Laparoscopic One-Anastomosis Gastric Bypass: Technique, Results, and Long-Term Follow-Up in 1200 Patients

Candidates were evaluated by a multidisciplinary medical unit and underwent preoperative psychological, nutritional, and comprehensive medical evaluations. A thorough assessment was performed of their general condition, associated illnesses, risk factors, mental status, motivations for BS, and ability to adhere to a postoperative regimen. Biochemical and radiological studies (abdominal ultrasound, chest x-ray, upper GI series), as well as endocrine and cardiopulmonary assessment, were performed. Supplementary tests related to existing comorbidities were carried out accordingly. Patients underwent respiratory physiotherapy and were encouraged to do physical exercise. In order to lose at least 10 % (20 % for BMI >50 kg/m²) of their EW preoperatively, they were submitted to 12 days of an 800-cal/day pure high-protein diet with micronutrients and vitamins (Vegefast-Complet®-Vegenat, Spain) followed by a complete liquid diet 5 to 7 days prior to operation. This was part of a prospective randomized controlled trial (RCT) comparing Vegefast-Complet® to a natural high-protein diet performed at our center [11]. Antithrombotic and antibiotic prophylaxis were given to all patients.

All procedures were performed by a single surgeon (MC) and essentially the same surgical team. The patient is placed in a modified lithotomy position with the surgeon standing between the patient’s legs, the camera operator on the right side, and an assistant on the left. Six trocars are normally used (Fig. 1). OAGB is diagrammatically shown in Fig. 2. The first step consists of locating the Treitz ligament and measuring the jejunal loop to be bypassed which initially was ∼200 cm. After publishing results of our first 209 patients [5], we started to measure the whole small bowel (SB) routinely (Treitz to ileocecal valve), to determine the extent of bypassed SB (afferent limb) and common channel (efferent limb); we select the midportion and thus their lengths are usually similar (from ∼250 to 350 cm). For increasing BMIs, we add 10–50 cm of bypassed SB (with no specific formula) but always maintain at least ∼250–300 cm of common channel. Therefore, the extent of bypassed SB was based on total SB length and BMI for the last ∼1000 patients in this series. The patient is then placed in a 30° anti-Trendelenburg position. The angle of His is identified and the fat pad at the esophago-gastric (EG) junction explicitly dissected in order to visualize the diaphragm’s left crus for optimal endostapler positioning at this critical location. With associated hiatal hernia (very common in MO), this step includes section of peri-esophageal adhesions and phreno-esophageal ligament to reduce the hernia. Hiatal closure is performed selectively (marked enlargement). An automated camera-holding system (Lap Man®-Medsys, Belgium) is then installed and operated through a laser remote control (Lapstick®-Medsys, Belgium). Ultrasound shears (Autosonix®-Covidien, USA) are used to section the lesser gastric curvature’s fat and blood vessels at the level of the crow’s foot to enter the lesser sac. An endoscopic stapler loaded with a 45-mm/3.5-mm cartridge (Endo-GIA®-Covidien, USA) is inserted through the created opening and applied, sectioning the stomach horizontally. A 36-Fr double-lumen orogastric tube (Ref 340.36®, Vygon, France) is inserted to calibrate the gastric reservoir. Fatty tissue and fibrous adhesions between the posterior gastric wall and pancreas are dissected. Then, an endoscopic stapler loaded with 60-mm/3.5-mm cartridges (Endo-GIA®-Covidien, USA) is consecutively applied (usually three times), sectioning the stomach vertically and completing the gastric reservoir. The latter should be long, narrow, well vascularized, and easy to move caudally. The orogastric tube is removed and the previously chosen SB mobilized upward placing it without tension in an antecolic, antegastric position. Bivalving of omentum is especially required in superobese patients and those with central obesity. A continuous reabsorbable no. 2-0 suture (Polisorb®-Covidien, USA) is sewn in a latero-lateral position, securing the SB loop to the gastric reservoir’s staple line along ∼8–10 cm. Enterotomy and gastrotomy (distal reservoir) are done with ultrasound shears. An endoscopic stapler loaded with a 30-mm/3.5-mm cartridge (Endo-GIA®-Covidien, USA) is partially inserted (∼75 %) and applied between both, thus creating a gastroenteric anastomosis ∼2–2.5 cm long. Incisions on the anterior anastomotic wall are sutured with reabsorbable no. 2-0 (Polisorb®-Covidien, USA) interrupted stitches. These are also used to fix the (afferent) biliopancreatic (BP) SB loop in an upward direction (∼8–10 cm) to the excluded stomach, and also the common (efferent) SB loop to the excluded gastric antrum. These measures unload anastomotic tension, improve its orientation, and reinforce the antireflux mechanism. Anastomosis integrity is verified with a pneumatic test. Fibrin glue (Tissucol®-Baxter, USA) or surgical glue (Glubran 2 ®-Gem, Italy) is applied to the anastomotic surface and stapled regions of the gastric reservoir and remnant, and the greater omentum is tucked and adhered to them. Lastly, a penrose drain is positioned under the left hepatic lobe and brought out through the 5-mm right subcostal incision.

