Oxaliplatin related lncRNAs prognostic models predict the prognosis of patients given oxaliplatin-based chemotherapy

• Published in: Cancer cell international Vol. 23; no. 1; p. 103
• Main Authors: Zhou, Qing-Nan, Lei, Rong-E, Liang, Yun-Xiao, Li, Si-Qi, Guo, Xian-Wen, Hu, Bang-Li
• Format: Journal Article
• Language: English
• Published: England BioMed Central 27-05-2023; BMC
• Subjects: Algorithms; Cancer; Cells; Chemotherapy; Colorectal cancer; Colorectal carcinoma; Datasets; Gastric cancer; Genes; Learning algorithms; Leukemia; lncRNAs; Machine learning; Machine learning algorithm; Medical prognosis; Non-coding RNA; Ovarian cancer; Oxaliplatin; Oxaliplatin sensitivity; Prediction models; Prognosis; Regression analysis; Statistical analysis; Support vector machines; Survival analysis; Tumor cell lines; China; lncRNAs; Oxaliplatin sensitivity; Machine learning algorithm; Cancer
• ISSN: 1475-2867; 1475-2867
• DOI: 10.1186/s12935-023-02945-3

Oxaliplatin-based chemotherapy is the first-line treatment for colorectal cancer (CRC). Long noncoding RNAs (lncRNAs) have been implicated in chemotherapy sensitivity. This study aimed to identify lncRNAs related to oxaliplatin sensitivity and predict the prognosis of CRC patients underwent oxaliplatin-based chemotherapy.
Data from the Genomics of Drug Sensitivity in Cancer (GDSC) was used to screen for lncRNAs related to oxaliplatin sensitivity. Four machine learning algorithms (LASSO, Decision tree, Random-forest, and support vector machine) were applied to identify the key lncRNAs. A predictive model for oxaliplatin sensitivity and a prognostic model based on key lncRNAs were established. The published datasets, and cell experiments were used to verify the predictive value.
A total of 805 tumor cell lines from GDSC were divided into oxaliplatin sensitive (top 1/3) and resistant (bottom 1/3) groups based on their IC50 values, and 113 lncRNAs, which were differentially expressed between the two groups, were selected and incorporated into four machine learning algorithms, and seven key lncRNAs were identified. The predictive model exhibited good predictions for oxaliplatin sensitivity. The prognostic model exhibited high performance in patients with CRC who underwent oxaliplatin-based chemotherapies. Four lncRNAs, including C20orf197, UCA1, MIR17HG, and MIR22HG, displayed consistent responses to oxaliplatin treatment in the validation analysis.
Certain lncRNAs were associated with oxaliplatin sensitivity and predicted the response to oxaliplatin treatment. The prognostic models established based on the key lncRNAs could predict the prognosis of patients given oxaliplatin-based chemotherapy.