Optic neuritis (ON) refers to an inflammatory injury of the optic nerve, which can manifest as a CIS or, in some cases, represent a harbinger for the diagnosis of MS. With an incidence of 1–5 per 100,000 per year [2, 11, 12], ON is a common cause of acquired vision loss in young adults and the best-characterized CIS [9]. In fact, 1 in every 5 MS patients presents with MS associated (MSON) as the first clinical manifestation of their disease [9]. Much of what we have come to understand in terms of the epidemiology and clinical presentation of typical MSON has been based on the initial experience and subsequent follow-up from the Optic Neuritis Treatment Trial (ONTT) [13, 14]. This randomized, multicenter study was initially designed to compare the benefits of treatment with either intravenous methylprednisolone (IVMP) (250 mg administered every 6 h for 3 days followed by oral prednisone [1 mg/kg/day] for 11 days), oral prednisone (1 mg/ kg/day), or oral placebo in 457 patients with acute ON [2]. From the ONTT, we learned that most MSON patients are Caucasian (85 %) women (77 %), with a mean age of 32 years [2, 13, 14]. In adults, the majority of ON cases are unilateral, but occasionally bilateral simultaneous vision loss is observed. Yet, in this setting, ON mimics need to be considered including: neuromyelitis optica (NMO), toxic–metabolic optic neuropathies, and Leber’s hereditary optic neuropathy (LHON).

From the point of view of their clinical presentation, MSON patients often report subacute onset vision loss that worsens over hours to days. Ninety-two percent of affected individuals experience pain within the 1st week of symptom onset, which is frequently provoked by eye movements [2]. Approximately one-third of individuals affected by MSON note flashes of light in the affected eye, known as photopsias or phosphenes [15, 16], albeit they may not divulge this information without prompting. When the diagnosis is suspected, there are several localizing features on initial examination, which are key to securing the diagnosis. Initially, the severity of vision loss in the affected eye (MSON eye) may range from mild (Snellen visual acuity equivalent of 20/20) to no light perception. In patients with unilateral optic nerve involvement, a relative afferent pupil defect (RAPD) will be apparent in the MSON eye. Visual field defects follow the topography of the retinal nerve fiber layer (RNFL). Cecocentral, altitudinal, and arcuate patterns of vision loss are often observed. Dyschromatopsia, or decreased color vision, is also common. This finding can be particularly helpful in localizing the diagnosis in patients with mild central vision loss and disproportionate deficits in color vision function [2]. In cases of retrobulbar MSON, the fundus examination is initially normal, whereas patients with anterior MSON (sometimes referred to as “papillitis”) manifest mild to moderate optic disc swelling acutely. The ONTT demonstrated that severe optic disc edema, vitreous cell, and hemorrhage are relatively uncommon findings in MSON patients and may herald a mimic such as neuroretinitis. Not surprisingly, these atypical fundus features are associated with a reduced risk of developing MS, potentially because the patient does not in fact have MSON. The prognosis for recovery after MSON is generally favorable, with approximately 95 % of patients achieving a visual acuity of 20/40 vision in their affected eye a year after clinical presentation [2, 11, 17]. Despite regaining “normal” vision, however, many MSON patients report persistent problems including fatigue and heat-induced (Uhthoff’s phenomenon) vision loss, altered motion and depth perception (Pulfrich phenomenon), and decreased spatial vision at low-contrast levels [2]. There is therefore discordance between what patients report versus what is captured with standard ophthalmic testing in the setting of post-acute MSON, indicating a need for more sensitive measures of vision loss in this patient population.

As a putative model of MS, the afferent visual pathway (AVP), with MSON as its relapse “prototype,” offers several potential advantages. First and foremost, there is the benefit of precise localization. In the setting of acute ON, the AVP model provides objective evidence of a symptomatic lesion in any optic nerve, which can be anatomically localized in the CNS [2]. As a second point of merit, more than 90 % of ON patients report pain at initial presentation. In light of the highly specialized nature of the afferent visual pathway, any perturbation in the system that interferes with visual perception, particularly central vision, will be noticed and reported by affected individuals. Functional recovery can therefore be monitored from a relatively precise point of onset in the AVP model of MS [2]. Thirdly, the afferent visual pathway is a functionally eloquent CNS system and deficits can be captured with reproducible measures of visual function including high- and low-contrast visual acuity, automated perimetry, and color vision testing. Furthermore, optical coherence tomography (OCT) provides structural measures of neuronal and axonal integrity. By pairing OCT measures with quantitative visual outcomes, it is possible to devise a structural–functional paradigm to elucidate the temporal evolution and relative contributions of inflammation, axonal loss, neuronal damage, and cortical compensation to post-MSON recovery in the AVP model of MS [2]. Finally, previous pathological studies have shown that tissue-specific injury in the AVP mirrors global CNS effects in MS patients [2, 18]. Simply put, the back of the eye is the front of the brain. At the time of an acute MSON event, cytokine release induces transient conduction block, likely induced by nitric oxide [2, 19]. When myelination and axonal integrity are intact, recovery ensues with the removal of inflammatory mediators. During recovery from MSON, remyelination improves saltatory conduction through sodium channels, which are distributed along the demyelinated optic nerve segment [9]. Cortical plasticity is also believed to play a role in optimizing function in the more chronic phases of recovery, albeit the timeline and mechanisms involved therein are not well understood. The AVP model can therefore be interrogated to monitor tissue-specific factors that underpin CNS injury and repair in MS patients [2].

