Introduction
Endoscopic full-thickness resection (eFTR) is an emerging innovative endoscopic resection
technique for complex colorectal lesions. With the advantage of enabling a transmural
resection, eFTR offers an alternative to radical surgery in lesions considered incurable
with current resection techniques such as endoscopic mucosal resection (EMR) or endoscopic
submucosal dissection (ESD).
In clinical practice, the main indications for eFTR are non-lifting lesions with severe
submucosal fibrosis, lesions that involve difficult anatomical locations such as the
appendiceal orifice or diverticula, and subepithelial tumors. Furthermore, eFTR is
gaining interest as a valid diagnostic and therapeutic treatment option for T1 colorectal
cancer (CRC), as it can provide high quality pathological specimens and exact histological
risk assessment [1 ].
Several studies have reported encouraging results on the short-term safety and efficacy
of eFTR for numerous indications [1 ]
[2 ]
[3 ]
[4 ]
[5 ]
[6 ]
[7 ]
[8 ]
[9 ]
[10 ]
[11 ]
[12 ]. However, firm conclusions on clinical results will require analysis of large prospective
series of patients in everyday clinical practice.
Since the introduction of the full-thickness resection device (FTRD; Ovesco Endoscopy
AG, Tübingen, Germany) in the Netherlands in 2015, we started a nationwide prospective
registry of all consecutive eFTR procedures to monitor patient outcomes and further
increase knowledge on clinical applicability and safety. In this prospective observational
multicenter study, we aimed to evaluate the technical and clinical success, and safety
of eFTR for colorectal lesions in current clinical practice.
Methods
Study design
In this study, all registered procedures between July 2015 and October 2018 were analyzed.
Results were prospectively recorded in 20 Dutch hospitals (5 academic and 15 non-academic).
Procedures were performed by eFTR-certified gastroenterologists. For eFTR certification,
experienced gastroenterologists attended a 1-day training course on eFTR, with thorough
theoretical background and hands-on training in ex vivo porcine models.
Because the data were collected as part of standard medical care, the Institutional
Review Board of the Amsterdam University Medical Centers decided that the study fell
outside of the legislation regarding Medical Research Involving Human Subjects Act
and therefore formal ethical approval was not deemed necessary. The study was registered
in the Dutch Trial Register: NTR6292 (http://www.trialregister.nl/ ).
All patients were informed about the eFTR procedure and the peri-procedural risks
by their treating physician and informed consent for the procedure was obtained. The
results of 58 eFTR procedures were published previously [13 ]
[14 ]. Coded data were collected into a secure online structured database provided by
Castor EDC, Amsterdam, the Netherlands [15 ].
Outcome
The primary outcome was the technical success of all initiated procedures, defined
as the proportion of macroscopic complete en bloc resections (judged by the endoscopist).
Secondary outcome measures included: clinical success, defined as the rate of histologically
confirmed radical (R0) resections (tumor-free lateral and deep resection margins);
rate of histologically confirmed full-thickness resection (presence of the muscularis
propria in the resection specimen); procedure-related adverse events; and recurrence
at first follow-up colonoscopy.
Study subjects
Eligibility for eFTR was judged by the treating physician after a complete colonoscopy
had been performed. Indications for eFTR were “difficult polyps,” including non-lifting
polyps (treatment naïve, recurrent, or incompletely resected polyps) and polyps involving
the appendiceal orifice or a diverticulum; primary resection for suspected T1 CRCs;
and secondary completion treatment after previous incomplete endoscopic resection
with a positive (≤ 1 mm) or non-assessable resection margin (R1/Rx resection). Other
indications for eFTR were subepithelial tumours. Because of the limited size of the
FTRD cap, lesions with an estimated diameter of ≥ 30 mm were considered unsuitable
for eFTR [16 ]. As this study used a prospective registry of current clinical practice, no explicit
exclusion criteria were formulated.
eFTR procedure and management
All patients received standard split-dose polyethylene glycol (PEG) bowel preparation.
