Traumatic Optic Neuropathy 

Updated: Sep 13, 2018
Author: Ryan S Jackson, MD; Chief Editor: Arlen D Meyers, MD, MBA 



Traumatic optic neuropathy (TON) refers to an acute injury of the optic nerve secondary to trauma. The optic nerve axons may be damaged either directly or indirectly and the visual loss may be partial or complete. An indirect injury to the optic nerve typically occurs from the transmission of forces to the optic canal from blunt head trauma. This is in contrast to direct TON, which results from an anatomical disruption of the optic nerve fibers from penetrating orbital trauma, bone fragments within the optic canal, or nerve sheath hematomas.

The image below depicts traumatic optic neuropathy.

Plain film radiograph of a 2-year-old girl followi Plain film radiograph of a 2-year-old girl following blunt periocular trauma. Although a left orbital wall fracture is not evident, a loculated pocket of intraorbital air is highlighted by the arrow.


Patients with traumatic optic neuropathy (TON) can present with a variable degree of vision loss (decreased visual acuity, visual field abnormalities, or loss of color vision). Most cases (up to 60%) present with severe vision loss of light perception (LP) or worse. In the acute phase, the optic nerve usually appears normal on funduscopic examination, but optic nerve atrophy is often seen 3-6 weeks after the injury.



The incidence of traumatic optic neuropathy (TON) in the setting of a closed traumatic head injury in various studies ranges from 0.5-5%. Using the National Trauma Data Bank, Gise et al found TON to be the most common (86.1%) pediatric visual pathway injury in the United States between 2008 and 2014.[1]

The vast majority of TON cases are seen in males (up to 85%), with a mean age of 34 years. Motor vehicle and bicycle accidents account for the majority of causes, followed by falls and assaults. TON has also been associated with penetrating orbital trauma (eg, stab wounds, pellet and gunshot wounds, foreign bodies) and recreational sports (eg, paint ball injury).


The pathophysiology of traumatic optic neuropathy (TON) is thought to be multifactorial, and some researchers have also postulated a primary and secondary mechanism of injury. In indirect TON cases, the injury to the axons is thought to be induced by shearing forces that are transmitted to the fibers or to the vascular supply of the nerve. Studies have shown that forces applied to the frontal bone and malar eminences are transferred and concentrated in the area near the optic canal. The tight adherence of the optic nerve’s dural sheath to the periosteum within the optic canal is also thought to contribute to this segment of the nerve being extremely susceptible to the deformative stresses of the skull bones. Such injury leads to ischemic injury to the retinal ganglion cells within the optic canal.

A secondary mechanism can result in optic nerve swelling after the occurrence of the acute injury. The optic nerve swelling can exacerbate retinal ganglion cell degeneration by further compromising the vascular blood supply, either through a rise in intraluminal pressure or reactive vasospasm. These secondary mechanisms, in theory, form the rationale for optic canal decompression via medical (ie, steroids) or surgical means (ie, bony decompression). Finally, the intracranial segment of the optic nerve may be damaged by forces delivered to the axons by the shifting of the brain following head trauma. With this mechanism, the nerve fibers may be injured against the falciform dural fold or through a shearing force where the nerve becomes fixed as it enters the intracranial opening of the optic foramen.

Relevant Anatomy

The human orbit is composed of seven bones that comprise 4 walls: the roof, lateral wall, medial wall, and floor. The 7 bones include: ethmoid, frontal, lacrimal, maxillary, palatine, sphenoid (lesser and greater wings), and zygomatic bones. The 4 orbital walls converge posteriorly at the apex of the orbit, and the optic canal is located within the lesser wing of the sphenoid, which is one of the bones that comprise the orbital roof.

The optic canal is separated from the superior orbital fissure by the optic strut of the sphenoid bone; hence the optic strut comprises the optic canal's lateral wall. The canal's medial wall separates it from the sphenoid sinus. Structures that pass through the optic canal include the optic nerve sheath and axons, their supportive glia, the ophthalmic artery, and branches of the carotid sympathetic plexus of the autonomic nervous system.

The optic nerve is a cable of retinal ganglion cells that enters the eye approximately 4 mm nasally and 0.8 mm superiorly from the center of the macula at the posterior pole. Inside the eyeball, the anterior optic nerve forms the optic disc, which is visible clinically with funduscopic examination as a pink, slightly elevated structure centered upon the opening in the posterior sclera.

