Combined Predictive Model for Endoscopic Third Ventriculostomy Success in Adults and Children

Abstract

The selection of patients in whom endoscopic third ventriculostomy (ETV) can be effective remains poorly defined. The endoscopic third ventriculostomy success score (ETVSS) and the presence of bowing of the third ventricle have been identified as independent factors for predicting success, each with limitations. The objective of this study is to elaborate a combined predictive model to predict ETV success in a mixed cohort of patients. Demographic, intraoperative, postoperative, and radiological variables were analyzed in all ventriculostomies performed consecutively at a single institution from December 2004 to December 2022. Qualitative and quantitative measurements of preoperative, immediate and late postoperative MRI were conducted. Univariate analysis and logistic regression models were performed. 118 ETV were performed in the selected period. 106 procedures met inclusion criteria. The overall success rate was 71.7%, with a median follow-up of 3.64 years [1.06;5.62]. The median age was 36.1 years [11.7;53.5]. 35.84% were children (median=7.81 years). Among the 80 patients with third ventricle bowing, success rate was 88.8% (p<0.001). Larger third ventricle dimensions on preoperative mid-sagittal MRI were associated with increased ETV success. The model with the best receiver operating characteristic (ROC) curves, with an area under the curve (AUC) of 0.918 (95% CI 0.856;0.979) includes sex, ETVSS, presence of complications and third ventricle bowing. The presence of bowing of the third ventricle is strongly associated with a higher ETV success rate. However, a combined predictive model that integrates it with the ETVSS is the most appropriate approach for selecting patients for ETV.