Background

Optic nerve (ON) involvement is frequent in multiple sclerosis (MS) and its inclusion as the fifth anatomical location for dissemination in space (DIS) for the diagnosis of MS has been proposed in 2016 by the Magnetic Resonance Group Imaging in Multiple Sclerosis (MAGNIMS). However, there was insufficient evidence to support this recommendation.

Objectives

To investigate the effect of including ON involvement assessed by visual evoked potentials (VEP) as the fifth location in DIS criteria for MS diagnosis, in patients with a typical clinically isolated syndrome (CIS).

Methods

We studied consecutive patients presenting with CIS between 2012 and 2019 from two Portuguese hospitals with complete initial evaluation, including brain and spine MRI and VEP.

McDonald 2017 criteria and a set of modified criteria that included ON involvement in DIS assessed by VEP were applied retrospectively. Performance of the two sets of criteria to predict development of clinically definite multiple sclerosis (CDMS) and/or MRI activity during follow-up was evaluated.

Results

76 patients were included, 25% of which had an ON CIS. ON asymptomatic involvement on VEP was found in 12.3% of non-ON CIS. 27 (35.5%) patients converted to CDMS and 37 (48.7%) had MRI activity during follow-up (mean=3.79 years, range 1.04 – 8.36 years). 59.2% of the patients began disease modifying treatment (DMT) before conversion to CDMS.

Modified DIS criteria in combination with dissemination in time (DIT) was slightly more sensitive to predict CDMS (77.8% vs 74.1%), but less specific (57.1% vs 61.2%). When the outcome was CDMS or MRI activity during follow-up, modified DIS+DIT criteria were more sensitive (73.2% vs 65.9%) with equal specificity (65.7% vs 65.7%), but these differences were not statistically significant.

Modified criteria allowed for the diagnosis of 3 additional patients at baseline (42/76 vs 39/76), in average 6 months before fulfilment of McDonald2017 criteria and subsequent initiation of DMT.

Conclusions

Although inclusion of ON involvement on DIS criteria led to accurate identification of more MS patients, in our sample it did not allow for statistically significant increase in sensitivity for MS diagnosis despite asymptomatic involvement of ON assessed by VEP at CIS diagnosis being frequent.