Potential of 6-mm-diameter fully covered self-expandable metal stents for unresectable malignant distal biliary obstruction: a propensity score-matched study
Article information
Abstract
Background/Aims
To date, only thinner-diameter metal stents have been evaluated for unresectable malignant distal biliary obstruction (UR-MDBO). This study investigated the outcomes and optimal cohorts for a 6-mm-diameter fully covered self-expandable metal stent (FCSEMS) compared with those for a 10-mm-diameter FCSEMS.
Methods
This single-center retrospective cohort study included patients who underwent initial transpapillary metal stenting for UR-MDBO. Propensity score matching (1:1) analysis was performed.
Results
Of 133/68 patients who underwent 6-mm/10-mm-diameter FCSEMS deployment, 59 in each group were selected. The median time to recurrent biliary obstruction was not significantly different between the groups (p=0.46). In contrast, use of the 6-mm-diameter FCSEMS resulted in a significantly reduced incidence of stent-related adverse events (AEs) (p=0.016), especially cholecystitis (p=0.032), and patients aged <70 years were particularly affected by this significant reduction. Among the patients in the end-stage cohort who were unable to continue chemotherapy after FCSEMS deployment, the free rate of stent-related events, including recurrent biliary obstruction and stent-related AEs, was significantly higher in the 6-mm group (p=0.027).
Conclusions
For UR-MDBO, a 6-mm-diameter FCSEMS can be an optimal and safe option in the younger cohort with a relatively high risk of AEs and in the end-stage cohort requiring safer drainage without interference from stent-related events during times of poor prognosis.
INTRODUCTION
Endoscopic drainage of unresectable malignant distal biliary obstruction (UR-MDBO) improves not only obstructive jaundice and cholangitis but also quality of life, and enables palliative chemotherapy.1-3 Self-expandable metal stents (SEMSs), especially fully covered SEMSs (FCSEMSs), are used as standard treatment for UR-MDBO because their stent patency is longer than that of plastic stents (PSs).4-6 Recently, larger-diameter FCSEMSs, such as 12-mm and 14-mm-diameter FCSEMSs, were developed to extend stent patency in UR-MDBO.7-11
However, clinically, stent-related adverse events (AEs) often occur post-FCSEMS deployment, which is reportedly strongly associated with a high risk of pancreatitis and cholecystitis due to pancreatic and cystic duct orifice (CDO) compression.12-15 Therefore, the use of larger-diameter FCSEMSs will likely increase the incidence of such AEs due to duct orifice overloading. Regarding malignant biliary obstruction drainage, it is important not only to lengthen the time to recurrent biliary obstruction (TRBO), which reflects stent function, but also to develop a safe biliary stent with a low risk of stent-related AEs. Relatively early stent-related AEs can be fatal, especially in pre-terminal patients with low tolerance.16
A randomized controlled trial17 found that 8-mm-diameter FCSEMSs were non-inferior to 10-mm-diameter FCSEMSs for UR-MDBO in terms of TRBO (hazard ratio [HR], 0.90). However, the thinner diameter was not associated with a reduction in AEs, and 10-mm-diameter FCSEMS deployment remains the standard for UR-MDBO.3,18 Nonetheless, this result may support a paradigm shift in future stent development, indicating that not only larger-diameter SEMSs should be considered for UR-MDBO. Given that the deployment of PSs, which are thinner than SEMSs, reduces the incidence of pancreatitis and other AEs in MDBO but has a significantly higher rate of stent occlusion,19-21 the deployment of 6-mm-diameter FCSEMSs could be considered a hybrid drainage method that maintains stent patency because of SEMS-specific expansion and is safer because of their thinner diameter (similar to PSs).
A recent study reported the clinical outcomes of 6-mm-diameter FCSEMS deployment for MDBO, including resectable MDBO.22 However, although TRBO lengthening is required in UR-MDBO and is the development target of larger-diameter FCSEMSs, the impact of thinner stents (namely, 6-mm-diameter FCSEMSs) on TRBO and the safety in UR-MDBO remain unknown. Additionally, in the UR-MDBO cohort, investigations of patients who are more suitable for 6-mm-diameter FCSEMS deployment (i.e., the optimal cohort) are needed. Therefore, we aimed to compare the outcomes of the 6-mm-diameter FCSEMS with those of the more commonly used 10-mm-diameter FCSEMS for UR-MDBO in a propensity score-matched (PSM) cohort and to identify the optimal cohort for 6-mm-diameter FCSEMS deployment.
