Objective: In this multicenter study, a new semiautomatic method for segmenting the cervical cord from C2 to C5 was used to investigate the correlation between cord atrophy and clinical disability in a large sample of patients with multiple sclerosis (MS).
Methods: T2 and 3-dimensional T1-weighted cervical cord scans and dual-echo brain scans were acquired from 143 healthy controls, 22 patients with clinically isolated syndromes (CIS), 101 patients with relapsing-remitting MS (RRMS), 79 patients with secondary progressive MS (SPMS), 58 patients with benign MS (BMS), and 75 patients with primary progressive MS (PPMS) in 3 European centers. Normalized cervical cord cross-sectional area (CSAn) was measured by an active surface cord model. Between-group comparisons were performed using linear mixed-effect models. A nonparametric kernel estimator was used to obtain smoothed plots of CSA along the cervical cord.
Results: Cord CSAn was significantly lower in PPMS vs healthy controls, BMS vs RRMS, SPMS vs BMS, and RRMS. From C2 to C5, a net separation and definition of the plots of patients with BMS, PPMS, and SPMS was seen with respect to those of the other study groups. CSAn was correlated with Expanded Disability Status Scale (r = -0.49, p < 0.0001), with a differential effect among disease clinical phenotypes: no association in either CIS or in BMS; association in RRMS (r = -0.30, p = 0.001), SPMS (r = -0.34, p = 0.001), and PPMS (r = -0.27, p = 0.01).
Conclusions: Cervical cord atrophy provides a relevant and useful marker for the characterization of clinical heterogeneity of patients with MS. The stability of this measure among different centers supports its use as potential outcome measure to monitor disease progression in multicenter trials.