Optimizing cut-offs for high neutrophil-to-lymphocyte ratio as poor survival prognostic factor in colorectal cancer.

e15111 Background: The incidence of colorectal cancer (CRC) in Peru has increased in the last decades. Needing to use all the possible tools for an accurate diagnosis and early treatment. Neutrophil-to-lymphocyte ratio (NLR) has been associated as poor prognostic factor in OS and PFS in CRC. There is no data that support this statement in Latin America. It is of a special utility in our country the detection of a simple and reproducible prognostic biomarker that guides the use of more advanced tests. Our objective was to explore the factors associated with OS in the local-locally advanced and metastatic settings. Methods: We retrospectively reviewed the electronic medical records of 609 patients with CRC from one specialized Peruvian cancer center between 2006 and 2016 Descriptive results for numeric variables were presented as means with standard deviation (SD) or medians with interquartile range (IQR), depending on their distributions; otherwise, we expressed the qualitative variables as numbers with percentages. We divided our population into two groups: Local-locally advanced (L-LA) (516 pts) and debut metastatic- recurrence (M-R) (108 pts). We performed a ROC curve analysis to determine an appropriate cut-off value for NLR in both groups (L-LA:NLR ≥3, M-R:NLR ≥5). A univariate survival analysis was performed with Kaplan Meier method, comparing the curves with Log Rank test. A multivariate analysis was performed using the Cox regression model with the statistically significant variables found in the univariate analysis. Results: Pts with high NLR had significantly shorter OS in L-LA [HR, 12.1; 95% CI,5.019-29.211; p < 0.001] M-R [HR, 5.382; 95% CI,2.835-10.217; p < 0.001] than pts with low NLR. In the multivariate model, NLR retained a significant association with OS in both groups. Cox regression demonstrated that in L-LA setting sex, histologic grade and lymph node involvement; and in M-R setting sidedness, histologic grade, LVI and metastasectomy performed were independently risk factors for a shorter OS. Conclusions: High NLR is associated with poor prognosis (with our cut-offs L-LA:NLR ≥3, M-R:NLR ≥5). There are other variables to be considered that affect the OS, as: sex, histologic grade and lymph node involvement, sidedness, histologic grade, LVI and metastasectomy performed.