Clinical characteristics, management and visual outcomes of traumatic optic neuropathy: a five-year retrospective study from a tertiary referral centre

Abstract

Background

Traumatic optic neuropathy (TON) is a potentially blinding complication of craniofacial trauma with an uncertain prognosis and no established standard of care. Evidence on outcomes and prognostic factors remains limited, particularly in Southeast Asian populations.

Methods

A retrospective observational study was conducted at a tertiary referral centre in Malaysia between September 2020 and August 2025. Patients diagnosed with TON were identified through medical record review. Visual acuity (VA) was categorised according to the World Health Organization (WHO) visual impairment classification. Visual outcomes at one-year follow-up were analysed using the Wilcoxon signed-rank test, and factors associated with visual improvement were evaluated using Fisher’s exact test.

Results

Fifty-eight patients were identified, of whom 54 had complete data for visual outcome analysis. The median age was 23.5 years, and most patients were male (82.8%). Road traffic accidents were the predominant aetiology, accounting for 86.2% of TON cases. TON occurred across all traumatic brain injury severities, with comparable distribution in mild and severe cases. At presentation, 53.8% had severe visual impairment, including 13.8% with no perception of light. There was a statistically significant overall improvement in visual acuity at one-year follow-up compared with baseline (p = 0.003). However, subgroup analysis showed no significant association between treatment modality and visual improvement. Visual improvement was also not significantly associated with age, baseline severity, timing of corticosteroid initiation, or the presence of orbital or skull fractures and intracranial haemorrhage (all p > 0.05).

Conclusion

Visual outcomes in TON are variable and unpredictable, highlighting the importance of early neuro-ophthalmic assessment, close visual monitoring, and appropriate counselling regarding the guarded visual prognosis. Due to the relatively small sample size of this cohort, these findings may not be generalisable to broader populations.

Background

Traumatic Optic Neuropathy (TON) is an uncommon but potentially devastating consequence of craniofacial trauma, particularly in the setting of traumatic brain injury (TBI), with an estimated global incidence ranging from 0.7 to 2.5% [1]. It most frequently occurs following blunt head injury, particularly in association with road traffic accidents (RTA), falls and assaults, and predominantly involves young adults [2–4]. Clinically, TON is characterised by acute optic nerve dysfunction resulting in partial or complete visual loss. Patients typically present with sudden reduction in visual acuity (VA), visual field defects, impaired colour vision, and a relative afferent pupillary defect, often in the absence of significant anterior or posterior segment ocular injuries [4, 5].

The pathophysiology of TON is complex and multifactorial. Primary injury involves mechanical damage to retinal ganglion cell (RGC) axons and their vascular supply through either direct or indirect mechanisms [5]. Direct injury may occur due to optic canal fractures during trauma, where displaced bone fragments compress or lacerate the optic nerve [6, 7]. In contrast, indirect TON, which is more common, results from the transmission of concussive forces from blunt head trauma to the optic canal, leading to shearing injury [4–6]. Secondary mechanisms, include optic nerve oedema, inflammation and ischaemia within the confined optic canal, further exacerbating neuronal damage through compartment-like effect [5].

Early therapeutic intervention for TON is primarily aimed at limiting secondary injury [5]. Treatment strategies have focused on reducing optic nerve oedema and improving microvascular perfusion, most notably through high-dose corticosteroids and optic nerve decompression (OND). Despite this rationale, the optimal management of TON remains controversial. The International Optic Nerve Trauma Study (IONTS) did not demonstrate a significant difference in visual recovery among patients treated with high-dose corticosteroids, OND surgery, combined therapy, or conservative observation [7]. Furthermore, concerns regarding increased mortality associated with high-dose corticosteroids in patients with concurrent TBI have been raised following the Corticosteroid Randomisation After Significant Head Injury (CRASH) trial [8]. Consequently, no universally accepted standard of care exists, and treatment decisions are often individualised.

Given the variability in clinical presentation and the lack of reliable prognostic indicators, further real-world data are needed. This study aimed to evaluate the demographic and clinical characteristics, treatment approaches, and visual outcomes of TON in a tertiary referral centre in Malaysia.

