Abstract
There is no immunohistochemical or molecular marker to confirm the histologic diagnosis of pancreatic ductal adenocarcinoma (PDAC). This is particularly important in a scenario of unknown primary. Molecularly, PDAC is characterized by a limited set of driver mutations, and new predictive and prognostic markers are needed to guide novel therapies. Recent data show that DNA methylation profiles combined with complex machine learning algorithms are ideal tools to improve the diagnosis of PDAC. In addition, DNA methylation can be used to gain a deeper understanding of PDAC pathogenesis and further stratify this entity. Furthermore, exciting technologies have emerged, such as nanopore sequencing, which can be used to move these diagnostic tools from the postoperative to the intraoperative setting, or even as a liquid biopsy approach.