Prediction of vestibular schwannoma growth: a novel rule based on clinical symptomatology.

ABSTRACT

OBJECTIVES: The aim of this study was to formulate a predictive rule for vestibular schwannoma growth during the initial observation period after diagnosis. METHODS: Logistic regression models were fitted, with tumor growth in the first year as the dependent variable and patient characteristics as the independent variables. Backward selection was used to eliminate superfluous predictors. The area under the receiver operating characteristic curve was taken as a measure of the model's discriminative power. RESULTS: Eventually, the model or rule consisted of 4 significant growth predictors localization (if extrameatal, +1; if intrameatal, 0), sudden sensorineural hearing loss (if present, -1; if absent, 0), balance symptoms (if present, +1; if absent, 0), and complaints of hearing loss for less than 2 years (if present, +1; if absent, or present for more than 2 years, 0). A higher score indicates a higher likelihood of tumor growth during the period of observation after diagnosis. If the total score is 0 or less, the likelihood of tumor growth during the first year after diagnosis is less than 10%. If the score is 3, the likelihood of growth during the first year after diagnosis is more than 70%. CONCLUSIONS: We were able to create a useful rule to predict vestibular schwannoma growth during the first year after diagnosis.