Introduction

Bariatric surgical procedures are classically classified along a spectrum from purely restrictive to malabsorptive procedures, although recent reports suggest that more complex hormonal, inflammatory, central nervous system, and gut microbial factors may also have an important impact on the effects of the operations. One anastomosis gastric bypass (OAGB) is a “combined procedure” and it has both a “restrictive” and a “malabsorptive” component [1, 2].

While the early results in terms of weight management and T2DM control appear promising, the previous position statement noted that there is a lack of long-term evidence for durability of effect, as well as long-term nutritional complications. The role of the procedure in the revisional setting was also noted to not be well defined with available literature at the time demonstrating that weight loss tended to be more modest in that setting with more complications documented [3].

Bile reflux and stomal cancer have been potential complications of the OAGB based on experience with the Billroth II (BII) reconstruction following subtotal gastrectomy, as well as the Mason “loop” gastric bypass [4,5,6,7]. The 2018 OAGB position statement found that the rate of biliary reflux appeared to be lower than expected, not exceeding 2% of all operated patients, and rate of gastric cancer did not appear to be reported more often than other bariatric surgery procedures. The IFSO 2018 taskforce therefore recommended that “bile reflux is either under reported or does not seem to be a major issue but remains a theoretical risk” [3].

OAGB appears extremely effective in reducing obesity related comorbidities, offering a good quality of life with an acceptable complication rate [8, 9]. The increasing numbers of OAGB in Europe and in Asia-Pacific recently brought this technique in third position in order of frequency, behind sleeve gastrectomy (SG) and Roux-en-Y gastric bypass (RYGB) [10, 11].

The 2020 task force undertook a new systematic review to summarize the evidence of the literature on the efficacy and safety of the OAGB procedure with the aim of providing an up-to-date information to guide practice.

Methods

Search Strategy and Quality Assessment

A systematic review of literature was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement [12]. A literature search was carried out in electronic databases (Cochrane, Embase, MEDLINE, Pubmed) in order to retrieve all papers related to OAGB in combination or not with RYGB and SG. The following search string was used: One Anastomosis Gastric Bypass or Mini Gastric Bypass or Single Anastomosis Gastric Bypass and Sleeve Gastrectomy or Gastric Bypass or Roux en Y Gastric Bypass and Bariatric Surgery. Two independent researchers (GP and GM) analyzed each article, first by title and abstract, and subsequently by the full text and extracted the relevant data. In case of disagreement a third researcher (MDL) was consulted. A manual search was conducted to identify further relevant studies. Papers that were not written in the English language, or without available full text, or letters to the editor were excluded. No time restriction has been addressed for the research and studies already addressed in the 2018 IFSO position statement were included. Both randomized and nonrandomized studies were included in the review for quantitative analysis, while review and meta-analysis were included for qualitative analysis. The studies were divided into the following categories: (A) OAGB vs another procedure or as a stand-alone procedure, (B) evaluation of long- and short-term complications of OAGB compared or not with RYGB or SG and (C) technical details of OAGB.

In order to reduce the risk of bias, the JADAD score [13] was used to assess the quality of randomized controlled trials (RCTs) and papers with a score of ≥ 3 were included in the analysis. The methodological quality of nonrandomized surgical studies was assessed with a MINORS score. A score ≥ 10 for noncomparative studies and ≥ 14 for comparative studies was fixed as a threshold for inclusion in the analysis [13, 14].

Inclusion Criteria

All types of study design were accepted. Only full text articles were included. For quantitative analysis studies with greater than 15 participants and with greater than 12 months follow-up were included.

Data Extraction

Two authors independently extracted data from the included studies using standardized electronic forms. A third author checked the extracted data for any errors and resolved disagreements between authors. Studies' information included year, degree of evidence of the study, group comparative, primary or revisional, study size, follow-up rate, demographics, technique, weight loss, comorbidities resolution (T2DM, HTN, OSAS, DS), and complications.

Results

Literature Search

The outcome of the search strategy is summarized in the PRISMA flow chart describing literature data screening process (Fig. 1).

