Optic neuritis (ON) is a general term for an optic neuropathy resulting from an idiopathic, inflammatory, infectious, or demyelinating etiology. If the optic nerve is swollen on ophthalmoscopy, then the term papillitis or anterior ON is used. If the optic nerve is normal on ophthalmoscopy, then it is called retrobulbar ON. In clinical practice, most ophthalmologists use the term optic neuritis to describe idiopathic or demyelinating ON.
What Are the Features of Typical Optic Neuritis?
Patients with idiopathic or demyelinating ON usually present with a “typical” clinical profile as shown in Table 2–1 (Beck, 1992a, 1993a,c–e, 1994a; Cleary, 1993; Frederiksen, 1991; Gerling, 1998a,b; Jin, 1999; Keltner, 1993a, 1993b; Optic Neuritis Study Group, 1991; Schneck, 1997; Slamovits, 1991a; Wakakuru, 1999b; Wall, 1998).
The clinical characteristics of 455 patients with ON enrolled in the Optic Neuritis Treatment Trial (ONTT), a study sponsored by the National Eye Institute conducted at 15 clinical centers in the United States between the years 1988 and 1991, are outlined in Table 2–2.
The majority of patients with ON with eye or ophthalmic trigeminal distribution pain or pain with eye movement have involvement of the orbital segment of the optic nerve (Kupersmith, 2002). The absence of pain, particularly with eye movement, suggests the disorder is limited to the canalicular or intracranial portion of the optic nerve (Kupersmith, 2002).
What Visual Field Defects Are Noted with Optic Neuritis?
Analysis of initial perimetry in the ONTT showed that the most common presenting pattern was a diffuse field defect (48%), with altitudinal/arcuate defects in 20%, and central/cecocentral loss in only 8%. Classic teaching (in the Goldmann and tangent perimetry era) indicated that central scotoma was the most common pattern of visual field loss in ON, and the finding of only 8% in this category at first seems surprising. However, many early studies involved Goldmann perimetry in which a central scotoma represented depressed sensitivity within the central 30 degrees. The pattern of diffuse field loss in the ONTT actually may represent this same pattern. If one includes both the diffuse and the central/cecocentral categories of the ONTT, this study actually confirms that central visual field loss in the most common defect in ON (Arnold, 1999). Moreover, the study of Fang et al regarding global field loss involvement in ON suggests that within the central 30 degrees even cases with focal (central scotoma, arcuate) defects usually show an element of superimposed general depression (Fang, 1999a). The report of Keltner et al, which incorporated central and peripheral visual field testing (the latter by Goldmann perimetry) in the ONTT, supports this concept; although 97.1% of patients initially showed defects within the central 30 degrees, only 69.9% had abnormal peripheral fields (Keltner, 1999).
Acute, usually unilateral loss of vision
Visual acuity (variable visual loss 20/20 to no light perception (NLP)
Visual field (variable optic nerve visual field defects)
A relative afferent pupillary defect (RAPD) in unilateral or bilateral but asymmetric cases
Periocular pain (90%), especially with eye movement (Gerling, 1998a)
Normal (65%) or swollen (35%) optic nerve head
A young adult patient (<40 years) but ON may occur at any age
Eventual visual improvement
Improvement over several weeks in most patients (90%) to normal or near-normal visual acuity
88% improve at least one Snellen line by day 15
96% improve at least one line by day 30
Visual recovery may continue for months (up to 1 year)
Patients may complain of residual deficits in contrast sensitivity, color vision, stereopsis, light brightness, visual acuity, or visual field (Cleary, 1993, 1997; Frederiksen, 1997b; Steel, 1998)
Although it is not unusual for patients with ON to have central loss without peripheral loss, it is rare for the peripheral field to be abnormal in the presence of normal central 30 degree fields. Keltner et al showed that in most cases peripheral testing does not increase sensitivity, with only 2.9% of eyes in the study having abnormal peripheral fields with normal central fields (Keltner, 1999). When the results obtained through Humphrey automated central static visual fields and Goldmann peripheral kinetic isopters are compared, the far periphery appears to recover more rapidly than the central field, at least in more severe cases of ON (Keltner, 1999). Thus, in most cases recovery in ON can probably be monitored effectively with automated perimetry of the central visual fields alone. However, in cases of severe loss of central field, a peripheral kinetic visual field obtained with a Goldmann perimeter may provide additional information about the patient’s vision in the far periphery (Keltner, 1999). Gerling et al noted that peripheral testing may better define defects that are diffuse in the central 30 degrees but are actually altitudinal when the nasal periphery is tested (Gerling, 1998b).
