摘要

Background: To better identify patients with high mortality risk, we developed a systematic inflammation index (IPI) based on neutrophil to lymphocyte ratio (NLR) and albumin. Methods: The performance of pretreatment IPI was evaluated in patients with surgically resected non-metastatic colorectal cancer. IPI was predefined and compared with Glasgow Prognostic Score (GPS)/ modified GPS in terms of discrimination and calibration abilities. Results: In multivariate analysis, patients with an IPI of 1 or 2 had 1.68(95%CI:1.15-2.44) or 3.56(95% CI:2.12-5.98)-fold increased cancer specific mortality risk(CSMR) respectively in comparison to patients with an IPI of 0. The prognostic significance was independent of tumor locations and nodal status. Compared with the GPS/mGPS, IPI had the higher c statistics and lower Akaike Information Criterion. IPI showed good calibration in predicting 1-year, 3-year and 5-year CSMR. Conclusions: IPI is readily available, independently prognostic and may reflect the host inflammation, immune and nutritional status that could have impact on cancer progression.