Glucose metabolism, advanced glycation end products, and chronic inflammation and the risk of pancreatic cancer in the European Prospective Investigation into Cancer and Nutrition (EPIC)

  • Topic: Pancreatic cancer
  • Institution: DKFZ (German Cancer Research Centre)
  • Country: Germany
  • Status: Completed

Scientific abstract

(View plain language abstract)


Obesity, pancreatitis, and type 2 diabetes are well-established risk factors for pancreas cancer. However, the mechanisms underlying the relationship of these risk factors with cancer risk have only been partly resolved. It has been hypothesised, that these factors may be related to increased risk of pancreas cancer through disturbances in glucose tolerance and inflammatory responses.


We conducted a nested-case control study within the European Prospective Investigation into Cancer and Nutrition (EPIC) to investigate the association of impaired glucose metabolism (HbA1c), of an advanced glycation end product and its receptor (CML and esRAGE), and of pro- inflammatory cytokines (CRP, IL-6, sTNF-R1 and 2) with pancreatic cancer risk. Serum and plasma levels of inflammatory markers and AGE products were measured in 466 pancreatic cancer cases and 466 individually matched controls using highly sensitive enzyme-linked immunoassay. HbA1c levels in erythrocytes were analysed using HPLC.


Using conditional logistic regression models, we observed a gradually increased risk of pancreatic cancer with increasing pre-diagnostic HbA1c levels up to an OR of 2.42, even for individuals with HbA1c levels within the non-diabetic range. Elevated CML levels tended to be associated with a reduction in pancreatic cancer risk (OR = 0.57 [95% CI, 0.32–1.01]), whereas no association was observed for esRAGE. None of the inflammatory markers were significantly associated with risk of pancreatic cancer overall, although a borderline significant association was observed for higher circulating sTNF-R2 (crude OR = 1.52 [95% CI 0.97 – 2.39]).


Independently of diabetes, intermediate hyperglycaemia is a risk factor for pancreatic cancer. By contrast, our results do not provide evidence for an association of higher CML, lower esRAGE, or higher CRP, IL-6, or sTNF-R1 levels with risk of pancreatic cancer, whereas sTNF-R2 might be a risk factor. The role of AGE/RAGE and the involvement of pro- inflammatory cytokines in the pathogenesis of exocrine pancreatic cancer would benefit from further investigations.

Plain language abstract

Background & Hypothesis

Smoking, obesity, pancreatitis, and diabetes are well-established risk factors for pancreatic cancer. However, the mechanisms underlying the relationship of these risk factors with cancer risk have only been partly resolved. Higher blood levels of glucose, advanced glycation endproducts (AGE, potentially harmful products formed of sugars and proteins mainly in the body), or of inflammatory markers have been proposed as possible mechanisms.


We conducted a study within the European cohort EPIC to address the question, whether blood glucose, AGE or inflammatory factors are involved in the development of pancreatic cancer. In the EPIC study, more than 10 years ago, blood was taken of participants who did not have any cancer or other disease at that time. Now, almost 500 participants were diagnosed with pancreatic cancer. Of these men and women we measured glucose, AGE and inflammatory markers in blood and, in addition, also of the same number of subjects of comparable sex and age who did not develop pancreatic cancer. We then calculated risk of pancreatic cancer if any of these blood markers were either higher or lower than average.

Key findings

We observed that men and women with the highest levels of glucose had a 60% higher risk of developing pancreatic cancer compared to men and women with the lowest levels. This increase in risk was independent of obesity or diabetes, meaning that regardless of weight or being diabetic, pancreatic cancer risk is higher with higher blood glucose levels. The results for AGE and inflammatory markers were not consistent. Therefore, no final conclusion can be drawn for their risk in- or decreasing potential with respect to pancreatic cancer. However, it seemed as if women with higher levels of a specific inflammation marker (receptor 1 of TNF-alpha) also had an almost twice as high risk to develop a pancreatic tumour than women with low levels of that marker. In addition, another related inflammatory marker (receptor 2 of TNF-alpha) seemed to act the opposite compared to glucose. This means that the likelihood to develop a pancreatic tumour is higher in overweight or diabetic individuals if this inflammatory marker is also elevated.

In conclusion, independently of diabetes or obesity, high glucose levels are a risk factor for pancreatic cancer. By contrast, our results do not clearly support a role of inflammatory markers or AGE in pancreatic cancer development. However, research is still young in this field and further studies are needed to either prove or disprove what we found in the European EPIC study.

Grant publications