Although it has long been postulated that ON tends to affect the papillomacular bundle with resultant central/cecocentral scotoma, the pattern loss in ON in the ONTT revealed pure papillomacular involvement in only 8%. Fang et al showed that ON affects the entire 30 degrees (global field involvement) even in patients who appear to have localized depression of visual threshold, indicating that ON does not have a true predilection for the papillomacular bundle, or any specific nerve fiber bundle (Fang, 1999a). In another study, Fang et al assessed specific nerve fiber group involvement by analyzing recovery of field within concentric field rings in the central 30 degrees and found that return of field function does not appear to differ between patients with diffuse or localized defects (Fang, 1999b). They postulate that reduced redundancy of axons in the periphery of the field compared with near fixation may be responsible for the greater recovery of threshold near fixation.
Age (years) (mean ± SD)
Mean days of visual symptoms before entry
Ocular pain present
Pain worsened by eye movement
Optic disc appearance
Optic disc swollen
Optic disc normal (retrobulbar)
Characteristics of swollen optic disc
Mild and focal
Mild and diffuse
Severe and focal
Severe and diffuse
Retinal or optic disc hemorrhage
On both disc and retina
More than trace cells
Present on or adjacent to disc
Present in the macula
20/20 or better
No light perception (NLP)
Visual field defects in involved eye
Altitudinal, arcuate, nasal step
Median visual field mean deviation (quartiles)
Visual function deficits in fellow eye
Visual field (mean deviation)
Abnormal MRI (one or more white matter lesion)
–23.02 (–31.90, –12.25)
Percents represent the percent of patients with the characteristic.
What Are the Features of Atypical Optic Neuritis?
Patients who meet the criteria listed in Table 2–1 are considered to have typical ON. Conversely, patients with the features listed in Table 2–3 have atypical ON (Beck, 1993a–e, 1994b; Biousse, 1999; Lee, 1998a; Moschos, 1990; Optic Neuritis Study Group, 1991). For example, the fundus features that should lead the examiner to consider an alternate diagnosis to ON include lipid maculopathy, very severe disc edema with marked hemorrhages, cotton wools spots, vitreous cells, pale optic disc edema, retinal arteriolar narrowing, and retinopathy.
What Disorders May Be Associated with Optic Neuritis?
Table 2–4 lists a number of disorders that may be associated with typical or atypical ON. The presence of one of these disorders is usually suggested by the historical or examination findings.
What Are the Clinical Features of Optic Neuritis in Children?
The clinical features of ON in children differ from those in adults. Table 2–5 summarizes these features. Brady et al reviewed 25 cases and concluded that pediatric ON is usually associated with visual recovery; however, a significant number of patients (22%) remain visually disabled. A normal magnetic resonance (MR) image of the brain may be associated with a better outcome. Younger patients are more likely to have bilateral disease and a better visual prognosis (Brady, 1999).
Bilateral simultaneous onset of ON in an adult patient
Lack of pain
Severe headache (e.g., sphenoid sinusitis)
Ocular findings suggestive of an inflammatory process
Posterior chamber inflammation more than trace
Macular exudate or star figure
Retinal infiltrate or retinal inflammation
Severe disc swelling
Lack of significant improvement of visual function or worsening of visual function after 30 days
Lack of at least one line of visual acuity improvement within the first 3 weeks after onset of symptoms
Age greater than 50 years
Preexisting diagnosis or evidence of other systemic condition
Inflammatory (e.g., sarcoidosis, Wegener’s granulomatosis, systemic lupus erythematosus)
Infectious disease (e.g., Lyme disease, tuberculosis, human immunodeficiency virus infection)
Severe hypertension, diabetes, or other systemic vasculopathy
Exquisitely steroid sensitive or steroid-dependent optic neuropathy
In another study of 47 children with multiple sclerosis, 38 (80.9%) had ON at least once, and 10 (21.3%) had two or more attacks of ON (Boiko, 2000). The presence of tumor necrosis factor α7 (TNF-α7) locus on chromosome 6 was proposed as a possible marker of early multiple sclerosis (MS) onset in these patients.
