Introduction
Device-assisted enteroscopy (DAE) allows for direct endoscopic access to the small
bowel with the option for tissue acquisition and therapeutic procedures [1 ]
[2 ]
[3 ]
[4 ]
[5 ]. However, deep enteroscopy is a challenging and time-consuming procedure and, in
particular, visualization of the entire small intestine is usually only achieved by
experts in enteroscopy using DAE techniques [6 ]
[7 ]
[8 ]
[9 ]
[10 ].
Motorized spiral enteroscopy (MSE), using the novel PSF-1 PowerSpiral Enteroscope
(Olympus Medical Systems Corporation, Tokyo, Japan), was recently introduced into
clinical practice and represents a new technology, namely “self-propelling enteroscopy,”
that is a technical refinement of the principle of spiral enteroscopy [11 ]. An integrated electric motor is used to rotate a short spiral overtube at the distal
part of the insertion section of the enteroscope. MSE was recently shown to be effective
and safe for antegrade deep enteroscopy in terms of diagnostic success rates, procedural
duration, and depth of maximum insertion (DMI) in an initial prospective pilot study
[12 ]. In an additional prospective study involving the same two European centers, the
novel technique achieved total enteroscopy using an antegrade or combined antegrade
and retrograde approach in 70 % of the cases [13 ]. However, these results were achieved at two tertiary referral centers that already
had vast experience in MSE after an initial learning period. Furthermore, patients
who had undergone major abdominal surgery and with surgically altered anatomy were
not included in these trials; these patients may be at higher risk of procedure-related
adverse events [3 ]
[14 ].
For these reasons, the current large multicenter study was designed with the primary
objective of assessing the safety of the technique in a real-life setting, including
appraisal of learning curve aspects and with investigations performed in patients
after previous abdominal surgery, including with altered gastrointestinal (GI) anatomy.
Methods
Study design
This international multicenter prospective observational study (SAfety and performance
of the MotorIzed Spiral Endoscope; SAMISEN) was conducted at ten European endoscopy
tertiary referral centers. Data were collected from September 2019 until February
2021. The study protocol was approved by the institutional review board at each center
prior to initiation of the study.
Study objectives
The aim of the study was to evaluate the safety, efficacy, and learning curve associated
with MSE in a large cohort of patients with an indication for deep enteroscopy in
a real-life setting. Antegrade, retrograde, and bidirectional procedures were included.
To reflect the different levels of experience with MSE, the ten participating European
reference centers/investigators were either already experienced in MSE, or experienced
in DAE enteroscopy but newly starting to use MSE. Once enrollment into the study had
reached the halfway point, the inclusion of patients who had undergone major abdominal
surgery, including those with altered GI anatomy, was allowed.
Inclusion and exclusion criteria
Patients with suspected small-bowel disease with either a positive or suggestive finding
on prior small-bowel imaging (video capsule endoscopy [VCE], radiology) or another
clinical indication for deep enteroscopy were enrolled, after informed consent had
been obtained. The indications for deep enteroscopy and the study exclusion criteria
are presented in [Table 1 ].
Table 1
The indications for deep enteroscopy using motorized spiral enteroscopy (MSE), and
the study exclusion criteria and secondary end points.
Indications for deep enteroscopy[1 ]
Exclusion criteria
Any contraindication to standard enteroscopy (e. g. severe coagulopathy or known coagulation
disorder, bowel obstruction/stenosis, stents or other instruments implanted in the
intestinal tract, suspected gastrointestinal perforation, esophageal or gastric varices,
eosinophilic esophagitis) as judged by the investigator after careful individual risk
assessment
Secondary end points
Diagnostic yield: percentage of procedures that either confirmed a diagnosis from
previous studies, or established a new definitive diagnosis at the anatomical location
identified in previous studies, or findings that could explain the clinical symptoms
User feedback and assessment of handling characteristics: number and rate of procedures,
subjectively assessed by the endoscopist who performed the procedure, as worse, similar,
or better than balloon-assisted enteroscopy in the following categories: (a) handling;
(b) insertion; (c) positioning; (d) procedural time needed; and (e) staff and resources
needed
1 Indications were not limited to this list.
Recruitment of patients
All consecutive patients with an indication for deep enteroscopy were registered at
10 European reference centers and screened for enrolment. Patients who did not meet
the inclusion criteria or refused to sign the informed consent form were excluded
from the study.