Optical coherence tomography (OCT) provides a noninvasive means of measuring neuroaxonal injury in the anterior visual pathway [20–22]. As the “optical analog” of ultrasound B-mode imaging, OCT facilitates in vivo imaging of retinal structures including the retinal nerve fiber layer (RNFL), ganglion cell layer (GCL), and inner nuclear layer (INL). Since the retina is typically devoid of myelin, OCT measures provide an ideal opportunity to capture the manifestations of CNS neurodegeneration, neuroprotection, and, potentially, neurorepair. Decrements in peripapillary RNFL thickness as measured by OCT have been interpreted to represent axonal damage [2]. Because the macula contains a large proportion of retinal ganglion cell neurons (about 34 % of total macular volume) [22], changes in macular volume and, more recently, GCL thickness have been interpreted to represent altered neuronal integrity in the afferent visual pathway. Widespread availability of spectral-domain OCT has allowed us to obtain high-resolution imaging (5–7 μm) of retinal structures 50-fold faster than previous time-domain OCT models [21, 22]. As OCT continues to evolve, future studies will be able to employ novel techniques including longer wavelength and swept source technology, ultrahigh-resolution OCT, en face imaging, and polarization-sensitive OCT that will enhance our ability to diagnose, monitor, and treat MS patients [23].

At the time of an acute MSON event, when vision loss is at its nadir, patients often manifest peripapillary RNFL measurements that are increased in their MSON eye relative to their non-MSON eye [2]. Correspondingly, the optic nerve in the MSON eye may be mildly edematous or hyperemic secondary to axoplasmic flow stasis. In contrast, at baseline OCT-measured macular volume and GCL measures are symmetric between MSON eyes and non-MSON eyes in the setting of acute ON [2]. Ganglion layer (GCL) thinning may be the first manifestation of retrograde neuronal loss detectable in the AVP model, because the ganglion layer “signal” is not obscured by superimposed edema [24]. In the ensuing 2–3 months, RNFL, GCL, and macular volume thinning evolve, with earliest signs of significant RNFL atrophy often manifesting in the temporal region [2]. These OCT-measured changes are commensurate with the evolution of optic disc pallor. Collectively the studies to date have shown that RNFL, GCL, and macular volume continue to decrease for 6–12 months after symptom onset, plateauing thereafter [2]. Yet, visual recovery 12 months after ON has not been linked to the extent of peripapillary RNFL swelling seen acutely but has instead been associated with the amount of RNFL, macular volume, and ganglion layer loss observed 6–12 months after symptom onset [2].

In the era of retinal segmentation with spectral-domain OCT, INL thickening and microcystic macular edema have been explored as potential markers of inflammation in ON patients [25, 26]. In a study by Kaushik and colleagues [26], INL thickening was shown to correlate with lower RNFL values and reduced GCL thickness in ON eyes. Thus, more severe inflammation in the retina may be associated with more severe neuroaxonal injury in the afferent visual pathway of ON patients.

As previously stated, the natural history of MSON is favorable, with the vast majority of patients achieving 20/20 or better Snellen acuity equivalent visual function in their MSON eyes a year after the acute event [11]. Kupersmith and colleagues [27] identified early functional and structural markers that may help predict recovery after MSON. Specifically, 1 month after presentation, high-contrast visual acuity values less than or equal to 20/50, contrast sensitivity measures less than 1.0 log units, and visual field mean deviation measures less than or equal to −15.0 dB (as determined by automated perimetry) predicted moderate to severe outcomes for each of these measures at 6 months. In a prospective study involving 25 MSON patients, Kupersmith [28] also demonstrated that early OCT-measured RNFL loss at 1 month was predictive of RNFL thinning at 6 months in MSON eyes [28]. Other studies have shown that a year after an isolated ON event, peripapillary RNFL measurements are reduced by approximately 20 % relative to the fellow eye [2]. In a meta-analysis of time-domain OCT studies [14 studies (2,063 eyes),] RNFL values were reduced from 5 to 40 μm (averaging 10–20 μm) in MSON eyes [29]. Furthermore, comparing MSON eyes with the eyes of healthy controls showed an estimated average RNFL loss of 20 μm [29]. Lower RNFL values have been shown to correlate with reduced visual acuity, visual field mean sensitivity, and color vision scores [2], reinforcing the notion that form and function are lightly linked in the AVP model of MS. For patients selected without recruitment bias, an OCT “cutoff” of 75 μm has been shown to represent a threshold of RNFL integrity that can predict the extent of visual recovery after MSON [30]. The findings to date suggest that MSON may be the “relapse prototype” in the AVP model through which the acute consequences of an optic nerve injury can be tracked, and structural–functional correlations can be established over a defined time period. To this end, the impact of patient-related factors including age, gender, and ethnicity on recovery from a relapse event may also be explored in the AVP model. In a prospective study involving 39 men and 105 women with acute ON, men showed more apparent loss of RNFL thickness in their ON eyes from baseline to 6 months than women [31]. These findings raise interesting questions about the potential influence of gender in MS, which may be explored in future studies.

