Abstract
Background Endoscopic endonasal skull base reconstructions have been associated with postoperative cerebrospinal fluid (CSF) leaks. Objective A repair protocol for endoscopic endonasal skull base reconstruction is presented with the objective of decreasing the overall leak rate. Methods A total of 180 endoscopic endonasal skull base reconstructions were reviewed. Reconstructions were classified I to IV according to the reconstruction method, determined by severity of intraoperatively encountered CSF leaks for types I to III, and planned preoperatively for type IVs, which required nasoseptal flap. Results A total of 11 patients(6%) had postoperative leaks: 0 in type I (0%), 2 in type II (5%), 7 in type III (18%), and 2 (4%) in type IV reconstruction. Type III leak rate was higher than all other reconstructions. Total 31 intraoperative and 16 postoperative lumbar drains were placed. More patients had lumbar drains placed postoperatively for type III and intraoperatively for type IV than all other groups. There were significant overall differences in postoperative CSF leaks and lumbar drain placement between the four reconstruction types. No patient with type III reconstruction and intraoperative lumbar drain had postoperative CSF leak. Conclusions A repair protocol for endoscopic endonasal reconstructions determined by intraoperative CSF leak and preoperative planning minimizes unnecessary repair materials and additional morbidity. Our experience leads to a routine prophylactic lumbar drain placement in all type III leak and reconstructions. We also favor the type III reconstruction for minor intraoperative leaks, and a more generous use of type IV reconstructions in expectation of significant intraoperative CSF leak. The option of rescue flap technique in type III leaks should be strongly considered.