2026 Poster Presentations
P258: REGIONAL SKULL BASE DEFECTS AS PREDICTORS OF ENCEPHALOCELE FORMATION IN SPONTANEOUS CSF LEAK PATIENTS
Joyce Jeong, BS1; Rachel Akers, MS1; Katie Sinchek, MD2; Bobby Tajudeen3; Aimee Szewka, MD2; 1Rush University Medical College; 2Department of Neurology, Rush University Medical Center; 3Department of Otorhinolaryngology - Head and Neck Surgery, Rush University Medical Center
Background: Spontaneous cerebrospinal fluid (CSF) leaks and encephaloceles are increasingly linked to chronic intracranial pressure changes and skull base anomalies. Identifying which bony defects are most associated with encephalocele formation could improve early diagnosis and surgical planning.
Methods: We conducted a retrospective chart review of 65 patients presenting with spontaneous CSF leaks. Imaging and operative findings were analyzed to identify encephalocele presence and location. Skull base anomalies were categorized by site of osteitic dehiscence (cribriform plate, ethmoid, central sphenoid, tegmen, lateral sphenoid, petrous apex, and others). Encephalocele presence was compared with dehiscence location using univariate statistics.
Results: Of 65 patients, 46 (70.8%) had encephaloceles. The most common osteitic dehiscence sites overall were the cribriform plate (40%), lateral sphenoid (33.8%), central sphenoid (29.2%), and tegmen (27.7%). When stratified by encephalocele presence, central sphenoid dehiscence was significantly more common in encephalocele patients compared with non-encephalocele patients (37.0% vs. 10.5%, p = 0.033). Anterior encephaloceles (involving the cribriform plate, planum sphenoidale, or ethmoid roof) were more frequently associated with central sphenoid dehiscence compared to lateral encephaloceles (64.3% vs. 25.0%, p = 0.019). Tegmen dehiscence was also more common in encephalocele patients (34.8% vs. 10.5%, p = 0.047), with lateral encephaloceles demonstrating a particularly strong association (50.0% vs. 0% in anterior, p = 0.002). In contrast, cribriform plate dehiscence was paradoxically less common among encephalocele patients (30.4% vs. 63.2%, p = 0.014). No significant group differences were observed at other sites, including the lateral sphenoid, petrous apex, and ethmoid.
Conclusion: In this cohort of patients with spontaneous CSF leaks, encephalocele formation was strongly associated with central sphenoid and tegmen skull base defects, whereas cribriform plate defects were more common in non-encephalocele patients. Importantly, these associations demonstrated anatomical specificity in which central sphenoid defects were linked to anterior encephaloceles, while tegmen defects were linked to lateral encephaloceles. Notably, some patients harbored multiple skull base defects or more than one encephalocele, highlighting the complexity of disease patterns and suggesting that multifocal weakening of the skull base may occur in the setting of chronically elevated intracranial pressure. These findings highlight that not all skull base anomalies confer equal risk and that defect location may not only predict encephalocele formation but also its subtype, with direct implications for early recognition, risk stratification, and surgical planning. Larger, prospective studies are needed to confirm these relationships and refine predictive models for clinical use.