Outcomes of Children With Optic Neuritis as Their First Demyelinating Event

Abstract

Introduction

Optic neuritis (ON) is an acquired demyelinating syndrome and the most common cause of acute optic nerve inflammation in children and adults. This study describes the clinical and diagnostic features, visual outcomes, and relapse risk of children presenting with their first ON episode.

Methods

We prospectively identified 50 children presenting with ON as their first demyelinating event. Patients underwent both serum myelin oligodendrocyte glycoprotein (MOG) and aquaporin-4 (AQP4) antibody screening (n = 42) or were diagnosed with AQP4-seropositive neuromyelitis optica spectrum disorder (NMOSD) without MOG testing (n = 8). We analyzed demographics, antibody status, magnetic resonance imaging (MRI) findings, treatments, relapses, and visual disability.

Results

Subjects were stratified by diagnosis into idiopathic ON (n = 6), MOG antibody disease (n = 20), multiple sclerosis (n = 11), and NMOSD (n = 13). Females comprised 66% of the cohort. The mean age at onset was 12 years, and Black patients represented 48% of the cohort. Decreased visual acuity was nearly universal (96%). The logMAR of the worst eye at onset was most severe in NMOSD (3.1) and mildest in idiopathic ON (1.3). Ninety percent received intravenous steroids as acute treatment. Visual recovery varied by diagnosis, with mean improvement of 1.5 logMAR. Half (n = 25) experienced relapses, most commonly ON (19 of 25) or longitudinally extensive transverse myelitis (9 of 25). NMOSD patients had the highest relapse rates and poorest visual outcomes.

Conclusion

Pediatric ON often leads to chronic demyelinating conditions. Visual recovery is overall good, but patients with NMOSD have worse visual outcomes and higher relapse rates.

Keywords

adolescents children demyelinating disease optic neuritis pediatric

Optic neuritis (ON) is an acquired demyelinating syndrome and the most common cause of acute optic nerve inflammation in both children and adults.¹ Although ON is overall less common in children than adults,² it represents up to a third of all demyelinating events in children.^3,4 The underlying disorder causing ON often determines the associated clinical features, recurrence risk, and long-term outcomes, emphasizing that proper diagnosis and treatment is vital to optimize outcomes in these patients. Near or complete visual recovery by 6 months from onset is another distinctive feature of pediatric ON.^5–8 However, observational and prospective studies in children with ON^5,7–9 have shown that 50% to 60% will go on to be diagnosed with an underlying demyelinating disease. Of these, multiple sclerosis (MS) is the most common diagnosis (14%-39%), although with the recent commercial availability of myelin oligodendrocyte glycoprotein (MOG) antibody testing, MOG antibody disease (MOGAD) has become comparably prevalent (29%-36%), followed by aquaporin-4 (AQP4)-seropositive neuromyelitis optica spectrum disorder (NMOSD) (11%-33%) and MOG-seronegative acute disseminated encephalomyelitis (ADEM) (7%-13%)^5,8-11. It is now well known that MOGAD can present with a variety of phenotypes. Phenotypes involving ON include isolated ON (bilateral, unilateral, or sequential), ON with ADEM (either simultaneous or sequential), ON with longitudinally extensive transverse myelitis (LETM), and ON with encephalitis.

Our understanding of pediatric ON has shifted significantly over the past 20 years with changes in diagnostic criteria. It has progressed from an assumption that most children will have a single idiopathic event that does not require long-term treatment to a more nuanced approach to long-term management, particularly in MOGAD patients. However, there are still significant gaps in our understanding of long-term prognosis in children. This study aims to describe the clinical features, visual outcomes, and relapse risk of all children and adolescents presenting with a first-time episode of optic neuritis.

Methods

This cohort of children and adolescents presenting with ON as their first demyelinating event was extracted from our Pediatric Onset Neuroimflammatory data set of patients with symptom onset before 18 years of age evaluated at a single tertiary center (Children's of Alabama) (n = 758). Subjects were enrolled retrospectively from 1991 to 1999 and prospectively since January 1, 2000. The neuroinflammatory data set was queried through December 31, 2024, for patients with ON as their first demyelinating attack (n = 150). The study cohort was subsequently limited to patients with both MOG and AQP4 serum antibody screening (n = 42) plus 8 AQP4-seropositive NMOSD patients tested only for AQP4 before MOG antibody testing became commercially available in 2018. These 8 AQP4-seropositive NMOSD patients were included in the study cohort because true seropositivity for both MOG and AQ4 antibodies is essentially nonexistent.¹² All AQP4 and MOG serum antibody testing was performed at Mayo Clinic, Rochester, MN using live cell-based assay with flow cytometry. Patients were classified by initial and current demyelinating phenotypes (idiopathic ON [ION], MOGAD, MS, and NMOSD) as defined by 2013 International Pediatric MS Study Group criteria, 2015 NMOSD criteria (defining both seropositive and seronegative NMOSD critiera), and 2023 MOGAD diagnostic criteria^9–15 (Figure 1). Diagnostic groups were analyzed with respect to demographics, serum antibody status, magnetic resonance imaging (MRI) findings, acute and long-term treatments, subsequent demyelinating events, and visual outcome comparing acuity at onset, at recovery within 1 year, and on last follow-up. In an attempt to gauge functional disability, only the most affected eye (“worst eye”) was evaluated in cases of bilateral optic neuritis, rather than all involved eyes. Visual acuity was obtained using Snellen chart, then converted to logarithm of the minimum angle of resolution (logMAR)¹⁶ (eg, Snellen 20/20 = logMAR 0, 20/200 = logMAR 1.0, hand motion = logMAR 2.4, light perception = logMAR 2.9, and no light perception = logMAR 4).

