Multidisciplinary approach of non-responders after bariatric surgery reduces the indication for revisional surgery: a retrospective cohort study

Background 20–30% of patients show a lack of response after bariatric surgery (BS). These non-responders may experience insucient weight loss or signicant weight regain. Based on previous research in our center, it has been identied that before the introduction of a multidisciplinary team (MDT), 68% of the non-responders underwent revisional surgery. This study describes the effect of an MDT on treatment strategy in non-responders after BS. Methods this retrospective study included non-responders that were reviewed in an MDT meeting. Patients were categorized as primary non- responders (1NR) or as secondary non-responders (2NR). Outcomes assessed were: I. MDT-based treatment (conservative versus operative), II. Weight loss, III. Complications after revisional surgery.


Abstract
Background 20-30% of patients show a lack of response after bariatric surgery (BS). These non-responders may experience insu cient weight loss or signi cant weight regain. Based on previous research in our center, it has been identi ed that before the introduction of a multidisciplinary team (MDT), 68% of the nonresponders underwent revisional surgery. This study describes the effect of an MDT on treatment strategy in non-responders after BS.
Methods this retrospective study included non-responders that were reviewed in an MDT meeting. Patients were categorized as primary non-responders (1NR) or as secondary non-responders (2NR). Outcomes assessed were: I. MDT-based treatment (conservative versus operative), II. Weight loss, III. Complications after revisional surgery. Results a total of 104 patients were included (n = 15 1NR, n = 89 2NR). In 73 patients, lifestyle and/or behavioral changes were indicated. Only eleven patients (13%) were re-operated in which one complication occurred.
Twenty patients did not show up at their appointment with the dietician, physical therapist and/or medical psychologist and were excluded from further analysis. Conservatively treated patients lost 2.1 kg < 12 months (SD = 7.29) and 0.8 kg < 24 months (SD = 5.08). Surgically treated patients lost 12.0 kg < 12 months (SD = 4.29) and 26.3 kg < 24 months (SD = 2.75).
Conclusions the rate of revisional surgery decreased after the introduction of an MDT. An explanation for this could be that an MDT drives more patients towards a conservative treatment since it identi es modi able lifestyle and/or behavioral factors contributing to non-response. Incorporation of an MDT may contribute to the selection of patients who might bene t from revisional surgery.

Background
Bariatric surgery (BS) has proven to be the most effective treatment for obesity in achieving long-term weight maintenance [1][2]. Unfortunately, 20-30% of patients do not respond well after BS as they experience insu cient weight loss or signi cant weight regain [3][4]. Popular criteria for describing insu cient weight loss are < 50% excess weight loss (EWL), or < 20% total weight loss (TWL) 1-2 years after surgery [5]. Hereafter, insu cient weight loss is described as primary non-response (1NR) while regaining an excessive amount of weight, after initial successful weight loss, is described as secondary non-response (2NR) [6]. The etiology for non-response can be lifestyle related but can also be found in anatomical alterations like a dilated pouch or gastroenterostomy, or gastro-gastric stula [7][8].
Revisional BS constitutes a popular solution for non-responders. Commonly offered revisional procedures are Roux-en-Y gastric bypass (RYGB), lengthening of biliopancreatic limb, band placement, revision of the gastric pouch and/or stoma or biliopancreatic diversion/ duodenal switch (BPD/DS) [7][8][9]. Up to now, the e cacy and safety of these procedures remain controversial as each procedure brings along speci c risks [10]. Lengthening of the biliopancreatic limb may result in severe malnutrition leading to the need for reoperations. Band placement may result in reoperation due to dysphagia and/or band migration. Revision of the pouch and/or stoma displayed unsatisfactory results with no BMI loss comparing the prerevision and 2-year post-revision situation [11].
In 1991, the National Institutes of Health Consensus Statement advocated that a multidisciplinary team (MDT) is required to optimize bariatric patient care [12]. This resulted in the formation of dedicated teams consisting of bariatric surgeons, obesity physicians, dieticians, physical therapists and medical psychologists. The bene cial effect of MDTs on surgical outcome has extensively been described in surgical oncology [13][14][15]. However, the effect of an MDT in non-responders after BS is relatively poorly understood. It was Srivastava et al. who demonstrated that a multidisciplinary lifestyle intervention combined with medication could improve early weight loss in non-responders [16].
Based on previous research in our center between 2012 and 2015, it has been identi ed that before the introduction of an MDT, 68% of the non-responders underwent revisional surgery [9]. This study included 65 patients with weight loss failure and weight regain who were consulted by a bariatric surgeon, though consultation by a dietician and physical therapist and/or medical psychologist was not scheduled routinely. Furthermore, no joint meeting by the MDT took place. It is likely that the role of an MDT needs further expansion in the management of non-responders after BS. The aim of this study was to describe the effect of an MDT, in a bariatric tertiary referral center, on treatment strategy in non-responders after BS.

