Xiaodan Zhong1,2,3, Ying Tao4, Jian Chang2, Yutong Zhang2, Hao Zhang1,3, Linyu Wang1,3, Yuanning Liu1,3. 1. College of Computer Science and Technology, Jilin University, Changchun, China. 2. Department of Pediatric Oncology, The First Hospital of Jilin University, Changchun, China. 3. Key Laboratory of Symbolic Computation and Knowledge Engineering, Ministry of Education, Jilin University, Changchun, China. 4. Department of Anesthesiology, China-Japan Union Hospital of Jilin University, Changchun, China.
Abstract
BACKGROUND: The prognostic value of immune-related genes and lncRNAs in neuroblastoma has not been elucidated, especially in subgroups with different outcomes. This study aimed to explore immune-related prognostic signatures. MATERIALS AND METHODS: Immune-related prognostic genes and lncRNAs were identified by univariate Cox regression analysis in the training set. The top 20 C-index genes and 17 immune-related lncRNAs were included in prognostic model construction, and random forest and the Least Absolute Shrinkage and Selection Operator (LASSO) regression algorithms were employed to select features. The risk score model was constructed and assessed using the Kaplan-Meier plot and the receiver operating characteristic curve. Functional enrichment analysis of the immune-related lncRNAs was conducted using the STRING database. RESULTS: In GSE49710, five immune genes (CDK4, PIK3R1, THRA, MAP2K2, and ULBP2) were included in the risk score five genes (RS5_G) signature, and eleven immune-related lncRNAs (LINC00260, FAM13A1OS, AGPAT4-IT1, DUBR, MIAT, TSC22D1-AS1, DANCR, MIR137HG, ERC2-IT1, LINC01184, LINC00667) were brought into risk score LncRNAs (RS_Lnc) signature. Patients were divided into high/low-risk score groups by the median. Overall survival and event/progression-free survival time were shortened in patients with high scores, both in training and validation cohorts. The same results were found in subgroups. In grouping ability assessment, the area under the curves (AUCs) in distinguishing different groups ranged from 0.737 to 0.94, better in discriminating MYCN status and high risk in training cohort (higher than 0.9). Multivariate Cox analysis demonstrated that RS5_G and RS_Lnc were the independent risk factors for overall and event/progression-free survival (all p-values <0.001). Correlation analysis showed that RS5_G and RS_Lnc were negatively associated with aDC, CD8+ T cells, but positively correlated with Th2 cells. Functional enrichment analyzes demonstrated that immune-related lncRNAs are mainly enriched in cancer-related pathways and immune-related pathways. CONCLUSION: We identified the immune-related prognostic signature RS5_G and RS_Lnc. The predicting and grouping ability is close to being even better than those reported in other studies, especially in subgroups. This study provided prognostic signatures that may help clinicians to choose optimal treatment strategies and showed a new insight for NB treatment. These results need further biological experiments and clinical validation.
BACKGROUND: The prognostic value of immune-related genes and lncRNAs in neuroblastoma has not been elucidated, especially in subgroups with different outcomes. This study aimed to explore immune-related prognostic signatures. MATERIALS AND METHODS: Immune-related prognostic genes and lncRNAs were identified by univariate Cox regression analysis in the training set. The top 20 C-index genes and 17 immune-related lncRNAs were included in prognostic model construction, and random forest and the Least Absolute Shrinkage and Selection Operator (LASSO) regression algorithms were employed to select features. The risk score model was constructed and assessed using the Kaplan-Meier plot and the receiver operating characteristic curve. Functional enrichment analysis of the immune-related lncRNAs was conducted using the STRING database. RESULTS: In GSE49710, five immune genes (CDK4, PIK3R1, THRA, MAP2K2, and ULBP2) were included in the risk score five genes (RS5_G) signature, and eleven immune-related lncRNAs (LINC00260, FAM13A1OS, AGPAT4-IT1, DUBR, MIAT, TSC22D1-AS1, DANCR, MIR137HG, ERC2-IT1, LINC01184, LINC00667) were brought into risk score LncRNAs (RS_Lnc) signature. Patients were divided into high/low-risk score groups by the median. Overall survival and event/progression-free survival time were shortened in patients with high scores, both in training and validation cohorts. The same results were found in subgroups. In grouping ability assessment, the area under the curves (AUCs) in distinguishing different groups ranged from 0.737 to 0.94, better in discriminating MYCN status and high risk in training cohort (higher than 0.9). Multivariate Cox analysis demonstrated that RS5_G and RS_Lnc were the independent risk factors for overall and event/progression-free survival (all p-values <0.001). Correlation analysis showed that RS5_G and RS_Lnc were negatively associated with aDC, CD8+ T cells, but positively correlated with Th2 cells. Functional enrichment analyzes demonstrated that immune-related lncRNAs are mainly enriched in cancer-related pathways and immune-related pathways. CONCLUSION: We identified the immune-related prognostic signature RS5_G and RS_Lnc. The predicting and grouping ability is close to being even better than those reported in other studies, especially in subgroups. This study provided prognostic signatures that may help clinicians to choose optimal treatment strategies and showed a new insight for NB treatment. These results need further biological experiments and clinical validation.
Authors: Nicolas Jacquelot; Jonathan M Pitt; David P Enot; Maria Paula Roberti; Connie P M Duong; Sylvie Rusakiewicz; Alexander M Eggermont; Laurence Zitvogel Journal: Oncoimmunology Date: 2017-03-07 Impact factor: 8.110
Authors: Jun S Wei; Igor B Kuznetsov; Shile Zhang; Young K Song; Shahab Asgharzadeh; Sivasish Sindiri; Xinyu Wen; Rajesh Patidar; Sushma Najaraj; Ashley Walton; Jaime M Guidry Auvil; Daniela S Gerhard; Aysen Yuksel; Daniel Catchpoole; Stephen M Hewitt; Paul M Sondel; Robert Seeger; John M Maris; Javed Khan Journal: Clin Cancer Res Date: 2018-05-21 Impact factor: 12.531
Authors: Lori S Hart; JulieAnn Rader; Pichai Raman; Vandana Batra; Mike R Russell; Matthew Tsang; Maria Gagliardi; Lucy Chen; Daniel Martinez; Yimei Li; Andrew Wood; Sunkyu Kim; Sudha Parasuraman; Scott Delach; Kristina A Cole; Shiva Krupa; Markus Boehm; Malte Peters; Giordano Caponigro; John M Maris Journal: Clin Cancer Res Date: 2016-10-11 Impact factor: 12.531
Authors: Dorian Stolk; Hans J van der Vliet; Tanja D de Gruijl; Yvette van Kooyk; Mark A Exley Journal: Front Immunol Date: 2018-09-21 Impact factor: 7.561