Cranial nerve II, the optic nerve, transmits afferent visual inputs from the retina that eventually reach higher cortical processing centers in the occipital lobe. In traumatic optic neuropathy (TON), complete or partial impairment of visual function follows orbital or head trauma due to damage of optic nerve fibers [7, 8].

Direct or indirect forces may damage the optic nerve. Blunt trauma to the head or orbit is the more common cause of TON, termed indirect TON [9]. The force of closed head trauma is transmitted through the surrounding soft tissue and bony optic canal to the optic nerve, disrupting optic nerve fibers [10, 11]. In direct TON, penetrating injury causes direct stress to the optic nerve with resultant anatomical disruption [9, 10, 12]. Direct TON is less common than indirect TON and has a worse prognosis for long-term visual potential [7].

There are four anatomical divisions of the optic nerve: intraocular (1 mm), intraorbital (24–28 mm), intracanalicular (9 mm), and intracranial (16 mm) (see Fig. 12.6) [11].

The intraocular ON is comprised of the optic disc, prelaminar area (anterior to lamina cribrosa), and laminar area. Beginning at the optic disc, 1.2 million retinal ganglion cell (RGC) axons exit the globe through the scleral canal to form the optic nerve. The RGC axons are physically supported by the lamina cribrosa and metabolically supported by interweaving astrocytes [9]. The lamina cribrosa is a system of 10 connective tissue plates integrated with the sclera, and axon bundles are transmitted through openings in the lamina cribrosa.

Posterior to the lamina cribrosa, the optic nerve fibers are myelinated by oligodendrocytes and surrounded by a meningeal sheath (pia mater, arachnoid mater, and dura mater). At this point, the optic nerve is within the muscle cone of the eye. The intraorbital part of the optic nerve is longer than the distance between the posterior globe and optic canal. This length difference allows for optic nerve laxity in the orbit and facilitates unrestricted globe rotation. Before entering the optic canal, the optic nerve passes through the annulus of Zinn.

The optic canal travels through the lesser wing of the sphenoid bone. In the optic canal, the dural sheath surrounding the optic nerve fuses with the periosteum and immobilizes the nerve.

At the transition from intracanalicular to intracranial segments, the optic nerve passes under a dural fold called the falciform ligament. The final intracranial portion of the optic nerve does not have a dural sheath. The optic nerve terminates at the optic chiasm where the neural fibers continue posteriorly as the optic tracts.

(see Fig. 12.7). The short posterior ciliary arteries, branches from the ophthalmic artery, provide vascular supply to the optic nerve head. The short posterior ciliary artery distribution has few anastomoses, so the optic nerve head is susceptible to ischemia in cases of poor perfusion [13]. The primary blood supply along the length of the optic nerve is from pial branches of the surrounding meninges. The branches are supplied by the posterior ciliary arteries, which are small branches from the ophthalmic artery. Although the central retinal artery and vein travel within the anterior 10–12 mm of the optic nerve, only a small portion of optic nerve vascular supply is from the central retinal artery (see Table 12.2 and Fig. 12.7).

The clinical picture of TON varies based on the anatomical site of injury. In posterior optic nerve injury, the most common injury, the optic nerve is damaged posterior to the entry point of the central retinal vessels [12]. In anterior optic nerve injury, a less common presentation, damage is anterior to where the central retinal artery enters the optic nerve. In this injury, visual loss is often associated with abnormalities in the retinal circulation [8].

The most common site of optic nerve injury is intracanalicular. In a study of 42 patients with TON, 71.4% of injuries occurred at this location [11, 14]. The dural sheath of the optic nerve is adherent to the periosteum in the optic canal, and such areas of dural sheath attachment are highly susceptible to trauma [14, 15]. Coup and contrecoup forces in blunt head trauma cause the greatest degree of injury at transition sites between mobile and fixed segments of the nerve [16]. Traumatic stress disrupts the pial vessels within the optic canal, and neuropathy may result from decreased vascular supply to the optic nerve [16]. In addition, the majority of blunt forces from frontal head injury are transmitted to the optic foramen [14, 17]. With the direct trajectory of force and the closed, inflexible space of the optic canal, there is higher risk of compression, shearing, contusion, and stretching injuries to the intracanalicular optic nerve [17].

The optic nerve is also susceptible to traumatic forces in the intracranial segment where the optic nerve passes under the falciform dural fold [8]. The falciform dural fold may impinge on the optic nerve and produce shearing forces during blunt trauma [18]. Still, the intracranial optic nerve is relatively shielded from injury by the surrounding soft tissue and bone. Shearing forces are first absorbed by the intracanalicular optic nerve and often do not reach the intracranial segment [17].

Comparatively, injury of the intraocular optic nerve segment is rare. The intraorbital segment of the optic nerve is usually spared from injury due to its laxity within the orbit and shielding by the surrounding fat and extraocular muscles [17].

There is an evolving understanding of the pathophysiology of TON, which is likely multifactorial [19, 20]. One perception of indirect TON describes primary and secondary injuries. With primary injury, the immediate outcome of trauma is contusion necrosis [8]. There is irreversible shearing of RGC axons and subsequent RGC degeneration [8, 12, 21]. Secondary injury is caused by optic nerve edema within the rigid optic canal, which produces a compartment syndrome [9]. Edema may be driven by the initial trauma or the resultant ischemia. The optic nerve is highly susceptible to edema from ischemia, inflammation, or compressive forces because its axonal transport system has a high-energy requirement and is easily disrupted. The compartment syndrome exacerbates ischemia by further limiting blood supply to the RGCs, and many RGCs undergo apoptosis [12]. Because primary injury is irreversible, TON treatments focus on limiting secondary injury. Evolving protocols include neuroprotection strategies and medical or surgical decompression to alleviate the edema [21–23].

