Medwrite Portfolio_20 Surgery Sample Articles

J. Ponce / Surgery for Obesity and Related Diseases 5 (2009) 104 –110 109introduced into the American market quite recently (Sep- Acknowledgmentstember 2007 [18]), it would be premature to compare it tothe Lap-Band in the American setting. The author would like to thank Medwrite Medical Com- munications (Maiden Rock, WI) for its substantial contri- In my own experience with the Lap-Band and Realize bution to the development of this report. Ethicon Endo-Band, both bands require careful placement technique and Surgery sponsored their services. The sponsor had no role inpostoperative management to ensure weight loss success the design, preparation, or approval of the manuscript.and minimal complications. Gastric prolapse secondary toband displacement used to be the most common surgical Disclosurescomplication seen in our practice. Now, with better tech-nique and careful follow-up in association with the wider J. Ponce is a consultant, speaker, and proctor for, andlow-pressure bands, which may increase stability, slippage has received research funds from, Ethicon Endo-Surgeryhas become more infrequent. The most important improve- and Allergan and uses both the Realize Band and the Lap-ments made to the Lap-Band have been the 360° balloon Band in his private practice.circumference, wider area of contact with the gastric wall,and low basal pressures. Both bands are now very stable, Appendixwith a wide range of adjustability and good circumferentialrestriction. Supplementary data The effectiveness and rate of complications for both The videos associated with this article can be found, inbands have been very similar in our experience, as well as the online version, at www.SOARD.org under “Multimediain other studies. Eight studies have directly compared the Library.”SAGB and Lap-Band [5,53–59]; 1 of these included a pro-spective comparison of the short-term SAGB/Lap-Band Referencesweight loss and complications and found no statisticallysignificant differences between these 2 bands [56]. Suter et [1] National Institutes of Health. NIH Consensus Development Confer-al. [15] described a prospective, randomized comparison of ence statement: gastrointestinal surgery for severe obesity. Obes Surgearly SAGB/Lap-Band results, in which they also discov- 1991;1:243–56.ered no difference in weight loss, food tolerance, or qualityof life at 18 months, and the same rates of late morbidity and [2] Fried M, Miller K, Kormanova K. Literature review of comparativereoperation. Each of the SAGB/Lap-Band comparative studies of complications with Swedish band and Lap-Band. Obesstudies independently concluded band comparability. Surg 2004;14:256 – 60. Important differences in the surgical technique required [3] Chapman AE, Kiroff G, Game P, et al. Laparoscopic adjustablefor use with the Realize Band versus the Lap-Band include gastric banding in the treatment of obesity: a systematic literature(1) band priming, (2) band passage through the posterior review. Surgery 2004;135:326 –51.tunnel, (3) band locking, (4) gastric plication, and (5) portplacement. First, the Realize Band requires leaving the [4] O’Brien PE, Dixon JB. Lap-Band: outcomes and results. J Laparoen-balloon completely unfilled with air or saline to prevent doscopic Adv Surg Tech 2003;13:265–70.early obstruction and allow for appropriate band passagethrough the posterior tunnel. Second, the Realize Band is [5] Steffen R, Biertho L, Ricklin T, Piec G, Horber FF. Laparoscopicpassed using a suture loop attached to the distal tab, in contrast Swedish Adjustable Gastric banding: a five-year prospective study.to the tubing used for the Lap-Band. Third, the Realize Band Obes Surg 2003;13:404 –11.is locked using the tabs. In contrast, the Lap-Band requires thatthe tubing be passed through the buckle to lock it. Fourth, the [6] Vertruyen M. Experience with Lap-band System up to 7 years. Obesplication tunnel for the Realize Band is wider to allow enough Surg 2002;12:569 –72.space for band inflation without causing any tension on thesutures. The final difference is in port placement. The Realize [7] Belachew M, Belva PH, Desaive C. Long-term results of laparoscopicport can be placed with a sutureless port fixation system, as adjustable gastric banding for morbid obesity. Obes Surg 2002;12:described in the surgical technique section. 564 – 8.Conclusion [8] O’Brien PE, Dixon JB, Brown W, et al. The laparoscopic adjustable gastric band (Lap-Band): a prospective study of medium-term effects The pars flaccida technique using the Realize Band and on weight, health and quality of life. Obes Surg 2002;12:652– 6.port placement has been standardized, is simple and repro-ducible, and can be performed within a short time. Adjust- [9] Maggard M, Shugarman LR, Suttorp M, et al. Meta-analysis: surgicalments need to be done frequently and gradually to achieve treatment of obesity. Ann Intern Med 2005;142:547–59.and maintain satiety and weight loss. [10] O’Brien PE, McPhail T, Chaston TB, Dixon JB. Systematic review of medium-term weight loss after bariatric operations. Obes Surg 2006; 16:1032– 4. [11] Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a systematic review and meta-analysis. JAMA 2004;292:1724 –37. [12] Ponce J, Haynes B, Paynter S, et al. Effect of Lap-Band-induced weight loss on type 2 diabetes mellitus and hypertension. Obes Surg 2004;14:1335– 42. [13] Frigg A, Peterli R, Peters T, et al. Reduction in co-morbidities 4 years after laparoscopic adjustable gastric banding. Obes Surg 2004;14: 216 –23. [14] Hell E, Miller KA, Moorehead MK, Samuels N. Evaluation of health status and quality of life after bariatric surgery: comparison of stan-

110 J. Ponce / Surgery for Obesity and Related Diseases 5 (2009) 104 –110 dard Roux-en-Y gastric bypass, vertical banded gastroplasty and [36] Weiner R, Wagner D, Blanco-Engert R, et al. Eine neue Technik zur laparoscopic adjustable silicone gastric banding. Obes Surg 2000;10: laparoskopischen Plazierung des steuerbaren Magenbandes (LAP- 214 –9. Band) zur Vermeidung eines Slippage. Chirurg 2000;71:1243–50.[15] Suter M, Giusti V, Worreth M, Heraief E, Calmes J-M. Laparoscopic gastric banding: a prospective, randomized study comparing the Lap- [37] Fielding GA, Allen JW. A step-by-step guide to placement of the Band and the SAGB: early results. Ann Surg 2005;241:55– 62. Lap-Band adjustable gastric banding system. Am J Surg 2002;184:[16] Weiner R, Blanco-Engert R, Weiner S, et al. Outcome after laparo- 26S–30S. scopic adjustable gastric banding— 8 years’ experience. Obes Surg 2003;13:427–34. [38] Ren CJ, Fielding GA. Laparoscopic adjustable gastric banding: sur-[17] U.S. Food and Drug Administration. New device approvals: LAP- gical technique. Laparoendosc Adv Surg Tech A 2003;13:257– 63. BAND© adjustable gastric banding (LAGB®) system—P000008: summary of safety and effectiveness data. June 5, 2001. [39] O’Brien PE, Dixon JB, Laurie C, et al. A prospective randomized trial[18] U.S. Food and Drug Administration. New device approvals. of placement of the laparoscopic adjustable gastric band: comparison REALIZE™ adjustable gastric band—approval #P070009. Summary of the perigastric and pars flaccida pathways. Obes Surg 2005;15: of safety and effectiveness data. September 28, 2007. 820 – 6.[19] Ponce J, Dixon JB. 2004 ABS consensus conference. Laparoscopic adjustable gastric banding. Surg Obes Relat Dis 2005;1:310 – 6. [40] Fried M, Kormanova K, Kasalicky M. Comparison of SAGB QC[20] Ponce J, Paynter S, Fromm R. Laparoscopic adjustable gastric band- “classic” titanium port and VelocityTM port: pilot phase prospective ing: 1,014 consecutive cases. J Am Coll Surg 2005;201:529 –35. randomized study on perioperative and short-term postoperative im-[21] Kothari SN, DeMaria EJ, Sugerman HJ, et al. Laparoscopic band plantation outcomes. Obes Surg 2006;16:716 –20. failure: conversion to gastric bypass and their preliminary outcomes. Surgery 2002;131:625–9. [41] Miller K, Pump A. Mechanical versus suture fixation of the port in[22] Suter M, Calmes JM, Paroz A, Giusti V. A 10-year experience with adjustable gastric banding procedures: a prospective randomized laparoscopic gastric banding for morbid obesity: high long-term com- blinded study. Surg Endosc 2008 Apr 4; [Epub ahead of print]. plication and failure rates. Obes Surg 2006;16:829 –35.[23] Doherty C, Maher JW, Heitshusen DS. Long-term data indicate a [42] Angrisani L, Favretti F, Furbetta F, et al. Lap-Band® Rapid Port™ progressive loss in efficacy of adjustable silicone gastric banding for system: preliminary results in 21 patients. Obes Surg 2005;15:936. the surgical treatment of morbid obesity. Surgery 2002;132:724 – 8.[24] Allen JW, Coleman MG, Fielding GA. Lessons learned from laparo- [43] Realize adjustable gastric band instructions for use. Cincinnati: Ethi- scopic gastric banding for morbid obesity. Am J Surg 2001;182: con Endo-Surgery, Inc.; 2007. 10 – 4.[25] Fielding GA, Duncombe JE. Clinical and radiological follow-up of a [44] Miller K. Operation primer: laparoscopic gastric banding with the cohort of laparoscopic adjustable gastric bands from 1998. Obes Surg Swedish Adjustable Gastric Band (SAGB Quick Close). Haslemere, 2005;15:634 – 40. Surrey, UK: Euromed Communications, Ltd.; 2006.[26] Szinicz G, Schnapka G. A new method in the surgical treatment of disease. Acta Chir Aust 1982;14(Suppl):43. [45] Shen R, Ren CJ. Removal of peri-gastric fat prevents acute obstruc-[27] Szinicz G, Muller L. A new method in the surgical treatment of tion after Lap-Band surgery. Obes Surg 2004;14:224 –9. pathologic obesity—results of animal experiments. Proceedings of the XI Annual Meeting of the European Society for Artificial Organs, [46] Realize injection port instructions for use. Cincinnati: Ethicon Endo- September 9 –12, 1984, Alpbach-Innsbruck, Austria. London: WB Surgery, Inc.; 2007. Saunders; 1984. p. 322.[28] Szinicz G, Muller L, Erhart W, et al. “Reversible gastric banding” in [47] Provost DA. Laparoscopic adjustable banding: an attractive option. surgical treatment of morbid obesity—results of animal experiments. Surg Clin North Am 2005;85:789 – 805. Res Exper Med (Berl) 1989;189:55– 60.[29] Hallberg D, Forsell P. Ballongband Vid Behandling av Massiv Over- [48] Shen R, Dugay G, Rajaram K, Cabrera I, Siegel N, Ren CJ. Impact of vikt (Balloon Band for Treatment of Massive Obesity). Svensk Kiru- patient follow-up on weight loss after bariatric surgery. Obes Surg rgi 1985;43:106. 2004;14:514 –9.[30] Kuzmak LI. A review of seven years’ experience with silicone gastric banding. Obes Surg 1991;1:40 – 8. [49] Ponce J. Laparoscopic gastric band adjustments. Bariatric Times[31] Catona A, Gossenberg M, La Manna A, Mussini G. Laparoscopic 2007;May. adjustable gastric banding: preliminary series. Obes Surg 1993;3: 207–9. [50] Forsell P, Hallberg D, Hellers G. A gastric band with adjustable inner[32] Broadbent R, Tracey M, Harrington P. Laparoscopic gastric banding: diameter for morbid obesity: preliminary studies. Obes Surg 1993;3: a preliminary report. Obes Surg 1993;3:63–7. 303– 6.[33] Belachew M, Legrand MJ, Defechereux TH, et al. Laparoscopic adjustable silicone gastric banding in the treatment of morbid obesity: [51] Dargent J. Pouch dilation and slippage after adjustable banding: is it a preliminary report. Surg Endosc 1994;8:1354 – 6. still an issue? Obes Surg 2003;13:111–15.[34] Belachew M, Legrand M, Vincent V, et al. Laparoscopic placement of adjustable silicone gastric band in the treatment of morbid obesity: [52] Cunneen SA, Phillips E, Fielding G, et al. Studies of the Swedish how to do it. Obes Surg 1995;5:66 –70. Adjustable Gastric Band and Lap-Band: systematic review and meta-[35] Forsell P, Hellers G. The Swedish Adjustable Gastric Banding analysis. Surg Obes Relat Dis 2008;4:174 – 85. (SAGB) for morbid obesity: 9-year experience and a 4-year follow-up of patients operated with a new adjustable band. Obes Surg 1997;7: [53] Miller K, Hell E. Laparoscopic adjustable gastric banding: a prospec- 345–51. tive 4-year follow-up study. Obes Surg 1999;9:183–7. [54] Nowara HA. Egyptian experience in laparoscopic adjustable gastric banding (technique, complications, and intermediate results). Obes Surg 2001;11:70 –5. [55] Vassallo C, Andreoli M, La Manna A, Turpini C. 60 Reoperations on 890 patients after gastric restrictive surgery. Obes Surg 2001;11: 752– 6. [56] Ponson AE, Janssen IM, Klinkenbijl JH. Laparoscopic adjustable gastric banding: a prospective comparison of two commonly used bands. Obes Surg 2002;12:579 – 82. [57] Bende J, Csiszar M, Ursu M. Modern surgical interventions in the treatment of morbid obesity. Orv Hetil 2003;144:2459 – 63. [58] Ceelen W, Walder J, Cardon A, et al. Surgical treatment of severe obesity with a low-pressure adjustable gastric band: experimental data and clinical results in 625 patients. Ann Surg 2003;237:10 – 6. [59] Dukhno O, Ovnat A, Levy I. Our experience with 250 laparoscopic adjustable silicone gastric bandings. Surg Endosc 2003;17:601–2.

Surgery for Obesity and Related Diseases 7 (2011) 8 –14 Original article Gastric plication: preclinical study of durability of serosa-to-serosa appositionHector J. Menchaca, M.D.a,*, Jason L. Harris, Ph.D.b, Suzanne E. Thompson, D.V.M., M.S.b, Mary Mootoo, R.V.T., S.R.S., B.B.A.b, Van N. Michalek, B.A.a, Henry Buchwald, M.D., Ph.D.a aUniversity of Minnesota, Minneapolis, Minnesota bEthicon Endo-Surgery, Incorporated, Cincinnati, Ohio Received May 6, 2010; accepted November 5, 2010Abstract Background: Vertical gastric plication is a novel surgical approach for reducing the stomachKeywords: capacity. Anterior surface plication and greater curvature plication are variations of vertical gastric plication that reduce the gastric capacity through infolding of the anterior surface or greater curvature of the stomach, respectively. These approaches have been tested, with positive results in a small number of preclinical and clinical trials. A key step toward greater investigation of vertical gastric plication as a viable bariatric procedure would be confirmation that the apposed serosal tissues can be securely and durably bonded. We compared the short-term durability of gastric plications and serosal bonds using a variety of fastening devices and techniques in a university hospital and private company setting. Methods: A total of 30 anterior surface plication and 8 greater curvature plication procedures were performed using an open or a laparoscopic surgical technique in 38 hound dogs. The fastening devices used were T-tags, buttressed T-tags, 2 types of suture, and 4 types of staple-based fasteners. The density of the fastening devices was varied among the dogs. With 3 exceptions (2 in keeping with the study design and 1 due to complications), the dogs were followed up for 8 weeks. Histologic examinations and tensile testing were performed postmortem. Results: All fastening devices created durable plication folds, except for 1 technique. The only technique that did not produce durable serosa-to-serosa adhesions was a staple/suture combination. Intentional abrasion of the apposed sites had no demonstrable effect on the bonds between the tissues in any group. Increasing the fastener density and number of rows increased the adhesions within the folds. Conclusion: Our findings have confirmed the short-term durability and development of dense fibrous appositions of the serosal folds for fastening approaches that create fixed serosal apposition. Additional studies are needed to identify the optimal fastening modality for anterior surface plication, greater curvature plication, and, ultimately, clinical applications of this procedure. (Surg Obes Relat Dis 2011;7:8 –14.) © 2011 American Society for Metabolic and Bariatric Surgery. All rights reserved. Vertical gastric plication; Anterior surface plication; Greater curvature plication; Gastric volume reduction; Serosal apposition; Bariatric surgery Presented, in part, at a poster session of the 95th Clinical Congress of the American College of Surgeons, October 13, 2009, and, in part, as an oralpresentation at the Annual Meeting of the American Society for Metabolic and Bariatric Surgeons, June 26, 2010. *Correspondence: Hector J. Menchaca, M.D., University of Minnesota, 420 Delaware Street Southeast, Mayo Mail Code 290, Minneapolis, MN 55455. E-mail: [email protected]/11/$ – see front matter © 2011 American Society for Metabolic and Bariatric Surgery. All rights reserved.doi:10.1016/j.soard.2010.11.002

H. J. Menchaca et al. / Surgery for Obesity and Related Diseases 7 (2011) 8 –14 9 The gastric restrictive procedure known as vertical gas- were not successful as long-term fastening devices. All hadtric plication (VGP) has shown promise as an addition to the corroded, were dislodged, or had exerted enough pressure tocurrently available bariatric approaches for the treatment of compromise vascular perfusion and cause localized tissuemorbid obesity. Its development was supported by evidence necrosis (although tissue healing around these regions pre-that patients who undergo the Nissen gastric fundoplication vented gastric leaks from occurring).for treatment of gastroesophageal reflux disease experiencea small, but significant, weight loss [1]. In VGP, the gastric This early experience resulted in our current experi-wall is infolded to reduce the gastric capacity. At least 2 ments, which assessed the suitability of 3 types of fasteningvariations of VGP have been studied clinically: plication of devices (T-tags, sutures, and staples). We compared thethe anterior surface and plication of the greater curvature outcomes in 38 dogs, each of which underwent 1 fastening[2–5]. VGP is performed without banding, partitioning, or technique. A secondary goal was to assess the value oftransection of the gastrointestinal tract; instead, the recon- deliberate abrasion of the serosa to stimulate an inflamma-figured stomach is stabilized with sutures or other fastening tory healing response to augment tissue bonding. The ten-devices applied in multiple longitudinal rows. The fasteners sile strength of the apposed serosal surfaces was assessed asmaintain serosal apposition, leading to long-term serosa-to- an indicator of short-term bonding.serosa healing. VGP has resulted in minimal anatomic dis-ruption and few complications [2–5]. Moreover, it can be Methodscombined with other therapies or more extensive proce-dures, if necessary, at a later date. VGPs were performed at 2 locations: the University of Minnesota (Minneapolis, MN; 6 male crossbred hounds) The findings of several preclinical investigations have and Ethicon Endo-Surgery (Cincinnati, OH; 32 femalesupported the viability of VGP as a clinical procedure. crossbred hounds). Figure 1 provides an overview of theFusco et al. [6,7] performed 2 studies of VGP in Wistar rats. surgical approach used in both settings. The range of theThe first study compared infolding of the greater gastric dogs’ weight was 19.5–30.0 kg. All dogs had been withoutcurvature with 2 control procedures [6]. The second study food for Ն12 hours before the procedure, before endoscopy,compared infolding of the greater gastric curvature to a and before the surgical re-examination (endpoint). Accesssimilar procedure that infolded the anterior gastric wall [7]. to water was permitted at all times.In both studies, infolding of the greater gastric curvatureresulted in significant reductions in weight gain that ex- Surgical procedureceeded the results for any comparison group. Another re-search group, also using Wistar rats, compared 2 methods All aspects of these studies were performed in accor-for infolding the greater curvature, resulting in a comparable dance with the animal research protocols approved by thereduction in gastric capacity [8]. Both procedures achieved Institutional Animal Care and Use Committees at each in-identical reductions in weight gain and food intake, with stitution. The animal care was in accordance with eachsustained reductions in gastric volume. institution’s standard operating procedures. Several early reports have highlighted the clinical prom- All procedures were performed with the dogs underise of VGP and, in particular, laparoscopic greater curvature general anesthesia. After anesthetic induction, the dogsplication (LGCP). In early testing in humans, Talebpour and were intubated and maintained on inhalation anesthesiaAmoli [2] reported the outcomes of LGCP performed in 100 (isoflurane in oxygen). A flexible endoscope was passedpatients in private hospitals in Iran. They reported a per- transorally into the gastric lumen, and the stomach wascentage of excess weight loss of 61% at 1 year and 57% at moderately insufflated. The stomach exterior was accessed3 years, similar to the rates after other restrictive methods, by midline laparotomy or laparoscopic access. One half ofwith minimal need for reoperation. Sales Puccini [4] per- the dogs underwent deliberate serosal abrasion in the area toformed a variation of LGCP in 100 patients in Colombia, be infolded using an electrosurgical tip cleaner pad (me-with 69.61% excess weight loss at 1 year. More recently, chanical) or ultrasonic energy.Ramos et al. [5] reported 60% EWL at 1 year from aninterim analysis of 10 LGCP patients in an ongoing study Three general fastening mechanisms were used:involving 30 patients in Brazil. In the United States, pre-liminary results have been reported by the Cleveland Clinic 1. T-tags: each T-tag was a 12-mm long stainless steelFoundation [3] for 15 patients who underwent VGP. The tube connected at midpoint to a single strand of 2-0findings at 3 months were considered encouraging for both braided polyester. The T-tags were placed to span theanterior and greater curvature plication procedures. full thickness of the gastric wall, from the serosal surface into the gastric lumen. In 14 dogs, the T-tags In a series of unpublished pilot experiments, our coop- were placed about the intended fold line and theerative group performed VGP in a live porcine model, using suture ends were tied to appose the serosal tissue. Incustomized intragastric clamps as fasteners. This achieved 4 dogs, the T-tags were buttressed endolumenallythe desired serosal fold formations; however, the clamps with a plastic washer between the T-tag and the mu- cosa (Fig. 2).

10 H. J. Menchaca et al. / Surgery for Obesity and Related Diseases 7 (2011) 8 –14Fig. 1. (A) Schema of anterior surface plication procedure completed with titanium wire formed into a ϳ5-mm square), or a2 rows of fasteners. Amount of gastric capacity that could be reduced using staple–suture combination (2 commercial box staplesthis approach was limited by size of anterior gastric wall. (B) Schema of approximated by a single tied strand of 2-0 braidedgreater curvature plication procedure completed with 2 rows of fasteners. polyester suture).Using this approach, gastric capacity could be maximally reduced. LC ϭlesser curvature; GC ϭ greater curvature; a ϭ initial or interior row of Each dog was treated using 1 of the described fasteningfasteners for maintaining apposition; b ϭ final load-bearing row of fasten- modalities. The fastener created the plication, infolding ei-ers. Schematics supplied by Ethicon Endo-Surgery. ther the anterior (ventral) surface or the greater curvature, starting distal to the esophagogastric junction, at or near the 2. Sutures: sutures were applied with interrupted mat- fundus, and extending toward the prepyloric antrum. For tress stitches and consisted of either 2-0 monofilament greater curvature plication, the gastroepiploic vasculature or 2-0 braided polyester. Commercially available au- was dissected away from the stomach and remained with the tomatic suturing devices were used in 4 dogs. Full- omentum. Alternatively, it was dissected from the omentum thickness or partial-thickness passage of suture varied and remained along the stomach. The fastening modality between the groups. was secured to create firm apposition of the serosal surfaces. 3. Staples: staples were applied using a commercially One to three rows of fasteners were placed, aligned to the available box stapler for laparoscopic hernia mesh lesser curvature. The spacing between the fasteners varied fixation (.021-in. diameter titanium wire formed into a from .5 to 3.0 cm. Differences in spacing reflected differ- 3.75-mm ϫ 5.5-mm rectangle), a prototype ring sta- ences in the contact areas of the various fasteners. The ple applier (.021-in. diameter stainless steel wire spacing was intentionally varied to identify trends linked to formed into a 5-mm circle), a prototype box staple durable apposition; however, the spacing was consistent for applier (.021-in. diameter stainless steel or .021-in. each treatment group. The depth of fastener penetration was also identified. Fasteners or fastener applicators capable of penetrating the entire gastric wall were characterized as full-thickness fas- teners. The full-thickness placement was confirmed by in- traoperative endoscopic observation, where possible. We intended to encompass the full thickness of the apposed gastric walls when applying any full-thickness fastening device. Intraoperative quantification of the penetration depth was not possible for the partial-thickness fasteners. Nevertheless, engagement of the submucosal layer was tar- geted with all partial-thickness fastening modalities. The infolded section of the stomach was imaged using the endoscope, and the general aspect and overall length and depth were recorded. Air was suctioned from the stomach, and the endoscope withdrawn. The incisions were closed routinely. Postoperative drug administration Analgesics were administered the day of the procedure and for 3 days postoperatively. Antibiotic prophylaxis was started on the day of the procedure and continued for 3–7 days postoperatively. At Ethicon Endo-Surgery, antiemetics were administered the afternoon after surgery and on post- operative day 1. Postoperative care and evaluation The dogs’ health was monitored twice daily for the first 7–14 days and daily thereafter. All the dogs received a 40 cal/kg/d liquid diet consisting of CliniCare canine liquid diet and canned dog food, with or without Test Diet Low Residue Canine Surgery Mix, for 14 days postoperatively, followed by a standard dog chow thereafter. The body weight was recorded weekly. The dogs were anesthetized

H. J. Menchaca et al. / Surgery for Obesity and Related Diseases 7 (2011) 8 –14 11Fig. 2. Schematic of (Left) T-tags and (Right) buttressed T-tags.periodically, and the gastric lumen was examined endoscop- The findings for the study cohorts are summarized inically to assess the status of the infolded stomach section. At Table 1. All fastening devices and techniques created dura-each examination, images were made for comparison with ble plication folds, except for 1. The only technique that didthe previous images of the infolded section. The dogs were not result in durable serosa-to-serosa adhesion as observedhumanely euthanized and necropsy was performed at the by endoscopic evaluation was the staple-suture combina-end of the survival period. At necropsy, images of the tion. Intentional abrasion of the apposed sites had no de-exterior surface of the stomach were taken, a section of the monstrable effect on the bonds between the tissues in anyinfolded tissue was submitted for histologic evaluation, and group. Intermittent point failures in serosal apposition oc-a section of infolded tissue was submitted for tensile testing. curred in those dogs that had received only 1 row of fas- teners; in regions of the fold not containing fasteners, theOutcome assessments serosal surfaces had not bonded (Fig. 3). Healing and characterization of the original geometry A cross section of a gastric fold at necropsy is shown inwere assessed by endoscopic inspection of the infolded Fig. 4A. The plication had healed, and new serosal tissuegastric wall. In early experiments, this inspection occurred had bridged the apposing surfaces. Histologic studies of theevery 10 –14 days; however, in the later experiments, the bridges (Fig. 4B) showed connective tissue networks andinspections were typically done 1 and 4 weeks postopera- angiogenesis.tively, with direct inspection at necropsy. With 3 excep-tions, all the dogs were euthanized at 56 –59 days after the In the mechanical tests, the durable plications were stron-procedure, at which point necropsy was performed by a ger than the surrounding tissue, with tissue failures oftenmedical or veterinary pathologist. The durability of the occurring in the tissues adjacent to the folds, but not withininfolded tissue was indicated by the persistence of the orig- them. In all cases, the presence of a fold indicated the foldinal fold geometry. The presence and degree of serosal was strong enough to withstand the in vivo stresses createdbonding was noted. In the histopathologic evaluations, con- within the gastric wall from eating, normal gastric func-nective tissue bridging and angiogenesis were considered tions, and vomiting (if present).indicators of serosa-to-serosa healing. Erosive lesions andinflammatory tissue were noted when present. Significant inflammatory responses were seen in those dogs that had received full-thickness braided polyester su- Tensile testing was performed using Instron tensile test- tures, a multifilament, nonabsorbable suture material placeding equipment. The sections of infolded tissue were securely into the nonsterile gastric lumen. Mild mucosal erosion wasgripped on either side of the fold and pulled apart at a seen with the metal T-tag ends potentially deployed in directconstant rate until failure occurred. The force, displacement, contact with the tissue walls. Mucosal defects without in-and description of the failure were recorded and imaged. flammation were associated with tags that were not yet overgrown with mucosa. Minimal or mild chronic inflam-Results mation and fibrosis surrounded the sutures and staples in both the partial-thickness and the full-thickness monofila- With 1 exception, all the dogs survived to the predesig- ment treatment groups.nated study endpoints (21 d for 2 dogs and 56 –59 d for 36dogs). One dog with a gastric plication created using staples Discussionwas euthanized at 25 days because of persistent anorexia,intermittent vomiting, and weight loss. Endoscopic imaging The results of our study have added to the evidence thatwas performed before death. No obvious stricture or ob- the available fastening modalities can achieve durable se-struction was observed. rosa-to-serosa appositions when appropriately spaced. All techniques tested were successful, except for the staple-

