Introduction
The ability to safely and easily perform a septotomy in the gastrointestinal (GI)
tract could dramatically improve the treatment of several pathological conditions,
including diverticula and post-surgical complications. Symptomatic epiphrenic esophageal
diverticulum (EED) is an example of such a condition, where the current standard surgical
treatment is associated with high rates of morbidity (0 %–75 %) and mortality (0 %–11.1 %)
[1] and the latest therapeutic approach (diverticular peroral endoscopic myotomy [D-POEM])
is technically demanding and preferred only for motility disorder-related EED [2].
In 2015, we proposed an alternative method to treat EED, consisting of the use of
a pair of magnets to induce a compression anastomosis between the diverticulum and
the esophagus. The compression induces ischemia, leading to necrosis, inflammation,
and ultimately to a healing process, allowing the creation of an endoluminal anastomosis.
During revisional endoscopy, the residual septum is cut using an electrical knife
[3]. The feasibility of this approach was confirmed by another team [4]. Although effective, this technique requires two interventions, the second one being
technically challenging.
Based on this previous experience, we have developed a new device that provides marsupialization
of an EED, called the MAGUS (MAgnetic Gastrointestinal Universal Septotome). It consists
of two magnets that are placed on each side of the septum of the diverticulum, which
are linked by a self-retractable suture wire ([Fig. 1]). By applying a constant pressure with both, the magnets induce a compression anastomosis
and the wire induces what we call “wire compression cutting” (WCC). After confirming
the method in proof-of-concept experiments in a pig model, we initiated a prospective
human study. We present here the results of this study in terms of conceptual design,
technical preclinical success, and early clinical outcomes of the first two patients
with an EED enrolled in the study.
Fig. 1 Diagram showing the system of magnets and wire that are used to cut the septum of
the diverticulum, achieving a marsupialization. Both the magnets and the wire cut
by applying a continuous pressure inducing pressure necrosis and wound healing with
fibrosis.
Methods
Magnet and wire septotome
The MAGUS device (BMDC, Brussels, Belgium) consists of three parts ([Fig. 2a]): a single magnet (18-mm wide and 4.2-mm thick); the self-retractable wire (a 0.4-mm
braided non-resorbable surgical wire); and a magnetic box (19-mm wide and 8.3-mm thick),
which contains two magnets that encompass the spring-activated wire winding system.
The device concept was patented in 2018 (EP2018/055106 – WO/2018/158395).
Fig. 2 Diagrams showing: a the spatial arrangement of the MAGUS magnetic device, which consists of a single
magnet (blue), a self-retractable wire (green), and magnetic box (casing shown in
black) that contains the winding system (orange) and two further magnets (blue); b the MAGUS delivery system used during the clinical trial allows distal attachment/proximal
release of the single magnet (1a/1b) and distal attachment/proximal release of the
magnetic box (2a/2b), with the self-retractable wire being activated when the magnetic
box is released.
Acceptable ranges of force and pressure applied by the magnets (i. e. the single magnet
and the magnetic box) and the wire have been calculated based on four magnetic devices
found in the literature. In this paper, the WCC and magnetic compression anastomosis
are considered as two similar physiological phenomena that are induced by pressure.
The four magnetic devices were the IMAS [5], the Magnamosis system [6], the Flourish device [7], and magnets created by Cook (Winston Salem, North Carolina, USA) and used in different
studies [3]
[8]. From pictures, the information present in these reports, and theoretical calculation
using the FEMM software (www.femm.info), the minimal force needed to avoid dislodgment
of the magnets is 2.2 N at 2 mm and the pressure needed to induce ischemia ranges
from 6.4 to 150 kPa at 2 mm.
The force (pressure) of attraction between the two magnets of the MAGUS ranges from
4 N (17 kPa) to 5 N (21 kPa) at 2 mm, while the mean pressure applied by the wire
is estimated to be between 50 and 70 kPa for a 40-mm high septum.
Once in place, the device is expected to achieve the entire cutting process within
28 days [3]
[5]
[6]
[7]
[8].
