = 4 studies)
Retrospective cohort 6- and 12- years follow-up groups
6 years follow-up group Total: 115 CIS: 36 RRMS: 68 PPMS: 11 12 years follow-up group Total: 79 CIS: 27 RRMS: 50 PPMS: 2 Median EDSS at baseline: 2.0
6 years follow-up group: 35.3 12 years follow-up group: 34.9
aPVV change at 2 years follow-up Algorithm used: SIENA and VIENA
EDSS at 6-and 12- years follow-up used as a continuous variable
aPVV change (from baseline to year 2) among 6-year follow-up group (UV analysis)Linear standardized β = 0.048, P = 0.609
•
Every 1 standard deviation increase in aPVV (BVL) over 2 years was associated with a 0.048 standard deviation higher EDSS score at 6-year follow up. However, the association was not statistically significant.
aPVV change (from baseline to year 2) among 12-year follow-up group (UV analysis)Linear standardized β = 0.064, P-value =0.576
•
Every 1 standard deviation increase in aPVV (BVL) over 2 years was associated with a 0.064 standard deviation higher EDSS score at 12-year follow up. However, the association was not statistically significant.
Ventricular enlargement was not associated with physical disability progression in MS patients
Total: 1465 CIS: 124 RRMS: 1089 SPMS: 217 PPMS: 35 Median EDSS at baseline: NR
Overall: NR DP converters: 47.6 DP non-converters: 44.3
aPVV change at 5 years follow-up Algorithm used: 1. Ventricular volume baseline-SIENAX 2. Change in ventricular volume- VIENA
EDSS at 5-year follow-up classified as DP converters: Increase in EDSS by ≥ 1.5, ≥ 1.0, and ≥ 0.5 from baseline <1, 1 to 5.5, and >5.5 points respectively DP non-converters: Patients not meeting the disease progression criteria
aPVV change and DP conversion (UV analysis)DP converters vs DP non-converters: 3.7% vs 3.4%Difference between DP converters and DP non-converters: 0.36%, 95% CI (–0.93 to 1.65); Cohen's d = 0.04; P = 0.89
•
There was no significant difference in aPVV change between DP converters and DP non-converters at 5 years of follow-up
aPVV change between DP converters and DP non-converters was not significantly different in MS patients.
Total: 261 CIS: 18 RRMS: 97 PPMS: 69 SPMS: 77 Median EDSS at baseline: 4.0 Median MSSS at baseline: 5.4
Overall: 43
aPVV change at 1-2 years follow-up Algorithm used: SIENAX and SIENA
EDSS and MSSS at 10 years used as a continuous variable
aPVV change and EDSS (MV analysis)Linear β = 0.07, 95% CI (0.03 to 0.11)
•
Every 1 per cent increase in central brain atrophy (measured using aPVV change) over 1-2 years was significantly associated with a 0.07-points higher EDSS score at 10 years.
aPVV change and MSSS (MV analysis)Linear β = 0.06, 95% CI (0.01 to 0.11)
•
Every 1 per cent increase in central brain atrophy (measured using aPVV change) over 1-2 years was significantly associated with a 0.06-points higher (worsening) MSSS score at 10 years.
Factors adjusted in MV analysis: Imaging protocol, disease type and EDSS at baseline
Early brain atrophy rates were related to subsequent long-term physical disability in MS patients.
Total: 54 RRMS: 43 PPMS: 11 Median EDSS at baseline: 2.0
Overall: 35
aPVV change at 2- and 5.5-years follow-up Algorithm used: SIENA
Annualized EDSS change from baseline to 2 years, 5.5 years and from 2 years to 5.5 years as a continuous variable and further classified as DP: An increase in EDSS of at least 1.0 points after 5.5 years of follow-up. No DP: Patients not meeting the criteria for DP
BVL and disability at 5.5 years (UV analysis)Median (IQR)
•
Central BVL (measured using aPVV change) from baseline to 2-years was significantly higher in the DP group compared to the no DP group over 5.5 years
•
There was no significant difference in aPVV change from 2 to 5.5-years between DP and no DP groups over 5.5 years
aPVV change and annual EDSS change (UV analysis)Spearman's r (P-value)
•
An increase in central BVL (measured using aPVV change) from baseline to 2-years was associated with an increase (worsening) in annualized EDSS change from baseline to 2-years follow-up.
•
Central BVL (measured using aPVV change) from baseline to 2-years was not associated with annualized EDSS change from 2 to 5.5-years follow-up.
•
Central BVL (measured using aPVV change) from 2 to 5.5-years was not associated with annualized EDSS change from baseline to 2 years or from 2 to 5.5-years follow-up.
An increase in central BVL (measured using aPLVV change) from baseline to 2-years was significantly associated with an increase (worsening) in annualized EDSS change from baseline to 5.5 years follow-up
Note: Result for aPVV change from 2 to 5.5 years and EDSS change from baseline to 5.5 years was not reportedaPVVchangeand DP after 5.5 years (MV analysis)Logistic OR = 1.17 95% CI (1.01 to 1.35); P = 0.035
•
Every 1% increase in central BVL (measured using aPLVV change) from baseline to 2 years was associated with a 1.17 times higher risk of developing DP at 5.5 years follow-up
Factors adjusted in MV analysis: Age, disease duration, onset type, sex, baseline EDSS score, presence of treatment, T1 lesion load, T2 lesion load, and gadolinium-enhancing lesion load at baseline, annual percentage brain volume change, annual percentage ventricular volume change within the first 2-years and the change in T1 lesion load and T2 lesion load during the first 2- years
The rate of early ventricular enlargement was a strong MRI predictor for medium-term physical disability progression in MS patients.
Association between lateral ventricular atrophy and physical disability in MS (n = 10 studies)
PLVV and aPLVV, changes at 4.8 years follow-up Algorithm used: SIENA
EDSS change at 4.8 years of follow-up classified as 1) DP: An increase in EDSS by ≥ 1.5, ≥ 1.0 and ≥ 0.5 points if baseline EDSS was ≥ 0.0, < 5.0 and ≥ 5.5 points respectively 2) Stable: Patients who do not meet the criteria of DP
PLVV change and DP (MV analysis)Linear β = –4.48, 95% CI (–5.36 to –3.59); P<0.001
•
Experiencing DP over one year of follow-up was significantly associated with a 4.48% increase in PLVV (increased BVL) from baseline to 4.8 years of follow-up
aPLVV change and DP (UV analysis)
•
Patients with DP had a significantly greater aPLVV change (4.1% vs. 2.3%, Cohen's d = 0.27, P < 0.001) from baseline to 4.8 years of follow-up when compared to stable patients.
Factors adjusted in MV analysis: Sex, age, and changes in MRI scanner strength
MS patients with physical disability progression had a significantly greater BVL than stable patients.
Every 1.0-unit higher EDSS at baseline was significantly associated with a 1.469 ml increase in LVV (increased BVL) during 9 years of follow-up
Factors adjusted in MV analysis: Age, sex, education, disease course (relapsing-remitting vs progressive), disease duration, use of disease-modifying therapy, trait conscientiousness, whether conscientiousness was measured from informant report, and MRI scanner field strength
A higher EDSS score at baseline was associated with an increased BVL from baseline to 9 years in MS patients.
Retrospective cohort Overall: NR DP: 4.8 years DI: 4.7 years Stable: 4.8 years
Total: 980 RRMS: 794 PPMS: 163 SPMS: 23 Median EDSS at baseline: NR
Overall: NR DP: 47.1 DI: 44.7
Stable: 45.4
PLVV and aPLVV changes at 4.7 years follow-up Algorithm used: 1. Baseline-SIENA or SIENAX 2. Follow-up- SIENA
EDSS change at 4.7 years of follow-up classified as 1) DP: An increase in EDSS by 1.0 or 0.5 points if baseline EDSS was >5.5 points 2) DI: A decrease in EDSS by 1.0 and 0.5 points if the baseline EDSS was 2.0 to 5.5 and >5.5 points respectively 3) Stable: Non-occurrence of disability progression or disability improvement.
PLVV change and physical disability progression (MV analysis)ANCOVA analysis
•
The physical disability progression group had a significantly higher PLVV change (BVL) compared with both disability improvement and stable groups between baseline and 4.7 years of follow-up
aPLVV change and physical disability progression (MV analysis)ANCOVA analysis
•
The physical disability progression group had a significantly higher aPLVV change (BVL) compared with both disability improvement and stable groups during 4.7 years of follow-up
*Bonferroni adjusted pair-wise comparisonFactors adjusted in MV analysis: The differences between the groups were calculated using ANCOVA adjusted for† Age at first MRI, race, T2-LV, corresponding baseline structural volume, and EDSS.
MS patients in the physical disability progression group had a significantly higher BVL than both disability improvement and stable groups over 4.7 years of follow-up.
Total: 870 CIS: 56 RRMS: 634 PPMS: 31 SPMS: 149 Median EDSS at baseline: 2.5
Overall: 49.8
PLVV and aPLVV changes at 4.6 years follow-up Algorithm used: NeuroSTREAM (2D FLAIR scans)
EDSS change at 4.6 years of follow-up classified as 1) DP: An increase in EDSS of ≥ 1.5, ≥1.0, and ≥0.5 points if the baseline EDSS was 0.0, ≥1.0 to 5.0, and ≥5.5 points respectively 2) No DP: Patients who do not meet the criteria for DP
BVL among DP and no DP groups* (UV analysis)
•
PLVV change and aPLVV change were not significantly different between the disease progression and no-disease progression groups from baseline to 4.6 years of follow-up
*P-value reported after including the covariate “MRI scanner change”
There was no significant difference in lateral ventricular atrophy between MS patients with and without disease progression.
