Abstract 3222
Background
Prognostic factors in resected colon cancer (CC) guide adjuvant therapy and intensity of follow up. Inflammation parameters as C-reactive protein and neutrophil counts in relation to lymphocyte number (neutrophil/lymphocyte ratio: NLR) have been correlated with poor prognosis in advanced tumors. To confirm the prognostic value of a prechemoterapy NLR in adjuvant setting, we performed a retrospective analysis of high risk stage II and stage III resected CC patients randomized into the TOSCA phase 3 trial comparing 3 or 6 months of adjuvant chemotherapy.
Methods
patients randomized in TOSCA trial with data available for NLR analysis before chemotherapy were included. A recursive partitioning analysis was performed to identify the best cut-off that better discriminates patients in terms of relapse-free survival (RFS). According to this cut-off, RFS and overall survival (OS) hazard ratios (HR) for NLR were calculated and adjusted for age, sex, treatment type (XELOX vs FOLFOX) and duration (3 vs 6 months), grade, stage, performance status (PS), site and CEA level.
Results
Out of 3759 subjects randomized in the TOSCA trial, 1590 were included in the efficacy analysis. Mean NLR was 2.1 (median 1.8; range 0.3-24.0). Among patients analysed, 17.4% relapsed and 12.2% died. The best NLR cut off detect in this analysis population is 2.33. However, only age, PS, stage III and CEA levels were associated with RFS and OS in multivariate analysis, but not NLR>2.33 (RFS: HR = 1.17, 95%CI 0.90-1.51; P = 0.24, OS: HR = 1.16 95%CI 0.84-1.61; P = 0.38). Site was also correlated with poor OS.
Conclusions
Prechemotherapy NLR is not significantly associated with poor prognosis in patients with CC undergoing adjuvant chemotherapy. Resection of primary tumor and the associated reduced inflammatory stimulus may explain the lack of correlation with prognosis of NLR. Baseline evaluation, before surgery, likely represents the better timing to collect this information in cancer patients.
Clinical trial identification
Legal entity responsible for the study
GISCAD.
Funding
Has not received any funding.
Editorial Acknowledgement
Disclosure
All authors have declared no conflicts of interest.