Traumatic Optic Neuropathy

Abstract

Traumatic optic neuropathy (TON) is a specific neurological sequence of traumatic brain injury (TBI). It has a different mechanism than other most neurologic complications of head trauma and its consequences can be devastating. The damage can be from direct penetrating trauma or bone fracture injuring the optic nerve directly or secondary to indirect blunt trauma (usually causing traction). The diagnosis of TON is based on the clinical history and examination findings indicative of optic neuropathy, especially the presence of defective pupillary light response. TON can cause only mild vision loss but, in some cases, severe vision loss is present. Imaging findings can support the diagnosis, and provide information on the mechanism as well as treatment options. The treatment options include observation alone, systemic steroids, erythropoietin, surgical decompression of the optic canal, or combination. The evidence base for these various treatment options is controversial and each treatment has its side effects and risks. Poor prognostic factors include poor visual acuity at presentation, loss of consciousness, no improvement in vision in the first 48 hours, and evidence of optic canal fractures on neuroimaging.

1. Introduction

Traumatic optic neuropathy (TON) is an acute injury of the optic nerve secondary to blunt or penetrating head injuries, in which dysfunction of the optic nerve, vision impairment, is caused secondary to direct or indirect trauma to the nerve. Traumatic optic neuropathy can be isolated but usually, it is part of more widespread head trauma. Historic studies report an incidence of 0.5–2% of head injuries [1, 2]. A recent national epidemiological survey of TON in the UK found a minimum prevalence in the general population of 1 in 1,000,000 [3].

TON might be present after vehicle accidents (car and bicycle), falls from heights, falling debris, assault, stab, and gunshot wounds (Figure 1). Iatrogenic injuries are also reported (mainly secondary to endoscopic sinus surgeries or suprasellar neurosurgeries) [3]. It is important to note that all patients with TON have head injuries and 66% of them have a significant head injury [4]. TON can be present after apparently otherwise mild injuries, but it is most common in the setting of craniofacial fractures [5].

Figure 1.

A 37-year-old female was admitted after gunshot wounds. Upon admission, she was intubated and unconscious. She had multi-compartment hemorrhages, parenchymal, subdural, and subarachnoid and secondary mass effect with midline displacement to the left (not shown). In plain radiograph (A) two metallic foreign bodies were seen. The bottom one is in the right maxillary sinus, with hyperdense small fragments in its track. The second one is adjacent to the right orbital apex. Lateral orbital wall fractures, as well as apical fractures, were noted (B + C). On the first ophthalmologic evaluation, she had pinpoint pupils (secondary to sedative agents) but still, a right RAPD was documented. It was not possible to have dilated fundus exam (DFE). She had no signs of orbital compartment syndrome (no proptosis, free eye movements, and normal intraocular pressure). Multidisciplinary consultation of neurosurgeon, oculoplastic surgeon and ENT, concluded that any surgical intervention aimed to remove the apical foreign body will have a high risk of bleeding and life-threatening complications—It was decided on conservative follow up. Endoscopic sinus surgery was done, removing the maxillary sinus foreign body. Upon regaining consciousness, the patient was evaluated again and had a vision of 20/30 and nonspecific visual field defects. DFE was unremarkable. On her 2 months follow-up, there was no RAPD, and vision improved to 20/20 with no visual field defects.

The vast majority of affected patients are young males (79–85%) in their early thirties. Children constitute a large portion as about 20% of patients are younger than 18 years [6]. In this age group, falls (50%) and motor vehicle accidents (40%) are the most common causes of TON.

TON is classified as direct when the nerve is injured directly by a projectile object that penetrates the orbit to damage the optic nerve, or indirect injury when it results from the non-penetrating effects of trauma.

The mechanism of TON is multifactorial. In the case of direct TON damage is caused directly to the nerve by laceration or impingement of the nerve from various causes, including penetrating a foreign body, displaced bone fragment, or optic canal fracture (Figure 1). In indirect trauma, compression forces from the superior orbital rim are transferred and concentrated in the orbital roof and optic canal, where the nerve is most vulnerable since it is fixed within the bony optic canal; Coup-contrecoup forces whip mobile portions of the optic nerve against fixed structures, causing injury [7, 8]. Shearing injury to the axons and microvasculature can also play a role, leading to necrosis [9, 10]. Violent rotation of the globe can also result in partial or complete optic nerve avulsion (Figure 2) [11]. Depending on the nature of the event, the shock wave can also fracture the optic canal and bone fragments can impinge on the nerve, [8] (occasionally being referred to as direct injury). Diffuse axonal injury is another mechanism that is thought to be involved in TON. As previously reported [12, 13], following head injury, axons of the brain white matter are deformed, swelling, cytoskeleton damage, and impaired axoplasmic transmission led to the disconnection of axons, regression, reorganization, and degeneration.

