The clinical prognosis of optic neuritis is surprisingly good. Of 457 patients, 91% to 95% had a visual acuity of at least 20/40 at 1 year after the ictus (Beck and Cleary 1993b), 92% at 10 and 15 years. In an earlier study, 68% recovered to 20/30 or better vision and 87% were above 20/180 (Hutchinson 1976). Another study found good recovery in 86% (Bradley and Whitty 1968). In 12 patients with no light perception, 8 recovered to 20/40 or better, but 11 patients had dyschromatopsia (Slamovits et al 1991). In other studies, 65% returned to normal within 1 year (Celesia et al 1990). Better recovery is associated with a short stretch of involved optic nerve on gadolinium-enhanced MRI, high visual evoked potential amplitudes, and an early, rapid rate of improvement in vision (Hickman et al 2004). Poor prognostic factors for visual recovery include initial low visual acuity, absence of pain, involvement of the intracanalicular optic nerve (Deschamps et al 2002), and decreased axial diffusivity on diffusion tensor imaging MRI. Those who do not recover should be investigated in more detail to rule out other diseases, including Devic disease.
The risk of developing multiple sclerosis is controversial. (The 1983 Poser criteria do not consider recurrence of optic neuritis, even in the fellow eye, as multiple sclerosis—unless it occurs more than 15 days from the initial episode.) Optic neuritis can be the first sign of multiple sclerosis.
In 39 studies before 1985, only one eighth to one third of patients with isolated idiopathic optic neuritis eventually developed clinical multiple sclerosis. Other studies had similar results; 12% of 110 Mexicans developed multiple sclerosis at 2 years (Corona-Vazquez et al 1997). Four percent of 23 Chileans monitored developed multiple sclerosis within 10 years after optic neuritis (Alvarez and Cardenas 1989). Of 88 Brazilians studied, 11% developed multiple sclerosis within 5 years, but the median time to multiple sclerosis was 1 year and lesions were frequently spinal (Lana-Peixoto and Lana-Peixoto 1991). In Pakistan, 15% developed multiple sclerosis at 5 years and 24% at 10 years (Kazim et al 2010). The low numbers in these 4 studies suggest a regional variation in prognosis. In 40 Italians older than 12 years of age, 25% developed multiple sclerosis (Mapelli et al 1991), and 14% of 50 Swedish optic neuritis patients developed multiple sclerosis at 1 year (Frederiksen et al 1989). In the optic neuritis treatment trial, 17% of 126 patients in the United States developed multiple sclerosis over 2 years (in the placebo group) (Beck et al 1993c). At 5 years, 30% of the patients had developed multiple sclerosis (Beck 1997); at 10 years, 38% of 388 patients (some had been treated with steroids) had multiple sclerosis (some, maybe 5%, had prior neuritis in fellow eye), and at 15 years, 50% had multiple sclerosis independent of MRI lesions at baseline (Optic Neuritis Study Group 2008).
Others find a higher rate of evolution to multiple sclerosis. Thirty-two percent of 82 Australians studied developed multiple sclerosis over 5 years (Hely et al 1986) and 40% at 15 years. Thirty-nine percent of Minnesotans developed the disease at 10 years, but 60% of those developed multiple sclerosis by 40 years of age. Perivenous sheathing and recurrent optic neuritis increased the risk (Rodriguez et al 1995). In a related series of Minnesotans, 30% developed multiple sclerosis after 10 years, 13% developed neuromyelitis optica, and these had more severe and more rapid onset. Early recurrence of optic neuritis increased the risk of multiple sclerosis (Pirko et al 2004). There was a nonsignificant trend for intravenous steroids to prevent conversion to multiple sclerosis (Pirko et al 2004). Other studies showed a higher incidence of progression to multiple sclerosis; 26% of 86 Swedes at 5 years, 34% at 10 years, and 38% over 13 years (Nilsson et al 2005), 42% of Italians at 10 years and 52% at 15 years, 42% of 36 Norwegian patients within 2 years (Gronning et al 1989), 48% of Londoners at 5 years, with more disability with enhancing or infratentorial MRI lesions (Swanton et al 2009), 57% of 30 Danes within 6 years (Anmarkrud and Slettnes 1989), 58% of 60 New Englanders over 15 years (Rizzo and Lessell 1988), and 60% of 48 Dutch over 2 years (Sanders and van Lith 1989). The risk was 57% in 101 British patients followed for 11.6 years, although 45 of the original 146 were not included in the follow-up.
An isolated episode of optic neuritis without other signs of multiple sclerosis should not be treated with disease-modifying therapy. However, if there are associated MRI lesions and CSF oligoclonal bands, the risk of multiple sclerosis is quite high, and this therapy is usually appropriate.
