Neutrophil-lymphocyte ratio (NLR) as a prognostic factor in metastatic breast cancer

Abstract

Background: Neutrophil-lymphocyte ratio (NLR) might be a surrogate marker of the tumor microenvironment immune balance and has been proposed as a prognostic factor for different tumors. A metanalysis of NLR for breast cancer (BC) showed that higher values at diagnosis were associated with lower survival. However, these results were mainly derived from the analysis of early BC cases: only three of twelve articles included women with metastatic breast cancer (MBC), sample size was small and no differential statistical analysis was performed for MBC. The aim of this work was to determine the prognostic value of NLR for MBC. Methods: We retrospectively collected clinical and analytical data from a series of consecutive MBC patients treated in one center between 2009 and 2016. NLR (=neutrophil count/lymphocyte count) was obtained from differential white blood cell count at diagnosis of metastasis, before starting any treatment. Non-parametric tests (Mann Whitney U, Kruskal-Wallis) were used to evaluate differences of NLR between groups. Kaplan-Meier curves and Cox regression models (univariate and multivariate) were used for overall survival analysis. Results: 265 consecutive patients with MBC were included, 117 of them (44%) with metastatic disease at diagnosis. Median age: 59 (19-95); ECOG 0-1 (69%), 2-3 (12%); site: bone only (37%), visceral only (18%), bone+visceral (30%); tumor subtype: HR (hormone receptor)+/HER2- (59%), HR+/HER2+(17%), HR-/HER2+(7%), HR-/ HER2- (14%); disease free interval in recurrent MBC: 24 months (100, 67%). Outcomes: 135 deaths; median overall survival (OS): 35 months (IC95%: 27.4-42.6). Median NLR was 2.31 (range: 0.70-44.33), with significant higher values in women with ECOG 2-3 (p=0.008) or negative estrogen receptors (p=0.03). Univariate Cox model of OS showed a HR=1.07 (95%: 1.03-1.11; p<0.001) for NLR as a continuous variable; using the median value as cut-off, HR=1.47 (95%CI: 1.05- 2.07; p=0.024). A multivariate Cox model showed the independent value of NLR for OS (Table). Conclusions: A higher NLR at diagnosis is a negative prognostic factor for overall survival in metastatic breast cancer. These data, if prospectively validated, may support the addition of NLR to MBC prognostic models and may lead to differential therapeutic approaches in patients with higher NLR.