Keywords
Endoscopy Upper GI Tract - Other focus (of reviewers) - GI surgery - Endoscopy Upper
GI Tract - Endoscopic resection (ESD, EMRc, ...) - Endoscopy Upper GI Tract - RFA
and ablative methods
Introduction
Roux-en-Y gastric bypass (RYGB) is one of the most commonly performed bariatric surgeries
to treat morbid obesity [1]. RYGB promotes around 56.7% to 67% excess weight loss (%EWL) within 2 years of the
procedure, significantly decreasing long-term obesity-related complications and mortality
[2]
[3]
[4]
[5]. However, weight recurrence (WR) is not uncommon after RYGB. The reported onset
of weight regain after RYGB varies between studies, but it positively correlates with
the number of years after surgery [6]. Two years after the procedure, approximately 17% of patients regain more than 15%
of their nadir weight [7]. In the longer term, one of every five patients will recover more than 40% of their
maximum weight loss [8]. With that, recidivism of comorbidities and worsening quality of life typically
ensue.
The etiology of WR after RYGB is multifactorial and includes psychological, behavioral,
hormonal, and anatomical components [9]
[10]. Anatomically, the gastrojejunostomy (GJ) size has been listed as an independent
risk factor [11]. In addition, the pouch length and volume inversely correlate with weight loss after
the surgery [12].
While surgical revision is invasive and carries a significant risk of adverse events
(AEs) [13]
[14], the transoral approach to the pouch and GJ is easy and safe. Consequently, several
centers have successfully employed endoscopic suturing to reduce the stoma size and/or
the pouch size to treat WR [15]
[16]
[17]
[18]. Currently, the most common technique involves mucosal ablation with argon plasma
coagulation (APC) around the gastrojejunal anastomosis (Supplemental Fig. 1) before
applying running stitches [19].
In addition to reducing the stoma size, our institution has recently adopted a novel
approach consisting of tubularization of the distal pouch with an O-shaped gastroplasty
suturing pattern. Preliminary data have demonstrated that the tubular TORe (tTORe)
carries better weight loss outcomes than the standard TORe [17]. This is the first study to describe the tTORe approach fully and to compare its
safety and efficacy to the standard procedure.
Patients and methods
This is a retrospective cohort study of a prospectively maintained database conducted
at Mayo Clinic, Rochester, Minnesota, United States. Patients with a previous RYGB
who underwent endoscopic revision by either TORe or tTORe between 2012 and 2020 were
included. Patients with gastro-gastric fistulas and those with incomplete weight loss
data at follow-up were excluded. Per institutional regulations, all patients have
consented to chart reviews and de-identified data collection for research purposes.
Procedures
All patients were under general anesthesia in a supine or left lateral decubitus position
for the endoscopic revision of the gastric pouch. First, the operator performed a
diagnostic esophagogastroduodenoscopy (EGD) to confirm eligibility. If no exclusion
criteria were found, the procedure started after administering prophylactic antibiotics
(which were given due to the potentially nonsterile nature of the full-thickness suturing).
Most patients received ablative procedures on the rim of the GJ before suturing at
the endoscopist’s discretion (APC). Then, the endoscopic suturing device (Overstitch,
Apollo Endosurgery, Austin, Texas, United States) was mounted onto a double channel
gastroscope (Olympus 2TH-180, Olympus America, Brooklyn Park, Minnesota, United States)
and used to perform the running stitches at the GJ. Different stitching patterns were
employed at the operator’s discretion based on a tailored evaluation of the pouch’s
anatomy. For the tTORe group, additional O-shaped stitches were placed on the gastric
pouch after finishing GJ revision. The decision of performing pouch revision was also
made at the discretion of the endoscopist based on pouch anatomy, primarily the pouch
diameter (large pouches, generally considered over 4 cm, were selected to undergo
tTORe).
Data collection and definitions
Researchers collected data from patient electronic medical records. Baseline characteristics
included age, gender, years after RYGB, time from RYGB to endoscopic revision, presurgical
weight, nadir weight, concomitant use of weight loss medications at the endoscopic
revision, preprocedural weight, and preprocedural stoma size and pouch diameter.
