Keywords Endoscopy Upper GI Tract - Precancerous conditions & cancerous lesions (displasia
and cancer) stomach - Endoscopic resection (ESD, EMRc, ...) - Polyps/adenomas/ ...
- Colorectal cancer
Introduction
Colorectal cancer (CRC) is the third most common cause of cancer worldwide and the
third leading cause of cancer deaths in Western countries. [1 ] Colonoscopy remains a commonly performed screening test for CRC as it has both diagnostic
and therapeutic capabilities. Studies have shown a strong association between screening
colonoscopy and a reduced risk of death from colorectal cancers [2 ]. Endoscopic mucosal resection (EMR) is the standard method for removing sessile
colorectal polyps larger than 10 mm [3 ]
[4 ]. Conventional EMR (C-EMR) involves filling the colonic lumen with air or CO2 followed by submucosal injection of fluid underneath the polyp to decrease the risk
of causing accidental transmural thermal injury leading to perforation [5 ]. It has been shown to have high therapeutic success rates and allows for a 90% reduction
in the need for long-term surgery [6 ]. However, with piecemeal C-EMR, local recurrence rates of > 30% have been described,
which is a clear disadvantage of the procedure, forcing the use of alternative techniques
to reduce these figures [7 ]
[8 ]. Furthermore, high recurrence rates have prompted the application of thermal ablation
to the EMR defect which adds both time and cost to the procedure [9 ].
Underwater EMR (U-EMR) is a technique, described a decade ago, in which the colon
is filled with water instead of air/CO2 , decreasing colonic wall tension and resulting in potential benefits that may improve
the opportunity for safe and complete en-bloc resection of mucosal lesions [10 ]. The technique eliminates the need for submucosal injection, because of the "floating"
effect of water submersion on the mucosa and submucosa, resulting in separation from
the muscularis propria [11 ]
[12 ]. Several systematic reviews and meta-analysis have attempted to compare outcomes
of U-EMR with C-EMR, however the strength of evidence remains low, as most of these
include observational, retrospective cohort studies [13 ]
[14 ]
[15 ]
[16 ].
Given these limitations of existing literature, we conducted an updated systematic
review and meta-analysis to evaluate the efficacy and safety of U-EMR in comparison
to C-EMR for colorectal lesions, assessing data only from published randomized controlled
trials (RCTs).
Methods
Search strategy
The literature was searched by a medical librarian for the concepts of RCTs, underwater
EMR and colorectal polyps. Search strategies were created using a combination of keywords
and standardized index terms. Keywords included “endoscopic mucosal resection,” “EMR,”
“colorectal lesions” and “underwater EMR” along with phrases associated with the procedure
such as “colonoscopy”. Searches were run on November 25, 2022 in Ovid Cochrane Central
Register of Controlled Trials (1991+), Ovid Embase (1974+), Ovid Medline (1946+ including
epub ahead of print, in-process & other non-indexed citations), Scopus (1823+), and
Web of Science Core Collection (Science Citation Index Expanded 1975+ & Emerging Sources
Citation Index 2015+). After limiting results to English language with some pediatric
and animal studies removed, a total of 178 citations were retrieved. Deduplication
was performed in EndNote following the Bramer method [17 ] leaving 85 citations. Manual search for studies of interest was performed in google
scholar by two authors (SC, DSD). Details of study selection are provided in PRISMA
Flow Chart – Supplementary Fig. 1 . [18 ] The full search strategy is available in Supplementary Appendix-1. The PRISMA checklist
was followed and is provided as Supplementary Appendix-2. Reference lists of evaluated
studies were examined to identify other studies of interest.
Study selection
In this meta-analysis, we included only RCTs where outcomes of C-EMR and U-EMR were
provided. Studies were included irrespective of inpatient/outpatient setting, follow-up
time, geography and whether published as full manuscripts or abstracts, as long as
they provided the clinical outcomes data needed for the analysis.
Our exclusion criteria were as follows: 1) cohort studies reporting outcomes of U-EMR
only, C-EMR only and/or U-EMR vs C-EMR; 2) studies that included less than 25 patients;
3)
studies performed in the pediatric population (age < 18 years); and 4) studies not
published in English language. In cases of multiple publications from a single research
group reporting on the same patient, same cohort and/or overlapping cohorts, data
from the
most recent and/or most appropriate comprehensive report were retained. The retained
studies
were decided by two authors (SC, BPM) based on the publication timing (most recent)
and/or
the sample size of the study (largest).
