The Efficacy and Safety of Adjustable Intragastric Balloon for Weight Loss: A Systematic Review and Meta-Analysis

Obesity has emerged as a global health priority due to its escalating prevalence and its association with a myriad of complications, encompassing but not limited to type 2 diabetes, cardiovascular disease, osteoarthritis, obstructive sleep apnea, and various malignancies [1]. According to published literature, more than 10% weight loss can improve lipid metabolism, reduce the compression of the respiratory tract by visceral fat, thereby decreasing the incidence of type 2 diabetes, hypertension, fatty liver disease, and obstructive sleep apnea and improving the quality of life for patients [1]. Lifestyle interventions targeting modifications in dietary habits and physical activity stand as the first approach to weight management, owing to their cost-effectiveness and the minimal risk of associated complications [2]. Nevertheless, achieving enduring weight loss has proven to be challenging even within the framework of rigorous clinical trials [3].

Endoscopic therapies provide a sustained alternative for weight loss, exhibiting less invasiveness than bariatric surgery and greater effectiveness than conventional lifestyle interventions [4]. The intragastric balloons (IGBs) have been in clinical use for over 3 decades, substantiating their safety and effectiveness in facilitating weight loss [5, 6]. However, as IGBs lack adjustability, most patients can only be inserted for 6–8 months [7]; 2–9% of patients encounter intolerance, leading to early balloon extraction [8]; and some patients experienced variability in response and reduced efficacy after 1–3 months [9]. Therefore, the FDA approved the first adjustable intragastric balloon (aIGB), Spatz3 (Spatz Medical), in 2021. The balloon design facilitates the removal of fluid in cases of intolerance or the addition of extra fluid to enhance weight-loss outcomes [10]. In the latest multicentre randomized controlled trial (RCT) about aIGB, the mean total body weight loss (TWL) at 32 weeks was 15.0%; clinical response was observed in 92% of patients, upward volume adjustment enabled an additional mean of 5.2% TWL, and downward volume adjustment allowed 75% patients to complete the full duration of therapy [11]. Overall, the adjustability of balloon volume permitted individualized therapy, optimizing both weight loss and tolerance.

To date, no meta-analysis has assessed the efficacy and safety of aIGB for weight loss. For these reasons, we comprehensively evaluated the effectiveness and safety of aIGB through systematic review and meta-analysis. Additionally, we provided a more referenced assessment by reviewing all case reports about aIGB.

This study followed the latest version of Meta-Analysis of Observational Studies in Epidemiology (MOOSE) and Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [12, 13]. The methodology of the research was recorded in PROSPERO (CRD42023478559).

A comprehensive search of the literature was performed on October 31, 2023, in multiple bibliographic databases (PubMed, Embase, and the Cochrane Database), without time or language restrictions. The primary items were searched in the Medical Subject Headings database and relevant articles, and the keywords finally included “adjustable intragastric balloon” OR “Spatz.” The search was applied to all fields. The references from the retrieved articles were also manually reviewed for any missed articles. Two investigators (Guowu Sun and Yinuo Wang) conducted the systematic search. Disagreements were resolved by consensus with the intervention of a third reviewer (Chuqi Xia).

Predefined inclusion criteria were retrospective, prospective or cross-sectional studies, cross-sectional or cohort studies, or RCTs with studies reporting weight loss results of inserting the latest aIGB (Spatz3) in adult patients, and the sample of studies was >80, the age of patients was over 18 years old. Exclusion criteria were animal studies, commentaries, reviews, overlapping patients, and publications in a language other than English. If multiple articles were on the same study sample with the same exposure and outcome, the publication with the largest sample size was retained. In addition, case reports about the latest aIGB were aggregated for description and were not included in the calculation. All studies were evaluated for quality independently by 2 investigators (Guowu Sun and Yinuo Wang). RCTs’ risk-of-bias assessment was done using the Cochrane Collaboration risk of bias tool, and case-control and cohort studies’ quality were assessed by the Newcastle-Ottawa scale (NOS) [14]. If studies received 6 out of 9 possible points in NOS, we consider their quality acceptable.