Procedures were performed with the patient under sedation with propofol or midazolam
and/or fentanyl, according to local practice. A single dose of prophylactic intravenous
antibiotic therapy was advised for all eFTR procedures in the early study period,
but since 2017 antibiotic prophylaxis was deemed unnecessary and was no longer advised,
except for appendiceal lesions without previous appendectomy. Patients on anticoagulation
therapy were advised to use their medication according to the Dutch guideline for
antithrombotic therapy in endoscopy: coumarins and direct-acting oral anticoagulants
(DOACs) were discontinued, single-agent antiplatelet agents could be continued [17 ]. The FTRD was used in all procedures, and these were performed using CO2 insufflation.
First, the target lesion was identified during a diagnostic endoscopy and marked by
creating coagulation marks with the dedicated probe or by other means, and the lesion
size was estimated at the discretion of the endoscopist. Thereafter, the colonoscope
was withdrawn, the FTRD was mounted and the colonoscope was advanced again to the
target area. A specialized grasping forceps was advanced through the working channel
to grasp and slowly pull the lesion into the cap. Once the lateral margins of the
lesion were pulled into the cap, the clip was deployed. Immediately thereafter, the
tissue was resected with the pre-loaded snare, while the clip secured the integrity
of the bowel wall. The endoscope with the resection specimen entrapped in the cap
was then withdrawn. After the specimen had been secured and the device demounted,
the endoscope was introduced once again to inspect the position of the clip and the
resection site. The specimen was stretched and pinned onto cork or paraffin before
immersion into formalin for histological analysis ([Fig. 1 ] and [Fig. 2 ]) [13 ]. To obtain the most accurate measurement of lesion and specimen size after eFTR,
we used the size measured at histology.
Fig. 1 Endoscopic full-thickness resection of a recurrent non-lifting adenoma (third recurrence).
a – c Endoscopic images showing: a,b a recurrent adenoma in the transverse colon with clear fibrosis and fold conversion;
c the resection site with the over-the-scope clip in place and exposure of the submucosal
tattoo. d Macroscopic appearance of the resected specimen pinned down with the serosal side
up (plus tattoo) on paraffin. e Histopathological appearance showing a full-thickness resection of a tubulovillous
adenoma with low grade dysplasia, with clear submucosal fibrosis and ink. Lateral
and deep resection margins were clear. Source for Fig. 1e: Lianne Koens.
Fig. 2 Completion endoscopic full-thickness resection after a previous incomplete resection
of a low risk T1 colorectal cancer. a – d Endoscopic images showing: a,b narrow-band imaging of the target lesion; c the mounted full-thickness resection device on the marked lesion; d the full-thickness resection site with the over-the-scope clip. e Histopathological appearance showing a R0 resection of a moderately differentiated
adenocarcinoma with deep submucosal invasion (SM3), without lymphovascular invasion
or tumor budding. Source for Fig. 2e: Lianne Koens.
Clinical admission for 24 hours to monitor signs of discomfort, bleeding, or perforation
was advised in the early study period, but from 2017 the advice was withdrawn and
this was left to the discretion of treating physician. The recommended dietary regimen
was a clear fluid diet for 24 hours, followed by a normal diet.
Adverse events and follow-up
All immediate procedure-related adverse events that resulted in prolonged admission
and/or an intervention (i. e. blood transfusion, re-colonoscopy, or surgery) were
recorded at the time of colonoscopy. Delayed procedure-related adverse events requiring
readmission or intervention (i. e. blood transfusion, colonoscopy, or surgery) were
recorded approximately 14 days after eFTR, when patients were contacted to discuss
the histopathological results. The severity of adverse events was graded according
to the American Society for Gastrointestinal Endoscopy [18 ].
Patients were scheduled for follow-up colonoscopy after 3 – 6 months. The eFTR scar
was inspected with high definition white-light endoscopy and (digital) chromoendoscopy
for macroscopic recurrent or residual tissue and the presence of the clip. Endoscopic
and/or histological findings compatible with granulation tissue or reactive changes
were left untreated. Residual polyp tissue was treated with conventional treatment
strategies if possible. When endoscopic resection of the residual polyp was considered
impossible or submucosal invasive cancer was suspected, a tattoo (SPOT) was placed
followed by case discussion in a colorectal multidisciplinary meeting. Follow-up colonoscopies
were scheduled depending on histological and endoscopic findings. Patients referred
for additional surgery after eFTR were excluded from scar surveillance.