Methods

Study design

This retrospective study was conducted at Hospital Pakar Universiti Sains Malaysia, a tertiary referral academic teaching hospital in Kelantan, Malaysia, with established multidisciplinary services including neurosurgery, neurology, neuroradiology, neurorehabilitation, and neuro-ophthalmology. Patients diagnosed with TON between September 2020 and August 2025 were identified through the hospital database, and their medical records were reviewed. The diagnosis of TON was established based on the presence of relative afferent pupillary defect (RAPD) in unilateral cases, along with clinical features of optic nerve dysfunction following ocular or craniofacial trauma, including reduced VA, impaired colour vision, visual field defects, or reduced red saturation. Treatment decisions, including corticosteroid administration, were based on the clinical diagnosis of TON and were not dependent on the availability of assessable visual acuity measurements.

All eligible cases diagnosed by an ophthalmologist during the study period were included using a total population sampling approach. Patients with significant associated ocular injuries that could independently impair vision, including open globe injury, hyphaema, vitreous haemorrhage and traumatic retinal detachment were excluded. Patients with pre-existing ocular comorbidities affecting visual function, such as significant cataract, retinopathies or glaucoma were also excluded.

Data collection

Data collected included demographic characteristics, mechanism of injury, treatment modality, time from injury to steroid initiation, associated ophthalmological finding and visual field patterns. Computed tomography (CT) scans of the brain and orbit were reviewed to document orbital and skull fractures as well as associated intracranial injuries.

VA at first ophthalmology review and at one-year follow-up was recorded. In medically unstable or intubated patients where VA assessment was not performed, it was recorded as not testable. VA was categorised according to the World Health Organization (WHO) visual impairment classification into mild (6/6–6/18), moderate (< 6/18–6/60), severe (< 6/60–3/60), and blind (< 3/60 to no perception of light, NPL).

Visual outcome was assessed by comparing baseline VA with VA at one-year follow-up. Visual improvement was defined as a gain of two or more lines on the Snellen acuity chart, improvement in the WHO visual impairment category, or improvement from NPL or perception of light (PL) to a measurable level of vision such as hand movements (HM), counting fingers (CF), or better. All data were collected using a standardized online google form.

Statistical analysis

Descriptive analyses were used to summarize demographic and clinical data. Continuous variables were presented as median with range, while categorical variables were expressed as frequencies and percentages. Visual outcome was analysed using the Wilcoxon signed-rank test. Factors associated with visual outcome were evaluated using Fisher’s exact test. Statistical significance was set at p < 0.05 (two-tailed).

Results

Demographic and clinical characteristics

A total of 58 patients met the inclusion criteria, and their demographic characteristics, clinical findings, and treatment patterns were analysed. However, four patients were excluded from paired visual outcome analysis due to incomplete VA documentation, leaving 54 patients for comparison of baseline and one-year visual acuity outcomes.

The median age was 23.5 years (range 4–79 years), and most patients were male (82.8%). RTA were the predominant cause of injury (86.2%). All patients were of Malay ethnicity, reflecting the demographic composition of the study area. Most patients were referred to the ophthalmology service within 24 h of injury (67.24%) and started on intravenous methylprednisolone (IVMP) therapy within first 72 h following trauma. These demographic data are presented in Table 1.

Table 1.

Demographic profile of patients with TON

Variable	Frequency, n	Percentage, %
Age
< 18	18	31.03
18–40	28	48.28
> 40	12	20.69
Sex
Male	48	82.76
Female	10	17.24
Race
Malay	58	100
Aetiology
RTA	50	86.21
Fall	5	8.62
Assaulted	1	1.73
Blast injury (Firecrackers)	2	3.44
Time of referral from injury
< 24 h	39	67.24
24–72 h	14	24.14
> 72 h	5	8.62
Time from diagnosis to treatment
< 24 h	10	17.24
24–72 h	38	65.52
> 72 h	10	17.24

Abbreviation: RTA, road-traffic accident

At presentation, more than half of the patients had severe visual impairment (VA worse than 6/60), including eight patients (13.8%) with NPL vision. Among these eight patients, five had mild TBI and their vision remained unchanged at one-year follow-up. The proportion of patients with good vision (6/6–6/18) increased from 29.3% to 44.8%, while the number of patients in the blind category decreased from 28 to 17. Overall, a significant improvement in visual impairment category was observed at one year compared with baseline (Wilcoxon signed-rank test, p = 0.003) (Table 2).

Table 2.

