Cervical spinal cord atrophy: An early marker of progressive MS onset

Abstract

Objective: To assess whether cervical spinal cord atrophy heralds the onset of progressive MS.

Methods: We studied 34 individuals with radiologically isolated syndrome (RIS) and 31 patients with relapsing-remitting MS (RRMS) age matched to 25 patients within a year of onset of secondary progressive MS (SPMS). Two raters independently measured (twice per rater) the cervical spinal cord average segmental area (CASA) (mm²) of axial T2-weighted images between C2 and C7 landmarks. The midsagittal T2-weighted image from the end of C2 to the end of C7 vertebra was used to measure the cervical spine (c-spine) length (mm). Sex, age at cervical MRI, number and location of cervical spinal cord lesions, c-spine length, and diagnoses were analyzed against the outcome measures of CASA and C2 and C7 slice segmental areas.

Results: Intrarater and interrater agreement was excellent (intraclass correlation coefficient >0.97). The CASA area (p = 0.03) and C7 area (p = 0.002) were smaller in SPMS compared with RRMS. The C2 area (p = 0.027), CASA (p = 0.004), and C7 area (p = 0.003) were smaller in SPMS compared with RIS. The C2 area did not differ between SPMS and RRMS (p = 0.09). The C2 area (p = 0.349), CASA (p = 0.136), and C7 area (p = 0.228) did not differ between RIS and MS (SPMS and RRMS combined). In the multivariable model, ≥2 cervical spinal cord lesions were associated with the C2 area (p = 0.008), CASA (p = 0.009), and C7 area independent of disease course (p = 0.017). Progressive disease course was associated with the C7 area independent of the cervical spinal cord lesion number (p = 0.004).

Conclusion: Cervical spinal cord atrophy is evident at the onset of progressive MS and seems partially independent of the number of cervical spinal cord lesions.

Classification of evidence: This study provides Class III evidence that MRI cervical spinal cord atrophy distinguishes patients at the onset of progressive MS from those with RIS and RRMS.

Figure. Cervical spinal cord area assessment

(A) C-spine length: first, the midsagittal T2-weighted image was identified. After adjustments for contrast and edge sharpness, the top of the C2-C3 disc space was established as the upper limit. The top of the C7-T1 disc space was established as the lower limit. A vertical line was drawn to connect each vertebra from the highest point in the disc space to the next. The 6 vertical line measurements from the end of C2 to the end of C7 vertebra were summed up to calculate the c-spine length (mm). (B) Cervical spinal cord area: T2-weighted axial MRIs were used for measuring the C2 cross-sectional area, C7 cross-sectional area, and CASA. To draw 2 horizontal lines perpendicular to the mid-spine vertical line at the upper limit and at the lower limit, the top of the C2-C3 disc space and the top of the C7-T1 disc space described in A were used. After the upper and lower limits of the axial images were identified, adjustments for contrast and edge sharpness were made. Then, these landmarks were used to measure the uppermost (C2) and lowermost (C7) cross-sectional areas (mm²). All available axial slices between these limits were measured and averaged for calculating CASA (mm²). CASA = cervical spinal cord average segmental area; c-spine = cervical spine.