Predictive factors of survival and treatment tolerance in older patients treated with chemotherapy and radiotherapy for locally advanced head and neck cancer

Purpose: To report outcomes and predictive factors of overall survival, hospitalization and treatment completion rates in elderly patients with locally advanced head and neck cancer treated with concurrent chemoradiotherapy (CRT). Material and methods: A retrospective analysis of patients aged 70 years or older treated with concurrent CRT for locally advanced head and neck cancer was conducted. Univariate and multivariate analysis as well as competing risk survival analysis were used to determine predictors of mortality. Logistic regression was used to predict for hospitalization and treatment completion rates. Results: In total, 129 patients were included. Median follow-up was 27 months (range: 1.7-125 months). Completion rate of combined CRT was 84%. Actuarial OS and DSS at 4 years were 56% and 75%. Hospitalization rate was 36%. On multivariate analysis, a Karnofsky performance status (KPS) <= 80 was predictive of mortality. Using competing risks, KPS <= 80 and weight loss >5% were predictive of cancer mortality whereas Charlson score >= 3 was predictive of mortality due to other causes. On logistic regression, patients with abnormal renal function and lower body mass index were more likely to be hospitalized during their treatment course. Charlson score and chemotherapy regimen were predictive of treatment completion. Conclusion: Concurrent CRT may be a feasible treatment option for healthier older patients at the cost of high hospitalization rates. Pre-treatment factors linked to physiological age such as KPS <= 80, Charlson score >= 3, abnormal renal function should be considered at the time of treatment decision. (C) 2015 Elsevier Ltd. All rights reserved.