Medwrite Portfolio_20 Surgery Sample Articles

OBES SURGTable 5 Metabolic outcomes for diabetic patients at 3 monthsOutcome Value Mean ± SD (range) N=23 Postop Mean change (% change) p valuea PreopAbsolute weight (kg) 137.6±31.3 (80.0–196.2) 111.6±26.2 (72.0–149.0) −26.0 (18.9) <0.01BMI (kg/m2) 50.8±8.9 (34.0–62.4) 41.3±7.9 (32.3–54.9) −9.5 (18.7) <0.01Systolic BP (mmHg) 123.3±5.8 (110.0–140.0) 120.0±10.0 (110.0–130.0) −3.3 (2.7) NSDiastolic BP (mmHg) 82.0±2.0 (60.0–90.0) 72.0±10.6 (60.0–80.0) NSGlycemia (mg/dL) 133.9±29.6 (93.0–234.0) 90.0±19.9 (62.0–116.0) −10.0 (12.2) <0.01OGTT (mg/dL) 138.1±46.4 (92.0–232.0) 87.3±20.3 (62.0–122.0) −43.9 (32.8) <0.05HbA1C (%) −50.8 (36.8) <0.001Cholesterol (mg/dL) 7.6±1.6 (6.2–13.5) 5.9±0.5 (5.3–7.2) NS (0.07) HDL (mg/dL) 217.4±47.6 (138.0–289.0) 198.9±35.4 (133.0–297.0) −1.7 (22.4) NS LDL (mg/dL) 46.4±9.8 (29.0–56.0) 43.6±5.7 (36.0–60.0) −18.5 (8.5) NS Triglycerides (mg/dL) 122.8±35.5 (69.0–184.0) 122.5±33.6 (72.0–172.0) NS (0.09) 239.6±157.6 (83.0–567.0) 132.6±45.2 (86.0–215.0) −2.8 (6.0) −0.3 (0.2) −107.0 (45.0)Preop preoperative, Postop postoperative, BMI body mass index, BP blood pressure, OGTT oral glucose tolerance test, HbA1C glycosylatedhemoglobin, HDL high-density lipoprotein, LDL low-density lipoproteina Paired samples t test assessing change from baselinereference to this problem, Himpens et al. [52], in their long- Seamguard, and suturing (PSD was not a comparator) [54].term study, had preoperatively excluded patients with Other studies have evaluated PSD efficacy in theirsignificant GERD and reported 23% to 26% of patients procedures and demonstrated minimization of leak [26,suffering from frequent episodes of GERD. They also claim 55–58]. In a prior randomized trial at our institution, RYGBthat the “neo-fundus” formation was responsible for both staple-line reinforcement with PSD mounted on the linearweight regain and GERD. With these indications, two stapler was shown to significantly reduce extraluminalpatients in their series were re-sleeved. In this classic paper, bleeding, obtain a dry operating field, abbreviate operatingno mention is made of crural dissection and repair. time, and also, reduce leak [26]. This evidence combined with that of prior SG studies with good results using PSD Although incidence of anastomotic leak is relatively low buttressing influenced our use of the same material toin SG, 2.2% (0% to 20%) [53], its potential to increase reinforce the transection line in SG. The incidence of leakmorbidity, length of hospital stay, and mortality is great, in the current study was 0. Although there was noand there is no agreed standard procedure for leak controlled comparison, we speculate that the use of PSDprevention. Kasalicky et al. report no leak at 18-month reinforcement was a central factor in this outcome. SinceSG follow-up without a reinforced staple line; their summarizing the current data, we have continued our seriestechnique incorporates waiting 30–60 s after stapler closure of reinforced LSGs, now totaling 250 patients, 42 withbefore firing, use of 1–2 interrupted stitches to control cruroplasty, maintaining 0 gastric leaks and no mortality.bleeding, and covering the staple line with 100% oxygen-ous cellulose to prevent residual bleeding [42]. However, A limitation of the study is that it was retrospective andthe 2009 international summit on SG reported that staple- lacked a controlled comparison group. Also, early stagedline reinforcement was employed by 65.1% of surgeon procedures in high-risk patients were combined withresponders; 42.1% used buttressing material, 50.9% over- primary operations, and comorbidity data other than forsewed, and 7.0% used both methods [34]. Surgeons have T2DM were not studied in detail.employed running or interrupted absorbable or nonabsorb-able sutures to oversew the transection line, as well as In conclusion, improved strategies for diagnosis andsegments of dehydrated bovine pericardium (Per-Strips management of hiatal hernia preoperatively and intraoper-Dry® [PSD], Synovis), absorbable polymer (Seamguard®, atively in SG are needed; crural repair pledgets and/orGore), and fibrin sealants. A randomized trial of buttressing mesh, and the associated procedure for their implementa-material used in SG by Dapri et al. found no significant tion, require standardization. The combination of meticu-difference in postoperative leak between no reinforcement, lous surgical technique, PSD reinforcement of the gastric staple line, and a low threshold of conversion to RYGB or

OBES SURGBPD/DS make SG a safe and effective operation in the 15. Brethauer SA, Hammel JP, Schauer PR. Systematic review ofshort term. Weight loss and T2DM outcomes for SG require sleeve gastrectomy as staging and primary bariatric procedure.confirmation over the long term. Surg Obes Relat Dis. 2009;5:469–75.Acknowledgments We thank Giuliana Vitolo and Maria Police of 16. Sánchez-Santos R, Masdevall C, Baltasar, et al. Short- and mid-the Department of Surgery, San Giovanni Bosco Hospital, Naples, term outcomes of sleeve gastrectomy for morbid obesity: theItaly, for their invaluable help in collecting the data, and for their experience of the Spanish National Registry. Obes Surg. 2009;19support in the management and follow-up of patients. This work was (9):1203–10.supported by Synovis Life Technologies, Inc. 17. Rubin M, Yehoshua RT, Stein M, et al. Laparoscopic sleeveConflict of Interest The authors declare that they have no conflict gastrectomy with minimal morbidity: early results in 120of interest. morbidly obese patients. Obes Surg. 2008;4:33–8.References 18. Serra C, Pérez N, Bou R, et al. Laparoscopic sleeve gastrectomy. A bariatric procedure with multiple indications. Cir Esp. 2006;79 1. Clinical Issues Committee of the American Society for Metabolic (5):289–92. and Bariatric Surgery. Updated position statement on sleeve gastrectomy as a bariatric procedure. Surg Obes Relat Dis. 19. Abu-Jaish W, Rosenthal RJ. Sleeve gastrectomy: a new surgical 2010;6(1):1–5. approach for morbid obesity. Expert Rev Gastroenterol Hepatol. 2010;4(1):101–19. Review. 2. North American Association for the Study of Obesity (NAASO) and the National Heart, Lung, and Blood Institute (NHLBI). The 20. Karamanakos SN, Vagenas K, Kalfarentzos F, et al. Weight loss, Practical Guide: identification, evaluation, and treatment of appetite suppression, and changes in fasting and postprandial overweight and obesity in adults. NIH Publication #00-4084, ghrelin and peptide-YY levels after Roux-en-Y gastric bypass and October 2000. sleeve gastrectomy: a prospective, double blind study. Ann Surg. 2008;247(3):401–7. 3. Regan JP, Inabnet WB, Gagner M, et al. Early experience with two-stage laparoscopic Roux-en-Y gastric bypass as an alternative 21. Himpens J, Dapri G, Cadière GB. A prospective randomized in the super-super obese patient. Obes Surg. 2003;13:861–4. study between laparoscopic gastric banding and laparoscopic isolated sleeve gastrectomy: results after 1 and 3 years. Obes Surg. 4. Almogy G, Crookes PF, Anthone GJ. Longitudinal gastrectomy as 2006;16(11):1450–6. a treatment for the high-risk super-obese patient. Obes Surg. 2004;14:492–7. 22. Omana JJ, Nguyen SQ, Herron D, et al. Comparison of comorbidity resolution and improvement between laparoscopic sleeve gastrecto- 5. Hamoui N, Anthone GJ, Kaufman HS, et al. Sleeve gastrectomy my and laparoscopic adjustable gastric banding. Surg Endosc. in the high-risk patient. Obes Surg. 2006;16:1445–9. 2010;24(10):2513–7. 6. Silecchia G, Boru C, Pecchia A, et al. Effectiveness of 23. Akkary E, Duffy A, Bell R. Deciphering the sleeve: technique, laparoscopic sleeve gastrectomy (first stage of biliopancreatic indications, efficacy, and safety of sleeve gastrectomy. Obes Surg. diversion with duodenal switch) on co-morbidities in super-obese 2008;18(10):1323–9. Epub 2008 Jun 6. high-risk patients. Obes Surg. 2006;16:1138–44. 24. Buchwald H, Oien D. Metabolic/bariatric surgery worldwide 7. Cottam D, Qureshi FG, Mattar SG, et al. Laparoscopic sleeve 2008. Obes Surg. 2009;19(12):1605–11. gastrectomy as an initial weight-loss procedure for high-risk patients with morbid obesity. Surg Endosc. 2006;20:859–63. 25. Moy J, Pomp A, Dakin G, et al. Laparoscopic sleeve gastrectomy for morbid obesity. Am J Surg. 2008;196(5):e56–9. 8. Lee CM, Cirangle PT, Jossart GH. Vertical gastrectomy for morbid obesity in 216 patients: report of two-year results. Surg 26. Angrisani L, Lorenzo M, Borrelli V, et al. The use of bovine Endosc. 2007;21:1810–6. pericardial strips on linear stapler to reduce extraluminal bleeding during laparoscopic gastric bypass: prospective randomized 9. Nocca D, Krawczykowsky D, Bomans B, et al. A prospective clinical trial. Obes Surg. 2004;14(9):1198–202. multicenter study of 163 sleeve gastrectomies: results at 1 and 2 years. Obes Surg. 2008;18:560–5. 27. World Medical Association (WMA) Declaration of Helsinki— ethical principles for medical research involving human subjects.10. Tucker ON, Szomstein S, Rosenthal RJ. Indications for sleeve Adopted by the 18th WMA General Assembly, Helsinki, Finland, gastrectomy as a primary procedure for weight loss in the June 1964, amended by the 59th WMA General Assembly, Seoul, morbidly obese. J Gastrointest Surg. 2008;12:662–7. October 2008. World Med J 2008;54(4):122–5.11. Felberbauer FX, Langer F, Shakeri-Manesch S, et al. Laparoscopic 28. Sauerland S, Angrisani L, Belachew M, et al. Obesity surgery: sleeve gastrectomy as an isolated bariatric procedure: intermediate evidence-based guidelines of the European Association for term results from a large series in three Austrian centers. Obes Endoscopic Surgery (EAES). Surg Endosc. 2005;19(2):200–21. Surg. 2008;18:814–8. 29. Fried M, Hainer V, Basdevant A, et al. Inter-disciplinary European12. Skrekas G, Lapatsanis D, Stafyla V, et al. One year after guidelines on surgery of severe obesity. Int J Obes. 2007;31:569–77. laparoscopic “tight” sleeve gastrectomy: technique and outcome. Obes Surg. 2008;18:810–3. 30. Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults; The Evidence Report.13. Langer FB, Bohdjalian A, Shakeri-Leidenmühler S, et al. Conversion Bethesda, MD; September 1998. NIH Publication: National Institute from sleeve gastrectomy to Roux-en-Y gastric bypass—indications of Health, National Heart, Lung and Blood Institute (NHLBA) in and outcome. Obes Surg. 2010;20(7):835–40. cooperation with The National Institute of Diabetes and Digestive and Kidney Diseases (NIDDKD).14. Uglioni B, Wölnerhanssen B, Peters T, et al. Midterm results of primary vs. secondary laparoscopic sleeve gastrectomy (LSG) as 31. Metropolitan height and weight tables. Stat Bull Metropol Life an isolated operation. Obes Surg. 2009;19(4):401–6. Epub 2009 Found 1983;64:3–9. Jan 24. 32. Miller RD, editor. Miller’s anesthesia. Philadelphia: Elsevier Churchill Livingstone; 2005. 33. Societa Italiana di Chirurgia dell’Obesita. Bariatric Surgery Survey, 2008. 34. Gagner M, Deitel M, Kalberer TL, et al. The Second International Consensus Summit for Sleeve Gastrectomy, March 19–21, 2009. Surg Obes Relat Dis. 2009;5(4):476–85. Epub 2009 Jun 13.

OBES SURG35. Bohdjalian A, Langer FB, Shakeri-Leidenmühler S, et al. Sleeve 47. Ovrebo KK, Hatlebakk JG, Viste A, et al. Gastroesophageal reflux gastrectomy as sole and definitive bariatric procedure: 5-year in morbidly obese patients treated with gastric banding or vertical results for weight loss and ghrelin. Obes Surg. 2010;20(5):535– banded gastroplasty. Ann Surg. 1998;228:51–8. 40. Epub 2010 Jan 22. 48. Dolan K, Finch R, Fielding G. Laparoscopic gastric banding and36. Abbatini F, Rizzello M, Casella G, et al. Long-term effects of crural repair in the obese patient with a hiatal hernia. Obes Surg. laparoscopic sleeve gastrectomy, gastric bypass, and adjustable 2003;13:772–5. gastric banding on type 2 diabetes. Surg Endosc. 2010;24 (5):1005–10. Epub 2009 Oct 29. 49. Frezza EE, Ikramuddin S, Gourash TR, et al. Symptomatic improvement in gastroesophageal reflux disease (GERD) following37. Rizzello M, Abbatini F, Casella G, et al. Early postoperative laparoscopic Roux-en-Y gastric bypass. Surg Endosc. 2002;16:1027– insulin-resistance changes after sleeve gastrectomy. Obes Surg. 31. 2010;20(1):50–5. Epub 2009 Nov 15. 50. Angrisani L, Iovino P, Lorenzo M, et al. Treatment of morbid obesity38. Weiner RA, Weiner S, Pomhoff I, et al. Laparoscopic sleeve and gastroesophageal reflux with hiatal hernia by Lap-Band. Obes gastrectomy—influence of sleeve size and resected gastric Surg. 1999;9(4):396–8. volume. Obes Surg. 2007;17:1297–305. 51. Soricelli E, Casella G, Rizzello M, et al. Initial Experience with39. Moon Han S, Kim WW, Oh JH. Results of laparoscopic sleeve laparoscopic crural closure in the management of hiatal hernia in gastrectomy (LSG) at 1 year in morbidly obese Korean patients. obese patients undergoing sleeve gastrectomy. Obes Surg. Obes Surg. 2005;15:1469–75. 2010;20:232–5.40. Gan SS, Talbot ML, Jorgensen JO. Efficacy of surgery in the 52. Márquez MF, Ayza MF, Lozano RB, et al. Gastric leak after management of obesity-related type 2 diabetes mellitus. ANZ J laparoscopic sleeve gastrectomy. Obes Surg. 2010;20:1306–11. Surg. 2007;77:958–62. 53. Dapri G, Cadière GB, Himpens J. Reinforcing the staple line41. Ou Yang O, Loi K, Liew V, et al. Staged laparoscopic sleeve during laparoscopic sleeve gastrectomy: prospective randomized gastrectomy followed by Roux-en-Y gastric bypass for morbidly clinical study comparing three different techniques. Obes Surg. obese patients: a risk reduction strategy. Obes Surg. 2010;20(4):462–7. Epub 2009 Dec 11. 2008;18:1575–80. 54. Stammberger U, Klepetko W, Stamatis G, et al. Buttressing the42. Kasalicky M, Michalsky D, Housova J, et al. Laparoscopic sleeve staple line in lung volume reduction surgery: a randomized three- gastrectomy without an over-sewing of the staple line. Obes Surg. center study. Ann Thorac Surg. 2000;70(6):1820–5. 2008;18:1257–62. 55. Himpens J, Dobbeleir J, Peeters G. Long-term results of laparoscopic43. Vidal J, Ibarzabal A, Romero F, et al. Type 2 diabetes mellitus and sleeve gastrectomy for obesity. Ann Surg. 2010;252:319–24. the metabolic syndrome following sleeve gastrectomy in severely obese subjects. Obes Surg. 2008;18:1077–82. 56. Shikora SA. The use of staple-line reinforcement during laparoscopic gastric bypass. Obes Surg. 2004;14(10):1313–20. Review.44. Tagaya N, Kasama K, Kikkawa R, et al. Experience with laparoscopic sleeve gastrectomy for morbid versus super morbid 57. Shikora SA, Kim JJ, Tarnoff ME. Comparison of permanent and obesity. Obes Surg. 2009;19:1371–6. nonpermanent staple line buttressing materials for linear gastric staple lines during laparoscopic Roux-en-Y gastric bypass. Surg45. Gill RS, Birch DW, Shi X, et al. Sleeve gastrectomy and type 2 Obes Relat Dis. 2008;4(6):729–34. Epub 2008 Jun 30. diabetes mellitus: a systematic review. Surg Obes Relat Dis. 2010;6 (6):707–13. 58. Daskalakis M, Berdan Y, Theodoridou S, et al. Impact of surgeon experience and buttress material on postoperative complications46. Shi X, Karmali S, Sharma AM, et al. A review of laparoscopic after laparoscopic sleeve gastrectomy. Surg Endosc. 2011;25(1):88– sleeve gastrectomy for morbid obesity. Obes Surg. 2010;20:1171–7. 97.

