BackgroundDespite recent advances in skull base reconstructive techniques, including the multilayer technique during the last decade, complete reconstruction of grade 3 intraoperative high-flow cerebrospinal fluid (CSF) leak remains challenging. This study was designed to investigate the role of injectable hydroxyapatite (HXA) used in the multilayer technique on the clinical outcome of skull base reconstruction for intraoperative high-flow CSF leak.Materials and MethodsThis study enrolled 187 patients who experienced intraoperative high-flow CSF leak after endoscopic endonasal surgery for anterior skull base or suprasellar pathologies between January 2014 and July 2021. All skull base defects were reconstructed using the conventional multilayer technique including a vascularized naso-septal flap (NSF, n = 141) and the combined use of HXA with the conventional multilayer technique (HXA group, n = 46). We retrospectively evaluated the efficacy of the HXA group by 1:2 propensity score matching analysis.ResultsOverall, 17 of 187 patients (9.1%) showed postoperative CSF leaks, resulting in second reconstruction surgery. There were no statistical differences in patient age, sex, body mass index, tumor location, tumor type, and degree of resection, except for the follow-up period between the two groups. The HXA group showed a significantly lower incidence of postoperative CSF leak than the control group (0% vs. 12.1%, p < 0.05). Postoperative lumbar drain (LD) was performed in 8.7% of the HXA group compared to 46.1% of the control group (p < 0.01). CSF leak-related infection rates showed a decreasing tendency in the HXA group compared to the control group (0 vs. 7.1%, p = 0.06). A total of 46 patients in the HXA group were well matched with the control group (92 patients) at a 1:2 ratio. In the propensity score-matched control group, there were higher rates of postoperative CSF leaks than in the HXA group.ConclusionThe use of HXA combined with the conventional multilayer technique completely reduced postoperative CSF leaks in this study. This technique resulted in reduced CSF leakage, even without postoperative LD, and decreased infection rates. Further randomized comparative studies are required to confirm our findings.