研究动态
Articles below are published ahead of final publication in an issue. Please cite articles in the following format: authors, (year), title, journal, DOI.

一种新型人工智能网络,可根据基因突变特征评估胃肠道癌症免疫治疗的预后。

A novel artificial intelligence network to assess the prognosis of gastrointestinal cancer to immunotherapy based on genetic mutation features.

发表日期:2024
作者: Bicheng Ye, Zhongyan Li, Qiqi Wang
来源: Frontiers in Immunology

摘要:

免疫检查点抑制剂 (ICIs) 彻底改变了胃肠癌治疗,但缺乏可靠的生物标志物阻碍了精确的患者反应预测。我们采用新型人工智能网络开发并验证了基因组突变特征 (GMS),以预测接受治疗的胃肠癌患者的预后ICI 治疗。随后,我们使用多组学数据探索了不同亚型的潜在免疫景观。最后,通过对癌症药物敏感性基因组学(GDSC)数据库的药物致敏数据进行分析,确定了UMI-77。使用细胞计数试剂盒8(CCK8)测定和平板克隆形成测定评估UMI-77对AGS和MKN45细胞系的敏感性。利用人工智能网络,我们开发了独立预测胃肠道预后的GMS癌症患者。 GMS 在三个公共队列中表现出一致的性能,并在接受者操作特征 (ROC) 曲线分析中表现出 6、12 和 24 个月总生存 (OS) 的高灵敏度和特异性。它优于传统的临床和分子特征。与高风险样本相比,低风险样本显示出更多的溶细胞免疫细胞,并且免疫原性潜力增强。此外,我们还鉴定出了小分子化合物 UMI-77。 UMI-77 的半数抑制浓度 (IC50) 与 GMS 呈负相关。值得注意的是,被归类为高风险的 AGS 细胞系对 UMI-77 显示出更高的敏感性,而被归类为低风险的 MKN45 细胞系则显示出较低的敏感性。这里开发的 GMS 可以可靠地预测胃肠道癌症患者的生存获益关于 ICI 治疗。版权所有 © 2024 Ye、Li 和 Wang。
Immune checkpoint inhibitors (ICIs) have revolutionized gastrointestinal cancer treatment, yet the absence of reliable biomarkers hampers precise patient response prediction.We developed and validated a genomic mutation signature (GMS) employing a novel artificial intelligence network to forecast the prognosis of gastrointestinal cancer patients undergoing ICIs therapy. Subsequently, we explored the underlying immune landscapes across different subtypes using multiomics data. Finally, UMI-77 was pinpointed through the analysis of drug sensitization data from the Genomics of Drug Sensitivity in Cancer (GDSC) database. The sensitivity of UMI-77 to the AGS and MKN45 cell lines was evaluated using the cell counting kit-8 (CCK8) assay and the plate clone formation assay.Using the artificial intelligence network, we developed the GMS that independently predicts the prognosis of gastrointestinal cancer patients. The GMS demonstrated consistent performance across three public cohorts and exhibited high sensitivity and specificity for 6, 12, and 24-month overall survival (OS) in receiver operating characteristic (ROC) curve analysis. It outperformed conventional clinical and molecular features. Low-risk samples showed a higher presence of cytolytic immune cells and enhanced immunogenic potential compared to high-risk samples. Additionally, we identified the small molecule compound UMI-77. The half-maximal inhibitory concentration (IC50) of UMI-77 was inversely related to the GMS. Notably, the AGS cell line, classified as high-risk, displayed greater sensitivity to UMI-77, whereas the MKN45 cell line, classified as low-risk, showed less sensitivity.The GMS developed here can reliably predict survival benefit for gastrointestinal cancer patients on ICIs therapy.Copyright © 2024 Ye, Li and Wang.