Data abstraction and quality assessment
Data on study-related outcomes from the individual studies were abstracted independently
onto a standardized form by at least two authors (SC, DR). Author DSD cross-verified
the collected data for possible errors and two authors (SC, JB) did the quality scoring
independently. The Cochrane Collaboration tool to assess risk of bias (Supplementary
Appendix-3) [19 ]. The quality of evidence presented in the RCTs was assessed using the Grading of
Recommendations Assessment, Development and Evaluation (GRADE) methodology (Supplementary Fig. 2 ) [20 ].
Outcomes assessed
Patients were classified into two cohorts based on the technique used – underwater
EMR (U-EMR) and conventional EMR (C-EMR).
The primary outcomes were: 1) pooled risk ratio and proportions of R0 resection, defined
as a complete en-bloc resection of a lesion with tumor-free lateral and vertical margins
[21 ]; 2) pooled risk ratio and proportions of en-bloc resection; and 3) pooled risk ratio
and proportions of lesion recurrence, defined as an adenoma or cancer at the resection
site on follow-up colonoscopy.
The secondary outcomes were: 1) pooled risk ratio and proportions of piecemeal resection;
2) pooled risk ratio and proportions of incomplete resection, defined as at least
one neoplastic tissue retrieved from the resection edge after polypectomy, presence
of any adenomatous or serrated pathology in the biopsy specimen or polyps requiring
complementation with thermal ablation after resection attempt; 3) mean difference
in resection time, defined as period between the start of submucosal injection (in
the C-EMR group) or intra-intestinal water injection (in the U-EMR group) and completion
of the colonoscopic resection; and 4) pooled risk ratio and proportions of adverse
events including perforation, immediate and delayed bleeding.
Statistical analysis
We used meta-analysis techniques to calculate the pooled estimates and 95% CIs (confidence
intervals) in each case following the methods suggested by DerSimonian and Laird using
the random-effects model [22 ]. When the incidence of an outcome was zero in a study, a continuity correction of
0.5 was added to the number of incident cases before statistical analysis [23 ]. The Mantel-Haenszel-type method was used to estimate the pooled odds ratio for
all outcomes. [24 ] Heterogeneity between studies was assessed by means of a χ
2 test (Cochran Q statistic) and quantified with the I2 statistics. In this, values of < 30%, 30% to 60%, 61% to 75%, and > 75% were suggestive
of low, moderate, substantial, and considerable heterogeneity, respectively. Publication
bias was ascertained, qualitatively, by visual inspection of funnel plot and quantitatively,
by the Egger test [25 ]. When publication bias was present, further statistics using the fail-Safe N test
and Duval and Tweedie’s ‘Trim and Fill’ test was used to ascertain the impact of the
bias [26 ]. P < 0.05 was considered statistically significant while comparing the two groups.
All analyses were performed using Comprehensive Meta-Analysis (CMA) software, version
3 (BioStat, Englewood, New Jersey, United States).
Results
Search results and population characteristics
From an initial pool of 468 studies, 257 records were screened after reduplication,
81 full-length articles were assessed. A total of seven RCTs with 1458 patients (U-EMR:
739, C-EMR: 719) were included in the final analysis. Overall, 865 (177, > 20 mm)
and 826 (192, > 20 mm) polyps were resected using U-EMR and C-EMR, respectively. Mean
age ranged from 55.1 to 70 years. Further details along with the population characteristics
are described in [Table 1 ] and [Table 2 ].
Table 1 Study details and population characteristics.
Study
Design
Total patients
Gender
Mean age
Polyp morphology
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR
C-EMR
NR, not reported; IQR, interquartile range; U-EMR, underwater endoscopic mucosal resection;
C-EMR, conventional endoscopic mucosal resection.