Figure 2.

A 39-year-old male suffered from severe trauma from a surfboard. On first evaluation, he had lacerations of the inferior and superior eyelids with limited eye movements in elevation, suppression and adduction. He had a dilated unresponsive pupil on the left and a vision of NLP in his left eye with RAPD+4. DFE revealed vitreous hemorrhage, and the optic nerve could not be identified in the posterior pole. Besides medial and inferior orbital wall fractures with down displacement into the maxillary sinus (B), the intraconal orbital fat appeared infiltrated and the left optic nerve was thickened and irregular in its course (A). Discontinuity of the posterior aspect of the globe, adjacent to the optic nerve head was noted. A diagnosis of optic nerve avulsion was made, and it was unfortunately impossible to regain vision in this case. Attempts were made to reconstruct the orbit by suturing the lacerations involving the medial cantal folds with excellent cosmetic results, which were important for a young patient.

Orbital compartment syndrome (OCS), in which acute severe bleeding is maintained within the orbit, is another specific subgroup of TON. The orbit is a confined, cone-shaped space, which is bound on all sides by bony walls, except anteriorly, where it is limited by the orbital septum and tarsal plates of the upper and lower eyelid. These structures have limited elasticity and thereby the orbit has limited compliance. Any increase in the orbital content, secondary to bleeding blood vessels, or trapped air, will result in an increase in intra-orbital pressure [14]. Besides the possibility of direct compressing on the nerve, it results in disturbing the orbital blood flow and thereby causes acute ischemic damage to the nerve [14, 15].

Post-injury, biochemical cascades exacerbate the initial damage and different treatment modalities are intended to limit this secondary injury. Most cases of visual loss are immediate; however, delayed visual loss is documented in 10% of cases.

2. Diagnosis

Since TON can develop even due to minor head trauma, it should be suspected if any evidence of impaired vision exists following head or facial trauma. In the conscious patient, a detailed history accounting for the mechanism of injury and previous visual status is mandatory. Visual function assessment and comprehensive eye examination should be carried out ruling out other causes of visual loss.

Clinical signs supporting the diagnosis of optic neuropathy include: 1) Impaired visual acuity: about 40–60% of patients present with only light perception vision, or worse [3, 16, 17], although, the visual acuity may range from normal to no light perception. Late deterioration of vision may occur secondary to intra-sheath hematoma and should raise again the diagnosis of TON [16]. 2) Relative afferent pupilar defect (RAPD), an asymmetrical pupillary response to light, which is a very specific sign of optic neuropathy, is very important in the assessment of TON patients; it may be the only subjective evidence at presentation in mild cases and more importantly in the unconscious patient and nonverbal children. A negative RAPD due to symmetric optic nerve injury should always be considered [2]. To note, in the settings of head trauma some patients may have dilated pupils which interferes with the pupillary examination. Alcohol, illicit drugs, narcotics, paralyzing agents, hypothermia, oculomotor nerve neuropathy, and sympathetic injury (Horner’s syndrome), can all interfere with pupillary testing [18]. 3) Color vision impairment indicating optic neuropathy.

Visual field defect with variable field defects is another helpful adjunct in the diagnosis and monitoring of TON patients. Unfortunately, in the acute setting, many trauma patients are unable to cooperate and have formal computerized visual field exams.

Visual evoked potentials (VEP) are another important tool in the diagnosis of TON in unconscious and nonverbal patients, providing evidence of visual pathway status and predicting the visual outcome. Patients with better VEP amplitudes have favorable visual outcomes [19].

The optic disc appearance in the early course of TON depends on the site of injury along with the optic nerve. When the injury is anterior to the entry site of the CRA a swollen optic disc with retinal hemorrhages is expected. However, in the majority of the patients, the disc appearance is normal since most of the cases have a more posterior injury. Late in the course of TON, a clinically evident optic atrophy usually develops within 4–6 weeks following the trauma. Kanamori et al. demonstrated that RNFL thinning and RGC complex loss began 2 weeks after trauma and plateaued at 20 weeks [20].

Radiologic studies (CT and MRI) may demonstrate bony fractures, optic sheath hematoma or intra-orbital air, and bleeding (orbital compartment). CT scan with coronal reconstruction images is an excellent imaging modality for demonstrating optic canal fractures [13, 14, 15, 21, 22, 23]. Multiplayer spiral computed tomographic (MSCT) with spatial stereo reconstruction is an advanced imaging tool that can better evaluate the optic canal status and diagnose combined injuries in other craniofacial tissues [24]. Nevertheless, the role of neuroimaging in the diagnosis of TON is still controversial since the majority of TON patients do not demonstrate relevant findings and the diagnosis can be made on clinical grounds only. However, in a patient with progressive visual loss, repeated neuroimaging is crucial for identifying surgical candidates and directing the surgical treatment.

