2025 Poster Presentations
P256: DELAYED CEREBROSPINAL FLUID LEAK FOLLOWING ENDOSCOPIC TRANSNASAL SKULL BASE SURGERY: ASSOCIATION BETWEEN RECONSTRUCTION METHOD AND RADIATION THERAPY
Motoyuki Umekawa, MD1; Hirotaka Hasegawa1; Yuki Shinya1; Masahiro Shin2; Nobuhito Saito1; 1The University of Tokyo Hospital; 2Teikyo University Hospital
Objectives: Endoscopic transnasal surgery (ETS) allows for minimally invasive, high-resolution panoramic views of deep intracranial space and has recently been applied to extended skull base surgery. Skull base reconstruction is essential for preventing cerebrospinal fluid (CSF) leaks, and we basically use non-vascularized multilayer fascia for this purpose. However, recurrent or refractory tumors such as chordomas and high-grade meningiomas often require radiation therapy (RT), raising concerns about the damage of the reconstructed tissue. Based on our experience with delayed CSF leaks following ETS and RT, we conducted a retrospective analysis to assess the risk of delayed CSF leaks.
Methods: Patients who underwent ETS at our institution from January 1991 to October 2023 and had a follow-up period of at least six months were included. Data collected included patient demographics, pathologies, number of ETS procedures, skull base reconstruction methods, and number of RT sessions. Delayed CSF leak was defined as any CSF leak occurring six months or more after the last ETS. We analyzed risk factors associated with delayed CSF leaks.
Results: A total of 287 patients were included, with a median follow-up period of 45 months. The primary diseases were pituitary adenoma (106 patients, 37%), chordoma (61 patients, 21%), and meningioma (43 patients, 15%). ETS was performed once in 218 patients (76%) and more than once in 69 patients (24%). Skull base reconstruction methods included simple closure with or without fat grafting or local sphenoid mucosal flap in 132 patients (46%), non-vascularized multilayer fascial closure in 141 patients (50%), and multilayered fascial closure with a pedicled mucosal flap in 14 patients (5%). RT was administered to 102 patients. Delayed CSF leaks occurred in five patients (1.7%), with a cumulative incidence rate from initial ETS of 0.6% at 3 years, 1.2% at 5 years, and 3.7% at 10 years. The incidence was significantly higher in the RT group compared to the non-RT group (2.6% at 5 years, 7.0% at 10 years vs. 0.0% at 10 years; log-rank test, p = 0.027). No delayed CSF leaks occurred in the group reconstructed with multilayered fascial closure and a pedicled mucosal flap. Bivariate analysis using Cox proportional hazard model revealed that chordoma (hazard ratio [HR] 9.48, 95% confidence interval [CI] 1.05–85.4, p = 0.045), number of ETS (HR 1.93, 95% CI 1.29–2.93, p = 0.001), and number of RT (HR 1.55, 95% CI 1.10–2.19, p = 0.009) were associated with incidence of delayed CSF leak.
Conclusion: In the non-RT group, delayed CSF leaks did not occur with simple closure or non-vascularized multilayered fascial closure. In contrast, while the incidence was low, delayed CSF leaks could occur in the RT group. These were successfully prevented with reconstruction using a pedicled mucosal flap so far, suggesting vascularized tissue may have resilience to RT.