Clinical characteristics in TON

Variable	Frequency, n	Percentage, %
Visual Acuity
At Presentation
6/6–6/18	17	29.31
< 6/18–6/60	7	12.07
< 6/60–3/60	2	3.45
< 3/60 – PL	20	34.48
NPL	8	13.79
Not testable*	4	6.90
At one-year follow-up
6/6–6/18	26	44.83
< 6/18–6/60	6	10.34
< 6/60–3/60	5	8.62
< 3/60 – PL	6	10.34
NPL	11	18.97
Not testable*	4	6.90
Visual Field at Diagnosis
Normal	10	17.24
Abnormal	12	20.69
Not testable*	36	62.07
Colour vision
Normal	11	18.97
Abnormal	29	50.00
Not testable**	18	31.03
Associated Ocular Injury
Traumatic mydriasis	17	29.31
Periorbital haematoma	46	79.31
Subconjunctival haemorrhage	31	53.45
Retrobulbar haemorrhage	3	5.17
Berlin oedema	2	3.44
Cranial nerve palsies	11
3rd	4	6.90
4th	3	5.17
6th	4	6.90
Conjunctival laceration	1	1.73
Iridodialysis	1	1.73
Corneal abrasion	3	5.17
Traumatic Brain Injury (Severity)
None	4	6.90
Mild	23	39.66
Moderate	6	10.34
Severe	25	43.10
Treatment
Corticosteroid (IVMP)	44	75.86
3 days	24	54.55
5 days	20	45.45
IVMP alone	10	22.73
IVMP + oral steroid continuation	31	70.45
IVMP + OND	3	6.82
Conservative***	14	24.14

Visual acuity at presentation and one-year follow-up (Wilcoxon signed-rank test, p = 0.003)Abbreviations: IVMP, intravenous methylprednisolone; NPL, no perception of light; PL, perception of light; OND, optic nerve decompression

*Not testable due to intubation, medical instability, or poor Glasgow Coma Scale (GCS)

**Not testable due to intubation, medical instability, or profoundly reduced vision (PL/NPL)

***Includes mecobalamin therapy (n = 5) and no medical treatment (n = 9)

Objective visual field assessment using Humphrey perimetry could only be performed in a limited proportion of patients due to the acute trauma setting. Among those tested, abnormalities included monocular hemianopic defects, homonymous hemianopia, arcuate scotoma, superior or inferior field defects, and central scotoma. Associated ocular findings included periorbital haematoma, subconjunctival haemorrhage, traumatic mydriasis, and gaze restriction secondary to third, fourth or sixth cranial nerve palsies. These findings are summarised in Table 2.

The severity of associated TBI varied among patients, with severe TBI being the most common (43.1%), followed by mild TBI (39.7%). Radiographic evaluation showed that orbital or skull fractures and intracranial haemorrhage were present in more than 80% of patients. Among patients with orbital fractures, the orbital floor and zygomatic complex were the most commonly involved structures, while skull fractures most frequently involved the frontal and temporal bones. Intracranial haemorrhage particularly subdural and extradural haemorrhage was commonly observed. Detailed findings are summarised in Table 3.

Table 3.

Radiographic findings in TON

Radiographic findings	Frequency, n	Percentage, %
Fracture pattern ( n  = 58)
Orbital only	17	29.31
Skull only	6	10.34
Orbital + Skull	24	41.38
No fracture	11	18.97
Orbital Wall fractures ( n  = 41)
Roof	11	26.83
Floor	28	68.29
Medial	13	31.71
Lateral	26	63.41
Zygomatic Complex	27	65.85
Posterior	3	7.32
Skull Bone Fractures ( n  = 30)
Basal Skull	11	36.67
Frontal	19	63.33
Parietal	7	23.33
Temporal	14	46.67
Occipital	2	6.67
Intracranial injury* ( n  = 58)
SAH	14	24.14
SDH	19	32.76
EDH	19	32.76
Cerebral contusion	4	6.90
Direct CCF	1	1.73
Others**	3	5.17
None	4	6.89

Abbreviation: SAH, subarachnoid haemorrhage; SDH, subdural haemorrhage; EDH, extradural haemorrhage; CCF, carotid cavernous fistula

*Patients may have more than one intracranial injury

**Other intracranial injuries include interventricular or interhemispheric bleed and cerebral oedema

Treatment and factors associated with visual improvement

A total of 44 patients received intravenous methylprednisolone (IVMP) 1 g/day administered in four divided doses for 3–5 days, including three patients who subsequently underwent OND. Visual improvement was observed in 14 patients (33.3%). Fourteen patients were managed conservatively. Five patients received adjunctive mecobalamin (500 mcg TDS for 3 months), of whom two showed visual improvement (Table 4).

Table 4.

