Aim To describe clinical features and outcome of a series of children with first-episode optic neuritis investigated in three paediatric neurology centres.
Methods Databases were searched to identify children (<16 years) with optic neuritis and life table analysis was used.
Results 44 children (female/male ratio 1.8) median age 10.9 years were followed up for median 1 year. Optic neuritis was unilateral in 43%. Maximal visual deficit was severe (<6/60) in 77%, with full recovery in 70%. Cumulative probability of developing MS (11/44) or NMO (3/44) at 2 years was 0.45. Relapsing optic neuritis was a strong predictor for development of MS or NMO. A positive MRI (>1 brain T2 hyperintense lesion) was a strong predictor for development of MS.
Discussion Childhood optic neuritis is associated with severe visual deficit with good recovery. An initial abnormal MRI brain scan or relapsing optic neuritis should alert the clinician to MS or NMO diagnosis.
Statistics from Altmetric.com
Childhood optic neuritis due to demyelination is a potentially treatable condition that may lead to visual impairment. A recent Canadian surveillance study showed an incidence of 0.2 per 100 000 children.1 However, incidence rates are likely to be geographically different and more studies are needed. Isolated optic neuritis may also be the first manifestation of multiple sclerosis or neuromyelitis optica (NMO). The natural history of childhood optic neuritis remains relatively unreported. Case series reported in the literature have tended to be single-centre studies, not including NMO as an outcome, and some have included cases of polysymptomatic clinically isolated syndromes (CIS) at presentation.2,–,7 We reviewed isolated first-episode childhood optic neuritis cases in three tertiary UK centres to evaluate clinical features and outcome. A secondary aim was to inform the design of a prospective national epidemiological surveillance study.
Computerised databases were searched for consecutive cases to indentify optic neuritis cases between 1 January 2002 and 31 March 2008. The age at presentation was limited from 1 month to less than 16 years of age. A text-based departmental database of patients seen in the paediatric neurology and ophthalmology departments was searched using the terms “optic neuritis”. Searches were conducted in three UK Children's Hospitals. Patients' notes were retrospectively analysed. We included cases with acute or subacute visual loss and one or more of the following: relative afferent pupillary defect in the affected eye in unilateral cases, visual field deficit or scotoma, impaired colour vision, optic disc oedema or abnormal visual evoked potentials. We excluded cases with other neurological findings, concurrent acute disseminated encephalomyelitis or other causative aetiology or previous central nervous system (CNS) inflammatory demyelinating episodes. Conditions excluded were anterior ischemic optic neuropathy, neuroretinitis, compressive or infiltrative optic neuropathies and Leber's hereditary optic neuropathy.
What is already known on this topic
▶ Optic neuritis may be the first presentation of multiple sclerosis or neuromyelitis optica (NMO).
▶ The natural history is underreported.
What this study adds
▶ Childhood isolated optic neuritis is associated with severe visual deficit with generally good recovery.
▶ Our cohort confirms that an initial abnormal MRI brain scan or relapsing optic neuritis should alert the clinician to multiple sclerosis or NMO diagnosis.
Any pre-existing ophthalmic conditions that had affected visual acuity were noted. At least two researchers in each centre carried out the search and decided on inclusion, and in case of disagreement, the first author was consulted. The primary outcome was visual acuity at recovery, and the secondary outcome was multiple sclerosis or NMO diagnosis on follow-up. A “positive MRI” was defined as showing one or more brain T2 hyperintense lesions. International paediatric multiple sclerosis study group consensus definitions8 were used for defining multiple sclerosis and other CNS demyelinating conditions.
An excel spreadsheet was designed to collect and record anonymised data in the three centres. Descriptive data are presented. Statistical analysis utilised life-table analysis by the Kaplan–Meier method performed using SPSS (version 17.0) software.
Searches indentified a total of 44 consecutive cases that were eligible for inclusion. Ages ranged from 2.0 to 15.8 years with a median of 10.9 years (IQR 9.0–13 years). The female/male ratio was 1.8 (p=0.07). Optic neuritis was unilateral in 43% and bilateral in 57%. Maximal visual deficit at presentation was severe (a snellen equivalent of <6/60 (LogMAR 1.00) in 77% of cases.
Cases were followed up for 2 months to up to 6 years (mean 1 year, 10 months, median 1 year). Visual recovery (to pre-existing visual acuity if known, or a snellen equivalent of 6/9 or better (LogMAR 0.18) occurred in 70% (41/59 eyes). Data on visual recovery was not available for 10 eyes. On follow-up, 32% were diagnosed with multiple sclerosis (11/44) or NMO (3/44). Cumulative probability of cases developing multiple sclerosis or NMO (see figure 1) by 2 years after optic neuritis onset was 0.45 (95% CI 0.25 to 0.65). Gender, age (younger or older than 10 years of age), unilateral/bilateral optic neuritis and visual acuity severity did not predict multiple sclerosis development. Relapsing optic neuritis (see figure 2) was a strong predictor for the development of multiple sclerosis or NMO (logrank p<0.001).
At presentation out of the 44 isolated optic neuritis cases, 38 (86%) had MRI brain imaging, of which 34% had white matter lesions other than in the optic nerve. A positive MRI for multiple sclerosis (see figure 3) was a strong predictor for development of multiple sclerosis (logrank p<0.001). One patient with an initial normal MRI developed multiple sclerosis.
