Introduction
Polypectomy during colonoscopy reduces colon cancer by approximately 50 % [1]. Small polyps are routinely resected during endoscopic intervention. However, resection
of large polyps (>20 mm) becomes more intricate and complicated. Among colonic polyps
detected during endoscopy, 5 % were lateral spreading lesions (LSLs) that can manifest
considerable growth along the bowel wall before progression to more invasive malignancy.
Moreover, large LSLs are recognized precursors of colorectal cancer but fortunately
can be treated with endoscopic mucosal resection (EMR) [2]. EMR is a standard technique for removal of large, sporadic, and laterally spreading
colorectal polyps. Previously, larger lesions would be an indication for surgery,
but endoscopic intervention has proven more effective in terms of significantly lowering
healthcare costs and lessening hospitalization days [3]
[4]. Unfortunately, EMR has a major drawback of adenoma recurrence. Despite complete
EMR of such lesions via the piecemeal approach, rates of adenoma reappearance at first
surveillance colonoscopy range from 15 % to 30 % [5]. The theory is that there exist unseen microadenomas at the margins of EMR, which
steadily grow to sizable adenomas given a month to year. Subsequently, adenoma recurrence
is one of the key limitations of EMR.
Snare tip soft coagulation (STSC) is a unique method initially used to control bleeding
[6], but the technique was later applied as adjuvant ablation at post-EMR margins following
standard resection of all visible adenoma islands. Klein et al [7] demonstrated efficacy in a randomized trial of adjuvant thermal ablation of post-EMR
mucosal defect margins via reduction of polyp recurrence rates. Likewise, that result
was later confirmed in a large multicenter study by Sidhu et al [8] illustrating the superior efficacy of adjuvant thermal ablation in real-world settings.
Currently, the literature regarding colon EMR with adjuvant thermal ablation consists
of three large studies—two multicenter RCTs and one large multicenter cohort study
along with a few retrospective studies [7]
[8]
[9]
[10]
[11]
[12]. Our meta-analysis will prompt a broader examination of the efficacy of STSC as
an adjuvant ablation method to post-EMR defects with regard to reduction in adenoma
reappearance.
Methods
Literature search and search strategy
This review is in accordance with Preferred Reporting Items for Systematic Reviews
and Meta-Analyses Statement (PRISMA) [13] (Supplement 1) with the studies reporting endoscopic mucosal resection of colorectal lesions and
adenoma recurrence decreasing technique by STSC of margins of EMR defect. Electronic
databases such as PubMed and the EMBASE (OVID) library were used for systematic literature
search (Supplement 2). Comprehensive literature search was assisted by an experienced librarian. The search
included key terms such as endoscopic mucosal resection, colorectal lesions, ablation
technique-snare tip soft coagulation, and adenoma recurrence. The search was run in
December 2021 and restricted to articles published in English. Ethical approval was
not sought as analysis included dataset ([Table 1]) [7]
[9]
[10]
[11]
[12].
Table 1
Baseline characteristics included in the studies.
|
Author, year
|
Study designs, country
|
Size
|
STSC
|
Non-STC
|
|
|
|
Patient (n)
|
Age (mean)
|
Female (n)
|
Polyp size, SD, rage (mm)
|
Patient (n)
|
Age (mean)
|
Female (n)
|
Polyp size, SD, range (mm)
|
|
Senada, 2020 (abs)
|
RCT multicenter, USA
|
> 20 mm
|
73
|
65.5
|
40
|
30 ± 11.1
|
75
|
66.6
|
40
|
33.3 ± 16.7
|
|
Wehbeh, 2020 (abs)
|
Retrospective, USA
|
> 20 mm
|
148
|
66.6 (10.8)
|
70
|
32.5 (13.7)
|
140
|
66.6 (10.8)
|
61
|
30.4 (10.9)
|
|
Perez, 2021 (abs)
|
Retrospective, USA
|
> 20 mm
|
43
|
65 (11)
|
23
|
NR
|
33
|
66 (9)
|
23
|
NR
|
|
Klein, 2019
|
RCT multicenter, Australia
|
> 20 mm
|
210
|
66.1 ± 11.6
|
109
|
30 (25–40)
|
206
|
67.0 ± 13.1
|
104
|
30 (25–45)
|
|
Kandel, 2019
|
Retrospective, USA
|
> 20 mm
|
60
|
66 (49–81)
|
35
|
28 ± 11; 25 (20–60)
|
60
|
65 (45–83)
|
29
|
28 ± 11; 25 (20–60)
|
STSC, snare tip soft coagulation; SD, standard deviation; RCT, randomized controlled
trial.
