Background: Patients with failed endoscopic retrograde cholangiopancreatography (ERCP) are conventionally offered percutaneous transhepatic biliary drainage (PTBD). While PTBD is effective, it is associated with catheter-related complications, pain, and poor quality of life. Endoscopic ultrasound-guided biliary drainage (EUS-BD) is a minimally invasive endoscopic option increasingly offered as an alternative to PTBD. We compare outcomes of EUS-BD and PTBD in patients with biliary obstruction at a single tertiary care center.
Methods: A retrospective review was performed in patients with biliary obstruction who underwent EUS-BD or PTBD after failed ERCP from June 2010 through December 2014 at a single tertiary care center. Patient demographics, procedural data, and clinical outcomes were documented for each group. The aim was to compare efficacy and safety of EUS-BD and PTBD and evaluate predictors of success.
Results: A total of 60 patients were included (mean age 67.5 years, 65 % male). Forty-seven underwent EUS-BD, and thirteen underwent PTBD. Technical success rates of PTBD and EUS-BD were similar (91.6 vs. 93.3 %, p = 1.0). PTBD patients underwent significantly more re-interventions than EUS-BD patients (mean 4.9 versus 1.3, p < 0.0001), had more late (>24-h) adverse events (53.8 % vs. 6.6 %, p = 0.001) and experienced more pain (4.1 vs. 1.9, p = 0.016) post-procedure. In univariate analysis, clinical success was lower in the PTBD group (25 vs. 62.2 %, p = 0.03). In multivariable logistic regression analysis, EUS-BD was the sole predictor of clinical success and long-term resolution (OR 21.8, p = 0.009).
Conclusion: Despite similar technical success rates compared to PTBD, EUS-BD results in a lower need for re-intervention, decreased rate of late adverse events, and lower pain scores, and is the sole predictor for clinical success and long-term resolution. EUS-BD should be the treatment of choice after a failed ERCP.
Keywords: Biliary drainage; Endoscopic retrograde cholangiopancreatography; Endoscopic ultrasound.