Introduction
Bariatric surgery (BS) remains the most effective treatment for morbid obesity, offering sustained long-term weight reduction with improvement and prevention of comorbidities and reduced mortality rates [
1,
2]. However, up to 25% of patients struggle with considerable weight regain after BS [
2‐
4]. In a NEJM article from 2017 [
3] with a patient follow-up of 12 years after Roux-en-Y gastric bypass (RYGB), median weight regain was 10 kg out of a mean maximal weight loss of 45 kg (i.e., a mean regain of weight after weight NADIR of 22.5%). An Indian study [
4] of 9617 patients followed up over 5 years after RYGB demonstrated a significant weight regain of more than 25% after weight NADIR in 14% of patients (i.e., 1336 of initial 9617 patients). It is well known that weight regain is associated with reappearance or worsening of obesity-associated comorbidities like hypertension and type 2 diabetes (T2D): 75% experience remission after 2 years and 51% after 12 years [
5]. Therefore, it is important to address weight regain after RYGB surgery, to avoid compromising the sustained improvement of obesity-associated comorbidities [
3] and to improve long-term patient satisfaction.
Reoperations are associated with serious complications in all surgery. Bariatric surgery is especially prone to late revisions due to insufficient weight loss or weight regain, which results in increased complications and higher mortality rates [
6‐
11]. Endosurgery for dilated upper anastomosis and weight regain was evaluated in a 2018 meta-analysis [
12] and considered a valid alternative to laparoscopic surgical reoperation for weight regain, as a much lower complication rate was recorded [
12] compared with laparoscopic reoperations [
6‐
11].
Pharmacotherapy with liraglutide, a glucagon-like peptide1 (GLP1) receptor agonist, has been successful in reducing weight in non-surgical patients with obesity in its 3.0 mg daily subcutaneous injection and was reported to result in a mean weight loss of about 9 kg after 1 year in non-surgical patients [
13]. Liraglutide, therefore, presents an interesting alternative to reoperation or endosurgery in the management of weight regain after BS. In a preliminary study of short duration (28 weeks) [
14] in RYGB patients, adjunctive high-dose liraglutide resulted in a median weight loss of 4.7 kg/m
2. Similarly, a study in RYGB or gastric banding patients [
15] resulted in a median weight loss of 7 kg after 8 months. Moreover, in a short-term randomized clinical trial of 28 weeks, patients with obesity were successfully and safely treated with subcutaneous liraglutide (1.8 mg daily) for persistent or recurrent T2D [
16] more than 1 year after bariatric surgery (RYGB or sleeve resection). Studies comparing medical, surgical (i.e., reoperation or endosurgery), or lifestyle interventions for managing insufficient weight loss or weight regain more than 6 years after RYGB surgery are scarce [
17] and need further medical validation.
Since more than 80% of weight regain has been previously observed within the first 6 years after the operation [
3], we included 95 consecutive patients who had undergone a RYGB at least 6 years previously presenting for yearly controls in the outpatient practice. These patients were assigned to different groups: initiation of liraglutide
. [
5,
13‐
15], endosurgery of the upper anastomosis [
12], or laparoscopic implantation of a Fobi-ring with pouch resizing [
11,
17]. Patients who kept their weight through lifestyle treatment alone [
18] served as control.
With the study, we aim to further elucidate whether pharmacotherapy with liraglutide is similarly effective in reversing weight regain more than 6 years after RYGB as revisional surgery aimed at restoring restriction.
Materials and Methods
Patients
Between January 2016 and March 2018, 95 consecutive unrelated Caucasian Swiss patients with formerly severe obesity, having been operated for RYGB at least 6 years earlier, were included in the study, independent of reported weight regain (Table
1). Inclusion criteria were as follows: RYGB at least 6 years ago, age older than 18 years, and regular follow-up at least every 6 months at the outpatient consultation of the obesity specialist (FH) after their primary RYGB operation. All patients were covered by public health insurance. Health insurance of individual patients was not a factor in deciding which treatment patients chose. Exclusion criteria were as follows: patients had received bariatric procedures other than RYGB or patients with a history of psychosis or alcohol or drug abuse.