Zhang, 2020
Prospective, parallel-group, open-label, randomized controlled trial (RCT), May 2019
to November 2019, multicenter, China
66
64
40/26
35/29
55.1 (11.2)
57.6 (9.8)
0-Is (50), 0-Ip (0), 0-IIa (21)
0-Is (54), 0-Ip (1), 0-IIa (16)
Yen, 2020
Prospective, randomized controlled trial (RCT), October 2016 to September 2018, single-center,
United States
128
127
123/5
125/2
64.4 (8.3)
64.6 (8.3)
Is 126, IIa 107, IIb 11
Is 115, IIa 88, IIb 4
Yamashina, 2019
Prospective, randomized controlled trial (RCT), multicenter, Japan
108
102
64/44
75/27
70 (43–86)
68 (42–95)
Ip 2, Is 41, IIa 64, IIc 1
Ip 0, Is 44, IIa 58, IIc 0
Nagl, 2021
Prospective, randomized controlled trial (RCT), August 2017 to October 2020, single-center,
Germany
81
76
51/30
52/24
68.1 (9.6)
66.3 (11.9)
Is 14, IIa 54, IIb 3, IIc 0, 0-lla/0-llc 0, 0-lla/0-ls 10
Is 20, IIa 49, IIb 2, IIc 1, 0-lla/0-llc 1, 0-lla/0-ls 3
Lenz, 2022
Prospective, randomized controlled trial (RCT), April 2017 to March 2022, multicenter,
Brazil
53
52
25/28
22/30
64.4
64
--
--
Hamerski, 2019 (Abs)
Prospective, randomized controlled trial (RCT), multicenter, United States and Italy
158
145
NR
NR
NR
NR
NR
NR
Rodríguez Sánchez, 2022
Prospective, randomized controlled trial, multicenter February 2018 to February 2020,
Spain
145
153
87/58
96/57
67.5 (10.4)
67.2 (9.8)
Is 74, IIa 65, IIc 10
Is 66, IIa 80, IIc 16
Table 2 Study outcomes.
Study
Total polyps
Polyp size (mm)
Outcomes
R0 resection
Incomplete resection
En-bloc resection
Resection time
Perforation
Immediate bleeding
Delayed bleeding
Piecemeal resection
Recurrence
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR
C-EMR
NR, not reported; IQR, interquartile range; U-EMR, underwater endoscopic mucosal resection;
C-EMR, conventional endoscopic mucosal resection.
Zhang, 2020
66
71
6.0 (IQR 5.0–8.0) [median]
5.0 (IQR 4.0–7.0) [median]
59/66
62/71
7/66
9/71
62/66
65/71
1.2 (IQR 0.8–1.4) [minute]
1.35 (IQR 0.9–1.8) [minute]
0/71
0/71
1/71
1/71
0/71
2/71
4/66
6/71
NR
NR
Yen, 2020
248
214
9.9 (6–40) [mean]
9.9 (6–45) [mean]
NR
NR
5/248
4/214
223/248
193/214
3.8 (+/- 0.34) [minute]
5.4 (+/- 0.35) [minute]
0/248
0/214
5/248
4/214
0/248
0/214
NR
NR
NR
NR
Yamashina, 2019
108
102
14 (IQR 7–25) [median]
13.5 (IQR 7–25) [median]
74/108
51/102
NR
NR
96/108
76/102
2.75 (IQR 1.95–4.6) [minute] {median}
2.91 (IQR 2.2–4.4) [minute] {median}
0/108
0/102
0/108
0/102
3/108
2/102
12/108
26/102
NR
NR
Nagl, 2021
75
73
25 (IQR 20–40) [median]
30 (IQR 20–40) [median]
26/75
12/73
NR
NR
27/75
14/73
7 (IQR 3–13) [minute] {median}
13 (IQR 7–25) [minute] {median}
0/81
0/76
19/76
11/76
1/81
2/76
48/75
59/73
8/50
15/64
Lenz, 2022
61
59
17.4 (IQR 8) [median]
17.5 (IQR 8) [median]
NR
NR
1/61
1/59
37/61
32/59
NR
NR
0/61
0/59
2/61
5/59
0/61
0/59
NR
NR
1/50
8/53
Hamerski, 2019 (Abs)
158
145
NR
NR
NR
NR
NR
NR
76/158
35/145
10.8 (+/- 10.2)
16.9 (+/- 13.2)
1/158
2/145
NR
NR
9/158
9/145
NR
NR
10/124
16/103
Rodríguez Sánchez, 2022
149
162
30 (IQR 25–40) [median]
30 (IQR 25 – 38.8) [median]
41/149
39/162
12/131
26/148
47/149
46/162
14 (IQR 6–30) [minutes]
23 (IQR 9.2–45) [minutes]
4/149
5/162
24/149
25/162
7/149
13/162
NR
NR
24/149
25/162
Characteristics and quality of included studies
All included trials in our analysis were prospective with two single-center [27 ]
[28 ] and five multicenter trials [29 ]
[30 ]
[31 ]
[32 ]
[33 ]. All trials originated from different geographical regions including Asia, North
America, and Europe.