An upper GI hydrosoluble contrast swallow (Gastrografin®-Bracco Diagnostics, Canada) was routinely performed the morning after to verify patency, rule out leaks, and provide a baseline postoperative “map.” The urine catheter was removed and the patient actively mobilized. A clear liquid diet was then started with swallows of ∼30 ml. The patients usually tolerated this regimen well, their drain was removed, and they were discharged ∼24 h postoperatively with specific indications regarding diet, activities, and medications. The drain was left in place and removed at the first office visit only in patients with unusual operative bleeding, a higher risk for postoperative bleeding (i.e., liver failure), and complex operative cases (i.e., reoperations and revisional BS).

Postoperative FU office visits for clinical and complete biochemical evaluation including macro- and micronutrients were scheduled 3, 6, 12, 18, and 24 months after BS, and yearly thereafter. In between these time frames, we kept close contact through telephone or e-mail. Due to our described referral pattern, we strove to keep an updated web page which included a blog and forum where patients could retrieve all kinds of information and maintain communication with any member of our team and with other patients as well. In regard to our electronic reports for our “remote” FU, concerning measurements and WL, we have tried to be as objective as possible and thus these include front and side full-length photographs with waist and hip measurements, and weight (either from an electronic scale report with name or a patient’s photo on a bathroom scale); moreover, the electronic FU report contains all information which would be evaluated during an in-office consultation including GI status, side effects and development of complications, eating and exercise habits, use of medications and supplements, and complete biochemical reports.

The patients followed a nutritional protocol designed at our center. Intestinal adaptation was usually completed within the first 3 months; however, the process could take up to 6 months or more, especially in cases of frequent dietary transgressions. From the 6th month onwards, the patients usually tolerated all kinds of food. They were administered proton pump inhibitors (PPIs) and sucralfate daily (first month), high-protein powder and calcium (first 3 months), and a daily mineral-vitamin supplement (first year), with maintenance doses recommended for life. Specific mineral and/or nutritional supplements were prescribed as needed for documented deficiencies (some starting even in the preoperative period).

Postoperative endoscopic (control) studies were planned for all patients completing a 5-year FU as well as for those referring persistent upper GI symptoms. Along with preoperative psychological evaluations, a QoL baseline assessment was applied using the Impact of Weight on QoL (IWQoL) survey [12]; this latter was repeated at periodic postoperative office visits and/or included in the electronic “remote” FU report.

There were 744 female (62 %) and 456 male (38 %) patients with a mean age of 43 years (range, 12–74). Mean preoperative BMI was 46 kg/m² (range, 33–86) and mean preoperative EW was 65 kg (range, 34–220). Our cohort was composed of different subgroups which included 697 patients (58 %) with no previous or simultaneous abdominal operations (subgroup 1). Another 273 (23 %) had had prior open abdominal operations and thus required adhesiolysis of variable complexity, and a total of 203 (17 %) had abdominal operations performed simultaneously, particularly gallbladder removal and/or hiatal or ventral hernia repairs (subgroup 2). Finally, in 27 (2 %), laparoscopic OAGB was performed as a revision of other (failed) bariatric procedures (subgroup 3) including previous laparoscopic gastric bands (n = 13), as well as open vertical banded gastroplasties (n = 14). Perioperative results are shown in Table 1. Operative time (OT) varied among different subgroups. Mean length of stay (LOS) was significantly lower for patients without complications. Other results demonstrating quick patient recovery are outlined, as well as a summary of overall morbidity and mortality (further specified in Table 2). Noteworthy is the fact that we did not reoperate or have a record of any reoperation elsewhere for failure of the procedure.

