The efficacy and safety of a duckbill‐type anti‐reflux metal stent as the initial metal stent for distal malignant biliary obstruction in unresectable pancreatic cancer

Abstract Background The usefulness of duckbill‐type anti‐reflux metal stent (DMS) in self‐expandable metal stent‐naïve pancreatic cancer (PC) patients has not been well‐studied. This study aimed to evaluate the efficacy and safety of DMS in such patients. Methods We analyzed consecutive patients with unresectable PC who received a covered metal stent (CMS) as the initial self‐expandable metal stent at our institution. Technical success, functional success, causes of recurrent biliary obstruction (RBO), time to RBO (TRBO), adverse events (AEs), and reintervention rates were compared between DMS and conventional CMS (c‐CMS). Results A total of 69 patients were included (DMS: 28, c‐CMS: 41). Technical success, functional success, and AEs were similar between groups. Tumor ingrowth was more common in the DMS group (18% vs. 0%, p = 0.009), while non‐occlusion cholangitis tended to be more common in the c‐CMS group (0% vs. 15%, p = 0.074). Median time to RBO was similar between groups (276 vs. 273 days, p = 0.915). The anti‐reflux valve of DMS was found torn in 56% of patients. Endoscopic reintervention was successful in all cases, despite failed stent removal in 88% of patients in the DMS group. Conclusions DMS was not associated with longer time to RBO compared to c‐CMS in self‐expandable metal stent‐naïve patients.


INTRODUCTION
Distal malignant biliary obstruction (MBO) is a common complication of pancreatic cancer (PC). Endoscopic biliary drainage using self -expandable metal stents Sludge formation and food impaction resulting from duodenobiliary reflux are major causes of recurrent biliary obstruction (RBO) after CMS placement. 4,5 Several types of anti-reflux metal stents (ARMSs) have been developed to prevent duodenobiliary reflux, with conflicting results. [6][7][8][9][10][11] Duckbill-type anti-reflux metal stent (DMS) is a novel fully covered laser-cut type ARMS. Several studies including ours have reported the efficacy and safety of this stent [12][13][14][15][16] and DMS was reported to be more effective than conventional CMS (c-CMS) when used in the preoperative setting 12 or reinterventions after CMS dysfunction. 15 However, the superiority of DMS over c-CMS when used as the initial SEMS for unresectable PC has not been elucidated.
We conducted this retrospective study to evaluate the efficacy and safety of DMS in comparison with c-CMS as the initial SEMS in patients with unresectable PC.

Patients
This was a retrospective study of consecutive patients with unresectable PC who received a CMS (DMS or c-CMS) as the initial SEMS for distal MBO at our institution between January 2020 and November 2021. Only patients who received a CMS across the papilla were included in this study. The following were excluded: (1) patients who had a history of biliary SEMS placement, (2) patients with SEMS placed above the papilla, (3) patients with surgically altered anatomy, and (4) patients with concomitant hilar biliary obstruction. The selection of the type of SEMS was mainly based on the time period in which a given patient underwent initial SEMS placement. In general, c-CMS was used between January 2020 and November 2020, while DMS was used between December 2020 and November 2021. Written informed consent for the procedure was obtained from every patient. This study was conducted in accordance with the Declaration of Helsinki and was approved by the ethics committee of our institution (Institutional Review Board number: 2021-GB-119).

DMS and c-CMS
The ARMS used in this study was a fully covered laser-cut type SEMS with a 12.5 mm duckbill-shaped anti-reflux valve (ARV) attached to the distal end (Kawasumi Duckbill Biliary Stent; Kawasumi Laboratories Inc., Tokyo, Japan; Figure 1). The ARV is usually closed to prevent duodenobiliary reflux but opens when the bile duct pressure increases. DMSs with diameters of 10 mm and lengths of 60 or 80 mm were used in all patients.

F I G U R E 1 The duckbill-type anti-reflux metal stent
The c-CMS used in this study was a fully covered braided type SEMS (HANAROSTENT Biliary; M.I.Tech, Seoul, Korea). All stents had diameters of 10 mm and lengths of 60, 70, or 80 mm.

