A pilot study of novel duodenal covered self-expandable metal stent fixation

Migration of duodenal covered self-expandable metal stents (C-SEMSs) is the main cause of stent dysfunction in patients with malignant gastric outlet obstruction (mGOO). Because endoscopic SEMS placement is frequently selected in patients with poor performance status, we concurrently focused on the safety of the treatment. This pilot study included 15 consecutive patients with mGOO who underwent duodenal partially covered SEMS (PC-SEMS) placement with fixation using an over-the-scope-clip (OTSC). Technical feasibility, clinical success for oral intake estimated by the Gastric Outlet Obstruction Scoring System (GOOSS) score, and adverse events including stent migration were retrospectively assessed. All procedures were successful, and clinical success was achieved in 86.7% (13/15). Mean GOOSS scores were improved from 0.07 to 2.53 after the procedure (P < 0.001). Median survival time was 84 days, and all patients were followed up until death. Stent migration occurred in one case (6.7%) at day 17, which was successfully treated by removal of the migrated PC-SEMS using an enteroscope. For fixation using an OTSC, additional time required for the procedure was 8.9 ± 4.1 min and we did not observe OTSC-associated adverse events. Poor performance status was associated with clinical success (P = 0.03), but we could provide the treatment safely and reduce mGOO symptoms even in patients with poor performance status. In conclusion, duodenal PC-SEMS fixation using an OTSC is feasible for preventing stent migration in patients with mGOO including those with poor performance status.

www.nature.com/scientificreports/ performance status who would be unable to tolerate surgical GJ may be an appropriate candidate for endoscopic SEMS placement. Clinicians should evaluate treatment options carefully to optimize care for patients. Endoscopic duodenal SEMS placement may be selected because it is minimally invasive. Uncovered SEMSs (U-SEMSs) and covered SEMSs (C-SEMSs) (including partially covered SEMSs [PC-SEMSs]) are available, and RCTs 6-10 have revealed their clinical features. The main cause of stent dysfunction with U-SEMSs is tumor ingrowth via the mesh of the stent, which occurs in 16-44% 6-9 of cases. Chemotherapy is reported as the only method to prevent tumor ingrowth 11,12 , but in actual clinical practice more than 70% of patients with mGOO are not eligible for chemotherapy due to disease progression 13 . C-SEMSs were designed to prevent tumor ingrowth, but the incidence of stent migration reportedly ranges from 6 to 32% [6][7][8][9][10][11][14][15][16][17][18][19][20][21][22][23][24] . Chemotherapy that results in a decrease of tumor volume and reduction of alimentary tract compression is recognized as a major predictive factor for stent migration 11 . Improvements in the shape of the stent have been attempted to prevent C-SEMS migration, but the results have been controversial and unsatisfactory.
Anchoring of C-SEMSs with devices has been proposed as another countermeasure against C-SEMS migration 25,26 . A recent systematic review 27 has reported that OTSC was used for stent fixation in only 20 patients. Although from the limited cases, the overall technical and clinical success rate was 100.0% and 80.0%, respectively. In the systematic review, the largest paper (n = 12) was presented by Mudumbi et al. 28 ; the major indication for endoscopic stenting with OTSC fixation was tracheo-esophageal fistula. We previously evaluated the safety and effectiveness of duodenal C-SEMS fixation using suturing and an over-the-scope-clip (OTSC) device (Ovesco Endoscopy, Tübingen, Germany) in an experimental model 26 . However, to the best of our knowledge, no clinical trial has been conducted using these methods for duodenal SEMS fixation. Endoscopic suturing device is not available for clinical use in our country; therefore, we conducted this pilot study to evaluate the technical feasibility and clinical effectiveness of duodenal PC-SEMS fixation using OTSC in patients with mGOO. Table 1 lists the demographic and clinical characteristics of all patients. Patients included 10 males (66.7%), and the median age of all patients was 78 years (range 43-96). Pancreatic cancer (n = 9, 60.0%) and gastric cancer (n = 3, 20.0%) were the two main etiologies. Fourteen patients (93.3%) had no oral intake (GOOSS score 0), and 3 patients (20.0%) were unable to care for themselves (Group C; KPS score [0 − 40]).