Total: 1651 CIS: 137 RRMS: 1219 PPMS: 255 SPMS: 40 Median EDSS at baseline: 2.5 Median MSSS at baseline: 3.7
Overall: NR MS: 45.6 CIS: 39.5
aLVV change at 5 years follow-up Algorithm used: Baseline LVV NeuroSTREAM software
Annual change in EDSS and MSSS as a continuous variable and EDSS was further classified as DP: An absolute change in EDSS from the first to most recent follow-up MRI with an increase in EDSS of ≥1.5, ≥1.0 and ≥0.5 points if the baseline EDSS was 0.0, 1.0 to 5.0, ≥5.5 points respectively was confirmed after at least 24 weeks. Non-DP/ Stable- Not defined
aLVV change and EDSS change (UV analysis)Linear mixed-effect model β = 0.098; P = 0.0001
•
Every one-point increase in EDSS score annually was significantly associated with a 0.098 ml increased BVL (measured using aLVV change) at 5 years follow-up
aLVV change and MSSS change (UV analysis)Linear mixed-effect model β = 0.078; P = 0.0001
•
Every one-point increase (worsening) in MSSS score annually was significantly associated with a 0.078 ml increased BVL (measured using aLVV change) at 5 years follow-up
aLVV change and DP (UV analysis)Linear mixed-effect model β = 0.105 (21.6%); P = 0.001
•
MS patients with DP had a 21.6% increased rate of BVL (measured using aLVV change) compared to those without DP at 5 years follow-up
Increased BVL was associated with physical disability progression in MS patients.
Total: 81 RRMS: 62 SPMS: 11 PPMS: 8 Median EDSS at baseline: 3.5
Overall: 42
LVV change from baseline to 5-year or from baseline to 10-year follow-up Algorithm used: 1. Normalized brain volume-SIENAX 2. Longitudinal brain volume changes- SIENAX multi-timepoint (SX-MTP)
EDSS at 5- and 10-year follow-up classified as DP: An increase of EDSS by 1.0 and 0.5 points if baseline EDSS was <6.0 and ≥6.0 respectively No DP: Patients who do not meet the criteria for physical disability progression
LVV change and EDSS (MV analysis)Quantitative estimates not reported
•
There was no significant association between lateral ventricular volume change and physical disability progression either at 5 years or 10 years (quantitative data not reported)
Factors adjusted in MV analysis: Age, sex, scanner type, baseline volumes of specific structures, T1-lesion volumes, T2-lesion volumes, MS subtype, disease duration and use of disease-modifying treatment.
Lateral ventricular volume change was not associated with physical disability progression in MS patients.
PLVV change measured at baseline, 6 months and yearly thereafter until 4 years follow-up. Algorithm used: Modified SIENAX multi-time point algorithm
EDSS at 4 years follow-up was classified as SDP: An increase of EDSS 1.0 and 1.5 if the baseline EDSS was > 0.0 and 0.0 sustained over at least 24 weeks after the end of the study, respectively. SDI: At least a 1.0-point decrease in EDSS. Conversion from CIS to CDMS used as a binary variable
BVL and EDSS (MV analysis)*P<0.05 for differences between SDP and stable plus SDI group in the confirmatory analysis that excluded MRI data between baseline and 6 months
•
There was a significant increase in PLVV (P< 0.001) among CIS patients who developed SDP compared to patients who remained stable or improved in their physical disability status over 4-years follow-up
•
Conversion from CIS to CDMS (MV analysis)
•
There was a significant increase in PLVV (P = 0.025) among converters to CDMS compared to non-CDMS converters over 4-years follow-up
Factors adjusted in MV analysis: Age, sex, time from the first event to baseline assessment, and treatment status
Lateral ventricular volume enlargement was associated with SDP and conversion to CDMS in patients with CIS over 4 years.
A longitudinal observational study of brain atrophy rate reflecting four decades of multiple sclerosis: a comparison of serial 1D, 2D, and volumetric measurements from MRI images.
Total: 37 RRMS: 16 PPMS: 4 SPMS: 17 Median EDSS at baseline: NR Median MSSS at baseline: NR
Overall: 42
aLVV change at 9.25 years follow-up Algorithm used: Semiautomatic tool in HERMES MultiModality, Region Growing method.
EDSS and MSSS at 9.25 years follow-up (Not clearly specified how EDSS & MSSS were categorized in logistic regression analysis)
aLVV change and EDSS (UV analysis)Exact logistic OR = 1.24, 95% CI (1.01 to 1.62); P = 0.037
•
A 0.01 cm3 increase in aLVV (BVL) was significantly associated with a 1.24 times higher risk of physical disability progression (measured by EDSS) over 9.25 years of follow-up
aLVV change and MSSS (UV analysis)Exact logistic OR = 1.46, 95% CI (1.09 to 2.21); P = 0.006A 0.01 cm3 increase in aLVV (BVL) was significantly associated with a 1.46 times higher risk of physical disability progression (measured by MSSS) over 9.25 years of follow-up
Lateral ventricular volume enlargement was associated with a higher risk of physical disability progression in MS patients.
Association between third ventricular atrophy and disability in MS (n = 2)
Total: 103 CIS: 27 RRMS: 34 SPMS: 26 MS-related deaths: 16 Median EDSS at baseline: NR
Overall: 31.7
Change in widths of the third ventricle at 5 years of follow-up – linear measure Algorithm used: Measured on axially acquired PD/T2-weighted MR scans.
EDSS at 30 years used as a continuous variable
Third ventricular atrophy and EDSS (MV analysis) Linear β = 0.2, 95% CI (0.013 to 0.384); P = 0.04 • Every 1 mm of third ventricular atrophy within the first 5 years was significantly associated with 0.2 points higher EDSS at 30 years Factors adjusted in MV analysis: Age, sex, EDSS change in the first 5 years
Brain atrophy early in the course of MS independently predicted progressive disease and physical disability after 30 years.
A longitudinal observational study of brain atrophy rate reflecting four decades of multiple sclerosis: a comparison of serial 1D, 2D, and volumetric measurements from MRI images.
Total: 37 RRMS: 16 PPMS: 4 SPMS: 17 Median EDSS at baseline: NR Median MSSS at baseline: NR
Overall: 42
aTVV change at 9.25 years follow-up Algorithm used: Semiautomatic tool in HERMES MultiModality, Region Growing method.
EDSS and MSSS at 9.25 years follow-up (Not specified how EDSS & MSSS were categorized in logistic regression analysis)
aTVV change and EDSS (UV analysis)Exact logistic OR = 1.36, 95% CI (0.99 to 1.95); P = 0.053
•
A 0.001 cm3 increase in aTVV (BVL) was associated with a 1.36 times higher risk of physical disability progression (measured by EDSS) over 9.25 years of follow-up, however, the association was not significant
aTVV change and MSSS (UV analysis)Exact logistic OR = 1.52, 95% CI (1.01 to 2.57); P = 0.044
•
A 0.001 cm3 increase in aTVV (BVL) was significantly associated with a 1.52 times higher risk of physical disability progression (measured by MSSS) over 9.25 years of follow-up
Annual third ventricular volume enlargement was associated with a higher risk of physical disability progression in MS patients.
3.3 Total ventricular atrophy and physical disability
•
Of the four studies reporting on the association between total ventricular volume change and EDSS, two found a significant relationship such that an increase in the total ventricular volume was predictive of physical disability progression (
One study reporting on the association between total ventricular volume change and MSSS found that early brain atrophy rate was significantly related to subsequent long-term physical disability (
3.4 Lateral ventricular atrophy and physical disability
•
Of the nine studies reporting on the association between lateral ventricular volume change and EDSS, six found that an increase in the lateral ventricular volume was significantly associated with physical disability progression (
A longitudinal observational study of brain atrophy rate reflecting four decades of multiple sclerosis: a comparison of serial 1D, 2D, and volumetric measurements from MRI images.
Two studies reporting on the association between lateral ventricular volume change and MSSS found that an increase in the lateral ventricular volume was significantly associated with physical disability worsening over time (
A longitudinal observational study of brain atrophy rate reflecting four decades of multiple sclerosis: a comparison of serial 1D, 2D, and volumetric measurements from MRI images.
Two studies reporting on the association between lateral ventricular volume change and conversion from CIS to CDMS found a significant increase in lateral ventricular volume among converters to CDMS compared to non-CDMS converters (
Of the two studies reporting on the association between third ventricular atrophy and EDSS, one found a significant association such that early ventricular atrophy was predictive of long-term physical disability progression (
One study reporting on the relationship between third ventricular volume change and MSSS found that third ventricle atrophy resulted in higher risk of physical disability progression (
A longitudinal observational study of brain atrophy rate reflecting four decades of multiple sclerosis: a comparison of serial 1D, 2D, and volumetric measurements from MRI images.
Table 4 presents an overview of the 13 primary studies reporting data on the association between gray matter (total gray matter, cortical gray matter, deep gray matter, and cortical thickness) atrophy and physical disability in MS. Of these 13 studies, nine were prospective cohort and four were retrospective cohort. The follow-up duration varied from one year to 10 years with a sample size ranging from 36 to 1214. The average age of patients across the 13 studies ranged from 31 to 45 years. The baseline median: i) EDSS reported in eight studies ranged from 1.5 to 4.0 points; ii) MSFC reported in one was –1.27. The studies in Table 4 have been arranged based on the type of gray matter atrophy, which was measured using the i) total gray matter volume (nine studies), ii) cortical gray matter volume (six studies), iii) deep gray matter volume (three studies), and iv) cortical thickness, a linear measure (three studies). The total exceeds 13 since eight studies provided data on multiple measures of gray matter atrophy. Under each head, the studies are arranged in reverse chronological order, starting with the most recently published.
