Optic Nerve Decompression for Traumatic Optic Neuropathy

Hippocrates noted the association of trauma just above the eyebrow and gradual vision loss. By the 18th century, the relationship between frontal trauma and vision loss with an absence of ocular injury was well appreciated. In 1879, Berlin described the first pathologic examination of the optic nerve after head trauma. In 1890, Battle first distinguished penetrating direct from nonpenetrating indirect optic nerve injuries. The 20th century saw significant progress in defining the classification, pathophysiology, and management of traumatic optic nerve injuries.

Historically, the 3 treatment paradigms advocated for traumatic optic neuropathy are observation, medical corticosteroid therapy, or optic canal decompression.[8] In the early 1900s, transcranial unroofing of the optic canal was the surgical procedure of choice for traumatic optic neuropathy treatment. This procedure was used sparingly because of the inherent risks of intracranial surgery. In the 1920s, Sewell performed a transethmoidal optic canal decompression by removing the lamina papyracea and medial wall of the optic canal. Although his technique was refined by progressive advances in transnasal, transantral, transorbital, and external paranasal sinus surgery, the technique was not performed routinely until the 1960s in Japan and the 1980s in the United States. During this period, systemic corticosteroid treatment was also extended to treatment of traumatic optic neuropathy.

Recent advances in endoscopic instrumentation and intranasal sinus surgical techniques have refined extracranial surgical approaches for traumatic optic neuropathy. Currently, endoscopic optic nerve decompression (OND) via an intranasal and transethmoidal or transsphenoidal approach has gained popular support.[9, 10]

Craniofacial trauma patients are at significant risk for visual disturbances due to a variety of mechanisms owing to the vulnerable anatomic location and rigidity of the bony orbit. Trauma can precipitate various pathophysiological conditions that ultimately manifest as visual dysfunction. Nonneuropathic ophthalmic injuries should be excluded with a thorough ophthalmic examination, which includes orbital and cranial imaging studies. This examination should clearly delineate the nature of any neuropathic vision loss. Trauma-induced injury to the optic nerve can occur anywhere along the nerve's intraorbital-to-intracranial length.

Direct traumatic optic neuropathy is the term used when the optic nerve is impinged, crushed, or transected. These injuries are usually the result of open craniofacial trauma, such as penetrating wounds (eg, from knives, BBs, pellets) or extensive crush injury with displaced cranio-orbital fractures.

Indirect traumatic optic neuropathy occurs in the absence of direct optic nerve injury and is more common than direct traumatic optic neuropathy.

Frequency

In the United States, incidence of indirect traumatic optic neuropathy is approximately 2.5% in patients with midface trauma and 2-5% in patients with closed head injury.

Internationally, incidence of indirect traumatic optic neuropathy in the Western world is reportedly 0.7-5%. Most clinical series in Western literature involve fewer than 40 patients. A higher incidence of indirect traumatic optic neuropathy is reported in some Japanese studies; however, the reason remains unclear.

The UK literature reports that the vast majority of traumatic optic neuropathy patients are young males (79-85%), with a mean age of 34 years.[11, 12]

Traumatic optic neuropathy is most commonly caused by motor vehicle and bicycle accidents (15-75% of cases, depending on the series). Falls (15-50% of cases) are the next most common cause, followed by physical violence and recreational sports.

The exact pathophysiology of traumatic optic neuropathy is poorly understood. Although optic nerve avulsion and transection, optic nerve sheath hematoma, and optic nerve impingement (from a penetrating foreign body or bony fracture) all reflect traumatic mechanisms of the optic nerve dysfunction, they are frequently considered entities independent of traumatic optic neuropathy. These less common forms of traumatic neuropathic vision loss are covered separately in Traumatic Optic Neuropathy.

