AUTHOR=Yan Xin , Wu Hua-Hui , Chen Zhao , Du Guo-Wei , Bai Xiao-Jie , Tuoheti Kurerban , Liu Tong-Zu 

TITLE=Construction and Validation of an Autophagy-Related Prognostic Signature and a Nomogram for Bladder Cancer

JOURNAL=Frontiers in Oncology

VOLUME=Volume 11 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2021.632387

DOI=10.3389/fonc.2021.632387

ISSN=2234-943X

ABSTRACT=Objective: Bladder cancer (BC) is one of the top ten cancers endangering human health. Accurate tools for BC patients risk stratification are still missing. We aimed to develop an autophagy-related signature which could predict the prognosis of BC. In order to provide clinical doctors with a visual tool which could precisely predict the survival probability of BC patients, we also attempted to establish a nomogram based on the risk signature.
Methods: We screened out autophagy-related genes (ARGs) combining weighted gene co-expression network analysis (WGCNA) and differentially expressed gene (DEG) in BC. Based on the screened ARGs, we performed survival analysis and Cox regression analysis to identify potential prognostic biomarkers. A risk signature based on the prognostic ARGs by multivariate Cox regression analysis was established, which was validated by using seven datasets. In order to provide clinical doctor with a useful tool for survival possibility prediction, a nomogram assessed by the ARG-based signature and clinicopathological features was constructed, which was verification by using four independent datasets.
Results: Three prognostic biomarkers including BOC (P = 0.008, HR = 1.104), FGF7(P = 0.030, HR = 1.066), and MAP1A (P = 0.001, HR = 1.173) were identified and validated. An autophagy-related risk signature was established and validated. This signature could act as an independent prognostic feature in patients with BC (P = 0.047, HR = 1.419). Then we constructed two nomograms with and without ARG-based signature, subsequent analysis indicated that the nomogram with ARG signature showed high accuracy for overall survival probability prediction of patients with BC (C-index = 0.732, AUC = 0.816). These results proved that the ARG signature improved clinical net benefit of the standard model based on clinicopathological features (age, pathologic stage).
Conclusions: Three ARGs were identified as prognosis biomarkers in BC. An ARG-based signature was established for the first time, showing strong potential for prognosis prediction in BC. This signature was proved to improve the clinical net benefit of the standard model. A nomogram was established by using this signature, which could lead to more effective prognosis prediction for BC patients.