2026 Poster Presentations
P248: POST-OPERATIVE SUPRASELLAR CSF LEAK IN NON-PITUITARY PATHOLOGY: A SINGLE INSTITUTION CASE SERIES
Kelly E Daniels, MD; Kristen L Zayan, MD; Georgios A Zenonos, MD; Paul A Gardner, MD; Carl H Snyderman, MD, MBA; Eric W Wang, MD; Garret W Choby, MD; UPMC
Background: Endoscopic endonasal surgery is being increasingly utilized over transcranial approaches for suprasellar lesions including meningiomas, Rathke cleft cysts, and craniopharyngiomas. Accessing these suprasellar tumors presents challenges over other locations in the sagittal plane as their superior extension thins the diaphragma or may erode into the third ventricle making intraoperative high flow leaks much more common and oftentimes unavoidable.
Objective: This study aims to use a decade-long experience with suprasellar pathology to identify etiology of post-operative CSF leaks as well as best practices for avoidance and management.
Methods: This is a single-institution retrospective case series reviewing post-operative CSF leaks associated with suprasellar defects for non-pituitary pathology from 2015-2025. Patients with other extended approaches in the sagittal plane, including transcribriform and transclival, were not included in this analysis. Patient demographics, operative and reconstructive techniques, and strategy for post-operative leak management were among the variables collected.
Results: A total of 20 patients met inclusion criteria. Pathology consisted of tuberculum meningioma (n=6, 30%), craniopharyngioma (n=6, 30%), suprasellar Rathke cleft cyst (n=3, 15%), planum meningioma (n=2, 10%) and other (n=3, 15%). 3 cases were for recurrent lesions, all of which were craniopharyngiomas. All patients had intra-operative leaks, with 3 (15%) having low flow and 17 (85%) having high flow leaks. The most common means of reconstruction was a dural substitute inlay (all patients), with or without fascia lata onlay (5 patients had fascia onlays, 15 patients did not), followed by a nasoseptal flap (17 out of 20 patients had nasoseptal flaps, 2 had free mucosal grafts, and 1 had a pericranial flap). At the time of initial surgery, 5 patients received lumbar or extraventricular drains and 15 (75%) did not. 8 patients presented with post-operative leak during their same admission, whereas 12 patients were readmitted for leak management. Of those with post-operative leaks all were managed by return to the operating room for exam under anesthesia; displacement of the reconstruction was noted in 9 patients, with 14 patients noted to have flap dehiscence, and 2 patients with flap necrosis or death. Both instances of flap necrosis were noted to be only partial necrosis, and only one of them required additional vascularized reconstruction via a lateral nasal wall flap. Otherwise, on return to the operating room, all patients were able to reuse their initial reconstruction after repositioning and bolstering. Lumbar drain or EVD were placed in 18/20 patients.
Conclusions: CSF leak remains an inherent risk following suprasellar endonasal approaches due to their propensity for intra-operative high flow leaks. A majority of these leaks result from displacement or dehiscence of the reconstruction, and as such can be salvaged with return to the operating room for leak identification and reconstruction repositioning and securing. Future efforts to reduce the incidence of suprasellar CSF leaks could focus on optimizing technique to avoid flap dehiscence and novel techniques to better secure the reconstruction.