Study population
Data about non-responders was collected in our center from January 2016 till December 2018 and was retrospectively analyzed. Patients were included if they regained > 5% weight with respect to the lowest postoperative weight after RYGB or sleeve gastrectomy (SG) (nadir weight). At the moment of inclusion, > 5% regain was a commonly used cut-off point [6]. Patients with a previous history of BS such as laparoscopic adjustable gastric banding or Mason gastroplasty were included as well. Patients were excluded in case of a banded gastric bypass, one anastomosis gastric bypass (OAGB) or BPD. The primary BS could have taken place in our center or elsewhere. In our center, all patients were screened for primary BS by an MDT. On top of the regular treatment, a preoperative individual treatment was given if indicated by the MDT. Patients were categorized as primary non-response (1NR) if the patients' excess weight loss (EWL) was less than 50% 12-24 months after primary surgery; patients were categorized as secondary non-response (2NR) if the EWL exceeded the 50% EWL threshold and a regain of > 5% was reported (nadir weight). All patients were discussed in an MDT meeting. The below section describes in more detail the process and approach taken before and during the MDT assessment.
Assessment before MDT meeting Initial assessment of the patient was done by the bariatric surgeon. Hereafter, the patient was referred to the dietician and physical therapist for assessment of nutritional habits and physical activity. The dietician focused on food intake, food choices, satiety, hunger and signs of emotional eating; the physical therapist focused on an activity habits. If indicated, consultation of a medical psychologist was offered. Gastrointestinal contrast studies and laboratory tests were only performed on indication.

MDT meeting
Once the patient had a consultation with all members of the team individually, the treatment strategy was discussed in a joint MDT meeting (weekly occurrence). These meetings included at least one member of the following elds of expertise: bariatric surgeon, nurse practitioner, dietician, physical therapist and medical psychologist; the meeting was chaired by an experienced bariatric surgeon. Notably, under certain circumstances the patient was discussed in the MDT despite that he/she was not seen by all members of the team. For example, in case of persistent lack of attendance at one of the appointments.

MDT treatment strategy
After evaluation of the patient by the team, a decision was taken whether the patient needed further lifestyle and/or behavioral intervention, or whether the patient was quali ed for revisional surgery. On behalf of this study, treatment options were divided into conservative and operative treatment ( Fig. 1). A conservative treatment consisted of a nutritional and/or physical intervention, summarized as "lifestyle". A nutritional intervention was indicated, for example, in case of unhealthy food choices and detrimental eating patterns. A physical intervention was indicated in case of a sedentary lifestyle with the goal to increase activity habits. A behavioral intervention was indicated when there were signs of emotional eating and problems in impulse control. If there were signs indicating an eating disorder, patients were referred to a psychiatric clinic specialized in treatment of such disorders. The indication for operative treatment was not based on the degree of non-response. The type of procedure depended on the index procedure, perioperative ndings and expert opinion.

Study outcomes
Patient demographics evaluated in this study include gender, age, BMI, weight loss, previous surgical history and relevant comorbidities. Outcome parameters were MDT-based treatment categorized as conservative or operative, weight loss in a period of 24 months after MDT meeting and complications after revisional surgery. Weight loss was described as %EWL and was calculated as follows: (initial weight − nal weight) / (initial weight − ideal body weight) × 100%. Ideal body weight is based on a BMI of 25 kg/m 2 . Initial body weight was the weight at the moment of screening. Optional, weight loss was also expressed in percent total weight loss (%TWL). %TWL was calculated as follows: ((preoperative weightnal weight) / preoperative weight)) × 100%. The percent of regain was calculated as percent kg gained after reaching lowest postoperative weight.

