Introduction
Endoscopic full-thickness resection (EFTR) has recently been used for epithelial and
subepithelial gastrointestinal lesions [1]
[2]. Several reports have been published on use of EFTR in which most lesions measured < 3 cm,
particularly subepithelial lesions such as gastrointestinal stromal tumors (GISTs),
but surgery is the standard for GISTs [3]
[4]. To date, in Japan, use of EFTR has been limited to gastric subepithelial lesions
< 3 cm as an advanced medical treatment, given the feasibility of oral retrieval and
security of defect closure using applicable closure devices.
Several endoscopic closure devices and techniques have been developed, such as over-the-scope
clip (OTSC) [2]
[5]
[6], purse-string suturing using endoloop and clips [1]
[7]
[8], hand suturing [9]
[10]
[11], O-ring and OTSC [12], and OverStitch [13], but these are complicated and require use of specific devices. For widespread use
of EFTR, application of easily available and daily-use devices for closure is ideal.
We developed a simple closure technique involving only clips and suture line, called
line-assisted complete closure (LACC) [14], which can be used to treat mucosal defects after endoscopic submucosal dissection
(ESD), and showed high rates of success and efficacy with it [15].
Thus, we conducted a pilot study using porcine models to evaluate the feasibility
of modified LACC for defects < 3 cm after gastric EFTR as a first step.
Patients and methods
Study approval
This study was conducted after approval from the Animal Care and Use Committee, Okayama
University (institutional number: OKU-2020730).
Animal experiment
Six live, three-way crossbred, 3-month-old female pigs were included in this study.
After environmental acclimation for 7 days, the pigs were fasted for 24 h before the
endoscopic procedures. On the day of the procedure (Day 0), 0.04 mg/kg medetomidine,
0.2 mg/kg midazolam and 5 mg/kg ketamine were injected intramuscularly, followed by
tracheal intubation and continuous inhalation anesthesia using 2 % to 5 % isoflurane.
EFTR and modified LACC as described below, were performed at the greater curvature
of the gastric antrum in three pigs, and at the anterior wall of the lower gastric
body in the other three pigs. The procedures for the greater curvature of the gastric
antrum were performed in forward view, and those for the anterior wall of the lower
gastric body were performed in retroflex view. After the procedures, the pigs were
allowed to drink water and were fed with a regular diet 1 day after the procedure
(Day 1). For 3 days including Day 0, an antibiotic (enrofloxacin 2.5 mg/kg) was injected
intramuscularly.
Until postoperative day 4 (Day 4), the pigs were observed carefully, and their eating
habits, body weight, and general conditions were documented. Blood samples were collected
on Days 0 and 4. Under general anesthesia, closure sites were endoscopically observed
on Day 4, and 50 mL 1 % methylene blue was poured into the closure sites to evaluate
leakage. A midline abdominal incision was performed, and the presence or absence of
peritonitis and leakage of fluid with methylene blue was evaluated visually. The peritoneal
side of the closure sites was also evaluated. All pigs were euthanized using a lethal
dose of potassium chloride.
Modified LACC for defects after EFTR
The modified LACC and EFTR procedures were performed by three endoscopists who had
experience with more than 100 gastrointestinal ESDs. A single-channel endoscope (PCF-Q240ZI;
Olympus Tokyo, Japan) and CO2 insufflation were used, and the endoscope was inserted into the stomach ([Video 1]). After irrigation of the stomach using physiological saline, a circular marking
(~3 × 3 cm) was created using a triangle tip electrosurgical knife (TT knife; Olympus)
and submucosal injection of saline was performed. ESD using the TT knife just outside
the marking was performed to create a mucosal defect and expose the muscle layer.
After endoscope withdrawal, a clip-and-line was applied. The endoscope was reinserted
and the center of the exposed muscle layer was captured using a clip-and-line to achieve
traction. A muscle layer (~2 × 2 cm) inside the defect was removed using the TT knife
and EFTR was achieved. When excess free air affected the circulation dynamics during
EFTR, peritoneal aspiration using a 20-gauge needle was performed.
Video 1 Endoscopic full-thickness resection (EFTR) and line-assisted complete closure (LACC)
of the defect. First, endoscopic submucosal dissection about 3 cm in size was performed.
Second, EFTR 2 cm in size was performed using the traction technique. Third, the precutting
method was used. Fourth, LACC was applied to minimize the defect. Fifth, a simple
clip was added to achieve complete closure.
