A bioinformatics-based immune-related prognostic index for lung adenocarcinoma that predicts patient response to immunotherapy and common treatments

Bioinformatics analysis of an immunotherapy responsiveness-related gene signature in predicting lung adenocarcinoma prognosis

Background

Immune therapy has become first-line treatment option for patients with lung cancer, but some patients respond poorly to immune therapy, especially among patients with lung adenocarcinoma (LUAD). Novel tools are needed to screen potential responders to immune therapy in LUAD patients, to better predict the prognosis and guide clinical decision-making. Although many efforts have been made to predict the responsiveness of LUAD patients, the results were limited. During the era of immunotherapy, this study attempts to construct a novel prognostic model for LUAD by utilizing differentially expressed genes (DEGs) among patients with differential immune therapy responses.

Methods

Transcriptome data of 598 patients with LUAD were downloaded from The Cancer Genome Atlas (TCGA) database, which included 539 tumor samples and 59 normal control samples, with a mean follow-up time of 29.69 months (63.1% of patients remained alive by the end of follow-up). Other data sources including three datasets from the Gene Expression Omnibus (GEO) database were analyzed, and the DEGs between immunotherapy responders and nonresponders were identified and screened. Univariate Cox regression analysis was applied with the TCGA cohort as the training set and GSE72094 cohort as the validation set, and least absolute shrinkage and selection operator (LASSO) Cox regression were applied in the prognostic-related genes which fulfilled the filter criteria to establish a prognostic formula, which was then tested with time-dependent receiver operating characteristic (ROC) analysis. Enriched pathways of the prognostic-related genes were analyzed with Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses, and tumor immune microenvironment (TIME), tumor mutational burden, and drug sensitivity tests were completed with appropriate packages in R (The R Foundation of Statistical Computing). Finally, a nomogram incorporating the prognostic formula was established.

Results

A total of 1,636 DEGs were identified, 1,163 prognostic-related DEGs were extracted, and 34 DEGs were selected and incorporated into the immunotherapy responsiveness-related risk score (IRRS) formula. The IRRS formula had good performance in predicting the overall prognoses in patients with LUAD and had excellent performance in prognosis prediction in all LUAD subgroups. Moreover, the IRRS formula could predict anticancer drug sensitivity and immunotherapy responsiveness in patients with LUAD. Mechanistically, immune microenvironments varied profoundly between the two IRRS groups; the most significantly varied pathway between the high-IRRS and low-IRRS groups was ribonucleoprotein complex biogenesis, which correlated closely with the TP53 and TTN mutation burdens. In addition, we established a nomogram incorporating the IRRS, age, sex, clinical stage, T-stage, N-stage, and M-stage as predictors that could predict the prognoses of 1-year, 3-year, and 5-year survival in patients with LUAD, with an area under curve (AUC) of 0.718, 0.702, and 0.68, respectively.

Conclusions

The model we established in the present study could predict the prognosis of LUAD patients, help to identify patients with good responses to anticancer drugs and immunotherapy, and serve as a valuable tool to guide clinical decision-making.

An immune-related gene prognostic index for predicting prognosis in patients with colorectal cancer

Background

Colorectal cancer (CRC) is one of the most common solid malignant burdens worldwide. Cancer immunology and immunotherapy have become fundamental areas in CRC research and treatment. Currently, the method of generating Immune-Related Gene Prognostic Indices (IRGPIs) has been found to predict patient prognosis as an immune-related prognostic biomarker in a variety of tumors. However, their role in patients with CRC remains mostly unknown. Therefore, we aimed to establish an IRGPI for prognosis evaluation in CRC.

Methods

RNA-sequencing data and clinical information of CRC patients were retrieved from The Cancer Genome Atlas (TCGA) and The Gene Expression Omnibus (GEO) databases as training and validation sets, respectively. Immune-related gene data was obtained from the ImmPort and InnateDB databases. The weighted gene co-expression network analysis (WGCNA) was used to identify hub immune-related genes. An IRGPI was then constructed using Cox regression methods. Based on the median risk score of IRGPI, patients could be divided into high-risk and low-risk groups. To further investigate the immunologic differences, Gene set variation analysis (GSVA) studies were conducted. In addition, immune cell infiltration and related functional analysis were used to identify the differential immune cell subsets and related functional pathways.

Results

We identified 49 immune-related genes associated with the prognosis of CRC, 17 of which were selected for an IRGPI. The IRGPI model significantly differentiates the survival rates of CRC patients in the different groups. The IRGPI as an independent prognostic factor significantly correlates with clinico-pathological factors such as age and tumor stage. Furthermore, we developed a nomogram to improve the clinical utility of the IRGPI score. Immuno-correlation analysis in different IRGPI groups revealed distinct immune cell infiltration (CD4⁺ T cells resting memory) and associated pathways (macrophages, Type I IFNs responses, iDCs.), providing new insights into the tumor microenvironment. At last, drug sensitivity analysis revealed that the high-risk IRGPI group was sensitive to 11 and resistant to 15 drugs.

Conclusion

Our study established a promising immune-related risk model for predicting survival in CRC patients. This could help to better understand the correlation between immunity and the prognosis of CRC providing a new perspective for personalized treatment of CRC.

Correlation Study on the Expression of INSR, IRS-1, and PD-L1 in Nonsmall Cell Lung Cancer

Objective

To study the expression and correlation of insulin receptor (INSR), insulin receptor substrate-1 (IRS-1), and programmed cell death ligand-1 (PD-L1) in nonsmall cell lung cancer (NSCLC).

Methods

45 lung cancer tissues and 30 adjacent normal tissues of NSCLC patients diagnosed in the Second Affiliated Hospital of Shandong First Medical University from June 2019 to August 2020 were selected. The expressions of INSR, IRS-1, and PD-L1 proteins in tumor tissues and adjacent tissues of NSCLC were detected by immunohistochemical staining.

Results

The expression of INSR and IRS-1 in NSCLC was significantly higher than that in adjacent normal lung tissue (P < 0.05). INSR expression had statistical significance with the degree of pathological differentiation of nonsmall cell carcinoma (P = 0.031), but had no significant association with age, gender, pathological type, TNM stage, and lymph node metastasis status (P > 0.05). There was no significant correlation between IRS-1 positive expression and NSCLC patients' age, gender, pathological typing, degree of differentiation, TNM stage, and lymph node metastasis (P > 0.05). PD-L1 positive expression was correlated with lymph node metastasis of NSCLC (P = 0.028), while there was no significant correlation with gender, age, pathological type, TNM stage, and pathological differentiation degree of NSCLC patients (P > 0.05). Spearman correlation analysis showed that PD-L1 protein expression had a significant positive correlation with IRS-1 protein expression (r = 0.373), but was not correlated with the expression of INSR protein.

Conclusion

IRS-1 may be involved in the regulation of PD-L1 expression and mediate the occurrence of tumor immune escape, which is expected to become a new target for NSCLC immunotherapy and provide new clinical evidence for immunosuppressive therapy.