Abstract
Introduction
Obesity is associated with reduced life expectancy and various comorbidities. Surgical interventions are effective but accompanied by the risk of serious complications. Less invasive endoscopic procedures mainly comprise the intragastric balloon (IB) and the duodenal-jejunal bypass liner (DJBL). A randomized, sham-controlled study comparing both procedures has not been undertaken so far.
Methods
We performed a randomized, patient- and assessor-blinded, controlled trial comparing weight loss in IB versus DJBL versus a sham procedure (2:2:1 ratio). Patients with a BMI >35 kg/m2 or >30 with obesity-related comorbidities were included. The IB was removed after 6 months and the DJBL after 12 months. The main objective was successful weight loss (>10% from baseline) 12 months after explantation of the devices. Secondary outcomes were changes in comorbidities, quality of life, and complications.
Results
Thirty-three patients were randomized. Recruitment has to be stopped suddenly in after the DJBL device lost its CE mark in Europe. In all, 11 patients received DJBL, 15 IB, and 7 were allocated to the sham group. Blinding was feasible in all patients. Weight decreased from baseline until explantation (DJBL: 129.4 ± 28.3 kg to 107.4 ± 16.7 kg; IB: 118.3 ± 22.8 kg to 107.4 ± 25.7 kg; sham: 134.6 ± 18.0 kg to 131.2 ± 14.3 kg), but patients regained weight almost to the baseline level 12 months after explantation. Only 1 patient in IB group reached the primary endpoint. Severe device-related complications were very rare.
Conclusion
Endoscopic bariatric procedures failed to achieve effective weight loss 12 months after explantation of the devices. The results of this trial need to be interpreted with caution due to its early termination.
Keywords: Obesity, Endoscopy, Diabetes, Weight loss, Intragastric balloon, Bypass liner
Introduction
The prevalence of obesity has been rapidly increasing since 1980 and more than 700 million people worldwide currently have obesity [1]. Calculations from representative population samples estimate that the number of people with obesity will reach 1.12 billion in 2030 [2]. Many studies have shown that a substantial loss of life expectancy and disease-free life years are associated with obesity-related comorbidities [3, 4], implying that obesity is not only a health problem but also a socioeconomic burden. Thus, effective approaches for the treatment of obesity are urgently needed [5].
However, managing obesity is often challenging as lifestyle alterations result in a mean total weight loss of only about 5 kg (5.8% of baseline weight) even when using commercial weight loss programs [6, 7]. Although even a moderate weight loss of 5–10% of body weight leads to a meaningful improvement in insulin resistance, blood pressure, and dyslipidemia [8, 9], only 20% of obese people are willing to accept treatment [3] and only two-thirds complete their respective programs even in the study setting [7]. More effective interventions comprise bariatric surgical and endoscopic treatments. Both sleeve gastrectomy and gastric bypass lead to sustainable weight loss and improvement of obesity-related comorbidities [10, 11], but these procedures are irreversible and harbor the risk of serious perioperative complications, malabsorption, and necessity of dietary supplementation. Furthermore, in more than 10% of patients who underwent bariatric surgery, repeated interventions are required [12, 13].
In contrast, bariatric endoscopic procedures are reversible, minimally invasive, less costly, and may offer a potentially lower risk approach compared to bariatric surgery [14]. These endoscopic approaches limit oral food intake and gastric exclusion, increase gastric emptying time, or evoke malabsorption by partially inhibiting the breakdown or absorption of nutrients [15]. Several devices have been designed, but intragastric balloons (IB) and the duodenal-jejunal bypass liner (DJBL) are among the most widely used and thoroughly studied devices. Different types of IB have been developed that are mainly endoscopically implanted, filled with fluids, and are usually removed after 6 months [14]. The DJBL is an impermeable, fluoropolymer tube reversibly anchored to the duodenal bulb. The device results in an effective prevention of digestion and absorption in the upper intestine and also has restrictive effects [16].
In spite of convincing weight loss in patients undergoing endoscopic bariatric procedures, there are also studies showing good results following a sham procedure [17, 18]. Moreover, randomized studies comparing IB and DJBL are lacking. Thus, we performed a multicenter, randomized, controlled trial to compare weight loss in endoscopically implanted DJBL versus IB versus sham procedures.
Methods
Patients
Patients were prospectively enrolled at the participating centers (University of Leipzig Medical Center, Sachsenhausen Clinic in Frankfurt and Bogenhausen Clinic in Munich) between April and November 2017. Inclusion criteria were 18 years of age or older with a body mass index (BMI) ≥35 kg/m2 or ≥30 kg/m2 and obesity-related comorbidities. Obesity-related comorbidities comprise arterial hypertension, diabetes, and associated diseases, cardiovascular diseases, dyslipidemia, arthritis, obstructive sleep apnea, and endocrinologic, psychiatric, and gastrointestinal diseases. Prior to enrollment, all participants attempted a conservative weight-loss therapy that was ineffective. Furthermore, all patients had to have a medical indication for the long-term use of proton pump inhibitors (PPIs) due to the fact that PPIs are mandatory for the implantation of DJBLs and could affect the primary endpoint differently from placebo. Additionally, a gastroduodenoscopy had to be available in all participants within 3 months before recruitment to exclude contraindications for participation. Exclusion criteria were malignant diseases, peptic ulcer, type 1 diabetes, large hiatal hernia, previous gastric surgery or bariatric endoscopic or surgical therapy, high risk of gastrointestinal bleeding, symptomatic gallstones, contraindication for general anesthesia, PPIs or the medical devices used, uncontrolled gastroesophageal reflux disease, and pregnancy or inability to provide informed consent.