Table 1
Baseline patient characteristics before RYGB and at the beginning of treatment of weight regain
Total | 95 | 11/84 | 54 ± 12 | 122 ± 18 | 45 ± 6 | 25.7 ± 3.8 | 18 (0–52) | 30.4 ± 5.1 | 9 ± 4 |
DC (controls) | 30 | 6/24 | 55 ± 10 | 124 ± 20 | 44 ± 9 | 25.5 ± 3.4 | 6 (0–9.7) | 27.1 ± 5.0 | 9 ± 4 |
LG (liraglutide) | 34 | 3/31 | 56 ± 10 | 120 ± 19 | 45 ± 8 | 25.5 ± 4.4 | 25 (11–46) | 31.2 ± 4.0 | 9 ± 5 |
ES (endosurgery) | 15 | 0/15 | 55 ± 12 | 115 ± 10 | 43 ± 4 | 24.9 ± 3.8 | 20 (11–52) | 31.0 ± 4.2 | 8 ± 4 |
FP (Fobi)§ | 16 | 3/13 | 43 ± 11£ | 132 ± 16£ | 46 ± 5 | 27.5 ± 3.3 | 27 (13–43) | 34.2 ± 4.9 | 8 ± 3 |
At the regular outpatient half-yearly bariatric control visit, at least 6 years after RYGB, patients were divided into 4 groups. Patients who had gained less than 10% from their respective measured weight NADIR (median 6% (range 0–9.7%), Table
1) were enrolled as controls. They were provided with dietary and lifestyle counseling at least every 3 months up to 24 months after study initiation (group DC,
n = 30) to address their eating behavior, physical activity, substance use, and weight control practice, similarly as suggested by King et al. [
18]. All patients who suffered from considerable weight regain (> 10% from weight NADIR) were offered one of the following three therapeutic possibilities in addition to dietary and lifestyle counseling [
18]:
(a)
Treatment with subcutaneous self-injection of the Glucagon-like peptide 1 receptor (GLP-1) agonist liraglutide (Novo Nordisk Pharma AG, Zurich, Switzerland), to a maximal dose of 3.0 mg daily, if tolerated, to enforce satiety (group LG,
n = 34) [
13]. Median weight regain from weight NADIR was 25% (range 11–46; Table
1).
(b)
Endosurgery, using Apollo’s Overstitch™ suturing system to reinstall restriction (group ES,
n = 15) [
12]. Median weight regain from weight NADIR was 20% (range 11–52; Table
1). For operative details, see below.
(c)
Surgical resizing of the gastric bypass pouch with additional implanting of a Fobi-ring to reinstall restriction (group FP,
n = 16) [
6,
11,
17]. Median weight regain from weight NADIR was 27% (range 13–43; Table
1). For operative details, see below.
Weight regain of 10% was chosen as a cut off for offering one of the three treatment modalities. The rationale for that was that internal evaluations of the quality of life questionnaires 8 years after BS revealed that weight regain of more than 10% was reported to have the most important negative impact on patients’ quality of life (unpublished data FH).
All patients were followed by the same obesity specialist (SCOPE Fellow, FH) at his private practice, Ärztezentrum Reichenburg, Switzerland. All weights reported in text and tables were measured at the outpatient consultation of the obesity specialist (FH).
Patients were fully informed and gave written consent to the study protocol and their participation in the study. All procedures performed were in accordance with ethical standards of the national research committee. The study complied with the Declaration of Helsinki 1964 and its later amendments.
Comorbidities and Their Reporting
Comorbidities listed were reported using the following criteria: Hypertension was defined as blood pressure ≥ 130/85 mmHg or taking antihypertensive drugs. Type 2 diabetes (T2D) was reported when HbA1c ≥ 6.5% or taking antidiabetic drugs; Dyslipidemia was reported with triglyceride levels > 200 mg/dl, HDL-cholesterol < 35 mg/dl, LDL-cholesterol levels > 100 mg/dl, or taking lipid-lowering drugs.