Statistical Analysis

The study cohort is characterized using descriptive statistics (mean ± SD, median, range). Time to first relapse among patients who had received a disease-modifying therapy (DMT) used time-to-event analysis methods to evaluate relapse-free survival following DMT initiation. Specifically, the Kaplan-Meier estimator was used to generate survival curves for each DMT efficacy group, representing the probability of remaining relapse-free over time. To statistically assess differences between the survival distributions of the 2 groups, the log-rank test was employed. The null hypothesis stated that there was no difference in survival functions between groups, whereas the alternative hypothesis proposed that at least 1 pair of survival functions differed significantly.

Permissions: The Pediatric Onset Neuroinflammatory Database has been continuously approved by the UAB Institutional Review Board since 2000, with waiver of consent granted for enrollment.

Results

Demographics

A total of 50 pediatric patients with ON were included, stratified into idiopathic ON (ION, n = 6), MOGAD (n = 20), MS (n = 11), and NMOSD (n = 13) (Table 1). Our cohort represents patients seen from 1991 to 2024 with a median year of onset in 2018 or 2019, except our NMOSD group, which was seen much earlier (median in 2012).

Table 1.

Demographic Data.

	Diagnosis
Characteristic	All(n = 50)	ION(n = 6)	MOGAD(n = 20)	MS(n = 11)	NMOSD(n = 13)
% of total	-	12	40	22	26
Sex, female, %	66	50	55	81.8	76.9
Age at first event, y (mean ± SD)	11.9 ± 4.0	14.4 ± 2.2	10.4 ± 3.5	14.7 ± 2.7	10.8 ± 4.7
Race/ethnicity, %
Black	48	16.7	30	72.7	69.2
White	40	83.3	55	9.09	23.1
Latino	8	0	5	18.2	7.7
Asian	4	0	10
Length of follow-up, y (mean ± SD)	5.3 (5.6)	2.7 (2.2)	3.7 (2.9)	3.3 (2.0)	10.7 (8.2)
Year of first ON event, median (range)	2018 (1991-2024)	2019 (2015-2020)	2018 (2011-2024)	2018 (2016-2021)	2012 (1991-2018)

Abbreviations: ION, idiopathic optic neuritis; MOGAD, myelin oligodendrocyte glycoprotein antibody disease; MS, multiple sclerosis; NMOSD, neuromyelitis optica spectrum disorder; ON, optic neuritis.

Females comprised 66% of the total cohort, with highest proportions in the MS (82%) and NMOSD (77%) groups. The mean age at first event across subgroups was around 12, with disease-specific means ranging from 10 years in MOGAD to 14 years in MS. Black patients were disproportionately represented in the cohort (48%), particularly in the MS (73%) and NMOSD (69%) subgroups, whereas ION and MOGAD cohorts were predominantly White (83% and 55%, respectively).

Serologic and Imaging Findings

Serum MOG antibodies were positive in 22 of 42 (52%) tested patients (by definition all MOGAD patients [20/20]), whereas serum AQP4 antibodies were exclusively detected in the NMOSD group (10/13) (Table 2). Three patients in the NMOSD group were seronegative for AQP4 antibodies (and MOG seronegative) but fulfilled 2015 NMOSD criteria.¹⁴ All 3 had long segment bilateral ON, but each had another coinciding core clinical characteristic: 1 with LETM, 1 with LETM and diencephalic syndrome, and 1 with ADEM and acute brainstem syndrome. Of the 10 AQP4 seropositive NMOSD patients, only 2 were tested for MOG antibodies (both negative). Two patients with positive serum MOG antibodies did not fulfill 2023 diagnostic criteria for MOGAD.¹⁵ One patient had a MOG titer of 1:40 with short-segment unilateral ON, lack of both perineural sheath enhancement and optic disc edema, and no brain lesions, and so it was classified as ION. The second patient had an initial MOG titer of 1:40 at ON onset with short segment unilateral ON, multiple T2 hyperintense juxtacortical, periventricular, and cerebellar lesions typical for MS, as well as 7 oligoclonal bands in CSF. Repeat MOG titer 2 years later was 1:100. As this patient's presentation fulfilled 2013 diagnostic criteria for pediatric MS,¹³ and a small proportion of MS patients can be MOG seropositive,¹⁷ this patient was considered to have MS and not MOGAD.

Table 2.