A large 2009 survey of patients in the UK estimates the minimum incidence of TON as 1 per one million people per year [24]. TON occurs secondary to head injuries (see Table 12.3). In prospective studies, the incidence of TON following head trauma varies between 0.5 and 5% of closed head injuries [8, 19, 20]. In adult patients, the predominant mechanisms of TON include falls, assault, and motor vehicle accidents (MVAs; see Table 12.3). In pediatric populations, mechanisms for TON are similar and primarily include falls from a height, MVAs, and sports injuries [25, 26]. TON is more frequently diagnosed in men and younger patients, averaging around 30 years old [7, 24, 27–29]

In patients with mid-facial fractures, 2.5% develop TON [30]. The likelihood of TON increases with more severe head trauma. There is 6–10% rate for patients with zygomaticomaxillary complex fractures [19, 31, 32], and patients with nasoethmoid complex fracture have 1.6 times greater chance of developing TON [27].

Less common causes of TON in adults are secondary to recreation-associated trauma during sports or iatrogenic [24]. Iatrogenic TON has been identified secondary to orbital surgery [33], osteotomies [34, 35], facial fracture repair [36, 37], and endoscopic sinus surgery [32, 38].

Other serious injuries often coexist, including orbital wall fractures, closed globe injuries (both anterior and posterior segment), ocular adnexal injuries to eyelid, extraocular muscles, orbital foreign body, skull fractures, and intracranial bleeding [24, 29]. However, it is difficult to predict the likelihood of TON solely based on the severity of the trauma [39].

The diagnosis of TON is primarily clinical [9, 10]. There are often delays to diagnosis if the patient presents with impaired consciousness or unstable condition secondary to multisystem trauma [10, 19, 48]. The history frequently reveals a mechanism of head trauma, as stated above, followed by subjective decrease in visual acuity. Detailed questions should be asked about the mechanism of trauma, including the potential for ocular foreign bodies with penetrating injury. Determine the time course for visual change and time when the injury occurred [11]. Also investigate the patient’s past ocular history, including any previous trauma or surgery to the involved eye.

If there is suspicion for TON, assess pupillary reactivity, visual acuity (VA), visual fields, and color vision, perform a funduscopic examination on all patients. Examination may be limited due to lid ecchymosis, edema, or other extensive facial trauma [49].

The most reliable early sign of optic nerve dysfunction is the presence of a relative afferent pupillary defect (RAPD) [20]. A RAPD will be present with the exception of bilateral and symmetric TON [10, 11]. This can be examined regardless of a patient’s level of consciousness, so it is an invaluable measure for the diagnosis of TON in unconscious patients [39].

Patients with TON often have an acute decrease in VA following trauma. Occasionally, decrease in VA is delayed over weeks secondary to inflammation and edema or development of an optic nerve sheath hematoma [8, 20]. Visual acuity is often 20/400 or less in the affected eye [10]. In studies, 40–60% of patients initially present with light perception or worse [24, 32, 44].

Variable visual field defects are observed following TON. Commonly observed visual field defects following TBI are scattered visual field loss (58%) and homonymous visual field defects (22%) measured by confrontation method [50, 51]. If there is partial optic nerve avulsion, visual field loss will correspond to the location of injury [15].

Color vision is often decreased in TON. In one report, 76% of patients with TON visualized 12/17 or fewer Ishihara color plates [24].

Findings of a RAPD and a normal fundus examination increase the likelihood of TON. In posterior optic nerve injury, the fundus appears normal during the early clinical stage [12, 24]. With an anterior optic nerve injury, there may be optic disc swelling, dilated and tortuous retinal veins, or retinal hemorrhages [11, 15, 24]. Optic atrophy develops over the following weeks, and becomes evident 4–6 weeks after injury [9, 11, 19]. Therefore, it is useful to monitor patients with repeated dilated fundus examinations following closed head trauma. Of note, the fundus examination may initially be difficult to complete if there is coexisting intraocular hemorrhage (IOH), thus a repeat fundus examination should be performed at a later date in these instances.

Assess ocular adnexa for other effects of trauma, including crepitus or step-off from orbital fractures [10]. Evaluate for oculomotor dysfunction, including ability to perform extraocular movements, strabismus, and accommodative and vergence defects [52]. Identify proptosis, which may be associated with intraorbital hemorrhage and can be difficult to diagnose if there is also periorbital swelling. Complete a full ophthalmic examination to identify additional ocular consequences of head trauma, such as corneal abrasion, traumatic cataract, and vitreous prolapse [53].

Imaging has uncertain clinical value. There is no consistent correlation between the finding of an optic canal fracture, the severity of visual loss, and the prognosis for visual recovery [17].

CT is the best imaging tool to identify optic canal fracture, intracranial mass lesion, such as a hematoma, subluxated lenses, or other specific pathology compromising the optic nerve [11, 19]. There is no clear protocol for imaging studies in the diagnosis of TON. Some clinicians use CT scan for all cases, while others reserve imaging based on clinical severity or when contemplating the need for therapeutic interventions [20]. For example, CT imaging is part of surgical planning for optic nerve decompression [10, 49].

MRI is an alternate imaging modality to assess intracranial mass lesions. It is used for patients with a clear history of trauma and no likelihood of metallic foreign bodies. However, MRI may be difficult to perform in critically injured patients [19].

B-scan ultrasonography is an inexpensive imaging modality that can be used in closed orbit injury [11]. It has demonstrated utility for early diagnosis of traumatic optic nerve avulsion. Optic nerve avulsion is often associated with concomitant vitreous hemorrhage that obstructs visualization of the fundus, and B-scan allows visualization despite intraocular bleeding [54].

