A Comparison of Sampling Techniques of Suspected Malignant Biliary Strictures

Abstract

BACKGROUND: It remains difficult to determine whether biliary strictures are malignant or benign. Tissue sampling by brush cytology and forceps biopsy during Endoscopic Retrograde Cholangiopancreatography (ERCP) has been the common practice. This study aims to compare sampling sensitivity of Fine Needle Aspiration (FNA) by Endoscopic Ultrasound (EUS) of a suspected malignant lesion with the yield from sampling strictures at ERCP. METHODS: In this retrospective study, 147 patients with indeterminate bile duct strictures, who underwent ERCP and tissue sampling and/or EUS with FNA of the suspected malignant lesion from September 2001 to November 2003, were evaluated. Final diagnoses of malignant strictures were confirmed histopathogically based on all sampling methods. Tissue specimens were reported as normal, suspicious, or malignant. Suspicious and malignant were considered positive samples. Comparisons were made between multimodal sampling techniques and bimodal or unimodal methods, and by neoplasm type. RESULTS: Of the 147 patients, 121 had cancer proven. Sixty-seven percent of the patients with cancer had positive tissue sampling by at least one of the three techniques. Tissue sampling sensitivity varied by cancer type with the highest yield seen in pancreatic cancer (79%) compared with cholangiocarcinoma (56%) and other cancer types (41%). FNA had the greatest yield: positive in 81% of malignancies compared with 42% by forceps biopsy and 33% by brushings. FNA also was most sensitive in pancreatic cancer: 89% vs. 50% in cholangiocarcinomas. Finally, there was no improved yield seen in triple-modal sampling with FNA, brushing, and biopsy, compared with FNA and brushing, brushing and biopsy, FNA alone, or brushing alone: 71%, 86%, 60%, 92%, 29%, respectively, in proven cancers. CONCLUSIONS: Tissue sampling sensitivity by fine needle aspiration of the suspected lesion is far superior to sampling of suspected malignant strictures by brush cytology or biopsy at ERCP. The combination of three methods of sampling does not increase yield compared with FNA alone. In addition, tissue sampling sensitivity varied by type of neoplasm. Therefore, in patients with biliary strictures, fine needle aspiration of the suspected malignancy itself is recommended.