Serologic and Imaging Findings.

	Diagnosis
	All(n = 50)	ION(n = 6)	MOGAD(n = 20)	MS(n = 11)	NMOSD(n = 13)
Serum MOG Ab, positive/tested	22/42	1/6	20/20	1/11	0/5
Serum AQP4 Ab, positive/tested	11/49	0/6	0/19	0/11	10/13
First brain MRI within 30 d of onset, % (positive/available)	86 (43/50)	100 (6/6)	90 (18/20)	82 (9/11)	69 (9/13)
Days to first brain MRI, mean (SD)	8.1 (8.0)	7.3 (9.5)	5.1 (4.4)	10.7 (8.5)	12.3 (10.3)
WM lesions on first brain MRI, %	63	67	33	100	89
Enhancing lesions on first brain MRI, %	45	50	18	89	56

Abbreviations: Ab, antibody; AQP4, aquaporin-4; ION, idiopathic optic neuritis; MOG, myelin oligodendrocyte glycoprotein; MOGAD, myelin oligodendrocyte glycoprotein antibody disease; MRI, magnetic resonance imaging; MS, multiple sclerosis; NMOSD, neuromyelitis optica spectrum disorder; WM, white matter.

As there was significant variability (up to years) in the time to the first brain MRI, only brain MRIs performed within 30 days of symptom onset were included (n = 43), with a mean time of 8 days to MR imaging (Table 2). T2-hyperintense brain lesions were identified in 86% of initial MRIs, including those of all MS patients and 89% of NMOSD patients, but even two-thirds of ION patients and only a third of MOGAD patients. Almost half of all patients had enhancing brain lesions, ranging from 18% of MOGAD patients, approximately half of NMOSD and ION patients, and almost all (89%) MS patients.

Features of Initial ON Event

Initial ON features differed across subgroups (Table 3). Bilateral ON occurred in 19 patients (38% of the cohort), most commonly in NMOSD (77%) and MOGAD (40%). Decreased visual acuity was nearly universal (96%). Fourteen patients had other neurologic symptoms with their initial ON event, none in the ION cohort. In the MS cohort, only 1 patient had other neurologic symptoms (cranial neuropathy). Eight MOGAD patients had other neurologic symptoms: 3 with LETM, 2 with altered mental status but with MRIs not consistent with ADEM, 2 with seizure activity within a few weeks of ON, and 1 with motor and sensory changes due to an internal capsule lesion. Five NMOSD patients had other neurologic symptoms in addition to bilateral ON, as follows: LETM, LETM and diencephalic syndrome, LETM and acute brainstem syndrome, cranial neuropathy, and ADEM with acute brainstem syndrome.

Table 3.

Initial ON Features and Treatment.

	Diagnosis
	All(n = 50)	ION(n = 6)	MOGAD(n = 20)	MS(n = 11)	NMOSD(n = 13)
Initial ON features
Bilateral ON, %	38	0	40	9.1%	76.9
Any VA loss, % (positive/known)	98 (45/46)	100 (5/5)	95 (18/19)	100 (10/10)	100 (12/12)
Worst eye logMAR at onset, mean (SD) (positive/known)	2.2 (1.9) (45/46)	1.3 (1.2) (5/5)	2.0 (1.4) (18/19)	2.3 (1.3) (10/10)	3.1 (1.1) (12/12)
Worst eye logMAR ≥1 at onset, % (positive/known)	78 (36/46)	40 (2/5)	74 (14/19)	80 (8/10)	100 (12/12)
Color desaturation, % (positive/known)	73.7 (28/38)	60 (3/5)	64.3 (9/14)	88.9 (8/9)	80 (8/10)
Eye pain, % (positive/known)	73.8 (31/42)	100 (5/5)	52.9 (9/17)	72.7 (8/11)	100 (9/9)
APD, % (positive/known)	73.3 (33/45)	83.3 (5/6)	78.9 (15/19)	50 (5/10)	88.9 (8/9)
Optic nerve edema, % (positive/known)	67.6 (23/34)	80 (4/5)	70.6 (12/17)	83.3 (5/6)	33.3 (2/6)
Other neurologic symptoms, %	26	0	35	9.09	38.5
Received acute treatment for first event, %	90	83.3	95	90.9	84.6
High-dose steroids, %	90	83.3	95	90.9	84.6
Time to steroids, d, mean (SD)	11.4 (16.2)	3.8 (2.7)	8.05 (8.7)	9.9 (7.8)	22 (27.9)
Steroid taper, %	68.9	40	68.4	80	72.7
IVIg, %	12	0	15	9.1	15.4
Time to IVIg, d, mean (SD)	21.2 (13.1)	-	20 (11.1)	42	12.5 (5.0)
Plasma exchange, %	18	0	25	9.10	23.1
Time to plasma exchange, d, mean (SD)	20.8 (12)	-	14.2 (7.9)	34	30.5 (13.4)

Abbreviations: APD, afferent pupillary defect; ION, idiopathic optic neuritis; IVIg, intravenous immunoglobulin; MOGAD, myelin oligodendrocyte glycoprotein antibody disease; MS, multiple sclerosis; NMOSD, neuromyelitis optica spectrum disorder; ON, optic neuritis; VA, visual acuity.