The optic nerve is a collection of axons (approximately 1.2 million) originating from the retinal ganglion cells. With its meningeal sheath, the optic nerve is 3-4 mm in diameter and its total length measures approximately 50 mm from the globe to the optic chiasm. The nerve is composed of intraocular (about 1 mm), intraorbital (20-30 mm), intracanalicular (5-11 mm), and intracranial (3-16 mm) segments. The axons transmit visual stimuli from the retina to the eight primary visual nuclei in the lateral geniculate nucleus.

The topographic organization of the axons, as arranged by the retina, is relatively preserved within the optic nerve. In addition to the visual input that it carries to the lateral geniculate nucleus, the optic nerve also carries afferent fibers that participate in the pupillary responses. Except for its intraocular segment, the axons of the optic nerve are myelinated. Similar to the cerebral white matter, the optic nerve is ensheathed by oligodendrocytes, which are responsible for the production of myelin. Astrocytes are also present within the glial septa, and they provide nutritional support to the axons.

Throughout its intraorbital course, the optic nerve remains surrounded by pia, arachnoid, and dura mater, which move along with the optic nerve during normal eye movements. Within the optic canal, the sheath of the optic nerve is fused to the sphenoid periosteum, and, as a result, the nerve and its sheath are tightly fixed to the bony canal within this confined space. At the posterior foramen of the optic canal, the optic nerve sheath fuses with an overlying falciform fold of dura that lines the calvaria.

Pial branches of the internal carotid, anterior cerebral, and anterior communicating arteries perfuse the intracranial optic nerve. Small pial branches from the ophthalmic artery supply the intracanalicular optic nerve. The intraorbital optic nerve is also supplied by perforating branches derived from the ophthalmic artery. The arterial circle of Zinn-Haller supplies the intraocular optic nerve with contributions from the posterior ciliary arteries, the pial arterial network, and the peripapillary choroidal vasculature.




Patients with traumatic optic neuropathy (TON) present with vision loss after blunt or penetrating trauma. Symptoms consist of acute unilateral decrease in vision or visual fields. Delayed diagnosis may occur secondary to other injuries related to the trauma, such as a traumatic brain injury. Such injuries would delay the presentation, and thus the evaluation, of any ophthalmologic injury. 

Physical Examination

Physical examination consists of a full ophthalmologic evaluation, including for obvious signs of trauma, such as ecchymosis, hematoma, lacerations, and bony step-offs. In addition, neurologic examination evaluating for decreased visual acuity and afferent pupillary defect should be carried out. Lastly, funduscopic examination should be performed; initial funduscopic assessment may reveal a normal optic nerve appearance, since optic atrophy may not be evident at the time of presentation.



Differential Diagnoses



Imaging Studies

Neuroimaging studies (computed tomography [CT] scanning or magnetic resonance imaging [MRI]) are an important part of the assessment when traumatic optic neuropathy (TON) is suspected. In the post-trauma setting, CT scanning is the preferred modality for demonstrating the presence of an optic canal fracture, a displaced bony fragment impinging upon the optic nerve, a metallic foreign body in the orbit, orbital emphysema, or an optic nerve sheath hematoma. A brain and orbit MRI may be useful in certain settings to delineate the extent of hemorrhage involving the neurovascular structures at the orbital apex or to rule out inflammatory or infiltrative causes for an optic neuropathy. The vast majority of patients with TON suffer an indirect injury to the optic nerve within the optic canal, and neuroimaging studies typically demonstrate no abnormalities of the anterior visual pathways, although a fracture in the region of the optic canal may be seen.

A retrospective study by Bodanapally et al indicated that, using diffusion-weighted imaging (DWI), hyperintensity of the optic nerve resulting from diffusion restriction can aid in the diagnosis of TON.[2]

A retrospective study by Reddy et al indicated that in patients with TON, CT scan findings of intraconal hematoma and hematoma along the optic nerve are risk factors for poor visual acuity at hospital admission.[3]

Other Tests

See the list below:

  • Visual field perimetry

    • Automated perimetry can be obtained only in patients who retain adequate vision/acuity. Patients with poor visual acuity (worse than 20/200) may be assessed with Goldmann perimetry or with confrontational visual field testing.

    • No visual field loss pattern is pathognomonic for traumatic optic neuropathy (TON), although a dense central scotoma is characteristic. Recovery of optic nerve function may be documented via serial visual field testing.