Comparison of visual outcomes across treatment groups

Group	Visual Improvement, n (%)	No Improvement, n (%)	Unable to assess, n (%)
Corticosteroid-based therapy
Corticosteroid* (n = 41)	14 (34.15)	26 (63.41)	1 (2.44)
Corticosteroid + OND (n = 3)	0 (0.00)	2 (66.67)	1 (33.33)
Conservative management
Observation (n = 9)	1 (11.11)	7 (77.78)	1 (11.11)
Mecobalamin therapy (n = 5)	2 (40.00)	2 (40.00)	1 (20.00)

Abbreviation: IVMP, intravenous methylprednisolone; OND, optic nerve decompression

*Corticosteroid group includes IVMP, with or without continuation oral prednisolone

Factors associated with visual improvement were further analysed (Table 5). Early initiation of steroid (< 72 h from injury) was not significantly associated with visual improvement compared with delayed treatment (p = 0.74). Age group, baseline VA, presence of orbital or skull fracture, and intracranial haemorrhage were also not significantly associated with visual improvement (all p > 0.05).

Table 5.

Factors associated with visual outcome

Variable	Visual Improvement, n (%)	No Improvement, n (%)	p-valuea
Age (n = 54)			0.489
≤ 40	14 (32.6)	29 (67.4)
> 40	5 (45.5)	6 (54.5)
Baseline visual impairment (n = 54)			0.252
Mild/Moderate	6 (25.0)	18 (75.0)
Severe/Blind	13 (43.3)	17 (56.7)
Treatment (n = 54)			0.732
Steroid	14 (33.33)	28 (66.67)
Non steroid	3 (25.0)	9 (75.0)
Time to steroid initiation (n = 42)*			0.740
< 72 h	12 (40.0)	18 (60.0)
> 72 h	4 (33.3)	8 (66.7)
Orbital/Skull fractures (n = 54)			0.727
Yes	15 (34.1)	29 (65.9)
No	4 (40.0)	6 (60.0)
Intracranial haemorrhage (n = 54)			1.000
Yes	14 (35.9)	25 (64.1)
No	5 (33.3)	10 (66.7)

Four patients were excluded from the visual outcome analysis because both baseline and follow-up visual acuity could not be reliably assessed (two corticosteroid-treated and two conservatively managed patients)

*Time-to-steroid initiation analysis was performed only among corticosteroid-treated patients with assessable visual outcomes (n = 42)

^a Fisher’s exact test: p < 0.05, significant

Discussion

TON continues to pose a significant clinical challenge due to its heterogeneous presentation, uncertain prognosis, and lack of consensus regarding optimal management. In this five-year cohort, the predominance of young male patients and RTA as the primary mechanism of injury mirrors global epidemiological patterns, reflecting higher exposure to high-impact trauma in this demographic [4, 5, 9–12]. Falls, assault and blast injuries were also observed and similarly involved young males exposed to occupational and environmental hazards. As these individuals are in their most economically productive years, TON may carry significant socioeconomic implications and adversely affect quality of life [13, 14]. This underscores the broader public health importance of increasing awareness and implementing preventive strategies, including road safety enforcement and early ophthalmic screening in patients involved in RTA.

Most of our patients presented with severe visual impairment, reflecting the high-impact mechanisms of injury typically associated with TON. Notably, TON occurred across the spectrum of TBI severity, with a comparable distribution between mild and severe TBI, suggesting that the severity of optic nerve injury does not necessarily correlate with overall severity of head trauma [9]. A key finding of this study is the significant overall improvement in VA at one year, indicating that meaningful visual recovery is possible even in patients with poor baseline vision. This observation is consistent with previous reports that spontaneous visual recovery may occur in up to half of TON cases [15]. However, the extent of recovery remained variable, with a substantial proportion of patients continuing to have poor visual outcomes, reinforcing the generally guarded prognosis associated with TON [10].

The lack of demonstrable benefit of corticosteroid therapy in this cohort is consistent with existing literature, including the IONTS and subsequent systematic reviews [5, 7, 16–19]. Although significant improvement in visual outcomes with corticosteroid therapy has been reported [20], this finding remains inconsistent. In this cohort, neither the use of high-dose corticosteroids nor the timing of initiation within 72 h showed a significant association with visual improvement, in keeping with previous reports [5, 17–19]. It should be noted that treatment timing was analysed relative to the time of TON diagnosis rather than the time of injury. In routine clinical practice, TON diagnosis may be delayed by factors such as delayed ophthalmology referral, particularly in patients with polytrauma requiring neuroimaging, multidisciplinary evaluation, and urgent neurosurgical intervention, limiting reliable evaluation of ultra-early treatment windows (< 8 or < 24 h from injury). This may reflect the occurrence of irreversible primary injury at the time of trauma, thereby limiting the potential for any pharmacological or therapeutic intervention to alter visual outcomes.

