Introduction
Acute nonvariceal upper gastrointestinal bleeding (NVUGIB) is a common and challenging
clinical presentation, with up to 10% mortality [1]
[2]. The most frequent cause of acute NVUGIB is gastroduodenal peptic ulcer disease
[1]
[2]
[3].
European Society of Gastrointestinal Endoscopy (ESGE) guidelines recommend standard
therapies (i.e. mechanical with through-the-scope [TTS] clip, thermal or sclerosing),
with diluted epinephrine injection, as first-line endoscopic treatment of actively
bleeding (Forrest Ia and Ib) peptic ulcers [1]
[4]. Given the high risk of rebleeding, ESGE recommends standard therapy with or without
epinephrine injection for Forrest IIa peptic ulcers, and adherent clot removal followed
by endoscopic hemostasis of underlying active bleeding or nonbleeding visible vessel
for Forrest IIb peptic ulcers [1]
[4]. Different standard endoscopic therapies are similarly effective in achieving hemostasis
[4]
[5]
[6]. However, unsuccessful immediate hemostasis and rebleeding are burdened with increased
mortality and may require angiographic or surgical treatment [7]
[8]
[9].
The over-the-scope (OTS) clip (OTSC; Ovesco Endoscopy AG, Tübingen, Germany) was originally
designed for the endoscopic closure of perforations, fistulas, and anastomotic leaks
[10]. OTS clips have been shown to be more effective than standard treatment as rescue
therapy for acute peptic ulcer rebleeding [11], and this indication has been included in the ESGE guidelines [1]
[4]. Recently, OTS clips have also been assessed as first-line endoscopic therapy of
NVUGIB [12]
[13]
[14]
[15], and randomized controlled trials (RCTs) have demonstrated that first-line treatment
with OTS clips is more effective than standard therapies in acute NVUGIB from several
different causes [14]
[15]
[16]. A recent meta-analysis on acute NVUGIB from different etiologies showed that, compared
with standard therapy, hemostasis with OTS clips is associated with a higher rate
of effective durable hemostasis and a lower rate of 30-day rebleeding [17]. We aimed to compare the efficacy of OTS clips with that of TTS clips as first-line
mechanical treatment in the specific setting of peptic ulcer bleeding.
Methods
Trial design
This was an international, multicenter, parallel, and open-label RCT (“TTS clip vs. OTS clip as first-line endoscopic treatment of Peptic ulcer bleeding, TOP Study). The study protocol was approved by the institutional
review board at each center, and the study was conducted in accordance with the declaration
of Helsinki. Written and informed consent was obtained from all patients before enrollment.
The study is reported following the recommendations of the Consolidated Standards
of Reporting Trials (CONSORT) Statement guidelines [18] (see Table 1s in the online-only Supplementary material).
Participating centers and endoscopists
The study was conducted at five European centers (nonacademic and academic, four in
Italy and one in Spain), since October 2018, with different starting times at each
center following approval from the local ethics committee. Before the start of the
study, a preliminary investigator meeting was organized with other endoscopy units
in Italy.
We did not define a minimum number of overall procedures performed as a requirement
for an endoscopist to participate in our RCT, as ESGE guidelines have not established
the minimum number of cases required to certify the ability to manage NVUGIB and to
correctly use endoscopic hemostatic devices [19]. However, only endoscopists who had been routinely performing treatment for acute
upper gastrointestinal bleeding (UGIB) for ≥5 years and had already placed >20 OTS
clips for nonbleeding indications and >10 OTS clips for UGIB could participate in
our RCT.
Patients
All patients with clinically suspected acute (<24 hour) NVUGIB were screened for enrollment.
Written informed consent was obtained prior to esophagogastroduodenoscopy (EGD). Inclusion
criteria were age ≥18 years, American Society of Anesthesiologists (ASA) score I–IV
before endoscopy, and ability to give informed consent. Exclusion criteria were age
<18 years, ASA score V, and pregnancy or breastfeeding.
