Independent Validation of the Colloid Cyst Risk Score to Predict Symptoms and Hydrocephalus in Patients with Colloid Cysts of the Third Ventricle

Abstract

The Colloid Cyst Risk Score (CCRS) was devised to identify patients with symptomatic colloid cyst and stratify risk of hydrocephalus. The CCRS considers patient age, presence of headache, colloid cyst diameter, fluid-attenuated inversion recovery hyperintensity, and location within the third ventricle. The purpose of this study was to independently evaluate the validity of the CCRS. Patients with a colloid cyst of the third ventricle were identified retrospectively from institutional billing records and radiology report archives. Patients without a confirmed diagnosis of colloid cyst of the third ventricle or magnetic resonance imaging of the brain were excluded. Data were collected via retrospective chart review. One hundred and fifty-six patients met inclusion and exclusion criteria. In our cohort, the CCRS stratified symptomatic patients and patients with hydrocephalus across all scores (P < 0.001). From CCRS 2 to 5, the percentage of symptomatic patients increased from 13% to 100%, whereas the percentage of patients with hydrocephalus increased from 8% to 83%. Simple logistic regression showed that total CCRS, headache, axial diameter, fluid-attenuated inversion recovery hyperintensity, and risk zone were all highly predictive of symptomatic status and hydrocephalus (P < 0.001). Logistic regression with receiver operating curves for the CCRS showed an area under the curve of 0.914 for symptomatic colloid cysts and an area under the curve of 0.892 for colloid cysts with hydrocephalus. Our data analysis validates the predictive value of the CCRS for both symptomatic status and hydrocephalus and supports the use of the CCRS in risk stratification and clinical decision making.