Abstract
BACKGROUND/OBJECTIVES: Autoimmune pancreatitis (AIP) is a diagnosis-challenging disease that often mimics pancreatic malignancy. Pancreatic resection is considered to be a curative treatment for pancreatic ductal adenocarcinoma (PDAC). This meta-analysis aims to study the incidence of AIP in patients who have undergone pancreatic resection for clinical manifestation of cancer. METHODS: A comprehensive search was conducted in three databases, PubMed, Embase and the Cochrane Library, using the terms 'autoimmune pancreatitis' and 'pancreatic resection' and supplemented by manual checks of reference lists in all retrieved articles. RESULTS: Ten articles were included in the final analysis. 8917 pancreatic resections were performed because of a clinical suspicion of pancreatic cancer. AIP accounted for 140 cases (1.6%). Type 1 AIP comprised the majority of cases, representing 94% (132 cases), while type 2 AIP made up the remaining 6% (eight cases) after further classification. AIP accounted for almost 26% of all cases of benign diseases involving unnecessary surgery and was overrepresented in males in 70% of cases compared to 30% in females. The mean age for AIP patients was 59 years. Serum CA 19 - 9 levels were elevated in 23 out of 47 (49%) AIP patients, where higher levels were detected more frequently in patients with type 1 AIP (51%, 22 out of 43) than in those with type 2 AIP (25%, 1 out of 4). The sensitivity of IgG4 levels in type 1 AIP was low (43%, 21/49 patients). CONCLUSION: Even with modern diagnostic methods, distinguishing between AIP and PDAC can still be challenging, thus potentially resulting in unnecessary surgical procedures in some cases. Serum CA 19 - 9 levels are not useful in distinguishing between AIP and PDAC. Work must thus be done to improve diagnostic methods and avoid unnecessary complicated surgery.