Prediction of outcome in breast cancer patients using test parameters from complete blood count

The aim of this study was to evaluate the prognostic effect of test parameters from pretreatment complete blood count (CBC) for predicting outcome in breast cancer patients. A total of 162 patients with breast cancer and a long follow-up were enrolled in this study. Red cell indices (RCIs) and neutrophil-lymphocyte ratio (NLR) from CBC prior to treatment, as well as related clinical data, were retrospectively collected. We evaluated the association of RCI and NLR with tumor size, clinical stage, histological grade, estrogen receptor, progesterone receptor, and human epidermal growth factor receptor 2 status. We further performed survival analysis and Cox multivariate analysis, stratified by RCI and NLR median values, to evaluate their prognostic effects. In the disease-free survival (DFS) analysis, patients in the higher mean corpuscular hemoglobin (MCH) and NLR groups exhibited shorter DFS times compared with those in the lower MCH and NLR groups (P=0.017 for MCH and P=0.039 for NLR). The univariate analysis revealed that both MCH and NLR were significantly associated with DFS. The Cox multivariate analysis demonstrated that only MCH was an independent predictor associated with disease relapse (hazard ratio = 1.975, 95% confidence interval: 1.118-3.487, P=0.019), whereas no index was associated with overall survival. Our results suggest that MCH prior to treatment may be a predictive marker associated with DFS in breast cancer.