Keywords
Endoscopy Lower GI Tract - Endoscopic resection (polypectomy, ESD, EMRc, ...) - Endoscopy
Upper GI Tract - POEM - Gastrostomy and PEJ
Introduction
Through-the-scope endoscopic clips (TTSCs) are used for a broad range of indications,
including closure of gastrointestinal defects, with decreasing closure efficacy for
larger defects [1]. Many types of TTSCs are available, and physicians often select clips based on physical
characteristics, cost, or functional characteristics [2]. Defect characteristics and TTSC capabilities should be considered when choosing
the closure device and technique [1]. Defect size is an important determinant of clip choice, i.e. small perforations
that occur during endoscopic polyp resection can successfully be closed using TTSCs
with favorable outcomes, while larger perforations are more expediently and effectively
closed with large-caliber over-the-scope clips (OTSCs) mounted on the outside of a
clear cap affixed to the tip of the endoscope [3].
A 2019 bench study tested five models of marketed TTSCs with respect to rotatability,
overshoot/whip, open/close precision, and tensile/closure strength in four different
endoscope configurations: (1) straight, (2) duodenal sweep, (3) full retroflexion,
and (4) across the duodenoscope elevator [2]. The authors mentioned that clip use in confined spaces or with thin tissue (e.g.,
Zenker’s diverticulectomy, endoscopic submucosal dissection-induced defects or bleeding)
probably calls for a smaller clip with precise open/close, while therapies in full
retroflexion or across the duodenoscope elevator probably require precise clip functioning
in the most strained configuration [2]. Perforation closure may require clips with enhanced ability to laterally manipulate
tissue to appose defect edges, and large ulcers or fibrous ulcers may require a clip
with higher compression force [2]. Because increased gripping ability might expand potential indications, we are testing
a TTSC with anchor prongs. Use of this clip has been reported for closure of complex
polyp resection defects in three patients [4]. For preliminary evidence of the functional capacity of the new TTSC, we present
an analysis of our first 50 cases.
Patients and methods
Study design
This was a case series of 50 consecutive patients with indication for endoscopic clipping,
who received treatment with the MANTIS clip (Boston Scientific Corporation, Marlborough,
Massachusetts, United States). This clip was cleared by the US Food and Drug Administration
in August 2022 [5]
[6]. The study was conducted at two US sites and one Canadian site.
Patient population
Patients of any age scheduled for an indicated endoscopic clipping per local standard
of practice were eligible to enroll. Exclusion criteria were enrollment in another
study that would directly interfere with the current study or investigator’s assessment
that the subject was at risk for study device or procedure-related complications.
All centers obtained approval from their respective local ethics committees. All patients
provided signed informed consent before nonemergent procedures. In emergent cases
(e.g. perforation or acute bleeding), consent was performed after the procedure, but
before data collection because preprocedural consent was unfeasible. Periprocedural
management of antiplatelet and anticoagulant agents was per endoscopist discretion.
Study visits
Index procedure and postprocedure follow-up
Baseline screening visit included informed consent, eligibility assessment, age, gender
and relevant medical history. Intervention was placement of a new endoscopic clip
in the gastrointestinal tract. During the index procedure, the type of clipping procedure,
number of clips placed, and lesion size were recorded. After the index procedure,
participants continued to receive medical care per standard of practice. Adverse events
(AEs) and device events were followed for 30 days. The last study visit was a telephone
interview at 30 days (± 5 days). For participants who did not complete this interview,
the reason was recorded.
Outcomes
We evaluated: 1) successful closure of defects defined as no submucosa visible and
clips < 1 cm apart; 2) delayed bleeding rate; 3) reinterventions; 4) rate of serious
AEs (SAEs) related to MANTIS clip or endoscopic portion of procedure; and 5) technical
difficulties.
Statistical analysis
Baseline characteristics, medical history, outcome measures, and AEs were summarized
using mean, median, standard deviation (SD), range for continuous variables (e.g.
age, procedure times), and proportions for categorical variables. All analyses were
performed in SAS version 9.4.
Results
Patient and procedure characteristics
Among 63 patients screened for study eligibility, 10 were excluded for not having
an indication for clip placement and three were excluded because the investigator
deemed them at risk for study device or procedure-related complications per the Instructions
for Use. Among the latter three patients, one defect was deemed too fibrotic, there
was concern about placing both jaws in the center of a hot endoscopic mucosal resection
(EMR) defect in the rectum of the second, and there was concern about placing both
jaws of the MANTIS in submucosa to manage a vessel with ongoing bleeding in EMR defect
in the third patient. Standard TTSCs were used in these cases. Among enrolled participants,
the mean age was 62.8 years (range 28.0–80.0 years) and most participants were male
(n = 32, 64.0%) and White (n = 44, 88.0%). Sixteen (32.0%) were taking nonsteroidal
anti-inflammatory drugs including aspirin, six (12.0%) were taking anticoagulants,
and two (4.0%) were taking antiplatelets at baseline ([Table 1]).