Randomization and Treatment Scheme
This study was conducted in a patient- and assessor-blinded manner. Therefore, an unblinded study team (implanters) kept personal lists with implanted devices, and a blinded team attended the study visits. After patients’ written informed consent, the patients were randomly assigned (1:2:2) using block randomization either to the sham group (endoscopy in sedation) or to one of the interventional groups (DJBL or IB), stratified with respect to diabetes status (yes/no) and center. The randomization was performed electronically and reported to the PI center by fax within 24 h.
Sham endoscopy and IB were performed under sedation, whereas DJBL was implanted under general anesthesia. To ensure blinding, all patients consented to full anesthesia. All patients underwent regular dietary consultations, clinical examinations, blood samples for analysis of comorbidities (e.g., fasting blood glucose, HbA1c, lipids, inflammatory markers), and bioelectric impedance analysis (BIA). Hepatic abscesses as a potential complication of DJBL were monitored by periodic abdominal sonography (US). In addition, study patients were evaluated for concomitant medication, obesity-related comorbidities, quality of life (SF36 quality of life questionnaire), and adverse events (AEs) or side effects. Study participation ended 24 months after implantation or baseline endoscopy. The randomization process and a study flowchart can be found in Figure 1 of the published study protocol [19].
Fig. 1.
The progress of patients in the WET trial is depicted. Two patients were randomized to the liner group on November 9 and 10, 2017. With the retraction of the CE mark for the EndoBarrier® (GI Dynamics, Boston, USA) on November 14, it was no longer legally permitted to implant the device. To provide these 2 patients’ medical treatment and allow them to remain in the trial, a balloon was implanted. For the purposes of the analysis of this trial (and in the flowchart), they are treated as balloon patients. WET, Weight-loss Endoscopy Trial.
Medical Devices
In this study, the EndoBarrier® system (endoscopically implanted DJBL, GI Dynamics, Dusseldorf, Germany, CE010311) and the Orbera Intragastric Balloon™ (Apollo Endosurgery, San Diego, USA, CE27493) were used. The DJBL was implanted over a period of 12 months and the IB for a period of 6 months as per manufacturer’s recommendation. All implanters were experienced endoscopists who had performed at least 10 IBs and 10 DJBLs prior to study participation.
Study Objectives
The primary objective of this study was to compare the success rates in weight loss (defined as ≥10% reduction from baseline) between the three procedures 1 year after removal. Secondary aims were the success rates as well as percentage of weight loss at 6 months, 12 months, and upon removal of the devices. We also assessed complication rates and changes in obesity-related comorbidities and quality of life. In addition, the feasibility of the blinding process was evaluated.
Adverse Events
Classification of AEs and serious AEs was defined in DIN EN ISO 14155:11. In addition, adverse device events (ADEs), serious ADEs, and unanticipated serious ADEs were determined according to Medical Devices Safety Regulation §2. In the case of SAE, SADE, or USADE, the unblinded team decided if unblinding was deemed necessary.
Statistics
The baseline characteristics of the patients were summarized as mean, standard deviation, or frequency, depending on the data. Successful weight loss 12 months after explantation was defined to be W12month/WBaseline ≤0.9. Primary and secondary study objectives were analyzed by using a generalized linear mixed model using a profiling method for confidence intervals (CIs), which can take into account the longitudinal structure of the data as well as missing data. The stratification attribute diabetes was included as a covariate in the model, and a closed-testing procedure was used. As a sensitivity analysis, the 2 × 3 contingency table was analyzed with a χ2 test. Tests are all two-sided, and the significance level is set at 5%. No interim analyses were performed. Additional information on the methods and statistical models can be found in the published study protocol of this trial [19]. Analyses were carried out with R version 4.1.1 and the package “lme4” for the mixed model [20].
Different definitions of success have been recommended, including excess weight loss [21]. However, EWL strongly depends on baseline BMI and thereby has considerable inaccuracies [22]. We estimated proportions of weight loss and the devices’ failures rates based on the previously published work [17, 23–28]. Assuming weight loss of 10.4% (DJBL), 6% (IB), and 3.9% (sham) for completers and accounting for expected failure rates of 15% (DJBL) and 10% (IB) and a dropout rate of 10%, we expected successful weight reduction in 40% (DJBL), 16% (IB), and 9% (sham) of patients. Our simulations showed that 150 patients were necessary for a power of 93% for the global test at a 5% significance level.
Results
Trial Recruitment and Patient Characteristics
Out of 181 screened individuals, we included 33 patients from three sites to the trial. Recruitment and randomization had to be stopped suddenly in November 2017 after the DJBL device (EndoBarrier®) lost its CE mark in Europe. Two patients were randomized to DJBL on the day of the announcement. To provide these 2 patients medical treatment and allow them to remain in the trial, a balloon was implanted. For the purposes of the analysis of this trial, they are treated as balloon patients.
The DJBL devices that were implanted before the withdrawal of the CE mark did not have to be removed, and patients could remain in the trial. The lead ethics committee, competent national authority, and Data Safety and Monitoring Board (DSMB) were all in favor of continuing with the original trial design in a modified legal setting. After 1 year of negotiating with GI Dynamics and despite their initial assurance that they would assist us, we were finally forced to close the trial due to lack of support from the manufacturer.
All of the 33 participants that were allocated to the trial are considered in the intention-to-treat analysis of the primary endpoint (Fig. 1). At the final follow-up visit, data from 21 of them were available with most of the losses coming from the balloon group due to retraction of consent and loss of contact. Patient characteristics at the baseline can be found in Table 1. There were 9 patients (27%) with diabetes mellitus type 2, but only four of them (12%) had HbA1c values greater than or equal to 6.5%.
Table 1.