12 H. J. Menchaca et al. / Surgery for Obesity and Related Diseases 7 (2011) 8 –14Table 1Fastening methods, relevant procedure variables, and outcomes for vertical gastric plication procedureFastening method Dogs (n) Serosal Rows (n) Full/partial Procedure type Approach Durable Location treatment thickness plication foldT-tag 3 Mechanical 3 Full Anterior Open Yes* MN 3 None 3 Full Anterior Open Yes* MNButtressed T-tags 4 Harmonic 3 Full Anterior Open Yes EES 4 None 3 Full Anterior Open Yes EES2-0 Ethibond sutures, interrupted 2 Mechanical 2 Full Anterior Open Yes EES pattern 2 None 2 Full Anterior Open Yes EES 2 Mechanical 3 Full Anterior Open Yes EESPartial-thickness, automated suturing 2 None 3 Full Anterior Open Yes EES device, interrupted pattern 2 Harmonic 1 Full Greater curvature Open Point failure† EES 2 None 3 Partial Anterior Laparoscopic Yes EES2-0 Prolene, interrupted patternFull-thickness, automated suturing 2 None 3 Full Anterior Open Yes EES 2 None 1 Full device, interrupted pattern Greater curvature Laparoscopic Point failure† EESBox stapler 1 PartialPrototype ring stapler 1 Harmonic 3 Partial Anterior Open Yes EESStaple–suture combination 2 Harmonic 3 Partial Anterior Open Yes EESPrototype box stapler 2 Harmonic 3 Partial Anterior Open No EES 2 Harmonic 3 Partial Greater curvature Laparoscopic Yes‡ EES None 3 Greater curvature Laparoscopic Yes EESMN ϭ Minnesota; EES ϭ Ethicon Endo-Surgery.* One dog was prospectively evaluated at week 3.† Fold intact, but some stitches had failed, with healing confined to fastened region.‡ One dog removed from study because of excessive weight loss at week 3.suture combination. Thus, our feasibility data suggest that ing the operative time and simplifying the surgical proce-other methods would be preferred. We sought to assess the dure.short-term (8-wk) durability of fastened, apposed tissues,not to compare the efficacy of the various devices. Given the The diversity of study groups that resulted in durablerelative success of the many fastening methods, we are plications at either 3 or 8 weeks suggests multiple ap-performing statistically powered studies of the fastening proaches to creating a durable plication exist. Failure of theapproaches that lend themselves to device applications that staple–suture combination—in which a suture was used tomaximize the safety and procedure outcomes while reduc- connect pairs of staples—sheds light on the requirements of successful apposition. This approach did not eliminate rel-Fig. 3. Histologic image of single row of suture at greater curvature. In ative motion between staples and the apposed serosal tis-regions of fold not containing fasteners, serosal surfaces did not bond. sues, because the suture could slide along the staples’ back-Black arrow indicates serosal space; red arrow, fibrous healing of plication. span/crown. Moreover, gaps existed between the backspanM ϭ mucosa; SM ϭ submucosa; TM ϭ tunica muscularis. Masson’s of each staple and the tissue to which it was attached. Thetrichrome stain. motions and loads of the stomach create tension, which increases these gaps, ultimately causing an opening between the gastric walls that would otherwise be apposed. Ulti- mately, the staples migrated out of the tissue into which they had been inserted. Brethauer et al. [3] used 2 rows of running suture to create the plications. In contrast, most of our plications incorporated 3 rows of fasteners. However, in 4 dogs, we used 2 rows of fasteners, and in 4 others we used a single row. Aside from the staple-suture combination, the only other point failures in serosal apposition occurred in those dogs that had received only 1 row of fasteners. In those dogs, fold geometry was generally maintained, with point suture failures observed (presumably at sites at which the tension stresses were increased). The serosal surfaces were not bonded in the regions of the fold not containing fasten- ers. Therefore, it appears that Ͼ1 row of fasteners is re-

H. J. Menchaca et al. / Surgery for Obesity and Related Diseases 7 (2011) 8 –14 13Fig. 4. (A) Three-week cross-sectional view of fold fastened with 3 rows quired to limit motion and distribute the tension on theof T-tags (Minnesota group). (B) Photomicrograph at 8 weeks. Distal apposed serosal surfaces. No clinically relevant outcome(pyloric) portion of plication site from dog 2 (Ethicon Endo-Surgery group) was linked to the point failures at the 8-week point; never-with no abrasion. Layers of stomach wall were folded. Apposed serosal theless, longer term evaluations of single-row plicationslayers healed together to form 1–2-mm fibrous connective tissue bridge. should be performed before their clinical use.Spaces with sutures shown (arrows). Masson’s trichrome stain. M ϭinternal tunica muscularis; m ϭ external tunica muscularis; S ϭ collagen Our early T-tag experiments investigated various pat-(stained blue). Muscle stained pink. terns of distance spacing between fasteners. No differences in outcomes were observed between the maximum and minimum spacings used (3 and 1.5 cm, respectively). Our technique evolved with time to the use of a small spacing (Յ2 cm) in the final, outermost load-bearing row. Larger spacings (Յ3 cm) were used in the interior rows. More rigorous investigations of the optimal spacing for the opti- mal fastener are needed. Several connections emerged between the surgical tech- niques and plication outcomes. All full-thickness fasteners applied with appropriate spacing yielded durable plications. However, fasteners containing braided suture with interstices were associated with heightened inflammation. Partial-thick- ness fastening modalities avoid this concern. The target tissue depth needed for durable bonding was not specifically ad- dressed; however, it was likely related to fastener spacing within a row and the number of rows used. We believe that deeper anchoring will provide more durable results. For staple-based approaches, the penetration depth will be determined by the final formed shape of the staple and apposition of the device to the tissue, before staple de- ployment. For approaches using suturing methods, the depth will be controlled through needle placement. We attempted to penetrate the submucosa when using the partial-thickness fastening modalities. Finally, our expec- tation that serosal abrasion might enhance the healing response was disproved; serosal abrasion had no effect on the durability of the plications. This was true for all fastener–spacing combinations tested. Our findings did not allow the postulation of a definitive model for serosa-to-serosa apposition for gastric plication. Three principles for success, however, were evident. First, the contact of the serosal surfaces must be continuous, without slippage. This tissue apposition can be provided by several fastening devices (e.g., T-tags, buttressed T-tags, sutures, and staples). Second, the anchoring points using the fasteners should be placed no further than 2.5 cm apart to allow a seamless bonding without separation gaps. Third, complete tissue strata penetration to the mucosa by the fasteners is preferable for fixation. Hand sewing with nonbraided, single- filament, permanent suture is the least expensive method to achieve apposition; however, the same results can be obtained using a mechanical suturing device or staples. Whatever the apposition method used, a longer and deeper imbrication will further decrease the gastric luminal volume. Because imbrica- tion does not involve gastric wall resection, the risk of leaks should be absent because the penetrating fasteners do not cause extensive tissue damage. The limitations of the present study included the techni-

14 H. J. Menchaca et al. / Surgery for Obesity and Related Diseases 7 (2011) 8 –14cal challenges in our analysis of 2 important parameters: the tirely or largely intact at 8 weeks after surgery. These datatensile strength of the bonds and the volume change of the suggest that the durability of the plication is dependent onstomach after surgery. We performed tensile testing on all continuous fixed serosal apposition by the fastening modal-test specimens. The location of failures varied between and ity at multiple points along the fold. Multiple rows ofwithin the groups, with most occurring near the jaws of the fasteners and fastener spacing Ͻ2.5 cm within a row pro-testing device and some occurring within the fold. The duced durable outcomes. Intentional serosal abrasion didlimitations of these tests made quantitative measurements not enhance outcomes for the fastener/spacing combinationsunreliable. The fastener and fold size influenced the width studied. Both full- and partial-thickness fastening modali-of the specimen (ϳ4 cm) needed to ensure a reasonably ties resulted in durable outcomes, but full-thickness fasten-homogeneous response. This width was nearly one third of ing was associated with an increased risk of localized in-the overall fold length and included a significant percentage flammation.of the circumference of the stomach in this region (e.g., fora complete anterior surface plication, the fold would ac- These findings support the early clinical evidence thatcount for ϳ50% of the gastric circumference). This left serosa-to-serosa apposition can generate durable gastric pli-minimal tissue to establish the width of the grip surface cations. They also suggest that VGP merits additional studyneeded to minimize edge effects. Therefore, failures at the for clinical efficacy. Additional studies are needed to deter-grips could not be reliably attributed to the fold strength. mine the optimal fastening modality for use in VGP. TheseMeasurements of the tissue stresses require precise assess- data have shed light on the requirements for an optimalments of force and tissue cross-sectional area. Such mea- device that maximizes secure tissue apposition while reduc-surements are challenging in gastric tissue owing to its ing the operative time.nonuniform thickness. Measurements of unconstrained tis-sue widths were complicated by the significant release of Disclosuresresidual stress after removal of the test specimen from therest of the stomach. This study was funded by Ethicon Endo-Surgery. H. Buchwald was a consultant to Ethicon Endo-Surgery; J. L. We were also unable to measure the changes in stomach Harris, S. E. Thompson, and M. Mootoo are employees ofcapacity. We attempted to measure volume by endoscopically Ethicon Endo-Surgery.placing a large pliable bag in the stomach, filling it to ameasured pressure, and then measuring the amount of fluid Referencesrequired to achieve this pressure. Technical challenges and alack of repeatability within the same dog caused us to abandon [1] Neumayer C, Ciovica R, Gadenstätter M, et al. Significant weightthis approach after the first group of experiments was com- loss after laparoscopic Nissen fundoplication. Surg Endosc 2005;pleted. 19:15–20. Other limitations were also present. Our study ended at 8 [2] Talebpour M, Amoli BS. Laparoscopic total gastric vertical plicationweeks; however, human patients would need the apposi- in morbid obesity. J Laparoendosc Adv Surg Tech A 2007;17:793– 8.tions to persist for many years. In addition, our study was afeasibility study. A statistically powered study is needed [3] Brethauer SA, Harris JL, Chand B, Kroh M, Rogula T, Schauer PR.and is being planned. Initial results of vertical gastric plication for severe obesity. Presented at the Society of American Gastrointestinal and Endoscopic Surgeons, Another important variable was the effect of dog behav- Phoenix, Arizona, April 22–25, 2009.ior on gastric loads (e.g., vomiting and eating large or smallamounts of food at a given time). The low number of dogs [4] Sales Puccini CE. Surset Gástrico de Sales: una alternativa para cirugíain our study prevented us from correlating these behaviors bariátrica restrictiva. Rev Colomb Cir 2008;23:131–5.with the outcomes. [5] Ramos AC, Galvao M, Behrens E, Montufar F, Zundel N. Tubular Finally, the lack of predictive animal models in the field of sleeve gastroplasty (TSG) as a new approach to bariatric treatment.bariatric surgery remains significant. Rodent models have pre- Presented at the 14th World Congress of the International Federationdicted weight loss but provide challenges of scale when ap- for the Surgery of Obesity, Paris, France, August 26 –29, 2009.plied to human metabolic disease. The early findings fromTalebpour and Amoli [2], Brethauer et al. [3], and Ramos et al. [6] Fusco PEB, Poggetti RS, Younes RN, Fontes B, Birolini D. Evaluation[5] have suggested that preclinical findings might translate to of gastric greater curvature invagination for weight loss in rats. Obesclinical results. A more complete discussion of human trials Surg 2006;16:172–7.can be found in the report by Brethauer et al. [9]. [7] Fusco PEB, Poggetti RS, Younes RN, Fontes B, Birolini D. Compar-Conclusion ison of anterior gastric wall and greater gastric curvature invaginations for weight loss in rats. Obes Surg 2007;17:1340 –5. The 3 basic fixation methods tested (i.e., T-tags, sutures,and staples) each created serosal plications that were en- [8] Johnson AC, Tyler KR, Harris JL, Greenwood-Van Meerveld B. A novel gastric invagination procedure produces weight loss in rats. Digestive Disease Week, May 30 –June 4 2009, Chicago, Illinois. [9] Brethauer SA, Harris JL, Kroh M, Schauer PR. Laparoscopic gastric plication for the treatment of severe obesity. Surg Obes Relat Dis 2010;7:15–22.

Surgery for Obesity and Related Diseases 4 (2008) 423– 429 Original articleGastric banding for the treatment of type 2 diabetes mellitus in morbidly obeseAnthony Brancatisano, B.Appl.Sc., M.B.B.S., Ph.D.a,*, Sara Wahlroos, B.Med.Sc.a, Slade Matthews, B.Med.Sc., Dip.Ed., Ph.D.b, Roy Brancatisano, M.B.B.S.a aInstitute of Weight Control, Sydney, Australia bBosch Institute of Medical Research, University of Sydney, Sydney, Australia Received May 14, 2007; revised July 22, 2007; accepted October 24, 2007Abstract Background: To assess the efficacy of the Swedish adjustable gastric band in the treatment of typeKeywords: 2 diabetes mellitus (T2DM), impaired glucose tolerance (IGT), and the metabolic syndrome (MS) in the morbidly obese. Methods: We identified all patients with T2DM, IGT, or the MS at surgery from our database of 905 consecutive patients who had undergone placement of the Swedish adjustable gastric band between January 2001 and April 2007. The patients were followed up by our multidisciplinary team, and their T2DM was managed by their treating primary care physician and/or endocrinologist. Results: A total of 682 patients had Ͼ6 months of follow-up. Of these, 78 patients had T2DM, 64 had IGT, and 100 had the MS. At a median follow-up of 12.5 months, patients with T2DM had a mean Ϯ SD excess weight loss of 38% Ϯ 15%. This was associated with hemoglobin A1c and fasting blood sugar levels decreasing from 8.0% Ϯ 1.7% to 6.1% Ϯ 1.0% (P Ͻ.0001) and from 9.6 Ϯ 3.4 mmol/L to 5.7 Ϯ 1.5 mmol/L (P Ͻ.0001), respectively. Remission and/or improvement in patients with T2DM was judged by the complete cessation and/or reduction in medication and normalization of laboratory values. This occurred in 81% of those taking oral hypoglycemic agents. Of the patients taking a combination of oral hypoglycemic agents and insulin, 43% ceased and/or reduced their oral hypoglycemic agents, and 93% ceased and/or reduced their insulin requirements. Of those on insulin only, 75% ceased and/or reduced their insulin. No patient with IGT developed diabetes or progressed to require medications. Remission and/or improvement in the MS occurred in 88% of patients. Remission of T2DM was dependent on both the magnitude of excess weight loss (P ϭ .008) and the duration of the pre-existing T2DM. Using binary logistic regression analysis, a duration of T2DM of Ͻ5 years before surgery was 6.5 times more likely to lead to resolution of T2DM after the weight loss (P ϭ .004). Conclusion: Weight loss after Swedish adjustable gastric band placement is an effective treatment of T2DM in morbidly obese patients, with early intervention offering the greatest chance of remission. It might even prevent the occurrence of T2DM in patients with IGT. (Surg Obes Relat Dis 2008;4: 423– 429.) © 2008 American Society for Metabolic and Bariatric Surgery. All rights reserved. Swedish adjustable gastric band; Bariatric surgery; Diabetes; Impaired glucose tolerance; Metabolic syndrome; Morbid obesity This research was supported by a grant from Johnson & Johnson It is well known that obesity is a major risk factor for theMedical, the manufacturer of the Swedish adjustable gastric band. development of type 2 diabetes mellitus (T2DM). The risk Presented at Annual Meeting of the American Society for BariatricSurgeons, June 11 to June 16, 2007, San Diego, California is proportional to the body mass index (BMI) and the *Reprint requests: Dr Anthony Brancatisano, Institute of Weight Con- duration of the obesity [1]. Approximately 47% of patientstrol, 495 Windsor Road, Baulkham Hills, Sydney, NSW 2153, Australia. with T2DM have a BMI Ͼ30 kg/m2 and 9% have a BMI Ͼ40 kg/m2 [2]. Obesity, when associated with impaired E-mail: [email protected] glucose tolerance (IGT), is a leading cause of T2DM and is associated with a 45–55% increase in mortality risk during a 5-year period [3]. A particular cluster of abnormalities,1550-7289/08/$ – see front matter © 2008 American Society for Metabolic and Bariatric Surgery. All rights reserved.doi:10.1016/j.soard.2007.10.011

424 A. Brancatisano et al. / Surgery for Obesity and Related Diseases 4 (2008) 423– 429namely abdominal adiposity, dyslipidemia, hypertension, Postoperative phase and follow-upand elevated fasting blood glucose, comprises the metabolicsyndrome (MS) [4]. The MS increases the relative risk of All patients began a fluid diet followed by transition to adeveloping atherosclerotic cardiovascular disease twofold solid food diet within a 4 – 6-week period. Each patient wasand T2DM fourfold [4]. followed up by the multidisciplinary team. Follow-up in- cluded monthly visits to a member of the multidisciplinary Effective weight management is an essential component team for the first 6 months, then every 3 months up to 1of the long-term treatment of T2DM and the MS. Weight year, and then biannually, or as required. We classifiedloss of about 5–10% improves glycemic control and reduces patients as lost to follow-up if they failed to attend forthe requirements for antidiabetic medication [5]. Modest consultation, tests, or band adjustments.weight loss also lowers blood pressure, improves dyslipide-mia, and is associated with a reduction in T2DM-related All adjustments to the band for the addition or removalmorbidity and mortality [6,7]. However, modest weight loss of sterile saline were performed under fluoroscopic imagingis difficult to achieve and maintain in obese diabetic pa- for the exact location of the port. Fluid was added if thetients. Conventional treatment has been disappointing be- patient reported inadequate feeling of satiety, inadequatecause of the failure of patients to comply with lifestyle restriction, or inadequate weight loss.modifications and oral hypoglycemic medication [2]. Fur-thermore, insulin treatment in these patients promotes Weight lossweight gain and worsens insulin resistance [8]. Body weight was measured using an electronic scale Bariatric surgery has been shown to produce a loss of (Tanita Corporation of America, model TBF-310). The BMIabout 48 –70% of excess weight [9]. Roux-en-Y gastric calculated by dividing the patient’s weight in kilograms bybypass has been proposed as the most effective therapy for the height in meters squared, as measured using a stadiom-T2DM, with resolution of T2DM in most patients [10]. eter. Weight loss was expressed as the percentage of excessHowever, gastric banding is minimally invasive, adjustable, weight loss (%EWL). The preoperative excess weight waspotentially reversible, and safe. Therefore, the aim of this calculated as the weight in kilograms at surgery minus thestudy was to report on the health outcomes in morbidly ideal body weight in kilograms, as measured by the Metro-obese patients with T2DM, IGT, and the MS using the politan Life tables. The %EWL was calculated by dividingSwedish adjustable gastric band. the weight change at follow-up by the excess weight pre- operatively multiplied by 100.Methods Co-morbid disease controlPatient selection The obese patients were stratified according to the Patients were admitted into the program if they fulfilled presence of T2DM, IGT, or the MS. The diagnosis ofthe following criteria: age Ͼ16 years, BMI Ͼ40 kg/m2 or T2DM and IGT were determined using the AmericanϾ35 kg/m2 in the presence of co-morbidities, and a history Diabetes Association criteria, and the MS was deter-of prolonged previous attempts at weight loss by other mined using the updated joint American Heart Associa-means. All patients provided informed consent. All patients tion and National Heart, Lung, and Blood Institute diag-were clinically evaluated by the surgeon, and a preoperative nostic criteria [12]. The preoperative fasting blood sugarassessment was conducted by a medical practitioner. They level, hemoglobin A1c (HbA1c) level, lipid levels, andwere followed up by the multidisciplinary team, which liver enzyme concentrations were collected prospec-included a nutritionist, a psychologist, and an exercise con- tively, and follow-up values were obtained from the pa-sultant. For the purposes of this study, all patients at 6 tient’s treating endocrinologist or primary care physicianmonths of surgical follow-up with T2DM, IGT, or the MS or as a part of the routine laboratory measurements per-were identified. formed by our clinic. We assessed the dose and frequency of all hypoglycemic and insulin requirements and anti-Gastric band surgery hypertensive and lipid-lowering medications pre- and postoperatively. We scored the patients according to re- The Swedish adjustable gastric band (Ethicon Endo- mission, improvement, or no change in their co-morbidSurgery, Johnson & Johnson Medical, Livingston, United illnesses. Remission of co-morbid illness was defined asKingdom) was inserted laparoscopically and placed around the patient no longer requiring medication after surgerythe fundus of the stomach immediately below the esopha- and when the laboratory values were within the normalgogastric junction. The basic procedure has been previously range. Improvement was considered if the medicationdescribed [11] but was modified to include excision of the dose was reduced and/or an improvement had occurred infat pad, repair of any hiatal defect, and high posterior lateral the laboratory values. If no remission or improvement,fundus-to-gastric pouch sutures. All surgical procedures the co-morbid illness was considered unchanged. Allwere conducted by a single surgeon. co-morbid illnesses were managed only by the treating

A. Brancatisano et al. / Surgery for Obesity and Related Diseases 4 (2008) 423– 429 425Table 1Patient characteristics (mean Ϯ SD)Characteristic Total cohort T2DM group (n ϭ 78) IGT group (n ϭ 64) (n ϭ 905) Before SAGB At 12.5-mo follow-up Before SAGB At 16-mo follow-upDemographics 44 Ϯ 8 52 Ϯ 8* 106.6 Ϯ 20.1† 47 Ϯ 11 97.8 Ϯ 20.0† Age (yr) 38 Ϯ 6† 36 Ϯ 7† Gender 199 (22) 33 (42) 11 (17) Males, n (%) 706 (78) 45 (58) 37.8 Ϯ 14.8 53 (83) 41.1 Ϯ 17.4 Females, n (%) 122 Ϯ 25 133.2 Ϯ 26.1* 26.6 Ϯ 16.6 122.8 Ϯ 24.0 25.0 Ϯ 12.3 Weight (kg) 44 Ϯ 8 47 Ϯ 8‡ 45 Ϯ 8 BMI (kg/m2) 5.7 Ϯ 1.5† 4.8 Ϯ 0.7† %EWL 9.6 Ϯ 3.4 6.1 Ϯ 1.0† 6.2 Ϯ 1.3 AWL (kg) 8.0 Ϯ 1.7Glucose metabolism Fasting BSL (mmol/L) HbA1c (%) T2DM ϭ type 2 diabetes mellitus; IGT ϭ impaired glucose tolerance; SAGB ϭ Swedish adjustable gastric band; n ϭ number of patients; BMI ϭ bodymass index; %EWL ϭ percentage of excess weight loss; AWL ϭ absolute weight loss; BSL ϭ blood sugar level; HbA1c ϭ hemoglobin A1c. * P Ͻ.001 compared with total cohort and IGT before SAGB. † P Ͻ.0001 compared with before SAGB. ‡ P Ͻ.001 compared with total cohort.endocrinologist or primary care physician. Neither the Resultssurgeon nor any allied healthcare professional from ourclinic were involved in any treatment decision to reduce Of the 905 consecutive patients who underwent laparo-or cease any medication. scopic placement of the Swedish adjustable gastric band since 2001, 682 patients had Ͼ6 months of follow-up. OfQuality-of-life assessment these patients, 78 had T2DM and a median follow-up of 12.5 months, 64 had IGT and a median follow-up of 16 The patients’ health-related quality of life was assessed months, and 100 patients had the MS and a median fol-using the Medical Outcomes Study Health Survey (SF-36) low-up of 14 months. The combined T2DM and IGT group[13], and the presence and severity of depression was as- represented 21% of the entire cohort of patients who under-sessed using the Beck Depression Inventory [14]. The went gastric banding using the Swedish adjustable gastricT2DM and IGT patients completed the SF-36 and Beck band. The average preoperative age, weight, and BMI forquestionnaire preoperatively and at follow-up. the entire cohort and the T2DM and IGT groups are listed in Table 1. The T2DM patients were older and weighed moreStatistical analysis (P Ͻ.001), with a greater proportion of male patients com- pared with either the entire cohort or the patients with IGT. All data were entered into and maintained on a comput- In the T2DM group, 1 band slip and 1 pouch dilationerized database (LapBase, AccessMed, Melbourne, Austra- occurred; both required revisional surgery. Two port infec-lia) using Microsoft Access. The data were collected pro- tions occurred, with one requiring port replacement. Nospectively. Data are expressed as the median and/or mean Ϯ complications occurred in the IGT group. No patient withthe standard deviation. A 2-sided Student t test was used to T2DM, IGT, or the MS were lost to follow-up.test for significant differences between 2 samples of quan-titative variables. Analysis of variance was used to test for Weight loss and effect on glycemic controldifferences in Ͼ2 groups, with the Leven test used to de-termine the homogeneity of variance among groups. If the For the T2DM and IGT group, laparoscopic gastric band-Leven test results were significant, the nonparametric post ing was associated with a %EWL of 38% Ϯ 15% andhoc Games-Howell test was used. If the results were not 41% Ϯ 17%, respectively (Table 1). The T2DM and IGTsignificant, the parametric post hoc Tukey’s honesty signif- patients achieved a smaller, but significantly lower, weighticant difference test was used. The influence of disease loss than that observed in the entire cohort, which averagedstatus on weight loss measures was examined using the 44% Ϯ 11% (P Ͻ.001) and 48% Ϯ 16% (P Ͻ.001) at 12logistic regression analysis. A P value Ͻ.05 was considered and 18 months, respectively.statistically significant. The Statistical Package for SocialSciences for Windows, version 15 (SPSS, Chicago, IL) was T2DM was generally poorly controlled in our patients onused for analysis. entering our program, despite treatment with oral hypogly- cemic (OH) medications and insulin (Table 1). Weight loss