Statistical analysis
Standard descriptive statistics were used. Variables are reported as mean with standard
deviation (SD) for continuous and normally distributed variables, as median and interquartile
range (IQR) for non-normally distributed continuous variables, and as percentages
for counts or categorical variables. Categorical variables were analyzed using chi-squared
or two-sided Fisher’s exact tests. A two-sided P value < 0.05 was considered statistically significant. Statistical analysis was performed
using SPSS 24 (SPSS, Chicago, Illinois, USA).
Results
Patients characteristics, procedural data, and technical success
A total of 367 eFTR procedures were performed in 362 patients between July 2015 and
October 2018 and were included for analysis (63 % men; mean age 69 years). The indications
were: difficult polyps (n = 133), primary resection of suspected T1 CRC (n = 71),
re-resection after previous incomplete resection of T1 CRC (n = 150), and subepithelial
tumors (n = 13). Procedures were performed by 37 certified endoscopists divided over
20 Dutch hospitals. Patient characteristics are shown in [Table 1 ].
Table 1
Characteristics of the 362 patients who underwent endoscopic full-thickness resection
(eFTR) and the 367 procedures performed.
sex, male, n (%)
227 (62.7)
mean age (SD), years
69 (8.2)
Indication, n (% of eFTR procedures)
221 (60.2)
71 (19.3)
150 (40.9)
133 (36.2)
85 (23.2)
28 (7.6)
5 (1.4)
15 (4.1)
3 (20.0)
13 (3.5)
Estimated median diameter of lesion (IQR), mm[* ]
10 (8 – 15)
Lesion location, n (% of eFTR procedures)
143 (39.0)
27 (7.4)
17 (4.6)
51 (13.9)
14 (3.8)
27 (7.4)
7 (1.9)
224 (61.0)
22 (6.0)
123 (33.5)
79 (21.5)
SD, standard deviation; IQR, interquartile range.
* Size estimated by endoscopist.
Technical success was achieved in 83.9 % of all procedures (n = 308/367) ([Table 2 ]). In 1.9 % of procedures (7/367), the target lesion could not be reached successfully
([Fig. 3 ]). The main reason for not reaching the lesion was failure to pass the sigmoid colon
with the FTRD in three procedures (42.9 %); for the other four procedures, no reason
was described. In 14 procedures (3.9 %), the eFTR was not performed because the lesion
could not be retracted into the cap owing to lack of mobility of the tissue. Of these
14 lesions, 12 were requiring re-resection after previous EMR.
Table 2
Technical success rates for all endoscopic full-thickness resection (eFTR) procedures
initiated and clinical success rates in those amenable to EFTR.
Overall
T1 CRCs
Difficult polyps
Subepithelial tumors
Initiated eFTR procedures, n
367
221
133
13
Technical success, n (%)
308 (83.9)
191 (86.4)
105 (78.9)
12 (92.3)
Procedures amenable to eFTR, n[1 ]
346
211
122
13
Resection, n (%)
285 (82.4)
186 (88.2)
86 (70.5)
13 (100)
288 (83.2)
176 (83.4)
100 (82.0)
12 (92.3)
Lesion diameter, median (IQR), mm[2 ]
12 (8 – 17)
13 (9 – 18)
12 (8 – 15)
9 (5 – 15)
23 (20 – 28)
23 (19 – 27)
23 (20 – 29)
26 (20 – 30)
Location lesion, n (%)[3 ]
131 (37.9)
55 (26.1)
76 (62.3)
0 (0)
215 (62.1)
156 (73.9)
46 (37.7)
13 (100)
Histology, n (%)
145 (41.9)
120 (56.9)
16 (13.1)
8 (61.5)
75 (21.7)
13 (6.2)
63 (51.6)
0 (0)
15 (4.3)
5 (2.4)
10 (8.2)
0 (0)
17 (4.9)
3 (1.4)
14 (11.5)
0 (0)
84 (24.3)
67 (31.8)
17 (13.9)
0 (0)
5 (1.4)
0 (0)
0 (0)
5 (38.5)
3 (0.9)
3 (1.4)
0 (0)
0 (0)
2 (0.6)
0 (0)
2 (1.6)
0 (0)
CRC, colorectal cancer; IQR, interquartile range; LGD, low grade dysplasia; HGD, high
grade dysplasia.
