Abstract
BACKGROUND: Inflammation is a hallmark of cancer, and systemic markers of inflammation are increasingly recognised as negative prognostic factors for clinical outcome. Neutrophil-to-lymphocyte ratio (NLR) is readily available from routine blood testing of patients diagnosed with cancer. METHODS: Peer-reviewed publications from PubMed/MEDLINE, Web of Science and EMBASE were identified according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines. Hazard ratios (HR) for overall survival (OS) and surrogate endpoints (SE; comprising disease-, recurrence- and progression-free survival) were pooled using a random effects model. Additional analysis was carried out to further investigate NLR as an independent prognostic factor and account for heterogeneity. RESULTS: Seventy-one eligible papers comprising 32,788 patients were identified. High NLR was associated with poor clinical outcomes. Significant publication bias was observed, and larger studies also adjusted for more covariates. Correcting for publication bias in multivariate studies brought our best estimate for true effect size to HR = 1.57 (95% CI 1.39-1.78; p < 0.0001) for OS and to HR = 1.38 (95% CI 1.16-1.64; p = 0.0003) for SE. CONCLUSIONS: NLR is confirmed as an easily available prognostic biomarker in colorectal cancer, despite the limitations of some studies previously reporting this finding. As such, it should be routinely collected in prospective clinical trials. While more standardised and rigorous large-scale studies are needed before high NLR can be fully assessed as an independent predictor of CRC progression and outcome, the data suggest that it may be used to highlight individuals with tumour-promoting inflammatory context.