The article I am focusing on draws a link between the risk of developing pancreatic ductal adenocarcinoma with new onset of diabetes (I assume this is type II diabetes) in male and female patients 50 years or older in Rochester, MN from the beginning of 1950 to the end of 1994.
Pancreatic cancer is often not detected until the cancer is in an advanced stage. When it is detected early, it is done so essentially “by luck” (because the patient has not yet started displaying cancer related symptoms). Early detection techniques such as an endoscopic ultrasonography, (a probe with a sonograph on the end of it, which uses high-frequency sound waves to produce a graphic representation of an area of interest. See this youtube video) and endoscopic retrograde cholangiopancreatography (inject dies into the pancreas and watch them travel using X-rays) (http://www.webmd.com/digestive-disorders/endoscopic-retrograde-cholangiopancreatogram-ercp) are the current early detection techniques used when fighting pancreatic cancer. While effective, these are rather invasive for a patient who is not at risk for pancreatic cancer. To screen for potential pancreatic cancer patients who are at an elevated risk, my article suggests using a biomarker that could hint toward cancer--researchers suggest using hyperglycemia and diabetes.
Patients with hyperglycemia and “new onset” diabetes are identified as high risk for developing diabetes. The problem with using diabetes as an early marker for pancreatic cancer is limited in the sense that many people have diabetes, yet not many have pancreatic cancer. While not every one with diabetes develops pancreatic cancer, the results are startling for the opposite. 80% of patients who have pancreatic cancer also have diabetes and hyperglycemia. Additionally after successful treatment of pancreatic cancer (if at all possible), diabetes and hyperglycemia often get better and even remit.
Finding patients to identify a possibly correlation between “new onset” of diabetes and pancreatic cancer is quite difficult. In fact, 1/3 of patients with type II diabetes are undiagnosed. Many patients that are potential study interests are those who were hospitalized due to diabetes or patients at a clinic, far too advanced to be described as new onset diabetes. This is because the initial onset of diabetes is unknown. (I must emphasize that this article draws correlations between ONLY new onset cases of diabetes and pancreatic cancer.)
Results: Of the 2122 patient who were eligible for the study, (were diagnosed for diabetes mellitus between 1950-1994 and were over the age of 50 at time of diagnosis) 18 diabetes subjects met the criteria for also having pancreatic adenocarcinoma within 3 years of diabetes diagnosis (0.85%). Of the 18, 11 did not have a history of diabetes in their family. Additionally, 12 of the patients were males (67%). This 3 year incidence of pancreatic cancer coupled with diabetes is 8 times higher than comparable people (age, sex
In my opinion, because the patients were not screened for diabetes nor pancreatic cancer, (these were discovered due to other ailments), using hyperglycemia or diabetes as a biomarker for pancreatic cancer cannot yet be viewed as causal evidence, simply correlation in this study. Additionally, the small sample size of patients who actually had pancreatic adrenocarinoma (n=18) is not a large enough sample size to begin to make the claims that the researches are making. What makes me most uncomfortable about this article is the fact that only Rochester, Minnesota is studied. Rochester, Minnesota is less than 200,000 residents now and at the start of the study only had 30,000 residents. Further more, the 2000 Rochester census reports that nearly 90% of the residents are non-Hispanic whites (75% white in all of the USA). This sample does not represent the country as a whole. Additionally, the awareness and treatment for diabetes differ greatly from 1950 to 1994. Further more, the diet of those in Rochester, Minnesota is different than those from the rest of the United States. I feel that this study is an interesting way to start looking at early detection techniques for pancreatic cancer yet does not have the breath to do so.
An interesting question raised by the researchers: Would screening for cancer at the time of detection of diabetes led to detection of resectable (surgically removable) pancreatic cancer?
--Michael
If you would like to read this fantastic article here is the link: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2377196/?tool=pubmed
