Enhanced detection of optic neuritis with contrast-enhanced 3D FLAIR in orbital MRI

Diagnosing optic neuritis can be challenging, particularly in atypical cases where conventional MRI may be inconclusive. This study evaluates whether adding CE FS 3D FLAIR to standard orbital MRI improves diagnostic performance.

This retrospective single-center diagnostic accuracy study included 48 patients (mean age, 60 years ± 18; 30 women) clinically suspected of acute optic neuritis who underwent orbital MRI between September 2016 and July 2024. All underwent FS 2D T2WI, CE FS 3D T1WI, and CE FS 3D FLAIR. Definite optic neuritis was defined by combined clinical and paraclinical findings. Segment-, nerve-, and patient-based analyses were performed using confidence scores (≥ 4.0 = abnormal). Diagnostic performance of the standard protocol (FS 2D T2WI + CE FS 3D T1WI) and advanced protocol (FS 2D T2WI + CE FS 3D T1WI + CE FS 3D FLAIR) was compared. For nerve-level analyses, including ROC analysis, patient-level cluster bootstrap resampling was applied to account for within-patient correlation between paired optic nerves, yielding cluster-corrected 95% confidence intervals and p values. Patient-level comparisons were evaluated using McNemar's test.

CE FS 3D FLAIR detected optic disc abnormalities in 18 of 55 symptomatic nerves (32.7%) compared with 11 of 55 (20.0%) on CE FS 3D T1WI. The advanced protocol yielded higher AUC (0.93 vs. 0.88, bootstrap p value = 0.003), accuracy (90.6% vs. 80.2%), sensitivity (89.1% vs. 69.1%), and NPV (86.4% vs. 69.6%) than the standard protocol. In atypical presentations, sensitivity for definite optic neuritis improved from 50.0% (13/26) to 76.9% (20/26).

Adding CE FS 3D FLAIR to conventional orbital MRI significantly improves diagnostic accuracy for optic neuritis, especially in atypical clinical presentations.