Surgery for Obesity and Related Diseases ] (2014) 00–00 Original articleComparison of laparoscopic sleeve gastrectomy leak rates in four staple-line reinforcement options: a systematic review Michel Gagner, M.D.a,*, Jane N. Buchwald, B.A.b aDepartment of Surgery, Hopital du Sacré Coeur, Montréal, QC, Canada bDivision of Scientific Research Writing, Medwrite Medical Communications, Maiden Rock, WI, U.S.Abstract Objective: The study compared laparoscopic sleeve gastrectomy (LSG) staple-line leak rates of 4 prevalent surgical options: no reinforcement, oversewing, nonabsorbable bovine pericardial stripsKeywords: (BPS), and absorbable polymer membrane (APM). Background: LSG is a multipurpose bariatric/metabolic procedure with effectiveness proven through the intermediate term. Staple-line leak is a severe complication of LSG for which no definitive method of prevention has been identified. Methods: The systematic review study design was employed using Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement screening guidelines. Inclusion criteria centered on variables potentially relevant to LSG leak: leak rate, age, gender, calibrating bougie size, distance between pylorus and gastric transection line, overall complication rate, and mortality. Analysis of variance models were used to explore differences in select demographic and surgical technique variables characterizing each reinforcement group. An omnibus χ2 test followed by independent Fisher’s exact tests were used to compare leak rates. Results: There were 659 articles identified; 41 duplicates removed. Of 618 remaining articles, 324 did not meet inclusion criteria. Of the 294 remaining articles, 206 were eliminated (kin studies, those not reporting staple-line or leak incidence, those reporting discontinued products). There were 88 papers included in the analysis. Statistically significant differences were found between groups across demographic and surgical variables studied (p o 0.001). There were 191 leaks in 8,920 patients; overall leak rate 2.1%. Leak rates ranged from 1.09% (APM) to 3.3% (BPS); APM leak rate was significantly lower than other groups (p o 0.05). Conclusion: Systematic review of 88 included studies representing 8,920 patients found that the leak rate in LSG was significantly lower using APM staple-line reinforcement than oversewing, BPS reinforcement, or no reinforcement. (Surg Obes Relat Dis 2014;]:00–00.) r 2014 American Society for Metabolic and Bariatric Surgery. All rights reserved. Bariatric; Metabolic; Laparoscopic sleeve gastrectomy; LSG; Leak; Staple line; Reinforcement; Systematic review Laparoscopic sleeve gastrectomy (LSG) has proven long-term data are accruing [1–5]. In 1993, Marceauhighly effective in achieving durable weight loss and co- reported a new hybrid procedure, the biliopancreatic diver-morbidity reduction over the short and intermediate terms; sion (BPD) with duodenal switch (DS) [6], an operation combining a parietal gastrectomy (later termed a SG) with a *Correspondence: Michel Gagner, MD., Clinical Professor of Surgery, modified Scopinaro BPD [7] and the DeMeester DS [8]. InChief, Bariatric and Metabolic Surgery, 315 Place D’Youville, Suite 191, 1999, SG was first performed laparoscopically as a portionMontréal, QC, Canada H2Y 0A4. of a DS by Gagner et al. [9,10]; after observing marked weight loss with LSG, the operation was proposed as a first- E-mail: [email protected]://dx.doi.org/10.1016/j.soard.2014.01.0161550-7289/r 2014 American Society for Metabolic and Bariatric Surgery. All rights reserved.

2 M. Gagner and J. N. Buchwald / Surgery for Obesity and Related Diseases ] (2014) 00–00step procedure for high-risk patients with a second step reinforcement by oversewing, reinforcement with nonab-BPD/DS or RYGB [11,12]. Subsequently, LSG was studied sorbable bovine pericardial strips, and reinforcement withas a definitive operation. The current procedure involves a absorbable polymer membrane.stapled mostly vertical transection of the stomach andremoval of the left gastric fundus body and proximal left Methodsantrum to create a tubular alimentary channel along thestomach’s lesser curvature. Food restriction, early satiety, Search strategy, inclusion criteria, variables of interestdecreased production of the appetite-stimulating hormoneghrelin, and increased production of GLP-1 and PYY-36 An electronic literature search of the National Library ofhave been associated changes after LSG [13,14]. Medicine’s PubMed database was performed. The search strategy followed the identification and screening guidelines An assessment of the global distribution of metabolic/ established by the Preferred Reporting Items for Systematicbariatric surgery procedures found that, after Roux-en-Y Reviews and Meta-Analyses statement (PRISMA) [24].gastric bypass (RYGB) SG/LSG was the most performed Articles were identified by the keywords: “sleeve gastrec-metabolic/bariatric procedure, trending upward from0% in tomy,” laparoscopic sleeve gastrectomy,” and “vertical gas-2003 to 5.3% in 2008 to 27.8% (94,689 procedures) in 2011 trectomy,” and were limited to research papers of all designs[15]. Between 2008 and 2011, the regional trend of SGs on human patients in the English language. The searchperformed increased dramatically by 571.8% in Europe, included records from no beginning date through March 2012.121.4% in the United States and Canada, 561.7% in SouthAmerica, and 2,255.4% in the Asia/Pacific region [15]. LSG Electronic results were screened by title to excludewas recently recognized as a primary metabolic/bariatric duplicate studies resulting from collecting e-publicationssurgical operation by the American Society for Metabolic that were followed several months later by print publica-and Bariatric Surgery and the American College of Sur- tions. The remaining records were screened by readinggeons [16,17]. abstracts. Studies that reported on LSG were included if the study included leak data and if the type of staple-line According to a recent meta-analysis, LSG has been reinforcement used, or lack thereof, was clarified. Com-associated with 57.6% mean excess weight loss (EWL) at 1 ments, Letters to the Editor, articles with no actual staple-year and 70.1% EWL at 3 years [1], superior to that of line reinforcement or leak data, or with no focus on LSG,restrictive procedures and comparable to malabsorptive- were excluded from eligibility in the analysis, as were caserestrictive and primarily malabsorptive procedures over the reports, animal studies, abstracts, series with o5 patients,intermediate term. Other advantages of LSG include its and review articles without accompanying data. Of eligiblerelative procedural simplicity; no need to introduce a foreign full-text articles, those not describing a form of reinforce-body; absence of an enteric anastomosis; maintenance of ment studied (e.g., a reinforcement material, oversewing, orgastrointestinal continuity; no or low risk of ulceration and no reinforcement); kin studies (i.e., reports discovered tointernal hernia; lower rates of dumping syndrome; lowered contain overlapping data, or the same author group report-ghrelin levels; improved quality of life; and no need to ing outcomes for similar periods of time); and those articlesperform postoperative adjustments [2,26,27]. The overall not focused on percentage leak incidence were excluded.complication rate in large medical centers is r15%; theirreversibility of LSG and risk of gastroesophageal reflux pose Data collection objectives centered on variables potentiallypotential issues [18]. Overall LSG mortality is .3% with a relevant to leak in LSG, including: prior bariatric/metabolicleak-related mortality of .1% [ 1]. The .0–5.5% incidence of surgery, gender, age, BMI, staple height, intraoperative andstaple-line leak is perhaps the most salient concern and target postoperative leak tests, the calibrating bougie size, thefor improvement associated with the procedure [16,17,20,21]. distance between the pylorus and gastric transection line, overall complications, mortality, and %EWL. In 1993, the most significant complication of BPD/DSwas gastric leak and fistula [6]; in 2013, leaks continue to Abbreviated termsdetract from LSG outcomes, at times resulting in a high rateof morbidity and revisional procedures [22]. Most LSG No reinforcement ¼ “N”. Reinforcement by oversewingstaple-line leaks are proximal (89%) [19] and may be alone ¼ “O”. Reinforcement with nonabsorbable bovinerelated to ischemic or mechanical issues caused by intra- pericardial strips (Peri-Strips Dry, Baxter Healthcare, St.luminal pressures sufficient to exceed tissue and suture-line Paul, MN) ¼ “BPS”. Reinforcement with absorbableresistance [23]. A variety of surgical options, including polymer membrane (GORE SEAMGUARD Reinforcement,staple-line reinforcement, suture invagination, and biolog- W. L. Gore & Associates, Elkton, MD) ¼ “APM”.ical sealand have been used to try and reduce leakincidence. The aim of the present study was to employ Statistical analysisthe systematic review (SR) study design to collect relevantevidence for comparison of the staple-line leak rates of 4 All data were extracted from original sources to fieldsprevalent surgical options in LSG: no reinforcement, within an Excel (Microsoft, Redmond, WA) database. Data

LSG Leak Rates in 4 Reinforcement Options / Surgery for Obesity and Related Diseases ] (2014) 00–00 3manipulation and analysis was conducted using JMP rates within each of the 4 reinforcement categories werestatistical software, version 8.0.1 (SAS Institute Inc., Cary, calculated. An omnibus χ2 test in conjunction with inde-NC), “R” open source statistical software, version 2.15.0 (R pendent Fisher’s exact tests were used to compare theFoundation for Statistical Computing, Vienna, Austria), and number of patients with and without leaks for the differentSPSS software, version 20 (IBM SPSS, Chicago, IL). reinforcement options. In addition, weighted-mean rates ofCriteria-based data were aggregated from selected studies overall complications and excess weight loss were calcu-representative of the 4 LSG reinforcement options of lated for each reinforcement method. For studies reportinginterest. Select demographic (i.e., age, % females, body weight loss in terms of BMI reduction, %EWL wasmass index [BMI, kg/m2]) and surgical technique (i.e., calculated as: [(BMI change)/(baseline BMI – 25)] X 100.bougie size, distance from pylorus) variables were summar- All statistical tests were 2-tailed and alpha was set atized using mean, SD, range, and the percentage of studies p o 0.05.reporting on each variable. One-way ANOVA tests wereperformed on study summary data using reinforcement Resultsmethod as the grouping variable. Post hoc pairwise t testswere used to determine where statistically significant differ- Study characteristicsences existed. Pearson’s χ2 test was used to determinewhether there were any differences in the distribution of Figure 1 presents the results of the systematic review ofgender by reinforcement method; a permutation test was the literature through record identification, screening, andused to identify where significant differences existed. The analysis using the PRISMA diagram format. A total of 659overall leak rate for LSG patients, as well as, patient leak articles were identified in the initial search. Forty-oneFig. 1. Results of the systematic review of the literature by article identification, screening, and analysis using the Preferred Reporting Items for SystematicReviews and Meta-Analyses statement (PRISMA) flow diagram format.

4 M. Gagner and J. N. Buchwald / Surgery for Obesity and Related Diseases ] (2014) 00–00duplicates were removed. Of the 618 articles remaining to 8.0% of studies were randomized controlled trials (RCTs). Four studies (4.0%) did not report study design. Countriesbe screened by title and abstract, 324 did not meet search in Europe were the origin of the greatest number of LSG articles (56.0%) relative to the Americas (28.0%), Asiastrategy criteria and were excluded. Of the remaining 294 (15.0%), and Oceania (1.0%). A method of LSG staple-linefull-text articles assessed for eligibility, 206 were eliminated reinforcement was the main study focus in 29.0% of articles included in the analysis. The mean study duration across allas either not presenting staple-line reinforcement informa- papers reviewed was 2.8Ϯ1.7 years (median 3.0 years).tion (n ¼ 107), not presenting leak data (n ¼ 69), kinstudies (n ¼ 17), those not focused on percentage leak Patient characteristicsincidence (n ¼ 11), or papers that reported the Duet TRSTMReload reinforcement product (Covidien, Mansfield, MA), The final analysis incorporated 8,920 LSG patientsrecalled in 2012 (n ¼ 2). A porcine small intestinal (71.1% females) with data on reinforcement methods andsubmucosa reinforcement product manufactured by Cook leak rate. Overall, mean patient age was 41.0 years, pooled SD: 10.9 years; and mean patient BMI was 45.6Ϯ7.3.Biotech (West Lafayette, IN), was not included in this Approximately 40.0% of all studies included patients withanalysis because there were no clinical publications describ- previous bariatric/metabolic surgery, with each study group comprised of r10.0% revision patients. In addition, aing its use in LSG. significant proportion of patients who had prior bariatric A final set of 88 papers was included in the SR. Of these, surgery underwent procedures that incurred fewer compli- cations than major revisional procedures, e.g., band32 papers were counted in the N category [23–56]; 42, in removals.the O category [3,18,20,26–29,47,48,57–89]; 9 in the BPScategory [26,28,48,57,59,90–93]; and 22 in the APM As reported in Table 2, ANOVA models comparing staple-category [25–27,46,48,58,91,94–109]. (In the analysis of line reinforcement methods along demographic variablescharacteristics of the above accepted studies [Table 1], indicated the presence of multiple significant differences.certain papers were counted in 41 category if they With regard to age, pairwise t tests revealed that LSG patientsincluded data on Z2 reinforcement options of interest.) receiving APM were significantly older than patients Table 1 displays the general characteristics of includedstudies by reinforcement option. The publication date ofstudies ranged from 2001 to 2012. The predominant designsacross LSG reinforcement options were prospective(48.0%) and retrospective (42.0%) case series. A total ofTable 1Characteristics of accepted studiesReinforcement method Publication date range Study design Continent of article origin Reinforcement as a main focus type ¼ N (%) Article N (%) N (%)No reinforcement 2001–2012 P ¼ 14 (44.0) Asia ¼ 10 (31.0) 9/32 (28.0) R ¼ 16 (50.0) Americas ¼ 4 (13.0) 9/42 (21.0)Oversewing (suture) 2005–2012 RCT ¼ 2 (6.0) Europe ¼ 18 (56.0) 6/9 (67.0) NR ¼ 0 (.0) Oceania ¼ 0 (0) 6/22 (27.0)Bovine pericardium 2009–2012 P ¼ 19 (45.0) Asia ¼ 4 (10.0) 30/105a (29.0) R ¼ 17 (40.0) Americas ¼ 11 (26.0)Absorbable membrane 2004–2012 RCT ¼ 4 (10.0) Europe ¼ 26 (62.0) NR ¼ 2 (5.0) Oceania ¼ 1 (2.0)Total 2001–2012 P ¼ 4 (44.0) Asia ¼ 0 (.0) R ¼ 5 (56.0) Americas ¼ 1 (11.0) RCT ¼ 0 (.0) Europe ¼ 8 (89.0) NR ¼ 0 (.0) Oceania ¼ 0 (.0) P ¼ 12 (55.0) Asia ¼ 2 (9.0) R ¼ 6 (27.0) Americas ¼ 13 (59.0) RCT ¼ 2 (9.0) Europe ¼ 7 (32.0) NR ¼ 2 (9.0) Oceania ¼ 0 (.0) P ¼ 49 (47.0) Asia ¼ 16 (15.0) R ¼ 44 (42.0) Americas ¼ 29 (28.0) RCT ¼ 8 (8.0) Europe ¼ 59 (56.0) NR ¼ 4 (3.0) Oceania ¼ 1 (1.0) NR ¼ not reported; P ¼ prospective; R ¼ retrospective; RCT ¼ randomized controlled trial. Asia includes articles from: Israel, India, Singapore, China, Saudi Arabia, Korea, Taiwan, and Japan. The Americas includes articles from: United States,Canada, Chile, Venezuela. Europe includes: France, Netherlands, United Kingdom, Austria, Greece, Italy, Czech Republic, Belgium, Switzerland, Sweden,Spain, and Germany. Oceania include: New Zealand. aA denominator of 105 was used for total percentage calculations (rather than the actual number of studies [88]), as several studies included data on Z 2reinforcement options of interest.

