Biliary strictures present a diagnostic and therapeutic challenge to clinicians due to unsatisfied accuracy of sampling modality. The major problem is very difficult to discern malignant from non-malignant strictures, such as patients with primary sclerosing cholangitis (PSC). With the poor prognosis and high mortality rate of advanced stage of hepatopancreaticobiliary malignancies, early and accurate diagnosis impacts patients' outcome and possible surgical candidacy. Therefore, a pre-operative determination of malignancy to help plan appropriate treatment is highly desirable.

Before 2000s, several diagnostic modalities, including laboratory tests, ultrasonography (US), computed tomography (CT) scan, cholangiography by percutaneous transhepatic cholangiography endoscopic (PTC) and endoscopic retrograde cholangiopancreatography (ERCP), and brushing cytology disclosed 13% to 24% false positive rate for suspicious malignant hilar strictures. Compared to recent studies, ERCP brushings still suffer from low sensitivity (41.6% ± 3.2% (99% CI)) and negative predictive value (58.0% ± 3.2% (99% CI)). In order to increase diagnostic accuracy, at least two sampling methods, including brushing cytology, biopsy, and fine-needle aspiration is therefore recommended. One article showed multimodal tissue-sampling (Brushing + Biopsy + Fine-needle aspiration) increased the sensitivity for diagnosis of malignant biliary stricture to 62%. However, no any literature demonstrate the best sequence of combined sampling modalities to yield the highest diagnostic accuracy. Besides, the role of stricture dilation before or after different tissue sampling modality is still uncertain.

In this study, the investigators want to compare stricture dilation before or after multimodal tissue-sampling, including brush cytology, intraductal suction and forceps biopsy for the diagnosis of malignant biliary stricture and also assess which kind of the sequence of combined tissue-sampling modalities could offer the highest diagnostic accuracy.

Inclusion Criteria:

• Clinically suspicious biliary stricture that required tissue sampling as medically indicated were considered for the study

Exclusion Criteria:

• Biliary stricture caused by extra-luminal compression, such as pancreatic cancer and lymphadenopathy
• Contraindication for ERCP study
• Age younger than 20 years
• Prior tissue sampling had yielded a diagnosis of malignancy
• A guidewire could not be passed through the stricture
• Less than 6-month follow-up was available for patients with negative tissue sampling