Comparison of the prognostic value of prognostic nutritional index in colorectal cancer deaths:a retrospective cohort study

Evidences showed that Prognostic Nutritional Index (PNI) is one important predictive factor for some cancers. However, the association between PNI and colorectal cancer is still unclear. In this study, we aim to explore the association between PNI and death in patients with locally advanced colorectal cancer. We enrolled 400 patients with colorectal cancer. The pathological results was used to confirm it. The patients received radical resection. The PNI at baseline was set as the exposure variable, death was the outcome variable, and other potential confounders were also analyzed. There was no association between PNI and the deaths of local advanced colorectal cancer and radical resection of colorectal cancer (HR=0.88; 95% CI, 0.78-1.00). Applying the cubic spline function, we found that it is not linear between PNI and death. Two-piecewise cox model showed that there was 68% decrease of risk of deaths when PNI change per 1 (HR=0.32; 95% CI, 0.13-0.79) on the left side of the inflection point (41.35). On the right of the inflection point, there was no association between PNI and death. PNI and deaths are linear correlation. Within a certain range (PNI=26-41), PNI is an independent protective factor for death risk. Practical Application: We found that there is not linear relationship between prognostic nutritional index (PNI) and deaths. In a specific range, PNI can be used to predict the death risk of colorectal cancer.