Baseline characteristics
| All patients (n = 33) | Liner (n = 11) | Balloon (n = 15) | Sham (n = 7) | |
|---|---|---|---|---|
| Number of females, n (%) | 17 (52) | 6 (55) | 7 (47) | 4 (57) |
| Age, years | 43.4±12.6 | 48.1±10.9 | 42.2±12.3 | 38.7±15.0 |
| BMI, kg/m2 | 41.5±6.1 | 43.0±7.6 | 39.3±4.2 | 44.1±6.2 |
| Smoking, n (%) | ||||
| Current smoker | 11 (33) | 3 (27) | 4 (27) | 4 (57) |
| Ex-smoker | 11 (33) | 3 (27) | 5 (33) | 3 (43) |
| Nonsmoker | 11 (33) | 5 (45) | 6 (40) | 0 (0) |
| Diabetes (%) | 9 (27) | 3 (27) | 4 (27) | 2 (29) |
| HbA1c | 5.8±1.1 | 5.9±0.6 | 5.8±1.6 | 5.8±0.9 |
| ≥6.5 | 4 (12) | 1 (9) | 1 (7) | 2 (29) |
| Retinopathy | 1 (3) | 0 (0) | 0 (0) | 1 (14) |
| Hypertension | 17 (52) | 6 (55) | 7 (47) | 4 (57) |
| Dyslipidemia | 26 (79) | 11 (100) | 9 (60) | 6 (86) |
| Hyperlipidemia | 23 (70) | 10 (91) | 7 (47) | 6 (86) |
| Varicose veins (legs) | 8 (24) | 5 (45) | 2 (13) | 1 (14) |
| Hypothyreosis | 4 (12) | 0 (0) | 3 (20) | 1 (14) |
| Fatty liver | 26 (79) | 9 (82) | 12 (80) | 5 (71) |
| Snoring | 27 (82) | 10 (91) | 11 (73) | 6 (86) |
| Sleep apnea | 7 (21) | 4 (36) | 1 (7) | 2 (29) |
| GERD | 12 (36) | 7 (64) | 4 (27) | 1 (14) |
| Asthma | 2 (6) | 1 (9) | 0 (0) | 1 (14) |
| Osteoarthritis | 8 (24) | 3 (27) | 3 (20) | 2 (29) |
| Chronic kidney disease | 7 (21) | 4 (36) | 2 (13) | 1 (14) |
| Depression | 6 (18) | 1 (9) | 3 (20) | 2 (29) |
| SF-36, n (%) | ||||
| Vitality | 45 [39, 51] | 50 [45, 55] | 40 [35, 44] | 45 [38, 60] |
| Physical functioning | 70 [53, 80] | 70 [49, 85] | 62 [48, 78] | 70 [70, 78] |
| Bodily pain | 73 [51, 88] | 62 [46, 79] | 74 [62, 96] | 62 [42, 87] |
| General health perceptions | 54 [42, 63] | 62 [44, 70] | 51 [39, 57] | 62 [44, 67] |
| Physical role functioning | 75 [50, 100] | 75 [50, 94] | 50 [50, 100] | 100 [88, 100] |
| Emotional role functioning | 100 [50, 100] | 100 [42, 100] | 83 [67, 100] | 100 [67, 100] |
| Social role functioning | 81 [62, 100] | 88 [75, 100] | 75 [53, 97] | 100 [69, 100] |
| Mental health | 68 [55, 80] | 72 [58, 80] | 66 [46, 78] | 72 [62, 80] |
| IPAQ activity category, n (%) | ||||
| Low | 2 (6) | 0 (0) | 1 (7) | 1 (14) |
| Moderate | 16 (48) | 5 (45) | 7 (47) | 4 (57) |
| High | 8 (24) | 3 (27) | 4 (27) | 1 (14) |
| Invalid/missing | 7 (21) | 3 (27) | 3 (20) | 1 (14) |
| BIA – fat-free mass (%) | ||||
| Manufacturer’s algorithm | 42.8±10.3 | 42.4±12.0 | 40.7±7.9 | 48.0±12.4 |
| According to Jimenez | 41.5±7.0 | 40.3±7.9 | 41.1±6.1 | 43.9±7.4 |
BIA, bioelectrical impedance analysis.
Success of Blinding
The patients and the treating physicians other than the endoscopists were blinded, but uncertainties about how effective blinding could be led us to include a question for both patients and physicians at each visit as to which group they believed was allocated. The majority of patients in the sham group and their physicians believed that they were in an intervention group in the first few weeks after the initial gastroscopy (online suppl. Fig. 1; for all online suppl. material, see https://doi.org/10.1159/000539816). While suspicion grew over time regarding allocation to the sham group, only 2/5 (40%) of these patients were guessed correctly by the physician at 24 months and 4/5 (80%) by the patients themselves.
Weight Loss
Percentage weight loss over time is shown by the randomization group in Figure 2a. The intervention groups had mean weight loss amounting to 10–13% of the initial weight during intervention and had success rates of about 50%, whereas the control group, despite mean weight loss of about 5%, had a success rate of 0%. In all three arms, there was substantial weight regain after removal (intervention) or after 6 months (control). At 1 year after removal, only a single patient (IB group) reached the primary endpoint of 10% weight loss (Fig. 2b). The percentage of patients with success is thus 0% (95% CI: 0–26%, DJBL), 7% (95% CI: 1–30%, IB), and 0% (95% CI: 0–35%, sham) with no indication for a difference between the groups. Percentage weight loss and success at different points in time can be found in Table 2 along with estimates for the difference between groups.
Fig. 2.
a Mean percentage weight loss over time is shown by the randomization group from a mixed model using imputed data. Times were shifted slightly according to the group to avoid overlapping. b Proportion of patients with successful weight loss (>10%), where whiskers represent 95% CI.
Table 2.
