To validate the 2024 McDonald inter-eye difference (IED) cutoff for optical coherence tomography (OCT) in a large multiple sclerosis (MS) cohort, to assess performance in detecting asymptomatic optic nerve lesions on MRI, and potential discrepancies between OCT and MRI

A total of 785 consecutive MS patients underwent spectral-domain OCT to assess peripapillary retinal nerve fiber layer (pRNFL) and ganglion cell–inner plexiform layer (GCIPL) thickness. A subset of 260 patients underwent 3T MRI within three months of OCT, including 3D double inversion recovery.

The final cohort included 740 patients (27% with a history of ON). MRI was analyzed in 240 patients. The 2024 McDonald IED cutoff (GCIPL ≥4 µm or pRNFL ≥6 µm) showed 83% sensitivity, 55% specificity, and 63% overall accuracy for identifying prior ON. In the MRI subgroup, comparable performance was observed for detecting optic nerve involvement on MRI in patients without clinical ON. MRI detected lesions in 30% of asymptomatic eyes from (46%) patients without a history of ON, while OCT-IED was positive in 27% of asymptomatic patients. Concordance between OCT and MRI was 77% overall and 90% in eyes with previous ON. Among asymptomatic patients 25 were OCT-negative/MRI-positive (11 with bilateral optic nerve involvement) and 15 were OCT-positive/MRI-negative, mostly with isolated GCIPL or pRNFL asymmetry.