Clinical utility of linear endosonography in patients with unexplained biliary dilatation and negative MRCP, with predictors for detection of neoplastic lesions

Abstract Background and study aims Unexplained biliary dilatation (UBD) frequently represents a diagnostic dilemma. Linear endosonography (EUS), with its close proximity to the pancreaticobiliary system and the ability for tissue acquisition, could have a role in revealing etiologies of UBD particularly in the context of negative other non-invasive modalities. However, in such patients, the decision for this semi-invasive procedure is usually debatable and often needs justification. Thus, we aimed to evaluate the diagnostic utility of linear EUS in patients with UBD after negative magnetic resonance cholangiopancreatography (MRCP) and to delineate predictors for EUS ability to detect neoplastic lesions. Patients and methods This was a prospective diagnostic observational study between 2018 and 2021. Included patients with evidence of biliary dilatation on abdominal ultrasound and negative MRCP underwent linear EUS examination. Results were compared to the final diagnosis relied on histopathology after EUS-guided biopsy or surgery for neoplastic lesions, while ERCP, diagnostic EUS criteria plus histopathology for autoimmune pancreatitis, Rosemont criteria for chronic pancreatitis, and/or follow-up for 6 months were the gold standard tests for non-neoplastic etiologies. Logistic regression was conducted to reveal predictors of neoplasm detection by EUS. Results Sixty-one patients (mean age 60 years, 32 females) were enrolled; 13.1% of them were asymptomatic. The final diagnosis was categorized into 29 patients with and 32 without neoplasms. Sensitivity, specificity, positive, negative predictive values (PPV, NPV), and accuracy of EUS-positive findings were 98.3%, 100%, 100%, 66.7%, and 98.3%, respectively. Corresponding results for neoplasm identification were 100%, 93.8%, 93.5%, 100%, and 96.7%, respectively. The most common neoplastic etiologies were small pancreatic and ampullary masses. Common bile duct (CBD) diameter > 12.9mm and CA19-9 > 37.1 u/ML were independent predictors for pancreaticobiliary neoplasm at multivariate regression analysis. Conclusion Linear EUS appears to have a high accuracy in detecting etiologies of UBD, with higher sensitivity for small pancreatic, ampullary lesions, and CBD stones. CBD diameter > 12.9 mm and elevated CA19-9 > 37.1 u/ML should raise a concern about the presence of occult pancreaticobiliary neoplasm, and then, EUS may be warranted even in asymptomatic patients.