The concept of autoimmune pancreatitis (AIP) was first reported in 1995 by Yoshida K et al. [1]. Even before this report, cases of AIP or chronic pancreatitis with other organ involvement such as the lacrimal glands, retroperitoneum, and biliary ducts had begun to accumulate in the literature [2–5]. Then in 2001, Hamano et al. [6] reported high serum IgG4 levels in patients with AIP, which is now a hallmark of type 1 AIP/IgG4-related disease (RD). In 2002, Yamamoto Y et al. [7] reported that patients with so-called Mikulicz’s disease, the disease ignored for nearly 50 years [8], also had high serum IgG4 levels. Then, through extensive discussion and communications among experts from different clinical fields, finally, a new concept of IgG4-RD was established by combining AIP, Mikulicz’s disease, and other diseases [9–12]. Since the concept of AIP was introduced in 1995 [1], cases of AIP have been identified not only in Japan but also in other countries, and interesting differences have been observed between AIP cases from Asia and Western countries [13–15]. In brief, AIP in Western countries occurs equally between males and females in a relatively young population and is frequently associated with inflammatory bowel disease. Importantly, no such cases had elevated serum IgG4 levels. These features are quite distinct from AIP with high serum IgG4 levels seen in Japan. After extensive discussion, a consensus statement on Asian diagnostic criteria for AIP [16] and international consensus diagnostic criteria (ICDC) for AIP [17] were created, and now, these two types of AIP are referred to as type 1 and type 2 AIP. Diagnosis of type 1 AIP is well established, whereas that of type 2 AIP mainly depends on histological findings, and none of specific serum markers have been identified. Fortunately, both types of AIP generally respond well to corticosteroid treatment. A consensus for treatment of AIP, however, has not yet been established. More importantly, the pathophysiology of AIP is largely unknown.

Although the pathogenesis of AIP remains a mystery, establishing both an accurate diagnosis and appropriate and effective treatments is currently the most important goal. Great progress has been made in diagnostic imaging such as computed tomography, endoscopic ultrasonography, and endoscopic retrograde pancreatography, but high serum IgG4 levels and dense infiltration of IgG4-positive plasma cells in the affected tissues are the core comprehensive diagnostic criteria for type 1 AIP/IgG4-RD [11, 17]. But, obtaining tissue samples is difficult in some patients with AIP; thus, diagnosis of type 1 AIP often depends on high serum IgG4 levels together with typical images and clinical manifestations. While a high serum IgG4 level is a very sensitive and specific marker for type 1 AIP, type 1 AIP patients with normal IgG4 and IgG levels are occasionally encountered. Thus, more universal serum markers for diagnosing type 1 AIP are needed, such as the specific autoantibodies for other autoimmune diseases. Although the histopathologic and clinical features are quite distinct for type 2 AIP [13–15], which is more prevalent in Western countries than in Asian countries, no characteristic immunologic changes have been detected in the serum or the tissues to date.

As for treatment, both types of AIP are very sensitive to corticosteroid administration, and indeed, many patients have been successfully treated with corticosteroids. There is currently no consensus, however, regarding the treatment regimen for induction or maintenance or therapeutic end points. Now, clinical trials are in progress in patients with IgG4-RD/type 1 AIP to establish a consensus for corticosteroid treatment regarding the initial dose, duration of treatment, and therapeutic end point. Importantly, it has been increasingly recognized that a considerable number of patients relapse after a decrease or cessation of corticosteroid administration [18, 19]. Therefore, at first, a treatment method for maintaining long-term remission must be established. Indeed, relapse may be prevented altogether by careful and long-term use of low-dose corticosteroids. Immunomodulators such as azathioprine and 6-mercaptopurine may have a role in maintaining remission in relapsing type 1 AIP. Recent reports also demonstrated a very potent therapeutic effect of anti-CD20 antibody, rituximab, on refractory IgG4-RD, including type 1 AIP [18–20]. Interestingly, the effect of anti-CD20 antibody appears to persist over time. This is a bit surprising but very interesting findings. Although the precise mechanism underlying the effect of CD20 antibody on pathophysiology of AIP is not clear, it may be noted that in addition to the decrease of B cells and IgG4 plasma cells [20], this antibody appears to induce significant reduction of T-cell infiltration in the affected tissues, suggesting a B-cell-T-cell interaction in the pathophysiology of type I AIP/IgG4-RD [21].