Data management and statistical analysis
All consecutive patients at the study centers fulfilling the inclusion criteria were
registered and enrolled after informed consent had been obtained. A PostgreSQLdatabase
was created using an XClinical platform (Munich, Germany) with an electronic case
report form (eCRF). Data entry was done by trained study nurses at each study center
and was verified by a physician.
Statistical analyses were carried out by a professional statistician (SCO:SSiS, Berlin,
Germany) using SAS, version 9.4 (SAS Institute Inc., Cary, North Carolina, USA). Continuous
measures are summarized with sample size, mean, median, SD, minimum and maximum, and
interquartile range (IQR). Categorical measures are presented with the counts and
percentages of subjects in each category, with 95 % Clopper–Pearson confidence intervals
where reasonable. Fisher’s exact test was used to compare categorical variables. P values less than 0.05 were considered statistically significant. All authors had
access to the study data, and reviewed and approved the final manuscript.
Study device
The novel motorized spiral enteroscope PSF-1 was approved in Europe with a CE mark
during the entire study period ([Fig. 1 ]). The MSE system and procedural steps have been described in detail in previous
publications [11 ]
[12 ]
[13 ]
[15 ] (Appendix 1 s, see online-only Supplementary material).
Fig. 1 Images of motorized spiral enteroscopy being performed using the PSF-1 PowerSpiral
enteroscope (Olympus Medical Systems Corporation, Tokyo, Japan), which is 1680 mm
in length, with an outer diameter of 11.3 mm at the insertion portion, and has an
integrated electric motor that is used to rotate a short disposable spiral overtube
(240 mm in length, 31.1 mm outer diameter of the soft spiral fins) that is attached
to a rotation coupler located 40 cm proximal to the endoscope’s tip. For antegrade
MSE, the study device is inserted through the mouth and advanced with the assistance
of motorized clockwise spiral rotation. Marking of the deepest point (depth of maximum
insertion) was done using ink dye injection and/or clipping. Therapeutic interventions
were usually performed during the withdrawal phase.
Study investigators and endoscopist requirements
All procedures were performed by one or two accredited endoscopists at each study
site. Each study endoscopist had vast experience in deep enteroscopy using a standard
device-assisted technique (double-balloon enteroscopy, single-balloon enteroscopy,
and/or manual spiral enteroscopy) and had successfully passed a dedicated theoretical
and practical hands-on training module on MSE prior to accreditation (Appendix 2 s).
The first five cases for each individual endoscopist were considered to be learning
curve cases and were allocated to the training phase of the study protocol. This was
not applicable for study endoscopists who had performed more than 20 documented cases
of MSE outside the study protocol prior to initiation. For these endoscopists, all
MSE procedures performed within the study were allocated to the core phase of the
study protocol.
Motorized spiral enteroscopy and periprocedural management
The MSE procedure was performed as an antegrade, retrograde, or combined bidirectional
procedure. The selection was made on the basis of the pre-investigational results
of VCE or other imaging methods. For patients in whom an indication for total enteroscopy
was present, a second MSE could be performed from the opposite direction (in the same
session or on another day), if the first approach remained incomplete.
Post-procedural measures
In this observational study, clinical investigations and blood sample analyses were
performed according to the local policies at each center. The final study visit was
completed before each patient was discharged from the hospital.
Study end points, outcome measures, and definitions
The primary end point of the study was the number of serious adverse events (SAEs;
the number of patients with at least one SAE) caused by MSE during or after the procedure.
