Introduction
Esophageal endoscopic submucosal dissection (ESD) is widely spread as a minimally
invasive treatment for superficial esophageal cancer [1]
[2], but it is often difficult to obtain adequate visualization due to its narrow lumen
compared to the stomach and the colon. In particular, intraoperative perforation is
reported to be relatively frequent in left-sided wall lesions [3], which may be due to the difficulty in getting enough traction by gravity to obtain
an adequate visualization, in addition to the accumulation of water and blood due
to gravity.
To solve this issue, some reports have suggested the effectiveness of traction, such
as clip-thread method [4]
[5]
[6]
[7]. However, the direction of traction is limited, and the traction direction is closer
to the tangential direction as the lesion diameter increases. Therefore, it is difficult
to obtain effective traction when the lesion diameter or circumference increases.
Given these difficulties, we developed an over-tube type dedicated traction device
(ENDOTORNADO). It is 400 mm long and the outside diameter is 21.0 mm. It has a 2.8-mm
diameter working channel to deliver various devices and allows the traction from any
directions by rotating itself ([Fig. 1]). We have reported it significantly shortened the procedure time of esophageal ESD
in live porcine experience [8]. Consequently, we investigated the feasibility and potential usefulness of this
new device in human esophageal ESD clinically.
Fig. 1 Appearance of ENDOTORNADO. ENDOTORNADO is 400 mm long and the outside diameter is
21.0 mm. It has a 2.8-mm diameter working channel to deliver various devices and a
soft inner tube to prevent mucosal damage when proceeded into the esophagus.
Patients and methods
Study design and patient population
This was a single-center retrospective observational study. We analyzed six consecutive
superficial esophageal cancer cases performed ESD by one operator using ENDOTORNADO
between January and March 2022 (tESD group). Twenty-three consecutive cases performed
conventional esophageal ESD as usual by the same operator from February 2018 to December
2021 (cESD group) were chosen as controls. This operator had over 400 cases of esophageal
ESD experience at the beginning of this study. Cases performed under general anesthesia
with intubation were excluded, and anesthesia was provided by Flunitrazepam or Dexmedetomidine
Hydrocholoride. Esophagogastric junction cancer were excluded, and when multiple lesions
were resected at the same time, the largest lesion was included in this study.
This study was performed in accordance with 2008 version of the Declaration of Helsinki.
We obtained informed patient consent to analyze the data, and the study protocol was
approved by the institutional review board of our institution (20190139).
Surgical equipment
ESD was performed using a high-vision therapeutic endoscope with a water jet function
(GIF-H290T; Olympus Medical Systems, Tokyo, Japan). Submucosal injections of a 10 %
glycerin solution (Glyceol; Chugai Pharmaceutical Co, Ltd, Tokyo, Japan) were performed
with a 25G needle (NeedleMaster, Olympus Medical Systems, Tokyo, Japan). For difficult
cases, 0.4 % sodium hyaluronate (MucoUp, Boston Scientific, Marlborough, Massachusetts,
United States) was used as required.
As an energy device, the operator used a needle type knife with injection function
(DualKnife J, Olympus Medical Systems, Tokyo, Japan) in all cases. In tESD group,
DualKnife J was only used for marking, mucosal incision and creation of small mucosal
flap, and a high-frequency scissors forceps (ClutchCutter, Fujifilm, Tokyo, Japan)
was used for submucosal dissection. DualKnife J was powered by a high-frequency electrosurgical
unit (VIO 3, ERBE Elektromedizin, Tübingen, Germany), which supported lesion marking
(soft coagulation function, effect 6.0), hemostasis with the knife tip (spray coagulation
function, effect 1.2), mucosal incision (dry cut function, effect 2.2), and submucosal
dissection (swift coagulation, effect 3.5). ClutchCutter was also powered by VIO3,
which supported submucosal dissection (Endocut I function [effect 1.0, duration 1,
interval 1] and forced coagulation [effect 3.0]).
Description of technique
After identifying the area of the lesion, markings were made around the lesion. A
full circumferential mucosal incision was made and a small mucosal flap was created
on the oral side of the lesion. Then, ENDOTORNADO was inserted and placed in the esophagus
with the scope as a guide. The tip hood was re-attached and the scope was re-inserted
to the oral side of the lesion. The snare forceps were delivered through the side
hole of ENDOTORNADO and the snare was expanded. The clip forceps was delivered through
the forceps hole of the scope, and the clip was set through the expanded snare. Then
the clip was deployed on the small mucosal flap created on the oral side of the lesion,
and the clip was grasped by tightening the snare. In this step, the assistant's task
was successfully relieved by using a clothespin to maintain the snare grasping state.
Thereafter, the traction from any directions was possible by the assistant making
a torque to ENDOTORNADO ([Fig. 2]). The submucosal dissection was performed using a high-frequency scissors forceps
(ClutchCutter, Fujifilm, Tokyo, Japan) while adjusting the traction direction appropriately
([Video 1]).
Fig. 2 Traction created by ENDOTORNADO. Arrows indicate the traction direction. The traction
can be obtained from any directions depending on the various situations.
