The risk of pancreatic cancer following pancreatitis: an association due to confounding?

BACKGROUND & AIMS Chronic pancreatitis has been suggested as a causal risk factor for pancreatic cancer in a recent study. The aim of this study was to clarify the relationship between chronic pancreatitis and pancreatic cancer. METHODS All patients in the Swedish inpatient Register with a discharge diagnosis of pancreatitis from 1965 to 1983 were identified. They were stratified into subcohorts as follows: (1) one episode of unspecified pancreatitis (n = 823); (2) one episode of acute pancreatitis (n = 24,753); (3) recurrent pancreatitis (n = 7328); and (4) chronic pancreatitis (n = 4546). We also identified those with associated diagnoses indicating gallbladder disease or alcoholism. The patients were followed up through record linkage to the nationwide Swedish Cancer Register, Death Register, and Migration Register. RESULTS After exclusion of cancers occurring in the first year, there were excess risks for pancreatic cancer in all subcohorts. However, the risks declined with time in all subcohorts. A persistent excess risk after 10 years was restricted to patients with associated alcohol abuse (standardized incidence ratio, 3.8; 95% confidence interval, 1.5-7.9). CONCLUSIONS The findings are not consistent with reports that pancreatitis is causally associated with a long-term risk of pancreatic cancer. Selection bias, alcohol consumption, and smoking may contribute to some of the patterns of risk that have been observed.