Related Experiment Video
Updated: Aug 14, 2025
Predictive Immune Modeling of Solid Tumors
Published on: February 25, 2020
Development and validation of a combined metabolism and immune prognostic model in lung adenocarcinoma
Yu Shi1, Shihui Dai1, Yu Lei1
1Department of Oncology, the First Affiliated Hospital of Anhui Medical University, Hefei, China.
Background:
Tumor metabolism and immune response can affect the biological behavior of tumor cells. There is an obvious relationship between glycolysis and immune response. However, the association between metabolism and immune response and prognosis in lung adenocarcinoma (LUAD) has not yet been examined in a comprehensive and detailed manner. The establishment of reliable models for predicting the prognosis of LUAD based on glycolysis ability and immune status is still highly anticipated.
Methods:
The expression of genes were obtained from online databases, and the differentially expressed genes in LUAD tissues and adjacent tissues were identified. We used LUAD samples in The Cancer Genome Atlas (TCGA) database as training set and the Gene Expression Omnibus (GEO) databases as validation sets. The best predictive model was constructed using least absolute selection and shrinkage operator (LASSO) regression and Cox regression. The receiver operator characteristic (ROC) curve is used to verify the accuracy of the model. The expression status of the Glycolysis-related genes (GRGs) and the status of the immune cells in LUCD patients were further analyzed. The protein levels of the 3 identified genes were then tested in LUAD patients.
Results:
We identified 3 GRGs and immune-related genes (i.e., fibroblast growth factor 2, hyaluronan-mediated motor receptor, and nuclear receptor 0B2) and constructed a stable comprehensive index of glycolysis and immunity (CIGI) prediction model. The validation results for this CIGI model were quite stable across different datasets and patient subgroups and the CIGI score can be included as an independent prognostic factor for LUAD patients. The area under the curve (AUC) values of 1-, 3- and 5-year of the finally established nomogram model are 0.767, 0.735 and 0.769. Further analysis showed that LUAD patients in the low-risk group had lower levels of glycolytic gene expression than those in the high-risk group and exhibited an immunosuppressed state. Finally, hyaluronan-mediated motor receptor may play a role in inhibiting cancer, while fibroblast growth factor 2 and nuclear receptor 0B2 may play roles in promoting cancer.
Conclusions:
In this study, we established a new prognostic prediction model for LUAD patients that combines glycolysis ability and immune status.