Table 4Association between relative gray matter (total gray, cortical gray or deep gray volumes or cortical thickness) atrophy and physical disability in MS.
Reference (Country); Quality rating
Study design / Duration
Study population
Average age (years)
BVL measurement & algorithm used
Physical disability measurement
Results
Conclusion
Association between total gray matter atrophy and disability in MS (n = 9 studies)
Total: 38 RRMS: 36 SPMS: 2 Median EDSS at baseline: 4.0
Overall: 36.8
PGMF and PGMV changes, measured annually for 3 years Algorithm used: Longitudinal volume changes assessed using longitudinal processing stream implemented in VBM (v 8)
EDSS at the third year was classified as EDSS worsening- An increase in 1.0 and 0.5 points, when baseline EDSS was <5.5 and ≥5.5 points respectively or sustained up until the 3-year follow-up time point. EDSS stable or improved: Definition not reported Baseline EDSS used as a continuous variable
PGMF loss from 2ndto 3rd-year and EDSS at 3rd-year follow-up (UV analysis)
•
Patients with EDSS worsening at 3 years follow-up had a significantly greater BVL (as measured using PGMF) from 2nd to 3rd -year follow-up than patients with stable or improved EDSS
PGMV loss from baseline to 1 year and EDSS at 3rd-year follow-up (UV analysis)
•
Patients with EDSS worsening at 3 years follow-up had a significantly greater BVL (measured using total gray matter volume loss) during the first-year follow-up than patients with stable or improved EDSS
PGMV loss from 1stto 2ndyear or 2ndto 3rdyear and EDSS at 3rd-year follow-up (UV analysis)
•
No significant associations were found between PGMV change during the second or third 1-year periods and EDSS worsening at 3rd -year (quantitative estimates not reported)
Change in PGMF or PGMV and baseline EDSS (UV analysis)Spearman's r = NR
•
Baseline EDSS was not correlated with change in PGMF or PGMV from baseline to 1st year or 1st to 2nd -year or 2nd to 3rd -year (quantitative data not reported)
Note: Results for the association between PGMF change from baseline to 1st year or 1st to 2nd -year and disability status at the third year were not reported.
MS patients with EDSS worsening at 3rd – year follow-up had significantly greater gray matter volume loss than patients with stable or improved EDSS.
Short-term MRI measurements as predictors of EDSS progression in relapsing-remitting multiple sclerosis: grey matter atrophy but not lesions are predictive in a real-life setting.
Change in total GMV from baseline and between 5-14 months follow-up Algorithm used: SIENA and FreeSurfer (v 5.2.0)
EDSS change at 3 years follow-up used as a continuous variable
Total GMV loss and EDSS change (MV analysis)Linear β = 0.129; P = 0.003
•
Every 1 per cent increase in BVL (measured using total GMV loss) from baseline to 5-14 months follow-up was significantly associated with a 0.129-points increase (worsening) in EDSS at 3 years of follow-up
Factors adjusted in MV analysis: Treatment status, sex, age, number of T1-, T2- and Gd-lesions, absolute change of lesion numbers and SIENAX volumes from visit 1 to visit 2, aPBV change and global atrophy measurements from the longitudinal FreeSurfer processing.
Short-term (5-14 months from baseline) gray matter atrophy was predictive of physical disability progression at 3 years of follow-up in RRMS patients.
PGMV change at 1-year follow-up Algorithm used: In-house built semi-automated algorithm implemented on SPM (v 8) and MRIcro software
EDSS at 4 years follow-up was classified as Confirmed worsening: An increase in 1.0 and 0.5 points if prior EDSS was ≤5.0 and ≥5.5 points respectively confirmed at a further visit with a time gap of at least 6 months Progression-free patients: Patients who do not meet the criteria for confirmed worsening at 4 years
PGMV loss and physical disability (UV analysis)
•
PGMV loss at 1-year follow-up was not significantly associated with confirmed EDSS worsening at 4-years follow-up
PGMV loss was not significantly associated with confirmed EDSS worsening at 4-years follow-up in RRMS patients.
PPMS: 36 Median EDSS at baseline: 4.0 Median MSFC at baseline: –1.27
Overall: 42.8
Rate of change in GMV at 5 years follow-up Algorithm used: SIENAX FSL (v 4.1.9)
EDSS and MSFC at 5 years follow-up, annualized EDSS (step-change) and MSFC change used as continuous variables (step-change details not reported)
Rate of GMV loss and EDSS (UV analysis)Pearson's r and P NR
•
There was no significant correlation between annual BVL (measured using the rate of annualized GMV loss) over 5 years and EDSS at 5 years or annualized EDSS step-change over 5 years (quantitative data not reported)
Rate of GMV loss and MSFC (UV analysis)Pearson's r and P NR
•
There was no significant correlation between annual BVL (measured using the rate of annualized GMV loss) over 5 years and annualized MSFC change over 5-years of follow-up (quantitative data not reported)
Note: Result for GMV loss and MSFC at 5 years was not reported.
There was no significant correlation between annual BVL and physical disability progression over 5 years in PPMS patients.
Total: 81 RRMS: 62 SPMS: 11 PPMS: 8 Median EDSS at baseline: 3.5
Overall: 42.0
GMV change from baseline to 5-year or from baseline to 10-year follow-up Algorithm used: 1. Normalized brain volume-SIENAX 2. Longitudinal brain volume changes- SIENAX multi-timepoint (SX-MTP)
EDSS at 5- and 10-year follow-up classified as DP: An increase of EDSS by 1.0 and 0.5 points if baseline EDSS was <6.0 and ≥6.0 respectively No DP: Patients who do not meet the criteria for physical disability progression
GMV loss and EDSS (MV analysis)Quantitative estimates not reported
•
There was no significant association between gray matter volume change and physical disability progression either at 5 years or 10 years (quantitative data not reported)
Factors adjusted in MV analysis: Age, sex, scanner type, baseline volumes of specific structures, T1-lesion volumes, T2-lesion volumes, MS subtype, disease duration and use of disease-modifying treatment.
There was no significant association between gray matter volume change and physical disability progression either at 5 years or 10 years in MS patients.
Overall: NR Clinically definite MS: 27.0 Stable CIS: 30.1
PGMV change measured at baseline, 6 months and yearly thereafter until 4 years follow-up. Algorithm used: Modified SIENAX multi-time point algorithm
EDSS at 4 years follow-up was classified as SDP: An increase of EDSS 1.0 and 1.5 if the baseline EDSS was > 0.0 and 0.0 sustained over at least 24 weeks after the end of the study respectively. SDI: At least a 1.0-point decrease in EDSS.
BVL and EDSS (MV analysis)*P<0.05 for differences between SDP and stable plus SDI group in the confirmatory analysis that excluded MRI data between baseline and 6 months
•
There was a significant decrease in the PGMV (P = 0.011) among CIS patients who developed SDP compared to patients who remained stable or improved in their physical disability status over 4-years follow-up
Factors adjusted in MV analysis: Age, sex, time from the first event to baseline assessment, and treatment status.
Increased gray matter atrophy was associated with sustained physical disability progression in CIS patients on standard disease-modifying therapy.
GMF change at 1-year follow-up Algorithm used: Semi-automated algorithm implemented on SPM (v 5) and MRIcro software.
EDSS at 1 and 2 years and EDSS change from 0-1 and 0-2 years used as a continuous variable
GMF loss and EDSS change (UV analysis)Spearman's r = –0.177; P = 0.025
•
Increased BVL (measured using GMF change) from baseline to 1-year follow-up was associated with an increase (worsening) in EDSS delta from baseline to 2 years follow-up
Note: Correlation between 1) GMF loss and EDSS at 1 or 2 years or EDSS change from 0-1 year were not reported.
Very early brain atrophy development was associated with increased physical disability from baseline to 2- years follow-up after a CIS.
Total: 39 Subtypes not mentioned Median EDSS at baseline: 4.0
Overall: 35.0
GMF change at 1-year follow-up Algorithm used: MATLAB's SPM (v 5) used for segmentation and T1-weighted sequences used to obtain values for GMF.
EDSS change from 0-1 year used as a continuous variable
GMF loss and EDSS (UV analysis)Spearman's r = NR
•
BVL (measured using GMF change) occurring during the first year of follow-up did not correlate with EDSS change during the first year of follow-up (quantitative data not reported)
BVL occurring during the first year of follow-up was not associated with physical disability progression in MS patients.
Overall: NR Stable CIS: 29 Progressing to CDMS: 28
PGMV change at 6 months follow-up Algorithm used: SIENAX
Conversion from CIS to CDMS used as a binary variable
PGMV loss in CIS and CDMS groups (MV analysis)
•
Relative change in PGMV over 6 months showed a consistent trend to decrease faster in the CDMS group, this trend did not reach the level of statistical significance (P>0.1, mixed models)
Factors adjusted in MV analysis: Age, sex, and treatment changes.
Very early (short term) BVL was not associated with the conversion of CIS patients to CDMS.
Association between cortical gray matter atrophy and disability in MS (n = 6 studies)
Total: 1214 CIS: 253 RRMS: 708 SPMS: 128 PPMS: 125 Median EDSS at baseline: NR
Overall: NR CIS: 33 RRMS: 38.2 SPMS: 48.2 PPMS: 48.5
Cortical gray matter volume loss per year Algorithm used: Regional volumes summarized using Neuromorphometrics protocol
Annual EDSS change used as a continuous variable
Cortical gray matter volume loss and EDSS (MV analysis)Linear β = NR; P = NR
•
There was no association between the rate of change in the cortical gray matter volume and the rate of change in the EDSS over time (quantitative data not reported)
Factors adjusted in Mv analysis: Time, scanner magnetic field, subject group, sex, age at baseline, and total intracranial volume at baseline; and the interactions of each of these with time.
