Developing methods for advancing the early detection of pancreatic ductal adenocarcinoma leveraging electronic medical records data

Pancreatic ductal adenocarcinoma (PDAC) is projected to become the second-most common cause of cancerrelated death in the United States. Detecting PDAC early has been challenging due to its rapid progression and late presentation. Recent collaborative efforts among NCI, PanCAN, and NIDDK have resulted in multiple studies utilizing electronic medical records (EMRs). They provide opportunities to advance the early detection of PDAC, since the approach taken in the research setting is identical to what would be used in clinical practice. Future implementation in EMRs could lead to earlier diagnostic testing or interventions, enable the rapid identification of cancer trends or risk factors, and allow for quicker action. These studies, including our own, have demonstrated that individuals with glycemically-defined New Onset Hyperglycemia and Diabetes (NOD) have a significantly higher risk of being diagnosed with PDAC within 3 years of meeting diabetes mellitus criteria. The Enriching New-Onset Diabetes for Pancreatic Cancer (ENDPAC) score further stratifies individuals with NOD based on their PDAC risk, considering factors such as weight changes, blood glucose levels, and age at diabetes onset. However, several challenges need to be addressed to ensure the success of PDAC early detection. Firstly, it is important to better define the at-risk population for PDAC, as NOD alone may not be sufficient. Individuals with new-onset prediabetes and those with worsening long-standing diabetes mellitus may also be viable targets for early detection of PDAC. Secondly, more effective enrichment strategies leveraging biomarkers are critical, as directly applying existing biomarkers may not yield optimal results. For example, the ENDPAC score, which was developed primarily in a predominantly white NOD population, should be adapted for other high-risk groups, such as individuals with NOPD, LSDM, and diverse demographic populations to ensure broader applicability. We will develop a comprehensive set of methods to address these challenges, aimed at identifying high-risk populations, exploring promising biomarkers using EMR data, and integrating them with advanced biomarkers in an efficient manner. Given the expertise of our team and the strength of our collaborative partnerships, the advancements in methodologies, and the potential clinical applications, we anticipate that our efforts will yield tangible clinical benefits in the near term.