Abstract
BackgroundPancreatic cancer (PC) has much weaker prognosis, which can be divided into diabetes and non-diabetes. PC patients with diabetes mellitus will have more opportunities for physical examination due to diabetes, while pancreatic cancer patients without diabetes tend to have higher risk. Identification of prognostic markers for diabetic and non-diabetic pancreatic cancer can improve the prognosis of patients with both types of pancreatic cancer.MethodsBoth types of PC patients perform differently at the clinical and molecular levels. The Cancer Genome Atlas (TCGA) is employed in this study. The gene expression of the PC with diabetes and non-diabetes is used for predicting their prognosis by LASSO (Least Absolute Shrinkage and Selection Operator) Cox regression. Furthermore, the results are validated by exchanging gene biomarker with each other and verified by the independent Gene Expression Omnibus (GEO) and the International Cancer Genome Consortium (ICGC). The prognostic index (PI) is generated by a combination of genetic biomarkers that are used to rank the patient’s risk ratio. Survival analysis is applied to test significant difference between high-risk group and low-risk group.ResultsAn integrated gene prognostic biomarker consisted by 14 low-risk genes and six high-risk genes in PC with non-diabetes. Meanwhile, and another integrated gene prognostic biomarker consisted by five low-risk genes and three high-risk genes in PC with diabetes. Therefore, the prognostic value of gene biomarker in PC with non-diabetes and diabetes are all greater than clinical traits (HR = 1.102, P-value < 0.0001; HR = 1.212, P-value < 0.0001). Gene signature in PC with non-diabetes was validated in two independent datasets.ConclusionsThe conclusion of this study indicated that the prognostic value of genetic biomarkers in PCs with non-diabetes and diabetes. The gene signature was validated in two independent databases. Therefore, this study is expected to provide a novel gene biomarker for predicting prognosis of PC with non-diabetes and diabetes and improving clinical decision.
Highlights
Pancreatic cancer (PC) is an aggressive cancer of the digestive system, which is becoming a serious health problem worldwide
With the deeper understanding of the relationship between PC patient with diabetes and non-diabetes, recent data suggests that diabetes and altered in glucose metabolism are the consequence of PC, and yet, the clinical presentation of the altered glucose metabolism in these patients vary considerably (Yalniz & Pour, 2005)
In pancreatic patients with diabetes cohort, gender is significantly associated with overall survival time
Summary
PC is an aggressive cancer of the digestive system, which is becoming a serious health problem worldwide. Overall survival for patients with pancreatic cancer is poor, mainly due to a lack of biomarkers to enable early diagnosis and a lack of prognostic markers that can inform decision-making, facilitating personalized treatment and an optimal clinical outcome (Siegel, Miller & Jemal, 2016). We predict that PC patients with diabetes and non-diabetes are different in their prognostic biomarkers. Identification of prognostic markers for diabetic and non-diabetic pancreatic cancer can improve the prognosis of patients with both types of pancreatic cancer. Both types of PC patients perform differently at the clinical and molecular levels. The conclusion of this study indicated that the prognostic value of genetic biomarkers in PCs with non-diabetes and diabetes. This study is expected to provide a novel gene biomarker for predicting prognosis of PC with non-diabetes and diabetes and improving clinical decision
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