Only 1 patient (MOGAD) had normal visual acuity despite painful eye movements, color desaturation, papilledema, as well as optic nerve enhancement. The logMAR of the worst eye at nadir was most severe in NMOSD (3.1) and mildest in ION (1.3). Color desaturation, eye pain, and afferent pupillary defect were frequently reported, particularly in NMOSD and MS. Optic nerve edema was reported in 68% of patients, highest in ION (80%) and MS (83%).

Acute treatment was administered in 90% of cases. High-dose intravenous (IV) steroids (20-30 mg/kg/d) were used in 88%, initiated earliest in ION (mean 3.8 days) and latest in NMOSD (22 days). Steroid tapers were employed in 65% of patients overall. Plasma exchange and IV immunoglobulin (IVIg) were administered to 18% and 12% of all patients respectively, most commonly in MOGAD (n = 5) and NMOSD (n = 3), on average 3 weeks after onset.

Long Term Treatment and Outcomes

Patients were followed on average for 5.3 years from initial ON event, but this varied substantially with the longest mean follow-up in NMOSD (10.7 years) and shortest in ION (2.7 years) (Table 4). The median number of relapses per patient was 0.5 (range 0-12), which was highest in NMOSD (3). Relapses involving ON specifically occurred most frequently in NMOSD patients with a median of 1 but a range up to 8.

Table 4.

Long-term Treatments and Outcomes.

	Diagnosis
	All(n = 50)	ION(n = 6)	MOGAD(n = 20)	MS(n = 11)	NMOSD(n = 13)
No. of relapses, median (range)	0.5 (0-12)	-	0 (0-7)	0 (0-2)	3 (0-12)
No. of relapses with ON, median (range)	0 (0-8)	-	0 (0-7)	0 (0-2)	1 (0-8)
Ever received DMT, % (positive/known)	60 (29/48)		40 (8/20)	100 (9/9)	92 (12/13)
Number of DMTs, median (range)	1 (0-6)	-	0 (0-3)	1 (0-3)	2 (0-6)
Receiving any DMT at last follow-up, % (positive/known)	48 (22/46)	-	25 (5/20)	78 (7/9)	91 (10/11)
First DMT: high efficacy, % (high/any DMT)	55 (16/29)	-	75 (6/8)	67 (6/9)	42 (5/12)
High-efficacy DMT at last follow-up, % (high/any DMT)	82 (18/22)	-	80 (4/5)	86 (6/7)	80 (8/10)

Abbreviations: DMT, disease-modifying therapy; ION, idiopathic optic neuritis; MOGAD, myelin oligodendrocyte glycoprotein antibody disease; MS, multiple sclerosis; NMOSD, neuromyelitis optica spectrum disorder; ON, optic neuritis.

Because of the small cohort, multivariable modeling by diagnosis for DMT was not possible, and all diagnostic groups were pooled. Overall, 60% of patients received a DMT at some point, ranging from all MS patients and 92% of NMOSD patients to 40% of MOGAD patients and no ION patients. There was a wide range (0-6) of the number of DMTs used for different patients, with a median of one DMT for all patients. At last follow-up, slightly less than half of all patients were actively receiving a DMT, including 25% of MOGAD, 78% of MS and 91% of NMOSD patients. Patients diagnosed prior to 2018 and treated with DMT (n = 14) typically received lower-efficacy therapies (azathioprine, fingolimod, interferon beta, glatiramer acetate, mycophenolate mofetil, prednisone), whereas patients diagnosed since 2018 and treated with DMT (n = 15) were started on high-efficacy agents (rituximab, cyclophosphamide, IVIg, ocrelizumab, satralizumab). In the MOGAD cohort, 8 patients received DMT after a second event. Three of 6 patients who were actively receiving DMT relapsed. One patient relapsed during an attempted prednisone taper, 5 months after starting rituximab. The other 2 patients were on lower efficacy DMT (azathioprine and mycophenolate mofetil) at the time of relapse and had no further relapses after changing to high efficacy DMT (rituximab).

More than half of NMOSD patients had been diagnosed by 2012 when our center began using rituximab as first line DMT for NMOSD. Thus, only 42% of NMOSD patients received high efficacy DMT as first-line therapy compared to 75% of MOGAD and 67% of MS patients. By the final visit, 82% of all patients were on high-efficacy treatment, including 80% of NMOSD patients. Although Kaplan-Meier analysis showed 29% relapse-free survival in the high-efficacy DMT group vs 8% in the low-efficacy group, this difference was not statistically significant (Figure 2).

