Analysis of Racial Survival Disparities in Head and Neck Cancer by Site of Primary Tumor

Abstract

Objectives:

Head and neck cancer carries a significant survival disparity by race and site of primary tumor. We performed survival analysis to determine how racial survival disparities differ by site.

Methods:

We selected 34,182 patients from the Surveillance, Epidemiology, and End Results (SEER) database with head and neck cancer (HNC) diagnosed from 2004-2009. Eighty-three point nine percent were white, 11.4% black, 4.28% Asian or Pacific Islander, and 0.36% American Indian. All patients with >1 primary and primaries outside the upper aero-digestive tract were excluded. American Joint Committee on Cancer (AJCC) staging in SEER was recorded with stages I and II as early stage, and stages III and IV as late stage. Hazardous ratio were calculated with and without stage as covariates.

Results:

Stage independent: hazard ratio (HR) for black vs. white survival was 1.637 (95% confidence interval [CI] = 1.455-1.842, P <.001) for mouth, 2.402 (95% CI = 2.182-2.644, P <.001) for oropharynx, 1.382 (95% CI = 1.275-1.530, P =<.001) for larynx, and 1.563 (95% CI = 01.337-1.827, P = <.001) for hypopharynx. Stage dependent: HR of 1.297 (95% CI = 1.152-1.460, P = <.001) for mouth, 2.349 (95% CI = 2.134-2.586, P <.001) for oropharynx, 1.124 (95% CI = 1.014-1.246, P = .025) for larynx, and 1.541 (95% CI = 1.318-1.802, P =<.001) for hypopharynx.

Conclusions:

We found racial survival disparities in HNC for all sites, with differing reasons by location. The mouth and larynx’s disparity is significantly affected by stage at presentation, but primaries of the oropharynx are not. Therefore, poor tumor recognition is a major reason for black survival disparities in early diagnostic areas. However, in inferior upper aero-digestive tract lesions the disparity is not due to a difference in tumor recognition and likely stems from differences in treatment or risk factors such as human papillomavirus status.