Accuracy of intraoperative frozen margins for sinonasal malignancies and its implications for endoscopic resection of sinonasal melanomas

Background: The main objective of endoscopic tumor surgery remains similar to open approaches, with the goal being total tumor resection with clear margins. Beyond cosmesis, endoscopes offer the advantage of limiting the size of the resection as well as aiding in the procurement of tissue margins in areas adjacent to critical structures or deep in the sinonasal cavity. Because of the close proximity of these tumors to critical structures and classic otolaryngology teaching with the goal sparing normal sinonasal mucosa, sinonasal tumor resection margins tend to be more conservative than those practiced for the same type of tumor in a different anatomic location. What is not uniformly agreed upon is the optimal margin of resection as well as the reliability of intraoperative frozen margins for the varied histologic subtypes seen in sinonasal malignancies. Methods: Retrospective review of malignant sinonasal tumors resected endoscopically by 1 surgeon at 2 institutions between 2006 and 2011. Results: Thirty-one patients with mixed histologies were identified, with the most common being mucosal melanoma (25.8%) and squamous cell carcinoma (23.3%). The overall false-negative rate for intraoperative frozen margins was 6.5%, with both false negatives associated with mucosal melanoma. The false-negative margin rate for mucosal melanoma was 25%. Conclusion: Intraoperative frozen margins for sinonasal tumors are reliable for most histologic subtypes, with the exception of those for sinonasal mucosal melanomas. This has implications for the size of margins needed for the resection of sinonasal melanomas as they may need to be larger than those for other tumors.