426 A. Brancatisano et al. / Surgery for Obesity and Related Diseases 4 (2008) 423– 429had a major effect on glycemic control in both T2DM Table 2patients and IGT patients. The HbA1c and blood glucose Liver enzymes and lipids before and after gastric banding in patientslevels decreased significantly in the T2DM group (8.0% Ϯ with T2DM and IGT at median follow-up of 15 months1.7% to 6.2% Ϯ 1.1% and 9.6 Ϯ 3.4 mmol/L to 5.7 Ϯ 1.5mmol/L, P Ͻ.0001, respectively; Table 1). In 31 (40%) of Variable Reference Before At P valuethe 78 patients with T2DM, the HbA1c normalized (refer- follow-upence range Ͻ6.1%). Only 10% of patients had a HbA1c range SAGBlevel Ͼ7%. Liver enzymes Ͻ30 36 Ϯ 24 20 Ϯ 8 Ͻ.001 In the T2DM group, of the 60 patients treated with OH ALT (U/L) Ͻ30 26 Ϯ 14 20 Ϯ 5 Ͻ.001medication, 51% no longer required it, 30% reduced their AST (U/L) 83 Ϯ 25 76 Ϯ 23medication by 56%, and 17% had no change. Only 1 patient ALP (U/L) 30–150 36 Ϯ 25 27 Ϯ 30 .03required an increase in their OH medication. Of the 14 GGT (U/L) Ͻ35 .017patients in the T2DM group who were taking a combinationof OH agents and insulin, 21.5% no longer required OH Lipids 3.5–5.4 4.7 Ϯ 0.9 4.5 Ϯ 0.8 NSagents, 21.5% reduced their OH medication by 64%, and Cholesterol (mmol/L) 0.1–2.0 2.1 Ϯ 1.0 1.6 Ϯ 0.8 .000243% had no change; 2 patients required an increase in their Triglyceride (mmol/L) Ͼ1.0 1.4 Ϯ 0.6 1.5 Ϯ 0.2 NSOH medication. Of those 14 patients, 57% no longer re- HDL cholesterol Ͻ4.5 2.5 Ϯ 0.8 2.5 Ϯ 0.9 NSquired their insulin, 36% reduced it by 56%, and 1 patient LDL cholesterolrequired an increase in the insulin dose. For the 2 patientswho required an increase of OH medication, the insulin dose ALT ϭ alanine aminotransferase; AST ϭ aspartate aminotransferase;was either no longer required or reduced. Of the 4 patients ASP ϭ alkaline phosphatase; GGT ϭ gamma-glutamyle transferase;previously taking only insulin, 1 patient no longer required HDL ϭ high-density lipoprotein; LDL ϭ low-density lipoprotein; NS ϭit, 2 patients reduced it by 36%, and 1 patient had no not significant; other abbreviations as in Table 1.change. No patient required an increased dose. Overall, 83%of patients had ceased and/or reduced their diabetic medi- 6.5% Ϯ 1.1%, P Ͻ.03). T2DM patients with resolutioncation. Of the 78 patients with T2DM, 43% had complete achieved a significantly greater %EWL (42% Ϯ 15%) com-resolution and normalization of their HbA1c (5.6% Ϯ 0.7%) pared with those without resolution (%EWL 34% Ϯ 13%, P ϭand had a significantly greater %EWL (43% Ϯ 15%) com- .008), irrespective of duration of follow-up post surgery. How-pared with the 57% of patients who did not have complete ever, a trend was found for a slightly greater excess weight lossT2DM resolution (%EWL of 34% Ϯ 13%, P Ͻ.001). No in the group with a duration T2DM of Ͻ5 years (%EWLsignificant difference was found in the time of follow-up 39% Ϯ 16%) compared with those with a duration Ͼ5 yearsbetween these 2 groups. (%EWL 35% Ϯ12%), but this did not reach statistical signif- icance. Of the 64 patients with IGT, weight loss was associatedwith the mean fasting blood sugar level decreasing signifi- Effect on MS, liver enzymes, and quality of lifecantly from 6.2 Ϯ 1.3 mmol/L to 4.8 Ϯ 0.7 mmol/L(P Ͻ.0001; Table 1). In this group, 51 patients were treated A total of 100 patients were classified as having the MS.with diet alone, and none had progressed to T2DM at a Apart from the improvement in glycemia, weight loss wasmedian follow-up of 16 months. Of the 13 IGT patients associated with improvements in dyslipidemia, hyperten-treated with OH medication, 38% no longer required it, sion, and liver enzyme test results (Table 2). Of the 871 patient reduced the dosage, 46% had no change, and patients with hypertension, 27 (31%) no longer required1 patient required an increase in their medication. medication, 22 (25%) had a reduced requirement, and 38 (44%) had no change in their medication requirement. Of Using binary logistic regression analysis, we assessed the 55 patients with hyperlipidemia, 18 (33%) no longerwhether the preoperative age, gender, weight, BMI, fasting required medication, 6 (11%) had a reduced requirement,blood sugar level or HbA1c level, %EWL, or duration of and 31 (56%) had no change in their medication require-pre-existing T2DM predicted for remission of T2DM. Pa- ment. Preoperatively, in the combined T2DM and IGTtients with a more recent onset of T2DM had a better chance group, 35% of patients had at least one abnormal liverof remission. Of those with Ͻ5 years of pre-existing T2DM, function test. Postoperatively, 8% patients had at least one73% had resolution compared with 27% for those with elevated liver enzyme result. When the results were exam-Ͼ5-year duration of T2DM. The duration of T2DM Ͻ5 ined in their entirety, resolution and/or improvement in theyears appeared to confer a 6.5 times (odds ratio) greater MS had occurred in 88 (88%) of 100 patients.likelihood of the T2DM resolving according to the model(pseudo r2 Cox and Snell ϭ 0.22, P ϭ .004). Also, the Weight loss also had a positive effect on patients’ health-duration of T2DM resulted in significant differences in the related quality of life and depression in both IGT andpostoperative HbA1c level. A more recent onset of T2DM (Ͻ5 T2DM. The quality-of-life results showed a statisticallyyr) was associated with a lower HbA1c (5.9% Ϯ 1.1% versus significant improvement at a median follow-up of 14 months, returning to community normal values in all 8 subscales (Table 3). Weight loss was also associated with a statistically significant improvement in the Beck depression inventory from a baseline of 18.8 Ϯ 10 to 7.2 Ϯ 5.2

A. Brancatisano et al. / Surgery for Obesity and Related Diseases 4 (2008) 423– 429 427Table 3 Pontiroli et al. [19] observed, in a 4-year case-controlledShort form-36 quality-of-life domain scores at baseline and follow-up in study, that gastric banding effectively prevented the pro-T2DM and IGT patients gression of IGT to T2DM in their morbidly obese patients compared with a nonsurgical group. Similar results were Baseline Median P value* Community obtained by Dixon and O’Brien [15]. Our results add to the body of evidence that weight loss through gastric banding (total cohort) follow-up level can prevent the progression to T2DM in patients with IGT. (14 mo) Our patients with T2DM and those with IGT achieved a smaller, but significantly lower, %EWL compared with then 460 75 Ͻ.0001 88 entire cohort at a median follow-up of 12 and 18 months.Physical function 54 Ϯ 25 80 Ϯ 21 Ͻ.0001 83 These results are similar to those observed by Dixon andPhysical role 46 Ϯ 38 82 Ϯ 31 Ͻ.0001 74 O’Brien [15] in their diabetic patients. The reason for this isPain 55 Ϯ 26 77 Ϯ 24 Ͻ.0001 74 unclear. Despite this, the weight loss brought about byGeneral health 34 Ϯ 16 66 Ϯ 24 Ͻ.0001 59 gastric banding led to disease improvement and/or remis-Energy 36 Ϯ 23 61 Ϯ 23 Ͻ.0001 83 sion.Social function 54 Ϯ 32 81 Ϯ 26 Ͻ.0001 80Emotional role 50 Ϯ 44 85 Ϯ 28 Ͻ.0001 73 Our results have demonstrated that remission of T2DMMental health 52 Ϯ 26 75 Ϯ 19 is dependant on the amount of excess weight lost and the duration of the preoperative T2DM. Remission of T2DMAbbreviations as in Table 1. was not influenced by patient age, gender, or preoperativeData presented as mean Ϯ standard deviation. weight or BMI. Previous studies have reported similar find-Community scores obtained from Ware et al. [11]. ings, suggesting that patients with a shorter duration of the* Paired Student’s t test (2 sided). disease have better outcomes and thus have advocated for the need for early surgical intervention [15–17]. A receiver(P Ͻ.001) in patients with T2DM and IGT after a median operating characteristic curve plotting the sensitivity versusfollow-up of 14 months. 1 Ϫ specificity for the detection of resolution at varying years with T2DM showed 5 years to be in the region ofDiscussion greatest discrimination. Our patients with a preoperative duration of T2DM of Ͻ5 years had lower postoperative Our results indicate that laparoscopic gastric banding is HbA1c levels than those with Ͼ5 years’ duration. Usingeffective in preventing and promoting remission of T2DM logistic regression analysis, a preoperative duration ofin morbidly obese patients. Weight loss was associated with T2DM of Ͻ5 years conferred 6.5 times greater likelihood ofnear normal HbA1c values, with resolution and/or improve- T2DM resolving according to the model. The likely mech-ment of T2DM in 83% of patients. Weight loss also resulted anisms responsible for the improved glycemic control arein normglycemia in our patients with IGT. No patient who improvement in insulin sensitivity due to weight loss and acontrolled their IGT with diet progressed to require medi- tendency for beta-cell function to improve, especially incation. We found that weight loss was also associated with those patients with a shorter duration of T2DM [15]. Recentimprovement in dyslipidemia and hypertension, with 88% studies by Camastra et al. [20] showed that beta-cell func-of patients having improvement and/or resolution of the tion can fully recover 2 years after malabsorptive bariatricMS. All these improvements occurred after a short fol- surgery. Whether purely restrictive surgery results in similarlow-up period, with a weight loss of about 25 kg, even findings has not yet been determined.though the patients were still obese. Although the United Kingdom Prospective Diabetes Our results are consistent with those of previous studies, Study showed that intensive interventions targeting multiplewhich showed improvement in T2DM, hypertension, and risk factors reduces the risk of cardiovascular disease anddyslipidemia after bariatric surgery with Roux-en-Y gastric microvascular complications in patients with T2DM [21],bypass, biliopancreatic diversion [9,10], or gastric banding the median HbA1c values at 10 years were 7% in the[15–19]. Gastric bypass and biliopancreatic diversion pro- intensively treated group, with one half of the intensivelyduce rapid weight loss and remission in Ͼ80% of patients, treated group failing to achieve this glycemic target. Thewith normal glycemic control resulting. Dixon and O’Brien treatment targets have still not been met even in a large[15] found resolution and/or improvement of T2DM in 90% proportion of patients attending an Australian tertiary refer-of patients within 1 year of gastric banding, Dolan et al. ral hospital [22]. Gastric banding provides a single thera-[16], found a 66% remission at 2 years, Pontiroli et al. [19] peutic modality. Our results have shown that the improve-found a 45% rate of resolution within 4 years, and Ponce ment in the MS after weight reduction with gastric bandinget al. [17] an 80% remission rate at 2 years. Of our T2DM can significantly reduce important cardiovascular risk fac-patients, 85% achieved resolution and/or improvement. tors in subjects who are morbidly obese with T2DM,Moreover, laparoscopic gastric banding has certain dis-tinct advantages compared with gastric bypass and bilio-pancreatic diversion, including lower surgery-relatedmorbidity and mortality, reversibility, and ease of adjust-ability.

428 A. Brancatisano et al. / Surgery for Obesity and Related Diseases 4 (2008) 423– 429thereby potentially decreasing total and T2DM-related mor- in non-obese patients. Finally intervening early with gastricbidity and mortality [6,7,21]. banding offers the greatest chance of remission of T2DM. Given all these findings, gastric band placement should be Another consequence of the MS is the development of considered as therapy for T2DM and pre-diabetes in thenonalcoholic steatohepatitis. Insulin resistance, systemic morbidly obese.hypertension, and increased alanine aminotransferase havebeen shown to be independent predictors of nonalcoholic Disclosuressteatohepatitis in patients with morbid obesity [23]. Ourresults have shown a significant improvement in all liver The authors have no commercial associations that mightenzymes (Table 2) after weight loss, with a likely improve- be a conflict of interest in relation to this article.ment in fatty liver disease, consistent with the data fromDixon and O’Brien [15]. References Our study had a number of limitations. It was an obser- [1] O’Brien K, Webbie K. Health, well being and body weight: character-vational study with data collected prospectively. The study istics of overweight and obesity in Australia, 2001. Australian Institute ofhad no control group. Also, no published randomized con- Health and Welfare, 2004; Cat. No. AUS 43, Bulletin No. 13.trolled trials comparing medical and surgical treatment ofT2DM are available. The closest is the Swedish Obese [2] National Health and Medical Research Council. Clinical PracticeSubjects Study [24], which was a nonrandomized compar- Guidelines for the management of overweight and obesity in adults.ative study of 1700 subjects during a 10-year period who Canberra ACT, Australia: National Health and Medical Researchunderwent bariatric surgery. They reported a 36% resolu- Council; 2003. p. 1–265.tion rate of T2DM compared with 13% in matched controlsubjects. O’Brien et al. [25] recently published the first [3] Cameron AJ, Welbourn TA, Zimmet PZ, et al. Overweight andrandomized controlled clinical trial comparing gastric band- obesity in Australia: the 1999 –12000 Australian Diabetes, Obesitying and an intensive medical program in patients with a and Lifestyle Study (AusDiab). Med J Aust 2003;178:427–32.BMI of 30 –35 kg/m2. They found that at 2 years the patientsundergoing gastric banding had greater weight loss (87.2% [4] Grundy SM. Does the metabolic syndrome exist? Diabetes CareEWL) compared with the nonsurgical group (21.8% EWL). 2006;29:1689 –92.Although only 15 patients had the MS at the start of thestudy in either group, at the completion of the trial, only 1 [5] Wing R, Koeske R, Epstein I, et al. Long-term effects of modestgastric banding, but 8 nonsurgical, patients had the MS. weight loss in type 2 diabetic patients. Arch Intern Med 1987;147:Given these data, we have no reason to suspect that the large 1749 –53.differences observed in our data were the result of con-founding factors or chance alone. [6] Lean ME, Powrie JK, Anderson AS, et al. Obesity, weight loss and prognosis in type 2 diabetes. Diabet Med 1990;7:228 –33. Various studies have shown that T2DM has a negativeeffect on health-related quality of life by compromising [7] Williamson DF, Thompson TJ, Thun M, et al. Intentional weight lossphysical function, psychological status in the form of de- and mortality among over-weight individuals with diabetes. Diabetespression, and social relationships [26 –28]. Our study dem- Care 2000;23:1499 –504.onstrated significant improvement in the Beck depressionindex, as well as in psychosocial well-being, as judged by [8] Kundlacek S, Schernthaner G. The effect of insulin treatment on HbA1c,the SF-36, in morbidly obese T2DM and IGT patients who body weight and lipids in type 2 diabetic patients with secondary failureunderwent gastric banding. Our results are consistent with to sulfonylureas. Horm Metab Res 1992;24:478 – 83.previous studies that showed improvement in physical andemotional health after bariatric surgery [9,15]. [9] Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a systematic review and meta-analysis. JAMA 2004;292:1724 –37.Conclusion [10] Pories WJ, Swanson MS, MacDonald KG, et al. Who would have Weight loss after laparoscopic gastric banding led to thought it? An operation proves to be the most effective therapy forimprovement in a number of metabolic and clinical vari- adult-onset diabetes mellitus. Ann Surg 1995;222:339 –52.ables. We found improved glycemia in patients with T2DMand IGT. Most T2DM patients had a decrease in, or discon- [11] Cadiere GB, Bruyns J, Himpens J, Favretti F. Laparoscopic gastro-tinuation of, their OH and insulin medication. Also, signif- plasty for morbid obesity. Br J Surg 1994;81:1524.icant improvement occurred in the control of hypertension,dyslipidemia, and abnormal liver enzymes with resolution [12] Grundy SM, Cleeman JI, Daniels SR, et al. Diagnosis and manage-and/or improvement of the MS. In addition, significant ment of metabolic syndrome: an American Heart Association/Na-improvement occurred in health-related quality of life, with tional Heart, Lung, and Blood Institute scientific statement— execu-values returning to levels comparable to community norms tive summary. Circulation 2005;112:e285–90. [13] Ware J. Validity norm-based interpretation. In: SF-36 health survey: manual and interpretation guide. Boston: The Health Institute, New England Medical Centre; 1997. p. 10.1–10.38. [14] Cash T. Multidimensional body-self relations questionnaire: MBSRQ user’s manual. Norfolk, VA: Old Dominion University; 1994. [15] Dixon JB, O’Brien PE. Health outcomes on severely obese type 2 diabetic subjects 1 year after laparoscopic adjustable gastric banding. Diabetes Care 2002;25:358 – 62. [16] Dolan K, Bryant R, Fielding G. Treating diabetes in the morbidly obese by laparoscopic gastric banding. Obes Surg 2003;13:439 – 43. [17] Ponce J, Haynes B, Paynter S, et al. Effect of Lap-Band® induced weight loss on type 2 diabetes mellitus and hypertension. Obes Surg 2004;14:1335– 42. [18] Sjöstrom L, Lindroos A, Peltonen M, et al. Lifestyle, diabetes, and cardiovascular risk factors 10 years after bariatric surgery. N Engl J Med 2004;351:2683–2753.

A. Brancatisano et al. / Surgery for Obesity and Related Diseases 4 (2008) 423– 429 429[19] Pontiroli AE, Vedani P, Folli F, et al. Laparoscopic gastric banding [24] Sjöstrom CD, Lissner L, Wedel H, et al. Reduction in incidence of prevents type 2 diabetes and arterial hypertension and induces their diabetes, hypertension and lipid disturbances after intentional weight remission in morbid obesity. Diabetes Care 2005;28:2703–9. loss induced by bariatric surgery: the SOS intervention study. Obes Res 1999;7:477– 84.[20] Camastra S, Frascerra S, Manco M, et al. ␤-Cell function in severely obese type 2 diabetic patients. Diabetes Care 2007;30:1002– 4. [25] O’Brien PE, Dixon JB, Laurie C, et al. Treatment of mild to moderate obesity with laparoscopic adjustable gastric banding or and intensive[21] UK Prospective Diabetes Study Group. Tight blood pressure control medical program: a randomized trial. Ann Intern Med 2006;144: and risk of macrovascular and microvascular complications in type 2 625–34. diabetes: UKPDS 38. BMJ 1998;317:703–13. [26] Rubin PR, Peyrot M. Psychosocial problems and intervention in[22] Bryant W, Greenfield JR, Chisholm DJ, et al. Diabetes guidelines: diabetes. Diabetic Care 1992;15:1640 –57. easier to preach than to practice? A retrospective audit of out patient management of type 1 and type 2 diabetes mellitus. MJA 2006;185: [27] Lloyd CE, Matthews KA, Wing RR, et al. Psychosocial factors and 305–9. complications of IDDM. Diabetes Care 1992;15:166 –72.[23] Dixon JB, Bhathal PS, O’Brien PE. Weight loss and non-alcoholic fatty [28] Lustman PJ, Anderson RJ, Freedland KE, et al. Depression and poor liver disease: falls in ␥-glutamyl transferase concentrations are associated glycemic control: a meta-analytic review of the literature. Diabetes with histologic improvement. Obes Surg 2006;16:1278 – 86. Care 2000;23:934 – 42.

Surgery for Obesity and Related Diseases 5 (2009) 588 –597 Original article Safety and effectiveness of REALIZE adjustable gastric band: 3-year prospective study in the United StatesEdward Phillips, M.D.a,*, Jaime Ponce, M.D.b, Scott A. Cunneen, M.D.a, Sunil Bhoyrul, M.D.c,Eddie Gomez, M.D.d, Sayeed Ikramuddin, M.D.e, Moises Jacobs, M.D.f, Mark Kipnes, M.D.g, Louis Martin, M.D.h, Robert T. Marema, M.D.i, John Pilcher, M.D.j, Raul Rosenthal, M.D.k,Richard Rubenstein, M.D.l, Julio Teixeira, M.D.m, Thadeus Trus, M.D.n, Natan Zundel, M.D.k aCedars-Sinai Medical Center, Los Angeles, California bHamilton Medical Center, Dalton, Georgia cScripps Clinic Medical Group, La Jolla, California dMercy Hospital, Miami, Florida eUniversity of Minnesota Medical Center, Minneapolis, Minnesota fMercy Hospital, Miami, Florida gDGD Research, a Cetero Research Co, San Antonio, Texas hLouisiana State University School of Medicine, New Orleans, Louisiana iUS Bariatric, Fort Lauderdale, Florida jMethodist Specialty and Transplant Hospital, San Antonio, Texas kCleveland Clinic Florida, Weston, Florida lBrookhaven Memorial Hospital Medical Center, Patchogue, New York mSt. Luke’s Roosevelt Hospital Center, New York, New York nDartmouth, Hitchcock Medical Center, Lebanon, New Hampshire Received October 9, 2008; revised December 11, 2008; accepted December 18, 2008Abstract Background: The effectiveness and safety of bariatric surgery using laparoscopic adjustable gastric bands have been demonstrated in numerous published studies. We present the results of the first U.S. multicenter trial of the REALIZE adjustable gastric band, a laparoscopic adjustable gastric band previously available only outside the United States as the Swedish adjustable gastric band. Methods: A total of 405 morbidly obese patients were screened at 12 different centers from May to November 2003 to participate in a prospective, single-arm study of the safety and effectiveness of the laparoscopically implanted REALIZE band. Changes in excess body weight, the parameters of diabetes and dyslipidemia, and the incidence of complications were assessed at 3 years of follow-up. Results: Of the 405 patients, 276 (78.3% women and 61.2% white) qualified for the study. The average age was 38.6 Ϯ 9.4 years (range 18 – 61), and the preoperative body mass index was 44.5 Ϯ 4.7 kg/m2. The mean hospital stay was 1.2 Ϯ 1.3 days. At 3 years, the average excess weight loss was 41.1% Ϯ 25.1% or a decrease in the body mass index of 8.2 kg/m2 (18.6%) (P Ͻ .001). In diabetic patients with a baseline elevated hemoglobin A1c level, the level decreased by 1% (P Ͻ .001). The total cholesterol, low-density lipoprotein cholesterol, and triglycerides decreased by 9%, 16%, and 50%, respectively (P Ͻ .001), and the high-density lipoprotein cholesterol increased by 25% (P Ͻ .001) in patients with abnormal baseline values. One patient required conversion to an open surgical technique. No 30-day mortality occurred. The complication frequencies were generally low and included esophageal dysmo- tility in 0.4%, late balloon failure in 0.4%, band erosion in 0.4%, slippage in 3.3%, esophageal dilation in 3.3%, pouch dilation in 3.6%, catheter kinking in 1.1%, port displacement in 2.5%, and port *Correspondence to: [email protected]/09/$ – see front matter © 2009 American Society for Metabolic and Bariatric Surgery. All rights reserved.doi:10.1016/j.soard.2008.12.007

E. Phillips et al. / Surgery for Obesity and Related Diseases 5 (2009) 588 –597 589Keywords: disconnection in 4.3%. Reoperations were required in 15.2% of the patients and involved 2 band replacements, 9 band revisions, 5 port replacements, 22 port revisions, and 4 explants. Conclusion: The results of our study have shown that the REALIZE adjustable gastric band is safe and effective in a diverse U.S. population of morbidly obese patients. Significant weight loss was achieved throughout the 3 years of follow-up, with corresponding improvements in the indicators of diabetes and dyslipidemia. (Surg Obes Relat Dis 2009;5:588 –597.) © 2009 American Society for Metabolic and Bariatric Surgery. All rights reserved. Morbid obesity; Bariatric surgery; Gastric band; Swedish adjustable gastric band; SAGB; Body mass index; Weight loss; REALIZE; REALIZE adjustable gastric band; Laparoscopic adjustable gastric band; LAGB Obesity is a global epidemic [1–3]. The World Health The commonly accepted assessments of safety and effec-Organization projections have indicated that the number of tiveness such as those described in the proceedings of theobese adults (body mass index [BMI] 30 –39.9 kg/m2) [4] American Society for Metabolic and Bariatric Surgerywill increase from approximately 400 million to 700 million meetings and previous Food and Drug Administration (FDA) trials [43,44] were used.by 2015 [5]. This health problem is having a major effect in The REALIZE and SAGB brand names refer to the samenearly every country and culture. In 2004, 32.2% of adults line of LAGB devices. The present study was undertakenin the United States were obese and 4.2% (Ͼ10 million) by the manufacturer (Ethicon Endo-Surgery, Cincinnati,were morbidly obese (BMI Ն40 kg/m2, or Ն35 kg/m2 with OH) to obtain FDA approval to market the device in theՆ2 co-morbidities) [6]. Numerous serious conditions are United States. Thus, this is the first report of the use ofassociated with obesity, including hypertension, dyslipide- this line of LAGBs in the United States. For consistency,mia, type 2 diabetes mellitus, and coronary artery disease. this report refers to the band and its related componentsObesity also increases the risk of endometrial, breast, pros- by the REALIZE name, under which they will be approvedtate, and colon cancer [7–14], and it is associated with a in the United States. Currently, these banding productssignificant decline in quality of life (QOL) [15–18] and life are available only under the name of SAGB outside theexpectancy [19,20]. United States. Bariatric surgery is recommended as a long-term treatment Methodsoption by the U.S. health advisories [4]. Unlike medicallysupervised diets and programs, bariatric surgery induces and Study centerssustains weight loss in the long term, resolves or improvesco-morbidities [21–23], and increases longevity in morbidly The present study was registered at the National Insti-obese patients [24 –27]. A recent systematic review of the tutes of Health Web site (www.clinicaltrials.gov identifier:published data concluded that all current bariatric operations NCT00166205). Twelve academic and/or private practiceachieve a major and durable weight loss in the medium term clinical sites were selected to participate in the present studyand that Roux-en-Y gastric bypass achieves greater weight loss according to laparoscopic surgery experience and multidis-than gastric banding at 1 and 2 years but with no difference at ciplinary team approach. All surgeons in the present studyϾ2 years [28]. Placement of a laparoscopic adjustable gastric had undergone didactic, case-study, and hands-on trainingband (LAGB) results in significant improvements in co-mor- in REALIZE band placement and band adjustment. On-sitebidities (e.g., type 2 diabetes mellitus, dyslipidemia) [29 –34] proctoring was conducted by surgeons experienced inand QOL [35–37]. LAGB has also been deemed the safest of SAGB implantation outside the United States. All studyall bariatric operations [38,39], and it is associated with only centers followed a common protocol approved by theirone-tenth the operative mortality rate of gastric bypass [40]. respective institutional review board.LAGB has an odds ratio for complications 3.5 times less thanthat of other laparoscopic bariatric operations [41], and it is Study populationmostly reversible. The study subjects included patients seeking weight loss Although an extensive body of clinical data is available surgery at the study sites. Male and female patients agedon the use of the Swedish adjustable gastric band (SAGB) 18 – 61, with a 5-year history of morbid obesity and docu-outside the United States [42], no data are available from mented weight loss failure, willing to commit to 3 years ofstudies performed in the United States. In the United States, follow-up were given the opportunity to participate if theyonly 1 device, the Lap-Band (Allergan, Irvine, CA) was did not have any of the following exclusions: presence ofavailable for LAGB placement before the recent availability inflammatory disease or congenital or acquired anomalies ofof the REALIZE adjustable gastric band. The aim of this the gastrointestinal tract, hiatal hernia, severe cardiopulmo-prospective, multicenter study was to characterize the safety nary disease, gastrointestinal bleeding disorders, cirrhosis,and effectiveness of the REALIZE band in inducing weight intestinal telangiectasia, esophageal dysmotility, gastric re-loss in a U.S.-based population of morbidly obese patients.