After completion of EFTR, modified LACC was performed as described below ([Fig. 1], [Fig. 2] and [Video 1]) [14]
[15]. Multiple small incisions just outside the defect (precutting method [16]) were made circumferentially using the TT knife as preparation for LACC. Precutting
was performed four on each side (in total eight precutting). A 3–0 silk line (Akiyama-Seisakusho
Co. Ltd., Tokyo, Japan) was tied to a clip (HX-610–090, EZ clip; Olympus) mounted
on an applicator (HX-110LR; Olympus), and the clip was retracted into the applicator
and inserted into the accessory channel. The clip tied a line was fully opened, and
one side of the claw of the clip was inserted into the incision on the oral side,
while the other side of the claw was placed inside the margin of the defect (at the
exposed muscle layer), and the claws were closed. Another clip (without a line) was
used to anchor the line to the opposite side. One side of the claw of another clip
was inserted into the anal-side incision, while the other side of the claw was placed
inside the margin (at the exposed muscle layer), and the claws were closed. Finally,
both clips were gathered together by gently pulling the line by hand. At this stage,
both mucosa and muscle layer were gathered as much as possible. A third clip was inserted
while threading a line through the gap at the base of the clip teeth, and the clip
was fully opened near the first and second clips, resulting in the line and clips
being fixed (locking technique) [17]. The third clip was closed to catch the mucosa. The TT knife was inserted, and the
line was cut by forced coagulation mode. The modified LACC procedure using regular
clips (EZ clip) was repeated to minimize the defects, and reopenable clips (Sure Clip;
MC Medical, Tokyo, Japan) were added to close the defect completely. The modified
LACC was a combination of the original LACC and, precutting and locking technique.
Fig. 1 Modified line-assisted complete closure (LACC) of the defect after endoscopic full-thickness
resection (EFTR). a The defect after EFTR. b The clip-and-line was applied to catch the mucosa of the oral side. c Another clip was used to anchor the line to the opposite side. d Both clips were gathered together. e The third clip fixed the line. f The third clip was closed to catch the mucosa. g The line was cut by triangle tip knife. h LACC procedure was repeated to minimize the defects. i Complete closure was achieved.
Fig. 2 Schema of modified LACC. a Multiple small incisions (basically eight incisions) just outside the defect were
made circumferentially. b A clip tied a line was applied. One side of the claw was inserted into the incision,
while the other side of the claw was placed at the exposed muscle layer, and the claws
were closed. c, d Another clip was used to anchor the line. One side of the claw was inserted into the
anal-side incision, while the other side of the claw was placed at the exposed muscle
layer, and the claws were closed. e By pulling the line, both clips were gathered. f, g A third clip was inserted from the accessory channel while threading a line through
the gap of the base of the clip teeth. h, i The clip was fully opened, resulting in the line and clips being fixed. j The clip was closed to catch the mucosa, and the line was cut by TT knife. k The modified LACC procedure was repeated to minimize the defects. l Reopenable clips were added to close the defect completely.
Measurement outcomes
The evaluated outcomes were the success rate of modified LACC, which was defined as
complete closure (whole defect was covered by the mucosa) on Day 0, closure procedure
time, modified LACC times, maintenance of defect closure on Day 4, presence of peritonitis
or leakage on Day 4, change of laboratory test, and clinical course. Clinical course
was evaluated on the basis of the weight-loss process and dietary intake.
Results
[Table 1] shows the location and defect size based on resected specimens and the outcomes
of modified LACC. Successful modified LACC of the defect after EFTR was achieved in
all six cases. Median closure procedure time was 43.5 minutes (range 37 to 65). Median
number of LACC were 3 times (range 3 to 4 times), and median total number of clips,
which included the clip used for modified LACC, was 24.5 (range 20 to 28).
Table 1
Outcomes of modified LACC after EFTR.
|
Case
|
Location
|
ESD defect size, mm
|
EFTR defect size, mm
|
Complete closure
|
Closure procedure time, min
|
Total no. clips
|
Modified LACC times
|
|
1
|
Antrum, Gre
|
26 × 25
|
18 × 16
|
Yes
|
45
|
20
|
3
|
|
2
|
Antrum, Gre
|
28 × 26
|
18 × 17
|
Yes
|
65
|
27
|
4
|
|
3
|
Antrum, Gre
|
27 × 25
|
20 × 19
|
Yes
|
38
|
20
|
3
|
|
4
|
Lower body, Ant
|
27 × 26
|
18 × 18
|
Yes
|
42
|
24
|
3
|
|
5
|
Lower body, Ant
|
29 × 28
|
19 × 18
|
Yes
|
62
|
28
|
4
|
|
6
|
Lower body, Ant
|
30 × 28
|
21 × 19
|
Yes
|
37
|
25
|
3
|
LACC, line-assisted complete closure; EFTR, endoscopic full-thickness resection; ESD,
endoscopic submucosal dissection.
Maintenance of defect closure on Day 4 was not achieved in all cases. Partial dehiscence
was observed in three cases at the greater curvature of the gastric antrum and complete
dehiscence was observed in the other three cases at the anterior wall of the lower
gastric body. However, there was neither peritonitis nor fluid leakage in any case.
In the gastric antrum, the defects were covered by greater omentum, and in the gastric
body, the defects were covered by adhesion by surrounding tissues including the liver
([Fig. 3]). All pigs ate well on Day 1 (700 g/day swine diet; Oriental Yeast Co. Ltd. Tokyo.)
and body weight increased on Day 4. The clinical course progressed well with no delayed
adverse events (AEs) ([Table 2]).