To assess subacute visual recovery following the initial ON event, only patients with a second vision examination within 12 months from onset (“second examination”) were evaluated (Table 5). Eighty-six percent of patients had an examination within 12 months, at a mean of 2.5 months from onset. Functional visual recovery varied by underlying diagnosis, with a mean logMAR of the worst eye of 0.7 and a change from onset of 0.5 in all patients (Figure 3). Patients with MS showed the most improvement within 1 year (logMAR 0.2), and patients with NMOSD the least (logMAR 1.9). Overall, 26% of patients were legally blind (defined in the United States as Snellen 20/200, logMAR ≥1) in the worst eye at the second examination, ranging from zero MS patients to 70% of NMOSD patients.

Table 5.

Vision Outcomes at Second and Final Examinations.

	Diagnosis
	All(n = 50)	ION(n = 6)	MOGAD(n = 20)	MS(n = 11)	NMOSD(n = 13)
Patients with second examination within 12 mo of onset, %	86 (43/50)	100 (6/6)	90 (18/20)	82 (9/11)	77 (10/13)
Time from onset to second examination, , mo, mean (SD)	2.5 (1.7)	2.4 (2.2)	1.8 (1.0)	2.4 (2.1)	3.2 (1.5)
Patients with worst eye logMAR ≥1 at second examination, % (positive/known)	26 (11/43)	33 (2/6)	11 (2/18)	0 (0/11)	70 (7/10)
Time from onset to final examination, , mo, mean (SD)	59.2 (61.1)	36.3 (24.4)	41.4 (28.7)	40.7 (23.6)	112 (95.2)
Patients with worst eye logMAR ≥1 at final examination, % (positive/known)	28 (13/46)	0 (0/5)	11 (2/19)	10 (1/10)	83 (10/12)

Abbreviations: ION, idiopathic optic neuritis; MOGAD, myelin oligodendrocyte glycoprotein antibody disease; MS, multiple sclerosis; NMOSD, neuromyelitis optica spectrum disorder.

At the last vision examination (almost 5 years from onset on average), the mean logMAR was 0.8 overall, again worst in NMOSD (2.3) and best in ION (0.1). The breakdown of patients legally blind in their worst eye at the final examination was quite variable among diagnoses: 0% of ION patients, 11% of MOGAD patients, 10% of MS patients, and 83% of NMOSD patients. The percentage of patients with legal blindness in their worst eye at the final examination was relatively stable, accounting for 28% of all patients. Four patients (all NMOSD) were legally blind in both eyes at their last follow-up.

Discussion

This study, which presents the clinical features and outcomes of 50 children with a first-time optic neuritis event at a single site, spans 33 years, one of the longest lengths of follow-up of any such study to date. Our cohort is similar to other reports in terms of a mean age at onset of 12 years and a female preponderance.^3,5,6,8 A notable difference in our population is a higher proportion of Black patients, representing half of all patients and more than two-thirds of MS and NMOSD patients, compared with 10% to 20% of patients in other studies.^3,5 Although some of this discrepancy is due to a higher Black population in Alabama and the Southeast region (25%-30%),¹⁸ this does not fully explain the disproportionate representation of Black patients in our cohort and warrants further explanation. The only patients allowed by our inclusion criteria not tested for serum MOG antibodies were those with positive serum AQP4 antibodies. This exception made our NMOSD population skew less recent than other groups, with a median year of onset in 2012 compared with 2018-2019 for other groups. As treatment practices have changed over time, this very likely affected the outcomes of our NMOSD group as described below.

We also found a higher proportion of children whose final diagnosis was other than ION: 88% in our population compared with 30% to 60% in other studies.^{5,7–9,19,20} This could be related to our significantly longer follow-up time, averaging more than 5 years compared with 6 months or 2 years in most other studies. The emergence of MOGAD as a distinct entity in the past 10 years, which includes patients with a monophasic disease course that previously would have been characterized as ION, is another potential explanation for this discrepancy with older studies.^6,8,9 Our proportion of MS patients was comparable to many studies (10%-20%),^{5,7,9,10,19,21} but our NMOSD cohort was higher at 26% compared with 7% to 15% reported elsewhere.^5,7,8,20

The clinical features of the first ON event were overall similar to those reported in other published reports,^{5,6,8–10,22} with more than a third of patients presenting with bilateral optic neuritis (including three-fourths of NMOSD patients and more than half of MOGAD patients). The majority of patients had vision loss, color desaturation, eye pain, afferent pupillary defect, and papilledema. A fourth of our patients had other neurologic symptoms along with optic neuritis, which has been described to occur in more than a third of children with ON.⁶ Our patients overall had severe vision loss at their nadir with a median logMAR of 2.4 (hand motion only) and almost 80% with a logMAR of 1 or worse. Many other studies show 20% to 50% of patients with nadir logMar of 1,^5,7,8 although some report up to 70% to 80%.^6,22

