The Use of Resorbable and Integrative Collagen Sponges to Obliterate the Intrasellar Space and Achieve a Fatless, Spherical Dural Reconstruction Following Transsphenoidal Surgery

Introduction: The repair of cerebrospinal fluid (CSF) leaks following transnasal transsphenoidal (TNTS) surgeries for sellar pathologies has developed considerably over the previous decades. Traditionally performed using autologous fat grafts, the advent of the nasoseptal flap (NSF) has allowed for the repair of even high flow leaks and larger skull base defects with great success. The development of improved allografts and fibrin sealants has allowed for unique ways to reconstruct the sella. However, it is unclear which materials are best suited for this task, and if autologous fat is still a necessary component of the repair.

Methods: Beginning in August of 2018, the primary author repaired grade I and II leaks without the use of autologous fat. Upon identification of a CSF leak following resection of the pituitary tumor, a piece of collagen sponge (Helistat) was cut to fill the expanded sellar cavity, while limiting extension into the suprasellar space. An oversized piece of spongy dural allograft (Duragen Plus) was then used to wrap around the intrasellar hemostatic agent, with particular care to ensure that a long length of allograft would span the entirety of the diaphragma sella. All edges were tucked intradurally and along the medial cavernous sinus and sellar floor. A Valsalva maneuver was performed to 30 mm Hg to ensure no brisk leakage of CSF was seen. A free mucosal graft or nasoseptal flap was then placed over the repair, and sealed with fibrin glue. This was then buttressed with nasopore.

Results: Eight patients from August of 2018 received sellar reconstruction in the abovementioned manner. Six were female, median age was 53 (39–68), and all had macroadenomas with an average maximum diameter of 3.1 cm (2–4.8 cm). There were two grade I leaks, and six grade II leaks. One patient had no mucosal grafting (grade I), one patient had a free mucosal graft (grade II), while all others had NSFs placed. In one case, Helistat was not available and hemostatic cellulose (Surgicel Fibrillar) was used. Tisseel was used once, while all other patients received Adherus. All patients had at least two months follow-up with no signs of CSF leak postoperatively. Postoperative MR studies demonstrate sellar lucency where the collagen sponge was placed.

Conclusion: We describe a method through which absorbable collagen sponge is used to abrogate CSF flow through the arachnoid in a semi-impermeable fashion. We believe this allows sufficient diminution of both the physical fluid pulsations transmitted from the intracranial space as well as the access of fluid to the dural repair, thereby allowing it time to seal and integrate. The use of an absorbable sponge to fill the sellar space is critical in this fatless technique as opposed to simply using dural allograft, as the latter may result in a thickened mass of dural substitute that may be challenging to traverse in cases of repeat surgeries. Future studies are needed to determine if this method is still viable without a vascularized flap or fibrin sealant.

Fig. 1

Fig. 2