Cortical gray matter volume loss was not associated with physical disability progression in MS patients.
Total: 38 RRMS: 36 SPMS: 2 Median EDSS at baseline: 4.0
Overall: 36.8
PBCV change measured annually for 3 years Algorithm used: Volume measure was obtained on the MPRAGE images using the longitudinal stream included in the FreeSurfer analysis suite and the VBM (v 8) approach implemented in the SPM (v 8) software.
EDSS at the third year was classified as EDSS worsening- An increase in 1.0 and 0.5 points, when baseline EDSS was <5.5 and ≥5.5 points respectively or sustained up until the 3-year follow-up time point. EDSS stable or improved: Definition not reported Baseline EDSS used as a continuous variable
PBCV change frombaseline to 1 year and EDSS at 3rd-year follow-up (UV analysis)
•
Patients with EDSS worsening at 3 years follow-up had a significantly greater BVL (measured using bilateral cortex) during the first-year follow-up than patients with stable or improved EDSS
PBCV change from 1stto 2ndyear or 2ndto 3rdyear and EDSS at 3rd-year follow-up (UV analysis)
•
No significant associations were found between PBCV change during the second or third 1-year periods and EDSS worsening at 3rd -year (quantitative estimates not reported)
PBCV change and baseline EDSS (UV analysis)Spearman's r = NR
•
Baseline EDSS was not correlated with change in PBCV from baseline to 1st year or 1st to 2nd -year or 2nd to 3rd -year (quantitative estimates not reported)
MS patients with EDSS worsening at 3-years follow-up had significantly greater bilateral cortical volume loss than patients with stable or improved EDSS.
Short-term MRI measurements as predictors of EDSS progression in relapsing-remitting multiple sclerosis: grey matter atrophy but not lesions are predictive in a real-life setting.
Change in cortex volume from baseline and between 5 – 14 months follow-up Algorithm used: SIENA and FreeSurfer (v 5.2.0)
EDSS change at 3 years follow-up classified as DP – yes: Increase in EDSS by ≥ 1.5, ≥ 1.0, and ≥ 0.5 points from baseline 0.0, 1.0 to 4.0, and >5.0 points respectively DP – no: Patients not meeting the disease progression criteria
Cortex volume loss and DP (MV analysis)Logit OR = 0.71, 95% CI (0.55 to 0.87); P = 0.016
•
Every 1 per cent increase in BVL (measured using cortex volume loss) from baseline to 5 – 14 months follow-up was associated with a 1.4 times significantly higher risk of physical disability progression over 3 years
Factors adjusted in MV analysis: Treatment status, sex, age, number of T1-, T2- and Gd-lesions, absolute change of lesion numbers and SIENAX volumes from visit 1 to visit 2, aPBV change and global atrophy measurements from the longitudinal FreeSurfer processing.
Short-term (5–14 months from baseline) loss of cortex volume was predictive of EDSS worsening (physical disability progression) at 3 years of follow-up in RRMS patients.
Cortical gray matter volume changes (regions: Precuneus, postcentral gyrus, and orbital gyrus) at 1-year follow-up Algorithm used: Images were analyzed using SPM5(v 958) running under the MATLAB 7.0 (R14) environment and processed using the VBM toolbox (v 1.03).
EDSS at 1-year follow-up classified as Progressive group: An increase of ≥ 1.0 point Stable group: No change MSFC (for any of the 3 components) at 1-year follow-up classified as Progressive group: ≥ 0.25% Stable group: No change of ≥ 0.25%
Right precuneus of the right parietal lobe volume lossand EDSS (UV analysis)Progressive group: cluster PFWE ˂0.05Stable group: minimum PFWE = 0.283Between-group differences: cluster Pcorrected ˂0.001
•
There was a significant cortical GMV loss as measured in the right precuneus of the right parietal lobe in the progressive group but not in the stable group over 1-year follow-up with significant between-group differences
Postcentral gyrus of the right parietal lobe volume loss and EDSS (UV analysis)Progressive group: Estimates not reportedStable group: Estimates not reportedBetween-group differences: cluster Pcorrected ˂0.001
•
There were significant between-group differences between the progressive and stable groups in the postcentral gyrus of the right parietal lobe over 1-year follow-up
Orbital gyrus of right frontal lobe volumeloss and MSFC (UV analysis)Progressive group: cluster PFWE ˂0.05Stable group: minimum PFWE = 0.848Between-group differences: minimum PFWE = 0.663
•
There was a significant cortical GMV loss as measured in the orbital gyrus of the right frontal lobe in the progressive group but not in the stable group over a 1-year follow-up, with no significant between-group differences
P FEW (Family-wise error) – probability of making at least one type-1 error among a specified group, or "family," of testsCluster P- clusters of voxels are tested for significanceP-corrected – P-value corrected for multiple comparisonsNote. Association between 1) orbital gyrus of the right frontal lobe and EDSS and 2) right precuneus and postcentral gyrus of the right parietal lobeand MSFC was not reported.
Cortical regional gray matter volume reductions were associated with physical disability progression in RRMS patients.
Total: 81 RRMS: 62 SPMS: 11 PPMS: 8 Median EDSS at baseline: 3.5
Overall: 42
CV change from baseline to 5-year or from baseline to 10-year follow-up Algorithm used: 1. Normalized brain volume-SIENAX 2. Longitudinal brain volume changes- SIENAX multi-timepoint (SX-MTP)
EDSS at 5- and 10-year follow-up classified as DP: An increase of EDSS by 1.0 and 0.5 points if baseline EDSS was <6.0 and ≥6.0 respectively No DP: Patients who do not meet the criteria for physical disability progression
CV loss and EDSS (MV analysis)At 5-year follow-up: Logistic β = −0.338; P <0.01At 10-year follow-up: Logistic β = −0.126; P = NS
•
Every 1 per cent increase in BVL (measured using CV change) was significantly associated with 1.4 times higher risk of physical disability progression at 5 years; however, the association was not significant at 10 years
Factors adjusted in MV analysis: Age, sex, scanner type, baseline volumes of specific structures, T1-lesion volumes, T2-lesion volumes, MS subtype, disease duration and use of disease-modifying treatment.
Early BVL was associated with physical disability progression in MS patients.
Overall: NR Clinically definite MS: 27.0 Stable CIS: 30.1
PCV change measured at baseline, 6 months and yearly thereafter until 4 years follow-up. Algorithm used: Modified SIENAX multi-time point algorithm
EDSS at 4 years follow-up was classified as SDP: An increase of EDSS 1.0 and 1.5 if the baseline EDSS was > 0.0 and 0.0 sustained over at least 24 weeks after the end of the study respectively. SDI: At least a 1.0-point decrease in EDSS.
BVL and EDSS (MV analysis)*P<0.05 for differences between SDP and stable plus SDI group in the confirmatory analysis that excluded MRI data between baseline and 6 months
•
There was a significant loss in the PCV (P = 0 .001) among CIS patients who developed SDP compared to patients who remained stable or improved in their disability status over 4-years follow-up
Factors adjusted in MV analysis: Age, sex, time from the first event to baseline assessment, and treatment status.
BVL (measured using PCV change) was associated with physical disability progression.
Association between deep gray matter atrophy and disability in MS (n = 3 studies)
Total: 1214 CIS: 253 RRMS: 708 SPMS: 128 PPMS: 125 Median EDSS at baseline: NR
Overall: NR CIS: 33 RRMS: 38.2 SPMS: 48.2 PPMS: 48.5
Deep gray matter volume loss per year Algorithm used: Regional volumes summarized using Neuromorphometrics protocol
Annual EDSS change used as a continuous variable
DGMV loss and EDSS (MV analysis)Linear β = –0.04, 95% CI (–0.02 to –0.06); P = 0.006
•
Every SD (Z-score) of DGMV loss per year was significantly associated with an annual increase of 0.04 units of EDSS
Factors adjusted in MV analysis: Time, scanner magnetic field, subject group, sex, age at baseline, and total intracranial volume at baseline; and the interactions of each of these with time.
DGMV loss was associated with physical disability progression in MS patients.
Total: 81 RRMS: 62 SPMS: 11 PPMS: 8 Median EDSS at baseline: 3.5
Overall: 42
Total subcortical DGMV change from baseline to 5-year or from baseline to 10-year follow-up Algorithm used: FIRST (v.1.2) was used to estimate absolute tissue volume changes.
EDSS at 5- and 10-year follow-up classified as DP: An increase of EDSS by 1.0 and 0.5 points if baseline EDSS was <6.0 and ≥6.0 respectively No DP: Patients who do not meet the criteria for physical disability progression
Total subcortical DGMV loss and EDSS (MV analysis)At 5-year follow-up: Logistic β = −0.113; P <0.05At 10-year follow-up: Logistic β = −0.072; P = NS
•
Every 1 per cent increase in BVL (measured using total subcortical DGM volume change) was significantly associated with 1.1 times higher risk of physical disability progression at 5 years; however, the association was not significant at 10 years
Factors adjusted in MV analysis: Age, sex, scanner type, baseline volumes of specific structures, T1-lesion volumes, T2-lesion volumes, MS subtype, disease duration and use of disease-modifying treatment.
Early BVL was associated with physical disability progression in MS patients
Total normalized PSDGMV change measured at baseline, 6 months and yearly thereafter until 4 years follow-up. Algorithm used: Percentage volume changes at each time point were estimated using FIRST.
EDSS at 4 years follow-up was classified as SDP: An increase of EDSS 1.0 and 1.5 if the baseline EDSS was > 0.0 and 0.0 sustained over at least 24 weeks after the end of the study respectively. SDI: At least a 1.0-point decrease in EDSS.