Data were extracted by 3 authors (Chuqi Xia, Guowu Sun, and Yinuo Wang) independently and then compared. The baseline information included (1) the first author’s name, (2) the year of publication, (3) the country, (4) the study design, (5) the setting of the research center, (6) patient demographics (age, gender, BMI, weight), (7) sample of patients, (8) inserting time. The outcome indicators included (1) TWL after treatment, (2) number of upward adjustment and additional weight loss after upward adjustment, (3) number of downward adjustment and number of alleviating intolerance after downward adjustment, (4) number of adverse events (early removal within 3 months, gastric ulcer, migration with intestinal occlusion, esophagitis, and balloon leaks). When data could not be extracted or calculated from the text, the corresponding author was contacted to obtain further information. Efficacy and safety were evaluated by calculating TWL and severe adverse event rates reported in the included studies.

The mean or standard deviation (sd) for continuous data (TWL) and number of events or total sample for each categorical factor were obtained or calculated when possible. The pooled TWL and proportions of the event were then computed using either fixed-effects or random-effects models, depending on study homogeneity or heterogeneity, and the statistical heterogeneity was evaluated using I² statistics and Cochran Q test values [15]. An I² value >50% was considered high statistical heterogeneity (I² > 50% and p < 0.05). For the main outcomes that failed the homogeneity test, we searched for potential sources of heterogeneity, including study type, study setting, average age, and sample and evaluated these via subgroup analysis, where TWL was the summary outcome. As fewer than 10 studies were included in this meta-analysis, Egger’s publication bias plot was generated to evaluate the possibility of publication bias. All statistical analyses were performed using Stata SE.

The flow diagram, Figure 1, summarizes the process of study selection. The initial search yielded 5,216 potentially relevant articles (2,228 articles from PubMed, 2,975 articles from Embase, and 13 articles from the Cochrane Database), of which 1,682 were duplicates. The remaining 3,241 articles underwent title and abstract review, then 3,206 irrelevant articles that did not fulfill the eligibility criteria were excluded. A total of 35 articles performed a full-length review, and 23 articles were excluded according to the exclusion criteria. Finally, 12 unique studies which are fulfilling the redefined inclusion criteria were included in this meta-analysis [16‒27]. One study with 2 cohorts’ results was pooled into one cohort [21]. The characteristics of the included studies are comprehensively described in Table 1. Twelve studies with a total of 4,981 patients were included in this meta-analysis: 2 were RCTs, 10 were observational studies without randomization, 7 were multicenter, and 5 were single-center. In terms of patients’ baseline information, 734 patients were men, 1,760 patients were women, the mean age was 41.4 years old, the mean baseline weight was 100.20 kg, the mean baseline BMI was 36.64 kg/m², and the average implantation time was 9.9 months.

Ten observational studies were assessed for bias using NOS; the maximum score of them is 9 points, while the minimum score is 6 (online suppl. Table 1; for all online suppl. material, see https://doi.org/10.1159/000542921). RCTs’ risk-of-bias assessment was done using the Cochrane Collaboration risk-of-bias tool (online suppl. Fig. 1). The quality of these articles was all considered to be acceptable.

Table 2 summarizes the efficacy of aIGB, outcome indicators including TWL after treatment, number of upward adjustment additional weight loss after upward adjustment, number of downward adjustment, and number of alleviating intolerance after downward adjustment. The TWL after aIGB treatment was reported in 6 studies [17, 18, 21, 23, 24, 26]. Pooled estimation of a meta-analysis showed the pooled mean TWL after aIGB treatment was 16.4% (95% CI: 0.153–0.175, I² = 91.2%, p < 0.001) (Fig. 2a), and the Egger’s test showed p = 0.252 (online suppl. Fig. 2a). All researchers claim to have provided health education to patients, including dietary intervention and exercise guidance, but the contribution of these lifestyle interventions to weight loss seems to be non-significant. In one of the studies, the researchers reported a TWL of 15% for the aIGB group, while the TWL for the group with lifestyle intervention alone was only 3% [17].

