Failed Sleeve Gastrectomy: Single Anastomosis Duodenoileal Bypass or Roux-en-Y Gastric Bypass? A Multicenter Cohort Study

The sleeve gastrectomy (SG), as derived from the first step of the duodenal switch procedure, has recently become the most performed bariatric procedure worldwide. It has especially gained popularity during the past decade because of its relative simplicity compared to, for example, the Roux-en-Y gastric bypass. In the short-term, the SG yields good results for weight loss and comorbidity resolution [1]. However, if weight loss is inadequate or patients experience weight regain, they are advised to undergo revisional surgery. This is especially apparent in those with a higher initial BMI before the SG [2, 3]. Other patients have satisfactory weight loss but suffer from functional complications such as severe gastroesophageal reflux disease (GERD) or dysphagia due to a stenosis. Perhaps more concerning are recent reports of patients developing Barrett esophagus as soon as 5 years post-SG, which theoretically increases the risk of esophageal carcinoma [4‐7]. To compound matters, these patients do not benefit enough from the SG as a stand-alone procedure and are advised to undergo revisional surgery.

As the duodenal switch is technically demanding and associated with a high rate of perioperative morbidity [8, 9], other procedures for revisional surgery are needed. The question still remains which bariatric procedure should be performed as revisional surgery after a SG. Two available options include the single anastomosis duodenoileal (SADI) bypass and RYGB. The SADI has been introduced by Sanchez-Pernaute, A et al. (2007) as a simplification of the biliopancreatic diversion and duodenal switch [10]. It is suggested that weight loss results are similar to those obtained after the duodenal switch, but complications rates and nutritional deficiencies might be less frequent [11‐13]. However, data available upon this matter are scarce, especially for revisional surgery. A second option is the RYGB, which has been used regularly for many years and has proven its effect in bariatric surgery as a safe and effective primary as well as revisional procedure [14‐17]. A major advantage for GERD patients is that in a RYGB, the restrictive function of the pylorus is bypassed, which is why this operation is the best option to reverse GERD symptoms [18]. However, questions have been raised regarding failure rates following RYGB [19, 20].

To date, studies on SADI following SG reported on only small sample sizes. Furthermore, a comparison with the RYGB as a second step has never been made. The aim of this study is to investigate the effectiveness of the SADI versus the RYGB on health outcomes in morbid obese patients who have undergone SG with up to 2-year follow-up.

One hundred forty morbidly obese patients underwent revisional surgery after primary SG. A SADI was performed on 66 patients and 74 patients were converted to a RYGB. All of the SADI patients were operated to improve weight loss. Indications for revision to a RYGB were insufficient weight loss in 39 (52.7%) patients, SG-related functional problems or reflux in 29 patients (39.1%), or a combination of both in 6 patients (8.1%). An overview of the baseline characteristics is given in Table 1. The SADI group had a higher average pre-operative BMI, was younger, had a shorter hospital stay, and underwent surgery later after a SG compared to the RYGB group.

The present study investigated the effectiveness of the single anastomosis duodenoileal (SADI) bypass versus the Roux-en-Y gastric bypass (RYGB) on health outcomes as a revisional procedure after a failed sleeve gastrectomy (SG). To our knowledge, this study, which included 66 SADI and 74 RYGB patients, is the first to compare the SADI and RYGB as a second step operation for insufficient weight loss or weight regain.

Two previous studies evaluated weight loss after a SADI procedure with a prior SG and found excellent results, with a percentage excess weight loss of 70–80% after 24 months [11, 13]. These findings are similar to our results, where a percentage excess weight loss of 78% was found, and adds evidence to the notion that a SADI after SG provides consistent weight loss, even in different populations. When the RYGB after SG is compared to the existing literature, there seems to be an agreement on a peak in weight loss after 12 months and a decline in weight loss or even regain after this period [26, 27]. The overall level of %TBWL reported in these papers seems to be slightly higher than those obtained in our study but does not exceed 20%. Another proposed procedure to improve weight loss after a prior SG is the revisional sleeve gastrectomy (Re-SG), which is particularly promising when the original sleeve has dilated. It is found that the overall percentage of excess weight loss following the Re-SG can reach up to 57% at 12 months and up to 60% at 20 months follow-up [28, 29]. These percentages seem to exceed weight loss reported for RYGB in the current study; however, contrary to these outcomes Alsabah, S. et al. found more weight loss following revisional RYGB than after a Re-SG [29]. Still, they are not as high as those obtained after a SADI.

A second outcome evaluated in the present study was the rate of complications after revisional surgery. The SADI was originally developed as a modification of the biliopancreatic diversion with duodenal switch (BPD-DS). A reduction to just one anastomosis in SADI might be the reason why there are less complications when compared to the BPD-DS. Few studies compared the BPD-DS with RYGB after SG and found a generally higher amount of complications following BPD-DS; however, not significant [30, 31]. The present study found a similar complication rate; however, it can only be speculated on whether the SADI is an actual improvement for complication rates due to scarcity of this topic in the existing literature.