1 Ability to obtain histology (lesion was reached and over-the-scope clip deployed).
2 Size measured by the pathologist at histopathology.
3 Proximal is defined as cecum to splenic flexure and distal as descending colon to
rectum.
4 The subgroup of T1 CRCs included a metastasis of gastric cancer (n = 1), signs of
an inflammatory disease (n = 1), and the presence of mucin fields (n = 1).
Fig. 3 Flowchart of the endoscopic full-thickness resection (eFTR) procedures included in
the study. OTSC, over-the-scope clip.
In 346 procedures (94.3 %), it was possible to perform eFTR (lesion reached and clip
deployed) and histology was obtained. Device malfunctions were reported in 32 of these
346 procedures (9.2 %). These were caused by snare malfunctions (n = 23; 6.6 %), grasper
dysfunction (n = 6; 1.7 %), and inability to release the clip (n = 3; 0.9 %). However,
it was possible to complete the resection in 22 of the snare malfunctions and five
of the grasper dysfunctions.
Clinical success
In the total cohort of procedures amenable to eFTR (n = 346), histological R0 resection
was achieved in 285 procedures (82.4 %) and full-thickness resection in 288 (83.2 %)
([Table 2 ]). There were 14 cases that were judged during the procedure to be technically unsuccessful
owing to macroscopic incomplete resection but were found to have tumor-free resection
margins (R0) at histology (4.0 %). The median diameters of the resected lesions and
the resected specimens at histology were 12 mm (IQR 8 – 17) and 23 mm (IQR 20 – 28),
respectively. R0 resection for lesions > 20 mm was achieved in 72.7 % vs. 78.3 % for
lesions ≤ 20 mm (P = 0.71).
eFTR in the subgroup of T1 CRCs (n = 221) was technical successful in 191 procedures
(86.4 %) and R0 resection was achieved in 186 procedures (88.2 %). For primary resections
of T1 CRCs, the R0 resection rate was 77.9 % (n = 53) compared with 93.0 % (n = 133)
for secondary treatment (Table 1s ).
After resection of the scar for previously incompletely resected T1 CRCs (R1/Rx),
only scar tissue was found at histology in 117 procedures (81.8 %). Residual adenocarcinoma
was found in 11 procedures (7.7 %). Of these, five patients were referred for additional
oncological surgery because of high risk features for lymph node metastasis (LNM)
or an incomplete eFTR. In one of these five patients, surgery was not performed owing
to significant comorbidities. Of the four operated patients, one surgical specimen
showed residual T2 cancer without LNM and one a T2 cancer with a positive lymph node.
In the two other patients, no residual cancer was found.
In the 68 patients with a primary resection for T1 CRC, 18 (26.5 %) were referred
for secondary surgery because of high risk features and/or incomplete resection. Of
all the surgical specimens, one case had residual cancer (T3N2) (Table 2s ).
In the subgroup of difficult polyps (n = 133), the procedure was technically successful
in 105 procedures (78.9 %) and an R0 resection was obtained in 86 procedures (70.5 %).
In the subepithelial lesions (n = 13), eFTR was technically successful in 12 cases
(92.3 %) and R0 resection was achieved in all cases ([Table 2 ]).
Adverse events
Overall, adverse events occurred in 34 procedures (9.3 %) ([Table 3 ]). Severe adverse events occurred in 10 procedures (2.7 %), including seven perforations
(1.9 %). In two patients, the perforation was noticed immediately; in both, the clip
had not been released appropriately before the lesion was resected. In five patients
delayed perforations occurred on days 1 – 8. In three of these, the lesion was located
in the sigmoid and one of these patients was on immunosuppressive therapy for inflammatory
bowel disease. The other two delayed perforations were located in the transverse colon
and cecum. All seven patients needed surgical repair. In addition, three patients
with an appendiceal lesion (n = 3/15; 20.0 %) developed secondary appendicitis (all
without previous appendectomy) and underwent an acute laparoscopic appendectomy (n = 2)
or a laparoscopic ileocecal resection (n = 1).
