Research Paper Volume 12, Issue 21 pp 21854—21873

Three-gene risk model in papillary renal cell carcinoma: a robust likelihood-based survival analysis

Yutao Wang1, , Kexin Yan2, , Jiaxing Lin1, , Jianfeng Wang1, , Zhenhua Zheng1, , Xinxin Li1, , Zhixiong Hua1, , Yuepeng Bu1, , Jianxiu Shi1, , Siqing Sun1, , Xuejie Li1, , Yang Liu1, , Jianbin Bi1,3, ,

  • 1 Department of Urology, The First Hospital of China Medical University, Shenyang 110001, Liaoning, China
  • 2 Department of Dermatology, The First Hospital of China Medical University, Shenyang 110001, Liaoning, China
  • 3 Joint Fund of Science and Technology Department of Liaoning Province and State Key Laboratory of Robotics, Shenyang 110001, Liaoning, China

Received: March 1, 2020       Accepted: August 14, 2020       Published: November 5, 2020
How to Cite

Copyright: © 2020 Wang et al. This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.


Background: Papillary renal cell carcinoma (PRCC) accounts for 15% of all renal cell carcinomas. The molecular mechanisms of renal papillary cell carcinoma remain unclear, and treatments for advanced disease are limited.

Result: We built the computing model as follows: Risk score = 1.806 * TPX2 - 0.355 * TXNRD2 - 0.805 * SLC6A20. The 3-year AUC of overall survival was 0.917 in the training set (147 PRCC samples) and 0.760 in the test set (142 PRCC samples). Based on the robust model, M2 macrophages showed positive correlation with risk score, while M1 macrophages were the opposite. PRCC patients with low risk score showed higher tumor mutation burden. TPX2 is a risk factor, and co-expression factors were enriched in cell proliferation and cancer-related pathways. Finally, the proliferation and invasion of PRCC cell line were decreased in the TPX2 reduced group, and the differential expression was identified. TPX2 is a potential risk biomarker which involved in cell proliferation in PRCC.

Conclusion: We conducted a study to develop a three gene model for predicting prognosis in patients with papillary renal cell carcinoma. Our findings may provide candidate biomarkers for prognosis that have important implications for understanding the therapeutic targets of papillary renal cell carcinoma.

Method: Gene expression matrix and clinical data were obtained from TCGA (The Cancer Genome Atlas), GSE26574, GSE2048, and GSE7023. Prognostic factors were identified using “survival” and “rbsurv” packages, and a risk score was constructed using Multivariate Cox regression analysis. The co-expression networks of the factors in model were constructed using the “WGCNA” package. The co-expression genes of factors were enriched and displayed the biological process. Based on this robust risk model, immune cells infiltration proportions and tumor mutation burdens were compared between risk groups. Subsequently, using the PRCC cell line, the role of TPX2 was determined by Cell proliferation assay, 5-Ethynyl-20-deoxyuridine assay and Transwell assay.


PRCC: papillary renal cell carcinoma; WGCNA: weighted gene co-expression network analysis; PPI: protein-protein interactions; TMB: tumor mutation burden.