Six studies reported data of upward adjustment [16, 19, 22, 24, 26, 27], and all reported additional weight loss with a mean of 6.3 (4.8–9.3) kg. Pooled prevalence of patients choosing upward balloon adjustment during treatment was 34.2% (95% CI: 0.220–0.485, I² = 96.5% p < 0.001) (Fig. 2b), and the Egger’s test showed p = 0.763 (online suppl. Fig. 2b).

Seven studies reported the data of downward adjustment [16, 17, 19, 22, 24, 26, 27]. Pooled prevalence of patients choosing downward balloon adjustment during treatment was 9.2% (95% CI: 0.065–0.119, I² = 73.9%, p = 0.001) (Fig. 2c), and the Egger’s test showed p = 0.487 (online suppl. Fig. 2c). All studies reported that downward adjustment could alleviate intolerance or relief symptoms (pain, nausea, vomiting, abdominal pressure, and so on); 3 reported a remission rate of 100% [16, 24, 27], and the minimum remission rate was 75% [17]. Pooled estimation of a meta-analysis showed a remission rate of 90.8% (95% CI: 0.817–0.974, I² = 53.4% p < 0.045) (Fig. 2d), and the Egger’s test showed p = 0.893 (online suppl. Fig. 2d).

Table 3 summarizes clinical studies and all reports of adverse events after aIGB. According to the descriptions in these articles, the health status of all reported patients prior to balloon implantation met the conditions required for the procedure, and no contraindications were present. Ten studies reported data of serious adverse events [16, 17, 19, 20, 22‒27], and there was no mortality. Eight studies reported the occurrence rate of intolerance and early removal within 3 months; the pooled estimate of occurrence rate was 5.7% (95% CI: 0.035–0.078, I² = 79.8% p < 0.001) (Fig. 3a), and Egger’s test showed p = 0.303 (online suppl. Fig. 2e). Eight studies reported the occurrence rate of stomach ulcer [17, 19, 20, 22‒26]; the pooled estimate of prevalence of stomach ulcers was 1.1% (95% CI: 0.008–0.014, I² = 5.1%, p = 0.390) (Fig. 3b), and Egger’s test showed p = 0.186 (online suppl. Fig. 2f). In addition, 3 studies reported the prevalence of balloon leaks was 7.3% (8/110) [16], 0.5% (1/187) [17], and 0.9% (1/110) [25], respectively. Two articles reported that the prevalence of migration with intestinal occlusion was 0.04% (1/2,349) [20] and 1.8% (2/110) [25]. Only 1 literature reported the prevalence of esophagitis was 14.0% (329/2,349) [20].

In addition, there are 6 articles reporting 9 cases of serious adverse events after aIGB (Table 3) [29‒34]. Out of these, 8 (88.9%) were female, the mean age of the patients was 41.9 (24–62) years, and the mean inserting time was 6.75 (0.25–12) months. Out of 9 patients, 3 (33.3%) presented with perforation, 2 (22.2%) had balloon hyperinflation, and the remaining 4 cases reported small bowel obstruction [29], gastric outlet obstruction [30], pancreatitis [32, 33], and tear of gastric wall separately [34]. In terms of treatment, 3 underwent laparotomy [29, 34], 2 underwent laparoscopic repair of gastric perforation [33], 2 removed the balloon [30, 32], 1 refilled the balloon [31], and 1 substituted with another balloon [31]. 88.9% (8/9) of patients recovered after treatment and even succeeded in losing weight. Only one 45-year-old female had an urgent laparotomy for a huge tear of the greater curvature of the stomach, but unfortunately passed away intraoperatively despite all resuscitative measures [34].

Publication bias was assessed for all 6 outcomes by Egger’s test. All 6 egger’s tests were not significant (p > 0.05), showing that there was no obvious publication bias detected for these outcome measures (online suppl. Fig. 2). There was significant heterogeneity present in TWL outcomes and most single rate outcomes. We investigated the influence of a single study on the overall percentage of TWL, remission rate after downward adjustment, and the rate of early removal by omitting 1 study at a time. The omission of any one study showed no clinically significant difference in the results, indicating that the results were statistically reliable (online suppl. Fig. 3). However, after omitting the study of Badurdeen et al. [18], the heterogeneity of TWL decreased to 67.3%, and the heterogeneity of this study may stem from its unique gender ratio as it is the only study involved in which more men than women participated. After omitting the study of Usuy and Brooks [24], the heterogeneity of TWL decreased to 33.7%, but the result (87.9%) was still stable.

