Endoscopic endonasal surgery for nonadenomatous, nonmeningeal pathology involving the cavernous sinus.

OBJECTIVE Surgery within the cavernous sinus (CS) remains a controversial topic because of the delicate and complex anatomy. The risk also varies with tumor consistency. Softer tumors such as pituitary adenomas are more likely to be surgically treated, while firm tumors such as meningiomas are often treated with radiosurgery. However, a wide range of pathologies that can involve the CS are amenable to surgery. The authors describe and analyze their results using endonasal endoscopic "medial-to-lateral" approaches for nonadenomatous, nonmeningeal tumors, in relation to the degree of invasion within the CS. METHODS A prospectively acquired database of consecutive endoscopic approaches for tumors with verified intraoperative CS invasion was reviewed. Pituitary adenomas and meningiomas were excluded. Degree of invasion of the CS was classified using the Knosp-Steiner (KS) grading system as well as the percentage of cavernous carotid artery (CCA) encasement. Extent of resection of the entire tumor and of the CS component was assessed by independent neuroradiologists using volumetric measurements of the pre- and postoperative MRI studies. Demographic data and complications were noted. RESULTS Fifteen patients (mean age 51.1 years who received endoscopic surgery between 2007 and 2013 met the selection criteria. There were 11 malignant tumors, including chordoma, chondrosarcoma, hemangiopericytoma, lymphoma, and metastatic cancer, and 4 benign tumors, including 3 cavernous hemangiomas and 1 dermoid. All cases were discussed before treatment in a tumor board. Adjuvant treatment options included chemotherapy and radiotherapy. The mean pre- and postoperative tumor volumes were 12.74 ml and 3.86 ml. Gross-total resection (GTR; ie, resection greater than 95%) was the goal in 13 cases and was achieved in 6 patients (46%) while in addition 5 patients had a greater than 80% resection. Gross-total resection in the CS was accomplished in 55% of the tumors with KS Grades 1-2 and in 16.6% of the tumors with KS grades 3-4, respectively. Likewise, GTR was accomplished in 55% of the tumors with CCA encasement under 75% and in 14.3% of the lesions with CCA encasement over 75%, irrespective of tumor volume and underlying pathology. There were 18 preexisting cranial neuropathies involving cranial nerves III-VI, of which 9 fully resolved, 4 improved, and 3 remained unchanged; 2 of these worsened with tumor recurrence. Surgical complications included 1 transient new cranial nerve VI palsy associated with Horner's syndrome and 1 case of panhypopituitarism. There were no postoperative CSF leaks and no infections. The mean extended follow-up was 34.4 months. CONCLUSIONS Endonasal endoscopic approaches can play a role in the management of nonmeningeal, nonadenomatous tumors invading the CS, either through biopsy, debulking, or GTR. An advantage of this method is the relief of preexisting cranial neuropathies with low risk for new neurological deficit. Extent of resection within the CS varies with KS grade and degree of carotid encasement irrespective of the underlying pathology. The goals of surgery should be clearly established preoperatively in consultation with radiation and medical oncologists.