Search

The improvement of treatment for patients with ‘driver‐gene‐negative’ lung adenocarcinoma (LUAD) remains a critical problem to be solved. We aimed to explore the role of methylation of N6 adenosine (m6A)‐related long noncoding RNA (lncRNA) in stratifying ‘driver‐gene‐negative’ LUAD risk. Patients negative for mutations in EGFR, KRAS, BRAF, HER2, MET, ALK, RET, and ROS1 were identified as ‘driver‐gene‐negative’ cases. RNA sequencing was performed in 46 paired tumors and adjacent normal tissues from patients with ‘driver‐gene‐negative’ LUAD. Twenty‐three m6A regulators and relevant lncRNAs were identified using Pearson's correlation analysis. K‐means cluster analysis was used to stratify patients, and a prognostic nomogram was developed. The CIBERSORT and pRRophetic algorithms were employed to quantify the immune microenvironment and chemosensitivity. We identified two clusters highly consistent with the prognosis based on their unique expression profiles for 46 m6AlncRNAs. A risk model constructed from nine m6A lncRNAs could stratify patients into high‐ and low‐risk groups with promising predictive power (C‐index = 0.824), and the risk score was an independent prognostic factor. The clusters and risk models were closely related to immune characteristics and chemosensitivity. Additional pan‐cancer analysis using the nine m6AlncRNAs showed that the expression of DIO3 opposite strand upstream RNA (DIO3OS) is closely related to the immune/stromal score and tumor stemness in a variety of cancers. Our results show that m6AlncRNAs are a reliable prognostic tool and can aid treatment decision‐making in ‘driver‐gene‐negative’ LUAD. DIO3OS is associated with the development of various cancers and has potential clinical applications.

Abstract Objective Examining the role of immune‐related genes (IRGs) in “driver gene negative” lung adenocarcinoma (LUAD) may provide new ideas for the treatment and study for this LUAD subgroup. We aimed to find the hub immune‐related gene pairs can stratify the risk of “driver‐gene‐negative” LUAD. Materials and Methods IRGs were identified according to ImmPort database based on RNA sequencing results of tumors and normal tissues from 46 patients with “driver gene negative” LUAD at The First Affiliated Hospital of Sun Yat‐sen University and cyclically singly paired as immune‐related gene pairs (IRGPs). Multivariate Cox analysis was used to construct an immune risk model and a prognostic nomogram combining was also been developed. Immune microenvironment landscape described by CIBERSORT and drug sensitivity calculated by pRRophetic algorithm were used to explore possible treatment improvements. Results A novel immune risk model with 5‐IRGPs (CD1A|CXCL135, CD1A|S100A7L2, IFNA7|CMTM2, IFNA7|CSF3, CAMP|TFR2) can accurately distinguish patients in the high‐ and low‐risk groups. Risk score act as an independent prognostic factor and is related to clinical stage. There are significant differences in tumor immune microenvironment and PD‐1/PD‐L1/CTLA‐4 expression between groups. The low‐risk patient may benefit more from the commonly used chemotherapy regimens such as gemcitabine and paclitaxel. Conclusion This study constructed 5‐IRGPs as a reliable prognostic tool and may represent genes pairs that are potential rationale for choice of treatment for “driver gene negative” LUAD.

Oesophageal squamous cell carcinoma (ESCC) remains a clinically challenging disease with high morbidity rates and poor prognosis. ESCC is also the most common pathological type of oesophageal cancer (EC) in China. Ras-related genes are one of the most frequently mutated gene families in cancer and regulate tumour development and progression. Given this, we investigated the Ras-related gene expression profiles and their values in ESCC prognosis, using data from the Genotype-Tissue Expression (GTEx) and The Cancer Genome Atlas (TCGA) databases. We found that we could identify three distinct oesophageal cancer clusters based on their unique expression profile for 11 differentially expressed Ras-related genes with each of these demonstrating some prognostic value when, evaluated using univariate Cox analysis. We then used multivariate Cox analysis to identify relevant independent prognostic indicators and used these to build a new prognostic prediction model for oesophageal cancer patients using these three Ras-related genes. These evaluations produced an area under the curve (AUC) of 0.932. We found that our Ras-related signatures could also act as independent factors in ESCC prognosis and that patients with low Ras scores showed a higher overall expression levels of various immune checkpoint genes, including TNFSF4, TNFRSF8, TNFRSF9, NRP1, CD28, CD70, CD200, CD276, METTL16, METTL14, ZC3H13, YTHDF3, VIRMA, FTO, and RBM15, as well as a higher CSMD3, FLG, DNAH5, MUC4, PLCO, EYS, and ZNF804B mutation rates, and better sensitivity to drugs such as erlotinib, paclitaxel, and gefitinib. In conclusion, we were able to use the unique expression profiles of several Ras-related genes to produce a novel disease signature which might facilitate improved prognosis in ESCC, providing new insight into both diagnosis and treatment in these cancers.