Prediction of recurrence free survival for esophageal cancer patients using a protein signature based risk model

Background: Biomarkers to predict the risk of disease recurrence in Esophageal squamous cell carcinoma (ESCC) patients are urgently needed to improve treatment. We developed proteins expression-based risk model to predict recurrence free survival for ESCC patients. Methods: Alterations in Wnt pathway components expression and subcellular localization were analyzed by immunohistochemistry in 80 ESCCs, 61 esophageal dysplastic and 47 normal tissues; correlated with clinicopathological parameters and clinical outcome over 86 months by survival analysis. Significant prognostic factors were identified by multivariable Cox regression analysis. Results: Biomarker signature score based on cytoplasmic β-catenin, nuclear c-Myc, nuclear DVL and membrane α-catenin was associated with recurrence free survival [Hazard ratio = 1.11 (95% CI = 1.05, 1.17), p < 0.001, C-index = 0.68] and added significant prognostic value over clinical parameters (p < 0.001). The inclusion of Slug further improved prognostic utility (p < 0.001, C-index = 0.71). Biomarker Signature Score(slug) improved risk classification abilities for clinical outcomes at 3 years, accurately predicting recurrence in 79% patients in 1 year and 97% in 3 years in high risk group; 73% patients within low risk group did not have recurrence in 1 year, with AUC of 0.76. Conclusions: Our comprehensive risk model predictive for recurrence allowed us to determine the robustness of our biomarker panel in stratification of ESCC patients at high or low risk of disease recurrence; high risk patients are stratified for more rigorous personalized treatment while the low risk patients may be spared from harmful side effects of toxic therapy.