METHODS
Study patients
This single-center retrospective cohort study enrolled patients with MDBO (≥2 cm below the hepatic bifurcation, including early bifurcation of the right posterior branch) who underwent endoscopic retrograde cholangiopancreatography (ERCP) with initial transpapillary metal stent deployment (including cases in which an FCSEMS was deployed for initial drainage, after PS use, or endoscopic nasobiliary drainage) at the National Cancer Center Hospital between October 2017 and August 2022 (expansion cohort of a previous report22). The exclusion criteria were the deployment of an 8-mm-diameter FCSEMS, deployment above the papilla, concurrent FCSEMS and PS deployment, and surgical resection post-FCSEMS deployment.
Equipment and procedures
ERCP was performed using a therapeutic duodenoscope (JF260 and TJF260; Olympus Medical Systems). The FCSEMSs were deployed transpapillarily after endoscopic sphincterotomy with a moderate-sized incision in all cases. The selected stent was sufficiently long to pass through the stenotic site and deployed across the papilla under fluoroscopic guidance. A 10-mm-diameter stent (BONASTENT; Standard Sci-Tech) was used between October 2017 and August 2019, and a 6-mm-diameter stent (HANAROSTENT [Boston Scientific] and EGIS biliary stent [Sumitomo Bakelite]) were used after September 2019. Both the 6-mm and 10-mm-diameter stents were braided-type FCSEMSs (Fig. 1). The use of nonsteroidal anti-inflammatory drugs (NSAIDs) for the prevention of post-ERCP pancreatitis (PEP) was determined by physicians. Aggressive periprocedural hydration to prevent PEP was not performed in all patients.
Clinical outcomes and definitions
Clinical success was defined as a normalized or 50% decrease in total bilirubin level within 14 days. TRBO was defined as the time from the stent deployment to the occurrence of RBO. RBO was defined as re-dilatation of the intrahepatic bile duct accompanied by an increase in biliary enzyme levels. When bile duct dilatation was confirmed by imaging, endoscopy or percutaneous intervention was urgently performed for RBO due to stent occlusion and stent migration (excluding gallbladder drainage for cholecystitis). AEs post-FCSEMS deployment were distinguished from RBO and defined as stent-related AEs; these included pancreatitis, cholecystitis, non-occlusion cholangitis, bleeding, and perforation, per the Tokyo Criteria 2014.23 AEs were graded according to the American Society for Gastrointestinal Endoscopy lexicon.24 AEs were divided into early (occurring <30 days) and late (occurring ≥31 days) events.
We also analyzed stent-related events, defined as a combination of RBO and stent-related AEs.25,26 Pancreatitis and cholecystitis were diagnosed according to the Cotton criteria27 and the Tokyo Guidelines,28 respectively. Tumor involvement in the CDO and pancreaticobiliary maljunction was evaluated using computed tomography, magnetic resonance cholangiopancreatography, and ERCP. Tumor involvement in the duodenal papillary region was diagnosed by endoscopy. The end-stage cohort included patients who did not receive active therapy, such as chemotherapy or radiation, and only received supportive care for advanced cancer after FCSEMS deployment, either because of personal preferences or non-feasibility due to general status, terminal stage, or other reasons.
Statistical analyses
Continuous variables are expressed as medians (interquartile ranges) and categorical variables as numbers (percentages). Qualitative and quantitative differences between groups were evaluated using the χ2 or Fisher exact test for categorical parameters and the Mann-Whitney U test for continuous variables. TRBO and the time to stent migration, stent occlusion, and stent-related events were calculated as the time from FCSEMS deployment to each event (death was censored) using the Kaplan-Meier method and compared between groups using the log-rank test. Univariate and multivariate Cox proportional hazards regression analyses were performed to calculate the HRs and 95% confidence intervals (CIs) for potential factors of RBO. Additionally, binary logistic regression was used to calculate odds ratios (ORs) with 95% CIs for potential factors of stent-related AEs. All reported p-values were two-sided, and a p-value of <0.05 was considered statistically significant.
PSM was conducted to minimize bias between groups. The propensity score was estimated using multiple logistic regression analysis based on age, sex, primary disease, metastatic status, prior PS deployment, prior endoscopic sphincterotomy, prior cholecystectomy, total bilirubin level, amylase level, tumor invasion to the duodenal papilla, and tumor invasion to the CDO. Using caliper matching, pairs (6-mm and 10-mm FCSEMS groups) on the propensity score logit were matched 1:1 within a range of 0.2 standard deviations.29 Propensity score logit distributions are shown in Supplementary Fig. 1. The covariate balance was assessed using standardized differences; >10% of the absolute value was considered significantly imbalanced. Data were analyzed using IBM SPSS ver. 27.0 (IBM Corp.).
