Introduction

Gastric bypass surgery is still one of the most effective therapeutic options for severe obesity. Bypass surgery improves obesity-associated medical complications markedly [1,2,3,4,5] and therefore influences obesity-related mortality in a positive manor [6].

Since the first gastric bypass was performed in 1954, the procedure has been modified several times over the last decades. The increasing knowledge of the human physiology and endocrine system has shifted the attention from old concepts of “restriction” and “malabsorption” to the hormonal and metabolic mechanisms triggered by the duodenal and small bowel exclusion [7]. However, an adequate and endurable postoperative weight-loss (WL) remains a matter of great concern. Giving the fact that weight regain (WR) or an insufficient weight loss (IWL) are issues that most of the bypass-patients have to face on the long term [8–9], a greater focus has been set on the variations of the limb-lengths of gastric bypass surgery in order to create the most effective and durable results with the lowest rate of adverse events [10].

Various operations with different limb lengths have been developed in order to achieve a better and longer lasting WL, especially in patients with a BMI > 50 Kg/m2 [11,12,13], or to correct IWL or WR after previous bariatric-metabolic surgery [10–11, 14–15]. The distal LRYGB, the very long limb RYGB and the biliopancreatic diversion are just examples that serve these purposes by creating a very short common-channel (CC) (50–100 cm) [16,17,18]. However, these operations have been criticized for causing severe vitamin deficiencies, bone loss, protein malnutrition, decreased fat-absorption and short bowel syndrome [11, 19,20,21,22, 23]. Many bariatric centers have therefore no longer performed these operations due to these serious side effects. The laparoscopic proximal gastric bypass (LRYGB) with a short Roux limb has been adopted as a standard procedure in many centers over the last decades. However, WR and IWL have been an issue using this technique in the long term follow up. We therefore introduced a modified “distal” gastric bypass (Fig. 1) in our bariatric center in order to create an alternative to the classical distal LRYGB, with a fairer compromise between metabolic impact and maintenance of WL. Its definition matches with the Type II distal LRYGB described in the DISCOURSE-Trial for revisional surgery [24] and with the very long roux limb LRYGB cited in other studies [25,26,27], except for the length of the CC, which we defined between 120 and 150 cm from the ileocecal valve, similarly to Risstad et al. und Salte et al. [21,22,23], in order to have less problems with malabsorption.

Fig. 1
figure 1

The laparoscopic distal Roux-en-Y Gastric Bypass (LRYGB) performed in our study

Full size image

We present our 5-year-results of distal gastric bypass compared to proximal gastric bypass surgery.

Materials and methods

Participants

We collected data from all patients that underwent a primary distal LRYGB between 2014 and 2015 in our bariatric center and analyzed them over 5 years. The data were compared to a cohort of patients that underwent a primary proximal LRYGB in the same time period in another bariatric center.

Inclusion criteria for the retrospective data collection were eligibility for bariatric-metabolic surgery according to the Swiss Society of Morbid Obesity (SMOB) and attendance to the regular follow-up visits within the first 5 years after surgery (at 3, 6, 9, 12, 24, 48 and 60 months). Patients that explicit declined their consent for data-collection were excluded from the analysis.

Our outcome reporting followed the guidelines of Brethauer et al. 2015 [28].

Operative technique

All operations were carried out laparoscopically by four experienced bariatric surgeons.

The distal LRYGB consisted of an intermediate short CC (120–150 cm), a BPL of 50–70 cm and a variable long RL. The proximal LRYGB consisted of a RL of 150 cm, a BPL of 50–70 cm and a variable long CC. All operations were performed using the same technique. The first step was to create the gastric pouch using a three cartridges (60 mm) of a linear stapler (one horizontal and two vertical). The BPL was then measured from the ligament of Treitz and connected to the gastric pouch. The gastrojejunostomy was performed using a circular stapler in all patients in both cohorts. The insertion site of the circular stapler was closed and the jejunojejunostomy was separated from the gastrojejunostomy using a linear stapler (jejunal segment resection). In the distal LRYGB group, the CC was subsequently measured with a length of 120–150 cm from the ileocecal valve and then connected with the BPL using a linear-stapled jejunojejunostomy in all patients in both cohorts. In the proximal LRYGB group 150 cm of RL was measured from the gastrojejunostomy and anastomosed to the BPL. The open mesenteric defects (Petersen and Brolin) have been closed using a nonabsorbable suture in all proximal LRYGB patients. The closure of the intermesentery space (Brolin-space) was not routinely performed in the distal LRYGB cohort in order to avoid a damage to the blood-supply of the small bowel, due to the large size of the defect.

