In the current era of vestibular schwannoma surgery, the rate-limiting step to gross total resection, especially in tumors ≥3.0 centimeters, is the course and integrity of the facial nerve.
Even with a combination of meticulous microsurgical technique and intraoperative electromyography, injury to the facial nerve can still occur.
In instances in which the nerve is anatomically intact, reanimation with anastomotic techniques should be deferred in favor of a course of postoperative observation to see how much function is regained.
Since its inception, surgery for vestibular schwannoma (VS) has striven for a balance between complete tumor removal and satisfactory patient outcomes. At the turn of the 20th century, mortality was the major obstacle to extirpation, but as the era and surgical technique progressed, mortality rates declined precipitously, and reducing loss of function, especially with regard to the facial nerve, became the focus of improvement. In the current era of VS surgery, the rate-limiting step to gross total resection, especially in tumors ≥3.0 centimeters, is the course and integrity of the facial nerve. Even with a combination of meticulous microsurgical technique and intraoperative electromyography (EMG), injury to the facial nerve can still occur. We present a case of facial nerve injury and discuss anatomic factors that lead to such an injury, as well as management strategies to take when it occurs.
A 40-year-old female presented with right-sided hearing loss, imbalance, and tinnitus, all of which had grown progressively worse over the preceding year. An audiogram showed sensorineural hearing loss on the right with a pure-tone average of 55 decibels and a word recognition score of 25%. An MRI revealed an approximately 3.1-cm tumor in the right cerebellopontine angle consistent with VS. Given the size and age of the patient, treatment was recommended with microsurgical resection through a retrosigmoid craniotomy. She underwent surgery 2 months after presentation.
The patient was placed under general anesthesia with endotracheal intubation. Intraoperative monitoring of cranial nerves V, VII, VIII, X, and XI was utilized. After positioning in a left lateral decubitus position, a semilunar incision was marked out approximately three fingerbreadths medial to the digastric notch. The scalp was incised and moved laterally while the nuchal musculature was mobilized inferiorly. A burr hole was made over the junction of the transverse and sigmoid sinus, and a bone flap was turned. The dura was opened in a semilunar fashion. The ninth, tenth, and eleventh cranial nerves were identified inferiorly as well as the inferior pole of the tumor, and the arachnoid over the tumor was released from inferior to superior. The posterior pole of the tumor was stimulated at 3.0 milliamps without response from the facial nerve. The tumor was then internally debulked and rolled superiorly. The eighth cranial nerve fibers were identified and cut because they were splayed out over the posterior portion of the tumor, which allowed visualization of cranial nerve VII at the brainstem. Further internal debulking of the tumor was performed, and as the tumor capsule was being mobilized from the lateral aspect of the cerebellum and brainstem, a significant amount of venous bleeding occurred from a large venous tributary. It was thought to be away from the course of the facial nerve and was controlled with bipolar cautery. Shortly after, a conduction block in the facial nerve occurred. The remaining tumor in the cerebellopontine angle was removed. The facial nerve was taking a course over the superior pole of the tumor before entering the acoustic meatus. Distal to the conduction block, the nerve stimulated at 0.2 milliamps. The posterior part of the acoustic meatus was drilled away, and the remaining portion of the tumor was removed for a gross total resection. Careful inspection showed that the facial nerve was intact but that it would not conduct proximal to the area of the bipolar injury. A 3-mm nerve support supplement tube was placed over this region and coated in dural sealant. The dura was closed primarily, the bone flap replaced, and the wound closed in anatomic layers.
Postoperatively, the patient had House-Brackmann VI right facial weakness and some mild right V3 numbness. She was discharged home on postoperative day 4. At 3 months postoperatively, the patient continued to have House-Brackmann VI right facial weakness, and an MRI showed gross total resection of the tumor. At 1 year postoperatively, the patient had improved to House-Brackmann IV right facial weakness. Her improvement started about 7 months from surgery with the right corner of her mouth and then with the right eye at about 9 months. She also noticed altered taste sensation and some watering of her eye with olfactory stimulation. Two and a half years from surgery, she continued with House-Brackmann IV right facial weakness, and MRI showed no evidence of recurrent tumor.
The early history of VS surgery was devoted to balancing patient operative survival and extent of resection. In 1913, using finger dissection, some of the earliest case series of VS removal reported operative mortality rates between 67% and 84%. Finding this unacceptably high, Cushing endorsed internal debulking of VS without removing the capsule and reported a mortality rate of 10% to 15%, though this increased to 54% at 5 years secondary to tumor recurrence. Dandy further refined VS surgery with intratumoral decompression followed by extracapsular dissection with an operative mortality rate of 10.9% in 1941. Not long after, using a similar technique, Horrax and Poppen reported a 5-year mortality rate of 12.7%.
While mortality rates declined, morbidity rates remained high due to cranial nerve injury, particularly of the facial nerve. Although the first successful gross total resection of a VS with preservation of the facial nerve was reported by Cairns in 1931, it could not be accomplished on a routine basis; it was not until the microsurgical revolution led by House, Hitselberger, Kurze, and Yasargil in the 1960s and the introduction of intraoperative EMG monitoring in 1979 that facial nerve preservation became routinely achievable. Currently, rates of mortality are less than 1%, and facial nerve dysfunction rates are between 3% and 43%, depending on tumor size.
Even with these advances, sometimes the surgeon and patient must choose between complete tumor resection and risk of permanent facial nerve dysfunction. Factors that influence this are tumor size, the degree to which the nerve fibers are spread out by the tumor, and the location of the facial nerve relative to the tumor. The most common location of the facial nerve is anterior to the tumor in over 75% of patients, which means the tumor bulk is posterior to the facial nerve and the nerve is not encountered until most of the tumor is removed. In addition, an anterior course of the nerve means that it does not have to be protected while dissecting the tumor away from the medial interface of the tumor and cerebellum. When the course of the facial nerve is on the posterior surface of the tumor, it is encountered earlier in the surgery and must be protected while removing the bulk of the tumor.
In the presented case, the facial nerve was taking a course over the superomedial portion of the tumor that put it in close relationship to two large veins: the veins of the cerebellopontine fissure and the middle cerebellar peduncle. This course is unusual and present in approximately 1% of tumors. After debulking the tumor, as the tumor capsule was being separated from the cerebellum and brainstem, heavy bleeding was encountered from one of these two aforementioned veins. It is very common in our experience that these veins become quite engorged with even medium-sized VS as they and the adjacent brainstem get compressed by the tumor, and the tumor may drain into these veins, increasing their flow and pressure. Despite the course of the facial nerve being known, inadvertent injury to the facial nerve with the bipolar occurred, resulting in a proximal conduction block. The rest of the tumor was safely removed, resulting in complete tumor resection and an anatomically intact, but nonfunctioning, nerve.
In instances where the facial nerve is completely severed during surgery, there is no question that reconstruction with an end-to-end anastomosis or interposition graft should be attempted during or shortly after surgery. However, there is some debate regarding an anatomically intact but electrically diminished or unresponsive nerve, with some endorsing resection of the damaged section and immediate reconstruction and others delaying reinnervation for at least 1 year to determine whether the nerve recovers. In our institutional series of 11 patients with anatomically intact but nonstimulating nerves, we found that 64% of patients improved to at least House-Brackmann grade III in less than 1 year from surgery. Because these results are on par or better than those seen with reinnervation, we do not perform immediate facial nerve reconstruction of any type until a year from surgery if the nerve is anatomically contiguous.