Subclassification of Koos grade 4 vestibular schwannoma: insights into tumor morphology for predicting postoperative facial nerve function.

OBJECTIVE:Koos grade 4 vestibular schwannoma (KG4VS) is a large tumor that causes brainstem displacement and is generally considered a candidate for surgery. Few studies have examined the relationship between morphological differences in KG4VS other than tumor size and postoperative facial nerve function. The authors have developed a landmark-based subclassification of KG4VS that provides insights into the morphology of this tumor and can predict the risk of facial nerve injury during microsurgery. The aims of this study were to morphologically verify the validity of this subclassification and to clarify the relationship of the position of the center of the vestibular schwannoma within the cerebellopontine angle (CPA) cistern on preoperative MR images to postoperative facial nerve function in patients who underwent microsurgical resection of a vestibular schwannoma. METHODS:In this paper, the authors classified KG4VSs into two subtypes according to the position of the center of the KG4VS within the CPA cistern relative to the perpendicular bisector of the porus acusticus internus, which was the landmark for the subclassification. KG4VSs with ventral centers to the landmark were classified as type 4V, and those with dorsal centers as type 4D. The clinical impact of this subclassification on short- and long-term postoperative facial nerve function was analyzed. RESULTS:In this study, the authors retrospectively reviewed patients with vestibular schwannoma who were treated surgically via a retrosigmoid approach between January 2010 and March 2020. Of the 107 patients with KG4VS who met the inclusion criteria, 45 (42.1%) were classified as having type 4V (KG4VSs with centers ventral to the perpendicular bisector of the porous acusticus internus) and 62 (57.9%) as having type 4D (those with centers dorsal to the perpendicular bisector). Ventral extension to the perpendicular bisector of the porus acusticus internus was significantly greater in the type 4V group than in the type 4D group (p < 0.001), although there was no significant difference in the maximal ventrodorsal diameter. The rate of preservation of favorable facial nerve function (House-Brackmann grades I and II) was significantly lower in the type 4V group than in the type 4D group in terms of both short-term (46.7% vs 85.5%, p < 0.001) and long-term (82.9% vs 96.7%, p = 0.001) outcomes. Type 4V had a significantly negative impact on short-term (OR 7.67, 95% CI 2.90-20.3; p < 0.001) and long-term (OR 6.05, 95% CI 1.04-35.0; p = 0.045) facial nerve function after surgery when age, tumor size, and presence of a fundal fluid cap were taken into account. CONCLUSIONS:The authors have delineated two different morphological subtypes of KG4VS. This subclassification could predict short- and long-term facial nerve function after microsurgical resection of KG4VS via the retrosigmoid approach. The risk of postoperative facial palsy when attempting total resection is greater for type 4V than for type 4D. This classification into types 4V and 4D could help to predict the risk of facial nerve injury and generate more individualized surgical strategies for KG4VSs with better facial nerve outcomes.