Flash visual evoked potential (FVEP) is a useful form of VEP because it can be performed in patients despite level of consciousness, lid hematomas, or refraction abnormalities [39]. VEP should be measured in both eyes, with the uninjured eye as a control [11]. The presence and amplitude of VEP predicts visual outcome in TON [55]. Deformed evoked potentials correlate with morphological changes of the optic nerve and measurements of reduced pupillary reactivity [39].

FA is used in the evaluation of anterior optic nerve injury, such as understanding the degree of retinal vessel involvement in optic nerve avulsion [56, 57].

In patients with indirect TON, there is a correlation between initial and final visual acuity. Patients who are NLP on presentation have little or no visual acuity improvement regardless of medical, surgical, or conservative management [7, 43, 58]. If there is no visual recovery after 48 h, the prognosis for final visual improvement is poor [20, 59]. Furthermore, younger patients have greater likelihood of visual recovery [60].

Measurements of the RAPD with neutral density filters indicate that patients with an initial RAPD of 2.1 log units or less can have visual acuity improve to 20/30 or better, while patients with an initial RAPD of 2.1 log units or greater have little visual improvement [17, 61].

Loss of consciousness following trauma is associated with a poor visual prognosis [20]. This finding may be related to a higher impact injury. Similarly, the presence of an orbital fracture is associated with smaller improvements from initial to final visual acuity due to a greater force of traumatic impact and greater severity of optic nerve trauma [7, 62]. Patients with anterior orbital fractures have more visual acuity recovery than patients with posterior orbital fractures [62]. Patients with optic canal fractures have varied success with visual recovery [19, 20].

VEP can also predict patient prognosis for visual function. FVEP amplitudes that are 50% of normal greater predict a good visual outcome of 20/30 or better (compared to the uninjured eye) [20]. Initial absence of visual evoked responses is associated with minimal, if any, visual recovery and final visual acuity of 20/300 or less [19, 55].

Compared to indirect TON, direct TON has a poor prognosis in all cases. There is acute, severe visual loss and to date no clinical interventions have demonstrated significant benefit for vision recovery [7, 20].

Direct trauma to the optic nerve may result from fractures of the optic canal, injury, or compression by bony fragments (Fig. 12.8), penetrating foreign bodies, or mass effect such as an expanding hematoma [63]. The causes of direct optic nerve injury can be distinguished with physical examination and imaging techniques. Generally, the diagnosis of direct TON is best detected radiographically with CT [63].

Treatment strategies are often limited. However, patients have been documented to regain vision [64]. A direct optic nerve injury may irreversibly damage one section of the involved nerve, but leave other areas with the potential for visual recovery [15, 64].

Optic nerve avulsion is a rarely reported form of anterior optic nerve injury [54]. Due to the difficulty with diagnosis, it is likely that the incidence of traumatic optic nerve avulsion is underreported. During trauma, the intraorbital optic nerve is forcibly separated from the globe at the lamina cribrosa, while the optic nerve sheath and adjacent sclera remain intact [54, 56, 65]. Avulsion may be partial or complete. In partial avulsion, there is segmental detachment of the optic nerve. In complete avulsion, there is full separation of the optic nerve from the retina, choroid, and vitreous and retraction of the lamina cribrosa from the scleral rim [56].

Optic nerve avulsion may follow a wide variety of injuries, including severe facial trauma, blunt injury of the orbit, and globe concussion [54, 65]. A frequent mechanism of optic nerve avulsion is direct trauma to the globe in which a small foreign body enters the space between the orbital rim and the globe to displace the eye [56, 66, 67]. For example, this can occur from a strike in the eye by a finger during contact sports such as basketball or boxing [66, 68].

Optic nerve fibers at the disc are susceptible to shearing forces as they pass through the lamina cribrosa because there is little supportive connective tissue in this location [69]. Suggested reasons for optic nerve tears at the lamina cribrosa refer to the traumatic forced rotation of the globe and resultant shearing forces, sudden increase in intraocular pressure driving the nerve posteriorly out of the scleral canal, or displacement of the globe concurrent with retropulsion of the nerve [9, 57, 69].

Acute, severe vision loss immediately follows orbital trauma [9]. Patients are often NLP on presentation with a RAPD. Because the anterior ciliary vasculature is maintained, the globe will have a normal appearance.

Fundus examination: The most specific means for diagnosis is fundoscopy (see Fig. 12.9). There is often a ring of hemorrhage encircling the visibly disrupted optic nerve head [8, 9, 15]. In the case of a partial avulsion, there are segmental depressions or excavation of the optic nerve head [56]. In complete avulsions, there is absence of the optic nerve head with a remaining “cavity” in its place [9, 54, 56]. There are often limitations to the fundus examination. The clinical diagnosis may be delayed in its earlier stages by extensive vitreous hemorrhage, hyphema, or retinal hemorrhage that obstruct a clear view of the fundus [56, 57, 68].

Ultrasonography: Positive findings with B-scan include visualizing that the optic nerve does not reach the optic disc, hyperlucency anterior to optic nerve, and posterior ocular wall defects in the region of the optic nerve head (see Fig. 12.10) [54, 65]. At 4–6 weeks after injury, optic nerve head defects and ingrowth of glial scar tissue have been identified [70]. B-scan is especially useful in cases of vitreous hemorrhage or other obscuration of fundoscopy. However, B-scan has demonstrated variable success in the diagnosis of optic nerve avulsion [67].

Fluorescein Angiography: The retinal vasculature may be intact or have varying degrees of interruption following either partial or complete optic nerve avulsion. There may be a spectrum of normal filling, delayed filling, partial filling, or complete absence of retinal circulation [56].

