Risk factors for cerebrospinal fluid leak after extended endoscopic endonasal surgery for adult patients with craniopharyngiomas: a multivariate analysis of 364 cases

OBJECTIVE Over the last decade, the extended endoscopic endonasal approach (EEEA) has evolved as a credible surgical alternative for removing craniopharyngiomas. However, postoperative cerebrospinal fluid (CSF) leak remains one of the most pressing concerns. Craniopharyngiomas often invade the third ventricle, resulting in a higher rate of third ventricle opening after surgery and potentially increasing the risk of postoperative CSF leak. Identifying the risk factors associated with CSF leak after EEEA for craniopharyngiomas may have more clinical value. Nevertheless, there is a lack of systematic studies on the topic. Previous studies yielded inconsistent results, probably due to heterogeneous pathologies or small sample sizes. Hence, the authors present the largest known single-institution case series of the use of purely EEEA for craniopharyngiomas to systematically study the risk factors for postoperative CSF leak. METHODS The authors retrospectively reviewed 364 cases of adult patients with craniopharyngiomas who were treated at their institution from January 2019 to August 2022, and they analyzed the risk factors for postoperative CSF leak. RESULTS The overall rate of postoperative CSF leak was 4.7%. In the univariate analysis, larger dural defect size (OR 8.293, 95% CI 3.711–18.534, p < 0.001) and lower preoperative serum albumin level (OR 0.812, 95% CI 0.710–0.928, p = 0.002) were associated with higher rates of postoperative CSF leak. Predominantly cystic tumors (OR 0.325, 95% CI 0.122–0.869, p = 0.025) were linked to decreased risk of postoperative CSF leak. However, postoperative lumbar drainage (OR 2.587, 95% CI 0.580–11.537, p = 0.213) and third ventricle opening (OR 1.718, 95% CI 0.548–5.384, p = 0.353) were not related to postoperative CSF leak. In the multivariate analysis, larger dural defect size (OR 8.545, 95% CI 3.684–19.821, p < 0.001) and lower preoperative serum albumin level (OR 0.787, 95% CI 0.673–0.919, p = 0.002) were identified as independent risk factors for postoperative CSF leak. CONCLUSIONS The authors’ repair technique yielded a reliable reconstructive outcome for high-flow CSF leak in EEEA for craniopharyngioma. Lower preoperative serum albumin level and larger dural defect size were identified as independent risk factors for postoperative CSF leak, potentially providing new insights into minimizing the risk of postoperative CSF leak. Third ventricle opening was not associated with postoperative CSF leak. Lumbar drainage may not be necessary for high-flow intraoperative leak, but this finding may require validation with a prospective randomized controlled trial in the future.