LSG Leak Rates in 4 Reinforcement Options / Surgery for Obesity and Related Diseases ] (2014) 00–00 5Table 2 Significant differences in the distance of the transectionDemographic data by reinforcement method line from the pylorus were also evident. The overall model mean distance (and pooled SD) was 4.9Ϯ1.5 cm. AlthoughReinforcement method Variable mean Ϯ SD the APM and O reinforcement groups were statistically (Rangea) similar, both showed significantly greater mean distances [% Reportedb] from the pylorus compared to the N and BPS groups, 5.4Ϯ0.9 [APM] and 5.4Ϯ1.3 [O] versus 4.1Ϯ1.5 [N] and Age % Female BMI 4.6Ϯ2.6 [BPS] (p o 0.001). Also, a significantly smaller distance between the transection line and pylorus was usedNo reinforcement 40.9Ϯ10.6 73.0Ϯ11.0 46.5Ϯ7.7 in the N group compared to all other groups (p o 0.001).Oversewing (suture) (27.0–47.1) (43.0–100.0) (28.5–68.0)Bovine pericardium [93.8] [90.6] [96.9] Mobilization of the greater curvature and fundus beforeAbsorbable 40.8Ϯ11.1 72.0Ϯ10.0 43.1Ϯ6.5 the resection was performed in all but three studies.P-value (33.9–54.0) (10.0–97.0) (34.0–58.2) Heterogeneity of staple heights between studies precluded [90.5] [95.2] [95.2] meaningful statistical analysis; however, the most common 38.9Ϯ10.3 67.0Ϯ9.0 48.5Ϯ8.2 stapling technique used across reinforcement methods (34.3–45.9) (51.0–76.0) (43.2–56.7) (approximately 50.0%) was a combination of staple car- [77.8] [77.8] [88.9] tridges—a green load (4.8 mm), used at the antrum, and a 43.0Ϯ10.9 66.0Ϯ22.0 50.0Ϯ8.8 blue load (3.5 mm), applied at the gastric corpus and (37.6–47.3) (21.0–93.0) (44.0–60.1) fundus. The most common form of intraoperative leak test [86.4] [86.4] [90.9] used across reinforcement methods studied was the meth- o .001 o .001 o .001 ylene blue dye test; approximately 60.0% of all studies employed this test. The second most commonly used testBMI ¼ body mass index (kg/m2); SD ¼ standard deviation. (approximately 30%) was the air-leak test.aRange ¼ minimum to maximum.bPercentage of studies reporting on variables. Staple-line leak ratecomprising the remaining three reinforcement groups, Approximately 80% of all studies reporting on post-43.0Ϯ10.9 [APM] versus 40.9Ϯ10.6 [N], 40.8Ϯ11.1 [O], operative leak determination performed a water-soluble38.9Ϯ10.3 [BPS] (all differences significant at p o 0.001). upper gastrointestinal contrast study; the most commonlyConversely, BPS patients were found to be significantly referenced contrast agent was gastrografin. In total, 191younger than all other patient groups (p o 0.001). leaks were reported in 8,920 patients (overall incidence: 2.1%). As reported in Table 4, leak rates after LSG ranged The overall mean percentage of female LSG patients from 1.1% (n ¼ 16/1,462) in patients reinforced with APM,across reinforcement methods was 71.1%. APM and BPS Table 3 Bougie size and distance from pylorus by reinforcement methodpatient groups were statistically similar in gender distribu- Reinforcement method Variable mean Ϯ SDtion (66.0% versus 67% female); however, both groups (Range)awere comprised of significantly fewer females than the N [% Reportedb](73.0%) and O (72.0%) groups (differences significant atp o 0.001). The N and O groups were not significantlydifferent in gender distribution. Significant BMI differences between the four reinforcementgroups were observed (all differences significant atp o0.001). The heaviest patients were in the APM groupwith a mean baseline BMI of 50.0Ϯ8.8; whereas, the O grouppresented with the lowest relative mean BMI (43.1Ϯ6.5). Bougie size (Fr) Distance from pylorus (cm)Surgical technique No reinforcement 36.5Ϯ8.8 4.1Ϯ1.5 Oversewing (suture) (32.0–60.0) (2.5–7.0) Table 3 presents summary data for relevant surgical Bovine pericardium [90.6] [53.1]technique variables specific to LSG staple-line reinforce- Absorbable membrane 44.4Ϯ10.6 5.4Ϯ1.3ment methods. Again, ANOVA indicated the presence of P-value (30.0–60.0) (2.5–7.0)significant differences among the reinforcement methods [95.2] [54.8]studied. With regard to bougie size (overall model mean: 37.3Ϯ4.6 4.6Ϯ2.640.2 [Fr], pooled SD: 8.9), O group patients had a (32.0–48.0) (3.0–9.0)significantly greater mean bougie size than patients in the [100.0] [67.0]remaining 3 reinforcement groups, 44.4Ϯ10.6 [O] versus 36.1Ϯ3.3 5.4Ϯ0.936.5Ϯ8.8 [N], 37.3Ϯ4.6 [BPS], 36.1Ϯ3.3 [APM] (all (29.0–60.0) (3.0–8.0)differences significant at p o 0.001). Conversely, a signifi- [90.9] [64.0]cantly smaller mean bougie size was used in the APM o .001 o .001group relative to the other reinforcement groups(p o 0.001). Fr ¼ French; SD ¼ standard deviation. aRange ¼ minimum to maximum. bPercentage of studies reporting on variables.

6 M. Gagner and J. N. Buchwald / Surgery for Obesity and Related Diseases ] (2014) 00–00Table 4Leak rates by reinforcement methodReinforcement method Leaks Number of Patients w/o leaks % Leaks P-value compared to APMaAbsorbable membrane 16 1,446 1.09 —Oversewing (suture) 86 4,128 2.04 .02No reinforcement 67 2,512 2.60 .001Bovine pericardium 22 3.30 .0006Total 191 643 2.14 — 8,729APM ¼ (Seamguard) absorbable polymer membrane; PSD ¼ (Peri-Strips Dry) bovine pericardial strips.aTwo-tailed Fisher’s exact test.to 3.3% (n ¼ 22/665) in patients reinforced with BPS. A 2 x significantly lower than the other reinforcement options4 contingency table χ2 test indicated that significant differ- studied. This was true despite the fact that the APM groupences in the leak rate existed among reinforcement methods was characterized by use of smaller bougie sizes and a(X2 ¼ 14.7, p o 0.005). Fisher’s exact tests demonstrated greater baseline BMI—factors that have been previouslythat the leak rate of 1.1% for the APM group was linked to higher leak rates [1,19]. The APM group was alsosignificantly lower than that attained by the 3 other found to be significantly older and to have had the LSGreinforcement options (all differences significant at transection performed at a distance from the pylorus (5.4p o 0.05). In addition, there was a marginally significant cm) Zto that of the other groups. In addition, the APMdifference in leak rate between the O and BPS groups (2.0% group was found to have the lowest overall weighted-meanversus 3.3%, p ¼ 0.046). rate of complications (5.5%).Overall complications During the past decade, the time frame in which almost all LSG evidence has been published, 90.0% of articles Overall mortality rate was .1% (n ¼ 9/8,920 patients). included in the current SR were prospective and retrospec-Total sample weighted-mean rate of overall complications tive case series, with 8.0% RCTs. The current analysiswas 7.01%. Independent weighted-mean overall complica- comprised data from 88 papers; yet, 187 LSG papers weretion rates for each reinforcement method are presented in excluded because they did not report staple-line leak ratesFigure 2. The APM group was found to have the lowest rate or key surgical techniques that may affect leak. Standardsof overall complications (5.5%), and the no reinforcement for reporting LSG do not yet exist; without them, it isgroup, the highest (8.9%). difficult to include a wide, and perhaps more representative, array of reports in a pooled analysis and to compare dataWeight loss across studies. Numerous individual LSG series included in this SR compared leak rates between no reinforcement, All reinforcement option groups experienced significant oversewing, and/or Z2 reinforcement options with con-weight loss at 1 year (450% EWL). Overall weighted- flicting results. Several reinforcement technique and productmean EWL across reinforcement methods was 64.5% at 1 categories are available, each with multiple specific brandyear. The oversewing group (lowest baseline BMI, 43.1) and composition options, further complicating pooledhad the highest weight loss at 1 year, with a weighted-meanEWL of 73.1%. The BPS group had the second highestweight loss, 65.7% EWL, followed by the no reinforcementgroup, 60.9% EWL, and the APM group (highest baselineBMI, 50.0%), 58.4%.Discussion Fig. 2. Weighted-mean overall complication rates for each of the four reinforcement options. APM (Seamguard) absorbable polymer membrane,Current LSG review findings and the literature PSD (Peri-Strips Dry) bovine pericardial strips. In this study, 88 systematically reviewed articles relatingLSG staple-line reinforcement methodology and leak rateswere analyzed. Within a total sample size of 8,920 patients,191 leaks were recorded for an overall leak rate of 2.1%.Four reinforcement options were studied: no reinforcement,oversewing, bovine pericardial strip reinforcement (BPS,Peri-Strips Dry), and absorbable polymer membrane rein-forcement (APM, Seamguard). APM patients were foundto have a leak rate of 1.1%, which was statistically

LSG Leak Rates in 4 Reinforcement Options / Surgery for Obesity and Related Diseases ] (2014) 00–00 7comparison within the reinforcement category. Identifying and staple-line reinforcement (or no reinforcement). Parikhdefinitive evidence regarding the safest and most effective et al. identified an overall leak rate of 2.2% (198 leaks intechniques for preventing staple-line leak is also problem- 8,922 patients), comparable to that of the current SRatic without observational studies that report LSG outcomes (2.1%). Reinforcing the staple line was not found to havein a standardized manner and a greater volume of RCTs. a significant effect on leak rate; however this may be because in their definitive analysis, all absorbable reinforce- To our knowledge, there are only three randomized trials ments were included, whereas in the present study, a(RTs) [27,58,110], 1 of which was controlled [27]. Three significant difference was found between absorbable andSRs [1,19,111] (2 of which are SRs with more rigorous nonabsorbable reinforcement. As observed in other studiesmeta-analysis [MA]) have focused on LSG staple-line focused on bougie size [114,115], Parikh et al. calculated areinforcement and may be compared to the current SR significant diminution of the leak rate up to 3-year follow-findings. While these conflict in their conclusions as to the up (with no impact on weight loss) when a Z40Fr bougieoptimal staple-line reinforcement option, they are in agree- was used to size the gastric sleeve. In addition, leak rate wasment that the leak rate for LSG is relatively low— not shown to be affected by the distance of the pylorus fromgenerally, o2.4%19 and o1.3% [112,113] in experienced the gastric transection line [1] . An SR by Aurora et al.hands—and that reinforcement does not significantly differ found a similar overall leak rate of 2.4% (115 leaks in 29from no reinforcement in reducing staple-line leak. studies encompassing 4,888 patients), higher in the super- obese (2.9%). Eighty-nine percent of leaks were identified An RCT reported by Dapri et al. in 2010 [27], and RTs at the esophagogastric junction. Like Parikh et al. andpublished in 2012 by Albanopoulos et al. [58] and unlike the present study, Aurora et al. combined reinforce-Gentileschi et al. [110], resulted in similar outcomes ment methods and concluded that reinforcement did notregarding prevention of staple-line leak in LSG. In a study have a significant effect on leak rate. Aurora et al. found, asof 75 comparable morbidly obese patients randomized to did Parikh et al, that the risk of leak was higher when theLSG with either no reinforcement (control), staple-line bougie size used was o40Fr19. Choi et al.’s SR/MAoversewing, or buttressing with APM, Dapri et al. noted a included 8 studies incorporating 1,345 patients. The ORsignificant reduction in operative blood loss with APM, but comparing staple-line reinforcement (either buttressing orno significant difference between the 2 options and the oversewing or both) to no reinforcement for hemorrhagecontrol in preventing postoperative staple-line leak[27]. was .559 (95 % CI, .247–1.266), and for leak, .425 (95%Albanopoulos et al. randomized 90 morbidly obese patients CI, .226–0.799). Contradictory to the previous 2 SRs, Choiwith like characteristics to either an APM or oversewing et al. found that reinforcement with oversewing and/ortreatment group. Comparable to the complication rate for buttressing material had a statistically significant effect onAPM in the current SR (5.5%), Dapri et al. reported a total reduction of leak rate and overall complications [111].complication rate of 6.2% in the APM group, with 2patients developing postoperative staple-line leak under Three sequential international consensus summitsthe gastroesophageal junction (4.2%). There were .0% [112,113,116] and 1 international expert panel summarycomplications in the oversewing group; however, there [117] published during the last decade ratify the value ofwas no statistically significant difference between the 2 LSG as a staged and stand-alone procedure. Although, basedmethods in preventing leak [58]. In a RT of 120 morbidly on the 2011 Oxford Centre for Evidence-Based Medicineobese patients randomized to an oversewing group, an APM “Levels of Evidence” [118], reports of expert opinion carrygroup, or a group that received a staple-line roofing matrix less evidentiary weight than RCTs, RTs, SRs, and MAs, the(developed from a bovine-derived gelatin matrix of throm- Third International Consensus Summit for Sleeve Gastrec-bin derived from humans mixed with additional compo- tomy convened experts with a substantial combined experi-nents; Floseal, Baxter Biosurgery, Deerfield, IL) used for ence of 19,605 LSG cases with a proximal leak rate of 1.3%the first time in this LSG study, Gentileschi et al. recorded, [112]. Of the 67.1% majority of expert surgeons who usedrespectively, 1 bleed and 1 leak; 1 bleed; and 1 leak, with reinforcement in LSG, 43.0% oversewed the staple lineno statistically significant difference between the methods (sutures applied seromuscularly or through-and-through),of leak prevention [110]. Gentileschi et al. remarked that and 57.0% used a buttressing product (21.0% BPS; 42.0%reinforcement of the staple-line significantly increases APM; 33.0% Duet [discontinued in 2012]). Mean bougiehemostasis, possibly strengthens the staple line, and may size was 36 Ϯ 4.8Fr (median 34.5Fr, range 32–60Fr).reduce leak, although there is no conclusive evidence tosupport reinforcement in LSG. In practice guidelines developed through consensus voting and published recently in the International Sleeve Three SRs published by Parikh et al. [1], Aurora et al. Gastrectomy Expert Panel Consensus Statement [113],[19], and Choi et al. [111] in 2012 focused on staple-line recommendations were derived from data and knowledgeleak in LSG. Parikh et al. reported results of a SR that based on a collective experience of 12,799 LSG cases inincluded 112 studies and employed a general estimating which the leak rate was 1.1%. With respect to reinforcingequation to analyze the leak odds ratio (OR) relative to the staple line, the Expert Panel achieved consensus onbougie size, distance of the transection from the pylorus,

