439 The ANS-7 Score: An Externally Validated Seven-Point System Predicting Poor Facial Nerve Outcome Following Resection of Vestibular Schwannomas

View at publisher Summarise with VIRAT
Authors: James Feghali, Sai Chandan Reddy, Ali Karim Ahmed, Melissa Canales, Shaan Bhandarkar, Patrick J. Kramer, Joe Ping Yiu Ha, Justin Liu, Lily H. Kim, Adam S. Vesole, Firas A. Houssein, Deepa J. Galaiya
+18 more Bryan Ward, Kofi D.O. Boahene, Jason C. Nellis, Wade W. Chien, Charles C. Della Santina, Howard Francis, Francis X. Creighton, Matthew Stewart, John Carey, Risheng Xu, Justin M. Caplan, Chetan Bettegowda, JON D. WEINGART, Henry Brem, Rafael J. Tamargo, Daniel Sun, Michael K. Lim, Christopher M. Jackson
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

INTRODUCTION: No externally validated predictive model for facial nerve injury after vestibular schwannoma (VS) resection exists, despite the utility of such a tool in patient counselling and early referral to facial reanimation surgery. METHODS: We identified patients who underwent microsurgical resection of VS via a retrosigmoid approach at Johns Hopkins Hospital between July 2016 and April 2024. Preoperative characteristics and intraoperative facial nerve stimulation properties were recorded. Following univariable analysis, a stepwise multivariable logistic regression model predicting poor facial nerve outcome (HB >2) at last follow-up was derived and used to create a scoring system. Predictive metrics were compared to the previously published facial nerve outcome score (FNOS). The model was also validated in two separate tertiary referral centers at University of Cincinnati and Stanford Health. RESULTS: There were 360 VS patients who underwent microsurgical resection with available follow-up data. There were 80 (22%) patients with poor facial nerve outcome (HB >2) at last follow-up (3.1 ± 2.3 years). Gross total resection was achieved in 304(84%) patients. the optimal logistic regression model included abnormal renal function (OR = 8.3, 95% CI [1.4-48.5], p < 0.001), neuroma size ≥2.2 cm (OR = 3.3, [1.7-6.6], p < 0.001), and stimulation threshold post-resection (weak stimulation OR 7.1, [3.1-16.2], p < 0.001). the parameter estimates were used to construct the ANS-7 score with patients demonstrating increased probability of poor outcome with increasing score values. the score outperformed FNOS in our cohort (AUC = 0.807 [0.742-0.872] vs FNOS AUC = 0.696 [0.622-0.771]). Upon external validation, AUCs of 0.713 [0.555-0.870] and 0.729 [0.547-0.911] were derived. A calculator encompassing the model was deployed at https://facialnervefunction.shinyapps.io/Facial_Nerve_Function_Calculator/. CONCLUSIONS: The relatively simple ANS-7 scoring system accurately predicts poor facial nerve outcome with reasonable external validity. The model can be used for patient counseling and for identifying high-risk patients for facial weakness.

Fields of Study & Subjects

Meningioma and schwannoma management Facial Nerve Paralysis Treatment and Research Neurofibromatosis and Schwannoma Cases

Publication details

Journal / source
Neurosurgery
Volume
72
Issue
Supplement_1
Pages
126–127
Publication date
26 March 2026
Language
EN