Combining endoscopic ultrasound and tumor markers improves the diagnostic yield on the etiology of common bile duct dilation secondary to periampullary pathologies

Background: Unexplained common bile duct (CBD) dilatation may be caused by many etiologies, such as periampullary tumors, a pancreatic neoplasm, choledocholithiasis or an inflammatory stenosis. The aim of this study is to evaluate the diagnostic yield of endoscopic ultrasonography (EUS) in patients with unexplained CBD dilatation, in combination with tumor markers, liver chemistry, symptoms, surgical history and whether there is dilatation of the pancreatic duct (PD). Methods: From January 2016 to July 2017, 115 patients were referred for EUS in our center for CBD dilatation of an unknown etiology. A treatment plan is made based on the EUS result combined with the other clinical information. The final diagnosis is determined by surgical histology or follow-up of at least 3 months. Results: The sensitivity, specificity and accuracy of EUS for patients with choledocholithiasis were 100.0% (10/10), 100.0% (105/105) and 100.0% (115/115), respectively. The sensitivity, specificity and accuracy of EUS for patients with periampullary tumor were 86.5% (32/37), 89.7% (70/78) and 88.7% (102/115), respectively. The sensitivity, specificity and accuracy of EUS for patients with inflammatory stenosis were 88.2% (60/68), 89.4% (42/47) and 88.7% (102/115), respectively. The overall accuracy of EUS for diagnosing an undetermined etiology for CBD dilatation was 88.7% (102/115) and was higher than the accuracy of ultrasound (US) (64.1%), computed tomography (CT) (66.2%), magnetic resonance imaging (MRI) (67.0%) or PET-CT (66.0%). The sensitivity, specificity, positive predictive value, negative predictive value and accuracy of EUS combined with tumor markers for patients with malignant dilatation of CBD were 91.9% (34/37), 97.4% (76/78), 94.4% (34/36), 96.2% (76/79) and 95.7% (110/115), respectively. PD dilation (P=0.026) and weight loss (P=0.035) had significant predictive values of malignancy. Conclusions: EUS is an effective diagnostic tool for determining the etiology of a CBD dilatation, and offers meaningful information for guiding a treatment plan. EUS used in conjunction with tumor markers has high yield in differentiating benign and malignant CBD dilatation. More attention should be paid to patients with PD dilation or weight loss to prevent misdiagnosis of malignant CBD dilation.