Objectives: Minor salivary gland carcinoma (MSGC) rarely occurs in the larynx. Prior large studies of MSGC have focused almost exclusively on oral cavity and oropharyngeal subsites, with analyses of laryngeal MSGC limited to a few single-institution studies. Surgical approaches to the larynx and the morphology and distribution of laryngeal minor salivary glands differ significantly from the oral cavity and the oropharynx. Consequently, laryngeal MSGC merits analysis as an independent entity. The study analyzes the epidemiology, patient outcomes, and prognostic factors of laryngeal MSGC with a large, population-based database.
Study Design: Cross-sectional analysis of the National Cancer Institute’s SEER database (Surveillance, Epidemiology, and End Results).
Methods: We reviewed the SEER database for all cases of MSGC of the larynx from 1973 to 2013. Relevant demographic, clinicopathologic, and survival variables were extracted and analyzed. Tumor histology was categorized as mucoepidermoid carcinoma, adenocarcinoma, adenoid cystic carcinoma, or rare carcinomas. The rare carcinomas group encompassed any histology that represented less than 5% of tumors. Cox multivariate regression was performed to identify prognostic factors.
Results: We identified 311 cases of MSGC of the larynx (mean age, 63 years; 69% male). The supraglottis (41.2%) was the most commonly involved subsite, followed by not otherwise specified subsites (23.8%), the glottis (22.8%), and the subglottis (12.2%). The most common histologic subtypes were adenocarcinoma (37.6%), adenoid cystic carcinoma (23.2%), rare carcinomas (22.8%), and mucoepidermoid carcinoma (16.4%). Five- and 10- year rates of disease-specific survival (DSS) were 61.8% and 52.5%, respectively. Survival varied by tumor histology with adenoid cystic carcinoma conveying a significantly better prognosis (5-yr DSS 77.3%) than adenocarcinoma (57.1%), rare carcinomas (56.2%), or mucoepidermoid carcinoma (52.7%) (P < .001). After controlling for other variables, however, tumor histology did not remain a significant predictor of survival. In multivariate models, independent prognostic factors included high grade (P < .001) and advanced stage (P<.001). Age and anatomic subsite were not independently associated with disease-specific survival.
Conclusion: This study represents the largest series of laryngeal MSGC to date. Tumor grade but not tumor histology or anatomic subsite were independently predictive of poor survival. Advanced stage was also a poor prognosticator.