Introduction
Endoscopic submucosal dissection (ESD) has emerged as a widely accepted treatment
for complex colon lesions, particularly lesions > 20 mm or those located in anatomically
difficult positions [1]. Resection of complex colon polyps is technically challenging and time consuming.
ESD is superior to endoscopic mucosal resection (EMR) in managing complex colon polyps
with higher en bloc and curative resection rates and a lower recurrence rate [2]. However, widespread adoption of ESD is limited due to the long procedure time,
technical difficulty, lack of training facilities, and higher risk of perforation
[3]
[4]. To expedite dissection, hybrid ESD was introduced as a rescue treatment modality
in difficult ESD cases [5]. Hybrid ESD is defined as circumferential mucosal incision and partial submucosal
dissection with use of a snare for the final resection step. It has the advantage
of shorter procedure time. However, the en bloc resection rate is lower than with
standard ESD [6].
Cap-assisted ESD (ESD) is considered the standard ESD technique, where the transparent
cap assists in better visualization by keeping distance between the endoscope end
and target lesion [7]. Introduction of ESD with the double-balloon endolumenal intervention platform (DEIP)
has been proposed to facilitate technically demanding ESD procedures and reduce dissection
time [8]. DEIP has no relationship to the double-balloon enteroscopy. DEIP is an add-on device
to the endoscope. It comprises a flexible oversheath with two manually inflatable
balloons ([Fig. 1], [Fig. 2]). The double-balloon endoluminal intervention device stabilizes the colon and creates
a therapeutic zone to ensure sufficient submucosal dissection. In addition, the fore
balloon can be used to create traction using a suture and clip, which can further
expedite the dissection process ([Fig. 3], [Fig. 4]) [9]. However, the DEIP technique is in its early adoption stages and has only been described
in a case report and a small case series of porcine models [8]
[10]. Beside its use in ESD, DEIP can be used in patients with a long and tortuous colon,
as it can reduce sigmoid looping and shorten the colon. This improves the reach and
control during complex polypectomy procedures.
Fig. 1 Double-balloon endoluminal intervention platform (DEIP).
Fig. 2 Endoscopic view of the fore-balloon part of DEIP.
Fig. 3 Diagram explaining the concept of dynamic retraction using the fore balloon, suture,
and a clip.
Fig. 4 Dynamic retraction is used for submucosal dissection.
We undertook this study with two key objectives: 1) to compare en bloc resection rate
and procedure time between standard ESD and DEIP; and 2) to compare the ability to
complete the ESD procedure without resorting to hybrid ESD or piecemeal resection
in standard ESD and DEIP.
Patients and methods
Study design and patient population
The study was approved by the Institutional Review Board of Baylor College of Medicine,
Houston, Texas. This study included all patients who underwent colon ESD at Baylor
St Luke’s Medical Center, Houston, Texas, United States between September 2016 and
October 2019. Patients were referred to our center for management of colonic polyps
that were found to be: 1) impossible to resect using the en bloc method; 2) located
in an anatomically challenging position; 3) fibrotic or to have a central depression;
or 4) to be residual after prior EMR.
ESD procedures
All ESD procedures were carried out by a single experienced endoscopist (M.O.), using
a single-channel video endoscope with water jet function Pentax EC38-i10 L (Pentax
America, Montvale, New Jersey, United States). M.O. started performing ESD in 2014
and had performed more than 500 ESDs at the time of writing this manuscript. DEIP
became available in our center in September 2017, and all colon ESDs done prior to
that time were performed using a standard ESD approach (18 lesions). Training using
the DEIP was completed by performing ESD using the device in an animal model for a
total of five procedures in two training sessions. Then, we started to perform ESD
using DEIP at Baylor College of Medicine. Two DEIP procedures were performed in 2017.
In 2018, 17 colonic lesions were removed with a standard colonic ESD approach versus
five lesions removed with DEIP. In 2019, 14 colonic lesions were removed with a standard
ESD approach and 53 lesions were removed by DEIP. The choice between the approaches
was left to the endoscopist’s preference based on the availability of the device or
endoscopy staff familiar with operating the device at the time of the procedure.
