Longitudinal Identification of Pre‐Lesional Tissue in Multiple Sclerosis With Advanced Diffusion MRI

BACKGROUND AND PURPOSE: Structural MRI (sMRI) is used in monitoring multiple sclerosis (MS) but lacks sensitivity in detecting clinically relevant damage to normal-appearing white matter (NAWM), that is, pre-lesional tissue, and specificity for identifying the underlying substrate of injury. In this longitudinal study, we identified pre-lesional tissue in MS patients and investigated its microstructure by modeling diffusion-weighted imaging (DWI) data using diffusion tensor imaging and neurite orientation dispersion and density imaging (NODDI). METHODS: We enrolled 18 patients with relapsing-remitting MS (10 females, 31.92 ± 8.09 years, disease duration 0.91 ± 1.81 years) and 18 healthy controls (10 females, 31.89 ± 8.15 years). Participants underwent two sMRI and DWI sessions (baseline and follow-up) with the same protocols. Average apparent diffusion coefficient (ADC), fractional anisotropy (FA), orientation dispersion index (ODI), and neurite density index (NDI) were estimated in data-driven regions of interest: nonpersistent lesional tissue (lesional tissue at baseline, resolved at follow-up), lesions that only existed at follow-up (pre-lesional tissue at baseline, lesions at follow-up), persistent lesional tissue (lesions at baseline and follow-up), and NAWM. RESULTS: Compared to NAWM, pre-lesional tissue showed lower ODI, and resolved lesional tissue showed higher FA and ADC and lower ODI and NDI. Over time, persistent lesional tissue showed a decrease in FA and ODI and an increase in NDI. Compared to nonpersistent lesional tissue, persistent lesional tissue showed higher ADC and lower NDI. CONCLUSIONS: DWI and, more particularly, NODDI, can reveal the unique microstructure of persistent, resolved, and pre-lesional tissue in MS.