Objectives: To identify factors associated with positive margins following surgical management of sinonasal squamous cell carcinoma (SCC).
Study Design: Retrospective analysis of the National Cancer Database (NCDB) 2004-2014.
Methods: Adult patients with sinonasal SCC clinically staged T1-T4a stage were identified from the NCDB. Factors associated with positive margins were analyzed using multivariable binary logistic regression. Cases from 2010-2014 had surgical approach (open versus endoscopic) available and were separately analyzed to assess the association of surgical approach with margin status. The association of margin status with overall survival (OS) was assessed via Kaplan-Meier log-rank tests and Cox multivariable regression, controlling for patient, tumor, and treatment characteristics.
Results: We identified 2,791 cases, of which 746 (26.7%) had positive margins. 1,696 (60.7%) were within the nasal cavity, 957 (34.3%) were within the maxillary sinus, and 139 (5.0%) were within the ethmoid sinus. On multivariable analysis, factors associated with positive margins include higher T stage (T2 vs T1: OR 1.380 [1.042-1.829]; T3 vs T1: OR 2.407 [1.820-3.185]; T4a vs T1: OR 2.805 [2.157-3.648]), less differentiated tumors (moderately-differentiated vs. well-differentiated: OR 1.336 [1.018-1.752]; poorly-differentiated vs. well-differentiated: OR 1.509 [1.128-2.019]), and tumors in the ethmoid sinus (versus nasal cavity; OR 2.009 [1.374-2.937]). Cases treated at facilities with higher volumes were associated with a lower likelihood of positive margins (OR 0.700 [0.564-0.869]). Analysis of cases with known surgical approach revealed similar trends. An endoscopic approach was not associated with an increased likelihood of obtaining positive margins (versus open; OR 1.111 [0.767-1.609]). Factors associated with endoscopic approach included tumors in the ethmoid sinus (versus nasal cavity; OR 3.206 [1.728-2.937]). However, factors associated with a decreased likelihood of endoscopic approach include high facility volume (OR 0.653 [0.448-0.950]) and clinically staged N1+ patients (OR 0.073 [0.010-0.540]). After controlling for patient, tumor, and treatment factors, positive margins were associated with decreased OS in the overall cohort (HR 1.590 [1.383-1.827]), surgical approach cohort (HR 1.462 [1.120-1.910]), and subgroup analyses of each primary site (nasal cavity: HR 1.394 [1.128-1.723]; maxillary sinus: HR 1.602 [1.308-1.961]; ethmoid sinus: HR 3.046 [1.575-5.889]).
Conclusion: In this study, we identified positive margin rates in sinonasal SCC. As expected, positive margins were associated with decreased OS even after controlling for patient, tumor, and treatment factors. We found that positive margins were less likely at cases performed at high-volume facilities, and more likely in the ethmoid sinus than other subsites. Importantly, there was no difference in likelihood of positive margins in endoscopic versus open surgery. Sinonasal SCC is rare, and patients may benefit from treatment at higher-volume facilities familiar with the procedure, particularly at subsites more at risk for positive margins. Endoscopic surgery may offer a safe, less-invasive alternative to open surgery for carefully selected patients.