Patients underwent EGD with a standard or therapeutic gastroscope. Endoscopy was performed
under deep sedation or after endotracheal intubation. Forrest classification was used
to characterize gastroduodenal peptic ulcers. In cases of endoscopically confirmed
active bleeding (Forrest Ia: spurting; Forrest Ib: oozing) or high risk (Forrest IIa:
visible vessel) or Forrest IIb (adherent clot) gastroduodenal peptic ulcer, patients
were enrolled and randomized by the treating medical team intraprocedurally to receive
mechanical endoscopic treatment with OTS or TTS clips.
Randomization
The random allocation sequence was generated in the coordinating center (Azienda USL
di Modena, Carpi Hospital, Italy) by means of computer-generated random numerical
series (https://wwwservizi.regione.emilia-romagna.it/generatore/). Randomization was
done in blocks with lists of 20 numbers per center, with “1” (odd number) encoding
for OTS clip and “2” (even number) for TTS clip. Subsequently, 1:1 randomization lists
were distributed in every center. Neither the patient nor the treating physicians
or endoscopists were blinded.
Interventions and definitions
In the OTS clip group, Forrest Ia, Ib, IIa, and any underlying active bleeding (Forrest
Ia or Forrest Ib) or nonbleeding visible vessel (Forrest IIa) identified after adherent
clot removal in Forrest IIb peptic ulcers, were treated with OTS clips. The OTS clip
systems available in the participating centers were 10/6a, 10/6t, 11/6a, 11/6t, 12/6a,
and 12/6t types, as defined by clip size (10, 11, or 12 mm), depth of cap (6 mm),
and shapes of clip teeth (i.e. traumatic [t, teeth with small spikes] or atraumatic
[a, blunt teeth]). The choice of OTS clip system was left to the discretion of the
endoscopist. After placing the bleeding lesion at the center of the applicator cap
and suctioning, the OTS clip was deployed to obliterate the bleeding point. There
were no limitations on the number of OTS clips used in each procedure.
In the TTS clip group, in accordance with ESGE guidelines [1]
[4], Forrest Ia and Ib peptic ulcers were treated with combined endoscopic therapy with
TTS clip and diluted epinephrine injection, while Forrest IIa peptic ulcers were treated
with TTS clip monotherapy or with combined treatment with diluted epinephrine injection.
For Forrest IIb peptic ulcers, any underlying active bleeding (Forrest Ia or Forrest
Ib) or nonbleeding visible vessel (Forrest IIa) identified after adherent clot removal
was treated as described above [1]
[4]. The endoscopist was allowed to decide which volume of diluted epinephrine to inject,
and which of the TTS clips available at each center to use. There were no limitations
on the number of TTS clips used in each procedure.
Successful initial hemostasis was defined as absence of bleeding after at least 1
minute of observation, verified using the timer on the endoscopy screen, after the
effective application of the assigned endoscopic therapy. Thirty day rebleeding was
defined as newly onset clinical and/or laboratory signs of acute UGIB with endoscopic
evidence of active bleeding from the previously treated peptic ulcer within 30 days
after successful initial hemostasis. Finally, overall clinical success was defined
as the composite of successful initial hemostasis and no evidence of 30-day rebleeding.
In cases of unsuccessful hemostasis or 30-day rebleeding, the endoscopist was allowed
to choose any rescue endoscopic treatment (TTS clip, OTS clip, hemostatic powders
or forceps, thermal therapy, sclerosing agents). In cases of further bleeding, patients
would be referred for angiographic embolization or surgery.
Pre- and post-treatment photographic documentation of all endoscopic procedures was
recorded.
Periprocedural management and follow-up
Before and after endoscopic treatment, patients were managed according to ESGE guidelines
regarding resuscitation, timing of endoscopy, proton pump inhibitor administration,
and anticoagulant/antiplatelet discontinuation [1]
[4].
Every patient enrolled was re-evaluated 30 days after the index EGD by phone call
or scheduled outpatient visit, to assess for clinical signs of UGIB, hospital admissions,
and repeat endoscopies.
Outcomes
The primary end point was the rate of 30-day rebleeding after successful initial hemostasis.
Secondary end points were successful initial hemostasis rate, repeat EGD due to clinical
and/or laboratory signs of rebleeding after successful initial hemostasis, overall
clinical success rate, need for blood transfusion and number of red blood cell units
transfused, length of hospital stay, 30-day mortality rate, and complications associated
with endoscopic therapy.
