Creation and Validation of a Survival Nomogram Based on Immune-Nutritional Indexes for Colorectal Cancer Patients.

Abstract

Nutritional and inflammatory status was associated with prognosis in various types of malignant cancer, including colorectal cancer (CRC). This clinical research was performed to estimate the prognostic role of immune-nutritional indexes CRC in patients and to set up a survival nomogram based on the significant immune-nutritional indexes. 1024 CRC patients underwent surgical resection from Wuhan Union Hospital were enrolled and divided into the test cohort (n = 717) and validation cohort (n = 307). A total of 19 immune-nutritional indexes were included into our analysis. The Cox regression analysis was utilized to identify the informative immune-nutritional indexes which were closely associated with overall survival (OS) and disease-free survival (DFS). Survival nomograms were created in the test set and further verified in the validation set. Td-ROC was curved to estimate the predictive performance of survival nomograms for CRC patients. Body mass index (BMI), chemotherapy, TNM stage, T stage, lactate dehydrogenase (LDH)/prealbumin (PA), monocytes (MON)/albumin (ALB), and prognostic nutritional index (PNI) were seven potent prognostic biomarkers of CRC patients. We created an OS-nomogram based on the seven risk indexes, and the predictive accuracy expressed with area under curve (AUC) was 0.826 for 1-year, 0.809 for 3-year, and 0.80 for 5-year OS rates in the test set and 0.795 for 1-year, 0.749 for 3-year, and 0.647 for 5-year OS rates in the validation set. TNM stage, T stage, LDH/ALB, and MON/ALB were risk factors for unfavorable DFS in CRC patients. We further built a DFS-nomogram based on the four risk factors, and the predictive performance presented with AUC was 0.806 for 1-year, 0.763 for 3-year, and 0.82 for 5-year DFS rates in the test set, and 0.704 for 1-year, 0.692 for 3-year, and 0.692 for 5-year DFS rates in the validation set. Our survival nomogram based on immune-nutritional indexes is a useful and potential prognostic tool in CRC patients.