Optical coherence tomography studies have also revealed evidence of afferent visual pathway involvement in patients with nonvisual CIS, which raises interesting questions regarding mechanisms of subclinical neuroaxonal injury in the CNS of MS patients. Oberwahrenbrock and colleagues [32] used spectral-domain OCT to compare 45 CIS patients with age/sex-matched controls. Patients were stratified into categories based upon the presence or lack thereof of prior MSON by history. The subgroups in this study included CIS eyes with clinical MSON, CIS eyes with suspected subclinical MSON and CIS eyes unaffected by MSON. Interestingly, the eyes of CIS patients with no evidence of clinical or subclinical MSON showed significant reduction of GCL and INL layer thicknesses, with a topography similar to that of MSON eyes. Moreover, CIS eyes with suspected subclinical MSON showed significant RNFL thinning, albeit the most pronounced thinning was noted in CIS eyes with an established history of MSON. The observation that GCL damage is detectable in CIS eyes without clinical history of MSON, was interpreted to support the concept that neuroaxonal loss can occur in the CNS of MS patients without preceding or concomitant demyelination. Thus, it is conceivable that OCT may be used to detect evidence of primary neurodegeneration in the AVP Model of MS. Finally, in this study, loss of GCL thickness in the eyes of nonvisual CIS patients also showed that neurodegeneration is not limited to advanced disease stages but a feature that occurs early in disease development [32]. Notably, the findings described by Oberwahrenbrock et al. [32] were in disagreement with prior reports, which have shown equivocal results perhaps owing to the limits of time-domain OCT technology [33, 34]. For example, in a time-domain OCT by Outteryck et al. [33], peripapillary RNFL values and macular volumes failed to distinguish CIS patients (n = 56, including 18 patients with MSON and 38 patients without MSON) from control subjects.

Few studies have explored the association between OCT measures in the afferent visual pathway at the time of an acute inflammatory relapse and future risk of MS. In a prospective study of 50 CIS patients with isolated MSON, there were no significant differences in RNFL thickness in either MSON eyes or non-MSON eyes observed between CIS patients who developed clinically definite MS (42 %) and those who did not develop MS (58 %) during the 2-year study period [35]. Moreover, a link between initial RNFL values and baseline evidence of CNS inflammation on magnetic resonance imaging (MRI) as future predictors for the development of MS in CIS patients has not been proven. In a previous study, 50 patients who experienced MSON as a CIS were followed for a mean period of 34 months with OCT testing [36]. RNFL values in MSON eyes and non-MSON eyes were compared between patients with MRI evidence of white matter lesions and patients with normal baseline MRI findings, over a 2-year period. Twenty-one patients (42 %) developed clinically definite MS (CDMS) during the study. After 2 years, temporal RNFL values were thinner in MSON patients with MRI lesions at baseline, but the results were not significant. In a time-domain OCT study, Outteryck [33] evaluated 56 CIS patients and 32 control subjects to investigate whether OCT measures of RNFL thickness and macular volume revealed early axonal loss in the afferent visual pathway. In this prospective case series, there was no link between RNFL and MRI evidence of CNS inflammation at baseline, disseminated CNS inflammation according to the revised McDonald criteria, gadolinium enhancement on initial MRI, multifocal CIS presentation, altered visual-evoked potentials, or development of “McDonald”-proven MS at 6 months. These investigators concluded that OCT does not predict conversion to MS at 6 months in CIS patients [33]. Yet, in a more recent spectral-domain OCT study performed by Perez-Rico et al. [37] involving 29 CIS patients without MSON, OCT-measured RNFL thickness was found to be an independent predictor of clinically definitive MS diagnosis at 12 months. Based upon their findings, the authors concluded that retinal axonal loss measured by OCT is an important prognostic factor of conversion to MS in patients with CIS in the absence of symptomatic MSON [37]. Again, there may be discrepancies between published reports that reflect differences in the CIS patient population being evaluated, as well as disparity between the generations of OCT technology being used. As OCT continues to evolve, it may be possible to pair RNFL, GCL, INL and other paraclinical measures of CNS neuroaxonal integrity to identify which CIS patients manifest subclinical dissemination of lesions in space and time that portend a greater risk for developing MS.

For most MS patients, the hallmark of the diagnosis is change, both in terms of neurological disability and the MRI-measured burden of disease over time. At this point, a role for OCT in complementing conventional tools used to diagnose and monitor disease activity is beginning to emerge.