Abstract
Objective. (1) Explain the need for an animal model to study intracranial injuries to the facial nerve. (2) Describe various techniques attempted to identify and crush the intracranial segment of the facial nerve in a rat model. (3) Describe in detail a successful rat model of intracranial facial nerve crush injury.
Study Design. Randomized controlled animal study.
Setting. Animal laboratory.
Subjects and Methods. Multiple attempts at surgical approaches to the cerebellopontine angle were attempted on cadaveric rats. Once a successful approach was derived, this was used on 19 live rats under anesthesia. Fourteen rats had a 1-minute facial nerve crush performed, and 5 had a sham surgery with complete surgical exposure of the facial nerve but no crush. Rats were followed for a 12-week duration evaluating immediate postoperative facial nerve function, complications, and survival.
Results. All 14 (100%) rats that underwent surgery with crush injury had complete facial paralysis postoperatively. Complete facial paralysis was defined as loss of eye-blink reflex, flat vibrissae, and lack of vibrissae movement. The 5 sham surgery rats had complete facial function postoperatively. Surgery was performed by 2 separate surgeons with no difference in outcome between the 2. Complications occurred in only 1 animal (1/19, 5.3%), which was a corneal abrasion requiring sacrifice.
Conclusion. Our group describes a consistent method for performing an intracranial crush injury in the rat. This new model and its applications in translational facial nerve research are promising, particularly with tumors or lesions at the cerebellopontine angle.
Keywords
Facial nerve preservation during vestibular schwannoma resection remains the most important consideration of patients undergoing microsurgery of the skull base. 1 The psychosocial ramifications of facial paresis or paralysis extend beyond the functional loss and inherent complications caused by injury to the facial nerve. 2 The facial nerve is particularly vulnerable and sensitive to injury at the cerebellopontine angle (CPA), with injury resulting from traction, hemorrhage, edema, vascular stasis, and ischemia.3,4 Although reestablishment of direct neural integrity through reanastamosis is the first-line therapy for transection injury, 4 there is no accepted treatment for less severe injury to the nerve.
Our proposed model uses a crush injury as opposed to an avulsion or transection injury. We chose this model because it gives a consistent injury that is reproducible and more clinically relevant.5,6 In addition, it eliminates the variability of recovery associated with a neurorrhaphy. An intracranial crush injury followed by treatment with nimodipine and an intracranial transection have been reported.7,8 Neither of these articles described the surgical approach in detail. In addition, no articles have discussed the challenges inherent to this surgery and the potential complications. Finally, no studies have evaluated the reproducibility of the crush injury vs a sham surgery to ensure that exposure of the nerve is possible without injury, thereby providing a true control group for study.
The purpose of this study is to present a rat model for intracranial facial nerve crush in detail. This article reviews the specific technique developed to obtain a successful approach to the adult male rat skull base to achieve consistent intracranial injury results. We also describe other techniques attempted but abandoned because of technical and anatomical limitations. This model is important to study potential treatments to speed recovery and improve final facial nerve outcome following injury to the facial nerve in the intracranial segment.
Materials and Methods
Male Sprague-Dawley rats were obtained from Harlan (Indianapolis, Indiana). Animals were maintained on a 12-hour light/dark cycle and given a standard rodent diet with water ad libitum. All surgical procedures were approved by the Institutional Animal Care and Use Committee and were performed in accordance with the National Institutes of Health guidelines.
An extended middle cranial fossa approach was carried out on 19 live animals. Each animal was anesthetized with an intraperitoneal injection of ketamine (100 mg/mL; 0.1 mL/100 g body weight) and xylazine (20 mg/mL; 0.025 mL/100 g body weight). Five of these animals were randomly determined to undergo sham surgery. Sham surgery was defined as complete exposure where the nerve was unroofed circumferentially but not crushed. In the surgical group of 14 animals, the entire surgical procedure was carried out including a 1-minute crush injury of the nerve. Immediately following surgery, all animals were allowed to recover from anesthesia before behavior analysis.
Attempted Approaches
Initial attempts involved removing the tympanic membrane and ossicles, allowing visualization of the facial nerve as it turned medially at the geniculate ganglion. Although this technique showed promise on cadaveric specimens, bleeding resulted in animal death. We attempted a technique similar to the final procedure, preserving the cochlea and eighth nerve, but the nerve injury results were inconsistent.
Final Operative Technique
Our final technique involved sacrifice of the eighth nerve and cochlea, allowing a more complete exposure of the facial nerve. The animal was positioned first with the head to the right, lying on its left side. A curvilinear incision was made on the shaved and prepped postauricular skin approximately 5 mm behind the ear ( Figure 1A ). The skin was elevated in the subdermal plane toward the ear canal. The extratemporal portion of the facial nerve was identified as it exited the stylomastoid foramen. The nuchal crest between the superficial temporal muscle anteriorly and the splenius capitis muscle posteriorly was identified ( Figure 1B ).
