Liu Jia, Lv Dekang, Wang Xiaobin, Wang Ruicong, Li Xiaodong
Department of Gynecology, Liaoning Cancer Hospital & Institute, Cancer Hospital of China Medical University, Shenyang, China.
Cancer Center, Institute of Cancer Stem Cell, Dalian Medical University, Dalian, China.
Front Oncol. 2021 Mar 8;11:622805. doi: 10.3389/fonc.2021.622805. eCollection 2021.
Alternative splicing (AS) is significantly related to the development of tumor and the clinical outcome of patients. In this study, our aim was to systematically analyze the survival-related AS signal in ovarian serous cystadenocarcinoma (OV) and estimate its prognostic validity in 48,049 AS events out of 21,854 genes. We studied 1,429 AS events out of 1,125 genes, which were significantly related to the overall survival (OS) in patients with OV. We established alternative splicing features on the basis of seven AS events and constructed a new comprehensive prognostic model. Kaplan-Meier curve analysis showed that seven AS characteristics and comprehensive prognostic models could strongly stratify patients with ovarian cancer and make them distinctive prognosis. ROC analysis from 0.781 to 0.888 showed that these models were highly efficient in distinguishing patient survival. We also verified the prognostic characteristics of these models in a testing cohort. In addition, uni-variate and multivariate Cox analysis showed that these models were superior independent risk factors for OS in patients with OV. Interestingly, AS events and splicing factor (SFs) networks revealed an important link between these prognostic alternative splicing genes and splicing factors. We also found that the comprehensive prognosis model signature had higher prediction ability than the mRNA signature. In summary, our study provided a possible prognostic prediction model for patients with OV and revealed the splicing network between AS and SFs, which could be used as a potential predictor and therapeutic target for patients with OV.
可变剪接（AS）与肿瘤的发生发展及患者的临床预后显著相关。在本研究中，我们旨在系统分析卵巢浆液性囊腺癌（OV）中与生存相关的AS信号，并在21,854个基因中的48,049个AS事件中评估其预后有效性。我们研究了1,125个基因中的1,429个AS事件，这些事件与OV患者的总生存期（OS）显著相关。我们基于7个AS事件建立了可变剪接特征，并构建了一个新的综合预后模型。Kaplan-Meier曲线分析表明，7个AS特征和综合预后模型能够有力地对卵巢癌患者进行分层，并使其具有独特的预后。从0.781到0.888的ROC分析表明，这些模型在区分患者生存方面具有高效性。我们还在一个测试队列中验证了这些模型的预后特征。此外，单变量和多变量Cox分析表明，这些模型是OV患者OS的优越独立危险因素。有趣的是，AS事件和剪接因子（SFs）网络揭示了这些预后性可变剪接基因与剪接因子之间的重要联系。我们还发现，综合预后模型特征比mRNA特征具有更高的预测能力。总之，我们的研究为OV患者提供了一个可能的预后预测模型，并揭示了AS与SFs之间的剪接网络，其可作为OV患者的潜在预测指标和治疗靶点。