  • Visual evoked potential (VEP)

    • VEP can be helpful to document the presence of TON in unresponsive patients or in cases with concomitant ocular injuries. Patients can also be followed with serial VEP examinations to document recovery of function when clinical parameters are equivocal.

    • VEP is not considered essential in making the diagnosis of TON, and logistically, the ability to perform a neurophysiological evaluation is often hindered by the inability to transport the trauma patient to the neurophysiology laboratory.

    • In unilateral cases of TON, a flash VEP amplitude ratio (affected side/normal side) greater than 0.5 appears predictive of a favorable, long-term visual outcome. Visual recovery is considered unlikely when VEP amplitudes are nonrecordable.

  • Retinal nerve fiber layer (NFL) imaging: Scanning laser polarimetry and optical coherence tomography can be used to assess and monitor retinal NFL axonal loss during the period of follow-up.

Diagnostic Procedures

Traumatic optic neuropathy (TON) is based on a clinical diagnosis of optic nerve dysfunction supported by a recent history of trauma to the head. In an acute setting following trauma, the diagnosis may be delayed if the patient is unconscious and a formal visual acuity assessment cannot be performed. In a conscious, cooperative patient, the diagnosis of TON should be verified by testing the patient for an abnormal visual acuity, an ipsilateral afferent pupillary defect (APD), impairment of color vision, and a visual field defect on formal perimetry.

Visual acuity

Even in an acute trauma setting, patients should have a visual acuity assessment as soon as possible. This can be performed with a Snellen eye chart (which measures distance vision) or a near vision card. If the patient cannot read the top letter on the eye chart, the visual acuity may be recorded with the following nomenclature (in order of decreasing visual acuity): counting fingers vision, hand motion perception, light perception (LP), or no light perception (NLP).

Pupil examination

The pupil examination assesses the size of both pupils, the response of the pupils to light and near stimulation, and an evaluation for a relative afferent pupillary defect or APD (swinging-flashlight test). The swinging-flashlight test compares the pupillary reaction to light between the two eyes; normally, both pupils should constrict equally to light, and the constriction should be maintained as the light is rapidly switched between the two pupils. An eye with a unilateral optic nerve injury will demonstrate an APD, verifying the presence of TON. In the rare case of a bilateral TON, a relative APD may not be seen if the injury is symmetric between the 2 sides, and both pupils may be dilated and nonreactive to light if the injury is profound.

Funduscopic examination

The funduscopic examination can be performed with the use of a direct ophthalmoscope, indirect ophthalmoscope, or the slit-lamp biomicroscope. Because the location of the injury in most TON cases is within the posterior orbit or the optic canal, the optic disc typically appears normal on funduscopic examination on initial diagnosis. Optic nerve atrophy usually appears 3-4 weeks after the traumatic event, and the disc acquires a diffuse pallor. Rarely, optic nerve changes can be seen with direct injuries to the retrobulbar section of the optic nerve, presenting as an avulsed optic nerve head or optic disc swelling with surrounding hemorrhage.

Histologic Findings

Cadaver studies have not demonstrated consistent histopathologic findings in traumatic optic neuropathy, suggesting multiple mechanisms of injury in patients with TON. Gross pathologic changes that have been reported include hematomas within the optic nerve sheath and occasionally areas of visible necrosis. Microscopic findings include interstitial hemorrhage, fibrosis of the pial septa, and a chronic inflammatory infiltrate by lymphocytes, plasma cells, and iron-laden macrophages. One case demonstrated degeneration of axons with loss of myelin in a triangular section consistent with damage to the penetrating vessels supplying the infracted region.

With time, loss of astrocyte support appears to occur within the optic nerve in TON injuries, followed by replacement of the axons with microglia. Based on the histopathologic findings in TON, axonal loss appears to occur either through shearing forces, hemorrhagic infarction, direct compression of the nerve fibers from an extrinsic nerve sheath hematoma, or a combination of these mechanisms.



Medical Therapy

The main treatment options for traumatic optic neuropathy (TON) include systemic corticosteroids and surgical optic nerve decompression, either alone or in combination. Review and analysis of the literature are complicated by the variety of therapeutic approaches and a lack of randomized, controlled studies on the use of these modalities for TON.[4]

The current knowledge base on the use of steroids for TON is based on small retrospective studies, anecdotal reports, and extrapolation from national traumatic brain and spinal cord injury studies. Steroid therapy for TON can be categorized as follows: moderate dose (60-100mg of oral prednisolone), high dose (1 gram of intravenous methylprednisolone/day), or mega dose (30 mg/kg loading dose of intravenous methylprednisone, followed by 5.4 mg/kg/h for 24 hours).

