Introduction
Bariatric procedures are vital to tackling the obesity pandemic and its accompanying
metabolic comorbidities. In this space, endoscopic bariatric therapy (EBT) has become
increasingly popular as a minimally invasive alternative to well-established surgical
bariatric procedures.
EBT includes a wide array of options, such as space-occupying devices (e. g., intragastric
balloon [IGB]) [1] and restrictive procedures (e. g., endoscopic sleeve gastroplasty [ESG] and primary
obesity surgery, endoluminal procedure) [2]. These procedures are in various stages of development, testing, and adoption, with
the IGB having the most long-term data and ESG gaining more recent traction with a
randomized clinical trial underway (MERIT trial, NCT03406975). The safety and efficacy
of EBT has been explored in multiple retrospective studies and landmark clinical trials
with findings thus far suggestive of significant weight loss outcomes with reasonable
safety profiles [3]
[4]
[5]
[6]
[7]. As the popularity of EBT and the data supporting its safety and efficacy grows,
it is imperative that clinicians have a data-driven perspective on the outcomes of
these procedure compared to laparoscopic bariatric therapy (LBT).
The elective nature of these procedures makes risk assessment of the utmost importance.
With the minimally invasive nature of EBT comes the assumption that the procedures
are inherently lower risk. Thus far, smaller studies have supported this hypothesis,
as reported above. While efficacy and proof of concept can be demonstrated in small
retrospective studies, real-world adverse events (AEs) are difficult to extrapolate
without large-volume data. It is imperative to perform larger comparative studies
to better understand the outcomes and to be able to guide patients in choosing from
the menu of options available for these elective bariatric procedures.
Endoscopic and surgical bariatric therapies have traditionally been difficult to compare
for many reasons. Patient populations can differ significantly, especially with regard
to comorbidities, given the eligibility criteria used for bariatric surgery [8]. In addition, most EBTs are only approved for use in patients with body mass index
(BMI) < 40 kg/m2, limiting overlap. Finally, data on EBTs are limited because of the novelty of the
procedures, and long-term data are particularly scarce. Overall, while informative,
the available data comparing EBT and LBT is suboptimal because it is retrospective,
performed at centers of excellence, and based on small cohorts.
The Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program
(MBSAQIP) is a unified national accreditation program for bariatric surgery centers
that is the result of a combined effort of the American College of Surgeons (ACS)
and the American Society for Metabolic and Bariatric Surgery (ASMBS). All accredited
participating centers report their outcomes to the MBSAQIP database. This study aimed
to use the MBSAQIP database to compare 30-day postoperative morbidity and mortality
outcomes of primary EBT vs LBT. According to the ASMBS, laparoscopic sleeve gastrectomy
(LSG) is currently the most commonly performed bariatric surgery [9]. For this reason, a subgroup analysis was also performed comparing EBT to LSG.
Patients and methods
The MBSAQIP 2019 database contains data from 206,570 cases from 868 centers and our
analysis was performed using this dataset. Patients between the ages of 18 and 80
years were included. Patients were included only if their BMI ranged from 35 to 40 kg/m2, because that is commonly the range of overlap of most endoscopic and surgical procedures.
Only endoscopic or laparoscopic primary weight loss procedures were included. Revision
or conversion procedures were excluded.
Propensity score matching 1:50 was performed for EBT versus LBT based on age, sex,
and BMI. The primary outcome was AEs, which were defined by the occurrence of any
one of the 26 recorded AEs in the database. These AEs included: coma > 24 hours, stroke,
myocardial infarction (MI), cardiac arrest, unplanned Intensive Care Unit admission,
unplanned intubation, pulmonary embolism, deep vein thrombosis, sepsis, septic shock,
Clostridium difficile infection, urinary tract infection, superficial incisional surgical site infection,
deep incisional surgical site infection, organ space surgical site infection, wound
disruption, pneumonia, peripheral nerve injury, acute renal failure, incisional hernia,
need for transfusion, drain present 30 days post-procedure, need for a ventilator
postoperatively, unplanned intubation, treatment for dehydration as outpatient, or
Emergency Department outpatient visit.
Secondary outcomes included readmission, reoperation, reintervention, and length of
stay. Multivariable logistic regression controlling for baseline comorbidities (obstructive
pulmonary disease, history of MI, hypertension, sleep apnea, diabetes) was used to
compare the two groups with respect to the occurrence of an AE. P < 0.05 was considered significant.
