Comparative Evaluation of Enhancement Patterns of Nasoseptal Flaps and Free Mucosal Grafts in Primary Endoscopic Pituitary Surgery and Association with CSF Leak

Background: The endoscopic transnasal transsphenoidal (TNTS) approach to the anterior skull base has become the technique of choice for addressing pituitary tumors. The Hadad-Bassagasteguy nasoseptal flap (NSF) revolutionized endoscopic skull base surgery by significantly reducing the rates of cerebrospinal fluid (CSF) leaks. This flap has been modified multiple times since and a variety of techniques using both vascularized grafts and free mucosal grafts (FMG) are used today for skull base reconstruction. Despite these advances, CSF leaks remain the most common complication associated with TNTS. There has been limited data evaluating the appearance of these various reconstruction techniques on MRI in the perioperative period to predict patterns of vascularization of skull base reconstruction, how this vascularization pattern changes over time, and if these findings may be associated with CSF leak rates.

Methods: Patients having undergone primary resection of pituitary adenoma and sellar reconstruction with either a FMG or a NSF via TNTS between 2012 and 2023 at a tertiary rhinology practice underwent chart and imaging review. Demographic data, operative details, and perioperative complications were reviewed. MRIs obtained within 10 days of surgery and 3 months after surgery were reviewed by a neuroradiologist to determine patterns of NSF or FMG enhancement patterns over time. Vascularization patterns were compared between NSF and FMG and differences in CSF leak rates were evaluated.

Results: In total, 52 patients (27 females [51.9%]; mean [SD] age, 50.9 [16.7] years) having undergone primary resection of pituitary adenoma via TNTS with 10-day and 3-month MRIs available for review were included. Twenty-seven (51.9%) patients had anterior skull base reconstruction via FMG harvested from the posterior septum and were compared to 25 (48.1%) patients having undergone NSF reconstruction. There were no significant differences in patient demographics between those having undergone reconstruction via FMG and NSF. Patients having undergone reconstruction with NSF had higher rates of intraoperative CSF leaks (n, %; 20, 80.0%) compared to those having undergone FMG reconstruction (2, 7.4%), p < 0.0001. None of the FMG reconstruction patients had graft enhancement on MRI within 10 days of surgery, but all had apparent graft enhancement at their 3-month MRI. The majority of patients (16, 64%) reconstructed via NSF demonstrated flap enhancement at 10 days after surgery and by 3 months all patients had enhancement of their reconstruction. Four (7.7%) patients experienced a CSF leak after surgery. Three (75%) of these patients had NSF reconstruction and had a grade 2 or greater intraoperative CSF leak. Two (66%) of NSF patients with leaks had flap enhancement at their first MRI. There was no difference in CSF leak rates based on NSF flap patterns of enhancement (p = 1.0).

Conclusion: There was no association between patterns of flap enhancement on MRI and CSF leaks in the perioperative period. FMGs demonstrated a predictable change in enhancement pattern over the course of 3 months that may represent neovascularization of their skull base reconstruction. NSFs without flap enhancement at 1 week demonstrated similar patterns of neovascularization on subsequent MRIs without increased risk of complications ([Fig. 1]).

Fig. 1

No conflict of interest has been declared by the author(s).

Publication History

Article published online:
05 February 2024