Postoperative CSF Leaks following Endoscopic Anterior Skull Base Surgery: Risk Factors and Management

Objective: Postoperative cerebrospinal fluid leak following endoscopic anterior skull base surgery is one of the leading causes of postoperative morbidity. The factors associated with increased incidence of postoperative CSF leaks have not been well described. In this study, we analyze data from a 10-year period of endoscopic anterior skull base surgery in our institution to identify characteristics associated with increased risk of postoperative cerebrospinal fluid leak. Management of these leaks is also described.

Methods: A retrospective chart review was performed at a tertiary care academic medical center to identify patients who underwent endoscopic anterior skull base surgery between August 2007 and January 2018. Data on age, sex, tumor histology, location and size, revision surgery, history of skull base radiation, medical comorbidities such as body mass index (BMI), diabetes mellitus, hypertension, heart disease, intraoperative CSF leaks, and methods of reconstruction were recorded.

Results: A total of 833 patients underwent endoscopic anterior skull base surgery from August 2007 through January 2018. The age range was 7 to 83 years. 25 patients (3%) developed postoperative CSF leaks, 13 females and 12 males. Mean age of these patients was 53.8 years (range: 16–78 years) and mean BMI was 29.8 (range: 22.5–46.2). Leaks occurred at a median of 9 days following surgery (range: 1–200 days). Tumor types in patients who developed postoperative CSF leaks included pituitary adenoma 46% (12), while 54% were other pathologies: meningioma (4), chordoma (1), craniopharyngioma (2), arachnoid cyst (1), fibrous dysplasia (1), Rathke’s cleft cyst (1), metastatic renal cell carcinoma (1), schwannoma (1), and granular cell tumor (1). The mean tumor size at initial resection was 2.6 cm (range: 0.3–4.1 cm). Two patients (8%) had prior radiation. Seven patients (28%) had prior surgery. Three (12%) had diabetes mellitus. Statistical analysis was performed using chi-square and Fisher’s exact tests. Tumor size and BMI were not associated with an increased incidence of postoperative leak (p = 0.076 and p = 0.919), respectively. Suprasellar extension and cavernous sinus invasion, as well as pathologies other than pituitary adenomas, were not associated with increased incidence of postoperative CSF leak (p = 0.723 and p = 0.081), respectively. Operative management, performed in 23 patients, included nasoseptal flap (primary and revision), abdominal fat and tensor fascia lata grafts, and middle turbinate flaps. Operative techniques will be demonstrated with video. Lumbar drains were utilized in addition to surgery in 17 patients. Two patients were managed with a lumbar drain without operative intervention, with resolution of the CSF leak in 3 to 4 days.

Conclusion: Postoperative CSF leaks occurred in 3% of our series following anterior endoscopic skull base surgery. Pathologies other than pituitary adenomas, tumor size, suprasellar or cavernous sinus extension, as well as other factors including history of prior surgery, prior radiation, elevated BMI, hypertension, diabetes mellitus, and cardiac disease were not associated with an increased incidence of postoperative CSF leaks. Techniques such as nasoseptal flap, abdominal fat, tensor fascia lata, dural substitutes, and turbinate flaps were useful for reconstruction.