Sometimes Diabetes Means Cancer

Walter Park, MD, an assistant professor of gastroenterology & hepatology, acknowledges that it will be many years before he recognizes the fruits of two of his current projects. The first is a large consortium targeting chronic pancreatitis funded by the National Institutes of Health through a U01 grant; one of its goals is to examine a relationship between newly diagnosed diabetes and pancreas cancer. The second is a biobank of pancreatic cyst fluid that he started eight years ago to help unlock some of the secrets of pancreas cancer.

Chronic pancreatitis is a debilitating and painful condition about which little is known. There are few treatments; patients have chronic pain; and it is a difficult disease to manage, especially as many patients are prescribed narcotics and often develop drug dependencies. It is also a risk factor for pancreas cancer.

When the National Institutes of Health announced that two of its institutes, the National Cancer Institute (NCI) and the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDKD), would fund a U01 grant to support a discrete project in chronic pancreatitis, Park and two Stanford colleagues — Aida Habtezion, MD, MSc, assistant professor of gastroenterology & hepatology, and Seung Kim, MD, PhD, professor of developmental biology — applied and were successful, along with nine other centers.  The 10 centers then formed a consortium.

Park explains that the two institutes “realized that this was a poorly understood area where new knowledge would be helpful: from the NCI perspective, particularly as a strategy to identify early cancer; from the NIDDKD perspective, to better understand the natural history of chronic pancreatitis.”

The focus of the U01, therefore, is on studying the natural history of chronic pancreatitis and its complications, specifically including the development of diabetes and pancreas cancer.

Many patients with chronic pancreatitis develop diabetes as a complication. Diabetes became known as an important factor following a study at the Mayo Clinic that looked at patients with pancreas cancer and found that many of them had newly diagnosed diabetes as well. This suggested that a recent diagnosis of diabetes could be connected in some way with pancreas cancer. And interestingly, says Park, “when some of these patients went to surgery because they had local resectable cancer, their diabetes went away after they removed the tumor. This stimulated a hypothesis that for some patients, diabetes is a signal, and the diabetes may have formed as an effect of the tumor in the pancreas.”

With this as background, Park describes the dual goals of the 10-center consortium: to amass a large enough sample size to make sense of the relationship between diabetes and chronic pancreatitis; and to study the natural history of chronic pancreatitis. “Two major cohorts are being developed,” he explains. “One is 2,000 patients with chronic pancreatitis, who we will follow over 10 to 20 years. The other is new-onset diabetics over the age of 50 who are otherwise well, and we’ll follow them with the expectation that in about one percent of the patients the diabetes is actually a reflection of cancer. We have to recruit 10,000 new-onset diabetic patients to get to 100 patients with pancreas cancer.”

The two other principal investigators in the Stanford group bring expertise in immunology and candidate biomarkers. Park describes the contributions his two co-PIs anticipate making to the study: “Aida Habtezion, who is an immunologist in our division, will enable us to better define certain immune profiles to try to predict cancer as well as to predict whose chronic pancreatitis is going to be worse. Seung Kim, who is a Howard Hughes investigator in the department of developmental biology, has identified a potential biomarker called Neuromedin U that could explain this tumor effect on diabetes and could be detected in the blood. In our proposal, we highlighted his work and suggested that we have some potential candidate biomarkers that we could use to try to identify whose diabetes might be related to the early onset of cancer.”

Because of the difficulty of enrolling large numbers of patients with either chronic pancreatitis or new-onset diabetes, a consortium was necessary. “Once patients are recruited,” Park says, “we’ll be collecting and banking biospecimens for biomarker evaluation and validation from a sample size large enough to allow us to develop some meaningful observations. This material becomes the substrate for all the different ideas each center has.”

Organizing a consortium of 10 centers, each with its own principal investigators, hypotheses, and expertise, is not an exercise for the faint of heart, and it takes time. Park describes it as having “a lot of chefs in the kitchen. There’s a process of consensus that takes a bit of time. But we’re almost done completing the study design for the prospective cohorts. We hope to launch these cohorts in January 2017 and to recruit all the patients we need in three years, ending in 2020. Then we follow them for as long as possible.”

“This study will probably take me through to the mid to end of my career.”

A Clinician for Patients with Pancreas Cancer

When not tending to the U01, Park devotes his clinical and research time to early detection strategies for pancreas cancer, which is one of the few cancers that are rising in incidence, lacking much progress in either screening or prevention.

Park has been focusing on pancreatic cysts, known precursor lesions for pancreas cancer. Thanks to the use of CT and MRI in clinical practice there have been many incidental findings on the pancreas, and these include pancreatic cysts. Park points out that it is important to recognize that “not all pancreatic cysts have potential to become cancer but approximately half do. As our imaging has gotten better, we are finding these at an increasingly alarming rate. And because we can’t reassure the patient that this is just a benign incidental finding, it has caused a lot of anxiety over the past 10 years.”

The way to calm the anxiety is to remove the cyst, but that is not without significant risk. “It carries a mortality rate of at least two percent in the hospital, and complications are quite common, as high as 30 percent,” says Park. Equally important, he continues, “what patients don’t realize is that the risk of cancer from many of these cysts is actually quite low. The risk of taking them to surgery is probably higher than the chance that it would become cancer in the next year.”

So, back in 2008, when he was finishing his fellowship at Stanford, he started collecting cyst fluid from patients during endoscopic procedures. “We’d send part of it for clinical care,” he says, “and the other part to our freezer. Since then we’ve maintained a database of these samples, and we have over 300 now, which is a wonderful resource for quickly identifying and validating potential promising biomarkers.”

Park also works with Stanford colleagues to try to discover new biomarkers. So far two successful collaborations have identified potentially new biomarkers which are currently being validated. One collaboration is with Gary Peltz, MD, professor of anesthesiology, perioperative and pain medicine, who is interested in metabolomics. And the other collaboration is with Anson Lowe, MD, associate professor of gastroenterology & hepatology, with whom Park is looking at another biomarker called amphiregulin.

Park is on a mission to fulfill some of the needs of patients with pancreatic cysts. “We need better biomarkers, better tools to help us discern which cysts have any potential to become cancer and then, more importantly, which of them have features that show that cancer may be imminent.”