Randomized sham-controlled trial of the 6-month swallowable gas-filled intragastric balloon system for weight loss

Accepted Manuscript

Randomized Sham-Controlled Trial of the Six-Month Swallowable Gas-Filled Intragastric Balloon System for Weight Loss Shelby Sullivan MD , James Swain MD , George Woodman MD , Steven Edmundowicz MD , Tarek Hassanein MD , Vafa Shayani MD , John C. Fang MD , Mark Noar MD , George Eid MD , Wayne J. English MD , Nabil Tariq MD , Michael Larsen MD , Sreenivasa S. Jonnalagadda MD , Dennis S. Riff MD , Jaime Ponce MD , Dayna Early MD , Eric Volkmann MD , Anna R. Ibele MD , Matthew D. Spann MD , Kumar Krishnan MD , Juan Carlos Bucobo MD , Aurora Pryor MD PII: DOI: Reference:

S1550-7289(18)31096-7 https://doi.org/10.1016/j.soard.2018.09.486 SOARD 3514

To appear in:

Surgery for Obesity and Related Diseases

Received date: Revised date: Accepted date:

9 May 2018 14 September 2018 25 September 2018

Please cite this article as: Shelby Sullivan MD , James Swain MD , George Woodman MD , Steven Edmundowicz MD , Tarek Hassanein MD , Vafa Shayani MD , John C. Fang MD , Mark Noar MD , George Eid MD , Wayne J. English MD , Nabil Tariq MD , Michael Larsen MD , Sreenivasa S. Jonnalagadda MD , Dennis S. Riff MD , Jaime Ponce MD , Dayna Early MD , Eric Volkmann MD , Anna R. Ibele MD , Matthew D. Spann MD , Kumar Krishnan MD , Juan Carlos Bucobo MD , Aurora Pryor MD , Randomized Sham-Controlled Trial of the Six-Month Swallowable Gas-Filled Intragastric Balloon System for Weight Loss, Surgery for Obesity and Related Diseases (2018), doi: https://doi.org/10.1016/j.soard.2018.09.486

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Highlights: 

Endoscopic Bariatric Therapies are a new class of obesity treatment options including devices that require endoscopy for placement or removal, including



CR IP T

intragastric balloons that are space occupying devices in the stomach Randomized controlled trials from two fluid filled intragastric balloons that are placed endoscopically and removed endoscopically have demonstrated

significantly more weight loss with the intragastric balloons plus lifestyle therapy



AN US

compared with lifestyle therapy alone.

This manuscript describes the randomized sham controlled trial for safety and efficacy of lifestyle therapy with or without a swallowable gas filled intragastric

AC

CE

PT

ED

M

balloon for weight loss.

1

ACCEPTED MANUSCRIPT

Randomized Sham-Controlled Trial of the Six-Month Swallowable Gas-Filled Intragastric Balloon System for Weight Loss Shelby Sullivan MD1,2, James Swain MD3, George Woodman MD4, Steven Edmundowicz MD1,2, Tarek Hassanein MD5, Vafa Shayani MD6, John C. Fang MD7,

CR IP T

Mark Noar MD8, George Eid MD9, Wayne J. English MD10, Nabil Tariq MD11, Michael Larsen MD12, Sreenivasa S. Jonnalagadda MD13, Dennis S. Riff MD14, Jaime Ponce MD15, Dayna Early MD1, Eric Volkmann MD7, Anna R. Ibele MD7, Matthew D. Spann

AN US

MD10, Kumar Krishnan MD11, Juan Carlos Bucobo MD16, and Aurora Pryor MD16 1

Washington University School of Medicine, St. Louis, MO; 2University of Colorado

School of Medicine, Aurora, CO; 3HonorHealth Research Institute, Scottsdale, AZ; 4

Midsouth Bariatrics, Memphis, TN; 5Southern California Research Center, Coronado,

M

CA; 6Bariatric Institute of Greater Chicago, Bolingbrook, IL; 7University of Utah Hospital, Salt Lake City, UT; 8Endoscopy Microsurgery Associates, Townson, MD; 9Alleghany

ED

Singer Research at West Penn, Pittsburgh, PA; 10Vanderbilt University, Nashville, TN; 11

PT

Houston Methodist Research Institute, Houston, TX;

12

Virginia Mason Medical Center,

Seattle, WA; 13Saint Luke's Hospital of Kansas City, Kansas City, MO;

14

Anaheim

CE

Clinical Trials, Anaheim, CA; 15Chattanooga Bariatrics, Chattanooga, TN; 16Stony Brook

AC

Medicine, Stony Brook, NY

Corresponding Author: Shelby Sullivan University of Colorado School of Medicine 12631 E 17th Ave

2

ACCEPTED MANUSCRIPT

Aurora, CO 80045 [email protected] Tele: 314-265-9277 Fax: 720-848-2749

CR IP T

Acknowledgements: We would like to thank Dr. Jie Yang for her statistical analyses. Running Title: The SMART Trial

AN US

Running Title: The SMART Trial Word Count: 4354 words

Clinical Trial Registration: clinicaltrials.gov NCT02235870

M

Funding: The SMART trial was funded by Obalon Therapeutics, Carlsbad, CA, USA. All authors received contracted research support through their institutions from Obalon

ED

Therapeutics.

PT

Abstract

Background: Obesity is a significant health problem and additional therapies are needed

CE

to improve obesity treatment.

AC

Objective: Determine the efficacy and safety of a six-month swallowable gas-filled intragastric balloon system for weight loss. Setting: 15 academic and private practice centers in the United States Methods: This was a double blind randomized sham-controlled trial of the swallowable gas-filled intragastric balloon system plus lifestyle therapy compared with lifestyle 3

ACCEPTED MANUSCRIPT

therapy alone for weight loss at 6 months in participants age 22-60 years old, BMI 30-40 kg/m2, across 15 sites in the US. The endpoints included: difference in percent total body weight loss (TBWL) in Treatment Group vs Control Group is > 2.1%, and a

CR IP T

Responder Rate of > 35% in the Treatment group. Results: 387 subjects swallowed at least one capsule. 93.3% of participants completed all 24 weeks of blinded study testing. Non-serious Adverse events occurred in 91.1% of patients, but only 0.4% were severe. One bleeding ulcer and one balloon deflation

AN US

occurred. In the completer analysis, the Treatment and Control Groups achieved

7.1±5.0% and 3.6±5.1% TBWL respectively, and a mean difference of 3.5%, p= 0.0085. Total body weight loss in Treatment and Control Groups were 7.1 ± 5.3kg and 3.6 ± 5.1kg (p <0.0001) and BMI change in the Treatment and Control groups were 2.5 ± 1.8

M

kg/m2 and 1.3 ± 1.8 kg/m2 (p<0.0001), respectively. The Responder Rate in the Treatment Group was 66.7%, p<0.0001. Weight loss maintenance in the Treatment

ED

group was 88.5% at 48 weeks.

PT

Conclusions: Treatment with lifestyle therapy and the 6-month swallowable gas-filled intragastric balloon system was safe and resulted in twice as much weight loss

CE

compared with a sham control, with high weight loss maintenance at 48 weeks.