The ACS Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program
and the centers participating in the ACS MBSAQIP are the source of the data used herein;
they have not verified and are not responsible for the statistical validity of the
data analysis or the conclusions derived by the authors.
Results
We matched 211 EBTs with 9,059 LBTs based on age, sex, and BMI. [Table 1] shows the demographic characteristics and comorbidities of the groups after matching.
Table 1
Demographic characteristics and comorbidities of EBT and LBT groups after matching.
|
LBT group
|
EBT group
|
P value
|
|
Age (years)
|
46.6 ± 11.6
|
46.6 ± 10.9
|
0.94
|
|
Sex (% females)
|
84.4
|
84.4
|
0.92
|
|
BMI (kg/m2)
|
37.6 ± 1.4
|
37.5 ± 1.4
|
0.23
|
|
Diabetes (%)
|
26.7
|
15.2
|
0.001
|
|
Hypertension requiring medications (%)
|
47.5
|
36.5
|
0.001
|
|
Sleep apnea (%)
|
35.7
|
25.1
|
0.001
|
|
COPD (%)
|
0.87
|
0.47
|
0.54
|
|
History of MI (%)
|
1.2
|
0.5
|
0.34
|
LBT, laparoscopic bariatric therapy; EBT, endoscopic bariatric therapy; BMI, body
mass index; COPD, chronic obstructive pulmonary disease; MI, myocardial infarction.
Operative length was significantly lower in the EBT group than in the LBT group (63.9,
95 % confidence interval [CI]: 57.9, 69.8 versus 81.1, 95 % CI: 80.1, 82.1). Length
of stay post-procedure was also significantly lower in the EBT group than the LBT
group (0.49 days, 95 % CI: 0.29, 0.69 versus 1.43 days, 95 % CI: 1.41, 1.45).
There was no difference between EBT and LBT in the odds of readmission (odds ratio
[OR] = 0.31, 95 % CI: 0.08, 1.25), reoperation (OR = 0.39, 95 % CI: 0.05, 2.84), and
reintervention (OR = 0.98, 95 % CI: 0.24, 3.99).
The odds of having any AE were lower in the EBT group than the LBT group (OR = 0.33,
95 % CI: 0.16, 0.68). After controlling for: chronic obstructive pulmonary disease
(COPD), sleep apnea, history of MI, hypertension requiring medications, and diabetes,
EBT continued to be associated with lower odds of having any AE than LBT, with an
OR of 0.34 (95 % CI: 0.16, 0.69).
Subgroup analysis
A total of 211 patients who underwent EBTs were matched with 8,541 who underwent LSGs
based on age, sex, and BMI. [Table 2] shows the demographic characteristics and comorbidities of the groups after matching.
Table 2
Demographic characteristics and comorbidities of EBT and LSG groups after matching.
|
LSG group
|
EBT group
|
P value
|
|
Age (years)
|
46.5 ± 11.7
|
46.6 ± 10.8
|
0.89
|
|
Sex (% females)
|
85.1
|
84.3
|
0.76
|
|
BMI (kg/m2)
|
37.5 ± 1.38
|
37.4 ± 1.39
|
0.38
|
|
Diabetes (%)
|
24.9
|
15.2
|
0.005
|
|
Hypertension requiring medications (%)
|
45.9
|
36.5
|
0.007
|
|
Sleep apnea (%)
|
34.5
|
25.1
|
0.004
|
|
COPD (%)
|
1
|
0.5
|
0.43
|
|
History of MI (%)
|
1.3
|
0.5
|
0.31
|
LSG, laparoscopic sleeve gastrectomy; endoscopic bariatric therapy; BMI, body mass
index; COPD, chronic obstructive pulmonary disease; myocardial infarction.
Operative length was significantly lower in the EBT group (63.9 minutes, 95 % CI:
57.9, 69.8) than the LBT group (69 minutes, 95 % CI: 68.3, 69.8), though the difference
of a few minutes may not be of clinical significance. Length of stay post-procedure
was also significantly lower in the EBT group than in the LSG group (0.49 days, 95 %
CI: 0.29, 0.69 in EBTs versus 1.38 days, 95 % CI: 1.36, 1.39 in LSG).
There was no difference between patients undergoing EBT and LBT in the odds of readmission
(OR = 0.41, 95 % CI: 0.11, 1.67), reoperation (OR = 0.47, 95 % CI: 0.06, 3.42), or
reintervention (OR = 2.14, 95 % CI: 0.51, 8.93).