Primary end point

The primary end point of our study was weight loss (WL) 5 years after surgery. WL was calculated as percent of excess of BMI loss (%EBMIL), loss of body mass index (Delta-BMI), percent of excess weight loss (%EWL) and percent of total weight loss (%TWL) at 3, 6, 9, 12, 24, 48 and 60 months postoperative. The ideal body weight was set at the weight corresponding to a BMI of 25 Kg/m2.

We defined successful weight loss (SWL) as %EWL > 50 at 18 months after primary surgery, while a %EWL < 50 at 18 months postoperative was considered as insufficient weight loss (IWL) [29].

Due to the heterogeneous definition of weight regain (WR), we chose to calculate it using four of the six calculations suggested by Voorwinde et al. 2020 [30] (Supplementary Table 1). We excluded the statements “any weight gain” or “weight gain with a BMI > 35 Kg/m2 after a successful weight loss”, reported in the afore mentioned publication, as too approximate.

Secondary end points

Further endpoints were postoperative complications and changes in the patients’ metabolic parameters after surgery, with particular focus on calcium-balance (parathormone, calcium, 25-OH-vitamin D), micronutrients (ferritin, albumin, zinc, vitamin B1 and vitamin B12) and hemoglobin A1c (HbA1c), especially in patients with type 2 diabetes or impaired glucose tolerance at baseline. All the laboratory measurements were collected at baseline and 5 years after surgery.

The complications were divided into major (serious, requiring intervention or re-operation) and minor (non-serious, requiring a conservative therapy or no therapy), as well as into early and late according to their onset (earlier or later than 30 days postoperative). Complications were also classified using the Clavien-Dindo Classification [31] and we calculated the Comprehensive Complication Index (CCI®) [32] for each patient having a complication within 30 days after surgery.

Patients’ information was collected from hospital or bariatric-centers database’s and recorded in a separate database using Microsoft® Excel® Spreadsheet Software, Version 16.16.27.

Statistics

Descriptive data were expressed as proportions for categorical variables, means with standard deviation (SD) for parametric and medians with interquartile ranges (IQR) for non-parametric data, unless otherwise specified. Kolmogorov-Smirnov test was used to test distribution of data. For comparison, the students t-test was performed for parametric and the Mann-Whitney-test for non-parametric data. Fischer’s exact test was used for categorical variates. A p value of < 0.05 was considered significant. Dichotomizations are based on medians. Multivariable analysis was performed using stepwise regression for the endpoints > 78% EBMIL-5y, > 61% EWL-5y, > 30% TWL-5y and > 12 Kg/m2 Delta-BMI-5y.

Statistical analysis was performed with JMP® Software Version 14.1 (SAS Institute, Cary, USA). Graphics were made with Graph Pad Prism® 8.4.3 (Graph Pad Software, La Jolla, CA, USA). We performed a univariable and multivariable analysis to investigate the impact of the following six factors on our primary outcome: age, sex, baseline-BMI, obesity-related medical problems, operative technique and post-operative adverse events. Significant p value was set at < 0.05.

We could not perform a power analysis to improve the quality of our work, since numerous comparative studies of good quality, such as Müller et al. 2008 [16], Risstad et al. [22–23], Salminen et al. 2018 [33] and Peterli et al. 2018 [34], either did not achieve statistically significant results or did not show their data distribution.

This study was carried out in accordance with the Swiss Association of Research Ethics Committees, Section Zürich (BASEC-Nr. 2019–02477).

Results

Demographics

One hundred and sixty-four patients underwent a proximal (N = 60) or distal (N = 104) LRYGB between January 2014 and December 2015 and were eligible for the retrospective data analysis. Four patients did not give their consent to the data collection und thus were excluded (one in the proximal and three in the distal group) (Fig. 2). Five years after surgery, we observed four lost-to-follow-up patients in the proximal and 23 in the distal LRYGB cohort. Only one patient died within five years after gastric bypass surgery in the distal LRYGB group due to a pancreatic cancer.

Fig. 2
figure 2

Study participants

Full size image

The baseline-characteristics of the two cohorts are given in Table 1. In both groups the majority were female (75% in the proximal vs. 78% in the distal LRYGB group) (p = 0.60). The median age was 45 (34–54) years in the proximal LRYGB group and 47 (39–55) years in the distal cohort (p = 0.18). At baseline, thirty-six patients (61%) had a BMI > 40 Kg/m2 in the proximal and 49 (49%) in the distal LRYGB group (p = 0.13).