AC

Key Words: Swallowable Gas-Filled Intragastric Balloon System, endoscopic bariatric therapy, intragastric balloon Introduction Obesity continues to be a significant medical problem in the United States, with a prevalence of 37.7% and an increase in overall obesity rates in women from 2005 to

4

ACCEPTED MANUSCRIPT

20141. Multiple etiologies are likely contributing to the increasing prevalence of obesity, including lack of treatment options that are acceptable to patients. Recently, endoscopically placed and/or removed bariatric therapies have emerged as viable treatment options. These approaches have less weight loss than bariatric surgery, but

CR IP T

are also associated with significantly lower complication rates and are approved for use in lower BMI categories2-5. Lowering complication rates further while maintaining

weight loss benefits may increase patient acceptance and increase obesity treatment

AN US

rates.

A new six-month swallowable intragastric balloon system (SGIBS, Obalon Balloon System, Obalon therapeutics, Carlsbad, CA), comprised of 3 thin walled 250 ml nitrogen-mix gas filled balloons which are swallowed successively over time has been

M

developed for weight loss. A pilot study of 17 patients (BMI 31.0±2.6 kg/m2, weight 85.2±11.1 kg, age 39.0±9.0, Male/Female 2/15) with a previous generation system

ED

implanted for 12 weeks, demonstrated weight loss of 5.0 kg or 5.9% total body weight

PT

loss (%TBWL). All adverse events reported were mild to moderate and no serious adverse events or unanticipated device events occurred6. These results suggested that

CE

treatment with the previous generation of the balloon system may result in successful

AC

weight loss and a low rate of complications. The purpose of the Six Month Adjunctive Weight Reduction Therapy (SMART)

Trial was to determine the safety and efficacy of the SGIBS in a prospective randomized sham controlled trial with all patients receiving the same lifestyle therapy program. We hypothesized that %TBWL and the Responder Rate (defined as ≥5% TBWL) would be higher in the subjects receiving the SGIBS compared with control subjects and that the 5

ACCEPTED MANUSCRIPT

rate of serious adverse events and non-serious adverse events rated severe would be low. Methods

CR IP T

Trial Design This was a multicenter 24-week prospective randomized double-blind sham controlled trial with 24-week follow-up and control arm cross over after unblinding

conducted across 15 centers in the United States from March 9, 2015 through May 19,

AN US

2016. After passing all screening assessments, participants were randomized in a 1:1 ratio to treatment with the SGIBS plus lifestyle therapy (Treatment Group) or lifestyle therapy alone (Control Group). Randomization was computer generated using a central

M

randomization scheme with random sized block design. Only un-blinded study team members had access to group allocation, unblinded study documents, and the

ED

electronic data capture system to ensure protection of the study blind. Both the study participants and registered dietitians performing follow-up visits were blinded to study

PT

assignment. The study was approved by each individual study site institutional review

CE

board, registered with FDA, and written informed consent was obtained from all participants in accordance with the guidelines of the Declaration of Helsinki. The study

AC

was registered on clinicaltrials.gov (NCT02235870). Participants

Adults age 22-64 years old weight stable for 12 months with a BMI between 3040 kg/m2, at least one attempt to lose weight through a medically or non-medically supervised weight loss program without success and willing to avoid non-commercial air 6

ACCEPTED MANUSCRIPT

travel or scuba diving during the study period were eligible for enrollment in the trial. Patients were recruited through local news reports, radio advertising and print advertising. Key exclusion criteria included: failure to swallow a test placebo capsule, use of medications known to cause weight loss or weight gain, use of nonsteroidal anti-

CR IP T

inflammatory drugs (NAIDS), history of structural or functional disorders of the esophagus, prior foregut surgery, hiatal hernia >2 cm diagnosed on upper

gastrointestinal series imaging, peptic ulcer disease, type 1 or type 2 diabetes requiring oral medications or insulin, poorly controlled hypertension, and severe organ

AN US

dysfunction.

Components of the Swallowable Gas-Filled Intragastric Balloon System The SGIBS consists of 3 nylon polyethylene blend balloons with a radiopaque

M

metal self-sealing valve attached to a small diameter (3 French) catheter and filled with

ED

250 ml of a nitrogen mix gas. Each balloon is folded within a hard gelatin US Pharmacopeia grade capsule. A catheter is attached to the capsule at the metal self-

PT

sealing valve for device administration and inflation. The balloons are inflated with the dispenser, which also contains a battery powered pressure gauge and regulator for

CE

pressure readings from the balloon and control of gas pressure, which was calibrated to

AC

the altitude of each study site (Figure 1.) Due to gas pressures in the balloons, patients must not permanently live in a location 2500 feet below or 4000 feet above administration altitude. Device administration and removal

7

ACCEPTED MANUSCRIPT

After passing all screening assessments and randomization, patients reported to the study site for initial capsule administration. The Treatment Group participants received capsules that contained a swallowable balloon. Control group participants received a capsule filled with a food grade sugar, which was similar in weight and

CR IP T

appearance to the swallowable balloon filled capsule. Prior to capsule swallow, the dispenser was prepared by attaching extension tubing and filling the extension tubing with a small amount of gas from the gas canister. All participants swallowed the capsule and tethered catheter per their study group assignment. In all cases,

AN US

radiographic images were obtained to confirm advancement of the capsule into the stomach with digital X-Ray or Fluoroscopy. The catheter was then attached to the extension tubing pre-filled with gas, and the second verification was performed by

M

measuring pressure inside the balloon with the pressure gauge on the dispenser with the gas pre-filled in the extension tubing. Once the capsule dissolved off of the balloon

ED

and the balloon was able to unfold without restriction (which can only happen in the stomach), the pressure dropped below 7.0 kilopascal (kPa). The dispenser valve was

PT

then opened and the nitrogen-mix gas inflated the balloon through diffusion from the

CE

gas container to the balloon. Once the balloon pressure equalized between 9-13 kPa, the catheter was detached from the balloon, and the valve self-sealed. After balloon

AC

inflation or sham balloon inflation, a radiographic image was obtained to verify balloon inflation. Further details about the device administration can be found in Supplementary Materials A. All subjects were treated prophylactically for balloon administration symptoms with an anti-emetic medication and anti-spasmodic medication, which were required to start the night before the administration and continue for 5 days after

8

ACCEPTED MANUSCRIPT

administration. Additional balloon or sham capsule administrations occurred at week 3 and at either week 9 if weight stable for three weeks or week 12 if weight loss occurred between weeks 6 and 9.

CR IP T

24 weeks after the initial balloon administration, all balloons were removed endoscopically under monitored anesthesia care or conscious sedation. Each balloon was deflated by aspirating the nitrogen-mix gas with a standard endoscopy injection needle, then secured with a rat-toothed alligator forceps and removed through the

AN US

mouth. This process was repeated for the other two balloons. Follow-up care

All subjects were treated with a proton pump inhibitor for the duration of 24

M

weeks of the randomized blinded portion of the trial, and in the Control Group participants who crossed over to receive balloon administrations. A single well-defined

ED

weight loss behavior modification program was designed prior to any participant enrollment and utilized for lifestyle therapy. The program was a moderate intensity

PT

program developed in accordance with AHA/ACC/TOS 2013 Guidelines for the Management of Overweight and Obesity in Adults with a goal weight loss of 1.5-2

CE

pounds per week7. The program was standardized across all sites and both the control

AC

and treatment groups to eliminate any treatment bias by any site or RD. Participants underwent lifestyle therapy sessions administered by a blinded registered dietitian every 3 weeks for the first 24 weeks. To ensure that all subjects received the same level of lifestyle therapy intervention, additional subject consultations throughout the study period were not permitted. Both randomization assignments received the same training and the same topics at each pre-specified week throughout the course of the 6-month 9

ACCEPTED MANUSCRIPT

blinded evaluation. All subjects were to be counseled at the pre-scheduled visits for 30-35 minutes and time of each consultation session was recorded. Treatment group participants continued the standardized lifestyle therapy after balloon removal and unblinded using the standardized program every 3 weeks for an additional 24 weeks (48

CR IP T

weeks total).

