Prognostic Value of Tumor Size in T2 Glottic Laryngeal Cancer Staging

Abstract

We report on the largest detailed analysis of T2 glottic laryngeal cancer using the National Cancer Database (NCDB) and identify patient subsets within the current T2 stage grouping with significantly different overall survival (OS). Adults with glottic laryngeal squamous cell carcinoma were identified in the 2004-2015 NCDB, grouped according to AJCC 8 staging as T2N0M0, classified by supra/sub-glottic extension (SGE) vs. impaired vocal cord mobility (VCM), stratified by tumor size and analyzed according to demographics, HPV status, disease extent and treatment received. Patients with missing data and non-squamous histologies were excluded. Propensity score-matched analysis (PSM) was utilized to create balanced cohorts between SGE and impaired VCM subgroups. Additionally, a second PSM was performed to create balanced cohorts between tumor size subgroups (≤ 2cm vs. >2cm). Kaplan-Meier statistics and Cox proportional hazards regression modeling were used for these two sets of matched cohorts. Of 11,375 patients identified, 4,679 met inclusion criteria; 3,862 (82.5%) male and 817 (17.5%) female. Median age 66 years (range, 20-90 years). Median follow-up 3.8 years (range, 0.02-12.8 years). Tumor size was documented in 1,213 cases (26%), with the majority (75%) of tumors ≤ 2cm. In propensity score matched analysis comparing T2N0 glottic cancers with impaired VCM (n=965) and those with SGE (n=965), no significant difference in 10-year OS was found (36% vs 37%, log-rank test, P= .93; HR, 1.03; 95% CI, 0.89-1.18; P= .72), while tumor size >2cm was a predictor for poorer OS at 10-years (HR, 2.11; 95% CI, 1.59-2.80; P< .001). 10-year OS remained statistically significant after PSM between tumor size ≤ 2cm and >2cm, favoring ≤ 2cm (38% vs 22%, log-rank test, P< .001; HR, 1.65; 95% CI, 1.29-2.11; P= .01). In T2N0 glottic laryngeal cancer, 10-year OS is superior with tumors ≤ 2cm compared to those >2cm. These results suggest that current AJCC grouping of these two subsets of patients into a single T2 stage group may obscure the more favorable prognosis of patients with smaller tumors. Our findings support the consideration of T2a and T2b subdivisions for glottic cancers based on tumor size in future staging iterations.