Keywords Endoscopy Lower GI Tract - Polyps / adenomas / ... - Tissue diagnosis - Endoscopic
resection (polypectomy, ESD, EMRc, ...)
Introduction
Colorectal polyps are the main source of colorectal cancer. Endoscopic removal can
significantly reduce colorectal cancer incidence [1 ]. However, 10% to 27% of colorectal cancers are caused by incomplete polyp resection;
complete removal of precancerous colorectal lesions thus is crucial for preventing
colorectal cancer. Most polyps found during colonoscopy are small (< 10 mm), with
extremely low malignancy rates (0.1%) [2 ]. Removing small colorectal polyps is a primary concern for clinicians and endoscopists.
Cold snare polypectomy (CSP) offers better efficacy and safety than with other methods
[3 ]
[4 ]
[5 ]
[6 ] and has become the gold standard for removal of small colorectal polyps (1–9 mm).
Electrocautery is not used; therefore, the incidences of postoperative bleeding, post-polypectomy
syndrome, and perforation are reduced and operation and resection times are shorter,
making it a popular technique in clinical practice. However, hot snare polypectomy
(HSP) is associated with a high R0 rate [7 ]
[8 ]
[9 ] of 47.3% to 98.2% [5 ]
[7 ]
[10 ]. The average submucosa depth was significantly lower in CSP specimens than in HSP
and endoscopic mucosal resection (EMR) specimens [11 ]. Moreover, there is an 11.3% to 36% chance of cold snare defect protrusion (CSDP)
after CSP, indicating polyp fragmentation and incomplete mucosal layer resection [12 ]
[13 ]. Therefore, traditional CSP requires further improvement, and more effective and
safe removal methods for small colorectal polyps are needed.
Traditional CSP uses air as the intestinal cavity expansion medium. However, because
of gravity, the true polyp shape may shrink into a ball, resulting in an unclear polyp
boundary display. Consequently, when capturing the polyp, the snare may not achieve
a clear view of the cutting-edge margin, which may cause residual tumor tissue on
the margin. Polypectomy under underwater endoscopy also has good therapeutic effects
[14 ]. Nevertheless, most studies have focused on larger lesions and require auxiliary
measures such as submucosal injection or electrocautery [15 ]. We hypothesized that underwater CSP (UCSP) could use the magnification and lifting
effect of water to clearly display the lesion boundary, remove more of the muscularis
mucosa, and improve the R0 rate for small colorectal polyps. Thus, in this study,
we aimed to compare UCSP and traditional CSP for small colorectal polyp removal regarding
the R0 rate.
Patients and methods
Ethics statement
This study was approved by the Institutional Review Board of Jianyang People’s Hospital
(approval number 2021 (08)) and registered in the Chinese Clinical Trial Registry
(ChiCTR2100044872). All patients provided informed consent.
Study design
In this prospective study 662 patients with colorectal polyps who underwent colonoscopies
at Jianyang People’s Hospital (Sichuan, China) between April 2022 and October 2023
were randomized.
Inclusion and exclusion criteria
Inclusion criteria were as follows: age 18 to 75 years, single or multiple colorectal
polyps measuring 4 to 9 mm in diameter (Paris classification: 0-Ip, 0-Is, 0-Isp, and
0-IIa), and compliance with narrow-band imaging international colorectal endoscopic
type 2 classification.
Exclusion criteria were as follows: age < 18 years; presence of familial polyposis
or inflammatory bowel disease; history of acute myocardial infarction within the past
6 months; severe heart, liver, kidney dysfunction, or mental disorder; use of aspirin,
warfarin, and other anticoagulants or presence of coagulation dysfunction; polyp diameter
≥ 1 cm; participation in other clinical trials within the past 60 days; and pregnancy
or sessile serrated lesions.
Equipment and preoperative preparation
All participants started a liquid diet 1 day before colonoscopy and underwent bowel
preparation using 2 L of polyethylene glycol electrolyte powder combined with simethicone.
Equipment included a high-definition magnifying endoscope system (CV-290, Olympus,
Tokyo, Japan), a conventional endoscope (CF-H290I, PCF-Q260AZI, and PCF-Q260JI, Olympus),
biopsy forceps (AMHBFE2.4×1800, Anrei Medical, Zhenjiang, China), and a snare (AMH-SNER241518,
Anrei Medical).