Weight loss “successful weight loss” (>30% weight loss) by randomization arm over time
| Liner | Balloon | Sham | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| weight loss (%) | proportion with success (%) | weight loss (%) | 95% CI for difference to liner in weight loss | proportion with success (%) | 95% CI for difference to liner in success | weight loss (%) | 95% CI for difference to liner in weight loss | proportion with success (%) | 95% CI for difference to liner in success | |
| 6 months | 10.2 | 55 | 9.8 | −4.3 to 5.6 | 47 | −30 to 43 | 5.2 | −1.2 to 10.8 | 0 | 11 to 79 |
| At removal | 11.8 | 55 | 9.8 | −2.6 to 7.7 | 47 | −30 to 43 | 5.2 | 0.1 to 12.7 | 0 | 11 to 79 |
| 12 months | 11.8 | 55 | 6.7 | 0.5 to 10.6 | 20 | −3 to 65 | 2.9 | 2.9 to 14.6 | 0 | 11 to 79 |
| 1 year after removal | 4.4 | 0 | 2.9 | −3.1 to 6.0 | 7 | −30 to 21 | 1.8 | −3.0 to 7.9 | 0 | −37 to 27 |
Bold lettering indicates that the 95% CI does not contain the value zero.
Secondary Endpoints
Nine patients had type 2 diabetes at baseline (3 DJBL, 4 IB, and 2 sham). At device removal, 2 DJBL, 2 IB, and 1 sham patients were known to have improved regarding diabetes. Improvement was defined as a decrease of HbA1c of at least 10% or reduction of diabetes medication. One year later, there was known improvement in 1 DJBL, 1 IB, and 2 sham patients and worsening in 1 DJBL patient relative to baseline. Hypertension, dyslipidemia, nonalcoholic fatty liver disease, snoring, and gastroesophageal reflux were the other most common comorbidities, most of which showed improvement in all three groups (see online suppl. Fig. 2).
Table 3 shows comparisons between the groups at different points in time for HbA1c, selected items from quality of life, physical activity, and bioelectric impedance analysis. There were only 4 patients with HbA1c ≥6.5% at baseline (1 DJBL, 1 IB, 2 sham) so that a meaningful comparative analysis of these patients is not possible. The raw HbA1c data in these patients can be found in online supplementary Figure 3 and the remaining subscales from SF-36 in online supplementary Table 1.
Table 3.
Selected secondary endpoints
| Liner | Balloon | Sham | |||
|---|---|---|---|---|---|
| estimate | estimate | difference to liner (95% CI) | estimate | difference to liner (95% CI) | |
| HbA1c, % | Negative differences indicate superiority | ||||
| Baseline | 7.03 | 6.35 | 6.62 | ||
| 6 months | 6.57 | 6.26 | 0.37 (0.06 to 0.68) | 6.54 | 0.38 (−0.00 to 0.77) |
| 12 months | 6.62 | 6.29 | 0.36 (−0.01 to 0.72) | 6.56 | 0.35 (−0.06 to 0.77) |
| 24 months | 6.81 | 6.46 | 0.33 (−0.05 to 0.72) | 6.76 | 0.36 (−0.06 to 0.78) |
| Physical functioning (SF-36) | Positive differences indicate superiority | ||||
| Baseline | 58 | 55 | 60 | ||
| 1 month | 64 | 74 | 13 (2 to 23) | 71 | 5 (−7 to 17) |
| 3 months | 65 | 71 | 9 (−2 to 19) | 72 | 5 (−6 to 17) |
| 6 months | 69 | 72 | 5 (−6 to 16) | 73 | 2 (−11 to 15) |
| 12 months | 72 | 68 | −1 (−13 to 10) | 69 | −5 (−18 to 8) |
| 18 months | 68 | 74 | 8 (−4 to 20) | 73 | 2 (−11 to 16) |
| 24 months | 66 | 69 | 6 (−6 to 18) | 84 | 16 (2 to 29) |
| Physical activity (IPAQ), MET-h/day | Positive differences indicate superiority | ||||
| Baseline | 6.1 | 5.8 | 5.3 | ||
| 6 months | 7.8 | 8.6 | 1.1 (−3.6 to 5.8) | 5.4 | −1.7 (−4.4 to 4.2) |
| 12 months | 9.5 | 7.3 | −1.9 (−7.3 to 3.3) | 9.0 | 0.2 (−6.1 to 6.5) |
| 24 months | 7.8 | 4.9 | −2.6 (−8.3 to 3.1) | 5.7 | −1.4 (−7.8 to 5.0) |
| Percentage fat according to BIA | Negative differences indicate superiority | ||||
| Baseline | 36.3 | 35.1 | 40.1 | ||
| 6 months | 34.1 | 30.4 | −2.5 (−5.8 to 0.7) | 37.0 | −1.0 (−4.7 to 2.8) |
| 12 months | 32.6 | 31.7 | 0.3 (−3.3 to 3.7) | 39.2 | 2.7 (−1.1 to 6.6) |
| 24 months | 36.1 | 35.0 | 0.1 (−3.6 to 3.8) | 40.4 | 0.4 (−3.7 to 4.5) |
Bold lettering indicates that the 95% CI does not contain the value zero.
BIA, bioelectrical impedance analysis; HbA1c, glycated hemoglobin; IPAQ, International Physical Activity Questionnaire; MET, metabolic equivalent of task; SF-36, Short Form (36) Health Survey.
Harms and Complications
There were no technical complications associated with the implantation of either IB or DJBL. At explantation, 2 patients had technical complications, all in the liner group, though all could be explanted at the foreseen time at 1 year. One patient had a dislocated device with the anchoring system entering the pylorus. The dislocation of the implanted liner was noticed during the regular study endoscopy conducted at 6 months. However, the patients did not report any complaints that were related to this observation. In addition, an adequate endoscopic passage through the liner was possible and the unblinded study team decided not to remove the liner prematurely. The other patient had bleeding and repeated aspiration after strong mucous production during the explantation procedure.