As a secondary safety end point, the overall frequency of adverse events was registered.
All adverse events were defined and classified using the most recent version of MedDRA
(Medical Dictionary for Regulatory Activities; www.meddra.org). All adverse events
were stratified by severity (mild, moderate, severe) [16 ] and by relation to the study treatment and/or the study device. All AEs were systematically
registered in the eCRF. Additionally, all SAEs were promptly reported via a paper
form (fax). The secondary end points are detailed in [Table 1 ].
Definition of analysis populations and subgroup analyses
The study enrolment was subdivided into two phases: training phase (as previously
defined) and core phase (all cases that were not training phase cases). To further
address the learning curve and reflect the increase in complexity, the core phase
was further subdivided according to the enrolment plan into: core phase 1 (CP1; first
half of the core phase population, which included only patients without previous major
abdominal surgery) and core phase 2 (CP2; second half of core phase population, which
also included patients who had undergone major abdominal surgery, including those
with altered GI anatomy).
Further information was prospectively registered and subsequent subgroup analyses
were performed, where reasonable, including subgroups of patients treated at experienced
MSE centers (previous experience of > 20 MSE cases) and new MSE centers (experience
of < 20 MSE cases, who therefore enrolled patients in the training phase population),
undergoing diagnostic and interventional procedures, who had undergone previous abdominal
surgery or had altered anatomy, with Crohn’s disease, who were taking aspirin during
the study (80–100 mg daily), and who underwent MSE with or without general anesthesia.
Data management, statistical analysis, and sample size calculation
The primary aim of this observational study was to evaluate the safety of MSE. The
SAE rate was used as a surrogate parameter. A technical review by the European Society
of Gastrointestinal Endoscopy (ESGE) reported SAE rates of up to 8 % or higher associated
with interventional DAE procedures, and up to 0.8 % for purely diagnostic procedures
[3 ]. In order to guarantee a reasonable number of subjects that needed to be enrolled
in the current study, a viable case number calculation was initially done to demonstrate
that the SAE rate was below an 8 % threshold as the upper limit. Therefore, a minimum
of 245 subjects for the core phase of the study was considered necessary for statistical
analysis. The precision of the SAE rate was estimated based on at least 245 patients
(width of the 95 %CI): n = 5, SAEs 2 % (0.7 %–4.7 %); n = 10, 4.1 % (2.0 %–7.4 %);
n = 15, 6.1 % (3.5 %–9.9 %); n = 20, 8.2 % (5.1 %–12.3 %). Taking into account an
expected dropout rate of 5 %, a total of 260 subjects was determined to be the minimum
total sample size required.
Shortly, after study initiation, a new joint guideline by the ESGE and United European
Gastroenterology (UEG) for the first time proposed upper limits for SAE rates of 1 %
and 5 % for diagnostic and therapeutic DAE procedures, respectively [14 ]. However, this recommendation was only based on moderate quality evidence and also
suggested that higher complication rates can be expected after previous abdominal
surgery and in patients with altered anatomy.
Results
Patient characteristics and procedural details
Between September 2019 and February 2021, 302 patients were enrolled in the study
[Fig. 2 ]. Four patients had to be excluded because no deep enteroscopy had been performed.
A total of 298 patients (120 women; median age 68 years, range 19–92) were eligible
for analysis, with 47 patients allocated to the training phase and 251 allocated to
the core study phase.
Fig. 2 Patient enrollment and allocation to study groups. Four patients had to be excluded
from the analysis, because no deep enteroscopy was performed: training phase (n = 1),
core phase 1 (n = 3).
