The clinical profiles of patients suffering from AIP seems to differ in the two forms of the disease [4]. Patients with type 1 AIP are older, with a higher prevalence of males and a more frequent involvement of other organs, both of the gastrointestinal tract and extra-gastrointestinal. The main clinical implication of this classification seems to be that type 1 AIP is a systemic relapsing IgG4-associated disease, whereas type 2 AIP is a serum IgG4-negative disease that does not relapse. However, the frequency of relapse(s) seems to be similar in the two forms of AIP, according to a recent paper from a French series [5].

AIP is a disease that quickly and fully responds to steroids [6–9]. The initial dosage of prednisolone ranges from 0.5 to 1 mg/kg/day, tapering to 2.5–5 mg every week.

The concept of the disease has changed over the time. Initially, AIP was considered to involve the entire pancreas, based on the Japanese experience. Later, the possibility of a segmental inflammatory involvement of the pancreas was proposed. Therefore, a clinically based classification of focal vs. diffuse AIP is now generally accepted [7, 10]. Focal AIP is defined as a segmental involvement of the pancreatic parenchyma with or without the presence of a low-density mass, as determined at imaging. In an Italian series, focal AIP was more frequent than diffuse AIP (63% vs. 37%) [7]. Compared to diffuse AIP, patients with focal AIP are older, more frequently males, and the main clinical presentation is jaundice. Since these findings are shared with pancreatic cancer, many patients with AIP undergo resective surgery (pancreaticoduodenectomy).

Indeed, this morphologically based classification has important clinical implications since in the presence of a focal disease, particularly if a low-density mass is detected at imaging, pancreatic cancer must be carefully and confidently excluded before steroid therapy is introduced [11]. The risk is, on the one hand, to operate on a patient with AIP that fully responds to steroids and, on the other, to treat a resectable cancer with steroids, delaying surgery and exposing the patient to the possibility of metastases or local invasion, both of which may preclude surgery. The diagnostic algorithm is therefore different in focal and diffuse forms of AIP.

In diffuse forms, a cholangiocarcinoma may be suspected in the presence of a single stenosis of the common bile duct. However, the probability of this being a malignancy (cholangiocarcinoma, peri-ampullar neoplasia) is very low and the differential diagnosis should mainly include acute pancreatitis (Figs. 17.3, 17.4).

By contrast, in the focal forms the likelihood of a cancer is very high. A recent review of the studies investigating the frequency of benign disease in patients who underwent pancreatic surgery for a resectable mass in the head of the pancreas found that 10% of these patients had an inflammatory pancreatic disease [11]. A recent study reported that half of these patients had a final diagnosis of LPSP. Therefore, in the presence of a pancreatic mass, only one patient in ten can be expected to have an inflammatory disease, while the large majority have a malignancy. The implication for clinical practice is that in the presence of a pancreatic mass pancreatic cancer needs to be excluded before patients are treated with steroids. Despite radiological, serological, and clinical findings highly suggestive of AIP, we strongly recommend that a biopsy of the pancreas be performed in such cases. For this purpose, fine-needle aspiration (FNA) biopsy is more accurate than core biopsy in the exclusion of cancer. Figure 17.5 shows the algorithm we propose in focal AIP with or without a hypodense mass at imaging. A trial with high-dose steroid therapy (1 mg/kg/day) may lead to a definitive diagnosis of AIP and should be made only if cytology is negative and the results of imaging, i.e., CT (Fig. 17.6), magnetic resonance imaging (MRI), endoscopic ultrasound (EUS), or contrast-enhanced EUS, are suggestive of AIP. At an international meeting held in Fukuoka, Japan, international consensus diagnostic criteria (ICDC) were recently proposed and can be used in this setting [12] (Figs. 17.5, 17.7). A second look after 3 weeks with the same imaging technique used at the basal examination is required. Normalization or significant improvement of the pancreatic morphology (disappearance of the hypodense mass, normalization of the pancreatic ductal system) is an important diagnostic finding. However, the decision to use steroids is difficult and should be made in an experienced tertiary center, after discussions among clinicians, radiologists, and surgeons.