Department of Chemistry, Organic & Biological Analytical Chemistry Group, University of Liège, Allée du 6 août, B6c, B-4000, Liège (Sart Tilman), Belgium.
GIGA Institute, Translational Gastroenterology and CHU de Liège, Hepato-Gastroenterology and Digestive Oncology, University of Liège, Quartier Hôpital, Avenue de l'hôpital 13, B34-35, B-4000, Liège, Belgium.
Metabolomics. 2020 Aug 13;16(8):88. doi: 10.1007/s11306-020-01707-w.
Colorectal cancer is one of the most diagnosed cancers, leading to numerous deaths. In addition to existing screening methods, metabolic profiling could help both to diagnose and to understand the various states of the disease.
Find specific candidate biomarkers (CB) in serum of patients with colorectal cancer (CRC), in comparison to the situation after remission (R-CRC), evaluated on distinct patients.
All serum samples were analyzed using comprehensive two-dimensional gas chromatography (GC × GC) coupled to high resolution time of flight mass spectrometry (TOF-MS) through an optimized and validated untargeted analytical method regulated by a quality control (QC) system. First, we used a specific multi-approaches data (pre)processing workflow to highlight, annotate and assess the performances of the most altered metabolites between CRC patients (n = 18) and healthy control samples (HC, n = 19) specifically matched for age and gender, two of the most influential confounding factors. On the contrary, due to the difficulty to control for all clinical and demographic traits when sampling small cohorts, the samples from patients in remission (n = 17) were not matched. Because of the consequent risk of bias, the usual null hypothesis significance tests (NHST) could not be applied reliably. Therefore, we compared the R-CRC samples to another specifically matched group of healthy controls (R-HC, n = 17), and used this comparison to indirectly address the difference between patients with colorectal cancer and patients in remission through a measure called effect size (ES) whose methodological aspects were investigated.
24 candidate biomarkers were found significantly altered and able to discriminate the CRC and HC samples efficiently (Receiver Operating Characteristic (ROC) area under the curve (AUC) of 0.86, sensitivity and specificity of 0.72 and 0.78). 10 of those were found to have signals close to healthy levels in the R-CRC samples and were therefore specific to colorectal cancer. In the point-biserial case studied here, r-like (strength of association) and d-like (standardized mean difference) ES were directly convertible and only linear and rank-based ES were different. We therefore used and recommend Hedges' g, Spearman's rho and Kendall's tau, along with an unstandardized ES. The confidence intervals, that quantify the uncertainty of the measure, were well represented through scatterplots and distribution curves.
The candidate biomarkers found, along with their specificity, could help for the detection of colorectal cancer, the diagnosis of remission, and for the understanding of its pathophysiology, after proper validation on independent cohorts. The effect size, here applied on a MS global profiling data set, is an ideal complement to NHST and a useful tool to compare and combine distinct cohorts, within a study as well as between studies (meta-analysis).
结直肠癌是最常见的诊断癌症之一，导致大量死亡。除了现有的筛查方法外，代谢组学还可以帮助诊断和了解疾病的各种状态。
在不同的患者中，寻找结直肠癌（CRC）患者血清中的特定候选生物标志物（CB），与缓解后（R-CRC）的情况进行比较。
使用全面的二维气相色谱（GC×GC）结合高分辨率飞行时间质谱（TOF-MS），通过优化和验证的无靶向分析方法对所有血清样本进行分析，并通过质量控制（QC）系统进行调节。首先，我们使用特定的多方法数据（预）处理工作流程，突出显示、注释和评估 CRC 患者（n=18）与年龄和性别特定匹配的健康对照样本（HC，n=19）之间代谢物变化最大的性能，这是两个最具影响力的混杂因素。相反，由于在采样小队列时很难控制所有临床和人口统计学特征，因此缓解期患者（n=17）的样本未进行匹配。由于随之而来的偏倚风险，通常的零假设显着性检验（NHST）不能可靠地应用。因此，我们将 R-CRC 样本与另一组特定匹配的健康对照组（R-HC，n=17）进行比较，并使用该比较通过称为效应量（ES）的度量值间接解决结直肠癌患者与缓解期患者之间的差异，其方法学方面进行了研究。
发现 24 种候选生物标志物发生明显改变，能够有效区分 CRC 和 HC 样本（接收器工作特征（ROC）曲线下面积（AUC）为 0.86，灵敏度和特异性分别为 0.72 和 0.78）。其中 10 种在 R-CRC 样本中发现信号接近健康水平，因此是结直肠癌特异性的。在本文研究的点双列情况下，r 型（关联强度）和 d 型（标准化平均差异）ES 是直接可转换的，只有线性和等级 ES 不同。因此，我们使用并推荐 Hedges'g、Spearman's rho 和 Kendall's tau，以及未标准化的 ES。通过散点图和分布曲线可以很好地表示置信区间，该置信区间量化了度量的不确定性。
在所发现的候选生物标志物中，其特异性有助于结直肠癌的检测、缓解的诊断以及对其病理生理学的理解，在对独立队列进行适当验证后。这里应用于 MS 全局分析数据集的效应量是 NHST 的理想补充，也是比较和组合不同队列的有用工具，无论是在一项研究中还是在多项研究（荟萃分析）中。
