BERLIN—Spinal cord volume deficits in patients with multiple sclerosis (MS) may explain clinical disability that appears out of proportion to lesion load on brain imaging, according to research presented at ECTRIMS 2018.
In a pool of 362 patients with mild to moderate MS-related disability and identical white matter lesion load identified by MRI, those with higher disability had significantly lower spinal cord volumes, compared with those with mild disability.
An Analysis of Patient Records
Some patients with MS have a relatively high level of disability, but a low burden of white matter intracerebral lesions on MRI. Little is known about spinal cord volume in patients with MS and a pronounced dissociation between intracerebral lesion load and disability, said Michaela Andelova, MD, of Charles University in Prague.
Dr. Andelova and her colleagues hypothesized that spinal cord volume would differ between patients with varying levels of disability despite identical white matter lesion load. To test this hypothesis, they looked at records for 1,245 patients with relapsing-remitting MS. They divided them into three groups by severity of clinical disability and extent of cerebral T2 hyperintense lesion load. The investigators identified 53 patients whose total volume of T2-weighted hyperintense lesions was less than 3 mL, but whose Expanded Disability Status Scale (EDSS) scores were at least 3.5. They called them the low lesion load–high disability (LLHD) group.
The researchers then identified another 71 patients with a volume of T2-weighted hyperintensities greater than 9 mL, but whose EDSS score was less than 1.5. They called them the high lesion load–low disability (HLLD) group.
The remaining 1,121 patients, who did not have these paradoxical associations, were analyzed separately.The investigators measured mean upper cervical cord area (MUCCA) for all patients. On the basis of images acquired by a 3-T MRI scanner, they used an in-house, semiautomated method to calculate MUCCA as the mean sum of spinal cord area in 21 slices centered at the C3–C4 intervertebral disk.
Mean Upper Cervical Cord Area Correlated With Disability
“Despite higher disability, LLHD patients demonstrated significantly higher normalized total brain volume, higher normalized volumes of the thalamus and callosum, and smaller lateral ventricles than the HLLD group,” said Dr. Andelova and her collaborators.
However, the LLHD patients had MUCCA values that were significantly lower than those of the other groups. The nonparadoxical group’s mean MUCCA was 84.02 mm2, while the HLLD group had a mean MUCCA of 85.75 mm2. The difference between these groups was not statistically significant. By contrast, the LLHD group’s mean MUCCA was significantly smaller than that of the other groups, at 80.40 mm2.
Next, Dr. Andelova and her colleagues compared 362 patients with moderate disability (ie, EDSS scores between 3.5 and 6.5) with matched patients who had mild MS-related disability (ie, EDSS score less than 3) and identical cerebral lesion loads. They found that MUCCA was significantly smaller in the group with moderate disability (78.86 mm2 vs 84.44 mm2).
In addition to having identical lesion loads, the mild and moderate disability groups had similar normalized total brain volume and regional brain volumes. The group with moderate disability had slightly less white matter volume, said Dr. Andelova. All differences between groups retained statistical significance after adjustment for potential confounders such as age, sex, and duration of disease.
“Reduced spinal cord volume may explain part of the clinical–radiologic paradox in patients who have high disability despite low intracranial lesion load,” said the researchers. “In line with this finding, relatively preserved spinal cord volume may be associated with functional reserve and less physical disability in patients with low disability despite high cerebral lesion load.”
Dr. Andelova and her collaborators plan to examine cerebral lesion distribution and perform quantitative MRI investigations of lesion distribution. They intend to seek potential associations between various distribution patterns and accelerated spinal atrophy.
—Kari Oakes