Background
During previous decades, the treatment and management of non-variceal upper gastrointestinal
bleeding (NVUGIB) have substantially improved, with endoscopic treatment being the
first-line modality. After the index endoscopy, rebleeding occurs in up to 20 % of
cases [1 ], with a mortality rate of 10 % [2 ]. Recurrent bleeding after endoscopic therapy is associated with significant mortality,
with a higher risk in older populations and those with multiple comorbidities. This
trend may be attributable to the rising comorbidity in NVUGIB patients and the increasing
use of antithrombotic drugs [3 ].
Therefore, there is a need to develop additional medical therapies that will improve
the maintenance of hemostasis. The European Society of Gastrointestinal Endoscopy
(ESGE) guidelines [4 ] recommend (strong recommendation, high-quality evidence) combining epinephrine injection
with a second hemostasis modality (thermal contact, mechanical therapy, or injection
of a sclerosing agent), especially for actively bleeding ulcers. The over-the-scope
clip (OTSC® , Ovesco Endoscopy GmbH, Tübingen, Germany) is a novel tool that can securely hold
a larger volume of tissue and to a greater depth with respect to the standard through-the-scope
clip (TTS) [5 ]
[6 ]
[7 ]. To the best of our knowledge, there are no comparative studies on the efficacy
of OTSC and other hemostatic methods for first-line hemostasis. Thus, we aimed to
compare first-line endoscopic hemostasis achieved using conventional techniques versus
that obtained using OTSC placement for NVUGIB.
Materials and methods
Study population
From January 2007 to March 2018, 793 consecutive patients underwent upper endoscopy
with the hemostasis procedure for NVUGIB. The inclusion criteria were as follows:
age > 18 years, NVUGIB related to ulcers, Mallory Weiss lesion, Dieulafoy lesion,
anastomotic bleeding, or angioectasia. The exclusion criteria were: incomplete clinical
information, other causes of bleeding (post-sphincterotomy bleeding, post-polypectomy
bleeding, malignancy, hemorrhagic gastritis, or watermelon stomach), or endoscopic
hemostasis with only epinephrine injection because the ESGE recommends (strong recommendation
and with high-quality evidence) that epinephrine injection therapy should not be used
as endoscopic monotherapy. We collected data with regard to the following variables:
age, sex, year of bleeding, number of major comorbidities (cardiac failure, ischemic
heart disease, asthma, chronic obstructive pulmonary disease, diabetes mellitus, rheumatoid
arthritis, liver failure, renal failure, disseminated malignancy, pneumonia, dementia,
recent major operation, cerebrovascular disease, hematological malignancy, hypertension,
trauma/burns, other cardiac disease, major sepsis, and/or other liver disease), anticoagulant/antithrombotic
therapy, site of bleeding (esophagus, stomach, duodenum, and/or anastomosis), Forrest
classification [8 ], hemostasis technique (epinephrine with/without TTS, OTSC, thermic device, or sclerosing
agent) for the most severe lesion according to the Forrest classification, adverse
events related to the hemostasis technique used, Rockall Score [9 ], Helicobacter pylori infection (assessed using biopsy or fecal antigen), rebleeding rate, rebleeding from
a different site, rescue hemostasis technique (endoscopic, arterial embolization,
or surgery), mortality rate within 30 days, and hospitalization (days). The study
was approved by the Ethics Committee of the University of Modena on 10 May 2018 (Prot
AOU 0011529/18).
Description of the procedure
All of the endoscopic procedures were performed in an inpatient setting, under anesthesia-assisted
deep sedation by a single, skilled operator. Hemodynamically unstable patients were
adequately resuscitated before they underwent upper endoscopy with crystalloid/colloid
infusion and erythrocyte concentrate transfusion if needed. Patients with a non-cirrhosis
related coagulopathy and with a prolonged prothrombin time with an international normalized
ratio (INR) > 2.0 were transfused with fresh frozen plasma. The use of prothrombin
complex concentrate infusions was preferred for patients with serious/life-threatening
bleeding. We performed upper endoscopy once the INR was < 2.5. Before endoscopy, the
patients received an intravenous bolus of proton pump inhibitor (pantoprazole 80mg),
followed, if needed, by constant infusion (8 mg/hour).