CT/MRI: CT or MRI is not the preferred imaging modality. Results are inconsistent for the diagnosis of ON avulsion [54, 65, 67].

No medical or surgical treatments have proven helpful in cases of optic nerve avulsion [9, 67].

Patients will have acute visual acuity loss and a RAPD. Early funduscopic examination is often normal, and optic nerve atrophy may be observed days following trauma (Fig. 12.12).

Diagnosis is largely based on clinical history of trauma, RAPD and acute vision loss with NLP visual acuity. CT scan may show a foreign body or bone fragment transecting the optic nerve (Figs. 12.13, 12.14 and 12.15) and VEP amplitudes are decreased [9].

There is currently no effective management for optic nerve transection [9]. If there is extensive fracture, surgical repair can be pursued; however the damage to the optic nerve is irreversible [72].

Hemorrhage with the potential to cause direct TON may involve the optic nerve sheath, the optic nerve, or the potential space between the optic nerve and optic nerve sheath. In an autopsy series following patients with closed head trauma conducted by Crompton et al., 83% of autopsies had evidence of optic nerve sheath hemorrhage and 36% had hemorrhage involving the interstitium of the optic nerve [18].

The hematoma or hemorrhage has a compressive effect on the vessels supplying the optic nerve and the optic nerve axons [15]. If blood is not evacuated promptly, there is risk of irreversible damage to the compressed optic nerve axons [73].

Patients may present with proptosis, RAPD, and a sudden onset monocular decrease in visual acuity [9, 74, 75]. The optic disc may appear healthy in early stages, or optic disc edema can be present [75, 76].

A CT scan can identify an optic nerve sheath hematoma (Fig 12.16). The affected optic nerve sheath will be expanded and hyperdense, and its increased size and density can be visualized [9, 74]. MRI is more specific than CT scan in detecting and localizing optic nerve hemorrhage [76].

Optic nerve sheath or intrasheath hemorrhage benefits from surgical drainage [15]. In optic nerve sheath hematoma, early nerve sheath fenestration to evacuate the clot is associated with greater recovery of visual acuity [9, 74, 77]. Megadose steroids have been used to treat optic nerve sheath hematoma, but their utility is unclear (See: Indirect TON Treatment section for discussion about corticosteroid use in TON) [75].

Orbital compartment syndrome describes an acute increase of intraorbital pressure. It may arise secondary to orbital hemorrhage or orbital emphysema. Because a fascial sheath and rigid bony walls enclose the orbital contents, small increases in intraorbital volume from blood or air can elevate the intraorbital pressure [37]. Arterial supply to the optic nerve can be compromised and lead to optic neuropathy [9, 15].

Orbital hemorrhage can follow facial injury or surgery. Injection of retrobulbar anesthesia is a well-documented iatrogenic cause of orbital hemorrhage, occurring in 0.44–3% of cases [15].

Orbital emphysema is a common complication of orbital fracture, occurring in 50% of cases [78]. Nose blowing, coughing, sneezing, vomiting, or straining while lifting heavy objects by a patient with a blowout fracture can force air from the sinuses into the orbital space [78, 79]. Orbital emphysema is largely a benign process. However, air may become trapped in the orbit if a small fracture creates a one-way ball-valve mechanism [9, 15]. The trapped air increases intraorbital pressure and has a compressive, mass effect on the optic nerve [78, 80].

Patients will experience acute onset proptosis, orbital pain, decreased visual acuity, RAPD, restricted extraocular muscle motility, and raised intraocular pressure [9, 15, 78, 80]. In ocular emphysema, palpation of the upper and lower eyelids may reveal tense underlying tissue [78].

Diagnosis is primarily clinical, in association with a history of facial trauma or recent orbital surgery. CT scan is useful to evaluate the fracture, identify presence of orbital emphysema or hemorrhage, and localize air pockets [80]. Facial radiographs may also reveal air within the orbit, but it is best confirmed with CT scan [80].

The goal of management is to rapidly decrease orbital pressure. Emergency surgical decompression by lateral canthotomy or cantholysis is essential [73, 78, 80]. The maximum potential for visual recovery follows surgical decompression within 2 h of injury [81]. In traumatic orbital compartment syndrome from retrobulbar hemorrhage, successful orbital decompression is achieved with lateral canthotomy and inferior cantholysis. When performed promptly, the patient may recover full visual function [82].

Indirect TON is the most common form of optic nerve injury following blunt closed head trauma [9, 14, 28]. It involves the damaging effect of shearing forces on optic nerve RGCs (see section: TON Pathophysiology).

The treatment of indirect TON is currently debated. Before any treatment is initiated for TON, other treatable causes of visual loss should be ruled out [10, 83]. Approaches to indirect TON include conservative management, corticosteroids, optic canal decompression, and combined surgical and steroid treatment. In addition, research for neuroprotection strategies designed to repair damaged RGCs is ongoing.

With conservative management for indirect TON, patients are observed for visual recovery. Interventions such as steroids or surgery are withheld [19]. Reports of spontaneous visual recovery following indirect TON are encouraging for conservative management, with a recovery rate of 30–60% [12, 20, 43, 83, 84]. The prognosis for spontaneous recovery is positively related to the patient’s presenting visual acuity [12, 20]. A prospective study is needed to more clearly define the natural course of indirect TON [8, 83].

There is mixed evidence regarding the use of corticosteroids to treat indirect TON. High-dose corticosteroids administered soon after trauma may relieve optic nerve edema within the optic canal and decrease risk of damage to RGCs [14]. However, research outcomes have shown unclear benefits with corticosteroid treatment. There is debate on the best dose and time course of treatment. In addition, the risks of long-term disability and mortality from high-dose steroid regimens must be carefully considered.