For standard ESD, a transparent plastic cap (Disposable Distal Attachment, Model D-201 – 15004;
Olympus America Inc., Center Valley, Pennsylvania, United States) was fitted to the
distal end of the colonoscope. For DEIP, a U.S. Food and Drug Administration-approved
commercially available double-balloon endoluminal intervention platform (DiLumen,
Lumendi, Westport, Connecticut, United States) was used. This platform comprises a
flexible sheath with two independently inflatable balloons (fore and aft balloons).
The endoscope passes through the sheath using a gel lubricant with a 1-cm endoscope
tip projecting out of the sheath. Endoscope tip stability is ensured by the aft balloon,
while flattening of the mucosal folds with lesion retraction is provided by the fore
balloon [8].
Submucosal injection of compound solution composed of 500 CC of HESPAN (6 % hetastarch
in 0.9 % sodium chloride), 1 cc of epinephrine, 1:10,000,0.1 mg/mL, and 3 cc of methylene
blue (1:20000 mL) was used to lift the lesion-bearing mucosa. Incision of three-quarters
of the circumference with a DualKnife (Olympus America Inc., Center Valley, Pennsylvania,
United States) to reach the submucosal plane of dissection was then performed to allow
adequate retention of the submucosal injectate. The incision was performed using Endocut
Q mode (3,3,3) of Erbe VIO 200 s generator (Erbe USA Marietta, Georgia, United States).
After identification of the submucosal plane, repeated injection and dissection was
performed using Swift Coag mode (Effect 3, watt: 35) followed by completing the circumferential
incision to ensure resection of the lesion in one-piece fashion (en bloc resection
method). Whenever needed, carbon dioxide insufflation was done using a CO2 EFFICIENT Endoscopic Insufflator STERIS (Mentor, Ohio, United States). For hybrid
ESD, a snare was used for the final resection step after performing partial submucosal
dissection with the DualKnife (Olympus America, Center Valley, Pennsylvania, United
States) using dry cut current and swift coagulation mode for dissection. A coagulation
grasper (Olympus America, Center Valley, Pennsylvania, United States) was used to
control intraprocedural bleeding. After the resection was completed, the lesion was
extracted and mounted over cardboard with small pins with careful examination of the
lesion’s border. The decision to resort to hybrid ESD was made by the operator (M.O.).
Reasons for switching to hybrid ESD were: 1) lack of traction of the dissected part
of the lesion; 2) inability to stabilize the endoscope with respect to target lesion
to allow sufficient dissection and 3) expediting the procedure in case of micro-perforation
or hemodynamic instability.
Study variables collected were patient age, gender, body mass index (BMI), size and
location of the lesion, successful completion of en bloc and curative resection (yes/no),
histopathological examination of the lesion, and ESD operative and postoperative complications.
Study outcomes
En bloc resection was defined as resection of the polyp in a one-piece fashion with
macroscopic tumor-free margins. R0 resection was defined as specimens with negative
deep and lateral resection margins for adenoma/cancer. Curative resection was defined
as en bloc, R0 resection of colon polyp with < 1000 µm of submucosal invasion and
favorable histologic features [11]. Unfavorable histologic features were defined by Rex et al. as: 1) margin between
the tumor and cautery line < 2 mm; 2) invasion of the stalk; 3) poor differentiation;
4) lymphovascular invasion; and 5) inadequate orientation of the histologic sections
for pedunculated lesions. For nonpedunculated lesions, unfavorable histologic features
were defined as: 1) piecemeal resection; 2) positive resection margins; 3) invasion
depth > 1000 µm; 4) poor differentiation; 5) lymphovascular invasion; 6) tumor budding;
and 7) inadequate orientation of the histologic sections [11].
Procedure time was defined as the time interval from scope insertion to withdrawal.
Operative complication was defined as micro-perforation recognized during the ESD
procedure. Postoperative complications were defined as: 1) abdominal pain necessitating
healthcare provider visit or hospital admission; 2) bleeding that required blood transfusion
or emergency endoscopic management; and 3) perforation evident on abdominal imaging
(x-ray or computed tomography) after the ESD procedure.
Statistical analysis
Baseline demographics and operative characteristics were described as mean (± SD).
Characteristics were compared using the chi-square test and student’s t-test. P < 0.05 was considered statistically significant.