What Is the Evaluation of Optic Neuritis?
In atypical cases, consideration should be given to doing a lumbar puncture and additional laboratory studies; in the ONTT, syphilis serology, antinuclear antibody, and chest x-ray were performed. The required evaluation depends on the history and examination, with specific attention to infectious or inflammatory etiologies as listed in Table 2–4. In addition, patients with inflammatory autoimmune ON often have progressive or recurrent steroid-responsive or steroid-dependent optic neuropathy (Beck, 1994a; Bielory, 1993; Riedel, 1998).
The association of acute or subacute loss of vision in one or both eyes caused by optic neuropathy preceded or followed by a transverse or ascending myelopathy is referred to as neuromyelitis optica (Devic’s disease). The clinical features of Devic’s disease are outlined in Table 2–6.
Guillain-Barré syndrome (Nadkarni, 1993; Ropper, 1991)
Miller Fisher syndrome (Chan, 2002)
Chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) (Kaufman, 1998; Lee, 1999)
Syphilis (Frohman, 1997)
Tuberculosis (Mansour, 1998)
Lyme disease (Arnold, 1993; Jacobson, 1991; Karma, 1995; Lesser, 1990; Winterkorn, 1990)
Bartonella henselae (Cat-scratch disease) (Brazis, 1986; Schwartzman, 1994, 1995)
Mycoplasma (Nadkarni, 1993; Salzman, 1992; Sheth, 1993)
Brucellosis (Abd Elrazek, 1991; McLean, 1992)
Meningococcus (Miller, 1995)
Propionibacterium acnes (Kouyoumdjian, 2001)
Histoplasmosis (Perry, 1999; Yau, 1996)
Cryptococcus (Golnik, 1991)
Rickettsiae (e.g., Q fever, epidemic typhus)
Toxoplasmosis (Banta, 2002; Falcone, 1993; Grossniklaus, 1990; Pierce, 1993; Rose, 1991; Song, 2002)
Toxocariasis (Komiyama, 1995)
Cysticercosis (Chang, 2001; Menon, 2000)
Hepatitis A (McKibbin, 1995)
Hepatitis B (Achiron, 1994)
Cytomegalovirus (CMV) (Harkins, 1992; Ho, 1995; Mansour, 1997; Patel, 1996; Roarty, 1993)
Chickenpox (Lee, 1997)
Herpes zoster (Deane, 1995; Greven, 2001; Gunduz, 1994; Lee, 1997; Miyashita, 1993; Mori, 1997; Nakazawa, 1999)
Herpes simplex virus 1 (Tornerup, 2000)
Epstein-Barr (EB) virus (infectious mononucleosis) (Anderson, 1994; Beiran, 2000; Corssmit, 1997; Straussberg, 1993)
Measles (Totan, 1999)
Mumps (Sugita, 1991)
HTLV-1 (Lehky, 1996; Merle, 1997; Yoshida, 1998)
Prions (Jakob-Creutzfeldt disease)
HIV (AIDS)-related (Friedman, 1991; Nichols, 1992)
Primary HIV-related optic neuritis (Burton, 1998; Malessa, 1995; Newman, 1992; Quicenco, 1992; Sadun, 1995; Sweeney, 1993)
Syphilis (McLeish, 1990)
Cat-scratch disease (Bartonella henselae) (Schwartzman, 1994, 1995)
Cryptococcus (Golnik, 1991)
Histoplasmosis (Yau, 1996)
Cytomegalovirus (CMV) (Mansour, 1997; Patel, 1996; Roarty, 1993)
Herpes zoster (Friedlander, 1996; Lee, 1998b; Litoff, 1990; Margolis, 1998; Meenken, 1998; Shayegani, 1996)
Toxoplasmosis (Falcone, 1993)
Postvaccination (Albitar, 1997; Hull, 1997; Kerrison, 2001; Linssen, 1997; Stewart, 1999; Topaloglu 1992; van de Geijn, 1994; Yen, 1991)
Bacille Calmette-Guérin (BCG)
Anthrax (Kerrison, 2002)
Trivalent measles-mumps-rubella vaccine
Mantoux tuberculin skin test
Focal infection or inflammation (Bath, 1998; Farris, 1990; Fujimoto, 1999; Moorman, 1999)
Malignant otitis externa
Systemic inflammations and diseases
Behçet’s disease (Vaphiades, 1999)
Inflammatory bowel disease (Hutnik, 1996)
Sarcoidosis (Beck, 1994; Case Records Massachusetts General Hospital, 1996; DeBroft, 1993; Haupert, 1997; Kosmorsky, 1996)