In 1990, the multicenter, randomized, double-blind controlled National Acute Spinal Cord Injury Study (NASCIS 2) examined the use of “megadose” methylprednisolone following acute spinal cord injury [85]. An experimental group of patients with acute spinal cord injury was administered a bolus methylprednisolone dose of 30 mg/kg followed by infusion at 5.4 mg/kg/h for 23 h, while a control group received a placebo. Neurologic recovery was assessed with measurements of motor function, sensation to pinprick, and sensation to touch. Patients were followed up to 6 months after injury. The researchers found that patients who received methylprednisolone within 8 h of their injury had significant neurologic recovery, and this recovery remained stable when assessed 6 months following injury. Neurologic recovery was not observed in the placebo group or patients who received methylprednisolone greater than 8 h following trauma [85].

In a subsequent multicenter, randomized, double-blind clinical trial, NASCIS 3, researchers examined the outcome of extending the duration of megadose methylprednisolone treatment. Patients were treated within 3–8 h of acute spinal injury, and the study maintained the bolus dose of 30 mg/kg that was used in NASCIS 2 [86]. However, half of the participants received the 5.4 mg/kg/h methylprednisolone infusion for a total of 24 h and half of the participants received the infusion for 48 h. Neurologic recovery was measured and patients were followed for 6 months. The study results confirmed the neurologic benefit of early megadose corticosteroids following acute spinal injury. The study concluded that 48 h treatment benefits patients, if treatment is initiated 3–8 h after injury. This group had improved motor recovery and self-reported functional independence with 48 h steroid therapy. However, there was also an increased likelihood of severe sepsis and severe pneumonia in the 48 hour steroid infusion group. For patients who were treated within 3 h of injury, there was no significant difference between 24 and 48 h treatment regimens, and it was advised to continue using 24 h therapy for this group to reduce the complications of steroids [86].

Following the NASCIS 2 and 3 reports of neurologic recovery with megadose corticosteroids, researchers have sought to apply the results to patients with indirect TON [8]. An increasing number of patients with TON were managed with high-dose corticosteroids [10]. The results of research studies have varied widely. Several studies found a positive relationship between high-dose corticosteroids and recovery from TON [14, 44]. However, some studies comparing corticosteroid managed and conservatively managed patients with TON did not yield significantly different visual outcomes [29, 43]. There were also studies comparing corticosteroid management with surgical management that did not find a difference in treatment approach [62].

The research regarding indirect TON treatment has been limited by several factors. Many studies had small sample sizes or a lack of control group. Ability to compare studies has been greatly limited [10, 43]. Studies differed on multiple factors including differences in the timeline of treatment, corticosteroid dosing, baseline visual acuities, follow-up frequency, inclusion criteria, and exclusion criteria. Furthermore, in several studies, patients received combined corticosteroid and surgical treatment, which made it difficult to study the efficacy of each individual treatment [42, 73].

In 1999, the International Optic Nerve Trauma Study (IONTS) was established to address the limitations of indirect TON treatment research in evaluating the effect of corticosteroid treatment [28]. The study was designed as a randomized controlled trial, but was later modified to an observational nonrandomized study due to difficulties with patient recruitment. The study followed three types of treatment groups: corticosteroid only, surgical decompression only, and conservative management. In the corticosteroid treatment group, 40% received megadose steroids consistent with the NASCIS trials, while 60% received lower doses (see Table 12.4). For all treatment groups, the initial visual exam was conducted within 3 days of injury and treatment was initiated within 7 days of injury. Visual outcomes were assessed by change in VA. After adjustment for baseline visual acuity, the researchers concluded that there was no difference in visual outcome between the treatment groups [28].

In addition, the researchers did not find that timing or dose of treatment had an effect on the visual outcome. The researchers concluded that the the type of treatment should be made on a case-by-case basis, and there is no clear evidence to recommend one treatment over another [28].

The IONTS has merit as the first prospective study to evaluate the treatment of indirect TON. The researchers used a standardized method of defining and comparing visual improvement between the treatment groups [28]. The IONTS was a large patient study, and patient enrollment was great enough to detect statistical differences between the corticosteroid and surgical treatment groups [9]. However, there were study limitations. There were fewer patients in the conservative management group, and the conclusions regarding this group do not have as much power. Furthermore, the study design was not randomized, controlled, or masked. Because it became an observational study, there may have been selection bias regarding choice of treatment, and some patients were treated with both corticosteroids and surgery [8]. There were differences in corticosteroid dose, timing of treatment, and criteria for optic canal decompression between the patients [11].

Since the IONTS, corticosteroid research for indirect TON remains limited. The 2013 Cochrane Eyes and Vision Group review of steroids for TON identified only one randomized double-blind controlled trial to address the use of high-dose corticosteroids for indirect TON [20]. The research, by Entezari et al. in 2007, randomized patients with indirect TON to either a high-dose corticosteroid group (1000 mg methylprednisolone/day over 72 h) or a placebo group [58]. Patients were diagnosed with indirect TON within 7 days of injury based on clinical presentation of RAPD and decreased visual acuity. On average, they were treated within 52 h of injury. The patients were followed for 3 months and visual improvement was measured by visual acuity. At 3 months, there was no significant difference in visual recovery between the treatment and control groups. This conclusion supports the results of IONTS.

In addition to questions about corticosteroid efficacy, there have been concerns about treatment risks [63]. NASCIS 2 identified more frequent gastrointestinal bleeds in the steroid-treated group, and NASCIS 3 identified more severe sepsis and pneumonia in the 48-hour steroid-treated group [85, 86]. In addition, wound infections, pancreatitis, and acute psychosis have been reported following high doses of methylprednisolone [11].