Results
Procedures characteristics
Our study cohort included 111 patients who underwent colonic ESD with 51 standard
ESD cases and 60 DEIP cases. The mean (± SD) age for the standard ESD group was 66.3
years (± 10.7) compared with 65.5 years (± 10.8) in the DEIP group. There were more
females (54.9 %) in the standard ESD group, while there were more males (53.3 %) in
DEIP group. Mean (± SD) BMI was 28.0 kg/m2 (± 4.4) in the standard ESD group compared with 29.3 kg/m2 (± 6.2) in the DEIP group. Among the standard ESD group, 72.5 % of lesions were in
the right colon and 27.5 % of lesions in the left colon with mean (± SD) lesion size
of 6.2 cm2 (± 5.5). Among the DEIP group, 81.7 % of lesions located in the right colon and 18.3 %
of lesions located in the left colon with mean (± SD) lesion size of 7.6 cm2 (± 6.0). There were no significant differences in age, gender, BMI, tumor location,
or size between the two groups ([Table 1]).
Table 1
Baseline characteristics of the study population according to assisted ESD.
|
Number (%)
|
|
|
Characteristic
|
Standard ESD (n = 51)
|
DEIP (n = 60)
|
P value
|
|
Mean (± SD) age at ESD, years
|
66.3 ± 10.7
|
65.5 ± 10.8
|
.7
|
|
|
23 (45.1)
|
32 (53.3)
|
.4
|
|
|
28 (54.9)
|
28 (46.7)
|
|
|
28.0 ± 4.4
|
29.3 ± 6.2
|
.2
|
|
|
6.2 ± 5.5
|
7.6 ± 6.0
|
.2
|
|
Histopathology
|
.4
|
|
|
22 (43.1)
|
33 (55.0)
|
|
|
5 (9.8)
|
5 (8.3)
|
|
|
6 (11.8)
|
7 (11.7)
|
|
|
2 (3.9)
|
1 (1.7)
|
|
|
2 (3.9)
|
2 (3.3)
|
|
|
8 (15.7)
|
10 (16.7)
|
|
|
0
|
1 (1.7)
|
|
|
6 (11)
|
1 (1.7)
|
|
Location of lesion
|
.3
|
|
|
37 (72.5)
|
49 (81.7)
|
|
|
14 (27.5)
|
11 (18.3)
|
ESD, endoscopic submucosal dissection; DEIP, double-balloon endoluminal intervention
platform; BMI, body mass index.
Study outcomes
Mean (± SD) procedure times were 81.9 minutes (± 35.4) and 96.4 minutes (± 42.2) in
standard ESD and DEIP groups, respectively ([Table 2]). Hybrid ESD was performed in 51.0 % and 36.7 % of patients in the standard ESD
and DEIP groups, respectively, P = .2. The en bloc resection rate was similar between the two groups with 76.5 % and
78.3 % in standard ESD and DEIP groups, respectively, P = .08. The curative resection rates was also similar between the two groups, with
68.6 % for standard ESD compared with 70.0 % for DEIP, P = 1.
Table 2
Study outcomes according to assisted ESD.
|
Number (%)
|
|
|
Outcome
|
Standard ESD (n = 51)
|
DEIP (n = 60)
|
P value
|
|
Mean (± SD) procedure time, minutes
|
81.9 ± 35.4
|
96.4 ± 42.2
|
.06
|
|
Post-operative complications
|
.2
|
|
|
46 (90.2)
|
56 (93.3)
|
|
|
2 (3.9)
|
3 (5.0)
|
|
|
3 (5.9)
|
0
|
|
|
0
|
1 (1.7)
|
|
ESD operative complications
|
1
|
|
|
51 (100.0)
|
59 (98.3)
|
|
|
0
|
1 (1.7)
|
|
|
39 (76.5)
|
47 (78.3)
|
.8
|
|
|
35 (68.6)
|
42 (70.0)
|
1
|
|
|
26 (51.0)
|
22 (36.7)
|
.2
|
ESD, endoscopic submucosal dissection; DEIP, double-balloon endoluminal intervention
platform.
Regarding complications, micro-perforation was encountered in one patient (1.7 %)
in the DEIP group and was successfully managed with clip closure during the procedure.
Post-ESD abdominal pain was encountered in 3.9 % of patients in the standard ESD group
compared with 5.0 % of patients in the DEIP group. Postoperative bleeding was found
in 5.9 % of patients in standard ESD group. However, no patients in the DEIP group
suffered postoperative bleeding.
To assess for improvement in procedure time in the DEIP group overtime, we divided
DEIP cases into two groups (first half and second half). The mean (± SD) procedure
time in the first half was 99.4 minutes (± 43.6) compared with 93.5 minutes (± 41.4)
in the second half, P = 0.6.