Adverse events were assessed prior to capsule administration, 24 hours after capsule administration, and at each 3-week follow-up visit. All non-serious adverse

AN US

events were categorized based on FDA guidance in the Code of Federal Regulations8 (Supplementary Material B): Outcomes

M

The study had two co-primary efficacy endpoints: 1) the difference in mean %TBWL between the Treatment group and Control group, which had to be at least 2.1%

ED

and 2) at least 35% of subjects in the Treatment achieving a weight loss response, with response defined as losing ≥5% TBWL at 24 weeks. The primary safety endpoint was

PT

evaluated with the percentage of subjects who experienced a serious adverse event

CE

that was attributed to device or procedure. Secondary endpoints included weight loss maintenance from week 24 to week 48 and changes in cardiometabolic risk factors from

AC

baseline to week 24 for the Treatment and Control groups. Calculations

Percent total body weight loss was calculated as:

%𝑇𝐵𝑊𝐿 =

𝑆𝑡𝑎𝑟𝑡𝑖𝑛𝑔 𝑊𝑒𝑖𝑔ℎ𝑡 (𝑘𝑔) − 𝐸𝑛𝑑𝑖𝑛𝑔 𝑊𝑒𝑖𝑔ℎ𝑡 (𝑘𝑔) × 100% 𝑆𝑡𝑎𝑟𝑡𝑖𝑛𝑔 𝑊𝑒𝑖𝑔ℎ𝑡 (𝑘𝑔)

10

ACCEPTED MANUSCRIPT

Percent Excess Weight Loss was calculated as: 𝑆𝑡𝑎𝑟𝑡𝑖𝑛𝑔 𝑊𝑒𝑖𝑔ℎ𝑡 (𝑘𝑔) − 𝐸𝑛𝑑𝑖𝑛𝑔 𝑊𝑒𝑖𝑔ℎ𝑡 (𝑘𝑔) × 100% 𝐸𝑥𝑐𝑒𝑠𝑠 𝑊𝑒𝑖𝑔ℎ𝑡 (𝑘𝑔)

CR IP T

%𝐸𝑊𝐿 =

𝐸𝑥𝑐𝑒𝑠𝑠 𝑊𝑒𝑖𝑔ℎ𝑡 (𝑘𝑔) = 𝑆𝑡𝑎𝑟𝑡𝑖𝑛𝑔 𝑊𝑒𝑖𝑔ℎ𝑡(𝑘𝑔) − 𝑊𝑒𝑖𝑔ℎ𝑡 (𝑘𝑔)𝑓𝑜𝑟 𝑎 𝐵𝑀𝐼 𝑜𝑓 25 𝑘𝑔⁄𝑚2

25 × 𝐻𝑒𝑖𝑔ℎ𝑡 (𝑐𝑚)2 10,000

AN US

𝑊𝑒𝑖𝑔ℎ𝑡 (𝑘𝑔) 𝑓𝑜𝑟 𝑎 𝐵𝑀𝐼 𝑜𝑓 25 𝑘𝑔⁄𝑚2 =

Sample Size

The sample size for the two co-primary endpoints was determined using SAS 9.3

M

software (SAS Institute Inc., Cary, NC) POWER Procedure (PROC POWER) for two sample means and one sample frequency, providing at least 80% power with a one-

ED

sided significance level of 0.025 and a randomization of 1:1, treatment to control.

PT

Assumptions included a mean difference of 3.4% TBWL between the treatment and control groups with a standard deviation of 4.0% and a responder rate of 46% in the

CE

treatment arm, with 304 subjects total required to meet 80% power for both co-primary endpoints. Assuming an attrition rate of 10% after randomization and 50% at screening,

AC

676 subjects were required to ensure enough subjects completed the trial to achieve statistical power of at least 80%. Statistical methods The modified intention to treat analysis population was defined as any subject who swallowed at least one capsule. The completer analysis population were all 11

ACCEPTED MANUSCRIPT

subjects who completed study testing through week 24. Analysis of Covariance (ANCOVA) was utilized to evaluate the mean difference in %TBWL with starting weight as a covariates, and difference in cardiometabolic change from Baseline to Week 24 between the treatment and control groups with corresponding baseline cardiometabolic

CR IP T

data was used as a covariate. The analyses included using the starting weight as a covariate for %TBWL. Binomial proportion with a 95% Clopper-Pearson exact

confidence interval was utilized for percentage of responder rate. Chi-square test was utilized for comparing the responder rates between the device arms. The weight loss

AN US

maintenance analysis was assessed with a paired t-test. For missing weight data, the Last observation Carried Forward (LOCF) imputation method and combining estimates across 10 imputed datasets by the method of Rubin9 were evaluated separately to

M

analyze the efficacy results. All means listed for outcome measures in the randomized controlled portion of the study are least squares means to adjust for the mean starting

ED

weight difference between groups with standard deviation with either standard deviation

analyses.

CE

Results

PT

or 95% confidence interval. SAS 9.3 software was used to perform all the statistical

AC

Subjects

Three hundred eighty-seven subjects were randomized, successfully swallowed

the first capsule and were included in the modified intention to treat analysis (mITT, Treatment Group n=198 and10- Control Group n=189)). N=350 subjects completed the 24-week randomized sham controlled study testing with three successful capsule swallowing (Treatment Group n=174 and Control n=176) and were included in the 12

ACCEPTED MANUSCRIPT

completer analysis (Figure 2). Baseline demographics and characteristics are seen in Table 1. N=160 patients in the Treatment Group continued study participation after unblinding through week 48, and 128 patients in the Control Group continued study participation after unblinding, received SGIBS treatment, and completed study testing

CR IP T

through week 48. In the Treatment Group, 68.6% of patients correctly guessed their study assignment at week 0 and 90.5% at week 24. In the Control Group, 35.5% of patients correctly guessed their study assignments at week 0 and 65.6% at week 24.

AN US

Procedures

Successful capsule swallow with the initial capsule administration was 91.7% and 93.1% in the Treatment and Control groups respectively, and one (0.7%) failed initial capsule administration in the unblinded crossover group. Success rates for the second

M

and third capsule administrations were 98.9% to 100.0% in the blinded and crossover

ED

groups. The balloon administration and inflation time was 9.8 ± 4.0 minutes. The threeballoon removal procedure time was 15.6±7.2 minutes, and 98.5% of removals occurred

PT

with conscious sedation or monitored anesthesia care using propofol.

CE

Weight Loss Outcomes

%TBWL in the mITT analysis at 24 weeks was 6.6±5.1% in the Treatment Group

AC

and 3.4±5.0% (p=0.0354) in the Control Group (Figure 3a, Table 2), with a difference of 3.2% (95% CI 2.2, 4.2) meeting the first co-primary endpoint. %TBWL in the completer analysis at 24 weeks was 7.1±5.0% in the Treatment Group and 3.6±5.1% in the Control Group (p=0.0085), with a difference of 3.5% (95% CI 2.5, 4.6; Figure 3b, Table 3).

