HSC Section 8_April 2017

R. B. CANNON ET AL.

and many studies document the safety of this procedure (22,34–36). This study aims to describe the outcomes of early surgical decompression via the MF approach and repair for traumatic facial nerve paralysis in patients with poor facial nerve prognosis in an unmatched series of patients. We also report the long-term postoperative outcomes of the MF facial nerve decompression with and without nerve grafting, particularly in regards to long- term facial function outcomes, safety, and hearing results. With Institutional Review Board approval, we searched our database for patients who underwent a MF craniotomy for the diagnosis of temporal bone fracture and facial paralysis. Patients with temporal bone fractures and facial paralysis treated by other surgical approaches were not included. The medical records of 23 consecutive patients treated from 1996 to 2014 at the University of Utah Hospital were retrospectively reviewed. There were five patients who were lost to follow-up within 1 year after surgery or did not meet the inclusion criteria and were excluded. All patients had immediate-onset, com- plete facial paralysis (HB VI), underwent a computed tomo- graphic (CT) scan, and then underwent ENoG testing. If the patients had > 90% degeneration on ENoG testing, no volun- tary EMG motor unit potentials, and presented within 14 days of the traumatic facial nerve paralysis, then surgical decom- pression via the MF approach with possible facial nerve repair was recommended. For patients electing to have surgical decompression, a standard MF approach to the internal auditory canal (IAC) was performed. The facial nerve was identified in the lateral IAC, exposed from the labyrinthine segment and meatal fora- men to the geniculate ganglion and a portion of the tympanic segment. The temporal bone fracture was identified, explored, and the facial nerve was fully exposed on either side of the fracture. For decompression, the fibrous ligament at the meatal foramen of the fallopian canal and exposed nerve sheath (epineurium) was incised. For irreversible injuries, nerve graft- ing was performed. A mastoidectomy was performed if exposure of the descending segment of the facial nerve was necessary. The nerve was cut back to healthy ends and then a great auricular nerve graft, or in one case a sural nerve graft, was harvested, trimmed, reversed, and placed into position spanning the gap between the ends. A sutureless repair was performed if feasible. The patient’s clinical course was followed, which included: HB grade, pre- and postoperative hearing results with a 4- frequency pure-tone average (PTA) using 0.5, 1, 2, and 3 kHz and word recognition score (WRS), and both immediate post- operative and long-term complications. Final HB grade was assessed at the clinic visit at 1 year after surgery, and the final hearing result was assessed at a clinic visit between 3 months and 1 year after surgery. Hearing outcomes compared the preoperative bone PTA and the postoperative air PTA because all patients had a significant preoperative conductive hearing loss and, thus, preoperative inner ear function was compared with their postoperative hearing result. The results were com- piled and statistically analyzed with the chi-square test and regression analysis: including p values, a best-fit linear trend- line, and R 2 value. The House-Brackmann (HB) facial nerve grading system is an ordinal variable and appropriate ordinal MATERIALS AND METHODS

all patients regaining normal or near normal facial func- tion (HB I or II) (3,11,14). Patients with immediate-onset complete facial paralysis (HB VI) that persists, show > 90% degeneration on ENoG, and have absent volitional nerve activity on EMG, have a poor chance of recovery if observed or managed medically (15). The site of facial nerve injury in patients with traumatic paralysis after a temporal bone fracture is peri-geniculate in approximately 90% of cases (16,17). In addition, the meatal foramen and labyrinthine seg- ment have been shown to be the narrowest portion of the bony facial canal and common sites of electric conduction blockage due to significant edema or bone fragments, which can result in constriction and injury of the intratemporal facial nerve (18,19). Therefore, any surgical intervention should address these sites. When hearing is intact, the middle fossa (MF) approach provides optimal access the meatal foramen, labyrinthine segment, and peri-geniculate areas (20,21). Alternatively, if irreversible hearing loss has occurred after trauma, a translabyrinthine approach may be used. Published rates of facial nerve recovery vary between surgical intervention and observation, (3,9,10,22–29) and the evidence for the role of surgery for traumatic facial nerve paralysis was recently found to be inconclusive by a systematic review (4). The current body of literature is heterogeneous and lacks standardization, with incomplete data reporting, vary- ing surgical criteria, and erratic electrical testing. Length of follow-up and type of surgery performed are also important variables that vary significantly in the currently published studies. Understanding of the natural history of traumatic paralysis in patients with poor prognosis after clinical evaluation and electrical testing is limited; however, designing a study with matched controls would be difficult to conduct ethi- cally because when a patient reaches the electrodiag- nostic criteria in the setting of trauma, it would be unethical to withhold treatment. In this situation, these patients are at high-risk of complete nerve section that will likely not improve without intervention, therefore, surgical treatment is appropriate and necessary, but the effectiveness of MF decompression and repair is not fully defined. Timing for surgical decompression is controversial. Several studies indicate surgery should be performed within 6 days to 2 weeks of the trauma (10,30–32), whereas, others suggest that good facial nerve out- comes can be achieved regardless of the timing of surgery, as long as it is performed within the first 3 months (22,31). In certain cases in which damage to the nerve is irreversible, nerve grafting with an autologous donor nerve may be performed in conjunction with decompression surgery. Many recipients of nerve graft- ing can recovery facial nerve function of a HB III, but this is the best reported HB score obtainable (33). For patients undergoing MF surgery, preserving hearing and perioperative morbidity is a significant concern,

Otology & Neurotology, Vol. 37, No. 6, 2016

191