Adaptive evaluation of gross total resection rates for endoscopic endonasal approach based on preoperative MRI morphological features of pituitary adenomas.

Objective: This study aims to define a set of related anatomical landmarks based on preoperative Magnetic Resonance Imaging (MRI) of patients with pituitary adenomas (PAs). It explores the impact of the dynamic relationships between different anatomical landmarks and the tumor on the resection rate and tumor progression/recurrence during the endoscopic endonasal approach (EEA).

Methods: A single-center institutional database review was conducted, identifying patients with PAs treated with EEA from December 2018 to January 2023. Clinical data were reviewed, and anatomical landmarks were categorized into two regions: the suprasellar region and the cavernous sinus region. Following basic statistical and univariate logistic regression analyses, patients were randomly divided into training and validation sets. A nomogram was then established through the integration of least absolute shrinkage and selection operator (LASSO) regression and multivariable logistic regression analysis. The clinical prediction model was evaluated using the area under the receiver operating characteristic curve (AUC), calibration curves, and decision curve analysis. Kaplan-Meier curves were plotted for survival analysis.

Results: A total of 626 patients with PAs were included in the study, with gross total resection (GTR) achieved in 570 cases (91.05%). Significant differences were observed in the distribution of age, Knosp grade, and tumor size between the GTR and near total resection (NTR) groups. LASSO regression identified 8 key anatomical landmarks. The resulting model demonstrated an AUC of 0.96 in both the training and validation sets. Calibration curves indicated a strong agreement between the nomogram model and actual observations. Survival analysis revealed that the extent of resection (EOR), age, Knosp grade, tumor size, and PAs extending beyond several anatomical landmarks identified were significantly associated with the progression or recurrence of PAs.

Conclusion: This study proposes a model for adaptively assessing the resection rate of PAs by delineating relevant anatomical landmarks. The model comprehensively considers instrument manipulation angles, surgical accessibility during EEA procedures, anatomical variations, and the displacement of related anatomical structures in pathological states. This approach can assist neurosurgeons in preoperative planning and developing personalized surgical strategies.