The sera of patients with AIP exhibited a polyclonal band in the rapidly migrating fraction of gel electrophoresis that contained γ-globulins; this resulted in the finding that β−γ globulin bridging was a characteristic of AIP. Immunoprecipitation assays revealed that this polyclonal band was the result of high serum concentrations of IgG4 [2]. IgG4 is typically a minor component in IgG fractions; it constitutes only 3–7 % of total serum IgG. However, the serum IgG4 concentrations in patients with AIP were over tenfold higher than those in healthy subjects. In addition, we found elevated serum IgG4 concentrations in 90 % of patients with AIP, but rarely in patients with other conditions, including pancreatic cancer, chronic pancreatitis, primary biliary cirrhosis (PBC), primary sclerosing cholangitis, and Sjögren’s syndrome (Fig. 9.1) [2, 11]. It has been reported that elevated serum IgG4 was found in a restricted number of conditions, including allergic disorders, parasite infestations, and pemphigus. Those results suggested that IgG4 represents a sensitive and specific marker for AIP, and it promised to be useful for the diagnosis of this disease [2]. The serum IgG4 concentration and the IgG4/IgG ratio were significantly reduced after corticosteroid therapy; this finding indicated that IgG4 might be a useful disease activity marker. Later, the clinical usefulness of IgG4 was assessed worldwide. Currently, serum IgG4 is considered a reliable marker for the diagnosis of AIP, and it has been included in various diagnostic criteria [4–10].

Patients with elevated IgG4 are considered to be in a highly active AIP disease state. Compared to patients with normal serum IgG4 levels, those with elevated IgG4 more frequently exhibit jaundice at AIP onset, diffuse pancreatic enlargement on imaging, significantly higher 18F-2-fluoro-2-deoxy-d-glucose uptake in pancreatic lesions, multiple extrapancreatic lesions, and a requirement for maintenance therapy [12, 13]. In addition, infiltration of IgG4-bearing plasma cells is a histological hallmark of AIP, and it is used in pathological diagnoses [14].

When the concept of AIP was first proposed, a high serum IgG concentration was listed as a characteristic laboratory finding [15]. Thus, IgG was considered a serological marker in the first diagnostic criteria proposed by the Japanese Pancreatic Society in 2002 [3]. However, the sensitivity and specificity of IgG are inferior for diagnosing AIP compared to IgG4. Therefore, IgG is currently used mainly as an activity marker to predict recurrence and estimate disease activity in clinical follow-ups of patients with AIP.

High serum IgE concentrations were detected in 30–40 % of patients with AIP, and the positive detection rate was also very high (86 %) [16, 17]. These findings suggested that an allergic mechanism may be contributing to the pathogenesis of AIP; thus, in some patients, AIP is complicated with an allergic response [16, 17]. However, the exact role or clinical significance of serum IgE elevation in AIP has not been fully elucidated. Although IgE does not necessarily reflect disease activity, the detection of elevated IgE might be a useful marker for the diagnosis of AIP in an inactive stage [17].

Interestingly, reduced IgA and IgM concentrations were detected in patients with AIP, in addition to increased IgG4 levels. It was reported that IgM was negatively correlated to IgG or IgG4 in patients with AIP. Moreover, the ratios of IgG:IgM and IgG:IgA in patients with AIP were significantly increased compared to those in patients with other diseases; thus, these ratios provided excellent diagnostic sensitivity and specificity in differentiating AIP from the other diseases. Those results suggested that IgG:IgM and IgG:IgA ratios may serve as novel diagnostic markers for differentiating AIP from other hepatopancreatic diseases [18].

ANA and RF are detectable in a wide range of autoimmune diseases, but their production may not be specifically related to those conditions. ANA and RF are typically detected in 30–50 % of AIP samples. However, this detection may represent a nonspecific, active disease state in immunological conditions [1]. After corticosteroid therapy, these autoantibodies promptly returned to undetectable levels.

Anti-SSA/Ro and anti-SSB/La autoantibodies are specific markers for Sjögren’s syndrome. Antimitochondrial antibody (AMA) is a specific marker for PBC. These disease-specific autoantibodies are seldom detected in patients with AIP [1, 19].

Carbonic anhydrase II (CA II) and lactoferrin are distributed in the ductal cells of the pancreas. Both proteins have been proposed as candidate target antigens in the pathogenesis of AIP, but the presence of autoantibodies to these antigens is not sufficiently specific or sensitive for an AIP diagnosis [20, 21].

Helicobacter pylori (H. pylori) infections may also trigger the occurrence of AIP, possibly as a result of molecular mimicry. There is substantial structural homology between the human CA II and the H. pylori alpha-carbonic anhydrase; the homologous segments contain the binding motif for the HLA molecule, DRB1*04:05, which is closely associated with AIP [22]. Those data led to the hypothesis that the DRB1*04:05-restricted peptide of CA II might be presented in genetically predisposed subjects; then, when reactive T cells and autoantibodies interact with the CA II of pancreatic ductal cells, they cause injury to pancreatic tissue [23]. Similarly, selected peptides from the plasminogen-binding protein (PBP) of H. pylori exhibited structural homology with the ubiquitin-protein ligase E3 component, n-recognin 2 (UBR2), which is highly expressed in pancreatic acinar cells. Antibodies against the PBP peptide were detected at high levels in patients with AIP, but they were barely detectable in patients with pancreatic cancer. It seems likely that the UBR2 in pancreatic acinar cells may be targeted by an autoantibody against the PBP of H. pylori in patients with AIP [24].