Development and validation of a novel prognostic score for elderly head-and-neck cancer patients undergoing radiotherapy or chemoradiation

Abstract

Background and purposeTo establish a clinically feasible prognostic score and nomogram based on easily accessible clinical data that will aid decision-making in elderly head-and-neck squamous cell carcinoma (HNSCC) patients undergoing (chemo)radiotherapy. Material and methods284 elderly HNSCC patients (≥65 years) undergoing curative (chemo)radiotherapy were included for the development of a score predicting overall survival (OS) based on the beta regression coefficients from significant parameters in a multivariate Cox regression analysis with p < 0.1 as inclusion criterion. A second, external cohort of 217 elderly HNSCC patients receiving (chemo)radiotherapy was used for validation. Using the aggregated data (n = 501), a nomogram was developed to predict 2- and 4-year OS. ResultsKarnofsky Performance Status (HR = 2.654; p < 0.001), Charlson Comorbidity Index (HR = 2.598; p < 0.001) and baseline C-reactive protein (CRP) level (HR = 1.634; p = 0.068) were prognostic for OS in the multivariate analysis. An OS score based on beta regression coefficients was created, in which reduced performance status, increased comorbidity burden and increased CRP levels were included, leading to 3 distinct survival groups. The median OS for the 3 groups amounted to 107, 28 and 6 months, respectively (p < 0.001). The developed score was able to significantly differentiate between a favorable (median OS = 130 months), intermediate (29 months) and unfavorable prognosis (9 months) also in the external validation cohort (p = 0.005). ConclusionWe propose a novel, validated prognostic score based on easily accessible clinical data allowing stratification between prognostic groups of elderly HNSCC patients receiving (chemo)radiotherapy. The derived nomogram for the prediction of 2-year and 4-year OS may aid decision-making for this vulnerable population.