Table 1 Demographics
Full size table

There was no significant difference in the overall incidence of obesity-related medical problems (p = 0.99) or in the rate of previous abdominal surgery (p = 0.67) between the two cohorts. However, the distribution of diabetes mellitus / impaired glucose tolerance, dyslipidemia and bone changes were more frequent in patients that underwent a proximal gastric bypass (p < 0.001).

Missing data

The BW and BMI measures 5 years after surgery were available for 55 from 59 patients after proximal LRYGB and for 77 from 101 patients after distal LRYGB.

The data to assess SWL and IWL at 18 months after surgery were complete for the proximal LRYGB cohort (59/59), but not for the distal group (92/101 patients).

Among the laboratory findings we had some missing data since blood sampling was not standardized before and after surgery. In addition to that, we could not compare Vitamin B12 levels in the two cohorts, since it was expressed as whole in the proximal group and as active Vitamin B12 (Holotranscobalamin) in the distal cohort.

Weight loss

Our weight loss results are shown in Table 2; Fig. 3.

Table 2 Weight Loss (WL), Weight Regain (WR) and Insufficient Weight Loss (IWR)
Full size table
Fig. 3
figure 3

Weight loss proximal vs. distal LRYGB

Full size image

The distal LRYGB resulted in significantly better long-term results than the proximal gastric bypass in terms of %EBMIL (median at 5 years: 83% vs. 65%, p = 0.001), %TWL (median at 5 years: 32% vs. 26%, p = 0.017) and %EWL (median at 5 years: 65% vs. 51%, p = 0.029). 5-years-Delta-BMI between the two cohorts was not significant (13 vs. 11, p = 0.09).

We observed a significantly higher rate of long-term WR after proximal compared to the distal LRYGB, according to WR-definition 1 (35% vs. 18%, p = 0.032) and WR-definition 4 (29% vs. 14%, p = 0.038).

Patients in the proximal LRYGB group were more likely to face IWL 18 months after surgery compared to the distal LRYGB group, although the difference was not significant (25% vs. 13%, p = 0.053).

The multivariable analysis (Table 3) showed a significant correlation between “female sex” and %EBMIL > 78 (p < 0.001), %EWL > 61 (p</=0.001), %TWL > 30 (p < 0.02) at 5 years postoperative, but not between “female sex” and long-term Delta-BMI > 12 Kg/m2 (p = 0.06). The impact of the operative technique (distal vs. proximal LRYGB) was only significant on the Delta-BMI at 5 years (p = 0.049). The presence of „obesity-associated medical problems at baseline” showed a positive correlation with long-term %EBMIL in the univariable (p = 0.013) but did not remain significant in the multivariable analysis (p = 0.54).

Table 3 Multivariable analysis (a-d)
Full size table

13 patients after proximal and 25 patients after distal LRYGB experienced major postoperative adverse events (Table 4) with a total of 18 major complications in the proximal and 26 in the distal group. Most of them occurred later than 30 days after surgery (12 after proximal and 21 after distal LRYGB). The median CCI for complications within 30 days was 33.7 (24–42) in the proximal LRYGB group and 20.9 (21–42) in the distal LRYGB cohort (p = 0.25).

Table 4 Postoperative adverse events
Full size table

Supplementary Table 2 shows the adverse events in detail. The most common serious complication was “internal hernia” (7 cases in the proximal and 17 in the distal cohort, p = 0.40).

According to the serum values displayed in Table 5, none of our patients showed a severe long-term hypoalbuminemia (Albumin < 30 g/l). Improvement in HbA1c (HbA1c < 6% after being pathological at baseline) was demonstrated in 7 out of 9 patients after proximal and in 10 out of 16 patients after distal LRYGB (p = 0.43). In the distal LRYGB group, significantly more patients had an increased parathyroid hormone level (PTH, > 6.9 pmol/l) than the proximal cohort at 5 years post-surgery (59/96 vs. 13/59, p < 0.001), but the two cohorts did not show a significant difference in the prevalence of secondary hyperparathyroidism (PTH > 6.9 pmol/l and calcium < 2.2 mmol/l) (12% vs. 8%, p = 0.65).