13

ACCEPTED MANUSCRIPT

The Responder Rate in the mITT analysis was 62.1% in the Treatment Group which was greater than the second co-primary endpoint of 35% (p<0.0001) and was greater than the Responder Rate in the Control Group which was 30.7% (p<0.0001).

compared with 31.8% in the Control Group, p<0.0001.

CR IP T

The Responder Rate in the Completer analysis was 66.7% in the Treatment Group

Weight loss impact may be magnified in lower BMI groups, thus the breakdown for BMI 30-34.9 and BMI 35-40 are broken down for the mITT cohort in Table 4 and for

AN US

the Completers in Table 5. For all groups, the difference between treatment and controls remains significant.

Weight loss maintenance at 48 weeks in subjects with any weight loss at week 24 in the Treatment Group was 88.5% (n=151, 7.8±4.4%TBWL at 24 weeks and

M

6.9±6.5% TBWL at 48 weeks, Figure 3c). Weight loss maintenance in the entire mITT

ED

Treatment group was 87.9%. Total body weight loss in the subjects in the Control Group who still qualified for and received the SGIBS after unblinding and completed

PT

week 48 testing was 3.6% ± 4.4% at week 24 and 7.0±6.2% at week 48. Patients in the Control Group who had a BMI <30 kg/m2 at the time of unblinding were not eligible to

CE

receive the SGIBS.

AC

Cardiometabolic Outcomes Several cardiometabolic outcomes improved more in the Treatment Group

compared with the Control Group at week 24. Decreases in systolic blood pressure, plasma total cholesterol concentration, plasma triglyceride concentration, and plasma

14

ACCEPTED MANUSCRIPT

glucose concentration were significantly greater in the Treatment Group Compared with the Control group (Table 6). Balloon Deflations and Early Removals

CR IP T

Out of 981 successfully administered balloons, one deflation (0.1%) occurred. The balloon was retrieved without sequela. Out of 336 patients who received the

SGIBS, early removals due to non-serious adverse events occurred in 3.3% of subjects.

AN US

Adverse Events

Adverse events are listed for all patients who received any swallowable balloons during the 48- week trial (n=336, Table 7). 99.5% of AEs were rated as mild or moderate. 75.6% of balloon residency adverse events required no intervention. 9.3% of

M

events were treated with peppermint teas or over-the-counter medications. The majority of the moderate events were medically managed with an increase in PPI

ED

dosage from 40 mg once a day to twice a day, extension of the use of the antispasmodic or anti-emetic medication, or an additional prescription of sucralfate (1

PT

gm four times a day) for the duration of the therapy to treat abdominal cramping and

CE

persistent nausea. Less than 3% of events related to balloon residency required additional diagnostic procedure or intervention outside of prescription medications, and

AC

0.9% of adverse events during balloon removals required additional endoscopic therapy. All adverse events for the Treatment Group compared with the Control Group through week 24 can be seen in Supplementary Material C and D. One serious adverse event occurred in the study including both the Treatment group and Control group subjects who crossed over after the blinded portion of the 15

ACCEPTED MANUSCRIPT

study (0.3%). One subject developed a bleeding gastric ulcer after taking protocol prohibited NSAIDs The subject received two units of packed red blood cells, and an upper endoscopy performed within 24 hours of the event revealed ulcers and erosions consistent with NSAID induced peptic ulcer disease, but no lesions actively bleeding or

CR IP T

requiring endoscopic therapy.

Discussion

AN US

Additional therapies with the possibility of widespread adoption are needed to improve obesity treatment rates. The present study was the first randomized sham controlled trial of a swallowable nitrogen-mix gas filled balloon system to treat obesity.

M

The results from this trial demonstrate the safety and efficacy of SGIBS for weight loss. The serious adverse event rate (single bleeding ulcer) was low (0.3%) and early balloon

ED

deflation without migration occurred in only 1 balloon for a rate of 0.1%. The balloons were well tolerated with only 0.4% of non-serious adverse events rated as severe.

PT

Moreover, there was twice as much weight loss in the Treatment Group compared with

CE

the Control group and weight loss maintenance was 88.5% at 48 weeks. Weight loss also significantly increased in the Control subjects that were allowed to receive the

AC

SGIBS after unblinding, despite the fact that subjects with a higher response to weight loss interventions (ie patients who achieved a BMI <30 kg/m 2) were excluded from the cross-over.

The balloons are easy to administer without the need for sedation, which allows for a large number of practitioners who could be trained to administer the balloons.

16

ACCEPTED MANUSCRIPT

Although balloon removal does require an endoscopist, the removal procedure is short and will require only minimal training as endoscopists are already trained to remove foreign bodies from the stomach. These factors could increase access for patients and decreased treatment costs. In addition, administering three balloons over time allows

CR IP T

for individualization of therapy to optimize patient weight loss and minimize

accommodative symptoms, potentially increasing patient acceptance of this therapy. Previous publications of the SGIBS used a prior generation of balloon did not use

AN US

the current gas formulation and was designed for 3 months use. In one open label pilot study, balloon administration was 98% successful and weight loss was 36% EWL at 3 months (5 kg, range 0-10.2 kg) with no device malfunctions or serious adverse events 6. A case series in 17 adolescent children treated with one or two 3-month SIGBS also

M

found a weight reduction of 95.8 ± 18.4 kg to 83.6 ± 27.1 kg (p < 0.05), but also extended the treatment time to 18.6± 2.4 weeks10. Additional publications of the earlier 3-month

ED

version of the SGIBS include a prospective analysis of 72 patients undergoing therapy

PT

who experienced 19.4% median excess weight loss (EWL) at 12 months, and demonstrated the highest median %EWL in patients with class I obesity11. Lastly, a

CE

case series described in a book chapter described treatment with the 3 month SGIBS in 62 patients (age 41.3±14.9 yrs, BMI 38.2±7.3 kg/m2)12. Unlike the present manuscript,

AC

only 21 (33.9%) patients received 2 balloons and 6 (9.7%) patients received all 3 balloons. In one patient the balloon was placed endoscopically, but it is important to note that this is not recommend. Only 6 patients required post-placement symptom treatment, with a single dose of butylscopolamine bromide. Percent EWL was 23.3±8% and no serious adverse events occurred. Superficial gastric erosions were seen in 4

17

ACCEPTED MANUSCRIPT

patients who had stopped their proton pump inhibitor therapy. 4 patients did not return at 3 months for removal but were evaluated between 5-6 months after placement and all balloons had passed out of the GI tract. These data align with the present study despite using a prior generation of balloon, a shorter duration of treatment, and in some cases

CR IP T

fewer balloons in the treatment period.

