A population-based evaluation of the natural history of chronic pancreatitis

Background: Based on reports from tertiary care centers, chronic pancreatitis (CP) is considered to be a painful and debilitating disease frequently requiring invasive interventions. Our primary aim was to assess the natural course of CP in a population-based cohort using endoscopic and surgical interventions as surrogates for disease aggressiveness. Methods: We identified all patients (n = 89, alcoholic [ACP = 46], non-alcoholic [NACP] = 43) with newly diagnosed definite CP from Olmsted County, Minnesota between 1977 and 2006. Patients were followed until death or censoring. Medical records were reviewed at time of diagnosis and during each follow-up. Both lifetime proportions and cumulative incidence since the initial manifestation of CP were estimated and compared. Survival was estimated with Kaplan-Meier methodology. Results: Median age at CP diagnosis was 56 years (IQR, 48–67) and 56% were male. During median follow-up of 10 years, 68 (76%) experienced pancreatic pain, but only 27 (30%) needed any invasive therapeutic intervention: 23% had endotherapy and 11% had pancreatic surgery. During the clinical course, when compared with NACP, ACP patients had significantly more (all p < 0.05) pain (87 vs. 65%), recurrent acute pancreatitis (44 vs. 23%), pseudocysts (41 vs. 16%), cumulative incidence of exocrine insufficiency (60 vs. 21%), and annual hospitalizations after CP diagnosis (0.79 vs. 0.25). The cumulative risk of diabetes, calcifications, surgery and overall survival was similar in ACP and NACP. Conclusions: Our study suggests that CP at a population level may have a milder course than that reported from tertiary centers. We confirm that ACP has a more severe phenotype than NACP.