Identification of a Signature Gene Set for Oxaliplatin Sensitivity Prediction in Colorectal Cancer

Xiaopeng ZhanXinyue LiYimin ChuYing XuFengli ZhouJi LiDaming YangChangping HuHaixia Peng^*Zhaoxia Wu^*

• Digestive Endoscopy Center, tongren hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China

Colorectal cancer (CRC) is among the most prevalent cancers worldwide, with oxaliplatin-based chemotherapy being a standard treatment option. Nevertheless, the emergence of resistance to oxaliplatin represents a pivotal challenge, as it undermines the efficacy of the treatment. In this study, we utilised large-scale datasets, encompassing both Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA), in conjunction with machine learning algorithms, to identify 14 genes that exhibited a potential association with oxaliplatin sensitivity. The predictive value of these genes was assessed using progression-free survival (PFS) analysis. Consequently, four genes (AXDND1, BAMBI, MAPK8IP2 and BMP7) were identified as being associated with PFS following oxaliplatin-based therapy. Multivariate logistic regression analysis in external validation datasets from GEO and GDSC (Genomics of Drug Sensitivity in Cancer) also confirmed the robustness of different combinations of these genes in oxaliplatin sensitivity prediction. The expression levels of four genes in external validation datasets and oxaliplatin-resistant cells were found to be significantly different. Silencing the expression of the four genes in CRC cells resulted in a significant enhancement of the toxicity of oxaliplatin, whereas the knockdown of BAMBI and BMP7 led to a substantial reduction in sensitivity to oxaliplatin. In conclusion, our findings have identified a gene set comprising four key genes that are associated with oxaliplatin sensitivity in CRC. This finding provides a promising basis for the development of prognostic biomarkers that can be used to guide personalised chemotherapy strategies in CRC.