Construction and Validation of a 9-Gene Signature for Predicting Prognosis in Stage III Clear Cell Renal Cell Carcinoma

Abstract

Purpose: Aim of this study was to develop a multi-gene signature to help better predict prognosis for stage III renal cell carcinoma (RCC) patients.Methods: Fourteen pairs of stage III tumor and normal tissues mRNA expression data from GSE53757 and 16 pairs mRNA expression data from TCGA clear cell RCC database were used to analyze differentially expressed genes between tumor and normal tissues. Common different expressed genes in both datasets were used for further modeling. Lasso Cox regression analysis was performed to select and build prognostic multi-gene signature in TCGA stage III kidney cancer patients (N = 122). Then, the multi-gene signature was validated in stage III renal cancer cases in Fudan University Shanghai Cancer Center (N = 77). C-index and time-dependent ROC were used to test the efficiency of this signature in predicting overall survival.Results: In total, 1,370 common different expressed genes were found between tumor and normal tissues in both datasets. After Lasso Cox modeling, nine mRNAs were finally identified to build a classifier. Using this classifier, we could classify stage III clear cell RCC patients into high-risk group and low-risk group. Prognosis was significantly different between these groups in discovery TCGA cohort, validation FUSCC cohort and entire set (All P < 0.001). Multivariate cox regression in entire set (N = 199) revealed that risk group classified by 9-gene signature, age of diagnosis, pN stage and ISUP grade were independent prognostic factor of overall survival in stage III kidney cancer patients.Conclusion: We developed a robust multi-gene classifier that can effectively classify stage III RCC patients into groups with low and high risk of poor prognosis. This signature may help select high-risk patients who require more aggressive adjuvant target therapy or immune therapy.

Identification of key genes associated with Idiopathic Pulmonary Fibrosis by combining multiple analytical methods

“…As a result, 13 genes including SERPIND1, CDH2, CP, WDR63, DNAH5, FGG, THBS1, VCAM1, COL1A2, POSTN, CXCL12, MMP13, SERPINE1 were selected through integrating the LASSO analysis and SVM-RFE analysis result. The LASSO (Least Absolute Shrinkage and Selection Operator) methods is an popular and important regularization in many regression analysis methods (e.g., COX regression, logistic regression) [17]. Compared with the linear model, LASSO model can reduce the variable numbers and effectively avoid over tting.…”

Influence of gene expression on survival of clear cell renal cell carcinoma

“…Candidate ccRCC-related genes were selected based on previous published literature., [9][10][11][12][13][14][15][16][17][18][19] including our review article 19 and pathways described in the Cancer Genome Anatomy project. A total of 388 candidate genes were evaluated in the discovery sets.…”

“…Therefore, the identification of reliable biomarkers for ccRCC progression is greatly needed. We and others reported candidate biomarkers, such as long non-coding RNAs, 9 gene expression signatures, [10][11][12][13][14][15] epigenetics 16 for ccRCC progression and/or survival. [17][18][19] However, there is currently no clinically accepted molecular biomarker for ccRCC progression.…”

The Genes—Candidates for Prognostic Markers of Metastasis by Expression Level in Clear Cell Renal Cell Cancer

“…Therefore, the need to predict their occurrence is very relevant. The presence of prognostic markers will create the ability to form groups for dynamic observation, apply therapeutic approaches to prevent metastasis, detect emerging metastases Diagnostics 2020, 10, 30 2 of 13 in time, and treat patients effectively [6]. At the same time, molecular markers of the tumor metastatic potential used in practice are still missing [7].…”