Predictive model of resection in endoscopic endonasal approach for pituitary adenomas based on anatomical limits

The aim of this project is to study several anatomical-radiological features of pituitary adenomas obtained from preoperative radiological images and to analyze their relationship with the extent of resection achieved through the endoscopic endonasal approach. The second objective was to create a prediction model of the extent of resection.

We retrospectively evaluated 105 patients. Tumor volume, Knosp grade, suprasellar-diaphragm coefficient and invasion of the posterior compartment have been analyzed. The extent of resection was assessed by analyzing the postoperative magnetic resonance. We created the predictive scale using statistically independent variables.

When each of the variables has been studied individually, a statistically significant value of all of them is appreciated to obtain a complete resection. However, only the Knosp grade and the suprasellar-diaphragm coefficient had a statistically significant value as independent variables. The sum of the Odds Ratio obtained from the Knosp scale, and the suprasellar-diaphragm coefficient gives the probability of complete resection. A new set of cases was employed to validate the scale.

The cavernous sinus invasion and the newly designed suprasellar diaphragm coefficient are directly related to the extent of resection in pituitary adenoma surgery performed by a transellar endoscopic approach. Moreover, based on both radiologic factors, a predictive scale may predict the probability of complete resection in a series of patients.