Department of General Surgery, Second Affiliated Hospital of Naval Medical University, Shanghai 200003, China.
Institutes of Biomedical Sciences, Fudan University, Shanghai 200032, China.
Pharmacol Res. 2022 Apr;178:106155. doi: 10.1016/j.phrs.2022.106155. Epub 2022 Mar 4.
The XELOX chemotherapy protocol that includes capecitabine and oxaliplatin is the routine treatment for colorectal cancer (CRC), but it can cause chemotherapy-related adverse events such as thrombocytopenia (TCP). To identify predictive biomarkers and clarify the mechanism of TCP susceptibility, we conducted integrative analysis using normal colorectal tissue (CRT), plasma, and urine samples collected before CRC patients received adjuvant XELOX chemotherapy. RNA-sequencing and DNA methylation arrays were performed on CRT samples, while liquid chromatography-mass spectrometry was performed on CRT, plasma, and urine samples. Differentially expressed features (DEFs) from each uni-omics analysis were then subjected to integrative analysis using Multi-Omics Factor Analysis (MOFA). Choline-deficiency in plasma and CRT was found as the most critical TCP-related feature. Based on bioinformatic analysis and literature research, we further concluded that choline-deficiency was the possible reason for most of the other TCP-related multi-omics DEFs, including metabolites representing reduced sphingolipid de novo synthesis and elevated solute carrier-mediated transmembrane transportation in CRT and plasma, DNA hypermethylation and elevated expression of genes involved in neuronal system genes. In terms of thrombocytopoiesis, these TCP-related DEFs may cause atypical maintenance and differentiation of megakaryocyte, resulting a suppressed ability of thrombocytopoiesis, making patients more susceptible to chemotherapy-induced TCP. At last, prediction models were developed and validated with reasonably good discrimination. The area under curves (AUCs) of training sets were all > 0.9, while validation sets had AUCs between 0.778 and 0.926. In conclusion, our results produced reliable marker systems for predicting TCP and promising target for developing precision treatment to prevent TCP.
XELOX 化疗方案包括卡培他滨和奥沙利铂,是结直肠癌(CRC)的常规治疗方法,但会引起化疗相关不良事件,如血小板减少症(TCP)。为了鉴定预测生物标志物并阐明 TCP 易感性的机制,我们使用 CRC 患者接受辅助 XELOX 化疗前采集的正常结直肠组织(CRT)、血浆和尿液样本进行了综合分析。对 CRT 样本进行了 RNA 测序和 DNA 甲基化阵列分析,同时对 CRT、血浆和尿液样本进行了液相色谱-质谱分析。然后,使用多组学因子分析(MOFA)对每个单组学分析中的差异表达特征(DEFs)进行综合分析。发现血浆和 CRT 中的胆碱缺乏是与 TCP 最相关的特征。基于生物信息学分析和文献研究,我们进一步得出结论,胆碱缺乏是大多数其他与 TCP 相关的多组学 DEF 的可能原因,包括代表 CRT 和血浆中从头合成的鞘脂减少和溶质载体介导的跨膜转运增加的代谢物、DNA 超甲基化和神经元系统基因参与的基因表达升高。就血小板生成而言,这些与 TCP 相关的 DEF 可能导致巨核细胞的非典型维持和分化,导致血小板生成能力受到抑制,使患者更容易发生化疗诱导的 TCP。最后,我们开发并验证了具有合理良好区分能力的预测模型。训练集的曲线下面积(AUCs)均>0.9,而验证集的 AUC 在 0.778 至 0.926 之间。总之,我们的结果为预测 TCP 提供了可靠的标志物系统,并为开发预防 TCP 的精准治疗提供了有前景的靶点。