Complications and side effects are depicted in Table 2. Conversions to an open approach due to intraoperative problems and open reoperations due to immediate early complications were all performed through a left subcostal mini-laparotomy (LSML); this has been described elsewhere [13], and we utilized it routinely in the open bariatric era [14]. Intraoperative complications requiring conversion to an open approach occurred in four patients (0.3 %). Intra-abdominal hemorrhage was not adequately controlled by laparoscopic means in two; bleeding was from a gastric reservoir’s artery in one and from a ruptured vein in the posterior gastric wall in the other. The remaining two cases included one EG junction perforation by the calibration tube and one incorrect gastric transection in a patient with severe cardio-esophageal inflammation; both were related to technical difficulty during our learning curve and required conversion to distal Roux-en-Y gastric bypass (RYGB) with esophago-ileal anastomosis.

Early major complications requiring reoperations occurred in 16 patients (1.3 %) and included intra-abdominal bleeding (9), leaks (3), and early SB obstruction (SBO) (2). Rare complications included necrosis of excluded anterior gastric wall in one patient and acute dilation of excluded stomach in another. The former was in a patient with a BMI >70 who required a tracheostomy for intubation; the tube was displaced and severe abdominal pain developed few hours after the operation. Re-laparoscopy disclosed extensive necrosis of excluded stomach which was resected through LSML. In the other patient, excluded stomach and afferent limb were completely filled with BP secretion. Re-laparoscopy ruled out SBO and verified anastomosis integrity. Decompression was obtained through a nasogastric tube positioned into the afferent loop (removed upon the patient’s discharge) and a gastrostomy tube (removed on an ambulatory basis). Successful conservative treatment of early major complications was achieved in 12 patients (1 %), including leaks (10), acute pancreatitis (1), and infected hematoma (1).

Sporadic clinical reflux was reported by 26 patients (2 %). The few episodes were associated with dietary transgressions (always at night). Endoscopic studies revealed the presence of some bile in the stomach with mild to moderate pouch gastritis, but did not document any esophagitis. Treatment included dietary and healthy life recommendations, continued FU by our nutritionists, PPIs (40 mg/day for 6 months), and sucralfate (1 g before every meal and before bedtime for 3 months, followed by 1 g before bedtime for another 3 months). An upper GI endoscopy was ordered afterwards; all patients healed or significantly improved. Depending on their course, FU endoscopic studies were performed annually in some patients. Moreover, at completion of a 5-year FU, we planned screening endoscopic studies for all patients and were able to obtain them in 265 (22 %). Most other patients were completely asymptomatic and did not want the study. There were no significant findings; importantly, there were no cases of esophageal reflux and/or esophagitis or signs of acute or chronic stomal or MU. Mild to moderate pouch gastritis was found in 21 (8 %) and presence of HP in 9 (3.4 %). These patients were treated as those with clinical reflux (see above) and HP eradication when needed.