Secondary analysis based on hybrid ESD status
We reclassified all patients according to hybrid ESD status. The hybrid ESD group
included 48 patients, while the non-hybrid ESD group included 63 patients. Age, gender,
BMI, tumor location and size, procedure time, curative resection, operative, and postoperative
complications were comparable between the two groups ([Table 3], [Table 4]). The rate of en bloc resection was significantly higher in the non-hybrid ESD group
(88.9 %) compared to the hybrid ESD group (62.5 %), P = 0.001.
Table 3
Characteristics of the study population according to hybrid-ESD.
|
Number (%)
|
|
|
Characteristic
|
Non-hybrid ESD (n = 63)
|
Hybrid-ESD (n = 48)
|
P value
|
|
Mean (± SD) age at ESD, years
|
65.8 ± 11.9
|
65.9 ± 9.2
|
.2
|
|
|
28 (44.4)
|
27 (56.3)
|
.2
|
|
|
35 (55.6)
|
21 (43.8)
|
|
|
28.7 ± 6.3
|
28.7 ± 4.2
|
.1
|
|
|
7.5 ± 6.1
|
6.2 ± 5.4
|
.4
|
|
Location of lesion
|
.3
|
|
|
46 (73.0)
|
40 (83.3)
|
|
|
17 (27.0)
|
8 (16.7)
|
ESD, endoscopic submucosal dissection; BMI, body mass index.
Table 4
Study outcomes for hybrid-ESD.
|
Number (%)
|
|
|
Outcome
|
Non-hybrid ESD (n = 63)
|
Hybrid-ESD (n = 48)
|
P value
|
|
Mean (± SD) procedure time, minutes
|
89.5 ± 38.9
|
90.5 ± 41.3
|
.9
|
|
Postoperative complications
|
.4
|
|
|
60 (95.2)
|
42 (87.5)
|
|
|
2 (3.2)
|
3 (6.3)
|
|
|
1 (1.6)
|
2 (4.2)
|
|
|
0
|
1 (2.1)
|
|
ESD operative complications
|
.4
|
|
|
63 (100.0)
|
47 (97.9)
|
|
|
0
|
1 (2.1)
|
|
|
56 (88.9)
|
30 (62.5)
|
.001
|
|
|
43 (68.3)
|
34 (70.8)
|
.8
|
ESD, endoscopic submucosal dissection.
Non-hybrid ESD
[Table 5] shows study outcomes between the standard ESD and DEIP groups after exclusion of
hybrid ESD from both groups. Study outcomes were comparable between the two groups.
En bloc resection was higher in the DEIP group (92.1 %) compared with standard ESD
group (83.3 %), but not statistically significant, P = 0.4.
Table 5
Study outcomes for standard ESD and DEIP after exclusion of hybrid ESD.
|
Number (%)
|
|
|
Outcome
|
Standard ESD (n = 24)
|
DEIP (n = 38)
|
P value
|
|
Mean (± SD) procedure time, minutes
|
82.6 ± 36.9
|
93.8 ± 39.9
|
.3
|
|
Postoperative complications
|
.2
|
|
|
23 (95.8)
|
36 (94.7)
|
|
|
0
|
2 (5.3)
|
|
|
1 (4.2)
|
0
|
|
|
0
|
0
|
|
ESD operative complications
|
–
|
|
|
24 (100.0)
|
38 (100.0)
|
|
|
0
|
0
|
|
|
20 (83.3)
|
35 (92.1)
|
.4
|
|
|
16 (66.7)
|
26 (68.4)
|
1
|
ESD, endoscopic submucosal dissection; DEIP, double-balloon endoluminal intervention
platform.
Discussion
This study demonstrated a similar procedure time, en bloc and curative resection rates
between DEIP and standard ESD. There is a lower chance of switching to hybrid ESD
in the DEIP group, though the difference did not reach statistical significance. Non-hybrid
ESD had a significantly higher en bloc resection rate compared with hybrid ESD. These
data underscore the importance of DEIP in decreasing the chance of switching to hybrid
ESD, which in turn increases the en bloc resection rate in complex colon polyps and
accordingly, the local recurrence rate [12]. En bloc resection is the goal of every ESD procedure, ensuring the removal of colonic
polyps with tumor-free margins and higher rates of curative resection [13].
