The efficacy and accuracy of endoscopic ultrasound for detecting common bile duct stones in intermediate to high‐risk patients with non‐diagnostic CT or MRCP

Abstract

Choledocholithiasis or common bile duct (CBD) stones are found in 15% to 20% of patients with symptomatic gallstone disease. Transabdominal ultrasound (TAUS), computed tomography (CT) and magnetic resonance cholangiopancreatography (MRCP) are the most commonly used imaging studies for patients with suspected CBD stones. However, the sensitivity of these modalities may decline in patients with microlithiasis (CBD stones smaller than 5 to 6 mm). From literature reviewed, endoscopic ultrasound (EUS) was as effective as MRCP and endoscopic retrograde cholangiopancreatography (ERCP), with higher accuracy in small CBD stones than MRCP and less invasive than ERCP. However, there is not much evidence in the past regarding the diagnostic value of EUS for detecting CBD stones in intermediate to high probability patients based on the American Society for Gastrointestinal Endoscopy (ASGE) guidelines, after non‐diagnostic CT or MRCP. A total of 92 patients with intermediate to high probability of choledocholithiasis based on ASGE guidelines were recruited in this study. We calculated the sensitivity, specificity, positive and negative predictive value, and positive and negative likelihood ratios of EUS for detecting CBD stones after non‐diagnostic CT or MRCP. By using univariate and multivariate analysis, we described the significance of patients' baseline characteristics and predictors that imply for further EUS study after inconclusive CT or MRCP. 35 (38%) out of 92 patients were found to have CBD stones on EUS. Among them, 30 had stones removed by ERCP, and three patients received cholecystectomy and choledocholithotomy with stones removal. Two patients with positive findings of CBD stones on EUS initially, ended up with negative findings on subsequent ERCP procedure. 57 (62%) patients with negative findings of CBD stones on EUS received regular follow‐up with a period of 9 to 24 months and showed no significant change in subsequent imaging and biochemical studies. The diagnostic accuracy of EUS for choledocholithiasis after non‐diagnostic CT or MRCP was 94.3%, with a sensitivity of 100% and specificity of 96.6%. Among all predictors for choledocholithiasis, the presence of clinical cholangitis was the only significant predictor in our multivariate analysis. In patients with intermediate to high probability of choledocholithiasis based on ASGE guidelines, EUS has shown a favorable accuracy and efficacy for detecting CBD stones when CT or MRCP is inconclusive. And the presence of clinical cholangitis is a significant predictor that indicates the necessity for EUS in patients with suspected CBD stones after non‐diagnostic CT or MRCP.