In total, 287 AEs were reported in 31 of the 33 patients. The mean number of AEs per patient is 9.2, 8.8, and 7.7 for the DJBL, IB, and sham groups, respectively. There was one death due to heart attack in the DJBL group and one bleeding upon explantation of the DJBL, classified as a severe AE, but that could be stopped endoscopically. In the IB group a single patient suffered nausea, vomiting, and abdominal pain classified as a severe AE. No AE was classified as severe in the sham group. Gastrointestinal disorders comprised the most common type of AE in every group with a mean of 3.4, 4.4, and 1.7 AE per person among 10 (91%), 12 (80%), and 3 (43%) patients in the DJBL, IB, and sham groups, respectively. More details on AE severity and type can be found in online supplementary Tables 2, 3, and 4.
Discussion
In our trial, patients with obesity were randomized to one of two endoscopic interventions (DJBL or IB) or to a sham group. We observed good weight loss in both intervention groups while on treatment and a weaker but notable one in the sham group. However, weight regain after explantation was rapid and of a similar rate for all three arms. Improvements in comorbidities such as diabetes and hypertension were seen all three arms. More gastrointestinal disorders were documented in the intervention arms compared to the sham arm.
Regarding weight loss at the end of treatment, a meta-analysis of four open-label randomized controlled trials (RCTs) comparing IB to conservative therapy found total weight loss of 9.3% in the IB group compared to 5.5% in controls at 6 months [29]. This agrees very well with our estimates of 9.8% for IB and 5.2% for our sham group. Four further RCTs not included in the meta-analysis for lack of specific data generally found BMI reductions that were numerically larger than a sham group and roughly comparable to those in our study, considering different lengths of treatment and populations [16, 17, 23, 24]. A recent study used network meta-analysis of RCTs to estimate the mean difference in percentage weight loss between IB and DJBL [30] and found them to be almost equivalent at explantation, also agreeing well with our data. Note, however, that the network meta-analysis of this endpoint was only able to include 5 IB and 2 DJBL studies, the latter of which had 3 and 6 months of treatment. A large RCT (n = 170 randomized) comparing DJBL with intensive medical therapy to intensive medical therapy alone found maximal weight reductions of 11.7 kg (DJBL) and 6.2 kg (control) in raw data [31], which is very similar to our DJBL and sham groups.
Not many data are available from RCTs for weight at least 1 year after explantation. One meta-analysis on endoscopic bariatric therapies included percentage weight loss and estimates it at 6.0% at 36 months for IB [21]. However, data originated from only 2 studies, both of which had serious methodological flaws [25, 32]. One study with a 4-year follow-up after beginning DJBL treatment found 2.2% weight loss (compared to 9.1% after treatment) and was likely biased since only 15 of 29 patients could be analyzed, 5 of whom provided data before bariatric surgery and not at 4 years [33]. A study with 16 diabetics found that weight 6 months after explantation was 2.4 kg below the baseline value, but 3.5 kg above the value upon explantation [34]. The RCT mentioned above that included intensive medical therapy found about 5 kg weight reduction in completers from both arms 1 year after treatment end [31]. It is important to bear in mind that even very small changes in weight or in blood parameters are associated with large differences in risk of obesity-related comorbidities [35]. In our data, we see a return to baseline weight 1 year after treatment end, but use methods to account for missing data, which is not true in most studies.
A meta-analysis on the impact of IB on obesity-related comorbidities included 10 RCTs and 30 observational trials and concluded that IB is “more effective than diet, [but that] the strength of the evidence is limited” [36]. This meta-analysis concluded that the mean difference between IB and control was a further reduction in fasting glucose by 12.7 mg/dL that the odds ratio for diabetes resolution was 1.4. Because of the indication for use of DJBL and its intended purposes, research has focused on its impact on obesity-related comorbidities in patients with type 2 diabetes, with particular attention on glycemic control. One meta-analysis included 5 RCTs and 9 observational trials and found a mean difference in HbA1c of 0.9% between DJBL and control groups [37]. A second meta-analysis based on 5 RCTs and 10 observational trials found a nonsignificant mean difference in fasting glucose of 67 mg/dL (−14–148) [38]. Our own data indicate improvements in obesity-related comorbidities in all three arms, some of which are immediately relevant such as snoring and GERD, though the small sample size and low prevalence of diabetes make comparisons with the literature difficult. The primary mechanism responsible for these improvements is likely weight loss. For example, those with improved hypertension at 6 months had lost 2 kg more weight than those without, which is consistent with the aforementioned small changes in weight being associated with large differences in comorbidities. However, other mechanisms are certainly plausible considering that diabetes, for example, can improve within days of bariatric surgery [39].
The safety of both IB and DJBL was analyzed in a systematic review including 82 articles on IB and 11 for DJBS. The most common AE in both devices was pain (34, 59% for IB, DJBL) followed by nausea/vomiting (29, 39%), whereas GERD was only seen in IB (18%). Early removal was necessary in 7.5% of IB and 4.2% of DJBL patients. Serious harms were rare, with migration (1.4%) and gastric perforation (0.1%) in IB and migration (4.9%), GI bleeding (3.9%), sleeve obstruction (3.4%), liver abscess (0.13%), cholangitis (0.13%), acute cholecystitis (0.13%), and esophageal perforation (0.13%) in DJBL [21]. However, a large multicenter study was terminated due to a high incidence of liver abscesses (3.5%) and an 11.7% rate of AEs requiring removal of the DJBL (NCT01728116). Our own data showed very high prevalence of gastrointestinal AEs in 91% of DJBL and 80% of balloon patients, though they were also reported in 43% of the sham patients. Since conducting this trial, evidence has accumulated, suggesting that the gastric balloon may be used safely for 12 months [40] (though caution regarding conflicts of interest may be necessary), but that harms in the case of the liner are more common after 6 months [41]. In our trial, serious AEs were only observed in 3 patients and the fatal heart attack is very unlikely to be related to the liner.