8 M. Gagner and J. N. Buchwald / Surgery for Obesity and Related Diseases ] (2014) 00–00most topics, notably, that staple-line reinforcement reduces Conclusionbleeding along the staple line, and that either buttressing oroversewing the staple line are valid options. No consensus Short- and mid-term evidence on the effectiveness ofwas reached on whether reinforcement reduces leak rate or LSG is promising. LSG is a multipurpose operation that hasshould be routinely performed; whether the smaller bougie the advantage of convertibility to alternative surgicalsizes create smaller sleeves and a greater incidence of leaks; solutions. The procedure’s safety overall is good; develop-and whether the gastric transection should begin 4–6 cm ing proven surgical strategies to minimize the risk offrom the pylorus. potentially life-threatening staple-line leak after LSG is of critical importance. In a 2009 review of reinforcement of the staple line inLSG, Chen et al. suggested that a sample size of nearly The current SR is aligned with findings of prior SRs, SR/10,000 LSG procedures would be required to detect a MAs, RTs, RCTs, and international consensus and expertstatistically significant difference between the typically low panel reports that suggest the need for longer-term (Z5-leak rates associated with reinforcement options [91]. The year) data; standardization of the LSG procedure and LSGcurrent SR sample (n ¼ 8,920) approaches this volume; reporting; and studies focused on demonstration of thehowever, the current finding that use of APM staple-line value of staple-line reinforcement options versus no rein-reinforcement results in a significantly lower leak rate forcement. The current systematic review of 88 studiescompared to 2 other reinforcement options and to no representing 8,920 patients found that APM staple-linereinforcement is contradictory to the findings of 5 of the reinforcement was associated with a significantly lower6 aforementioned RTs and SRs, but is in agreement with leak rate than oversewing, BPS reinforcement, and noChoi et al, who found that reinforcement had a statistically reinforcement.significant effect on leak rate reduction. These contradictoryfindings may be the result of confounding by factors that Disclosuresrelate significantly to leak and/or to factors inherent in thedesign or conduct of the studies. If further study confirms This work was supported by W. L. Gore & Associates.definitively that there is no difference between APM (oranother reinforcement option) and no reinforcement, it may Referencesalso be the case that reinforcement is a superior choice inLSG for certain patients. The international LSG Consensus [1] Parikh M, Issa R, McCrillis A, Saunders JK, Ude-Welcome A,Groups’ and Expert Panels reports that reinforcement of the Gagner M. Surgical strategies that may decrease leak after laparo-LSG staple line is not only a valid option, but the method scopic sleeve gastrectomy: a systematic review and meta-analysis ofpracticed by the majority of surgeons; reinforcement may 9991 cases. Ann Surg 2013;257:231–7.have been historically aimed at leak prevention but mayprove to more definitively minimize hemorrhage. [2] Brethauer S, Hammel J, Schauer P. Systematic review of sleeve gastrectomy as staging and primary bariatric procedure. Surg ObesLimitations Relat Dis 2009;5:469–75. The analytic power of the current SR was limited by the [3] Himpens J, Dobbeleir J, Peeters G. Long-term results of laparo-paucity of experimental studies (RCTs) and the lack of scopic sleeve gastrectomy for obesity. Ann Surg 2010;252:319–24.reporting standards in the LSG literature. Variables essentialto accurately and completely report LSG procedures, [4] Todkar JS, Shah SS, Shah PS, Gangwani J. Long-term effects ofparticularly details surrounding the staple-line reinforce- laparoscopic sleeve gastrectomy in morbidly obese subjects withment method selected, distinctions between oversewing type 2 diabetes mellitus. Surg Obes Relat Dis 2010;6:142–5.techniques, bougie size, distance of the transection linefrom the pylorus, and leak data, were absent from a marked [5] Sarela AI, Dexter SP, O'Kane M, Menon A, McMahon MJ. Long-number of studies identified, limiting the sample of evalu- term follow-up after laparoscopic sleeve gastrectomy: 8–9-yearable records and the power of review results. Future studies results. Surg Obes Relat Dis 2012;8:679–84.that employ a multivariate analysis controlling for potentialconfounding variables would further our understanding of [6] Marceau P, Biron S, Bourque RA, Potvin M, Hould FS, Simard S.the differences between reinforcement materials. Achieving Biliopancreatic diversion with a new type of gastrectomy. Obes Surgconsensus regarding optimal surgical technique requires 1993;3:29–35.completeness and clarity of reported variables as summar-ized by SRs and MAs; these designs require the capacity to [7] Scopinaro N, Gianetta E, Civalleri D, et al. Biliopancreatic bypasspool standardized data from many studies; when not for obesity: initial experience in man. Br J Surg 1979;66:618–20.available, the opportunity to provide guidance for safer,more effective LSG procedures is reduced. [8] DeMeester TR, Fuchs KH, Ball CS, et al. Experimental and clinical results with proximal end-to-end duodenjejunostomy for pathologic duodenogastric reflux. Ann Surg 1987;206:414–24. [9] Ren CJ, Patterson E, Gagner M. Early results of laparoscopic biliopancreatic diversion with duodenal switch: a case series of 40 consecutive patients. Obes Surg 2000;10:514–23, discussion 524. [10] Gagner M, Patterson E. Laparoscopic biliopancreatic diversion with duodenal switch. Dig Surg 2000;17:547–66. [11] Regan JP, Inabnet WB, Gagner M. Early experience with two-stage laparoscopic Roux-en-Y gastric bypass as an alternative in the super- super obese patient. Obes Surg 2003;13:861–4. [12] Gumbs AA, Gagner M, Dakin G, Pomp A. Sleeve gastrectomy for morbid obesity. Obes Surg 2007;17:962–9.

LSG Leak Rates in 4 Reinforcement Options / Surgery for Obesity and Related Diseases ] (2014) 00–00 9[13] Marceau P, Cabanac M, Frankham PC, et al. Accelerated satiation [34] Mognol P, Chosidow D, Marmuse J-P. Laparoscopic sleeve after duodenal switch. Surg Obes Relat Dis 2005;1:408–12. gastrectomy as an initial bariatric operation for high-risk patients: initial results in 10 patients. Obes Surg 2005;15:1030–3.[14] Serra C, Pérez N, Bou R, et al. Laparoscopic sleeve gastrectomy. A bariatric procedure with multiple indications. Cir Esp 2006;79: [35] Kasalicky M, Michalsky D, Housova J, et al. Laparoscopic sleeve 289–92. gastrectomy without an over-sewing of the staple line. Obes Surg 2008;18:1257–62.[15] Buchwald H, Oien D. Metabolic/bariatric surgery worldwide 2011. Obes Surg 2013;23:427–36. [36] Hakeam H, O'Regan P, Salem A, Bamehriz F, Jomaa L. Inhibition of c-reactive protein in morbidly obese patients after laparoscopic[16] ASMBS Clinical Issues Committee. Updated position statement on sleeve gastrectomy. Obes Surg 2009;19:456–60. sleeve gastrectomy as a bariatric procedure. Received in revised form 14 March 2012 published online 13 February 2012. [37] Mui W, Ng E, Tsung B, Lam C, Yung M-Y. Laparoscopic sleeve gastrectomy in ethnic obese Chinese. Obes Surg 2008;18:1571–4.[17] Hutter MM, Schirmer BD, Jones DB, et al. First report from the American College of Surgeons Bariatric Surgery Center Network: [38] Gan S, Talbot M, Jorgensen J. Efficacy of surgery in the manage- laparoscopic sleeve gastrectomy has morbidity and effectiveness ment of obesity-related type 2 diabetes mellitus. ANZ J Surg positioned between the band and the bypass. Ann Surg 2011;254: 2007;77:958–62. 410–20, discussion 420–2. [39] Taller J, Wisbach G, Bertucci W. Single-incision laparoscopic sleeve[18] Himpens J, Dapri G, Cadiere G. A prospective randomized study gastrectomy for morbid obesity: video technique and review of first between laparoscopic gastric banding and laparoscopic isolated 10 patients. Surg Obes Relat Dis 2010;6:559–60. sleeve gastrectomy: results after 1 and 3 years. Obes Surg 2006;16: 1450–6. [40] Sabbagh C, Verhaeghe P, Dhahri A, et al. Two-year results on morbidity, weight loss and quality of life of sleeve gastrectomy as[19] Aurora AR, Khaitan L, Saber AA. Sleeve gastrectomy and the risk first procedure, sleeve gastrectomy after failure of gastric banding of leak: a systematic analysis of 4,888 patients. Surg Endosc and gastric banding. Obes Surg 2010;20:679–84. 2012;26:1509–15. [41] Nienhuijs SW, deZoete JP, Berende CAS, De Hingh IHJT, Smulders[20] Burgos AM, Braghetto I, Csendes A, et al. Gastric leak after JF. Evaluation of laparoscopic sleeve gastrectomy on weight loss laparoscopic-sleeve gastrectomy for obesity. Obes Surg 2009;19:1672–7. and co-morbidity. Int J Surg 2010;8:302–4.[21] Ferrer-Marquez M, Belda-Lozano R, Ferrer-Ayza M. Technical con- [42] Magee CJ, Barry J, Arumugasamy M, Javed S, Macadam R, troversies in laparoscopic sleeve gastrectomy. Obes Surg 2012;22:182–7. Kerrigan DD. Laparoscopic sleeve gastrectomy for high-risk patients: weight loss and comorbidity improvement—short-term[22] Moszkowicz D, Arienzo R, Khettab I, et al. Sleeve gastrectomy results. Obes Surg 2011;21:547–50. severe complications: is it always a reasonable surgical option? Obes Surg 2013 Feb 12 [Epub ahead of print]. [43] Goitein D, Goitein O, Feigin A, Zippel D, Papa M. Sleeve gastrectomy: radiologic patterns after surgery. Surg Endosc 2009;23:[23] Baker RS, Foote J, Kemmeter P, et al. The science of stapling and 1559–63. leaks. Obes Surg 2004;14:1290–8. [44] Lee W-J, Hur KY, Lakadawala M, Kasama K, Wong SKH, Lee Y-[24] Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group. C. Gastrointestinal metabolic surgery for the treatment of diabetic Preferred Reporting Items for Systematic Reviews and Meta- patients: a multi-institutional international study. J Gastrointest Surg Analyses: The PRISMA Statement. PLoS Med 2009;6:e1000097. 2012;16:45–52.[25] Consten ECJ, Gagner M, Pomp A, Inabnet WB. Decreased bleeding [45] Lee SY, Lim CH, Pasupathy S, et al. Laparoscopic sleeve after laparoscopic sleeve gastrectomy with or without duodenal gastrectomy: a novel procedure for weight loss. Singapore Med J switch for morbid obesity using a stapled buttressed absorbable 2001;52:794–8. polymer membrane. Obes Surg 2004;14:1360–6. [46] Simon T, Scott J, Brockmeyer J, et al. Comparison of staple-line[26] Casella G, Soricelli E, Rizzello M, et al. Nonsurgical treatment of leakage and hemorrhage in patients undergoing laparoscopic sleeve staple line leaks after laparoscopic sleeve gastrectomy. Obes Surg gastrectomy with or without Seamguard. Am Surg 2011;77:1665–8. 2009;19:821–6. [47] Musella M, Milone M, Bellini M, Leongito M, Guarino R, Milone[27] Dapri G, Cadière GB, Himpens J. Reinforcing the staple line during F. Laparoscopic sleeve gastrectomy. Do we need to oversew the laparoscopic sleeve gastrectomy: prospective randomized clinical study staple line? Ann Ital Chir 2011;82:273–7. comparing three different techniques. Obes Surg 2010;20:462–7. [48] Stamou KM, Menenakos E, Dardamanis D, et al. Prospective[28] Basso N, Casella G, Rizzello M, et al. Laparoscopic sleeve comparative study of the efficacy of staple-line reinforcement in gastrectomy as first stage or definitive intent in 300 consecutive laparoscopic sleeve gastrectomy. Surg Endosc 2011;25:356–60. cases. Surg Endosc 2011;25:444–9. [49] Berende CAS, de Zoete JP, Smulders JF, Nienhuijs SW. Laparo-[29] Ser K-H, Lee W-J, Lee Y-C, Chen J-C, Su Y-H, Chen S-C. scopic sleeve gastrectomy feasible for bariatric revision surgery. Experience in laparoscopic sleeve gastrectomy for morbidly obese Obes Surg 2012;22:330–4. Taiwanese: staple-line reinforcement is important for preventing leakage. Surg Endosc 2010;24:2253–9. [50] Mukherjee S, Devalia K, Rahman MG, Mannur KR. Sleeve gastrectomy as a bridge to a second bariatric procedure in super-[30] Bernante P, Foletto M, Busetto L, et al. Feasibility of laparoscopic obese patients-a single institution experience. Surg Obes Relat Dis sleeve gastrectomy as a revision procedure for prior laparoscopic 2012;8:140–4. gastric banding. Obes Surg 2006;16:1327–30. [51] Gadiot RP, Biter LU, Zengerink HJ, et al. Laparoscopic sleeve[31] Iannelli A, Schneck A, Noel P, Amor I, Krawezykoski D, gastrectomy with an extensive posterior mobilization: technique and Gugenheim J. Re-sleeve gastrectomy for failed laparoscopic sleeve preliminary results. Obes Surg 2012;22:320–9. gastrectomy: a feasibility study. Obes Surg 2011;21:832–5. [52] Topart P, Becouarn G, Ritz P. Comparative early outcomes of three[32] Han S, Kim W, Oh J. Results of laparoscopic sleeve gastrectomy laparoscopic bariatric procedures: sleeve gastrectomy, Roux-en-Y (LSG) at 1 year in morbidly obese Korean patients. Obes Surg gastric bypass, and biliopancreatic diversion with duodenal switch. 2005;15:1469–75. Surg Obes Relat Dis 2012;8:250–4.[33] Fuks D, Verhaeghe P, Brehant O, et al. Results of laparoscopic [53] Lakdawala MA, Muda NH, Goel S, Bhasker A. Single-incision sleeve gastrecctomy: a prospective study in 135 patients with morbid sleeve gastrectomy versus conventional laparoscopic sleeve gastrec- obesity. Surgery 2009;145:106–13. tomy—a randomised pilot study. Obes Surg 2011;21:1664–70.