Another swallowable balloon, which is not FDA-approved has been described in a feasibility study of 135 patients13. That swallowable balloon is saline filled, used for 4-

AN US

months, and is intended to self-deflate and pass out of the GI tract. In that series 3 patients required endoscopy for early balloon removal and one patient required laparoscopy for small bowel obstruction. Weight loss was 15.1% TBWL. Initial published results with the 6 month SGIBS described in this manuscript

M

include two published abstracts from our trial group14,15, which support the data

ED

presented in this manuscript. Weight loss compared to sham control in the present study is similar to a saline-filled dual balloon available in the US studied in a similarly

PT

designed randomized sham-controlled trial 3. The weight loss in both of the multi-center sham-controlled studies of IGB’s in the US (the present study, SMART Trial, and the

CE

afore mentioned REDUCE Trial) were less than the weight loss seen in the randomized

AC

controlled trial of the saline-filled single balloon4. A possible reason for the difference in weight loss is the effect a blinded sham design has on lowering weight loss in treatment arm. This was demonstrated in the ESSENTIAL trial of the primary obesity surgery endoluminal (POSE) procedure where plications are placed in the stomach for weight loss. This was a multi-center randomized sham-controlled trial, however 31 subjects were not randomized due to training purposes, completed one year of therapy, and 18

ACCEPTED MANUSCRIPT

were called the Lead-In group. Knowledge of the assigned group was the only difference between the Lead-In and Treatment groups, but that knowledge increased weight loss by 42%16. A detailed comparison of all the currently FDA approved

CR IP T

intragastric balloon system can be seen in a recently published review17. Treatment subjects achieved greater improvements in several cardiometabolic risk factors including systolic blood pressure, fasting plasma total cholesterol,

triglycerides, and glucose. These improvements are most likely related to the difference

AN US

in weight loss that was seen in the Treatment compared to the Control groups. The changes were modest, but larger changes would not be expected as the majority of subjects had normal blood pressure, fasting lipids, and fasting glucose at baseline. The percent of weight loss that was maintained from week 24 to week 48 weight

M

loss was higher in this study than previously reported US intragastric balloon trials3,4.

ED

The mechanism for this difference is not clear, but could be related to the rate of weight loss during balloon implantation in the stomach. Studies have demonstrated that most

PT

of the weight loss with fluid filled balloons occurs in the first 3-4 months after balloon placement18. In the SMART trial, rate of weight loss increased after the 3rd balloon was

CE

administered in the Treatment Group, and patients were still losing weight at 24 weeks

AC

when the balloons were removed. Only 1/981 balloons administered during this study was found to be partially

deflated in the stomach at the time of retrieval in the current study, and there were no reports of bowel obstruction. Outside of this trial, there is one published report of a balloon deflation with the previous 3-month SGIBS version that did require endoscopic removal19. The true incidence of deflation in the 6 month SGIBS is unknown but these 19

ACCEPTED MANUSCRIPT

data suggest that the deflation rate is lower than earlier air filled balloon20,21 and other currently FDA-approved balloon systems3,22. Adverse events (including non-serious and serious adverse events) occurred in

CR IP T

91.1% of patients, however 99.5% of these were non-serious and rated as mild to moderate with a combined rate of serious adverse event and non-serious adverse

events rated as severe of 0.5% in the current study. The low rate of severe adverse events may be related to progressive administration of balloons with a smaller volume

AN US

over time. This administration strategy appears to reduce gastric accommodative

symptoms. Moreover, no subjects required intravenous fluid hydration for dehydration due to nausea or vomiting. In addition, only one serious adverse event, bleeding ulcer, occurred in one subject using non-steroidal medications prohibited in the study protocol.

M

The direct contributions of the balloons and the NSAIDs in this case are not completely clear, but the prohibited study medications likely played an important role in the

ED

development of the gastric ulcer.

PT

This study has several limitations. The study was not powered to determine effects on metabolic outcomes and most subjects had normal metabolic parameters at

CE

baseline testing before treatment. Therefore, large changes in metabolic outcomes

AC

would not have been seen in this study population, but might be present in a different population. In addition, although weight loss was significantly greater in the Treatment Group compared to Control Group, it was less impactful than a surgical therapy with an average 15.6 lb weight loss in the treatment group at 6 months. This is in part attributed to the non-personalized, moderate intensity lifestyle therapy which limited the ability of dietitians to provide additional support to patients who had less weight loss in 20

ACCEPTED MANUSCRIPT

contrast to treatment in clinical practice and the sham-controlled trial design, which has been shown to reduce weight loss in a multi-center clinical trial as noted above16. This likely resulted in less weight loss than would be seen in clinical practice with a more individualized treatment focus. Over 85% of participants in this study were women. 3-5

. However, men in the

CR IP T

This is consistent with other endoscopic bariatric therapy trials

Treatment group lost 9.1±5.4 %TBWL, so it is likely that men will lose at least as much weight as women in clinical practice. In addition, this study only followed the original treatment group for 6 months after SGIBS removal. Although weight loss maintenance

AN US

was high at 88.5%, weight loss maintenance beyond 6 months post removal is

unknown. Lastly, this study did not investigate the mechanism of action of weight loss. The rate of weight loss increased after the third balloon placement, which suggests

M

volume of the balloons was important for weight loss. This contrasts the fluid filled balloons, in which a recent meta-analysis found no correlation between filling volume

ED

and weight loss with a single fluid filled balloon

.

PT

Conclusions

23

Treatment with the SGIBS resulted in two times more weight loss compared with

CE

lifestyle therapy alone in a randomized sham controlled trial with a low rate of serious

AC

adverse events and non-serious adverse events rated as severe. In addition, weight loss maintenance 6 months after balloon system removal was high. The weight loss and safety profile seen in this trial support the SGIBS as an effective safe treatment option for patients with Class I and Class II obesity. Disclosures:

21

ACCEPTED MANUSCRIPT

The following authors have conflicts of interest to disclose: Shelby Sullivan: contracted research for Obalon Therapeutics, Aspire Bariatrics, Allurion Technologies, Elira, and BARONova; consulting for Aspire Bariatrics, GI Dynamics, USGI Medical, Obalon Therapeutics, Spatz FGIA consulting and stockholder for Elira Therapeutics; James Swain: consultant for Obalon

CR IP T

Therapeutics; George Woodman: contracted research for Obalon Therapeutics, Allurion Technologies and BARAnova; Steven Edmundowicz: stockholder and advisory board for Elira Therapeutics, Motus GI, and EndoStim, contracted research for Medtronic; Tarek Hassanein: contracted research for Obalon Therapeutics; Vafa Shayani: contracted research for Obalon

AN US

Therapeutics and Apollo Endosurgery (Orbera); John C. Fang: consultant for Obalon

Therapeutics and Merritt Health and Wellness, chief medical officer for Veritract; Mark Noar: no disclosures; George Eid: speaker for Medtronic and consultant for Apollo Endosurgery (Orbera); Wayne J. English: contracted research for Obalon Therapeutics and Allurion

M

Technologies; Nabil Tariq: no disclosures; Michael Larsen: no disclosures; Sreenivasa S. Jonnalagadda: data safety monitoring board for Spatz FGIA; Dennis S. Riff: no disclosures;

ED

Jaime Ponce: contracted research for Obalon Therapeutics and Allurion Technologies, consultant and speaker for Gore Medical and Olympus Medical, consultant for ReShape

PT

Medical, Medtronic, and Aspire Bariatrics; Dayna Early: no disclosures; Eric Volkmann: consultant for Obalon Therapeutics; Anna R. Ibele: contracted research for Obalon

CE

Therapeutics; Matthew D. Spann: no disclosures; Kumar Krishnan: no disclosures; Juan Carlos Bucobo: no disclosures; and Aurora Pryor: speaker for Medtronic, Ethicon Endo-surgery,

AC

Stryker, Merck, and Gore, contracted research for Obalon Therapeutics and BAROnova, stockholder for TransEnterix

References

22

ACCEPTED MANUSCRIPT

7.