In children compared to adults, optic neuritis more often follows a virus infection, is more often bilateral (42% vs. 6%) (de la Cruz and Kupersmith 2006; Wilejto et al 2006), is associated with papillitis, and traditionally has a poor visual prognosis. However, a study of 36 children with optic neuritis showed full recovery in 83% of eyes, even though one half had brain MRI lesions at presentation and one third developed multiple sclerosis over a 2.4-year follow-up (Wilejto et al 2006). Optic neuritis, transverse myelitis, or cerebellitis are commonly present at the onset of multiple sclerosis in children (Boutin et al 1988).
Only 8% of 39 children developed a recurrence of optic neuritis and 15% developed multiple sclerosis after 9 years (Kriss et al 1988). In 94 children at the Mayo Clinic with optic neuritis, only 13% had progressed to multiple sclerosis at 10 years, and only 26% by 40 years (Lucchinetti et al 1997). In Philadelphia, Pennsylvania, 18 children with optic neuritis were followed for more than 2 years, and 3 of the 18 developed multiple sclerosis (Bonhomme et al 2009). All 3 children had brain positive MRI at onset, but 11 children with negative MRI did not develop multiple sclerosis. In Brazil, 10 of 27 children with optic neuritis had bilateral problems, but only 1 out of 27 developed multiple sclerosis over 13 months (Lana-Peixoto and Andrade 2001). In Turkey, 8 of 31 children developed multiple sclerosis after 2.2 years and were more likely to be older girls with unilateral optic neuritis (Cakmakli et al 2009). One study found that optic neuritis is likely to evolve into multiple sclerosis in children when it is associated with cerebrospinal fluid inflammation and bacterial or virus infections or vaccinations (Riikonen et al 1988). However, a larger study found that optic neuritis preceded by infections was less likely to lead to multiple sclerosis (Lucchinetti et al 1997).
Predictions of the risk of developing multiple sclerosis should not be based on linear models. Onset of multiple sclerosis is most frequent in the first 1 to 5 years after optic neuritis than at later times (Bradley and Whitty 1968; Ebers 1985; Riikonen et al 1988; Cendrowski 1991). Multiple sclerosis is more likely to develop after optic neuritis in individuals aged 21 to 44 years (Hutchinson 1976; Rizzo and Lessell 1988; Mapelli et al 1991), in women (Rizzo and Lessell 1988), and in patients with bilateral optic neuritis as adults (Hutchinson 1976) (though some reports disagree), rather than as children (Parkin et al 1984). Uhthoff symptom (Scholl et al 1991), retinal venous sheathing (Lightman et al 1987), pain, no optic disc swelling, severe visual acuity loss, prior optic neuritis in the fellow eye, and prior nonspecific neurologic symptoms are also predictors of multiple sclerosis (Beck 1997), as is onset of optic neuritis in January through March. Postvaccinal optic neuritis (Riikonen et al 1988), a DR3 or DR2 background (Francis et al 1987; Riikonen et al 1988), serum antibodies to proteasomes (Beyer et al 2007), potentially pathogenic (non-Leber) mitochondrial DNA variants (Bosley et al 2007), cerebrospinal fluid cells or oligoclonal bands (Cendrowski 1991; Ghezzi et al 1996; Nilsson et al 2005), cerebrospinal fluid IgG (Jacobs et al 1997), disseminated lesions on MRI (Miller et al 1988b; Sandberg-Wollheim et al 1990; Beck 1997), or abnormal evoked potentials also suggest that multiple sclerosis will develop (Cendrowski 1991). Unless the lesions are concomitant, recurrent attacks of optic neuritis (either eye) do not increase the chance of developing multiple sclerosis (Hutchinson 1976; Francis et al 1987; Rizzo and Lessell 1988), although 1 report disagrees (Beck 1997). Visual function is less likely to recover with successive episodes. A history of optic neuritis in a multiple sclerosis patient actually improves the long-term prognosis compared to patients without optic neuritis (Weinshenker et al 1991).
Optic neuritis recurs in either eye in 35% of patients at 10 years (Beck et al 1993c). Recurrent optic neuritis is more common in patients who develop multiple sclerosis (48%) than in those who don’t (24%) (Beck and Group 2004). Bilateral simultaneous optic neuritis led to multiple sclerosis in only 1 of 11 adults after an interval of up to 30 years. Sequential optic neuritis, however, led to multiple sclerosis in 8 of 20 (Parkin et al 1984). In children, 1 of 17 with recurrent optic neuritis developed multiple sclerosis. Antibodies to myelin oligodendrocyte glycoprotein are at high titers in children with recurrent optic neuritis (Rostasy et al 2012).
Recurrent optic neuritis is typical in Devic disease. An antibody to aquaporin 4 (NMO-IgG) has high sensitivity (73%) and specificity (91%) for NMO (Lennon et al 2004). NMO-IgG is positive in many cases of the Asian type of multiple sclerosis (optic nerve and spinal cord predominance), but not in classic Western-type multiple sclerosis. This marker will cause re-evaluation of multiple sclerosis as a disorder with multiple subtypes—with different courses (more devastating in NMO), disease associations (autoimmune with NMO), and therapy (IVIG and plasmapheresis in NMO).