BVL and EDSS (MV analysis)*P-value >0.05 for differences between SDP and stable plus SDI group in the confirmatory analysis that excluded MRI data between baseline and 6 months
•
There was no significant decrease in the total normalized PSDGMV among CIS patients who developed SDP compared to patients who remained stable or improved in their physical disability status over 4-years follow-up
Factors adjusted in MV analysis: Age, sex, time from the first event to baseline assessment, and treatment status.
There were no significant differences in total normalized PSDGMV change among CIS patients across physical disability categories over 4-years follow-up
Association between cortical thickness loss and disability in MS (n = 3 studies)
Total: 86 RRMS: 75 Progressive MS: 11 Median EDSS at baseline: 2.0
Overall: 40.4
Percentage cortical thinning (left superior temporal gyrus, left lateral orbitofrontal cortex, right middle temporal gyrus) over 1 year- linear measure Algorithm used: Thickness measures assessed on lesion-filled 3D T1-weighted images using the longitudinal pipeline implemented in the FreeSurfer (v 6.0) analysis suite.
EDSS at 1 year used as a continuous variable
Cortical thinning and EDSS (MV analysis)GLM correlationLeft superior temporal gyrus: rmax* = −0.38Left lateral orbitofrontal cortex: rmax* = −0.28
•
Higher cortical atrophy (measured using cortical thinning in the left superior temporal gyrus and left lateral orbitofrontal cortex) from baseline to 1-year follow-up was significantly associated with a higher EDSS score at 1 year
Right middle temporal gyrus: rmax = −0.21; P = NR
•
No association was found between cortical atrophy (measured using cortical thinning in the right middle temporal gyrus) from baseline to 1-year follow-up and EDSS score at 1 year
R max = highest partial correlation coefficient inside the cluster*Clusters surviving at p<0.05, cluster-wise Monte Carlo null-z threshold simulationFactors adjusted in MV analysis: Sex, age, site, and EDSS at baseline.
Parieto-temporal cortical thinning over time was associated with 1-year physical disability worsening in MS patients.
Total: 243 RRMS: 180 SPMS: 51 PPMS: 12 Median EDSS at baseline: 3.0
Overall: 44.5
Annual cortical thickness change over 6 years – linear measure Algorithm used: ANIMAL
Annual change in log EDSS over 6 years used as a continuous variable
Cortical thickness change and EDSS (MV analysis)Whole cohortLinear mixed-effect model* - mean t-values, rangeRight side of the brain: –2.22 to –3.54Left side of the brain: –2.39 to –3.80
•
Increased atrophy in large extended bilateral cortical regions, predominantly in the right temporal and left frontal and parietal lobes was significantly associated with an increased physical disability (measured using log[EDSS] change) over 6 years
RRMSpatientsLinear mixed-effect model* - mean t-values - NR
•
Increased atrophy in large, extended bilateral cortical regions was significantly associated with an increased physical disability (measured using log[EDSS] change) over 6 years
SPMS patientsLinear mixed-effect model - mean t-values - NR
•
No association was found between cortical atrophy and physical disability (measured using log[EDSS] change) over 6 years
PPMS patientsLinear mixed-effect model* - mean t-values - NR
•
Increased atrophy in small clusters, predominantly in the bilateral superior parietal gyri, the left precentral gyrus, the left middle frontal gyrus, and the left postcentral gyrus was significantly associated with an increased physical disability (measured using log[EDSS] change) over 6 years
*Correction with the false discovery rate approach for multiple comparisons set at q < 0.05Factors adjusted in MV analysis: Age and disease duration.
Increased cortical thinning was associated with physical disability progression in RRMS and PPMS patients but not in SPMS patients.
Total: 38 RRMS: 36 SPMS: 2 Median EDSS at baseline: 4.0
Overall: 36.8
PCT change measured annually for 3 years – linear measure Algorithm used: Thickness measure was obtained on the MPRAGE images using the longitudinal stream included in the FreeSurfer analysis suite and the VBM (v 8) approach implemented in the SPM (v 8) software.
EDSS at the third year was classified as EDSS worsening- An increase in 1.0 and 0.5 points, when baseline EDSS was <5.5 and ≥5.5 points respectively or sustained up until the 3-year follow-up time point. EDSS stable or improved: Definition not reported Baseline EDSS used as a continuous variable
PCT change from baseline to 1 year and EDSS at 3rd-year follow-up (UV analysis)
•
Patients with EDSS worsening at 3 years follow-up had a significantly greater cortical atrophy (measured using a decrease in bilateral mean cortical thickness) during the first-year follow-up than patients with stable or improved EDSS
PCT change andbaseline EDSS (UV analysis)Spearman's r = NR
•
Baseline EDSS was not correlated with change in cortical thickness from baseline to 1st year or 1st to 2nd -year or 2nd to 3rd -year (quantitative data not reported)
Note: Results for the association between PCT loss from 1st to 2nd or 2nd to 3rd -year and disability status at the third- year were not reported.
MS patients with EDSS worsening at 3-years follow-up had a significantly greater cortical atrophy than patients with stable or improved EDSS.
3.6 Total gray matter atrophy and physical disability
•
Of the eight studies reporting on the association between total gray matter volume change and EDSS, four found a significant relationship such that an increase in the total gray matter volume loss was associated with physical disability progression (
Short-term MRI measurements as predictors of EDSS progression in relapsing-remitting multiple sclerosis: grey matter atrophy but not lesions are predictive in a real-life setting.
One study each reporting on the association between total gray matter volume change and MSFC/conversion from CIS to CDMS found no association between them (
3.7 Cortical gray matter atrophy and physical disability
•
Of the six studies reporting on the association between cortical volume change and EDSS, five found that cortical gray matter volume loss was significantly associated with physical disability progression (
Short-term MRI measurements as predictors of EDSS progression in relapsing-remitting multiple sclerosis: grey matter atrophy but not lesions are predictive in a real-life setting.
One study reporting on the association between cortical volume change and MSFC found that cortical volume reductions were significantly associated with physical disability progression (
3.8 Deep gray matter atrophy and physical disability
•
Of the three studies reporting on the association between deep gray matter volume change and EDSS, two found a significant association such that an increase in deep gray matter volume loss was associated with physical disability progression (
Three studies reporting on the relationship between cortical thickness and EDSS found that cortical thinning was significantly associated with physical disability progression (
Table 5 presents an overview of the six primary studies reporting data on the association between limbic system structures (amygdala, hippocampus, bilateral cingulate gyrus, and thalamus) volume loss and physical disability in MS. The studies in the table are arranged in reverse chronological order, starting with the most recently published. Of these six studies, five were prospective cohort and one was retrospective cohort. The follow-up duration varied from two years to 10 years with a sample size ranging from 26 to 229. The average age of patients across the six studies ranged from 38 to 44 years. The baseline median: i) EDSS reported in four studies ranged from 1.5 to 4.0 points; ii) MSSS reported in one was 4.08; and iii) MSFC reported in one was –1.2.
•
Of the two studies reporting on the association between limbic system structure volume loss and EDSS, one found a significant relationship such that an increase in the thalamus volume loss was predictive of worsening of physical disability over time (
). The other study found a significant trend of association between decreased volume of the amygdala and physical disability progression in MS patients (
) while one study reporting on the association between thalamus volume change and conversion from CIS to CDMS found a significant decrease in thalamus volume among converters to CDMS compared to non-CDMS converters (
One study reporting on the relationship between bilateral cingulate gyrus volume loss and MSFC found a significant association between the two such that higher rate of bilateral cingulate cortex volume loss was associated with greater physical disability as measured using MSFC (
Table 5Association between limbic system structures (amygdala, hippocampus, bilateral cingulate gyrus, and thalamus) volume and physical disability in MS.
Total: 229 RRMS: 179 SPMS: 50 Median EDSS at baseline: NR
Overall: 44.4
Annual rate of volume loss in thalamus and ventral lateral nucleus (part of thalamic subnuclei- thalamus) over 6 years Algorithm used: MAGeT
Annual change in EDSS over 6 years classified as DP: An increase in EDSS of 1.0 and 0.5 points if the baseline EDSS was ≤5.5 and >5.5 points respectively No DP/stable - Not defined
Annual rate of thalamic volume loss and DP (MV analysis)Cox HR = 0.73, 95% CI (0.63 to 0.84); P < 0·001
•
Every 1% increase in the annual rate of thalamic volume loss was significantly associated with an extra 27% risk to develop disease progression in the following year
Annual rate of ventral lateral nucleus loss and DP (MV analysis)Cox HR = 0.80, 95% CI (0.70 to 0.91); P < 0·001
•
Every 1% increase in the annual rate of volume loss was significantly associated with an extra 20% risk to develop disease progression in the following year
Factors adjusted in MV analysis: Age at baseline, sex, disease duration at baseline, MS type, normalized total brain volume, white matter lesion volume
Volume loss in the thalamus and ventral lateral nucleus was predictive of physical disability worsening as assessed by EDSS in MS patients.
RRMS: 26 Mean EDSS at baseline: 2.42 Mean MSSS at baseline: 4.08
Overall: 38.1
2-year percentage change in the volume of thalamus Algorithm used: Volumes obtained using FIRST at each time point were normalized relative to whole head volume for each subject.
Changes in EDSS and MSSS between baseline and 2 years used as a continuous variable
Percentage change in the volume of thalamus and changes in EDSS and MSSS (UV analysis)Pearson's r (P-value):
•
Changes in the volume of thalamus was not correlated with the change in EDSS/MSSS over 2 years
Change in thalamus volume was not associated with change in EDSS/MSSS over 2 years in RRMS patients.