The greatest limitation in our trial is certainly the lack of power due to early termination because of the liner losing its CE mark. The CE mark is a quality standard and legal prerequisite to be allowed onto the European market and was rescinded because of production quality issues according to the manufacturer. Moreover, the blinding may not have been as successful as in trials in which the balloon was simply removed from the application cartridge and this could have impacted on the weight loss in our sham group. Finally, due to misunderstandings of the protocol, Figure 1 shows that not all patients in the sham group received an endoscopy as intended.
Despite the early termination, the addition this trial makes to available evidence is substantial. For one, there is a direct comparison of DJBL, IB, and sham and the inclusion of a generally obese population and not only one with type 2 diabetes. Considering the prevalence of obesity, the modest successes of conservative therapies, and the irreversible nature and complications of surgery, it is important to explore various other avenues. Furthermore, the correct treatment of missing data is essential in this field and the literature on IB and DJBL treatment is replete with examples where estimates are likely highly biased because of failure to do so. Finally, it is also important to note that we have no competing interests – this too is not the standard in the DJBL literature to date.
Regarding future implication of endoscopic treatment in obesity, the single and unique use of an IB or a DJBL device will probably not result in long-term weight loss after an application of 6 or 12 months. A valuable option for IB is a bridge to surgery in superobese patients with a high risk of AEs to improve the preoperative conditions [42]. In patients with comorbidities making them unfavorable for bariatric surgery, an IB or DJBL could be taken into consideration as an individual treatment option. In addition, the repeated use of IB over a period of more than 12 months could help to maintain or increase the IB induced weight loss [43]. Based on press releases, the manufacturer of the DJBL aims for a recovery of CE mark. However, the results of the recruiting Step-1-study have to be awaited until local authorities will eventually revise their decision. In the meantime, new concepts of transduodenal barriers that line the duodenum (e.g., removable transduodenal bypass) were developed and prospective trials will be conducted [44].
In conclusion, despite lack of power, our data strongly suggest that IB and DJBL lead to comparable weight loss while implanted, which is superior to a sham procedure. However, these procedures fail to achieve effective weight loss 12 months after explantation. These results need to be interpreted with caution due to the trial’s early termination.
Acknowledgments
Many thanks to Georg Kähler, Andreas Pfeiffer, and Johannes Hüsing for their excellent advice within the Data Monitoring Committee.
Statement of Ethics
This prospective, randomized, and patient- and assessor-blinded trial was registered (DRKS00011036) and approved by the Ethics Committee of the Medical Faculty of the University of Leipzig and all study centers. Written informed consent was obtained from all participants to be allocated to the study. The final study protocol was approved by the Ethics Committee of the Medical Faculty of the University of Leipzig (Kaethe-Kollwitz-Str. 82, 04109 Leipzig, Germany; Trial registration number: DRKS00011036 on DRKS at January 16, 2017) in accordance with the declaration of Helsinki, the “Medical Association’s Professional Code of Conduct,” and the principles of ICH-GCP guidelines (issued in June 1996, ISO14155 from 2012). In addition, the German Medical Devices Act (MPG, §§20-23a) was followed. Furthermore, local legal and regulatory authorities as well as the medical secrecy and the Federal Data Protection Act were followed. All local ethics committees of the participating centers consented to the master ethics committee approval. Prior to enrollment, each patient was given detailed information about the aims, scope, and possible consequences of the trial by a physician. No diagnostic or interventional procedures required for the clinical trial were performed without obtaining written consent from the patient.
Participating centers and ethics committees outside of Leipzig are as follows:
Clinic for Internal Medicine; Division of Gastroenterology; Sachsenhausen Clinic – Ethics committee of the Medical Faculty of the State Authorisation Association for Medical Issues; Hesse, Im Vogelsang 3, 60488 Frankfurt, Germany.
Clinic for Gastroenterology, Hepatology and Gastrointestinal Oncology; Bogenhausen Clinic; Munich – Ethics committee of the Medical Faculty of the Technical University Munich; Ismaninger Str. 22, 81675 Munich, Germany.
Conflict of Interest Statement
The authors declare no conflicts of interest related to this study.
Funding Sources
The trial was funded by the German Research Foundation (DFG; HO 2493/4-1), which was not involved in the database management (collection, analysis, interpretation of data) and had no access to randomization codes. The funding body did not participate in designing the study or writing the manuscript. The study protocol has undergone a peer-review process by the funding body.
Author Contributions
Conception and design: M.H., C.P., D.P., and A.H. Acquisition of data: M.H., J.F., F.G., J.S., and A.H. Analysis and interpretation of data: M.H., J.F., C.P., D.P., and A.H. Drafting the manuscript: M.H., D.P., and A.H. Revising the manuscript: C.P., F.G., W.S., J.S., and J.F. Final approval to the submitted version: M.H., A.H., C.P., D.P., F.G., W.S., J.S., and J.F. Funding: A.H., C.P., and D.P. All authors agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All authors read and approved the final manuscript.
Funding Statement
The trial was funded by the German Research Foundation (DFG; HO 2493/4-1), which was not involved in the database management (collection, analysis, interpretation of data) and had no access to randomization codes. The funding body did not participate in designing the study or writing the manuscript. The study protocol has undergone a peer-review process by the funding body.
Data Availability Statement
The data that support the findings of this study are not publicly available due to their containing information that could compromise the privacy of research participants but are available from the corresponding author (M.H.) upon reasonable request.
Supplementary Material.