Early endoscopy (within 24 hours) was performed in all cases with either a diagnostic
(9.2-mm) or a therapeutic (10-mm) endoscope (Pentax Medical, Tokyo, Japan). In order
to achieve endoscopic hemostasis, in addition to epinephrine injection, we used thermal
modalities (argon plasma coagulation, ERBE, VIO® , Tuebingen, Germany), mechanical therapy with TTS clip (QuickClip2, Olympus® , Tokyo, Japan; Resolution Clip, Boston Scientific® , Natick MA, USA; DuraClip, ConMed® , Greenwood, USA; SureClip, Micro-Tech® , Anna Arbor, Mi, USA) and sclerosing agents, based on the choice of the endoscopist.
When we used the OTSCs, the endoscope was extracted and equipped with the OTSC system.
The OTSC size (11 or 12 mm) and type were chosen by the endoscopist. The 11-mm and
12-mm OTSCs were used with both “diagnostic” (9.2 mm, working channel 2.8 mm) gastroscope
and “therapeutic” (10 mm, working channel 3.7 mm) gastroscope. The OTSC was deployed
on the lesion either with suction or after tissue retraction into the cap with an
anchor device. In addition, an injection of epinephrine solution was allowed (but
not mandatory) before or after OTSC deployment.
Outcomes and clinical data
All of the data were retrospectively collected from medical records. The primary outcome
was the rebleeding rate (defined as occurrence of hematemesis, aspiration of blood
from the nasogastric tube, instability of arterial blood pressure of cardiac frequency,
and a fall of > 2 g/dL in the hemoglobin level) within 24 hours or within 30 days
after hemostasis (immediate or late bleeding). Secondary outcomes were the mortality
rate within 30 days and adverse events related to the hemostasis technique used.
Statistical analyses
A univariable analysis was conducted for all of the baseline characteristics presented
in [Table 1 ] and [Table 2 ]. Variables that differed significantly between other hemostasis techniques and OTSC
were used to create a propensity score so as to match the “conventional” group patients
with the OTSC group (1:1). A propensity score is the probability that a unit with
certain characteristics will be assigned to the treatment group (as opposed to the
control group). The scores can be used to reduce or eliminate selection bias in observational
studies by balancing the covariates between the treatment and control groups. Propensity
score matching [10 ] creates sets of participants for the treatment and control groups. A matched set
consists of at least one participant in the treatment group and one in the control
group with similar propensity scores. The goal is to approximate a random experiment.
Covariates to be included in the model were related to the outcome but not to the
exposure so as to increase the precision of the estimated exposure effect without
increasing the bias [11 ]. Calculating a propensity score is an iterative process. The t test (Stata statistical software, version 13, StataCorp, College Station, TX, United
States) was used to determine whether each covariate was balanced within each block.
Patients were matched using the nearest-neighbor method without replacement and with
a caliper width equal to 0.1.
Table 1
Characteristics of the patients in the unmatched and matched cohorts.