Ethical statements
This study was approved by our institutional review board of the National Cancer Center Hospital on February 26, 2018 (approval no. 2018-149). The requirement for informed consent was waived owing to the retrospective design of the study.
RESULTS
Entire cohort
Of the 352 patients who underwent initial metal stenting for MDBO at our hospital, 133 and 68 were enrolled in the 6-mm and 10-mm groups (before PSM), respectively, after applying the exclusion criteria (Supplementary Fig. 2).
The baseline and treatment characteristics of the cohort are shown in Tables 1 and 2, respectively. Clinical success (94.7% vs. 95.6%, p=1.0) and the median TRBO (HR, 0.97; 95% CI, 0.59–1.61, stratified by log-rank test: p=0.913) (Fig. 2A) were not significantly different between the groups. The incidence of stent-related AE was significantly lower in the 6-mm group than in the 10-mm group (12.0% vs. 30.9%, p=0.001), and each component event was lower in the 6-mm group without significant differences (Table 3).
Matched cohort
The PSM allocated 59 patients to each group (n=118). No significant differences in baseline characteristics were found between the two groups (Table 1). However, NSAID use for PEP prevention was more frequent in the 6-mm group than in the 10-mm group (86.4% vs. 67.8%, p=0.016), owing to clinical differences during the time gap, and the lengths of the 6-mm and 10-mm-diameter FCSEMSs were different (Table 2). The median TRBO was not significantly different between the 6-mm (287 days; 95% CI, 108–466 days) and 10-mm groups (286 days; 95% CI, 183–389 days) (HR, 1.28; 95% CI, 0.67–2.46; p=0.458), and the multivariate adjusted HR was 1.28 (95% CI, 0.66–2.48; p=0.463) (Fig. 2B, Table 4). The cumulative incidence of RBO due to stent occlusion was not significantly different between the groups (log-rank test, p=0.236), although the incidence due to stent migration was significantly higher in the 6-mm group than in the 10-mm group (log-rank test, p=0.003) (Supplementary Fig. 3A). No AEs due to migrated stents, such as perforations, were observed. However, RBO due to stent migration was not observed in the 6-mm group of the end-stage cohort (0/18 cases) (Supplementary Fig. 3B). The incidence of stent-related AEs was significantly lower in the 6-mm group than in the 10-mm group (13.6% vs. 32.2%, p=0.016), especially cholecystitis (1.7% vs. 13.6%, p=0.032). Pancreatitis and non-obstructive cholangitis were also less common in the 6-mm group, although the differences were not significant (p=0.717 and 1.00, respectively) (Fig. 3A). Multivariate analysis of stent-related AEs revealed age <70 years (OR, 5.20; 95% CI, 1.71–15.8; p=0.004) and guidewire to the pancreatic duct (OR, 3.41; 95% CI, 1.12–10.4; p=0.031) as risk factors, and 6-mm-diameter FCSEMS deployment (OR, 0.31; 95% CI, 0.11–0.86; p=0.024) was a risk suppressor compared with 10-mm-diameter FCSEMS (Table 5).
High-risk cohort
The area under the curve for age in predicting stent-related AEs was 0.73 (95% CI, 0.62–0.84). Patients were allocated to one of two age groups using an optimal cutoff value of 70 years (Supplementary Fig. 4). In the high-risk cohort of patients aged <70 years (67 cases), there was no significant difference in TRBO between the groups (HR, 0.81; 95% CI, 0.35–1.92; p=0.638). However, the stent-related AE rate was significantly lower in the 6-mm group than in the 10-mm group (18.2% vs. 47.1%, p=0.012), and the incidence of each component event was similar or lower in the 6-mm group, without significant differences (Fig. 3B). In contrast, in the cohort of patients aged ≥70 years (51 patients), the TRBO (HR, 1.93; 95% CI, 0.69–5.40; p=0.204) and stent-related AE incidences did not differ significantly between the groups (7.7% vs. 12.0%, p=0.668) (Fig. 3C).