To develop more precise diagnostic methods and treatment options, elucidation of the pathophysiology of the disease is essential. Infection, allergic reaction, and autoimmune mechanisms are proposed to be involved in the pathophysiology of AIP. Despite intensive research, however, the pathogenesis and pathophysiology of both types 1 and 2 AIP remain unknown in most part. But, there are several clues in considering its pathophysiology particularly for type 1 AIP/IgG4-RD. A hallmark of type 1 AIP/IgG4-RD is a very high serum IgG4 concentration with dense IgG4-positive plasma cell infiltration in the affected tissues. In contrast to the generalized hypergammaglobulinemia associated with many other inflammatory or autoimmune diseases, type 1 AIP/IgG4-RD is characterized by high IgG1 and high IgE levels with normal or even low serum IgM and IgA levels in the serum [22]. Also, although IgG4 is known to have a critical role in the pathophysiology of some diseases, even in such diseases, the total IgG4 levels are not high [23, 24]. Therefore, distinct mechanisms for an IgG4 class switch must be involved in this disease. Early studies demonstrated an increase in Th2 cytokines such as interleukin (IL) 4 and transforming growth factor β in type 1 AIP/IgG4-RD [25], but recent reports demonstrated increase in several other cytokines as well, including IL10, a proliferation-inducing ligand (APRIL), and B-cell activating factor (BAFF), which might contribute to the increase in IgG4 plasma cells and high serum IgG4 levels [26, 27]. The data may also suggest the involvement of regulatory T cells and innate immune cells [25]. Nevertheless, it is unknown why these cytokines in particular are increased in this disease.

Whether the high IgG4 has a pathogenic role is an interesting question. In this regard, IgG4 autoantibodies have a critical pathogenic role in primary membranous glomerulonephropathy [23] and pemphigus vulgaris [24]. In contrast, because IgG4 is incapable of inducing complement activation [28], it is generally believed that IgG4 does not have a pathogenic role. However, the fact that IgG4 is deposited together with IgG1 and C3 and occasionally with C1q and C4 in the tubular basal membrane in IgG4-related nephropathy [29, 30] suggests a pathogenic role of IgG4 in this disease. Curiously, many patients with type 1 AIP/IgG4-RD have low C3 and C4 levels [30], suggesting complement activation. Alternatively, because IgG4 binds other immunoglobulins through their Fc regions [31], IgG4 deposition in the affected tissues might merely reflect its binding to other IgG subclasses such as IgG1 within the tissue.

Careful observation of patients with type 1 AIP/IgG4-RD reveals several interesting characteristic features as well. First, similar to Behcet’s disease or sarcoidosis, this disease rarely overlaps with other inflammatory or autoimmune diseases. For example, different from type 2 AIP, type 1 AIP is rarely associated with ulcerative colitis (UC) or primary sclerosing cholangitis (PSC). Although colitis is reported to be associated with type 1 AIP/IgG4-RD [32], this appears to be a manifestation of the disease itself rather than an overlap with other diseases [33]. Another point is that few familial cases have been reported [34]. This fact might contradict the idea of the involvement of genetic or environmental factors in the development of the disease, but the association of some genetic factors including MHC class II gene with type 1 AIP/IgG4-RD has been reported [35]. As recently reported, B-cell depletion by rituximab, a CD20 antibody, has a potent therapeutic effect on refractory type 1 AIP/IgG4-RD [19, 21], and this effect is associated with a reduction of not only serum IgG4 and infiltrating IgG4 plasma cells in the affected tissues but also T cells infiltrated in the tissues. These results might suggest not only an important regulatory function of B cells on T cells but also critical roles of T cells in the pathophysiology of type 1 AIP/IgG4-RD. In relation to this, whether or not tissues infiltrated with lymphocytes may serve as an effector site or an induction site requires clarification.

Another interesting issue is the relationship between type 1 AIP/IgG4-RD and cancer. Yamamoto et al. [36] and Shiokawa et al. [37] reported a high risk of cancer in patients with Mikulicz’s disease and AIP, respectively. Indeed, Shiokawa et al. found that among 108 AIP patients, 15 patients (13.9 %) had cancer during a median follow-up of 3.3 years, with a relative risk of cancer of 4.9 at the time of AIP diagnosis. It is well established that inflammation promotes cancer development, such as gastric cancer and liver cancer [38]. On the other hand, several cancers are known to induce paraneoplastic syndrome, such as dermatomyositis and pemphigus vulgaris. Notably, however, Shiokawa et al. found no pancreatic cancer in their patients. Moreover, the risk of cancer was highest within the first year after AIP diagnosis, and only few cases of AIP recurred after successful treatment of coexisting cancers, while relapse often occurred in patients without cancer cure. These data may favor the notion that some cases with AIP develop as a paraneoplastic syndrome, but further investigation is clearly required to obtain a final conclusion. Interestingly, previous studies suggested important roles of IgG4 autoantibodies in the pathophysiology of some paraneoplastic syndromes [23, 24].