Hypoglossal Nerve Transfer for Facial Nerve Paralysis: A Systematic Review and Meta-Analysis

Background: Hypoglossal-facial nerve (12-7) anastomosis can restore symmetry and voluntary movement on the face in patients with facial nerve paralysis. Traditional 12-7 transfer includes direct end-to-end nerve anastomosis, sacrificing the entire hypoglossal nerve. Contemporary, end-to-side anastomosis, or split anastomosis techniques limit tongue morbidity by preserving some hypoglossal nerve. Direct outcome comparisons between these techniques are limited. Objective: To compare reported outcomes of facial movement, tongue, speech, and swallow outcomes among the different types of hypoglossal-facial nerve anastomosis schemes. Evidence Review: For this systematic review and meta-analysis, a comprehensive strategy was designed to search PubMed, Scopus, and the Cochrane Database from inception to January 2021, in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis, reporting guideline yielding 383 results. Any participant who underwent 12-7 transfer using any of the three techniques, with or without an interposition graft, and had documented preoperative and postoperative evaluation of facial nerve function with a validated instrument such as House-Brackmann (HB), was considered for inclusion. Secondary outcomes of synkinesis, tongue atrophy, and speech or swallowing dysfunction were also compared. Forty-nine studies met inclusion criteria, representing data from 961 total patients who underwent 12-7 transfer. Results: The proportion of good HB outcomes (HB I-III) did not differ by anastomosis type: End-to-side and end-to-end anastomosis (73% vs. 59%, p = 0.07), split and end-to-end anastomosis (62% vs. 59%, p = 0.88), and end-to-side anastomosis and split anastomosis (73% vs. 62%, p = 0.46). There was no difference in reported synkinesis rates between the anastomosis types. However, end-to-side anastomosis (z = 6.55, p < 0.01) and split anastomosis (z = 3.58, p < 0.01) developed less tongue atrophy than end-to-end anastomosis. End-to-side anastomosis had less speech/swallowing dysfunction than end-to-end anastomosis (z = 3.21, p < 0.01). Conclusion: End-to-side and split anastomoses result in similar HB facial nerve outcomes as the traditional end-to-end 12-7 anastomosis. End-to-side anastomosis has decreased complications of tongue atrophy and speech/swallow dysfunction compared to end-to-end anastomosis. In addition, split anastomosis has decreased rates of tongue atrophy compared to end-to-end anastomosis.