BACKGROUND Pancreatic cancer (PC) is the fourth leading cause of cancer death because of its subtle clinical symptoms in the early stage. To discover particular serum metabolites as potential biomarkers to differentiate pancreatic carcinoma from benign disease (BD) is on urgent demand. METHOD To comprehensively analyze serum metabolites obtained from 14 patients with PC, 10 patients with BD and 10 healthy individuals (normal control, NC), we separated the metabolites using both reversed-phase liquid chromatography (RPLC) and hydrophilic interaction liquid chromatography (HILIC). The data were acquired on a high-resolution quadrupole time-of-flight mass spectrometer operated in negative (ESI-) and positive (ESI+) ionization modes, respectively. Differential metabolites were selected by univariate (Student's t test) and multivariate (orthogonal partial least squares-discriminant analysis (OPLS-DA)) statistics. Sequential window acquisition of all theoretical spectra (SWATH) analysis was further utilized to validate the metabolites found in discovery stage. The receiver operator characteristics (ROC) curve analysis was performed to evaluate predictive clinical usefulness of 8 metabolites. RESULTS A total of 8 metabolites including taurocholic acid, glycochenodexycholic acid, glycocholic acid, L-glutamine, glutamic acid, L-phenylalanine, L-tryptophan, and L-arginine were identified and relatively quantified as differential metabolites for discriminating PC, BD and NC. The 8 metabolites and their combination discriminated PC from BD and NC with well-performed area under the curve (AUC) values, sensitivity and specificity. CONCLUSION Bile acids (especially taurocholic acid) performed to be potential biomarkers in PC diagnosis. Other amino acids (such as L-glutamine, glutamic acid, L-phenylalanine, L-tryptophan, and L-arginine) in serum samples from PC patients might provide a sensitive, blood-borne diagnostic signature for the presence of PC or its precursor lesions.

中文翻译：

---

基于LC-MS / MS和SWATH的血清代谢组学可实现胰腺癌中生物标志物的发现。

背景技术胰腺癌（PC）由于其早期的细微临床症状而成为癌症死亡的第四大主要原因。迫切需要发现特定的血清代谢物作为区分胰腺癌与良性疾病（BD）的潜在生物标记。方法为了全面分析从14例PC患者，10例BD患者和10例健康个体（正常对照组，NC）获得的血清代谢物，我们使用反相液相色谱（RPLC）和亲水相互作用液相色谱（HILIC）分离了代谢物。数据分别在高分辨率（ESI-）和正离子（ESI +）电离模式下运行的四极杆飞行时间质谱仪上获得。通过单变量选择差异代谢物（Student' st检验）和多元（正交偏最小二乘判别分析（OPLS-DA））统计数据。所有理论光谱的连续窗口采集（SWATH）分析被进一步用于验证发现阶段发现的代谢物。进行接收者操作者特征（ROC）曲线分析以评估8种代谢物的预测临床实用性。结果共鉴定出牛磺胆酸，糖基去氧胆酸，糖胆酸，L-谷氨酰胺，谷氨酸，L-苯丙氨酸，L-色氨酸和L-精氨酸等8种代谢产物，作为鉴别PC，BD和NC的鉴别代谢产物。 。这8种代谢物及其组合将PC与BD和NC区别开来，曲线下面积（AUC）值，灵敏度和特异性均表现良好。结论胆汁酸（尤其是牛磺胆酸）在PC诊断中可能是生物标志物。PC患者血清中的其他氨基酸（例如L-谷氨酰胺，谷氨酸，L-苯丙氨酸，L-色氨酸和L-精氨酸）可能会为PC或其前体的存在提供敏感的血源性诊断标记病变。