Biopsy diagnosis of type 1 autoimmune pancreatitis: Does it bring a conclusion or confusion?

Abstract

A biopsy‐based diagnosis of type 1 autoimmune pancreatitis (AIP) is now feasible via an endoscopic ultrasound‐guided fine‐needle biopsy, but there are potential issues to address. The benefits of acquiring large tissue samples include more successful immunostaining for Immunoglobulin G4 and more identifications of storiform fibrosis, obliterative phlebitis, and the ductal lesions of type 1 AIP. However, storiform fibrosis may not be present in all the type 1 AIP lesions. An interobserver agreement study revealed only slight‐to‐moderate agreement among pathologists diagnosing the histological findings of type 1 AIP. Potential reasons for disagreement are the different time phases of the inflammation (which result in heterogeneous histological pictures), a focal appearance of the typical histological findings, and the different definitions used by pathologists. We have thus devised guidance for diagnosing type 1 AIP based on biopsy tissues. In this guidance, we define each histological finding of type 1 AIP, for example, storiform fibrosis as a swirling arrangement of inflammatory cells, spindle‐shaped cells, and delicate collagens as a unit. The necessity of elastic stains for identifying obliterative phlebitis is explained, with examples of mimickers. Another important purpose of a biopsy in type 1 AIP cases is differentiation from pancreatic ductal adenocarcinoma (PDAC). In this situation, acinar‐ductal metaplasia observed in type 1 AIP is a mimicker of PDAC and should not be confused. For the resolution of potential disagreements among pathologists, a multi‐disciplinary approach with the collaboration of clinicians, radiologists, and pathologists is necessary to avoid confusion.

Autoimmune Pancreatitis and Immunoglobulin G4-related Sclerosing Cholangitis: Past, Present, and Future

“…To differentiate AIP from pancreatic cancer, the MRI features favoring AIP are diffuse enlargement (diagnostic OR, 75), a capsule-like rim (52), multiple strictures of the main pancreatic duct (47), and homogeneous delayed enhancement (46). 47 In contrast, the MRI features favoring pancreatic cancer are target-type enhancement (diagnostic OR, 41) and a discrete pancreatic mass (35).…”

“…The key histopathological features of type 1 AIP include 1) diffuse lymphoplasmacytic infiltration and fibrosis, 2) numerous IgG4-positive cells, 3) storiform fibrosis, 4) obliterative phlebitis, and 5) ductal lesion (periductal infiltrates with fibrosis). 16,41,52 EUS-guided sampling from the pancreas can provide a definite diagnosis of AIP. Endosonographers use FNB needles for patients with suspected AIP because acquiring sufficient amounts of tissue is essential for an AIP diagnosis.…”

“…Endosonographers use FNB needles for patients with suspected AIP because acquiring sufficient amounts of tissue is essential for an AIP diagnosis. 41,52 A recent meta-analysis showed that the pooled diagnostic yields for level 1 or 2 histology criteria of AIP were 55.8% for FNA and 87.2% for FNB (p=0.030). 41 Therefore, FNB needles (≥22 G in size) should be used when a pancreatic biopsy is required to differentiate AIP from pancreatic cancer.…”

“…Regarding the pathological examination of samples, agreement among pathologists about the diagnosis of type 1 AIP has been reported to be suboptimal. 52 The observer-dependent nature in interpreting storiform fibrosis and obliterative phlebitis is problematic. 41,52 For a better evaluation of obliterative phlebitis, the addition of elastic stains, such as Elastica van Gieson staining, should be considered.…”

“…52 The observer-dependent nature in interpreting storiform fibrosis and obliterative phlebitis is problematic. 41,52 For a better evaluation of obliterative phlebitis, the addition of elastic stains, such as Elastica van Gieson staining, should be considered. 41 IgG4…”