A prospective intra-individual study to compare the image quality of magnetic resonance (MR) pancreatography at 3.0 T and 1.5 T in patients with autoimmune pancreatitis.

Autoimmune pancreatitis (AIP) is a unique form of chronic pancreatitis caused by an autoimmune mechanism that responds well to steroid therapy. One of the most important issues on AIP is to distinguish it from pancreatic cancer as the treatments are totally different from each other. An accurate differentiation of AIP from pancreatic cancer is therefore crucial.

Two most important image findings of AIP are pancreatic enlargement and pancreatic ductal stricture. When CT shows typical diffuse sausage-like swelling of the pancreas and peripancreatic hypodense rim, it is easy to differentiate AIP from pancreatic cancer. However, those typical cases are not very common and, moreover, 30% of AIP manifest as focal mass/enlargement of the pancreas, making a differential diagnosis very difficult. When pancreatic feature is atypical at CT, it is important to find diffuse or multifocal stricture of the main pancreatic duct that is characteristic feature of AIP. In AIP, a diffusely attenuated pancreatic duct is thinner than normal, and this does not appear at CT. Pancreatography is therefore necessary.

Two current imaging tools to demonstrate the pancreatic duct are endoscopic retrograde pancreatography (ERP) and MR pancreatography (MRP). ERP provides high resolution images using different projections and enables various procedures including aspiration/biopsy and stent insertion. However, the use of diagnostic ERP in diagnosing AIP has been debated as ERP is an invasive procedure, having potential complications including pancreatitis, perforation of the stomach or duodenum. Moreover, it is difficult to perform endoscopic procedure in patients who underwent gastric surgery. Whereas, MRP can noninvasively image the pancreatic and biliary systems at the same time without risks of procedure-related complications and can evaluate other intrabdominal organs on cross-sectional images. The relatively lower spatial resolution of MRP using 1.5 T compared with ERP images may make it difficult to demonstrate fine changes of the pancreatic duct in AIP and sometimes make false positive or negative findings.

The superiority of 3.0 T over 1.5 T MR systems has been observed in several studies. However, only a few studies using the 3.0 T MR systems in the pancreaticobiliary tract have been reported and, furthermore, the usefulness of 3.0 T MRP for the diagnosis of AIP has not yet been investigated.

The purpose of this study is to prospectively compare the image quality of MRP at 3.0 T and 1.5 T in patients with AIP using ERP as the reference standard.

Inclusion Criteria:

• Typical CT findings (diffuse sausage-like pancreatic swelling or multifocal pancreatic swelling with or without peripancreatic rim, multifocal biliary stricture, renal lesion, or retroperitoneal fibrosis)
• Serum level of immunoglobulin G fraction 4 > 135mg/dL

Exclusion Criteria:

• Patients under 20 years of age
• Women who are pregnant, lactating or who are of childbearing potential
• Patients with any physical or mental status that interferes with the signing of informed consent
• Patients with a contraindication for MRP or ERP examination

Patients with probable autoimmune pancreatitis based on clinical and CT findings