Clinical Features and Prognostic Factors in Anti-Myelin Oligodendrocyte Glycoprotein Antibody Positive Optic Neuritis

ABSTRACT

In order to review the clinical features of anti-myelin oligodendrocyte glycoprotein antibody positive optic neuritis (MOGON), we investigated the clinical characteristics, visual function, optical coherence tomography findings, and magnetic resonance imaging of 31 patients (44 eyes). MOGON was more common in middle age without sex difference and was characterised by pain on eye movement and optic disc swelling. Magnetic resonance imaging lesions tended to be long with inflammation around the optic nerve sheath; longer lesions were associated with worse visual acuities at onset. Recurrence was significantly associated with retinal nerve fibre layer thinning, and thus, it is important to reduce recurrence as much as possible.

Introduction

Optic neuritis is a disease that causes acute loss of visual function and may result in residual serious visual dysfunction. Recently, measurements of anti-aquaporin 4 (AQP4)¹ and anti-myelin oligodendrocyte glycoprotein (MOG) antibodies² have become available, making it possible to evaluate the pathogenesis of the disease.

Clinical features of anti-MOG antibody positive optic neuritis (MOGON) include ocular pain and optic disc swelling and a good response to steroid pulse therapy. Moreover, MOGON is considered to have a better prognosis as compared to anti-AQP4 antibody positive optic neuritis.³ However, there are a few reported cases in which patients have been left with severe visual impairment.⁴ Thus, it is important to clearly understand each of the clinical features and prognostic factors when managing optic neuritis.

In the present study, we investigated the clinical features and prognostic factors of a series of patients with MOGON.

Methods

Study design and patients

We reviewed the medical records of patients with optic neuritis who visited the Inouye Eye Hospital from January 2010 to October 2022 and met the diagnostic criteria for optic neuritis by Petzold et al.⁵ We included 49 MOGON patients who were positive for serum anti-MOG antibodies and met the criteria for MOGAD by Banwell et al.⁶ Among them, cases with visual impairment due to other retinal diseases or optic nerve disorders, cases with neurological impairment preceding optic neuritis, cases with follow-up observations of less than 6 months, and cases with unclear progression details were excluded. We were left with 44 eyes from 31 cases that developed optic neuritis (affected eyes) and 18 eyes that did not develop optic neuritis (non-affected eyes). The study was approved by the Ethics Committee of Inoue Eye Hospital.

Clinical assessment

Optic neuritis was diagnosed based on the ophthalmological work-up including the medical history and ophthalmological examination (best corrected visual acuity [BCVA], intraocular pressure, relative afferent pupillary defect status, fundus examination, visual field examination, and magnetic resonance imaging [MRI]). Anti-MOG antibody measurements were performed after obtaining informed consent, and blood samples were taken before treatment or during the follow-up period. After collection, the blood samples were kept cold on ice and transported to the Department of Neurology, Niigata University School of Medicine, or Tohoku University School of Medicine for anti-MOG antibody testing. A cell-based assay was used to qualitatively determine the anti-MOG antibody positivity. Recurrent optic neuritis was defined as acute visual dysfunction lasting more than 24 h and accompanied by clinical symptoms such as ocular pain and optic disc swelling, or findings on MRI, after 1 month of onset in anti-MOG antibody-positive cases.

BCVA was assessed using the Snellen decimal visual acuity, and off-chart visual acuity was converted as follows: 0.01 for count fingers; 0.005 for hand motion⁷; and 0 for the presence or absence of perception of light. We examined the nadir of the BCVA in the acute phase, and the last BCVA in the chronic phase.

Visual field testing was performed with the Humphrey Field Analyser (HFA: 30-2 Program) and the Goldmann visual field test was used to assess patterns of visual field loss. The mean deviation (MD) values were calculated by the HFA.

