EUS and/or EUS-guided FNA in patients with CT and/or magnetic resonance imaging findings of enlarged pancreatic head or dilated pancreatic duct with or without a dilated common bile duct.

Background: Incidental findings of an enlarged head of pancreas (HOP) or dilated pancreatic duct (PD) with or without a dilated common bile duct (CBD) on CT or magnetic resonance imaging (MRI), in patients without obstructive jaundice, raise suspicion for a pancreatic neoplasm, but their clinical significance has not been established.

Objective: To determine the prevalence of pancreatic neoplasm in this patient group.

Design: Retrospective analysis of a prospective database.

Setting: Tertiary-care university hospital.

Patients: Patients without obstructive jaundice at initial presentation, who underwent EUS and/or EUS-guided FNA (EUS-FNA) for an abnormal CT and/or MRI with an enlarged HOP (n = 67) or a PD with or without a dilated CBD (n = 43). The final diagnosis was based on definitive cytology, surgical pathology, and clinical follow-up.

Interventions: An EUS examination was performed by using a radial echoendoscopy followed by a linear echoendoscopy, if a focal pancreatic lesion was identified. Fine-needle aspirates were stained with Diff-Quik and Papanicolaou's methods, and were immediately assessed by an attending cytopathologist.

Main Outcome Measurements: (1) The prevalence of pancreatic neoplasms and (2) performance characteristics of EUS-FNA for identifying malignant neoplasm, in this patient group.

Results: In 110 study patients, the final diagnosis included adenocarcinoma (n = 7), pancreatic intraepithelial neoplasia (n = 1), neuroendocrine tumor (n = 1), tumor metastasis (n = 1), and benign cyst (n = 3). Thirty-two patients had EUS evidence of chronic pancreatitis, and, in the remaining 65 patients, the pancreas was normal. The accuracy of EUS and EUS-FNA for diagnosing pancreatic neoplasm in these patients was 99.1%, with 88.8% sensitivity, 100% specificity, 99% negative predicative value, and 100% positive predictive value.

Limitation: A retrospective design and surgical confirmation in only a small number of study patients.

Conclusion: A pancreatic neoplasm is seen in a clinically significant number of patients with "enlarged HOP" or "dilated PD with or without a dilated CBD" but without obstructive jaundice. EUS-FNA seems highly accurate for diagnosing pancreatic neoplasm in these patients.