The Validity of the Koos Classification System With Respect to Facial Nerve Function

BACKGROUND: The Koos classification of vestibular schwannomas is designed to stratify tumors based on extrameatal extension and compression of the brainstem. Our prior study demonstrated excellent reliability. No study has yet assessed its validity. OBJECTIVE: To present a retrospective study designed to assess the validity of the Koos grading system with respect to facial nerve function following treatment of 81 acoustic schwannomas. METHODS: We collected data retrospectively from 81 patients with acoustic schwannomas of various Koos grades who were treated with microsurgical resection or stereotactic radiosurgery. House-Brackmann (HB) scores were used to assess facial nerve function and obtained at various time points following treatment. We generated Spearman's rho and Kendall's tau correlation coefficients along with a logistic regression curve. RESULTS: We found no significant difference in the presence or absence of facial dysfunction by Koos classification when looking at all patients. There was a positive but fairly weak correlation between HB score and Koos classification, which was only significant at the first postoperative clinic appointment. There was a statistically significant difference in the presence or absence of facial dysfunction between patients treated with surgery vs radiation, which we expected. We found no statistically significant difference when comparing surgical approaches. Logistic regression modeling demonstrated a poor ability of the Koos grading system to predict facial nerve dysfunction following treatment. CONCLUSION: The Koos grading system did not predict the presence of absence of facial nerve dysfunction in our study population. There were trends within subgroups that require further exploration.