Unmatched cohort
Matched cohort
Characteristics
Other (215)
OTSC (112)
P value
Other (84)
OTSC (84)
P value
n[1 ]
95 %CI
n[1 ]
95 %CI
n[1 ]
95 %CI
n[1 ]
95 %CI
Age
71 ± 15
69 – 73
72 ± 14
69 – 75
0.5
70 ± 14
67 – 73
70 ± 14
67 – 73
0.8
Sex (male)
151 (70)
63 – 86
88 (79)
70 – 86
0.1
66 (79)
68 – 87
66 (79)
68 – 87
1
Year of bleeding
2012 ± 3
2015 ± 1
0.0001
2012 ± 3
2015 ± 1
0.0001
Major comorbidities
0.06
0.1
39 (18)
13 – 24
14 (13)
10 – 20
14 (17)
9 – 26
14 (17)
9 – 26
37 (17)
12 – 23
12 (11)
6 – 18
20 (24)
15 – 34
9 (11)
5 – 19
60 (28)
22 – 34
28 (25)
17 – 34
20 (24)
15 – 34
18 (21)
13 – 32
36 (17)
12 – 22
28 (25)
17 – 34
16 (18)
11 – 29
23 (27)
18 – 38
43 (20)
15 – 26
30 (26)
19 – 36
14 (17)
9 – 26
20 (24)
15 – 34
Antithrombotics/anticoagulants
0.3
0.8
58 (27)
21 – 33
26 (23)
16 – 32
21 (26)
16 – 36
24 (29)
19 – 39
64 (29)
24 – 36
40 (34)
27 – 45
26 (32)
21 – 42
28 (35)
23 – 44
12 (6)
3 – 9
3 (3)
1 – 8
5 (6)
2 – 13
2 (2)
1 – 8
6 (3)
1 – 6
3 (3)
1 – 8
1 (2)
1 – 6
3 (4)
1 – 10
15 (7)
4 – 11
2 (2)
1 – 6
4 (5)
1 – 12
2 (2)
1 – 8
31 (14)
10 – 20
22 (20)
13 – 29
13 (15)
8 – 25
13 (15)
8 – 25
25 (12)
8 – 17
14 (13)
10 – 20
12 (12)
8 – 24
11 (11)
7 – 22
4 (2)
1 – 5
2 (2)
1 – 6
2 (2)
1 – 8
1 (2)
1 – 6
Cause of bleeding
0.0001
0.8
31 (14)
10 – 20
4 (4)
1 – 9
7 (8)
3 – 16
4 (5)
1 – 12
116 (54)
47 – 60
88 (79)
70 – 86
61 (73)
62 – 82
63 (75)
64 – 84
35 (16)
12 – 22
10 (9)
4 – 16
9 (11)
5 – 19
10 (12)
6 – 21
15 (7)
4 – 11
0
0
0
0
0
0
18 (9)
5 – 13
10 (9)
4 – 16
7 (8)
3 – 16
7 (8)
3 – 16
Site of bleeding
0.02
0.9
39 (18)
13 – 24
12 (11)
6 – 18
9 (11)
5 – 19
11 (11)
7 – 22
66 (31)
25 – 37
25 (22)
15 – 31
26 (31)
21 – 42
23 (28)
18 – 38
92 (43)
36 – 50
67 (60)
51 – 69
42 (50)
39 – 61
44 (54)
41 – 63
18 (8)
5 – 13
8 (7)
3 – 14
7 (8)
3 – 16
6 (7)
3 – 15
Forrest classification
0.0001
0.2
10 (5)
2 – 8
11 (10)
5 – 17
4 (5)
1 – 12
10 (12)
6 – 21
100 (47)
40 – 53
48 (43)
33 – 53
42 (50)
39 – 61
35 (42)
31 – 53
68 (32)
25 – 38
53 (47)
39 – 57
38 (45)
34 – 57
39 (46)
36 – 58
30 (13)
10 – 19
0
0
0
0
0
0
5 (2)
1 – 5
0
0
0
0
0
0
2 (1)
1 – 3
0
0
0
0
0
0
Other hemorrhagic lesions
71 (33)
27 – 40
38 (34)
25 – 43
0.9
31 (37)
27 – 48
29 (35)
24 – 46
0.7
Rockall score
6.3 ± 1.7
6.1 – 6.5
7.0 ± 1.7
6.7 – 7.3
0.0007
6.3 ± 1.5
5.9 – 6.6
6.8 ± 1.7
6.4 – 7.2
0.06
ASA, acetylsalicylic acid; NSAID, non-steroidal anti-inflammatory drugs; OTSC, over-the-scope
clip; DAPT, dual antiplatelet therapy; LMWH, low molecular weight heparin; NOAC, novel
oral anticoagulants; VKA, vitamin K antagonist.
1 Values expressed as absolute number (%) or mean ± SD.
Table 2
Characteristics of the procedures and outcomes of the unmatched and matched cohorts.