Table 5 Blood serum measurements
Full size table

Discussion

Our long-term results of distal gastric bypass in patients with a BMI between 37 and 44 kg/m2 are encouraging. In comparison to proximal laparoscopic Roux-en-Y gastric bypass, the distal LRYGB resulted in a significantly better %EBMIL, %EWL and %TWL, without a notable increase in adverse events or metabolic problems five years after surgery. Our study findings differ from those of other similar studies, specifically the conclusions drawn by Müller et al. (2008) [16] and Risstad et al. / Salte et al. (2016, 2021) [22–23], which evaluated BMI-decrease and did not find a significant difference between the two techniques. The missing difference in the latter study [22–23] might be explained by the inclusion of patients with BMI higher than 50 kg/m2. Delta-BMI and %TWL depend less on the initial BMI, while %EWL heavily does. Our Delta-BMI analysis did not yield significant results, emphasizing the limitations of this parameter in accurately reporting WL. This finding reinforces the vital need for standardization in reporting outcomes for bariatric-metabolic surgery, which is essential to maintain credibility and enable consistent comparisons across different studies.

In accordance with our results, the rate of adverse events in the above-mentioned studies [16, 22–23] was not significantly different. Our primary concern after distal LRYGB surgery, was the incidence of internal hernia, since we did not consistently close Brolin’s space. Internal hernias were indeed the most frequent major complication in the distal LRYGB group but the incidence between the two cohorts was not significantly different (p = 0.40). The high incidence of internal hernias in both groups, if compared with the current literature [16, 18, 20, 33,34,35], underlines the need for a more careful closure of the mesenteric defects.

Fortunately, we did not observe any cases of severe malnutrition, gastrointestinal side effects or other significant metabolic issues in the distal gastric bypass group. This might be due to the length of the CC, which was set at 120–150 cm in this study. However, we only reported 5-year follow up data and potential side effects can occur in a later stage. Our data revealed a 5-year prevalence rate of 62% for altered parathyroid hormone (PTH) levels in the distal LRYGB cohort, which was significantly higher than the 22% observed in the proximal group (p < 0.001). This finding indicates the need for a careful monitoring of calcium metabolism in these patients to prevent the onset of secondary hyperparathyroidism. We have to consider that severe malnutrition and metabolic complications often occur later than five years after the operation. Therefore, we cannot conclude that this procedure without any risk on the long term.

The significantly higher prevalence of diabetes mellitus/impaired glucose tolerance observed at baseline in our proximal LRYGB cohort compared to the distal LRYGB cohort (20%/54% vs. 12%/26%, p < 0.001) might be seen as a bias, suggesting that these patients had a worse predisposition to lose weight. However, our long-term WL results for this cohort were comparable to other studies, including Nergaard et al. 2014 [35] and Peterli et al. 2018 [34] (WL reported as %EBMIL) or Zhang et al. 2014 [36] and Salminen et al. 2018 [33,34,35,36] (WL reported as %EWL) and our multivariate analysis showed that the impact of obesity-associated medical problems on WL was not significant.

This work has clearly the limitations of a retrospective, non-matched cohort study. Despite our best efforts to minimize reporting and collection biases by strictly adhering to the guidelines for standardized outcome reporting in bariatric surgery suggested by Brethauer et al. [28] and performing a multivariable analysis, we cannot rule out the possibility of information bias due to the retrospective design and the presence of some missing data. Nevertheless, we believe that this study provides important information for the bariatric-metabolic surgery community. Our evidence supports some ongoing prospective studies like the DUCATI study [25,26,27] in considering a primary very long Roux-limb gastric bypass as a superior alternative to the standard LRYGB in order to achieve a more adequate and long-lasting WL. Elongation of the CC together with a long RL, while keeping the BPL constant, as we did in our procedure, might be critical in mitigating issues related to malabsorption and thus reduce the skepticism regarding the distal LRYGB. A primary distal gastric bypass operation may be technically more challenging than the standard proximal gastric bypass, but could play an important role in a patient-tailored approach that addresses diverse needs and objectives within bariatric-metabolic surgery. Furthermore, we could only provide five-year data of our patients due to the loss of follow up and the study design. However, obesity affects patients over decades. Weight regain and metabolic/malnutritive complications can occur a later stage ten years after the operation. Longer follow up data would be of great value and might alter our results. Better evidence with longer follow up periods is needed to support these considerations.

Conclusion

Our findings suggest that a primary modified distal laparoscopic Roux-en-Y gastric bypass (LYRGB) procedure with a common channel length of 120–150 cm and a biliopancreatic limb of 50–70 cm is a feasible and safe therapeutic option. It appears to result in superior long-term weight loss and weight loss maintenance compared to the standard proximal LYRGB technique. The metabolic complications and adverse events five years after surgery are reasonable.