8. 9. 10. 11.

12.

13. 14.

15.

AC

16.

CR IP T

6.

AN US

5.

M

4.

ED

3.

PT

2.

Flegal KM, Kruszon-Moran D, Carroll MD, Fryar CD, Ogden CL. Trends in Obesity Among Adults in the United States, 2005 to 2014. JAMA. 2016;315(21):2284-2291. Chang S, Stoll CT, Song J, Varela J, Eagon CJ, Colditz GA. The effectiveness and risks of bariatric surgery: An updated systematic review and meta-analysis, 2003-2012. JAMA Surgery. 2014;149(3):275-287. Ponce J, Woodman G, Swain J, et al. The REDUCE pivotal trial: a prospective, randomized controlled pivotal trial of a dual intragastric balloon for the treatment of obesity. Surgery for Obesity and Related Diseases. 2015;11(4):874-881. Courcoulas A, Abu Dayyeh BK, Eaton L, et al. Intragastric balloon as an adjunct to lifestyle intervention: a randomized controlled trial. Int J Obes. 2017;41(3):427-433. Thompson CC, Abu Dayyeh BK, Kushner R, et al. Percutaneous Gastrostomy Device for the Treatment of Class II and Class III Obesity: Results of a Randomized Controlled Trial. Am J Gastroenterol. 2017;112(3):447-457. Mion F, Ibrahim M, Marjoux S, et al. Swallowable Obalon® gastric balloons as an aid for weight loss: a pilot feasibility study. Obesity surgery. 2013;23(5):730-733. Jensen MD, Ryan DH, Apovian CM, et al. 2013 AHA/ACC/TOS guideline for the management of overweight and obesity in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and The Obesity Society. Circulation. 2014;129(25 Suppl 2):S102-138. FDA. 21 CFR Ch. I Sec 803. In: Government Publishing Officde; 2011:42-58. Rubin D. Multiple Imputation for Nonresponse in Surveys. New York, NY: John Wiley & Sons; 1987. De Peppo F, Caccamo R, Adorisio O, et al. The Obalon swallowable intragastric balloon in pediatric and adolescent morbid obesity. Endoscopy International Open. 2017;5(1):E59-E63. Almuhanna Y, Alasfar F, Alotaibi F. Obalon Gastric Balloon: Kuwait Experience. 2016; https://www.sages.org/meetings/annual-meeting/abstracts-archive/obalon-gastric-balloonkuwait-experience/. Accessed June 30, 2018, 2018. Genco A, Maselli R, Casella G, Cipriano M, Redler A. Intragastric Balloon Treatment for Obesity. In: Agrawal S, ed. Obesity, Bariatric and Metabolic Surgery: A Practical Guide. Cham: Springer International Publishing; 2016:485-492. Alsabah S, Al Haddad E, Ekrouf S, Almulla A, Al-Subaie S, Al Kendari M. The safety and efficacy of the procedureless intragastric balloon. Surg Obes Relat Dis. 2018;14(3):311-317. Sullivan S, Swain JM, Woodman G, et al. 812d The Obalon Swallowable 6-Month Balloon System is More Effective Than Moderate Intensity Lifestyle Therapy Alone: Results From a 6- Month Randomized Sham Controlled Trial. Gastroenterology. 2016;150(4):S1267. Pryor A, Swain J, Woodman G, et al. A 6-Month Swallowable Balloon System Results In Sustainable Weight Loss At 1 Year: Results from A Prospective, Randomized Sham-Controlled Trial. Surgery for Obesity and Related Diseases. 2016;12(7):S26-S27. Sullivan S, Swain JM, Woodman G, et al. Randomized sham-controlled trial evaluating efficacy and safety of endoscopic gastric plication for primary obesity: The ESSENTIAL trial. Obesity (Silver Spring). 2017;25(2):294-301. Sullivan S, Edmundowicz SA, Thompson CC. Endoscopic Bariatric and Metabolic Therapies: New and Emerging Technologies. Gastroenterology. 2017;152(7):1791-1801. Al-Momen A, El-Mogy I. Intragastric Balloon for Obesity: A Retrospective Evaluation of Tolerance and Efficacy. Obesity surgery. 2005;15(1):101-105. Vlachou E, Direkz S, Murino A, et al. Small bowel obstruction caused by a migrated Obalon gastric bariatric balloon: nonsurgical management by antegrade double-balloon panenteroscopy. Endoscopy. 2016;48(S 01):E403-E404.

CE

1.

17. 18. 19.

23

ACCEPTED MANUSCRIPT

20. 21. 22.

AC

CE

PT

ED

M

AN US

CR IP T

23.

Benjamin SB. Small bowel obstruction and the Garren-Edwards gastric bubble: an iatrogenic bezoar. Gastrointest Endosc. 1988;34(6):463-467. Zeman RK, Benjamin SB, Cunningham MB, et al. Small bowel obstruction due to Garren gastric bubble: radiographic diagnosis. AJR Am J Roentgenol. 1988;150(3):581-582. Abu Dayyeh BK, Kumar N, Edmundowicz SA, et al. ASGE Bariatric Endoscopy Task Force systematic review and meta-analysis assessing the ASGE PIVI thresholds for adopting endoscopic bariatric therapies. Gastrointest Endosc. 2015;82(3):425-438 e425. Kumar N, Bazerbachi F, Rustagi T, et al. The Influence of the Orbera Intragastric Balloon Filling Volumes on Weight Loss, Tolerability, and Adverse Events: a Systematic Review and MetaAnalysis. Obesity surgery. 2017;27(9):2272-2278.

24

AC

CE

PT

ED

M

AN US

CR IP T

ACCEPTED MANUSCRIPT

Figure 1. Components of the swallowable gas-filled intragastric balloon system. Shown are the components of the swallowable gas-filled intragastric balloon system including the capsule containing intragastric balloon attached to catheter (Panel A), the Dispenser (Panel B), the intragastric balloon fully inflated (Panel C), and a fluoroscopic image of all three intragastric balloons inflated in a patient’s stomach (Panel D) 25

AC

CE

PT

ED

M

AN US

CR IP T

ACCEPTED MANUSCRIPT

26

AC

CE

PT

ED

M

AN US

CR IP T

ACCEPTED MANUSCRIPT

Figure 2. Consort Subject Flow Diagram for both the randomized-sham control portion of the trial and the cross over follow-up period.

27

ED

M

AN US

CR IP T

ACCEPTED MANUSCRIPT

PT

Figure 3. % Total Body Weight Loss at 24 weeks in Treatment and Control groups and week 48 in the Treatment group. Shown are the %total body weight loss in the modified

CE

Intention to Treat Analysis (Panel A), the %total body weight loss in the completer

AC

analysis (Panel B) and the %total body weight loss in the weight loss maintenance in Treatment patients with any weight loss at 24 weeks who completed 48 weeks of treatment (Panel C). I bar indicate standard errors. Data in Panel A and Panel B are listed as Least Squares-Means.

