Efficacy and limitations of the histological diagnosis of type 1 autoimmune pancreatitis with endoscopic ultrasound-guided fine needle biopsy with large tissue amounts.

OBJECTIVES: We examined the efficacy and limitations of acquiring large specimens by endoscopic ultrasound-guided fine needle biopsy (EUS-FNB) for diagnosing type 1 autoimmune pancreatitis (AIP). METHODS: Patients from 12 institutions with non-neoplastic diseases or pancreatic ductal adenocarcinoma (PDAC) with large EUS-FNB specimens were investigated. Slides stained with hematoxylin-eosin, elastic, IgG4, and IgG stains were evaluated. The IgG4- and IgG-positive cell numbers were counted in three foci. The diagnoses were based on the Japan Pancreas Society 2011 (JPS 2011) criteria and the International Consensus Diagnostic Criteria (ICDC). RESULTS: We analyzed 85 non-neoplastic (definite type 1 AIP in 73/85 based on the ICDC) cases and 64 PDAC cases. IgG4-positive cells were numerous (>10 in 85.9%), and the IgG4/IgG ratios were high (>40% in 81.2%). Plasma cell crushing by an artifact caused unsuccessful immunostaining, notably in smaller samples. Tissue lengths were an important factor for the presence of storiform fibrosis and obliterative phlebitis, but storiform fibrosis was equivocal even in large tissues. A definite or possible histological diagnosis was achieved in 45.9% (39/85) and 41.2% (35/85), respectively, and contributed to the definite final diagnosis of type 1 AIP in 33.3% (ICDC) and 55.6% (JPS 2011) in cases with segmental/focal lesions. In the PDAC group, >10 IgG4-positive cells was rare (2/58), but elastic stains revealed fibrous venous occlusions in 10.3% (6/58). CONCLUSIONS: EUS-FNB with large tissue amounts was useful for diagnosing type 1 AIP, notably by facilitating successful IgG4 immunostaining, but definite diagnosis may not be achieved even in cases with large specimens.