In 2004, the Corticosteroid Randomization After Significant Head Injury (CRASH) trial investigated the safety of megadose corticosteroids administered for head injury [87]. The CRASH study was an international, multicenter, double-blind randomized placebo-controlled trial that enrolled 10,008 patients with head trauma. The patients were randomized within 8 h of injury. One group received a 2 g methylprednisolone over 1 hour in a 100 mL infusion, followed by a maintenance dose of 0.4 g/h for 48 h in a 20 mL/h infusion and the other group received a placebo. The doses were set to correspond with the NASCIS 3 trial. The study recruitment was stopped early when unmasked results indicated higher mortality rates in the steroid-treated group. At 2-week follow-up, there was a significantly greater risk of death from all causes for the patients in the corticosteroid group [87]. At 6-month follow-up, the researchers found a significantly higher risk of mortality and severe disability in the steroid-treated group [88]. This effect did not change based on injury severity or time from injury. The researchers concluded that corticosteroids are not recommended for cases of head trauma. The study results can extrapolate to the treatment of indirect TON because many of these patients also experienced head trauma [8].

Concerns about corticosteroid risk have been substantiated by animal studies. Steinsapir and coworkers examined the effect of very high-dose corticosteroids on injured optic nerves in rats [89]. Following crush injuries to the optic nerve, rats either received very high-dose methylprednisolone or saline placebo. When the optic nerves were examined six weeks after crush injury, the rats that underwent steroid treatment lost more axons than the placebo group. The study indicates that methylprednisolone may be directly toxic to a damaged optic nerve.

Further research with double-blind RCTs is needed to elucidate how steroids should be used in the treatment of indirect TON. Based on the IONTS and the study by Entezari et al., there is no clear evidence that steroids provide an additional benefit in the treatment of indirect TON. Future research regarding other doses and timeframes of corticosteroid treatment may alter these conclusions. Given the CRASH trial results, steroids should not be given to patients with head injuries. Other patients should be carefully evaluated on a case-by-case basis.

Surgical management of indirect TON is focused on reducing secondary injury to the optic nerve (See: Pathophysiology section). Optic nerve edema primarily affects the intracanalicular segment of the optic nerve. Optic canal decompression relieves the secondary intracanalicular compartment syndrome by providing more space for the edematous optic nerve and limiting post-injury ischemia [8, 11, 29, 90]. Early surgical intervention is believed to augment the potential for visual recovery [19, 91].

Surgical indications for TON are limited by research studies, which are primarily small and retrospective [19, 23]. The criteria are well defined for optic nerve hematoma, but surgical indications are less clear with optic canal fracture or indirect TON.

There is strong evidence to support surgical management in cases of hematoma (See: Direct TON-hematoma section). With retrobulbar hematoma or optic nerve sheath hematoma, urgent surgical decompression and evacuation of the hematoma is indicated [11, 58, 63]. Surgical evacuation of hematoma with optic nerve sheath fenestration for intrasheath hemorrhage quickly relieves optic nerve compression [24]. In case reports of patients with radiological evidence of enlarged optic nerve sheaths and gradual visual loss, there is a positive response to optic nerve sheath decompression with a high potential for visual recovery [8].

With optic canal fracture, surgical management with optic nerve decompression has been recommended. In particular, surgery has been recommended for comminuted canal fracture or radiologic evidence that the fractured optic canal segment impinges on the optic nerve [11, 19, 24, 28, 45, 58]. However, optic canal fracture has been identified as a poor prognostic factor, regardless of intervention, and researchers question if surgery is beneficial in all cases of optic canal fracture [20]. Poor outcome is related to optic nerve transection. If the bone fragments transected the optic nerve, the injury to RGCs is irreversible and optic canal decompression would not have an effect on visual function [19]. As a result, case studies indicate that surgical decompression for patients with optic canal fractures has low potential for significant visual improvement [92].

The considerations for indirect TON are not as well defined. Given the pathophysiology of indirect TON, optic nerve swelling is a commonly identified reason for optic canal decompression [8]. Intraneural edema, diagnosed radiologically and suggestive of indirect TON, has been cited as an indication for surgical management [11].

Surgical candidates are identified based on the likehood of visual recovery. Indirect TON patients with gradual visual decline and VA better than NLP are considered the best candidates for surgical decompression [49, 93]. Gradual visual change following injury is considered to indicate a reversible cause of vision loss, whereas sudden onset, acute and complete vision loss is more likely to indicate a nonreversible process such as optic nerve transection [14]. Given their poor prognosis, it is unclear if there is a benefit to treating patients with total, sudden vision loss following trauma [42, 94]. In retrospective studies, few patients (0–17%) with sudden vision loss who were NLP at presentation regained vision after surgical decompression [14, 95]. In comparison, the same retrospective studies indicated that the majority of patients with vision better than NLP experienced some degree of visual recovery either after surgical decompression or conservative management [95].

Because the guidelines for optic canal decompression are unclear, the surgery should only be considered for conscious patients [15, 83]. These patients are able to undergo a full ophthalmic exam to first rule out other causes of visual loss [42]. They can also be counseled regarding the risks and benefits of surgery, and actively participate in the decision-making process [10, 90].

It is widely understood that earlier management of indirect TON yields better visual outcomes. Several reports define effective surgical management as occurring within 1 week of trauma [15, 49, 60, 96].

