Barrier-Limited Multimodality Closure for Reconstruction of Wide Sellar Openings

Background: Although the endonasal endoscopic approach (EEA) allows for increased bone removal and thus a wider dural exposure, it also presents the challenge of skull base reconstruction in the face of an intraoperative cerebrospinal fluid (CSF) leak. Obtaining a watertight reconstruction using a fat graft with wide sellar exposures can be challenging, and with no barrier in place, carries the risk of reinstating mass effect. The alternative, a vascularized nasoseptal flap, may require several days to heal and has an approximate CSF leak rate of 6%.

Objective: The purpose of this study is to introduce and assess the efficacy of a barrier-limited multimodality (BLMM) closure technique for intraoperative CSF leaks obtained during an EEA, consisting of an autograft fat-based watertight seal, a limiting membrane barrier, and the vascularized nasoseptal flap.

Methods: Between 2008 and 2010, 196 patients underwent an EEA procedure for lesions involving the sellar, parasellar, and suprasellar regions at our institution. We performed a retrospective review of 27 consecutive patients who experienced an intraoperative CSF leak (Grade 1-11, Grade 2-9, and Grade 3-7) requiring repair with the BLMM technique. The membrane barrier is first created by securing a piece of absorbable collagen sponge to the dural edges with titanium clips. This is followed by an autologous fat graft that is typically buttressed in place using bone harvested during the exposure. The vascularized nasoseptal flap is then rotated to cover the entire construct. The results of 43 prior reconstructions with intraoperative CSF leaks repaired using a nasoseptal flap only were analyzed as a comparison group.

Results: There were no postoperative CSF leaks for the patients who underwent reconstruction using the BLMM closure technique. There were no complications attributable to the BLMM reconstruction. The CSF leak rate for the comparison group with nasoseptal flap repair only was 19%.

Conclusion: The BLMM closure may further decrease the incidence of postoperative CSF leaks compared with that obtained with a predominant reliance on a vascularized nasoseptal flap. The membrane barrier allows for a watertight inner closure and prevents herniation of the fat autograft into the resection cavity. An outer layer nasoseptal flap provides a living barrier for optimal long-term defense.