Table 3
Safety of the 367 endoscopic full-thickness resection (eFTR) procedures.
Adverse events, n (%)
34 (9.3)
Mild adverse events[1 ]
16 (4.4)
8 (2.2)
2 (0.5)
1 (0.3)
1 (0.3)
2 (0.5)
1 (0.3)
1 (0.3)
Moderate adverse events[2 ]
8 (2.2)
7 (1.9)
1 (0.3)
Severe adverse events[3 ]
10 (2.7)
7 (1.9)
3 (0.8)
Indication for surgery, n (%)
65 (17.7)
13 (3.5)
11 (3.0)
18 (4.9)
10 (2.7)
7 (70.0)
3 (30.0)
7 (1.9)
6 (1.6)
Median hospital stay (IQR), days
1.0 (1 – 1)
CRC, colorectal cancer; IQR, interquartile range.
1 Mild: unplanned hospital admission or prolongation of hospital stay for ≤ 3 days
or post-procedure medical consultation or procedure aborted (or not started) because
of an adverse event.
2 Moderate: unplanned hospital admission or prolongation of hospital stay for 4 – 10
days or intensive care admission for 1 night or radiological intervention or transfusion
or repeat endoscopy for an adverse event or unplanned anesthesia/ventilation support,
i. e. endotracheal intubation during conscious sedation.
3 Severe: unplanned hospital admission or prolongation of hospital stay for > 10 days
or intensive care admission for > 1 night or surgery for an adverse event or permanent
disability [18].
4 Surgical resection was due to a synchronous CRC (n = 2), another polyp that could
not be endoscopically resected (n = 1), preference for a surgical resection (n = 1),
and the presence of mucin fields (n = 1). One patient participated in a trial and
also received chemotherapy.
Moderately severe adverse events were observed following eight procedures (2.2 %).
In seven (1.9 %), post-procedural bleeding was observed that required re-admission
or repeat endoscopy. In 16 (4.4 %), a mild adverse event was observed. In two cases
(0.5 %), this was due to a perforation. The first perforation, which was seen immediately
after the procedure, could be successfully clipped with an over-the-scope clip (OTSC).
The second case, located in the sigmoid, occurred a few days after an uneventful eFTR
and was treated conservatively with antibiotics and re-admission for 3 days. One patient
with a history of appendectomy developed a peri-appendiceal abscess after eFTR for
an appendiceal lesion and was treated conservatively.
Surgery
Overall, 65 patients (17.7 %) underwent surgery after the initial eFTR procedure ([Table 3 ]). In 13 (3.5 %), surgery was required because eFTR could not be performed (lesion
not reached or clip not deployed); in 10 (2.7 %), emergency surgery was performed
because of an adverse event. Elective oncological surgery was performed in 29 patients
(7.9 %): in 11 (3.0 %) after incomplete eFTR (R1 /Rx), and in 18 (4.9 %) because of
high risk features for LNM. Of these elective surgery cases, histology was available
in 26 (89.7 %). Residual adenocarcinoma was found in three (11.5 %) (Table 2 s ).
Follow-up
Endoscopic follow-up data were available for 187 procedures (63.4 %) after eFTR and
without additional surgery. Median time to follow-up was 4 months (IQR 3 – 6). In
31 procedures (16.6 %), the clip was still in situ at follow-up. In 100 procedures
(33.9 %), endoscopic surveillance was still pending or not recorded. Furthermore,
in the remaining eight cases (2.7 %), no follow-up was scheduled. Reasons to refrain
from follow-up were co-morbidities (n = 3), no indication for follow-up (n = 2), or
unclear reasons (n = 3). Residual/recurrent lesions were seen in 12 patients (6.4 %).
No adenocarcinomas were detected (Table 3s ).
