Association of blood inflammatory/immune markers with outcomes in (neo)adjuvant breast cancer: A large single institutional study.

Abstract

e12540 Background: An inflammatory state in various cancer populations may correlate with mortality. Neutrophil-lymphocyte ratio is a surrogate marker of an inflammatory state. A recent meta-analysis showed the predictive value of neutrophil to lymphocyte ratio in breast cancers, but series are generally small. We aim to study the associations of Neutrophil-Lymphocyte Ratio (NLR) with outcomes in stage I–III breast cancer in patients who received neo-adjuvant chemotherapy (NAC) or upfront surgery. The endpoints are overall survival (OS) and breast cancer-specific survival (BCSS). In the NAC group, association with pathological complete response (PCR) rate was also studied. Methods: Data of patients with stage I–III breast cancer treated from 2011–2017 were extracted from a prospectively maintained registry and merged with full blood count (FBC) results from a clinical management software. FBC were performed within one-month pre-chemo for NAC patients and one-month pre-surgery for upfront surgery patients. PCR is defined as stage ypT0/isN0M0. OS is defined as death from any cancer from diagnosis date, censored at last follow-up. BCSS is defined as death from breast cancer from diagnosis date, censored at last follow up or death from any cause. The NLR values with the maximal Youden’s indexes calculated for each outcome were used as cut-off, logistic regression was used to determine PCR association and cox regression and log rank for OS and BCSS. Results: A total of 2,479 eligible patients were analysed. Overall, treatment compliance was high (87.6% of ER+ patients received endocrine treatment, and 94.1% of HER2+ NAC patients had targeted therapy). In the NAC group (n = 357), 23% achieved PCR. NLR did not show any statistically significant association with PCR. In unadjusted analysis, high NLR was associated with worse BCSS (log-rank p = 0.003 figure 1). In multivariable analysis (MVA), only triple negative and HER-enriched cancers were significantly associated with PCR. In NAC patients, NLR was associated with OS (cut-off 2.63; OR 1.6, p = 0.077) and BCSS (cut-off 3.58; OR 2.2, p = 0.003) in MVA. In patients treated with up-front surgery (n = 2122), unadjusted analysis showed high NLR was associated with worse BCSS (OR 1.55, p = 0.05; figure 2). In MVA, NLR (cut-off 2.13, OR 1.57, p = 0.005), triple negative histology, stage and age were significant predictors of OS. BCSS was not significantly associated with NLR (OR 1.38 95%CI 0.90-2.12 p = 0.145). Conclusions: Using a large cohort of patients, a high NLR was found to be associated with worse outcomes in NAC and upfront surgery breast cancer patients.