Meta-analysis outcomes
Primary outcomes
Overall pooled rate of R0 resection was higher with U-EMR, 58.05% (CI 29.75–81.89,
I2 96%) vs 44.57% (CI 17.38–75.45, I2 96%) with C-EMR. The difference between the two was found to be approaching statistical
significance, RR 1.25 (CI 0.99–1.59, I2 68%), P = 0.07 ([Fig. 1 ]).
Fig. 1 Forest plot, RR, R0 resection.
Overall pooled rate of successful en-bloc resection was significantly higher with
U-EMR than C-EMR, 70.17% (CI 46.68–86.34, I2 97%) vs 58.14% (CI 31.59–80.68, I2 97%), respectively, RR 1.21 (CI 1.01–1.44, I2 77%), P = 0.02 ([Fig. 2 ]). Based on three trials, for polyps > 20 mm in size, we found no statistically significant
difference in rates of en-bloc resection between the two techniques, 37.8% (CI 29.77–46.63,
I2 65%) vs 29.14% (CI 24.29–34.51, I2 0%), RR 1.24 (CI 0.83–1.84, I2 57%), P = 0.3 ([Table 3 ]).
Table 3 Study outcomes for polyps > 20 mm in size.
Study
Polyps > 20 mm in size
Incomplete resection
En-bloc resection
Resection time (minutes) (SD)
Perforation Immediate bleeding
Delayed bleeding
Piecemeal resection
Recurrence
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR
C-EMR
U-EMR, underwater endoscopic mucosal resection; C-EMR, conventional endoscopic mucosal
resection; IQR, interquartile range.
Yen, 2020
16/248
16/214
1/16
0/16
4/16
7/16
7.3 (0.62)
9.5 (0.66)
0/16
0/16
3/16
2/16
0/16
0/16
12/16
9/16
NR
NR
Lenz, 2022
12/61
14/59
NR
NR
NR
NR
NR
NR
0/12
0/14
NR
NR
0/12
0/14
NR
NR
0/11
5/14
Hamerski, 2019 (Abs)
NR
NR
NR
NR
68/150
39/139
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
Rodríguez Sánchez, 2022
149/149
162/162
12/131
26/148
47/149
46/162
14 (IQR
23 (IQR 9.2–45)
4/149
5/162
24/149
25/162
7/149
13/162
NR
NR
24/149
25/162
Fig. 2 Forest plot, RR, en-bloc resection.
Overall pooled rate of polyp recurrence at surveillance colonoscopy was significantly
lower with U-EMR, 7.88% (CI 5.16–11.85, I2 51%) as compared to C-EMR, 15.95% (CI 12.48–20.17, I2 14%), RR 0.62 (CI 0.41–0.94, I2 0%), P = 0.005 ([Fig. 3 ]).
Fig. 3 Forest Plot, RR, polyp recurrence.
Secondary outcomes
Overall pooled rate of piecemeal resection was significantly lower with U-EMR than
C-EMR, 19.16% (CI 4.35–56.65, I2 97%) vs 33.8% (CI 7.62–75.97, I2 97%), respectively, RR 0.66 (CI 0.43- 1.02, I2 36%), P = 0.05 (Supplementary Fig. 3 ).
While the overall pooled rate of incomplete resection was lower with U-EMR, 5.12%
(CI 1.98–12.60, I2 83%) than 6.15% (CI 1.79–19.1, I2 90%) with C-EMR, the difference between the two was found to be approaching statistical
significance, RR 0.67 (CI 0.41- 1.08, I2 0%), P = 0.07 (Supplementary Fig. 4 ).
Overall resection times were not significantly different comparing U-EMR and C-EMR,
standardized mean difference –1.21 min (–2.57, 0.16, I2 99%), P = 0.08 (Supplementary Fig. 5 ).
Overall adverse events
Pooled rates of perforation, immediate and delayed bleeding with U-EMR vs C-EMR were,
0.2% (CI 0.01–2.74, I2 0%) vs 0.26% (CI 0.02–3.89, I2 0%), 3.54% (CI 0.89–13.1, I2 87%) vs 3.95% (CI 1.22–12.06, I2 78%), and 1.14% (CI 0.28–4.61, I2 0%) vs 2.05% (CI 0.72–5.66, I2 6%), respectively. These events were comparable between U-EMR and C-EMR, RR 0.75
(CI 0.24–2.35, I2 0%), P = 0.6, RR 1.16 (CI 0.80–1.69, I2 0%), P = 0.4 and RR 0.73 (CI 0.41–1.29, I2 0%), P = 0.3, respectively (Supplementary Fig. 6 , Supplementary Fig. 7 , Supplementary Fig. 8 ).