590 E. Phillips et al. / Surgery for Obesity and Related Diseases 5 (2009) 588 –597flux, Barrett’s esophagus, chronic pancreatitis, terminal ill- postoperative inflammation and to allow for a more com-ness, previous malabsorptive or restrictive bariatric surgery, fortable transition from a liquid to a solid diet. The initialimpaired mental status (using the Diagnostic and Statistical band adjustment (3– 4 cc) was made no sooner than 28 daysManual of Mental Disorders, 4th edition, criteria), use of after surgery, if the weight loss was within a 1–2 lb/wkprescription medications or over-the-counter supplements (.5–1 kg/wk) range and in the absence of symptoms offor weight reduction, acute or chronic infection (localized or obstruction or pain with eating. Subsequent band adjust-systemic), and a known or suspected allergy to silicone. ments were made according to clinical progress (i.e., weightWomen of childbearing age who were pregnant, lactating, loss, dietary and behavioral changes, and absence of com-or not practicing birth control were also excluded. plaints) at the discretion of the investigator, but at intervals of Ն28 days. Although the port location was performedSurgical technique either under fluoroscopy or by palpation, the level of stoma restriction was always verified fluoroscopically (i.e., limited The REALIZE band was implanted laparoscopically using upper gastrointestinal series).the pars flaccida technique, a method created for the im-plantation of the SAGB that has been instrumental in re- Study variablesducing short- and longer term gastric banding morbidity[42,45]. The patients were placed in a mild reverse Tren- The percentage of excess weight loss (%EWL) was thedelenburg position and supine or lithotomy position. The primary effectiveness variable and was calculated as theliver was retracted, and the gastroesophageal and diaphrag- difference in the baseline and postoperative weight dividedmatic regions were inspected. At the angle of His, a limited by the difference in the baseline weight and ideal bodydissection was performed until exposure of the left crus was weight (i.e., upper limit value of the medium frame rangeobtained, creating space for an articulating blunt dissector to from the Metropolitan Tables for Life Insurance, 1983),emerge later in the procedure. Occasionally, a large epigas- multiplied by 100. To minimize site-to-site variability intric fat pad was removed, if necessary, for placement of the weight measurements, all study sites used a calibrated dig-band. An opening in the lesser omentum in the avascular ital scale of the same model and capacity (Tanita Corpora-pars flaccida region was created, the right crus was identi- tion of America, Arlington Heights, IL).fied, and blunt dissection was performed posteriorly near thejunction of the crura. The dissection plane was kept away Other assessments included changes in BMI, absolutefrom the lesser curvature perigastric fat. A retrogastric tun- weight, parameters of co-morbidity (i.e., glycosylated hemo-nel was created by passing an articulating blunt dissector globin [HbA1c], total, low-density lipoprotein, and high-den-from the right crus to the angle of His. The articulating sity lipoprotein cholesterol, and triglycerides), and QOL.dissector tip was then deflected anteriorly to emerge in the Changes in QOL were evaluated by the 36-item short-formsmall opening created previously at the angle of His. Before Health Survey and the Beck Depression Inventory II, 2ndintroducing the REALIZE band into the abdominal cavity, all edition. The QOL results will be the subject of a separate reportair was aspirated from the balloon and a knot was tied at the and have not been presented in the present study. All labora-end of the tubing. A suture loop was placed at the band tory tests (including a serum pregnancy test) were performedlock-end flap using nonabsorbable 0-0 or 2-0 sutures. The by a central laboratory (Quintiles Laboratory, Atlanta, GA).band was introduced through a 15-mm working port. The safety of the REALIZE band was assessed according to The band was pulled through the retrogastric tunnel incidence of device-related complications. Study data wereusing the suture loop secured in the articulating dissector tip collected preoperatively, on the day of surgery, during hos-slot, taking care to orient the balloon toward the stomach pitalization at 4 – 6 weeks, and at 2, 4, 6, 8, 10, 12, 15, 18,wall. The band’s tabs were locked and secured with 2 21, 24, 28, 32, and 36 months.sutures. An anterior gastrogastric plication was performed,using 2 or 3 permanent sutures, from the anterior fundus Statistical analysisdistal to the band to the anterior gastric pouch, creating aloose gastric tunnel to accommodate inflation of the band. The primary effectiveness one-sided hypothesis was formu-The tubing was exteriorized through an upper subcostal lated according to the 36.2% EWL (95% confidence intervalworking port and attached to the connector end of the port. 32.6 –39.8%) at 3 years reported in the FDA submission for theThe band’s injection port was secured to the anterior rectus previously marketed Lap-Band system. The difference in theabdominal muscle fascia with nonabsorbable sutures. mean %EWL (36.2%) and the noninferiority margin (3.6%) resulted in a target primary endpoint of 32.6% EWL at 3 yearsBand adjustments and required a sample size of 276 patients. The %EWL was computed using a 1-sided t test. The ␣ was set at P Ͻ .05 for The injection port was located either by fluoroscopy or all analyses. For the effectiveness analyses, the patients servedpalpation. The band was kept empty during the first 6 as their own controls, and it was assumed that patients’ bodypostoperative weeks to avoid overrestriction resulting from weight and parameters of co-morbidity would not change sig- nificantly without intervention. The analyses of effectiveness

E. Phillips et al. / Surgery for Obesity and Related Diseases 5 (2009) 588 –597 591Fig. 1. Flowchart of study patients.parameters were conducted using 2-sided paired t tests. The Resultsdegree of site-to-site variability and other possible factors, suchas preoperative weight, gender, age, and ethnicity, which could Of the 405 patients screened at 12 sites, 276 qualified forpossibly influence %EWL, was assessed using analysis of the study and underwent surgery with REALIZE band place-variance. ment from May 2003 to November 2006 (Fig. 1). The reasons for screening failure included impaired mental sta- Sensitivity analyses were performed for 3 study popula- tus in 62, consent withdrawn or refused by 37, failedtions: (1) intent-to-treat (ITT): all patients implanted with weight/BMI eligibility criteria in 10, miscellaneous ineligi-the REALIZE band; (2) evaluable: only patients completing bility criteria for 12, and unknown for 8. The preoperativethe 3-year follow-up period; and (3) per protocol: only characteristics of the ITT population (n ϭ 276) are listed inpatients completing the 3-year follow-up who did not have Table 1.major protocol deviations that would affect the analysis ofdevice effectiveness. In multiyear studies, it is expected that Of the 276 enrolled patients, 228 (83%) completed 3some patients will not attend the study-required (scheduled) years of follow-up and 48 patients discontinued participa-visits; therefore, to determine the potential bias of missing tion before study completion for the following reasons:data in the conclusions, the following sensitivity analyses closing of 1 study site, as a consequence of Hurricanewere performed, imputing the missing weight values: (1) Katrina, in New Orleans (n ϭ 26), lost to follow-up (n ϭlast observation carried forward: the missing values were 16), complications (n ϭ 3, of whom 2 had the band ex-replaced by the corresponding value from the previous visit; planted), withdrawal of consent (n ϭ 2), and death (n ϭ 1).(2) worst observation carried forward: the missing %EWL One patient required intraoperative conversion to an openvalues were replaced by the patient’s worst %EWL value procedure because of dense intra-abdominal adhesions fromobtained during the 3-year follow-up period; and (3) zero: previous surgery.missing %EWL values were assumed to be 0. Because theresults for all 3 populations did not significantly differ, only The average operative time was 73.8 Ϯ 21.8 minutesthe ITT (last observation carried forward) results have been (range 35–184), and the mean hospital stay was 1.2 Ϯ 1.3included in this report, except as specified otherwise. days (range 0.2–21.2). Compliance with postoperative visit attendance was 72.1–96.4%. On average, patients received Statistical analyses were performed using Statistical 6 Ϯ 3.3 band adjustments within 3 years. The medianAnalysis Systems, version 8.2, for Windows (SAS Institute, number of adjustments during the first, second, and thirdCary, NC), and nQuery Advisor (Statistical Solutions, Sau- year of follow-up was 4 (range 1–11), 2 (range 1–7), and 2gus, MA) was used for sample size estimation. (range 1–9), respectively.

592 E. Phillips et al. / Surgery for Obesity and Related Diseases 5 (2009) 588 –597Table 1 effectiveness in producing significant reductions in excessPreoperative characteristics (n ϭ 276) body weight. Table 2 summarizes the results of the sensi- tivity analyses for weight loss. Because the results for all 3Variable Value populations did not differ significantly, only the ITT with last observation carried forward results were presented.Age (yr) 38.6 Ϯ 9.4 Mean Ϯ SD 18–61 The mean %EWL at 3 years was 41.1% Ϯ 25.1% (P Ͻ Range .001). This translated to an absolute decrease in the BMI of 276.5 Ϯ 40.8 8.2 Ϯ 5.1 kg/m2. Of all the patients, 24 (10.5%) achievedAbsolute weight (lb) 193.6–415.4 Ն75% EWL, 80 (35.1%) Ն50%, 143 (62.7%) Ն33%, and Mean Ϯ SD 176 (77.2%) Ն25%. However, 5 patients (2.2%) gained Range 146.5 Ϯ 13.4 Ͼ5% excess weight, and 6 (2.6%) had no change (i.e., Ϫ5% 121–197 to 5% EWL). No significant differences were found inIdeal body weight (lb) weight loss by age group or ethnicity (Table 3). However, Mean Ϯ SD 130.0 Ϯ 33.1 the absolute average excess weight lost by female patients Range 61.6–233.8 was 8% (P ϭ .024) greater than that of the male patients. When EWL was compared for subgroups of patients definedExcess body weight (lb) 44.5 Ϯ 4.7 by preoperative BMI, the average %EWL was significantly Mean Ϯ SD 35.0–58.1 greater for patients with a BMI of 35–39 kg/m2 (50.9%, Range P ϭ .013) and significantly lower for the super-obese 60 (21.7) (35.0%, P ϭ .013). The site-to-site variation in %EWL wasBody mass index (kg/m2) 216 (78.3) 15.3% Ϯ 28.5% to 61.2% Ϯ 28.4%. Mean Ϯ SD Range 169 (61.2) The site with the fewest patients (n ϭ 10) also had the 67 (24.3) lowest %EWL (20.0% Ϯ 25.9%); however, among theGender (n) 33 (12.0) remainder of the sites, no relationship was found between Male 4 (1.4) the volume of patients and the %EWL. Female 3 (1.1) 47 (17.8) On multivariate analyses, gender (F ϭ 25.7), study siteEthnicity (n) 52 (18.8) (F ϭ 3.8), preoperative weight (F ϭ 46.3), and follow-up White, non-Hispanic 74 (26.8) visit month (F ϭ 447.8) were all significant co-variables of Hispanic 43 (15.6) the %EWL (P Ͻ .0001). Black, non-Hispanic 20 (7.2) Asian/Pacific Islander 81 (29.3) The course for weight loss is shown in Fig. 2. The Other 118 (42.8) %EWL increased between the 4 – 6-week and 28-month 95 (34.4) points and remained relatively stable between the 28- andDiabetes mellitus (n) 24 (8.7) 36-month points. At 12 months, the average %EWL wasHypercholesterolemia (n) 51 (18.5) 38.0% Ϯ 17.7% and at 24 months was 44.7% Ϯ 22.4%.Sleep apnea (n)Osteoarthritis (n) Effects on co-morbidityCholelithiasis (n)Gastroesophageal reflux (n) Of the 45 patients with a documented history of diabetes,Hypertension (n) 31 had elevated HbA1c (7.99% Ϯ 1.39%) preoperatively. AtDepression (n) 3 years after implantation, the HbA1c in these patientsUnspecified hyperlipidemia (n) averaged 6.72% Ϯ 1.02% (P Ͻ .001), an absolute decreaseAsthma (n) Percentage of ideal body weight and excess body weight determinedusing Metropolitan Weight Tables for Life Insurance, 1983. Data in parentheses are percentages.Weight loss Sensitivity analyses assessing the effect of using differ-ent patient denominators and different methods of handlingmissing data supported the finding of the REALIZE band’sTable 2Percentage of excess weight loss at 3 years—sensitivity analysisVariable ITT LOCF ITT WOCF ITT zero Evaluable Per protocolPatients (n) 276 276 276 228 164Mean 41.1 37.2 35.3 42.8 39.6SD 25.1 26.4 28.2 25.4 23.9Median 39.2 34.5 34.3 40.5 38.8Range Ϫ26.5 to 129.3 Ϫ27.1 to 129.3 Ϫ25.6 to 129.3 Ϫ25.6 to 129.3 Ϫ25.6 to 107.995% CI 38.1–44.1 34.1–40.3 32.0–38.7 39.5–46.1 35.9–43.3P value* Ͻ.001 Ͻ.001 Ͻ.001 .002 .054 LOCF ϭ last observation carried forward; WOCF ϭ worst observation carried forward; zero ϭ missing %EWL values assumed to be 0; evaluable ϭpatients completing 3-year follow-up period; per protocol ϭ patients completing 3-year follow-up without major protocol deviations that would affect analysisof device; CI ϭ confidence interval; %EWL ϭ percentage of excess weight loss. * P value using t test for testing H0: ␮ Յ32.6% versus Ha: ␮ Ͼ32.6%.

E. Phillips et al. / Surgery for Obesity and Related Diseases 5 (2009) 588 –597 593Table 3 P-valuePatient weight loss Ͻ.001* Ͻ.001†EWL category 3-yr Postoperative findings Ͻ.001† Patients (n ϭ 276) Mean Ϯ SD Range NS‡ .024‡%EWL 276 41.1 Ϯ 25.1 Ϫ26.5 to 129.3EBW reduction (lb) 276 51.3 Ϯ 30.8 Ϫ34.2 to 133.2 NS‡BMI reduction (kg/m2) 276 8.2 Ϯ 5.1 Ϫ5.9 to 21.3%EWL by age group (yr) .013‡ 48 36.4 Ϯ 22.8 Ϫ26.5 to 89.0 Ͻ30 157 41.5 Ϯ 25.6 Ϫ25.6 to 129.3 30–44 71 43.4 Ϯ 25.2 Ϫ24.5 to 119.2 Ͼ45%EWL by gender 60 34.7 Ϯ 20.2 Ϫ5.8 to 96.7 Men 216 42.9 Ϯ 26.0 Ϫ26.5 to 129.3 Women%EWL by ethnicity 169 42.5 Ϯ 27.7 Ϫ26.5 to 129.3 White, non-Hispanic 67 38.5 Ϯ 21.2 Ϫ5.8 to 109.3 Hispanic 33 37.3 Ϯ 19.3 Black, non-Hispanic 7 48.3 Ϯ 15.1 5.5 to 77.0 Other 31.9 to 77.7%EWL by preoperative BMI (kg/m2) 42 50.9 Ϯ 27.1 35–39 102 40.1 Ϯ 26.7 3.9 to 129.3 40–44 88 40.6 Ϯ 22.3 Ϫ25.6 to 109.3 45–49 44 35.0 Ϯ 21.4 Ϫ26.5 to 85.5 Ͼ50 Ϫ3.2 to 73.7EWL ϭ excess weight loss; EBW ϭ excess body weight; BMI ϭ body mass index; NS ϭ nonsignificant.* Primary effectiveness variable: single-sided t test using clinically relevant mean value.† Secondary effectiveness variable: paired 2-sided t test relative to baseline measures.‡ According to analysis of variance model with gender subgroup and study site as factors.of 1.3 percentage points. Of the 179 patients with no doc- line. High-density lipoprotein cholesterol increased 25% and 20% in patients with low and normal preoperativeumented history of diabetes at baseline, 17 had an elevated levels, respectively (P Ͻ .001). Triglycerides decreasedHbA1c (6.75% Ϯ 0.69, which had decreased to 5.81% Ϯ 50% and 24% in patients with elevated and normal preop-0.24% (P Ͻ .001) after 3 years (Table 4). The HbA1c erative triglyceride levels, respectively (P Ͻ .001; Table 4).remained essentially unchanged in patients with normal Safetypreoperative values. The most frequent perioperative (Ͻ30-d) adverse events Patients with elevated total cholesterol and low-density were nausea (23.6%), pain (22.8%), and vomiting (10.9%). In the postoperative period (Ͼ30 d), vomiting (40.6%),lipoprotein cholesterol preoperatively had experienced a 9%and 16% decrease, respectively (P Ͻ .001) at 3 years post-operatively. The changes in these parameters were not sta-tistically significant in patients with normal values at base- Fig. 2. %EWL with time.

594 E. Phillips et al. / Surgery for Obesity and Related Diseases 5 (2009) 588 –597Table 4Mean changes from baseline in co-morbidity parameters at 3 yearsBefore implantation Patients* (n) Follow-up Mean Ϯ SD (range) Change (mg/dL) Ϫ1.27†Diabetes 31 Baseline 7.99 Ϯ 1.39 (6.3–11.8) HbA1c elevated 14 36 mo 6.72 Ϯ 1.02 (5.3–9.5) 0.21 Baseline 5.51 Ϯ 0.34 (5.0–6.1) HbA1c normal 17 36 mo 5.73 Ϯ 0.26 (5.4–6.2) Ϫ0.94† 162 0.07No diabetes Baseline 6.75 Ϯ 0.69 (6.2–8.4) HbA1c elevated 105 36 mo 5.81 Ϯ 0.24 (5.3–6.1) Ϫ21 (9%)† 110 Baseline 5.39 Ϯ 0.33 (4.7–6.1) 1.5 (1%) HbA1c normal 36 mo 5.46 Ϯ 0.38 (4.5–6.7) 65 11 (25%)†Total cholesterol‡ 153 Baseline 233.8 Ϯ 25.7 (201–306) 10.4 (20%)† Elevated 36 mo 212.8 Ϯ 34.3 (138–310) 83 Baseline 175.1 Ϯ 18.1 (134–200) Ϫ24.4 (16%)† Normal 127 36 mo 176.6 Ϯ 29.9 (115–279) 0.6 (Ͻ1%)HDL cholesterol§ 24 Baseline 34.0 Ϯ 4.3 (23–39) Ϫ201.8 (50%)† Decreased 192 36 mo 45.0 Ϯ 10.2 (23–72) Ϫ34.2 (24%)† Baseline 51.0 Ϯ 8.2 (40–73) Normal 36 mo 61.4 Ϯ 13.7 (37–109)LDL cholesterol¶ Baseline 156.0 Ϯ 20.5 (131–209) Elevated 36 mo 131.6 Ϯ 32.5 (60–219) Baseline 103.3 Ϯ 18.1 (55–130) Normal 36 mo 103.9 Ϯ 27.3 (44–171)Triglyceridesʈ Baseline 407.6 Ϯ 180.0 (256–940) Elevated 36 mo 205.8 Ϯ 125.3 (65–541) Baseline 141.9 Ϯ 50.5 (47–248) Normal 36 mo 107.7 Ϯ 42.97 (34–288)HbA1c ϭ glycosylated hemoglobin (normal range 4.3– 6.1%); LDL ϭ low-density lipoprotein; HDL ϭ high-density lipoprotein.* Patients with laboratory parameters at baseline and 3 yr postoperatively.† P Ͻ .001 by t test analysis.‡ Normal reference range: 130 –200 mg/dL.§ Normal reference range: 40 – 80 mg/dL.¶ Normal reference range: 0 –130 mg/dL.ʈ Normal reference range: 45–250 mg/dL.constipation (17.8%), gastric reflux (17.8%), and nausea investigator until 5 months after the hurricane (17 months(15.6%) were the most frequent adverse events. The inci- after the hernia repair) when the patient underwent surgerydence of device-related complications was low (Table 5). Of to replace and relocate the port. At 12 hours after thisthe 276 patients, 42 (15.2%) required reoperations: band surgery, the patient became acutely ill. The patient died ofrevision in 9, band replacement in 2, injection port replace- multiorgan system failure 89 hours after the latest surgery.ment in 5, injection port revision in 22, and band explana- The perioperative clinical course and subsequent postmor-tion in 4. Of the 4 explantations, 1 was due to esophageal tem analysis identified a gastric perforation causing intra-reflux, 1 to band erosion, and 2 to patient dissatisfaction abdominal sepsis as a major contributing factor to the pa-with their weight loss. tient’s death. This might have been an acute process, as suggested by the postmortem findings; however, an unrec- One patient in the study died. The causality was probably ognized band erosion, resulting in a chronically dischargingrelated to port replacement surgery. The patient was seen wound site, might have been a contributing factor. Theseapproximately 23 months after implantation with an incar- events would not have occurred in the absence of gastriccerated trocar site hernia in close proximity to the port site. band implantation.The patient underwent an uneventful hernia repair. After thehernia repair, the patient developed a chronically draining Discussionwound at the trocar site that did not respond to standardantibiotic therapy. In the aftermath of Hurricane Katrina and This is the first trial of the REALIZE band conducted in a U.S.the subsequent disruption of healthcare in Louisiana, the population. Although participating centers had varying degreespatient was unable to receive follow-up care by the study

E. Phillips et al. / Surgery for Obesity and Related Diseases 5 (2009) 588 –597 595Table 5 changes in co-morbid conditions. In terms of complications,Incidence of complications and reoperations in present study and both studies used different medical terminology coding sys-Lap-Band A trial tems for incidence trending (Coding Symbols for Thesaurus of Adverse Reaction Terms was used in the Lap-Band trialEvent Present study Lap-Band [44] and Medical Dictionary for Regulatory Activities was used (n ϭ 276; ITT) (n ϭ 299) in the SAGB trial).Band related Ͻ1 (1) 10 (29)* However, since the completion of the Lap-Band A trial, Esophageal dysmotility 3.3 (9) many improvements in the surgical technique (e.g., pars Esophageal dilation Ͻ1 (1) Ͻ1 (1) flaccida) and postoperative care have been introduced. Such Band leak Ͻ1 (1) 1 (3) improvements are the most probable source of the differ- Erosion 3.3 (9) ences in complications seen in both studies (summarized in Band slippage 3.6 (10) 24 (72)* Table 5). The most pronounced difference was perhaps the Pouch dilation 4.3 (12) incidence of slippage/pouch dilation and esophageal dys- Soma obstruction 9.4 (26) 14 (41) motility, which could be attributed to the pars flaccida Dysphagia 19.2 (53) 9 (26) technique. It has been well documented in the published Gastroesophageal reflux 34 (103) data that when the analysis of complications with LAGB 1.1 (3) has controlled for the use of the pars flaccida technique, theInjection port/catheter related 2.5 (7) NR rates of these complications will be reduced by as much as Kinking of catheter 4.3 (12) 6 (18) 50% [42]. However, the low-pressure properties of the Displacement 6.5 (18) SAGB might also have contributed to the lower incidence of Post disconnection Ͻ1 (2) NR events such as dysphagia and reflux. Injection port-site pain† 9 (26) Port leaks 1.5 (4) 3 (8) Although the participating surgeons had varying degrees Ͻ1 (2) of laparoscopic and gastric banding experience, the opera-Reoperations 3.3 (9) 15.4 (46) tive morbidity was minimal. The incidence of explantation Band explants 1.8 (5) 3 (9) was 1.5% (.7% because of device-related complications), Band removed and replaced 5 (16) significantly lower than the 3– 6% rate reported in large Band revised/not replaced 8 (22) series [47]. Only 1 patient in our study died (.4% mortality). Port removed and replaced 4.3 (13) These initial U.S. results of the use of the REALIZE band are Port revised/not replaced 4.3 (13) in the range of the current European, Australian, and South American LAGB weight-loss outcomes, as evidenced in the NR ϭ not reported. world bariatric meta-analyses of Buchwald et al. [48] and Data presented as percentages, with numbers in parentheses. Maggard et al. [49]. Finally, the levels of glycemia and * Complications reported under single category in Lap-Band A trial. dyslipidemia both improved coincident with weight loss in † All Reveal band device-related complications occurred in postopera- the REALIZE band patients. These improvements would betive period (Ͼ30 d postoperatively), except for 15 (5.4%) of injection expected to have a positive effect on cardiovascular mor-port-site pain, which were reported perioperatively. bidity and mortality because diabetes and lipid abnormali- ties are both well-established independent risk factors forof laparoscopic and gastric banding experience, as confirmed coronary artery disease.by the analysis of variance findings, significant weight loss wasachieved and morbidity was low at 3 years after surgery. ConclusionAnalyses of an ITT population are intended to avoid variousmisleading artifacts that can arise in intervention research. For The results of the present study have demonstrated theexample, if people who have a more refractory or serious safety and effectiveness of the laparoscopic REALIZE band inproblem tend to withdraw at a greater rate, even an ineffective achieving substantive weight loss and improving patienttreatment could appear to be providing benefits if only those co-morbidities. This evidence, together with the previouswho finished the treatment period were analyzed. In the present evidence supporting LAGB overall, including, most re-study, the analysis of the ITT population found that the REALIZE cently, a controlled trial of LAGB versus conventionalband was equally effective for all ages and ethnicities studied, weight loss therapy [46], suggests that bariatric surgerywith women attaining greater %EWL than men. The 35–39- using LAGB techniques might be a preferred approach tokg/m2 BMI group experienced the most weight loss, signifi- weight loss for selected morbidly obese patients.cantly better than that attained by the super obese. A Ͼ50%EWL was attained by 35.1% of patients, and most patients Acknowledgments(62.7%) lost Ն33% EWL. The slight weight regain evidencedafter 28 months is typical of the weight-loss curve for banding The authors would like to thank Mario Gutierrez, MSHS,and bypass and is generally followed by weight stabilization Sheryl Helsinger, Shana Zink, Kara Haas, MD, MPH (Ethi-[44,46]. The mean %EWL at 3 years achieved with the SAGB(41.1% Ϯ 25.1%) was 5% greater than that reported for theLap-Band A trial for FDA approval (36.2% Ϯ 24.6%). Thechange in the BMI, however, was the same in both studiesat 8 kg/m2 [37,44]. The Lap-Band A trial did not include

596 E. Phillips et al. / Surgery for Obesity and Related Diseases 5 (2009) 588 –597con Endo-Surgery, Cincinnati, OH), and Medwrite Medical [14] National Institutes of Health. Third report of the National CholesterolCommunications (Maiden Rock, WI) for their substantial Education Program Expert Panel on the detection, evaluation, andcontributions to the development of this report. treatment of high blood cholesterol in adults (Adult Treatment Panel III). NIH publication no. 02-5215. Bethesda: National Institutes ofDisclosures Health; 2002. This study was undertaken by the manufacturer of the [15] Fontaine KR, Cheskin LJ, Barofsky I. Health-related quality of lifeREALIZE adjustable gastric band, Ethicon Endo-Surgery, Inc. of obese persons seeking treatment. J Fam Pract 1996;43:265–7.(Cincinnati, OH) to obtain Food and Drug Administrationapproval to market it; therefore, the study was financed in [16] Tolonen P, Victorzon M, Makela J. Impact of laparoscopic adjustableits entirety by the sponsor. Additionally, some authors of gastric banding for morbid obesity on disease-specific and health-this report are consultants for Ethicon Endo-Surgery, Inc. related quality of life. Obes Surg 2004;14:788 –95.References [17] Karlsson J, Sjöström L, Sullivan M. Swedish obese subjects (SOS): an intervention study of obesity: 2-year follow-up of [1] National Institutes of Health. Gastrointestinal surgery for severe obe- health-related quality of life (HRQL) and eating behavior after sity: consensus statement 1991. Bethesda, MD: NIH Consensus De- gastric surgery for severe obesity. Int J Obes Relat Metab Dis velopment Conference; 1991;9:1–2. 1998;22:113–26. [2] Deitel M. Overweight and obesity worldwide now estimated to in- [18] Tolonen P, Victorzon M. Quality of life following laparoscopic ad- volve 1.7 billion people. Obes Surg 2003;13:329 –30. justable gastric banding—the Swedish band and the Moorhead-Ardelt questionnaire. Obes Surg 2003;13:424 – 6. [3] Mokdad AH, Serdula MK, Dietz WH, et al. The spread of the obesity epidemic in the United States, 1991–1998. JAMA 1999; [19] Allison DB, Fontaine KR, Manson JF, Stevens J, VanItallie TB. 282:1519 –22. Annual deaths attributable to obesity in the United States. JAMA 1999;282:1530 – 8. [4] National Heart, Lung and Blood Institute and National Institute of Diabetes and Digestive and Kidney Diseases. Clinical guidelines on [20] Mokdad AH, Marks JS, Stroup DF, et al. Actual causes of death in the the identification, evaluation, and treatment of overweight and obesity United States, 2000. JAMA 2004;291:1238 – 64. in adults: the evidence report. NIH publication no. 98-4083. Be- thesda: National Institutes of Health; 1998. [21] Pories WJ, Swanson MS, MacDonald KG, et al. Who would have thought it? An operation proves to be the most effective therapy for [5] World Health Organization. Obesity and Overweight. Fact Sheet No. adult-onset diabetes mellitus. Ann Surg 1995;222:339 –52. 311, September 2006. Available from: www.who.org. Accessed Oc- tober 17, 2007. [22] Goodrick GK, Foreyt JP. Why treatments for obesity don’t last. JAMA 1991;91:1243– 47. [6] Ogden CL, Carroll MD, Curtin LR, et al. Prevalence of overweight and obesity in the United States, 1999 –2004. JAMA 2006;295: [23] O’Brien PE, Dixon JB, Laurie C, et al. Treatment of mild to 1549 –55. moderate obesity with laparoscopic adjustable gastric banding or an intensive medical program: a randomized trial. Ann Intern Med [7] North American Association for the Study of Obesity and National 2006;144:625–33. Heart, Lung, and Blood Institute. The practical guide: identification, evaluation, and treatment of overweight and obesity in adults. NIH [24] Sjöström L, Narbro K, Sjöström CD. Effects of bariatric surgery publication no. 00-4084. Bethesda: National Institutes of Health; on mortality in Swedish obese subjects. N Engl J Med 2007;357: 2000. 741–52. [8] Buchwald H, for the Consensus Conference Panel. Bariatric sur- [25] Christou NV, Sampalis JS, Liberman M, et al. Surgery decreases gery for morbid obesity: health implications for patients, health long-term mortality, morbidity, and health care use in morbidly obese professionals, and third-party payers. J Am Coll Surg 2005;200: patients. Ann Surg 2004;240:416 –23. 593– 604. [26] Adams TD, Gress RE, Smith SC, et al. Long-term mortality after [9] Ellison MM, Mulcahy HE. Obesity: weighing up the cardiovascu- gastric bypass surgery. N Engl J Med 2007;357:753– 61. lar risks. Br J Cardiol 2001;8:61– 4. [27] Flum DR, Dellinger EP. Impact of gastric bypass operation on[10] Committee Report. Report of the Expert Committee on the Diagnosis survival: a population-based analysis. J Am Coll Surg 2004;199: and Classification of Diabetes Mellitus. Diabetes Care 2002;26(suppl 543–51. 1):S5-S2. [28] O’Brien PE, McPhail T, Chaston TB, Dixon JB. Systematic review of[11] National Institutes of Health. Sixth Report of the Joint National medium-term weight loss after bariatric operations. Obes Surg 2006; Committee on Prevention, Detection, Evaluation and Treatment of 16:1032– 4. High Blood Pressure. NIH publication no. 98-4080. Bethesda: Na- tional Institutes of Health; 1997. [29] Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a systematic review and meta-analysis. JAMA 2004;292:1724 –37.[12] American Association of Clinical Endocrinologists. American Asso- ciation of Clinical Endocrinologists medical guidelines for clinical [30] Ponce J, Haynes B, Paynter S, et al. Effect of Lap-Band-induced practice for the diagnosis and treatment of dyslipidemia and preven- weight loss on type 2 diabetes mellitus and hypertension. Obes Surg tion of atherogenesis, 2002, amended version. Endocrine Pract 2000; 2004;14:1335– 42. 6:19 –52. [31] O’Brien PE, Dixon JB, Smith A. The laparoscopic adjustable gastric[13] Centers for Medicare and Medicaid Services. National coverage anal- band: a prospective study of medium term effects on weight, health ysis (NCA): continuous positive airway pressure (CPAP) therapy for and quality of life. Obes Surg 2002;12:652– 6. obstructive sleep apnea. OSA no. CAG– 00093R. N Engl J Med 1996;334:99 –104. [32] Dixon JB, O’Brien PE. Changes in comorbidities and improvements in quality of life after Lap-Band placement. Am J Surg 2002;184: 51S– 4S. [33] Frigg A, Peterli R, Peters T, et al. Reduction in co-morbidities 4 years after laparoscopic adjustable gastric banding. Obes Surg 2004;14: 216 –23. [34] Dolan K, Bryant R, Fielding GA. Treating diabetes in the morbidly obese by laparoscopic gastric banding. Obes Surg 2003;13:439 – 43. [35] Hell E, Miller KA, Moorehead MK, Samuels N. Evaluation of health status and quality of life after bariatric surgery: comparison of stan- dard Roux-en-Y gastric bypass, vertical banded gastroplasty and

