Integrative approaches to identify novel markers for early detection of pancreatic cancer

Accumulating evidence has shown the survival benefit of early detection for highly fatal cancer.

By utilizing novel approaches assessing prediagnostic imaging, circulating biomarker, and familial data in three large-scale pancreatic cancer projects in Sweden and the USA, this proposal aims to identify high-risk individuals and detect pancreatic cancer at a curable stage.

First, with computed tomography data from a healthcare records-based registry in the USA, we will assess the exclusive predictive value of pancreatic cancer using prediagnostic change in weight and body composition by deep-learning-based strategy.

Second, leveraging prediagnostic blood samples from five cohorts in the USA, we will test the role for early diagnosis of pancreatic cancer using a protein expression panel on circulating tumor-derived extracellular vesicles.

Third, using Swedish unique registry-based biopsy data, we will determine the risk of pancreatic cancer in individuals with first-degree relatives having precursor lesions of this disease.

Our findings will reveal innovative strategies for early detection of pancreatic cancer and will shed light on a future integrated framework of translational pancreatic cancer research in human.

With further validation, these state-of-the-art approaches can be implied clinically in Sweden to efficiently identify pancreatic cancer at an operatable stage and improve cancer survival.