Sample size calculation
Based on the results of previous studies [13]
[14], upon estimating a ≥20% difference in 30-day rebleeding rate in favor of OTS clip
vs. TTS clip, with 80% power, 10% dropout rate, and a <0.05 significance P value, we calculated that a minimum of 49 patients per treatment arm were required
(total number of patients required = 98).
Data management and statistical analysis
Data were recorded on web-based electronic case report forms by the participating
physicians and extracted by the coordinating study center (Azienda USL di Modena)
for analysis. Intention-to-treat analysis was set as the primary analysis. Thus, patients
were analyzed in the intervention groups to which they were randomized, regardless
of the intervention that they eventually received [20]. Per-protocol analyses of 30-day rebleeding and repeat EGD due to clinical and/or
laboratory signs of rebleeding after successful initial hemostasis rates were also
performed.
Continuous variables were expressed as median with range or mean with SD. Categorical
variables were reported as frequencies and percentages unless stated otherwise. For
continuous variables, differences were determined using two-sample Wilcoxon rank sum
(Mann–Whitney test), whereas chi-squared test or Fisher’s exact test was used for
categorical variables. Relative risk (RR) was calculated using a generalized linear
model with a log-binomial link. We used maximum likelihood estimation for estimating
the parameters and a robust estimation of variance with a dichotomous covariate to
indicate the treatment. A P value of <0.05 was considered significant. Kaplan–Meier curve with log-rank test
was used for 30-day rebleeding analyses. No interim analyses were planned.
Statistical analyses were performed with STATA/SE 16.1 for Windows software (StataCorp,
College Station, Texas, USA).
All authors had access to the study data, and reviewed and approved the final manuscript.
Protocol amendments
“Overall clinical success” was included as a secondary outcome. This amendment ensued
data reappraisal in view of the recommendations by Laine et al. [21]. While original end points of our RCT were formulated during a consensus meeting
among the participating investigators, this composite end point reflects subsequent
scientific evidence on this topic, is patient centered, and may have relevant generalizability
[22]. No changes to the database were made after this outcome was added.
Results
Patient and lesion characteristics
Between October 2018 and October 2022, 251 patients were screened. Among these, 112
patients who met the inclusion criteria and had Forrest Ia–IIb gastroduodenal peptic
ulcer at EGD were enrolled and randomized to treatment with OTS clips (n = 61) or
TTS clips (n = 51) ([Fig. 1]). After clot removal, all the 15 Forrest IIb peptic ulcers identified turned out
to be Forrest Ia-b or IIa peptic ulcers. Patient and lesion characteristics are shown
in [Table 1] and [Table 2], respectively. All patients were treated with high dose intravenous proton pump
inhibitors periprocedurally.
Fig. 1 Consolidated Standard of Reporting Trials (CONSORT) flow diagram showing enrollment
and analysis. 1Arteriovenous vascular malformation, black esophagus, Cameron lesion, Los Angeles
grade C–D esophagitis, Mallory–Weiss tear, nasogastric tube-associated pressure ulcer,
ulcerated subepithelial lesion. MO, Azienda USL di Modena; NI, Niguarda Hospital,
Milan; CB, Campus Bio-Medico University Hospital, Rome; SP, ASST San Paolo Hospital,
Milan; MD, Madrid University Hospital.
Table 1 Patient characteristics.
|
|
OTS clip (n = 61)
|
TTS clip (n = 51)
|
|
OTS, over-the-scope; TTS, through-the-scope; ASA, American Society of Anesthesiologists;
NSAID, nonsteroidal anti-inflammatory drug; DOAC, direct oral anticoagulant; UGIB,
upper gastrointestinal bleeding; INR, international normalized ratio.