Overall, 80.9 % of the patients had positive findings on previous VCE (n = 151; 50.7 %)
or in other imaging modalities (n = 90; 30.2 %). There were 116 patients (38.9 %)
who were enrolled at two experienced MSE centers and 182 (61.1 %) who were enrolled
at new centers. Among the 298 patients, 337 MSE procedures were performed: (antegrade,
241; retrograde, 75; combined [single session], 21). Among the core phase patients,
54 (21.5 %) had had previous abdominal surgery, resulting in surgically altered GI
anatomy in 25 patients (10 %). Also in the core phase group, one-third of the patients
were initially planned for total enteroscopy (81/251; 32.3 %) ([Table 2 ]).
Table 2
Characteristics of the 298 patients entered into the study who underwent deep enteroscopy
using motorized spiral enteroscopy (MSE).
Patients in training phase/core phase, n
47/251
Sex, male/female, n
178/120
Age, mean/median (range), years
68/64.4 (19–92)
Body mass index mean/median (IQR), kg/m2
26.1/25.2 (22.8–29.0)
ASA classification, n (%)
33 (11.1 %)
124 (41.6 %)
129 (43.3 %)
12 (4.0 %)
Previous abdominal surgery, n (% of core phase group)
54 (21.5 %)
Surgically altered gastrointestinal anatomy n (% of core phase group)
25 (10.0 %)
1
3
8
3
2
1
6
1
Previous positive imaging as indication for MSE, n (%)
241 (80.9 %)
151 (50.7 %)
90 (30.2 %)
Patients planned for total enteroscopy, n (%)
98 (32.9 %)
81 (32.3 %)
IQR, interquartile range; ASA, American Society of Anesthesiologists.
Safety analysis
The population for safety analysis was comprised of all 298 patients (including the
47 training patients). Eight SAEs were reported in seven patients. Therefore, the
overall SAE rate per patient was 2.3 % (95 %CI 0.9 %–4.8 %). The upper limit of the
95 %CI was below the predefined threshold of 8 % and also below the 5 % threshold
suggested by the latest European guideline for therapeutic procedures [14 ]. In the core safety population (training patients excluded), the SAE rate was slightly
lower at 2.0 % (95 %CI 0.6 %–4.6 %). In the training phase group (47 patients), two
SAEs occurred, giving an SAE rate of 4.3 % (95 %CI 0.5 %–14.5 %). Details of the SAEs
are given in Table 1 s .
The overall AE rate was 11.1 % per patient (33/298) and 11.0 % per procedure (37/337).
Without the training phase cases, the overall AE rate was 9.6 % per patient (24/251).
Most of the reported mild AEs were related to clinically asymptomatic mucosal lacerations
at the level of the esophagus, the cardia, and the small bowel, and transient mild
abdominal pain.
Subgroup analyses for the primary endpoint
The SAE rates (per patient) were 0.8 % (1/126; 95 %CI 0.02 %–4.34 %) and 3.5 % (6/172;
95 %CI 1.29 %–7.44 %) for diagnostic procedures and when therapeutic interventions
were performed during MSE, respectively. The SAE rate was 1.6 % (4/257) when general
anesthesia was used and 3.8 % (3/80) when deep sedation was used (P = 0.24). All but one SAE occurred during antegrade MSE procedures.
SAE rates for further subgroups were as follows: after previous abdominal surgery,
1.9 % (1/53; the only event occurred in a patient with altered anatomy [4 %; 1/25]);
in known or newly diagnosed Crohn’s disease, 4.8 % (1/20); patients taking aspirin
during the study (80–100 mg daily), 0 % (0/86).
Procedural success, insertion depth, and procedure time
The anatomical region of interest could be reached in 295 of 337 procedures in the
entire study (87.5 %, 95 %CI 83.5 %–90.9 %) and in 250 of 284 procedures in the core
group (88.0 %, 95 %CI 83.7 %–91.6 %). Total enteroscopy was achieved in half of the
patients that were initially planned for a total enteroscopy (42/81). Procedural details
are shown in [Table 3 ].
Table 3
Details of the motorized spiral enteroscopy (MSE) procedures performed.