E. Phillips et al. / Surgery for Obesity and Related Diseases 5 (2009) 588 –597 597 laparoscopic adjustable silicone gastric banding. Obes Surg 2000;10: [43] Bariatric Surgery Review Committee (BSRC). BSRC clarifies 214 –9. requirements (September 18, 2006). Available from: http://www.[36] Brancatisano A, Wahlroos S, Brancatisano R. Improvement in co- surgicalreview.org/news_page.aspx?itemϭBSRCClarifiesRequirements. morbid illness after placement of the Swedish Adjustable Gastric Band. Surg Obes Relat Dis 2008;4(Suppl):S39 – 46. Accessed October 27, 2006.[37] Martin LF, Smits GJ, Greenstein RJ. Treating morbid obesity with [44] U.S. Food and Drug Administration. New device approvals. Lap- laparoscopic adjustable gastric banding. Am J Surg 2007;194:333– 43.[38] Fried M, Miller K, Kormanova K. Literature review of comparative Band® adjustable gastric banding (LAGB®) system—P000008: sum- studies of complications with Swedish band and Lap-Band. Obes mary of safety and effectiveness data. June 5, 2001. Available from: Surg 2004;14:256 – 60.[39] Chapman AE, Kiroff G, Game P, et al. Laparoscopic adjustable www.fda.gov/cdrh/pdf/p000008.html. Accessed June 1, 2008. gastric banding in the treatment of obesity: a systematic literature [45] Allen JW, Coleman MG, Fielding GA. Lessons learned from review. Surgery 2004;135:326 –51.[40] O’Brien PE, Dixon JB. Lap-Band: outcomes and results. J Laparoen- laparoscopic gastric banding for morbid obesity. Am J Surg 2001; dosc Adv Surg Techniq 2003;13:265–70. 182:10 – 4.[41] Parikh MS, Laker S, Weiner M, Hajiseyedjavadi O, Ren CJ. Objec- [46] Dixon JB, O’Brien PE, Playfair J, et al. Adjustable gastric banding tive comparison of complications resulting from laparoscopic bariat- and conventional therapy for type 2 diabetes: a randomized controlled ric procedures. J Am Coll Surg 2006;202:252– 61. trial. JAMA 2008;299:316 –23.[42] Cunneen SA, Phillips E, Fielding G, et al. Studies of the Swedish [47] Fielding GA, Ren CJ. Laparoscopic adjustable gastric band. Surg Clin adjustable gastric band and Lap-Band: systematic review and meta- North Am 2005;85:129 – 40. analysis. Surg Obes Related Dis 2008;4:2:174 – 85. [48] Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a systematic review and meta-analysis. JAMA 2004;292:1724 –37. [49] Maggard M, Shugarman LR, Suttorp M, et al. Meta-analysis: surgical treatment of obesity. Ann Intern Med 2005;142:547–59.

3-Year Real-World Outcomes with theSwedish Adjustable Gastric Band™ inFranceG. Ribaric, J. N. Buchwald, G. d’Orsay &F. DaoudObesity SurgeryThe Journal of Metabolic Surgery andAllied CareISSN 0960-8923OBES SURGDOI 10.1007/s11695-012-0765-2 1 32

Your article is published under the Creative Commons Attribution license which allows users to read, copy, distribute and make derivative works, as long as the author of the original work is cited. You may self- archive this article on your own website, an institutional repository or funder’s repository and make it publicly available immediately.1 23

OBES SURGDOI 10.1007/s11695-012-0765-2 CLINICAL REPORT3-Year Real-World Outcomes with the Swedish AdjustableGastric Band™ in FranceG. Ribaric & J. N. Buchwald & G. d’Orsay & F. Daoud &for the French Health Technology Assessment Body(Haute Autorité de Santé [HAS]) Swedish Adjustable Gastric Band (SAGB™) Study Group# The Author(s) 2012. This article is published with open access at Springerlink.comAbstract The study objective was to ascertain outcomes social health insurance system. SAGB results at 3-yearwith the Swedish adjustable gastric band (SAGB) on an follow-up are reported. The noncomparative, observational,intention-to-treat basis in multiple centers across the French prospective, consecutive cohort study design sought a 500-The French Health Technology Assessment Body (Haute Autorité de Santé[HAS]) Swedish Adjustable Gastric Band (SAGB) Study Group: J-F. Ain1, L.Arnalsteen2, R. Arnoux3, E. Attal4, R. Barei5, P. Bergevin6, J. Cady7, P.Campan8, J-M. Catheline9, J-M. Chevallier10, J. Dargent11, B. Dehaye12, C.Deseguin13, J. Flaisler14, G. Fromont15, H. Johanet5, G. Juglard16, F. Labbé17, A.Legris18, P. Lointier19, F. Meaux20, E. Nini17, P. Noël21, F. Pattou2, G. Pugnet22,D. Quinaux23, J.L. Riqué24, S. Rossi25, M. Sodji26, J. Tussiot27, C. Vaudois281Clinique JB Denis, 71000 Mâcon, France 2CHRU Lille, Hopital C. F. DaoudHurriez, 59 000 Lille, France 3Clinique du Tondu, 33000 Bordeaux, France Data Management & Biometrics, Medextens SARL,4Polyclinique St Côme, 60200 Compiègne, France 5Clinique Sainte Marie, Paris, France95520 Osny, France 6Centre Hospitalier Privé du Montgardé, 78410Aubergenville, France 7Clinique Geoffroy St Hilaire, 75005 Paris, France8CHU Conception, 13005 Marseille, France 9CH Avicenne, 93009 Bobignycedex, France 10Hôpital Européen Georges Pompidou, 75015 Paris, France11Polyclinique de Rillieux, 69165 Rillieux-La-Pape, France 12CH de Meaux,77100 Meaux, France 13Clinique St Louis, 34190 Ganges, France 14Poly-clinique du Grand Sud, 30932 Nîmes Cedex 9, France 15Polyclinique deBois Bernard, 62320 Bois Bernard, France 16Clinique Lamartine, 74200Thonon-les-Bains, France 17CHG Antoine Gayraud de Carcassonne, 11890Carc. 9, France 18Clinique du Dr Priollet, 51000 Châlons en Champagne,France 19Clinique de la Plaine, 63100 Clermont Ferrand, France 20CliniqueAmbroise Paré, 62 660 Beuvry, France 21Clinique de la Casamance, 13400Aubagne, France 22Clinique Ambroise Paré, 31100 Toulouse, France23Clinique des Ursulines, 17 rue Raymond Poincaré, 10000 Troyes, France24Clinique Notre Dame de l’Espérance, 66100 Perpignan, France 25Cliniquede l’Europe, 76100 Rouen, France 26Clinique des Emailleurs, 87000 Li-moges, France 27Clinique Hoffmann, 93110 Rosny-sous-Bois, France28Polyclinique des 4 Pavillons, 33310 Lormont, FranceG. RibaricEthicon Endo-Surgery (Europe), European Surgical Institute,Hamburg, GermanyJ. N. Buchwald G. Ribaric (*)Division of Scientific Writing, Medwrite Medical Clinical and Medical Affairs, Ethicon Endo-Surgery (Europe)Communications, GmbH, MD&D EMEA (Europe, Middle East, Africa),Maiden Rock, WI, USA Johnson & Johnson, Hummelsbütteler Steindamm 71,G. d’Orsay 22851 Norderstedt, GermanyJohnson & Johnson Medical, Ethicon SAS, e-mail: [email protected], France

OBES SURGpatient minimum recruitment geographically representative the USA and Brazil [1]. As estimated in 2007, the majority ofof continental France. Safety (adverse events [AEs], device- bariatric procedures performed in France, 87.3 %, were AGBsrelated morbidity, and mortality) and effectiveness (change [2]. As part of its responsibility to the social insurance systemin body mass index [BMI, kilograms per square meter], per- [3, 4], the French government commissioned safety and effi-centage excess weight loss, comorbidities, quality of life cacy studies to obtain insurance data with which to evaluate[QoL]) were assessed. Adjustable gastric band survival was the service to patients provided by the proprietary AGBs usedcalculated. Thirty-one surgeons in 28 multidisciplinary teams/ in France (e.g., Swedish Adjustable Gastric Band™sites enrolled patients between September 2, 2007 and April [SAGB™]). The objective of the current study, requested by30, 2008. SAGB was successfully implanted in 517 patients: the Commission d’Evaluation des Produits et Prestations (a88.0 % female; mean age, 37.5 years; obesity duration, branch of the French Health Technology Assessment Body,15.3 years (baseline: mean BMI, 41.0; comorbidities, 773 in Haute Autorité de Santé [HAS]) [5, 6] and sponsored by the74.3 % of patients; Bariatric Analysis and Reporting Outcome manufacturer of the SAGB, Ethicon Endo-Surgery Europe,System (BAROS), 1.4; EuroQoL 5-Dimensions (EQ-5D), GmbH, was to prospectively ascertain outcomes typical of the0.61; EuroQoL–visual analog scale (EQ–VAS), 52.3). At French experience with the SAGB. The study protocol re-3 years: BMI, 32.2 (mean change, −9.0; p<0.0001); excess quired recruitment of at least 500 consecutive SAGB patientsweight loss, 47.4 %; comorbidities, 161 in 27.2 %; BAROS, (with <20 % attrition) by a representative sample of SAGB-3.6 (+2.2, p<0.0001); EQ-5D, 0.84 (+0.22, p<0.0001); EQ– implanting surgeons (i.e., both very experienced and lessVAS, 73.4 (+21.4, p<0.0001). SAGB-induced weight loss experienced) in urban and rural centers across regions ofwas associated with substantially improved QoL. One death continental France [7]. We present SAGB cohort outcomesoccurred and was unrelated to the treatment. No AE was collected in 30 surgical sites between September 2, 2007 andreported in 68.3 % of patients, and no confirmed device- November 20, 2011 analyzed on an intent-to-treat (ITT) basisrelated AE in 77.0 %. Overall AE rate was 0.19 per patient with 3-year follow-up.year. Device retention was 87.0 %. Analysis of patients lost tofollow-up showed a nonsignificant effect on overall study Methodsresults. In a prospective, consecutive cohort, “real-world”,nationwide study, the Swedish Adjustable Gastric Band was Study Protocol and Conductfound safe and effective at 3-year follow-up.Keywords Laparoscopic adjustable gastric banding . In 2007, at the request of the French HAS [5, 6], with theObesity . National outcomes . Swedish adjustable gastric objective of data acquisition to assess reimbursement of theband . SAGB . Quality of life . BAROS . EQ-5D . QALY SAGB product in France, Ethicon Endo-Surgery Europe, GmbH undertook sponsorship of a nationwide study ofIntroduction SAGB outcomes in participating centers (registered in the Clinical Trials Web database, #NCT01183975 [7]). TheLong-term effectiveness and safety of adjustable gastric band- sponsor developed the protocol and case report form toing (AGB) is of ongoing importance to patients, surgeons, and implement HAS requirements and good clinical practicesinsurers. The most recent global survey of bariatric surgery, (GCPs) defined in ISO EN 14155-1 and 2 [8, 9], withincorporating 36 respondent nations/national groupings, found consideration for real-life study constraints. The protocolthat AGB utilization is in its ascendancy in the USA and was approved by HAS, the Comité Consultatif sur le Traite-Canada, having increased in this grouping 944.2 % between ment de l’Information en matière de Recherche dans le2003 and 2008 (from 9,270 to 96,800 procedures annually). domaine de la Santé, and the Commission Nationale deEven in Europe, where AGB use dropped 20.5 % relative to l’Informatique et des Libertés (CNIL) to ensure patientother bariatric procedures, the number of AGB operations welfare in the study methodology and the ethical conductper annum increased 34.2 % from 21,496 to 28,843 in the of the study. The study was monitored by Contract Researchsame time frame [1]. Selection of AGB as a bariatric Organization, Medextens SARL, Paris, France.treatment remains high and with several hundred thousandpatients currently in long-term follow-up, intermediate and All patients provided written informed consent prior tolong-term national AGB outcome data are needed for in- surgery in accord with GCP guidelines and the Declarationsurance valuations and also as benchmarks for bariatric of Helsinki [10]. Treatment payments for patients weresurgery outcome improvement. covered by French National Health Insurance [6]. An inde- pendent study-monitoring committee consisting of a medi- Between 2003 and 2008, France performed the third great- cal nutritionist, a nonparticipating bariatric surgeon, and aest number of bariatric procedures annually (n013,722) after pharmacologist audited study conduct and the interim report.

OBES SURGDesign, Setting, Recruitment weight divided by excess weight, calculated by Miller’s formulas [13–17] for ascertaining ideal weight, which cor-The study was of a prospective, multicenter, noncompara- respond to the midpoint value of the medium-frame rangetive observational design, with consecutive recruitment of on the Metropolitan Life Insurance Height and WeightSAGB patients by participating surgeons. Study objectives Tables, multiplied by 100), health-related QoL, and propor-specified by HAS were the evaluation of (1) safety by tional changes in comorbidities.occurrence of adverse events, (2) clinical effectiveness byweight reduction, and (3) comorbidity and quality of life QoL Instruments(QoL) improvement. Beyond the protocol prescribing stan-dardized safety and weight data collection and assessment, The Bariatric Analysis and Reporting Outcome Systemno other standardization of physical examinations, laborato- (BAROS), introduced by Oria and Moorehead in 1998ry work, radiology studies, or follow-up procedure(s) (e.g., [18, 19], is a well-validated bariatric-specific QoL assess-band adjustments) was feasible or appropriate, given the ment that integrates weight loss data, comorbidity changes,independent, real-life assessment of participating centers. and subjective QoL, complications, and reoperations. A maximum of three points are recorded within domains of The study was designed to record safety and effective- weight loss, comorbidity improvement, and subjective QoL;ness in morbidly obese patients who underwent SAGB points are deducted for complications and reoperations. Theimplantation in French hospitals. Patients were eligible for final score classifies outcomes as failure (≤1), fair (>1 to 3),surgery when presenting with a body mass index (BMI, good (>3 to 5), very good (>5 to 7), or excellent (>7 to 9).kilograms per square meter) ≥40 or ≥35 to <40 with comor-bidities after failure of medical treatment in the absence of The EuroQoL 5-Dimensions (EQ-5D) was selected as thecontraindications in accord with published guidelines [11, generic complement to the BAROS due to its usefulness in12]. Patients in whom SAGB implementation was a bariatric calculating quality adjusted life years (QALYs). Specifical-re-intervention were also part of consecutive enrollment. A ly, the EQ-5D has proven effective in the estimation of theminimum recruitment goal of 500 patients with <20 % loss relative cost effectiveness associated with obesity interven-to follow-up at 3 years was targeted. tions, including AGB [20]. The EQ-5D is a generic health- related QoL assessment comprised of five items and a Euro-Inclusion Criteria QoL–visual analog scale (EQ–VAS) [21–24] yielding a patient health profile along five dimensions mobility, self-Surgeons were selected in public, academic, and private care, usual activities, pain/discomfort, and anxiety/depres-institutions geographically distributed across continental sion. Each dimension is represented by one item with threeFrance that had varying degrees of bariatric surgery volume. response options no problem, some problems, and severeThe goal was to include surgeons and patients who would problems. Responses to these five items can be normativelyprovide SAGB outcomes representative of the broad French weighted to derive an EQ-5D utility score with a rangereal-life practice. In compliance with GCP standards, all of −0.594 to 1 (10ultimate health). A difference of ≥0.07surgeons underwent training in the study protocol. in EQ-5D utility has been identified as clinically important [24]. The EQ–VAS component represents a single-item All patients consecutively undergoing SAGB implantation global QoL assessment in which patients are asked to rateduring the recruitment period by participating investigators their current health on a scale from 0 (worst imaginable) towere included in the cohort, providing that they resided in 100 (best imaginable) [25].continental France and had signed informed consent. Accord-ing to routine practice requirements, these patients were Technique, Band Adjustmentexpected to meet French guidelines for bariatric surgery [5](similar to National Institutes of Health [11] and the European Gastric banding was performed via standardized pars flac-Guidelines on Surgery of Severe Obesity [12]), and to be cida technique [26]. SAGB model options were 2100-Xeligible based on the HAS criteria [5, 6]. (with locking ring and injection port); 2200-X Quick- Close; and BD2XV Quick-Close with Velocity™ injectionVariables port. Band adjustments were performed according to the discretion of participating surgeons.Primary safety variables were frequency of adverse events,mortality, and medical device-related morbidity. The prima- Data Collectionry effectiveness variable was ITT BMI change over thecourse of 3 years. Secondary effectiveness variables were Baseline patient characteristics were collected (e.g., genderabsolute weight, excess weight, percentage excess weight and age); weight, obesity-related comorbid disease, andloss (%EWL; i.e., difference in preoperative and follow-up

OBES SURGQoL were recorded preoperatively, and assessed at 1, 3, 6, records in the database, 31 of which actively contributed to12, 18, 24, and 36 months postoperatively. Comorbidity the cohort (note: five surgeons reconsidered their ability todata were obtained via questionnaire with diagnoses estab- effectively participate in the study and voluntarily withdrewlished according to each investigator’s usual practice. during the patient recruitment phase; one surgeon was ex- cluded from the study during the data validation process due Remote entry of clinical data was performed by inves- to marked protocol noncompliance; the 63 patient recordstigators or administrative support personnel via password- created by the six surgeons were deemed invalid). These 31protected access to the Medextens-Medalliance eCRF surgeons were part of 28 multidisciplinary teams in fiveManager (v.1.3) web database (CNIL registered, www. university hospitals, three general hospitals, and 23 privatemedalliance.fr). Patients were informed of their rights of institutions. Patient screening and enrollment began Sep-access, correction, and refusal to participate. tember 2, 2007 and concluded April 30, 2008. The final 3- year follow-up visit was due on April 30, 2011 but wasStatistical Analysis extended to November 20, 2011 to accommodate patient availability.Stata/MP software (v.11.2, StataCorp LP, College Station,TX, USA) was used to perform all statistical analyses. The 31 surgeons who contributed to the cohort performedStatistical command syntax was programmed as a repeatable a total of 1,239 bariatric procedures during the study period,and auditable script. Statistical analysis was performed of which 560 cases were recorded in the online database asaccording to International Conference on Harmonization consecutive SAGB patient records. Among the 560 records,E9 guidance. No statistical justification for sample size is 43 were deemed invalid because they contained data entrypresented herein, as determination of the number of consec- errors, were duplicates, SAGB implantation was not docu-utive patients required was pre-established by HAS specifi- mented, or because patients withdrew. Overall, 517 validcations. Continuous demographic variables were reported as consecutive patients records were included in the databasemean, standard deviation (SD), and 80 % interpercentile analysis. Participating surgeons performed an average ofrange. Categorical demographic variables (including preop- 16.7 SAGB implants. The majority of cases (514/517,erative BMI subgroups) were reported as number and per- 99.4 %) were performed laparoscopically; conversions tocentage. Comorbidity data and adverse events were also laparotomy were reported in two cases (0.39 %); 99.6 %reported as number and percentage. Implant-survival rate (515/517) used pars flaccida technique; 76.4 % (395/517)and associated 95 % confidence interval (CI) was calculated employed band fixation; and 49.13 % (254/517) used portusing the Kaplan–Meier method. Continuous outcome var- fixation. Immediate band filling occurred in just 0.58 % (3/iables were generally reported as mean, SD, and 95 % CI. A 517) of cases. Overall, 66.54 % (344/517) of SAGBqualitative assessment of comorbidities before and after implants were performed using model 2200X; modelSAGB implantation was carried out. Fisher’s exact test BD2XV was used in 33.27 % (172/517) of cases.was used to investigate relationships between categoricalvariables. Between-group comparisons along continuous Safety and weight data met completeness and accuracymeasures were conducted using two-tailed, independent- criteria required by the protocol. Three-year clinical out-sample t tests. In addition, a series of two-tailed, single- come data were available in 85.9 % (444/517) of the cohort;sample t tests were performed to assess within-patient the objective that <20 % of patients lost to follow-up atchange from baseline, adopting a null hypothesis of no 3 years was reached. Due to the observational nature of thissignificant mean change in weight and QoL variables study with significant differences between surgeons in terms(H000, based on the nature of the disease of obesity without of their routine practice, comorbidity and QoL data didintervention). The Shapiro–Wilk test for normality was con- not meet protocol-targeted completeness; comorbidity da-ducted on each variable distribution prior to t test. Linear ta was available for 78.0 % of the SAGB cohort at 3-yearregression was applied in the analysis of relationships be- follow-up, while QoL data availability at 3 years rangedtween QoL measures, as well as in the relationship between from 39.0 % to 78.0 %, depending on the complexity ofBMI and QoL. Statistical significance was set at p≤0.05. the instrument. Baseline Patient CharacteristicsResults In the 517 consecutive patients with confirmed SAGB implants, gender and age were reported in 516 patients (99.8 %); diseaseInitially, 51 surgeons were identified as trained to implant duration in 504 patients (97.5 %); height, weight, excessthe SAGB based on previous year records of performing ≥1 weight, and BMI in 513 patients (99.2 %). The cohort consistedbariatric procedure weekly and, specifically, 35 AGB pro- of 88.0 % (n0455) female and 11.8 % (n061) male, with acedures annually. Forty surgeons enlisted, 37 created patient mean age of 37.5±10.9 years, obesity duration of 15.3±

OBES SURGTable 1 Baseline characteristics Characteristic Valuefor 517 total patients receivingSAGB implant Gender 61 (11.8) Male, N (%) 455 (88.01)SD standard deviation, IPR Female, N (%) 37.5±10.9 (23.4 to 53.7)interpercentile range (the 80th 15.3±8.4 (6.0 to 28.0)IPR indicates variable values Age, mean±SD, years (80 % IPR) 1.7±0.1 (1.6 to 1.8)ranging from 10th to 90th per- Duration of obesity, mean±SD, years (80 % IPR) 111.4±16.8 (93.0 to 133.0)centile), BMI body mass index, Height, mean±SD, m (80 % IPR) 52.1±14.1 (37.8 to 71.0)BAROS Bariatric Analysis and Absolute weight, mean±SD, kg (80 % IPR) 41.0±4.9 (36.1 to 47.3)Reporting Outcome System, Excess body weight, mean±SD, kg (80 % IPR) 291 (56.3)EQ-5D EuroQoL 5-Dimensions, BMI, mean±SD, kg/m2 (80 % IPR) 193 (37.3)EQ–VAS EuroQoL-Visual Ana-log Scale BMI ≥40, kg/m2, N (%) 29 (5.6)aN=302 for BAROS data BMI ≥35 and<40, kg/m2, N (%) 384 (74.3)recorded at 30 days postoperative BMI <35, kg/m2, N (%) 343 (66.3)bN=449 with EQ-5D data at At least 1 comorbidity, N (%) 475 (91.9)inclusion History of family obesity, N (%) 471 (91.1)cN=446 with EQ–VAS data at 1-year specialized medical care and multidisciplinary assessment, N (%) 503 (97.3)inclusion First bariatric intervention, N (%) 1.4±1.4 (−0.3 to 3.0) Preoperative psychological evaluation, N (%) 0.61±0.31 (0.09 to 0.88) BAROSa, mean±SD (80 % IPR) 52.3±18.4 (30.0 to 80.0) EQ-5Db, mean±SD (80 % IPR) EQ–VASc, mean±SD (80 % IPR)8.4 years, BMI of 41.0±4.9 kg/m2, absolute weight of 111.4± care prior to SAGB. Mean BAROS was 1.4±1.4, mean EQ-16.8 kg, and excess body weight of 52.1±14.1 (Table 1). 5D was 0.61±0.31, and mean VAS was 52.3±18.4. AGB was the first bariatric surgery in 471 patients Adverse Events(91.1 %) and a re-intervention in 43 patients (8.3 %). Therewere 291 patients (56.3 %) with a BMI of ≥40; 193 (37.3 %) The majority of SAGB patients (68.3 %, 353/517) experi-with a BMI of ≥35 to <40; 29 patients (5.6 %) presented with enced no adverse events (AEs) of any kind. Also, no con-a BMI of <35. Baseline BMI could not be calculated in four firmed device-related AE or serious adverse events (SAEs;patients (0.77 %) due to missing weight and/or height data. life-threatening or disabling event requiring intervention) was reported in 77.0 % (398/517). Confirmed AEs and/or SAEs Patients with ≥1 comorbidity numbered 384 (74.3 %), associated with the SAGB procedure totaled 290 in 164with a total of 781 recorded comorbidities. A history of patients (31.7 %) intraoperatively and up to 3-year follow-family obesity was reported in 343 patients (66.3 %) and up. There were 153 (29.6 %) AEs and 137 (26.5 %) SAEs475 patients (91.9 %) had ≥1 year of specialized medical constituting an overall rate of 0.19 adverse event per patient year. There was one death caused by an accident unrelated toFig. 1 Time distribution of adverse events or re-interventions the procedure. Figure 1 shows the distribution of the 290 confirmed events with a peak event frequency of 42 events during the first quarter following SAGB implant; thereafter, frequency was approximately 10 events per quarter. Among 290 confirmed events, there were 143 device- related events in 119 patients (23.0 %); thus, the device- related event rate was 0.09 event per patient year. The most frequent device-related AEs were 22 (4.3 %) band slippages, 14 (2.7 %) esophageal dilations, 13 (2.5 %) pouch dilations, 13 (2.5 %) port malpositions, 9 (1.7 %) port rotations, and 7 (1.4 %) port disconnections, 5 (1.0 %) food intolerances, 5 (1.0 %) dysphagia, 4 (0.8 %) band infections, 3 (0.6 %) port infections, 3 (0.6 %) band-related blockages, and 3 (0.6 %)