Table 1 Baseline patient characteristics (n = 50).
|
Characteristic
|
Mean ± SD (range) or n/N (%)
|
|
POEM, peroral endoscopy myotomy.
*Each patient had one or more of the listed medical conditions; rows are not mutually
exclusive.
|
|
Mean age, yrs
|
62.8±11.9 (28.0, 80.0)
|
|
Gender: Male
|
32/50 (64.0%)
|
|
Ethnicity
|
|
|
45/50 (90.0%)
|
|
|
1/50 (2.0%)
|
|
|
4/50 (8.0%)
|
|
Race
|
|
|
44/50 (88.0%)
|
|
|
3/50 (6.0%)
|
|
|
1/50 (2.0%)
|
|
|
1/50 (2.0%)
|
|
|
1/50 (2.0%)
|
|
Medical history*
|
|
|
36/50 (72.0%)
|
|
|
9/50 (18.0%)
|
|
|
5/50 (10.0%)
|
|
|
4/50 (8.0%)
|
|
|
4/50 (8.0%)
|
|
|
1/50 (2.0%)
|
|
|
1/50 (2.0%)
|
|
|
1/50 (2.0%)
|
|
|
1/50 (2.0%)
|
|
|
1/50 (2.0%)
|
|
|
1/50 (2.0%)
|
|
Actively taking NSAIDs
|
16/50 (32.0%)
|
|
Actively taking anticoagulants
|
6/50 (12.0%)
|
|
Apixaban
|
3/50 (6.0%)
|
|
Lovenox
|
2/50 (4.0%)
|
|
Warfarin
|
1/50 (2.0%)
|
|
Actively taking antiplatelets (clopidogrel)
|
2/50 (4.0%)
|
Types of procedures performed
Post-EMR prophylactic hemostasis was the reason for clipping in 34 defects. EMR was
performed for 33 colonic lesions and one gastric lesion. Some patients had multiple
lesions and thus underwent EMR of multiple lesions or a combination EMR/hot snare
polypectomy. Sixteen colon polyps were removed from 12 patients via polypectomy without
EMR – 15 using hot snare and one using cold snare. These resection defects were then
clipped. Clipping was performed in two patients (4.0%) for hemostasis of active bleeding.
Clipping was performed in one patient (2.0%) each for gastrocutaneous fistula closure
[5], post peroral endoscopic myotomy (POEM) closure of the mucosal incision, or anchoring
of a feeding tube ([Table 2]).
Table 2 Procedure details (N = 50).
|
Characteristic
|
Mean ± SD or median
(range) or % (n/N patients)
|
|
SD, standard deviation; EMR, endoscopic mucosal resection; POEM, peroral endoscopic
myotomy; APC, argon plasma coagulation.
*In the post-EMR category, one patient had three lesions, four patients had two lesions,
and 28 patients had one lesion (total 39 lesions). Patients in all other categories
had one lesion/clipping procedure each.
†APC was used for margin and base ablation of a colonic EMR defect [6]
[7] and combined with ligature device for gastrocutaneous fistula closure [5].
|
|
Reason for clipping procedure
|
|
|
66.0% (33/50)
|
|
|
24.0% (12/50)
|
|
|
4.0% (2/50)
|
|
|
2.0% (1/50)
|
|
|
2.0% (1/50)
|
|
|
2.0% (1/50)
|
|
Additional modalities used during procedure
|
|
|
86.0% (43/50)
|
|
|
4.0% (2/50)
|
|
|
4.0% (2/50)
|
|
|
4.0% (2/50)
|
|
|
2.0% (1/50)
|
|
|
18.0 (1.2, 60.0)
|
|
|
10.0 (1.0,40.0)
|
|
|
2.6 ± 1.8 (1.0,9.0)
|
|
|
1.0 ± 1.9 (0.0,9.0)
|
|
|
3.6 ± 2.6 (1.0,11.0)
|
Study outcomes
Successful defect closure
Successful defect closure was completed in 53 of 55 lesions (96.4%) in 47 total patients
in this study. The case in which endoscopic clipping was used for successful feeding
tube anchoring is excluded here. Colonic post-EMR defects (33) were the most common
lesions. All clips were placed per labeled directions for use. A post-EMR defect measuring
30 mm in maximum diameter x 5 mm in minimum diameter in the cecum was unable to be
completely closed due to difficult positioning. A rectal polypectomy defect measuring
20 × 20 mm was incompletely closed due to inability to appose tissue.
