dNLR (neutrophil count /(total white cell - neutrophil count)) has shown prognostic utility for overall survival (OS) in a number of cancers. The CCTG/AGITG CO.17 and CO.20 studies examined the EGFR inhibitor, cetuximab (C), versus best supportive care, and in combination with the VEGF/FGFR inhibitor brivanib (B) or placebo, respectively, in previously treated patients with mCRC. Here, we examine dNLR in CO.17 and CO.20 for its utility to predict progression free survival (PFS), response rate (RR) and OS.
Association between dNLR status and PFS, OS and RR was analysed using multivariate regression models adjusting for baseline (BL) disease and patient (pt) characteristics. Interaction between treatment and dNLR status was studied by including an additional interaction term in the models.
CO.17 recruited 572 pts; 570 pts had available BL dNLR data. CO.20 recruited 750 pts; 718 had BL data available. In both, pts with high BL dNLR (≥2) were significantly more likely to have poorer ECOG status, received more lines of therapy, abnormal alkaline phosphatase levels, ≥ grade 1 anaemia, ascites, presented with lung metastases, have ≥2 sites of metastases. Multivariate analyses results are summarised in Table 1. Both studies showed pts with high dNLR had significantly shorter OS. Significant association of dNLR status with PFS was only found in CO.20 but no significant association with RR was found in either study. Significant association was found with OS in all four treatment groups, but for PFS, in C + B arm of CO.20 only. In both studies, no significant interaction was found between dNLR status and treatment for either OS or PFS.
Multivariate analysis of dNLR in CO.17 and CO.20