Socio-demographic variables influence outcome in Radiation Therapy Oncology Group head and neck trials

Abstract

6043 Background: The specific aim of this study was to evaluate the influence of socio-demographic factors on local-regional (LR) control and survival of patients treated on RTOG head and neck cancer clinical trials. Methods:Patients treated on RTOG 9003, RTOG 9111 and RTOG 9703 were utilized. The model was stratified by protocol and a step-up procedure using a 0.05 significance level was initially used to build the model. The Cox proportional hazards model was used to determine if any tumor or socio-demographic variables (age, gender, race, Karnofsky Performance Status (KPS), T-stage, N-stage, marital status, number of persons in household, education level, household income prior to cancer diagnosis, and tobacco use) are independent prognostic variables for LR control and/or survival. Results:A total of 1901 patients were entered onto the 3 RTOG trials of which 1822 were analyzable. Fifty-nine percent of the analyzable patients were from 9003, 28% from 9111 and 13% from 9703. In multivariate analyses T and N-Stage, and KPS were significant for both survival and LR control. Variables found to have prognostic value independent of disease related variables were gender, age, marital status and income for survival and marital status and race for LR control. A significant interaction between marital status and gender existed for both endpoints. Male patients currently married or a live-in partner, did significantly better than the other male patients, while no such effect was seen in the female patients. Tobacco use at protocol entry was evaluated in the multivariate model with only RTOG 9003 and 9111, and was significant for both endpoints. Conclusions: The socio-demographic variables race and marital status were found to have independent prognostic value for predicting LR control. Income and marital status were found to have independent prognostic value for predicting survival. These variables may identify patients requiring additional support and early interventions during the course of treatment to improve outcome. No significant financial relationships to disclose.