(Right = anterior, top = superior) (A) Postauricular incision. (B) Muscle exposure. (C) Skull exposure with craniotomy drawn. (D) Craniotomy performed. Crani, craniotomy site with dura exposed; dotted line, approximate area of initial craniotomy; FN-ET, facial nerve extratemporal; MP, mastoid process of the petrous temporal bone; NC, nuchal crest; OC, occipital squama; SpCM, splenius capitis muscle (r = resected); STM, superficial temporal muscle (r = resected); TS, transverse sinus.
Next, these 2 muscles were incised and dissected off the skull. The splenius capitis muscle was resected posteriorly to the skull base, and a small portion of the temporalis muscle was excised to provide improved exposure to the occipital squama and the mastoid process of the temporal bone ( Figure 1C ). The resected muscle was stored in saline to be used as a muscle plug at the end of the procedure to prevent cerebrospinal fluid leak from the craniotomy defect.
Through use of a high-speed drill with a 2-mm diamond burr under microscopic visualization, the nuchal crest was thinned to reveal the transverse sinus. Next, a craniotomy was made into the occipital squama until dura was seen, with the transverse sinus as the anterior limit ( Figure 1D ). The animal was then rotated 90 degrees with the head away from the surgeon. Using both epinephrine-soaked pellets and a small retractor, the parafloccular lobe of the cerebellum was retracted superiomedially. The drill was used to widen the craniotomy superiorly until the inferior cerebellar vein and mastoid cavity were identified. Posteriorly, the craniotomy was widened to the end of the occipital squama bone. This created a 3 × 3-mm craniotomy through which the petrous portion of the temporal bone was visualized ( Figure 2A ).
(Left = superior, top = anterior) (A) Identifying petrous ridge. (B) Exposing the eighth nerve. (C) Drilling anterior to the seventh nerve. (D) Unroofing the seventh nerve. Crani, craniotomy site; 8th Nerve, vestibulocochlear nerve (s = sacrificed); FN-ET, facial nerve extratemporal; FN-IC, facial nerve intracranial segment; PTB, petrous temporal bone; SpCMr, splenius capitis muscle (resected); STMr, superficial temporal muscle (resected); TS, transverse sinus.
With a 1-mm diamond burr, deeper dissection was performed to further define the petrous portion of the temporal bone. Staying posteriorly, the vestibulocochlear nerve was identified ( Figure 2B ) and sacrificed. The vestibulocochlear canal was followed laterally until the cochlea was identified and entered. This area was widened anteriorly until the posterior wall of the facial nerve was skeletonized. A thin portion of bone was left over the facial nerve to protect it until final unroofing. Drilling was then performed anterior to the facial nerve ( Figure 2C ) to allow room for forceps. This allowed for a circumferential view of the labyrinthine segment of the facial nerve.
Once the facial nerve was sufficiently skeletonized, a Rosen needle was used to remove the remaining bone from the surface of the nerve ( Figure 2D ). Because the nerve was anchored at the first genu, it was extremely susceptible to traction damage from microscopic movements during the crush injury. To solve this dilemma, we ensured that the nerve was mobile within the canal so that microscopic movements could be absorbed by movement of the brainstem. Curved jeweler’s forceps were then inserted into the craniotomy site, and the intracranial portion of the facial nerve between the brainstem and meatal foramen was crushed firmly with consistent pressure for 1 minute. Care was taken to avoid suction or manipulation of the nerve immediately after the crush to avoid transection.
A small muscle plug was inserted into the craniotomy site to prevent cerebrospinal fluid leak. A large muscle plug was placed over the craniotomy site to serve as a second layer of closure. The skin was closed in layers with suture, which was removed 10 to 14 days postoperatively.
Results
All crush animals (14/14; 100%) had evidence of complete facial nerve injury postoperatively. Complete facial paralysis was defined as complete lack of eye blink, flat vibrissae orientation, and absent vibrissae movement. 9 All sham animals (5/5; 100%) had normal facial nerve function postoperatively. All animals suffered expected vestibular loss with the sacrifice of the vestibulocochlear nerve on approach. This was evidenced by postoperative rotational motion and head tilt toward the side of injury. Rotational motion resolved slowly over the first postoperative week, but head tilt remained a permanent deficit.
There was no difference in surgical outcomes between 2 surgeons. Our animals were followed for at least 12 weeks postoperatively, and no signs of significant weight loss or physical distress were observed. We observed no evidence of cerebrospinal fluid leak, meningitis, stroke, or loss of any motor function besides the facial and vestibulocochlear nerves. The only complication (1/19; 5.3%) encountered was an apparent corneal abrasion and resulting infection that resulted in sacrifice of the animal.