Data at 24-month follow-up were reported as follows: improved, improvement in measured values at 2 years compared with baseline or reducing dosage of respective drug treatment; impaired, impairment in measured values at 2 years compared with values before study start or increasing dosage of respective drug treatment; remission, normalization of measured values and stop of all medication to treat comorbidities.
Follow-up Schedule
The patients had routine follow-up appointments at the outpatient private practice with the same certified SCOPE Fellow at 4 weeks, 8–12 weeks, 6, 9, 12, 18, and 24 months after the procedure or liraglutide initiation, with more frequent consultations as indicated by patient symptoms. Patients’ weight and vitals were taken at each visit using the same equipment. All post-operative procedures were recorded.
Statistics
Mean ± SD or median (range), where appropriate, is presented throughout the manuscript unless specified otherwise. Differences between groups were calculated using analysis of variance (ANOVA), ANOVA for repeated measures, Chi-squared test, Kruskal-Wallis, or Friedman test, where appropriate. All analyses were performed using SAS (version 9.3) or IBM SPSS (version 22).
Discussion
Long-term weight regain after bariatric surgery (i.e., after more than 6 years) is a frequent, bothersome clinical problem for patients [
2,
3]. The present study demonstrates that liraglutide administered subcutaneously at a daily dose up to 3.0 mg when used as an adjunct to lifestyle counseling results in similar mean weight loss as in non-operated patients treated with liraglutide in a previous study [
13] with minimal side effects. Interestingly, no patient-reported gastroesophageal reflux disease occurred during the 24 months of observation. Moreover, patients with liraglutide treatment lost 15 ± 5% of their body weight (i.e., 13 ± 4 kg) after 24 months, achieving a similar weight to their weight NADIR after RYGB (Table
2). This finding is in line with a recent short-term study of 20 patients with a loss of mean weight of 9.7% in 28 weeks [
14] or 7 kg after 8 months [
15]. It is of interest to note that in the present study, the mean change in body weight was also stable after 24 months of treatment as long as patients continued liraglutide treatment (weight at 30 months (
n = 23); 74 ± 10 kg and 36 months (
n = 18) 73 ± 10 kg).
In a 2017 retrospective analysis, Stanford et al. [
19] demonstrated that pharmacotherapy-induced weight loss might be a useful adjunct in patients with inadequate weight loss or weight regain after bariatric surgery. In that study, 54% of patients lost 5% of their total weight with concomitant medication after surgery; 30% lost more than 10% of their weight, and 15% more than 15% [
19]. In our study population, using liraglutide resulted in a mean weight loss of 15 ± 8% during the 24 months of treatment: 50% of patients lost more than 15%, 76% lost more than 10%, and 89% lost more than 5%. Interestingly, the percentage of RYGB patients in the present study losing less than 5% of the initial weight was only 11%, whereas in liraglutide, treated non-bariatric patients with a similar initial BMI, more than 35%, did not lose 5% of initial weight [
13]. Therefore, liraglutide seems to be the most efficient drug to date to treat inadequate weight loss or weight regain after bariatric operations [
19]. Moreover, liraglutide appears to be at least as effective as weight loss adjunctive in bariatric patients as compared with similarly treated non-bariatric patients with obesity [
13]). Whether earlier treatment of weight regain (i.e., before 9 years after RYGB) would result in similar weight loss needs further investigation. However, from a patient’s perspective, we would speculate that an earlier intervention might lead to better long-term preservation of quality of life.