OBES SURG 48 months and the last recorded failure date was 50.23 months following implantation. At 3 years, SAGB implant survival rate was 87.0 % (95 % CI: 83.7, 89.6). Weight LossFig. 2 Kaplan–Meier band survival curve Three-year postoperative outcomes were available in 85.9 % (444/517) of patients. Intention-to-treat BMI change overband-related records of frequent vomiting. Device-related the course of 3 years was available in 423 (81.8 %) patients,SAEs included 67 (13.0 %) band removals, 23 (4.4 %) including 373 patients with SAGBs still in place and 50abdominoplasties, 22 (4.3 %) port re-interventions without patients whose bands were removed. BMI change could notport removal, 8 (1.6 %) port removals, and 3 (0.6 %) band be calculated in 94 patients, including 73 lost to follow-upre-interventions without removal. with unknown outcomes, 13 with band removals, 7 with incomplete data, and 1 death. Mean absolute weight wasImplant Survival Rate Function 87.1 ± 17.6 kg (85.4, 88.8) compared to 111.6± 16.9 kg (110.0, 113.2) at baseline (Table 2). This represented a meanSAGB removals occurred at different intervals post implanta- absolute weight reduction of 24.5±14.6 kg (23.1, 25.9;tion. The risk of failure (SAGB removal) versus the chances of t[422]034.4, p<0.0001). Mean excess body weight wassuccess (SAGB retention) over time was estimated with 28.0±15.8 kg (26.5, 29.5) compared to 52.5±14.3 kgKaplan–Meier cumulative survival function. All reported (51.1, 53.8) at baseline, a reduction in excess weight ofremovals were included in the analysis (Fig. 2). Patients with 24.5±14.7 kg (23.1, 25.9; t[421]034.3, p<0.0001)SAGB in place at the date of database freeze (November 20, corresponding to a %EWL of 47.4±32.1 (44.3, 51.0). Mean2011) were censored. The 517 valid implanted patients were BMI was 32.2±5.8 (31.6, 32.7), down approximatelyexposed to a time-at-risk of 23,386 months. Sixty-seven fail- 22.0 % from 41.2±5.0 (40.7, 41.7). This change in patientures were confirmed; band removal was typically associated BMI represented a mean reduction of 9.0±5.3 (8.5, 9.5;with the most severe adverse events. Median time at risk was t[422]035.0, p<0.0001). Figure 3 depicts the cohort’s evo- lution in mean BMI, beginning with a maximum adult BMI of 42.8±4.9 (42.3, 43.2). A significant difference in preoperative mean BMI was observed between patients receiving SAGB as a first inter- vention and those receiving SAGB as a re-intervention (41.4 ± 4.4 vs 37.1 ± 7.4, p < 0.001); however, by 3-year follow-up, mean BMI was not significantly different (32.1 ±5.8 vs 32.8±5.6; p00.51) between these two groups of patients. In order to assess the risk of bias in estimating theTable 2 Three-year weight outcomes and mean change assessmentWeight variable Value P valueb Mean±SD (95 % CI) N 0 423a Preoperative 3 Years Mean changeAbsolute weight, kg 111.6±16.9 (110.0, 113.2) 87.1±17.6 (85.4, 88.8) −24.5±14.6 (−25.9, −23.1) <0.0001Excess body weight, kg 52.5±14.3 (51.1, 53.8) 28.0±15.8 (26.5, 29.5) −24.5±14.7 (−25.9, −23.1) <0.0001BMI, kg/m2 41.2±5.0 (40.7, 41.7) 32.2±5.8 (31.6, 32.7) −9.0±5.3 (−9.5, −8.5) <0.0001% EWL – 47.4±32.1 (44.3, 51.0) – –BMI body mass index, EWL excess weight loss, SD standard deviation, CI confidence intervala N0423 for absolute weight and BMI within-patient change calculations; N0422 for excess weight within-patient change and %EWL calculationsb P values obtained from single-sample two-tailed t tests assessing mean weight change in patients with complete preoperative and 3-year follow-updata

OBES SURGFig. 3 SAGB cohort body mass index evolution Arthropathy (38.0 %), hypertension (22.0 %), dyslipide- mia (20.0 %), gastroesophageal reflux disease (20.0 %),primary endpoint on an ITT basis, patients with and without obstructive sleep apnea (12.0 %), and diabetes (10.0 %)3-year follow-up weight data were compared on several were the most frequently cited comorbidities at baseline,variables that could potentially influence BMI change. each experienced by at least 10 % of the cohort. QualitativeMean preoperative BMI of patients with complete 3-year analysis suggested a continued reduction in the overallweight data relative to those without were not significantly number of comorbidites over time and a gradual increasedifferent (41.2±5.0 vs 40.3±4.6; p00.10). Age at baseline, in number of patients with no reported comorbidities. Usinggender, duration of disease, maximum adult BMI, SAGB as recorded weight as an indicator of data integrity, 773 comor-first intervention or re-intervention, SAGB model type, and bidties were reported at baseline in 74.3 % (381/513) offrequency of band removal were also not significantly dif- patients with concomitant weight data, while no comorbid-ferent between groups. ity was reported in 25.7 % (132/513). At 3-year follow-up, 161 comorbidities were reported in 27.2 % (116/426) of In addition, the magnitude of the effect of including patients patients with concomitant weight data, while no comorbid-with band removal in the calculation of BMI change over ity was reported in 72.8 % (310/426). In addition, the3 years was also assessed. As expected, a significant difference number of patients with multiple comorbid conditions wasin mean BMI change was observed between patients with band reduced at 3 years and the mean number of comorbiditiesablation vs patients with intact SAGBs at 3 years (−6.2±5.5 per patient fell from 1.5 at baseline to 0.4 at 3-year follow-vs −9.4±5.2, p<0.001); however, the inclusion of patients with up. Analysis also suggested a slight increase in the averageband ablation (in conformance with ITT analysis) had a non- number of comorbidities per patient from year 2 (visit 7) tosignificant effect on the overall cohort mean change value year 3 (visit 8), paralleling the modest weight regain over(−9.0±5.3 [−9.7, −8.4] vs −9.4±5.2 [−10.1, −8.7]). Overall, the same time period.BMI reduction was observed across visits with most of thereduction occurring during the first postoperative year, con- Quality of Lifetinuing through the second year, followed by a negligible BMIincrease during the third year (Fig. 3). Although patient QoL assessment participation did not meet the protocol objective for data completeness, there wereComorbidities sufficient data to carry out quantitative analyses. [Note: Patient participation diminished with the complexity of theAdhering to the study’s observational design, the protocol did three QoL instruments.] All measures of QoL over timenot seek to modify current diagnostic or therapeutic practices: were significantly improved (Table 3). At 3 years, forno protocol-driven diagnostic tests were required. At each patients with complete data, mean BAROS was 3.6±2.2follow-up visit, comorbidities were generally reported as (3.2, 4.0) compared to 1.4±1.3 (1.2, 1.6) at pseudo-“present” or “absent.” Significant variation in diagnostic baseline (i.e., 30 days post SAGB). This positive changemethodology, terminology, and reporting regularity was not- in bariatric-specific QoL represented a mean increase of 2.2ed. Investigators, themselves, often did not assess comorbid- ±2.2 (1.8, 2.6; t[133]011.9; p<0.0001). Figure 4 depicts theities, but rather, reported assessment results provided to them cohort’s upward trend in mean QoL as assessed by BAROSby third parties. As a result, it was determined that an under- over the course of the study. Mean EQ-5D utility score wasreporting bias was a distinct possibility, thus limiting the 0.84±0.21 (0.82, 0.86) relative to a baseline value of 0.62±reliability of a quantitative estimate of change in prevalence. 0.31 (0.59, 0.65). This represented a mean within-patient QoL improvement of 0.22±0.32 (0.19, 0.25; t[346]0 12.6; p<0.0001), thus, a mean utility gain of 0.66 QALY over 3 years. Mean EQ–VAS score was 73.4±17.1 (71.6, 75.2) compared to 52.0±18.4 (50.1, 53.9), a QoL in- crease of 21.4±22.8 (19.0, 23.7; t[353]017.7; p<0.0001). Regression analysis indicated a significant association between weight loss and the EQ-5D (although weight per se is not a focus of the EQ-5D questionnaire). Using EQ-5D individual change scores as the response variable while controlling for baseline BMI, BMI reduction over the 3- year period following SAGB intervention was significantly related to an increase in global QoL (adjusted R200.03; F(2, 339)06.93; p<0.01; Fig. 5). In addition, individual change

OBES SURGTable 3 Three-year quality of life outcomes and mean change assessmentQoL variable Value P valueb Mean±SD (95 % CI) Baseline 3 Years Mean changeBAROS 1.4±1.3a (1.2, 1.6) 3.6±2.2 (3.2, 4.0) 2.2±2.2 (1.8, 2.6) <0.0001EQ-5D 0.62±0.31 (0.59, 0.65) 0.84±0.21 (0.82, 0.86) 0.22±0.32 (0.19, 0.25) <0.0001EQ–VAS 52.0±18.4 (50.1, 53.9) 73.4±17.1 (71.6, 75.2) 21.4±22.8 (19.0, 23.7) <0.0001BAROS Bariatric Analysis and Reporting Outcome System, EQ-5D EuroQoL 5-Dimensions, EQ–VAS EuroQoL–visual analog scale, SD standarddeviation, CI confidence intervala BAROS data at 30 days postoperative functions as pseudo-baselineb P values obtained from single-sample two-tailed t tests assessing mean QoL change in patients with complete preoperative and 3-year follow-updata (i.e., N0134 for BAROS, N0347 for EQ-5D, N0354 for EQ–VAS)scores on the BAROS and the EQ-5D were shown to cor- year mortality, 0.07 % [29]. Risk related to this treatmentrelate at r00.43, p<0.01. Overall QoL results suggest that consisted of adverse events or SAEs requiring interventionEQ-5D, a patient-reported QoL outcome measure with min- in 31.7 % of patients, including device-related events re-imal response burden, may be a sufficient measure of QoL quiring intervention in 23.0 %. Band removal occurred infollowing SAGB surgery. just 13.0 % of patients and was usually associated with the most severe adverse events or lack of procedure effective-Discussion ness. At 3-year follow-up, device retention was 87.0 %. Analysis suggested a reduction in overall comorbidites andThe current study was representative of the SAGB standard an increase in the number of patients with no reportedof care in France and met the objectives defined in the comorbidity.protocol in line with CEPP requirements. Results weresupportive of the clinical safety and effectiveness of the A significant and sustained BMI reduction of 9.0 wasSAGB, as concluded in prior studies. Clinical outcomes observed, corresponding to 47.4 % EWL. BMI reductionwere available for 86 % of the SAGB cohort; analysis was significantly associated with an increase in patient QoL,indicated that patient data lost to follow-up over 3 years as has been shown in other AGB studies [30]; the BAROShad no significant effect on treatment outcomes. scale showed a significant 2.2-point improvement from 30- day visit to 3-year follow-up. On average, SAGB patients No treatment-related mortality was reported. AGB oper- tended to move from a BAROS QoL rating of “fair” toative mortality is typically <0.1 % [27, 28] and 30-day to 2- “good.” Also, SAGB patient EQ-5D Qol-related utility score significantly improved by 0.22 points (three times the reported minimally important difference) representing a utility gain of 0.66 QALY. In addition, the SAGB cohort’sFig. 4 Evolution of SAGB cohort quality of life scores as measured by Fig. 5 Relationship between body mass index reduction and quality ofthe Bariatric Analysis and Reporting Outcome System (BAROS®) life changes as measured by EuroQoL 5-dimensions (EQ-5D). v10 baseline; v803-year follow-up

OBES SURGEQ-5D mean utility score of 0.84±0.21 at 3 years was (GILB) [58]. At baseline, patients had a mean BMI of 43.7slightly higher than that recorded for a UK general practice and 34.1 at 3 years post-AGB treatment, a reduction of 9.6reference group of normal BMI patients (n0782), EQ-5D0 with >70 % follow-up (%EWL was not reported). These0.80±0.22 [31]. Overall, patient QoL change scores on results are similar to those of the current French SAGBBAROS and EQ-5D were shown to be significantly corre- cohort where baseline BMI was 41.0, and 32.2 at 3 years,lated (r00.43, p<0.01) and both instruments were sensitive a reduction of 9.0 (p<0.0001) with 81.8 % follow-up. Into BMI reduction over time. Results suggested that the EQ- 2012, a subsequent nationwide AGB study of the same5D, a self-report generic QoL measure with minimal re- GILB database, a comparison of the perigastric (PG) andsponse burden, may be a sufficient measure of QoL change pars flaccida (PF) techniques was undertaken by Di Lorenzofollowing SAGB surgery. Finally, at 3-year follow-up, trend et al. [26] in 2,549 Lap-Band patients (baseline BMI, 46.4):analysis indicated a modest and nonsignificant weight in- 1,206 (47.3 %) were operated via PG, 1,343 (52.7 %) viacrease (p00.14) beginning after year 2. Also, a corresponding PF, approach. In patients eligible for minimum 3-yeardecrease in QoL and a related increase in the average number follow-up, respective mean BMI was 33.8 (a reduction ofof comorbidities suggested that SAGB patients may require 12.6) and 32.4 (−14.0), and mean EWL was 47.2 % andmore support after year 2. 48.9 %. These results, as in the earlier GILB cohort [58] are essentially equivalent to those of the current study. As of April 25, 2012, more than 2,300 peer-reviewedarticles addressed the topic of AGB treatment [32]. The The third nationwide AGB report, by Phillips et al., apreponderance of these studies are of an observational de- prospective US Food and Drug Administration trial of thesign, in addition to several dozen randomized controlled SAGB (the “Realize® Band” in the USA) was conducted intrials (RCTs), most conducted in bariatric surgery “centers 12 academic and private centers with surgeons versed inof excellence” (COEs)—experienced, high-volume medical laparoscopic surgery but with varied AGB experience [75].centers that incorporate a multidisciplinary team approach Of 405 patients screened, 276 qualified for the study and[33] and compulsory pre- and post-operative protocols con- underwent band surgery. Mean baseline BMI was 44.5 withforming to professional [34–36] and national [5, 11, 12] a reduction of 8.2 at 3 years and 83.0 % follow-up; EWL inguidelines. Outcomes achieved in COE-standard sites may the REALIZE trial was 41.1 % compared to 47.4 % in thenot be comparable to those achieved in less-experienced, current French SAGB cohort.lower-volume sites. Yet, non-COE sites represent a sizeablenumber of centers performing bariatric surgery, a fact im- Two nationwide AGB-related government-commissionedportant to insurers assessing the value of AGB treatment on surveys have been conducted in France. Chevallier et al. [71]a national scale. Thus, while seven high-volume single- published a prospective, consecutive series, statistical analysiscenter AGB studies over the last 10 years reported of factors predictive of AGB outcomes from a national public≥55.0 % EWL at ≥3 years [37–43], integrated COE- and insurance perspective. Although this survey lacked 3-yearnon-COE nationwide %EWL results are somewhat lower data, it found that EWL was <50.0 % at 1–2 years in the[44, 45]. majority of 1,079 morbidly obese adults who had undergone the AGB procedure [71]. Basdevant et al. [2] studied current The current study’s findings should be discussed in the bariatric surgery practices across procedures in the samecontext of evidence from other national AGB studies with French National Medical Insurance Service registry as thatintermediate-term follow-up, yet such reports are few. In the studied by Chevallier et al. They showed a BMI reduction oflast decade, fewer than three dozen national bariatric surgery 9.9 and EWL of 46.0 % at 2 years, roughly equivalent to thestudies (16 large sample, retrospective, database analyses French SAGB cohort 3-year outcomes. The single prior[46–61], 14 prospective surveys [1, 2, 62–73], and one French clinical trial of an AGB with 3-year follow-up, a studyongoing [multi-article], nonrandomized, prospective con- of the MIDBand (MID, Dardilly, France), was conducted introlled study [74]) have been published in the peer- 13 centers (two academic, one public general hospital, 10reviewed literature, 14 of these featuring European private clinics) in 262 morbidly obese adults, 193 of whomcountries. In the seven nationwide studies with ≥2-year were included in the weight-loss analysis (26 % lost to follow-follow-up data in morbidly obese patients (five observation- up). Mean baseline BMI was 41.8 with a reduction of 8.2–al AGB or AGB-inclusive studies and two French AGB 30.7 at 3 years; median EWL was 61.0 % [76]. Percentagesurveys [2, 26, 58, 71, 74–76]), weight loss was less than EWL ranges reported at 2–3 years in the three prior Frenchor comparable to that of the current SAGB cohort. AGB studies, taken together with the French SAGB cohort study, are likely representative of current AGB weight loss In 2003, Angrisani et al. published the Italian Lap-Band outcomes in France.experience with 1,863 patients in 27 centers, one ofthe earliest and largest nationwide AGB studies with Chevallier et al. also found that, independent of patientintermediate-term follow-up based on a retrospectively an- characteristics, the key predictor of AGB success was surgicalalyzed national database, the Italian Group for Lap-Band volume (i.e., multidisciplinary team typically performing >2

OBES SURGbariatric procedures per week) [71]. Interestingly, although the comparing AGB and gastric bypass found 55.2 % EWL incurrent French SAGB study incorporated centers that per- the AGB group (n07,383, 18 studies) at 3 years [45]. Whileformed <2 bariatric procedures per week, its mean %EWL 3-year mean %EWL in the recent three reviews was greaterexceeded or equaled that of five of the six other national AGB than that of the current nationwide study, as would bestudies summarized. expected with a greater proportion of higher-volume sites and inclusive of RCTs, the mean %EWL 95 % confidence A final prominent, long-running, nationwide clinical tri- interval (95 % CI: 44.0, 51.0) of the current study was foundal, the Swedish Obese Subjects (SOS) study, also reported to overlap with 95 % CIs reported in the two most recentBMI reduction less than that of the current 3-year French and largest meta-analyses, suggestive of relatively compa-study, although it is typical for weight loss to diminish rable outcomes.somewhat after reaching its peak between 1 and 3 years.The SOS study is an ongoing, nonrandomized, prospective, Most intermediate-term AGB studies demonstrate highcontrolled study that has followed 480 primary healthcare rates of patient attrition, up to 86.2 % at >3-year follow-upsites and 25 public surgical departments in Sweden since the [45], and incomplete data collection. Although the currentrecruitment phase (September 1, 1987 to January 31, 2001). study satisfied the <20 % loss to follow-up with respect toBariatric surgery (gastric bypass, vertical banded gastro- weight and safety data, it was somewhat limited by issuesplasty, or AGB [n0156 at baseline]) was performed on related to incomplete and missing data, particularly in the2,010 morbidly obese patients and 2,037 contemporaneous- areas of comorbidity and QoL data collection. Indeed, duely matched morbidly obese controls received usual care in to a likely underreporting bias, a reliable quantitative esti-the Swedish primary care system. In 2004 (10-year follow- mate of change in prevalence of any specific comorbidityup), the reported mean BMI in the AGB subgroup (n0156) could not be calculated. Missing QoL data was evident evenhad decreased by 5.4 (12.8 %, p<0.05) to 36.5 from 41.9± at baseline, ranging from 13.0 to 42.0 %, again, depending4.2 (total surgical group—AGB alone, not reported) at base- on the complexity of the instrument. In addition, althoughline [74]. 91.1 % of patients in this cohort satisfied eligibility criteria, in some cases, eligibility requirements were varied accord- In addition to the small number of nationwide AGB ing to individual investigator discretion: Under the protocolstudies with ≥3 years of follow-up, the current outcomes requirement of consecutive recruitment, 29 patients (5.6 %)should be compared with those of systematic reviews or presented with a BMI <35 who either had a severe concom-meta-analyses, which typically include a wide sample of itant comorbidity or maximum adult BMI of ≥45 and werestudies, possibly limiting bias toward COE “center effects.” implanted with a SAGB as a re-intervention for a failed orIn considering AGB studies prior to 2003, a systematic previously complicated bariatric procedure. In the subsetreview by Chapman et al. concluded that the quality of analysis of re-intervention patients, it was shown that theAGB studies with intermediate follow-up (defined as 2– inclusion of these patients in the overall cohort calculations4 years) was only moderate, and long-term effectiveness had a nonsignificant effect on mean BMI change; indeed, no(>4 years), unproven [77]. Five years later, a systematic significant difference in mean BMI was observed at 3-yearreview by Tice et al. (sponsored by US insurer Blue Shield follow-up between re-intervention patients and SAGB first-of California comparing 1966–2007 AGB and Roux-en-Y intervention patients (32.81±5.6 vs 32.13±5.8, p00.51).gastric bypass outcomes) found similarly that, despite beinga popular procedure, AGB results beyond a single postopera- As longer-term AGB data accumulate and bariatric sur-tive year were seldom reported [78]. Four other systematic geons and insurers seek to analyze the “big picture” ofreviews and/or meta-analyses published in the last decade bariatric surgery, specifically, the clinical effectiveness ofreviewed %EWL and/or BMI change at ≥1 year, though these the AGB procedure, the opportunity to review publishedecho conclusions by Chapman et al. and Tice et al. that ≥3- findings from the full range of sites performing bariatricyear AGB outcomes were few, and limited in most studies, by surgery will be important. This report presents the firstmarked patient loss to follow-up [44, 45, 79, 80]. government-commissioned clinical trial results for the Swedish Adjustable Gastric Band in routine use in multiple In a 2004 meta-analysis by Buchwald et al. that included centers across France.1,848 patients with AGB %EWL data (1995–2003), EWLof 47.5 % at >2 years was nearly identical to that of the SAGB treatment was safe and effective. At 3-yearcurrent French SAGB study EWL, 47.4 % at 3 years [44]. A follow-up, 87.0 % of SAGB implants were fully functional2006 systematic review by O’Brien found 54.8 % weighted and a significant BMI reduction of 9.0 kg/m2 with amean EWL at 3-year follow-up (AGB n03,104, 12 studies) corresponding 47.4 % EWL was observed. The cohort’s[79]; a 2008 meta-analysis by Cunneen et al. comparing EQ-5D QoL-related utility score improved significantlySAGB (n04,274, 33 studies) and Lap-Band (n024,707, and was found to be essentially equivalent to that of patients104 studies) found respective 3-year EWL of 56.4 % and with normal BMI. These findings corroborate prior French50.2 % [80]; and a 2009 meta-analysis by Garb et al. countrywide AGB insurance surveys and compare favorably

OBES SURGwith other national AGB outcome studies, systematic reviews, clinicaltrials.gov: http://clinicaltrials.gov/ct2/show/NCT01183975?and meta-analyses. More long-term nationwide studies of term0ethicon+endo-surgery&rank015AGB outcomes are needed to further clarify reimbursement 8. Clinical investigation of medical devices for human subjects—partcriteria and refine treatment standards. 1: general requirements. International Standard Ref. ISO 14155– 1:2003(E).Acknowledgments We thank N Moore, head of the Department of 9. Clinical investigation of medical devices for human subjects—partClinical Pharmacology, Hôpital Pellegrin, Université Bordeaux, 1: clinical investigation plans. International Standard Ref. ISOFrance; J Mouiel, Bariatric Department, Hôpital St Roch, France; and 14155–2:2003(E). First edition May 15, 2003.O Ziegler, head of the Department of Nutrition, Hôpital Jeanne d'Arc, 10. World Medical Association. 7 Declaration of Helsinki: ethicalFrance, for their valued assistance as independent expert consultants to principles for medical research involving human subjects, 22 Oc-the trial. We are grateful for expert surgical consultation from M Fried, tober 2008. www.wma.net/en/30publications/10policies/b3/First Faculty of Medicine, Charles University, Prague, Czech Republic; index.html.study advice from K Haas, Medical Education, Ethicon Endo-Surgery, 11. NIH. Gastrointestinal surgery for severe obesity. NIH ConsensusCincinnati, OH, USA; and statistical consultation from J Jokinen, Statement 1991; Mar 25–27;9(1):1–20.Johnson & Johnson Consumer Products, Fort Washington, Pennsylva- 12. Fried M, Hainer V, Basdevant A, et al. Inter-disciplinary Europeannia, USA; and TW McGlennon, Director, Statistical Analysis and guidelines on surgery of severe obesity. Int J Obes (Lond). 2007;31Quality of Life Assessment, McGlennon MotiMetrics (M3), WI, USA. (4):569–77. 13. WHO: Physical status: the use and interpretation of anthropometry.Disclosure G Ribaric and G d’Orsay are employees of Johnson & Report of a WHO Expert Committee. WHO Technical ReportJohnson (J&J). JN Buchwald and F Daoud are Contract Research Series 854. Geneva: World Health Organization, 1995.Organization employees under contract with J&J. The study was 14. Lorentz FH. Ein neuer Konstitionsinde. Klin Wochenschr.financially supported by Ethicon Endo-Surgery Europe, GmBH, a 1929;8:348–51.subsidiary of J&J, according to a request by the HAS France for 15. Lorentz FH. Ein neuer Konstitionsinde der Frau. Klin Wochenschr.countrywide SAGB evidence generation. 1929;16:734–6. 16. Miller MA. A calculated method for the determination of idealOpen Access This article is distributed under the terms of the Crea- body weight. Nutr Support Serv. 1985;5(3):31–3.tive Commons Attribution License which permits any use, distribution, 17. Deitel M, Greenstein RJ. Recommendations for reporting weightand reproduction in any medium, provided the original author(s) and loss. Obes Surg. 2003;13(2):159–60.the source are credited. 18. Oria HE, Moorehead MK. Bariatric analysis and reporting out- come system (BAROS). Obes Surg. 1998;9:947–54.References 19. Nini E, Slim K, Scesa JL, et al. Evaluation de la chirurgie coelio- scopique de l’obésité par le score BAROS. Ann Chir. 2002;127 1. Buchwald H, Oien DM. Metabolic/bariatric surgery worldwide (2):107–14. 2008. Obes Surg. 2009;19:1605–11. 20. Ackroyd R, Mouiel J, Chevallier JM, et al. Cost-effectiveness and budget impact of obesity surgery in patients with type 2 diabetes in 2. Basdevant A, Paita M, Rodde-Dunet M-H, et al. A nationwide 3 European countries. Obes Surg. 2006;16:1488–503. survey on bariatric surgery: two years prospective follow-up. Obes 21. EuroQoL Group. EuroQoL: a new facility for the measurement Surg. 2007;17:39–44. of health-related quality of life. Health Policy. 1990;16: 199–208. 3. Avis de la CEPP du 1er septembre 2004 modifié le 20 octobre 22. Dolan P, Gudex C, Kind P, et al. A social tariff for EuroQoL: 2004 SAGB (anneau gastrique ajustable suédois), implant annu- results from a UK general population survey, in York Centre for laire ajustable pour gastroplastie. Modèles 2100-X et 2200-X. Health Economics Discussion Paper. University of York: York, http://www.santedev.org/biblio/pdf2/has/cepp/2004/10/20/ UK; 1990. pp020367.pdf. Accessed: March 12, 2012 23. Kind P, Dolan P, Gudex C, et al. Variations in population health status: results from a United Kingdom national questionnaire sur- 4. Avis de la CEPP 8 mars 2006 SAGB (anneau gastrique ajustable vey. BMJ. 1998;316:736–41. suédois) Quick-Close fourni avec le site VELOCITY et son appli- 24. Dixon S, Farina C, McEwan P, et al. Evaluation of the association cateur, implant annulaire ajustable pour gastroplastie. Modèle between health-related utility and obesity in hospital treated sub- BD2XV. jects. Value Health. 2004;7:331. 25. Brazier J, Roberts J, Tsuchiya A, et al. A comparison of the EQ-5D 5. Association Française d’Etudes et de Recherches sur l’Obésité, and SF-6D across seven patient groups. Health Econ. 2004;13: Association de Langue Française pour l’Etude du Diabète et des 873–4. Maladies Métaboliques, Société de Nutrition et de Diététique de 26. Di Lorenzo N, Furbetta F, Favretti F, et al. Laparoscopic gastric Langue Française. Recommandations pour le diagnostic, la pré- banding via pars flaccida versus perigastric positioning: technique, vention et le traitement des obésités en France. Cah Nutr Diet. complications, and results in 2,549 patients. Surg Endosc. 1998;33:10–42. 2010;24:1519–23. 27. Nguyen NT, Slone JA, Nguyen XM, et al. A prospective randomized 6. Haute Autorité de Santé: Obésité: prise en charge chirurgicale chez trial of laparoscopic gastric bypass versus laparoscopic adjustable l’adulte. Recommandations de bonnes pratiques professionnelles. gastric banding for the treatment of morbid obesity: outcomes, qual- January 2009. http://www.has-sante.fr/portail/jcms/c_765529/ ity of life, and costs. Ann Surg. 2009;250(4):631–41. obesite-prise-en-charge-chirurgicale-chez-l-adulte. Accessed: March 28. Himpens J, Cadière GB, Bazi M, et al. Long-term outcomes of 12, 2012 laparoscopic adjustable gastric banding. Arch Surg. 2011;146 (7):802–7. 7. Prospective National Cohort Study on Swedish Adjustable Gastric 29. Buchwald H, Estok R, Fahrbach K, et al. Trends in mortality in Band (SAGB) for Gastroplasty (Étude de Cohorte Nationale Pro- bariatric surgery: a systematic review and meta-analysis. Surgery. spective de l’Implant Annulaire Ajustable Pour Gastroplastie 2007;142(4):621–32. discussion 632–5. SAGB). PROTOCOL N°: 05-FR-004 1.0. Protocol available at:

OBES SURG30. Champault A, Duwat O, Polliand C, et al. Quality of life after 50. Morino M, Toppino M, Forestieri P, et al. Mortality after bariatric laparoscopic gastric banding: prospective study (152 cases) with a surgery: analysis of 13,871 morbidly obese patients from a nation- follow-up of 2 years. Surg Laparosc Endosc Percutan Tech. al registry. Ann Surg. 2007;246(6):1002–7. discussion 1007–9. 2006;16(3):131–6. 51. Ricciardi R, Town RJ, Kellogg TA, et al. Outcomes after open31. Sach TH. The relationship between BMI and HR-QoL: comparing the versus laparoscopic gastric bypass. Surg Laparosc Endosc Percu- EQ-5D, EuroQol VAS and SF-6D. Int J Obes. 2007;31:189–96. tan Tech. 2006;16(5):317–20.32. National Library of Medicine. Electronic PubMed database search 52. Davis MM, Slish K, Chao C, et al. National trends in bariatric for adjustable gastric banding. http://www.ncbi.nlm.nih.gov/ surgery, 1996–2002. Arch Surg. 2006;141(1):71–4. discussion 75. pubmed?term0adjustable%20gastric%20banding. Accessed: March 20, 2012. 53. Carbonelli AM, Lincourt AE, Matthews BD, et al. National study of the effect of patient and hospital characteristics on bariatric33. Apovian CM, Cummings S, Anderson W, et al. Best practice surgery outcomes. Am Surg. 2005;71(4):308–14. updates for multidisciplinary care in weight loss surgery. Obesity (Silver Spring). 2009;17(5):871–9. Epub 2009 Feb 19. 54. Nguyen NT, Root J, Zainabadi K, et al. Accelerated growth of bariatric surgery with the introduction of minimally invasive sur-34. American Society for Metabolic and Bariatric Surgery (ASMBS). gery. Arch Surg. 2005;140(12):1198–202. discussion 1203. http://asmbs.org/2012/03/unified-national-accreditation-program- for-bariatric-surgery-centers-announced-by-american-college-of- 55. Santry HP, Gillen DL, Lauderdale DS. Trends in bariatric surgical surgeons-and-american-society-for-metabolic-and-bariatric-surgery/. procedures. JAMA. 2005;294(15):1909–17. Accessed: March 20, 2012 56. Trus TL, Pope GD, Finlayson SR. National trends in utilization35. American College of Surgeons National Surgical Quality Improve- and outcomes of bariatric surgery. Surg Endosc. 2005;19(5):616– ment Program (NSQIP). http://site.acsnsqip.org/about/business-case/ 20. Epub 2005 Mar 11. . Accessed: March 15, 2012 57. Poulose BK, Holzman MD, Zhu Y, et al. National variations in morbid36. Surgical Review Corporation (SRC). http://www.surgicalreview.org/ obesity and bariatric surgery use. J Am Coll Surg. 2005;201(1):77–84. coe-programs/bariatric/. Accessed: March 15, 2012 58. Angrisani L, Di Lorenzo N, Favretti F, et al. The Italian Group for37. Rubin M, Spivak H. Prospective study of 250 patients undergoing LAP-BAND: predictive value of initial body mass index for laparoscopic gastric banding using the two-step technique: a tech- weight loss after 5 years of follow-up. Surg Endosc. 2004;18 nique to prevent postoperative slippage. Surg Endosc. 2003;17 (10):1524–7. Epub 2004 Aug 26. (6):857–60. 59. Belle SH, Berk PD, Courcoulas AP, et al. Longitudinal Assessment38. Naef M, Naef U, Mouton WG, et al. Outcome and complications of Bariatric Surgery Consortium Writing Group. Safety and effi- after laparoscopic Swedish adjustable gastric banding: 5-year results cacy of bariatric surgery: longitudinal assessment of bariatric sur- of a prospective clinical trial. Obes Surg. 2007;17(2):195–201. gery. Surg Obes Relat Dis. 2007;3(2):116–26.39. Steffen R, Biertho L, Ricklin T, et al. Laparoscopic Swedish 60. Kakarla VR, Nandipati K, Lalla M, et al. Are laparoscopic bariatric adjustable gastric banding: a five-year prospective study. Obes procedures safe in superobese (BMI ≥50 kg/m2) patients? A NSQIP Surg. 2003;13(3):404–11. data analysis. Surg Obes Relat Dis. 2011;7(4):452–8.40. Fielding GA. Laparoscopic adjustable gastric banding for massive 61. Demaria EJ, Winegar DA, Pate VW, et al. Early postoperative superobesity (>60 body mass index kg/m2). Surg Endosc. 2003;17 outcomes of metabolic surgery to treat diabetes from sites partic- (10):1541–5. ipating in the ASMBS bariatric surgery center of excellence pro- gram as reported in the Bariatric Outcomes Longitudinal Database.41. Fox SR, Fox KM, Srikanth MS, et al. The Lap-Band system in a Ann Surg. 2010;252(3):559–66. discussion 566–7. North American population. Obes Surg. 2003;13(2):275–80. 62. World Health Organization: WHO Global Infobase: country col-42. Ponce J, Paynter S, Fromm R. Laparoscopic adjustable gastric lected data, 2010. Accessible at: https://apps.who.int/infobase/ banding: 1,014 consecutive cases. J Am Coll Surg. 2005;201 comparestart.aspx. (4):529–35. 63. Burns EM, Naseem H, Bottle A, et al. Introduction of laparoscopic43. Jan JC, Hong D, Pereira N, et al. Laparoscopic adjustable gastric bariatric surgery in England: observational population cohort banding versus laparoscopic gastric bypass for morbid obesity: a study. BMJ. 2010;341:c4296. single-institution comparison study of early results. J Gastrointest Surg. 2005;9(1):30–41. 64. Harbottle L. Audit of nutritional and dietary outcomes of bariatric surgery patients. Obes Rev. 2011;12(3):198–204.44. Buchwald H, Estok R, Fahrbach K, et al. Trends in mortality in bariatric surgery: a systematic review and meta-analysis. Surgery. 65. Schauer DP, Arterburn DE, Livingston EH, et al. Decision modeling 2007;142(4):621–32. discussion 632–635. to estimate the impact of gastric bypass surgery on life expectancy for the treatment of morbid obesity. Arch Surg. 2010;145(1):57–62.45. Garb J, Welch G, Zagarins S, et al. Bariatric surgery for the treatment of morbid obesity: a meta-analysis of weight loss out- 66. Martin M, Beekley A, Kjorstad R, et al. Socioeconomic disparities in comes for laparoscopic adjustable gastric banding and laparoscop- eligibility and access to bariatric surgery: a national population-based ic gastric bypass. Obes Surg. 2009;19(10):1447–55. analysis. Surg Obes Relat Dis. 2010;6(1):8–15. Epub 2009 Jul 17.46. Nguyen NT, Hohmann S, Slone J, et al. Improved bariatric surgery 67. Stroh C, Birk D, Flade-Kuthe R, et al. A nationwide survey on outcomes for Medicare beneficiaries after implementation of the medi- bariatric surgery in Germany—results 2005–2007. Obes Surg. care national coverage determination. Arch Surg. 2010;145(1):72–8. 2009;19(1):105–12. Epub 2008 Oct 22.47. Sánchez-Santos R, Masdevall C, Baltasar A, et al. Short- and mid- 68. Santos LM, de Oliveira IV, Peters LR, et al. Trends in morbid term outcomes of sleeve gastrectomy for morbid obesity: the obesity and in bariatric surgeries covered by the Brazilian Public experience of the Spanish National Registry. Obes Surg. 2009;19 Health System. Obes Surg. 2010;20(7):943–8. Epub 2008 Jun 7. (9):1203–10. Epub 2009 Jul 2. 69. Christou NV, Efthimiou E. Bariatric surgery waiting times in48. Sundbom M, Karlson BM. Low mortality in bariatric surgery 1995 Canada. Can J Surg. 2009;52(3):229–34. through 2005 in Sweden, in spite of a shift to more complex procedures. Obes Surg. 2009;19(12):1697–701. 70. Dallal RM, Datta T, Braitman LE. Medicare and Medicaid status predicts prolonged length of stay after bariatric surgery. Surg Obes49. Schilling PL, Davis MM, Albanese CT, et al. National trends in Relat Dis. 2007;3(6):592–6. Epub 2007 Oct 23. adolescent bariatric surgical procedures and implications for sur- gical centers of excellence. J Am Coll Surg. 2008;206(1):1–12. 71. Chevallier J-M, Paita M, Rodde-Dunet M-H. Predictive factors of Epub 2007 Oct 18. outcome after gastric banding: a nationwide survey on the role of center activity and patients’ behavior. Ann Surg. 2007;246 (6):1034–9.

OBES SURG72. Tsai WS, Inge TH, Burd RS. Bariatric surgery in adolescents: banding with the MIDBAND™. Obes Surg. 2012;22(4): recent national trends in use and in-hospital outcome. Arch Pediatr 572–81. Adolesc Med. 2007;161(3):217–21. 77. Chapman AE, Kiroff G, Game P, et al. Laparoscopic adjustable gastric banding in the treatment of obesity: a systematic literature73. Santry HP, Chin MH, Cagney KA, et al. The use of multi- review. Surgery. 2004;135:326–51. disciplinary teams to evaluate bariatric surgery patients: results 78. Tice JA, Karliner L, Walsh J, et al. Gastric banding or bypass? A from a national survey in the USA. Obes Surg. 2006;16 systematic review comparing the two most popular bariatric pro- (1):59–66. cedures. Am J Med. 2008;121:885–93. 79. O’Brien PE, McPhail T, Chaston TB, et al. Systematic review of74. Sjostrom L, Lindroos A-K, Peltonen M, et al. Lifestyle, diabetes, medium-term weight loss after bariatric operations. Obes Surg. and cardiovascular risk factors 10 years after bariatric surgery. 2006;16:1032–40. NEJM. 2004;351(26):2683–93. 80. Cunneen SA, Phillips E, Fielding G, et al. Studies of Swedish adjustable gastric band and Lap-Band: systematic review and75. Phillips E, Ponce J, Cunneen SA. Safety and effectiveness of meta-analysis. Surg Obes Relat Dis. 2008;4(2):174–85. Epub REALIZE adjustable gastric band: 3-year prospective study in 2008 Feb 1. Review. the United States. Surg Obes Rel Dis. 2009;5:588–97.76. Gouillat C, Denis A, Badol-Van Straaten P, et al. Prospec- tive, multicenter, 3-year trial of laparoscopic adjustable gastric

Surgery for Obesity and Related Diseases 4 (2008) S56 –S62 Laparoscopic adjustable gastric banding in an ambulatory surgery center Brad M. Watkins, M.D., F.A.C.S.*, Jessie H. Ahroni, Ph.D., A.R.N.P., Robert Michaelson, M.D., F.A.C.S., Kevin F. Montgomery, M.D., F.A.C.S., Ron E. Abrams, M.D., D.A.B.A., Marc D. Erlitz, M.D., D.A.B.A., Jim E. Scurlock, M.D., D.A.B.A. Synchrony Health, Northwest Weight Loss Surgery, Everett, WashingtonAbstract Background: In the management of morbid obesity by laparoscopic adjustable gastric bandingKeywords: (LAGB), careful patient preparation and attentive follow-up have been shown to produce the best long-term results. Methods: Between November 2002 and August 2007, prospective data were collected on 2,411 consecutive patients, 84% of whom underwent LAGB at our freestanding outpatient surgery center, staffed by our dedicated multidisciplinary bariatric team. Outcomes reported include changes in mean body mass index (BMI), percentage excess body weight loss (%EBWL), and incidence of complications at 1 year, as well as the slippage rate up to 3 years. Results: A total of 2,003 (83%) female and 409 (17%) male patients with a mean age of 44.1 years (range 15–76 yrs) and a BMI (mean Ϯ SD) of 45.7 Ϯ 7.9 kg/m2 (range 29.1– 83.1 kg/m2) underwent LAGB. In 2,027 patients (84%), LAGB was performed as an outpatient procedure, with 1 (0.04%), conversion from a laparoscopic to an open procedure. The majority of operations were completed in less than 1 hour, using the pars flaccida technique. One-year weight loss data were available in 1,144 patients (47%). BMI decreased to 36.9 Ϯ 7.4 kg/m2 (Ϫ8.8). Mean %EBWL at 1 year (Ϯ 60 days) was 41.0% Ϯ 18.1% (range 0.7%–113.9%). Complications occurred in 241 of 2,411 (10%) patients. There was 1 death (0.04%). Cumulative slippage at 1, 2, and 3 years, respectively, was 0.4%, 2.4%, and 10%. There were 56 (2.3%) port-related problems, and 13 band explantations (0.54%). Conclusions: With extensive staff experience and patient preparation, LAGB can be performed safely as an outpatient procedure for select patients. Close follow-up is crucial in order to optimize LAGB outcomes for the long term. (Surg Obes Relat Dis 2008;4:S56 –S62.) © 2008 American Society for Metabolic and Bariatric Surgery. All rights reserved. Gastric banding; Morbid obesity; Lap-Band; Laparoscopic adjustable gastric banding; Ambulatory; OutpatientIntroduction with comorbidities [12]). Increased use of the pars flaccida technique has contributed to a significant reduction in Laparoscopic adjustable gastric banding (LAGB) has long-term LAGB complications [13–16]. LAGB is aproven to be a safe [1– 6] and effective [7–11] operation to purely restrictive weight–loss-inducing procedure thatinitiate and sustain weight loss for the morbidly obese has many advantages, including adjustability, reversibility,patient (body mass index [BMI] Ն40 kg/m2 or Ն35 kg/m2 and maintenance of normal physiologic absorption of nutrients and medications [17]. A failed band may be replaced, or a *Reprint requests: Brad M. Watkins, M.D., F.A.C.S., Synchrony different bariatric procedure (e.g., vertical banded gastro-Health, Everett, WA 98203. plasty, gastric bypass, biliopancreatic diversion/duodenal switch) may be used [18 –20]. E-mail: [email protected]/08/$ – see front matter © 2008 American Society for Metabolic and Bariatric Surgery. All rights reserved.doi:10.1016/j.soard.2008.04.008

B. M. Watkins et al. / Surgery for Obesity and Related Diseases 4 (2008) S56 –S62 S57 As is appropriate with all bariatric operations, LAGB is Table 1best thought of as a tool to support a process of successful Significant and relative contraindications for an outpatientweight loss rather than as an end in itself. In addition to LAGB procedurelimiting food intake, a properly positioned gastric bandmarkedly increases both fasting and postprandial feelings of Contraindicationssatiety, with no significant change in premeal or post-mealblood glucose, insulin, ghrelin, or leptin levels, [21]. The Significant cardiac comorbiditybest LAGB results are achieved when the patient receives Myocardial infarctionsupport before, during, and for life after the operation in a Coronary artery diseasemultidisciplinary program [22,23]. Congestive heart failure Significant valvular disease The paramount features of a successful bariatric pro- Previous valve repair/replacementgram, in addition to the operation itself, include develop- Abnormal stress testment of a well-trained, experienced staff dedicated to care Significant arrhythmia requiring postop monitoringof the morbidly obese patient, with special emphasis ongeneral anesthesia, management of obstructive sleep apnea Significant pulmonary comorbidity(OSA) and hepatic steatosis (as it relates to operative expo- Emphysemasure), bariatric-grade equipment, nutrition education, phys- COPDical education, psychological support, and proper band ad- Severe restrictive defectjustment. Providing these key elements to patients seeking Significant dyspnea on exertionLAGB in an outpatient rather than an inpatient setting also Poorly controlled asthmarequires careful patient selection and clearly defined pro-cesses that offset the potential risks of performing major Poorly controlled diabetessurgery without on-site hospital resources [24 –31]. Anticoagulant therapy Known coagulopathy In 2002, we dedicated our practice exclusively to gastric Impaired mobility requiring wheelchair or walkerbanding [32,33]. The majority of these procedures were Patients that live alone with inadequate access to supportcompleted in our freestanding, ambulatory surgical center, Relative contraindicationsone in which most of the employees are gastric-band pa-tients themselves. We are also fortunate to work in this Previous coronary artery bypass graft or stentsetting with dedicated bariatric anesthesiologists who play a Any patient taking digitalis or clopidogrelmajor role in optimizing outcomes. This report describes DVT/PE historyour experience with patients who underwent gastric banding Diabetic patients with BMI Ͼ60 kg/m2in our center and, in particular, those who received bands inour ambulatory program, the tenets of which we delineate COPD ϭ chronic obstructive pulmonary disease; DVT ϭ deep veinhere. thrombosis; PE ϭ pulmonary embolism.Methods sidered to be an appropriate candidate for an outpatient procedure in general. In our experience, the majority of We reviewed prospective data on 2,411 patients who patients are candidates for an outpatient procedure. Ourunderwent LAGB between November 2002 and July 2007 super-obese patients (BMI Ͼ50 kg/m2) [36] did not havewith up to 48-months follow-up. All patients in this series significant associated comorbidities that would precludereceived a Lap-Band (Allergan Inc., Irvine, CA); recent safe outpatient management, and thus, most of these candi-patients in the series received the Lap-Band Advanced Plat- dates were accepted for outpatient surgery.form (AP-Band; Allergan Inc.). We would not recommend outpatient banding early in a National Institutes of Health (NIH) criteria were used as bariatric program’s experience. Table 1 details our contra-guidelines for patient selection [12,34]. Ideal body weight indications for outpatient LAGB. Patients with these con-(IBW) and excess body weight (EBW) were determined by ditions were scheduled to have their procedure in the hos-a formula that corresponds to the midpoint of the range for pital.a medium frame on the Metropolitan Height and Weighttables, as described by Deitel and Greenstein [35]. All OSA screeningprocedures were performed by 1 of 3 surgeons (K. M.,R. M., B. W.). OSA screening is required of all patients with a Dixon et al. BASHIM (BMI, age, symptoms, HgbA1C, insulin lev-Contraindications to outpatient LAGB els, male sex) score of Ն3 [37]. Patients with witnessed apnea, neck circumference Ͼ43 cm, and men with diabetes A patient was considered for outpatient LAGB if, after are screened. If, upon screening, patients are found to havehealth screening and surgical consultation, he/she was con- either moderate or severe OSA, we require treatment with continuous positive airway pressure (CPAP) therapy before surgery. We do not consider severe sleep apnea as a con- traindication to outpatient treatment as long as it is treated with appropriately titrated CPAP and the patient compli- antly uses his/her CPAP machine.

S58 B. M. Watkins et al. / Surgery for Obesity and Related Diseases 4 (2008) S56 –S62Hepatic steatosis prevention After thorough pre-oxygenation, patients are induced with propofol, and endotracheal intubation is achieved using Similar to the experience of others [38,39], we are firm succinyl choline. Cricoid pressure is generally used to pre-believers that preoperative weight loss diminishes the size vent aspiration, although none of our patients have shownof the liver and substantially reduces the difficulty of the any signs of gastric regurgitation. Anesthesia is maintainedLAGB operation. To reduce the size of the liver, we require with desflurane and a propofol infusion, which also containspatients to lose 10 lbs for each of the following: BMI Ͼ50 2 ␮g/mL fentanyl and 1 mg/mL ketamine. Rocuronium iskg/m2, male sex, and diabetes. To assist with this weight used for intraoperative muscle relaxation.loss, patients are offered Optifast (Novartis Nutrition, Fre-mont, MI) liquid protein meal replacement shakes. After the intraperitoneal portion of the surgery is com- plete and the port is being secured, the propofol infusion isEquipment requirements and the challenge of discontinued and the concentration of inhalation anestheticpatient ambulation is significantly reduced to help ensure rapid recovery from anesthesia with minimal respiratory depression and rapid The weight limit of most of our equipment, including return of spontaneous airway control. The effect of rocuro-chairs, toilets, gurneys, wheelchairs, and stretchers, is 500 nium is reversed with physostigmine and atropine sulfate.lbs. Our operating table can accommodate 1,100 lbs. Pa-tients are only moved when they are transferred from the Surgical techniqueoperating table to the gurney, in which case, the HoverMatt(HoverTech International, Bethlehem, PA) is used; patients Our surgical technique has been previously describedwalk themselves to the operating room, from the recovery [26]. When a hiatal hernia is encountered, we repair it byroom gurney to the recliner, and from the recliner to their reapproximating the crura with 2-O Ethibond suture, mostlycar. anteriorly, with figure-of-8 suture. For large hiatal hernias, we occasionally reinforce the repair with a Crurasoft patch In addition, all patients receive extensive multidisci- (Bard, Inc., Helsingborg, Sweden). We do not routinelyplinary (i.e., medical, nutritional, psychological, and surgi- order barium esophagography in outpatients unless postop-cal) screening and education before surgery. Patients with a erative dysphagia to liquids occurs, which is rare. We havehistory of fenfluramine use and cardiac symptoms are re- a fluoroscopy unit in our center for this purpose.ferred for echocardiography. For prophylaxis of deep veinthrombosis, patients are required to stop estrogen-contain- All patients are extubated in the operating room and areing medications before surgery; heparin 5,000 U is admin- generally responsive to verbal stimulation within the first 5istered subcutaneously 1 hour before surgery, and compres- minutes of their PACU stay. Most patients are dischargedsion boots are applied during the procedure. from the facility within 2 hours of the completion of sur- gery. Patients return for a routine examination at 1 to 2Anesthetic techniques for obese patients weeks postoperatively. The general anesthetic techniques we use for obese pa- Postoperative visits and band adjustmenttients have been previously published [26]. Since postoper-ative nausea and vomiting have profoundly undesirable con- The first band adjustment is performed after 6 weeks ofsequences both surgically (i.e., nausea can result in band healing, and as needed thereafter. The majority of adjust-slippage) and in relation to outpatient management, a mul- ments are done by direct palpation, in the office, and withtifaceted approach to nausea prophylaxis is used, including fluoroscopy for ports that are difficult to locate if the patientthe perioperative use of a scopolamine patch, intravenous complains of heartburn or if the clinical picture is unclear.administration of metoclopramide, ranitidine, ondansetron, Patients are instructed to call the center if they have not lostand, in many cases, a small dose of droperidol (0.625 mg). weight in 3 weeks. Patients are not required to attend officeEmetogenic agents, such as nitrous oxide, and high doses of visits on a regular schedule, as the need for band adjustmentnarcotics are avoided to the greatest extent possible. While varies considerably among patients. If we have not heardsmall doses of fentanyl are used for perioperative analgesia, from a patient in over a month, we call or e-mail thatwe rely more heavily on meticulous infiltration at the sur- patient.gical sites with bupivacaine, as well as preemptive intrave-nous ketorolac tromethamine and dexamethasone sodium Statistical analysisphosphate. Patients are started on an oral suspension ofhydrocodone and acetaminophen in the post anesthesia care Continuous data were reported as mean Ϯ standard de-unit (PACU) as soon as they are able to take oral medica- viation (SD) and ranges, where applicable; categorical datations. were reported as number and percentage. A 2-sided paired t-test was used to assess a significant reduction in BMI from In the operating suite, an intravenous line is established, baseline. Differences in %EWL between subgroups (i.e.,and patients are given a small defasciculating dose of pan- female vs male; Ͻ50 yrs vs Ն50 yrs; Ͻ50 kg/m2 vs Ն50curonium bromide and the adjuvant agents listed above. kg/m2) were assessed via independent samples t-tests. An