Inclusion and exclusion
Our meta-analysis included studies of human subjects published in English that addressed
an adenoma recurrence technique called STSC on post-EMR defect margins in polyps >20 mm.
The comparator arm consisted of polyps resected by EMR without treatment with STSC.
Included were randomized controlled trials (RCTs) and observational studies. Conference
abstracts were also included in the study if they contained usable data. Adenoma recurrence
data on first surveillance colonoscopy (SC1) had to be reported to be eligible. Resection
of malignant polyps, polyps < 20 mm, non-English studies, and studies including pediatric
age groups (< 18 years) were excluded.
Study selection
Study eligibility was assessed by (MH and DY) by going through study titles, full
text, and conference abstracts. Any discrepancies between reviewers were resolved
by discussion with a third reviewer (PK).
Outcome
The primary outcome of the study was the adenoma recurrence rate at first surveillance
colonoscopy (SC1) with the use of STSC at the margins of mucosal defect compared to
standard EMR. Both histology and endoscopic diagnosis were performed for adenoma recurrence
assessment. Endoscopic diagnosis of adenoma recurrence is also highly accurate [14] for recurrence assessment. Data on post-procedure complications such as delayed
bleeding (post-EMR bleeding) were collected as a secondary outcome ([Table 2]) [7]
[9]
[10]
[11]
[12].
Table 2
Outcomes data.
|
STSC
|
Non-STSC
|
|
Author, year
|
Adenoma recurrence (n)
|
Post-EMR bleeding (n)
|
Adenoma recurrence (n)
|
Post-EMR bleeding (n)
|
|
Senada, 2020 (abs)
|
9
|
9
|
23
|
19
|
|
Wehbeh, 2020 (abs)
|
9
|
NR
|
21
|
NR
|
|
Perez, 2021 (abs)
|
2
|
NR
|
16
|
NR
|
|
Klein, 2019
|
10
|
49
|
37
|
47
|
|
Kandel, 2019
|
7
|
9
|
18
|
12
|
STSC, snare tip soft coagulation.
Statistical analysis
We calculated the risk ratio (RR) for adenoma recurrence comparing STSC of margins
of mucosal defect compared to standard EMR, non-STSC. The ORs of individual studies
were pooled in this meta-analysis using ReVman software (version 5) with a random
effects model. Random effects models are used to estimate an average effect. The variability
of the effects represented by their average may have consequences for the clinical
interpretation of the findings [15]
. Therefore, prediction interval was included in the forest plot. The heterogeneity
of studies was assessed using I2 (inconsistency) statistic. Heterogeneity values 25 %, 50 %, and 75 % indicate low,
moderate, and high level of heterogeneity [16] based on a previously published validated tool. In addition, we also calculated
the OR of post-EMR delayed bleeding in the STSC group and non-STSC standard EMR group. Statistical
analysis was conducted with Review Manager version 5.3.5 and JMP software version
14.
Risk-of-bias assessment
Observational studies differ from controlled trials in regard to validity problems
and can be prevented by the study design, e. g., by randomization, concealed allocation,
and masking [17]. We have acknowledged limitations of non-randomized studies compared to the gold
standard randomized studies. Studies were selected through meticulous search to avoid
selection bias. There is always a risk of misclassification of groups/data in a meta-analysis.
To prevent that, we included the most accurate measurement of available data and carefully
categorized data according to the proper group. Sensitivity analysis was conducted
only for randomized trials and non-randomized studies were excluded. The risk of bias
(ROB) for each study outcome was assessed using existing validated tools ([Table 3]) [7]
[9]
[10]
[11]
[12]. For RCTs, we used a revised Cochrane risk-of-bias tool for the RCT checklist (ROB
2) [18] and the methodical index for non-RCTs (MINORS) checklist [19]. For RCTs, the outcomes were assessed as a low and high ROB using the ROB algorithm.
For non-RCTs, the MINORS checklist was used providing scores ranging from 0 to 24.
Scores of 0 to 8 were considered high ROB, 9 to 16 were considered of some concern,
and 17 to 24 as low ROB. For publication bias we used a funnel plot to evaluate asymmetry.
However, given the small number of included studies, proper additional statistical
analysis was not performed [20].
Table 3
Risk-of-bias assessment.
|
Author, year
|
Design
|
No. in intervention group (STSC)
|
No. in control group (non-STSC)
|
Risk of bias for adenoma recurrence
|
Risk of bias for delayed bleeding
|
|
Senada, 2020 (abs)
|
RCT
|
73
|
75
|
Low
|
Low
|
|
Wehbeh, 2020 (abs)
|
Retrospective
|
148
|
140
|
Some concerns
|
Some concerns
|
|
Perez, 2021 (abs)
|
Retrospective
|
43
|
33
|
Some concerns
|
Some concerns
|
|
Klein, 2019
|
RCT
|
210
|
206
|
Low
|
Low
|
|
Kandel, 2019
|
Retrospective
|
60
|
60
|
Low
|
Low
|
STSC, snare tip soft coagulation; RCT, randomized controlled trial.