Results
Technical and clinical outcomes. Table 2 lists treatment outcomes and adverse events. The technical success rate was 100.0% (15/15), and successful application of the OTSC for fixation was accomplished in all patients. The mean procedure time required for OTSC placement was 8.9 ± 4.1 min, and no adverse events were related to the fixation method. The total median procedure time for PC-SEMS placement with fixation was 32.1 ± 8.5 min. Clinical success was achieved in 13 patients (86.7%), and mean GOOSS scores were improved from 0.07 to 2.53 after PC-SEMS placement with fixation (P < 0.001). Eight patients (53.3%) were able to have full diets. The clinical success rate as estimated by the GOOSS score was significantly associated with KPS status (Groups A and B, 100.0% [12/12]; Group C, 33.3% [1/3]; P = 0.03). Chemotherapy was performed after SEMS placement in 4 patients (26.7%) as appropriate. All selected regimens were S-1 monotherapy.  ) and all patients were followed up until death. All patients died of their underlying carcinoma. Figure 1 presents the cumulative stent patency curve estimated by Kaplan-Meier analysis.

Discussion
These results demonstrated that duodenal PC-SEMS placement and fixation with an OTSC for mGOO was successful in all cases and that 86.7% of the patients achieved clinical success. In particular, 53.3% of patients with mGOO could have full diets. The additional time required for the procedure was 8.9 ± 4.1 min, which may be within the permissible range. Moreover, no adverse events were related to the fixation method. Poor performance status was associated with clinical success as estimated by the GOOSS score (P = 0.03). Stent migration occurred in one case (6.7%) with no other adverse events. No asymptomatic stent migration was observed on periodic abdominal X-ray imaging. Stent migration is a major adverse event of duodenal C-SEMSs. Kim et al. 6 reported that stent migration was significantly associated with chemotherapy after stent placement. Isayama et al. 17 recommended the use of longer stents to prevent stent occlusion caused by tumor in-or overgrowth at the uncovered portion. According to this recommendation, in this pilot study, we used the longest PC-SEMS (120 mm) available in our country. We were fortunate to experience no cases of tumor in-or overgrowth. Once a duodenal C-SEMS has migrated into the jejunum, clinicians hope it will exit via the rectum or remain in the body without causing obstruction symptoms. Otherwise, with obstruction symptoms, surgical removal is required, and this exhausts patients with advanced cancer. In this study, we experienced one case with stent migration. The patient (case no.2) did not receive chemotherapy. Seventeen days after the procedure, patient was admitted to the emergency ward with abdominal distension and vomiting. As the position of migrated metal stent did not change, we decided to remove it using enteroscope (day 19). But the enteroscope could not pass the duodenal stenosis, we placed a second duodenal SEMS. The second SEMS was fully expanded at day 21, and we could pass the duodenal stenosis. Fortunately, we were able to remove the migrated PC-SEMS using an enteroscope without surgery 29 . It is important to accommodate asymptomatic migration, which is not included in stent dysfunction. Although this event fortuitously may not cause symptoms related to gastrointestinal obstruction, it potentially results in intestinal obstruction. Table 3 lists some published articles about endoscopic duodenal covered metal stenting for mGOO [6][7][8][9][10][11][14][15][16][17][18][19][20][21][22][23][24] . The literature review assessed more than 1000 cases of endoscopic duodenal C-SEMS stenting and found that chemotherapy was administered after C-SEMS placement in 41.2% (range 11.4-78.6%), and that stent migration occurred in 14.5% (range 6.0-32.3%) of cases. Not all of the clinical studies carried out periodic abdominal X-ray imaging or follow-up endoscopy, so the real rate of stent migration might be higher because asymptomatic stent migration could have been missed. Kim et al. 6 performed a routine 8-week follow-up endoscopy and found that 63% of cases with a distally migrating C-SEMS occurred in patients without any obstructive symptoms. Their prospective study confirmed stent migration in almost one-third of patients during total follow-up. Another important issue is that the rate of patients receiving chemotherapy varies in published articles. The chemotherapy regimen continues to change and evolve, and some prospective studies including RCTs excluded mGOO patients with poor performance status 10,17 . Patients with poor performance status tend to avoid receiving chemotherapy, so the rate of patients with mGOO receiving chemotherapy might be lower in real-world practice. We did not www.nature.com/scientificreports/ exclude patients with poor performance status, so 26.7% of patients in our study cohort