Efficacy of lumen-apposing metal stents for the management of benign gastrointestinal stricture: a systematic review and meta-analysis

Introduction

Benign gastrointestinal (GI) luminal strictures can occur as a result of peptic ulcers, caustic ingestion, chronic inflammation, radiation and surgical anastomosis [1]. Endoscopic dilatation using balloons or bougies has typically been used to treat benign GI strictures. Even with the inclusion of steroid injection, endoscopic dilatation therapy necessitates multiple sessions and is associated with stricture recurrence, apart from the risk of bleeding and perforation [2]. Fully covered self-expandable metal stents have gained popularity in refractory and recurrent strictures, and are used mostly in the esophagus. However, they have a high incidence of migration, even when fixed with clips or sutures [3,4].

Lumen-apposing, fully covered, self-expanding metal stents (LAMS) were first developed to facilitate endoscopic transluminal drainage and are commonly used to treat pancreatic fluid collections [5,6]. They subsequently received approval for bile duct drainage in failed endoscopic retrograde cholangiopancreatography, malignant distal biliary obstruction cases, and gallbladder drainage in non-surgical patients [7]. In practice, the application of LAMS has expanded beyond the indications provided in the device’s labeling, including enteric anastomosis, drainage of post-surgical collections, and benign GI luminal strictures [7,8]. The LAMS provides several benefits, such as 3 different diameters (10 mm, 15 mm, and 20 mm), a saddle-shaped design that provides an anchorage via its wide flanges, reducing migration risk, and a simple stepwise deployment technique that enables great technical success [7]. There are still limited data regarding the clinical performance of LAMS placement in benign GI strictures, with only a few case series and small-sized studies available. Hence, we conducted this meta-analysis to determine the efficacy, safety and outcomes of LAMS in treating various benign GI strictures.

Materials and methods

Results

Baseline study characteristics and quality assessment

A total of 2318 records were identified from the search strategy and 18 studies were included in the final analysis. Fig. 1 shows the PRISMA flowchart for the study selection and inclusion process. Table 1 shows the baseline characteristics of the included studies. The majority of the studies were from the USA, and only 2 studies were prospective [11,25]. Tables 2 and 3 summarize the clinical characteristics and clinical outcomes of the included studies, respectively. The proportion of patients with anastomotic strictures varied from 40-100%. A majority of the patients had failed prior endoscopic therapy, in the form of endoscopic balloon dilatation with or without steroid injection, fully-covered self-expanding metal stent insertion or stricturotomy. A 15×10-mm LAMS was the most commonly used stent, with median stent dwell time varying from 60-119 days. Our study quality assessment found that 11 were of medium quality [14-18,21,22,24,26-28] and 7 were of low quality [11-13,19,20,23,25] (Supplementary Table 1).

Figure 1 PRISMA flowchart for study identification and selection process

Table 1 Baseline characteristics of the included studies

Table 2 Clinical characteristics of the included studies

Table 3 Clinical outcomes of included studies

Adverse events (AEs) and risk of migration

The incidence of AEs was reported by 14 studies (n=483). The pooled incidence of AEs with LAMS for benign GI strictures was 13.5% (95%CI 8.6-18.5) (Supplementary Fig. 4). Table 4 summarizes the details of AEs along with their pooled incidence rates. Overall, 16 studies (n=511) reported on the outcome of LAMS migration. The pooled incidence of LAMS migration in benign GI strictures was 10.6% (95%CI 6.0-15.2) (Supplementary Fig. 5).

Table 4 Summary of adverse events associated with lumen apposing metal stents for benign luminal strictures

Publication bias and sensitivity analysis

Visual assessment of the funnel plots did not show any evidence of publication bias for the outcomes, except for short-term and long-term clinical success (Supplementary Fig. 7). The “leave-one-out” sensitivity analysis did not show any significant difference in the event rate of various outcomes. We separately analyzed the outcome of LAMS for anastomotic strictures, which did not show any significant difference in the pooled event rates (Table 5).

Table 5 Summary of findings for use of lumen apposing metal stents for benign luminal strictures

Discussion

Treating benign GI luminal strictures can be challenging using conventional endoscopic modalities, such as endoscopic balloon dilation, intralesional steroid injection, needle-knife incision, or a combination of these. Refractory benign GI strictures are now more frequently treated with stents like fully covered self-expandable metal stents, which are effective in dilating the stricture but have a high migration rate. In 2012, the FDA approved the first LAMS for treating pancreatic pseudocysts endoscopically [29]. In practical experience, around one third of patients undergo LAMS placement for off-label use in benign strictures due to the broader flange providing anchorage and reducing the risk of migration [30]. This meta-analysis provided an excellent opportunity to assess the technical and clinical outcomes and safety of LAMS in benign GI strictures.

