Identification of a 3-gene signature for predicting the prognosis of stage II colon cancer based on microsatellite status

Xiangxiong Huang; Heyang Xu; Yujie Zeng; Qiusheng Lan; Lu Liu; Wei Lai; Zhonghua Chu

doi:10.21037/jgo-21-405

Identification of a 3-gene signature for predicting the prognosis of stage II colon cancer based on microsatellite status

J Gastrointest Oncol. 2021 Dec;12(6):2749-2762. doi: 10.21037/jgo-21-405.

Authors

Xiangxiong Huang^#¹, Heyang Xu^#¹, Yujie Zeng¹, Qiusheng Lan¹, Lu Liu¹, Wei Lai¹, Zhonghua Chu¹

Affiliation

¹ Guangdong Provincial Key Laboratory of Malignant Tumor Epigenetics and Gene Regulation, Department of Gastrointestinal Surgery, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China.

^# Contributed equally.

Abstract

Background: Though colon cancer (CC) is one of the most malignant tumors across the world, CC patients with microsatellite instability-high (MSI-H) in stage II seem to have a better prognosis. However, the molecular mechanisms underlying the phenomena haven't been elucidated yet.

Methods: This study enrolled 322 CCs with known microsatellite status from GSE143985, GSE39582 and GSE92921 in the Gene Expression Omnibus (GEO) database. Robust rank aggregation (RRA) analysis, univariate Cox regression analysis and multivariate Cox stepwise regression analysis were performed to identify genes and construct risk score signature. Kaplan-Meier and receiver operating characteristic (ROC) curves analyses were used to evaluate the prognostic value of the signature. The potential mechanisms underlying this signature were assessed in the Metascape database, gene set enrichment analysis (GSEA) and immune infiltration analysis.

Results: RRA analysis identified 40 differently expressed genes (DEGs). A 3-gene risk score signature (MKQ signature) associated with disease-free survival (DFS) was generated. DFS was significantly longer in CC patients with lower than higher scores (P=0.0046). The areas under curves (AUCs) of the time-dependent ROC curves of MKQ signature at 1-, 3- and 5-year DFS were 1, 0.963 and 0.961 respectively. Recurrence-free survival (RFS) was significantly longer in patients in GSE39582 with lower than higher risk scores (P=0.032). The AUCs for 1-, 3- and 5-year RFS in GSE39582 were 0.63, 0.618 and 0.583, respectively, validating the value of the MKQ signature. Functional annotation and GSEA revealed that the MKQ signature was associated with multiple immune-related pathways. Immune cell infiltration was found to differ in patients differing in the MKQ signature.

Conclusions: Gene expression and microsatellite status identified a 3-gene signature (MKQ signature) that could facilitate risk-stratified management in patients with stage II CC. Dysregulation of MSMB, KRT23, and QPRT can serve as prognostic markers in stage II CC.

Keywords: Colon cancer (CC); gene set enrichment analysis (GSEA); gene signature; immune infiltration; microsatellite status.