Traumatic optic neuropathy, in its most common form, is an indirect event that occurs during or shortly after blunt trauma to the superior orbital rim, lateral orbital rim, frontal area, or cranium. The most widely held belief maintains that compression forces from the trauma are transmitted via the orbital bones to the orbital apex and optic canal. Laser interferometry studies demonstrate that forces applied to the frontal bone are concentrated and transferred to the orbital apex and anterior foramen of the optic canal. Elastic deformation of the sphenoid then allows transfer of the force to the intracanalicular segment of the optic nerve. Contusion of the intracanalicular optic nerve axons and pial microvasculature produces localized optic nerve ischemia and edema. The edematous ischemic axons result in further neural compression within the fixed-diameter bony optic canal, precipitate a positive feedback loop, and trigger the development of an intracanalicular compartment syndrome.

Although ischemia is considered the secondary event that gives rise to the neuropathy, the cellular and subcellular events that constitute the mechanism of neural damage are only now being realized. The roles of oxygen free radicals, enzymes, cytokines, intracellular calcium, and other forms of reperfusion damage are slowly being uncovered through basic science research.

A less common form of traumatic optic neuropathy that involves the intracranial optic nerve results from forces delivered by the brain's shift at the moment of impact. The intracranial optic nerve is sheared as it moves against the falciform dural fold as it overlies the sphenoid plane.

The rationale for medical and surgical treatment of indirect traumatic optic neuropathy stems from the belief that trauma creates a mechanical shearing on a proportion of retinal ganglion cell axons and subsequent edema of the optic nerve. This swelling within the rigid confines of a bony optic canal causes further trauma to the previously undamaged retinal ganglion cells, perpetuating the vision loss; therefore, in theory, decreasing this swelling may halt further damage and limit secondary damage to the optic nerve.

The diagnosis of traumatic optic neuropathy is clinical. Patients with midfacial and cranial trauma should elicit a high index of suspicion for traumatic optic neuropathy. Although patients with traumatic optic neuropathy may have serious and obvious craniofacial, neurosurgical, and other comorbidities, they may also have no visible signs of injury. In addition, although 40-60% of patients with traumatic optic neuropathy present with a visual acuity of light perception or no light perception, nearly 20% of patients have a visual acuity of 20/200 or better.

Forces required to cause traumatic optic neuropathy put the patient at risk for significant head injury and mortality. In one large Canadian study, this was as high as two thirds, and up to 14%, respectively.[13] Features of traumatic optic neuropathy include unilateral or bilateral ocular involvement, reduced visual acuity, afferent pupillary defects, impairment of color vision, variable visual-field defects, and changes in the optic disc appearance on funduscopy.

Assume optic nerve dysfunction when a loss of best-corrected visual acuity or visual field is accompanied by an ipsilateral afferent pupillary defect (eg, Marcus Gunn pupil). This, of course, is not reliable when the patient has bilateral insult. Obtain a detailed medical history and identify premorbid ocular conditions that may limit vision recovery. If the patient's clinical situation limits detailed communication, query the patient's family, paramedics, or witnesses to the trauma about the details of the injury.

Perform a comprehensive ophthalmic examination on all patients in whom traumatic optic neuropathy is suspected and include the following assessments:

• Ocular adnexa: Examination may reveal orbital rim and wall fractures, orbital edema, proptosis or enophthalmos, or extraocular muscle dysfunction. Signs of penetrating injuries, such as protruding foreign bodies, extruding orbital contents, or conjunctival laceration, may range from obvious to subtle.

• Visual acuity: Assess visual acuity immediately upon presentation. Perform a second assessment within 24 hours of the first to discern cases of delayed optic neuropathy (< 10% of traumatic optic neuropathy cases).

• Pupillary reaction: An afferent pupillary defect (APD) is a necessary condition for the diagnosis of traumatic optic neuropathy. Normally, light in one eye causes equal constriction of both pupils (direct and consensual pupillary light reflex). With APD, light in the affected eye causes only mild constriction of both pupils. Light in the unaffected eye causes normal constriction in both pupils. Pupillary reaction is evaluated with the swinging flashlight test (ie, briskly alternating a flashlight beam from one eye to the other). Alterating the light between each eye every 2-3 seconds, the pupil of the affected eye will dilate with direct light and constrict with light in the unaffected eye.

• Intraocular pressure: Increased intraocular pressure may accompany an orbital hematoma, diffuse orbital hemorrhage, orbital emphysema, or soft tissue edema.

