Elective Neck Dissection in cT1-4 N0M0 Head and Neck Spindle Cell Carcinoma

Objective To investigate the survival benefit of elective neck dissection (END) over neck observation in surgically resected cT1-4 N0M0 head and neck spindle cell carcinoma (HNSpCC). Study Design Retrospective cohort study. Setting The 2006 to 2018 hospital-based National Cancer Database (NCDB). Methods Patients with surgically resected cT1-4 N0M0 HNSpCC were selected. Linear, binary logistic, Kaplan-Meier, and Cox proportional hazards regression models were implemented. Results Of 815 patients satisfying inclusion criteria, a high proportion were male (72.4%) and white (86.0%) with disease of the larynx (43.8%) classified as high grade (87.7%) and cT1-2 (74.8%). In total, 235 (28.8%) patients underwent END. END utilization between 2006 and 2018 increased for cT1-2 disease (4.0% vs 30.5%, R-2 = 0.731) and for cT3-4 disease (15.4% vs 84.6%, R-2 = 0.606). In total, 58 (24.7%) ENDs detected occult nodal metastases (ONMs). The 5-year overall survival (OS) of patients undergoing neck observation and END was 62% and 54%, respectively (P = .215). Among patients undergoing END, patients with ONM had worse 5-year OS than those without ONM (38% vs 60%, P < .001). On multivariable Cox regression, END was not associated with OS (adjusted hazard ratio [aHR] 0.74, 95% CI 0.68-1.32, P = .735); ONM (hazard ratio [HR] 2.01, 95% CI 1.29-3.12, P = .002) was associated with worse OS. Conclusion END is performed in a high proportion (30%) of patients with cN0M0 HNSpCC but is not associated with higher OS. The rate of ONM approaching 25% and the association between ONM and worse OS, however, justify consideration of END in HNSpCC.