Fig. 1
figure 1

PRISMA flow chart

Full size image

Identification

4296 articles were identified from search strategy (from 1946 to June 2020), 4259 from databases (Cochrane, Embase, Medline, PubMed) and 37 from other sources (manual search, gray literature).

Screening

3392 titles and abstract were screened after duplicates removal, and 355 full text were analyzed.

Eligibility

110 full texts articles were assessed for eligibility (244 excluded because not full text, letter to the editor, not in English language, not focused on bariatric/metabolic surgery and/or OAGB complications, insufficient follow-up, insufficient number of patients, and fraudulent article).

Included

110 total number of articles full text reading and included in the final analysis, 95 included in quantitative synthesis (randomized and nonrandomized studies) and 15 included in qualitative synthesis (review and meta-analysis).

The included studies were grouped according to the following categories: (1) OAGB vs another procedure or a stand-alone procedure (Table 1); (2) evaluation of long- and short-term complications of OAGB compared or not with other surgical techniques (Table 2); and (3) technical detail of OAGB (Table 3).

Table 1 Study data on OAGB weight loss and comorbidities
Full size table
Table 2 Study data on OAGB short- and long-term complications
Full size table
Table 3 Study data on the OAGB operative technique
Full size table

Overall Summary

A number of 110 good quality studies, 9 randomized and 86 nonrandomized quantitative studies and 15 reviews and meta-analysis of qualitative studies were analyzed. All of them reported number of patients > 15 and follow-up >12 months.

Regarding the 95 studies included in the quantitative analysis, 80 studies analyzed OAGB outcomes (39 focused on OAGB as primary procedure, 41 as revision/secondary procedure or mixed,) with a total of 23,341 patients; 7 of these studies were also included in the group focused on complications, for a total of 22 studies. The follow-up rate of the above mentioned studies were 86.4% at 1 year, 65.1% at 3 years, and 55.4% at 5 years.

The average preoperative body mass index (BMI) ranged from 25.3 to 67 kg/m2, with a mean study BMI of 44.52 ± 5.54 kg/m2.

In the 80 studies that reported data on weight loss, the average observed %EBMIL was 79.14 ± 14.8 (including revision operations), and 83.77 ± 13.41 (only primary operations) at a mean time of 3.2 ± 4.4 years.

In the 19 studies that reported diabetes remission the average observed remission was 75.8% ± 12.2 at a mean time of 2.9 ± 3.4 years. In the 13 studies that reported hypertension remission the average observed remission was 61.2% ± 13.3 at a mean time of 3.1 ± 3.4 years. In the 8 studies that reported dyslipidemia remission the average observed remission was 70.4% ± 8.4 at a mean time of 3.8 ± 2.8 years. In the 5 studies that reported OSAS remission the average observed remission was 79.9% ± 12,3 at a mean time of 3.9 ± 3.2 years.

Outcomes After OAGB

We summarized 9 randomized controlled trials, 17 prospective nonrandomized studies, 53 retrospective studies, and 1 case series study. Forty-one studies reported OAGB as primary procedure, 12 studies as revisional procedure, and 29 studies as mixed procedure.

Weight Loss (WL)

Data on weight loss are reported in Table 1.

Nine randomized controlled trials reported weight loss collecting 431 patients with “regular” primary OAGB, 40 patients with banded OAGB and 30 patients with “distal” OAGB. Mean follow-up was 25.33 months, mean calculated %EWL 67.85, mean calculated %EBMIL 87.54 and mean Change in BMI (ΔBMI) is 13.9.

Outcomes were reported after 1 year in three studies, with OAGB patients achieving %EWL of 66.9 ±23.7 (%EBMIL 65.7, ΔBMI 16.1) [98], 66.9 ± 10.9 [104] and 63.1 ± 8.7 (in this study, the “distal” OAGB group achieved %EWL 69.4 ± 15.4, %EBMIL 69.5, ΔBMI 16.0) [33].