Randomized monitoring
It was impossible to determine whether patients had 4- to 9-mm colorectal polyps before
colonoscopy; therefore, a sealed opaque envelope containing a computer-generated random
code was opened by the assistant. The doctor was informed of the polypectomy method
upon encountering a polyp that met the criteria during the colonoscopy. When multiple
eligible polyps were identified in the same patient, the same method was consistently
applied for all polypectomies. Lesions that did not meet the inclusion criteria were
removed using conventional endoscopic treatment. The pathologist was blinded to the
method of endoscopic treatment.
Endoscopic procedure
All operators performing endoscopic procedures were experienced professional endoscopists.
Each operator had conducted > 5,000 colonoscopies and was proficient in techniques
such as CSP, EMR, and endoscopic submucosal dissection. The UCSP procedure was as
follows. After finding the target lesion, the digestive tract cavity was filled with
sterile water (22 ± 2°C). The polyp was maneuvered to the 6 o’clock position to the
extent possible, and a snare was used to capture it (with approximately 1 to 2 mm
of normal surrounding mucosa). The snare was gradually reduced until closure was complete.
No electrocoagulation was required; instead, the polyp was directly removed by gently
pressing the snare against the intestinal wall, which facilitated complete removal
(multiple snare resections were not allowed). Subsequently, the endoscope suction
button was removed and the suction hole was covered with a finger to prevent tissue
damage caused by the suction button. The excised polyp tissue was then suctioned through
the endoscope suction channel and collected in a specimen bottle for pathological
examination. Following this, biopsies were performed at the base and edge of the polypectomy
site in four quadrants—one sample from the base and one from each of the four symmetrical
edge points. If a CSDP was observed at the base, it was sampled as a basal biopsy
specimen and placed in a separate specimen bottle for examination ([Fig. 1 ]). Polyps in the CSP group were removed using conventional CSP, and pathological
specimens were obtained following the same protocol as in the UCSP group.
Fig. 1 Underwater cold snare polypectomy procedure. a .
A 5-mm colonic polyp is seen under white light. b After the
polyp is submerged in water, the surrounding mucosa is smooth, the edges are clear,
the
light reflection disappears, and the microstructure observed by narrow-band imaging
becomes clear. c The polyp and the surrounding normal mucosa
are captured by a snare. d Biopsy specimens are taken from the
edges and base of the incision to determine whether any tumor remains. e The pathological diagnosis is low-grade dysplasia, which contains more
muscularis mucosa. f Pathology of the excisional margin biopsy
shows no residual tumor cells.
Basic characteristics of patients and polyps
We evaluated the following indicators: general patient characteristics (age and sex),
intestinal score, polyp characteristics (size, location, and Paris classification),
and pathological diagnosis results.
Baseline characteristics of patients included in the UCSP and CSP groups were similar
in terms of number of polyps detected, sex, age, and Boston Bowel Preparation Scale
score ([Table 1 ]). The assigned polyps exhibited similar background characteristics.
Table 1 Baseline patient characteristics.
UCSP
CSP
P value
BBPS, Boston Bowel Preparation Scale; CSP, cold snare polypectomy; SD, standard deviation;
UCSP, underwater cold snare polypectomy.
Patients (n)
89
83
Sex, n (%)
0.175
54 (60.7)
57 (68.7)
35 (39.3)
26 (31.3)
Median age (range, years)
56 (51–62)
57 (51–63)
0.385
BBPS, median (range)
7 (6–7)
7 (6–7)
0.529
Approximately 60% of polyps were classified as type 0-I, whereas 40% were classified
as type 0-II. The number of polyps was higher in the left colon than in the right
colon. Average size was similar, with polyps measuring 6.2 mm and 5.7 mm in the UCSP
and CSP groups, respectively. Over 95% of biopsy types in both groups were low-grade
dysplasia. Postoperative pathological examination of biopsy specimens showed that
both groups had high-grade dysplasia (HGD; UCSP, n = 2; CSP, n = 5) and HPs (UCSP,
n = 3; CSP, n = 2), with no significant differences between the groups ([Table 2 ]).
Table 2 Polyp characteristics.
UCSP
CSP
P value
CSP, cold snare polypectomy; HGD, high-grade dysplasia; LGD, low-grade dysplasia;
SD, standard deviation; UCSP, underwater cold snare polypectomy.