Overall group
Core group
Number of patients
298
251
MSE approach (per patient), n (%)
229 (76.8 %)
200 (79.7 %)
27 (9.1 %)
22 (8.8 %)
48 (16.1 %)
36 (14.3 %)
12 (4.0 %)
11 (4.4 %)
21 (7.0 %)
15 (6.0 %)
Total number of procedures performed
337
284
MSE route, n (% of procedures)
241 (71.5 %)
211 (74.3 %)
75 (22.3 %)
58 (20.4 %)
21 / (6.2 %)
15 (5.3 %)
Procedural success rate, n (% of procedures)[1 ]
295 (87.5 %)
250 (88.0 %)
Total enteroscopy rate, n (% of patients)
46 (15.4 %)
42 (16.7 %)
46 (46.9 %)
42 (51.9 %)
Procedure time (median/IQR), minutes
39.0 (27–54)
38.0 (25–54)
59.5 (45–79)
58.5 (45–79)
32.0 (20–50)
29.5 (18–40)
48.0 (33–69)
44.5 (28–65)
Type of anesthesia used, n (%)
General anesthesia
Sedation
–
202 (83.8 %)
39 (16.2 %)
–
35 (46.7 %)
40 (53.3 %)
–
20 (95.2 %)
1 (4.8 %)
Diagnostic yield per patient, n (%)
251 (84.2 %)
208 (82.9 %)
Therapeutic yield per patient, n (%)
172 (57.7 %)
151 (60.2 %)
IQR, interquartile range; DMI, depth of maximum insertion.
1 Anatomical region of interest reached.
Diagnostic yield
In 251 of 298 patients, the procedures with MSE either confirmed a diagnosis from
previous studies, or established a new definitive diagnosis at the anatomical location
identified in previous studies or findings that could explain the clinical symptoms.
Therefore, the diagnostic yield was 84 % ([Table 3 ]).
Therapeutic yield
In 172 of 298 patients, at least one therapeutic intervention (other than biopsies)
was performed. The therapeutic yield per patient was 57.7 %. The time needed for interventions
was a mean of 7.8 minutes and a median of 3.0 minutes (IQR 1.0–10.0) ([Table 3 ]).
Learning curve analysis
Only one SAE occurred at an experienced center (1/116; 0.9 %), whereas the overall
SAE rate was 3.3 % at new MSE centers (6/182; P = 0.25). The SAE rate, which was 4.3 % (2/47) in the training phase, decreased to
2.4 % (3/124) in core phase 1 and 1.6 % (2/127) in core phase 2 ([Fig. 3a ]). As expected, the overall number of SAEs in the entire study population was too
low for further subgroup analyses in terms of a learning curve effect.
Fig. 3 Analysis of the learning curve with respect to: a study phase; b center experience.
In all study phases, the rate of procedures that reached the anatomical region of
interest (procedural success rate) was high: training phase, 84.9 % (45/53); CP1,
89.0 % (129/145); and CP2, 87.1 % (121/139; P = 0.70). The procedural success rate was not significantly different between procedures
done at experienced (89.1 %; 123/138) and new MSE centers (86.4 %; 172/199; P = 0.51). The diagnostic success rate (per patient) was constantly high throughout
all study phases: training phase, 91 % (43/47); CP1, 77 % (96/124); CP2, 88 % (112/127;
P = 0.03). The overall diagnostic yield was 76 % (88/116) and 90 % (163/182) at experienced
and new MSE study centers (P = 0.002), respectively. However, the rate of positive imaging tests prior to MSE
was lower at the experienced centers (55 % vs. 76 %).
Total enteroscopy was achieved in 19 % (22/116) and 13 % (24/182) of experienced and
new MSE centers, respectively (P = 0.19). The rate of therapeutic MSE procedures (therapeutic yield) increased slightly
throughout the study phases: training phase, 44.7 % (21/47); CP1, 59.7 % (74/124);
CP2, 60.6 % (77/127; P = 0.14). Procedures performed at experienced centers (69.8 %; 81/116) had a higher
overall therapeutic yield than new centers (50.0 %; 91/182; P < 0.001) ([Fig. 3b ]).