Video 1 Esophageal endoscopic submucosal dissection (ESD) using ENDOTORNADO. Esophageal ESD
using ENDOTORNADO for superficial esophageal cancer after radiation therapy for thymoma.
Although the lesion was located on the scar after radiotherapy and severe fibrosis
of the submucosal layer was observed, en bloc resection was achieved safely and smoothly
without any adverse events.
Outcome measures
We collected data, such as age, sex, lesion location, occupied circumference, size,
macroscopic type, preoperative invasion depth, en bloc resection rate, R0 resection
rate, adverse events (intraprocedural perforation and delayed bleeding), specimen
size and total procedure time from patient medical records. Total procedure time was
divided into the time required for mucosal incision and submucosal dissection respectively,
and in tESD group, the time required for device setup was also included. The length
of mucosal incision line and the resection area were calculated by assuming the specimen
to be an ellipse in shape using the largest and smallest diameters of the specimen.
The mucosal incision speed was calculated by dividing the length of mucosal incision
line with the mucosal incision time. The submucosal dissection speed and the total
procedure speed were also calculated by dividing the resection area with the submucosal
dissection time and the total procedure time respectively.
Statistical analysis
We compared the clinical outcomes between cESD group and tESD group. Chi-squared test
or Fisher’s exact test were used for comparison of categorical variable, and Wilcoxon's
rank sum test was used for comparison of continuous variable. All statistical analyses
were performed using JMP version 16.2.0 (SAS Institute, Inc., Cary, North Carolina,
United States). P < 0.05 was considered statistically significant.
Results
Procedures were completely achieved in all six patients in tESD group. The ENDOTORNADO-related
adverse event was only one minor mucosal peeling of the cervical esophagus associated
with its insertion into the esophagus.
Comparing the two groups, there were no significant differences in patient and lesion
characteristics ([Table 1]). En bloc resection was achieved in all cases in both groups. R0 resection was achieved
in all cases in tESD group. No intraoperative perforation or postoperative bleeding
was observed in both groups.
Table 1
Clinical characteristics of the analyzed cases (n = 29).
|
Factors
|
cESD
(n = 23)
|
tESD
(n = 6)
|
P value
|
|
Age
|
Mean ± SD
|
69.1 ± 6.0
|
69.2 ± 11.2
|
0.91
|
|
Sex
|
Male
|
19 (82.6 %)
|
5 (83.3 %)
|
1.00
|
|
Location
|
Lt
|
13 (56.5 %)
|
3 (50.0 %)
|
1.00
|
|
Others
|
10 (43.5 %)
|
3 (50.0 %)
|
|
Horizontal location
|
Left wall
|
4 (17.4 %)
|
1 (16.7 %)
|
1.00
|
|
Others
|
19 (82.6 %)
|
5 (83.3 %)
|
|
Occupied circumference
|
< 50 %
|
17 (73.9 %)
|
6 (100.0 %)
|
0.30
|
|
≥ 51 %
|
6 (26.1 %)
|
0 (0.0 %)
|
|
Lesion size
|
Median [IQR], (mm)
|
25 [15–35]
|
23 [12–30]
|
0.57
|
|
Macroscopic type
|
0-IIc
|
16 (69.6 %)
|
3 (50.0 %)
|
0.63
|
|
Others
|
7 (30.4 %)
|
3 (50.0 %)
|
|
Preoperative invasion depth
|
EP/LPM
|
15 (65.2 %)
|
5 (83.3 %)
|
0.63
|
|
Others
|
8 (34.8 %)
|
1 (16.7 %)
|
Lt, lower thoracic esophagus; EP, epithelium; LPM, lamina propria mucosa.
The total procedure time was significantly shorter in tESD group than in cESD group
even though it included the setup time in tESD group. In particular, submucosal dissection
time was significantly reduced to about one-quarter in tESD group. The mucosal incision
speed, the submucosal dissection speed, and the total procedure speed were also significantly
increased in tESD group ([Table 2]).
Table 2
Clinical outcomes of esophageal ESD (n = 29).
|
Factors
|
cESD
(n = 23)
|
tESD
(n = 6)
|
P value
|
|
En bloc resection
|
yes, N (%)
|
23 (100.0 %)
|
6 (100.0 %)
|
–
|
|
R0 resection
|
yes, N (%)
|
20 (87.0 %)
|
6 (100.0 %)
|
1.00
|
|
Intraprocedural perforation
|
yes, N (%)
|
0 (0.0 %)
|
0 (0.0 %)
|
–
|
|
Delayed bleeding
|
yes, N (%)
|
0 (0.0 %)
|
0 (0.0 %)
|
–
|
|
Specimen size
|
Median [IQR], (mm)
|
45 [33–55]
|
41 [28–49]
|
0.29
|
|
Mucosal incision time
|
Median [IQR], (min)
|
18 [10–22]
|
8 [5–12]
|
0.010*
|
|
Setup time
|
Median [IQR], (min)
|
0 [0–0]
|
7 [7–8]
|
< 0.001*
|
|
Submucosal dissection time
|
Median [IQR], (min)
|
42 [18–57]
|
11 [4–16]
|
0.004*
|
|
Total procedure time
|
Median [IQR], (min)
|
58 [29–82]
|
26 [17–37]
|
0.015*
|
|
Mucosal incision speed
|
Median [IQR], (mm/min)
|
8 [6–11]
|
13 [12–20]
|
0.003*
|
|
Submucosal dissection speed
|
Median [IQR], (mm2/min)
|
32 [26–44]
|
76 [69–142]
|
< 0.001*
|
|
Total procedure speed
|
Median [IQR], (mm2/min)
|
23 [17–28]
|
30 [26–40]
|
0.046*
|
*Statistically significant.