All patients eventually underwent MR imaging of their brain, but some with a delay of years from their ON event. Of the 43 patients who had MR imaging of their brain within 30 days of the initial attack, two-thirds had white matter (WM) lesions and almost half had enhancing lesions. All patients with a final diagnosis of MS had WM abnormalities on their first MRI, 89% with contrast enhancement, which has been well described in other reports.^6,10,23 Eighty-nine percent of our NMOSD patients had WM lesions compared with 25% to 70% reported elsewhere.^24–26 Surprisingly, two-thirds of the ION patients also had WM lesions, half of which enhanced with contrast, and conversely the MOGAD patients were less likely to have WM lesions (33%) or enhancement (18%). Interestingly, 3 of the 4 ION patients with WM lesions had MS-like periventricular lesions, but no lesions in other areas fulfilling dissemination in space criteria for MS.¹³ One patient had no repeat brain MRI available, the other 2 had no new lesions on their most recent brain MRIs (1 at 6 months, 1 at 6 years from initial imaging). These patients had no other clinical or diagnostic features suggestive of other central nervous system inflammatory conditions. Of note, under the recently published 2024 changes to the MS diagnostic criteria,²⁷ these patients would now qualify for a diagnosis of MS, despite having a monophasic disease course and stable neuroimaging.

The vast majority of patients received appropriate acute treatment with high-dose steroids within 2 weeks of onset. Of the 4 patients who did not receive any acute treatment, 1 had a diagnosis of MS, 1 of MOGAD, and 2 of NMOSD. These latter 2 patients did not receive acute treatment for their second attacks of optic neuritis as well. Although the ONTT argued that acute treatment with IV steroids does not affect long-term visual outcomes but only hastens recovery,^28–30 a recent study in adult ON³¹ showed that a delay of 3 days in acute treatment portends worse visual outcomes in NMOSD, whereas MOGAD patients can tolerate a delay in treatment up to a week. Most patients received a steroid taper after high-dose steroids, although both the ONTT and more recent studies do not demonstrate an effect on visual outcomes.^28,32 A minority of patients received IVIg or plasma exchange (12% and 18%), in general a few weeks after symptom onset. There are conflicting reports in the literature on the effectiveness of IVIg in ON: Roed et al³³ showed no difference in treatment with IVIg vs placebo, whereas Tselis et al³⁴ found that most patients (78%) who received both IV steroids and IVIg had good visual recovery compared with those who received IV steroids only (23%). The effectiveness of plasma exchange in pediatric demyelinating syndromes is more clear,^35,36 although data in ON specifically are lacking.

As mentioned earlier, our follow-up range was quite high: from 2 weeks to 33 years, with a mean of over 5 years. In terms of relapses, our NMOSD patients were again outliers: although other groups had a median of no relapses, they had a median of 3 relapses, including a median of 1 ON relapse.

A little more than half of our patients were treated with a DMT at some point—ranging from none of the ION patients to 92% of NMOSD patients. There was a wide range in the number, types, and lengths of long-term treatment. Most patients received a median of 1 DMT, with our NMOSD population receiving the highest at 2. Slightly less than half of all patients were still receiving a DMT at their last follow-up—the least in MOGAD (25%) and highest in MS (78%) and NMOSD (91%). A high-efficacy DMT was chosen as first-line in 55% of all patients: 75% of MOGAD patients but only 42% of NMOSD patients. However, at the last follow-up, 82% of patients overall (including 80% NMOSD) were receiving high-efficacy DMT. This is reflective of treatment practice changes over the past 30 years from an escalation approach to a more intensive induction approach. In an attempt to distinguish the effect of high- vs low-efficacy DMT and the natural disease course of different demyelinating disorders, survival analysis showed that 92% of patients on low-efficacy DMT had a relapse after starting treatment compared to 71% on high efficacy DMT, although notably this difference was not statistically significant. At our site, rituximab was first used in NMOSD as second-line treatment in 2006 and first line in 2012 and became first-line treatment for MS in 2018. After this practice change, we subjectively noted a dramatic decrease in relapses overall, although this is not reflected in our survival analysis.

More than half of patients had full visual recovery from their first event, although another quarter were legally blind at both their second and final examinations. Our NMOSD group had a mean logMAR of 1.9 at the second examination and 2.3 at their last examination, and 83% were legally blind in their worst eye at last follow-up. This is similar to reports in other studies,^6,22,31,37 although as discussed above, it is not clear how much of the divergence in our NMOSD population can be attributed to shifts in clinical practices, the natural course of the disease, or potentially other unidentified factors.

This study's strengths include an unusually long follow-up period compared with other similar studies, which allowed us to track the disease course and long-term recovery over a much longer period. Our study represents a distinctive population, with all but one patient residing in the Southeast United States, and as discussed before, a disproportionately high Black population that deserves further inquiry.

Our study is limited in that it represents data from a single site with a relatively modest cohort size. This study was a prospective observational study, and thus dependent on the clinical data available for each patient. There was a wide variability in the timing of imaging and serologic studies, acute and long-term treatments and modalities, as well as clinical and examination follow-up time frames. Despite these limitations, this study provides meaningful information on the clinical characteristics and long-term outcomes of children presenting with a first-time optic neuritis event.

