Prognostic value of neutrophil-to-lymphocyte ratio (NLR) on overall survival (OS), progression free survival (PFS) and disease control rate (DCR) in patients with metastatic colorectal cancer (mCRC) from the RECOURSE study.

Abstract

744 Background: Elevated neutrophil-to-lymphocyte (NLR) ratios may represent markers of a suboptimal host immune response to cancer and have been shown to correlate with prognosis in multiple tumor types. Trifluridine/tipiracil (FTD/TPI also known as TAS-102) compared with placebo significantly improved OS (median: 7.1 vs.5.3 months, hazard ratio (HR) = 0.68, 95% confidence interval (CI) [0.58;0.81], p < 0.0001), PFS, and DCR in the phase 3 RECOURSE study, conducted in patients with refractory mCRC. A post-hoc analysis was conducted to assess correlation between clinical outcomes and baseline NLR (in blood) in RECOURSE. Methods: A retrospective review on 782/800 patients, with available NLR, was performed on OS, PFS, and DCR in two subgroups of patients with low ( < 3) or high (≥ 3) NLR at baseline. Based on literature, the cutoff value of 3 was chosen as the number of patients was similar in each NLR subgroup. Between-group comparison was performed using a stratified Cox’s proportional hazard model for OS and PFS, and Fisher’s Exact test for DCR. Results: The median value of NLR at baseline was similar in each group, FTD/TPI (N = 519): 4.4 ± 6.3, placebo (N = 263): 4.7 ± 6.5. Almost 60% of patients in each group had high NLR. In the low NLR vs. high NLR subgroup, there were differences at baseline for ethnic origin (Asian: 45% vs.29%), ECOG PS 0 (69% vs. 49%), and number of metastatic sites ≥ 3 (32% vs. 46%). The median OS [95 % CI] was statistically significantly higher in the low vs. high NLR subgroup: 8.4 [7.8;9.5] vs. 5.3 [4.7;5.7] months, HR = 0.49, 95% CI [0.41;0.59], p < 0.0001. Irrespective of NLR, all patients benefited from FTD/TPI vs. placebo. The results of PFS and DCR were statistically significantly better in the low NLR vs. high NLR subgroup. The multivariate Cox analysis for OS with the interaction test between treatment groups and NLR showed that NLR is not a predictive factor (p = 0.15). Conclusions: In this retrospectively analyzed mCRC population, NLR was shown to be an independent prognostic factor. Further research is warranted to assess if NLR can be a stratification factor in mCRC clinical trials. Clinical trial information: NCT01607957.