High-resolution CT scans of the intraorbital and intracanalicular optic nerve are helpful to visualize optic nerve pathology and evaluate the need for surgical management [49]. CT findings that indicate a need for surgical management include an enlarged optic nerve sheath suggestive of hemorrhage, optic nerve edema, or optic canal fracture [42, 49]. After surgical need is identified, CT scans are further utilized in surgical planning [42]. For example, imaging the orbital apex identifies the relationship of the optic nerve and the carotid artery, which is critical to reduce intraoperative complications [90].

There are a variety of surgical techniques for optic nerve decompression. In all approaches, a wall of the optic canal is removed to provide room for the edematous optic nerve [45]. The earliest technique for optic canal decompression, developed in the 1960s, takes an intracranial approach to reach the optic canal [15, 92, 97]. With a craniotomy, surgeons are able to directly visualize a large area of the brain and potentially discover coexistent areas of brain damage following trauma [45]. However, the intracranial approach has significant risks to the patient and outcomes are overall poor [14, 49]. It is an invasive procedure and requires an external incision. Possible consequences of intracranial optic nerve decompression include frontal lobe edema due to the need to retract the frontal lobe, and the loss of olfaction [91, 94, 96].

Extracranial approaches to reach the optic canal were developed as a less invasive way to achieve optic canal decompression [49]. There are a variety of extracranial approaches, and all surgical techniques have the same final step of removing a wall of the optic canal to achieve decompression. The extracranial transethmoidal approach is most frequently used [10, 42]. Approaches that have been reported include transconjunctival [73, 98], transantral [92], transethmoidal [60, 92, 99, 100], transnasal [10], and lateral facial [101]. Combined techniques include transethmoidal/transorbital [90], transethmoidal/transsphenoidal [91], endonasal transsphenoidal [96, 102–104] and transconjunctival/intranasal endoscopic approaches [94].

Extracranial approaches offer several advantages over intracranial optic canal decompression. Because the approach is less invasive, it can be completed under general anesthesia [45, 92]. There is no external incision, no risk of frontal lobe edema or decreased olfaction, a faster recovery time, and decreased morbidity compared to intracranial techniques [91, 94, 96].

Although safer than intracranial surgery, extracranial optic canal decompression has several reported complications. All of the extracranial approaches risk cerebrospinal fluid leak, meningitis, intraorbital infections, and accidental dural exposure [11, 23, 91]. For example, in the IONTS, 3 out of 33 patients treated with extracranial decompression experienced CSF leak, and one developed meningitis postoperatively [28]. Enophthalmos may result from expansion of orbital volume [90, 94].

There is risk of damage to the surrounding anatomy in the orbit and the sellar region [9]. Direct, thermal, or ischemic injury to the optic nerve can occur intra-operatively [14, 90, 105]. There may be injury to the lacrimal apparatus, ophthalmic artery, or surrounding brain tissue [11, 91, 92, 94]. With a transconjunctival approach, there is risk for medial rectus entrapment and injury to the globe [94]. All approaches to optic canal decompression carry the serious risk of injury to the carotid artery because the optic nerve travels close to the carotid artery near the sphenoid sinus [10]. There is high mortality associated with damage to the carotid artery due to difficulty controlling hemorrhage, subsequent intracerebral hypoperfusion, and extravascular dissection of the expanding hematoma [90]. If carotid artery laceration occurs, immediate management should include sinus pressure packing tamponade, carotid artery ligation, and possible endovascular tamponade [90].

Ultimately, selection of surgical method should be individualized. Surgeon experience, location of pathology, and patient anatomy should all be considered. The intracranial approach still has use if intracranial pathology secondary to trauma also needs to be addressed [90]. When pathology is localized to the optic nerve, an extracranial approach is safer for the patient.

As with corticosteroid treatment, the decision to manage indirect TON with surgical decompression has not been standardized and the choice to perform surgery should also be made on an individual basis. Several studies have reported success with optic nerve decompression; however, these are frequently uncontrolled, small, retrospective case studies so it is difficult to make clear conclusions about surgical indications or timing [49, 60]. Furthermore, it is not possible to compare published studies because a variety of intracranial and extracranial surgical decompression techniques are employed, and frequently corticosteroid treatment is given to surgical patients in non-standardized regimens [11, 91]. There is also a recruitment bias prevalent in the literature, with selection of patients who have worse baseline visual acuities or have already failed corticosteroid treatment [11, 12]. Based on a Cochrane Review examining Optic Nerve Decompression for TON, there are no randomized controlled trials in which any form of surgical intervention for TON, either on its own or in combination with steroids, was compared to steroids alone or no treatment [23]. The IONTS (see: Corticosteroid Treatment section) could not find a reason to recommend surgical management for indirect TON over corticosteroids or conservative management [62]. Without clear evidence that surgical decompression benefits patients with TON, and given the well-known risk factors of the surgery, the decision to perform optic nerve decompression will remain controversial until further research is conducted [23].

The goal of experimental treatment studies for TON is neuroprotection. Following trauma, irreversibly damaged RGCs release neurotransmitters and inflammatory factors that induce apoptosis of nearby, undamaged RGCs. New treatment strategies aim to prevent the receptive secondary apoptosis and minimize total damage to the ON [11, 19, 83]. For example, glutamate inhibitors prevent apoptosis in animal models [106]. Neurotrophic factors including fibroblast growth factor [2, 107], brain-derived neurotrophic factor [108, 109], and ciliary neurotrophic factor [110, 111] combat growth factor deficiencies following TON and promote survival of remaining RGCs. Additional neuroprotective research includes the anti-inflammatory small heat-shock protein crystalline [112], nitric oxide synthase inhibitors [113], tumor necrosis factor-alpha inhibitors [114], calcium channel blockers [115], the immunosuppressant FK50 [6, 116], and erythropoietin [117]. These interventions are promising, but still in early stages of investigation.

