2026 Poster Presentations
P249: NATURE ABHORS A CAVITY: THE RELATIONSHIP BETWEEN INTRADURAL DEAD SPACE AND RISK OF CEREBROSPINAL FLUID LEAK FOLLOWING RESECTION OF OLFACTORY GROOVE MENINGIOMAS
Kelly E Daniels, MD; Michael Xie; Georgios A Zenonos; Paul A Gardner; Garret W Choby, MD; Eric W Wang; Carl H Snyderman; UPMC
Background: Rates of cerebrospinal fluid (CSF leak) following endoscopic endonasal resection of olfactory groove meningiomas (OGM) range between 10-30% and remain a significant challenge in managing these tumors due to the propensity for further morbidity. Despite improvements in reconstructive techniques with multilayer reconstruction and vascularized flaps, persistent CSF leak remains a troublesome complication.
Objective: The authors aim to determine whether increased tumor volume, and subsequently greater intracranial dead space following resection, is associated with an increased risk of post-operative CSF leak. A secondary aim is to identify high-risk patients and propose adjustments to the reconstructive algorithm to prevent postoperative CSF leak.
Methods: This is a retrospective case series at a single institution. All patients who underwent EEA for OGM between 2014-2025 who had imaging available for review were included. Pre-operative MRI was reviewed to measure maximal tumor dimensions in anterior-posterior (AP), lateral (L), and cranial-caudal (CC) dimensions. Post-operative CT head was used to measure AP and L dimensions of the bony defect, as a proxy for the dural defect. Other perioperative details were reviewed including lumbar drain usage, initial reconstructive technique, and management of any post-operative leaks. Univariate analysis was performed.
Results: 53 patients met inclusion criteria. The overall CSF leak rate was 21.1%. 51 patients were reconstructed with a dural substitute inlay, a fascia lata onlay, and a nasoseptal flap. 1 patient was reconstructed with a pericranial flap, and 1 patient was reconstructed with temporalis fascia. Gross total resection was achieved in 34 patients. 14 patients had a subtotal resection with >90% of tumor resected, and 4 patients had partial resections with <90% of tumor resected. On multivariate analysis, no variables including age, gender, maximum tumor dimension, tumor volume, bony defect area, nor use of a lumbar drain were associated with increased risk of post-operative CSF leak.
Conclusions: We conclude that pre-operative volume of tumor does not predict risk of CSF leak. Reconstruction of these tumors remains a challenge, and each patients’ independent risk factors should be considered when planning a reconstructive approach. We propose that a prospective root cause analysis process would be able to better identify the cause of post-operative CSF leak. Importantly, patients should be counseled appropriately about the higher risk of post-operative CSF leak with OGM compared to other tumors accessed endonasally.