Conclusion

In conclusion, most children presenting with an initial ON event were diagnosed with a chronic demyelinating disease, although half did not have recurrence of ON or another demyelinating event. Over a third of children had bilateral ON, and although most had severe vision loss at their nadir, visual recovery was excellent in more than half, with a quarter having limited recovery. Children with NMOSD had notably worse outcomes, including significant long-term visual disability and more frequent relapses. This difference highlights both the comparatively more severe natural course of NMOSD, as well as the changes in practice norms for treating these patients. Notably, pediatric ON is less likely to be a self-limited monophasic disease than is often presumed.

Footnotes

ORCID iDs

Lydia Marcus

Andrew Van

Miguel Arias

Hazel Quesada

Immaculada Aban

Author Contributions

LM (first author): conceptualization, methodology, writing, reviewing, editing, supervision. AV: methodology, writing, reviewing, editing. TC: methodology, writing, reviewing, editing. Sanford Williams: methodology, writing, reviewing, editing. MA: methodology, writing, reviewing, editing. HQ: methodology, statistical analysis, reviewing. IA: methodology, statistical analysis, reviewing, supervision. RGT: conceptualization, methodology, reviewing. JN: conceptualization, methodology, writing, reviewing, editing, supervision.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Pau

Al Zubidi

Yalamanchili

Plant

Lee

. Optic neuritis. Eye. 2011;25(7):833‐842. doi:10.1038/eye.2011.81

Rodriguez

Brien

PCO

Kurland

. Optic neuritis: a population-based study in Olrnsted County, Minnesota Neurology. 1995;45(2):244‐250.

Langer-Gould

Zhang

Chung

Yeung

Waubant

Yao

. Incidence of acquired CNS demyelinating syndromes in a multiethnic cohort of children. Neurology. 2011;77(12):1143-1148.

Banwell

Kennedy

Sadovnick

, et al. Incidence of acquired demyelination of the CNS in Canadian children. Neurology. 2009;72(3):232‐239. doi:10.1212/01.wnl.0000339482.84392.bd

Pineles

Henderson

Repka

, et al. The Pediatric Optic Neuritis Prospective Outcomes Study: two-year results. Ophthalmology. 2022;129(8):856‐864. doi:10.1016/J.OPHTHA.2022.03.021

Wilejto

Shroff

Buncic

Kennedy

Goia

Banwell

. The clinical features, MRI findings, and outcome of optic neuritis in children. Neurology. 2006;67(2):258‐262. doi:10.1212/01.WNL.0000224757.69746.FB,

Pineles

Repka

Liu

, et al. Assessment of pediatric optic neuritis visual acuity outcomes at 6 months. JAMA Ophthalmol. 2020;138(12):1253‐1261. doi:10.1001/JAMAOPHTHALMOL.2020.4231

Wan

Adebona

Benson

Gorman

Heidary

. Visual outcomes in pediatric optic neuritis. Am J Ophthalmol. 2014;158(3):503‐507.e2. doi:10.1016/J.AJO.2014.05.036

Park

Yang

Han

, et al. Characteristics of optic neuritis in South Korean children and adolescents: a retrospective multicenter study. J Ophthalmol. 2022;2:1‐8. doi:10.1155/2022/4281772

10.

Waldman

Stull

Galetta

Balcer

Liu

. Pediatric optic neuritis and risk of multiple sclerosis: meta-analysis of observational studies. J Am Assoc Pediatr Ophthalmol Strabismus. 2011;15(5):441‐446. doi:10.1016/J.JAAPOS.2011.05.020

11.

Yeh

Graves

Benson

Wassmer

Waldman

. Pediatric optic neuritis. Neurology. 2016;87(9):S53‐S58. doi:10.1212/WNL.0000000000002822

12.

Kunchok

Chen

McKeon

Mills

Flanagan

Pittock

. Coexistence of myelin oligodendrocyte glycoprotein and aquaporin-4 antibodies in adult and pediatric patients. JAMA Neurol. 2020;77(2):257‐259. doi:10.1001/jamaneurol.2019.3656

13.

Krupp

Banwell

Tenembaum

. Consensus definitions proposed for pediatric multiple sclerosis and related disorders. Neurology. 2007;68(16 suppl 2):S7-S12. doi:10.1212/01.WNL.0000259422.44235.A8

14.

Wingerchuk

Banwell

Bennett

, et al. International consensus diagnostic criteria for neuromyelitis optica spectrum disorders. Neurology. 2015;85(2):177‐189. doi:10.1212/WNL.0000000000001729

15.

Banwell

Bennett

Marignier

, et al. Diagnosis of myelin oligodendrocyte glycoprotein antibody-associated disease: International MOGAD Panel proposed criteria. Lancet Neurol. 2023;22(3):268‐282. doi:10.1016/S1474-4422(22)00431-8

16.

Tomaiuolo

Woreta

, et al. Open-globe injury repairs in the American Academy of Ophthalmology IRIS® Registry 2014-2018: Incidence, risk factors, and visual outcomes. Ophthalmology. 2023;130(8):812‐821. doi:10.1016/j.ophtha.2023.03.002

17.

Gaudioso

Mar

Casper

, et al. MOG And AQP4 antibodies among children with multiple sclerosis and controls. Ann Neurol. 2023;93(2):271‐284. doi:10.1002/ANA.26502

18.

