Cuproptosis-Related LncRNA-Based Prediction of the Prognosis and Immunotherapy Response in Papillary Renal Cell Carcinoma

Abstract

Cuproptosis, a new cell death pattern, is promising as an intervention target to treat tumors. Abnormal long non-coding RNA (lncRNA) expression is closely associated with the occurrence and development of papillary renal cell carcinoma (pRCC). However, cuproptosis-related lncRNAs (CRLs) remain largely unknown as prognostic markers for pRCC. We aimed to forecast the prognosis of pRCC patients by constructing models according to CRLs and to examine the correlation between the signatures and the inflammatory microenvironment. From the Cancer Genome Atlas (TCGA), RNA sequencing, genomic mutations and clinical data of TCGA-KIRP (Kidney renal papillary cell carcinoma) were analyzed. Randomly selected pRCC patients were allotted to the training and testing sets. To determine the independent prognostic impact of the training characteristic, the least absolute shrinkage and selection operator (LASSO) algorithm was utilized, together with univariate and multivariate Cox regression models. Further validation was performed in the testing and whole cohorts. External datasets were utilized to verify the prognostic value of CRLs as well. The CRLs prognostic features in pRCC were established based on the five CRLs (AC244033.2, LINC00886, AP000866.1, MRPS9-AS1 and CKMT2-AS1). The utility of CRLs was evaluated and validated in training, testing and all sets on the basis of the Kaplan–Meier (KM) survival analysis. The risk score could be a robust prognostic factor to forecast clinical outcomes for pRCC patients by the LASSO algorithm and univariate and multivariate Cox regression. Analysis of Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) data demonstrated that differentially expressed genes (DEGs) are primarily important for immune responses and the PI3K-Akt pathway. Arachidonic acid metabolism was enriched in the high-risk set by Gene Set Enrichment Analysis (GSEA). In addition, Tumor Immune Dysfunction and Exclusion (TIDE) analysis suggested that there was a high risk of immune escape in the high-risk cohort. The immune functions of the low- and high-risk sets differed significantly based on immune microenvironment analysis. Finally, four drugs were screened with a higher sensitivity to the high-risk set. Taken together, a novel model according to five CRLs was set up to forecast the prognosis of pRCC patients, which provides a potential strategy to treat pRCC by a combination of cuproptosis and immunotherapy.

A Model to Predict Prognosis of Renal Cell Clear Cell Carcinoma Based on 3 Angiogenesis-related Long Non-coding RNAs

“…Compared to existing signatures including several lncRNAs (more than 6) [25,26], our 3-lncRNA model contains the fewer genes and should be easy to apply in the clinic. Different functional lncRNA models, specifically ferroptosis/cuproptosis-associated lncRNA signatures and tumor-infiltrating lymphocyte-related lncRNA models, have been reported to predict ccRCC prognosis [27][28][29]. However, these lncRNAs in models are closely related to immunotherapy and drug response but not to ccRCC prognosis.…”

Predicting colorectal cancer prognosis based on long noncoding RNAs of disulfidptosis genes

“…Relevant risk curves and heatmaps were then plotted using the “survival,” “glmnet,” “survminer,” and “timeROC” R packages[ 23 , 24 ]. The following equation was then used to obtain the risk score[ 25 ]: Risk score = Σi = lnCoef (i) × Expr (i). The risk expression (Expr(i)) and risk coefficient (Coef(i)) functions represent the corresponding risk coefficient and risk expression.…”

A disulfidptosis-related lncRNAs signature in hepatocellular carcinoma: prognostic prediction, tumor immune microenvironment and drug susceptibility

“…Multivariate cox regression analysis was applied to build a two-lncRNA predictive model. The calculated formula used for disulfidptosis-related prognostic risk score was based on the published documents 22 , 38 . The model of risk score was calculated as follows: risk score = ∑i = LnCoef(i) × EXP(i).…”