Free rate of stent-related events
The crude free rates of stent-related events in the entire cohort were 69.8% and 65.1% at 90 days in the 6-mm and 10-mm groups, respectively (HR, 0.69; 95% CI, 0.45–1.04; p=0.072). Among the 47 patients in the end-stage cohort, the crude free rates of stent-related events at 30 days were 90.2% and 57.3% in the 6-mm and 10-mm groups, respectively (HR, 0.25; 95% CI, 0.07–0.81; p=0.013) (Supplementary Fig. 5).
In the matched cohort, the adjusted free rates of stent-related events were 59.2% and 64.8% at 90 days in the 6-mm and 10-mm groups, respectively (HR, 0.88; 95% CI, 0.52–1.50; p=0.632). Among the 32 patients in the end-stage cohort, the adjusted free rates of stent-related events at 30 days were 94.4% and 57.3% in the 6-mm and 10-mm groups, respectively. The free stent-related event rate was significantly higher in the 6-mm group than in the 10-mm group (HR, 0.19; 95% CI, 0.04–0.97; p=0.027) (Fig. 4).
DISCUSSION
Herein, we not only confirmed the outcomes and AEs of 6-mm and 10-mm-diameter FCSEMS deployment for UR-MDBO, but also investigated the optimal cohort for 6-mm-diameter FCSEMS deployment. The 6-mm-diameter FCSEMS deployment was associated with a slight increase in total RBO due to an increase in stent migration; however, the stent-related AE incidence was significantly lower with 6-mm-diameter FCSEMS deployment than with 10-mm-diameter FCSEMS deployment. In addition, a 6-mm-diameter FCSEMS was a potentially safer option, particularly for (1) the younger cohort with a relatively high risk of stent-related AEs and (2) the end-stage cohort with less tolerance. These new perspectives may lead to better outcomes for patients with UR-MDBO.
Owing to the recent trend toward the deployment of larger-diameter metal stents, there are few reports on the use of thinner-diameter stents, especially 6-mm-diameter SEMSs, for MDBO. Reportedly, stent patency at 3 months was significantly higher with 6-mm-diameter FCSEMSs (83.5%) than with PSs (45.3%) for preoperative MDBO (p=0.021).30 Another study found that the stent occlusion rate was significantly higher for 6-mm-diameter uncovered SEMSs than for 10-mm-diameter uncovered SEMSs (p=0.02).31 The stent occlusion rate was lower in this study than in previous studies because of the use of 6-mm-diameter “fully covered” SEMSs. Accordingly, there was no significant difference in the stent occlusion rate between the 6-mm and 10-mm groups; however, RBO due to stent migration was greater in the 6-mm group. This may be explained by the deployment of thinner FCSEMSs and chemotherapy-induced biliary stenosis relief.32 However, the HR for TRBO was only 1.28 for 6-mm-diameter FCSEMSs. Furthermore, the reduced incidence of stent-related AEs may outweigh the disadvantages associated with increased stent migration.
Risk factors for post-deployment pancreatitis include SEMS (vs. PS) deployment33 due to pressure on the pancreatic duct orifice by SEMS expansion.12,14,15 Thus, the pancreatitis risk is reduced if the pressure on the pancreatic duct orifice is reduced using thinner-diameter SEMSs.20 In this study, the incidence of pancreatitis was lower with 6-mm-diameter FCSEMS deployment than with 10-mm-diameter FCSEMS deployment in the entire and matched cohorts, although the differences were not significant. However, there was a significant difference in the use of NSAIDs even in the matching cohort, owing to clinical differences arising during the time gap in this comparison, emphasizing the importance of considering this factor when interpreting the results. Rather, the risk of cholecystitis was reduced the most by 6-mm-diameter FCSEMS deployment, despite the use of longer stents (leading to deployment across the CDO) in the 6-mm group. Previously reported risk factors for cholecystitis after SEMS deployment include pressure on the CDO by SEMS expansion.13,16,34 Thus, the etiology of the main stent-related AEs overlaps with overloading of the orifice of the two ducts due to stronger SEMS expansion, and 6-mm-diameter FCSEMS use, with less pressure on the two ducts, may reduce the risk of these two events. Consistent with this, the 6-mm-diameter FCSEMS deployment significantly reduced the overall incidence of stent-related AEs and was also identified as a risk suppressor in the multivariate analysis. Therefore, it is suggested to be a safer biliary stent for UR-MDBO.