Fig. 3 Closure site on Day 4. a Partial dehiscence was observed (greater curvature of the antrum). b Methylene blue was poured into the closure sites (luminal side). c The defects were covered by greater omentum (peritoneal side). (D) Complete dehiscence
was observed (anterior wall of the lower body). e Methylene blue was poured into the closure sites (luminal side). f The defects were covered by surrounding tissue (peritoneal side).
Table 2
Clinical course after line-assisted complete closure.
|
Case
|
WBC, /µL
|
WBC, /µL
|
BW, kg
|
BW, kg
|
No. remaining clips (Day 4)
|
Maintaining closure (Day 4)
|
Peritonitis
|
Leakage
|
Clinical course
|
|
(Day 0)
|
(Day 4)
|
(Day 0)
|
(Day 4)
|
|
1
|
17000
|
35100
|
25.9
|
27
|
9
|
Partially dehiscent
|
No
|
No
|
Being well
|
|
2
|
19000
|
23000
|
23.4
|
26.2
|
6
|
Partially dehiscent
|
No
|
No
|
Being well
|
|
3
|
19700
|
25200
|
23.8
|
26.2
|
7
|
Partially dehiscent
|
No
|
No
|
Being well
|
|
4
|
17500
|
15700
|
30.8
|
31.8
|
0
|
Entirely dehiscent
|
No
|
No
|
Being well
|
|
5
|
19900
|
20300
|
22.5
|
23.4
|
0
|
Entirely dehiscent
|
No
|
No
|
Being well
|
|
6
|
15700
|
17400
|
26
|
26.4
|
0
|
Entirely dehiscent
|
No
|
No
|
Being well
|
WBC, white blood cell; BW, body weight.
Discussion
Our study using porcine models showed that modified LACC for defect closure after
gastric EFTR was feasible without any AEs, but the closure was only partial or entirely
dehiscent on Day 4. The clinical course was good in all cases due to the defect being
rapidly covered by adhesion from the peritoneal side, but permanent closure was theoretically
ideal. Thus, some improvements in LACC are needed to achieve more secure closure after
EFTR.
To date, LACC has mainly been used for mucosal defect closure after ESD, although
there are several reports of successful application of LACC for large iatrogenic perforations
during endoscopic treatment [18]
[19]. In this study, we first evaluated the feasibility of modified LACC for defect closure
after EFTR. The defect after EFTR was completely closed in all cases on Day 0, but
closure was not maintained. There may be several reasons for closure dehiscence. First,
the gastric mucosa is thick and has few folds compared with other organs, which exert
tension, resulting in disruption of the closure site. Second, dietary control was
impossible in the porcine model. It is usual in clinical practice to fast a few days
after EFTR. The fasting period might influence dehiscence. Third, the strength of
closure by modified LACC seemed insufficient to maintain closure. A previous study
suggested that the closure strength of purse-string suturing using endoloop and simple
clip closure was weaker than that of hand suturing [20]. Besides, even if hand suturing were applied, permanent closure in all cases after
gastric ESD is difficult [9]
[10]. We assumed that the closure strength of modified LACC was equivalent to that of
purse-string suturing or simple clip closure; thus, we applied a precutting and locking
technique for LACC to reinforce the closure. However, the closure was dehiscent. Maintenance
of closure after gastric EFTR is still challenging and the rate of closure maintenance
after gastric EFTR has not been fully evaluated for other methods. Improvement in
LACC (e. g., application of cinching devices or tissue adhesives) in future might
be a good option to strengthen the closure.
In comparison with previously reported methods, modified LACC was easily available
without special devices, because hand suturing and purse-string require a two-channel
endoscope, and expensive devices are required for OTSC and OverStitch. However, median
closure time for modified LACC (43.5 minutes) was not shorter than that of other reported
methods for similar defect size in porcine or human studies: hand suturing, median
25 to 49.5 minutes [9]
[10]; O-ring and OTSC, median 35.5 minutes [12]; and pure-string suture, mean 14.6 minutes [8]. Further developments are needed to reduce procedure time.
This study had several limitations. First, small numbers of porcine models were used
for the pilot study to minimize animal suffering. Second, the follow-up period was
limited to Day 4. Long-term outcome of modified LACC after EFTR was unclear, although
most delayed AEs occur within 2 days after endoscopic resection. However, administration
of an antibiotic during most of the study period might mask peritonitis. Third, the
closure site and the presence of leakage were evaluated macroscopically. The site
was carefully observed using methylene blue, but there was no pathological evaluation.
Fourth, the LACC time varied, although the size of the defect treated by modified
LACC was limited. Although three experts performed the procedures, the stability and
reproducibility of modified LACC after EFTR was questionable, based on results from
this study. Fifth, many more clips were used for modified LACC than we anticipated,
because there are few folds in the gastric of porcine models and endoscopy maneuverability
was poor, which was not cost-effective.
Conclusions
In conclusion, we have shown that modified LACC can be used to close the defect after
gastric EFTR in porcine models, although the closure is not maintained. Fortunately,
the clinical course after the procedures was good in all cases, given the rapid adhesion,
but improvement is needed to maintain closure before using the technique in human
EFTR.