PPMS: 36 Median EDSS at baseline: 4.0 Median MSFC at baseline: –1.2
Overall: 42.8
Rate of change in bilateral cingulate gyrus at 5 years follow-up Algorithm used: SIENAX FSL 4.1.9
MSFC at 5 years follow-up as a continuous variable
Bilateral cingulate gyrus volume loss and MSFC (MV analysis)Pearson's partial r=0.51; P = 0.001
•
Higher annual BVL (measured using the annualized rate of volume loss in the bilateral cingulate cortex) over 5 years was significantly associated with greater physical disability (measured using MSFC score) at five years
Factors adjusted in MV analysis: Baseline normalized GMV
Higher rate of bilateral cingulate cortex volume loss was associated with greater physical disability (measured using MSFC score) in PPMS patients.
Total: 81 RRMS: 62 SPMS: 11 PPMS: 8 Median EDSS at baseline: 3.5
Overall: 42
Hippocampus, thalamus, and amygdala volume changes from baseline to 5-year or from baseline to 10-year follow-up Algorithm used: FIRST (V.1.2) used to estimate absolute tissue volumes.
EDSS at 5- and 10-year follow-up classified as DP: An increase of EDSS by 1.0 and 0.5 points if baseline EDSS was <6.0 and ≥6.0 respectively No DP: Patients who do not meet the criteria for physical disability progression
Hippocampus, thalamus, and amygdala volume changes and EDSS (MV analysis)Quantitative estimates of binary logistic regression analysis not reported
•
There was a trend of association between decreased volume of amygdala, and physical disability progression at 5 years (P<0.05) but not at 10 years, while no association was reported between thalamus and hippocampus volume changes and physical disability progression either at 5 years or 10 years
Factors adjusted in MV analysis: Age, sex, scanner type, baseline volumes of specific structures, T1-lesion volumes, T2-lesion volumes, MS subtype, disease duration and use of disease-modifying treatment.
There was a trend of association between decreased volume of amygdala and physical disability progression at 5 years while there was no association between thalamus and hippocampus volume changes and physical disability progression in MS patients.
Overall: NR Clinically definite MS: 27.0 Stable CIS: 30.1
Percentage thalamus volume change measured at baseline, 6-months and yearly thereafter until 4-years follow-up. Algorithm used: Volumes at each time obtained using FIRST.
EDSS at 4 years follow-up was classified as SDP: An increase of EDSS 1.0 and 1.5 if the baseline EDSS was > 0.0 and 0.0 sustained over at least 24 weeks after the end of the study respectively. SDI: At least a 1.0-point decrease in EDSS.
BVL and EDSS (MV analysis)*P-value >0.05 for differences between SDP and stable plus SDI group in the confirmatory analysis that excluded MRI data between baseline and 6 months
•
There was no significant decrease in the percentage thalamus volume among CIS patients who developed SDP compared to patients who remained stable or improved in their physical disability status over 4-years follow-up
Factors adjusted in MV analysis: Age, sex, time from the first event to baseline assessment, and treatment status
There were no significant differences in the percentage thalamus volume change among CIS patients across physical disability categories over 4-years follow-up
Percentage thalamus volume change measured at baseline, 6 months and yearly thereafter until 2 years follow-up. Algorithm used: NR
Conversion from CIS to CDMS used as a binary variable
Conversion from CIS to CDMS (MV analysis)
•
There was a significant decrease in the percentage thalamus volume (P = 0.009) among converters to CDMS compared to non-CDMS converters over 2-years follow-up
Factors adjusted in MV analysis: age, time from the first event to baseline assessment, and change in treatment status over the 2-year follow-up
Thalamic atrophy was associated with conversion to CDMS in patients with CIS over 2 years.
Table 6 presents an overview of the three primary studies reporting data on the association between basal ganglia (accumbens, caudate, globus pallidus, putamen, and left pallidum) volume loss and physical disability in MS. The studies in the table are arranged in reverse chronological order, starting with the most recently published. Of these three studies, two were prospective cohort and one was retrospective cohort. The follow-up duration varied from two years to 10 years with a sample size ranging from 28 to 81. The average age of patients across the three studies ranged from 37 years to 42 years. The median EDSS at baseline reported in all the three studies ranged from 2.42 to 4.0 whereas the median MSSS reported in one study was 4.08.
•
Of the three studies reporting on the association between basal ganglia volume loss and EDSS, only one study found a trend of association between decreased volumes of caudate, putamen, and pallidus and physical disability progression (
One study reporting on the association between caudate, putamen, and globus pallidus volume loss and MSSS found no significant relationship between them (
Total: 38 RRMS: 36 SPMS: 2 Median EDSS at baseline: 4.0
Overall: 36.8
PLPV change measured annually for 3 years Algorithm used: Obtained on the MPRAGE images by means of the longitudinal stream included in the FreeSurfer analysis suite (v5.1.0)
EDSS at the third year was classified as EDSS worsening - An increase in 1.0 and 0.5 points, when baseline EDSS was <5.5 and ≥5.5 points respectively or sustained up until the 3-year follow-up time point. EDSS stable or improved: Definition not reported Baseline EDSS used as a continuous variable
PLPV change from baseline to 1 year and EDSS at 3rd-year follow-up (UV analysis)
•
Patients with EDSS worsening at 3 years follow-up did not have a significantly greater BVL (measured using PLPV loss) during the first-year follow-up than patients with stable or improved EDSS
PLPV change from 1stto 2ndyear or 2ndto 3rdyear and EDSS at 3rd-year follow-up (UV analysis)
•
No significant associations were found between PLPV change during the second or third 1-year periods and EDSS worsening at 3rd -year (quantitative estimates not reported)
Change in PLPV and baseline EDSS (UV analysis)Spearman's r = NR
•
Baseline EDSS was not correlated with change PLPV from baseline to 1st year or 1st to 2nd -year or 2nd to 3rd -year (quantitative data not reported)
In MS patients, PLPV loss was not significantly different between patients with EDSS worsening and those with stable or improved EDSS.
RRMS: 26 Mean EDSS at baseline: 2.42 Mean MSSS at baseline: 4.08
Overall: 38.1
2-year percentage changes in the volumes of caudate, putamen, and globus pallidus Algorithm used: 1) Volumes obtained using FIRST at each time point were normalized relative to whole head volume for each subject 2) SIENA
Changes in EDSS and MSSS between baseline and 2 years used as a continuous variable
Percentage changes in the volumes of caudate, putamen, and globus pallidus and changes inEDSS and MSSS (UV analysis)Pearson's r (P-value):
•
Changes in the volumes of caudate, putamen, and globus pallidus were not correlated with the change in EDSS/MSSS over 2 years
Correlations between caudate, putamen, and globus pallidus volume changes and change in EDSS/MSSS over 2 years were not statistically significant in RRMS patients.
Total: 81 RRMS: 62 SPMS: 11 PPMS: 8 Median EDSS at baseline: 3.5
Overall: 42
Caudate, putamen, pallidus, and accumbens from baseline to 5-year or from baseline to 10-year follow-up Algorithm used: FIRST (V.1.2) used to estimate absolute tissue volumes.
EDSS at 5- and 10-year follow-up classified as DP: An increase of EDSS by 1.0 and 0.5 points if baseline EDSS was <6.0 and ≥6.0 respectively No DP: Patients who do not meet the criteria for physical disability progression
Caudate, putamen, pallidus, and accumbens volume changes and EDSS (MV analysis)Quantitative estimates not reported
•
There was a trend of association between decreased volume of caudate, putamen, and pallidus, and physical disability progression at 5 years (P<0.05) but not at 10 years, while no association was reported between accumbens volume change and physical disability progression either at 5 years or 10 years
Factors adjusted in MV analysis: Age, sex, scanner type, baseline volumes of specific structures, T1-lesion volumes, T2-lesion volumes, MS subtype, disease duration and use of disease-modifying treatment.
There was a trend of association between decreased volumes of caudate, putamen, and pallidus and physical disability progression at 5 years while there was no association between accumbens volume change and physical disability progression in MS patients.
BVL: Brain volume loss; DP: Disability progression; EDSS: Expanded Disability Status Scale; FIRST: FMRIB's Integrated Registration And Segmentation Tool; MPRAGE: Magnetization-Prepared Rapid Gradient-Echo; MS: Multiple sclerosis; MSSS: Multiple Sclerosis Severity Score; MV: Multivariable; NR: Not reported; PLPV: Percentage left pallidum volume; PPMS: Primary progressive multiple sclerosis; Q: Quality rating of study; RRMS: Relapsing-remitting multiple sclerosis; SIENA: Structural Image Evaluation, using Normalization, of Atrophy; SPMS: Secondary progressive multiple sclerosis; UV: Univariate.
Table 7 presents an overview of the six primary studies reporting data on the association between white matter volume loss and physical disability in MS. The studies in the table are arranged in reverse chronological order, starting with the most recently published. Of these six studies, five were prospective cohort and one was retrospective cohort. The follow-up duration varied from two years to 10 years with a sample size ranging from 39 to 217. The average age of patients across the six studies ranged from 31 to 42 years. The median EDSS at baseline was only reported in three studies and ranged from 1.5 to 4.0.