Age <70 years and a guidewire to the pancreatic duct (cannulation method) were risk factors for stent-related AEs. We hypothesized that the reduced risk of stent-related AEs in the younger cohort could be attributed to several factors. Reportedly, the risk of developing PEP decreases with age due to factors such as decreased pancreatic exocrine function, and younger patients are at a higher risk of pancreatitis.35-37 Furthermore, the incidence of post-deployment cholecystitis was lower in the older cohort than in the younger cohort, likely because the risk of cholecystitis after stent deployment increased because of gallstone retention, which is common in middle-aged patients,16 and a history of cholecystectomy was more common in the older cohort. Additionally, another study reported a higher percentage of post-ERCP AEs in patients aged <80 years.36 These results suggest that younger patients are at a higher risk of stent-related AEs than older patients. Therefore, we further analyzed the usefulness of the 6-mm-diameter FCSEMS in this high-risk cohort (age <70 years) and found a larger risk reduction in stent-related AEs with the 6-mm-diameter FCSEMS deployment in the younger cohort than in the older cohort, whereas TRBO was not significantly different. Thus, patients aged <70 years may be the optimal cohort for 6-mm-diameter FCSEMS deployment, benefiting from the safety advantages of thinner diameters.
In the end-stage cohort, for whom the drainage period is limited and patients are less tolerant, safer drainage without interference from stent-related events during the time of poor prognosis is more important than prolonged TRBO. Therefore, we defined RBO- and stent-related AEs as stent-related events based on previous reports25,26 and analyzed the event-free rates in the end-stage cohort. The event-free rate was significantly higher in the 6-mm group than in the 10-mm group before and after PSM because 6-mm-diameter FCSEMS deployment reduced not only the risk of stent-related AEs but also that of RBO due to stent migration. The stent migration risk was likely reduced in the end-stage cohort because the patients did not undergo chemotherapy and consequent stenosis relief, which has been reported as a risk factor for stent migration.32 In fact, stent migration was not observed during the observation period in the 6-mm group of the end-stage cohort, although mortality was a competing risk. These results suggest that 6-mm-diameter FCSEMS deployment is the more optimal drainage method for the end-stage cohort, with both a low risk of stent migration, which is a disadvantage of 6-mm-diameter FCSEMS, and a low incidence of stent-related AEs, which are relatively early events. However, reducing stent migration in the entire cohort has been challenging, and the development of new stents such as 6-mm-diameter partially bare stents, double bare stents, and laser-cut covered stents is expected in the future.
This study had some limitations. First, the 10-mm group included patients historically observed from October 2017 to August 2019, resulting in clinical differences during this time gap. Therefore, we used PSM and adjusted the data to minimize bias. Although the stent length was different between the two groups in the matching cohort, its difference could not be completely adjusted only by an ≥8 cm or <8 cm length on the univariate analysis. Second, we analyzed the time to event (RBO, stent-related events, and migration) using the Kaplan-Meier method and Cox regression analysis. However, overassessment may have occurred because other events and death were competing risks. Therefore, we confirmed the same significant trend in all analyses using the Gray test and Fine-Gray hazard model. Third, the data of this analysis partly overlap with those of a previous report.22 Nevertheless, the safety profile of 6-mm-diameter FCSEMS for UR-MDBO needs to be confirmed using PSM analysis in terms of sensitivity analysis, and this study provides new perspectives on the optimal indications for UR-MDBO. Finally, this study was retrospective in nature, which raises the possibility of unmeasured confounders. Thus, a prospective randomized controlled trial comparing the 6-mm and 10-mm-diameter FCSEMSs is necessary.
In conclusion, we found that 6-mm-diameter FCSEMSs are safer than 10-mm-diameter FCSEMSs for UR-MDBO while maintaining stent patency. Patients aged <70 years and the end-stage cohort may be the optimal cohorts for 6-mm-diameter FCSEMS deployment. However, further studies are required to validate these results.
Supplementary Material
Supplementary materials related to this article can be found online at https://doi.org/10.5946/ce.2024.044.
Notes
Conflicts of Interest
The authors have no potential conflicts of interest.
Funding
This work was supported by The National Cancer Center Research and Development Fund (2022-A-16).
Acknowledgments
We would like to thank the members of the Endoscopy Team of the Department of Hepatobiliary and Pancreatic Oncology, National Cancer Center Hospital, for their support of this research.
Author Contributions
Conceptualization: DY, SH; Data curation: DY, SH; Formal analysis: DY; Funding acquisition: SH; Investigation: DY, SH; Methodology: DY, SH; Project administration: SH; Resources: SH, YN, YM; Software: DY; Supervision: SH, YS, TO; Validation: all authors; Visualization: DY; Writing–original draft: DY, SH; Writing–review & editing: all authors.