Endoscopic Endonasal Approach for Craniopharyngiomas with Intraventricular Extension: Case Series, Long-Term Outcomes, and Review

Background

Traditionally, craniopharyngiomas with intraventricular extension were approached transcranially; however, endoscopic approaches are now increasingly used. We sought to study the endoscopic endonasal approach (EEA) in the setting of complex craniopharyngiomas with intraventricular extension and to compare it with existing literature.

Methods

Patients undergoing EEA for resection of craniopharyngioma with ventricular involvement from 2002 to 2015 were retrospectively reviewed. Outcomes were compared with previously published EEA and transcranial approach (TCA) studies for all craniopharyngioma locations.

Results

Sixty-two patients were included. Average tumor and intraventricular volume were 13.93 cm³ and 2.61 cm³, respectively. Patients presented with visual impairment, endocrinopathy, and, headache. Gross total resection (GTR) was achieved in 47% of all cases and increased to 77% after 2012 Approximately 98% experienced improvement or stability of vision. Postoperative cerebrospinal fluid (CSF) leak and meningitis rates were 19% and 8.1%, respectively. However, nasoseptal flap (NSF) use reduced CSF leak rate to 10%. Six (9.6%) patients required shunting before resection and 25% were shunted postoperatively. Seven of 10 patients (70%) treated before NSF use required shunting, whereas only 7 of 46 (15%) required shunting with NSF reconstruction. Review demonstrated similar outcomes between the present cohort and EEA or TCA for all craniopharyngioma locations. TCA had a greater GTR, however, with large study variation. EEA showed improved visual outcomes but also increased CSF leaks.

Conclusions

EEA for craniopharyngiomas with intraventricular extension shows similar outcomes to TCA and EEA for all craniopharyngiomas, expanding this anatomic limit. Given ventricular involvement, CSF leak rates are expectedly high. GTR increased and CSF leak rates dramatically decreased with time, suggestive of the steep learning curve to complex resection.

Introduction

Although histologically benign, craniopharyngioma's relatively deep location and tendency to adhere to critical structures makes total resection challenging. Size and degree of spread outside the suprasellar space drives choice of surgical approach. Thus, given their propensity for suprasellar spread, open, transcranial approach (TCA) often was used.1, 2, 3 However, craniopharyngiomas also may arise from the sella or ventricle. The relationship of craniopharyngiomas to the optic chiasm and the narrow corridor above the pituitary gland initially precluded the transsphenoidal approach, but with growing application and experience, the endoscopic endonasal approach (EEA) has increasingly been used. EEA provides superior visualization to the undersurface of the chiasm and hypothalamus—a relative blind spot of the TCA.⁴ EEA may produce similar, and in some regard superior, outcomes for craniopharyngioma resection.5, 6, 7, 8, 9, 10

Tumors with ventricular involvement may be approached via open craniotomy; however, limited suprasellar access, risk of forniceal, pericallosal artery, and/or internal cerebral vein injury are serious limitations.⁴ Yet, craniopharyngiomas with intraventricular extension occur frequently, given involvement of the adjacent infundibulum and hypothalamus.¹¹ Thus, the ventricular cavity is considered an anatomic limit of EEA¹² and is often a surrogate for hypothalamic involvement. Kassam et al.¹³ proposed an anatomic classification system for craniopharyngiomas as: preinfundibular (type I), transinfundibular (type II), retroinfundibular (III), retroinfundibular/third ventricular (type IIIa), retroinfundibular/interpeduncular cistern (type IIIb), and purely third ventricular (type IV).^12,13 However, multiple other classification schemes exist, and ventricular extension also can be closely described either as primarily intraventricular, infundibulotuberal, secondarily intraventricular, and pseudointraventricular.¹⁴ These nuanced schemes suggest the technical modifications necessary in resecting these lesions based on relationship to the third ventricular floor. These endoscopic resections require appropriate endoscopic expertise. Thus, the limits of EEA can be widened to the surgeon's experience, tools available, and then anatomic boundaries.

The anatomic limits of EEA have gradually expanded since its initial application.⁸ This study describes outcomes after EEA for craniopharyngiomas that extend into the third ventricle. These outcomes are compared with a review of other studies regarding EEA and TCAs for resection of craniopharyngiomas of all locations.