Guidewire-bile duct angle as a determinant of stent delivery success in EUS-guided hepaticogastrostomy: a prospective double-lumen dilator cohort with propensity score-matched historical controls.

Abstract

Background: Failure of stent delivery remains a major technical limitation of endoscopic ultrasound-guided hepaticogastrostomy (EUS-HGS), often necessitating additional tract dilation and increasing procedural complexity and cost.

Objectives: The aim of this study was to evaluate whether the guidewire-bile duct angle is associated with stent delivery success in EUS-HGS, and to explore whether a double-lumen dilator (DLD) is associated with procedural improvement consistent with an angle-mediated mechanism rather than direct mechanical advantage.

Design: Prospective observational study with a propensity score-matched historical control.

Methods: This was a prospective single-arm observational study of patients undergoing EUS-HGS using a DLD, compared with a propensity score-matched historical (retrospective) cohort that underwent conventional dilation. The main outcome was successful stent delivery without additional tract dilation.

Results: A total of 25 patients in the DLD group were compared with 25 matched controls. Successful stent delivery without additional dilation was significantly higher in the DLD group than in the conventional group (92% vs 48%, p = 0.001). The guidewire-bile duct angle measured immediately before stent delivery-system insertion was significantly larger in the DLD group than in the conventional group (153.4° vs 123.3°, p < 0.001). On multivariable analysis, a larger guidewire-bile duct angle remained independently associated with successful stent delivery without additional tract dilation (odds ratio per 10-degree increase, 5.49; 95% confidence interval, 2.79-10.83; p < 0.001), whereas DLD use itself was not independently associated with success after adjustment for angle. Device-related costs were lower in the DLD group due to reduced need for additional dilation. The guidewire-bile duct angle showed excellent apparent discrimination in this cohort, with an exploratory threshold of 126° (AUC 0.969), requiring external validation before clinical implementation.

Conclusion: In conclusion, the improvement associated with DLD use was consistent with an angle-mediated mechanism rather than a direct effect of dilation force; these findings should be regarded as hypothesis-generating mechanistic evidence rather than mechanistic proof.