Optical coherence tomography (OCT) was performed to obtain disc images, and the circumpapillary retinal nerve fibre layer (cpRNFL) was classified into superior, temporal, inferior, and nasal areas. The mean value of each nerve fibre layer area thickness was evaluated by the device software. Images of the macula were also taken by OCT, and the ganglion cell layer (GCL) thickness was measured around the macula using ganglion cell analysis, with the average value recorded.

When orbital magnetic resonance imaging (MRI) was performed, the optic nerve was divided into the pre-orbit, retro-orbit, canalicular, and intra-cranial portions as well as the optic chiasm and optic tract (Figure 1).⁸ The approximate length of the optic nerve lesion scored from 0 to 6. Optic nerve segments with gadolinium contrast on fat-suppressed T1-weighted imaging or with high signal on short tau inversion recovery imaging with apparent laterality in the acute phase were regarded as acute optic nerve lesions. The interpretations of the optic nerve lesions were performed by neuro-ophthalmologists (A.Y., M.I., and M.W.) with reference to the reports of the radiologists. We evaluated the features of the MRI lesions and classified the MRI findings into two or fewer lesion segments (short lesion group), and three or more lesion segments (long lesion group) and then compared the visual function and OCT findings between the two groups.

Figure 1.

Definition of each segment of the magnetic resonance imaging lesion. The optic nerve was divided into the pre-orbit, retro-orbit, canalicular, and intra-cranial portions as well as the optic chiasm and optic tract.

Differences in visual function and OCT findings between the recurrence and no-recurrence groups were also compared. The group with one or more recurrences was defined as the recurrence group, while the group with no recurrence was defined as the no-recurrence group.

Statistical analysis

Statistical analysis was performed using Statistical Package for the Social Sciences software (IBM SPSS Statistics ver. 22.0). Continuous data were reported as median (minimum, maximum) and mean ± standard deviation. The Mann-Whitney U-test was used to evaluate non-parametric data with no correspondence (comparison of visual function tests and OCT findings between the affected and non-affected eyes, between the short and long lesion groups based on the length of the MRI lesions, between the recurrence and no-recurrence groups, and comparison of the age and myopic degree between the recurrence and no-recurrence groups). Spearman’s rank correlation coefficient was used to determine the correlations between the visual function tests at the first and last visit, and the OCT findings at the first and last visit. The association was also evaluated between other parameters (age, duration from onset to the first visit) and visual function or OCT findings using Spearman’s rank correlation coefficient. Wilcoxon signed rank sum test was used for the correlated data (difference between visual function at the first and final visit, and between the OCT findings at the first and final visit). A p value of < .05 was considered statistically significant.

Results

Clinical characteristics

Table 1 summarises the clinical characteristics. Age at onset was 44 (20, 70) years, with 17/31 (54.8%) of the patients being female. The duration from onset to the first visit was 11 (1, 37) days and the follow-up period was 51.5 (6, 195) months. Bilateral simultaneous onset was observed in 9/31 (25.8%) patients. Recurrence during the follow-up period was observed in 7/31 patients (10/44 eyes), with pain during eye movement present in 39/44 (88.6%) eyes. Optic disc swelling was observed in 42/44 (95.5%) eyes at the onset. The pattern of visual field defects at the initial presentation was central scotoma in 55.8%, temporal hemianopia in 20.9%, superior or inferior hemianopia in 16.3%, and nasal hemianopia in 7.0%. During the acute phase, methylprednisolone pulse therapy was administered in 39/44 eyes, while 5/44 eyes did not undergo treatment.

Table 1.

Clinical characteristics at the initial presentation.