28

ACCEPTED MANUSCRIPT

Table 1. Study subject Characteristics Baseline Characteristic

mITT Analysis Treatment Control (n = 198) (n = 189)

p-value

Completer’s Analysis Treatment Control p-value (n = 174) (n = 176)

AC

CE

PT

ED

M

AN US

CR IP T

Demographic Age, mean (SD), y 42.7 (9.6) 42.5 (9.3) 0.8705 43.0 (9.5) 42.7 (9.3) Sex, women, No. (%) 171 (86.4) 170 (89.9) 0.2762 148 (85.1) 158 (89.8) Race, white, No. (%) 165 (83.3) 155 (82.0) 0.7310 144 (82.8) 145 (82.4) General Measures, mean (SD) BMI, kg/m2 35.2 (2.7) 35.5 (2.7) 0.2599 35.1 (2.7) 35.3 (2.7) Weight, kg 98.1 (13.2) 98.8 (11.9) 0.6049 98.7 (13.4) 98.6 (12.0) Height, cm 167 (8) 166 (7) 0.9159 167 (7) 167 (8) Blood pressure, mm Hg Systolic 129 (13) 128 (12) 0.2881 129 (13) 127 (12) Diastolic 82 (8) 81 (8) 0.7816 82 (8) 81 (9) Fasting laboratory values Cholesterol, mean (SD), mg/dL Total 202 (37) 201 (43) 0.8913 203 (37) 200 (43) LDL 126 (30) 125 (38) 0.7919 128 (30) 124 (38) HDL 56 (14) 57 (15) 0.4912 55 (14) 57 (15) Triglycerides, mean (SD), mg/dL 128 (74) 124 (66) 0.5318 132 (77) 124 (66) Glucose, mean (SD), mg/dL 94 (12) 95 (13) 0.8383 95 (12) 95 (12) Hemoglobin A1c, mean (SD), % 5.3 (0.4) 5.3 (0.5) 0.3530 5.3 (0.4) 5.3 (0.5) Medical conditions, No. (%) Pre-diabetes 3 (1.5) 0 (0.0) 0.2485 3 (1.7) 0 (0.0) Hypertension 31 (15.7) 28 (14.8) 0.8179 29 (16.7) 28 (15.9) Sleep Apnea 7 (3.5) 5 (2.6) 0.6137 6 (3.4) 5 (2.8) Current Smoker 11 (5.6) 13 (6.9) 0.5997 8 (4.6) 13 (7.4) BMI = Body Mass Index, HDL = high density lipoprotein, LDL = low density lipoprotein, mITT = modified intention to treat, No. = number, SD = Standard deviation

29

0.7475 0.1834 0.9269 0.4941 0.9266 0.5682 0.1830 0.5719

0.4610 0.3287 0.2694 0.2576 0.5394 0.5506 0.1218 0.8478 0.7447 0.2720

ACCEPTED MANUSCRIPT

Table 2. Changes in weight loss metrics from baseline to week 24 in the Treatment Group compared with the Control Group with the mITT Cohort (n=387)

Within-Group Difference in Least-Square Means (95% CI)

TBWL (%)

Treatment Control

6.6 ± 5.1 (5.9, 7.3) 3.4 ± 5.0 (2.7, 4.1)

EWL (%)

Treatment Control

23.9 ± 19.2 (21.3, 26.5) 12.4 ± 18.8 (9.7, 15.1) 6.6 ± 5.3 (5.9, 7.3) 3.3 ± 5.1 (2.6, 4.1)

0.0354*

11.5 (7.8, 15.3)

<0.0001

3.2 (2.2, 4.2)

<0.0001

2.3 ± 1.8 (2.1, 2.6) BMI Change Treatment 1.1 (0.8, 1.5) 2 (kg/m ) Control 1.2 ± 1.8 (1.0, 1.5) BMI = body mass index, EWL = excess weight loss, TBWL = total body weight loss

<0.0001

M

*%TBWL Difference is > 2.1%

P-value Between Groups

3.2 (2.2, 4.2)

AN US

Weight Loss Treatment (kg) Control

Between-Group Difference in LeastSquare Means (95% CI)

CR IP T

Group Treatment Weight Loss (n=198) Metric Control (n=189)

Group Treatment Weight Loss (n=174) Metric Control (n=176)

ED

Table 3. Changes in weight loss metrics from baseline to week 24 in the Treatment Group compared with the Control Group with the Completers Cohort (n=350)

Between-Group Difference in LeastSquare Means (95% CI)

P-value Between Groups

Treatment Control

7.1 ± 5.0 (6.4, 7.9) 3.6 ± 5.1 (2.9, 4.4)

3.5 (2.5, 4.6)

0.0085*

Treatment Control

26.0 ± 19.0 (23.2, 28.8) 13.0 ± 19.0 (10.3, 15.8)

13.0 (9.1, 16.9)

<0.0001

7.1 ± 5.3 (6.3, 7.9) 3.6 ± 5.1 (2.8, 4.3)

3.5 (2.5, 4.6)

<0.0001

AC

TBWL (%)

CE

PT

Within-Group Difference in Least-Square Means (95% CI)

EWL (%)

Weight Loss Treatment (kg) Control

2.5 ± 1.8 (2.2, 2.8) BMI Change Treatment 1.2 (0.9, 1.6) 2 (kg/m ) Control 1.3 ± 1.8 (1.0, 1.5) BMI = body mass index, EWL = excess weight loss, TBWL = total body weight loss *%TBWL Difference is > 2.1% 30

<0.0001

ACCEPTED MANUSCRIPT

Table 4. Comparison of Percent Total Body Weight loss in the Modified Intention to Treat analysis: BMI 30.034.9 kg/m2 and 35-40.0 kg/m2. Treatment n

Control

LS Mean (95% CI)

n

30.0-34.9 97 6.8% (5.8%, 7.9%) 83 35.0-40.0 101 6.4% (5.3%, 7.4%) 106 BMI = body mass index, LS = least squares

LS Mean (95% CI) 3.6% (2.4%, 4.7%) 3.3% (2.3%, 4.3%)

Treatment - Control Interaction: Device & BMI pEstimate (95% CI) p-value value 3.3% (1.8%, 4.8%) <0.001 0.8375 3.1% (1.7%, 4.5%) <0.001

CR IP T

BMI (kg/m2)

Table 5. Comparison of Percent Total Body Weight loss in the Completer analysis: BMI 30.0-34.9 kg/m2 and 35-40.0 kg/m2 BMI (kg/m2)

Treatment LS Mean (95% CI)

n

LS Mean (95% CI)

30.0-34.9 84 7.6% (6.4%, 8.7%) 80 35.0-40.0 90 6.4% (5.6%, 7.8%) 96 BMI = body mass index, LS = least squares

3.7% (2.6%, 4.9%) 3.5% (2.5%, 4.6%)

Treatment - Control Interaction: Device & BMI pEstimate (95% CI) p-value value 3.8% (2.3%, 5.4%) <0.001 0.5510 3.2% (1.7%, 4.7%) <0.001

AN US

n

Control

M

Table 6. Changes in cardiometabolic outcomes from baseline to week 24 in the Treatment Group compared with the Control Group

-3 (-5 to -1)

0.0138

ED

Mean (95% CI)

P-value for Within Group

PT

Outcome Variable

Changes from Baseline to Week 24 Mean (95% CI)

