Adding the Optic Nerve in Multiple Sclerosis Diagnostic Criteria: A Longitudinal, Prospective, Multicenter Study

BACKGROUND AND OBJECTIVES: The optic nerve is not one of the areas of the CNS that can be used to demonstrate dissemination in space (DIS) within the 2017 McDonald criteria for the diagnosis of multiple sclerosis (MS). Objectives were (1) to assess whether optic nerve-MRI (ON-MRI), optical coherence tomography (OCT), and visual evoked potentials (VEP) detect optic nerve involvement in clinically isolated syndrome (CIS) and (2) to evaluate the contribution of the optic nerve topography to the current diagnostic criteria in a prospective, multicenter cohort. METHODS: MAGNIMS centers were invited to provide prospective data on patients with CIS who underwent a visual assessment with at least 2 of 3 investigations (ON-MRI, OCT, or VEP) within 6 months of onset. Modified DIS criteria were constructed by adding the optic nerve topography, defined by each investigation separately and any combination of them, as the fifth area of the CNS. A risk assessment analysis and the performance of the different DIS criteria were analyzed using the diagnosis of MS according to the 2017 McDonald criteria as the primary outcome and new T2 lesions and/or a second relapse as the secondary outcome. RESULTS: We included 157 patients with CIS from 5 MAGNIMS centers; 60/157 (38.2%) patients presented with optic neuritis. Optic nerve involvement on ON-MRI was found in 40.2% patients at study entry and in 72.5% of those with optic neuritis.At follow-up (mean 27.9 months, SD 14.5), 111/157 patients (70.7%) were diagnosed with MS according to the 2017 McDonald criteria. Fulfilling either 2017 DIS or any modified DIS criteria conferred a similar high risk for reaching primary and secondary outcomes. The modified DIS criteria had higher sensitivity (92.5% [with ON-MRI] vs 88.2%), but slightly lower specificity (80.0% [with GCIPL IEA ≥4 µm] vs 82.2%), with overall similar accuracy (86.6% [with ON-MRI] vs 86.5%) than 2017 DIS criteria. Consistent results were found for secondary outcomes. DISCUSSION: In patients with CIS, the presence of an optic nerve lesion defined by MRI, OCT, or VEP is frequently detected, especially when presenting with optic neuritis. Our study supports the addition of the optic nerve as a fifth topography to fulfill DIS criteria.

Detection of lesions in the optic nerve with magnetic resonance imaging using a 3D convolutional neural network.

BACKGROUND: Optic neuritis (ON) is one of the first manifestations of multiple sclerosis, a disabling disease with rising prevalence. Detecting optic nerve lesions could be a relevant diagnostic marker in patients with multiple sclerosis. OBJECTIVES: We aim to create an automated, interpretable method for optic nerve lesion detection from MRI scans. MATERIALS AND METHODS: We present a 3D convolutional neural network (CNN) model that learns to detect optic nerve lesions based on T2-weighted fat-saturated MRI scans. We validated our system on two different datasets (N = 107 and 62) and interpreted the behaviour of the model using saliency maps. RESULTS: The model showed good performance (68.11% balanced accuracy) that generalizes to unseen data (64.11%). The developed network focuses its attention to the areas that correspond to lesions in the optic nerve. CONCLUSIONS: The method shows robustness and, when using only a single imaging sequence, its performance is not far from diagnosis by trained radiologists with the same constraint. Given its speed and performance, the developed methodology could serve as a first step to develop methods that could be translated into a clinical setting.