Institutional Insight on Fatless Reconstruction of the Skull Base for CSF Leak Repair following Endoscopic Transnasal Transsphenoidal Surgery: 44 Case Series

Introduction: Reconstruction of the anterior skull base for a CSF leak following endonasal transsphenoidal surgery has commonly employed fat or fascia lata grafts. Institutionally, we have transitioned to a combination of more modern allografts, dural sealants, and free and vascularized mucosal grafts. We find that this closure is comparable to traditional fat-based CSF leak repair methods, while saving patients from a separate incision and the associated 1 to 3.7% donor-site complication rate.

Methods: Single-institution retrospective review of all CSF leaks after an eTNTS repaired without a fat graft, from two neurosurgeons, from May 2014 to present. Primary outcome measures are postoperative CSF leak, return to OR, and lumbar drain placement.

Results: We performed this fatless reconstruction on a wide variety of patients (Table 1). Similar to our previously published principles, we used nasoseptal flaps (NSFs) for higher grade leaks (2 and 3) and mucosal free flaps for low-grade ones (Fig. 1). In line with multilayer closure techniques previously described, multiple layers of repair were addressed to a variable degree in each case, including arachnoid closure with Anastoclips, sellar space obliteration with Helistat or Gelfoam, sellar floor reconstruction with bone or absorbable plates, and episellar buttressing (coude catheter, NasoPore packing, or Merocel-filled gloves). Dural sealants (Adherus, DuraSeal, and Tisseel) were used in 48% of cases. Dural substitutes (DuraMatrix, DuraGen, and DuraGen Plus) were used 55% of the time. Two patients had prophylactic lumbar drains placed intraoperatively, and eight patients were placed on acetazolamide postoperatively. Only one patient had a CSF leak postoperatively requiring surgical repair, despite lumbar drain trial. The leak was grade 2, repaired with a nasoseptal flap, a dural substitute, Anastoclip, NasoPore packing, and no dural sealant. The tumor measured 25 mm, was soft, nonadherent, without cavernous sinus involvement, and there was no history of prior TNTS. Complete pseudocapsular dissection was achieved.

Conclusion: Similar CSF leak repair rates can be achieved without using a fat graft for intrasellar reconstruction. There is unlikely one ideal alternative and we demonstrate multiple permutations within our series; however, the principles of multilayer reconstruction remain the same. Successful repair relies on reducing the rate of CSF flow (arachnoid closure or sellar obliteration), more permanent dural and mucosal reconstruction (allograft, free or pedicled mucosa), and for higher flow leaks, a pedicled NSF with or without buttressing the repair. Further studies are required to determine the most effective and cost conscientious fat graft free methods that achieve the optimal degree of repair while continuing to minimize patient morbidity.