Visual acuity typically is decreased. Occasionally, optic neuritis can cause pain without noticeable visual loss. Perimetry shows scotomata. In 448 patients, visual loss was diffuse in 48%, central in 8%, and attitudinal in 20%; there were other defects in 23% (Keltner et al 1993). Other studies find central scotomata in 68% of patients (Gerling et al 1998). Sixty-nine percent of patients have visual field involvement in the fellow eye at baseline.
Some patients with multiple sclerosis have no history of optic neuritis and have normal visual acuity. In these patients, subclinical optic tract lesions are detected with visual evoked potentials (82%), contrast sensitivity tests (73%), pupillary light reflex (52%), flight of colors tests (36%), and color vision tests (Ishihara plates) (32%) (van Diemen et al 1992a). Loss of color saturation is traditionally most pronounced for red hues; but acutely, blue-yellow defects may be more common (Katz 1995). The 25-item National Eye Institute Visual Function Questionnaire (VFQ-25) is modestly correlated with these tests. Color vision is highly correlated with retinal nerve fiber layer thickness on OCT.
It was initially difficult to detect acute optic neuritis on MRI because of the small size of the nerve and because it is surrounded by fat that interferes with conventional T1 and T2 MRI. Short tau inversion recovery (STIR) MRI is more sensitive and reveals high-signal lesions in 85% of affected nerves and 20% of unaffected nerves (Miller et al 1988a; Gass and Moseley 2000). In the asymptomatic eye, there is sometimes no slowing of visual evoked potentials, even when there are MRI lesions (Hornabrook et al 1992). Diffusion tensor imaging (DTI) shows lesions undetectable with conventional MRI (Naismith 2008, personal communication). More radial diffusivity on DTI correlates with decline in vision. Fast spin echo (FSE) and fluid attenuated inversion recovery (FLAIR) with fat suppression further improves imaging.
Fat-suppressed T2 MRI (STIR) shows swelling of the nerve and dilatation of the anterior subarachnoid space, and fat-suppressed T1 shows enhancement of the optic nerve sheath (Hickman et al 2005).
Enhancement with dye indicates optic nerve inflammation or demyelination and is not seen with ischemic lesions. Triple-dose Gd T1 MRI shows the medial lesion length is 30 mm (range, 0 to 39 mm), and there is no chiasmal enhancement (Hickman et al 2004). Lesion length correlates with defects in visual fields and with slowing of visual evoked potentials. The mean duration of enhancement is 63 days (range, 0 to 113). Optic nerve atrophy on MRI correlates with low visual acuity and color vision, retinal nerve fiber thinning, and reduced visual evoked potential amplitude, but not latency.
MRI studies of patients presenting with optic neuritis show multiple cerebral white matter lesions in 23% of 24 patients (Stadt et al 1990), 30% of 418 patients (Beck et al 1993a), 50% of 8 patients (Jacobs et al 1986), 55% of 42 patients (Feinstein et al 1992), 61% of 37 patients (Ormerod et al 1986), 62% of 50 patients (Frederiksen et al 1989), at least 66% of 38 patients (Paty et al 1988), and 75% of 28 patients (Hornabrook et al 1992). Thirty-one of 32 optic neuritis patients had abnormalities on MRI, compared to 15% of patients with anterior ischemic optic neuropathy (Rizzo et al 2002). N-acetyl aspartate levels are lower in involved areas. 3T MRI picks up 25% more associated brain lesions than 1.5 T MRI. Scattered brain lesions may be more common when there is a high epidemiologic incidence of multiple sclerosis. Remote from the optic nerve inflammation, the magnetization transfer ratio is reduced in the visual cortex, suggesting transsynaptic degeneration in the visual pathways (Audoin et al 2006). However, the major determinant in another study was the length of the optic nerve lesion and not any pathology in the optic radiations or occipital cortex (Jenkins et al 2010).
Optical coherence tomography (OCT) with near-infrared light shows significant reduction of retinal nerve fiber layer (RNFL) thickness and macular volume (possible disappearance of retinal ganglion cells) (Trip et al 2005). Severity of retinal nerve fiber layer loss is as follows: neuromyelitis optica (NMO) > optic neuritis > normal fellow eye in multiple sclerosis > normal fellow eye in ADEM > normal eyes. Damage is worse in men than women. RNFL degeneration and axonal loss is more likely after prolonged visual evoked potentials, impaired color vision, and poor low-contrast visual acuity (Henderson et al 2011).