Four studies reported data after 2 years of follow-up. OAGB patients achieved %EWL of 64.4 ± 8.8 (%EBMIL 83.3, ΔBMI 16.5) [37], %EBMIL 94.29 ± 23.63, ΔBMI 14.0 (“banded” OAGB, compared to %EBMIL of 77.90 ± 29.25 in “regular” OAGB, ΔBMI 12.3) [109], %EBMIL 87,9 [34] and %EWL 74.6 ± 11.8 [39].

One RCT reported %EWL 66.48 after 3 years of follow up [40].

After 5 years, one study reported %EBMIL 134.6, ΔBMI 6.9 [8].

Four prospective studies [41,42,43,44] reported data on 1713 patients who underwent a primary OAGB procedures. Mean follow-up was 31.1 months, mean calculated %EWL was 76.12, mean calculated %EBMIL was 73.15 and mean ΔBMI 14.4.

Twenty-four retrospective studies [24, 27, 30, 31, 45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64] reported data on 8531 primary OAGB patients. Mean follow-up was 27.9 months, mean calculated %EWL was 75.59, mean calculated %EBMIL was 83.41 and mean ΔBMI 14,2.

There are currently no RCT reporting on the outcomes of OAGB in the revisional setting .

One case series [65] reported %EMBIL 95.32 ± 24.91 after 18 months of follow-up and ΔBMI 15.2.

Two prospective studies reported data on 60 patients who underwent revisional OAGB. Mean follow up was 12 months; the reported %EBMIL varied from 64.8 (ΔBMI 9.6) [66] to 81.6 [67].

Ten retrospective studies [32, 35, 68,69,70,71,72,73,74,75] reported data on 711 revisional OAGB procedures. The mean follow-up was 18.5 months, mean %EWL was 71.85, mean %EBMIL was 65.03 and mean ΔBMI 10.6.

Data on weight loss from the studies reporting mixed procedures are available in Table 1 [1, 5, 36, 38, 76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96, 102, 103, 107, 108].

Type 2 Diabetes Mellitus (T2DM)

Diabetes or metabolic syndrome was reported upon as a comorbidity of interest in 52 papers, among them this outcome was reported in all RCT [8, 33, 34, 37, 39, 40, 98, 104, 109].

As reported in the previous statement [3], in the Lee trial comparing OAGB to RYGB, there was 100% resolution of the metabolic syndrome at 2 years [37]. At 12 months no other treatment for T2DM was required for 50% of T2DM participants in a study [98] and no other treatment was required for 84% of T2DM participants in another study [104]. At 5 years, 60% of participants with T2DM at baseline had a HbA1c < 6.5% without medications in the low BMI trial focusing on change in diabetes [8].

In the subsequent RCTs, T2DM remission rates at 1 year were available in 3 studies 86.63% [104], 64.3% [33], and 52.6% [109]. At 2 years follow-up, the remission rate was 95.7% and 60% [34]. At 3 years a remission rate of 89.6 % has been reported in one study [40].

In terms of progression of HbA1c, plasma values went down, reporting a value of % of HbA1C<7 in 90% of patients at 3 years in one study [53] and reduced by 2.3% over the course of 2 years in another study [34]. This significant reduction occurred in the subgroup of diabetics as well.

In the remaining studies, major improvement in T2DM management was reported (Table 1).

Hypertension (HTN)

HTN was reported upon as a comorbidity of interest in 36 papers, and among them this outcome was reported in 5 RCT [33, 40, 98, 104, 109].

Darabi et al. report HTN resolution in 2 out of 3 patients at 1 year of follow-up [98].

Seetharamaiah et al. report resolution of HTN at three months in 35,84% of affected patients [104].

Shivakumar et al. report a remission rate of 64.15% at 1 year, 67.31% at 2 years, and 74% at 3 years [40].

Nabil et al. report a resolution rate of 40% 1 year after OAGB-MGB [33], and Ospanov et al, RCT study, report remission rate of 60% at 2 year [109].

In 30 studies (no RCTs), improvement in HTN management was reported (Table 1).

Obstructive Sleep Apnea Syndrome (OSAS)

OSAS was addressed in 20 of the analyzed papers, but none of them was RCTs (Table 1).