• Ophthalmoscopy: Perform ophthalmoscopy with the aid of a short-acting mydriatic agent (pupillary dilation) on all stable patients. Evaluate the retinal and choroidal circulation, optic nerve head morphology, and the presence of ring-shaped hemorrhage adjacent to the optic nerve head.

Medical or surgical intervention or a combination of both may be indicated for patients with indirect traumatic optic neuropathy. Indications for treatment are based on clinical judgment. Absolute indications for intervention, including optic canal decompression, have not been validated by controlled outcome studies; currently, physicians must decide on therapy for traumatic optic neuropathy without a consensus on standard of care.

In the Cochrane Database of Systematic Reviews, no randomized controlled trials were identified for either the use of corticosteroids or surgical treatment for traumatic optic neuropathy. Citing reports of visual recovery rates of 40-60% with conservative management, the authors conclude that the decision to proceed with surgery or high-dose corticosteroids depends on the clinical judgment and surgical skills of the surgeon as well as informed consent of the patient to appreciate the benefits and risks of both treatments.[14, 15]

The International Optic Nerve Trauma Study was organized to compare corticosteroids or surgery and corticosteroids, but after failure to enroll sufficient numbers of patients, the study was transformed into an observational study. Comparing no treatment (observation) versus corticosteroids or surgical decompression, the authors found no difference in the final visual acuity and commented that the decision to treat or not treat should be made on an individual patient basis.[16] A recently published study from Iran reported a randomized placebo-controlled trial of the use of intravenous high-dose corticosteroids versus saline in 31 patients with traumatic optic neuropathy; the authors found no statistically significant improvement in visual acuity in the 2 groups.[17]

Since publication of the CRASH (Corticosteroid Randomisation After Significant Head injury) study,[18] the role of systemic corticosteroids in the treatment of traumatic optic neuropathy has become more limited, and, in selected cases of traumatic optic neuropathy, surgeons may opt to bypass the use of steroid therapy and proceed to canalicular decompression.

In one paradigm, treatment is based on visual acuity and response to intravenous corticosteroid therapy. Patients must have a history of traumatic head injury, subnormal visual function, and an afferent pupillary defect. Patients who consent to the risks of treatment are initially started on systemic methylprednisolone therapy (dosage is available in the Medscape Reference article Traumatic Optic Neuropathy). Optic canal decompression is indicated (1) if visual acuity does not improve to 20/400 or better despite 24-48 hours of steroid therapy or (2) if visual acuity is 20/200 or better but deteriorates during or after completion of steroid therapy.[19]

The general anatomy of the optic nerve and its surrounding structures is outlined in Traumatic Optic Neuropathy. The features within this section emphasize specific anatomical structures the surgeon must understand to evaluate neuroimaging of the orbital apex and to perform surgical decompression of the optic canal.

If the anterior face of the sphenoid sinus is oriented vertically, the optic canal is likely adjacent to the lateral wall of the sphenoid sinus. If the anterior face is tilted obliquely, the optic canal may be adjacent to either the sphenoid or posterior ethmoid cells or both. When the canal is adjacent to the posterior ethmoid cells, the cells are known as Onodi cells, which may be pneumatized up to and even around the optic canal. In up to 25% of cases, a bony dehiscence may occur along the canal; always use care when working in this anatomic region.

The internal carotid artery is intimately associated with the optic nerve near the posterior foramen of the optic canal. Generally, the artery lies inferolateral to the nerve, away from the area of decompression. However, tortuosity in the carotid siphon, which brings the siphon closer to the zone of surgical decompression, can occur. Study of the carotid artery on preoperative imaging and the use of intraoperative computerized navigation assist appreciation of this anatomic variant.

Within the optic canal, the ophthalmic artery courses along the inferolateral aspect of the optic nerve. The artery does not enter the nerve until both the artery and the nerve are well anterior to the orbital apex. Carefully consider this anatomic relationship when surgically fenestrating the optic nerve sheath. Always perform surgical opening of the sheath in the quadrant medial and superior to the optic nerve.

