Diagnostic accuracy of optical coherence tomography inter‐eye percentage difference for optic neuritis in multiple sclerosis. (9th October 2017)

• Record Type: Journal Article
• Authors: Coric, D.; Balk, L. J.; Uitdehaag, B. M. J.; Petzold, A.
• Is Part Of: European journal of neurology. Volume 24:Number 12(2017:Dec.)
• Page Start: 1479
• Page End: 1484
• Publication Date: 2017-10-09
• Subjects: diagnostic criteria -- multiple sclerosis -- optic neuritis -- optical coherence tomography; Neurology -- Periodicals; Nervous system -- Diseases -- Periodicals; 616.8
• DOI: 10.1111/ene.13443
• Languages: English
• ISSNs: 1351-5101

Abstract

Abstract : Background and purpose: Multiple sclerosis‐associated optic neuritis (MSON) causes atrophy of the inner retinal layers, which can be quantified by optical coherence tomography. It has been suggested that the inter‐eye percentage difference (IEPD) of atrophy may be of diagnostic value in MSON. Methods: This was a prospective, cross‐sectional study in patients with multiple sclerosis and healthy controls (HCs). Spectral‐domain optical coherence tomography of both eyes was performed, followed by automated retinal layer segmentation of the peri‐papillary retinal nerve fibre layer (pRNFL) and macular ganglion cell and inner plexiform layer (mGCIPL). Receiver operator characteristics curves were plotted and the area under the curve was calculated for group comparisons of the IEPD of the pRNFL and mGCIPL. Results: There were 39 patients with bilateral MSON, 62 patients with unilateral MSON, 106 patients without MSON and 63 HCs. Diagnostic accuracy (area under the curve) of the IEPD was 0.73–0.86 for the pRNFL and 0.75–0.94 for the mGCIPL. The diagnostic sensitivity of the mGCIPL IEPD was 70% with a specificity of 97% for distinguishing unilateral MSON from HCs. For the comparison of bilateral MSON with HCs, sensitivity was 86% with a specificity of 97%. Conclusions: The IEPD of the pRNFL, and more particularly the IEPD of the mGCIPL, is a useful diagnostic measure for MSON. The IEPD is a dimensionless unit and may therefore contribute to overcome device and proprietary segmentation algorithm limitations.