Among the 55 defects, median maximum defect size was 18.0 mm (range 1.2–60.0 mm) and
median minimum defect size was 10.0 mm (range 1.0–40.0 mm). The mean number of study
clips used was 2.6 ± 1.8 (range 1.0–9.0) and the mean number of other (non-study)
clips used was 1.0 ± 1.9 (range 0–9.0) per procedure. The mean total number of clips
used during a procedure was 3.6 ± 2.6 (range 1.0–11.0).
For colon post-EMR defects, 32 of 33 defects were completely closed. The median minimum
defect diameter was 10 mm (range 5–40 mm) and the median maximum defect diameter was
25 mm (range 8–45 mm) for these defects. For post-EMR defects of the colon, the mean
number of clips placed was 4.4 ± 2.6 (range 1 to 11, including an average of 3.0 study
clips and 1.3 non-study clips). Due to some defects lying over folds and other factors,
the defect size was smaller than the actual lesion removed in some cases.
For non-colon EMR defects (excluding using a study clip to affix a jejunal tube),
21 of 22 were successfully completely closed. The median minimum diameter was 8 mm
(range 1–40 mm) and the median maximum diameter was 10 mm (range 1.2–60 mm). For all
non-EMR of colon defects, the mean number of clips placed was 2.7 ± 2.4 (range 1 to
11, including an average of 2.1 study clips and 0.6 non-study clips).
Delayed bleeding prophylaxis
In 41 (82.0%) patient cases, prophylaxis to reduce risk of delayed bleeding was reported
as the indication for endoscopic clipping. Delayed bleeding from clipped intervention
sites occurred in three of 56 (5.4%) total interventions in 50 patients in the study.
For post-EMR colon defects specifically, delayed bleeding occurred in one of 33 (3.0%)
clipped defects. The one case of delayed bleeding was the previously mentioned incompletely
closed cecal defect. For all other defects, delayed bleeding occurred in two of 22
clipped defects (9.0%). These cases were of the previously mentioned incompletely
closed rectal polypectomy defect and the post-POEM defect.
Serious adverse events related to the MANTIS clip
Of the 50 enrolled patients, none of the patients had AEs or SAEs related to the MANTIS
clip nor any other clip.
Study clip technical success
There was one instance of the clip deploying prematurely. There were zero instances
of the clip failing to deploy, failing to release from the catheter, being difficult
to release from the catheter, failing to open or close, being unable to rotate, failing
to advance from the sheath, being difficult to deploy, having the clip arms or anchor
prongs bent, having poor bite of tissue, or any other events that could be seen as
technical failure of the clip itself.
Serious adverse events related to the endoscopic procedure and reinterventions
Five patients had a total of four SAEs and one nonserious AE related to the endoscopic
portion of the procedure within a median of 12 days (range 0–14 days) after the index
procedure ([Table 3]). The first patient had a POEM procedure with a related SAE and two reinterventions.
The patient was found to have esophageal leak on imaging on Day 0 but the site was
contained so the patient was treated with antibiotics and nothing by mouth. This patient
had two repeat upper endoscopies. The patient had bleeding from the esophageal mucosectomy
site and an esophageal ulcer. The first repeat endoscopy performed for bleeding on
Day 13 after the procedure showed a clot near the clipping site; the study clip was
removed for better visualization and the site was treated with a hemostatic spray
and placement of a non-study clip. Two days later, this patient had another procedure
to treat an esophageal ulceration using hemostatic spray. No blood transfusion was
required. The second patient had a related SAE and reintervention due to bleeding
on Day 12 after EMR of a cecal lesion that was incompletely closed due to difficult
access. This patient underwent colonoscopy. The endoscopist surveilled the affected
site, but no treatment was needed. No blood transfusion was required. A third patient
who had a rectal hot snare polypectomy site that was not completely closed also had
a bleed on post-procedure Day 6 that resolved (nonserious AE with a reintervention).
Subsequently, colonoscopy was performed and no intervention was needed. No blood transfusion
was needed. Notably, a fourth patient had an unrelated SAE with a reintervention after
melena occurred on Day 1 after the procedure. An upper endoscopy demonstrated bleeding
not related to the previous endoscopic procedure nor the study clip. This bleeding
was treated with a non-study clip. A fifth patient had a related SAE of post-polypectomy
electrocoagulation syndrome that did not require reintervention.
