The value of tests evaluating visual functions in detecting overt or subclinical optic neuritis in multiple sclerosis

Retinal layer thinning as a biomarker of long-term disability progression in multiple sclerosis

BACKGROUND

Peripapillary retinal nerve fibre layer and macular ganglion cell plus inner plexiform layer thinning are markers of neuroaxonal degeneration in multiple sclerosis.

OBJECTIVE

We aimed to investigate the value of peripapillary retinal nerve fibre layer and ganglion cell plus inner plexiform layer thinning for prediction of long-term disability.

METHODS

This is a 6-year prospective longitudinal study on 93 multiple sclerosis patients. Optical coherence tomography scans were performed at baseline, after 1, 2 and 6 years. Primary endpoint was disability progression after 6 years, defined as expanded disability status scale worsening and/or cognitive deterioration. Univariate and multivariate analysis was used to investigate the value of peripapillary retinal nerve fibre layer and ganglion cell plus inner plexiform layer to predict the primary endpoint.

RESULTS

A total of 57 (61.3%) patients had disability worsening, 40 (43.0%) expanded disability status scale worsening and 34 (36.6%) cognitive deterioration. Mean peripapillary retinal nerve fibre layer and ganglion cell plus inner plexiform layer baseline thickness were 93.0 and 75.2 µm, and mean annualised peripapillary retinal nerve fibre layer and ganglion cell plus inner plexiform layer thinning rates over 6 years were 1.3 and 1.6 µm, respectively. Univariate and multivariate analysis revealed lower peripapillary retinal nerve fibre layer and ganglion cell plus inner plexiform layer baseline thickness and higher annualised thinning rates in patients with disability progression after 6 years. Effects were more pronounced for ganglion cell plus inner plexiform layer and expanded disability status scale worsening than for peripapillary retinal nerve fibre layer models and cognitive deterioration.

CONCLUSION

Ganglion cell plus inner plexiform layer and peripapillary retinal nerve fibre layer measurements depict neurodegeneration and predict disability progression in multiple sclerosis.

Low Contrast Visual Acuity Might Help to Detect Previous Optic Neuritis

Optic neuritis (ON) has been considered to be an important factor in the diagnosis of multiple sclerosis (MS) and neuromyelitis optica spectrum disorder (NMOSD), making ON detection increasingly critical for early diagnosis. Furthermore, subclinical ONs presenting no distinct decrease in visual acuity can be missed. Low contrast visual acuity (LC-VA) is known to be able to capture visual loss not seen in conventional high-contrast visual acuity (HC-VA) in MS. Therefore, to increase the sensitivity of ON detection, we investigated the advantage of LC-VA over conventional HC-VA. One hundred and eight patients with demyelinating disease (35 MS, 73 NMOSD) with ON at least 3 months prior and 35 controls underwent neuro-ophthalmic evaluation, including best-corrected conventional high contrast visual acuity (HC-VA) and 2.5% and 1.25% low contrast visual acuity (LC-VA). Receiver operating characteristic (ROC) curve analysis and the area under the curve (AUC) of various visual functions were used to determine the most relevant visual function test for the detection of optic nerve involvement. Additionally, the optimal cutoff point was obtained from the Youden index (J-index) as the points with the best sensitivity-specificity balance. When distinguishing ON from non-ON, the area under the ROC curve (AUC) was highest for the 2.5% LC-VA (0.835, P < 0.001; sensitivity 71.5%, specificity 88.6%), while it was 0.710 (P < 0.001) for the HC-VA and 0.770 (P < 0.001) for the 1.25% LC-VA. In discriminating between controls and ON, the AUC was also highest for the 2.5% LC-VA 0.754 (P < 0.001; sensitivity 71.5%, specificity 78.5%), while it was 0.719 (P < 0.001) for HC-VA and 0.688 (P < 0.001) for 1.25% LC-VA. In eyes with a history of ON (n = 137), the HC-VA and 2.5% LC-VA were abnormal in 64.2 and 71.5%, respectively (P < 0.001), with their combination detecting abnormalities in approximately 85.4% (P < 0.001). The 2.5% LC-VA was superior to HC-VA in detecting ON when distinguishing ON from non-ON or control. The 2.5% LC-VA might be a useful, feasible, and rapid method to detect ON. Furthermore, combining 2.5% LC-VA with conventional HC-VA would be better for detecting optic nerve involvements.