Unmatched cohort
Matched cohort
Characteristics
Other (n = 215)
OTSC (n = 112)
P value
Other (n = 84)
OTSC (n = 84)
P value
n[1 ]
95 %CI
n[1 ]
95 %CI
n[1 ]
95 %CI
n[1 ]
95 %CI
Hemostatic technique
171 (80)
78 (93)
44 (20)
6 (7)
53 (47)
44 (52)
49 (44)
32 (38)
10 (9)
8 (10)
H. pylori infection
0.6
0.2
21 (10)
6 – 14
30 (27)
19 – 36
33 (39)
29 – 51
22 (26)
17 – 37
65 (30)
24 – 37
14 (12)
7 – 20
8 (10)
4 – 18
12 (14)
8 – 24
129 (60)
53 – 67
68 (61)
51 – 70
43 (51)
40 – 62
50 (60)
48 – 70
Adverse events (procedure related)
1 (1)
0
1
0
0
0.3
Rebleeding
32 (15)
10 – 20
13 (12)
6 – 19
0.5
17 (20)
12 – 30
7 (8)
3 – 16
0.02
Rebleeding from other site
2 (1)
0 – 3
4 (4)
1 – 9
0.2
1 (1)
0 – 6
3 (4)
1 – 10
0.6
Rescue hemostasis
0.8
0.2
19 (59)
41 – 76
8 (62)
32 – 86
12 (70)
44 – 90
3 (42)
10 – 82
6 (19)
7 – 36
2 (15)
2 – 45
3 (18)
4 – 43
2 (29)
4 – 71
7 (22)
9 – 40
3 (23)
5 – 54
2 (12)
2 – 37
2 (29)
4 – 71
Mortality (death within 30 days)
20 (9)
6 – 14
8 (7)
3 – 14
0.6
5 (6)
2 – 13
2 (2)
1 – 8
0.4
Recovery, days
13 ± 19
10 – 16
15 ± 12
13 – 17
0.4
11 ± 10
9 – 13
15 ± 3
14 – 16
0.03
OTSC, over-the-scope clip.
1 Values expressed as absolute number (%) or mean ± SD.
2 Thermal modalities, sclerosing agent.
Continuous data are expressed as mean ± standard deviation values, while the categorical
variables are presented as frequencies (%). Continuous variables were compared using
the Student’s t test, and categorical variables were compared with the χ2 or the Fisher’s exact test, as appropriate. Stata version 13 was used for the statistical
analyses.
Results
In total, 327 patients were eligible for inclusion. Of these, 112 patients had OTSC
placement, and 215 underwent conventional hemostasis (epinephrine with/without TTS
or thermic device or sclerosing agent). The OTSC group and the “conventional” group
differed with respect to the year of bleeding, cause of bleeding, site of bleeding,
Forrest classification, and Rockall score. The OTSC device was implemented in our
endoscopic service ([Fig. 1 ]) from 2012 with an increase in the percentage until 2017 (for 2018, there is only
partial data until March 2018). Utilization of hemostasis with OTSC was more frequent
in duodenal ulcers with Forrest Ia to IIa and in patients in the “conventional” group
with a higher Rockall score. Adverse events related to the procedure were only reported
in a case with thermal (APC) hemostasis that worsened the bleeding. In order to mitigate
the effects of measurable baseline confounders, patients were matched into 84 pairs
using propensity score matching. Covariates included in the model were age, sex, number
of comorbidities, cause of bleeding, site of bleeding, Forrest classification, and
Rockall score. The population flow chart is presented in [Fig. 2 ], and the baseline patient characteristics of the matched cohort are summarized in
[Table 1 ] and [Table 2 ]. The kernel distribution of propensity scores before and after matching is shown
in [Fig. 3a,b ].
Fig. 1 Number of procedures performed per study year. Only partial data available for 2018.
OTSC, over-the-scope clip.
Fig. 2 Flow chart for the enrolled patients. OTSC, over-the-scope clip.
Fig. 3 Kernel distribution of propensity scores: a before matching; b after matching.
Matched cohort
The majority of the patients in both groups were men (79 %) aged 70 ± 14 years (mean
± SD). At the time of bleeding, antithrombotics or anticoagulants were being used
by most of the patients (74 % of the conventional group and 71 % of the OTSC group).
