The aim of this study is to address the effect of local inflammatory activity in brain and retinal neuro-axonal injury along MS course.

Neuroprotection in multiple sclerosis (MS) needs a correct understanding of the time interplay between inflammation and neurodegeneration.

168 patients with CIS or MS from MS-Visualpath cohort (February 2011-May 2018) were assessed for eligibility. We excluded nine and thirty participants for retinal and brain models, respectively due to absence of follow-up data, disease duration>30 years, poor image quality. Participants underwent neurological, ophthalmological, optical coherence tomography (OCT) and magnetic resonance imaging (MRI) examinations yearly up to three years of follow-up and biannually thereafter. We modelled dynamics of retinal thinning (peripapillary retinal nerve fiber layer–pRNFL–; ganglion cell plus inner plexiform layer–GCIPL– and inner nuclear layer-INL) and brain volume loss (whole-brain, gray-matter and thalamus) throughout MS course. We evaluated the effect modification of local inflammatory activity (active ≥ relapse or lesion) on retinal and brain changes over disease duration in mixed effect models additionally including age, sex, use of disease-modifying therapies, use of steroids (brain) and previous optic neuritis (retina).

159 patients [110 (69.2%) women; median (interquartile range: IQR) age of 40.7 (33.7-48.8) years] were included. Rates of retinal and brain changes were more pronounced during the 5 first years of the disease for all estimates except INL that was constant over MS course. Participants with active MS displayed significantly greater GCIPL thinning (p-value=0.002), whole-brain (p-value=0.038) and thalamus (p-value=0.001) volume loss than stable patients during first 5 years. Thereafter, active and stable MS displayed overlapping trends of steady rates of neuro-axonal injury.

Local inflammatory activity rules neurodegeneration in early MS. Best neuroprotective strategy for MS is haltering inflammatory activity during first years. Other approaches may be explored to curve late neurodegeneration.