This meta-analysis revealed a high technical success rate (99.9%) irrespective of stricture etiology, type, site in the GI tract and/or stent size. Simple deployment techniques and expert endoscopists placing the LAMS attributed to high technical success [7,31]. Overall, short-term and long-term clinical success rates were 93.9% and 72.8%, respectively. High clinical success rates can be attributed to the anchoring mechanism with a dumbbell-shaped design, a fully covered nature of the stent, and a larger diameter, producing apposition between the 2 lumens, thereby reducing the likelihood of stent migration, enabling a longer indwell duration and increasing patient tolerance [31,13]. In a previous meta-analysis on the outcome of endoscopic stenting in Crohn’s disease (CD)-related GI strictures (both anastomotic and non-anastomotic), the pooled clinical success rate was 71.3% [32]. In another meta-analysis by Mohan et al, LAMS demonstrated statistically better outcomes regarding clinical success and stent migration compared with SEMS [8].

Stent migration is a widely recognized limitation of conventional SEMS in treating benign GI strictures. In a previous systematic review and meta-analysis on the outcome of stent placement in refractory benign esophageal strictures, the migration rate was reported to be 28.6% (95%CI 21.9-37.1) [33]. The reported migration rate of SEMS in patients with CD-related stricture also remains as high as 32% [32]. The migration rate reported with LAMS in the present meta-analysis was 10.6% (95%CI 0.6-15.2). Hence, based on a high clinical efficacy and lower reported migration rate, LAMS may be a preferred option over SEMS for short-segment stricture. However, there is a need for further comparative studies before recommending one above the other.

Despite the high short-term clinical success, the overall reintervention rate with LAMS was as high as 23%. After the use of fully covered self-expandable metal stents for refractory benign esophageal strictures, the reintervention rate was 38.9% [33]. Tan et al [34] conducted a meta-analysis to evaluate the efficiency and safety of LAMS in treating benign strictures, including 6 studies and 144 patients, while the present analysis included a total of 18 studies with 527 patients. Tan et al reported high pooled technical success (98.3%) and clinical success (73.8%) rates, which were similar to our results. We also evaluated the pooled short-term clinical success and reintervention rates, which were not reported in the previous meta-analysis [34]. Tan et al reported an AE rate of 30.6% in their meta-analysis. Overall, AEs were seen in 13.5% of patients in the present study. Thus, with increasing expertise, the incidence of AEs with the use of LAMS is declining.

Our study showed high technical (99.9%) and short-term (94.6%) success rates, with few AEs (14.4%), when LAMS were used for anastomotic strictures. Although revisional surgery has a high clinical success rate for anastomotic strictures, it is associated with high postoperative morbidity and mortality [28]. Therefore, LAMS placement can be viewed as an alternative option after failed endoscopic dilatation in anastomotic strictures before surgery. Patients with recalcitrant strictures who are not surgical candidates, or who have had unsuccessful revisional surgery, can also be considered for LAMS placement.

The majority of the studies used an Axios stent with a short length of 10 mm. Hence, the utility of LAMS for longer strictures remains a matter of debate. Most studies have used the 15 mm diameter LAMS, while only a few have reported the use of 10-mm and 20-mm diameter Axios stents. The larger diameter stents have smaller flanges. While there are no data to suggest that larger diameter stents may have a higher risk of migration and reintervention than smaller diameter stents, further studies are required to compare the outcomes with different stent sizes for refractory strictures of the GI tract.

The strength of the present analysis was the absence of significant heterogeneity concerning a majority of the outcomes. Nevertheless, we recognized several limitations to the present analysis. First, there were no randomized control trials, and the majority of studies were retrospective. Second, details of the etiologies (malignant vs. benign) when resection and anastomosis were performed were unavailable for the majority of the studies. Third, we could not compare LAMS with other available therapies for managing benign GI strictures. Fourth, we could not perform a cost-effectiveness analysis of LAMS, because the relevant data were not available. Lastly, we could not analyze the optimal stent dwell time that would yield optimal long-term clinical success.

In conclusion, LAMS can be considered as an option for treating refractory benign GI strictures of short length. The present meta-analysis showed a high pooled technical and short-term clinical success rate, with an acceptable safety profile. However, the present level of evidence for the use of LAMS in benign GI strictures remains low. Further studies of the use of LAMS in patients with refractory strictures will allow us to determine the appropriate duration of stent therapy and help improve long-term outcomes.