The mean Rockall scores were 6.3 ± 1.5 and 6.8 ± 1.7 for the conventional and OTSC
groups, respectively.
The most frequent site of bleeding was the duodenum, with the bleeding mainly being
related to peptic ulcers with Forrest class Ia to IIa. Other hemorrhagic lesions with
minor significance compared with the main bleeding source were diagnosed in 37 % and
35 % of cases in the “conventional” and OTSC groups, respectively. Among traditional
hemostatic procedures, epinephrine + TTS was used in 93 % of cases. On the other hand,
OTSC was applied alone in 52 % of all of the cases and in 38 % of cases after the
injection of epinephrine. No adverse events were reported. The H.pylori status was
not assessed in > 50 % of patients in both groups. Rebleeding events were less common
in the OTSC group (20 % vs 8 %, 95 % confidence interval (CI) 12 – 30 vs 3 – 16; P = 0.02); however, the mortality rate of the two groups was not significantly different
(6 % vs 2 %; 95 %CI 2 – 13 vs 1 – 8; P = 0.4). Rescue hemostasis was mainly managed with another endoscopic procedure (42 %
vs 70 %) and less frequently with arterial embolization (29 % vs 18 %) or surgery
(29 % vs 12 %) in the OTSC and “conventional” groups, respectively. Rescue hemostasis
achieved with a second endoscopic procedure was managed with epinehprine + TTS, except
for three patients, managed with other modalities (epinehprine + sclerosing agent)
. The length of hospital stay (days) was longer in the OTSC group than in the “conventional”
group (15 ± 3 vs 11 ± 10 days, 95 %CI 14 – 16 vs 9 – 13; P value 0.03).
Discussion
Despite the major advances in NVUGIB management over the past decade, including the
prevention of peptic ulcer bleeding and high-dose proton pump inhibition, considerable
morbidity, mortality, and health economic burdens persist. Of particular note are
the rebleeding rates, one of the most crucial predictive factors of morbidity and
mortality that has not significantly improved as evident from longitudinal data in
the past 15 years [12 ]
[13 ]
[14 ]. Although several types of endoscopic treatment for NVUGIB have been described,
including injection therapy, thermal coagulation, hemostatic clips, fibrin sealant
(or glue), argon plasma coagulation, and combination therapy (typically injection
of epinephrine combined with another treatment modality), relatively few comparative
trials have been performed. Currently, most patients are being treated with either
thermal coagulation therapy or hemostatic clips, with or without the addition of injection
therapy.
In this study, we aimed to compare first-line endoscopic hemostasis, achieved using
conventional techniques, with OTSC placement for NVUGIB in a matched cohort of patients.
The OTSC system as a preliminary experience has been successfully used in patients
with severe bleeding or deep wall lesions, or perforations of the gastrointestinal
tract [14 ].
To date, clinical data on OTSC treatment for upper gastrointestinal bleeding is limited
to case series and retrospective studies [5 ]
[7 ]
[12 ]
[13 ]
[14 ]
[15 ]
[16 ]
[17 ]
[18 ]
[19 ]
[20 ]
[21 ]
[22 ]
[23 ]. The technical success rate varies from 77.8 % to 100 % when OTSC is used as first-line
therapy, with a rebleeding rate of 7.4 – 13.6 %. However, patient populations differ
widely with respect to the bleeding source, bleeding severity, and previous therapy.
Most studies include a limited sample size and lack a control group. In our study,
we tried to overcome these limitations by using propensity score analysis to balance
the confounding factors between the two groups. Our study showed the efficacy of OTSC
as a first-line therapy in the management of NVUGIB. The rebleeding rate was acceptably
low (8 %) in the OTSC group and was lower than that in the “conventional” group (P = 0.02). This finding is consistent with the previously reported rebleeding rates
of 0 – 22 % [24 ] and with the observation that rebleeding following OTSC placement occurs in up to
35 % of patients receiving antithrombotic therapy [14 ]. In addition, data from the first randomized study found that OTSC application was
more effective than standard hemostasis (TTS or thermal plus epinephrine) techniques
as a rescue therapy in patients with recurrent bleeding peptic ulcer [6 ]. The bleeding-related mortality rate was 2 % in the OTSC group, lower than that
(8 %) in the control group and occurred in patients with important comorbidities.