Discussion
MSE was recently introduced into clinical practice for deep enteroscopy in Europe
and parts of Asia. The novel technology using a motorized, self-propelling endoscope
has demonstrated favorable outcomes for deep enteroscopy in terms of insertion depth,
procedural duration, and efficacy of diagnostic and therapeutic interventions in patients
without previous abdominal surgery at expert centers [12 ]
[13 ]. In the current study, MSE was applied to a potentially more vulnerable population
of patients who had undergone major abdominal surgery, including those with altered
GI anatomy. In addition, when a new technology becomes available, it often involves
a learning curve with a potentially higher risk of associated AEs early on. Therefore,
safety analysis was chosen as the primary end point of this study. Evaluation was
done in a real-life setting with an internal control group, as not only expert centers
for MSE but also other centers, which contributed their learning curve experience,
were included.
True complication rates for deep enteroscopy are difficult to estimate because of
the limited number of available studies that were primarily designed to evaluate AE
rates. Therefore, reported complication rates in the literature mainly derive from
the secondary end points of studies with different primary objectives and consecutive
meta-analyses.
The latest European guideline on performance measures for small-bowel endoscopy, for
the first time, suggested thresholds of 1 % and 5 % for diagnostic and therapeutic
DAE procedures, respectively, in unselected populations [14 ]. In this context, the overall SAE rate in our study of 2.0 % shows that MSE can
be safely performed in this real-life and prospectively scrutinized scenario. Even
when training phase patients were included in the analysis, the overall SAE rate was
only 2.3 %. The distinct SAE rates for diagnostic and therapeutic procedures in our
study were 0.8 % and 3.5 %, respectively. While these SAE rates are below the thresholds
proposed by the ESGE guideline, the study was not powered for these subgroup analyses
and therefore results have to be interpreted with caution and cannot be generalized.
However, the current study clearly confirms the findings from a previous large prospective
pilot trial, which reported an SAE rate of 1.5 % [12 ].
Data regarding the use of spiral enteroscopy in patients after major abdominal surgery
and with surgically altered anatomy are limited. The available studies therefore do
not report an increased rate of AEs [17 ]
[18 ]
[19 ]
[20 ]; however, there is potential concern about an increased rate of AEs using the motorized
technique. Recently, a study found no increase in the AE rate for MSE in patients
who had undergone previous surgery [21 ]. In the current study, 21.5 % of the patients had previous abdominal surgery, with
10 % having surgically altered GI anatomy (40.8 % and 19.2 % of the CP2 group, respectively).
The SAE rate in this subgroup of patients was only 1.9 %. Remarkably, only one SAE
occurred in a patient with altered anatomy.
The most common complications of standard DAE are perforation, bleeding, and pancreatitis.
Looking into the details of the current study, only two bleeding-associated SAEs that
required endoscopic intervention occurred and a single perforation that required laparoscopic
suturing was observed, meaning the rates for both categories were within the anticipated
range. Pancreatitis has been reported to occur in 0.3 % of DAE procedures [3 ]. Remarkably, in the current study, no acute pancreatitis following MSE was registered,
indicating that the lower risk for post-DAE pancreatitis reported with spiral enteroscopy
is maintained for MSE [6 ].
Rotation of the spiral overtube, that is needed for movement of the endoscope, depends
on the functionality of the integrated electric motor and its peripherals. Thus, there
is concern among users about the durability of the MSE system and consequences of
a major equipment failure, like a motor breakdown, when the spiral is fixed deep within
the small bowel. Therefore, before starting a new procedure (after start-up of the
system), the operator is obliged to follow an inspection protocol to ensure normal
functionality of the system. However, the manufacturer provides a specific emergency
protocol (use of CO2 insufflation, fluid irrigation, and repetitive tip deflection under fluoroscopic
guidance to free the entrapped small bowel from the endoscope without spiral movement),
that has been successfully validated in animal models prior to first-in-human application.
