Background

Magnetic resonance imaging (MRI) findings in patients with neuromyelitis optica spectrum disorder (NMOSD) have not previously been studied with data from a prospective, randomized controlled study. During N-MOmentum, longitudinal MRIs were performed systematically.

Objectives

To characterize MRI findings in patients with NMOSD in the N-MOmentum study of inebilizumab. .....................

Methods

MRIs of the spinal cord, optic nerve and brain were performed at baseline, within 8 days of an NMOSD attack and at the end of the randomized controlled period (RCP; month 6.5). MRIs were read centrally by two independent, blinded-to-treatment neuroradiologists for new gadolinium-enhancing (Gd)-T1 enhancement events. Attacks were adjudicated by an expert committee.

Results

Complete MRI data were available for 192 (83%) of 230 participants, 42 of whom had an adjudicated attack (22 myelitis, 14 optic neuritis, 6 multi-domain). The remaining 38 patients did not have valid post-baseline MRI scans available for analysis. Inter-rater agreement between the two neuroradiologists for gadolinium-enhancing lesions was 98% for brain, 95% for spinal cord and 90% for optic nerve.

At the time of acute adjudicated NMOSD attacks, new Gd-T1 MRI enhancement corresponding to the affected clinical domain was present in 19/22 myelitis attacks (86%) and 11/14 optic neuritis attacks (79%). At the time of acute optic neuritis attacks, asymptomatic, new Gd-T1 enhancement was simultaneously observed in 4/14 spinal cord MRIs (29%) and 1/14 brain MRIs (7%). At the time of acute myelitis attacks, asymptomatic, new Gd-T1 enhancement was simultaneously observed in 6/22 optic nerve MRIs (27%) and 3/22 brain MRIs (14%).

In the 150 participants without an adjudicated attack, new Gd-T1 MRI enhancements compared with baseline readings were observed in the brain, spinal cord and optic nerve in 3%, 18% and 51% of patients at the end of the RCP, respectively.

Conclusions

At the time of attack, MRI enhancements were highly correlated to the clinical presentations. However, asymptomatic Gd-T1 enhancements were detected outside the symptomatic attack domain in about one-third of cases. Furthermore, subclinical Gd-T1 enhancements were observed in many patients who did not experience clinically overt attacks. Subclinical blood–brain barrier breakdown, particularly in the optic nerve, may be a frequent phenomenon in patients with active NMOSD.