It is of interest to note that liraglutide-treated bariatric patients lost a similar amount of weight (13 ± 8 kg or BMI 4.7 ± 2.9 kg/m
2) as did surgically revised patients (17 ± 7 kg or BMI 5.5 ± 2.9, NS vs liraglutide, Table
2) during the 24 months of study. This interesting finding raises the question of whether treatment of weight regain after RYGB surgery should primarily be treated with liraglutide, and as a consequence, revisional surgery only be considered for patients not responding to liraglutide treatment. This question is of obvious clinical importance since reoperations after bariatric surgery are associated with a relatively high rate of complications [
6‐
11]. This is also demonstrated in the present study with a rate of 33% complications in surgically revised patients [
20]. Moreover, reports on long-term outcomes after surgical gastric pouch revisions are rare with only small patient series [
21‐
23]. Besides the relatively high rate of major complications in these patients, weight loss appears to be inconsistent from zero up to 70% excessive weight loss with a substantial rate of major complications [
21‐
23] pointing to liraglutide as first-line treatment of weight regain after RYGB rather than surgery.
A third of patients in the FP group experienced complications that lead to hospitalization in two cases or endoscopic dilatation of the upper anastomosis (up to 32 times) due to an inability to eat and vomiting. In our study, gastrointestinal disorders were common and mostly transient side effects in all groups that received revisional surgery (group FP and ES). In contrast, liraglutide-treated patients demonstrated a safety profile that is consistent with findings in previous reports for non-operated patients [
13,
14,
24,
25]. Other then reported for non-operated patients with obesity [
5,
13,
24‐
26], the liraglutide dose in our study population could not be increased to the maximal daily dose of 3.0 mg in 50% of patients because of the persistence of gastrointestinal side effects. A 2018 study reported that postprandial GLP-1 concentrations increase after RYGB [
27]. This finding could potentially explain the need for a smaller dose of liraglutide than that recommended for non-operated patients with obesity.
Gallbladder-related events and pancreatitis were not seen in patients with RYGB and treated with liraglutide in the present study. This is in contrast to non-operated patients, who experience more frequent gallbladder-related events and pancreatitis with liraglutide treatment than with placebo [
24‐
26]. No serious cardiovascular event was observed during the study period in either the liraglutide or in the surgically treated groups. Whether more frequent side effects would occur in larger patient groups treated with liraglutide after bariatric surgery is to be verified in further studies.
As was the case for liraglutide-treated patients, patients treated with endosurgery for weight regain experienced only minor side effects such as reflux disease, which was easily treatable with proton pump inhibitors. However, in contrast to a recent meta-analysis [
12] and another study comparing drug treatment with endosurgery [
28], in the present study, endosurgery only prohibited further weight regain (Table
2), thus not allowing the patients to significantly lose weight [
29]. As a consequence, all patients in the ES group were not satisfied with this treatment modality and stopped the study after 12 months (Table
2). Whether using other or more frequent suturing techniques to reverse weight regain after bariatric surgery as demonstrated by Patel et al. [
30] would allow similar results to reverse weight regain after RYGB, as achieved with liraglutide or Fobi-ring implantation with pouch resizing, awaits further study.
As expected and as described before [
2,
3] for non-bariatric patients with obesity, liraglutide was able to improve hypertension, T2D, and dyslipidemia impressively as a result of weight loss in RYGB patients treated with liraglutide or Fobi-ring implantation with pouch resizing for weight regain more than 6 years after initial RYGB. In contrast, RYGB patients treated with lifestyle treatment (DC) and patients treated with endosurgery (ES) experienced mostly no change in comorbidities, possibly due to the non-significant weight change during 24 months of study.
There are several limitations to this study, foremost being the selection and ascertainment bias. Additionally, one could criticize the relatively small sample size in each group; however, the fact that the retention rate was 100% in all groups after 2 years strengthens the validity of the results. Furthermore, the study was conducted in only one obesity center. Whether the presented results can be extrapolated to other centers remains open. However, this could be also seen as a positive point that adds to the comparability of our observations, as it excludes all variations in study conduct and analysis. We are aware that randomization to the treatment groups would have been the best approach to increase the scientific validity of our research findings and to answer the proposed question, how to best treat weight regain after bariatric surgery. However, we believe that a randomization approach would have severely decreased the interest to participate in the study, increased the drop-out rate, and most certainly reduced adherence to the treatments. Hence, we would have had less valid data to report and share.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.