%EWL was defined as the percent of excess body weight compared to the ideal weight
for a body mass index of 25 kg/m2. Percent total body weight loss (%TBWL) was defined as the percent body weight loss
relative to the patient’s weight before each procedure. Procedure characteristics
included procedure time, number of sutures used, and final stoma diameter. Patients
were followed over 1 year, and patient weight after the procedure was documented at
3, 6, 9, and 12 months. Information on the need for an additional endoscopic evaluation
and AEs also was collected.
The primary outcome was the %TBWL after the endoscopic revision. Secondary outcomes
entailed AEs, serious AEs (SAEs), and procedure time.
Statistical analysis
Continuous variables were reported as means and standard deviations, and categorical
ones as frequencies or percentages. Either chi-squared or Fisher’s exact test was
used to compare categorical variables and the Student’s t-test to compare continuous ones. The analysis of variance for repeated measures (ANOVA
test) was used to analyze and compare weight trends between cohorts. Within each cohort,
we ran a separate analysis for patients on and off weight loss medications.
Results
Baseline characteristics
One hundred twenty-eight patients (85 tTORe, 43 TORe) fulfilled eligibility and were
included in the study. The whole cohort had a mean age of 46.2± 12.0 with a predominance
of female sex (86.7%). The average time from surgery to endoscopic revision was 11.4±
6.6 years. At the nadir weight, patients presented a mean %EWL of 83.9± 25% and a
mean %TBWL of 38.3± 11.5%. At the time of the endoscopic revision, the mean stoma
size and pouch were 28.9±7.0 mm and 5.7±3.4 cm, respectively. All baseline characteristics
were similar between both groups ([Table 1]). The tTORe procedure required more sutures than the standard TORe (P = 0.03).
Table 1 Baseline characteristics of the 128 patients undergoing endoscopic revision for WR
after RYGB.
|
Variable
|
Overall
n=128
(mean±SD)
|
Tubular TORe
n=85
(mean±SD)
|
TORe
n=43
(mean±SD)
|
P value
|
|
WR, weight recurrence; RYGB, Roux-en-Y gastric bypass; TORe, transoral reduction;
tTORe tubular transoral reduction; EWL, excess weight loss; TBWL, total body weight
loss; IQR, interquartile range; SD, standard deviation; APC, argon plasma coagulation.
|
|
Age, years
|
46.2±12.0
|
46.4±11.6
|
45.7±12.9
|
0.77
|
|
Sex, % female
|
86.7%
|
83.5%
|
93.0%
|
0.55
|
|
Nadir %EWL
|
84.0±25.0
|
82.4±25.1
|
87.0±24.7
|
0.33
|
|
Nadir %TBWL
|
38.3±11.5
|
37.7±12.0
|
39.5±10.7
|
0.41
|
|
Years from RYGB
|
11.4±6.6
|
11.4±6.7
|
11.3±6.5
|
0.92
|
|
Pre-revision stoma size, mm
|
28.9±7.0
|
28.8±7.3
|
29.2±6.4
|
0.77
|
|
Pre-revision pouch length, cm
|
5.7±3.4
|
5.9±2.1
|
5.2±5.2
|
0.30
|
|
Stitching pattern, %
|
|
|
8.6%
|
5 (5.9%)
|
6 (14.0%)
|
0.29
|
|
|
28.9%
|
29 (34.1%)
|
8 (18.6%)
|
|
|
22.7%
|
20 (23.5%)
|
9 (20.9%)
|
|
|
15.6%
|
7 (8.2%)
|
13 (30.2%)
|
|
|
15.6%
|
18 (21.2%)
|
2 (4.7%)
|
|
|
11%
|
6 (7.1%)
|
5 (11.6%)
|
|
APC performed, %
|
75.8%
|
69.4%
|
88.4%
|
0.02
|
|
Final stoma diameter, mm
|
8.2±1.1
|
8.1±1.2
|
8.2±0.9
|
0.51
|
Weight loss
At 3 and 6 months, the tTORe and TORe cohorts presented similar %TBWL (3 months: 8.5±4.9
vs. 7.3±6.0, P = 0.27 and 6 months: 8.1±7.4 vs. 6.8±5.6, P = 0.44). At 9 months, there was a trend toward greater weight loss in the tTORe cohort
(9.7± 8.6% vs. 5.1± 6.8%, P = 0.053). Finally, at 12 months, the %TBWL was significantly higher in the tubularization
group compared to the standard group (8.2±10.8 vs. 2.3±7.3%, P = 0.01). %TBWL increased from 6 to 9 months in the tTORe cohort but decreased in
the TORe group. Therefore, we found a significant difference in weight loss trend
over time between groups (P = 0.03). [Table 2] summarizes weight loss data for the whole cohort and both groups and [Fig. 1] depicts weight loss trends over the 12-month follow-up period. In a subanalysis
whereby non-APC patients were removed, at 3 and 6 months, the tTORe and TORe cohorts
presented similar %TBWL (3 months: 9.0±4.9 vs. 7.8±6.3, P = 0.32 and 6 months: 9.0±7.4 vs. 7.1±5.7, P = 0.29). However, at 9 and 12 months, the %TBWL was significantly higher in the tubularization
group compared to the standard group (9 months: 10.9±8.6 vs. 5.1±7.4%, P = 0.03, 12 months: 9.0±10.8 vs. 2.5±7.2). Supplemental Table 1 summarizes weight
loss data for both groups after removing non-APC patients, and Supplemental Fig. 2
depicts weight loss trends over the 12-month follow-up period in the APC-only cohort.