10 M. Gagner and J. N. Buchwald / Surgery for Obesity and Related Diseases ] (2014) 00–00[54] Gagniere J, Slim K, Launay-Savary MV, Raspado O, Flamein R, [74] Uglioni B, Wolnerhanssen B, Peters T, Christoffel-Courtin C, Kern Chipponi J. Previous gastric banding increases morbidity and gastric B, Peterli R. Midterm results of primary vs. secondary laparoscopic leaks after laparoscopic sleeve gastrectomy for obesity. J Visc Surg sleeve gastrectomy (LSG) as an isolated operation. Obes Surg 2011;148:e205–9. 2009;19:401–6.[55] Goitein D, Feigin A, Segal-Lieberman G, Goitein O, Papa MZ, [75] Skrekas G, Lapatsanis D, Stafyla V, Papalambros A. One year after Zippel D. Laparoscopic sleeve gastrectomy as a revisional option laparoscopic “tight” sleeve gastrectomy: technique and outcome. after gastric band failure. Surg Endosc 2011;25:2626–30. Obes Surg 2008;18:810–3.[56] Bellanger DE, Greenway FL. Laparoscopic sleeve gastrectomy, 529 [76] Dapri G, Vaz C, Cadiere G, Himpens J. A prospective randomized cases without a leak: short-term results and technical considerations. study comparing two different techniques for laparoscopic sleeve Obes Surg 2011;21:146–50. gastrectomy. Obes Surg 2007;17:1435–41.[57] Daskalakis M, Berdan Y, Theodoridou S, Weigand G, Weiner R. [77] Tucker O, Szomstein S, Rosenthal R. Indications for sleeve Impact of surgeon experience and buttress material on postoperative gastrectomy as a primary procedure for weight loss in the morbidly complications after laparoscopic sleeve gastrectomy. Surg Endosc obese. J Gastrointest Surg 2008;12:662–7. 2011;25:88–97. [78] Srinivasa S, Hill L, Sammour T, Hill A, Babor R, Rahman H. Early[58] Albanopoulos K, Alevizos L, Flessas J, et al. Reinforcing the staple and mid-term outcomes of single-stage laparoscopic sleeve gastrec- line during laparoscopic sleeve gastrectomy: prospective randomized tomy. Obes Surg 2010;20:1484–90. clinical study comparing two different techniques. Preliminary results. Obes Surg 2012;22:42–6. [79] Foletto M, Prevedello L, Bernante P, et al. Sleeve gastrectomy as revisional procedure for failed gastric banding or gastroplasty. Surg[59] Jurowich C, Thalheimer A, Seyfried F, et al. Gastric leakage after Obes Relat Dis 2010;6:146–51. sleeve gastrectomy-clinical presentation and therapeutic options. Langenbecks Arch Surg 2011;396:981–7. [80] Lakdawala MA, Bhasker A, Mulchandani D, Goel S, Jain S. Comparison between the results of laparoscopic sleeve gastrectomy[60] Serra C, Baltasar A, Andreo L, et al. Treatment of gastric leaks with and laparoscopic Roux-en-Y gastric bypass in the Indian population: coated self-expanding stents after sleeve gastrectomy. Obes Surg a retrospective 1 year study. Obes Surg 2010;20:1–6. 2007;17:866–72. [81] Acholonu E, McBean E, Court I, Bellorin O, Szomstein S, Rosenthal[61] Rubin M, Yehoshua R, Stein M, et al. Laparoscopic sleeve RJ. Safety and short-term outcomes of laparoscopic sleeve gastrec- gastrectomy with minimal morbidity early results in 120 morbidly tomy as a revisional approach for failed laparoscopic adjustable obese patients. Obes Surg 2008;18:1567–70. gastric banding in the treatment of morbid obesity. Obes Surg 2009;19:1612–6.[62] Lalor P, Tucker O, Szomstein S, Rosenthal R. Complications after laparoscopic sleeve gastrectomy. Surg Obes Relat Dis 2008;4: [82] Arias E, Martinez PR, Ka Ming Li V, Szomstein S, Rosenthal RJ. 33–8. Mid-term follow-up after sleeve gastrectomy as a final approach for morbid obesity. Obes Surg 2009;19:544–8.[63] Roa P, Kaidar-Person O, Pinto D, Cho M, Szomstein S, Rosenthal R. Laparoscopic sleeve gastrectomy as treatment for morbid obesity: [83] Kakoulidis TP, Karringer A, Gloaguen T, Arvidsson D. Initial technique and short-term outcome. Obes Surg 2006;16:1323–6. results with sleeve gastrectomy for patients with class I obesity (BMI 30–35 kg/m2). Surg Obes Relat Dis 2009;5:425–8.[64] Merchant A, Cook M, Srinivasan J, Davis S, Sweeney J, Lin E. Comparison between laparoscopic paraesophageal hernia repair with [84] Lombardo V, Baratta R, Giannone G. Laparoscopic sleeve gastrec- sleeve gastrectomy and paraesophageal hernia repair alone in tomy for morbid obesity: our initial experience. Ann Ital Chir morbidly obese patients. Am Surg 2009;75:620–5. 2010;81:17–21.[65] Himpens J, Schepper M, Dapri G. Laparoscopic conversion of [85] Boza C, Salinas J, Salgado N, et al. Laparoscopic sleeve gastrectomy adjustable gastric banding to sleeve gastrectomy: a feasibility study. as a stand-alone procedure for morbid obesity: report of 1,000 cases Surg Laparosc Endosc Percutan Tech 2010;20:162–5. and 3-year follow-up. Obes Surg 2012;22:866–71.[66] Felberbauer F, Langer F, Shakeri-Manesch S, et al. Laparoscopic [86] Alevizos L, Lirici MM. Laparo-endoscopic single-site sleeve gas- sleeve gastrectomy as an isolated bariatric procedure: intermediate- trectomy: results from a preliminary series of selected patients. term results from a larger series in three Austrian centers. Obes Surg Minim Invasive Ther Allied Technol 2012;21:40–5. 2008;18:814–8. [87] D’Hondt M, Vanneste S, Pottel H, et al. Laparoscopic sleeve[67] Braghetto I, Korn O, Valladares H, et al. Laparoscopic sleeve gastrectomy as a single-stage procedure for the treatment of morbid gastrectomy: surgical technique, indications, and clinical results. obesity and the resulting quality of life, resolution of comorbidities, Obes Surg 2007;17:1442–50. food tolerance, and 6-year weight loss. Surg Endosc 2011;25: 2498–504.[68] Leyba J, Aulestia S, Llopis S. Laparoscopic Roux-en-Y gastric bypass versus laparoscopic sleeve gastrectomy for the treatment of [88] Behrens C, Tang BQ, Amson BJ. Early results of a Canadian morbid obesity. A prospective study of 117 patients. Obes Surg laparoscopic sleeve gastrectomy experience. Can J Surg 2011;54: 2011;21:212–6. 138–43.[69] Nocca D, Krawczykowsky D, Bomans B, et al. A prospective [89] Lacy A, Obarzabal A, Pando E, et al. Revisional surgery after sleeve multicenter study of 163 sleeve gastrectomies: results at 1 and 2 gastrectomy. Surg Laparosc Endosc Percutan Tech 2010;20:351–6. years. Obes Surg 2008;18:560–5. [90] Slater BJ, Bellatorre N, Eisenberg D. Early postoperative outcomes[70] Dapri G, Cadière GB, Himpens J. Feasibility and technique of and medication cost savings after laparoscopic sleeve gastrectomy in laparoscopic conversion of adjustable gastric banding to sleeve morbidly obese patients with type 2 diabetes. J Obes 2011;2011: gastrectomy. Surg Obes Relat Dis 2009;5:72–6. 350–523.[71] Langer F, Reza Hoda M, Bohdjalian A, et al. Sleeve gastrectomy [91] Chen B, Kiriakopoulos A, Tsakayannis D, Wachtel MS, Linos D, and gastric banding: effects on plasma ghrelin levels. Obes Surg Frezza EE. Reinforcement does not necessarily reduce the rate of 2005;15:1024–9. staple line leaks after sleeve gastrectomy. A review of the literature and clinical experiences. Obes Surg 2009;19:166–72.[72] Melissas J, Koukouraki S, Askoxylakis J, et al. Sleeve gastrectomy —a restrictive procedure? Obes Surg 2007;17:57–62. [92] Kehagias I, Spryopoulos C, Karamanakos S, Kalfarentzos F. Efficacy of sleeve gastrectomy as sole procedure in patients with[73] Tagaya N, Kasama K, Kikkawa R, et al. Experience with laparo- severe obesity (BMI o50 kg/m2). Surg Obes Relat Dis 2012 Jan 13 scopic sleeve gastrectomy for morbid versus super morbid obesity. [Epub ahead of print]. Obes Surg 2009;19:1371–6.

LSG Leak Rates in 4 Reinforcement Options / Surgery for Obesity and Related Diseases ] (2014) 00–00 11 [93] Angrisani L, Cutolo PP, Buchwald JN, et al. Laparoscopic rein- [106] Chopra A, Chao E, Etkin Y, Merklinger L, Lieb J, Delany H. forced sleeve gastrectomy: early results and complications. Obes Laparoscopic sleeve gastrectomy for obesity: can it be considered a Surg 2011;21:783–93. definitive procedure? Surg Endosc 2012;26:831–7. [94] Saber A, Elgamal M, Itawi E, Rao A. Single incision laparoscopic [107] Nath A, Leblanc KA, Hausmann MG, Kleinpeter K, Allain BW, sleeve gastrectomy (SILS). a novel technique. Obes Surg 2008;18: Romero R. Laparoscopic sleeve gastrectomy: our first 100 patients. 1338–42. JSLS 2010;14:502–8. [95] Moy J, Pomp A, Dakin G, Parikh M, Gagner M. How I do it. [108] Ayloo S, Buchs NC, Addeo P, Bianco FM, Giulianotti PC. Robot- Laparoscopic sleeve gastrectomy for morbid obesity. Am J Surg assisted sleeve gastrectomy for super-morbidly obese patients. J 2008;196:e56–9. Laparoendosc Adv Surg Tech A 2011;21:295–9. [96] Zhang F, Strain G, Lei W, Dakin G, Gagner M, Pomp A. Changes in [109] Gluck B, Movitz B, Jansma S, Gluck J, Laskowski K. Laparoscopic lipid profiles in morbidly obese patients after laparoscopic sleeve sleeve gastrectomy is a safe and effective bariatric procedure for the gastrectomy (LSG). Obes Surg 2011;21:305–9. lower BMI (35.0–43.0 kg/m2) population. Obes Surg 2011;21:1168–71. [97] Lewis CE, Dhanasopon A, Dutson EP, Mehran A. Early experience [110] Gentileschi P, Camperchioli I, D’Ugo S, Benavoli D, Gaspari AL. with laparoscopic sleeve gastrectomy as a single-stage bariatric Staple-line reinforcement during laparoscopic sleeve gastrectomy procedure. Am Surg 2009;75:945–9. using three different techniques: a randomized trial. Surg Endosc 2012;26:2623–9. [98] Jacobs M, Bisland W, Gomez E, et al. Laparoscopic sleeve gastrectomy: a retrospective review of 1- and 2-year results. Surg [111] Choi YY, Bae J, Hur KY, Choi D, Kim YJ. Reinforcing the staple Endosc 2010;24:781–5. line during laparoscopic sleeve gastrectomy: does it have advan- tages? A meta-analysis. Obes Surg 2012;22:1206–13. [99] Kockerling F, Schug-Pass C. Gastroscopically controlled laparo- scopic sleeve gastrectomy. Obes Facts 2009;2(Suppl 1):15–8. [112] Deitel M, Gagner M, Erickson AL, Crosby RD. Third international summit: current status of sleeve gastrectomy. Surg Obes Relat Dis[100] Diamantis T, Alexandrou A, Nikiteas N, Giannopoulos A, Papal- 2011;7:749–59. ambros E. Initial experience with robotic sleeve gastrectomy for morbid obesity. Obes Surg 2011;21:1172–9. [113] Rosenthal RJ. International Sleeve Gastrectomy Expert Panel, et al International sleeve gastrectomy expert panel consensus statement:[101] Diamantis T, Alexandrou A, Pikoulis E, et al. Laparoscopic sleeve best practice guidelines based on experience of 412,000 cases. Surg gastrectomy for morbid obesity with intra-operative endoscopic Obes Relat Dis 2012;8:8–19. guidance. Immediate peri-operative and 1-year results after 25 patients. Obes Surg 2010;20:1164–70. [114] Gagner M. Leaks after sleeve gastrectomy are associated with smaller bougies: prevention and treatment strategies. Surg Laparosc[102] Gentileschi P, Camperchioli I, Benavoli D, Di Lorenzo N, Sica G, Endosc Percutan Tech 2010;20:166–9. Gaspari AL. Laparoscopic single-port sleeve gastrectomy for morbid obesity: preliminary series. Surg Obes Relat Dis 2010;6: [115] Parikh M, Gagner M, Heacock L, et al. Laparoscopic sleeve 665–9. gastrectomy: does bougie size affect mean %EWL? Short-term outcomes. Surg Obes Relat Dis 2008;4:528–33.[103] Saber AA, El-Ghazaly TH, Eliam A. Single-incision transumbilical laparoscopic sleeve gastrectomy. J Laparoendosc Adv Surg Tech A [116] Deitel M, Crosby RD, Gagner M. The first international consensus 2009;19:755–9. summit for sleeve gastrectomy (SG), New York City, October 25– 27, 2007. Obes Surg 2008;18:487–6.[104] Chowbey PK, Dhawan K, Khullar R, et al. Laparoscopic sleeve gastrectomy: An Indian experience—surgical technique and early [117] Gagner M, Deitel M, Kalberer TL, Erickson AL, Crosby RD. The results. Obes Surg 2010;20:1340–7. second international consensus summit for sleeve gastrectomy, March 19–21, 2009. Surg Obes Relat Dis 2009;5:476–85.[105] Nguyen NT, Smith BR, Reavis KM, Nguyen X-M T, Nguyen B, Stamos MJ. Strategic laparoscopic surgery for improved cosmesis in [118] Oxford Centre for Evidence-Based Medicine. Available from: http:// general and bariatric surgery: analysis of initial 127 cases. J www.cebm.net/index.aspx?o = 5653 (accessed March 30, 2013). Laparoendosc Adv Surg Tech A 2012;22:355–61.

CLINICAL SCIENCE 405Croat Med J. 2014;55:405-15 Goran Ribaric1, Janedoi: 10.3325/cmj.2014.55.405 Buchwald2Gastric band is safe and 1European Surgical Institute,effective at three years in a Ethicon Endo-Surgery (Europe),national study subgroup of Hamburg, Germanynon-morbidly obese patients 2Division of Scientific Writing, Medwrite MedicalAim To analyze the 3-year outcomes of lower body mass Communications, Maiden Rock,index (BMI) (<35 kg/m2) adjustable gastric band (AGB) re- WI, USAcipients across multiple sites in the French health insur-ance system. Received: April 30, 2013 Accepted: May 11, 2014Methods From prospectively collected data on a co- Correspondence to:hort of 517 morbidly obese Swedish Adjustable Gas- Goran Ribarictric Band® (SAGB) patients (Clinical Trials Web database, Director, Regional Safety Officer#NCT01183975), a retrospective analysis of a subgroup of EMEA29 low-BMI patients was conducted. Patients had a severe Medical Devices & Diagnosticsobesity-related comorbidity, had undergone a prior bariat- Johnson & Johnsonric procedure requiring reintervention, or had a maximum Ethicon Endo Surgery (Europe)adult BMI≥40. Safety (mortality, adverse events) and ef- GmbHfectiveness (BMI change, excess weight loss [EWL, %], to- Hummelsbütteler Steindamm 71tal body weight loss [%TBWL], quality of life [QoL], and co- 22851 Norderstedt, Germanymorbidities) were evaluated. [email protected] Multiple surgical teams/sites enrolled patients www.cmj.hrand performed SAGB procedures between September 2,2007 and April 30, 2008. Of 29 low-BMI patients (mean age,41.3 ± 10.3 years), 89.7% were female, and obesity durationwas 13.6 ± 7.3 years. Mean BMI was 31.5 ± 3.7; there were37 comorbidities in 15/29 patients. At 3-year follow-up,BMI was 29.4 ± 4.9 (mean change, -2.3 ± 6.2; P = 0.069); to-tal cohort EWL, 7.3 ± 74.8%; TBWL, 6.2 ± 18.8%; BMI≥30 to<35 EWL, 38.8 ± 48.0%; there were 7 comorbidities in 15/29patients (P < 0.031). There were 20 adverse events in 13 pa-tients (44.8%); SAGBs were retained in 25/29 (86.2%) at 3years.Conclusions In a retrospective analysis of a subgroupof BMI<35 kg/m2 patients, some following a prior bariat-ric procedure, SAGB was found to be safe and effective at3-year follow-up.