•
Of the three studies reporting on the association between white matter volume loss and EDSS, only one study (in RRMS) found a significant relationship such that an increase in white matter volume loss was associated with worsening of physical disability over time (
One study reporting on the association between white matter volume loss and conversion from CIS to CDMS found no significant relationship between the two (
aPWMV (cut-off: – 2.49%) changes at 1-year follow-up Algorithm used: In-house built semi-automated algorithm implemented on SPM (v 8) and MRIcro software
EDSS at 4 years follow-up was classified as Confirmed worsening: An increase in 1.0 and 0.5 points if prior EDSS was ≤5.0 and ≥5.5 points respectively confirmed at a further visit with a time gap of at least 6 months Progression-free patients: Patients who do not meet the criteria for confirmed worsening at 4 years
aPWMV loss and disability (UV analysis)Cox HR = 4.246, 95% CI (1.27 to 9.20); P = 0.015
•
Patients losing more than 2.49% of WMV between baseline and 1 year had 4.3 times higher risk of developing confirmed EDSS worsening at 4-years follow-up
Increased BVL in the first year of interferon-beta therapy was predictive of subsequent physical disability progression in RRMS patients.
Total: 81 RRMS: 62 SPMS: 11 PPMS: 8 Median EDSS at baseline: 3.5
Overall: 42
WMV change from baseline to 5-year or from baseline to 10-year follow-up Algorithm used: 1. Normalized brain volume-SIENAX 2. Longitudinal brain volume changes- SIENAX multi-timepoint (SX-MTP)
EDSS at 5- and 10-years follow-up classified as DP: An increase of EDSS by 1.0 and 0.5 points if baseline EDSS was <6 and ≥6 respectively No DP: Patients who do not meet the criteria for physical disability progression
WMV loss and EDSS (MV analysis)Quantitative estimates not reported
•
There was no significant association between white matter volume change and physical disability progression either at 5 years or 10 years (quantitative data not reported)
Factors adjusted in MV analysis: Age, sex, scanner type, baseline volumes of specific structures, T1-lesion volumes, T2-lesion volumes, MS subtype, disease duration and use of disease-modifying treatment.
White matter volume loss was not associated with physical disability progression at either 5 or 10 years in MS patients.
PWMV change measured at baseline, 6 months and yearly thereafter until 4 years follow-up. Algorithm used: Modified SIENAX multi-time point algorithm.
EDSS at 4 years follow-up was classified as SDP: An increase of EDSS 1.0 and 1.5 if the baseline EDSS was > 0.0 and 0.0 sustained over at least 24 weeks after the end of the study respectively. SDI: At least a 1.0-point decrease in EDSS.
PWMV loss and EDSS (MV analysis)*P-value >0.05 for differences between SDP and stable plus SDI group in the confirmatory analysis that excluded MRI data between baseline and 6 months
•
There was no significant decrease in the PWMV among CIS patients who developed SDP compared with patients who remained stable or improved in their physical disability status over 4-years follow-up
Factors adjusted in MV analysis: Age, sex, time from the first event to baseline assessment, and treatment status
There were no significant differences in white matter volume change among CIS patients across physical disability categories over 4-years follow-up.
WMF change at 1-year follow-up Algorithm used: Semi-automated algorithm implemented on SPM (v5) and MRIcro software
EDSS at 1 and 2 years and EDSS change from 0–1 and 0–2 years used as a continuous variable
WMF loss and EDSS (UV analysis)Spearman's r = No correlation, P = No correlation
•
No statistically significant correlations were found between WMF changes from baseline to 1- year follow-up and EDSS scores at 1-year or 2-years or EDSS change from baseline to 1- or 2- years follow-up (quantitative data not reported)
WMF change was not associated with physical disability progression in MS patients.
Total: 39 Subtypes not mentioned Median EDSS at baseline: 4.0
Overall: 35.0
WMF change at 1-year follow-up Algorithm used: MATLAB's SPM (v5) used for segmentation and T1-weighted sequences used to obtain values for WMF.
EDSS change from 0–1 year used as a continuous variable
WMF loss and EDSS (UV analysis)Spearman's r = NR
•
BVL (measured using WMF change) occurring during the first year of follow-up did not correlate with EDSS change during the first year of follow-up (quantitative data not reported)
BVL was not associated with physical disability progression during the first year of follow-up in MS patients.
Overall: NR Stable CIS: 29 Progressing to CDMS: 28
PWMV change at 6 months follow-up Algorithm used: SIENAX
Conversion from CIS to CDMS used as a binary variable
PWMV loss in CIS and CDMS groups (MV analysis)
•
Relative change in PWMV over 6 months showed a consistent trend to decrease faster in the CDMS group, this trend did not reach the level of statistical significance (P>0.1, mixed models)
Factors adjusted in MV analysis: Age, sex, and treatment changes.
Very early white matter volume change was not associated with conversion from CIS to CDMS.
aPWMV: Annualized percentage white matter volume; BVL: Brain volume loss; CDMS: Clinically definite multiple sclerosis; CI: Confidence interval; CIS: Clinically isolated syndrome; DP: Disability progression; EDSS: Expanded Disability Status Scale; HR: Hazard ratio; MATLAB: MATrix LABoratory; MRI: Magnetic resonance imaging; MS: Multiple sclerosis; MV: Multivariable; NR: Not reported; PPMS: Primary progressive multiple sclerosis; PWMV: Percentage white matter volume; Q: Quality rating of study; RRMS: Relapsing-remitting multiple sclerosis; SD: Standard deviation; SDI: Sustained disability improvement; SDP: Sustained disability progression; SIENA: Structural Image Evaluation, using Normalization, of Atrophy; SIENAX: Structural Image Evaluation, using Normalization, of Atrophy—Cross-Sectional SPM: Statistical parametric mapping; SPMS: Secondary progressive multiple sclerosis; UV: Univariate; WMF: white Matter fraction; WMV: White matter volume.
The corpus callosum includes axons projecting between the hemispheres. Relative loss of volume is commonly associated with MS.
Table 8 presents an overview of the four primary studies reporting data on the association between corpus callosum atrophy and physical disability in MS. The studies in the table are arranged in reverse chronological order, starting with the most recently published. Of these four studies, three were retrospective cohort and one was prospective cohort. The follow-up duration varied from two years to nine years with a sample size ranging from 29 to 217. The median age of patients across the four studies ranged from 28 years to 40 years. The median EDSS at baseline was only reported in two studies and ranged from 2.5 to 3.35.
•
Of the four studies reporting on the association between corpus callosum atrophy and EDSS, three found a significant relationship such that increased corpus callosum atrophy was associated with worsening of physical disability over time (
Similarly, of the two studies reporting on the relationship between corpus callosum atrophy and conversion from CIS to CDMS, both found that CIS patients who had converted to CDMS had a significantly greater corpus callosum atrophy than those who had not converted to CDMS over time (
Annualized change in CCA - linear measure Algorithm used: Image analysis was performed using OsiriX DICOM Viewer on Mac OS
EDSS at 5 years and annualized change in EDSS as a continuous variable Conversion from CIS to CDMS used as a binary variable
aCCA loss and EDSS (MV analysis)Linear β = –0.02, 95%CI (–0.04 to 0.00); P = 0.023
•
Every 1.0-unit higher EDSS at 5 years was significantly associated with a 0.02 cm2 annualized loss in CCA
aCCA loss and annualized EDSS change (MV analysis)Linear β = –0.07 (–0.17 to 0.03); P = 0.14
•
Every annual increase of EDSS by 1.0 unit was associated with a 0.07 cm2 annualized loss in CCA, however, this association was not statistically significant
aCCA loss and conversion to CDMS (MV analysis)Linear β = –0.07 (–0.14, –0.01); P = 0.035
•
CIS patients who had converted to CDMS by 5 years had a significantly greater annual loss in CCA by 0.07 cm2 than those who had not converted to CDMS
Factors adjusted in MV analysis: Age, sex, study site, 5-year intracranial area and whether participants were having a relapse at the time of their 5-year disability measure.
Increased corpus callosum atrophy was associated with physical disability progression in CIS patients converting to MS.
Change in CCI at 3 years follow-up – linear measure Algorithm used: From T1-weighted sagittal slices, CCI was obtained by summing the thickness of the anterior, posterior, and superior segments of the corpus callosum and dividing the result by the anteroposterior diameter.
Baseline EDSS used as a continuous variable
Change in CCI and EDSS (UV analysis)Spearman's r = 0.17; P = 0.27
•
There was no significant association between corpus callosum atrophy (measured using change in CCI) at 3 years follow-up and baseline EDSS score
There was no significant association between corpus callosum atrophy and physical disability in RRMS patients.
Overall: NR Stable CIS: 29 Progressing to CDMS: 28
PCCA change at 6 months follow-up – linear measure Algorithm used: Measured and averaged in seven 3-dimensional reconstructions of T1-weighted sagittal slices per patient using an automated procedure.
EDSS at 2 years used as a continuous variable Conversion from CIS to CDMS used as a binary variable
PCCA loss and EDSS (MV analysis)
•
Those with a faster CCA decrease showed marginally more severe physical disability (identified by EDSS) - detailed quantitative data not reported
PCCA loss in CIS and CDMS groups (MV analysis)
•
Decrease in PCCA over 6 months was significantly more pronounced in the CDMS group than in the CIS group (approximately –3.9% vs. –1.3% over 2 years, respectively; P<0.001 mixed model)
Conversion from CIS to CDMS (MV analysis)
•
Of the patients with a 6-month decrease in PCCA >=1% vs. <1%, 69% vs. 41% reached CDMS, respectively (hazard ratio = 1.8, 95%CI [1.1 to 2.9])
Factors adjusted in MV analysis: Age, sex, and treatment changes.
Six-month corpus callosum atrophy predicted individual risk of developing clinically definite multiple sclerosis in patients with CIS.
PCCA change at 1-year follow-up -linear measure Algorithm used: Measured on T 1 -weighted 3D images, using reconstructions of the sagittal slices.