Discussion
Intracranial facial nerve injury carries significant functional and psychosocial ramifications, necessitating further investigation. Multiple studies show continued advances in microsurgical techniques and outcomes compared to the early pioneers of skull base surgery.10-13 In 1 series of 71 large (3 cm or larger in diameter) vestibular schwannoma resections, 80% of patients demonstrated normal or near-normal facial function (House-Brackmann grade I or II) 6 months to 1 year after surgery, leaving 20% with moderate to severe facial nerve dysfunction. 12 Despite continued microsurgical advances, a significant percentage of patients will still suffer facial nerve dysfunction following surgery.
Both the compressive nature of CPA lesions, as well as their surgical dissection, can contribute to postoperative facial paresis/paralysis. Facial dysfunction can be a presenting symptom in up to 40% of patients, correlating most closely with the size and location of the lesion. 12 Multiple factors influence the ability to preserve the functional integrity of the facial nerve during vestibular schwannoma resection. These include tumor size, tumor type, prior surgery or irradiation, surgical approach, intraoperative facial nerve monitoring, and experience of the surgical team. 1 Manipulation of the intracranial segment of the facial nerve during CPA dissection can cause neuropraxia, axonotmesis, or neurotmesis neuronal-type injuries as well as disrupt local microcirculation. Because of its unique anatomy, the facial nerve lacks an epineurial protective layer during its intracranial pathway prior to entering the porous acousticus. 14 The nerve is therefore more susceptible to traction and/or compression injuries in this segment.
The surgical resection of CPA lesions leaves a significant portion of patients with facial paresis or paralysis. However, in a large study of 1000 cases of vestibular schwannoma resection, the facial nerve was severed in only 6% of cases, making this type of injury exceedingly rare. 13 Compression injury is distinct from transection injury, and we believe that the crush injury model is a more accurate approximation of common facial nerve injuries that can occur along the skull base both from tumor compression and dissection. Aside from the more clinical application of a crush injury model, this injury model is consistent and removes the variability of facial nerve reanastamosis after transection, which would not be possible in the rat because of limited exposure. 6 Our crush was performed for a total of 1 minute, which has been shown in previous models to produce a significant degree of nerve injury. 15 In our view, studies of facial nerve transection at the CPA are not as applicable to human nerve injury as those with a less severe nerve injury.
Prior peripheral nerve injury studies have shown minimal if any neuronal cell loss following a peripheral nerve crush injury.16,17 Our laboratory has shown by comparing extratemporal vs intratemporal facial nerve crush injuries that a more proximal injury results in greater neuronal cell body death. 6 These findings advocate the need for a more in-depth investigation of the intracranial segment of the facial nerve. Potentially, moving the injury even closer to the neuronal cell body would result in more cell body death. There are limited studies reported on both the approach to the intracranial facial nerve in rats and the effect and recovery of injury to this segment. In addition, there is a paucity of studies of nonsurgical treatments that could aid in recovery of the facial nerve after injury at the CPA. This research could have significant clinical ramifications.
The limitations of this study include the small sample size and difficulty in the surgical approach. Temporal bone anatomy in the rat is very different from the human, requiring a significant review of multiple resources prior to surgery. The surgical operative time, approximately 1.5 hours per case, required an additional dose of anesthesia, adding to the surgical risk, although no complications occurred in our study. This study was also completed in a single laboratory, which does not allow for proof of reproducibility among institutions.
After several attempts at exposure, the final successful operative approach was performed by 2 separate surgeons with similar results, displaying the surgical reproducibility within the same laboratory. All predetermined sham animals underwent complete unroofing of the facial nerve without crush injury. All of these animals awoke from anesthesia with full facial nerve function, verifying that drilling and exposure can be accomplished safely. It also demonstrates that the 1-minute crush injury is the only source of nerve damage. There were minimal complications from the procedure with no additional unexpected neurological sequelae. The only observed complication was related to vestibular injury ataxia, leading to a corneal abrasion and subsequent conjunctivitis from the animal’s bedding. After this encounter, all animals were treated more aggressively with ophthalmic lubrication postoperatively, and no further incidents occurred.
Conclusion
We believe that our operative approach is valuable for future studies evaluating intracranial facial nerve lesions and injury. This segment of facial nerve has significant clinical relevance to the fields of neurotology and neurological surgery. Although technically demanding, the results of our procedure are successful with acceptable rates of animal survival and minimal complication rates. Future directions of our lab include exploring new therapeutic methods of enhancing neuronal cell survival and increasing neural regeneration following intracranial nerve injury.
Author Contributions
Disclosures
Footnotes
Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.