Systemic lupus erythematosus (Ahmadieh, 1994; Galindo-Rodriguez, 1999; Giorgi, 1999a,b; Ninomiya, 1990; Ohnuma, 1996; Rosenbaum, 1997)
Mixed connective tissue disease
Rheumatoid arthritis (Agildere, 1999)
Acute posterior multifocal placoid pigment epitheliopathy (APMPPE) (Wolf, 1990)
Autoimmune optic neuropathy (Bielory, 1993; Riedel, 1998)
Familial Mediterranean fever (Lossos, 1993)
Bee or wasp sting (Berrios, 1994; Choi, 2000; Maltzman, 2000; Song, 1991)
Snake bite (Menon, 1997)
Postpartum optic neuritis (Leiba, 2000)
Retrobulbar optic neuritis with retinitis pigmentosa sine pigmento (Hatta, 2000)
Neuromyelitis optica (Devic’s disease) (Ahasan, 1994; Al-Deeb, 1993; Barkhoff, 1991b; Hainfellner, 1992; Hershewe, 1990; Igarishi, 1994; Jain, 1994; Jeffrey, 1996; Khan, 1990; Mandler, 1993, 1998; O’Riordan, 1996; Piccolo, 1990; Ramelli, 1992; Silber, 1990)
Recurrent optic neuromyelitis with endocrinopathies (Vernant, 1997)
More likely to be bilateral
More likely to have papillitis
May have worse presenting vision (later presentation?)
More likely to be associated with viral/parainfectious etiology
Less likely to be associated with multiple sclerosis
Age: Typically younger patients
Gender: Affects men and women equally
May be more common in African Americans who develop ON (Phillips, 1998)
May be more common in Asians who develop ON (Sakuma, 1999)
Recurrent optic neuromyelitis with endocrinopathies in eight Antillean women from Martinique and Guadeloupe (Vernant, 1997)
Familial cases: rare (Yamakawa, 2000)
Pathology: Differs from multiple sclerosis (MS)
Cerebellum is almost never affected
Excavation of affected tissue with formation of cavities common in Devic’s but rare in MS
Gliosis characteristic of MS absent or minimal with Devic’s
Arcuate fibers in cerebral subcortex relatively unaffected in Devic’s but severely damaged in MS
May have prodrome of fever, sore throat, and headache
May precede or follow paraplegia
Usually bilateral (hours, days, or rarely weeks between eyes)
Rapid and usually severe (complete blindness not uncommon)
Central scotoma most common visual field defect
Majority have mild disc swelling of both discs but may be normal
Occasional severe swelling with dilation of veins and extensive peripapillary exudates
May have slight narrowing of retinal vessels
Usually some recovery of vision
Often recovers within weeks to months
Some cases severe and permanent
Paraplegia (transverse myelitis)
Usually sudden and severe
Often recover to some degree but may be permanent complete paralysis
Spinal cord MRI often shows abnormality extending over three or more segments
May have Lhermitte’s symptom, paroxysmal tonic spasms, and radicular pain
Course: monophasic or relapsing
Rarely associated with demyelinating peripheral neuropathy
Rarely associated with HIV-1 infection, systemic lupus erythematosus, antiphospholipid antibody syndrome, and pulmonary tuberculosis
Often cerebrospinal fluid (CSF) pleocytosis (e.g., >50 WBC, often polymorphonuclear cells)
Oligoclonal bands uncommon
Rare increased intracranial pressure
Possible response to IV steroids
IV gamma globulin
Mortality less than 10 to 33%
Source: Ahasan, 1994; Al-Deeb, 1993; Barkhoff, 1991b; Blanche, 2000; Filippi, 1999; Hainfellner, 1992; Hershewe, 1990; Igarishi, 1994; Jain, 1994; Jeffrey, 1996; Khan, 1990; Mandler, 1993, 1998; O’Riordan, 1996; Phillips, 1998; Piccolo, 1990; Ramelli, 1992; Silber, 1990; Vernant, 1997; Wingerchuk, 1999; Yamakawa, 2000.