Fig. 1 Comparison of % TBWL trend between tTORe and TORe over a 1-year period. At 12 months,
P = 0.01. P value was non-significant at 3, 6, and 9 months.
Table 2 Weight loss data for tTORe and TORe groups.
|
%TBWL
|
Overall
(mean±SD)
|
Tubular TORe
(mean±SD)
|
Sample
(n=85)
|
TORe
(mean±SD)
|
Sample
(n=43)
|
P value
|
|
3 months
|
8.0±5.3
|
8.5±4.9
|
73
|
7.3±6.0
|
39
|
0.278
|
|
6 months
|
7.6±6.8
|
8.1±7.4
|
45
|
6.8±5.6
|
24
|
0.445
|
|
9 months
|
7.9±8.1
|
9.7±8.6
|
29
|
5.1±6.8
|
19
|
0.053
|
|
12 months
|
5.7±10.0
|
8.2±10.8
|
40
|
2.3±7.3
|
23
|
0.013
|
Secondary outcomes
The overall mean procedure time, including the diagnostic EGD, was 58.1 minutes. There
was a statistically significant difference between the tTORe and TORe groups (60.5
vs. 53.4 minutes, P = 0.03). Weight loss of patients under concomitant weight loss medications also did
differ from that of those not receiving pharmacotherapy ([Table 3]). However, at 12 months, there was trend toward a more significant loss in tTORe
patients receiving medications at 12 months (13.8±12.0 vs. 6.6±10.1 for tTORe + medications
and tTORe alone, respectively, P = 0.08).
Table 3 Weight loss data comparing outcomes of concomitant pharmacotherapy within groups.
|
|
tTORe
(n=85)
|
Mean difference
(%TBWL)
|
P value
|
TORe
(n=43)
|
Mean difference
(%TBWL)
|
P value
|
|
Pharmacotherapy
|
Pharmacotherapy
|
|
Yes
|
No
|
Yes
|
No
|
|
tTORE, tubular transoral outlet reduction; TORe, transoral outlet reduction; %TBWL,
percent total body weight loss.
|
|
3 months
|
10.0±5.6
|
8.1±4.7
|
1.9
|
0.20
|
7.4±6.2
|
6.9±5.6
|
0.5
|
0.86
|
|
6 months
|
8.4±8.4
|
8.0±7.1
|
0.4
|
0.86
|
9.5±5.6
|
6.2±5.5
|
3.3
|
0.29
|
|
9 months
|
13.0±8.5
|
8.3±8.4
|
4.7
|
0.18
|
3.9±6.6
|
7.1±7.2
|
3.2
|
0.33
|
|
12 months
|
13.8±12.0
|
6.6±10.1
|
7.2
|
0.08
|
1.6±8.4
|
2.6± 7.2
|
1.0
|
0.80
|
There was a 7.8% overall AE rate (10/128), which was similar between groups (8.2%
vs. 7.0% for tTORe and TORe, respectively, P = 0.61). The AEs included refractory nausea and vomiting requiring endoscopic balloon
dilation in three patients and hematemesis requiring endoscopic clipping in one patient.
Ten patients presented to the Emergency Department (ED) within 30 days of the procedure
with nausea/vomiting (n=5) followed by abdominal pain (n=3) as the main reasons. Three
of these 10 patients were managed conservatively and discharged on the same day, while
seven required hospitalization (7/128, 5.4%). There were no related deaths.