Multiple preplanned subgroup analyses were performed to assess the stability of these findings and identify potential sources of heterogeneity. Study setting, age, sample, and study type did not explain study heterogeneity (online suppl. Fig. 4). There was only 1 study that had a significant difference in baseline BMI from the other ones [18]; when omitting this study, the result was still statistically reliable.

This meta-analysis included 12 studies with a total of 4,981 patients. The results demonstrate that aIGB therapy is capable of achieving significant weight loss. The average insertion time was 9.9 months, with a pooled mean total weight loss (TWL) of 16.4%. Adjusting the balloon volume during treatment effectively managed both intolerance and weight loss plateaus.

Specifically, 34.2% of patients opted for upward balloon adjustments, reporting additional weight loss with a mean of 6.3 kg. Furthermore, 9.2% of patients chose downward balloon adjustments, with 90.8% of these patients experiencing alleviation of intolerance. The incidence of early intolerance and removal within 3 months was 5.7%, and the incidence of stomach ulcers was 1.1%. No significant publication bias was detected for these outcomes.

In summary, aIGB not only demonstrates excellent overall weight loss but also significantly improves patient tolerance and weight loss outcomes through volume adjustments. These findings are statistically reliable, indicating that aIGB is an effective weight loss and management strategy.

Despite a three-decade history and various iterations of intragastric balloons (IGBs), their broad-based adoption as a remedy for the burgeoning obesity epidemic has been significantly hindered by two critical issues: patient intolerance and weight loss plateau [14]. In response to these issues, the design of the aIGB (Spatz3) incorporates a feature allowing for control over its volume not merely at the point of implantation but continuously throughout the treatment duration. However, to achieve its adjustable functionality, the Spatz3 deviates from the completely smooth surface characteristic of conventional designs, the site for insertion of the filling valve forming a sort of “tail.” Currently, the efficacy and safety of aIGB have not been well described. Therefore, we conducted the first meta-analysis and systematic review to assess this issue.

Numerous studies have demonstrated the effectiveness of IGBs in weight loss, and the current models of IGBs include Orber, Reshape Duo, Obalon, Ellipse, etc. [35, 36]. This was confirmed in a recent RCT: 45.6% of Orbera patients achieved 10% TWL [15]. For aIGBs, our pooled result found that the inserting time of aIGB was 7–12 months, and the mean TWL was 16.4%, indicating significant and sustained weight loss, and even more favorable outcomes when compared to other IGBs. Additionally, compared to other weight-loss surgery methods, aIGB also has certain advantages. A clinical study involving 227 patients showed that patients who underwent endoscopic sleeve gastrectomy had a TWL of 16.6% 1 year post-surgery [16], similar to the weight-loss effect of aIGB (TWL 16.4%) we compiled. Since aIGB does not change the anatomical structure and does not require the removal of any tissue, it is undoubtedly a better option than endoscopic sleeve gastrectomy, when the weight-loss effects are comparable, as it inflicts less damage on the patient.

Meanwhile, regarding safety, the aIGB (Spatz3) has demonstrated substantial improvements over its predecessors. Espinet-Coll et al. [28] conducted a multicentre study on Spatz2 (predecessor of Spatz3) and found that out of 225 patients, 4 serious complications, 34 gastric ulcers, 7 migration, and 1 acute pancreatitis, but the complication rate of Spatz2 was lower when compared to other IGBs. Our data also indicated a very low rate of adverse event rates: 4% of early balloon removal, and 1.22% of gastric ulcers. In comparison, studies have shown that approximately 9%–12% of patients who undergo weight-loss surgeries such as RYGB and SG experience one or more adverse events within the first 5 years after surgery [17]. Therefore, aIGB is undoubtedly a safer method for weight loss. It is worth noting that although the effectiveness and safety of aIGB have been proven, this assistance is likely to be temporary. From the data we retrieved, the longest implantation time was 12 months, and longer follow-up data are lacking. However, compared to other weight-loss surgeries, aIGB still demonstrates superior effectiveness.