Change in EDSS from baseline to 9 years and EDSS at 9 years used as a continuous variable
PCCA loss and EDSS change (MV analysis)Linear β= –3.8; P = 0.004
•
Every 1 per cent increase in CCA loss (BVL) from baseline to 1 year was associated with a 3.8-points increase (worsening) in EDSS delta from baseline to 9-years
PCCA loss and EDSS (MV analysis)Linear β = –3.8; P = 0.004
•
Every 1 per cent increase in CCA loss (BVL) from baseline to 1 year was associated with a 3.8-points higher EDSS score at 9 years
Factors adjusted in MV analysis: Patient age, disease duration and baseline EDSS.
Increased corpus callosum atrophy in the initial year of disease-modifying therapy is a simple and reliable individual predictor of future progression of physical disability in RRMS patients.
aCCA: Annualized corpus callosum area; BVL: Brain volume loss; CCA: Corpus callosum area; CCI: Corpus callosum index; CDMS: Clinically definite multiple sclerosis; CI: Confidence interval; CIS: Clinically isolated syndrome; DICOM: Digital Imaging and Communications in Medicine; EDSS: Expanded Disability Status Scale; MS: Multiple sclerosis; MV: Multivariable; NR: Not reported; PCCA: Percentage corpus callosum area; Q: Quality rating of study; RRMS: Relapsing-Remitting multiple sclerosis; UV: Univariate.
Table 9 presents an overview of one primary study reporting data on the association between change in the medullary width and physical disability in MS. This was a prospective cohort study that included CIS patients (n=103) with a follow-up duration of 30 years. The average age of patients was 32 years while the median EDSS at baseline was not reported. Medullary atrophy was measured using change in the width of medulla oblongata (medullary width measurements were performed on sagittal scout images). The physical disability outcome was measured using EDSS and the development of SPMS- or MS-related death.
•
This study, reporting on the association between change in the width of the medulla oblongata and physical disability found that medullary atrophy was significantly associated with a higher (worse) EDSS and an increased risk of SPMS- or MS-related death.
Table 9Association between change in the width of medulla oblongata and physical disability in MS.
Change in the width of the medulla oblongata at 5 years of follow-up - linear measure Algorithm used: Width was measured as the dorsoventral diameter of the medulla on a mid-sagittal image normal to the craniocaudal cord orientation. The level of medullary measurement was determined by the craniocaudal pontine length mirrored caudally from the inferior pontine notch. Medullary width measurements were performed on sagittal scout images.
EDSS at 30 years used as a continuous variable Development of SPMS or MS-related death during the 30 years used as a binary variable
Medullary atrophy and EDSS (MV analysis)Linear β = 0.6, 95% CI (0.37 to 0.82); P = 0.0001
•
Every mm of medullary atrophy within the first 5 years was significantly associated with 0.6 points higher (worse) EDSS at 30 years
Medullary atrophy and development of SPMS or MS-related death (MV analysis)Logistic regression OR = 5.83, 95% CI (1.74 to 19.61); P<0.005
•
Every mm of medullary atrophy within the first 5 years resulted in approximately 6 times significantly increased risk of SPMS or MS-related death by 30 years
Factors adjusted in MV analysis: Age, sex, EDSS change in the first 5 years
Brain atrophy early in the course of MS independently predicted progressive disease and physical disability after 30 years.
Of the 50 studies assessed using the NOS checklist, 65% (n=33) were of medium quality (NOS score of 6-7) and 35% (n=17) were of low quality (NOS score of ≤5).
4. Discussion
4.1 Summary of results
Across all BVL measures, there was considerable heterogeneity in studies regarding the underlying study population (type of MS, baseline EDSS etc.), the definitions of BVL and image analysis methodologies, the physical disability measure used, the measures of association reported and whether the analysis conducted was univariable or multivariable. As a result, direct study-to-study comparisons and a potential meta-analysis were not feasible for the current review, which makes it difficult to quantify the overall magnitude of the association between different BVL measures and physical disability in MS. Notwithstanding these limitations, the following findings are noteworthy.
A total of 36 primary studies providing data on the association between whole BVL and physical disability in MS collectively suggest that whole brain atrophy is associated with greater physical disability progression in MS patients. Similarly, a total of 15 primary studies providing data on the association between ventricular atrophy and physical disability in MS collectively suggest that ventricular atrophy is associated with greater physical disability progression in MS patients. Along similar lines, the existing evidence based on a total of 13 primary studies suggests that gray matter atrophy is associated with greater physical disability progression in MS patients. Only four primary studies provided data on the association between corpus callosum atrophy and physical disability in MS. While the data are limited, they suggest that corpus callosum atrophy is associated with greater physical disability progression in MS patients. Only six primary studies provided data on the association between white matter volume loss and physical disability in MS. While the data are relatively limited, the majority of the existing evidence suggests no association between white matter atrophy and physical disability in MS. Only six primary studies provided data on the association between limbic system structures volume loss and physical disability in MS. In totality, the existing evidence on the association between limbic structure atrophy and physical disability in MS is inconsistent.
Finally, it is difficult to assign a relationship between basal ganglia volume loss and physical disability as well as medulla oblongata width and physical disability in MS due to very limited data (based on three and one primary studies, respectively).
Fig. 4 provides a high-level summary of the relationships between different brain volume loss measures and disability in MS.
Fig. 4Relationship between brain volume loss and disability in multiple sclerosis.
Management of patients with MS has been hampered by the absence of validated, easily implementable biomarkers of neurodegeneration, and predictors of future disability (
Comparing longitudinal brain atrophy measurement techniques in a real-world multiple sclerosis clinical practice cohort: towards clinical integration?.
). A total of 36 primary studies providing data on the association between whole BVL and physical disability in MS collectively suggest that whole brain atrophy is associated with greater physical disability progression in MS patients. This study can help define future imaging biomarkers for physical disability progression and treatment monitoring in MS. Earlier identification of patients who are likely to have worse clinical and economic outcomes over time is important in MS, given the wide spectrum of available treatments that have the potential to reduce the rate of brain atrophy (
Comparing longitudinal brain atrophy measurement techniques in a real-world multiple sclerosis clinical practice cohort: towards clinical integration?.
). Knowing several years in advance which patients will have more severe disease progression can aid treatment decision-making, helping to minimize healthcare costs (
). There seems to be a potentially long window of opportunity to modify the rate of brain atrophy before it manifests itself clinically. Treatments designed to slow or prevent progressive disease should be considered far in advance of its clinical onset (
The major overall limitation of this review is the presence of heterogeneity across studies with respect to multiple dimensions such as study design, patient population, BVL and image analysis techniques, physical disability assessment, as well as statistical methods and measures reported.
The publications were based on either a retrospective or a prospective cohort design, with considerable variability in the underlying study populations. Only 20% studies included all subtypes of MS patients (CIS, RRMS, PPMS and SPMS), while the remainder included either one type of MS patients (RRMS, CIS or PPMS) or a combination of different subtypes (e.g., RRMS+PPMS or RRMS+SPMS). Baseline disability also varied across studies: 18 studies reported a baseline EDSS score of ≤3.0, 10 reported a baseline score >3.0 and ≤4.0, and one study reported a baseline score of >4.0 (21 studies did not report the average EDSS score at baseline). The studies also varied with respect to the BVL outcome measures, for example, some studies reported a change/percentage change in brain volume (
Short-term MRI measurements as predictors of EDSS progression in relapsing-remitting multiple sclerosis: grey matter atrophy but not lesions are predictive in a real-life setting.
)). There was also substantial variability in the way physical disability (EDSS) was measured; with some studies reporting EDSS at a particular timepoint (e.g. at baseline (
), and some classifying patients into different categories based on a certain EDSS threshold (e.g., physical disability progression defined as an EDSS score increase of 1.5, 1.0 or 0.5 points over the time horizon versus no physical disability progression/stable
With respect to the image analysis technique, SIENA was the most commonly used algorithm, followed by SIENAX; although other algorithms were also used, such as VIENA (
Comparing longitudinal brain atrophy measurement techniques in a real-world multiple sclerosis clinical practice cohort: towards clinical integration?.
). The majority (∼70%) of the studies reported multivariable analysis, however, the potential confounders adjusted for in the analyses varied considerably. While most studies adjusted for age, sex, disease duration and baseline EDSS, some also adjusted for other key confounders such as MS subtype (
). Similarly, the measures of association used also varied across studies: most studies reported either a correlation coefficient (∼25%) or a linear regression coefficient (∼20%), while a few studies reported other measures such as odds ratio or hazard ratio.
Some study-specific limitations included small sample size (
Short-term MRI measurements as predictors of EDSS progression in relapsing-remitting multiple sclerosis: grey matter atrophy but not lesions are predictive in a real-life setting.
). Additionally, some studies did not adjust for biologic confounding factors that may have had an impact on brain atrophy, such as dehydration, diurnal brain volume changes, menstrual cycle, pseudoatrophy due to the use of DMTs; cardiovascular, environmental, and genetic risk factors; and comorbidities (
). An additional potential confounding factor in studies based on a limited number of observations that include periods before and after the initiation of DMTs is the accelerated BVL associated with a short-term treatment response (“pseudo-atrophy”;
). Accounting for this effect demands “re-baselining”, which is often done by assessing disease-associated BVL only from a year or more after any the start of anti-inflammatory treatments (
). However, this approach was not generally reported in the studies reviewed here. In addition, the existing data is largely suggestive of an association, as opposed to causation, between BVL and physical disability. This is because the majority of the studies included in this review (40 out of 50) evaluated changes in BVL and physical disability concurrently over the same follow-up period. Only ten studies (
00218-3&id=ga1.jpg){kind=link}
00218-3&id=gr1.jpg){kind=link}
00218-3&id=gr2.jpg){kind=link}
00218-3&id=gr3.jpg){kind=link}
00218-3&id=gr4.jpg){kind=link}