Several case reports [6, 7, 10–15] and two retrospective studies [8, 9] have documented synchronous and metachronous pancreatic cancer in AIP (Table 26.2). In six (40 %) of the 15 cases reported, the pancreatic cancer was detected within a year after the diagnosis of AIP, and histological examination of the cancers in all six cases revealed considerable IgG4-positive plasma cell infiltration around areas of invasive ductal adenocarcinoma but only a few IgG4-positive plasma cells in the noncancerous areas of the pancreas [11, 13, 15]. Pancreatic intraepithelial neoplasia (PanIN) lesions are common precursors of pancreatic cancer and are characterized by columnar to cuboidal cells with varying degrees of atypia [37–39]. KRAS mutations have been detected in more than 95 % of invasive pancreatic ductal adenocarcinomas (PDACs), and more than 90 % of PanIN lesions have been found to harbor a KRAS mutation, suggesting that KRAS activation is one of the earliest genetic events in PanIN [37, 40–43]. Kamisawa et al. [44] demonstrated KRAS mutations not only in the pancreas but also in the tissues of organs involved by IgG4-RD, including the major duodenal papilla, common bile duct, gallbladder epithelium, gastric mucosa, and colonic mucosa. A retrospective analysis of the surgically resected pancreata of AIP patients revealed PanIN lesions, including PanIN 2 and PanIN 3 lesions, in 82 % of AIP patients [8]. These findings support the notion that KRAS mutations and PanIN lesions are associated with the development of pancreatic cancer in AIP. In fact, it is questionable whether persistent chronic inflammation due to LPSP induces cancer in AIP. While persistent subclinical inflammation without any signs of active AIP may have been present before the diagnosis of pancreatic cancer, many cases in which the diagnosis of pancreatic cancer was made at the same time as the diagnosis of AIP have been reported. Nevertheless, LPSP may not play a significant role in pancreatic carcinogenesis, and LPSP is more likely to be a paraneoplastic syndrome.

Several possible mechanisms to explain the association between AIP and malignant neoplasms have been suggested. The first possible mechanism that has been suggested is aging, because older age is a risk factor of cancer and AIP is found commonly among the elder population. The second possible mechanism is that widespread persistent IgG4-related fibroinflammation of the associated organs with abundant infiltration by T lymphocytes and Foxp3-positive cells may cause carcinogenesis in the various organs [11, 45, 46]. The third possible mechanism is that tumor cells may cause AIP by activating IgG4-related immune responses because tumor cells release various cytokines [47–50]. The fourth possible mechanism is that the dysregulation of B-lymphocytes associated with autoimmune disease may lead to abnormal B-lymphocyte proliferation and lymphoma [51]. Takahashi et al. [29] found that the SIR for malignant lymphoma in patients with IgG4-related disease was 16.0 (95 % CI: 3.3–45.5), suggesting that IgG4-RD is a risk factor for malignant lymphoma.

Although the mechanism underlying the development of malignant neoplasms in AIP patients has yet to be identified, it is important to examine AIP patients for malignant neoplasms when the diagnosis of AIP is made and during follow-up examinations.

In conclusion, all of the evidence that has accumulated thus far indicates that it is more likely that cancer induces the development of LPSP, that is, that LPSP develops as a paraneoplastic syndrome, than that LPSP induces cancer. Nevertheless, careful long-term follow-up of AIP patients is recommended because of their high incidence of malignant neoplasms.