Endoscopic interventions
All patients underwent endoscopic retrograde cholangiopancreatography (ERCP) using a therapeutic duodenoscope (JF260, TJF260, TJF-Q290V; Olympus Medical Systems, Tokyo, Japan) under conscious sedation. Endoscopic sphincterotomy was generally performed before SEMS placement and prophylactic rectal nonsteroidal anti-inflammatory drugs (NSAIDs) were used at the discretion of the endoscopist. Some of the patients had previously undergone plastic stent placement at the referring institution. A SEMS was deployed under fluoroscopic and endoscopic guidance in the first session when the diagnosis of PC had already been confirmed and cholangitis was not present, otherwise, an endoscopic nasobiliary drainage (ENBD) tube was placed in advance and a SEMS was deployed in the next session. Before December 2020, we performed ERCP first and subsequently performed endoscopic ultrasound-guided fine-needle aspiration as needed in obstructive jaundice cases without pathological confirmation of cancer. From December 2020 onward (when DMS was generally used), we switched to a strategy of performing endoscopic ultrasound-guided fine-needle aspiration before ERCP unless the patient's condition made it preferable to perform ERCP first. Achieving a pathological diagnosis of cancer before ERCP allowed for SEMS placement in a single session, instead of requiring biliary drainage with a plastic stent or ENBD before SEMS placement. The length of SEMS was selected based on cholangiographic findings, and the distal end of the metal part of DMS was placed 5-10 mm below the papilla to completely expose the ARV into the duodenum.

Outcome measures
Outcomes of SEMS placement were generally defined according to Tokyo Criteria 2014. 17 In this study, nonocclusion cholangitis was considered RBO when endoscopic biliary drainage was required to treat cholangitis, while it was considered an adverse event (AE) when it improved conservatively. The primary outcome was time to RBO (TRBO), which was defined as the time from stent placement until RBO occurrence. Patients who were lost to follow-up or alive at the end of the study period, underwent stent removal due to AEs, underwent conversion surgery, or died without RBO were treated as censored cases at the time of the last follow-up, stent removal, conversion surgery, or death, respectively. The secondary outcomes were technical success, functional success, causes of RBO, AEs, and overall survival (OS). Technical success was defined as the successful deployment of a SEMS at the intended location, while functional success was defined as a 50% decrease in, or normalization of, serum bilirubin level within 14 days after SEMS placement. When serum bilirubin level was normal at the time of SEMS placement due to prior biliary drainage, functional success was defined as no exacerbation of serum bilirubin level after SEMS placement. The severity of AEs was graded according to the American Society of Gastrointestinal Endoscopy lexicon guidelines. 18 The duodenal invasion was diagnosed based on the endoscopic findings at the time of SEMS placement. The amount of ascites was evaluated using the most recent computed tomography scan before SEMS placement and was categorized according to the Japanese Classification of Gastric Carcinoma 19 : none, ascites undetected by computed tomography;mild,ascites localized in only one area such as the pelvic cavity; moderate, ascites neither mild nor severe; and severe, ascites throughout the abdominal cavity.Follow-up data was confirmed until September 30, 2022.

Statistical analysis
Continuous and categorical variables are expressed as medians with ranges and absolute numbers with proportions, respectively. Categorical variables were compared using the Chi-square test or Fisher's exact test as appropriate, while continuous variables were compared using the Mann-Whitney U test. Time to RBO and OS were estimated using the Kaplan-Meier method and were compared using the log-rank test. The cumulative incidence of RBO was estimated using the competing risk analysis and was compared using Gray's test. 20 Stent removal due to AEs or conversion surgery and death without RBO were considered competing events. p-Values < 0.05 were considered statistically significant. All statistical analyses were carried out using the EZR software version 1.40. 21