received chemotherapy after the procedure. We found that poor performance status was associated with clinical success, defined as relief of GOOSS score ≥ 2 within 1 week. Although the result could be due to underlying disease, symptoms related to mGOO were relieved in all cases (at least 1 GOOSS score improvement). Endoscopic duodenal PC-SEMS placement with fixation might be beneficial even in patients with poor performance status. Previously, because there were no designated duodenal C-SEMSs, clinicians applied esophageal C-SEMSs for relief of mGOO symptoms 16 . Due to the bended anatomy of the duodenum, stent fracture and gastrointestinal perforation became a serious problem. Recently, SEMSs with low axial force 30 (i.e., those that follow the gastrointestinal tract) have been preferred and widely used. As shown in Figs. 2C and 3D, a SEMS with low axial force resulting from the devised knit pattern easily follows the gastrointestinal tract. Both ends of the uncovered portion were designed to stick to the tumor to prevent migration. Furthermore, the range of proximal flare was widened ( Fig. 2A). Even with these ingenious improvements, previous studies using the PC-SEMS (flared-ComVi stent) 10,22 could not prevent stent migration with complete satisfaction. In comparison, although our pilot study included a limited number of cases, we observed favorable results (23.1% and 12.1% [flared-ComVi stent only] vs 6.7% [flared-ComVi stent with OTSC fixation]). In previous attempts to prevent stent migration, Choi et al. 23 used a C-SEMS with large star-shaped flaps and Choe et al. 24 used a 40-mm diameter funnel shape. Those stents could prevent distal migration (no cases in both studies), but proximal migration occurred in 11.1% and 16.7% of cases, respectively. Thus, stent migration is still a challenging problem, but our fixation method that anchors C-SEMSs to the gastrointestinal tract may contribute to approaching the ideal endoscopic SEMS placement.
Not enough is known about the role of anchoring of duodenal C-SEMSs in preventing migration. One previous study 25 reported the usefulness of endoscopic clips generally used for closing perforations or controlling bleeding to prevent C-SEMS migration. We previously practiced this technique, but still encountered cases of C-SEMS migration, and even without stent migration the clips were often out of place as assessed by routinely performed abdominal X-rays. Therefore, we evaluated the feasibility and safety of anchoring methods using experimental models 26 . The OTSC and suturing system had a significantly higher gripping force compared with the clipping system (OTSC vs. clip: 13.  www.nature.com/scientificreports/ Based on pathological findings in a porcine model, OTSC compressed the submucosal layer but not the muscle layer, and we concluded that the method may be safe even for preventive use. The results revealed that SEMS fixation with an OTSC and suturing method is feasible compared with the clipping method. Because an endoscopic suturing device is not available for daily clinical use in our country, we conducted duodenal PC-SEMS fixation using OTSC in patients with mGOO. In this pilot study, we did not observe adverse events, such as gastrointestinal perforation, related to OTSC placement. We also confirmed that the position of PC-SEMSs and OTSCs was unchanged until the last follow-up in 93.3% of cases. Our technical point of the OTSC application is that we suction edge of the placed SEMS at the center part of the OTSC system. We recommend to grab normal mucosa at one side of the OTSC with enough suction. Further large-scale research using our fixation method will be required to establish higher quality evidence. The one problem with this fixation method is the additional cost (79,800 Japanese yen; approximately 770 US dollars) for an OTSC. Further studies should be planned considering medical costs. This study had some limitations. First, it was a pilot study with a limited number of cases, and was conducted at a single institution. Further large-scale multicenter prospective research will be required to verify the superiority of our method over other conventional treatments and assess medical costs. Second, we selected an OTSC for fixation to anchor the PC-SEMS; other methods such as an endoscopic suturing system should be evaluated. Furthermore, a novel specific clip fixation device (stentfix OTSC, Ovesco Endoscopy AG, Tübingen, Germany) has been introduced and reported 31 . Currently, these system and device are not available in our country; we hope it will be accepted as a daily clinical procedure in the future.
In conclusion, an OTSC used for fixation of duodenal PC-SEMSs was safe and feasible for preventing stent migration in patients with mGOO. The treatment may bring benefits even for patients with poor performance status. We believe our efforts will contribute to establishing an anti-migratory method for duodenal PC-SEMSs.

