Oncotarget

Research Papers:

Serum metabolomics differentiating pancreatic cancer from new-onset diabetes

Xiangyi He, Jie Zhong, Shuwei Wang, Yufen Zhou, Lei Wang, Yongping Zhang and Yaozong Yuan _

PDF  |  HTML  |  Supplementary Files  |  How to cite  |  Order a Reprint

Oncotarget. 2017; 8:29116-29124. https://doi.org/10.18632/oncotarget.16249

Metrics: PDF 1648 views  |   HTML 2187 views  |   ?  


Abstract

Xiangyi He1, Jie Zhong1, Shuwei Wang2, Yufen Zhou3, Lei Wang1, Yongping Zhang1, Yaozong Yuan1

1Department of Gastroenterology, Ruijin Hospital, Shanghai Jiaotong University School of Medicine, Shanghai, China

2Department of Anesthesia, Ruijin Hospital, Shanghai Jiaotong University School of Medicine, Shanghai, China

3Department of Gastroenterology, Ruijin Hospital North, Shanghai Jiaotong University School of Medicine, Shanghai, China

Correspondence to:

Yaozong Yuan, email: yyz28@medmail.com.cn

Keywords: pancreatic cancer, metabolomics, serum, new-onset diabetes

Received: May 31, 2016     Accepted: February 20, 2017     Published: March 16, 2017

ABSTRACT

To establish a screening strategy for pancreatic cancer (PC) based on new-onset diabetic mellitus (NO-DM), serum metabolomics analysis and a search for the metabolic pathways associated with PC related DM were performed. Serum samples from patients with NO-DM (n = 30) and patients with pancreatic cancer and NO-DM were examined by liquid chromatography-mass spectrometry. Data were analyzed using principal components analysis (PCA) and orthogonal projection to latent structures (OPLS) of the most significant metabolites. The diagnostic model was constructed using logistic regression analysis. Metabolic pathways were analyzed using the web-based tool MetPA. PC patients with NO-DM were older and had a lower BMI and shorter duration of DM than those with NO-DM. The metabolomic profiles of patients with PC and NO-DM were significantly different from those of patients with NO-DM in the PCA and OPLS models. Sixty two differential metabolites were identified by the OPLS model. The logistic regression model using a panel of two metabolites including N_Succinyl_L_diaminopimelic_acid and PE (18:2) had high sensitivity (93.3%) and specificity (93.1%) for PC. The top three metabolic pathways associated with PC related DM were valine, leucine and isoleucine biosynthesis and degradation, primary bile acid biosynthesis, and sphingolipid metabolism. In conclusion, screening for PC based on NO-DM using serum metabolomics in combination with clinic characteristics and CA19-9 is a potential useful strategy. Several metabolic pathways differed between PC related DM and type 2 DM.


Creative Commons License All site content, except where otherwise noted, is licensed under a Creative Commons Attribution 3.0 License.
PII: 16249