Another anatomic consideration during decompression is the fused fibrous origin of the 4 rectus muscles (annulus of Zinn, Zinn ring). The optic nerve, ophthalmic artery, and fibers of the sympathetic nervous system emerge within this annulus at the anterior foramen of the optic canal. Some authorities believe that the annulus represents a nonosseous region that restricts optic nerve sheath distention and promotes optic nerve compression. Some proponents of optic canal decompression in traumatic optic neuropathy believe this annulus must be lysed in select clinical cases.

Patients with traumatic optic neuropathy may experience nonocular comorbidities such as closed head injury or multiorgan trauma. Basic and advanced life support is the primary objective until the patient is stabilized. The consultation for visual system evaluation should be prompt but must be triaged among the multiple consultations and ongoing critical care needs required for the individual patient. If neurosurgical concerns prevent pupillary dilation, a comprehensive examination of the posterior sclera, choroid, ciliary body, retina, and optic nerve head may be compromised or delayed. The evaluation and treatment of traumatic optic neuropathy begins after all other life-threatening injuries have been stabilized and basic lifesaving protocols have been fulfilled.

The use of corticosteroids in the treatment of traumatic optic neuropathy should be judicious in patients who are at risk (eg, those with diabetes mellitus, gastric ulcers, osteoporosis). In addition, a randomized controlled trial on the use of corticosteroids in patients with acute traumatic brain injury, the CRASH study discussed further in the traumatic optic neuropathy section, found a higher risk of death in the steroid group, leading investigators to prematurely terminate the trial.[18] Although the mechanism of higher mortality in the patients who received steroids remains to be elucidated, this should be considered in the decision to treat patients with traumatic optic neuropathy and head injury with corticosteroids.[15]

Hemostasis is essential during optic canal decompression. Obtain the following tests as suggested by the patient's medical history:

• Hemoglobin/hematocrit

• Platelet count

• Prothrombin time (PT)/activated partial thromboplastin time (aPTT)

• Bleeding time

Perform thin-slice CT scanning of the nose, sinuses, and orbits. CT scanning provides adequate imaging of orbital soft tissue and is better than MRI at delineating bony defects. A thin-section CT scan also provides an intraoperative road map for the surgeon in patients who require surgical decompression.

The decision for surgical decompression should still be based primarily on the clinical examination findings and not the CT scan findings. Small-review series have concluded that the extent of bony canal injury documented at surgery was underestimated by CT scan findings.

In polytraumatized patients with poor awareness, CT scanning with clinical exploration is the most important method for the assessment of traumatic optic neuropathy in the acute emergency setting. Fractures through the optic canal can be best depicted with thin-section CT scanning (eg, 1.5-mm cuts with 1-mm intervals).

Surgeons who wish to perform image-guided optic canal decompression need to obtain a special-order CT scan that is formatted to their computerized stereotactic localizing system.

Diffusion tensor magnetic resonance imaging (DT-MRI) may provide valuable information for evaluating the fibers of the optic nerve in traumatic optic neuropathy.[2, 3]

Patients suspected of sustaining traumatic optic neuropathy should undergo visual field testing. Although no visual field defects are pathognomonic of traumatic optic neuropathy, quantification of visual field defects is useful to assess convalescent visual improvements. Simple visual field screening can be accomplished at the bedside for unstable patients, but formally assess patients who can be evaluated in the clinic setting.

Multifocal visual-evoked potential (VEP), multifocal electroretinography (mfERG), and optical coherence tomography are 3 promising techniques in the future diagnoses of subclinical vision loss. Some of these tests are already used in neuro-ophthalmology for the studies of the retina and glaucoma. Although none of these techniques should replace a careful history and clinical examination, these techniques might be important as adjunct procedures in the evaluation of the unconscious patient or patients with bilateral optic neuropathy. Flash visual-evoked potential (FVEP) was studied in patients with traumatic optic neuropathy with calculation of a ratio of the amplitude of the injured to the uninjured eye. A ratio of greater than 50% was associated with favorable visual outcome.[20]

Histopathology is not integral to the clinical management of traumatic optic neuropathy. Clinicopathologic studies, however, have anecdotally demonstrated several features of traumatic optic neuropathy, as follows:

• Blood within the optic nerve sheath

• Interstitial optic nerve hemorrhage

• Fibrosis of the pial septa

• Lymphoplasmacytic infiltration

• Iron-laden macrophages

• Triangular-shaped axonal degeneration with loss of myelin

• Ischemic necrosis

The time-dependent histopathologic changes of the optic nerve following indirect trauma have not been adequately described.