Morphology, n (%)
0.733
37 (30.3)
35 (33.3)
25 (20.5)
26 (24.8)
3 (2.5)
2 (1.9)
57 (46.7)
42 (40)
Location, n (%)
0.264
3 (2.5)
2 (1.9)
10 (8.2)
17 (16.2)
24 (19.7)
27 (25.7)
22 (18.0)
15 (14.3)
44 (36.1)
27 (25.7)
19 (15.6)
17 (16.2)
Size (mm), n (%)
0.212
33 (27.0)
10 (9.5)
38 (31.1)
32 (30.5)
24 (19.7)
39 (37.1)
7 (5.7)
16 (15.2)
14 (11.5)
6 (5.7)
6 (4.9)
2 (1.9)
Size median (range, mm)
5 (5–7)
6 (5–6)
0.319
Histopathological diagnosis, n (%)
0.172
118 (96.7)
100 (95.2)
2 (1.6)
0 (0.0)
2 (1.6)
5 (4.8)
Outcomes
The primary study endpoint was R0 polyp resection, meaning that pathological results
of the biopsy specimens at the base and edge of the wound after polypectomy were all
negative [9 ]. When one or more tumor tissues were found in the biopsy specimen at the base or
edge of the polyp specimen, it was defined as an incomplete resection (non-R0 resection).
Secondary endpoints included resection time (UCSP group: from water injection initiation
to specimen retrieval; CSP group: from snare insertion to specimen retrieval), perforation
rate (intraoperative and postoperative), bleeding rate (intraoperative [continuous
bleeding for ≥ 30 seconds after polypectomy] and postoperative [visible hematochezia
within 30 days]), polyp retrieval rate, polyp integrity rate, and CSDP occurrence
rate. Resected polyps and biopsy specimens were independently assessed by two pathologists
using a blinded method. For disagreements, the pathology was reviewed until a consensus
diagnosis was reached. All patients were followed up at our Department of Gastroenterology
outpatient clinic 1 week and 1 month after the operation, during which postoperative
bleeding and other adverse events (AEs) were recorded.
Sample size calculation and statistical analysis
The reported range of CSP R0 resection rates for 4- to 9-mm colorectal polyps is 47.3%
to 98.2%. Considering clinical reality and expected improvement, we set the CSP R0
resection rate at 72% and hypothesized that the UCSP group could improve to 90%. With
an α error level set at 0.05 and controlling the β error level within 0.10, we calculated
that at least 98 polyps per group were required. In addition, to account for potential
data loss or polyps not meeting the analysis criteria (assuming a 10% exclusion rate),
we conservatively increased the sample size for the two groups to a total of 218 polyps
to ensure adequacy and robustness of the study.
Numerical data are presented as means ± standard deviations or medians and interquartile
ranges. The χ2 test was used to examine relationships between categorical variables,
while t -tests or Mann-Whitney U tests were employed to assess
differences between continuous variables. All statistical analyses were performed
using SPSS
Statistics for Windows (version 29.0. IBM Corp., Armonk, New York, United States).
A
significance level of 0.05 was set.
Results
Patient recruitment and screening process
Between April 2022 and October 2023, 662 patients who underwent colonoscopy were enrolled.
Of them, 180 had 235 polyps that met the inclusion criteria. Of these, 92 patients
with 125 polyps were assigned to the UCSP group and 88 patients with 110 polyps were
assigned to the CSP group. However, three polyps were excluded from the UCSP group
because they were hyperplastic. In the CSP group, five polyps were excluded because
of retrieval failure (n = 2), sessile serrated adenoma/polyps (n = 1), or hyperplastic
polyps (HPs; n = 2). The final analysis included 89 patients with 122 polyps in the
UCSP group and 83 patients with 105 polyps in the CSP group ([Fig. 2 ]).
Fig. 2 Study flowchart. CSP, cold snare polypectomy; UCSP, underwater cold snare
polypectomy.
Study endpoints
The R0 resection rate (primary endpoint) was significantly higher in the UCSP group
than in the CSP group (96.7% vs. 86.7%, P = 0.005), with a higher R0 rate at the base (100% vs. 91.4%, P = 0.001); this indicated superior efficacy of UCSP for complete polyp tissue removal.
