Differentiating malignant from benign bile duct strictures is a conundrum, since no diagnostic test is highly sensitive for diagnosing cancer. While ERCP is effective in palliating obstructive jaundice, standard diagnostic tools in ERCP have a low diagnostic sensitivity and confirm the stricture's etiology in <50% of cases. During the first ERCP, standard practice is to obtain routine cytology (RC) using a single brush sample. If this is not diagnostic, patients often undergo repeat ERCP, endoscopic ultrasound or other, increasing health care costs. The incremental yield of performing alternate ERCP-based diagnostic tools during the first ERCP including fluorescence in situ hybridization (FISH), cholangioscopy w/biopsy and multiple brushes for routine cytology is currently unknown. There are no studies quantifying the amount of testing utilized to firmly diagnose the etiology of the stricture, or the most efficient combination of diagnostic tools during the first ERCP. These are important knowledge deficiencies since a definitive tissue diagnosis during the first ERCP could reduce the need for downstream tests and expedite treatment, thereby improving patient-centered and economic outcomes. The added costs of using multiple tools during the first ERCP may be offset by these benefits.

Among patients with indeterminate bile duct strictures, the investigators hypothesize that a multimodality approach will be more sensitive without a significant reduction in specificity compared to multiple brush samples for routine cytology. The investigators will test this hypothesis using an experimental trial design by randomizing patients during their first ERCP to multiple brushing samples for cytology vs. a single brush sample for cytology + FISH + cholangioscopy w/biopsy. To obtain preliminary data for a definitive multi-center trial, the investigators propose a pilot and feasibility study to compare the performance characteristics of each approach by evaluating the prospective clinical course, including treatment delay, quality of life, and life expectancy for each enrolled patient. If our hypothesis is validated in a subsequent definitive study, the standard approach to tissue sampling during the first ERCP may be altered.

Inclusion Criteria:

• Extrahepatic BDS with no discrete mass on CT/MRI (either or both)
• A BDS is defined as a segmental narrowing of the bile duct > 50% of the proximal or distal unaffected duct.
• Biochemical evidence of cholestasis (increase in alkaline phosphatase ≥ 2x upper limit of normal ± total bilirubin ≥2.0mg/dL)

Exclusion Criteria:

• No clinical suspicion for malignancy
• Associated mass seen on CT or MRI
• Age ≤18, pregnancy, incarceration, inability to give informed consent
• Inability to undergo standard ERCP (e.g., postsurgical anatomy)
• Previous ERCP with sampling of BDS, other than a single brushing specimen sent for routine cytopathology