The most widely accepted contemporary treatments for traumatic optic neuropathy have included observation, steroids, and surgical decompression, but concerns about the use of corticosteroids in patients with acute brain trauma have led to recommendations to not treat traumatic optic neuropathy with steroids.[21] Lack of a prospective large-scale clinical trial perpetuates controversy as to the optimal treatment for traumatic optic neuropathy.[16] The timing and type of decompression procedure and selected use and optimal dosing of perioperative corticosteroids have also been widely reported but have not been validated by controlled outcome trials.[22]

A revision of the Cochrane review on this topic found only a single double-masked, randomized, controlled trial comparing placebo to high-dose intravenous steroids for traumatic optic neuropathy but concluded that no convincing evidence suggests steroids provided any additional benefit to vision.[23]

Nonetheless, a survey of members of the American Society of Ophthalmic Plastic and Reconstructive Surgery and the North American Neuro-Ophthalmology Society found that out of 165 respondents, 60% treated traumatic optic neuropathy with corticosteroids. This included 67% of oculoplastic surgeons and 47% of neuro-ophthalmologists. Twenty-three percent of respondents managed traumatic optic neuropathy with surgery.[24]

A more in-depth discussion of steroid therapy for traumatic optic neuropathy can be found in the Medscape Drugs & Diseases article Traumatic Optic Neuropathy.

Surgical optic nerve decompression (OND) is a reasonable and reported treatment for traumatic optic neuropathy. Evidence suggests that initial visual acuity (IVA) of no light perception is the most significant determinant of outcome in traumatic optic neuropathy. Patients with IVA of no light perception treated surgically within 7 days of injury had a better improvement degree than patients managed medically.

Various surgical approaches for decompression of the optic canal include transfrontal craniotomy, extranasal transethmoidal, transnasal ethmoidal, lateral facial, and endoscopic procedures.[4] An intranasal endoscopic approach is favored because of the proximity of the optic nerve to the sphenoid sinus and Onodi cell. Advantages of this approach include lack of external scars, preservation of olfaction, decreased morbidity, and faster recovery time.[5, 6, 7]

Chen et al reported that endoscopic optic nerve decompression can be safely and effectively achieved via a direct sphenoidotomy performed through the sphenoid ostium, in patients with high sphenoidal pneumatization and no supersphenoethmoidal air cells. The study involved five cases of traumatic optic neuropathy, with a 45°-angled endoscope used to reach the optic nerve canal.[25]

Emanuelli et al reported a relatively good risk-benefit ratio in patients with posttraumatic optic neuropathy when a protocol was followed in which patients received endovenous steroid therapy no more than 8 hours after injury, with endoscopic endonasal decompression of the intracanalicular segment of the optic nerve performed within 12-24 hours after the start of medical treatment. The study involved 26 patients, with a maximum 41-month follow-up period.[26]

Obtain imaging studies to delineate the exact anatomical relationship of the optic nerve and carotid artery to the posterior ethmoid cells and sphenoid sinus. If receiving megadose systemic corticosteroids, the patient may continue these drugs at a tapered dosage. If the patient has completed a preoperative corticosteroid trial, administer a loading dose of dexamethasone 1.5 mg/kg (or equivalent) a few hours preoperatively. The steroid's anti-inflammatory effect reduces the inflammation induced by surgery. Preoperative systemic antibiotics may be initiated once surgery is scheduled to suppress any preexisting chronic rhinosinusitis.

The authors have been successfully using endoscopic optic nerve decompression for the past decade. Although the use of a computerized surgical navigation system is not mandatory, the systems offer anatomical assistance. The operation is performed entirely with a zero-degree endoscope.