A total of 17 patients had cerebrospinal fluid (CSF) oligoclonal band testing (table 1). CSF oligoclonal band (OGB) had poor sensitivity (50%) but good specificity (89%) and did not reliably predict progression to multiple sclerosis (Fisher exact test p=0.1, negative likelihood ratio=0.56). Two patients who later developed NMO had negative CSF OGB. All three patients later diagnosed with NMO had NMO antibody testing after they had relapsed with transverse myelitis (after 0.3, 0.4 and 2 years) and diagnostic criteria were met.
Out of the 44 cases, 32 (73%) were documented to have received treatment with steroids. Administration and dosage varied widely, and the favoured regime was 3 days of intravenous methylprdenisolone (30 mg/kg, maximum 1 g) with various lengths (0−6 weeks) and doses of a prednisolone weaning regime. Children who had clinical CNS demyelinating relapses, and later diagnosed with multiple sclerosis by a paediatric neurologist, were treated with disease-modifying treatment (Rebif, Betainterferon or Copaxone).
Case series reported in the literature have tended to be single-centre or smaller studies, not including NMO as an outcome and some have included cases of polysymptomatic CIS at presentation.2,–,7 Our multi-centre study design utilised two sources for maximal case ascertainment, although capture−recapture analysis was not suitable due to problems with dependence and heterogeneity. Clear case definitions for inclusion and strict exclusion criteria were used at the outset. The retrospective nature, as well as ascertainment being from tertiary centres, raises the possibility of bias of case severity. The variable length of time to events, and follow up of the study, was adjusted for by applying life table statistical analysis methods.
Acute optic neuritis in children differs from the typical adult form.4 6 9 Bilateral disease is thought to be more common in children, as is severe loss of visual acuity (6 or worse in 84% of eyes in one series, with recovery of visual acuity to 6/9 or better in 76% of patients).4 In another recent case series involving 36 children, optic neuritis was unilateral in 58% and bilateral in 42%. Maximal visual deficit was severe in 69%, but full recovery occurred in 39 of 47 affected eyes (83%).7 In our series, optic neuritis was associated with severe visual deficit (<6/60 in 77%) with generally good visual recovery (70% in 41/59 eyes). Bilateral optic neuritis (57%) was not significantly more frequent than unilateral cases.
Age at presentation in our cohort (median 10.9 years) and a female predominance (ratio 1:1.8) was comparable to other studies.2,–,7 Neither variable (age greater than 10 years and female predominance) reached significance levels.
As in adults, the condition is predictive of the subsequent risk of multiple sclerosis, although estimates of risk vary among studies. Optic neuritis may also represent the first attack of Devic's NMO, an inflammatory, demyelinating condition in which clinical disease is referable to the optic nerves and spinal cord, without involvement of the remaining CNS white matter.10 11 Devic's NMO is associated with a high mortality (30%) in adults,10 but it is possible that the same process may have milder manifestations in children.12 Recently, a serum antibody that targets the aquaporin 4 molecule has been identified in Devic's NMO.13 In one longitudinal study, multiple sclerosis was diagnosed in 13% of children within 10 years after the first episode of optic neuritis, and in 19% within 20 years.6 In a Canadian cohort, 36% were diagnosed with multiple sclerosis at 2 years. In our study, we used lifetable analysis due to varying follow-up lengths and differing time points to relapse and multiple sclerosis or NMO diagnosis. The risk of multiple sclerosis or NMO development after isolated optic neuritis was high, with a cumulative probability of 0.45 at 2 years.
In our series, a positive MRI at optic neuritis onset was a strong predictor for development of multiple sclerosis. Only one patient with a normal MRI developed multiple sclerosis. This data is comparable with that of the other cohort studies. In terms of other predictors, relapsing optic neuritis was also a strong predictor for multiple sclerosis or NMO development. Two case series have implicated bilateral,6 7 as opposed to unilateral, optic neuritis5 as a predictor for development of multiple sclerosis, while one smaller series indicated that female sex and an older age5 indicated a higher multiple sclerosis risk after optic neuritis onset. Our study showed that gender, age (younger or older than 10 years of age), unilateral/bilateral optic neuritis and visual acuity severity did not appear to predict multiple sclerosis or NMO development.
Effects of corticosteroid treatment and other therapies on the recovery of visual function and on the risk of multiple sclerosis in children have not been established by randomised trials, but on the basis of data in adults, treatment with intravenous methylprednisolone is generally recommended if visual loss is unilateral and severe or is bilateral.9 Interferon therapy is considered in children with abnormal MRI scans of the brain, but data regarding the efficacy of therapy to prevent multiple sclerosis are lacking in this population. In our cohort, 77% of cases received corticosteroid treatment with the favoured regime being 3 days of intravenous methylprednisolone with various lengths of a prednisolone weaning doses.
Our data suggests that childhood optic neuritis is associated with severe visual deficit with good recovery. The risk of development of multiple sclerosis or NMO was high. An initial abnormal MRI brain scan or relapsing optic neuritis should alert the clinician to multiple sclerosis or NMO diagnosis. A UK- and Ireland-population-based prospective epidemiological surveillance study with multi-source case ascertainment is now underway to evaluate incidence rates and short-term outcomes (at 2 years). Evaluating potential risk factors for early CNS demyelinating relapse and multiple sclerosis or NMO diagnosis may guide treatment guidelines and inform future longitudinal therapeutic trials.
The authors thank the Ophthalmology Departments at the Birmingham Children's Hospital for help with retrieving records.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.