What Were the Results of the Optic Neuritis Treatment Trial (ONTT)?
The ONTT was developed to evaluate the efficacy of corticosteroid treatment for acute ON and to investigate the relationship between ON and MS (Beck, 1992a, 1993a–e, 1995a). The ONTT was sponsored by the National Eye Institute as a randomized, controlled clinical trial that enrolled 457 patients at 15 clinical centers in the United States between the years 1988 and 1991. The ONTT entry criteria specified that patients be between the ages of 18 and 46 years, that they have a relative afferent pupillary defect as well as a visual field defect in the affected eye, and that they were examined within 8 days of the onset of visual symptoms of a first attack of acute unilateral ON. Patients were excluded if they had previous episodes of ON in the affected eye, previous corticosteroid treatment for ON or MS, or systemic disease other than MS that might be a cause of the ON (Beck, 1992a, 1993a–e, 1995a). The clinical features of the ONTT patients are outlined in Table 2–2.
In the ONTT, all patients underwent testing for collagen vascular disease (antinuclear antibody [ANA]), serologic testing for syphilis (fluorescent treponemal antibody absorption [FTA-ABS]), and a chest radiograph for sarcoidosis. Lumbar puncture was optional. An ANA test was positive in a titer less than 1:320 in 13% of patients, and 1:320 or greater in 3%. Only one patient was eventually diagnosed with a collagen vascular disease.
Visual and neurologic outcomes in these patients were no different from those of the other ONTT patients. The FTA-ABS was positive in six patients (1.3%), but none had syphilis. A chest radiograph did not reveal sarcoidosis in any patient. However, in a separate report, Jacobson et al described 4 of 20 patients with isolated ON with a positive serology for Lyme disease (Jacobson, 1991). These authors recommended serologic testing for Lyme disease in patients with ON, with or without the typical rash of erythema migrans, who live in or have visited Lyme endemic areas. Cerebrospinal fluid (CSF) analysis was recommended for patients with positive serology and intravenous (IV) antibiotic therapy for unexplained pleocytosis (Jacobson, 1991). We do not order Lyme titers for patients with ON from nonendemic regions (class IV, level C).
The evaluation recommendations of the ONTT study group for patients with typical acute ON are listed in Table 2–7.
What Are the Neuroimaging Findings in Optic Neuritis?
Periventricular white matter signal abnormalities on magnetic resonance imaging (MRI) consistent with MS (Baumhefner, 1990; Jacobs, 1991) have been reported in 40 to 70% of cases of isolated ON (Christiansen, 1992; Feinstein, 1992; Francis, 1995; Frederiksen, 1991a; Jacobs, 1991; Morrissey, 1993). MRI with gadolinium may show enhancing lesions in 26 to 37% of patients with isolated ON (Christiansen, 1992; Merandi, 1991) and may increase the detection of disease activity (Guy, 1990; Merandi, 1991; Thompson, 1990).