Patient characteristics
A total of 69 patients were included in this study (28 and 41 patients in the DMS and c-CMS groups, respectively). The baseline and procedural characteristics of the two groups are summarized in Table 1.A significantly lower proportion of patients underwent biliary drainage before SEMS placement in the DMS group (64% vs. 95%,p = 0.002).Other baseline characteristics including duodenal invasion, concomitant duodenal metal stent, presence of moderate to severe ascites, and peritoneal dissemination were not different between the two groups. Stents used in each group were generally 8 cm in the DMS group and 7 cm in the c-CMS group. Endoscopic sphincterotomy was performed before SEMS placement in all patients, except for one patient in the c-CMS group who underwent endoscopic papillary balloon dilation because of a previously deployed duodenal metal stent placed across the papilla. Prophylactic NSAIDs were administered more frequently in the DMS group (75% vs. 29%, p < 0.001). Table 2 summarizes the outcomes of each SEMS. Technical and functional success rates were not different between the two groups.

Outcomes of SEMS
AEs occurred in four patients in the DMS group (moderate pancreatitis: 3, moderate pancreatitis and cholecystitis: 1) and three patients in the c-CMS group (moderate pancreatitis: 2, moderate non-occlusion cholangitis: 1; 14% vs. 7%, p = 0.430). Of the five patients who developed pancreatitis, four patients underwent SEMS removal on the following day (two patients in each group).
Overall RBO rates were not significantly different between the two groups (43% vs. 46%, p = 0.810). Tumor ingrowth was more frequently observed in the DMS group (18% vs.0%,p = 0.009),while non-occlusion cholangitis tended to be more frequent in the c-CMS group (0% vs. 15%, p = 0.074). Of the five patients who presented with tumor ingrowth in the DMS group, tumor ingrowth was evident at a median follow-up period of 202 days.
Kaplan-Meier curves of OS and TRBO are shown in Figure 2. Both median OS (332 days vs. 365 days, p = 0.852) and TRBO (276 days vs. 273 days, p = 0.915) were similar between the two groups. The cumulative incidence of RBO was also not different between the two groups (hazard ratio 1.06, 95% confidence TA B L E 1 Baseline and procedural characteristics of patients who received duckbill-type anti-reflux metal stent (DMS) or conventional covered metal stent (c-CMS) as the initial self -expandable metal stent (SEMS) for distal malignant biliary obstruction  interval, 0.53-2.12, p = 0.860), even when accounting for competing risks (Figure 3).

Reintervention after RBO
Twelve patients in the DMS group experienced RBO, of which nine resulted from stent occlusion and three from complete distal stent migration.At the time of reintervention, the ARV was found torn in five of the nine patients (56%) with stent occlusion.Stent removal was attempted in eight patients, but was successful in only one patient (13%). Of the seven unsuccessful cases, tumor ingrowth was present in five and DMS was torn during the removal attempt in two. Biliary cannulation was achieved in all seven unsuccessful cases through the stent mesh (one patient) or the torn ARV of DMS (three patients), or by trimming the distal part of DMS (three patients) using argon plasma coagulation and/or a loop cutter (Olympus Medical Systems). Reintervention was successfully per-formed in all 12 patients. Seven patients whose stents could not be removed underwent CMS placement in a stent-in-stent method. Three patients underwent CMS replacement after successful stent removal or stent migration, one underwent plastic stent placement in a stent-in-stent method (stent removal of the initial SEMS was not attempted in this patient due to a giant duodenal ulcer), and one underwent plastic stent placement after successful stent removal. Nineteen patients in the c-CMS group experienced RBO, of which nine resulted from stent occlusion, four from stent migration, and six from non-occlusion cholangitis. Stent removal was successful in all attempted cases (stent removal was not attempted in one patient due to a combined type 2 biliary and duodenal obstruction). Reintervention was successfully performed in all 19 patients. Fourteen patients underwent CMS replacement, one underwent simultaneous duodenal stenting and endoscopic ultrasound-guided choledochoduodenostomy, one underwent endoscopic TA B L E 2 Outcomes of initial self -expandable metal stent (SEMS) placement for distal malignant biliary obstruction  ultrasound-guided hepaticogastrostomy, one underwent plastic stent placement, and two underwent endoscopic nasobiliary drainage.

