Limits of endoscopic endonasal approach for cranio-vertebral junction tumors

The endoscopic endonasal approach has been recently proposed for cranio-vertebral junction lesions. The more common indication for this sagittal extension of the endonasal route is represented by odontoidectomy for irreducible ventral brainstem compression due to congenital or degenerative conditions. However, in an increasing number of studies its adoption for tumors involving the cranio-cervical junction has been reported. The aim of this study is to consider retrospectively our surgical series, focusing on the advantages and limits of this approach.

Each consecutive case of tumor involving the cranio-vertebral junction since 2007 to 2017 treated through an endoscopic endonasal approach has been included. All patients undergone preoperative neurological examination and neuroimaging (magnetic resonance imaging [MRI] and computed tomography angiography). These examinations were repeated after 3 months and then annually. Complementary treatments, recurrence rate and clinical status at mean follow-up of 18±7.3 months were considered.

Seven patients have been included in this study, mean age was of 47±17 years; male-to-female ratio was of 3:4. Series is composed by 6 chordomas and one foramen magnum meningioma. One patient had been already posteriorly stabilized for cranio-vertebral instability. Gross tumor removal was achieved in two cases, in the others a subtotal removal was demonstrated at postoperative MRI. One patient presented a transitory worsening of CN XII palsy, resolved within 3 months. For preoperative dysphagia and inhalation pneumonia, one case undergone tracheostomy and was fed with oro-gastric tube for 10 days. Three patients died for chordoma progression and at follow-up one presented a local recurrence.

Despite our experience is preliminary, the endoscopic endonasal approach has resulted safe for cranio-cervical junction tumor with a reduced number of complications. It can give a straight and direct trajectory to this deep region. We suggest that lateral extension of the tumor beyond the plane of cranial nerves is a limit for this approach, as well as an inferior expansion caudal to C1. Larger series and longer follow-up are required to assess the proper indications of this approach.