In our experience, death occurred in patients in whom the OTSC placement was impossible
or who experienced rebleeding. Unfortunately, the fatal events occurred despite the
interventional radiology or surgical approaches that were performed. This would appear
consistent with mortality reported in other recent studies [7 ]
[19 ]
[23 ]
[25 ].
In this study, we found and confirmed that, because of their lower rebleeding rate,
OTSC devices are suitable for patients with duodenal ulcers with high Forrest classification
status (Forrest Ia to IIa) and a high Rockall score.
The OTSC system, being a very contractile, super-elastic nickel titanium alloy, provides
tissue apposition that is far superior to that of traditional clipping. Hemostasis
is achieved by a combination of the following two mechanisms: (1) sealing the blood
vessel; and (2) closing off an ulcer. However, the main mechanism appears to be that
of “tissue compression” that occurs by compressing the surrounding tissue around the
vessel. Although it is possible to close an ulcer by applying the OTSC directly on
a bleeding vessel, it is believed that the above-mentioned “tissue compression” mechanism
better explains the hemostatic mechanisms.
These characteristics enable us to overcome the limitations of TTS used in the compression
of limited amounts of tissue, especially in the presence of scarred and hardened tissue
or inflammatory mucosa with a hemostatic effect not sufficient for large-sized vessels.
In addition, standard clips often detach from these lesions and induce more bleeding
by lacerating the vessel. Nevertheless, we acknowledge the limitations of the OTSC
system and agree with Asokkumar et al. [26 ] who have identified the following three common patterns that result in OTSC failure:
(1) delayed closure of OTSC occurring in lesions with large caliber arteries and those
with a deep fibrotic base; (2) shallow placement of OTSC resulting from inadequate
suction or premature clip deployment; and (3) misplacement of OTSC because of poor
visualization, difficult anatomy, and unstable endoscope position.
This study has certain limitations. First, because of the retrospective design, selection
bias and confounding factors could affect the study validity. In order to limit the
selection bias, all of the consecutive patients who underwent a hemostasis procedure
for NVUGIB were considered for enrollment. However, owing to the retrospective nature
of the study, complete information for all of the eligible patients was not available.
Thus, 30 % of the eligible patients were excluded. In order to reduce the confounding
factors in the two cohorts, propensity score analysis was applied to the study design.
This strategy to reduce the confounding has an unavoidable limitation because it reduces
the sample size. In our study, the post-hoc power (84 patients in each group, with
a 0.05 alpha error and rebleeding rates of 20 % in the conventional group and 8 %
in the OTSC group) is 61.2 %, which is slightly underpowered in comparison to the
standard reference value (80 %). Second, propensity score matching may lead to increased
covariate imbalance called the propensity score paradox [27 ]. Despite progressive pruning of the matched sets, the application of a caliper width
of 0.1 should avoid pruning near the lowest region of the imbalance trend. However,
it is possible that the nested variables influence the outcome. For example, it was
not possible to assess the severity of the comorbidities, and this could affect the
outcome. Moreover, the difference between the length of recovery in the two groups
could be associated with the severity of comorbidities in the patients which was not
scored in this study. Furthermore, the temporal relationship with the rebleeding rate
was difficult to assess because, after the introduction of the OTSC device in our
center in 2012, we preferred OTSC over TTS because of its observed, although empiric,
major efficacy in hemostasis. Thus, most of the control patients were those treated
before 2012, and during the study period, the skill level of the operators could have
changed. Thus, we suggest using OTSC as the first-line treatment for lesions with
a high risk of rebleeding in patients with a high risk Rockall score. Finally, randomized
controlled trials and a formal cost-effectiveness analysis are needed to determine
the impact of the first-line use of OTSC in patients with high risk NVUGIB.