Correspondingly, the resection rate of the muscularis mucosa was significantly higher
in the UCSP group than in the CSP group (68.9% vs. 43.8%, P ≤ 0.0001). There was no significant difference in the submucosal resection rate between
the two groups (19.7% vs. 16.2%, P = 0.307; [Table 3 ]). Furthermore, procedure time was comparable between the UCSP and CSP groups (109.5
seconds [86.8–134.3] vs. 110.0 seconds [83.5–143.5], P = 0.890). There were no significant between-group differences in terms of AEs, including
intraoperative bleeding, postoperative bleeding, and perforation, and no serious AEs
occurred. The polyp retrieval rate in the UCSP group was not significantly different
from that in the CSP group (100% vs. 98.2%, P = 0.218). This indicated that UCSP excels in retrieving resected tissue and preserving
the integrity of the polyp specimen, facilitating pathological analysis. There was
no significant between-group difference in the proportion of CSDPs (20.5% vs. 14.3%,
P = 0.147); this indicated that UCSP did not increase risk of polyp fragmentation and
residue ([Table 4 ]).
Table 3 Primary endpoint.
UCSP
CSP
Risk difference
(95%CI)
P value
CI, confidence interval; CSP, cold snare polypectomy; MM, muscularis mucosa; SM, submucosal;
UCSP, underwater cold snare polypectomy.
R0 resection, n (%)
118 (96.7)
91 (86.7)
4.5 (1.4–14.3)
0.005
122 (100)
96 (91.4)
0.4 (0.38–0.51)
0.001
83 (68.9)
46 (43.8)
2.7 (1.6–4.7)
< 0.0001
24 (19.7)
17 (16.2)
1.3 (0.6–2.5)
0.307
118 (96.7)
99 (94.3)
1.7 (0.5–6.5)
0.285
Table 4 Secondary endpoints.
UCSP
CSP
Risk difference
(95%CI)
P value
CI, confidence interval; CSP, cold snare polypectomy; CSDPs, cold snare defect protrusions;
UCSP, underwater cold snare polypectomy.
Polyp retrieval rate, n (%)
125 (100)
108 (98.2)
0.5 (0.4–0.5)
0.218
Polyp integrity rate, n (%)
1.6 (0.5–4.8)
0.285
116 (95.1)
97 (92.4)
6 (4.9)
8 (7.6)
CSDPs, n (%)
25 (20.5)
15 (14.3)
1.5 (0.8–3.1)
0.147
Median resection time (s)
109.5 (86.8–134.3)
110.0 (83.5–143.5)
0.890
Adverse events, n (%)
7 (5.7)
5 (4.8)
1.2 (0.4–4.0)
0.491
0
0
1.000
0
0
1.000
Discussion
Our findings indicate that UCSP has a higher R0 rate and overall efficacy than conventional
CSP for colorectal adenomas, suggesting that UCSP can serve as a viable modality for
resecting colorectal adenomas measuring 4 to 9 mm.
The low R0 resection rate in CSP can be primarily attributed to its inability to remove
sufficient muscularis mucosa [16 ]. In specimens removed by CSP, the removal rate for the muscularis mucosa ranges
from 35.3% to 57.0% [11 ]
[17 ], whereas the submucosa inclusion rate is 9% to 29% [11 ]
[18 ]; both rates are lower than those for HSP and EMR [11 ]. In this study, the R0 resection rate was lower in the CSP group than in the UCSP
group, and the CSP group mainly had residual tumors at the base. This result is similar
to that of the study by Maruoka et al [17 ], wherein UCSP improved the R0 resection rate by removing more muscularis mucosa.
We speculated that because of the lower density of the polyp relative to that of water,
polyp immersion in water resulted in a similar “buoyancy” effect on the mucosal layer
and submucosa surrounding the lesion. This allowed the snare to remove more mucosa
and reduce the residual lesion at the base. Moreover, we observed that after water
injection, intestinal peristalsis slowed, and the intestinal mucosa around the polyp
was stretched by the water pressure, causing the polyp to be “lifted” by the water.
This resulted in a clear display of the polyp boundary and a comprehensive view of
the polyp, thereby preventing misjudgments caused by polyp curling and incomplete
snaring resulting in residual lesions. In UCSP, the snare can clearly capture an ample
amount of normal mucosa surrounding the polyp, thereby minimizing likelihood of residual
tissue at the cutting edge. Therefore, compared with CSP using air as the intestinal
expansion medium, UCSP can reduce residue at the base and cutting edge, consequently
improving the R0 resection rate.
