A 15-Gene Signature and Prognostic Nomogram for Predicting Overall Survival in Non-Distant Metastatic Oral Tongue Squamous Cell Carcinoma

Muyuan Liu; Litian Tong; Bin Liang; Xuhong Song; Lingzhu Xie; Hanwei Peng; Dongyang Huang

doi:10.3389/fonc.2021.587548

A 15-Gene Signature and Prognostic Nomogram for Predicting Overall Survival in Non-Distant Metastatic Oral Tongue Squamous Cell Carcinoma

Front Oncol. 2021 Mar 9:11:587548. doi: 10.3389/fonc.2021.587548. eCollection 2021.

Authors

Muyuan Liu¹, Litian Tong², Bin Liang³, Xuhong Song³, Lingzhu Xie³, Hanwei Peng¹, Dongyang Huang^{3

4}

Affiliations

¹ Department of Head and Neck, Cancer Hospital of Shantou University Medical College, Shantou, China.
² Department of Anesthesiology, Cancer Hospital of Shantou University Medical College, Shantou, China.
³ Department of Cell Biology and Genetics, Key Laboratory of Molecular Biology in High Cancer Incidence Coastal Chaoshan Area of Guangdong Higher Education Institutes, Shantou University Medical College, Shantou, China.
⁴ Research Center of Translational Medicine, Second Affiliated Hospital of Shantou University Medical College, Shantou, China.

Abstract

Background: Oral tongue squamous cell carcinoma (OTSCC) is a devastating tumor with poor prognosis. There is an urgent need for reliable biomarkers to help predict prognosis and guide treatment for OTSCC. In the current study, we aimed to develop a robust multi-gene signature and prognostic nomogram to predict the prognosis of patients with non-distant metastatic OTSCC.

Methods: OTSCC-related differentially-expressed genes were screened from The Cancer Genome Atlas (TCGA) database. Univariate Cox regression based on 1,000 bootstrap replicates, LASSO regression and stepwise multivariate Cox regression were utilized to develop a novel multi-mRNA signature for predicting overall survival in OTSCC. The concordance index, area under receiver operating characteristic (ROC AUC) and calibration curve were employed to assess the prediction capacity of the novel multi-gene model. In addition, a prognostic nomogram was constructed to facilitate the clinical use of the fitted model. The Kaplan-Meier with log-rank test was employed to assess differences in overall survival.

Results: We successfully established a novel 15-mRNA prognostic model for predicting overall survival of non-distant metastatic OTSCC, involving ADTRP, ITGA3, RFC4, CCDC96, CYP2J2, NELL2, SPHK1, SPAG16, HBEGF, S100A9, EGFL6, ADGRG6, PDE4D, ABCA4, and CTTN. The prediction ability of this 15-gene signature was independent of other clinicopathological factors, with an HR of 11.5 (95% CI: 4.70-28.3). Moreover, internal validation by bootstrap analysis yielded a C-index of 0.849, with a 3-year AUC of 0.907 and 5-year AUC of 0.944, which implied excellent prediction accuracy of the fitted model. In addition, external validation by using the GEO dataset (GSE41116) yielded a C-index of 0.804, with a 3-year AUC of 0.868 and 5-year AUC of 0.855, which also indicated good prediction ability of the 15-gene model. Finally, a prognostic nomogram integrating risk group, grade, T stage and N stage was established.

Conclusion: Our results demonstrate our 15-gene signature was independently associated with overall survival in non-distant metastatic OTSCC. Moreover, the prognostic nomogram integrating the 15-gene signature and clinicopathological factors has potential to be developed as a prognostic tool.

Keywords: gene signature; nomogram; oral tongue squamous cell carcinoma (OTSCC); overall survival (OS); prognosis.