Abstract
BACKGROUND: Peripheral blood inflammation factor neutrophil-lymphocyte ratio (NLR), platelet count (PLT) and nutritional factor serum albumin (ALB) have been proposed as prognostic markers of head and neck squamous carcinoma cancer (HNSCC) in recent years. In the current study, nomogram predict models based on pre-treatment hematological parameters and a modified risk-stratified score system have been built. METHODS: A total of 197 patients with oropharyngeal, hypopharyngeal and laryngeal cancers receiving multimodality treatment between 2012 and 2014 were included. The pre-treatment ALB, neutrophil, lymphocyte and platelet count (PLT) were detected. Cancer-specific survival and locoregional recurrence (LRC) by 5 years' follow-up in the cases were obtained. To integrate clinical characteristics, we propose a modified risk-stratified score system. Kaplan-Meier method, proportional hazards COX model, logistic models were used to establish nomograms within external validation. RESULTS: Five-year LRC was decreased (p=0.004) for 140 patients with pre-treatment NLR <2.77. Five-year LRC and 5-year cancer-specific survival were decreased (p=0.031, p=0.021) with pre-treatment PLT ≥248×10(9)/L. Comparison of univariate parametric models demonstrated that pre-treatment NLR evaluation and PLT>248×10(9)/L were better among tested models. On Bayesian information criteria (BIC) analysis, the optimal prognostic model was then used to develop nomograms predicting 3- and 5-year LRC. The external validation of this predictive model was confirmed in 57 patients from another hospital. CONCLUSION: Pre-treatment NLR elevation and PLT>248×10(9)/L are promising predictors of prognosis in patients with operable HNSCC. Nomograms based on the pre-treatment hematological markers and modified risk-stratified score system provide distinct risk stratifications. There results provided the feasibility of anti-inflammatory and antiplatelet treatments for HNSCC patients.