Discussion
In our retrospective single-center study of 128 patients, we demonstrated that tubularization
of the distal pouch enhanced the durability of weight loss compared to traditional
TORe. While there was no difference within the first 9 months of follow-up, subjects
undergoing tTORe experienced more significant weight loss than those receiving standard
TORe at 1 year. Remarkably, weight loss trends are different between groups due to
a dichotomization starting at the 6- to 9-month timeframe. That seems to be the most
relevant effect of adding pouch tubularization to the TORe procedure in our cohort.
In 2016, Kumar et al. reported TORe outcomes in 150 patients with post-RYGB weight
regain. Patients presenting with pouch dilation underwent concomitant pouch reduction
with endoscopic suturing. However, there is no description of the pouch reduction
technique, no sample distinction between TORe and tTORe samples, and no definition
for pouch dilation. Therefore, their data lack clinical applicability and highlight
the importance of the present study [20].
Interestingly, weight results in our TORe cohort seem worse than previously published
data. For example, in 2018, Jirapinyo et al. reported a 9.6% TBWL at 6 months and
8.4% at 12 months following a purse-string TORe associated with APC or ESD [21]. In another retrospective study investigating TORe with adjunctive therapies, patients
had a similar %TBWL (8.5%) at 1 year [22]. In contrast, the %TBWL of our TORe cohort was 6.8% and 2.3% at 6 and 12 months,
respectively. In our center, highly specialized healthcare professionals recorded
all documented weight values during clinic visits, thus the values were not self-reported.
Our weighing protocol avoids self-enhancement bias and may have accounted for part
of the difference between our data and data in the literature [23].
As secondary outcomes, we compared the mean procedure time, the impact of concomitant
pharmacotherapy, and AEs. Because tTORe inherently involves adding a pouch tubularization
to the standard TORe, increased procedure time is inevitable. However, only 7 minutes
were added (60.5 vs. 53.4 minutes), which is probably cost-effective and logistically
feasible.
Remarkably, adjunct pharmacotherapy did not positively correlate with weight loss.
While this may initially seem illogical, it is consistent with a previous study that
demonstrated a lower %TBWL in patients receiving adjunct pharmacotherapy [22]. Because this was a retrospective study with no standardized guidelines for introducing
weight loss medications, selection bias may have played a central role. Those patients
losing less weight were probably more frequently receiving weight loss medications.
Possibly, a type 2 error may have also influenced our results. The difference in %TBWL
between tTORe patients with and without adjunct pharmacotherapy is striking: 13.8±12.0
versus 6.6±10.1 for a mean difference of 7.2. Therefore, we probably failed to detect
a real difference due to an underpowered sample.
Our overall AE rate (7.8%) compared to the literature is within the expected ranges.
Dhindsa et al. recently published a systematic review showing a pooled AE rate of
11.4 % for TORe, with abdominal pain being the most common (4.2%). Nonetheless, the
SAE rate was as low as 0.57 %, contrary to our 5.4% rate led by seven cases requiring
hospitalization. Our center is a high-volume center with a low threshold for hospital
admission for post-procedure ED visits because most of our patients come from other
states or countries with no nearby household. In non-academic non-referral centers,
some individuals would have probably been treated as outpatients, which could have
led to a lower SAE rate.
Our study has some limitations. First, the study was retrospective and restricted
to a single center of excellence, which may limit its generalizability. However, this
resulted in an advantage as the weight reporting was consistent and performed at our
institution by highly specialized healthcare professionals. Second, there was a heterogenous
distribution of adjunct APC to perform tTORe and TORe, with more APC cases in the
later cohort. Sound evidence shows that APC enhances weight loss during stoma revision
[19]. Ultimately, that fact highlights the tTORe results, as one would expect an even
more significant difference in weight loss had the cohorts equally received APC. Finally,
patients with larger pouches were the ones undergoing concomitant pouch reduction.
In this sense, it remains unknown how the pooled weight loss would behave if tTORe
were a routine rather than a tailored procedure.
Conclusions
In conclusion, distal tubularization of the pouch seems an attractive adjunct to the
standard TORe procedure. It may improve the durability of weight loss outcomes without
adding significant procedure-related risks. However, controlled studies are warranted
to confirm our findings.