Total number		31 patients (44 eyes)
Age at onset	Mean	44.2 ± 14.2 years
	Median	44 (20, 70) years
Number of females		17 patients (54.6%)
Bilateral simultaneous onset		8/31 patients (25.8%)
Pain during eye movement		39/44 eyes (88.6%)
Optic disc swelling		42/44 eyes (95.4%)
Time from onset to first visit	Mean	12.0 ± 7.63 days
	Median	11.0 (1, 37) days
Follow-up period	Mean	56.02 ± 48.52 months
	Median	51.5 (1, 195) months
Recurrence during follow-up		10/44 eyes (22.7%)
Visual field defects
Central scotoma		24/43 eyes (55.8%)
Temporal hemianopia		9/43 eyes (2.9%)
Superior or inferior hemianopia		7/43 eyes (16.3%)
Nasal hemianopia		3/43 eyes (7.0%)
Acute therapy
Intravenous methylprednisolone		39/44 eyes (88.6%)
No medication		5/44 eyes (11.4%)

Assessment of visual function and OCT findings

Table 2 shows the visual function at the first and last visit. At the first visit, the median BCVA was 0.30 (0.3, 1.2), with median MD values of −18.3 (−34.1, −2.26) dB. At the last visit, the median BCVA was 1.2 (0.9, 1.2) which was significantly better than that observed at the first visit (p < .001), with an MD value of −1.80 (−6.85, 2.84) dB, which had also significantly improved (p < .001). There was no correlation found between the BCVA at the initial and last visit (r = −0.012, p = .935), or between the MD values at the initial and last visit (r = 0.274, p = .15). Compared with the 18 non-affected eyes at the last visit, there were also no significant differences found for the BCVA and MD values of the 44 affected eyes (BCVA: affected eyes versus non-affected eyes, 1.2 [0.9, 1.2] versus −0.079 [1.0, 1.2], p = .1080; MD value; affected eyes versus non-affected eyes, −1.80 [−6.85, 2.84] versus −0.93 [−4.30, 2.91], p = .17).

Table 2.

Visual function and optical coherence tomography data for affected and non-affected eyes.

		Clinical data			p value
		Affected eye		Non-affected eye	Affected eye at the initial visit versus affected eye at the last visit	Affected eye at the last visit versus non-affected eye at the last visit
		Initial visit	Last visit	Last visit
BCVA	Mean	0.41 ± 0.41	1.18 ± 0.05	1.2 ± 0.03	p = .935(S) (r = −.012)	p = .1080(M)
	Median	0.3 (0.3, 1.2)	1.2 (0.9, 1.2)	1.2 (1.0, 1.2)
MD value (dB)	Mean	−18.3 ± 10.0	−2.07 ± 2.23	−1.06 ± 2.02	p = .15(S) (r = .274)	p = .1667(M)
	Median	−19.7 (−34.1, −2.26)	−1.80 (−6.85, 2.84)	−0.93 (−4.30, 2.91)
cpRNFL thickness (μm)	Mean	138.31 ± 59.77	75.92 ± 12.62	91.68 ± 13.88	p = .767(S) (r = .054)	p = .0002(M)*
	Median	116.37 (58.25, 288)	77.75 (51, 98)	96.75 (54, 106)
GCL thickness (μm)	Mean	76.43 ± 9.23	65.28 ± 9.11	76.2 ± 13.0	p = .011(S)* (r = .739)	p = .0012(M)*
	Median	78 (59,89)	64 (40, 86)	80 (42, 88)

BCVA = Best-corrected visual acuity; cpRNFL = Circumpapillary retinal nerve fibre layer; GCL = Ganglion cell layer; MD = Mean deviation.

Statistical analysis: affected eye at the initial visit versus affected eye at the last visit, Spearman’s rank correlation coefficient (S); affected eye at the last visit versus non-affected eye at the last visit, Mann-Whitney U test (M); *Significant difference, p < .05.