Dyslipidemia (DL)

DL was addressed in 26 of the analyzed papers, including 3 RCTs [33, 34, 98].

Darabi et al. report a remission of DL in 4 of the nine-patient affected during the follow-up of 1 year [98].

Nabil et al. report DL remission in 82.6% of the subjects after 1 year [33].

In 22 studies (no RCTs), major improvement in DL management was reported (Table 1).

Quantitative Analysis of Complications (Table 1–Table 2)

Complications reported in the studies focused on outcomes are shown in Table 1.

Complications were divided into early complications (≤30 days after surgery) and late complications (>30 days after surgery).

Early complications included anastomotic leak, wound infection, hematomas, hemorrhage, anastomotic stricture, and organ perforation/peritonitis. A total of 1082 early complications were reported among 19,051 OAGB procedures, with an overall incidence of 5.51% and a reoperation rate of 1.01%.

Primary procedures incidence rate 4.91%, revisional procedures incidence rate 7.9%.

Eleven cases of early death (overall incidence of 0.056%) were reported. Four of them for pulmonary embolism. Three cases for complications related to leak after revisional OAGB.

Late complications included marginal ulcers, bowel obstruction, malnutrition, and gastroesophageal reflux including biliary reflux. A total of 1025 late complications were described among 18763 OAGB procedures, with an overall incidence of 5.46% and a reoperation rate of 1.34%.

Primary procedures incidence rate 6.30%, revisional procedures incidence rate 5.58%.

A total of 25.76% of total reoperations were conversions to RYGB.

In 18.34% of the cases, late reoperations were required for bile reflux and 36.46% for malnutrition.

Eight cases of late death (overall incidence of 0.042%) were reported.

Table 2 reports studies focused on complications specifically related to OAGB such as reasons for readmission [15, 16, 20, 25, 32], malnutrition, anemia, and bile reflux.

Ten studies reported malnutrition as a direct effect of OAGB [17, 19, 24, 27, 29, 30, 33, 34, 110, 111]. Malnutrition appears to be the main cause of revisional surgery after OAGB [17, 19], and may be worse with the OAGB as compared to sleeve gastrectomy or RYGB [24].

Five studies [29, 30, 33, 110, 111] reported data on different biliopancreatic limb lengths and their effect on malnutrition and nutritional deficiencies. Two studies, one RCT consisting of 60 patients, [33], and one retrospective study consisting of 101 patients [110], confirmed a correlation between BPL length and the incidence of malnutrition. Another study that collected data on 36,952 individuals found that 92.3% of reoperations for malnutrition were associated with BPL length of over 200 cm [111].

Along the same lines, another analysis found a prevalence of 3.7% of hypoalbuminemia in patients with a BPL of 200 cm [29]. Interestingly, age > 40 years and low preoperative levels of plasma albumin (<4.0 g/dl) may worsen this condition [27]. Similarly, a fixed BPL of 200cm can be accompanied by hypovitaminosis, particularly vitamin A and vitamin D3 [30].

Regarding anemia [18, 31], one study [31] compared OAGB with RYGB, demonstrating a nonstatistically significant higher risk of being anemic following OAGB compared to RYGB (16.6% vs 12.7% after 2 years, p<0.55).

Several studies reported data on bile reflux [21, 22, 28, 91, 103, 107, 108], a major cause of revision after OAGB [17] although one study demonstrated no difference in histologically proven bile reflux between OAGB and RYGB [28].

Operative Technique (Table 1–Table 3)

The description of operative technique was reported in 28 of the analyzed studies, including 5 RCTs.

