Sources of residuals after endoscopic transsphenoidal surgery for large and giant pituitary adenomas.

BACKGROUND Giant and large pituitary adenomas (PA) constitute a specific subset of PAs, with gross total resection (GTR) rates frequently not exceeding 50%. Both an anatomical inaccessibility and an inadequate tumor visualization are thought to play a role. This study analyzes risk factors for postoperative residuals after endoscopic transsphenoidal pituitary surgery for large and giant pituitary adenomas. METHODS A retrospective analysis of patients with giant and large PA operated between 2015 and 2018 was performed. RESULTS Forty patients (13 females, 27 males) were included in the analysis (30 large and 10 giant PAs). The mean MRI follow-up time was 5.9 ± 6.54 months. Overall, GTR was achieved in 29 patients (72.5%), subtotal resection in 9 (22.5%), and the inconclusive result was in 2 (5%). Unexpected residuals represented 7 (77.7%) of all 9 residual tumors. The most frequent intraoperative factor associated with unexpected residual tumors was improper identification of residual tumor due to obstruction of view in 2 (28.5%) cases and inability to distinguish normal tissue from tumor in the other two (28.5%). Sub-analysis based on tumor size revealed that with large PAs, GTR was achieved in 25 (83.3%), STR in 4 (13.3%), and inconclusive in 1 (3.3%) patient. In patients with giant PAs, GTR was achieved in 4 (40%), STR in 5 (50%), and inconclusive in 1 (10%). Analysis of preoperative factors showed a significant association of residual tumors with larger suprasellar AP distance (p = 0.041), retrosellar extension (p = 0.007), and higher Zurich Score (p = 0.029). CONCLUSION Large and giant PAs are challenging lesions with high subtotal resection rates. Suprasellar AP distance, retrosellar extension, and higher Zurich Score seem to be significant predictors of degree of resection in these tumors. Improving the intraoperative ability to distinguish tumor from a normal tissue might further decrease the number of unexpected residuals.