Similarly, the role of OND remains uncertain. OND is considered a minimally invasive surgery done endoscopically aimed at relieving optic nerve compression within the optic canal [6]. The small number of patients who underwent endoscopic OND in our cohort precludes meaningful conclusions regarding its efficacy. While traditionally considered in cases refractory to corticosteroid therapy or with radiological evidence of optic canal fracture, its early use has also been extended to selected cases without clear radiographic compression [21]. Emerging evidence suggests that early OND, performed promptly after presentation, may yield better outcomes compared with a delayed approach following initial corticosteroid therapy [21]. Likewise, the limited number of patients who received adjunctive mecobalamin therapy precludes meaningful evaluation of its therapeutic effect. Although visual improvement was observed in two of the five treated patients, the retrospective nature of this observation and the very small sample size prevent any inference regarding treatment efficacy.

Recent studies have identified several potential prognostic factors in TON, with poor initial VA consistently emerging as the strongest negative predictor [22]. Other factors, including higher RAPD grades, optic canal fractures, associated intracranial injuries, lack of improvement after 48 h of early high-dose corticosteroid treatment and a gradual pattern of VA loss, have also been linked to poorer outcomes [23]. However, in our cohort, neither imaging findings nor treatment-related variables were significantly associated with visual improvement, highlighting the limited predictive value of current clinical and radiological parameters. Similarly, although a variety of visual field abnormalities were observed, no consistent relationship was identified between specific visual field patterns and the location of orbital fractures, skull fractures, or intracranial haemorrhage. From an ophthalmic perspective, awareness of TON, early ophthalmic referral, and thorough neuro-ophthalmic assessment in patients with head trauma, even in the absence of obvious ocular injury, remain essential.

This study has several limitations. Its retrospective design is inherently subject to incomplete or inconsistent documentation. The relatively small sample size, particularly within treatment subgroups, limits statistical power and precludes multivariable analysis. In addition, the acute trauma setting restricted comprehensive ophthalmic assessments, including visual field testing, in a proportion of patients. Despite these limitations, this study provides valuable real-world data on TON from a Southeast Asian population, contributing to the existing body of evidence on TON. Future prospective, multicentre studies with standardised diagnostic criteria and treatment protocols are needed to better define prognostic factors and guide evidence-based management. Incorporation of advanced imaging modalities and functional assessments may also improve understanding of disease mechanisms and recovery patterns.

Conclusion

Visual recovery in TON remains variable and unpredictable, and no clear prognostic factors or superior treatment modality were identified in this cohort. Due to the relatively small sample size, these findings may not be generalisable to broader populations. Management of TON should therefore emphasise early recognition, prompt neuro-ophthalmic assessment, close visual monitoring, and careful selection between corticosteroid therapy, surgical decompression, or conservative management based on the individual patient’s clinical condition and associated risks. Appropriate counselling regarding the guarded visual prognosis and uncertainty of treatment outcomes is essential. Further well-designed prospective studies with larger cohorts are needed to better define prognostic indicators and establish evidence-based management strategies for TON.

Abbreviations

TON

Traumatic optic neuropathy

TBI

Traumatic brain injury

RTA

Road-traffic accident

VA

Visual acuity

RGC

Retinal ganglion cells

OND

Optic nerve decompression

IONTS

International Optic Nerve Trauma Study

RAPD

Relative afferent pupillary defect

CT

Computed tomography

WHO

World Health Organization

PL

Perception of light

NPL

No perception of light

CF

Counting fingers

HM

Hand movement

IVMP

Intravenous methylprednisolone

Author contributions

S. A. M. conceived and designed the study, supervised the work, performed data analysis, and contributed to manuscript drafting and critical revision. M. S. Y. performed data analysis and drafted the manuscript. S. S. N. A. collected data, contributed to data analysis and manuscript drafting. A. M. R. contributed to manuscript drafting and critical revision. W. H. W. H. contributed to critical revision of the manuscript. All authors have read and approved the final manuscript.

Funding

None.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request, subject to institutional and ethical approval.

Declarations

Ethical approval and consent to participate

This study was conducted in accordance with the principles of the Declaration of Helsinki. Ethics exemption for this retrospective analysis was granted by the Universiti Sains Malaysia Human Research Ethics Committee (USM/JEPeM/KK/26050418). The requirement for informed consent was waived due to the retrospective nature of the study and the use of anonymized data.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.