Using Inflammatory Blood Markers as Useful Prognostic Factors in Patients Undergoing Neo-Adjuvant Therapy for Stage II-III Breast Cancer.

Abstract

PURPOSE/OBJECTIVE(S)\nThe prognostic value of inflammatory blood markers (IBMs) in breast cancer patients have been an area of growing interest. Most of the studies took parameters from one defined point, usually the pretreatment ones. Current study aims to investigate the potential prognostic significance of IBMs among patients with stage II-III breast cancer receiving neo-adjuvant therapy (NAT).\n\n\nMATERIALS/METHODS\nPatients with stage II-III breast cancer who received NAT and postoperative radiotherapy (RT) at a single institute from Jan 2009 to Dec 2016 were retrospectively reviewed. The neutrophil, lymphocyte, platelet and monocyte count were collected at different stages of treatment, after NAT but before surgery, before and after RT, respectively. LMR and SII was calculated for each patient. The breast cancer specific survival (BCSS) and locoregional recurrence free survival (LRRFS) were defined from the date of the start of RT until last follow up. Univariate and multivariate Cox regression analysis was performed to identify independent predictors for BCSS and LRRFS. Model performance was evaluated by the concordance index (c-index).\n\n\nRESULTS\nIn total, 205 patients were enrolled. After a median follow-up of 65 months, 19 BCSS events were identified. The optimal cutoff value of SII and LMR after NAT but before surgery by receiver operating characteristic curve were 467.76\u202f×\u202f109 (AUC\u202f=\u202f0.63) and 1.57 (AUC\u202f=\u202f0.65). The five-year BCSS of the group with lower LMR (LMR≤1.57) and higher LMR (LMR > 1.57) were 74.2% and 93.7%, respectively (P\u202f=\u202f0.001). The 5-year risk of BCSS in patients with high SII value (SII > 467.76\u202f×\u202f109) and low SII value (SII≤467.76\u202f×\u202f109) were 81.5% vs 96.4%, respectively, P\u202f=\u202f0.003. The multivariate analysis found that in addition to the traditional prognostic factors including hormone receptor, ypT, ypN stage and lymphovascular invasion, both post-NAT LMR (HR\u202f=\u202f0.28, 95% CI\u202f=\u202f0.10-0.81, P\u202f=\u202f0.02) and SII (HR\u202f=\u202f4.06, 95% CI\u202f=\u202f1.37-12.00, P\u202f=\u202f0.01) were independent predictors for BCSS. The C-index of the multivariate model of Neo-Bioscore for BCSS was 0.71(95% CI\u202f=\u202f0.83-0.59), which was improved to 0.80((95% CI\u202f=\u202f0.94-0.67) by adding LMR and SII. The prognosis of pre-RT and post-RT SII on LRRFS was analyzed in124 patients. The optimal cutoff value of SII before and after RT were 412.2\u202f×\u202f109 (AUC\u202f=\u202f0.71) and 571.48\u202f×\u202f109 (AUC\u202f=\u202f0.69) respectively. Patients with pre-RT SII > 412.2\u202f×\u202f109 or post-RT SII > 571.48\u202f×\u202f109 had a higher rate of locoregional failure (P\u202f=\u202f0.02 and P\u202f=\u202f0.01). After adjusting for cT, cN stage and age, post-RT SII remained predictive of locoregional failure (HR\u202f=\u202f10.22, 95% CI\u202f=\u202f1.25-83.45, P\u202f=\u202f0.03).\n\n\nCONCLUSION\nIBMs after NAT was useful prognostic factor for stage II-III breast cancer patients. Adding LMR and SII to Neo-Bioscore can further stratify the prognosis of BCSS. High level of SII after RT had a negative impact on local-regional control, whether it reflects preclinical studies of post-RT inflammation-mediated tumor recurrence needs further study.