Statistical Analyses
Descriptive statistics were computed for demographic and medical characteristics. Quantitative data are presented as mean with standard deviation or median with interquartile range; categorical data are expressed in percentages. A paired t-test was performed for comparing pre-and posttreatment weight.

Results
Description of the population A total of 119 non-responders were assessed of which fteen patients were excluded for having a banded gastric bypass, OAGB, gastric band or BPD. Therefore, 104 patients were included in the study. The group of 1NR consisted of 15 patients and the group of 2NR consisted of 89 patients (Table 1). In the group of 2NR, 19 patients received an individual treatment, while in the group of 1NR none of the patients received an individual treatment. Moreover, 40% of the patients with 1NR had a history of gastric banding, while only 16% had this history in the group of 2NR.    Data presented as number (%), mean (standard deviation) or median (interquartile range)

Description of MDT-based treatment
Prior to MDT meeting, 98% was referred to a dietician, 97% was referred to a physical therapist and 28% was referred to a medical psychologist ( Table 2; Fig. 2). Twenty patients did not show up at the appointment with the dietician and/or physical therapist and/or medical psychologist. Therefore, no optimal treatment could be advised and consequently these patients were excluded from further analysis.
In total, 87% received a conservative treatment and 13% an operative treatment. When focusing on the conservative treatment, 86% received a nutritional intervention, 63% received a physical intervention and 25% received a behavioral intervention. Operative treatment, no.

Alternation of the limb length
Revision of the gastric pouch and/or stoma Data presented as number (%) Effect of MDT-based treatment on weight loss outcomes Table 3 displays the effect on weight loss within 12 months and 24 months after start of the treatment. As a result of conservative treatment, patients with 1NR lost 1.2 kg within 12 months (SD = 4.3) and 2.8 kg within 24 months (SD = 2.7). Patients with 2NR lost 2.4 kg within 12 months (SD = 7.8) and 0.4 kg within 24 months (SD = 5.4). These results did not reach signi cance compared to their baseline weight. Additional analysis showed that within the group of conservatively treated patients with 2NR 49% did not gain any more weight or lost weight, while 51% gained weight within 24 months. The results of the surgically treated patients with 2NR are more pronounced. On average, these patients lost 12.0 kg within 12 months (SD = 10.8) and 26.3 kg within 24 months (SD = 1.1). The weight loss within 12 and 24 months was statistically signi cant (p < 0.05). The results were recalculated in %TWL. When using ≥ 15% TWL as a cut-off point for successful weight loss, six patients were categorized as 1NR and 98 patients as 2NR. This gives an enlargement of the group of the 2NR without affecting weight loss outcomes (conservatively treated patients lost − 2.4 kg < 12 months SD = 7.6 and − 1.7 kg < 24 months SD = 5.4, data not shown).

Effect of MDT-based operative treatment on surgical outcomes
Eleven patients were selected for operative treatment since no major nutritional-, physical-, or behavioral changes were identi ed. These re-operated patients were all categorized as 2NR. The index procedure was RYGB in 9 patients and SG in 2 patients. Four different surgical procedures were performed in these patients which are described in Table 2. In detail, 2 patients with prior SG underwent conversion to RYGB;