Discussion
This is a first study to investigate the clinical feasibility and safety of ENDOTORNADO
in human esophageal ESD. It always provided the good traction, so each time was shortened
and each speed was increased. The significantly shorter mucosal incision time might
be due to the influence of treatment strategy in tESD group, in which a full circumferential
mucosal incision was made at the beginning, when sufficient tension was more likely
to be obtained. Considering that submucosal dissection time accounted for a large
proportion of the total procedure time, we assume that the reduction in submucosal
dissection time had a greater contribution to the reduction in total procedure time.
In addition, the operator of all cases in this study had been already an expert, but
nevertheless, the use of this new device improved treatment outcomes compared to conventional
ESD. This fact implies more potential usefulness of this device. These results suggest
the clinical feasibility and safety of this device.
Conventional ESD requires close approach to the lesion in order to obtain adequate
visualization, so it is difficult to grasp the whole image of the lesion. In contrast,
ENDOTORNADO allows the operator to obtain adequate visualization from a distance.
In other words, this device does not require moving the scope to create adequate visualization,
so it is suitable for scissors type energy devices because they also do not require
moving the scope to dissect the tissue. In fact, the operator did not need additional
submucosal injection during submucosal dissection. These facts also contributed to
the shortened total procedure time. In addition, the adjustable traction direction
is very effective in esophageal ESD. In the clip-thread method, which is one of the
most common traction method, traction is directed toward the center of the lumen due
to the traction provided by the threaded clip and the counter traction provided by
the tip hood. The clip is often placed at the center and the most oral side of mucosal
flap so traction is relatively easy in the central part, however, it is difficult
to get effective traction on the lateral edges especially for large circumferential
lesions. In addition, for lesions that are longer in the long axis direction, traction
by threaded clip tends to be in the tangential direction, so it is difficult to obtain
sufficient traction. On the other hand, ENDOTORNADO, which allows adjustment of the
traction direction freely according to the dissection area, can overcome these weaknesses
of the conventional traction method.
One of the major features of ESD using this device is that it is largely dependent
on the skill and ability of the assistant because the assistant must appropriately
adjust the traction direction. In other words, even if the operators have less experience,
they would get preferable outcome if sufficient support is given from the assistant.
Therefore, ESD using this device might contribute standardize the procedure by bridging
the gap among endoscopists with various skill level.
There are several issues with using this device. First, the created strong traction
may result in adverse events, which sometimes pull the muscle layer in a tent-like
shape, and may increase the risk of muscle layer injury and perforation especially
at the early phase of submucosal dissection when the mucosal frap is tightly attached
to the muscle layer. However, because the good traction improves visibility of the
area to be dissected and the assistant can adjust the traction force by changing the
strength of the torque, this is not a major issue by avoiding blind manipulation and
observing the running of the muscle layer closely. Second, this new device is more
expansive than traditional traction methods. Nevertheless, we would like to emphasize
that there is still room for using this device for specific selected cases, such as
difficult cases with fibrosis, circumferential or large lesions, or cases operated
by trainees, even considering its high cost.
This study has several limitations. First, this was a single-center, retrospective
study. Because the study period extended over about 3 years and was based on historical
control, there were differences in methods and characteristics of patient and lesion,
even though they are not significant differences. Second, we examined a small sample
size, so selection bias cannot be denied. Third, the operator was limited to a single
expert. In addition, the assistant had experience with more than 2500 ESDs at the
beginning of this study, so he had the ability to appropriately adjust the traction
direction. Fourth, in tESD group, the operator learning curve in use of this new device
was undeniable because he experienced several cases in a short period of time. However,
he was already an expert with sufficient experience with ESD, and the learning curve
in his usual ESD skills can be considered irrelevant. Fifth, the operator used DualKnife
J for submucosal dissection in cESD group, but he used ClutchCutter in tESD group,
so the influence of the difference in device could not be ruled out. The improved
treatment outcome was not due to ENDOTORNADO alone, but to the combination with ClutchCutter.
Sixth, we could not demonstrate the superiority over the clip-thread method because
the cESD group did not consist of cases in which the clip-thread method were used.
Conclusions
In conclusion, ENDOTORNADO creates adjustable traction from any direction and may
have clinical feasibility. It could be an option for human esophageal ESD.