United States Census Bureau. Published 2025. https://www.census.gov.

19.

Bonhomme

Waldman

Balcer

, et al. Pediatric optic neuritis: brain MRI abnormalities and risk of multiple sclerosis. Neurology. 2009;72(10):881‐885. doi:10.1212/01.WNL.0000344163.65326.48

20.

Song

Zhou

Yang

, et al. Clinical characteristics and prognosis of myelin oligodendrocyte glycoprotein antibody-seropositive paediatric optic neuritis in China. Br J Ophthalmol. 2019;103(6):831‐836. doi:10.1136/BJOPHTHALMOL-2018-312399

21.

Lee

Han

Yang

. Population-based incidence of pediatric and adult optic neuritis and the risk of multiple sclerosis. Ophthalmology. 2020;127(3):417‐425. doi:10.1016/J.OPHTHA.2019.09.032

22.

Sudhakar

Kini

. Clinical features and visual outcome of pediatric optic neuritis. Indian J Ophthalmol. 2023;71(6):2603‐2604. doi:10.4103/IJO.IJO_440_23

23.

Heussinger

Kontopantelis

Rompel

Paulides

Trollmann

. Predicting multiple sclerosis following isolated optic neuritis in children. Eur J Neurol. 2013;20(9):1292‐1296. doi:10.1111/ENE.12184

24.

Cortese

Carrasco

Tur

, et al. Differentiating multiple sclerosis from AQP4-neuromyelitis Optica spectrum disorder and MOG-antibody disease with imaging. Neurol ®. 2023;100(3):308‐323. doi:10.1212/WNL.0000000000201465

25.

Pittock

Lennon

Krecke

Wingerchuk

Lucchinetti

Weinshenker

. Brain abnormalities in neuromyelitis optica. Arch Neurol. 2006;63(3):390‐396. doi:10.1001/ARCHNEUR.63.3.390,

26.

Kim

Paul

Lana-Peixoto

, et al. MRI characteristics of neuromyelitis optica spectrum disorder: an international update. Neurology. 2015;84(11):1165‐1173. doi:10.1212/WNL.0000000000001367

27.

Montalban

Lebrun-Frénay

, et al. Diagnosis of multiple sclerosis: 2024 revisions of the McDonald criteria. Lancet Neurol. 2025;24(10):850‐865. doi:10.1016/S1474-4422(25)00270-4

28.

Beck

Cleary

Anderson

, et al. A randomized, controlled trial of corticosteroids in the treatment of acute optic neuritis. The Optic Neuritis Study Group. N Engl J Med. 1992;326(9):581‐588. doi:10.1056/NEJM199202273260901

29.

Beck

Moke

Trobe

. The 5-year risk of MS after optic neuritis. Neurology. 1998;51(4):1237‐1238. doi:10.1212/wnl.51.4.1237-a

30.

Gal

Brodsky

Nazarian

, et al. Visual function 15 years after optic neuritis. A final follow-up report from the Optic Neuritis Treatment Trial. Ophthalmology. 2008;115(6):1082.e5. doi:10.1016/j.ophtha.2007.08.004

31.

Min

Moon

Kwon

, et al. Prognostic factors of first-onset optic neuritis based on diagnostic criteria and antibody status: a multicentre analysis of 427 eyes. J Neurol Neurosurg Psychiatry. 2024;95(8):753‐760. doi:10.1136/jnnp-2023-333133

32.

Wilf-Yarkoni

Feldmann

Rubarth

, et al. Effectiveness of oral prednisone tapering following intravenous methylprednisolone for acute optic neuritis in multiple sclerosis. PLoS One. 2023;18(12):1‐11. doi:10.1371/journal.pone.0288366

33.

Roed

Langkilde

Sellebjerg

, et al. A double-blind, randomized trial of IV immunoglobulin treatment in acute optic neuritis. Neurology. 2005;64(5):804‐810. doi:10.1212/01.WNL.0000152873.82631.B3,

34.

Tselis

Perumal

Caon

, et al. Treatment of corticosteroid refractory optic neuritis in multiple sclerosis patients with intravenous immunoglobulin. Eur J Neurol. 2008;15(11):1163‐1167. doi:10.1111/J.1468-1331.2008.02258.X,

35.

Bigi

Banwell

Yeh

. Outcomes after early administration of plasma exchange in pediatric central nervous system inflammatory demyelination. J Child Neurol. 2015;30(7):874‐880. doi:10.1177/0883073814545883,

36.

Koziolek

Mühlhausen

Friede

, et al. Therapeutic apheresis in pediatric patients with acute CNS inflammatory demyelinating disease. Blood Purif. 2013;36(2):92‐97. doi:10.1159/000354077,

37.

Chen

Zhao

Huang

, et al. Clinical characteristics of pediatric optic neuritis with myelin oligodendrocyte glycoprotein seropositive: a cohort study. Pediatr Neurol. 2018;83:42‐49. doi:10.1016/j.pediatrneurol.2018.03.003