Results
A total of 534 patients completed the first surveillance colonoscopy (SC1) and were
included in the meta-analysis ([Table 1]
). We included five studies (Kandel [9], Kelin [7], Senada [12], Wehbeh [10], Perez [11]) for meta-analysis, all of which used STSC at mucosal defect margins after EMR as
an intervention and adenoma recurrence as the primary outcome at first surveillance
colonoscopy (SC1). Two studies were multicenter RCTs. The remaining three were retrospective
cohort studies. First surveillance colonoscopy interval (SC1) ranged up to 12 months.
Size of polyps included was > 20 mm in both groups. Mean age was > 60 years. The total
number of patients in the STSC group was 534 and in the non-STSC group was 514.The
adenoma recurrence rate at SC1 in the STSC group was 6 % (confidence interval [CI],
5 %–9 %) and 22 % (CI 18 %–26 %) in the non-STSC group. Overall pooled OR for adenoma
recurrence in the STSC group was 0.26 (95 % CI, 0.16–0.41) compared to the non-STSC
group ([Fig. 1]) [7]
[9]
[10]
[11]
[12]. Heterogeneity between included studies was assessed with I2
= 23 %, which is considered as low. The heterogeneity decreased to zero, I2
= 0 when only randomized trials were included ([Fig. 2]) [7]
[9]
[10]
[11]
[12].
Fig. 1 Adenoma recurrence [7]
[9]
[10]
[11]
[12].
Fig. 2 Adenoma recurrence (RCT only) [7]
[9]
[10]
[11]
[12].
Only three studies reported post-procedural delayed bleeding as a complication. The
rate of delayed bleeding in the STSC group was 19 % (CI,15 %-24 %) and 22 % (CI,18 %-27 %)
in the non-STSC group. The overall pooled odds ratio (OR) for delayed post-EMR bleeding
in the STSC group was 0.82 (95 % CI, 0.57–1.18) compared to the non-STSC group ([Fig. 3]) [7]
[9]
[12]. The heterogeneity between included studies was assessed with I2 = 40 %, which is considered as low.
Fig. 3 Post-EMR delayed bleeding [7]
[9]
[12].
Discussion
To the best of our knowledge, we present the first meta-analysis showing the effectiveness
of EMR with adjunctive STSC at margins of mucosal defects to reduce adenoma recurrence.
This meta-analysis demonstrated that use of STSC of post-EMR mucosal defects decreases
adenoma recurrence in first surveillance colonoscopy. Among patients who had undergone
EMR and had STSC as an adjunctive treatment of mucosal defect margins, 74 % were less
likely to have adenoma recurrence. There was no difference in rates of post-EMR delayed
bleeding, suggesting that use of STSC as an adjunctive treatment is safe.
The field of endoscopy has undergone multiple advances that ultimately have reduced
the prevalence of colorectal cancer (CRC). CRC has minimal lymph node metastases as
the colonic mucosa lacks lymphatic vessels; thus, endoscopic resection of malignant
tissue is essentially curative [21]. Consequently, EMR is an established therapeutic modality that can be performed
on an outpatient basis and garner outstanding long-term results. In the treatment
of large colorectal lesions, the rate of adenoma recurrence is as high as 30 % and
remains a major hindrance to EMR [2]. The CARE study mirrored similar results: a 23.3 % rate of biopsy-proven residual
adenomas in incompletely resected lesions measuring 15 to 20 mm [22]. Along with lesion size > 40 mm, there are various other risk factors for adenoma
recurrence, such as utilization of argon plasma coagulation (APC) to ablate malignant
tissue, histological evidence of high-grade dysplasia, insufficient submucosal lifting
leading to incomplete resection, and intraprocedural bleeding [23]
[24]
[25].