Validation of meta-analysis results
Sensitivity analysis
To assess whether any one study had a dominant effect on the meta-analysis, we excluded
one study at a time and analyzed its effect on the main summary estimate. Sensitivity
analysis revealed no significant difference in the pooled outcomes for polyp recurrence.
Upon exclusion of the study by Zhang et al, we found that the pooled RR for R0 resection
was statistically significant in favor of U-EMR, RR 1.39 (CI 1.11–1.75), P = 0.005. This can likely be explained by the fact that that this study only included
colorectal polyps less than 10 mm in size and reported that the rate of R0 resection
was comparable between U-EMR and C-EMR, P = 0.706. As a result, its exclusion most likely resulted in favorable outcomes for
U-EMR during our sensitivity analysis.
Heterogeneity
We assessed dispersion of the calculated rates using the CI and I2 percentage values. The CI gives an idea of the range of the dispersion and I2 tells us what proportion of the dispersion is true vs chance [34 ]. Overall, low to moderate heterogeneity was noted in the pooled risk ratios of adverse
events as well as incomplete and piecemeal resection. Considerable to substantial
heterogeneity was noted in pooled risk ratios of R0 and en-bloc resection. This is
likely due to the inclusion of polyps in different locations and of variable sizes.
Bias assessment
Based on GRADE assessment of bias, overall quality of evidence was graded as moderate
(Grade B). This was primarily because due to the inherent design of the trials, the
performing endoscopist could not be blinded to the resection technique. Additionally,
publication bias was not ascertained as the number of studies included in our analysis
were less than 10.
Discussion
Our analysis, based on data from RCTs, shows that U-EMR achieves a higher rate of
successful en-bloc and R0 resections as well as lower rates of incomplete resection
for colorectal polyps compared to C-EMR. We found that for polyps greater than 20
mm in size, both techniques are able to achieve similar rates of en-bloc resection.
While U-EMR and C-EMR appear to have a similar safety profile in terms of immediate
and delayed bleeding as well as perforation, U-EMR is also associated with comparable
resection time and lower rates of polyp recurrence at surveillance colonoscopy.
U-EMR, described in 2012, has been well described by experts for lesions ≥ 20 mm [35 ]. It is easily learned does not require additional or new equipment [36 ]. Visualization of a polyp is enhanced underwater due to the “magnification” effect
that occurs owing to the higher index of refraction of water compared with air. Water
immersion also minimizes luminal distension, flexure angulation, and loop formation,
allowing for better maneuverability of the endoscope [37 ]. Compared to C-EMR, with U-EMR, larger mucosal surface area can be captured in a
snare as the colon is not distended or stretched. This feature could provide the possibility
to increase en-bloc resection rates, even for lesions larger than 20 mm [38 ]. In our analysis, we included a total of 369 polyps > 20 mm in size. In this group,
while U-EMR had a higher rate of successful en-bloc resection, it did not reach statistical
significance, U-EMR 38% vs C-EMR 29%, P = 0.3. This is likely due to the fact that only three of the included trials reported
outcomes for polyps greater than 20 mm size. We believe that for polyps between 20
mm and 30 mm in size, both EMR and endoscopic submucosal dissection (ESD) can be considered.
While EMR is acceptable if en-bloc resection is feasible, ESD could be a better alternative
if endoscopic appearance of the polyp i.e., pit pattern, is worrisome for high-grade
dysplasia or submucosal invasion. Furthermore, ESD may also be the preferred method
of resection if the anatomic location of the lesion is the rectum or morphologically
it has large nodules or depressed areas [39 ]
[40 ]
[41 ].
Interval cancer after colonoscopy can occur due to several reasons including adenoma
miss rate, which is about 17% for larger polyps (≥ 10 mm) [42 ]
[43 ]. An additional risk for development of interval CRC is incomplete resection rate
(IRR) of polyps at index colonoscopy. Malignant lesions have been shown to arise in
prior polypectomy sites in as many as 19% to 27% cases [44 ]. Furthermore, IRR is known to be significantly higher for sessile serrated adenomas/polyps
particularly those between 10 mm and 20 mm in size [45 ]. It is believed that R0 resection of colon polyps should be a key performance indicator
for endoscopists performing polypectomy [46 ]. In our analysis, outcomes of R0 resection were only reported in four trials. We
found that U-EMR resulted in higher rates of successful R0 resection and lower IRR.
