Algorithm for Sellar Reconstruction following Transsphenoidal Pituitary Surgery: A Review of 342 Cases

Background: While endoscopic transsphenoidal pituitary surgery has become a widely accepted approach to sellar and parasellar pathology, there is little consensus regarding reconstruction. Many algorithms have been proposed taking into consideration intraoperative findings and patient or tumor specific features. A multitude of techniques have been proposed, ranging from no reconstruction to vascularized pedicled flaps. Herein, we describe a minimalistic approach to sellar reconstruction and review the associated outcomes.

Methods: A retrospective review of 342 consecutive patients who underwent endoscopic pituitary surgery for pituitary adenoma at a tertiary academic center was performed. Patient demographics, intraoperative, and postoperative data were collected. Patients were grouped based on three closure techniques: Surgicel (S, n = 81), synthetic dural substitute with the use of dural sealant without the use of a nasoseptal flap (DRDG, n = 225), and primary dural repair with a vascularized nasoseptal flap (F, n = 36).

Results: There was no statistically significant difference between the three groups with respect to patient age, gender, body mass index (BMI), or a history of obstructive sleep apnea (OSA). With respect to tumor characteristics, the Surgicel group were smaller tumors (S = 1.98 ± 1.33 cm, DRDG = 2.53 ± 0.96 cm, F = 2.46 ± 0.94 cm, p < 0.001), more likely to be microadenomas (S = 17%, DRDG = 4%, F = 3%, p < 0.001), and were less likely to be an invasive tumor (S = 20%, DRDG = 36%, F = 56%, p < 0.001). Intraoperatively, significant differences were found between closure techniques when assessing intraoperative CSF leaks (S = 0%, DRDG = 30%, F = 94%, p < 0.001), anterior diaphragmatic sellar defect (S = 0%, DRDG = 4%, F = 11%, p = 0.11), high-flow leak (S = 0%, DRDG = 0%, F = 6%, p = 0.011), and patulous diaphragm (S = 5%, DRDG = 41%, F = 3%, p < 0.001). There was no significant difference in the type of resection between the three groups. During the postoperative period, significant differences in complications were found for sinusitis requiring antibiotics (S = 16%, DRDG = 23%, F = 44%, p = 0.004) and postoperative crusting (S = 7%, DRDG = 11%, F = 56%, p < 0.001). There were no significant differences in postoperative CSF leak rates (S = 1%, DRDG = 2%, F = 3%, p = 0.857). Our overall postoperative leak rate was 1.8%.

Conclusion: Based on this large series, we propose the following algorithm for sellar reconstruction: Surgicel alone for cases of small tumors or microadenomas without an intraoperative CSF leak; synthetic dural substitute for larger tumors with a diaphragma sellae leak or patulous diaphragm; and the use of a vascularized flap reserved for only large tumors with intraoperative leaks in combination with other ‘risk factors’ or a high flow intraoperative leak.