Preoperative nutritional deficits were found in some patients including iron (∼10 %), vitamin D (∼15 %), and calcium (∼4 %). After OAGB, a few patients developed excessive WL and/or nutrient deficits (usually within the first 2–3 postoperative years). Table 3 depicts the percentage of patients with specific deficiencies (values below normal range) at different points in time during our FU. Most were controlled and treated on an ambulatory basis and recovered with dietary recommendations and once intestinal adaptation was complete. However, a total of 14 patients (1.2 %) required further treatment for hypoalbuminemia; all received high-protein enteral supplements and pancreatic enzymes (Kreon®-Abbott, Germany) 10,000–25,000 IU with each meal during at least 3 to 6 months; 2 were readmitted and managed with IV albumin. Iron deficiency was rather common, especially in fertile women with copious menstrual bleeding. Up to one third required oral supplements beyond the expected time for intestinal adaptation, and 15 patients (1.3 %) required parenteral iron. Among liposoluble vitamins, vitamin D insufficiency was present in more than half of our patients at 3 years and one third in the long term; this required continuous supplementation in ∼20 % of them. Longer supplementation was also needed for vitamins A and K in ∼3 and 0.5 %, respectively. Deficits in hydrosoluble vitamins were basically found in B₉ (folic acid) and B₁₂; supplements were needed in ∼15 % at 3 years and 2 % in the longer term. An initial higher rate found in B₉ (Table 3) was probably a reflection of preoperative deficiency. Calcium deficit was found in ∼8 % during the first 2 years and decreased thereafter, but persisted in ∼2 % in the longer term; supplements were especially recommended to post-menopausal women. Zinc and copper were needed in ∼5 and 3 % in the long term only in women. Specific long-term phosphorus, magnesium, and manganese supplementation has not been needed.

As in other malabsorptive procedures, soft stools, increased bowel gas, and a bad fecal odor were present in most patients, especially those consuming fatty and pure carbohydrate foods. Bismuth salts, activated carbon, and simeticone usually controlled symptoms well, and they usually improved progressively with intestinal adaptation. These nutritional complications and side effects were not significantly different when comparing our first 209 patients [5] to the more malabsorptive group of patients operated on thereafter (see “Operative Technique”).

The 30-day readmission rate was 0.8 % (10 patients). Two had SBO and were reoperated on. All others were successfully treated conservatively during 24–48 h; 6 had persistent nausea and vomiting, 1 had a single episode of hematemesis and had a stress ulcer at endoscopy, and 1 had an abstinence syndrome from psychiatric medication with extreme anxiety. Late readmissions were required in 13 patients (1 %) for stomal stenosis (6), GI bleeding due to MU (5), and malnourishment (2). Two patients died in this series (0.16 %); both had superobesity, multiple comorbidities, and risk factors. One suffered a pulmonary thromboembolism 3 days after BS (without warning symptoms or additional postoperative complications). The other suffered gastric wall necrosis, was reoperated on, and developed refractory nosocomial pneumonia. Both deaths occurred during the initial part of our series [9], and there were no other casualties in >1000 patients operated on thereafter.

Table 4 depicts the number of patients operated on per year; a progressive increase was observed from 2002 to 2008. Also shown is accrual of patients for FU either as in-office visits or through electronic means. Although (as expected) there is a gradual decrease in number available for this long-term FU, at least 50 % of those operated on >12 years ago were being followed up, and we had information from 7 out of 10 among the whole group of 1200 patients. Our “remote” electronic FU reports were as objective as possible (see “Patients and Methods”); overall, outcomes for this group were not significantly different from those found in patients seen directly at our office. During this FU period, 15 (1.25 %) patients died from unrelated causes; these included aviation and auto accidents (4), lung cancer (2), surgical complications from unrelated operations several years after OAGB (2), family abandonment and suicide (2), myocardial infarction 5 years after OAGB (1), complicated appendicitis (1), complicated Guillain-Barre syndrome (1), unspecified autoimmune disease (1), and prostate cancer (1).

Depending on initial (preoperative) EW, patients lost a mean of ∼15–20 kg in the first month and ∼30–40 kg in the first trimester. Table 5 outlines evolution of WL expressed in various forms at different point intervals. The number and percentage of patients followed up at each time interval are included; only from 13 % (at 6 years) to 30 % (at 12 years) of our cumulative number of patients were lost for FU. Substantial WL was documented for most patients; through time, there was a slight weight increase in a few, which was not clinically relevant. Thus, EWL was maintained in most of our patients and according to Reinhold’s classification our results ranged from good (EWL >50 %) to excellent (EWL >75 %), and a long-term successful treatment (EWL >50 %) was achieved in almost all patients.

