Combining prognostic nutritional index (PNI) and controlling nutritional status (CONUT) score as a valuable prognostic factor for overall survival in patients with stage I–III colorectal cancer

BACKGROUND AND AIMS: This study compared the prognostic significance of various nutritional and inflammatory indicators such as neutrophil-to-lymphocyte ratio, lymphocyte-to-monocyte ratio and platelet-to-lymphocyte ratio, prognostic nutritional index, and controlling nutritional status score. In addition, we aimed to establish a more accurate prognostic indicator. METHODS: We retrospectively evaluated 1112 patients with stage I–III colorectal cancer between January 2004 and April 2014. The controlling nutritional status scores were classified as low (0–1), intermediate (2–4), and high (5–12) scores. The cut-off values for prognostic nutritional index and inflammatory markers were calculated using the X-tile program. P-CONUT, a combination of prognostic nutritional index and the controlling nutritional status score, was suggested. The integrated areas under the curve were then compared. RESULTS: The multivariable analysis showed that prognostic nutritional index was an independent prognostic factor for overall survival, whereas the controlling nutritional status score, neutrophil-to-lymphocyte ratio, lymphocyte-to-monocyte ratio, and platelet-to-lymphocyte ratio were not. The patients were divided into three P-CONUT groups as follows: G1, controlling nutritional status (0–4) and high prognostic nutritional index; G2, controlling nutritional status (0–4) and low prognostic nutritional index; and G3, controlling nutritional status (5–12) and low prognostic nutritional index. There were significant survival differences between the P-CONUT groups (5-year overall survival of G1, G2, and G3 were 91.7%, 81.2%, and 64.1%, respectively; p < 0.0001). The integrated areas under the curve of P-CONUT (0.610, CI: 0.578–0.642) was superior to those of the controlling nutritional status score alone (bootstrap integrated areas under the curve mean difference=0.050; 95% CI=0.022–0.079) and prognostic nutritional index alone (bootstrap integrated areas under the curve mean difference=0.012; 95% CI=0.001–0.025). CONCLUSION: Prognostic effect of P-CONUT may be better than inflammatory markers such as neutrophil-to-lymphocyte ratio, lymphocyte-to-monocyte ratio and platelet-to-lymphocyte ratio. Thus, it could be used as a reliable nutritional risk stratification tool in patients with colorectal cancer.