B. M. Watkins et al. / Surgery for Obesity and Related Diseases 4 (2008) S56 –S62 S59analysis of variance (ANOVA) was used to test for differ- Table 3 Weight loss (n ϭ 1,144) P-valueences in %EWL among ethnic groups. Statistical signifi- Weight loss at 1 year 36.9 Ϯ 7.4 (20.9–66.6) Ͻ.001*cance was set at P Ͻ .05. Filemaker software (FileMaker,Inc., Santa Clara, CA) was used for database development Variable %EBWLand management. Mean BMI, kg/m2,Results mean Ϯ SD (range)Patient characteristics Overall mean %EBWL Ϯ 41.0% Ϯ 18.1% (0.7%–113.9%) — SD (range) .034† Between November 2002 and August 2007, prospective 41.4% Ϯ 18.2% .003†data were collected on 2,411 consecutive patients who un- Sex 38.1% Ϯ 17.0% Ͻ.001†derwent LAGB. Patients were predominantly white (95%) Females, mean Ϯ SDand ranged in age from 15 to 76 years (mean 44.1 Ϯ 10.8 Males, mean Ϯ SD 42.0% Ϯ 18.4%yrs) (Table 2). Mean BMI was 45.7 Ϯ 7.9 kg/m2 (range 38.7% Ϯ 17.1%29.1– 83.1 kg/m2) for the total group. Mean body weight Agewas 283.1 Ϯ 58.2 lbs (range 156.2–593.5 lbs). Five hundred Ͻ50 yrs, mean Ϯ SD 43.1% Ϯ 18.5%ninety-five patients (24.7%) had a BMI Ͼ50 kg/m2. Ն50 yrs, mean Ϯ SD 34.1% Ϯ 14.7%Surgery BMI Ͻ50 kg/m2, mean Ϯ SD In 2,027 of 2,411 (84%) patients, LAGB was performed Ն50 kg/m2, mean Ϯ SDin our outpatient center. The majority of outpatient proce-dures were completed in less than 1 hour using the pars BMI at baseline for subset of patients with available weight-loss data ϭflaccida technique, and patients were discharged, on aver- 45.7 Ϯ 7.9 kg/m2. ANOVA results indicated no significant %EWL differ-age, within 2 hours of the procedure. ences among ethnicities.Weight loss * Paired t-test. Weight-loss data were available in 1,144 patients (47%) † Independent samples t-test.(Table 3). Mean BMI decreased by 9.0 kg/m2 to 36.8 Ϯ 7.4kg/m2 (P Ͻ .001). Average percentage excess body weight BMI Ն50 kg/m2 (43.1% Ϯ 18.5% vs 34.1% Ϯ 14.7%,loss (%EBWL) at 1 year (Ϯ 60 days) was 41.0% Ϯ 18.1% respectively, P Ͻ .001).(range 0.7%–113.9%) (P Ͻ .001). %EBWL lost by womenand men was 41.4% Ϯ 18.2% and 38.1% Ϯ 17.0%, respec- Complicationstively (P ϭ .034). Patients Ͻ50 years of age lost 42.0% Ϯ18.4% of their %EBWL; those aged Ͼ50 years lost slightly Complications occurred in 241 patients (10%) and areless EBW: 38.7% Ϯ 17.1% (P ϭ .003). Patients with a BMI detailed in Table 4.Ͻ50 kg/m2 lost significantly more EBW than those with a One patient (0.04%) was converted to an open procedure when we noticed significant bleeding from the spleen. The bleeding was controlled, with a loss of Ͻ2 units of blood. The patient had significant greater-omental adhesions of the spleen. Retraction of the greater omentum to expose the angle of His was felt to be the cause. There was 1 death (0.04%) in the series in a patient with pronounced emphysema. He experienced significant post- operative gastric edema and liquid dysphagia and died of pneumonia 2 weeks later.Table 2 Total n ϭ 2,411 Table 4 Total n ϭ 2,411, n (%)Patient characteristics Total complications and mortality 2,003 (83) 241 (9.9)Characteristic 408 (17) Event 1 (0.04) 44.1 Ϯ 10.8 (15–76) 1 (0.04)Sex 65.9 Ϯ 3.3 (56–78.5) Total complications 6 (0.25) Females, n (%) Mortality 1 (0.04) Males, n (%) 283.1 Ϯ 58.2 (156.2–593.5) Conversion to open procedure 40 (1.7) 45.7 Ϯ 7.9 (29.1–83.1) Superficial wound infection 56 (2.3)Age, yrs, mean Ϯ SD (range) Pulmonary embolusHeight, in, mean Ϯ SD (range) 2,289 (95) Gastric edema 124 (5.1)Weight, lbs, mean Ϯ SD (range) 39 (1.6) Port problem 13 (0.54)BMI, kg/m2, mean Ϯ SD (range) 35 (1.4) Slip/pouch dilation 4 (0.2)Ethnicity, n (%) 26 (1.1) Band explantation (total) 3 (0.1) 18 (0.7) 3 (0.1) White 4 (0.2) Removed by other physician 3 (0.1) African American Bowel perforation requiring removal Hispanic Psychological intolerance Native American Erosion Asian Pacific Other

S60 B. M. Watkins et al. / Surgery for Obesity and Related Diseases 4 (2008) S56 –S62 Six patients (0.25%) in this series had a superficial dysphagia. The on-call surgeon, instead of accessing thewound infection that was treated effectively with oral port and withdrawing saline, removed her band completely.cephalexin. The pulmonary embolism (0.04%) in this series Another patient who was losing weight slowly saw a dif-occurred in a patient with a BMI of 35 kg/m2 who had a ferent surgeon, who converted her to a gastric bypass. Oneperipheral intravenous catheter for postoperative rehydra- patient’s band became too tight and she presented to ation due to gastric edema. We assume that a clot from the different emergency room where the on-call surgeon, in-catheter traveled to her lung when the catheter was re- stead of withdrawing saline, took her to the operating roommoved, because vascular ultrasound noted the clot in the and converted her to a gastric bypass.same upper extremity as the catheter and there was no clotin her legs. Of the 3 bands removed due to bowel perforation, 1 patient represented our first slip repair, and was early in our All cases of gastric edema (40, 1.7%) were treated with experience. Instead of the typical chronic inflammation seensteroids, intravenous rehydration, and time. Typically, the with slips, this patient experienced excessive retching afterassociated liquid dysphagia resolved completely within 2 to surgery and had acutely inflamed tissue. Attempts to re-7 days. With the newer AP-Band, stoma occlusion postop- move the old gastro-gastric sutures resulted in a gastric teareratively was extremely rare; when it occurred, we accessed due to friable tissue; the band was removed out of concernthe port and withdrew the priming saline, which resulted in that it would become infected. Another patient had an un-complete resolution. eventful surgery and presented to the emergency room sev- eral days after surgery with peritonitis. Exploration revealed Port problems (56, 2.3%), such as tubing, or the port not blunt injury to the descending colon, presumably from tro-holding saline, or port malposition, were easily repaired car insertion. A 10-mm circular area of contusion on thewith the patient under local anesthesia in the outpatient unit. colon wall with a 1-mm perforation in the center was found.Several port complications represented infected ports, Another patient that had laparoscopic adhesiolysis inciden-which were repaired by removing the port, allowing the tal to the band operation presented several days later withsubcutaneous infection to resolve, and, after laparoscopic peritonitis; it was noted that she had a cautery injury to theretrieval, reconnecting the tubing to a new port in a different transverse colon, with a tiny perforation.location. There have been no re-infections to date. Three bands in this series were removed due to psycho- In our early experience, in the first year postoperatively, logical intolerance. There were 3 erosions leading to bandthe cumulative slip rate was 0.4%; by 2 years, it rose to removal. One of these patients had chemotherapy, and it is2.4%; and by year 3, it was 10%, although this rate applies unclear whether this played a role. Another erosion patientonly to a relatively small group of patients. Our total slip presented with free air and perforation of a normal sizerate is 5.1% (124/2,412 patients). “Slipped” bands tend to gastric pouch. It is unclear whether a retained tablet mightoccur in patients with excessive regurgitation/vomiting as a have played a role. We did not perform a preoperativeresult of keeping the band too tight, and in those patients endoscopy on this patient, and a diagnosis of erosion waswho chronically tried to curb the band’s restriction. Few of presumed.our cases presented the classic picture of the lower stomachslipping up through the band, as has occurred with the use Discussionof the perigastric operative technique. The majority of our“slips” would be best described as pouch dilation. In all Obesity is a complex disease that requires complex treat-cases, we repositioned the band laparoscopically, with ex- ment. Successful bariatric surgery is best thought of as acellent resolution of symptoms and return to active weight weight-loss program, not an operation; the operation is aloss. small yet significant part of the equation. Weight loss at 1 year in our outpatient center (41.0% Ϯ 18.1%) compares Of 13 total explantations (0.54%), 4 were of unknown favorably with the world literature for LAGB inpatientetiology and were removed by another surgeon, 3 were due procedure outcomes at 2 to 3 years (i.e., from 34.7% [40] toto bowel perforation, 3 to psychological intolerance, and 3 47.5% [41]). We believe that the total program is responsi-to erosion of the band. These are discussed in further detail, ble for the safety and effectiveness achieved in the outpa-as follows. tient setting. One patient presented to an emergency room not associ- We are fortunate to work with medical staff and anes-ated with our center several weeks postoperatively and had thesiologists who are dedicated to caring for bariatric pa-her band removed by the on-call surgeon, who thought the tients. We have performed every outpatient gastric bandband was infected. No intraoperative cultures were taken. operation with the same 3 anesthesiologists. A comprehen-Another patient was an oncology nurse who had been self- sive difficult airway cart is in the room at all times. Alladjusting her band. When we accessed her port in the office, coexisting medical conditions are controlled, with particularit always contained more saline than our records indicated it emphasis on issues associated with hypertension, asthma,should have. She adjusted herself too tightly on vacation and diabetes. The entire team is committed to using pain-once and could not reaccess the port to deflate the band, andshe presented to an emergency room with complete liquid

B. M. Watkins et al. / Surgery for Obesity and Related Diseases 4 (2008) S56 –S62 S61control strategies that eliminate or at least minimize the an adequate support network to assist in his/her weight-lossneed for narcotic analgesics. efforts after surgery. Many patients will see her postopera- tively for one-on-one sessions, and she provides free groupMediating slippage therapy at monthly support classes. All of our receptionists are successful band patients themselves, which provides In our experience, a “slipped band,” or dilated gastric another measure of support to patients. In addition, per-pouch, occurs with increasing frequency over time. In an sonal, telephone, and e-mail support are continuously avail-effort to reduce the slip rate associated with our early able to patients from the receptionists. After-hours calls areexperience, we made several changes in our technique. covered by full-time LAGB surgeons; if a band is too tight,These included using a newly available lower-pressure, our patients can have the band deflated as needed, 24 hourscircumferential balloon, the AP-Band. We were also more a day, 7 days per week.diligent to search for and repair hiatal hernias at the primaryoperation; nearly every slip repair is associated with a hiatal Band adjustments. If the band is too loose, the patientshernia, though it remains unclear whether the hiatal hernia describe hunger, poor volume control, and say, “The band iscontributes to the slip or the slip to the hernia. We also now not working,” or “It feels like I’m on a diet.” If the band isincorporate the anti-slip stitch, whereby the last stitch is too tight, patients complain of heartburn and liquid dyspha-used to close the defect between the stomach brought ante- gia. They develop soft-calorie syndrome, where the onlyrior to the band and the lesser curvature below the band. It thing that will pass through the band is unhealthy food (e.g.,is too early to tell whether any of these changes will alter the fried or greasy food, ice cream). Esophageal dilation on theslip rate. We continue to collect data and examine these fluoroscopy examination is also a sign that the band is tootrends for future reports. tight. When the band is perfectly adjusted, patients easily lose 1 to 2 lbs per week, are full and satisfied, and notLong-term postoperative management hungry for hours after a meal. As bariatric surgeons, it is our responsibility to properly Conclusionimplement and manage all aspects of postoperative care.Outsourcing these disciplines is adequate, as long as legit- With careful planning and management, outpatient gas-imate contributions to weight loss and maintenance of pa- tric banding can be accomplished in the bariatric populationtient safety are ensured. Optimized outcomes cannot be with good outcomes. At our institution, LAGB is an effec-achieved with inadequate staff. tive procedure with a low complication rate in the treatment of morbidly obese patients, and it can be performed in the Our postoperative management concentrates on 4 disci- outpatient setting in select cases. Our patients achieve ac-plines: nutrition education, physical education, psychologi- ceptable weight loss in the first year and continue to losecal support, and band adjustments. weight. When developing the program for outpatient bari- atric surgery, a safety-focused, expertly trained staff, usingNutrition education. In our office, education in nutrition is highly specialized bariatric equipment and emphasizingprovided by 5 full-time banded nurses (band patients, them- very early ambulation, is a necessity.selves) who provide extensive band-specific nutritionalcounseling before and after surgery. Each band adjustment Disclosuresappointment is 30 minutes long and includes trackingweight loss and the degree of current restriction based upon Dr. Watkins is a consultant for Ethicon Endo-Surgery;history and weight loss. The band is adjusted and the patient his expenses for the symposium and an honorarium wereis provided with counseling, as needed. Many nutrition paid by Ethicon Endo-Surgery, Inc., the manufacturer of theteaching tools (e.g., charts, graphs, plastic food, portion Realize Band (also known as the Swedish Adjustable Gas-demonstrations, handouts) are available to the patients in tric Band). Drs. Watkins and Michaelson are proctoringthe examination room and nurses’ offices. surgeons for Ethicon Endo-Surgery. The remaining authors have no commercial associations that might be a conflict ofPhysical education. We encourage our patients to “burn interest in relation to this article.calories.” We have negotiated discounts at most of the localgyms and encourage our patients to attend. Patients have Referencescoordinated walking parties in their local community andbefore support group meetings. We frequently have a phys- [1] Parikh MS, Fielding GA, Ren CJ. U.S. experience with 749 laparo-ical therapist as a speaker at our twice-monthly support scopic adjustable gastric bands: intermediate outcomes. Surg Endoscclasses. Many of our patients seek personal trainers on their 2005;19:1631–5.own. [2] Ponce J, Paynter S, Fromm R. Laparoscopic adjustable gastric band-Psychological support. Our full-time mental health profes- ing: 1,014 consecutive cases. J Am Coll Surg 2005;201:529 –35.sional is a banded patient herself. She screens for untreateddepression preoperatively and ensures that each patient has

S62 B. M. Watkins et al. / Surgery for Obesity and Related Diseases 4 (2008) S56 –S62 [3] O’Brien P, Dixon J. Weight loss and early and late complications–the [23] Favretti F, O’Brien PE, Dixon JB. Patient management after LAP- international experience. Am J Surg 2002;184:42–5S. BAND placement. Am J Surg 2002;184:38 – 41S. [4] Favretti F, Segato G, Ashton D, et al. Laparoscopic adjustable gastric [24] Gutschow CA, Collet P, Prenzel K, Hölscher AH, Schneider PM. banding in 1,791 consecutive obese patients: 12-year results. Obes Long-term results and gastroesophageal reflux in a series of laparo- Surg 2007;17:1– 8. scopic adjustable gastric banding. J Gastrointest Surg 2005;9:941– 8. [5] Adams TD, Gress RE, Smith SC, et al. Long-term mortality after [25] Wasowicz-Kemps DK, Bliemer B, Boom FA, de Zwaan NM, van gastric bypass surgery. N Engl J Med 2007;357:753– 61. Ramshorst B. Laparoscopic gastric banding for morbid obesity: out- patient procedure versus overnight stay. Surg Endosc 2006;20: [6] Sjöström L, Narbro K, Sjöström CD. Effects of bariatric surgery on 1233–7. mortality in Swedish obese subjects. N Engl J Med 2007;357:741–52. [26] Watkins BM, Montgomery KF, Ahroni JA, Erlitz MD, Abrams RE, [7] Angrisani L, Favretti F, Furbetta F, et al. Italian Group for Lap-Band Scurlock JE. Adjustable gastric banding in an ambulatory surgery System: results of multicenter study on patients with BMI Ͻ or ϭ35 center. Obes Surg 2005;15:1045–9. kg/m2. Obes Surg 2004;14:415– 8. [27] McCarty TM, Arnold DT, Lamont JP, Fisher TL, Kuhn JA. Optimiz- [8] Steffen R, Biertho L, Ricklin T, Piec G, Horber FF. Laparoscopic ing outcomes in bariatric surgery: outpatient laparoscopic gastric Swedish Adjustable Gastric banding: a five-year prospective study. bypass. Ann Surg 2005;242:494 –501. Obes Surg 2003;13:404 –11. [28] Huerta S, Heber D, Sawicki MP, et al. Reduced length of stay by [9] Vertruyen M. Experience with Lap-band System up to 7 years. Obes implementation of a clinical pathway for bariatric surgery in an Surg 2002;12:569 –72. academic health care center. Am Surg 2001;67:1128 –35.[10] Belachew M, Belva PH, Desaive C. Long-term results of laparoscopic [29] Ramos AC, Neto G, Passos M, et al. Ambulatory adjustable gastric adjustable gastric banding for morbid obesity. Obes Surg 2002;12: banding surgery. Obes Surg 2003;13:582. 564 – 8. [30] De Waele B, Lauwers M, Nieuwenhove Y, et al. Outpatient laparoscopic[11] Zehetner J, Holzinger F, Triaca H, Klaiber C. A 6-year experience gastric banding: initial experience. Obes Surg 2004;14:1108 –10. with the Swedish Adjustable Gastric Band: prospective long-term audit of laparoscopic gastric banding. Surg Endosc 2005;19:21– 8. [31] Kormanova K, Fried M, Hainer V, et al. Is laparoscopic adjustable gastric banding a day surgery procedure? Obes Surg 2004;14:[12] Hubbard VS, Hall WH. Gastrointestinal surgery for severe obesity: 1237– 40. National Institutes of Health Consensus Development Conference Statement. Obes Surg 1991;1:257– 66. [32] Watkins BM, Montgomery KF, Ahroni JA. Laparoscopic adjustable gastric banding: early experience in 400 consecutive patients in the[13] Fielding GA, Duncombe J. Clinical and radiological follow-up of USA. Obes Surg 2005;15:82–7. laparoscopic adjustable gastric bands, 1998 and 2000: a comparison of two techniques. Obes Surg 2005;15:634 – 40. [33] Ahroni JA, Montgomery KF, Watkins BM. Laparoscopic adjustable gastric banding: weight loss, co-morbidities, medication usage and[14] O’Brien PE, Dixon JB, Laurie C, et al. A prospective randomized trial quality of life at one year. Obes Surg 2005;15:641–7. of placement of the laparoscopic adjustable gastric band: comparison of the perigastric and pars flaccida pathways. Obes Surg 2005;15: [34] Clinical Guidelines on the Identification, Evaluation, and Treatment 820 – 6. of Overweight and Obesity in Adults; The Evidence Report. Be- thesda, Md; NIH Publication: National Institute of Health, National[15] Dargent J. Pouch dilatation and slippage after adjustable gastric Heart, Lung and Blood Institute (NHLBA) in cooperation with The banding: is it still an issue? Obes Surg 2003;13:111–5. National Institute of Diabetes and Digestive and Kidney Diseases (NIDDKD). September 1998.[16] Chevallier JM, Zinzindohoue F, Douard R, et al. Complications after laparoscopic adjustable gastric banding for morbid obesity: experi- [35] Deitel M, Greenstein RJ. Recommendations for reporting weight loss. ence with 1,000 patients over 7 years. Obes Surg 2004;14:407–14. Obes Surg 2003;13:159 – 60.[17] Ledoux S, Msika S, Moussa F, et al. Comparison of nutritional [36] Montgomery KF, Watkins BM, Ahroni JA, et al. Outpatient laparo- consequences of conventional therapy of obesity, adjustable gastric scopic adjustable gastric banding in super-obese patients. Obes Surg banding, and gastric bypass. Obes Surg 2006;16:1041–9. 2007;17:711– 6.[18] Westling A, Ohrvall M, Gustavsson S. Roux-en-Y gastric bypass [37] Dixon JB, Schachter LM, O’Brien PE. Predicting sleep apnea and after previous unsuccessful gastric restrictive surgery. J Gastrointest excessive day sleepiness in the severely obese. Chest 2003;123: Surg 2002;6:206 –11. 1134 – 41.[19] Weber M, Muller M, Michel JM. Laparoscopic Roux-en-Y gastric [38] Fris RJ. Preoperative low energy diet diminishes liver size. Obes Surg bypass, but not rebanding, as rescue procedure for patients with failed 2004;14:1165–70. laparoscopic gastric banding. Ann Surg 2003;238:827–33. [39] Lewis MC, Phillips ML, Slavotinek JP, Kow L, Thompson CH,[20] Kothari SN, DeMaria EJ, Sugerman HJ. Laparoscopic band failure: Toouli J. Change in liver size and fat content after treatment with conversion to gastric bypass and their preliminary outcomes. Surgery Optifast very low calorie diet. Obes Surg 2006;16:697–701. 2002;131:625–9. [40] Maggard M, Shugarman LR, Suttorp M, et al. Meta-analysis: surgical[21] Dixon AF, Dixon JB, O’Brien PE. Laparoscopic adjustable gastric treatment of obesity. Ann Intern Med 2005;142:547–59. banding induces prolonged satiety: a randomized blind crossover study. J Clin Endocrinol Metab 2005;90:813–9. [41] Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a systematic review and meta-analysis. JAMA 2004;292:1724 –37.[22] Shen R, Dugay G, Rajaram B, Cabrera I, Siegel N, Ren C. Impact of patient follow-up on weight loss after bariatric surgery. Obes Surg 2004;14:514 –9.

OBES SURG (2013) 23:1915–1921DOI 10.1007/s11695-013-1062-4 REVIEW ARTICLEPreventing Staple-Line Leak in Sleeve Gastrectomy:Reinforcement with Bovine Pericardium vs OversewingGeorges Nabih Al Hajj & Johnny HaddadPublished online: 22 August 2013 Keywords Bariatric surgery . Laparoscopic . Sleeve# Springer Science+Business Media New York 2013 gastrectomy . Morbid obesity . Gastric leak . Staple-line reinforcement . Buttress material . Bovine pericardiumAbstract One of the most serious, potentially life-threateningcomplications of laparoscopic sleeve gastrectomy (LSG) is Introductionstaple-line leakage. Oversewing the LSG staple line vsbuttressing it with bovine pericardial strips (BPS) to reduce Sleeve gastrectomy (SG) is a restrictive procedure initiallyperioperative bleeding and postoperative gastric leak was developed as part of a staged approach for high-risk, super-evaluated. From 2006 through 2011, 160 patients underwent obese patients [1, 2]. Since 2006, SG has been increasinglyLSG with suturing as the only staple-line reinforcement performed as a stand-alone bariatric procedure with good(Group A). From March 2010 through August 2012, 84 weight loss and resolution of obesity-related comorbiditiesLSG patients had BPS incorporated into their last two stapler [3–17].firings (Group B). Staple lines were evaluated perioperativelyfor bleeding, and patients were monitored for indications of SG, whether performed by open surgery or laparoscopystaple-line leaks (peritonitis, abnormal output from the drain). (LSG), involves the creation of a small gastric reservoir basedIn preoperative Group A and B, there were 117 (73.1 %) vs 56 on the gastric lesser curvature over an orogastric tube, in(66.7 %) females; mean age, 35.2 years (18.0–68.0) vs addition to removal of a large portion of the greater curvature.33.8 years (15.0–64.0); mean body mass index (BMI, kilograms LSG produces a decrease in ghrelin levels for up to 1 year,per square meter), 42.5 (27.0–76.0) vs 42.0 (30.0–58.0). Three which may reduce the desire for food [18, 19]. Notable ad-months after surgery, mean BMI for Group A was 37.3 (−5.9); vantages of LSG include low rates of complications (3–24 %)Group B, 35.2 (−7.3); at 6 months, 32.7 (−10.8) and 31.5 and mortality (0.39 %), the ease of performing the procedure,(−11.3; p <0.001). Although there was no significant difference preservation of the pylorus, maintenance of physiologicalin perioperative blood loss, oversewn staple lines in Group A food passage, and avoidance of foreign material [11–13, 20].often required electrocautery to stanch bleeding; this was notrequired for Group B. In Group A, 15 patients (9.4 %) devel- The most commonly reported complications of LSG areoped complications; in Group B, five (6.0 %; p =0.46). Gas- bleeding, stenosis, and leaks. Bleeding can occur from gastrictric fistula, verified by barium swallow, occurred in eight blood vessels during dissection of the greater curve of theGroup A patients (5.0 %); in Group B, one (1.2 %; p =0.17). stomach. Most bleeding problems associated with LSG occurRelative to oversewing, staple-line buttressing with bovine from the staple line after transection of the stomach [9, 21].pericardium was readily accomplished, safe, and associated This bleeding is most likely a result of the large staples usedwith a lower staple-line leak rate. for the thick tissue in the distal stomach. Large staples are not adequate to seal small vessels [22].G. N. Al Hajj (*) : J. Haddad Gastric leak after LSG is a serious complication, and reportsMiddle East Institute of Health (MEIH), Bsalim, Lebanon of its incidence in numerous studies range from 0 % to 5.7 %e-mail: [email protected] of patients [6, 9, 23–30]. Many surgeons have investigated the reinforcement of staple lines as a means of reducing gastricJ. Haddad leaks after LSG. These efforts have included staple-linee-mail: [email protected] oversewing, covering with omentum or jejunum, applying

1916 OBES SURG (2013) 23:1915–1921fibrin glue, and buttressing the staple line with a material that with the surgeon positioned between the patient's legs, thehas been preloaded onto the stapler gun [31–38]. first assistant on the patient's right, and the second assis- tant on the patient's left. Closed CO2 pneumoperitoneum In this study, we investigated the efficacy of treated bovine was induced by Veress needle insertion 20 cm below thepericardial strips (BPS) as the buttressing material used in the xiphoid process in the midline. Five trocars were positionedlast two stapler firings administered during construction of the in the usual manner.LSG staple line. We compared complication rates with aspecial focus on fistula in two groups of LSG patients: an We proceeded with a standard SG starting at the distalinitial group of 160 whose staple lines were oversewn, and 84 greater curvature and using the Ligasure as a vessel-sealingwho received BPS staple-line buttressing. A follow-up device. Serosal attachments of the posterior gastric wall to thequalitative analysis profiled characteristics of patients who pancreatic capsule were kept in order to prevent subsequentdeveloped postoperative gastric fistula. gastric rotation. The stomach was calibrated with the help of a 36-Fr orogastric tube.Methods All of the resection was accomplished by firing greenPatient Eligibility and Informed Consent endoscopic linear stapler cartridges (Powered Multifire Endo GIA™ 60 stapler with the Powered Multifire Endo SGIA™Patients seen at the Middle East Institute of Health in Bsalim, 60–4.8 knifeless single-use loading unit; Covidien). BetweenLebanon for treatment of morbid obesity underwent a multi- five and eight cartridges were used in each operationdisciplinary evaluation and preparatory process for bariatric depending on the length of the resected stomach. In Groupsurgery by the various clinicians of the Institute's Obesity A, after electrocautery was used to control small venousConsult Center. Patients were considered eligible for LSG if oozing along the nonbuttressed transection line, the line ofthey met the indications for bariatric surgery of the Interna- staples was reinforced with a manual, running, nonabsorbable,tional Federation for the Surgery of Obesity (IFSO) and the Ti-Cron 2–0 seroserosal suture. In Group B, PSD BPS wasNational Institute of Health (NIH) [39, 40]. Informed consent applied to the Endo GIA™ in the last two firings, and thewas obtained from all patients. nonbuttressed caudal part of the transected stomach was reinforced and not imbricated in the same fashion as withStudy Design Group A.In this retrospective analysis of prospectively collected data, No nasogastric tube was left in place. In rare instances, a244 patients were studied who qualified to receive LSG be- methylene blue dye test for determination of staple-line integ-tween 2006 and 2012. Group A was comprised of 160 patients rity was performed. A 14-Fr thoracic drain was placed underwho underwent LSG with continuous oversewing as the only and along the remaining stomach. The resected stomach wasform of staple-line reinforcement (Ti-Cron™ 2–0 suture, removed through the 15-mm trocar, usually without elongat-Covidien, Mansfield, MA, USA). Group B was comprised ing the incision. No fascial defects were closed.of 84 patients who, beginning in March 2010, underwent LSGwith oversewing of the staple line and BPS [using Peri-Strips Data CollectionDry® (PSD) with Veritas® collagen matrix, Synovis/BaxterHealthcare, St. Paul, MN, USA] incorporated into the last two Baseline patient characteristics, including age, gender, height,Endo GIA® stapler firings (using the Endo GIA™ Ultra weight, BMI, and comorbid conditions, were recorded in theUniversal Stapler, Covidien, Mansfield, MA, USA). LSG program's bariatric database. Operative time, blood loss, andprocedures were performed by the same surgeon and surgical complications were recorded. In addition, ease of BPS useteam in an operating room dedicated to obesity procedures. and, if visible, bleeding from the staple lines was noted.Surgical Technique After discharge, all patients were followed per the protocol established by the Institute's Obesity Consult Center, includ-Antibiotic prophylaxis was administered 30 min prior to the ing initial surgical, medical, and dietetic consultation 10 dayssurgical incision and in three doses postoperatively. All pa- after hospital discharge and regular follow-up appointmentstients received venous stasis prophylaxis by external pneu- every 3 weeks thereafter. Although postoperative radiographicmatic compression at the time of surgery and low-molecular or endoscopic studies were not routinely performed, theseheparin 6–8 h after surgery for the duration of hospitalization. studies were conducted on patients who returned for a follow-up appointment with any gastrointestinal symptoms, Under general anesthesia, the patient was placed in reverse such as pain, vomiting, or inadequate weight loss. DataTrendelenburg lithotomic position, arms and legs abducted, from the BPS patients were compared retrospectively to those for the patients who underwent the same procedure without BPS.