Delivery procedure
The MAGUS magnetic device is attached to a custom delivery system that keeps both
magnets apart and the self-retractable wire system inactivated during the delivery
procedure. The proximal and distal magnets can be released independently from one
another through a dedicated handle ([Fig. 2b]).
The main steps of the procedure are presented in [Fig. 3]. First, the delivery system is mounted on a pre-installed guidewire. Next, the endoscope
is inserted alongside and a forceps is used to mobilize the proximal magnet. After
delivery, it is pushed by the endoscope to the bottom of the diverticulum. The distal
magnet is then pulled to meet with the proximal magnet across the septum. Once the
meeting is confirmed, the distal magnet is released, which also activates the self-retractable
wire.
Fig. 3 The steps in the delivery of the MAGUS system are: a insertion of the device; b mobilization of the proximal magnet; c pulling of the catheter; d the meeting of the magnets; e release of the distal magnet and activation of the self-retractable wire.
Animal studies
The feasibility of WCC was assessed in animal experiments conducted on pig models:
one Yucatan (male, 1-year old, 51 kg) and three Pietrain x Landrace (females, 3–4-months
old, 36–42 kg).
An artificial septum was created surgically in the stomach of the four different animals,
resulting in a double stomach layer septum, with apposition of the two serosae. After
2 weeks, the magnets were implanted on this septum through a gastrotomy to assess
the feasibility of the WCC [9].
For pig #4, it was not possible to implant the magnets in a secure position that would
avoid migration. It was therefore decided to implant them in a bowel loop, through
an enterotomy. This created a septum between the afferent and efferent limbs, as illustrated
in [Fig. 4]. The magnets were implanted before activating the cutting wire manually.
Fig. 4 Diagrams and photographs from pig #4 showing: a implantation of the MAGUS device (external picture, with internal and cut-view schematics);
b the resulting cut (the yellow circle shows the cutting line); c macroscopic histological view; d microscopic histological view, with fusion of the layers and new mucosa growing on
top and no significant fibrosis present.
Because magnetic compression anastomosis has been proven to work in esophageal diverticula
[3], we did not consider the specific histology of the tissue when creating the artificial
septa.
The animals were fed from day + 1 after surgery. Follow-up fluoroscopy was performed
on day + 1 and day + 7. The animals were sacrificed at day 14.
Clinical trial
A safety and feasibility clinical trial was initiated at Erasme Hospital, Université
Libre de Bruxelles, Brussels in February 2020, after agreement from the ethical committee
of the hospital and the Belgian health authority, for patients presenting with a symptomatic
EED or anatomical conditions associated with a septum.
With the patient under general anesthesia, the MAGUS was implanted as described in
[Fig. 3] and shown in [Video 1]. A transparent cap was used to facilitate magnet manipulation during positioning.
A follow-up endoscopy was performed at 24 hours, as per protocol. Clinical and radiological
follow-up was scheduled at 14, 28, and 90 days. The primary endpoints included the
technical success of implantation across the septum, evacuation of the device within
28 days, and serious adverse events during the period of device intervention (up to
30 days). Symptomatic improvement (including Eckardt score [10]) was also monitored at 1 and 3 months.
Video 1 The steps in the MAGUS delivery procedure are as follows: after the diverticulum
has been identified, the proximal magnet is mobilized by the endoscope and placed
at the bottom of the diverticulum; the distal magnet is then pulled back via the delivery
catheter until both magnets meet at the bottom of the septum; after good positioning
of the magnets has been confirmed on a fluoroscopic image, the distal magnet is released
and the self-retractable wire is activated.
Results
Animal studies
Cutting of the created gastric septum was confirmed within 1 week for pig #1 (specimens
explanted early for analysis [9]) and occurred between days 7 and 28 in pig #2, after cutting of the septum was confirmed
at autopsy. For pig #3, dislodgement before cutting occurred within 24 hours. This
was likely due to the fact that the double stomach thickness of the pig limited the
force of the initial apposition of the magnets.
