Endoscopic Repair of Idiopathic CSF Leaks: An Institutional Approach for Postoperative Management

Objective: Endoscopic repair of skull base CSF leaks have been known to be successful, with rates published in literature to be above 90%. However, patients with an idiopathic etiology have been reported to have a higher number of recurrent CSF leaks requiring additional revision surgeries or multiple adjunct treatments. Strict protocols on postoperative management of such patients have not clearly been defined; however, institutional paradigms have been reported. Such management strategies include placing all patients on medications to decrease ICP, to placement of ventriculoperitoneal (VP) shunts. In our institution, patients who have undergone an endoscopic repair of an idiopathic CSF leak do not get treated postoperatively with any medication or shunt placement unless symptomatic for pseudotumor cerebri symptoms. A more selective postoperative approach is taken without routine use of diuretics or shunts. Our objective in this review is to study the long-term natural history spontaneous CSF leaks after primary single layer endoscopic repair, without the routine use of medications or shunting, unless symptomatic.

Study Design: Retrospective review.

Methods: A retrospective chart review was conducted for patients endoscopically treated for idiopathic CSF leaks at the University of Miami Health System from 2010 to 2016. Data on demographics, symptomology, location of skull base defect, BMI, repair type, and clinical follow-up was included.

Results: Twenty-six patients who underwent endoscopic repair of spontaneous CSF leaks were included in the study. Twenty-one (81%) were females. Average age was 47 years. 58% of patients were obese (BMI > 30 kg/m²). Average BMI of the study population was 32.55 kg/m². The location of the skull base defect was primarily at the cribriform plate (61%) and sphenoid sinus (35%). Endoscopic repair was performed as the initial surgical procedure in all cases and was successful in 96% (25 patients) of the cases. Average follow-up time for this study was 46.7 months (range of 9–99 months). Acetazolamide was used as a postoperative adjunct medication for only three patients (11%) based on pseudotumor cerebri symptoms, with persistent headaches being the most common complaint. VP shunt was also performed in only three patients and all of those patients had the shunt placed previously before the repair for a history of pseudotumor cerebri.

Conclusion: Endoscopic repair of idiopathic CSF leaks was found to have a high rate of success long-term in our study, even without the routine use of diuretics or VP shunts. Post operatively only a minor subset of patients required additional measures of acetazolamide to manage their symptoms No patients required the use of VP shunt postoperatively. Treating patient’s reflexively with additional medications and shunts can expose them to unwanted side effects and surgical risks. This study highlights that patients who have undergone an endoscopic repair of a spontaneous CSF leak be monitored in a close multidisciplinary approach between the otolaryngologist and neurologist. Additional treatment should be considered only in symptomatic patients or those who fail the initial repair. Weight loss should be part of the proposed postoperative management plan.