AC

CE

Systolic Blood Pressure, mm Hg Treatment 124 (122Baseline 126) Week 121 (11924 123) Control 125 (123Baseline 127) Week 126 (12424 128) Diastolic Blood Pressure, mm Hg Treatment Baseline 78 (76-79) Week 78 (77-79) 24

1 (-1 to 3)

0.2897

1 (-1 to 2)

0.5272 31

BetweenGroup Difference in Least-Square Means1 (95% CI)

P-value Between Group

-4 (-7 to -2)

0.0020

-1 (-3 to 1)

0.3360

ACCEPTED MANUSCRIPT

Outcome Variable

Mean (95% CI)

Changes from Baseline to Week 24 Mean (95% CI)

P-value for Within Group

BetweenGroup Difference in Least-Square Means1 (95% CI)

P-value Between Group

-7 (-13 to -1)

0.0214

-5 (-10 to 1)

0.0775

1 (0 to 3)

0.0802

-15 (-25 to -5)

0.0049

AC

CE

PT

ED

M

AN US

Baseline 77 (76-79) 2 (0 to 3) 0.0383 Week 79 (78-80) 24 Total Cholesterol, mg/dL Treatment 203 (197Baseline 208) -9 (-13 to -5) <0.0001 Week 193 (18824 199) Control 201 (194Baseline 207) -1 (-7 to 4) 0.6587 Week 199 (19324 205) Low Density Lipoprotein, mg/dL Treatment 127 (123Baseline 132) -6 (-10 to 0.0008 3) Week 121 (11624 126) Control 124 (119Baseline 130) 0.9654 0 (-5 to 5) Week 124 (11924 129) High Density Lipoprotein, mg/dL Treatment Baseline 55 (53-57) 3 (1 to 4) 0.0001 Week 58 (56-60) 24 Control Baseline 57 (55-59) 1 (-1 to 2) 0.3301 Week 58 (56-60) 24 Triglycerides, mg/dL Treatment 132 (121Baseline 143) -15 (-23 to 0.0006 7) Week 117 (10824 125)

CR IP T

Control

32

ACCEPTED MANUSCRIPT

Outcome Variable

Mean (95% CI)

Changes from Baseline to Week 24 Mean (95% CI)

P-value for Within Group

123 (114133) 126 (114138)

3 (-5 to 10)

0.4942

-5 (-7 to -2)

<0.0001

BetweenGroup Difference in Least-Square Means1 (95% CI)

P-value Between Group

Baseline Week 24

95 (93-97) 90 (88-92) 95 (93-96) 0 (-2 to 2)

95 (92-97)

5.3 (5.35.4) 5.4 (5.35.4)

0.1 (0.0 to 0.1)

ED

Control 5.3 (5.35.4) Week 5.4 (5.324 5.5) Alanine Transaminase, IU/L Treatment Baseline 25 (23-27) Week 20 (18-21) 24 Control Baseline 23 (21-25) Week 20 (18-22) 24

0.0008

0.0 (-0.1 to 0.1)

0.7787

1 (-3 to 1)

0.4373

0.8399

0.0009

0.1 (0.0 to 0.1)

0.0056

-5 (-6 to -3)

<0.0001

AC

CE

PT

Baseline

-5 (-7 to -2)

AN US

Week 24 Glucose, mg/dL Treatment Baseline Week 24 Control Baseline Week 24 Hemoglobin A1C Treatment

M

Baseline

CR IP T

Control

-4 (-5 to -3)

0.0024

33

ACCEPTED MANUSCRIPT

Table 7. Adverse events in all subjects treated with the SGIBS

Total adverse events Serious adverse events Non-serious adverse events Device related adverse events Device related gastric mucosal or esophageal damage events Removal procedure related adverse events Specification of individual adverse events Gastrointestinal Abdominal Pain Nausea Vomiting Dyspepsia Abdominal Distension

Patients (%) N = 336 306 (91.1) 1(0.3) 305 (90.8) 301 (89.6)

Events

33 (9.8)

35

40.0

51.4

8.6

45 (13.4)

53

75.5

22.6

1.9

16.9 16.0 16.1 21.1 30.4 9.3

0.5 0.2 0.0 0.0 0.0 0.0

0.0 3.3 48.0 5.9

0.0 0.0 4.0 5.9

1213 1 1212 1160

Severity: % of Events Mild Moderate Severe 82.2 17.3 0.5 0.0 0.0 100.0 82.3 17.3 0.4 82.5 17.1 0.4

CR IP T

Event

1191 494 311 71 69 54

31 (9.2) 28 (8.3) 24 (7.1) 17 (5.1)

37 30 25 17

82.6 83.8 83.9 78.9 69.6 90.7 100. 0 96.7 48.0 88.2

14 (4.2)

14

78.6

21.4

0.0

12 (3.6)

12

25.0

0.0

9 (2.7) 9 (2.7)

10 9

0.0 33.3

0.0 0.0

8 (2.4) 6 (1.8)

8 6

0.0 66.7

0.0 0.0

3 (0.9)

4

0.0

0.0

Chest Pain Gastric Ulcer

3 (0.9) 3 (0.9)

3 3

0.0 66.7

0.0 33.3

Sore Throat Device Intolerance

2 (0.6) 2 (0.6)

3 2

0.0 50.0

0.0 50.0

Hiccups Dry Heaving Food Passage Difficulty

1 (0.3) 1 (0.3) 1 (0.3)

2 1 1

75.0 100. 0 66.7 100. 0 33.3 100. 0 100. 0 0.0 100. 0 0.0 100. 0 100.0 100.0

0.0 0.0 0.0

0.0 0.0 0.0

CE

Excessive Gas Esophagitis

PT

Constipation Difficulty in Sleeping

AC

Hypersalivation

M

ED

Eructation Diarrhea Gastric Irritation Gastric Bleeding/Abrasion Esophageal Bleeding/Abrasion Esophagogastric Bleeding/Abrasion

AN US

305 (90.8) 244 (72.6) 188 (56.0) 58 (17.3) 57 (17.0) 49 (14.6)

34

ACCEPTED MANUSCRIPT

1

100.0

0.0

0.0

1 (0.3) 1 (0.3) 1 (0.3) 1 (0.3) 10 (3.0) 4 (1.2) 2 (0.6) 2 (0.6) 1 (0.3) 1 (0.3) 8 (2.4) 6 (1.8) 1 (0.3) 1 (0.3) 1 (0.3) 1 (0.3) 2 (0.6) 1 (0.3) 1 (0.3)

1 1 1 1 10 4 2 2 1 1 9 7 1 1 1 1 2 1 1

0.0 100.0 100.0 0.0 50.0 50.0 100.0 0.0 0.0 100.0 66.7 71.4 100.0 0.0 0.0 0.0 100.0 100.0 100.0

0.0 0.0 0.0 100.0 50.0 50.0 0.0 100.0 100.0 0.0 33.3 28.6 0.0 100.0 100.0 100.0 0.0 0.0 0.0

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

CR IP T

1 (0.3)

AC

CE

PT

ED

M

AN US

Fullness Peptic Ulcer Disease (Bleeding) Retaining Food & Fluid Shoulder Pain Syncope Respiratory Oxygen Desaturation Shortness of Breath Vocal Chord Spasm Asthma Coughing Metabolic/Nutritional Headache/Migraine Dizziness Fatigue Cardiovascular Hypertension Other Allergic Reaction Swollen Lips

35