Endoscopic (ERCP) management of malignant biliary obstruction — from imaging workup to stent patency surveillance
Painless jaundice with a dilated biliary tree is the classic presentation of malignant biliary obstruction. Cross-sectional imaging is obtained before any endoscopic intervention to identify the cause, localize the stricture level, and stage the tumor for resectability — all of which determine whether the patient goes to surgery, chemotherapy, or endoscopic palliation with a stent.
Malignant biliary obstruction arises when a tumor compresses or invades the bile duct, blocking bile flow into the duodenum. The differential is dominated by four entities:
• Pancreatic head adenocarcinoma — the most common cause (~60–70% of cases). The tumor arises adjacent to the intrapancreatic common bile duct (CBD) and typically produces a smooth-to-irregular distal CBD stricture, often with the classic "double-duct sign" (simultaneous dilation of the pancreatic duct and CBD) on cross-sectional imaging.
• Cholangiocarcinoma — arises from the bile duct epithelium itself. Perihilar (Klatskin) tumors at the confluence of the right and left hepatic ducts are the most common subtype and are notoriously difficult to resect; distal CBD cholangiocarcinoma behaves more like pancreatic cancer in terms of stent strategy.
• Ampullary carcinoma — arises at the ampulla of Vater where the CBD and pancreatic duct join the duodenum. Often presents earlier (smaller tumors causing obstruction sooner) and carries a better prognosis than pancreatic or cholangiocarcinoma.
• Metastatic lymphadenopathy — porta hepatis nodal metastases (from gastric, colorectal, or other primaries) can extrinsically compress the CBD without a primary biliary or pancreatic tumor.
CT pancreas protocol (dual-phase, thin-slice) and MRCP are complementary and usually both obtained:
• CT pancreas protocol: arterial and portal-venous phase imaging defines tumor size, delineates the interface between tumor and the superior mesenteric artery (SMA), superior mesenteric vein (SMV), portal vein, and celiac axis. Greater than 180° of contact/encasement of the SMA or celiac axis, or unreconstructible venous occlusion, generally defines unresectable disease.
• MRCP (magnetic resonance cholangiopancreatography): a non-invasive, contrast-free "roadmap" of the entire biliary tree. It defines the stricture level (proximal/hilar vs. mid-duct vs. distal), the length of the stricture, and the presence of intrahepatic duct dilation — information that directly determines whether an uncovered or covered stent will later be selected and whether unilateral or bilateral hepatic duct drainage is needed for hilar lesions.
• Laboratory workup: total and direct bilirubin quantify the degree of cholestasis; CA 19-9, while non-specific (elevated in benign cholestasis too, and absent in the ~10% of patients who are Lewis-antigen negative and cannot produce it), is used for post-treatment trend-following rather than diagnosis alone.
The distinction between a resectable and locally advanced/unresectable tumor is decided at this stage — before ERCP — because it changes the entire stenting strategy: a potentially resectable patient should generally receive a removable plastic or covered stent rather than an uncovered SEMS, to avoid compromising the surgical field.
Endoscopic retrograde cholangiopancreatography (ERCP) is performed with a side-viewing duodenoscope. The major papilla is identified, the bile duct is selectively cannulated, and contrast injection produces a cholangiogram that confirms the imaging findings and reveals the classic radiographic signature of a malignant stricture before any stent is placed.
The duodenoscope is advanced to the second portion of the duodenum and positioned en-face to the major papilla. A sphincterotome loaded with a guidewire is used to selectively cannulate the common bile duct — distinguishing it from the pancreatic duct orifice is the technical crux of the procedure and the main source of post-ERCP pancreatitis risk when the pancreatic duct is inadvertently and repeatedly cannulated.
Once the wire is in the duct, contrast is injected under fluoroscopy to generate a cholangiogram. Malignant strictures characteristically appear as:
• Irregular, asymmetric narrowing with "shouldering" (abrupt transition rather than gradual taper) • A "rat-tail" appearance — a long, tapering, eccentric narrowing • Abrupt cutoff with proximal ductal dilation often more than double the normal 6mm caliber
This contrasts with benign strictures (e.g., chronic pancreatitis, post-surgical), which tend to be smooth, symmetric, and gradually tapered — though overlap exists and tissue sampling remains essential.
A hydrophilic guidewire is advanced across the stricture under fluoroscopic guidance — often the most technically demanding step, particularly for tight or angulated hilar strictures. Once the wire has traversed the lesion, tissue diagnosis is typically pursued before or concurrent with stenting:
• Brush cytology: a cytology brush is passed over the wire and repeatedly advanced/withdrawn across the stricture to exfoliate cells, which are then smeared for cytopathology. Sensitivity alone is only ~45%, reflecting the desmoplastic, hypocellular nature of pancreaticobiliary malignancy.