In order to standardize the review, we analyzed 3 steps of OAGB that may impact on the weight loss and potential intra and post-operative complications, in particular:

  • Pouch and bougie size

  • Gastro-jejunostomy fashioning

  • Limb length

Pouch and bougie Size

Pouch and bougie size have been addressed in 22 studies, including 3 RCTs [34, 40, 109]. In the RCTs, the reported bougie size was either 36 [40, 109] or 37 French [34]. In the cohort studies, most teams use a 36 French bougie, but sizes ranged from 28 French [58] to 42 French [75]. Concerning the size of the pouch, there is a great disparity in reporting both in RCTs and in cohort studies. In the RCTs, only 2 report on the pouch size: one in terms of pouch length, i.e., 20 cm (n=180) [109], and the other in terms of pouch volume, i.e., 50–60 cc, (n=80) [104]. In the cohort studies, when mentioned, the pouch is mostly sized using anatomical landmarks (incisura angularis or crow’s foot and Angle of His).

Gastro-jejunostomy

The gastro-jejunostomy technique has been reported in 23 of the analyzed studies, including 4 RCTs [34, 40, 104, 109]. In the RCTs the technique of gastro-jejunostomy was mainly mechanical using the linear stapler [34, 40, 109]. Stapler load size (45 mm) was reported in one study with 101 patients [104].

Limb Length

Limb length has been reported in 25 of the quantitative analyzed studies, including 3 RCTs [34, 40, 109]. In the RCTs, the length of the BPL was 200 cm in 227 patients [34, 109], 150 to 180 cm in 101 individuals [104], and depending on total bowel length in 180 individuals, which translated in an average BPL length of 279 cm [30, 46]. In the remaining literature there is no homogeneity regarding this item.

Qualitative Analysis

The list and description of the selected review articles and metanalysis can be found in Table 4.

Table 4 OAGB qualitative analysis
Full size table

In 3 metanalyses [112,113,114] comparing a total of 12,866 OAGB patients versus 8804 RYGB, weight loss and T2DM remission at 1, 2, and 5 years were significantly better for OAGB. One study [112] found more malnutrition after OAGB and more intestinal obstructions after RYGB. No other significant differences in outcomes were registered in these 3 meta-analyses.

In one meta-analysis comparing a total of 1998 OAGB patients versus 1864 LSG patients, weight loss (EWL) at 1 and 5 years, as well as T2DM remission and HTN remission, was significantly better in the OAGB cohort [115].

In one meta-analysis observing the outcomes in up to 12 years follow-up in 12,807 OAGB patients, EWL at 5 years was 76.6%, and T2DM remission was obtained in 83.7% and HTN remission in 66.94%. Marginal ulcer rate was 2.7%, anemia rate 7.0%, and 0.71% developed malnutrition. Approximately 2.0% of patients reported postoperative gastroesophageal reflux [112].

In a systematic review involving 318 superobese (BMI ≥ 50) patients, early mortality was 0.31% (1 patient) and major complications were 2.2 % (7 patients). The leak rate was 0%, and the mean% EWL at 60 months was 90.75% [116].

In a systematic review of 69 publications (4 RCTs, 11 review articles, 54 clinical studies) with a total of over 38,000 patients, OAGB was demonstrated to have a short operative time, low complication rate, and excellent weight loss outcomes. The longer-term issues of nutritional deficiencies and bile reflux could not be addressed due to a paucity of long-term follow up data. [117].

In a network meta-analysis of 25 eligible RCTs, covering nonsurgical treatments and 8 different surgical procedures, including 1211 patients, it appeared that BPD and OAGB achieved higher T2DM remission rates than other bariatric procedures. However, the trials regarding OAGB and BPD were in the minority of the RCTs included in the meta-analysis (2 trials and 1 trial, respectively). Moreover, the number of patients included in studies of these procedures was small [118].

Discussion

The analysis of quantitative and qualitative studies demonstrates the effectiveness of the OAGB as a weight loss procedure. The magnitude of weight loss appears to be at least equivalent to RYGB and potentially superior to SG and GB. Quantitative analysis demonstrated that weight loss was greater for patients who had OAGB as their primary operation when compared to patients who had the OAGB as revision surgery after SG and GB. There appear to be differences in terms of weight loss in relation to the different surgical techniques used, particularly in relation to the length of the BPL, but this latter finding cannot be affirmed with certainty because of the limited number of studies focusing on this particular detail.

OAGB appears to have favorable effects on T2DM remission, as detected in the quantitative analysis, as well as in the qualitative analysis, both in short-term and in medium-term follow-up studies, i.e., after 24 months and 60 months, respectively.