For pig #4, migration of the device was confirmed after 14 days and the intestinal
specimen was explanted. Macroscopically, the cutting of the artificially created septum
was complete. Histological analysis showed that, after only 2 weeks, mucosa was growing
on the cut site and that a full-layer healing scar was visible ([Fig. 4]).
Clinical trial
Two men of 73 and 56 years of age with a symptomatic EED were enrolled. From fluoroscopic
images, the diverticulum and septum measured 52.5 mm and 25 mm, respectively, in the
first patient, and 58 mm and 40 mm, respectively, in the second patient. Technical
implantation was successful in both patients. Device insertion took 12 and 15 minutes
for the first and second patients, respectively. No serious adverse events were reported.
For the first patient ([Fig. 5]), the MAGUS migrated spontaneously between days 14 and 28. The Eckardt score dropped
from 2 to 1 between baseline and 28 days after the first intervention.
Fig. 5 Images from the first patient showing: a,b the implanted device: a at the end of the intervention endoscopic; b on radiological view; c,d barium swallow from: c baseline; d 3 months after insertion of the MAGUS, with almost complete disappearance of the
septum.
For the second patient, the cutting wire seemed to be blocked between the magnets.
This happened because the distal magnet was still loosely attached to the catheter
and was able to turn on itself, blocking the wire between the two magnets. Because
the self-retraction of the wire was documented to be sufficient to induce its release,
based on our previous experience, we decided to leave the device as it was. The magnets
had not migrated after 28 days and endoscopy showed that the magnetic anastomosis
was complete, but the wire was still blocked between the magnets. The wire was activated
by pulling on it, and migration was confirmed after a further 12 days. The patient’s
Eckardt score dropped from 6 to 2 between baseline and 28 days after the first intervention.
Discussion
The results of this study show that the MAGUS device is a rapid, single-stage, and
potentially safe and effective procedure that can be used to perform endoscopic sectioning
of a septum in the GI tract. The achievement of this during a single procedure has
become feasible because of the addition of the self-retractable wire to the magnets.
This wire applies a predefined range of pressure on the residual septum besides the
position where the magnets are placed. The biological process following the localized
ischemia induced by pressure necrosis is an inflammatory reaction that allows for
progressive healing. This process is biological and not physical and does not occur
in an explanted organ (data not shown). We have called this phenomenon “wire compression
cutting”.
EED is a rare condition that we selected for our first test in humans. The treatment
of this entity has been evolving over the last few years. Although D-POEM seems to
be a promising procedure compared with surgery, especially for EED associated with
motility disorders like achalasia [2], the use of magnets has also been reported but requires a second step to cut the
residual septum left after migration of the magnets [3]
[4]. This preliminary experience with a device that includes two magnets and a self-retractable
wire shows that the endoscopist can successfully marsupialize an EED in an intended
single-step procedure.
The main advantage of this device is that it offers a less technically demanding procedure,
based on the ease of use for the endoscopist and the application of the WCC concept.
Nevertheless, the activation of the cutting wire required a second endoscopy in one
patient. This problem can be solved at the initial insertion, if there is any doubt,
by ensuring the immediate activation of the self-retractable wire by mobilizing it
with a biopsy forceps.
Although the concept of the MAGUS device was originally tested in the setting of EED
treatment, this limited human experience suggests that the safety profile of the procedure
can allow us to consider other applications that may benefit from marsupialization.
For example, Zenker diverticulum, candy cane syndrome (a postoperative complication
of the afferent loop, for example post-bypass or after total gastrectomy) [11]
[12], or anastomotic refractory strictures, in which a septum could be responsible for
symptoms, represent potential and/or current areas of investigation.
Further clinical trials and follow-up are needed to assess the safety and efficacy
of the treatment, as we have here presented only feasibility results. Clinical endpoints
for further trials could focus on different adverse events, such as measurement of
the risks of leakage, recurrence, or stenosis due to fibrotic scarring.
Further development of the device should focus on the improvement of the delivery
catheter and procedure. This could ensure that the wire is properly activated at the
end of the procedure and improve the learning curve of the procedure.