DISCUSSION
This retrospective study evaluated the efficacy and safety of DMS in comparison with c-CMS for distal MBO in SEMS-naïve unresectable PC patients. DMS was not associated with longer TRBO compared to c-CMS (276 vs. 273 days, p = 0.915). Reasons for RBO differed between the two groups, with a higher rate of tumor ingrowth in the DMS group (18% vs. 0%, p = 0.009) and a tendency of a higher rate of non-occlusion cholangitis in the c-CMS group (0% vs. 15%, p = 0.074). Overall rates of AEs were similar between the two groups, with acute pancreatitis being the most common AE. The ARV of DMS was found torn in five of the nine patients (56%) who experienced RBO due to stent occlusion. Although stent removal of DMS was unsuccessful in most patients (88%) mainly as a result of tumor ingrowth, endoscopic reintervention was successful in every patient using the stent-in-stent method. Covered SEMS, which was developed to prevent stent occlusion due to tumor ingrowth, has been considered the standard first-line treatment for distal MBO in unresectable PC 3 . Sludge formation and stent migration remain the main causes of stent dysfunction of CMSs. 3 Development of stents with anti-migration properties (e.g., flared ends) 22,23 has contributed to the reduction of stent migration. Several braided-type ARMSs with various types of ARVs 6-11 were developed to prevent duodenobiliary reflux, in turn reducing the risks of sludge formation and food impaction. In a recent meta-analysis, braided-type ARMSs were reported to have a lower rate of stent occlusion but a higher rate of stent migration compared to conventional SEMSs, 24 highlighting the need for the development of ARMSs with effective anti-migration systems. The DMS used in this study is the first laser-cut type ARMS, which may have longer TRBO compared to braided-type ARMS due to the presumed lower risk of stent migration. 25 Indeed, the rate of DMS migration in this study (11%) was much lower than that of braided-type ARMS with a long funnel-type valve (31%) which was reported by Hamada et al. 11 Evidence to date suggests that DMS offers longer TRBO compared to c-CMS, especially when used in the preoperative setting 12 or reinterventions after CMS dysfunction. 15 However, evidence is limited regarding the efficacy of DMS in comparison with c-CMS in SEMS-naïve unresectable patients.
In the present study, DMS was not associated with longer TRBO compared to c-CMS in SEMS-naïve unresectable PC patients. Although the rate of stent migration was similar between the two groups (11% vs. 10%) and the rate of non-occlusion cholangitis was lower in the DMS group (0% vs. 15%, p = 0.074), the rate of stent occlusion was somewhat higher in the DMS group (32% vs. 22%, p = 0.408), resulting in similar overall RBO rates between the two groups. The negative results of our study could be due to several reasons including patient characteristics and the durability of the ARV and covered membrane. First, patients who would most benefit from ARMS are those who are at high risk of duodenobiliary reflux, including those with duodenal invasion and those who have indwelling duodenal stents. 26,27 The proportion of such patients was low in this study, indicating that only a few patients benefited from the use of ARMS. Second, the ARV of DMS was found to be torn in 56% of RBO cases, suggesting that the present ARV of DMS may not be durable enough to prevent duodenaobiliary reflux for a long period of time. Since a previous in vitro study showed that a duodenal pH environment caused a morphological change of ARV, 27 the development of a new ARV that is not affected by the duodenal pH environment might be necessary to improve the durability of the ARV. Third, the rate of tumor ingrowth was high in this study (18%) compared to previous studies using DMS. 12,13,15,28 This could be explained by the difference in the proportion of SEMS-naïve patients and the duration of the indwelling period of SEMS. 29 We speculated that the external force applied to the covered membrane of a laser-cut type SEMS may become strong enough to cause tumor ingrowth, especially when used in SEMS-naïve patients with a long indwelling time. Thus, modification of the design of the covered membrane might also be required to further prolong the TRBO of DMS, especially when used in SEMS-naïve PC patients.
Reintervention after RBO of the initial CMS is becoming an important issue, as the prognosis of PC patients has improved due to recent advances in chemotherapy.