Background: Early diagnosis of pancreatic cancer holds the promise for improving the prognosis of this disease. We aimed to identify individual or combination of metabolites in pre-diagnostic blood that can be used as early diagnostic markers to differentiate pancreatic cancer from normal controls.

Methods: Using targeted approach, we conducted a two-phase case-control metabolomics study in the Women's Health Initiative Study using pre-diagnostic fasting blood. Each cancer-free control was individually matched to each pancreatic cancer case by age (± 1 year), ethnicity, history of treated type 2 diabetes, body mass index (± 5 kg/m2), study arm, current use of hormone treatment, blood draw time in year (± 1), blood draw time in hour (± 3). All samples from cases were collected less than one year before the diagnosis of pancreatic cancer. We used triple quadrupole mass spectrometer LS/GC-MS to quantitate 480 metabolites in sera of 30 cases and 30 controls. The selected candidate metabolites were validated in sera of additional 18 cases and 18 controls. Each metabolite was normalized against the standards. We used the paired t-test to compare the concentrations of each metabolite among all cases and controls. Unadjusted P value less than 0.25 was used for selecting candidate metabolites for validation and false discovery rate (FDR) adjusted Q-value < 0.25 was considered statistical significant in the validation phase. Log fold change (FC) was used to indicate the magnitude of difference between cases and controls.

Results: Blood samples were collected 241 days (SD: 94 days) before pancreatic cancer diagnosis on average. A total of 181 metabolites were detected in the study samples with 28 metabolites found to be significantly differentially expressed in cases versus controls with unadjusted P value < 0.25. Further validation in the pooled 48 cases and 48 controls found the metabolites 1-methyltryptophan was significantly higher in controls compared to cases with pancreatic cancer (FC: -0.23; FDR Q-value = 0.01). The levels of acadesine (FC: 0.29), aspartic acid (FC: 0.44), and nicotinamide (0.34) were higher among cases than controls and the level of citrulline (FC: -0.22), sn-glycerol 3-phosphate (FC: -0.25), and ketoglutarate (FC: -0.20) were significantly higher among controls than cases (FDR Q-value < 0.25).

Conclusion: Metabolites associated with perturbed amino acid metabolism were detected by metabolomics approach in pre-diagnostic blood of postmenopausal women who developed pancreatic cancer. In particular, depletion of tryptophan along kynurenine pathway has been associated with immunosuppression in tumor microenvironment in pancreatic cancer. Further study should determine the use of the metabolites of amino acid metabolism in early diagnosis of pancreatic cancer, in addition to lipid and choline metabolites that were identified by previous studies.

Citation Format: Li Jiao, Cristian Coarfa, Kimal Rajapakshe, Suman Maity, Liang Chen, Feng Jin, Vasanta Putluri, Lesley Tinker, Haleh Sangi-Haghpeykar, Hashem B. El-Serag, Nagireddy Putluri. Novel metabolism pathway in pancreatic cancer: findings from targeted metabolomics study. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 4300.