The biggest advantage of aIGB is that the volume can be adjusted during the treatment period. On the one hand, many patients experienced intolerance in the early stage of inserting IGBs, such as pain, nausea, vomiting, abdominal pressure, and so on. FDA trials of Reshape Duo and Orbera IGBs reported 14 and 22% early extraction rates, respectively [37]. In this study, 9.2% of patients underwent a downward balloon adjustment, resulting in 87.79% of patients avoiding intolerance and early removal, and completing treatment. On the other hand, most of the patients had a weight loss plateau after the initial strong balloon effect and requested a second round of rapid weight loss [38]. In this study, 34.2% of patients chose to upward balloon adjustment, which led to additional weight loss in all patients, with an average loss of 6.3 kg, a maximum of 9.3 kg, and a minimum of 4.8 kg. Consequently, we believe that the aIGB is capable of effectively managing both intolerance and weight-loss plateaus, maintaining the efficacy of weight loss in the process.

AIGB also has postoperative adverse events. These adverse events mainly include perforation, obstruction, and inflammation, which are primarily related to the compression of the gastric wall by the balloon. An oversized balloon volume and improper positioning may be factors contributing to the occurrence of these adverse events [31]. In addition, overexpansion of the balloon is also a relatively common adverse event. Overexpansion occurs when, for some reason, the internal pressure of the balloon exceeds the set upper limit and compresses the surrounding tissue. The two reported cases of balloon overexpansion associated with aIGB were both caused by bacterial contamination [31], which suggests that clinicians need to ensure strict aseptic techniques when implanting aIGB and adjusting the balloon’s capacity. Early diagnosis and timely intervention are relatively effective methods to prevent the occurrence of malignant complications. When patients experience unexplained abdominal pain, bloating, nausea, vomiting, and other symptoms after the implantation of an aIGB, it is important to promptly check the status of the balloon and consider reducing the balloon's volume or even removing the balloon [29].

Our study has several limitations. First, the sample size was relatively small and the number of RCTs reporting TWL was insufficient. Second, substantial heterogeneity based on I² statistics was present in many outcomes. This may be due to differences in baseline patient characteristics and treatment regimens in some of the included studies. Third, the long-term effects are still unclear as only a few studies reported results at inserting 12 months. Fourthly, although we have conducted thorough and comprehensive searches and provided the most extensive summary to date, our research still has room for improvement in studying complications due to limited data. Some situations, such as compliance with treatment recommendations, medication use, medical history, and family history, were not included in the data we found. According to related research, the occurrence of complications could also be influenced by adherence to management recommendations, general health status of patients, and genetic factors [38, 39]. We look forward to more abundant data in the future to address these gaps.

In summary, this meta-analysis demonstrated that aIGB is a reasonably safe and effective device and could become a practical tool for weight loss. However, given its potentially serious complications, we also need to keep an eye on the patient’s condition. Nevertheless, the limited quality of the available literature leads to several constraints discussed above. To more comprehensively evaluate its effectiveness compared to other IGBs, further prospective studies, ideally RCTs, are warranted.

An ethics statement is not applicable because this study is based exclusively on published literature.

The authors have no conflicts of interest to declare.

This study is funded by the National Natural Science Foundation of China (8236030102, Daoming Liang), Joint Special Fund of Applied Fundamental Research of Kunming Medical University granted by Science and Technology Office of Yunnan (202301AY070001-025, Daoming Liang) (202401AY070001-363, Chuqi Xia), and Yunnan Revitalization Talent Support Program (XDYC-MY-2022-0100, Daoming Liang).

Chuqi Xia, Yinuo Wang, and Guowu Sun contributed equally to this study and performed analysis and interpretation of the data. Yinuo Wang and Guowu Sun contributed to the drafting of the article. Chuqi Xia, Wanyang Lei, and Daoming Liang contributed to the conception, design, and critical revision of the article for important intellectual content. All authors gave final approval of the article.

Chuqi Xia, Yinuo Wang, and Guowu Sun contributed equally to this work.Guarantor of the article: Daoming Liang, MD.

All data generated or analyzed during this study are included in this article and its supplementary material files. Further inquiries can be directed to the corresponding author.