553P - The derived neutrophil lymphocyte ratio (dNLR) as a biomarker in advanced colorectal cancer (aCRC)

Date 29 September 2014
Event ESMO 2014
Session Poster Display session
Topics Colon Cancer
Rectal Cancer
Translational Research
Presenter Jenny Seligmann
Citation Annals of Oncology (2014) 25 (suppl_4): iv167-iv209. 10.1093/annonc/mdu333
Authors J.F. Seligmann1, P. Hall1, H. Wilson2, S.D. Richman2, J. Barrett3, M. Seymour1, P. Quirke2
  • 1Department Of Oncology, University of Leeds, LS9 7TF - Leeds/GB
  • 2Pathology And Tumour Biology, Leeds Institute of Molecular Medicine, Leeds/GB
  • 3Department Of Biostatistics, Leeds Institute of Molecular Medicine, Leeds/GB



Links between inflammation, the immune response and cancer outcomes are compelling. The dNLR, calculated from a standard blood count, is a simple measure of inflammatory response, and in retrospective series high dNLR has been reported to be associated with poorer outcomes in aCRC. We examined dNLR as a prognostic marker, and assessed its predictive utility for two treatments in large phase III trials in aCRC.


2484 patients (pts) were studied, from the FOCUS (Lancet 370:143, 2007) and PICCOLO trials (Lancet Oncol 14:749, 2013). dNLR was calculated from the pre-randomisation blood count and prospectively defined as “high” (≥2) or “low” (<2); end-points were PFS and OS. dNLR was assessed first as a prognostic marker, then as a predictive marker for more versus less intensive therapy in two settings: (1) doublet chemotherapy (irinotecan/5FU or oxaliplatin/5FU) versus single agent 5FU as 1st-line treatment; (2) panitumumab plus irinotecan (IrPan) versus irinotecan (Ir) alone as 2nd line treatment of KRAS-wt pts. Finally, we asked if the change in dNLR after 6 weeks of chemotherapy is predictive of RECIST response at 12 weeks.


52.8% pts had high dNLR; 47.2% had low dNLR. High dNLR was independently associated with reduced OS in both trials (FOCUS: n = 1810 HR 1.51, p < 0.001; PICCOLO: n = 674, HR 1.59, p < 0.001). In patients with high dNLR, 1st-line doublets gave improved OS (HR = 0.82, p = 0.008) while in those with low dNLR they did not (HR = 0.96, p = 0.46), however interaction testing was non-significant so the predictive utility of dNLR was not confirmed. dNLR did not predict benefit from IrPan compared with Ir,. A dNLR rise of >25% after 6 weeks of chemotherapy was associated with a higher risk of disease progression at 12 weeks (OR = 1.6, p = 0.02) and inferior PFS (HR = 1.34, p < 0.001) and OS (HR = 1.38, p < 0.001) compared to stable dNLR, independent of baseline dNLR.


The dNLR provides prognostic information and may provide an early warning when chemotherapy is failing. It might also help identify patients who could be treated safely with less intensive 1st line chemotherapy, but confirmatory data are required. Given that dNLR is readily and freely calculated from routine clinical data, its integration into routine patient evaluation should be considered following further validation.


All authors have declared no conflicts of interest.