At the first visit, the OCT results of affected eyes showed that the average cpRNFL thickness was 116.37 (58.25, 288) μm and the average GCL thickness was 78 [59, 89] μm. The average cpRNFL and GCL thicknesses at the last visit were 96.75 (54, 106) μm and 64 (40, 86) μm, respectively. There was significant thinning for both of these as compared to that observed for the first visit (average cpRNFL thickness: p < .05; average GCL thickness: p < .05). The non-affected eyes at the last visit exhibited no significant changes in the average cpRNFL thickness (96.75 [54, 106] μm) and average GCL thickness (80 [42, 88] μm). Compared with the non-affected eyes, there was significant thinning in the affected eyes for the average cpRNFL (p < .01) and GCL thickness (p = .012) at the last visit. Examination of the correlation between the OCT findings at the first and last visit showed that the average cpRNFL thickness exhibited no correlation (r = −0.054, p = .767), while there was a correlation observed for the average GCL thickness (r = .739, p = .011).

There was no correlation observed between the age or duration after disease onset and the visual function parameters at baseline and during the follow-up, and between the age or time after disease onset and the OCT findings at baseline and during the follow-up (Table 3).

Table 3.

Correlation of age, duration from onset to first visit with visual function and optical coherence tomography findings.

	Age		Duration from onset to first visit
	Initial	Last	Initial	Last
BCVA	p = .38 (r = .13)	p = .26 (r = −.18)	p = .47 (r = .12)	p = .13 (r = −.25)
MD value (dB)	p = .54 (r = −.11)	p = .089 (r = −.28)	p = .50 (r = −.13)	p = .67 (r = −.078)
cpRNFL thickness (μm)	p > .90 (r = −.007)	p = .40 (r = −.14)	p = .20 (r = .23)	p = .20 (r = .23)
GCL thickness (μm)	p = .71 (r = −.10)	p = .50 (r = −.12)	p = .28 (r = .31)	p = .077 (r = .33)

BCVA = Best corrected visual acuity; cpRNFL = Circumpapillary retinal nerve fibre layer; GCL = Ganglion cell layer; MD = Mean deviation.

Statistical analysis: Spearman’s rank correlation coefficient; *Significant difference, p < .05.

Assessment of MRI findings

Information on the orbital MRI at the initial visit was available for 37 eyes (Table 4). The percentages according to the lesion location were pre-orbital 32/37 eyes (86.4%), retro-orbital 29/37 eyes (78.3%), canalicular 23/37 eyes (62.1%), intra-cranial 19/37 eyes (51.3%), chiasm 2/37 (5.4%), and optic tract 0/37 (0%). There were 16/37 eyes (43.2%) that had lesions of two segments or less and 21/37 eyes (56.8%) that had lesions longer than three segments. Inflammatory spread to the optic nerve sheath was observed in 24/37 eyes (64.8%). Lesions involving areas other than the optic nerve were seen in two patients, with a spinal cord lesion in one patient, and a brainstem lesion in one patient.

Table 4.

Orbital magnetic resonance imaging findings at the initial presentation.

Position of lesion
Pre-orbit		32/37 eyes (86.4%)
Retro-orbit		29/37 eyes (78.3%)
Canalicular		23/37 eyes (62.1%)
Intra-cranial		19/37 eyes (51.3%)
Chiasm		2/37 eyes (5.4%)
Optic tract		0/37 eyes (0%)
Length of lesions	Mean	2.83 ± 1.23 segments
	Median	3 (1, 5) segments
≤2 segments		16/37 eyes (43.2%)
≥3 segments		21/37 eyes (56.8%)
Other lesions		2/31patients (6.4%)

After further classifying the 37 eyes with MRI findings into two or fewer lesion segments (short lesion group: 16 eyes) and three or more lesion segments (long lesion group: 21 eyes), we then compared the visual function and OCT findings between the two groups (Table 5). There was a significantly greater disturbance in the BCVA at the initial visit for the long lesion group (BCVA: short lesion group versus long lesion group, 0.55 [0.01, 1.2] versus 0.09 [0, 1.2], p < .01). However, there were no significant differences for the MD values at the first visit (MD values: short lesion group versus long lesion group, −13.0 [−30.9, −2.26] dB versus −19.7 [−31.9, −4.1] dB, p = .253). At baseline on OCT, there was a trend for thicker cpRNFL thickness in the long lesion group (cpRNFL thickness: short lesion group versus long lesion group, 116.27 [58.25, 249] μm versus 130.25 [107.25, 288] μm, p < .071). There were no significant differences observed for BCVA, MD, cpRNFL thickness, and GCL thickness at the last visit between the long and short MRI lesion groups at baseline. Comparing 32 eyes with pre-orbit MRI lesions and five eyes without pre-orbit MRI lesions, the cpRNFL was thicker with the presence of pre-orbit MRI lesions, but not significantly so (with pre-orbit lesions versus without pre-orbit lesion: 94.0 [58.3–249.0] μm versus 118.00 [69.75–288.0] μm, p = .145).