The odds of having any AE were lower in the EBT group than in the LBT group (OR = 0.38,
95 % CI: 0.19, 0.78). After controlling for the comorbidities previously listed, EBT
continued to be associated with a lower risk of of having any AE than LSG, with an
OR of 0.39 (95 % CI: 0.19, 0.79).
Discussion
This is the first study to compare 30-day safety outcomes of EBT versus LBT from a
large and validated quality improvement database, the MBSAQIP. Overall, our findings
suggest that EBT is associated with a significantly lower rate AEs than LBT in the
30-day post-procedure period. Rates of readmission, reoperation, and reintervention
were not significantly different between these two groups.
Only a handful of studies have been performed thus far comparing the outcomes of EBT
and LBT. Of them, there has only been one study with large-volume data. This study
used the MBSAQIP database to compare an IGB to bariatric surgeries using the MBSAQIP
database, which included 145,408 patients undergoing IGB and 144,627 patients undergoing
laparoscopic gastric bypass in 2018 [10]. Propensity-matched analysis revealed a higher overall AE rate with IGB when compared
to LBT (5.0 % versus 2.6 %, P = 0.024). Other published studies have mostly focused
on comparing ESG to LSG in particular and their findings are suggestive of better
safety outcomes with ESG, but superior weight loss outcomes with LSG [6]
[11]. One retrospective analysis comparing ESG, LSG, and laparoscopic adjustable gastric
banding in 279 patients with obesity found that LSG achieved the highest percent total
body weight loss (%TBWL) (29.28 vs 13.30 vs 17.57 %, respectively). However, ESG was
found to have a significantly lower rate of morbidity and length of stay when compared
to the other techniques [11]. A case-matched retrospective analysis performed by our group also compared ESG
and LSG outcomes [6]. A total of 54 patients undergoing ESG were matched with 83 patients undergoing
LSG. Lower rates of AEs were associated with ESG compared to LSG (5.2 % vs 16.9 %,
P < 0.05). Interestingly, higher rates of de-novo gastroesophageal reflux disease
were encountered with LSG compared to ESG (14.5 % vs 1.9 %, P < 0.05). Our current study suggests that overall, for any type of EBT, there appears
to be better short-term safety outcomes as compared to LBT. When comparing EBT to
LSG alone, the short-term safety results were similar.
The strength of this study is that it was based on prospectively collected data from
multiple institutions, which allowed them to serve as an accurate representation of
real-world outcomes and it adds to the repertoire of information needed to provide
patient-centered, personalized care.
In addition, we controlled for differences between the groups in age, sex, BMI, and
comorbidities with propensity score matching and multivariable regression. As expected,
the LBT group had significantly more comorbidities, namely sleep apnea, diabetes,
and hypertension. Interestingly, however, history of MI and COPD were not significantly
different between the two groups. Finally, the study provides additional information
on operative length and length of stay. Length of stay was significantly lower in
the EBT group and was notably < 1 day as compared to LBT, which had a mean length
of stay > 1 day, suggesting that overnight observation is required post-procedure.
The authors would like to acknowledge the limitations of this study. One is that the
procedure type for EBT had not been recorded in the database, limiting our ability
to subcategorize and study different procedures. This is an inherent limitation of
the database, which we hope can be amended to allow more informed research and quality
improvement for EBT. There is likely a disproportionate representation of IGB over
other EBTs in these data, which must be acknowledged. Nevertheless, our results remain
a fair representation of the current landscape of EBTs. Another limitation is the
absence of long-term safety data, which limits our ability to make broader statements
about the overall safety of these procedures. However, with the striking difference
between these procedures seen in this analysis, the results remain significant and
should be further explored in future prospective studies. When reviewing the results
of this study, consideration must be given to the novelty of EBT compared to LBT,
which has been well established. It is likely that the safety profile of EBT will
further improve with increasing individual and collective experience.
A discussion comparing EBT to LBT is only complete after comparing weight loss, effects
on comorbidities, and long-term safety outcomes as well. However, this study focuses
only on short-term safety data. Future studies must continue to compare these different
facets of treatment and prospective data is much needed. The finding of shorter duration
of procedure and length of stay in EBTs vs LBTs may also have strong implications
regarding the relative cost of these procedures. Future research is needed to compare
the cost-effectiveness of these procedures.
Conclusions
There is currently a menu of bariatric surgical options available for patients to
choose from and information about EBT must continue to expand. Shorter duration of
procedure, shorter length of stay, and lower rates of post-procedural AEs all contribute
to making EBT more palatable for both patients and physicians.