|
|
Age, median (range), years
|
76 (19–93)
|
75 (46–95)
|
|
Sex, n (%)
|
|
|
44 (72.1)
|
33 (64.7)
|
|
|
17 (27.9)
|
18 (35.3)
|
|
ASA score
|
|
|
3 (4.9)
|
3 (5.9)
|
|
|
13 (21.3)
|
15 (29.4)
|
|
|
39 (63.9)
|
23 (45.1)
|
|
|
6 (9.9)
|
10 (19.6)
|
|
History of peptic ulcer, n (%)
|
13 (21.3)
|
9 (17.7)
|
|
History of NSAIDs, n (%)
|
11 (18.0)
|
10 (19.6)
|
|
Cardioaspirin, n (%)
|
10 (16.4)
|
12 (23.5)
|
|
Dual anti-platelet therapy, n (%)
|
5 (8.2)
|
1 (2.0)
|
|
Warfarin, n (%)
|
7 (11.5)
|
5 (9.8)
|
|
DOACs, n (%)
|
6 (9.8)
|
7 (13.7)
|
|
Admission due to UGIB, n (%)
|
41 (67.2)
|
34 (66.7)
|
|
Bleeding during hospital stay, n (%)
|
20 (32.8)
|
17 (33.3)
|
|
Syncope, n (%)
|
15 (24.6)
|
13 (25.5)
|
|
Hemodynamic instability at randomization, n (%)
|
29 (47.5)
|
24 (47.1)
|
|
INR, mean (SD)
|
1.3 (0.4)
|
1.3 (0.5)
|
|
Baseline hemoglobin, mean (SD), g/dL
|
8.6 (2.0)
|
8.7 (2.3)
|
Table 2 Lesion characteristics
|
|
OTS clip (n = 61)
|
TTS clip (n = 51)
|
|
OTS, over-the-scope; TTS, through-the-scope.
|
|
Peptic ulcer location, n (%)
|
|
|
17 (27.9)
|
11 (21.6)
|
-
|
6
|
3
|
-
|
4
|
3
|
-
|
4
|
3
|
-
|
2
|
1
|
-
|
1
|
1
|
|
|
44 (72.1)
|
40 (78.4)
|
-
|
27
|
29
|
-
|
11
|
10
|
-
|
6
|
1
|
|
Peptic ulcer size, median (range), mm
|
12 (5–40)
|
12 (3–40)
|
|
|
48 (78.7)
|
37 (72.6)
|
|
|
13 (21.3)
|
14 (27.4)
|
|
Forrest classification, n (%)
|
|
|
3 (4.9)
|
0 (0)
|
|
|
13 (21.3)
|
12 (23.5)
|
|
|
37 (60.7)
|
32 (62.8)
|
|
|
8 (13.1)
|
7 (13.7)
|
|
Helicobacter pylori status, n (%)
|
|
|
13 (21.3)
|
11 (21.6)
|
|
|
14 (23.0)
|
14 (27.5)
|
|
|
34 (55.7)
|
26 (51.0)
|
Endoscopic treatment
A single OTS clip was used for each lesion in all but one patient, who was successfully
treated with two OTS clips. In the TTS clip group, a median of 2 (range 1–8) TTS clips
were used for each lesion. Injection of diluted epinephrine was performed before mechanical
therapy in 12/12 Forrest Ib, 14/32 Forrest IIa, and 2/7 Forrest IIb TTS clip-treated
gastroduodenal peptic ulcers.
Primary outcome
All patients completed the 30-day follow-up. The rates of 30-day rebleeding after
successful initial hemostasis were 1.6% (1/61) and 3.9% (2/51) for patients treated
with OTS clips and TTS clips, respectively (Kaplan–Meier log-rank, P = 0.46) ([Fig. 2]
a, [Table 3]), with RR = 0.42 (95%CI 0.04–4.53, P = 0.47). In a per-protocol analysis restricted to patients with successful initial
hemostasis, 30-day rebleeding rates were 1.7% (1/60) and 5.0% (2/40) for OTS clip
and TTS clip groups, respectively (Kaplan–Meier log-rank, P = 0.35) (Table 2s).
Fig. 2 Kaplan–Meier curves in patients treated with over-the-scope (OTS) or through-the-scope
(TTS) clips. a 30-day rebleeding. b Repeat esophagogastroduodenoscopy due to clinical signs of rebleeding.