ESD is a technically challenging and time-consuming procedure, necessitating ongoing
innovations to overcome its complexity. DEIP has emerged as a new tool to ease the
challenging nature of ESD through counter-traction to expose the dissection field.
However, published data are scare as it is a newly developed platform [8]. Our experience revealed the importance of DEIP in straightening the mucosa around
colonic polyps, especially in redundant intestinal segments and anatomically difficult
and poorly visualized lesions. We believe such factors constitute a major share of
the ESD procedure complexity. One of the specific limitations of colonic ESD is the
ability to obtain a stable position, which allows steady submucosal dissection. Pocket-creation
ESD was proposed as a solution to the instability associated with colonic ESD. In
a study of 887 colonic lesions treated by ESD, the pocket-creation method (PCM) was
associated with higher en bloc (100 %) and R0 Resection rates (91 %) compared with
standard ESD [14]. DEIP provides the stability required to perform ESD without the need for creating
a pocket, which may require greater expertise in certain locations such as lesions
behind folds or in fibrotic lesions. Dynamic retraction using a clip to attach the
dissected distal lesion margin to a suture mounted on the fore balloon of the device
was crucial in expediting dissection in fibrotic cases and large lesions ([Video 1]).
Video 1 Sessile serrated adenoma with severe fibrosis due to prior EMR and tattooing removed
with DEIP using dynamic retraction method.
No added benefit of DEIP vs. standard ESD regarding procedure time was observed in
our study. This was attributed to a longer time taken to inflate the fore and aft
balloons to set up the tissue traction. Once the double-balloon endoluminal intervention
platform is settled, dissection is easy to perform with improved visualization and
mucosal stability.
Regarding operative and postoperative complications, rates of abdominal pain and micro-perforation
were slightly higher in DEIP compared with standard ESD. These could be explained
by the counter-traction imposed on the submucosal tissue by the double-balloon endoluminal
intervention platform exposing the vulnerable muscle layer to micro-perforation during
dissection [8]. Moreover, post-ESD electrocoagulation syndrome may be the underlying cause for
abdominal pain after the procedure [15].
In a Japanese retrospective study, the en bloc resection rate was lower in hybrid
ESD compared with non-hybrid ESD at 66.7 % and 94.2 %, respectively [6]. The Japanese en bloc resection rates were slightly higher than our results (62.5 %
in hybrid ESD and 88.9 % in non-hybrid ESD). Our outcomes were lower compared with
the Japanese outcomes. The difference could be explained by a steep learning curve
for colonic ESD, high BMI in our populations making colonoscopy and endoscopic dissection
more challenging, and the high prevalence of fibrosis in referred lesions in the west
due to aggressive sampling, tattooing directly under the lesions or attempts that
resulted in incomplete polypectomies. On the other hand, our procedure time was shorter
than in the Japanese study for non-hybrid ESD. The above-mentioned Japanese cohort
procedure time for non-hybrid ESD was 122 minutes (± 72.2) compared with 89.5 minutes
(± 38.9) in our study. They defined procedure time from first mucosal incision to
completion of hemostatic treatment after submucosal dissection. However, we defined
the procedure time from scope insertion to withdrawal. It is worth mentioning that
hybrid ESD was carried out as a rescue procedure in long ESD procedures with inadequate
dissection, resulting in lower rates of en bloc resection [6].
Our study has limitations associated with being retrospective and non-randomized.
Potential bias may be present given the preference for DEIP in overweight and obese
patients, anatomically challenging polyps, and patients with a history of difficult
colonoscopies due to redundant intestinal segments. This study is the first to report
resection outcomes in patients who have undergone DEIP compared with standard ESD.
It is worth mentioning that DEIP is associated with a higher cost for ESD. However,
DEIP could become cost-effective if it enables the endoscopist to perform ESD safely
and decreases the number of referrals for surgical resection of colon polyps. Further
data are needed on the cost-effectiveness of this novel device.
Conclusion
In conclusion, similar procedure time, en bloc and curative resection rates were observed
in both the DEIP and standard ESD groups. The odds of switching to hybrid ESD in DEIP
were lower compared to standard ESD. The introduction of DEIP may create a unique
opportunity for reducing the complexity of ESD and encouraging its widespread use.
A multicenter, randomized, controlled trial is needed to demonstrate impactful outcomes
with DEIP.