Although computed tomography (CT) scan of the head may also show abnormalities in MS and ON, CT has been relatively insensitive to the detection of MS plaques compared to MRI. MRI is a very sensitive test for detecting lesions consistent with MS (Baumhefner, 1990). Paty reported 19 cases of clinically definite MS (CDMS) out of 200 consecutive patients with suspected MS comparing predictive value of MR scanning with CT scanning, evoked potentials (EPs), and CSF analysis for oligoclonal bands (Paty, 1988). Eighteen of these 19 (95%) patients had MR scans that were “strongly suggestive of MS” at first evaluation. Fourteen of 19 (74%) patients had positive oligoclonal bands. Ten of 19 (53%) patients had abnormal somatosensory EPs, 9 of 19 (47%) patients had abnormal visual EPs (VEPs), and 9 of 19 (47%) patients had abnormal CT scans. Combining multiple reports, the risk of developing MS within 1 to 4 years is about 30% (range 23–35%) in patients with isolated ON and an abnormal MR scan (Beck, 1993a; Frederiksen, 1992; Jacobs, 1997; Söderström, 1998). Morrisey et al reported 89 patients (44 with ON, 17 with brainstem involvement, and 28 with spinal cord involvement) with an acute clinical demyelinating attack (Morrisey, 1993). Of these 89 patients, 57 (64%) had one or more MR scan abnormalities and 32 had no MR scan abnormalities. Only one of the 32 patients with normal MR scans developed MS, versus development of MS in 37 of 57 patients (65%) with an abnormal MR scan. Of the three isolated clinical syndromes (optic nerve, brainstem, spinal cord), ON with an abnormal MR scan had the highest rate of progression to MS—82%. Jacobs et al reported 42 patients with isolated monosymptomatic optic neuritis (Jacobs, 1991). During 5.6 years of follow-up, 21 patients developed MS. Of these 21 patients, 16 (76%) had abnormal MR scans and 5 had normal MR scans (Jacobs, 1991).
No laboratory studies or lumbar puncture required for typical optic neuritis
Potential testing for atypical optic neuritis
Collagen vascular disease screen
Complete blood counts
Lyme serology in patients from endemic areas
MR imaging of the brain for all optic neuritis (class I–II, level B)
Consider MR of head and orbit with fat suppression views to examine optic nerve course, especially in atypical optic neuritis
Söderström et al performed a prospective study of 147 consecutive patients with acute monosymptomatic ON (Söderström, 1998). Of 116 patients examined with MR scans, 64 (55%) had three or more high signal lesions, 11 (9%) had one or two high signal lesions, and 41 (35%) had a normal MRI. Among 146 patients undergoing CSF studies, oligoclonal bands were demonstrated in 103 (71%) patients. During the 6-year study period, 53 patients (36%) developed CDMS. Three or more MS lesions on MR scan or CSF oligoclonal bands were strongly associated with MS. Jacobs et al found that 42 of 74 (57%) patients with isolated monosymptomatic ON had 1 to 20 brain lesions by MR scans (Jacobs, 1997). All of the brain lesions were clinically silent and had characteristics consistent with MS. During 5.6 years of follow-up, 21 patients (28%) developed CDMS. Sixteen of the 21 converting patients (76%) had abnormal MR scans; the other 5 (24%) had MR scans that were normal initially (when they had ON only) and normal in 4 of the 5 when repeated after they had developed clinical MS. Of the 53 patients who had not developed CDMS, 26 (49%) had abnormal MR scans and 27 (51%) had normal MR scans. The authors concluded that the findings of an abnormal MR scan at the time of ON was significantly related to the subsequent development of MS. The interpretation of the strength of that relationship must be tempered by the fact that some of the converting patients had normal MR scans and approximately half of the patients who did not develop clinical MS had abnormal MR scans. Thus, it should be emphasized that MS is a clinical diagnosis that cannot be made on the basis of MR scan abnormalities alone (Guy, 1994; Paty, 1993), and the absence of MR scan abnormalities does not protect against the future development of MS (Beck, 1993d; Jacobs, 1991).