Table 5.

Comparison of visual function and optical coherence tomography findings by magnetic resonance lesion length at the initial presentation.

		At initial visit		At last visit
		Short lesion (16 eyes)	Long lesion (21 eyes)	Short lesion (16 eyes)	Long lesion (21 eyes)
BCVA	Mean	0.58 ± 0.41	0.27 ± 0.36	1.17 ± 0.09	1.20 ± 0
	Median	0.55 (0.01, 1.20)	0.09 (0, 1.20)	1.20 (0.90, 1.20)	1.20 (1.20, 1.20)
	p value	p = .010*		p = .100
MD value (dB)	Mean	−15.3 ± 10.7	−19.9 ± 9.4	−1.73 ± 1.47	−2.05 ± 2.36
	Median	−13.0 (−30.9, −2.26)	−19.7 (−31.9, −4.1)	−1.23 (−4.44, 0.20)	−2.11 (−5.90, 2.84)
	p value	p = .0619		p = .562
cpRNFL thickness (μm)	Mean	111.31 ± 52.49	166.27 ± 58.4	78.23 ± 12.87	77.72 ± 11.0
	Median	116.27 (58.25, 249)	130.25 (107.25, 288)	82 (53.5, 98)	80.13 (54.25, 92.25)
	p value	p = .0004*		p = .5430
GCL thickness (μm)	Mean	72.8 ± 9.41	80.7 ± 7.6	64.7 ± 11.5	67.0 ± 7.8
	Median	80 (59, 82)	83 (68, 89)	65 (40, 86)	65.5 (55, 80)
	p value	p = .0614		p = .6049

BCVA = Best corrected visual acuity; cpRNFL = Circumpapillary retinal nerve fibre layer; GCL = Ganglion cell layer; MD = Mean deviation.

Statistical analysis: Mann-Whitney U test; *Significant difference, p < .05.

Differences between visual function and OCT findings in the recurrence group

Visual function at the last visit was compared between the recurrence (10 eyes) and no-recurrence (34 eyes) groups, with no significant difference found Table 6. Regarding the OCT findings at the last visit, the recurrence group exhibited a significant thinning of average cpRNFL and GCL thickness compared with the no-recurrence group (cpRNFL thickness: no-recurrence group versus recurrence group, 81.3 [53.5, 98] μm versus 68.3 [51, 83] μm, p = .024; GCL thickness: no-recurrence group versus recurrence group, 67 [40, 86] μm versus 61 [55, 72] μm, p = .043). There were no significant differences in age and degree of myopia between the recurrence and no-recurrence groups (age: no-recurrence group versus recurrence group, 46.0 [20, 70] versus 36.5 [30, 69], p = .888; degree of myopia: no-recurrence group versus recurrence group, −1.6 [−9.0, 4.0] dioptres versus −0.5 [−9.0, 1.7] dioptres, p = .594).

Table 6.

Comparison of final visual function and optical coherence tomography findings with and without recurrence.

