Beyond Technical Failure: Development and Preliminary Prospective Validation of a Mechanism-Informed Recurrence Score Following Endoscopic Cerebrospinal Fluid Leak Repair.

backgroundDespite excellent success rates following endoscopic cerebrospinal fluid (CSF) leak repair, postoperative recurrence remains a clinically important challenge, particularly in spontaneous leaks. Increasing evidence suggests that recurrence reflects the interaction of clinical susceptibility, anatomical vulnerability, and reconstructive strategy rather than technical failure alone. However, no prospectively evaluated mechanism-informed model is currently available to support individualized recurrence risk stratification.

objectiveTo identify independent predictors of postoperative recurrence following endoscopic CSF leak repair and to develop and preliminarily prospectively validate a clinically applicable mechanism-informed model for individualized recurrence risk assessment.

methodsA mixed retrospective-prospective cohort study was conducted at a tertiary referral center. Sixty-eight consecutive patients who underwent endoscopic endonasal CSF leak repair between 2014 and 2025 were included in the derivation cohort. Independent predictors of recurrence were identified using multivariable logistic regression and incorporated into a mechanism-informed recurrence model integrating etiological, anatomical, and reconstructive determinants. Preliminary prospective validation was performed in an independent cohort of 11 consecutive patients managed using the same surgical protocol.

resultsThree variables were independently associated with postoperative recurrence: spontaneous etiology (OR 4.6; p < 0.001), defect size > 10 mm (OR 3.3; p = 0.003), and reconstruction using two layers or fewer (OR 2.7; p = 0.018). The proposed model demonstrated good discrimination (AUC 0.79; 95% CI 0.65-0.92) and stratified patients into clinically distinct risk categories. During preliminary prospective validation, no recurrences occurred in the low-risk group, whereas recurrence was observed among intermediate- and high-risk patients. Given the limited validation cohort, these findings should be considered exploratory and require confirmation through larger external multicenter studies.

conclusionPostoperative recurrence after endoscopic CSF leak repair appears to result from the combined influence of etiological, anatomical, and reconstructive factors rather than technical failure alone. The proposed mechanism-informed model provides a practical framework for individualized recurrence risk stratification and perioperative decision-making. Although preliminary prospective validation demonstrated encouraging performance, external multicenter validation is required before routine clinical implementation.