The effect of OAGB on comorbidities is shown in Table 6. The number of patients carrying each related disease preoperatively is included and varied widely. Figures in the columns “remission” and “improvement” represent what was recorded at the last patient evaluation (either as in-office or electronic FU). Severe metabolic comorbidities such as type II diabetes mellitus, insulin resistance, hypertension, and sleep apnea either totally resolved or substantially improved (most from the first day after BS). Remission was also demonstrated in most patients for other metabolic conditions like hyperlipidemia and liver steatosis when the first biochemical tests were ordered at the 3rd postoperative month. Benefits were also evident for patients with “mechanical” complications related to MO (osteoarthritis, urinary incontinence, and respiratory insufficiency). Noteworthy is the fact that 53 % of our patients had gastroesophageal reflux disease (GERD) of some degree and all were relieved after the operation. Significant improvement in QoL and in all but one IWQoL survey parameter was found 3 months after BS. The item “comfort with food” also improved markedly but only 6 months after the operation; moreover, problems with it were referred to as “insignificant” at the 1-year office visit, when no patients reported food intolerance of any kind. This improvement in QoL remained through time.

Advantages ascribed to the procedure [4, 42‐44] were evident from the perioperative period in our patients (Table 1). Recovery was usually quick and uneventful with rapid ambulation, flatus passage, and almost no need for analgesics; this led to a reduced LOS, which for our non-complicated cases (97 %) has been one of the shortest reported thus far [42, 44]. Although OAGB provides reduced difficulty compared to complex procedures, we believe it should not be promoted as a very simple, easy, and rapid operation. Our OT varied for each subgroup (Table 1), and was similar to other reports [42], but definitely fell short compared to the original MGB’s “thirty minute case” [4, 21]. Besides the fact that speediness is not among our goals, several modifications in OAGB prove time consuming.

As others [21‐26, 42‐44], we found OAGB to be a rather safe operation. There were 32 (2.7 %) early and 12 (1 %) late major complications (Table 2). Almost 70 % of the former were treated conservatively or through re-laparoscopy, and all late complications were treated conservatively. Specific analysis led to interesting findings.

Poor FU is definitely the “Achilles heel” in all BS. Since most patients came from different parts of our country, direct in-office consultation is difficult to accomplish. Nonetheless, we have been able to physically see >1/2 and >1/5 of our patients, 6 and 12 years after the operation, respectively (Table 4). Since our referral pattern was evident to us from the beginning, we strove to keep close contact with our patients through telephone or electronic means; also, “remote” reports have been as objective as possible and include all information that would be obtained directly at the office including photographs, questionnaires, and full biochemical tests. Certification by the IFSO-EAC as a Centre of Excellence and their rigorous and increasing requisites for recertification have validated our methods and continually support and guide our efforts to improve overall FU. The IFSO-EAC database indeed serves as the registry some colleagues have strongly recommended in order to have better insight in regard to this operation [47]. With all these resources, we have been able to gather information from ∼70 % of our patients through this long-term FU (Table 4).

A highly criticized concept, when initially proposed, is steadily being accepted worldwide. Centers of more than 25 countries are now performing it [81]. This has brought about a controversy regarding the name for the procedure, which has led to unfortunate confusions [82]. We recently presented our arguments on why we support views of others in considering OAGB as the only standing alternative name for MGB, in order to reconcile terms and facilitate issues in publishing future related courses and communications [83]. This was further supported at the 2015 IFSO meeting in Vienna where the academic MGB-OAGB International Club was founded [84]. We call again on the various bariatric teams that are performing the original MGB or our modified version, the OAGB, to aid in the dissemination and acceptance of this procedure by presenting and publishing their experiences and standardizing the name (to MGB/OAGB), in order for all of us to be recognized as a whole. Now that many of its controversies are being surpassed and the bariatric surgical community is accepting the procedure as a rational alternative in the bariatric repertoire, we should make all efforts in order to conciliate in regard to the name, avoid new disagreements, and work towards making MGB/OAGB mainstream in obesity and metabolic surgery.