Accuracy and Interrater Reliability of CISS Versus Contrast-Enhanced T1-Weighted VIBE for the Presence of Optic Canal Invasion in Tuberculum Sellae Meningiomas

Abstract

The magnetic resonance imaging sequence used to assess optic canal invasion by tuberculum sella meningiomas (TSMs) has not been standardized. Both constructive interference in steady state (CISS) and contrast-enhanced T1-weighted volume-interpolated breath-hold examination (VIBE) sequences are frequently used. The aim of the present study was to compare the accuracy and interrater reliability of these sequences in predicting optic canal invasion by TSMs. In the present retrospective study of 27 patients (54 optic canals) who had undergone endoscopic transtuberculum transplanum resection of TSMs, images from preoperative CISS and contrast-enhanced T1-weighted VIBE sequences were assessed by 5 neuroradiologists who were unaware of the operative findings. The readers evaluated the optic canal in 4 quadrants at 2 locations (the posterior tip of the anterior clinoid process and the optic strut). A quadrant was considered positive for tumor invasion if invasion was present at either of these 2 locations. The reference standard was intraoperative observation of gross optic canal invasion. The interrater agreement was good for the presence or absence of tumor involvement in a particular quadrant (CISS, 0.635; VIBE, 0.643; 95% confidence interval for the difference,-0.086 to 0.010). The mean sensitivity and specificity for optic nerve invasion were 0.643 and 0.438 with CISS and 0.643 and 0.454 with VIBE, respectively. No significant differences were seen between the sequences in terms of reader accuracy when the intraoperative findings were used as the reference standard. CISS and VIBE sequences both have good accuracy in predicting for optic canal tumor invasion by TMEs.