Table 3 Serious adverse events related to the endoscopic portion of the procedure.
|
Number
of SAEs
|
Number
of patients (n/N) (%)
|
|
SAE, serious adverse event.
|
|
Any serious adverse event
|
5
|
3/50 (6.0%)
|
|
Bleeding
|
2
|
2/50 (4.0%)
|
|
Esophageal perforation
|
1
|
1/50 (2.0%)
|
|
Post-polypectomy electrocoagulation syndrome
|
1
|
1/50 (2.0%)
|
|
Ulceration
|
1
|
1/50 (2.0%)
|
Descriptions of how the clip was used
“Grasp-and-drag” technique for a standard EMR defect
After hot snare piecemeal EMR of a lateral spreading lesion in the colon, several
non-MANTIS clips had already been placed. [Video 1] demonstrates the MANTIS clip was opened and normal mucosa on one edge of the lesion
was grasped. With the MANTIS clip closed, the colonoscope was used to lift the MANTIS
clip with grasped tissue to the opposite side of the lesion. The MANTIS clip was then
opened, with the anchor prongs maintaining the grasp of tissue. With the jaws open,
the MANTIS clip was gently pushed to the mucosa to maximize use of its jaws, then
closed approximating defect edges, then deployed ([Video 1] – Part 1).
Part 1 “Grasp-and-drag” technique to close a defect after hot snare piecemeal endoscopic
mucosal resection of a lateral spreading tumor in the colon.
Part 2 “Open-jaw” technique to prophylactically close a scarred defect of a hot endoscopic
mucosal resection of a 50-mm scarred granular lateral spreading tumor in the ascending
colon.
Part 3 Repeated “open-jaw” technique to prophylactically close a broad-based defect after
piecemeal endoscopic mucosal resection of a 45-mm granular lateral spreading tumor
in the cecum.Video 1
“Open-jaw” technique for a scarred defect
The lesion was a 50-mm granular lateral spreading tumor (G-LST) in the ascending colon
that had undergone a previous attempt at resection before referral to our center.
The lesion was assessed and resected with piecemeal hot EMR. Evidence of the previous
resection attempt was seen at the haustral folds as scarred-down tissue extending
across the defect. Hot avulsion was used to remove flat and fibrotic polyp tissue.
Snare tip soft coagulation was performed to the edges of the defect.
The MANTIS clip's anchor prongs were used to drag tissue to approximate the defect
without closing and reopening the jaws (“open-jaw technique”) or slipping from the
tissue as standard TTSCs might when used to close a wide defect. This technique is
useful when the defect is approached tangentially. Here, the MANTIS clip was rotated
so that when opened, the orientation was vertical with the top jaw between 10 and
2 o'clock. The bottom anchor prongs were used to grasp normal tissue at the distal
(anal side) edge of the lesion and the colonoscope was advanced to lift and push the
mucosa on the anal side of the defect toward the mucosal edge on the defect’s cecal
side. The top anchor prongs were then used to grasp normal tissue at the proximal
edge of the lesion. The clip was pushed gently to "bury" the clip and use the full
length of the jaws before deployment. This sequence was repeated. The defect was
closed with a total of six MANTIS clips and one standard TTSC. The MANTIS clip was
easily maneuvered for placement between previously deployed clips. Alternatively,
fewer MANTIS clips can be used to bring the defect edges closer, followed by closure
of the remaining defect using standard TTSC ([Video 1] – Part 2).
“Open-jaw” technique for a broad-based defect
This shows a large submucosal defect after piecemeal hot EMR of a 45-mm G-LST in the
cecum. The "open-jaw” technique was again utilized to grasp normal tissue on the anal
side of the lesion and move it to the cecal edge of lesion. Three MANTIS clips were
used followed by three standard TTSC to completely close this lesion ([Video 1] – Part 3).
Discussion
A new TTSC with anchor prongs had a high rate of success for defect closure, low rate
of delayed bleeds, and high rate of technical success regarding clip deployment in
this case series of 50 patients across three centers in the United States and Canada
in a variety of cases. There were zero AEs related to the clip itself and it was technically
successful in all but one instance because of premature deployment. The low rate of
related SAEs was consistent with published safety data on endoscopic clips [8]. We demonstrated two techniques for successful use of this technique.
Due to the tissue apposition ability of the MANTIS clip, it may be able to reduce
the number of standard TTSCs used during defect closure or allow for closure of larger
defects without OTSC or suturing devices. Recently, another novel TTSC has been introduced
– the dual action tissue (DAT) clip (Micro-Tech Endoscopy, USA, Ann Arbor, Michigan,
United States) [9]. This clip features two independent arms and is efficacious in closing large resection
defects. Because it cannot be rotated, it requires manipulation of the scope dials
for positioning and operating the independent jaws can be difficult with endoscope
looping [10]. The MANTIS is operated similarly to a standard TTSC. It may have a shorter learning
curve for both endoscopist and technician than alternative devices. Comparative studies
for endoscopic defect closure between these clips and other devices such as suturing
are warranted.
Data sharing
The data, analytic methods, and study materials for this study may be made available
to other researchers in accordance with the Boston Scientific Data Sharing Policy
(http://www.bostonscientific.com/en-US/data-sharing-requests.html).