3. Treatment

Currently, the most common treatments used for TON include 1) systemic steroids; 2) surgical optic nerve decompression; 3) a combination of steroids and surgical decompression. Nevertheless, the treatment of TON remains controversial as spontaneous visual improvement occurs in about 25–50% of TON patients [25, 26]. On the other hand, no medical or surgical treatment was proved until now with a clear advantage over observation only.

Systemic steroid therapy for TON has been extrapolated from the National Acute Spinal Cord Injury Study (NASCIS), a randomized controlled trial comparing steroid treatment to placebo for acute spinal injury, in which increased recovery of neurological function was seen in patients treated with methylprednisolone [27]. The rationale for their use in the treatment of TON is their anti-inflammatory and anti-oxidative effect. Typically, a regimen of very high (megadose) intravenous methylprednisolone is used to treat TON followed by oral steroid with tapering down. The International Optic Nerve Trauma Study (IONTS) is the largest comparative study comparing systemic steroids to surgical decompression and observation using different dosing and timing regimens failed to show any significant difference between the three groups [16]. In addition, there was no clear benefit for any timing or dosage regimen of corticosteroids on the final visual outcome. Lai et al. analyzed the risk factors for visual outcome in a small series of 20 TON patients with initial visual acuity of NLP and found that patients treated with methylprednisolone less than 24 hours from the injury showed better final visual acuity [28]. In another series, Sitaula et al. compared observation to oral prednisone (1 mg/kg for 7 days with 6 weeks taper) and high-dose intravenous methylprednisolone (1 g/d for 3 days followed by oral prednisone taper) found significant visual improvement in the high-dose group compared to the other two groups [29].

Other authors compared different regimes of steroids to observation only did not find any significant difference between their study groups [30, 31, 32, 33, 34]. Accounting for their side effects, in the lack of clear effects on visual recovery, steroids should be used judiciously by clinicians. Moreover, a large percentage of TON patients have a concomitant head injury; the CRASH study tested the effectiveness of corticosteroids following acute head injury was terminated prematurely due to an increased mortality rate in patients treated with steroids [35].

3.1 Systemic steroids

Surgical treatment for TON aims to reduce compression on the optic nerve exerted by edema, hematoma, or bony fragments. Currently, the main surgical approaches performed are medial transorbital external ethmoidectomy; transcranial surgery; and endoscopic transnasal surgery [36, 37, 38]. The transorbital ethmoidectomy and transcranial approach provide excellent surgical access to the optic canal, however, both are abandoned by many surgeons due to the high rate of complications and undesired cosmetic results. The endoscopic transnasal approach is more commonly used by surgeons. It provides adequate exposure for the medial bony wall of the optic canal through the sphenoid sinus with a less invasive procedure and more acceptable cosmetic results. However, it is a high-risk procedure due to the proximity of the internal carotid artery to the optic canal; and it should be done only by highly experienced surgeons in this kind of surgery. Several studies reporting the results of surgical decompression with or without steroids had been published in the literature with visual improvement rates ranging from 18–81% [39, 40, 41, 42, 43, 44, 45, 46]. Since primary reports failed to show obvious benefits for surgical decompression over observation, efforts are still made by researchers to optimize the visual outcomes following surgery. Reasonably, the authors presented the following indications for surgical treatment [23, 42, 43, 44, 45, 46, 47]: 1) history of traumatic head or face injury; 2) presence of hematoma or bony fragments compressing the optic verve; 3) poor response to initial medical therapy; 4) progressive visual loss not explained by other nontraumatic ocular pathology; 5) lack of evident damage to ocular tissue or intracranial optic nerve; 6) prolonger latency or reduction of amplitude in preoperative VEP scan. The optimal timing for surgery has been evaluated by several authors based on their own experiences. While some authors reported better visual outcomes in the early surgical intervention (<3 days) [48, 49, 50], others reported comparable visual improvement in the late surgical intervention (>7 days) [51]. Yu et al. compared immediate (within 3 days) to delayed (>3 days) optic canal decompression, and found that 73.5% of patients in the immediate surgical decompression group showed improved vision versus 46.9% in the delayed group [36]. Even though, the benefit of surgical decompression or the combination of surgical and medical treatment remains uncertain due to the lack of large comparative controlled studies. Moreover, several surgical complications, including CSF leak, infection, and bleeding, have been reported [37, 52, 53, 54]. Therefore, in the lack of clear evidence, the potential benefits and drawbacks of surgical treatment should be discussed in detail with the patient before surgery.