• Forceps biopsy: intraductal biopsy forceps can obtain a tissue core with somewhat higher yield than brushing for cholangiocarcinoma.
• Fluorescence in situ hybridization (FISH): applied to brushing specimens to detect chromosomal polysomy, raising combined sensitivity to roughly 60% when added to routine cytology.
A tissue diagnosis is not mandatory before palliative stenting in a patient with a classic malignant-appearing stricture and typical imaging, but it is essential if neoadjuvant chemotherapy or enrollment in a clinical trial is being considered.
Once the stricture is crossed, the interventionalist must choose between a plastic biliary stent and a self-expanding metal stent (SEMS). This is not a trivial preference — it is a guideline-driven decision (ASGE, ESGE) balancing expected patient survival, resectability, stricture location, and whether a tissue diagnosis is still pending.
The single most important variable driving stent choice is expected patient survival, because it determines how many times a stent will need to be exchanged before the patient's disease course is complete:
• Expected survival >3 months: SEMS is preferred. Plastic stents reliably occlude by ~3 months from biofilm and sludge deposition, so a patient likely to live longer than that would face at least one, and often several, repeat ERCPs for stent exchange if managed with plastic. A SEMS placed once can remain patent for a median of 6–8 months, meaningfully reducing the number of procedures, sedation exposures, and cholangitis episodes over the remaining disease course.
• Expected survival ≤3 months, or survival highly uncertain: plastic is often reasonable, particularly when procedural cost and simplicity matter more than long-term patency, or when initial decompression is being used as a bridge while staging and prognosis are still being finalized.
• Potentially resectable disease / awaiting surgery: a plastic stent (or a fully covered/removable SEMS) is favored so as not to place a permanent, tissue-embedded prosthesis into a duct that will shortly be resected and reconstructed. An uncovered SEMS becomes embedded in the duct wall by tissue ingrowth within weeks, which can complicate the surgical dissection and biliary anastomosis.
• Tissue diagnosis still pending: some endoscopists prefer plastic as a temporizing measure until brush cytology/FISH results return and the definitive treatment plan (surgery vs. palliation) is confirmed.
This slider-driven simulation applies the ASGE/ESGE rule directly: set "Expected Survival" above 3 months and the simulator recommends a SEMS; at or below 3 months it recommends plastic — mirroring the actual bedside decision algorithm used in ERCP practice.
When a metal stent is chosen, a second decision follows — covered or uncovered — and this is driven largely by stricture location:
• Distal CBD strictures (pancreatic head, distal cholangiocarcinoma, ampullary): a covered (or partially covered) SEMS is commonly used. Distal placement carries a lower migration risk than historically feared with modern flared-end designs, and covering the mesh prevents tumor ingrowth, keeping the option of removability open should the treatment plan change.
• Hilar (perihilar/Klatghtin) strictures: uncovered SEMS is generally preferred. A covered stent placed across the hepatic duct confluence risks occluding the contralateral hepatic duct or the cystic duct takeoff, precipitating contralateral cholangitis or cholecystitis. Uncovered mesh allows bile from side branches to flow through the interstices of the stent while still relieving the dominant stricture.
• Stricture length (the second simulator slider) determines the required stent length: the deployed stent must span the entire stricture with roughly 1–2cm of healthy duct margin proximally and distally to prevent early tumor overgrowth at the stent ends and to reduce the chance of the stricture being incompletely covered.
With the stricture crossed and the stent choice finalized, the device is advanced over the indwelling guidewire and deployed under fluoroscopic guidance. Correct positioning — full coverage of the stricture with adequate margins — is the technical determinant of both immediate success and long-term patency.
Deployment proceeds methodically over the wire that was left in place after cannulation and tissue sampling:
1. A delivery catheter (for plastic) or constrained delivery system (for SEMS, pre-loaded with the compressed nitinol mesh) is advanced over the guidewire under fluoroscopic visualization, using the previously obtained cholangiogram and radiopaque markers on the stricture (from the earlier contrast injection or endoscopic clips) as landmarks.
2. The device is positioned so its mid-portion straddles the tightest point of the stricture, with the proximal and distal ends extending 1–2cm into grossly normal, dilated duct on the upstream side and normal-caliber duct on the downstream side.
