Date of Award


Document Type

Open Access Thesis

Degree Name

Medical Doctor (MD)

First Advisor

Joanne Weidhaas

Second Advisor

Susan Higgins

Third Advisor

Douglas Ross, Roy Decker


We hypothesized that lymph node involvement in minor salivary gland cancers is associated with clinical and pathological factors commonly available to the clinician after a typical initial workup. Our aim was to identify these factors using a dataset that allowed us to compile the largest series of minor salivary gland cancers in the published literature. Using this dataset we also aimed to characterize the distribution of histological types by primary site, identify the predictors of the use of external beam radiation therapy and neck dissection, and examine the effect of lymph node involvement on survival. Using the SEER database, we identified 2667 minor salivary gland cancers with known lymph node status from 1988 to 2004. Univariate and multivariate analyses were conducted to identify factors associated with the use of neck dissection, the use of external beam radiation therapy, and the presence of cervical lymph node metastases. Kaplan Meier survival curves were constructed to examine the effect of lymph node involvement on survival. 426 (16.0%) patients had neck nodal involvement. Factors associated with neck nodal involvement on univariate analysis included increasing age, male gender, increasing tumor size, high tumor grade, T3-T4 stage, adenocarcinoma or mucoepidermoid carcinomas, and pharyngeal site of primary malignancy. On multivariate analysis, four statistically significant factors were identified, which included male gender, T3-T4 stage, pharyngeal site of primary malignancy, and high-grade adenocarcinoma or high-grade mucoepidermoid carcinomas. The proportions (and 95% confidence intervals) of patients with lymph node involvement for those with 0, 1, 2, 3 and 4 of these prognostic factors were 0.02 (0.01-0.03), 0.09 (0.07-0.11), 0.17 (0.14-0.21), 0.41 (0.33-0.49), and 0.70 (0.54-0.85) respectively. Grade was a significant predictor of metastasis for adenocarcinoma and mucoepidermoid carcinoma but not for adenoid cystic carcinoma. Overall survival was significantly worse at 5, 10, and 15 years for patients with lymph node involvement on presentation. A prognostic index using the four clinicopathological factors listed above can effectively differentiate patients into risk groups of nodal metastasis. The precision of this index is subject to the limitations of SEER data and it should be validated in further clinical studies.