DataSheet_1_Plasma metabolomics, lipidomics and cytokinomics profiling predict disease recurrence in metastatic colorectal cancer patients undergoing liver resection.docx

PurposeIn metastatic colorectal cancer (mCRC) patients (pts), treatment strategies integrating liver resection with induction chemotherapy offer better 5-year survival rates than chemotherapy alone. However, liver resection is a complex and costly procedure, and recurrence occurs in almost 2/3rds of pts, suggesting the need to identify those at higher risk. The aim of this work was to evaluate whether the integration of plasma metabolomics and lipidomics combined with the multiplex analysis of a large panel of plasma cytokines can be used to predict the risk of relapse and other patient outcomes after liver surgery, beyond or in combination with clinical morphovolumetric criteria.Experimental designPeripheral blood metabolomics and lipidomics were performed by 600 MHz NMR spectroscopy on plasma from 30 unresectable mCRC pts treated with bevacizumab plus oxaliplatin-based regimens within the Obelics trial (NCT01718873) and subdivided into responder (R) and non-R (NR) according to 1-year disease-free survival (DFS): ≥ 1-year (R, n = 12) and < 1-year (NR, n = 18). A large panel of cytokines, chemokines, and growth factors was evaluated on the same plasma using Luminex xMAP-based multiplex bead-based immunoassay technology. A multiple biomarkers model was built using a support vector machine (SVM) classifier.ResultsSparse partial least squares discriminant analysis (sPLS-DA) and loading plots obtained by analyzing metabolomics profiles of samples collected at the time of response evaluation when resectability was established showed significantly different levels of metabolites between the two groups. Two metabolites, 3-hydroxybutyrate and histidine, significantly predicted DFS and overall survival. Lipidomics analysis confirmed clear differences between the R and NR pts, indicating a statistically significant increase in lipids (cholesterol, triglycerides and phospholipids) in NR pts, reflecting a nonspecific inflammatory response. Indeed, a significant increase in proinflammatory cytokines was demonstrated in NR pts plasma. Finally, a multiple biomarkers model based on the combination of presurgery plasma levels of 3-hydroxybutyrate, cholesterol, phospholipids, triglycerides and IL-6 was able to correctly classify patients by their DFS with good accuracy.ConclusionOverall, this exploratory study suggests the potential of these combined biomarker approaches to predict outcomes in mCRC patients who are candidates for liver metastasis resection after induction treatment for defining personalized management and treatment strategies.

在转移性结直肠癌（mCRC）患者中，将肝脏切除术与诱导化疗相结合的治疗策略相较于单独化疗，能够显著提高患者的5年生存率。然而，肝脏切除术是一项复杂且昂贵的手术，近三分之二的患者会出现复发，这表明有必要识别出那些处于更高风险的患者。本研究旨在评估将血浆代谢组学和脂质组学整合，并结合大panel血浆细胞因子多联分析，是否能够用于预测肝脏手术后复发的风险以及其他患者结局，这超出了或与临床形态体积标准相结合。实验设计：通过600 MHz核磁共振光谱学对来自Obelics试验（NCT01718873）中接受贝伐珠单抗联合奥沙利铂方案治疗的30例不可切除mCRC患者的血浆进行代谢组学和脂质组学分析，并根据1年无病生存期（DFS）将患者分为应答者（R，n = 12）和非应答者（NR，n = 18）。在同一血浆样本上使用Luminex xMAP基于的多联微球免疫分析技术评估了大量细胞因子、趋化因子和生长因子。通过支持向量机（SVM）分类器构建了多生物标志物模型。结果：通过对在可切除性确立时进行反应评估时收集的样本的代谢组学分析获得的稀疏偏最小二乘判别分析（sPLS-DA）和负载图显示，两组之间的代谢物水平存在显著差异。两种代谢物，3-羟基丁酸和组氨酸，显著预测DFS和总生存期。脂质组学分析证实了应答者（R）和非应答者（NR）之间的明确差异，表明NR患者的脂质（胆固醇、甘油三酯和磷脂）显著增加，反映了非特异性的炎症反应。实际上，NR患者血浆中促炎细胞因子的显著增加得到了证实。最后，基于术前血浆中3-羟基丁酸、胆固醇、磷脂、甘油三酯和IL-6水平的组合的多生物标志物模型能够以良好的准确性正确分类患者的DFS。结论：总体而言，这项探索性研究提示了这些联合生物标志物方法在预测接受诱导治疗后进行肝脏转移切除术的mCRC患者结局中的潜力，从而为定义个性化的管理和治疗策略提供依据。