In studies comparing nutritional deficiencies after SADI, generally a higher percentage of deficiencies is described than observed in our patients [11, 13]. A difference in common channel length might have played a role in this, as previous studies included some cases with a common channel length of 200 cm, whereas the length was 250 cm for all but four of our patients. The authors mentioned that patients with a shorter common channel length were also more likely to develop deficiencies. Homan, J. et al. (2015) compared nutritional values for RYGB and BPD-DS after SG and found more deficiencies after BPD-DS (82% vs. 57%); however, not significant due to a small sample size [30]. The observation of a similar amount of nutritional deficiencies found in our sample provides some support for the hypothesis that deficiencies are less common after a SADI than after the traditionally performed BPD-DS.

The underlying mechanism for the difference in weight loss between the SADI and RYGB after SG might be explained by the difference in common channel and biliopancreatic limb length. It is assumed that a common channel length of 250 cm (with the exception of four common channels of 300 cm) in combination with a longer biliopancreatic limb length in SADI increases the malabsorptive component when compared to the RYGB leading to far better weight loss. Besides, a decline in weight loss or stabilization in weight 1 year after RYGB surgery is shared by several studies and can be considered as a failure of the procedure [32, 33]. Therefore, it may be questioned whether the RYGB should still be considered as an option for patients seeking to improve their weight loss after SG. Yet, many surgeons and patients are reluctant to choose a more invasive malabsorptive procedure such as the duodenal switch or SADI because of the disadvantages in terms of complications and deficiencies. It is argued that these disadvantages outweigh the benefits of more weight loss following these procedures when compared to the RYGB [8]. Perhaps these arguments might have led to a form of selection bias, in which patients with a higher pre-operative BMI are more likely to undergo a SADI, as can be seen in our data by the difference in baseline BMI before revisional surgery. However, after evaluation of the current results, it is reasonable to conclude that the disadvantages have been overemphasized, as was also mentioned previously by Sánchez-Pernaute, A. et al. [11]. In addition, identical operative times were observed and SADI patients experienced a shorter hospital stay after surgery. However, it should be noted that patients who presented with a functional problem are advised to undergo a RYGB. These patients were perhaps more prone to complications, leading to a higher rate of complications and a longer hospital stay following a RYGB.

Another concern associated with bariatric procedures that have larger malabsorptive characteristics, such as a SADI, is an increase in the occurrence of nutritional deficiencies. In the present study, the similarity of post-operative deficiencies found in both groups is likely related to sufficient supplementation, as every patient is advised to take specialized multivitamins to meet their daily requisite of vitamins after surgery and to prevent nutritional deficiencies from occurring. Moreover, patients are under strict guidance to meet their daily protein intake and multiple laboratory check-ups in the first year and yearly after that should ensure early detection of deficiencies and suitable treatment. As such, compliance to a strict vitamin regime is mandatory. It has been previously addressed that the super-morbid obese population is characterized for their extreme non-compliance [34], which might negatively impact the issue. This finding emphasizes the importance of regular follow-up for blood tests and a strict program by the clinic in which patients are treated.

A RYGB can be performed as revisional surgery if patients present with functional problems after a SG, such as dysphagia due to a stenosis or GERD. A narrow sleeve can be the cause of both of these problems; therefore, a SADI will most likely not solve the problem as the sleeve is left untouched. When a SG is converted into a RYGB, the sleeve is dissected to form a gastric pouch with a Roux-en-Y construction. As a result, problems attributed to a previous sleeve should resolve by bypassing the pylorus and promoting gastric emptying. In the present study, all patients but two were free of complaints after RYGB as a revisional procedure.

We acknowledge that our study has several limitations. Firstly, this includes the retrospective nature of our study, despite of the prospectively collected data in medical records. Secondly, an important factor after bariatric surgery which was not taken into account in the present study because of missing data is quality of life. Even though the SADI group did lose more weight, this does not necessarily lead to a better quality of life. Finally, as mentioned before, the super-morbid obese patients are known for their non-compliance. This was noticeable in our data by the significant amount of missing data. However, because of the frequent follow-up protocol, longitudinal data was mostly available. Points that contributed to the strength of our study were the relatively high number of patients included when compared to previous research. Furthermore, the present study combined results of multiple centers, which improves external validity of the results. Additionally, the SADI procedure as a second step has not been around for long and is still viewed as experimental; therefore, results that yield up to and including two-year follow-up are warranted [35].

In conclusion, revisional surgery following SG into RYGB or SADI are both feasible options, with a similar risk for complications and nutritional deficiencies. For cases of GERD or functional problems after SG, a conversion to RYGB is preferred. However, conversion into a SADI offers significantly more weight loss without increased short- and long-term morbidity and may therefore be considered the recommended operation for patients seeking weight improvement.