Although the role of EUS in the diagnosis of pancreatic cancer is well established, there is lack of evidence in the definite diagnosis of AIP. It is unclear if EUS imaging alone can reliably suggest definite diagnosis of AIP and differentiating between AIP and pancreatic cancer, since there are no pathognomonic findings suggestive of AIP on EUS. The lack of pathognomonic features and diverse spectrum of EUS imaging features limits the utility of EUS imaging alone [34]. The reported EUS finding for AIP is diffuse hypoechoic pancreatic enlargement, sometimes with coarse hyperechoic inclusions, and a mass lesion mimicking pancreatic cancer may be observed [16, 34, 35]. In AIP, on EUS, ductal wall thickening commonly spreads continuously from intrapancreatic CBD to the suprapancreatic middle bile duct (Fig. 12.3). In contrast, on EUS finding of patients with pancreatic head cancer, bile duct wall thickening is observed only at the intrapancreatic area. Hoki et al. reported that the findings of diffuse hypoechoic areas, diffuse enlargement, bile duct wall thickening, and peripancreatic hypoechoic margins are more frequent in AIP than in pancreatic cancer [36]. On the other hand, a focal hypoechoic area and focal enlargement are significantly more common in pancreatic cancer.

Real-time EUS can guide cytology/biopsy, allowing distinction of benign from malignant lesions for the pancreatic mass [22]. The EUS-guided fine-needle aspiration (FNA) provides sensitivity of about 80–90 %, specificity of about 95–100 %, and accuracy of about 90–95 % in the differential diagnosis between benign pancreatic disease and pancreatic cancer [34, 37, 38]. Although diagnostic sensitivity of EUS-FNA for pancreatic cancer is very high, some of the difficulty may exist in cases with well-differentiated carcinoma, those with extensive necrosis, and those with a background of chronic pancreatitis [39–41]. We should remember that a negative result of cytology/biopsy is not a guarantee of benign disease; hence repeat EUS-FNA or short-term follow-up imaging to assess steroids responsiveness is necessary [8, 23]. Actually, AIP is much less common than pancreatobiliary malignancies such as pancreatic cancer or cholangiocarcinoma. It should be emphasized that repeat EUS-FNA is warranted in patients with high suspicion of pancreatobiliary malignancies despite negative results at initial EUS-FNA [39]. If the steroid responsiveness is negative or equivocal, a definitive diagnosis always should be pursued by surgical exploration or resection. In the ICDC, negative work-up for pancreatobiliary malignancies is a prerequisite for a steroids trial [1].

For the histological diagnosis of AIP, EUS-guided trucut biopsy (TCB) is necessary. The primary role of EUS-FNA in a diagnosis of AIP may be to exclude malignancy rather than to provide definitive histological evidence for AIP [42]. While EUS-FNA with a small caliber (22/25 gauge) provides only cytological samples, a TCB needle (19 gauge) can acquire larger tissue specimen preserving tissue architecture and can permit a nonsurgical diagnosis of AIP [34, 43, 44]. For establishing a diagnosis of pancreatic cancer, EUS-TCB does not offer advantage over EUS-FNA; however, EUS-TCB should be considered when details of tissue architecture and immunostaining are required to establish a specific diagnosis [5, 34, 43]. EUS-TCB is particularly useful for diagnosing type 2 AIP patients who are seronegative and lack of OOI [22]. Until now, EUS-TCB is only available in a few specialized tertiary-care centers and is often not technically feasible according to a location of mass/enlargement in the pancreas [1, 34, 45]. However, EUS-TCB is expected to become more widespread with the availability of newly developed core biopsy needles with enhanced flexibility and improved tissue acquisition [44]. Newly developed core biopsy needles can access all areas in the pancreas. This may attain two goals of histologic examination in one sitting with the use of the same needle: exclusion of pancreatic cancer and histological diagnosis of AIP.

To overcome the limitations related to conventional EUS imaging, endosonographers have used several techniques including the characterization of tumor vascularization and estimation of elasticity distribution in normal and pathologic areas in the pancreas [46]. These imaging techniques can make EUS less operator dependent and improve the diagnostic yield of EUS-guided tissue sampling by selecting the most suspicious area of a tumor.

Elastography noninvasively gather informations about the relative hardness of the examined lesions compared with the surrounding tissues [47–49]. The premise is that malignant tumors are of harder consistency than benign ones. In a study by Dietrich et al., elastographic imaging of pancreatic cancer showed a markedly firmer area confined to the site of the low-echoic mass area, while in AIP cases, the hard area was not restricted to the mass lesion (Fig. 12.4).