Beyond Removability: Can Multi-Hole Covered Metal Stents Overcome the Remaining Challenges in Malignant Hilar Biliary Obstruction?

Abstract

Malignant hilar biliary obstruction (MHBO) remains one of the most technically demanding conditions in therapeutic pancreatobiliary endoscopy. Although bilateral drainage using uncovered self-expandable metal stents (UMSs) has conventionally been favored because of their prolonged patency and preservation of side-branch drainage, the non-removability of UMSs represents an increasingly important limitation. Improvements in systemic therapy have prolonged survival in patients with biliary tract and other malignancies, and recurrent biliary obstruction (RBO) after UMS placement is therefore encountered more frequently. Once the UMS becomes embedded in the bile duct wall due to tumor ingrowth or tissue hyperplasia, removal is generally impossible, and reintervention must be performed through or alongside the existing stent. This can restrict selective access to an undrained or newly obstructed segment, increase the complexity of repeated procedures, and leave a permanent metallic framework that may hinder subsequent modification of the drainage strategy or complicate conversion surgery. In this sense, removability is not simply a technical convenience: it preserves the option to reset the biliary anatomy at the time of reintervention.

Inoue et al. recently compared suprapapillary stent-by-stent deployment of 6-mm slim fully covered metal stents (FCMSs) with conventional UMS deployment for unresectable MHBO in a multicenter retrospective study [1]. After propensity score matching, 52 patients were included in each group. Technical and clinical success rates were comparable, whereas the FCMS group had a significantly lower incidence of RBO than the UMS group (32.0% vs. 60.8%) and a significantly longer time to RBO. Importantly, FCMSs could be removed in all patients undergoing reintervention. The reintervention time was shorter, and both the total number of reinterventions and the proportion of patients requiring multiple reinterventions were lower with FCMSs. These findings are important because they link removability not only to device retrieval but also to the cumulative burden of longitudinal biliary management.

The study therefore raises a broader question: should MHBO drainage primarily pursue “permanent drainage,” with the ideal goal of zero reintervention, or “removable drainage,” which accepts that reintervention may occur but seeks to make it predictable and technically straightforward? These aims should not be viewed as mutually exclusive or as a competition between devices. Rather, they represent different priorities along a patient's anticipated clinical course. Durable drainage with the lowest possible probability of another procedure may be the dominant objective in a patient with limited life expectancy or high procedural risk. By contrast, removability may be especially valuable in patients expected to receive prolonged systemic therapy, those likely to require repeated interventions, or those in whom conversion surgery remains a realistic possibility. The goal should therefore shift from identifying a universally superior stent to selecting a drainage strategy that retains sufficient patency while preserving the future options that are most relevant for the individual patient.

This perspective also clarifies the differences between plastic stents (PSs) and removable metal stents. Intraductal PSs can be electively exchanged or removed, with reliable removability demonstrated in patients with MHBO [2]. Their small caliber and relatively low radial force may reduce the consequences of branch coverage and make sequential reconstruction comparatively simple; however, they are more susceptible to clogging and generally require more frequent planned or on-demand exchange. In contrast, removable FCMSs provide a larger luminal diameter and have demonstrated longer stent patency than PSs in patients with MHBO [3]. The trade-off is that FCMSs exert greater radial force and their covering can obstruct side branches; migration, overgrowth at the stent ends, and tissue reaction can also complicate later removal. Thus, “removable” is not a binary property shared equally by PSs and FCMSs: PSs favor ease and repeatability of exchange, whereas FCMSs seek to combine removability with metal-stent caliber and durability.

What, then, determines good removability? First, the covering must remain mechanically intact and sufficiently impermeable to prevent tumor or hyperplastic tissue from fixing the stent to the duct wall. Covering material and architecture are therefore central, although comparative evidence specific to MHBO remains limited. Second, the stent should achieve enough radial force and anchoring to resist migration without creating excessive tissue reaction at uncovered ends or anchoring elements. Stent diameter may influence this balance: a slim FCMS can facilitate multistent deployment in the hilum, but diameter alone cannot guarantee removability. Third, proximal and distal positioning matters because an exposed edge at a bifurcation or within reactive tissue may become difficult to retrieve even when the body of the stent is fully covered. Finally, indwelling time is likely relevant. Data from non-malignant biliary strictures suggest that prolonged dwell time can increase FCMS removal failure through hyperplastic overgrowth [4], although the magnitude of this effect in malignant hilar disease is unknown. These considerations argue for evaluating removability as a time-dependent endpoint—successful atraumatic extraction after the clinically relevant dwell period—rather than as an intrinsic label attached to a stent at the time of deployment.