Thinning of the RNFL layer correlates with other imaging measures of axonal loss and visual dysfunction. Loss of 1.5 lines of low-contrast visual acuity is equivalent to a 4.5 µm reduction in RNFL thickness. Axonal loss is reflected in reduced visual evoked potential amplitude, but not with conduction slowing. Macular volume loss, indicating death of neurons, correlates with poor color vision. Retinal nerve fiber layer thickness after optic neuritis does not predict the chance of developing multiple sclerosis. Zeiss Stratus OCT III is less sensitive (60%) than visual evoked potentials (81%) at detecting prior (>6 months) optic neuritis, so evoked potential remains the preferred test (Naismith et al 2009).
Pattern reversal visual evoked potentials (VEP) are often, or always, slowed in the affected eye. Thirty-five percent of 47 patients with optic neuritis returned to normal within 2 years (Heinrichs and McLean 1988), but latencies in the other patients were prolonged and did not improve, even when vision had returned to normal. Evoked amplitudes are normal or only mildly reduced in optic neuritis. Three-dimensional visual evoked potentials are moderately more sensitive than conventional visual evoked potentials for detecting post-chiasmal lesions (Towle et al 1991). Multifocal visual evoked potentials (mfVEP) can potentially isolate and follow small sectors of an affected optic nerve but take longer to administer than the conventional evoked potentials. They correlate well with optical coherence tomography measures of the retinal nerve fiber layer (r=0.8) and are possibly more sensitive (Klistorner et al 2009). Electroretinogram detects outer retinal lesions. It typically shows abnormal P50 (early, A) and N95 (late, B) waves in retinal disorders, but only an abnormal N95 component in optic nerve disorders. However, the P50 can be abnormal in active optic neuritis. It does not correlate with visual evoked potentials.
Critical flicker frequency is abnormal in 100% of affected eyes at onset, but returns to normal in over 90% of patients with recovery (Woung et al 1993). Elevated body temperature amplifies abnormalities in functional tests; normal eyes are much less sensitive to temperature increases.
The spinal fluid in optic neuritis sometimes contains elevated protein, mild lymphocytosis, 35% (Hutchinson 1976), 38% (Frederiksen et al 1992), or 48% (vs. 52% of multiple sclerosis) (Soderstrom 1995), free kappa-light chains (63%) (Rudick et al 1986), an elevated IgG index (20% to 36%) (Frederiksen et al 1992; Sellebjerg and Frederiksen 1993), and oligoclonal bands (56% to 69%) (Frederiksen et al 1992; Sellebjerg and Frederiksen 1993; Soderstrom 1995). The presence of oligoclonal bands correlates with MRI lesions, and the presence or absence of bands tends to remain constant over time (Soderstrom 1995). Myelin basic protein levels are usually normal. In Devic disease, bands are less common (27%) and tend to disappear.
In children, cerebrospinal fluid changes are variable. Only 2 or 3 of 30 children with idiopathic optic neuritis had pleocytosis (Kennedy 1960; Meadows 1969), or elevated protein or gamma globulin (Kennedy 1960). In contrast, another study showed frequent pleocytosis, excess IgG, oligoclonal bands, and antiviral antibodies in 21 children with optic neuritis, but 16 of the children had preceding bacterial or virus infections or vaccinations (Riikonen et al 1988).
The basic workup for optic neuritis should consist of funduscopy, visual acuity to document the degree of visual loss, plus a neurologic exam and MRI and lumbar puncture to rule out associated diseases, especially multiple sclerosis. Orbital MRI with fat suppression is important in atypical optic neuritis--patients older than 45 years of age, bilateral onset, no pain, vertical hemianopsia, swollen optic nerves, retinal exudates, progression over more than 2 weeks, and recent sinusitis (Chan 2002). An ophthalmologic exam detects associated ocular abnormalities in approximately 20% of patients. When there are relevant clues, sedimentation rate, antinuclear antibodies, angiotensin converting enzyme levels, and tests for Lyme disease and syphilis are needed but are of little value in typical cases (Beck et al 1992; 1997).
In the past, some authors recommended essentially no investigation for optic neuritis (Glaser 1990). However, the presence of multiple sclerosis should be documented because it can be ameliorated by therapy. Serum NMO-IgG should be evaluated in patients with a history of severe or recurrent optic neuritis, as therapy differs between multiple sclerosis and Devic disease. CNS Sjögren disease, most common in young black women, overlaps with Devic disease (Javed et al 2008). A combination of visual tests to assess severity and predict prognosis could be used as a “visual disability severity scale,” including low-contrast visual acuity, OCT retinal nerve fiber layer thickness, optic nerve diameter, multifocal visual evoked potentials, low-contrast multifocal visual evoked potentials, and diffusion tensor MRI (Pula 2009, personal communication).