Risk stratification in head and neck carcinomas: A dynamic scoring scheme to predict prognosis

Risk adapted treatment is a well proven strategy in several cancers. However such a strategy requires identification of proven prognostic variables as an initial requisite. In this study we sought to identify such prognostic factors and the feasibility of developing a risk stratification scheme. 212 proven head and neck cancers were enrolled for the study. Stage appropriate treatment as per national protocols instituted and study cohort followed for a period of 2 years. Prospective patient, disease related and treatment data including disease failure collected, an assortment of 15 potential prognostic variables were identified based on published evidence. These variables were subjected to univariate analysis and those showing significance (eight) subjected to multivariate analysis. At multivariate analysis PS, treatment related weight loss, AJCC stage IV, histological grade 3 and treatment interruption emerged as prognostic variables. Arisk stratification scheme incorporating these factors segregated the study cohort into favorable, low and high risk groups was defined. A survival analysis of these risk groups predicted poor survival for the high risk group P=0.001 (log rank test).