Exploring New Imaging Biomarkers: 3D Conformational Analysis Of The Medulla And Upper Cervical Spinal Cord In People With Relapsing Multiple Sclerosis
Tatum M Moog1, Janneke Van Wingerden2, Bruno Padrazzi3, Svend S Geertsen4, Andreea M Rawlings4, Darin T Okuda1
1The University of Texas Southwestern Medical Center, Department of Neurology, Neuroinnovation Program, Multiple Sclerosis & Neuroimmunology Imaging Program, Dallas, TX, USA, 2Sanofi, Amsterdam, The Netherlands, 3Sanofi, Paris, France, 4Sanofi, Cambridge, MA, USA
Objective:
Explore three-dimensional (3D) medulla-upper cervical spinal cord (MUCSC) conformational imaging metrics as biomarkers in multiple sclerosis (MS).
Background:
Measures of spinal cord atrophy are associated with MS disease progression but have yet to emerge as reliable biomarkers of treatment efficacy. 3D MUCSC conformational analysis is a new technique that may provide additional information about neurodegenerative changes in MS beyond conventional magnetic resonance imaging metrics. In the placebo-controlled Phase 3 TEMSO trial (NCT00134563), teriflunomide 14 mg/day lowered the risk of disability worsening and reduced brain volume loss in participants with relapsing MS. 3D MUCSC conformational analysis of the TEMSO dataset has not been explored.
Design/Methods:
TEMSO baseline and 2-year end of study follow-up 3D T1-Magnetization Prepared Rapid Gradient Echo (MP-RAGE) images acquired primarily at 1.5T were assessed for quality, then segmented and analyzed using in-house proprietary software. Change from baseline in MUCSC volume, local curvature and surface texture were summarized with descriptive statistics.
Results:
Quality assessments led to the exclusion of 481 of the 799 TEMSO participants, comprising 64.0%, 62.1%, and 54.5% of participants in the placebo, 7 mg/day, and 14 mg/day arms, respectively. Mean age and disease duration were lower for placebo participants versus teriflunomide arms. No between-group differences were apparent for mean (SD) change from baseline in MUCSC volume (-86±236, -85±218, and -75±218 mm3 for placebo, 7 mg/day, and 14 mg/day arms, respectively), mean curvature (0.01±0.06, 0.02±0.07, and 0.02±0.07 mm), or surface texture (-15.6±28.0, -21.3±26.6, and -19.0±24.2 triangle count).
Conclusions:
3D conformational analysis of MP-RAGE images from the TEMSO trial showed no discernable between-group differences for MUCSC surface features or volume; however, imbalanced exclusion of participants during quality assessment likely introduced attrition bias. Further development of 3D conformational metrics as MS biomarkers may better illuminate mechanisms related to neurodegenerative changes and treatment efficacy.