Study strengthens link between diabetes and pancreatic cancer

Patients with diabetes mellitus are up to 2.3 times more likely than those without diabetes to develop pancreatic cancer, findings from a retrospective Veterans Affairs (VA) study of over 500,000 patients show. Although the study did not compare the incidence in 1 outcome of pancreatic cancer for patients who received adequate treatment for diabetes, the study’s investigator believes that the results provide sufficient evidence for a strong link between pancreatic cancer and diabetes. In an interview with the head of the Pancreas Interest Group, consultant in the Division of Gastroenterology and Hepatology, and associate professor at the Mayo Clinic College of Medicine in Rochester, Minnesota, applauded the findings for supporting prior research that indicates an increased risk of pancreatic cancer among diabetics.

This study nicely confirms what is known from previous studies. However, long-standing diabetes as a marker for pancreatic cancer is likely to have limited clinical utility. This is because the strength of the association is modest at best and some, but not all studies show a decrease in risk with longer diabetes duration. Moreover, while the number of persons with pancreatic cancer in the population is small, the number of those with long-standing diabetes is large. Set out to document the incidence of pancreatic cancer among veterans, using the Vas database, to this end, he conducted a retrospective case-control study, using data from the VISN 16 VA database of 501,350 patients who were treated at 10 VA centers in four states between 1998 and 2004. The mean age of the subjects was 61 years, and 92 percent were male. Doctor conducted multiple logistic regression analyses, controlling for the known effects of age, alcohol use, smoking, body mass index and gallstone disease on the risk of developing pancreatic cancer.

We found that 106,825 patients (21 percent) in the database had been diagnosed with diabetes, and 199 (0.19) of these patients had pancreatic cancer. Among the 394,525 patients without diabetes, only 279 (0.07 percent) had pancreatic cancer. Multivariate statistical analysis confirmed the association, revealing that diabetics were 2.3 times more likely than non-diabetics to develop pancreatic cancer (95 percent confidence interval [CI], 1.91-2.77; P<0.001). The research also showed that the presence of gallstones is significantly associated with a risk for pancreatic cancer, with gallstone-positive patients 4.63 times more likely than gallstone-negative patients to develop pancreatic cancer (95 percent CI, 3.23-6.65; P<0.001).

In a presentation of these data at the 2004 annual meeting of the American College of Gastroenterology, we noted that his data have limitations. However, even in the face of these limitations, his findings still hold.

We did not account for therapeutic interventions, disease duration and control of diabetes in general. Furthermore, risks from family history, chronic pancreatitis and other factors known to increase the risk of pancreatic cancer were not accounted for in the study. Nevertheless, the large sample size of the database compensates for the effects of some of these limitations.