Intraductal US is a useful adjunct to ERCP for distinguishing malignant from benign biliary strictures

Background: Distinguishing malignant from benign biliary strictures remains a challenge. This prospective study assessed intraductal US as an adjunct to endoscopic retrograde cholangiography and tissue sampling for diagnosis of malignant and benign biliary strictures. Methods: Sixty-two patients were enrolled who had biliary strictures suspected to be malignant but with negative tissue sampling by endoscopic retrograde cholangiography, or suspected biliary strictures based on clinical manifestations and/or cross-sectional imaging. Intraductal US was performed with an over-the-wire 2.4-mm diameter 20 MHz catheter US probe. The diagnostic accuracy of endoscopic retrograde cholangiography plus tissue sampling with and without intraductal US was compared with surgical findings or clinical outcome at a 12-month follow-up (benign/malignant) in nonsurgical cases. Tissue sampling results were reported as malignant, suspicious for malignancy, atypical, or normal. Analysis was by intention-to-treat. Results: Two patients (3%) were excluded from analysis because the stricture could not be traversed with the intraductal US probe. Of the 60 remaining patients (37 men, mean age 64 years, range 27-89 years), 31 had malignant strictures (12 cholangiocarcinoma, 8 pancreatic, 5 metastatic, 3 gallbladder, 3 papilla), and 29 had benign strictures. Sphincterotomy was not required in any case to facilitate intraductal US. Fourteen patients (23%) underwent subsequent surgery including 11 with a preoperative diagnosis of resectable tumor. Endoscopic retrograde cholangiography/tissue sampling (atypia considered equivalent to benign) correctly identified 15 of 31 malignant strictures (p = 0.001) and all 29 benign strictures (p = 0.16) (accuracy 73%, sensitivity 48%, specificity 100%). The addition of intraductal US correctly identified 28 of 31 malignant strictures and 27 of 29 benign strictures (accuracy 92%, sensitivity 90%, specificity 93%). Of 11 patients with tumors who came to surgery, intraductal US correctly staged 4 (36%), understaged 5 (45%), missing metastatic lymph nodes in all cases and vascular invasion in 1 patient, and overstaged 2 (18%), with a false-positive diagnosis of metastatic lymph nodes in 1 and tumor mass in 1 patient who had no cancer at surgery. Conclusion: Technically easy, intraductal US is a valuable adjunct to endoscopic retrograde cholangiography/tissue sampling that increases the ability to distinguish malignant from benign strictures. Intraductal US is unsuitable for assessing lymph nodes associated with malignant strictures. (Gastrointest Endosc 2002;56:681-7.)