406 CLINICAL SCIENCE Croat Med J. 2014;55:405-15For over 2 decades, since publication of the 1991 National gery (Europe) GmbH sponsored and performed a country-Institutes of Health (NIH) consensus conference statement wide health insurance study to assess reimbursement of(1), the cutoff point for bariatric surgery has been morbid the SAGB product in France. The study was registered (Clin-obesity (body mass index [BMI, kg/m2]≥40 or ≥35 with co- ical Trials Web database, #NCT01183975) (14) and a spon-morbidities), also termed class II obesity by the World Health sor-developed protocol and case report form developedOrganization (WHO) (2). This demarcation of access to bar- to direct implementation of HAS requirements and goodiatric surgery was based on the observation that an increase clinical practices (GCPs) (ie, patient welfare in study design,in BMI leads to an increase in the risk of comorbid illness and ethical study conduct), defined by ISO EN 14155-1 and -2premature death. Yet, investigation of the potential value of (15,16). Ethical approval and protocol approval were giv-bariatric surgery as a safe and effective treatment for over- en by HAS, the Commission Nationale de l’Informatique etweight (BMI 25-<30) and obesity class I (BMI≥30 to <35) pa- des Libertés, and the Comité Consultatif sur le Traitementtients has been under way since the publication of the NIH de l’Information en matičre de Recherche dans le domainestatement. In 1992 and 1995 landmark studies (3,4), Pories et de la Santé. Treatment payments were covered by nationalal theorized that bariatric procedures might be safe and as health insurance (13).beneficial for weight loss and comorbidity reduction in non-morbidly obese patients as it was in the morbidly obese (5). A contract research organization, Medextens SARL, Paris,In the last half decade, the least-invasive, lowest-risk restric- France, and an independent monitoring committee con-tive procedures, such as adjustable gastric banding (AGB), sisting of a non-participating bariatric surgeon, a pharma-have been employed at the forefront of exploring surgical cologist, and a medical nutritionist supervised the study’soptions for the <35 BMI patient. progress and prepared an interim report for review by HAS and the sponsor. Patients were required to provide writtenAdjustable gastric banding comprised the vast majority, informed consent before surgery per Declaration of Hel-nearly 90%, of bariatric procedures performed in morbidly sinki (17) and GCP guidelines.obese patients in France prior to 2008 (6). To assess the na-tional social insurance-supported use of the Swedish Ad- Design and settingjustable Gastric Band (SAGB) (7,8), the French governmentcommissioned a prospective, 31-center, “real-life,” obser- The prospective, multicenter, noncomparative study de-vation of SAGB safety and effectiveness in class II and III sign aimed to facilitate observation and reporting of out-obese patients (9). Between September 2, 2007 and April comes in a morbidly obese study cohort, of which the30, 2008, patients were selected and underwent SAGB im- current low-BMI cohort was a subgroup (9). Primary HASplantation in rural and urban centers. SAGB weight-loss objectives were to asses SAGB safety (mortality, adverseeffectiveness analyzed on an intent-to-treat basis at the event [AE] occurrence) and clinical effectiveness (changes3-year study endpoint was comparable to that of AGB find- in weight loss, quality of life [QoL], comorbid illness) in var-ings summarized by global meta-analyses (10,11). Under ied French hospital settings.the “real-world” SAGB study protocol requirement of con-secutive recruitment and surgeon discretion, 29 patients In order to incorporate “real-life” practice experiences(5.6% of 517) were included in the national SAGB study across geographically diverse regions of France, surgeonswho presented with a BMI<35 and a severe obesity-related were selected from academic, private, and public institu-comorbidity, and/or had experienced a prior complicated tions with differing bariatric surgery volumes. Per GCPbariatric surgery requiring revision, and/or had previously standards, surgeons were required to undergo training insustained a maximum adult BMI≥40. With the aim of con- the protocol; selected surgeons recruited SAGB patientstributing safety and effectiveness findings to the growing consecutively.<35 BMI evidence base, we report 3-year outcomes for theFrench low-BMI SAGB study group. Inclusion criteriaMethods Study eligibility for the primary trial was based on a recruit- ment goal of >500 patients with <20% loss to follow-upStudy protocol after 3 years. General inclusion criteria stipulated patients with morbid obesity after failed medical treatment and During 2007, the French Health Technology Assessment no contraindications in accord with French (12), Europe- Body (HAS) (12,13) requested that Ethicon Endo-Sur-www.cmj.hr

Ribaric and Buchwald: Three-year national band outcomes in non-morbidly obese 407an (18), and American NIH bariatric surgery guidelines (1). tional study design. Baseline characteristics (eg, gender,French residents with a BMI<35 were permitted inclusion age) were collected; weight, obesity-related comorbidin the consecutive SAGB study enrollment if they had an disease, and QoL were recorded on the day of surgery,adult maximum BMI≥40, and/or had a severe obesity-re- and at 1, 3, 6, 12, 18, 24, and 36 months postoperatively.lated comorbidity (thus, were receiving SAGB as a primary Comorbidity data were sought via questionnaire; diagno-intervention in the current study [an “index SAGB”]), or if ses were established and recorded consonant with indi-they required reintervention following a complicated prior vidual investigators’ typical practice via the password-pro-bariatric procedure [a “PBP+SAGB”] (SAGB as a secondary tected Medextens-Medalliance eCRF Manager (v.1.3) webintervention). database.Variables For the current low-BMI subgroup study, data were re- trieved from the original HAS archive and sorted by scriptSafety variables analyzed for the low-BMI cohort were mor- for the known 29 target patients. Coded variables that ad-tality and frequency of AEs. Effectiveness variables were dressed identified study topics were chosen and manuallyevaluated as change over 3 years in absolute weight (AW); exported to a dedicated SPSS database.BMI; excess body weight (EW); percentage EW loss (%EWL),ie, baseline AW – follow-up AW/EW, calculated by Miller’s Techniqueformulas (19-23) for identifying ideal weight, correspond-ing to the midpoint value of the medium-frame range on SAGB procedures were performed via pars flaccida tech-the Metropolitan Life Insurance Height and Weight Tables nique (31), and band adjustments were accomplished at×100; and percentage total body weight loss (%TBWL), the discretion of the surgeons. Three SAGB model options(ie, baseline AW – follow-up AW/baseline AW × 100) (24). were available: 2200-X Quick-Close; 2100-X (with lockingHealth-related QoL and changes in comorbid illness were ring and injection port); and the BD2XV Quick-Close withalso analyzed. VelocityTM injection port.Quality of life instruments Statistical analysisThe generic EuroQoL 5-Dimensions (EQ-5D), a psychomet- Statistical analyses were performed using the SPSS® soft-ric instrument valued for its utility in calculating quality ad- ware package (ver. 20, IBM SPSS, Chicago, IL, USA). Quan-justed life years as well as the relative cost-effectiveness of titative demographic variables were generally reported asobesity interventions, such as AGB (25), was used as a mea- median and interquartile range (IQR); qualitative variablessure of QoL. The EQ-5D is a health-related QoL evaluation (demographic and outcome) were reported as numberwith 5 items and a visual analogue scale (EQ-VAS) (26-29) and percentage. Adverse events were also reported asthat provides a 5-dimensional profile: mobility, usual activi- number and percentage. Quantitative measures of changeties, self-care, anxiety/depression, and pain/discomfort. Di- from baseline at 3 years were analyzed using the related-mensions are presented as 1 item with 3 response options: samples Wilcoxon signed rank test; between-group com-severe problems, some problems, and no problems. Item parisons were made with the Mann-Whitney U test. Theresponses can be weighted normatively to derive a utility Fisher exact test was used to investigate relationships be-score (range -0.594 to 1, where 1 = ultimate health). A clini- tween qualitative variables. Multivariate modeling, linearcally important difference has been identified at ≥0.07 on regression, and logistic regression were used to explorethe EQ-5D scale (29). The EQ-VAS module is a single-item relationships between patient characteristics, weight loss,global QoL evaluation in which patients rate their current and QoL. Alpha was set at P < 0.05.health (scale from 0 = worst imaginable to 100 = best imag-inable) (30). ResultsData collection Screening and enrollment of patients occurred between September 2, 2007 and April 30, 2008. The last follow-upProtocol-prescribed safety and weight data collection and visit at 3 years, due on April 30, 2011, was extended toassessment measures were the only standardized require- November 20, 2011 to accommodate patients’ sched-ments for the surgical centers, per the “real-life” observa- ules. All low-BMI cases were treated laparoscopical- www.cmj.hr

408 CLINICAL SCIENCE Croat Med J. 2014;55:405-15ly using pars flaccida technique and port fixation with no patients indicated that they had undergone prior AGB,conversions to laparotomy. while 1 reported prior sleeve gastrectomy, and 1, prior gastric balloon; 15/17 reported having had good resultsBaseline patient characteristics (ie, weight loss and comorbidity reduction) before expe- riencing poor weight loss and a variety of complicationsThe SAGB BMI<35 sample consisted of 89.7% (N = 26) fe- and subsequently selecting SAGB as their reinterventionmale and 10.3% (N = 3) male patients with a median age of treatment. This subjective reporting was corroborated38.8 years, obesity duration of 12.0 years, AW of 87.0, EW of by baseline data analysis that indicated that PBP+SAGB28.5, and median BMI of 33.1 (Table 1). Nine (31%) patients patients, compared to index SAGB patients, had a signifi-had a baseline BMI<30 and 20 (69%) had a BMI≥30 to <35. cantly lower median (IQR) number of comorbidities (0.0 [0.0-0.5] vs 2.0 [1.3-4.0]; P = 0.001), significantly higherFifty-two percent (15/29) of patients presented with at global QoL [EQ-VAS] (70.0 [50.0-80.0] vs 40.0 [25.0-55.0];least one comorbidity. Median EQ-5D was 0.7 and me- P = 0.003), and significantly lower median BMI (30.1 [27.7-dian VAS, 50.0. A history of family obesity was reported in 33.2] vs 34.6 [34.0-34.8]; P = 0.002). Indeed, 8/9 BMI<3019 patients (65.5%). SAGB was the first bariatric surgery SAGB patients (89%) were PBP+SAGB patients; where-in 12 patients (41.4%), referred to subsequently as index as 45% (9/20) of BMI≥30 to <35 SAGB patients wereSAGB patients, and a reintervention following a prior bar- PBP+SAGB patients. With respect to results and inter-iatric procedure that involved serious complications in pretation presented herein, BMI<35 PBP+SAGB patients17 (58.6% PGP+SAGB patients). Fifteen of 17 PBP+SAGB were, largely, former class-III (≥40 kg/m2) morbidly obese patients. Maximum adult BMI for PBP+SAGB patients wasTable 1. Preoperative patient characteristics* significantly greater than that for index SAGB patients (40.4 [38.1-42.6] vs 35.8 [34.2-37.5]; P = 0.001).Characteristic Median (IQR), N = 29 Adverse eventsGender: Fifty-five percent (16/29) of SAGB patients presenting withMale, N (%)   3 (10.3) a BMI<35 experienced no adverse events (AEs) over 3-year follow-up. There was an 86.2% overall rate of band survival,Female, N (%) 26 (89.7) that is, bands that remained implanted. There were 20 con- firmed AEs in 13 patients (44.8%): 1 in 7 patients (24.1%);Age (yrs) 38.8 (33.9-50.4) 2 in 5 patients (17.2%); and 3 in 1 patient (3.5%). An over- all rate of 0.23 confirmed adverse events per patient-yearDuration of obesity (yrs) 12.0 (10.0-17.5) was observed. Confirmed AEs in order of frequency were: band removal 4 (14%), port rotation 3 (10.3%), band slip-Height (m)   1.7 (1.6-1.7) page 2 (7%), esophageal dilation 2 (7%), food intolerance 2 (7%), abdominoplasty 2 (7%), dysphagia 1 (3.5%), GERDAW (kg) 87.0 (76.0-94.5) 1 (3.5%), port malposition 1 (3.5%), port reintervention (no removal) 1 (3.5%), and port dysfunction/removal 1 (3.5%).Ideal body weight (kg)† 60.9 (58.2-63.3) PBP+SAGB patients had a significantly higher median number of AEs than index SAGB patients (1.0 [0.0-2.0] vsEW (kg) 28.5 (16.3-33.0) 0.0 [0.0-0.0]; P = 0.030). In fact, 17/20 (85%) confirmed AEs, and all 4 confirmed band removals (ablations), occurred inBMI (kg/m2): 33.1 (28.8-34.6) the PBP+SAGB group. Conversely, 83.3% (10/12) of index- SAGB patients experienced no AE.<30, N (%)   9 (31.0) Weight loss≥30 and <35, N (%) 20 (69.0) Three-year postoperative weight outcomes for BMI<35Intervention: SAGB patients were available in 86.2% (25/29) of patients. Median AW was 80.0 (72.0-88.0) compared to 87.0 (76.0-PBP + SAGB‡, N (%) 17 (58.6)Index SAGB, N (%) 12 (41.4)At least 1 comorbidity, N (%) 15 (51.7)History of family obesity, N (%) 19 (65.5)EQ-5D   0.7 (0.3-0.8)EQ-VAS 50.0 (40.0-74.0)*Abbreviations: IQR – interquartile range; BMI – body mass index; AW– absolute weight; EW – excess weight; PBP – Prior bariatric procedurebefore Swedish Adjustable Gastric Band [SAGB] implantation in cur-rent study; index SAGB – SAGB as first and only bariatric procedure;EQ-5D – EuroQoL 5-Dimensions; EQ-VAS – EuroQoL-Visual AnalogueScale.†Ideal body weight derived from the Metropolitan Weight Tables forLife Insurance, 1983.‡SAGB implant in this study was either an index intervention, or aSAGB following a complicated prior bariatric procedure (PBP) beforethe current study. 58.6% of the BMI<35 cohort were reinterventionpatients that satisfied the consecutive recruitment condition.www.cmj.hr

