CRC-113 gene expression signature for predicting prognosis in patients with colorectal cancer

// Minh Nam Nguyen 1, * , Tae Gyu Choi 1, * , Dinh Truong Nguyen 2 , Jin-Hwan Kim 1 , Yong Hwa Jo 1 , Muhammad Shahid 1 , Salima Akter 1 , Saurav Nath Aryal 1 , Ji Youn Yoo 1 , Yong-Joo Ahn 1 , Kyoung Min Cho 1 , Ju-Seog Lee 3 , Wonchae Choe 1 , Insug Kang 1 , Joohun Ha 1 , Sung Soo Kim 1 1 Department of Biochemistry and Molecular Biology, Medical Research Center for Bioreaction to Reactive Oxygen Species and Biomedical Science Institute, School of Medicine, Kyung Hee University, Seoul, Republic of Korea 2 School of Biotechnology, Tan Tao University, Long An, Vietnam 3 Department of Systems Biology, Division of Cancer Medicine, The University of Texas MD Anderson Cancer Center, Houston, Texas, USA * These authors have contributed equally to this work Correspondence to: Sung Soo Kim, e-mail: sgskim@khu.ac.kr Keywords: colorectal cancer, microarray analysis, gene expression profile, risk prediction Received: June 26, 2015 Accepted: August 28, 2015 Published: September 11, 2015 ABSTRACT Colorectal cancer (CRC) is the third leading cause of global cancer mortality. Recent studies have proposed several gene signatures to predict CRC prognosis, but none of those have proven reliable for predicting prognosis in clinical practice yet due to poor reproducibility and molecular heterogeneity. Here, we have established a prognostic signature of 113 probe sets (CRC-113) that include potential biomarkers and reflect the biological and clinical characteristics. Robustness and accuracy were significantly validated in external data sets from 19 centers in five countries. In multivariate analysis, CRC-113 gene signature showed a stronger prognostic value for survival and disease recurrence in CRC patients than current clinicopathological risk factors and molecular alterations. We also demonstrated that the CRC-113 gene signature reflected both genetic and epigenetic molecular heterogeneity in CRC patients. Furthermore, incorporation of the CRC-113 gene signature into a clinical context and molecular markers further refined the selection of the CRC patients who might benefit from postoperative chemotherapy. Conclusively, CRC-113 gene signature provides new possibilities for improving prognostic models and personalized therapeutic strategies.