The effects of OAGB on HTN, OSAS and dyslipidemia also seemed favorable, but the data are too few and follow up is too short to be conclusive.

The current literature suggests that OAGB is a safe procedure. The percentage of early perioperative complications is low, even in comparison with RYGB. Major perioperative complications, such as need for blood transfusion, return to the operating room and/or prolonged length of hospital stay, are seen in 2-3% of the patients and perioperative mortality is <0.5%.

There are only a few studies with long-term (10 years) follow-up reporting on long-term outcomes. Some studies do report medium (5 years) follow-up, but in the majority of the studies follow up is limited to <5 years. Given the short-term nutritional outcomes and the anatomical changes that are made with this operation, it will be important in the future to document the risk of protein malnutrition, anemia and hypovitaminosis. On the basis of current knowledge, the impact of factors such as the BPL length and patient starting BMI will be important variables to consider.

Bile reflux is described as a reason for conversion from OAGB to RYGB in several studies, however this does not seem to be a frequent problem in the quantitative and qualitative studies analyzed.

In the quantitative studies we reviewed, no cases of cancer were reported, although this may represent inadequate endoscopic follow up and the lag time between the procedure and the cancer developing. It is hoped that the recent IFSO position statement recommending 2–3 yearly gastroscopies following OAGB will provide the opportunity for early detection should there be an issue, as well as better data in the future to help address this important issue [119].

It’s interesting to compare the findings of this current systematic review to a recent IFSO publication presenting data gathered from 101 experts on OAGB from 39 countries who voted on 55 statements in areas of controversy or variation associated with OAGB. In this study, all of the experts felt that OAGB was an acceptable mainstream surgical option. Approximately 94% of the experts felt that the construction of the gastric pouch should start in the horizontal portion of the lesser curvature. There was a consensus of 82, 84, and 85% for routinely supplementing iron, vitamin B12, and vitamin D, respectively [26].

Recommendation of the IFSO OAGB Taskforce

  1. 1.

    OAGB should remain the official IFSO-approved identifier for this procedure. MGB and OAGB may be different operations in the mind of some authors, but both operations are similar as they both have a single gastro-jejunal anastomosis and furthermore, most studies report on an operation that is not a pure MGB or a pure OAGB, but rather a mixed technique. We have learned from other procedures (e.g., RYGB, BPD, DS) that it is essential to define each type of intervention under a single name. For these reasons, IFSO selected OAGB as the approved identifier for this procedure.

  2. 2.

    The outcomes from OAGB are promising in terms of short operative time, low perioperative complication rate, good weight loss and good comorbidities remission (T2DM, HTN, OSAS, and DL) and appear at least equivalent to other bariatric surgery procedures.

  3. 3.

    OAGB in the primary setting provides better weight loss, comorbidity reduction and fewer complications when compared to the outcomes when OAGB is performed as a revisional procedure. Patients should be aware of these differences if undertaking OAGB as revisional procedure.

  4. 4.

    At this stage, bile reflux does not seem to be a major issue for patients who have undergone OAGB and there have not been increased reports of esophageal or gastric cancer. Due to the risk of under reporting and the time lag for carcinogenesis following OAGB, we recommend that patients should remain under the care of their multidisciplinary bariatric team and have regular endoscopic examinations as per the IFSO position statement on endoscopy [119].

  5. 5.

    While the data is currently lacking, long-term nutritional deficiencies due to the malabsorptive nature of OAGB procedures should be considered and patients should have at least an annual nutritional review and appropriate micro- and macronutrient supplementation. Long-term observational studies should be undertaken, potentially using national registries, to better understand nutritional requirements. Given that BPL length appears to be an important variable, RCTs are warranted to address this important issue.

  6. 6.

    The majority of the studies included in this review are case studies with <5 years follow up. In order to increase the body of evidence, patients should be encouraged to remain in long-term multidisciplinary care, and clinics should be encouraged to participate in registries, as well as longer-term follow-up studies, and possibly RCTs.