Suprapapillary placement may further improve the performance of slim FCMSs. Avoiding direct exposure of the stent lumen to duodenal contents could reduce duodenobiliary reflux and sludge formation, while the suprapapillary position may decrease distal migration into the duodenum. Nevertheless, the Inoue study also illustrates that removability does not eliminate the inherent disadvantages of covering the hilar biliary tree. Two patients in the FCMS group developed posterior-segment bilomas within 2 months, necessitating stent removal. In both cases, the proximal end of the right-sided FCMS was positioned at the bifurcation of the anterior and posterior sectoral ducts. Although a definite obstruction was not demonstrated, coverage of the posterior sectoral branch or disturbance of localized bile flow was strongly suspected. Thus, the proximal end of a covered stent may be as important as the stricture itself.

Branch occlusion can manifest not only as overt cholangitis but also as segmental cholangitis, liver abscess, or biloma. These events may be difficult to predict because apparently adequate contrast drainage at the end of the procedure does not necessarily guarantee sustained drainage after full stent expansion. Placing the proximal stent end precisely at a major bifurcation may be particularly hazardous. When both the anterior and posterior sectoral ducts remain patent, either complete bridging of the bifurcation or separate drainage using three stents may be preferable to leaving the covered stent edge at the bifurcation. Preprocedural assessment of the target liver volume and detailed interpretation of three-dimensional biliary anatomy are therefore essential. Preservation of segmental drainage is especially important when prolonged survival is anticipated because a small branch-related problem that might be clinically tolerable over weeks can become consequential over months of systemic therapy.

The branch-related adverse events observed with conventional slim FCMSs in the Inoue study provide a hypothesis for further stent design, rather than evidence for multi-hole fully covered self-expandable metal stents (MHSEMSs) themselves. Inoue et al. did not evaluate MHSEMSs. Conceptually, however, MHSEMSs incorporate multiple small holes into the covering membrane, with the aim of allowing bile to flow from adjacent branches while retaining many advantages of a covered stent. The holes may also permit limited tissue anchoring and thereby reduce migration. Successful suprapapillary stent-in-stent and side-by-side deployment of MHSEMSs for MHBO has been described, demonstrating technical feasibility while maintaining potential removability [5, 6]. A recent porcine hilar-obstruction model also suggested preservation of side-branch drainage and successful removability after MHSEMS placement [7]. These observations make the multi-hole concept attractive, but its clinical value in MHBO remains to be validated prospectively.

Clinical data from malignant distal biliary obstruction provide further proof of concept, while also illustrating the design trade-offs. In a multicenter retrospective study of 111 patients, the migration rate after MHSEMS placement was only 1.9%, and successful removal was achieved in 88.9% of attempted cases [8]. These findings suggest that a multi-hole design might address two important FCMS-related problems: side-branch obstruction and migration. In the hilar setting, communication through the holes might preserve drainage of a sectoral duct even when the stent edge or covered membrane overlaps a bifurcation. However, the same holes that permit branch drainage and tissue anchoring may allow tumor or hyperplastic tissue to grow into the stent. Indeed, ingrowth has been reported after MHSEMS placement for distal malignant biliary obstruction [8], and another comparative study in pancreatic cancer found a shorter time to RBO with MHSEMSs than with conventional FCMSs [9]. Reduced migration therefore does not necessarily translate into superior overall patency or easier late removal.

The ideal strategy for MHBO should therefore balance four longitudinal objectives: durable drainage, preservation of segmental bile flow, technical feasibility of multisegmental deployment, and the ability to revise or remove the stent when the clinical course changes. Conventional UMSs prioritize patency and branch preservation but sacrifice removability; PSs maximize exchangeability but usually at the cost of smaller caliber and more frequent reintervention; conventional FCMSs combine larger caliber with removability but may occlude branches or migrate; and MHSEMSs represent a rational attempt to reconcile these competing properties, although their benefits and trade-offs still require clinical validation. Future trials should compare these strategies using not only time to RBO but also successful removal at prespecified dwell periods, segment-specific adverse events, success of selective reintervention, cumulative number and duration of procedures, unplanned hospitalization, compatibility with subsequent surgery, and patient-reported burden. The next advance in MHBO drainage may depend less on whether a stent is simply covered or uncovered—or even removable or non-removable—than on how well the initial drainage plan anticipates the patient's likely oncological trajectory and preserves options for what comes next.

Hirofumi Kogure: conceptualization; literature review; writing – original draft; writing – review and editing. The author approved the final version of the manuscript.

The author has nothing to report.

The author declares no conflicts of interest.

This article is linked with T. Inoue et al. To view this article visit https://doi.org/10.1111/den.70031.

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.