Begin the procedure with a total sphenoethmoidectomy using a modified Messerklinger/Stammberger/Christmas technique. At the beginning of the operation, slowly inject 1% Xylocaine with 1:100,000 units epinephrine near the pterygopalatine and anterior ethmoid areas to facilitate intraoperative hemostasis. Injections within the posterior septum are made later when operative exposure is available. To take advantage of this hemostatic window of opportunity, complete the operation within 2 hours. Do not use intraoperative electrocautery because of its potential to damage the optic nerve and major vessels. To prevent inadvertent use of cautery, no electrocautery is connected within the operating room during the procedure.

The powered microdebrider (shaver) is used extensively for all parts of this operation, including bone removal. For thin bone and soft tissue, use a 12° angled shaver blade. Totally remove the uncinate process in its inferior two thirds and leave the maxillary antrostomy untouched. After a standard ethmoidectomy, insert the microdebrider through the natural ostium of the sphenoid and shave laterally to remove the sphenoid face.

After the sphenoid sinus is opened and the lamina papyracea is clearly delineated, make a hole through the thin lamina bone with a small curette, 1 cm anterior to the sphenoid face and immediately anterior to the bulge in the lateral sphenoid/posterior ethmoid wall caused by the optic nerve. Use care to avoid damage to the periorbita. (Opening the periorbita at this stage would hinder subsequent intraoperative visualization because of the prolapse of orbital fat into the ethmoid cells.) Remove the lamina bone posterior to the opening with the use of curettes and Blakesley forceps. As the bone removal moves posteriorly, the bone becomes thicker.

Expose the optic nerve and its sheath for a distance of approximately 10-15 mm. Thin the thick bone of the medial wall of the optic canal with the powered microdebrider fitted with a 3- or 4-mm straight, spherical, or angled router burr. Curettes may be used to complete the bony removal. The bony opening should expose at least 120° of the circumference of the nerve. Longitudinal opening of the Zinn rings and optic nerve sheath may now be performed. Clinical indications for opening these structures have not been elucidated clearly. The authors perform this fenestration only when preoperative visual acuity is at light perception level or worse. If the procedure is performed, longitudinally incise the optic nerve sheath with a very sharp sickle-shaped blade. The importance of a sharp blade cannot be overemphasized, since tractional force—even that induced by the cutting motion of the blade—on the optic nerve and sheath must be minimized.

Once the bony canal has been removed and the sheath has been incised, the procedure is concluded. No intranasal packing is placed.

Continue the systemic steroid therapy started preoperatively every 8 hours for 24 hours. If the patient had preexistent nasal or sinus mucosa inflammation, the steroids may be converted to oral prednisone and continued at a tapered dosage for 1-2 weeks. Postoperative antibiotic therapy has no known role, except perhaps in patients with preoperative chronic rhinosinusitis.

On the first postoperative day, start the patient on bulb-syringe saline nasal irrigations 3 times per day. Continue these irrigations for at least 1-2 weeks, until normal mucociliary function again is active.

Objectively define recovery of visual function based on serial assessment of multiple visual function parameters (eg, visual acuity, visual field, quantitation of afferent papillary defect, assessment of abnormal color vision). Perform daily follow-up evaluations immediately after trauma, during megadose methylprednisolone therapy, and immediately after surgical therapy. Less frequent examinations (q4-7d) are warranted during the intermediate period following surgery. Long-term follow-up is appropriate at a point 3 months or longer from the date of injury to document the final level of visual function.

For excellent patient education resources, visit eMedicineHealth's Eye and Vision Center. Also, see eMedicineHealth's patient education article Black Eye.

Injury to the globe, optic nerve, and extraocular muscles can result if the periorbita is breached during the removal of the lamina papyracea. A study of optic canal decompression on cadaveric specimens found microscopic evidence of injury to the optic nerve, pia, and an extraocular muscle, and dural fraying in the specimens.[27]

Injury to the anterior ethmoidal artery can produce an orbital hematoma and further aggravate the optic nerve status.

CSF leaks, meningitis, pneumocephalus, and death may occur after trauma to the cranial floor and dura.

Massive intracranial bleeding and stroke may follow injury to the intracranial internal carotid artery.

