The impact of comorbidity on treatment (chemoradiation and laryngectomy) of advanced, nondistant metastatic laryngeal cancer: a review of 16 849 cases from the national cancer database (2003-2008).

Objective: To investigate whether patients treated with laryngectomy had less comorbidity than those treated with chemoradiation, which could help explain the improved survival for the laryngectomy cohorts in recent studies.

Design: Observational cross-sectional study.

Patients: Patients receiving diagnoses of primary invasive advanced squamous cell carcinoma of the larynx between 2003 and 2008 were selected from the National Cancer Database, which collects information from more than 1400 facilities accredited by the American College of Surgeons' Commission on Cancer. Patient-level independent variables included age at diagnosis, sex, diagnosis year, race/ethnicity, primary payer status, and zip code-level education.

Main outcome measures: Primary treatment information. The association between treatment and patient clinical, sociodemographic, and facility-level and zip code-level socioeconomic status variables were analyzed using univariate statistics and multivariate models. Charlson Deyo Comorbidity and The Washington University Head and Neck Comorbidity Index scores were calculated from the hospital face sheet.

Results: The study demonstrated that receipt of treatment (chemoradiation vs total laryngectomy) was significantly associated with comorbidity. Treatment was not significantly associated with insurance status, race/ethnicity, or age. Patients with comorbidity were less likely to receive chemoradiation than subtotal or total laryngectomy, with a risk ratio (RR) of 0.84 (95% CI, 0.81-0.87) for patients with 1 or more comorbidities compared with those without any comorbidity, after controlling for factors such as tumor stage, age, race/ethnicity, insurance, and socioeconomic status. Patients were also less likely to receive chemoradiation than total laryngectomy if they had stage IV disease (RR, 0.81; 95% CI, 0.79-0.83) and if they had been diagnosed at a teaching or research institution (RR, 0.80; 95% CI, 0.77-0.84). Patients were more likely to receive chemoradiation if they were diagnosed after 2003 (RR, 1.37; 95% CI, 1.30-1.45) or if they lived in a zip code with a high percentage of high school graduates (RR, 1.1; 95% CI, 1.05-1.15).

Conclusions: This is the first study, to our knowledge, that demonstrates that patients with advanced laryngeal cancer with 1 or more comorbidities are more likely to receive surgery than chemoradiation compared with patients without any comorbidity, independent of numerous clinical and nonclinical variables among a large national cohort. A limitation of this study is the use of comorbidity data from the National Cancer Database, which gathers its information from hospital discharge face sheets. We recognize that the National Cancer Database may be an imperfect system for the collection of comorbidity data and encourage discussion on different methods to improve the system, including incorporating comorbidity data from the Surveillance, Epidemiology, and End Results Medicare Database and medical chart-based comorbidity data collection by cancer registrars.