3.2 Surgical treatment

In cases suspected of OCS, the examination must be performed as soon as possible, so as not to delay treatment. The conscious patient will complain about severe pain and vision loss. The patient will usually have marked eyelid swelling, proptosis, chemosis, and even subconjunctival hemorrhage. Severe ophthalmoplegia will be noted and even digital ocular palpation will demonstrate resistance to retropulsion and a firm globe indicating an elevated IOP. CT may be helpful in establishing the diagnosis in milder cases where there is uncertainty and vision remains intact, but when the clinical findings are suggestive and vision is markedly impaired, treatment should not be postponed until after imaging is performed. In cases of OCS urgent surgical decompression is the mainstay of treatment. A bedside, lateral canthotomy and cantholysis (LC/C) is the first-line approach for reducing intra-orbital pressure [55]. Bony orbital decompression can be considered an adjuvant procedure if an adequate response is not achieved after LC/C [56]. In these cases, reviewing orbital imaging is important to locate the hematoma or other causative pathology (air, abscess), and to plane the surgical approach (endonasal medial wall decompression, anterior orbitotomy via an eyelid crease incision, or transcranial approach) [14].

3.3 Erythropoietin

In the last few years, erythropoietin was suggested as a potential treatment for TON due to its anti-inflammatory and antiapoptotic effects, based on studies of CNS trauma patients [57]. Primary studies reported better visual outcomes with EPO treatment [58, 59]. Recently, the TONTT, a phase 3, a large comparative study compared erythropoietin to steroids and observation in 100 TON patients [60]. All three study groups demonstrated significant visual improvement compared to the baseline BCVA. However, no significant difference was found between the study groups. Of note, color vision improvement was also observed in all three study groups even though it was significant only in the erythropoietin group.

3.4 Experimental treatments

In the last two decades, research is ongoing to develop new therapies aiming to encourage neuroprotection and axonal regeneration. Stem cell transplantation is gaining more progress in the treatment of optic nerve damage due to their multidirectional differentiation. In a mice model, stem cells transplanted in the subretinal space differentiated into photoreceptor and retinal cells [61].

Recently, a prospective single-center prospective phase 1 study, investigated mesenchymal stem cell (MSC) transplantation in 20 patients with traumatic optic neuropathy. Optic canal decompression with mesenchymal stem cell implantation compared to optic canal decompression alone. Both groups showed significant improvements in vision compared with the baseline; however, there was no statistically significant difference between the study groups [62].

Investigations on other potential therapies, including anti-TNF, brain-derived neurotrophic factor (BDNF), and RNA, aiming to reduce retinal ganglion cell loss and encouraging axonal regeneration are in progress [63, 64, 65].

Currently, no standard of care therapy exists in addressing TON.

4. Prognosis

Spontaneous visual recovery of about 40–60% has been reported in TON patients treated conservatively [9]. The final visual acuity following TON has a wide range from 20/20 to NLP [28, 29]. The baseline visual acuity is the most important prognostic factor for recovery since it reflects the degree of damage to the optic nerve. Patients with better visual function are presumed to have more functioning retinal ganglion cells, while patients with no residual vision and poor visual acuity at a presentation associated have less functioning retinal ganglion cells. Hence, better visual acuity at presentation predicts a better final visual outcome, on the contrary, patients with no residual vision (NLP) have lower final visual outcomes.

According to some reports, patients who present with NLP are unlikely to improve at all [29, 66, 67, 68], while other reports indicated some rate of improvement even in a patient with no residual vision at presentation [66]. In patients with residual vision, those with lower visual acuity at presentation sometimes show more visual improvement rates [68, 69, 70]. Other negative prognostic factors presented by different authors include lack of improvement within the first 48 hours, optic canal fracture, absence of VEP responses, loss of consciousness, higher degrees of RAPD, age over 40 years, intra-sheath hematoma, and blood within the posterior ethmoidal cells [19, 71, 72, 73].

In cases of OCS, if treated within 2 hours, most patients will achieve a final visual acuity better than 20/40, though approximately 15% will be worse. Patients treated after 2 hours have poorer reported outcomes. In the case of delayed presentation, considering orbital decompression is still reasonable since there are reports of visual recovery even after delayed intervention and even with no decompression at all [74].

It is important to emphasize that recovery of vision after any kind of treatment modality mentioned, is not always immediate. There may be an ongoing improvement in VA for a few weeks post-intervention. Furthermore, most reports published provide a limited follow-up period and it is reasonable to deduce that long-term follow-up may show better outcomes. It should also be noted that even in cases of good VA acuity, some patients suffer from severe visual field defects.