Risk of Postoperative Cerebrospinal Fluid Leak in Reused Nasoseptal Flaps

Introduction: The nasoseptal flap takedown technique is an attractive option for reoperations or staged endoscopic endonasal skull base surgeries. However, it is often used in the setting of larger defects, and may suffer from retraction and less pliability of the flap, which may lead to a higher incidence of postoperative CSF leaks.

The objective of this study is to evaluate outcomes of reused nasoseptal flaps in comparison to primary nasoseptal flaps (NSF) and what factors among reused NSFs would be associated with a higher incidence of leaks.

Methods: A retrospective review of patients undergoing endoscopic endonasal skull base surgery between 2010 and 2015 in a single center was performed. Only cases where the NSF was reused for the first time and those with intraoperative CSF leak were included.

Results: 601 primary NSF were harvested and 43 developed postoperative CSF leak (7%). In the same period, 33 NSF were reused and 7 of those had CSF leaks (21%, p = 0.011). Only one case had meningitis.

Of these 33 cases, 25 underwent surgery for tumor recurrence and 8 were staged surgeries. Median interval between primary NSF harvest and reuse was 25 months (1 week-77months). Diagnoses included 11 pituitary adenomas, 1 Rathke’s cleft cyst, 7 craniopharyngiomas, 7 meningiomas, 5 chordomas, 1 chondrosarcoma and 1 trigeminal schwannoma. Median age was 56 (20–84) and median follow-up was 22 months (1–70 months).

Patients less than 60 years had more leaks (p = 0.032). None of the patients over 60 years old (n = 12) had CSF leaks.

Risk of CSF leak was not related to location of the lesion (p = 0.28 for posterior fossa and p = 1 for other locations) although 40% (n = 2) of the cases in the posterior fossa developed CSF leak postoperatively.

2 of 5 chordomas (40%) had CSF leaks, although tumor histology did not correlate with CSF leak. No significant difference was found between the sizes of the defects of those that developed and did not develop CSF leaks (p = 0.248).

5 of the 33 cases (15%) had intraoperative damage to the flap (small perforation, tear, tapering or damage of part of the flap blood supply), of those, 40% (n = 2) developed leaks (p = 0.194).

82.7% (n = 24) of those flaps that covered the defect completely did not develop CSF leaks. Among those that did not provide complete coverage, 50% had leaks (p = 0.189) even with additional grafts, such as acellular dermal matrix allograft.

Body mass index, diabetes mellitus, lumbar drain placement, radiotherapy before or after surgery, or time between harvesting the primary flap and its reutilization were not related to CSF leaks.

Conclusion: Although this study is limited by its sample size, reused NSF have a significantly higher risk of postoperative CSF leak, especially in patients younger than 60. Defects in the posterior fossa, following resection for chordoma, with intraoperative damage to the flap or incomplete coverage of the defect may require additional graft or close postoperative follow-up.