The past school of thought was to extend the EMR margin several millimeters within
normal-appearing mucosa with the objective of eliminating neoplasia. Unfortunately,
a multicenter study described no difference in terms of adenoma recurrence between
typical and extended EMR [26]. This was an important finding in EMR because there exist unseen residual neoplastic
cells at the margins of the snare trajectory. Intuitively, other interventions targeting
the EMR margins have also been introduced in endoscopy practice along with extended
EMR [26], such as precutting EMR [27], APC [28]
[29], and STSC [7]
[9]. In general, these interventions decrease recurrence rates by as much as 63 % [30], with APC and STSC demonstrating superior efficacy in recurrence reduction compared
to extended EMR and precutting EMR as per subgroup analysis. This corresponds with
a large retrospective multicenter cohort study in 2019 with only 4.5 % recurrence
in APC [31], and only 5.2 % [7] and 12 % [9] recurrence in STSC. Between the two cauterization techniques, however, STSC was
associated with decreased recurrence at 7.8 % compared with 10 % regarding the APC
group [32]. Regarding the current transmission from catheter to tissue, there is a heavy reliance
on fluctuating arching and it is thus operator dependent. Arching is variable, sporadic,
and is very challenging to visualize when confirming eradication of tissue [9]. In addition, there is always an increase in cost with the use of APC due to the
need for a special generator and catheter. Adverse events (AEs) such perforation occur
in approximately 0.5 % if cases with APC, although they are rare [33]
[34]. Thus, STSC seems to improve outcomes with EMR compared to APC.
The Australian cohort performed the first-ever large multicenter RCT assessing the
efficacy of post-EMR thermal ablation with STSC. It showed a 4-fold reduction in adenoma
recurrence at SC1 (5.8 % in the STSC group versus 20.2 % in the non-STSC arm) [7]. Those results were confirmed by a multicenter trial from the United States [12]. There was significant reduction in adenoma recurrence in the STSC group compared
to the non-STSC group (12 % STSC group vs 34 % non-STSC group, P = 0.001).
Moreover, a smaller non-randomized controlled trial regarding an educational intervention
of STSC demonstrated 12 % recurrence rate in the STSC group versus 30 % in the non-STSC
group [9]. These results show superior efficacy in a controlled setting. In addition, results
recently have been published of a large cohort study from the same group (Australia),
which showed that thermal ablation of post-EMR defect mucosal margins was achieved
in 95.4 % of cases (n = 989 large nonmalignant colorectal polyps, LNPC). Ninety-four
percent (n = 755, LNPC) underwent SC1 (median time 6 months), and only 1.4 % had adenoma
recurrence on follow up [8]. Thus, STSC remained more precise because the application distance is standardized
and garnering consistent results.
Our meta-analysis demonstrated a pooled adenoma recurrence rate 6 % in the STSC arm
compared to 22 % in the non-STSC arm. These results were statistically significant
with a narrow CI, underscoring the exceptional effectiveness in STSC. Most adenoma
recurrences are identified within 6 months post-EMR [23]
[28]
[35]; thus, that timeframe was selected for surveillance colonoscopy (SC1) across all
our studies. Subsequently, the study endpoints were homogenous with an I2 = 23%, indicating low heterogeneity. Likewise, baseline demographic characteristics
and polyp size were similar across studies, thus minimizing avenues of confounding
bias. In addition, a funnel plot seems symmetric, suggesting no publication bias ([Fig. 4]).
Fig. 4 Funnel plot of comparison 1 STSC vs non-STSC, outcome 1.1 adenoma recurrence.
Only three studies had reported delayed bleeding as a complication. Numerical results
were similar; however, the wide CIs suggest statistical insignificance (RR 0.82; 0.57,
1.18). Although statistical insignificant, STSC has a clear benefit in prevention
of delayed bleeding. Delayed bleeding is one of the most common complications post-EMR
[36]
[37]
[38]
[39]. Delayed bleeding is defined as any bleeding occurring within a month after completion
of the procedure and that requires Emergency Department presentation, hospitalization,
or reintervention (repeat endoscopy, angiography, or surgery) [37]. Along with the reduction in adenomas, there are logistical and financial benefits
associated with STSC. For example, STSC can be performed with the same snare that
was used for resection, hence it does not significantly lengthen procedure times.
Thus, STSC remains an extremely safe procedure with minimal AEs.
We acknowledge that our meta-analysis had limitations. First, our study only included
five studies including abstracts, which can be interpreted as not quite adequate.
However, we were able to include the usable data in the meta-analysis. Second, there
was variation among studies in recording morphology of polyps such as NBI and Paris
classification. If that information was available, we would have included it in the
study. Third, the generator settings were not available for all studies. However,
STSC effect in terms of adenoma recurrence is similar among all studies. This further
validates that STSC is effective, despite use of different generator settings.
Conclusions
In conclusion, our study illustrates that thermal ablation of post-EMR defects with
STSC significantly reduces adenoma recurrence at first surveillance colonoscopy. This
safe and simple technique can improve outcomes in such patients and should be integrated
into routine EMR practice. With lower rates of adenoma recurrence, follow-up intervals
post-EMR potentially can be prolonged, which can gradually lessen the burden on overall
healthcare resources.