	Non recurrence group (10 eyes)		Recurrence group (34 eyes)
BCVA	Mean	1.19 ± 0.06	1.20 ± 0
	Median	1.20 (0.90, 1.20)	1.20 (1.20, 1.20)
	p value	p = .4378
MD value (dB)	Mean	−2.2 ± 2.1	−1.6 ± 2.7
	Median	−1.9 (−6.85, 1.8)	−1.5 (−5.9, 2.8)
	p value	p = .6908
cpRNFL thickness (μm)	Mean	78.8 ± 12.4	67.8 ± 9.8
	Median	81.3 (53.5, 98)	68.3 (51, 83)
	p value	p = .0231*
GCL thickness (μm)	Mean	66.8 ± 9.7	60.9 ± 5.3
	Median	67 (40, 86)	61 (55, 72)
	p value	p = .0412*

BCVA = Best corrected visual acuity; cpRNFL = Circumpapillary retinal nerve fibre layer; GCL = Ganglion cell layer; MD = Mean deviation.

Statistical analysis: Mann-Whitney U test; *Significant difference, p < .05.

Discussion

In the present study, we investigated the clinical characteristics and prognostic factors of MOGON. Most patients were middle-aged, and the male-to-female ratio was almost the same. Ocular pain during eye movement and optic disc swelling were present in approximately 90% of the cases. An epidemiological study of optic neuritis in Japan reported that the age of onset of MOGON was 47 [3–82] years, with a male-to-female ratio of 24:25. In 77% of the cases, ocular pain during eye movement was observed, while 76% exhibited optic disc swelling,³ which was, in general, similar to that found in our present study. Other MOGON⁹ and MOGAD¹⁰ studies have reported that the onset of the disease was more common in middle-aged subjects, there was no difference between men and women, and the pain observed during eye movement and the presence of optic disc swelling was more likely to be associated with the disease. At our institution, paediatric cases are referred to the paediatric department of the general hospitals, so we have not been able to assess paediatric cases. Regarding the characteristics of paediatric MOGON, Moon et al. reported that paediatric cases exhibit fewer symptoms of eye pain and optic perineuritis compared with adults.¹¹ Additionally, in the context of MOGAD, Sechi et al. reported a higher prevalence of acute disseminated encephalomyelitis and multiple central nervous system disorders.¹² Although the final diagnosis may be based on an antibody measurement, it is important to take these clinical features of MOGON into consideration in order to administer prompt treatment in these patients.

MRI lesions in MOGON tend to present as long lesions with three or more segments and are more common in the anterior part of the optic nerve. In an epidemiological study of optic neuritis in Japan, MOGON was associated with more anterior intraorbital lesions (44%) with a majority of the lesions involving more than half the length of the optic nerve (61%).³ Other studies have also shown that there are more intraorbital lesions and fewer optic chiasm or optic tract lesions,^9,13 which is consistent with the results of our study. The inflammation around the optic nerve sheath that was found in our present study was present in a relatively high number of patients (64.8%), which is in agreement with the results reported by Chen et al.⁹ The prevalence of eye pain and optic disc swelling, as well as the tendency to be associated with long lesions and peri-optic sheath inflammation on MRI, were consistent with the findings reported by Banwell et al.⁶ MOGON has been reported to be more common in the anterior part of the optic nerve, while optic neuritis caused by neuromyelitis optica spectrum disorders (NMOSD) is more common in the posterior part of optic nerve.³ Furthermore, in addition to the inflammation around the optic nerve sheath, these findings can be used to distinguish MOGON from other forms of optic neuritis.¹⁴

There was a decrease in the BCVA and swelling of the cpRNFL at the initial visit when the MRI lesion was longer than three segments. However, there were no significant differences in long-term visual function or cpRNFL thickness associated with the differences in the MRI lesion length. Akaishi et al. reported that in optic neuritis caused by NMOSD, differences in the optic nerve lesion length were associated with the final visual outcome.⁸ Denis et al. reported a long-term follow-up study that showed that thinning of the cpRNFL thickness occurred in idiopathic optic neuritis and optic neuritis caused by NMOSD.¹⁵ Compared with NMOSD, although MOGON has a longer MRI lesion length at the initial presentation, it is considered to have a better long-term visual prognosis.¹⁶ A comparison between MOGAD and NMOSD additionally showed that there were significantly worse results in the NMOSD group for the modified Rankin Scale and the gait disability sequelae at the final follow-up after myelitis.¹⁷ These factors suggest that, unlike NMOSD, MOGON may be less likely to lead to final axonal damage and less likely to cause residual visual disturbance, even if there is a long MRI lesion at the initial presentation.