Adie’s tonic pupil results when postganglionic parasympathetic nerves to the iris sphincter and ciliary muscle are lost. The loss of this innervation results in a fixed dilated pupil that reacts poorly to light, but will constrict to near stimuli. When the near stimuli are removed, the characteristic tonicity becomes apparent as the pupil slowly relaxes. Additionally, Adie’s tonic pupil is known to be supersensitive to weak miotic agents such as pilocarpine due to cholinergic denervation [118]. Adie’s pupil can be associated with findings such as decreased deep tendon reflexes and the presence of anhidrosis. This triad of findings is known as Ross syndrome [119].

Adie’s tonic pupil is often idiopathic. Women are more often affected with a mean age of 32 years [120]. The tonic pupil initially presents unilaterally in 80% of cases, and the chance of second pupil involvement is 4% per year [120]. Underlying causes that may be responsible for the parasympathetic denervation include traumatic, iatrogenic, viral, and neoplastic etiologies.

Horner syndrome involves disruption of the sympathetic efferent fibers that normally cause the pupil to dilate, assist the lid in retraction, and enable perspiration of the forehead. The sympathetic pathway involved begins in the hypothalamus and, over the course of three neuron synapses, descends down the cervical cord, exits through the ventral rami of C8, T1, and T2, and joins the paraverterbral sympathetic chain to ascend some levels before continuing along the wall of the carotid artery and exiting the carotid wall to innervate its targets [121]. Dysfunction along the pathway, as depicted in Fig. 12.17, results in the typical findings of Horner syndrome: ptosis, miosis, and anhidrosis.

Spontaneous internal carotid artery dissection has been reported to present as Horner syndrome in approximately 28–41% of cases [122]. Internal carotid artery dissection occurs with an incidence of 2.6–3.0 in 100,000, and with equal occurrence rates in males and females, most often of young and middle-ages [123].

In one study of combat-associated ocular trauma, Horner syndrome due to upper chest and neck injury represented approximately 1% of the total cases. Similarly, a series of TBI patients demonstrated Horner syndrome in 0.9% of cases [124]. In one rare case, Horner syndrome was reported as a direct consequence of first rib fracture, itself a rare occurrence, suffered in a motor vehicle accident without carotid dissection [125].

Iatrogenic Horner syndrome can be a rare complication of different surgical procedures involving the head and neck, as described below. Even with closely associated procedures, such as conventional thyroidectomy, the occurrence of Horner syndrome has been reported to be as low as 18 cases between 1993 and 2015 [126].

Pediatric Horner syndrome, although rare, can be acquired or congenital. One study determined the pediatric incidence to be 1.42 per 100,000 patients younger than 19 years, and congenital syndrome to occur in 1 in 6250 births [127]. In a series of 73 pediatric patients with Horner syndrome, 42% were considered congenital or caused by birth trauma, 15% were acquired without surgical intervention such as by neoplasm or trauma, and 42% were acquired with surgical intervention as the cause [128].

Horner syndrome consists of the three classic findings of ptosis, miosis, and anhidrosis. Though these three findings make up the syndrome, dysfunction along the third-order, postganglionic neuronal fibers that run along the internal carotid artery will involve only the fibers responsible for mydriasis and lid retraction, thus resulting in a Horner syndrome with spared facial sweating [123].

In Horner syndrome caused by carotid artery dissection, common symptoms include ipsilateral headache in 68–92% of cases, focal cerebral deficits in 49–67% of cases, and cranial nerve palsies in 12–14% of cases [123].

Often, the syndrome occurs due to cervical-carotid dissection that can be caused by a motor vehicle accident (MVA), penetrating trauma to the upper chest or neck or iatrogenically [122, 129]. Such damage to the motor neuron pathway along its anatomical course will result in the signs and symptoms of Horner syndrome.

Iatrogenic Horner syndrome occurs with multiple types of procedures due to the extensive pathway of the sympathetic system along the spinal cord and carotid artery. Procedures involving manipulation of the neck, such as in chiropractic manipulation or endotracheal tube placement have the risk of causing iatrogenic Horner syndrome. The association with thyroid surgery described above is thought to be due to the contiguous localization of the oculosympathetic ganglion with neurovascular supply to the thyroid gland, complicating the isolation of thyroid structures during removal without causing Horner syndrome [126]. Other possible etiologies include postoperative hematoma compressing the sympathetic pathway, ischemia to the neurons, and stretching during retraction [130]. Minimally invasive video-assisted thyroidectomy is also thought to be associated with the occurrence of Horner syndrome, due to the limited mobility afforded to the endoscopic tools used in the procedure by the region of anatomy [126].

One example of surgical intervention resulting in a pediatric case of Horner syndrome is from a case report of pulmonary hydatid cyst removal in a 10-year-old girl whose ptosis recovered in 3 months and anisocoria recovered in 6 months [131].

Treatment of Horner syndrome secondary to cervical-carotid dissection with antiplatelet or anticoagulant medication can reduce the risk of stroke. Often, heparin bridged with warfarin to an INR of 2.0–3.0 for 3–6 months is used until the dissection resolves of its own accord, which occurs in 85–90% of cases [123]. Antiplatelet drugs such as aspirin or clopidogrel are also used frequently, either alone or in combination [132]. Between the two therapies, a recent randomized controlled trial of 250 patients with carotid or vertebral dissection determined that the recurrence of stroke on either therapy at 3 months was lower than found in previous studies, and that there was no significant difference between antiplatelet or anticoagulant medication in regards to the rare outcomes of recurrent ipsilateral stroke or death [132]. Patients who fail medical management of carotid artery dissection may be closely monitored or undergo surgical intervention, such as intravascular stenting [123].