References
- 1. GBD 2015 Obesity Collaborators; Afshin A, Forouzanfar MH, Reitsma MB, Sur P, Estep K, et al. Health effects of overweight and obesity in 195 countries over 25 years. N Engl J Med. 2017;377(1):13–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Kelly T, Yang W, Chen CS, Reynolds K, He J. Global burden of obesity in 2005 and projections to 2030. Int J Obes. 2008;32(9):1431–7. [DOI] [PubMed] [Google Scholar]
- 3. Haslam DW, James WP. Obesity. Lancet. 2005;366(9492):1197–209. [DOI] [PubMed] [Google Scholar]
- 4. Nyberg ST, Batty GD, Pentti J, Virtanen M, Alfredsson L, Fransson EI, et al. Obesity and loss of disease-free years owing to major non-communicable diseases: a multicohort study. Lancet Public Health. 2018;3(10):e490–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Ward ZJ, Bleich SN, Cradock AL, Barrett JL, Giles CM, Flax C, et al. Projected U.S. State-level prevalence of adult obesity and severe obesity. N Engl J Med. 2019;381(25):2440–50. [DOI] [PubMed] [Google Scholar]
- 6. Yumuk V, Tsigos C, Fried M, Schindler K, Busetto L, Micic D, et al. European guidelines for obesity management in adults. Obes Facts. 2015;8(6):402–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Jebb SA, Ahern AL, Olson AD, Aston LM, Holzapfel C, Stoll J, et al. Primary care referral to a commercial provider for weight loss treatment versus standard care: a randomised controlled trial. Lancet. 2011;378(9801):1485–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Jensen MD, Ryan DH, Apovian CM, Ard JD, Comuzzie AG, Donato KA, 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):S102–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Garvey WT, Mechanick JI, Brett EM, Garber AJ, Hurley DL, Jastreboff AM, et al. AMERICAN association OF clinical endocrinologists and AMERICAN college OF endocrinology comprehensive clinical practice guidelines for medical care OF patients with obesity. Endocr Pract. 2016;22(Suppl 3):1–203. [DOI] [PubMed] [Google Scholar]
- 10. Zhang C, Yuan Y, Qiu C, Zhang W. A meta-analysis of 2-year effect after surgery: laparoscopic Roux-en-Y gastric bypass versus laparoscopic sleeve gastrectomy for morbid obesity and diabetes mellitus. Obes Surg. 2014;24(9):1528–35. [DOI] [PubMed] [Google Scholar]
- 11. Hofsø D, Fatima F, Borgeraas H, Birkeland KI, Gulseth HL, Hertel JK, et al. Gastric bypass versus sleeve gastrectomy in patients with type 2 diabetes (Oseberg): a single-centre, triple-blind, randomised controlled trial. Lancet Diabetes Endocrinol. 2019;7(12):912–24. [DOI] [PubMed] [Google Scholar]
- 12. Hammer HF. Medical complications of bariatric surgery: focus on malabsorption and dumping syndrome. Dig Dis. 2012;30(2):182–6. [DOI] [PubMed] [Google Scholar]
- 13. Chang SH, Stoll CR, Song J, Varela JE, Eagon CJ, Colditz GA. The effectiveness and risks of bariatric surgery: an updated systematic review and meta-analysis, 2003-2012. JAMA Surg. 2014;149(3):275–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Sullivan S, Edmundowicz SA, Thompson CC. Endoscopic bariatric and metabolic therapies: new and emerging technologies. Gastroenterology. 2017;152(7):1791–801. [DOI] [PubMed] [Google Scholar]
- 15. Abu Dayyeh BK, Thompson CC. Obesity and bariatrics for the endoscopist: new techniques. TherapAdvGastroenterol. 2011;4(6):433–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. de Moura EG, Lopes GS, Martins BC, Orso IRB, Coutinho AMN, de Oliveira SL, et al. Effects of duodenal-jejunal bypass liner (EndoBarrier®) on gastric emptying in obese and type 2 diabetic patients. Obes Surg. 2015;25(9):1618–25. [DOI] [PubMed] [Google Scholar]
- 17. Martinez-Brocca MA, Belda O, Parejo J, Jimenez L, del Valle A, Pereira JL, et al. Intragastric balloon-induced satiety is not mediated by modification in fasting or postprandial plasma ghrelin levels in morbid obesity. Obes Surg. 2007;17(5):649–57. [DOI] [PubMed] [Google Scholar]
- 18. Mathus-Vliegen EMH, Eichenberger RI. Fasting and meal-suppressed ghrelin levels before and after intragastric balloons and balloon-induced weight loss. Obes Surg. 2014;24(1):85–94. [DOI] [PubMed] [Google Scholar]
- 19. Hollenbach M, Prettin C, Gundling F, Schepp W, Seufert J, Stein J, et al. Design of the Weight-loss Endoscopy Trial (WET): a multi-center, randomized, controlled trial comparing weight loss in endoscopically implanted duodenal-jejunal bypass liners vs. intragastric balloons vs. a sham procedure. BMC Gastroenterol. 2018;18(1):118. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Bates D, Mächler M, Bolker B. Fitting linear mixed-effects models using lme4. J Stat Software. 2015;67:1–48. [Google Scholar]
- 21. Abu Dayyeh BK, Kumar N, Edmundowicz SA, Jonnalagadda S, Larsen M, Sullivan S, 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):425–38.e5. [DOI] [PubMed] [Google Scholar]
- 22. Petroff D, Hoffmeister A. Excess weight loss should not be used to define success for bariatric endoscopy. Gastrointest Endosc. 2016;83(6):1306. [DOI] [PubMed] [Google Scholar]
- 23. Mathus-Vliegen EM, Tytgat GN. Intragastric balloon for treatment-resistant obesity: safety, tolerance, and efficacy of 1-year balloon treatment followed by a 1-year balloon-free follow-up. Gastrointest Endosc. 2005;61(1):19–27. [DOI] [PubMed] [Google Scholar]