Ribaric and Buchwald: Three-year national band outcomes in non-morbidly obese 40994.5) at baseline (Table 2), representing a median AW re- Total cohort median changes in weight-related obesity indi-duction of 3.0 (-5.5-16.5: P = 0.126), corresponding to a cators were not significant at 3 years; however, a high levelmedian %TBWL of 2.9 (-6.6-17.5). Median EW was 22.1 of individual variation in weight-loss outcomes was noted.(10.0-26.9) compared to 28.5 (16.3-33.0) at baseline, repre- While some BMI<35 patients lost significant weight, somesenting a median EW reduction of 3.0 (-5.5-16.5: P = 0.126), gained weight as indicated by negative %TBWL and %EWLcorresponding to a median %EWL of 8.8 (-28.7-54.3). Me- values. Subgroup analyses, by intervention type and BMIdian BMI was 30.1 (25.9-32.9), compared to a preoperative category, were carried out. As detailed in Table 2, index SAGBmedian BMI of 33.1 (28.8-34.6). This change represented patients experienced significantly greater median %EWLan overall median BMI reduction of 1.0 (-2.0-6.0; P = 0.123). than PBP+SAGB patients (51.1 [8.8-92.4] vs -20.0 [-68.8-18.6];Median BMI evolution over 3 years by type of SAGB inter- P = 0.001). In fact, while PBP+SAGB patients actually gainedvention (PBP+SAGB vs index SAGB) for the BMI<35 cohort a median 4.0 kg of AW over 3 years, corresponding to a BMIis presented in Figure 1. increase of 1.6 kg/m2, index SAGB patients experienced sig-Table 2. Weight loss*† Median (IQR), N = 25Total group Preoperative 3-y Median change P-value‡ 87.0 (76.0-94.5)AW (kg) 33.1 (28.8-34.6) 80.0 (72.0-88.0) 3.0 (-5.5-16.5) 0.126BMI (kg/m2) 28.5 (16.3-33.0) 30.1 (25.9-32.9) 1.0 (-2.0-6.0) 0.123EW (kg) — 22.1 (10.0-26.9) 3.0 (-5.5-16.5) 0.126TBWL (%) — 2.9 (-6.6-17.5) — —EWL (%) 8.8 (-28.7-54.3) —Subgroup 1 comparison 82.0 (74.5-89.3)PBP+SAGB (n = 14) 31.0 (28.0-33.1) 85.0 (78.0-90.8) -4.0 (-9.3-5.8) 0.401AW (kg) 22.6 (13.5-29.3) 31.6 (29.0-33.3) -1.6 (-3.2-2.0) 0.421BMI (kg/m2) — 24.6 (17.6-28.6) -4.0 (-9.3-5.8) 0.401EW (kg) — -5.0 (-12.5-6.1) — —TBWL (%) -20.0 (-68.8-18.6) — —EWL (%) 93.0 (87.0-100.0)Index SAGB (n = 11) 34.5 (34.0-34.8) 76.0 (65.0-83.0) 16.0 (3.0-31.0) 0.010AW (kg) 28.5 (23.0-32.9) 5.9 (1.0-11.0) 0.010BMI (kg/m2)EW (kg) 32.1 (29.4-34.1) 15.3 (2.5-22.1) 16.0 (3.0-31.0) 0.010TBWL (%) — 17.0 (2.9-32.3) — —EWL (%) — 51.1 (8.8-92.4) — —Subgroup 2 comparisonBMI<30 (n = 7)AW (kg) 76.0 (68.0-85.0) 85.0 (80.0-88.0) -9.0 (-15.0 to -4.0) 0.042BMI (kg/m2) 26.9 (25.7-28.7) 30.1 (27.4-33.4) -3.2 (-5.0 to -1.7) 0.042EW (kg) 13.5 (13.1-15.9) 22.5 (15.9-33.1) -9.0 (-15.0 to -4.0) 0.042TBWL (%) — -11.8 (-17.6 to -5.9) — —EWL (%) — -66.5 (-153.3 to -28.3) — —BMI≥30 (n = 18)AW (kg) 90.0 (85.3-97.0) 78.0 (67.3-90.8) 10.5 (-0.5-19.8) 0.006BMI (kg/m2) 34.0 (32.3-34.6) 29.6 (25.3-32.9) 3.7 (-0.2-7.6) 0.006EW (kg) 31.1 (25.8-33.7) 20.0 (7.1-26.8) 10.5 (-0.5-19.8) 0.006TBWL (%) — 11.0 (-0.7-23.0) — —EWL (%) — 33.1 (-2.5-73.8) — —*Abbreviations: IQR – interquartile range; BMI – body mass index; AW – absolute weight; EW – excess weight; PBP – Prior bariatric procedure beforeSwedish Adjustable Gastric Band [SAGB] implantation in current study; EWL – EW loss; TBWL – total body weight loss.†Calculations based on patients with complete preoperative and 3-y follow-up data.‡Related Samples Wilcoxon Signed Rank Test. www.cmj.hr

410 CLINICAL SCIENCE Croat Med J. 2014;55:405-15nificant changes over time in median weight-loss indica- a baseline BMI≥30 to <35 had significantly greater mediantors: AW decreased by 16 kg (P = 0.010), corresponding to a %EWL at 3 years than did those with a BMI<30 (33.1 [-2.5-BMI decrease of 5.9 kg/m2 (P = 0.010). Finally, patients with 73.8] vs -66.5 [-153.3 to -28.3]; P = 0.001). The BMI≥30 to <35 patient subset was further subdivided into PBP+SAGB (N = 8)Figure 1. Evolution in median body mass index (BMI, kg/m2) vs index SAGB patients (N = 10). Figure 2 depicts the evolu-over 3 years in Swedish Adjustable Gastric Band (SAGB) cohort tion of %TBWL and %EWL for BMI≥30 to <35 patients aswith baseline BMI<35 as moderated by intervention type moderated by whether they were a reintervention or index(prior bariatric procedure [PBP+SAGB] vs first intervention [in- SAGB. PBP+SAGB patients experienced somewhat irregulardex SAGB]). Error bars represent the ~ 95% confidence interval median weight outcomes over time; whereas, index SAGBbracketing the median. patients exhibited progressive, sustained weight loss: At 3 years, PBP+SAGB median BMI was reduced by 1.3 (-1.1-4.7, P = 0.030), median %TBWL was 3.9 (-3.1-14.8), median %EWL was 11.8 (-8.8-49.0); index SAGB median BMI was significant- ly reduced by 6.1 (0.5-12.1, P = 0.001), median %TBWL was 17.4 (1.5-35.4), median %EWL was 51.0 (4.1-100.4). In the development of a multivariate regression model exploring preoperative clinical variables significantly re- lated to %TBWL (ie, presence of comorbidity [r = 0.493, P = 0.012], QoL [EQ-VAS] [r = -0.482, P = 0.027], type of SAGB operation (prior bariatric procedure or index SAGB proce- dure) [r = 0.591, P = 0.002], and BMI [r = 0.631, P = 0.001]), only baseline BMI was found to be an independent predic- tor of 3-year %TBWL in the BMI<35 SAGB cohort. Results of simple linear regression of %TBWL on baseline BMI in the form of a scatterplot and regression line are presented in Figure 3. Baseline BMI and 3-year %TBWL correlated atFigure 2. Median weight-loss trends to 3 years post Swedish Figure 3. Scatter plot and regression line illustrating directAdjustable Gastric Band (SAGB) procedure for patients with relationship between preoperative body mass index (BMI,preoperative body mass index (BMI, kg/m2)≥30 to <35 as kg/m2) and percentage total body weight loss (%TBWL) formoderated by type of operation (prior bariatric procedure BMI<35 patients following Swedish Adjustable Gastric Band[PBP+SAGB] vs first intervention [index SAGB]) expressed in (SAGB) procedure at 3 years. Intersecting reference linespercentage total body weight loss (%TBWL) and percentage represent the point on the BMI axis (BMI = 30) above which aexcess weight loss (%EWL). Note: Follow-up rate at 18 and 24 positive %TBWL is predicted to occur at 3-year SAGB follow-up.months was not sufficient for reliable assessment.www.cmj.hr

Ribaric and Buchwald: Three-year national band outcomes in non-morbidly obese 411r = 0.631 (P = 0.001). A logistic regression model using BMI SAGB patients with a baseline BMI≥30 to <35. This subsetas the lone predictor was shown to correctly classify 83.3% of patients was the most successful in terms of weight loss:of patients into their respective “weight-loss” vs “weight- median baseline BMI was 34.7 (34.1-34.8) at 3-year follow-gain” groups. Logistic results are presented in the form of up, median BMI fell significantly (7.7) to 26.7 (20.2-29.4;a probability curve in Figure 4 (BMI odds ratio = 1.53 [95% P = 0.001); %TBWL was 22.3 (15.2-41.7); %EWL was 64.5CI: 1.1, 2.2]; beta coefficient = 0.428, P = 0.048; model con- (44.4-116.3).stant = -13.2, P = 0.008). ComorbiditiesOverall, 56.0% (14/25) of the SAGB BMI<35 cohort withcomplete weight data at 3 years achieved and main- Adhering to the study’s observational design, no diagnos-tained weight loss. Ninety-three percent (13/14) of those tic tests for comorbidity assessment were required. At eachcomprising the weight-loss group were patients who pre- visit, comorbidities were reported as present or absent. Sig-sented with a BMI≥30 to <35, and 57.0% (8/14) were index nificant variation in diagnostic methodology, terminology, and reporting regularity was noted. Despite this limitation,Figure 4. Probability curve depicting the likelihood of a qualitative analysis indicated a continued reduction in thepatient with a given preoperative body mass index (BMI, kg/ overall number of comorbidities over time and a gradu-m2) to experience weight loss (ie, positive percentage total al increase in those with no reported comorbidities. Atbody weight loss, %TBWL) at 3 years after Swedish Adjustable baseline, there were 37 comorbidities in 15/29 BMI<35 pa-Gastric Band (SAGB) procedure. Intersecting reference lines tients. At 3-year follow-up, comorbidities were significantlyrepresent 3 sample patients with baseline BMIs of 30.0, 31.0, reduced to 7 (P = 0.031); median number of comorbiditiesand 34.9 whose corresponding probability of weight loss at 3 per patient fell significantly from 1.0 (0.0-2.0) at baseline toyears following SAGB procedure are calculated to be 0.40, 0.50, 0.0 (0.0-0.0), P = 0.002.and 0.86, respectively. Quality of lifeTable 3. Quality of life Median (IQR) Three-year postoperative QoL outcomes were available in 58.6% of patients (17/29). Median EQ-5D utility score wasQol Median 0.8 (0.7-1.0) compared to 0.7 (0.3-0.8) at baseline (Table 3).Variable Baseline 3-y change P-value† This represented a significant within-patient median QoLEQ-5D   0.7 (0.3-0.8)   0.8 (0.7-1.0) 0.2 (0.0-0.3) 0.028 improvement of 0.2 (0.0-0.3) (P = 0.028), greater than 2.5EQ-VAS 50.0 (40.0-73.8) 75.0 (55.0-90.0) 5.0 (-15.0-40.0) 0.214 times the accepted clinically important difference. Median*Abbreviations: QoL – Quality of life; EQ-5D – EuroQoL 5-Dimensions; EQ-VAS was 75.0 (55.0-90.0) compared to 50.0 (40.0-73.8)EQ-VAS – EuroQoL-Visual Analogue Scale; IQR – interquartile range. at baseline; the median increase of 5.0 (-15.0-40.0) was†P-values obtained from related-samples Wilcoxon Signed Rank Tests not significant (P = 0.214). Regression analysis indicated aassessing median QoL differences in patients with complete preop- significant association between weight loss and QoL im-erative and 3-y follow-up data (ie, N = 16 for EQ-5D, N = 17 for EQ-VAS). provement. Using EQ-5D individual change scores as the response variable while controlling for baseline BMI, BMI reduction was significantly related to increasing EQ-5D utility scores (adjusted R2 = 0.30; F(2,13) = 4.3; P = 0.037). Discussion Results suggest that the SAGB was safe and effective in French patients with a baseline BMI<35. There was no mor- tality and the AE rate was 0.23 AEs per patient-year, approx- imately similar to the 0.19 AE rate found in the main HAS cohort study. Adverse events were primarily confined to PBP+SAGB patients; whereas, 83.3% of index SAGB pa- tients experienced no AE. SAGB device survival rate was also comparable to that found in the main co- www.cmj.hr

412 CLINICAL SCIENCE Croat Med J. 2014;55:405-15hort study (86.2% vs 87.0%). QoL was improved and a re- patients should be an option for carefully selected patients.duction in overall number of comorbidities was observed. O’Brien et al (2006), for example, published a randomizedBMI reduction was significantly related to positive changes controlled trial of AGB vs medical therapy in BMI 30-35 pa-in patient health status. On balance, weight loss trended tients (2 groups of 40 patients each) that demonstratedtoward significance at 3 years; however, some patients equivalent weight loss at 6 months; at 2 years, the medicaldemonstrated weight gain. For example, those present- therapy group had regained most of their weight, whereas,ing with a BMI<30 (89.0% PBP+SAGB) experienced a me- the surgical group had an 87.2% EWL (-20 kg) (38). Also, thedian 9.0-kg AW gain (TBWL = -11.8%). Conversely, patients recently reported randomized controlled “Surgical Therapywith BMI≥30 to <35 experienced significant AW loss (10.5 and Medications Potentially Eradicate Diabetes Efficiently”kg), median 33.1% EWL, and median 11.0% TBWL – more (STAMPEDE) trial (2012) showed the effectiveness of sleevethan double the 5.0% TBWL threshold associated with sig- gastrectomy and RYGB in BMI≥27 patients in reducingnificant comorbidity improvement (32). First-intervention weight and treating type 2 diabetes mellitus (39).BMI≥30 to <35 patients experienced a median EWL of51.0% (TBWL = 17.4%). In addition, within the BMI<35 co- Evidence for lowering the BMI cutoff for surgery comeshort, logistic regression modeling suggested that a base- from multiple observational studies as well, particularlyline BMI≥30 was the point above which weight loss was with respect to the AGB procedure. Angrisani et al (2004)likely to occur 3 years post SAGB surgery. reported the Italian experience in 210 AGB patients with a mean preoperative BMI of 33.9. At 60-month follow-up,Although weight-loss findings for the BMI≥30 to <35 first- mean BMI was 29.2 (40). Parikh et al (2006) described a 26-time SAGB patients derive from a very small subgroup kg weight loss at 2 years in low-BMI AGB patients that was(N = 10), their median weight-loss outcomes over 3 years sustained at 3-year follow-up (41). In 2009, Sultan et al re-were comparable to those of the 517 morbidly obese pa- ported 53 AGB patients with a mean baseline BMI of 33.1tients of the original HAS cohort (median BMI change, 6.1 who attained a BMI of 25.8 and EWL of 69.7% at 2 yearsvs 7.9; EWL, 51.0% vs 49.3%). The subgroup outcomes sug- along with substantial improvement in comorbidities (42).gest that surgical weight loss in patients in the BMI≥30 to Both Choi et al (2010) and Varela et al (2011) compared<35 category follows a pattern similar to that in patients low-BMI and morbidly obese cohorts undergoing AGBwith BMI>35. The observation lends support to the idea and found the procedure comparably safe and effectivethat lowering the 1991 NIH (1) bariatric surgery cutoff to in both weight categories (43,44); Varela et al also noted30 may be reasonable. In addition, obesity-related health that low-BMI patients had shorter operative times and lessrisks, such as type 2 diabetes mellitus and cardiovascular blood loss.disease, tend to arise at lower BMIs in certain non-Cauca-sian populations (eg, Asian Indians) due to a higher per- In the current SAGB study, in which median weight losscentage and central distribution of body fat (33-35). The in the BMI≥30 to <35 subgroup was significantly greaterAsian Indian Consensus Group, for example, has moved than in the BMI<30 subgroup, neither group lost an exces-to evaluate weight-related health risk with alternatives to sive amount of weight; in fact, mean AW increased slightlythe BMI metric in these patients and to lower the BMI cut- in the BMI<30 group (mostly prior bariatric procedure pa-off for bariatric surgery to BMI>32.5 with a comorbidity or tients), as is typical for bariatric surgery patients after theBMI>37.5 without comorbidities (36). point of their greatest weight loss. In 2007, Scopinaro et al found in their study of low-BMI biliopancreatic diversionThe American Society for Metabolic and Bariatric Surgery (BPD) patients that, although the mildly obese group lost(ASMBS) Position Statement on BMI 30-35 concluded in late nearly twice the weight of the overweight group, weight2012 that class 1 obesity leads to other serious comorbid loss was not excessive in either low-BMI category (45). Oth-illnesses and a lowered life expectancy, and that there was er surgical studies, including those using BPD, BPD withno evidence of clinical or cost-effectiveness, ethics, or eq- duodenal switch, AGB, sleeve gastrectomy, and RYGB, haveuity that should exclude the BMI 30-35 group from bariat- observed the same phenomenon (46-48). Weight loss ap-ric surgical treatment (37). The Statement recommended pears to stabilize within the postoperative year at a BMI>25that, at a minimum, certain procedures (ie, gastric band- regardless of whether the procedure falls into the restric- tive, malabsorptive/restrictive, or primarily malabsorptive ing, sleeve gastrectomy, Roux-en-Y gastric bypass [RYGB]) surgical category (49), and regardless of the preoperative that have been shown safe and effective in short and BMI. A homeostatic mechanism may exist that facilitates mid-term randomized controlled trials in BMI 30-35www.cmj.hr