Disseminated MRI lesions at the onset of optic neuritis increase the risk of developing multiple sclerosis, but the lesion burden does not correlate perfectly. The MRI at baseline (disseminated lesions or not) is more often normal in optic neuritis than in other clinically isolated syndromes. The MRI may be the most important predictor of developing multiple sclerosis (Tintore et al 2005). With isolated idiopathic optic neuritis, 34% (Miller et al 1988b), 50% (Jacobs et al 1986), 58% (Beck 1997), 64% (Morrissey et al 1993), to 70% (Dalton et al 2003) of patients have disseminated lesions on MRI that are similar to the lesions seen in multiple sclerosis. Single non-enhanced MRIs are unable to definitively show dissemination over time, but lesions of different intensities do suggest temporal dispersion. One year after optic neuritis, 12 of 34 (35%) patients with disseminated lesions at onset developed clinical signs of multiple sclerosis (Miller et al 1988b). Of 19 patients without MRI lesions, none developed clinical multiple sclerosis, although 3 had new MRI lesions. In a second study with a 5.6-year follow-up of 74 patients with optic neuritis, 22% developed clinical multiple sclerosis (Jacobs et al 1997); 76% of these had abnormal MRIs initially. Of note, half of the patients who did not develop clinical multiple sclerosis also had abnormal MRIs at onset. In a third study, optic neuritis with no MRI lesions led to multiple sclerosis at 5 years in 16%, versus 51% of patients with 3 or more MRI lesions (Beck 1997). The 10-year data for this study show multiple sclerosis in 22% without MRI lesions, and in 56% in those with 1 or more MRI lesions. The amount of disability is unrelated to baseline lesion load. In a fourth study, optic neuritis with no MRI lesions led to multiple sclerosis at 5.5 years in 6% of 44 optic neuritis patients, versus 82% of patients with an abnormal MRI (Morrissey et al 1993). With no baseline MRI lesions, a patient has a 25% chance of developing multiple sclerosis in 15 years (Pula and Reder 2009). With 1 lesion, the cumulative probability is 60%. With 3 or more lesions, the likelihood is 78%. Optic nerve enhancement itself does not count as a lesion. In a fifth study of 102 Italian patients, optic neuritis with no MRI lesions did not lead to multiple sclerosis at 8 years, whereas 57% of patients with an abnormal MRI did develop the disease (Ghezzi et al 2000). Despite the correlation between MRI lesions and development of multiple sclerosis, 19 to 31 years after the first episode of optic neuritis, 20 out of 30 patients had brain MRI lesions, yet had no clinical manifestations (Nilsson et al 2005). Spinal cord MRI helps determine dissemination in space in only 1% to 4% of optic neuritis patients who have normal brain MRIs (Dalton et al 2003).
Widespread MRI changes are more common when there is a history of optic neuritis in the fellow eye. However, when the fellow eye is involved at onset, patients are not more likely to have MRI or clinical evidence of multiple sclerosis (Beck et al 1993a). Long lesions in the posterior intracanalicular segment of the nerve cause slow and incomplete recovery (Gass and Moseley 2000; Hickman et al 2004).
Multiple sclerosis is more likely to develop over a year in patients with optic neuritis and delayed visual evoked potentials (36%) than in patients with normal evoked potentials (0%) (Fraser et al 2006).
Oligoclonal bands add to prediction of future course, and CSF should be obtained when there is clinical uncertainty about prognosis. Oligoclonal bands as a predictor of multiple sclerosis after monosymptomatic optic neuritis have 90% sensitivity, 60% specificity, and an odds ratio of 34, but there is heterogeneity between 10 studies (Skov et al 2011). The combination of central nervous system MRI lesions plus oligoclonal bands augers a high risk for multiple sclerosis; 94% at 4 years (Tumani et al 1998). Patients with negative MRI yet positive bands are 10 times more likely to develop multiple sclerosis than those with negative MRI and no bands (Cole et al 1998). A profile of negative MRI (specific) plus no oligoclonal bands (sensitive) is a strong predictor of not developing multiple sclerosis (0% at 2 years, 17% at 4 years) (Tumani et al 1998). In another series, the risk of developing multiple sclerosis was 66% at 4 years with 3 or more brain MRI lesions plus oligoclonal bands, but only 9% with neither (Jin et al 2003).
Development of multiple sclerosis is more likely with prior optic neuritis, other neurologic symptoms, white race, and a family history of multiple sclerosis. Patients with optic neuritis and HLA-DR3, especially in combination with DR2, are more likely to develop multiple sclerosis (Francis et al 1987). A multiple sclerosis-associated retrovirus has been associated with development of multiple sclerosis after an attack of optic neuritis. Good prognostic factors are male gender, simultaneous bilateral onset, and atypical features for optic neuritis that suggest another etiology for visual loss, such as poor vision, severe disc swelling, hemorrhages, and lack of pain.