- 24. Genco A, Cipriano M, Bacci V, Cuzzolaro M, Materia A, Raparelli L, et al. BioEnterics Intragastric Balloon (BIB): a short-term, double-blind, randomised, controlled, crossover study on weight reduction in morbidly obese patients. Int J Obes. 2006;30(1):129–33. [DOI] [PubMed] [Google Scholar]
- 25. Genco A, Lopez-Nava G, Wahlen C, Maselli R, Cipriano M, Sanchez MMA, et al. Multi-centre European experience with intragastric balloon in overweight populations: 13 years of experience. Obes Surg. 2013;23(4):515–21. [DOI] [PubMed] [Google Scholar]
- 26. Schouten R, Rijs CS, Bouvy ND, Hameeteman W, Koek GH, Janssen IMC, et al. A multicenter, randomized efficacy study of the EndoBarrier Gastrointestinal Liner for presurgical weight loss prior to bariatric surgery. AnnSurg. 2010;251(2):236–43. [DOI] [PubMed] [Google Scholar]
- 27. Escalona A, Pimentel F, Sharp A, Becerra P, Slako M, Turiel D, et al. Weight loss and metabolic improvement in morbidly obese subjects implanted for 1 year with an endoscopic duodenal-jejunal bypass liner. AnnSurg. 2012;255(6):1080–5. [DOI] [PubMed] [Google Scholar]
- 28. Dumonceau JM. Evidence-based review of the Bioenterics intragastric balloon for weight loss. Obes Surg. 2008;18(12):1611–7. [DOI] [PubMed] [Google Scholar]
- 29. Tate CM, Geliebter A. Intragastric balloon treatment for obesity: review of recent studies. Adv Ther. 2017;34(8):1859–75. [DOI] [PubMed] [Google Scholar]
- 30. Jung SH, Yoon JH, Choi HS, Nam SJ, Kim KO, Kim DH, et al. Comparative efficacy of bariatric endoscopic procedures in the treatment of morbid obesity: a systematic review and network meta-analysis. Endoscopy. 2020;52(11):940–54. [DOI] [PubMed] [Google Scholar]
- 31. Ruban A, Glaysher MA, Miras AD, et al. A duodenal sleeve bypass device added to intensive medical therapy for obesity with type 2 diabetes: a RCT. Southampton (UK): NIHR Journals Library; 2020. [PubMed] [Google Scholar]
- 32. Dastis NS, François E, Deviere J, Hittelet A, Ilah Mehdi A, Barea M, et al. Intragastric balloon for weight loss: results in 100 individuals followed for at least 2.5 years. Endoscopy. 2009;41(7):575–80. [DOI] [PubMed] [Google Scholar]
- 33. van Rijn S, Roebroek YGM, de Jonge C, Greve JWM, Bouvy ND. Effect of the EndoBarrier device: a 4-year follow-up of a multicenter randomized clinical trial. Obes Surg. 2019;29(4):1117–21. [DOI] [PubMed] [Google Scholar]
- 34. Cohen R, le Roux CW, Papamargaritis D, Salles JE, Petry T, Correa JL, et al. Role of proximal gut exclusion from food on glucose homeostasis in patients with Type 2 diabetes. Diabet Med. 2013;30(12):1482–6. [DOI] [PubMed] [Google Scholar]
- 35. Mariam A, Pantalone KM, Iyer N, Misra-Hebert AD, Milinovich A, Miller-Atkins G, et al. Associations of weight loss with obesity‐related comorbidities in a large integrated health system. Diabetes Obes Metab. 2021;23(12):2804–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Popov VB, Ou A, Schulman AR, Thompson CC. The impact of intragastric balloons on obesity-related Co-morbidities: a systematic review and meta-analysis. Am J Gastroenterol. 2017;112(3):429–39. [DOI] [PubMed] [Google Scholar]
- 37. Jirapinyo P, Haas AV, Thompson CC. Effect of the duodenal-jejunal bypass liner on glycemic control in patients with type 2 diabetes with obesity: a meta-analysis with secondary analysis on weight loss and hormonal changes. Diabetes Care. 2018;41(5):1106–15. [DOI] [PubMed] [Google Scholar]
- 38. Rohde U, Hedbäck N, Gluud LL, Vilsbøll T, Knop FK. Effect of the EndoBarrier Gastrointestinal Liner on obesity and type 2 diabetes: a systematic review and meta-analysis. Diabetes Obes Metab. 2016;18(3):300–5. [DOI] [PubMed] [Google Scholar]
- 39. Lingvay I, Guth E, Islam A, Livingston E. Rapid improvement in diabetes after gastric bypass surgery: is it the diet or surgery? Diabetes Care. 2013;36(9):2741–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Wiggins T, Sharma O, Sarfaraz Y, Fry H, Baker J, Singhal R. Safety and efficacy of 12-month intra-gastric balloon-series of over 1100 patients. Obes Surg. 2024;34(1):176–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Karlas T, Petroff D, Feisthammel J, Beer S, Blüher M, Schütz T, et al. Endoscopic bariatric treatment with duodenal-jejunal bypass liner improves non-invasive markers of non-alcoholic steatohepatitis. Obes Surg. 2022;32(8):2495–503. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Ball W, Raza SS, Loy J, Riera M, Pattar J, Adjepong S, et al. Effectiveness of intra-gastric balloon as a bridge to definitive surgery in the super obese. Obes Surg. 2019;29(6):1932–6. [DOI] [PubMed] [Google Scholar]
- 43. Dumonceau J-M, François E, Hittelet A, Mehdi AI, Barea M, Deviere J. Single vs repeated treatment with the intragastric balloon: a 5-year weight loss study. Obes Surg. 2010;20(6):692–7. [DOI] [PubMed] [Google Scholar]
- 44. Finze A, Otto M, Kähler G. Removable transduodenal bypass: development of a new endoscopic concept. Visc Med. 2023;39(3–4):87–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are not publicly available due to their containing information that could compromise the privacy of research participants but are available from the corresponding author (M.H.) upon reasonable request.