Numerous published reports are available that detail the clinical outcome of treatment modalities for traumatic optic neuropathy. Unfortunately, many of these reports are retrospective, have a limited number of patients, or suffer from one or more biases or deficiencies.

In 1999, Levin and the International Optic Nerve Trauma Study published the results of a multicenter, comparative, nonrandomized study of 133 patients with traumatic optic neuropathy.[28] To date, this is the largest series on steroid treatment available. The purpose of this study was to compare the visual outcome of traumatic optic neuropathy treated with corticosteroids, optic canal decompression surgery, or observation without treatment. Treatment, when undertaken, was initiated within 7 days of the injury. Seventy-six investigators in 16 countries collected the data between 1994 and 1997. Treatment decisions followed the investigators' customary practice, and no specific protocols for corticosteroid treatment or surgical technique were followed.

The study showed that visual acuity improved by 3 lines or better in 32% of patients treated with surgery, 52% of patients treated with corticosteroids, and 57% of patients in the untreated group. No clear benefit was found for either corticosteroid therapy or optic canal decompression. The study also found that the dosage or timing of corticosteroid treatment or the timing of optic canal decompression was not associated with an increased probability of improved visual acuity. Within the limitations of the study design, the authors concluded that neither corticosteroid therapy nor optic canal decompression should be considered the standard of care for patients with traumatic optic neuropathy. The authors suggested that whether to initiate treatment on an individual patient basis is reasonable for clinicians to decide.

In 1996, Cook et al performed a retrospective metaanalysis of all published English-language cases and selected non–English-language cases of traumatic optic neuropathy.[29] The authors found that vision recovery in treated patients was significantly better than in nontreated patients, but the authors found no difference in vision improvement among patients treated with steroids alone, surgical decompression alone, or combined steroid and surgical decompression.

Contrary to the findings of the International Optic Nerve Trauma Study, Kountakis et al (in a retrospective study of traumatic optic neuropathy patients treated from 1994-1998) showed that patients treated with surgical decompression following failed megadose steroid therapy fared significantly better than patients treated with megadose steroids alone.[19]

The role of delayed optic nerve decompression (OND), (defined as surgical decompression undertaken 2 weeks to several months after injury) in traumatic optic neuropathy remains unclear. However, the limited studies point to some benefit when this treatment is used as salvage therapy on patients who are not completely blind after steroid therapy failed.

Although the need for a large-scale, prospective, randomized, controlled treatment trial is evident, many individuals believe such a trial is unlikely, given the low incidence of traumatic optic neuropathy and difficulties that exist in the randomization of patients. To date, little evidence exists to guide the management of traumatic optic neuropathy. The only basis for medical treatment for traumatic optic neuropathy has been extrapolated from the randomized trials for treatment of spinal cord injury. Newer studies, however, point to increased complications in patients who received high-dose corticosteroid treatment after spinal cord injury or acute head injury, as reviewed by Steinsapir in 2006.[21] The risks associated with the use of high-dose corticosteroids and the risks of surgery along with a lack of evidence of clear benefit of either treatment must be considered in the management of traumatic optic neuropathy.

A better understanding of the cellular and biochemical mechanisms that involve normal and traumatized axons, glia, and myelin sheaths may eventually lead to better interventions for traumatic optic neuropathy. A better delineation of surgical indications and the standardization of operative technique will be welcomed advances. The role of adjuvant neuroprotective agents will probably be advanced. Although corticosteroids are but one type of neuroprotective agent, the recent expansion in basic science research regarding other neuroprotective agents may lead to a redirection in future therapy for this condition.

Researchers have been examining the role of A2A adenosine receptor as a potential therapeutic target for halting further retinal glial cell degeneration.[30] A recent murine study using transcorneal electrical stimulation in traumatic optic neuropathy found that this treatment may provide some delay in and protection from neuronal death in select responder groups of mice.[31] Further studies need to be done to define responders and nonresponders, but this article presents an interesting development on traumatic optic neuropathy treatment. Thus far, these and similar studies have been conducted in animal models but there is great hope that such endeavors will lead to improved treatment of traumatic optic neuropathy.