Novel Signatures Based on the Lymphocyte-to-C-Reactive Protein Ratio Predict the Prognosis of Patients with Early Breast Cancer: A RetrospectivStudy

Abstract

Background: The value of the lymphocyte-to-C-reactive protein (CRP) ratio (LCR) in early breast cancer (BC) is unclear. We explored the correlation between the LCR and survival of patients with early BC and established effective LCR-based prognostic signatures for predicting prognosis. Methods: In this retrospective study, we randomized 623 patients with early-stage BC diagnosed in December 2010 to October 2012 at the Sun Yat-sen University Cancer Center to training and verification datasets. The median follow-up of all patients was 109 months. The survival differences were calculated by Kaplan–Meier method using the Log rank test. For overall survival (OS) and disease-free survival (DFS), the independent factors in the training dataset were identified using univariate and multivariate Cox analyses, in which two-tailed P-values <0.05 were considered statistically significant. Based on this, we respectively constructed novel signatures for survival prediction and validated the efficiency of signatures through the concordance index (C-index), calibration and receiver operating characteristic (ROC) curves in both datasets. Results: The LCR, lymphatic vessel invasion (LVI), progesterone receptor (PR) status, and Ki67 index were independent prognostic factors of OS. And the LCR and LVI are associated to DFS too. High LCR was associated with better OS and DFS. We constructed the prediction signatures based on those independent prognostic factors and calculated the risk scores. Patients in the training dataset with higher risk scores had significantly worse prognosis (P <0.001). The signature had excellent discrimination capacity, with an OS C-index of 0.785 [95% confidence interval (CI): 0.713–0.857] and 0.750 (95% CI: 0.669– 0.832) in the training and verification datasets, respectively. The time–ROC curves also suggest accurate prediction by the signature. Conclusion: The LCR was a significant prognostic predictor of OS and DFS in early BC. The LCR-based prognostic signatures could be a useful tool for individualized therapeutic guidance.