Steroids have been used in TON since the early 1980s because of their perceived benefits in various animal models of central nervous system injury. Steroids were thought to provide neuroprotection in traumatic central nervous system injury through their antioxidant properties and inhibition of free radical-induced lipid peroxidation.

In 1990, Bracken and colleagues published their findings on the use of mega dose corticosteroid therapy in the National Acute Spinal Cord Injury Study 2 (NASCIS 2).[5] The NASCIS 2 was a multicenter clinical trial that evaluated patients with acute spinal cord injury treated with placebo, methylprednisolone, or naloxone. The study showed that methylprednisolone (30 mg/kg loading dose, followed by 5.4 mg/kg/h for 24 hours) started within 8 hours of injury was associated with a significant improvement in both motor and sensory function compared with patients treated with a placebo.

The findings of the NASCIS trials significantly influenced clinical practice and led to an increased use of steroids in treating TON. However, the clinical improvement was modest in these studies, and concern existed that the clinical benefit demonstrated for those patients treated in the first eight hours with mega dose steroids was the result of a statistical bias, since the analysis was performed post hoc rather than prospectively.

In 2005, the results of the Corticosteroid Randomization After Significant Head Injury (CRASH) trial raised concerns regarding the use of mega dose steroids (same dose as given in the NASCIS 2 study) in traumatic brain injury.[6] This study was the largest randomized study that evaluated steroids in patients with traumatic brain injury and was stopped early due to the significantly increased risk of death in patients that received mega dose steroids at their 6-month follow-up when compared with the placebo group (25.7% vs 22.3%; RR 1.15 CI 1.07 to 1.24; p=0.0001). Although the etiology of the increased risk of death was not determined, the findings of this study should be taken into consideration when managing cases of TON with concurrent traumatic brain injury.

Furthermore, concerns have been raised regarding the extrapolation of data from spinal cord injury studies to TON. There are important histologic distinctions between the spinal cord and optic nerve; for example, the optic nerve is a pure white matter tract and the spinal cord is a mixed gray and white matter tract. As a result, significant biologic differences may exist between the repair mechanisms of the optic nerve axons and insults to the spinal cord.

The International Optic Nerve Trauma Study (IONTS) was a nonrandomized intervention trial that compared visual outcomes for patients with TON treated with observation, systemic steroids, or optic canal decompression.[7] Published in 1999, the study included 133 patients who were evaluated and treated within 7 days of the traumatic event, with most of the patients being treated with either corticosteroids (n=85) or surgical decompression of the optic canal (n=33). Follow-up results showed that visual acuity increased by more than 3 lines in 32% of the surgery group, 52% of the corticosteroid group, and 57% of the observation group. However, the study was nonrandomized and uncontrolled, and the small numbers of patients in the observation group (n=9) limited the strength of the study’s statistical power.

Recent animal studies have also found variable results with steroid therapy. These animal models involve a direct, crush injury mechanism for inducing TON in rats, and any extrapolation of the data to humans with indirect insults to the optic nerve must be made with caution. However, no study has demonstrated a beneficial effect for steroid therapy in animals with TON, and one study, in particular, found that steroids exacerbated axonal loss as evidenced by a dose-dependent decline in axonal counts with increasing doses of steroids.[8] Such findings may suggest that steroids can exert a negative effect on ganglion cell survival, especially at higher, mega dose levels, due to their suppression of endogenous neuroprotective pathways.

Based on the current evidence, a therapeutic role for corticosteroids in the management of TON is unsubstantiated. If steroids are considered for TON, they should not be used in cases with concomitant traumatic brain injury or in patients that present 8 hours or more after initial injury. Whether clinicians should use mega dose rather than lower doses of steroids for selected cases of TON is also not clearly defined by the literature. The NASCIS studies used mega dose steroids in their protocol to demonstrate a beneficial effect in a subset of their patients, but the CRASH study identified several serious complications associated with their use in the trauma setting. Additionally, animal studies have demonstrated an association between increasing doses of steroids and retinal ganglion cell death.[9, 10]

Surgical Therapy

The rationale for surgical therapy in indirect traumatic optic neuropathy (TON) is to decompress the optic nerve at the site of injury, which is often the intracanalicular segment. Surgical decompression is thought to help reduce optic nerve compression and subsequent vascular compromise that may occur as a result of the indirect injury. Additionally, surgery has been postulated to remove bone fragments that may be impinging on the optic nerve within the optic canal.

