Optical coherence tomography: A useful tool for identifying subclinical optic neuropathy in diagnosing multiple sclerosis

Care of patients with multiple sclerosis (MS) has evolved in the last decade to shorten the time to confirm the diagnosis of clinically definite MS and to start disease-modifying drugs as soon as possible to decrease permanent disability and the risk of progressive disease. Such trends have been achieved by improving our understanding of the role of paraclinical tests such as the presence of oligoclonal bands on CSF examination, visual evoked potentials, or MRI imaging of the brain or spinal cord. Recently, the evidence that retinal damage detected with optical coherence tomography (OCT) is a sensitive test for identifying MS-related damage in the anterior visual pathway has been added to the diagnostic armamentarium.^1,2 However, symptomatic or subclinical optic nerve lesion as evidence of dissemination in space has not been included in the 2017 MS diagnostic criteria because the available data concerning the diagnostic specificity and sensitivity were considered to be insufficient to justify its inclusion at that time.³ Despite the fact that optic neuritis is one of the most characteristic findings of MS, the optic nerve has not been designated as a fifth lesion site³ because imaging and other data were unavailable to support its inclusion. Recently, Brownlee et al.⁴ found that the inclusion of symptomatic optic nerve involvement in patients with optic neuritis from their clinically isolated syndrome (CIS) cohort improved the accuracy of 2017 MS diagnostic criteria. This study did not include OCT. Today, evidence of optic neuritis may be confirmed clinically by visual evoked potentials or by imaging with MRI or OCT. Subclinical or oligosymptomatic cases can be confirmed with these complementary tests, and new technologies such as OCT can help to characterize the underlying conditions producing optic neuritis. Further refinement of the diagnosis of optic neuritis is now possible with the introduction of the aquaporin 4, glial fibrillary acid protein, and myelin oligodendrocyte protein antibodies, which help to exclude variants of optic nerve inflammation. Therefore, it is time to re-evaluate the inclusion of optic nerve as a fifth lesion site for dissemination in space for the diagnosis of MS.

In this issue of Neurology®, Outteryck and Lopes⁵ provide evidence that the presence of asymmetric retina atrophy by OCT is accurate (close to 90%) for identifying subclinical (and symptomatic) lesions in the optic nerve in patients with a first relapse of MS. They studied a prospective cohort of 130 patients with CIS (30% as optic neuritis) and identified optic nerve lesions (symptomatic and subclinical) with optic nerve MRI (3D double-inversion-recovery) as the gold standard. First, they found a significant frequency (30%) of cases with asymptomatic optic nerve lesions on MRI in patients with CIS with or without optic neuritis. They also highlight the use of the intereye asymmetry of the retinal ganglion cell layer thickness as the best method for assessing retina atrophy, as previously validated by the International Multiple Sclerosis Visual System Consortium (IMSVISUAL).¹ They found that assessment of the thickness of the ganglion cell layer plus inner plexiform layer (GCIPL) is the technique that provides the highest accuracy for diagnosing optic nerve lesions (compared with the retinal nerve fiber layer and low-contrast letter acuity). They also found 2 cutoffs for diagnosing optic nerve lesions using an intereye difference of the GCIPL: (1) in the case of symptomatic optic neuritis, an intereye asymmetry of 2.83 μm had a sensitivity of 88% and specificity of 83% in establishing an optic nerve lesion; and (2) for the detection of asymptomatic optic nerve lesions, an intereye asymmetry of 1.42 μm had a sensitivity of 89% and specificity of 73%. It is important to note that other causes of optic neuritis were excluded (including idiopathic optic neuritis and anti-aquaporin or anti–myelin oligodendrocyte glycoprotein antibodies). Therefore, the patients included in this study had an optic neuritis (symptomatic and subclinical) that was typical of MS.

However, there are some limitations. At present, the resolution of OCT is in the 2- to 3-µm range, and the test-retest variability is also in the 2- to 3-μm range. Finally, healthy controls may have this level of intereye difference. However, as the authors acknowledge, prospective studies aimed at defining the risk of conversion to clinically definite MS are warranted. Furthermore, advances in OCT technology continue to improve its resolution and the analysis using machine learning tools⁶ and may make our measurements even more precise.

The alternative of identifying subclinical lesions in the optic nerve by MRI has limitations because the noise produced by the bone and fat can limit interpretation. A 3D double-inversion-recovery magnetic resonance sequence of the orbit is the most sensitive technique for detecting optic nerve lesions⁷ but can be time-consuming compared to OCT and therefore less practical and efficient in the clinical setting.

OCT has been shown to capture relevant tissue damage in the anterior optic pathway that can predict clinical outcome after optic neuritis or in chronic MS.^8,–,10 OCT is now being proposed as a paraclinical tool to help establish evidence of dissemination in space for the diagnosis of MS. OCT is accessible at most of the MS centers or in collaboration with ophthalmologists. It is well tolerated by patients, is relatively inexpensive, and typically takes <5 minutes to perform. This valuable tool is now being integrated into the evaluation of patients with demyelinating diseases such as MS and neuromyelitis optica to speed the diagnosis and to support the decision-making process.

• © 2020 American Academy of Neurology