Transsphenoidal surgery: accuracy of an image-guided neuronavigation system to approach the pituitary fossa (sella turcica).

Abstract

To determine the accuracy of locating the pituitary fossa with the Brainsight neuronavigation system by determining the mean target error of the rostral (tuberculum sellae) and caudal (dorsum sellae) margins of the pituitary fossa. Experimental cadaveric study. Ten canine cadavers. Computed tomography (CT) and MRI were performed on each cadaver with fiducials in place. Images were saved to the neuronavigation computer and used to plan the drilling approach. The cadavers were placed in the surgical head clamp of the Brainsight system and positioned for a transsphenoidal approach. On the basis of the planning, 2 localization points were drilled, 1 each at the rostral and caudal margins of the pituitary fossa, and CT was repeated. Error was assessed from the difference in millimeters between the targets identified during Brainsight planning and the actual location of the 2 points drilled on each cadaver skull as identified by postdrilling CT. The rostral and caudal margins of the pituitary fossa provided 2 target points per cadaver. The median target error (interquartile range) for all target sites (n = 20) was 3.533 mm (range, 2.013-4.745). This stereotactic system allowed the surgeon to locate the rostral and caudal margins of the pituitary fossa with clinically acceptable accuracy and confidence. Using the Brainsight neuronavigation system for localization during transsphenoidal hypophysectomy may decrease morbidity and surgical time.