Methods
Patients. Eighteen patients with mGOO were assessed for eligibility at Nagoya City University Graduate School of Medical Sciences between October 2018 and April 2020. Three patients were declined to participate in the study. In total, 15 patients underwent duodenal PC-SEMS placement and fixation with an OTSC for mGOO ( Supplementary Fig. 1). Procedural and clinical data were collected and analyzed retrospectively from a prospectively maintained endoscopy database. The inclusion criteria were obstruction of the stomach or duodenum causing nausea, vomiting that reduced oral intake, or weight loss; and unresectable malignant pyloroduodenal Table 3. Migration rate of published endoscopic gastroduodenal covered stenting. C-SEMS covered selfexpandable metal stent, NA not available (or no details), OTSC over-the-scope-clip, RCT randomized controlled trial.

Author and reference
Year Study type Devices. All PC-SEMSs used in this study were flared-ComVi stents (Taewoong Medical, Seoul, Korea), 20 mm in diameter and 120 mm long. The stents had an uncovered flare (15 mm long) at both ends. The proximal uncovered flare was 25 mm in diameter and was expected to prevent stent migration ( Fig. 2A). The type of OTSC was 12/6 t, which has an OTSC cap diameter of 17.5 mm and clip width of 10 mm (Fig. 2B)  patients under deep sedation using midazolam (5-10 mg) and pethidine hydrochloride (17.5-35 mg). Duodenal stent placement was performed using a direct-viewing scope (CF-HQ290ZI; Olympus Medical Systems, Tokyo, Japan) or a side-viewing duodenoscope (TJF260V; Olympus Medical Systems, Tokyo, Japan). The endoscope was first positioned close to the gastric or duodenal stenosis site, and the GOO was evaluated endoscopically (Fig. 3A). Contrast medium was injected under fluoroscopic guidance to identify the site and length of the obstruction. The obstruction was negotiated using a 0.025-inch standard biliary guidewire and an endoscopic retrograde cholangiopancreatography (ERCP) catheter. After confirming the position of the stenosis with the ERCP catheter, we deployed the PC-SEMS and placed it under endoscopic and fluoroscopic guidance (Fig. 3B). Subsequently, the OTSC system for fixation was loaded onto the scope (PCF-Q260AI; Olympus Medical Systems, Tokyo, Japan), and part of the upper rim of the SEMS was suctioned into the transparent cap before releasing the OTSC, grasping both the SEMS and the gastric or duodenal wall. One OTSC was placed per patient (Fig. 3C,D). Data analysis and follow-up. The baseline information collected included age, gender, diagnosis, site of obstruction, Karnofsky performance status (KPS) score 32 , Gastric Outlet Obstruction Scoring System (GOOSS) 33 score, presence of ascites/liver metastasis/peritoneal dissemination, and chemotherapy after SEMS placement. The KPS is an assessment administered by a healthcare provider to assign a patient to one of 11 categories (ranging from 0 [dead] to 100 [normal activity, no evidence of disease]). According to the assessment results, patients are divided into three groups; Group A (80-100) can independently perform daily activities, Group B (50-70) can perform daily activities with help, and Group C (0-40) requires continuous assistance and progressively approaches death. The GOOSS is a scoring system to classify a patient's level of oral intake as follows: 0, no oral intake; 1, liquids only; 2, soft solids; 3, low-residue or full diet 33 . The presence of ascites/liver metastasis/peritoneal dissemination was evaluated by computed tomography before the procedure.

Chemotherapy after C-SEMS placement (%) Migration rate (%)
The primary endpoint was technical success. The secondary endpoints were clinical success, changes in GOOSS score from before SEMS placement to 1 week after SEMS placement, and adverse events including SEMS migration. Technical success was defined as satisfactory SEMS placement and precise positioning at the obstruction site, and ability to deploy the OTSC for SEMS fixation. Clinical success was defined as a GOOSS score of ≥ 2 Figure 3. A 60-year-old female with pancreatic cancer. The third portion of the duodenum was obstructed by the tumor (A), and a partially covered self-expandable metal stent (PC-SEMS) was deployed (B). The over-thescope-clip (OTSC) system was loaded onto the endoscope, and part of the upper rim of the metal stent was suctioned into the transparent cap. The OTSC was released to grasp both the metal stent and duodenal wall (C). A fluoroscopic image after OTSC and PC-SEMS placement (D). www.nature.com/scientificreports/ and relief of GOO symptoms within 1 week after SEMS insertion. Procedure time for OTSC placement was calculated from inserting the endoscope that loaded the OTSC to endoscope removal after OTSC placement. Stent migration was defined as movement out of the stricture, which was diagnosed on endoscopy and radiography. We routinely perform abdominal X-rays on days 1, 3, 7, and 14 and every 2 weeks after the procedure to detect symptomatic and asymptomatic stent migration until the patient's death.

Statistical analysis.
Values are reported as medians with ranges. Categorical data were compared using Fisher's exact test. Procedure time and changes in GOOSS scores are expressed as mean (± standard deviation) and were analyzed using Wilcoxon's signed rank test. Differences were considered to be significant at P < 0.05. The cumulative time to stent dysfunction was evaluated using Kaplan-Meier analysis. All statistical analyses were performed using SPSS software (version 19; IBM Corporation, USA).