Furthermore, in MSE, compared with manual spiral enteroscopy, the tactile feedback
of spiral rotation is replaced by a graphical rotation force indicator that displays
the direction and the resistance of the spiral rotation to the operator. The system
continuously monitors the current that is needed for spiral rotation as a surrogate
for the applied force. Automatic motor stops occur when a certain threshold is exceeded.
The system functionality check also includes a mandatory test of this “limit function”.
Rotation of the spiral overtube, which is needed for movement of the endoscope, depends
on the functionality of the integrated electric motor and its peripherals. Therefore,
there is concern among users about the durability of the MSE system and the consequences
of a major equipment failure (Appendix 3 s ). No failures were observed during any of the 337 MSE procedures. Premature disassembly
of the spiral overtube from the rotation coupler would result in the inability to
further apply spiral rotation and might lead to a total loss of the overtube within
the patient’s small bowel. In the current study, this situation occurred once, when
the overtube was already in the patient’s esophagus during the withdrawal phase and
it could be manually extracted. This situation may have occurred because a standard
mouthpiece was used instead of the approved mouthpiece with a larger diameter. Furthermore,
it has become evident that strict adherence to the manufacturer’s procedure guideline
is needed.
In the previous prospective trials using MSE, diagnostic success rates were reported
to be as high as 74.2 % for antegrade MSE [12 ] and 80 % when applying an antegrade and/or retrograde approach [13 ]. In the current study, where 87.5 % of the procedures successfully reached the region
of interest, the overall diagnostic yield was 84 %. This clearly indicates that the
high diagnostic success rates of MSE can also be reproduced in the real-life setting
of this multicenter study.
Looking at the learning curve, we found a higher SAE rate of 4.3 % in the training
phase, compared with 2.0 % in the entire core phase and 1.6 % in core phase 2 only,
when postsurgical patients, including those with altered anatomy, were also enrolled.
Although these results must be interpreted with caution because of the limited number
of patients and only two events in the training phase population, this indicates a
trend towards a (short-term) learning curve effect. Only one SAE occurred during a
procedure that was done at a center with previous MSE experience.
Remarkably, procedural and diagnostic success rates remained constantly high throughout
all study phases, including the training phase, and were not inferior at new MSE centers
compared with those with more experience. There was a trend for higher rates of total
enteroscopies at experienced centers (19.0 % vs. 13.2 %). Furthermore, although the
rate of therapeutic interventions slightly increased throughout the study phases (from
44.7 % to 60.6 %), no increase in the complication rate was noted.
Although our study represents the first large-scale international multicenter prospective
evaluation of the novel technique of MSE, it also has limitations. Firstly, the study
was powered for the overall rate of SAEs. Important subgroup analyses (i. e. training
phase, postsurgical and altered anatomy patients, or patients with specific small-bowel
diseases) are limited by small numbers of subjects in the respective groups. Secondly,
no control group was included in the study. In addition, the heterogeneous composition
of the study centers, with different levels of experience in terms of MSE, may be
seen as a limitation. However, it may also represent a strength of our study, as it
serves as an internal control. The involvement of centers with two differing levels
of MSE experience and different study phases related to the complexity of patients
reflects real-life clinical practice and addresses the learning curve effect.
In conclusion, this prospective multicenter study showed that MSE was feasible and
safe in a large cohort of patients in a real-life setting at centers with experience
in deep enteroscopy after a short learning curve. MSE was shown to be feasible in
postsurgical patients, including those with altered anatomy, without an increase in
the rate of AEs. These results justify further evaluation of MSE in further prospective
studies for various indications, including biliopancreatic interventions in postsurgical/altered
anatomy patients, preferably with the inclusion of a control group.