As a new method of endoscopic treatment, UCSP requires more water-injection steps
than CSP, which potentially prolongs procedure time. However, our study revealed no
significant difference in the procedure time between the UCSP and CSP groups, possibly
because water flows through the intestinal cavity, and some polyps were close to one
another. With a single water injection, multiple polyps can be consecutively removed.
Moreover, once resected underwater, the polyp floats, making retrieval more efficient.
The recovery rate for colorectal polyp specimens after CSP was 97.0% to 98.1% [3 ]
[19 ], and it was challenging to fully recover specimens < 10 mm in diameter. Incomplete
and unrecovered specimens cannot provide an accurate pathological diagnosis [20 ]. Studies have reported that 4% to 60.3% of polyp specimens are fragmented after
CSP [12 ]
[20 ]
[21 ]
[22 ]. Studies have indicated that sampling colorectal polyp specimens directly through
the endoscope working channel may lead to tissue damage due to improper handling,
thereby affecting accuracy of histological assessment of polyp margin [23 ]. In our practical experience, we have observed that some polyps tend to become lodged
at the suction button during aspiration, increasing risk of incomplete specimens.
To address this issue, we retrieved our specimens, we removed the suction button and
used the fingertip occlusion method to retrieve the specimens while trying to keep
them as intact as possible. Polyp recovery and integrity rates were higher in the
UCSP group than in the CSP group. For polyps with unclear edges or unrecovered specimens,
doctors suggest repeat colonoscopy. This situation was reduced in the UCSP group,
sparing patients from undergoing unnecessary colonoscopies.
The intraoperative bleeding rate for CSP was 0.4% to 9.6% [4 ]
[5 ]
[10 ], and delayed bleeding was significantly reduced after CSP compared with HSP [4 ]. UCSP did not require high-frequency electrocautery to remove polyps; therefore,
it did not cause more complications, such as delayed bleeding and perforation, than
HSP [5 ]. To ensure a negative cutting-edge margin, we removed 1 to 2 mm of normal mucosa
surrounding the polyp. However, this method did not increase risk of intraoperative
or postoperative bleeding, perforation, or other complications. This was consistent
with the results of Abe et al., who found that CSP with an enlarged cutting-edge margin
did not increase patient risk [24 ]. No patient in the study required endoscopic hemostasis, and none experienced perforations.
Intraoperative and postoperative bleeding rates were similar between the UCSP and
CSP groups, indicating no increase in likelihood of complications. Furthermore, incision
bleeding was promptly and clearly identified underwater. In cases in which bleeding
persisted, endoscopic metal clip closure of the wound could be performed promptly
to effectively stop bleeding and reduce risk of delayed bleeding. Therefore, our results
demonstrate that UCSP is safe.
CSDPs often occur during polyp cold snare resections. When using the CSP method to
remove colorectal polyps, probability of CSDP occurrence is 11.3% to 36%, and presence
of CSDPs indicates polyp fragmentation and incomplete mucosal layer resection [12 ]
[13 ]. Because of the buoyancy effect of water on polyps, UCSP removes a deeper mucosal
layer, which may increase the occurrence rate of CSDPs. Our findings showed that incidence
of CSDPs was not significantly higher in the UCSP group than in the CSP group. Furthermore,
CSDPs did not contain tumor tissue, and their occurrence did not reduce the R0 resection
rate. Although the fragmentation rate of the CSDP specimens was higher, the R0 resection
rate did not differ between CSDP and non-CSDP specimens [12 ]. Therefore, UCSP does not increase risk of tumor residues or the CSDP occurrence
rate.
This study had some limitations. First, it was a single-center, prospective study.
Therefore, it is necessary to increase the sample size, expand the research scope,
conduct multicenter randomized controlled trials, and perform long-term clinical follow-up.
Second, UCSP requires a large amount of water injection, and the total procedure time
is similar to that for CSP. More efficient water-injection equipment could reduce
procedure time.
Conclusions
In conclusion, UCSP had a higher R0 resection rate than conventional CSP for colorectal
adenomas measuring 4 to 9 mm. In addition, this endoscopic method did not increase
the rate of intraoperative and postoperative bleeding, perforation, or other complications.
Therefore, UCSP is a safe and effective technique for resecting small colorectal polyps.
Bibliographical Record
