Endoscopic Ultrasound with Fine Needle Aspiration (EUS-FNA) Is Valuable in the Diagnosis of Side-Branch Intraductal Papillary Mucinous Neoplasia (SB-IPMN) of the Pancreas

Abstract

Background: Pancreatic SB-IPMN often are discovered incidentally by imaging and mistaken for pseudocysts or other cystic neoplasia. Recent experience suggests that SB-IPMN may cause abdominal pain or pancreatitis and can lead to adenocarcinoma. Thus accurate diagnosis of this condition is important. Currently, no large series exists on the use of EUS-FNA in diagnosing SB-IPMN. Aims: Our goal was to assess the usefulness of EUS-FNA in the evaluation of SB-IPMN in a large university practice. Methods: Using an EUS database, patients referred for evaluation of pancreatic cystic lesions between October 2002 and October 2006 were identified. Using EUS-FNA features, CT or MRI and pathology, patients were assigned as probable or definite SB-IPMN. “Definite” diagnoses had surgical resections. EUS tapes were reviewed to determine the presence of main pancreatic duct communication (MPDC) if not mentioned in the EUS report. Patients with “probable” diagnoses had strong evidence suggesting the diagnosis. Patients with combined main and SB-IPMN were excluded. Results: 31 patients (15 male, mean age 68.5) with definite or probable SB-IPMN were identified. The mean lesion size was 20.9 mm (range 5-47), and 11 had multiple lesions. 28 cases (90%) were identified by EUS-FNA features alone while 3 (10%) were identified by EUS-FNA features and dedicated radiologic imaging. In 21 cases (68%) both MPDC and viscous fluid consistent with mucin were seen. In 13 of these 21 cases, cytology and CEA (median 316 ng/mL) suggested SB-IPMN. In 8 cases (26%) MPDC alone was seen, but elevated cyst CEA (median 429 ng/mL) or mucin on cytology were seen in 6, suggesting SB-IPMN. In 2 cases (6.5%) viscous fluid alone was seen, but cyst amylase and CEA were elevated (698 IU/L, 30,640 ng/mL) in 1 suggesting MPDC and thus SB-IPMN. Dedicated CT or MRI suggested SB-IPMN in only 18 patients (58%) and was significantly less accurate than EUS-FNA findings (p = 0.0078). Cytology suggested a mucinous neoplasm in 17 patients (55%). All 13 patients with surgical resections had MPDC seen and viscous fluid aspirated on EUS-FNA. Of these 13 cases, cytology and radiologic imaging were suggestive of SB-IPMN in 6 and 8 cases, respectively. Conclusions: In our experience, key EUS-FNA features (MPDC and viscous fluid) were highly valuable in diagnosing SB-IPMN. Elevated cyst CEA and cytology further enhanced the utility of EUS-FNA. Elevated cyst amylase implies MPDC and helps differentiate SB-IPMN from other mucinous cystic neoplasia when MPDC is not visualized. Radiologic imaging may aid in the diagnosis of some SB-IPMN, but EUS-FNA was superior in this study and should be the diagnostic modality of choice.