3. Plastic stents are pushed into final position with a pusher catheter; the pigtail or flap ends anchor the stent and resist migration.
4. SEMS are released from their constraining sheath — the nitinol mesh self-expands to its nominal 8–10mm diameter over several minutes to hours, exerting continuous radial force that further opens the stricture even after deployment.
5. Contrast is re-injected to confirm free flow of contrast (and subsequently bile) through the stent into the duodenum, with prompt drainage of the proximal, previously dilated system.
Technical success — defined as stent deployment spanning the stricture with demonstrated bile/contrast flow — is achieved in approximately 90–95% of cases in experienced centers. Failure usually stems from an inability to cross a very tight or angulated stricture, in which case percutaneous transhepatic biliary drainage (PTBD) or EUS-guided biliary drainage are used as rescue strategies.
Immediate post-procedural monitoring watches for:
• Post-ERCP pancreatitis (~3–5% overall risk, higher with difficult cannulation or pancreatic duct instrumentation) • Cholangitis, if drainage is incomplete or if contrast is injected proximal to an undrained segment without adequate stenting • Bleeding from sphincterotomy, typically self-limited • Perforation, rare (<1%) but potentially serious, related to the sphincterotomy or wire manipulation
Clinical response is then tracked biochemically over the following one to two weeks — a falling bilirubin trend is the earliest and most reliable indicator that drainage is adequate.
Placing the stent is not the end of biliary palliation — it is the start of a surveillance period defined by the expected patency of the chosen device. Understanding how and when stents fail, and recognizing the clinical signs of occlusion early, is what allows durable palliation of malignant biliary obstruction for the remainder of a patient's disease course.
Both plastic and metal stents eventually fail, but by different mechanisms and on different timelines:
• Plastic stents occlude by biofilm formation: bacteria adhere to the polyethylene surface within days of placement, form a glycocalyx biofilm, and progressively deposit calcium bilirubinate and cholesterol sludge that narrows and eventually occludes the lumen. Median patency is approximately 3 months, which is why routine, scheduled stent exchange (rather than waiting for occlusion) is standard practice for patients expected to need a plastic stent for an extended period.
• SEMS occlude more slowly and by different mechanisms: tumor ingrowth through the mesh interstices (mainly uncovered SEMS), tumor overgrowth beyond the stent margins, sludge/biofilm within the lumen (mainly covered SEMS, which behave somewhat like plastic once covered), or rarely stent migration. Median patency is 6–8 months, roughly double to triple that of plastic — the central rationale for using SEMS in patients expected to survive beyond 3 months.
Occlusion presents clinically as recurrence of the original obstructive picture, sometimes complicated by infection:
• Recurrent jaundice and pruritus, with rising conjugated bilirubin on labs — the most common presentation, reflecting gradually reduced bile flow without infection • Cholangitis — ascending bacterial infection of an obstructed biliary system, classically presenting with Charcot's triad: fever, right-upper-quadrant pain, and jaundice. Reynolds' pentad (triad plus hypotension and altered mental status) signals septic shock and requires urgent decompression
Management of occlusion depends on stent type: a occluded plastic stent is removed and exchanged for a new plastic stent or upgraded to a SEMS if survival now appears longer than initially estimated. An occluded SEMS is generally managed by placing a second stent through the lumen of the first ("stent-in-stent," typically a plastic or a second SEMS) rather than attempting removal of an uncovered, tissue-embedded metal stent, which can be difficult or impossible.
Clinical decompression — defined as bilirubin trending toward normal — is achieved in roughly 75–85% of patients by two weeks post-stenting. Failure to see this trend should prompt re-evaluation for occlusion, stent migration, or an undrained segment (e.g., an isolated intrahepatic duct in hilar disease) rather than simply waiting longer.
For patients living long enough to reach or exceed the expected patency window of their stent, a structured surveillance plan is used rather than waiting for a clinical crisis:
• Liver function tests are checked periodically (e.g., every 4–8 weeks) to detect a rising bilirubin/alkaline phosphatase trend before overt jaundice or cholangitis develops • Patients and caregivers are counseled to seek urgent evaluation for fever, jaundice, or right-upper-quadrant pain (Charcot's triad) rather than waiting for a scheduled visit • For patients with a plastic stent and an evolving prognosis of longer survival than initially estimated, proactive exchange for a SEMS is reasonable rather than repeated plastic exchanges every ~3 months • Multidisciplinary reassessment (oncology, surgery, interventional radiology, gastroenterology) continues throughout, since a change in resectability status or a new systemic therapy response can change the stenting strategy at any point in the disease course