Primary cancer of the sphenoid sinus--a GETTEC study.

BACKGROUND Primary involvement of the sphenoid sinus occurs in 2% of all paranasal sinus tumors and is associated with dismal prognosis. Optimal management remains debatable. METHODS A total of 23 patients were treated for a primary cancer of the sphenoid sinus from 1988 to 2004. Charts were reviewed for patient-, tumor-, and treatment-related parameters. Univariate and multivariate analyses were conducted to identify prognostic factors for locoregional control and survival. RESULTS Cranial neuropathies were present in 12 patients. Pathologic findings included adenoid cystic carcinoma, adenocarcinoma, lymphoma, squamous cell carcinoma, sarcoma, neuroendocrine carcinoma, melanoma, and malignant hemangiopericytoma. All but 2 patients had stages III to IV cancer. Radiotherapy was performed in 18 patients and chemotherapy in 12. Of 10 patients undergoing surgery, total excision with grossly negative margins was achieved in 4 patients and subtotal resection in 6. Median locoregional control and overall survival were 12 and 41 months, respectively. On multivariate analysis, cranial neuropathy was associated with worse locoregional control and survival. Surgery was rarely complete because of advanced stages at presentation, but it yielded better outcomes than other treatments without surgery in non lymphoma-cases. CONCLUSION Early CT and MRI should be performed when facing aspecific, rhinological, or neuro-ophtalmological symptoms. Cranial neuropathies indicate a worse prognosis. Surgery, including debulking surgery, may be preferred to combined modality treatments without surgery. Its apparently favorable impact on prognosis would need to be tested in homogenous histological groups of patients, which is impossible because of the rarity of the disease. Highly conformal radiotherapy (adjuvant or definitive) should be encouraged and optimized with concurrent chemotherapy in advanced stages. Aggressive multidisciplinary management including surgery, chemotherapy, and radiotherapy should be encouraged and adapted on histology and tumor extensions. Progress is still warranted to improve outcomes.