Heterogeneity in prognostic value of the neutrophil-to-lymphocyte ratio: a meta-analysis

High pre-treatment values of the neutrophil-to-lymphocyte ratio (NLR) are strongly associated with poorer survival outcomes in cancer patients. Here, we assess heterogeneity in the magnitude of this association and the prognostic potential of the NLR between patient subgroups. We conducted a random effects meta-analysis of 228 published studies (N=75,555 patients) relating NLR with overall survival across 18 cancer types. Cochrans Q test and Higgins I2 statistic were used to assess study heterogeneity. Pooled hazard ratios were compared between groups of studies classified by cancer type, geographic region, therapy type, and cut-off for high NLR to identify study-level characteristics associated with increased prognostic potential of the NLR. Pooled hazard ratios are highest in studies of melanoma and breast cancer and lowest in studies of brain cancer and lung cancer. Radiation as primary treatment also demonstrates a large pooled effect size as compared to other therapies. The NLR has greater prognostic value in certain cancer types and therapeutic regimens. Efforts are needed to comprehensively examine populations in which NLR has maximum prognostic power. Clinically meaningful thresholds for risk stratification should be identified within these patient subgroups to permit prospective validation of the prognostic potential of the NLR.