Discussion
In this study, we report the initial results of 367 eFTR procedures prospectively
recorded in an eFTR registry in the Netherlands, representing the largest patient
cohort published to date. While the interest and adoption of eFTR continues to grow
amongst endoscopists, outcomes of large prospective series on its safety and efficacy
are needed. In this multicenter collaboration, we aimed to investigate the clinical
outcomes of eFTR for complex colorectal lesions.
In our study, the technical success rate for eFTR was 83.9 % and R0 resection was
achieved in 82.4 % of the procedures. The results of our study are comparable with
the results of the previous prospective study by Schmidt and colleagues, and to several
other retrospective series [1 ]
[2 ]
[3 ]
[4 ]
[5 ]
[6 ]
[7 ]
[8 ]
[9 ]
[10 ]
[11 ]
[12 ]. In a meta-analysis comparing ESD and EMR for colorectal polyps, the R0 resection
rate was 80.3 % for ESD and 42.3 % for EMR [19 ]. Therefore, R0 resection rates of eFTR are comparable with ESD and more favorable
than EMR [20 ]
[21 ]
[22 ]
[23 ]
[24 ]
[25 ]. However, it is important to consider that this cohort mainly includes complex colorectal
lesions, considered unsuitable for conventional endoscopic en bloc resection techniques.
En bloc excision is always preferred as it provides a higher quality pathological
specimen and lower recurrence rates. This advantage should however be balanced against
a higher perforation risk [24 ]. eFTR can tackle the challenge posed by submucosal fibrosis, which makes safe en
bloc EMR or ESD difficult.
Although the learning curve for eFTR is acknowledged to be shorter than for ESD, eFTR
remains a challenging procedure with several limitations [3 ]. At present, a consensus on appropriate patient selection for eFTR is lacking. The
main limitation for colorectal eFTR is the limited lesion size that can be treated.
The FTRD system accommodates lesions with a maximum diameter of 30 mm [16 ]. In contrast to Schmidt and colleagues, we did not find a significant difference
in the R0 resection rate between lesions > 20 mm (72.7 %) vs. lesions ≤ 20 mm (78.3 %;
P = 0.71) [3 ]
[11 ]. However, our subgroup of lesions larger > 20 mm is relatively small and might have
biased our results. Considering these results together with an average size of 23 mm
for the resected specimen, we recommend lesions with a maximum estimated diameter
of 20 mm appropriate for eFTR.
Further technical limitations of the device are the impaired visibility and tip flexibility
due to the long cap. Using a “dummy” cap (prOVE cap; Ovesco Endoscopy AG) when it
is suspected that introduction will be difficult or there is doubt about wall mobility
could help to optimize patient selection. In our study, we report a relatively high
rate of device malfunction (9.2 %). However, in the majority of cases (84 %), the
procedure could be completed with clinical success. For this matter, we believe this
rate of device malfunction is acceptable.
The most frequent indication for eFTR in our cohort was T1 CRC (n = 221; 60.2 %) [3 ]. Although literature is scarce, our R0 resection rate for T1 CRCs of 88.2 % compares
favorably with previous studies [1 ]
[3 ].
In the group of patients with primary eFTR for T1 CRC, the R0 resection rate was 77.9 %
in our study. The retrospective study of Kuellmer and colleagues reported a lower
R0 resection rate of 60.9 % [1 ]. This difference in R0 resection rate could be explained by differences in the average
size of the resected lesion, 20 mm in the German study vs. 14 mm in ours. Besides,
in contrast to the prospective inclusion of lesions suspicious for T1 CRC in our study,
Kuellmer and colleagues included mainly non-lifting lesions that were initially classified
as benign but at histopathology were diagnosed as T1 CRC, suggesting the inclusion
of more complex lesions [1 ].
We believe eFTR, having several benefits compared with ESD, can be considered a valid
diagnostic and potentially therapeutic primary treatment option for T1 CRC. Besides
the advantage of demanding less procedural skill and time than ESD, obtaining a transmural
resection with eFTR optimizes histological assessment. Furthermore, eFTR enables safe
and radical endoscopic resection of deep submucosal invasive cancer. Recent studies
have reported a very low risk of LNM (1 % – 2 %) in T1 CRC, with deep submucosal invasion
as the only risk factor for LNM [26 ]
[27 ]. These findings underscore the need for clinical trials to assess the role of eFTR
in T1 CRC treatment.