Additionally, polyp recurrence at surveillance colonoscopy, as reported in four trials,
was seen in up to 15.9% patients with C-EMR as opposed to 7.9% in patients with U-EMR.
The statistically significant lower rate of recurrence with U-EMR may eventually lead
to lower rates of post colonoscopy CRC. There was insufficient data for us to calculate
the pooled rates of incomplete resection and recurrence for polyps > 20 mm in size.
However, based on two studies [31 ]
[47 ], a total of 18 of 148 polyps (12.2%) with U-EMR and 32 out of 157 (20.4%) with C-EMR
were incompletely resected. Additionally, nine of 140 patients (6.4%) undergoing U-EMR
and 22 of 145 patients (15.2%) undergoing C-EMR had recurrence at follow. Importantly,
post resection thermal ablation was not applied in any of the trials, to avoid bias
in the outcomes.
One of the concerns with U-EMR technique is possibly a higher risk of perforation
or deep muscle injury, as reported in a few cases. Contrary to C-EMR, in U-EMR, a
submucosal lifting agent separating the muscle layer from the mucosal layer is not
employed [35 ]
[48 ]. In our analysis, the pooled rates of perforations were comparable between the two
techniques. With C-EMR, while the use of electrocautery is thought to minimize intraprocedural
bleeding, higher rates of delayed bleeding have been reported in up to 5.1% patients.
[49 ] We found that the cumulative rates of delayed bleeding were comparable between U-EMR
and C-EMR, 1.14% vs 2.05%. Overall, our analysis confirms our previously reported
findings that U-EMR outperforms C-EMR in terms efficacy with comparable safety profile.
There are several strengths to our review including systematic literature search with
well-defined inclusion criteria, careful exclusion of redundant studies, inclusion
of good quality studies with detailed extraction of data, rigorous evaluation of study
quality, and statistics to establish and/or refute the validity of the results of
our meta-analysis. The study outcomes and definitions were uniform across all the
trials. We analyzed efficacy outcomes for all colorectal lesions and en-bloc resection
rates separately for polyps > 20 mm in size, and found that in the latter case, the
two techniques have comparable outcomes. Polyp recurrence was clearly defined as histologically
proven adenomas taken from biopsy samples of the resection scar at surveillance colonoscopy,
rather than visual appearance of the post-polypectomy site only. Only one of the trials
included in our analysis was published in abstract form [32 ]. We included updated data from one trial [47 ] while all others were published as full-length manuscripts. Finally, the majority
of RCTs included in our meta-analysis were multicenter experiences, performed at different
geographic locations around the world – Europe, the United States, and Asia — thus
making our results more generalizable.
There are also several limitations to this study, most of which are inherent to any
meta-analysis. First and foremost, one of the included studies in our analysis was
only published as a conference abstract [32 ]. Second, in two of the included trials, the mean/median polyp size was < 10 mm [27 ]
[29 ], which could have influenced our overall pooled rates of resection. Third, we found
that while pooled rates of R0 resection were higher with U-EMR, those of en-bloc resection
were comparable. This may be due to the fact that R0 resection rates were assessed
from only four trials. Fourth, we included colorectal lesions of different sizes and
morphology and were unable to characterize our outcomes based on these variables.
In two of the included trials, information regarding blinding of participants and
outcome assessment was not reported [31 ]
[32 ]. In all other trials, while the pathologist was blinded to the resection technique,
the performing endoscopists were not blinded, which likely resulted in risk of bias
in outcomes assessment. Finally, due to the inherent design of the included trials,
the performing endoscopists were not blinded to the resection technique, which could
have influenced the outcomes. Additionally, while in most studies, a group of experienced
endoscopists performed the procedures, in the study by Yen et al [27 ], all resections were performed by a single endoscopist, which may have resulted
in bias. Nevertheless, our study is the most up-to-date review, based on well-designed
RCTs across various geographical locations, comparing the efficacy and safety of U-EMR
and C-EMR.
Conclusions
Based on our results, we conclude that in terms of successful R0 and en-bloc resection,
U-EMR outperformed C-EMR. Lower rates of polyp recurrence, piecemeal and incomplete
resection, further validate the efficacy of U-EMR for colorectal lesions. We found
that rates of en-bloc resection for lesions > 20 mm size as well as adverse events
are similar between the two techniques.
