2017 GRANTEE: Yu-Xiao Yang, MD, MSCE, FACP
University of Pennsylvania
Research Project: Validation of a Model Predicting Pancreatic Cancer in New-onset Diabetes
Award: 2017 Pancreatic Cancer Action Network Early Detection Targeted Grant
Award Period: July 1, 2017 – June 30, 2019
Dr. Yang is an associate professor of medicine (with tenure) and epidemiology in the gastroenterology division, department of medicine, and a senior scholar in the Center for Clinical Epidemiology and Biostatistics (CCEB) at the University of Pennsylvania. He is also a staff gastroenterologist at the Corporal Michael J. Crescenz Veterans Affairs (VA) Medical Center and a core investigator of the Center for Health Equity Research and Promotion. He is board certified in internal medicine and gastroenterology. Dr. Yang has had extensive formal training in epidemiology and clinical research methodology and holds a Master of Science degree in clinical epidemiology from the University of Pennsylvania. He earned his MD at New York University School of Medicine.
The vast majority of pancreatic cancer cases are diagnosed at late stages and carry an abysmal prognosis. Thus, promoting early detection is the key to improving survival among pancreatic cancer patients. Approximately half of all patients with pancreatic cancer develop diabetes mellitus about two to three years before their cancer diagnosis. On the other hand, among those with newly diagnosed diabetes, the probability of developing pancreatic cancer in the ensuing three years is 0.85 percent. This seemingly low risk level is actually eight times the expected risk in the general population.
Based on current knowledge, Dr. Yang and his research team believe that diabetes that comes on shortly before the diagnosis of pancreatic cancer is different from the typical adult-onset diabetes (also known as type 2 diabetes mellitus); it is likely caused by an early-stage pancreatic cancer that has not progressed far enough to be discovered. Patients with this pancreatic cancer-associated diabetes are likely different in a number of ways from patients with the typical adult-onset diabetes.
The overarching goal of this investigation is to develop a statistical prediction model that captures the differences between those with pancreatic cancer-associated diabetes and those with regular diabetes. Dr. Yang recently built such a tool in a large patient population from the United Kingdom. It performed quite well in distinguishing the two types of diabetes in that population. In addition, it carries virtually no cost or risk because it uses information routinely available at the time of diabetes diagnosis, such as patients’ weight, weight changes, smoking status, glucose level, cholesterol level, medication use, etc.
However, before this tool can be considered for wider use in clinical practice, Dr. Yang and his colleagues must determine whether it performs equally well in another patient population. Therefore, the aim of the current project is to test and refine the performance of the prediction tool in a patient population other than the one from which it was derived. The study will include all U.S. veterans with newly diagnosed diabetes mellitus after the age of 35. Successful completion of this project will pave the way for the formulation of a targeted pancreatic cancer screening strategy in the new-onset diabetes population that begins with applying this risk prediction tool. By using this approach, Dr. Yang hopes to allow earlier detection of a substantial burden of pancreatic cancers by as much as three years.