Resection of scars after previously incomplete T1 CRC resection accounted for a large
subgroup in our registry. Positive or indeterminate resection margins (R1/Rx) are
associated with residual cancer [28 ]
[29 ]. Therefore, current guidelines advise oncological surgery after R1 /Rx resection,
even in the absence of histological risk factors for LNM [17 ]
[30 ]. eFTR allows transmural resection of the scar with an R0 resection rate of 93.0 %,
which is comparable with the 87.5 % rate in Kuellmer’s study [1 ]. Despite the fact only scar tissue and no residual cancer was found in 81.8 % at
histopathology, we believe eFTR can confirm local radicality by enabling a transmural
scar excision or serve as a potentially curative completion treatment where there
is residual cancer. However, currently the ability to histologically confirm complete
scar excision is lacking, and therefore in clinical practice one is relying on the
macroscopic completeness of scar resection. However, long-term outcomes are lacking
and further data are therefore warranted.
The second most prevalent indication in our cohort was “difficult polyps,” consisting
of non-lifting polyps or polyps that were involving difficult locations. Overall,
the R0 resection rate for this subgroup was 70.5 %, which was lower compared with
the previously reported R0 rate of 77.7 % in the study by Schmidt et al. [3 ]. However, we included mainly recurrent lesions after previously incomplete resections.
As mentioned, submucosal scarring can limit adequate tissue mobilization into the
cap [6 ]
[11 ]. From our experience, mobilization of recurrent lesions after previous EMR can be
challenging. In 14 patients (3.9 %), eFTR could not be performed because the lesion
could not be mobilized into the cap. Of these, 12 were scarred lesions previously
treated with EMR. In certain cases, a dummy cap may help to improve patient selection.
Adverse events occurred in 9.3 %, with 2.7 % being classified as severe. This severe
adverse event rate requiring emergency surgery is comparable with previous studies
(Schmidt et al., 2.2 %; Kuellmer et al., 3.8 %) [1 ]
[3 ]. Nevertheless, future efforts will hopefully further decrease complication rates.
The most feared complication is a delayed perforation, which occurred in 1.6 % of
procedures and predominantly in the left-sided colon. Owing to partial wall excision,
a relative stenosis can occur at the level of the clip. As fecal content is more solid
in the distal colon, high pressures at the level of the clip could contribute to tissue
disintegration or rupture. None of the three patients with delayed sigmoid perforation
in our registry received post-procedural stool softeners. Although debatable, we believe
prescribing post-procedural laxatives might reduce the delayed perforation risk in
left-sided lesions.
Another cause of severe adverse event was appendicitis, occurring in 20.0 % of procedures
for appendiceal lesions without previous appendectomy. Closure of the remaining appendix
by the clip carries a risk of acute appendicitis, most likely from retained mucus
[13 ]. Schmidt and colleagues reported a lower secondary appendicitis rate of 8.8 % [3 ]. There is no clear explanation for this discrepancy, although both are based on
small subgroups. Larger cohorts with adequate follow-up are required to elucidate
the position of eFTR in appendiceal lesions. In the meantime, patients should be well
informed about the risk of secondary appendicitis.
Our study has several limitations that should be addressed. First, because this study
is based on registry data, we cannot exclude potential selection bias and must rely
on accurate data recording in all 20 participating centers. Second, follow-up data
were not complete because not all surveillance colonoscopies had been performed or
recorded yet. However, this registry started in a strong and transparent collaboration
of 20 academic and non-academic centers and represents a good overview of eFTR for
colorectal lesions in current clinical practice.
In conclusion, eFTR is an exciting innovative resection technique that is clinically
feasible and safe for complex colorectal lesions (≤ 20 mm), with the potential to
obviate the need for surgical resection. Further efficacy studies on eFTR as a primary
and secondary treatment option for T1 CRC are needed, focusing on both the short-
and long-term oncological results.