Discussion
There is no standardized approach in the treatment of non-responders following bariatric surgery. While ongoing studies about the e cacy of revisional surgery are being reported in literature, studies about the effect of a multidisciplinary approach fall behind. To our knowledge, this report is rst in describing an MDT approach for non-responders after BS. The most interesting nding is that the minority of patients (13%) underwent revisional surgery, while all other patients (87%) were considered as not eligible for surgery. Before the introduction of an MDT evaluation in our center, the rate of revisional surgery was much higher (68% versus 13%) [9]. Based on these ndings, it is likely to assume that an MDT approach reduced the indication for revisional surgery. An explanation for this is that an MDT approach elicits important lifestyle and/or behavioral factors driving patients towards a conservative treatment.
When reviewing weight loss outcomes, revisional surgery was most effective in managing non-response. The nine re-operated patients achieved a weight loss of 12.0 kg (SD = 10.8), recalculated as 35.9% EWL (SD = 36.2) and 29% TWL (SD = 5.2) 24 months after surgery. This result differs from the weight loss that was found earlier in forty re-operated patients lacking an MDT approach (17.6% EWL, SD = 28.3) [9]. This difference suggests that weight loss outcomes of surgically treated non-responders have been improved due to the introduction of an MDT. One can argue that an MDT approach gives a more accurate way of selecting candidates for revisional surgery. This should however be interpreted with caution since the cohort of re-operated patients in this study was very small, different surgical procedures were performed by different surgeons leading to performance bias and the results showed high variation in weight loss outcomes.
It is noteworthy to mention that the 50% of conservatively treated patients did not gain weight and the other 50% only gained minor percentages. Therefore, the question is raised why should we offer evaluation by an MDT. Perhaps, the MDT underselected appropriate candidates for revisional surgery.
Although it is di cult to conclude on this based on this study, it could be presumed that an MDT is helpful in identifying lifestyle and behavioral factors contributing to non-response. When ignoring these factors while performing surgery, patients may be at risk for surgical morbidities. For further practices, a checklist of lifestyle and behavioral factors was made when evaluating candidates for revisional surgery (Table 4). The observed complication rate within this study is in line with other literature. Complication rate after pouch/anastomosis revision is estimated at 3.5% and after limb alteration 12% [10]. Revision of the gastric pouch and/or stoma was most frequently performed in this study. None of the seven patients that underwent this procedure suffered from a complication. Yet, one patient suffered from an ulcerative stenosis at the gastroenterostomy after limb length alteration, which is a frequently reported complication in literature [17].
An interesting nding in this study are the differences between primary and secondary non-responders. Previous studies did not report treatment outcomes using this classi cation, though it is proposed as standardized terminology [6]. This study supports this classi cation, since it was found that only patients with 2NR were treated surgically, indicating a different treatment strategy than patients with 1NR. A note of caution is necessary as there were signi cant demographic differences between the two groups. Speculatively, 1NR might be a manifestation of insu ciently treated or even untreated eating behaviors and/or psychological problems, whereas 2NR might point towards anatomical problems. This theory should be further explored to prove that different treatments strategies might be applicable for the two groups of non-responders.
This retrospective study has multiple limitations that should be mentioned. Not all patients received their primary surgery in our center and therefore information about preoperative screening and treatment could be missed. It is questionable whether certain lifestyle and/or behavioral factors were present prior to primary surgery and if so, if the patients' adjustments met the standards used in our center. Another limitation could be the possibility that the multidisciplinary team experienced a learning curve as the multidisciplinary evaluation of non-responders was introduced in 2016 in our center. It is possible that the different members of the team experienced a change in their evaluation and treatment of non-response. Moreover, it is important to bear in mind that pharmacological therapy has not been considered in this study cohort, while it has shown bene cial results in the treatment of non-responders [18]. Finally, it should be mentioned that results regarding weight loss and complication rates are limited due to the small selected group of patients that underwent revisional surgery. Patients that failed to attend follow-up appointments (48%) and/or investigations (19%) similarly limited the results. An explanation for this lack of follow-up could be the patients' embarrassment.

Conclusions
This study demonstrated that the indication for revisional surgery decreased after the introduction of an MDT for assessment of non-responders after bariatric surgery. An MDT is helpful in identifying modi able lifestyle and behavioral factors contributing to non-response. Incorporation of an MDT may contribute to the selection of patients who might bene t from revisional surgery and protect others from unnecessary surgical morbidity. Larger prospective studies are needed to evaluate the effectiveness of an MDT and might con rm that an MDT should be a mandatory part of the treatment of non-responders.

Declarations
Ethics approval of consent to participate For this type of study formal consent was not required from all individual participants. Ethical approval has been obtained from the Medical Ethics Committee of the Máxima MC. Reference number: N19.054 (L19.065).

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.   A caption was omitted by the authors in this version of the paper.