Table 3 Analysis of main study outcomes.
|
Outcome
|
OTS clip (n = 61)
|
TTS clip (n = 51)
|
P value
|
|
OTS, over-the-scope; TTS through-the-scope; EGD, esophagogastroduodenoscopy.
|
|
Successful initial hemostasis, n (%)
|
60 (98.4)
|
40 (78.4)
|
0.001
|
|
Reasons for unsuccessful initial hemostasis, n (%)
|
|
|
0 (0)
|
7 (13.7)
|
|
|
|
1 (1.6)
|
3 (5.9)
|
|
|
|
0 (0)
|
0 (0)
|
|
|
|
0 (0)
|
1 (2.0)
|
|
|
Repeat EGD due to clinical signs of rebleeding, n (%)
|
9 (14.8)
|
7 (13.7)
|
0.88
|
|
30-day rebleeding, n (%)
|
1 (1.6)
|
2 (3.9)
|
0.46
|
|
Overall clinical success, n (%)
|
59 (96.7)
|
38 (74.5)
|
0.001
|
Rescue endoscopic treatment was performed successfully in patients who experienced
30-day rebleeding. One duodenal post-bulbar Forrest Ib peptic ulcer, which re-bled
1 day after OTS clip first-line treatment, was treated with TTS clip placement. One
duodenal bulb anterior wall Forrest IIa peptic ulcer, which re-bled 2 days after TTS
clip first-line treatment, was treated with OTS clip placement. One gastric Forrest
IIb peptic ulcer, which re-bled 29 days after TTS clip first-line treatment, was treated
with thermal therapy.
Secondary outcomes
Rates of successful initial hemostasis were 98.4% (60/61) and 78.4% (40/51) in patients
treated with OTS clips and TTS clips, respectively (P = 0.001) ([Table 3], [Fig. 3]). Generalized linear model with log-binomial distribution confirmed that OTS clips
had a 25% higher efficacy compared with TTS clips in achieving successful hemostasis
(RR = 1.25, 95%CI 1.08–1.45, P = 0.003).
Fig. 3 Examples of successful first-line treatment of peptic ulcers with over-the-scope (OTS)
clip. Endoscopic images showing Forrest IIa duodenal bulb peptic ulcers: a, b, c before OTS clip placement; d, e, f after OTS clip placement.
In the OTS clip group, unsuccessful hemostasis occurred in one patient with a Forrest
IIb 15-mm peptic ulcer in the lesser curvature of the gastric body; failure was due
to misplacement, and bleeding was successfully managed during the same EGD with diluted
epinephrine injection and TTS clip placement. In the TTS clip group, unsuccessful
hemostasis occurred in seven fibrotic Forrest IIa peptic ulcers, three duodenal bulb
posterior wall (one Forrest Ib and two Forrest IIa) peptic ulcers, and one duodenal
bulb anterior wall Forrest Ib peptic ulcer with a large visible bleeding vessel. Among
these, 10 cases were successfully treated with OTS clips, and one was successfully
treated with a combination of OTS clip, thermal therapy, and hemostatic powder ([Table 3], Table 3s).
Among patients with successful initial hemostasis, 17 (10 and 7 initially treated
with OTS clip and TTS clip, respectively) underwent repeat EGD within 30 days from
the initial treatment. Repeat EGD was performed due to clinical signs of rebleeding
with endoscopic evidence of active bleeding from the previously treated peptic ulcer
(OTS clip group 1/61, TTS clip 2/51), clinical signs of rebleeding with no endoscopic
evidence of rebleeding (OTS clip 8/61, TTS clip 5/51), and other indication (OTS clip
1/61).
EGD was repeated due to clinical and/or laboratory signs of rebleeding after successful
initial hemostasis (with or without evidence of active bleeding from the previously
treated peptic ulcer) in 14.8% (9/61) and 13.7% (7/51) of patients treated with OTS
clips and TTS clips, respectively (Kaplan–Meier log-rank, P = 0.88) ([Table 3], [Fig. 2]
b). Per-protocol analysis showed that the rates of repeat EGD due to clinical signs
of rebleeding after successful initial hemostasis were 15.0% (9/60) and 17.5% (7/40)
for the OTS clip group and TTS clip group, respectively (Kaplan–Meier log-rank, P = 0.74) (Table 2s).
