Objectives: To describe the appearance and changes in the vascularized pedicle nasoseptal flap (NSF) after endoscopic skull base surgery over time.
Study Design: Retrospective review of patients undergoing endoscopic skull base surgery of the sella, tuberculum sella, or planum sphenoidale at a tertiary academic skull base surgery program.
Methods: A total of 23 patients were identified who underwent endoscopic skull base surgery from July 2018 through July 2020, all of which had intraoperative NSF use skull base reconstruction. All patients received immediate and delayed (>3 months) postoperative MRI scans. Features evaluated include enhancement of pedicle and flap, presence of pneumocephalus, flap thickness (in mm), re-expansion of the pituitary gland, and flap adherence to skull base for both immediate and delayed post-operative MRIs. Two separate physicians independently reviewed each scan to ensure inter-observer agreement.
Results: All 23 patients had the C-shaped NSF over operative defects. 19 patients had intraoperative CSF leaks (11 high flow and 8 low flow). There were no post-operative CSF leaks in this cohort. Immediately post-op, 17 (74%) pedicles and 15 flaps (65%) brightly enhanced, 4 (17%) pedicles and flaps weakly enhanced, and 2 (9%) pedicles and 4 (17%) flaps did not enhance. On delayed imaging, 20 (87%) pedicles and 19 (83%) flaps brightly enhanced, 2 (9%) pedicles and 4 (17%) flaps weakly enhanced, and 1 (4%) pedicle did not enhance. There were no significant differences in pedicle (p = 0.317) or flap enhancement (p = 0.074) over time. Of the 10 patients with change in enhancement over time, 6 had increased enhancement of both pedicle and flap, 3 had increased enhancement of flap only, and 1 had decreased enhancement of pedicle and flap. In this cohort, mean time to change in enhancement was 3.8 months. Mean NSF thickness on immediate and delayed post-op scans were 3.78 and 3.82 mm, respectively. 13 (56%) flaps changed thickness over time, with a mean increase of 0.61 mm (standard deviation: 0.583) over time. 7 and 6 flaps increased and decreased in thickness, respectively. On paired analysis, there was no significant difference in NSF thickness over time (p = 0.862). The NSF adhered directly to the skull base defect in 10 (44%) and 21 (91%) of patients on immediate and delayed imaging, respectively, and this change was significant (p = 0.001). Pneumocephalus was observed in 14 (61%) immediate post-op images, which resolved in all 14 scans on delayed imaging at 3 months. The pituitary gland reexpanded in 14 (61%) MRIs on delayed imaging.
Conclusion: Our findings demonstrate significant heterogeneity of NSF appearance after skull base reconstruction. Flap enhancement, thickness, and adherence to skull base can change over time. While it is important for surgeons and radiologists to evaluate for variations in flap appearance, the absence of enhancement and lack of adherence to skull base on immediate imaging does not appear to affect postoperative CSF leak rates.