In contrast, eyes that experienced recurrence had significantly lower cpRNFL thickness at the last visit compared with that for the no-recurrence eyes. The results are consistent with previous reports of relapse-dependent progression of neuropathy in MOGAD.¹⁸ However, there were no significant differences observed in our present study for the final visual function between patients with and without relapse. According to Matsumoto et al., in cases of anti-AQP4 antibody-positive individuals, GCA was believed to involve not only morphological thinning but also functional impairment.¹⁹ In the case of MOGON, functional impairment does not persist, and the thinning is primarily attributed to morphological remodelling. However, Matsuda et al. reported that MOGON patients had significantly more residual visual field defects compared with that found for antibody-negative patients.²⁰ In contrast to the results of our present study, Zhao et al. reported a worse final BCVA (20/30).¹³ In our present study, 10/44 eyes (6/31 patients) experienced recurrence, whereas Matsuda et al. reported 11/18 cases (including five cases with four or more recurrences), while Zhao et al. reported 10/20 cases (six cases with three or more recurrences), which seemed that there were more relapses compared with our study. It is important to avoid recurrence as much as possible, as cpRNFL thinning is more likely to be observed during recurrence. Moreover, in severe cases with repeated recurrence, there is a possibility that the visual dysfunction may eventually become permanent.

There were several limitations for this study. First, because the study was conducted in a single ophthalmic hospital, this may have resulted in the observation of less severe cases due to the difficulty in providing treatments such as plasma exchange therapy. The bias in the disease severity may have resulted in the relatively favourable visual function prognosis that was observed. Second, the retrospective study design needs to be considered when analysing the results, including those related to the different tests, follow-ups, and treatments. In the present study, complete data on the visual acuity, visual fields, OCT, and MRI were not available with the same measurement condition for all the patients. In addition, the follow-up periods and maintenance therapy were not standardised, with these differences potentially influencing the results of the visual function and OCT findings. Third, the present study did not impose limits on spherical power. It has been reported that high myopia can impact OCT findings.²¹ In this study, out of the 44 eyes examined, 4 had more than 6 dioptres of high myopia, and no significant differences in myopia were observed in OCT findings between the two groups. For a more detailed analysis, it would be necessary to include eyes without high myopia. Fourth, in most cases, maintenance therapy was often initiated at a dose of 0.5 mg/kg/day of prednisolone and gradually reduced over approximately 6 months. However, cases with minimal visual impairment or those with a longer time elapsed since the initial onset did not require maintenance therapy or had a more rapid reduction. The assessment of disease activity and the approach to reducing prednisolone dosage depended on the attending physician’s judgement, which represents a limitation of this study. Prospective randomised controlled trials with a larger number of patients will need to be undertaken in order to better understand the clinical features of MOGON.

Our study investigated the clinical features and prognostic factors of MOGON. As the recurrences can cause thinning of the cpRNFL, it is important to reduce the recurrences as much as possible.

Funding Statement

The authors reported there is no funding associated with the work featured in this article.

Disclosure statement

Toshiyuki, T received research grants from Cosmic Corporation Co Ltd. and Medical & Biological Laboratories CO Ltd.

Data availability statement

All data generated or analysed during this study are included in this article. Further enquiries can be directed to the corresponding author.

Statement of ethics

The research has been complied with all the relevant national regulations, and institutional policies and in accordance with the tenets of the Helsinki Declaration and has been approved by the Inouye Eye Hospital’s institutional review board (202307–5). Informed consent was obtained as an opt-out on the website, allowing the participants to refuse participation in the study at any time.