Overall clinical success rates were 96.7% (59/61) and 74.5% (38/51) in patients treated
with OTS clips and TTS clips, respectively (P = 0.001) ([Table 3]), with a higher RR in the OTS clip group (RR = 1.30, 95%CI 1.10–1.53, P = 0.002).
Blood transfusion was performed in 29.5% (18/61) and 33.3% (17/51) of patients treated
with OTS clips and TTS clips, respectively (P = 0.66). The median number of red blood cell units transfused per patient was 2 (range
0–10) in the OTS clip group and 2 (range 0–12) in the TTS clip group (P = 0.85).
The median length of hospital stay was 7 days (range 2–68) in the OTS clip group and
6 days (0–52) in the TTS clip group (P = 0.92).
Overall 30-day mortality rates were 1.6% (1/61) and 7.8% (4/51) in patients treated
with OTS clips and TTS clips, respectively (P = 0.18). Causes of death were not related to acute UGIB in any patient. No deaths
were observed in patients who experienced 30-day rebleeding. One patient treated with
an OTS clip died from septic shock due to pneumonia. In the TTS clip group, one patient
died from septic shock due to pneumonia, one from urosepsis and heart failure, one
from pulmonary edema secondary to end-stage renal failure, and one from metastatic
hepatocellular carcinoma.
No complications associated with endoscopic therapy were reported in either group.
Discussion
This first head-to-head RCT on OTS clips vs. TTS clips as first-line endoscopic treatment
for peptic ulcer bleeding showed that OTS clips were not superior to TTS clips regarding
30-day rebleeding rate. However, OTS clips showed higher efficacy than TTS clips in
terms of successful initial hemostasis and overall clinical success rates.
During the past few years, the OTS clip has been evaluated as first-line treatment
for acute NVUGIB [12]
[13]
[14]
[15]
[16]
[17]
[23]. Jensen et al. showed that OTS clips are more effective than standard endoscopic
therapy as first-line treatment of peptic ulcer and Dieulafoy lesions in terms of
rebleeding and severe complication rates [14]. The STING-2 RCT reported that first-line therapy with OTS clips is superior to
standard treatment with TTS clips or thermal therapy in achieving successful durable
hemostasis in acute NVUGIB with high rebleeding risk [15]. Finally, Lau et al. showed that initial treatment of acute NVUGIB with OTS clips
is associated with a lower probability of 30-day rebleeding compared with standard
therapy [16]. Conversely, another RCT reported that first-line treatment of ≥15-mm peptic ulcer
with OTS clips is not different from standard therapy in terms of the 30-day rebleeding
rate [23].
We observed that, once successful initial hemostasis was achieved, 30-day rebleeding
rates were remarkably low in both treatment groups, especially compared with recently
published RCTs on OTS clips as first-line endoscopic treatment of NVUGIB [14]
[15]
[16]. This is likely to be related to the stringent design of our study, as, in addition
to clinical and laboratory signs of acute NVUGIB, our definition of 30-day rebleeding
included subsequent endoscopic evidence of active bleeding from the previously treated
peptic ulcer. Overall 30-day mortality rates were 1.6% and 7.8% in the OTS clip and
TTS clip treatment groups, respectively. These data are in keeping with other studies
on endoscopic treatment of acute NVUGIB [11]
[15]
[24]. Moreover, in our study, those who died during the follow-up period were ASA IV
patients with multiple comorbidities and none of the causes of death were related
to bleeding.
The rate of successful initial hemostasis was higher in patients treated with OTS
clips compared with those receiving TTS clips (98.4% vs. 78.4%). Although in our study
most failed TTS clip cases were Forrest IIa peptic ulcers, which are not actively
bleeding at the time of the index EGD, our definition of “successful initial hemostasis”
is in line with previously published recommendations on methodology for RCTs on NVUGIB,
stating that “prevention of further bleeding is the primary clinical goal in patients
with NVUGIB” and that “therapy is required for patients with ulcers with active bleeding
or nonbleeding visible vessels” [21]. Furthermore, the terminology of our outcome definition and our inclusion criteria
are in keeping with recently published RCTs on endoscopic treatment of NVUGIB [15]
[16]
[25]. Most cases of unsuccessful hemostasis with TTS clips occurred either in fibrotic
or posterior duodenal bulb wall peptic ulcers, both representing well-established
issues for TTS clip application. It is worth noting that the OTS clip exerts a circumferential
and even compression on the treated tissue, with the bear claw design providing effective
anchoring to fibrotic tissue [26]
[27]. Moreover, the OTS clip applicator cap on the tip of the endoscope may help maintain
a more stable position even in difficult and potentially distorted locations such
as the posterior duodenal bulb wall. Indeed, ESGE guidelines state that OTS clips
should be considered as first-line therapy of peptic ulcers in cases of difficult
location or fibrotic features [4], although this was labeled as a weak recommendation with low-quality evidence.