However, no randomized, controlled studies have been performed to evaluate the role of surgery in TON. As mentioned previously, one of the largest series is from the IONTS, which did not provide any convincing evidence that surgical decompression of the optic canal in TON is superior to observation or corticosteroid therapy. Additionally, in cases in which bony fragments are impinging on the optic nerve within the canal, the prognosis of visual recovery is extremely poor because the bony fragments are more likely to have anatomically disrupted the optic nerve axons, leading to irreversible visual loss.

Interpreting the efficacy of surgical decompression when reviewing the published studies is difficult because they consist of small, retrospective series that have variable methodologies and inclusion criteria (ie, degree of vision loss, timing of surgery, whether concomitant steroids were used). In addition, the probability of selection bias for TON patients who elect to have surgical decompression cannot be ignored because patients with the worse visual function at presentation or those who have failed steroid therapy tend to be included in these surgical series. The timing of surgery and the preferred surgical approach is also controversial.

The variety of surgical approaches used in optic nerve decompression include intracranial, extracranial, orbital, transethmoidal, endonasal, and sublabial approaches, and the selection of the technique tends to be based on the surgeon’s training, background, and experience. Patients with profound vision loss and a visualized bone fragment impinging on a segment of the intracanalicular optic nerve on neuroimaging have been considered to be the best candidates for surgical intervention. Although anecdotal reports of impressive visual recovery exist for such patients, most cases with direct injuries to the optic nerve do not improve, and the risk of possible surgical complications such as cerebrospinal fluid leak or postoperative bleeding cannot be ignored.

A study by Yu et al reported that in children with indirect TON, endoscopic trans-ethmosphenoid optic canal decompression (ETOCD) produced better results in patients with residual vision prior to surgery than in those who had no light perception, with the investigators finding improvement rates of 69.7% and 37.9%, respectively.[11]

At this time, conclusive evidence that surgical decompression has a beneficial role for most patients with TON does not exist. Because a significant rate of spontaneous recovery is found in cases of TON, a randomized controlled trial comparing observation with optic canal decompression is the only reliable way to evaluate the therapeutic benefit for surgical intervention. The decision to perform surgical decompression of the optic canal should be made on a case-by-case basis, with the patient being informed that, to date, surgery has not been shown to improve the prognosis over observation alone.

Outcome and Prognosis

Most studies show a significant association between initial and final visual acuities. Patients with no light perception (NLP) likely have little to no recovery in vision. However, studies show that up to 50% of patients with traumatic optic neuropathy (TON) can have some improvement in vision, with or without treatment, although most of the time improvement is minimal. No well-designed study has shown whether surgical decompression or steroids has any better outcome than observation alone. In fact, the rate of minimal but spontaneous visual improvement in indirect TON is relatively high, ranging from 20-57% in published series.

Studies have shown that TON with concomitant orbital fractures tends to have more severe visual loss.[12, 13] Up to 85% of the patients with an orbital fracture (29 out of 34) presented with NLP in one particular study.[13] The presence of an orbital fracture implies a greater transmission of force to the optic canal, and hence, a greater injury to the optic nerve.

Future and Controversies

Traumatic optic neuropathy (TON) can lead to profound visual loss from either indirect or direct mechanisms. The diagnosis can be made with accurate history taking and clinical examination, based on the presence of visual loss (with accompanying loss of color vision and possible visual field defects) and an accompanying relative afferent pupillary defect (APD).

The optimal treatment for TON, however, remains debated among physicians. A review of the available literature, especially the IONTS and CRASH studies, provides insufficient evidence to conclude that corticosteroid therapy and/or optic canal surgery provides a therapeutic benefit over observation alone in patients with TON. Patients with TON treated with systemic steroids appear to have similar rates of visual recovery as untreated patients, and both animal and human studies suggest that under certain conditions, systemic steroids may actually be harmful, particularly at higher doses.

Therefore, corticosteroids should not be used in cases with concomitant traumatic brain injury or in patients who present 8 hours or more after initial injury. Based on the available evidence, surgical decompression of the optic canal in cases is not routinely recommended in TON. If treatment with either steroids or surgical intervention is considered, appropriate counseling should be given to the patient and their family about their potential benefits and risks in order to help them make an informed decision.