In keeping with our successful initial hemostasis results, the rate of overall clinical
success was higher in patients treated with OTS clips compared with those receiving
TTS clips (96.7% vs. 74.5%). We believe that this is a very important outcome, as
the ultimate goal of endoscopic treatment in patients with NVUGIB is to maintain hemostasis
in the long term.
Unlike previously published data, our study has important strengths, including the
fact that we performed a head-to-head comparison between two mechanical therapies,
with no other endoscopic treatments permitted besides diluted epinephrine injection,
and a more stringent study design. Whereas in other RCTs OTS clips were tested on
heterogeneous types of bleeding lesions, including severe esophagitis, Mallory–Weiss
tears, and Dieulafoy lesions [15]
[16], only patients with Forrest Ia–IIb gastroduodenal peptic ulcers were enrolled in
our study.
Similarly to other recently published RCTs on OTS clips vs. TTS clips in acute NVUGIB
[11]
[15]
[16]
[23], a limitation of our RCT is that, following TTS clip failure, crossover treatment
with OTS clips was allowed. However, restricting possible endoscopic rescue therapies
to other standard treatments would raise ethical concerns, as the OTS clip has already
been shown to be superior to standard treatment as rescue therapy for recurrent peptic
ulcer bleeding [1]
[4]
[11]. Furthermore, in keeping with recently published RCTs on endoscopic therapy of NVUGIB
[14]
[16], we did not set a minimum number of TTS clips to be used before deciding that treatment
had failed, because in the guidelines there are no recommendations on the minimum
number of TTS clip applications needed for effective hemostasis, and also to give
priority to the clinician’s judgement. Indeed, TTS clip failures in our RCT were mainly
in peptic ulcers in difficult locations (gastric lesser curvature, posterior bulb)
or those with fibrosis, where it would have been clear from the first TTS clip applied
that this treatment would be ineffective regardless of the number of TTS clips used,
either due to nonadherence or to impossibility of applying the assigned therapy.
We were unable to demonstrate a statistically and clinically significant difference
between OTS clips and TTS clips with regard to the primary outcome of 30-day rebleeding,
and higher patient numbers may be required to further assess our primary outcome.
However, it is important to highlight the higher efficacy of OTS clips compared with
TTS clips in achieving successful initial hemostasis, despite this being a secondary
outcome, and this evidence could be considered as hypothesis generating [28]
[29]. Taken together, the results of our RCT suggest that OTS clips display higher efficacy
over TTS clips as first-line therapy of acute peptic ulcer bleeding at the time of
the index EGD, probably due to the inability to place TTS clips in some fibrotic and/or
duodenal posterior peptic ulcers, with consequent insufficient endoscopic hemostatic
treatment. Widespread use of OTS clips over TTS clips as first-line treatment of bleeding
peptic ulcers may raise concerns related to possible overtreatment and cost-effectiveness,
as one OTS clip generally costs 5–10 times more than one TTS clip; however, the potential
pharmacoeconomic benefit of TTS clips may become negligible upon considering the lower
efficacy of TTS clips, with substantial growth in indirect costs and required resources
[30].
In conclusion, our RCT provides important evidence on first-line endoscopic mechanical
treatment of acute peptic ulcer bleeding. The rates of 30-day rebleeding were low
in both treatment arms, and OTS clips were not superior to TTS clips in prevention
of 30-day rebleeding. However, we observed that OTS clips have higher efficacy than
TTS clips in achieving successful initial hemostasis, especially in fibrotic peptic
ulcers or those located in the posterior duodenal bulb wall.