Ribaric and Buchwald: Three-year national band outcomes in non-morbidly obese 413weight loss in proportion to procedure-specific caloric 11Polyclinique de Rillieux, 69165 Rillieux-La-Pape, Franceabsorption capacity (5). An integrative analysis of the 16 12CH de Meaux, 77100 Meaux, Francethen-existing bariatric surgery studies in low-BMI patients 13Clinique St Louis, 34190 Ganges, Francedetected the same pattern of lesser weight loss in patients 14Polyclinique du Grand Sud, 30932 Nîmes Cedex 9, Francewith BMI<30 than in those with a BMI≥30, suggesting a 15Polyclinique de Bois Bernard, 62320 Bois Bernard, Franceblunting of the weight-loss cascade at around 30 BMI. 16Clinique Lamartine, 74200 Thonon-les-Bains, France 17CHG Antoine Gayraud de Carcassonne, 11890 Carc. 9, FranceAlthough the current study was limited by a restricted 18Clinique du Dr Priollet, 51000 Châlons en Champagne, Francepopulation of 29, the findings represent a small addition to 19Clinique de la Plaine, 63100 Clermont Ferrand, Francethe evidence base for bariatric surgery in the BMI<35 pa- 20Clinique Ambroise Paré, 62 660 Beuvry, Francetients. As in results for the primary HAS “real-world” cohort 21Clinique de la Casamance, 13400 Aubagne, Francestudy, the current low-BMI report contains an underreport- 22Clinique Ambroise Paré, 31100 Toulouse, Franceing bias partially due to data recording by numerous surgi- 23Clinique des Ursulines, 17 rue Raymond Poincaré, 10000 Troyes, Francecal teams across diverse locations in France; calculating a 24Clinique Notre Dame de l’Espérance, 66100 Perpignan, Francequantitative measure of change in specific comorbidities 25Clinique de l’Europe, 76100 Rouen, Francewas, therefore, not possible. 26Clinique des Emailleurs, 87000 Limoges, France 27Clinique Hoffmann, 93110 Rosny-sous-Bois, FranceAs early as 1997, Mason et al noted the dramatic trend to- 28Polyclinique des 4 Pavillons, 33310 Lormont, Franceward increasingly higher weights in bariatric surgery can-didates. They hypothesized that escalating obesity and Funding The study was financially supported by Ethicon Endo-Surgery Eu-life-threatening comorbidities should be prevented rather rope, GmbH, a subsidiary of J&J, Hamburg, Germanythan treated in their full expression (50). Current study out- Ethical approval given by the French Health Technology Assessment Bodycomes and those of a growing evidence base appear to (HAS), the Commission Nationale de l’Informatique et des Libertés, and thesupport the value of lowering the BMI access point for bar- Comité Consultatif sur le Traitement de l’Information en matičre de Recher-iatric surgery to permit earlier intervention in appropriate che dans le domaine de la Santé. The trial (#NCT01183975) was registeredpatients. Similar to findings in morbidly obese SAGB pa- in the Clinical Trials Web Database.tients at 3 years, SAGB treatment for low-BMI patients in Declaration of authorship GR planned and supervised the study uponFrance, particularly those with BMI≥30 to <35, was found which the current subset study was based. GR and JNB planned and devel-safe and effective. oped the outline for the current subset study. JNB supervised the analysis of the sample, contributed to data analysis, and prepared the manuscript.Acknowledgments We thank T. W. McGlennon, Director, Statistical Analy- Both authors reviewed and approved all content.sis and Quality of Life Assessment, McGlennon MotiMetrics (M3), WI, USA, Competing interests All authors have completed the Unified Competingfor performing the statistical analysis, and F. Daoud, Director, Data Manage- Interest form at www.icmje.org/coi_disclosure.pdf (available on requestment & Biometrics, Medextens SARL, Paris, France, for statistical consulta- from the corresponding author) and declare: GR is an employee of John-tion. We are grateful to the following surgeons who performed the banding son & Johnson (J&J). JNB is a CRO employee of Medwrite LLC under con-procedures and collected the patient data for the published French govern- tract with J&J.mental Health Technology Assessment evidence study summarizing resultsin the total cohort: ReferencesThe French Health Technology Assessment Body (Haute Autorité de San- 1 National Institutes of Health Consensus Development Panel.té [HAS]) Swedish Adjustable Gastric Band (SAGB) Study Group: J-F. Ain,1L. Arnalsteen,2 R. Arnoux,3 E. Attal,4 R. Barei,5 P. Bergevin,6 J. Cady,7 P. Campan,8 Gastrointestinal surgery for severe obesity. Ann Intern Med.J-M. Catheline,9 J-M. Chevallier,10 J. Dargent,11 B. Dehaye,12 C. Deseguin,13J. Flaisler,14 G. Fromont,15 H. Johanet,5 G. Juglard,16 F. Labbé,17 A. Legris,18 P. 1991;115:956-61. Medline:1952493 doi:10.7326/0003-4819-115-Lointier,19 F. Meaux,20 E. Nini,17 P. Noël,21 F. Pattou,2 G. Pugnet,22 D. Quinaux,23J.L. Riqué,24 S. Rossi,25 M. Sodji,26 J. Tussiot,27 C. Vaudois28 12-9561Clinique JB Denis, 71000 Mâcon, France 2 Obesity WHO. preventing and managing the global epidemic.2CHRU Lille, Hopital C. Hurriez, 59 000 Lille, France Report of a WHO Consultation on Obesity, Geneva, June 1997.3Clinique du Tondu, 33000 Bordeaux, France Geneva: WHO; 1998.4Polyclinique St Côme, 60200 Compičgne, France 3 Pories WJ, MacDonald KG Jr, Flickinger EG, Dohm GL, Sinha5Clinique Sainte Marie, 95520 Osny, France MK, Barakat HA, et al. Is type II diabetes mellitus (NIDDM) a6Centre Hospitalier Privé du Montgardé, 78410 Aubergenville, France surgical disease? Ann Surg. 1992;215:633-42. Medline:16326857Clinique Geoffroy St Hilaire, 75005 Paris, France doi:10.1097/00000658-199206000-000108CHU Conception, 13005 Marseille, France 4 Pories WJ, Swanson MS, MacDonald KG, Long SB, Morris PG, Brown9CH Avicenne, 93009 Bobigny cedex, France BM, et al. Who would have thought it? An operation proves to be10Hôpital Européen Georges Pompidou, 75015 Paris, France the most effective therapy for adult-onset diabetes mellitus. Ann Surg. 1995;222:339-50. Medline:7677463 doi:10.1097/00000658- 199509000-00011 5 Fried M, Ribaric G, Buchwald JN. Metabolic surgery for the treatment of type 2 diabetes in patients <35kg/m2: an www.cmj.hr

414 CLINICAL SCIENCE Croat Med J. 2014;55:405-15 integrative review of early studies. Obes Surg. 2010;20:776-90. 16 Clinical investigation of medical devices for human subjects – Medline:20333558 doi:10.1007/s11695-010-0113-3 Part 1: Clinical investigation plans. International Standard Ref. ISO6 Basdevant A, Paita M, Rodde-Dunet MH, Marty M, Nogučs 14155-2:2003(E). F, Slim K, et al. A nationwide survey on bariatric surgery: two years prospective follow-up. Obes Surg. 2007;17:39-44. 17 World Medical Association. Declaration of Helsinki: Medline:17355767 doi:10.1007/s11695-007-9004-7 ethical principles for medical research involving human7 Avis de la CEPP du 1er septembre 2004 modifié le 20 octobre 2004 subjects, 22 October 2008. Available from: www.wma.net/ SAGB (anneau gastrique ajustable suédois), implant annulaire en/30publications/10policies/b3/index.html. Accessed: March 18, ajustable pour gastroplastie. Modčles 2100-X et 2200-X [in 2014. French]. Available from: http://www.santedev.org/biblio/pdf2/has/ cepp/2004/10/20/pp020367.pdf. Accessed: March 12, 2012. 18 Fried M, Hainer V, Basdevant A, Buchwald H, Deitel M, Finer N, et8 Avis de la CEPP 8 mars 2006 SAGB (anneau gastrique ajustable al. Inter-disciplinary European guidelines on surgery of severe suédois) Quick-Close fourni avec le site VELOCITY et son obesity. Int J Obes (Lond). 2007;31:569-77. Medline:17325689 applicateur, implant annulaire ajustable pour gastroplastie. Modčle BD2XV [in French]. 19 WHO. Physical status: the use and interpretation of anthropometry.9 Ribaric G, Buchwald JN, d’Orsay G, Daoud F; French Health Report of a WHO Expert Committee. WHO Technical Report Series Technology Assessment Body. (Haute Autorité de Santé [HAS]) 854. Geneva: WHO; 1995. Swedish Adjustable Gastric Band (SAGB™) study group. 3-year real-world outcomes with the Swedish Adjustable Gastric Band™ in 20 Lorentz FH. Ein neuer Konstitionsinde [in German]. Klin France. Obes Surg. 2013;23:184-96. Medline:23054572 doi:10.1007/ Wochenschr. 1929;8:348-51. doi:10.1007/BF01721823 s11695-012-0765-210 Garb J, Welch G, Zagarins S, Kuhn J, Romanelli J. Bariatric surgery 21 Lorentz FH. Ein neuer Konstitionsinde der Frau [in German]. Klin for the treatment of morbid obesity: a meta-analysis of weight Wochenschr. 1929;16:734-6. doi:10.1007/BF01738018 loss outcomes for laparoscopic adjustable gastric banding and laparoscopic gastric bypass. Obes Surg. 2009;19:1447-55. 22 Miller MA. A calculated method for the determination of ideal Medline:19655209 doi:10.1007/s11695-009-9927-2 body weight. Nutritional Support Services. 1985;5:31-3.11 Cunneen SA, Phillips E, Fielding G, Banel D, Estok R, Fahrbach K, et al. Studies of Swedish adjustable gastric band and Lap- 23 Deitel M, Greenstein RJ. Recommendations for reporting Band: systematic review and meta-analysis. Surg Obes Relat Dis. weight loss. Obes Surg. 2003;13:159-60. Medline:12760387 2008;4:174-85. Medline:18243061 doi:10.1016/j.soard.2007.10.016 doi:10.1381/09608920376446711712 Association Française d’Etudes et de Recherches sur l’Obésité, Association de Langue Française pour l’Etude du Diabčte et des 24 American Society for Bariatric Surgery Standards Committee. 2004- Maladies Métaboliques, Société de Nutrition et de Diététique 2005: Guidelines for weight calculations and follow-up in bariatric de Langue Française. Recommandations pour le diagnostic, la surgery. Surg Obes Relat Dis. 2005;1:67-8. Medline:16925214 prévention et le traitement des obésités en France [in French]. Cah doi:10.1016/j.soard.2004.12.005 Nutr Diet. 1998;33:10-42.13 Haute Autorité de Santé. Obésité: prise en charge chirurgicale chez 25 Ackroyd R, Mouiel J, Chevallier JM, Daoud F. Cost-effectiveness l’adulte. Recommandations de bonnes pratiques professionnelles. and budget impact of obesity surgery in patients with type 2 January 2009. Available from: http://www.has-sante.fr/portail/ diabetes in 3 European countries. Obes Surg. 2006;16:1488-503. jcms/c_765529/obesite-prise-en-charge-chirurgicale-chez-l- Medline:17132416 doi:10.1381/096089206778870067 adulte. Accessed: March 12, 2012.14 Prospective National Cohort Study on Swedish Adjustable Gastric 26 EuroQoL Group. EuroQoL. A new facility for the measurement Band (SAGB) for Gastroplasty. (Étude de Cohorte Nationale of health-related quality of life. Health Policy. 1990;16:199-208. Prospective de l’Implant Annulaire Ajustable Pour Gastroplastie Medline:10109801 doi:10.1016/0168-8510(90)90421-9 SAGB). PROTOCOL N°: 05-FR-004 1.0. Available from: http:// clinicaltrials.gov/ct2/show/NCT01183975?term=ethicon+endo- 27 Dolan P, Gudex C, Kind P, Williams A. A social tariff for EuroQoL: surgery&rank=15. Accessed: March 18, 2014. results from a UK general population survey. In: York Centre for15 Clinical investigation of medical devices for human subjects – Part Health Economics Discussion Paper. York: University of York; 1990. 1: General Requirements. International Standard Ref. ISO 14155- 1:2003(E). 28 Kind P, Dolan P, Gudex C, Williams A. Variations in population health status: results from a United Kingdom national questionnaire survey. BMJ. 1998;316:736-41. Medline:9529408 doi:10.1136/bmj.316.7133.736 29 Dixon S, Farina C, McEwan P. Evaluation of the association between health-related utility and obesity in hospital treated subjects. Value Health. 2004;7:331. doi:10.1016/S1098-3015(10)62409-6 30 Brazier J, Roberts J, Tsuchiya A, Busschbach J. A comparison of the EQ-5D and SF-6D across seven patient groups. Health Econ. 2004;13:873-4. Medline:15362179 doi:10.1002/hec.866 31 Di Lorenzo N, Furbetta F, Favretti F, Segato G, De Luca M, Micheletto G, et al. Laparoscopic gastric banding via pars flaccida versus perigastric positioning: technique, complications,www.cmj.hr

Ribaric and Buchwald: Three-year national band outcomes in non-morbidly obese 415 and results in 2,549 patients. Surg Endosc. 2010;24:1519-23. 42 Sultan S, Parikh M, Youn H, Kurian M, Fielding G, Ren C. Early U.S. Medline:20354885 doi:10.1007/s00464-009-0669-y outcomes after adjustable gastric banding in patients with a body32 American Society for Bariatric Surgery Standards Committee. 2004- mass index less than 53 kg/m2. Surg Endosc. 2009;23:1569-73. 2005. Guidelines for weight calculations and follow-up in bariatric Medline:19263156 doi:10.1007/s00464-009-0341-6 surgery. Surg Obes Relat Dis. 2005;1:67-8. Medline:16925214 doi:10.1016/j.soard.2004.12.005 43 Choi J, Digiorgi M, Milone L, Schrope B, Olivera-Rivera L, Daud33 Shah SS, Todkar JS, Shah PS, Cummings DE. Diabetes remission A, et al. Outcomes of laparoscopic adjustable gastric banding and reduced cardiovascular risk after gastric bypass in Asian in patients with low body mass index. Surg Obes Relat Dis. Indians with body mass index <35 kg/m2. Surg Obes Relat Dis. 2010;6:367-71. Medline:20185374 doi:10.1016/j.soard.2009.09.021 2010;6:332-8. Medline:19846351 doi:10.1016/j.soard.2009.08.00934 Wen CP, David Cheng TY, Tsai SP, Chan HT, Hsu HL, Hsu CC, et 44 Varela JE, Frey W. Perioperative outcomes of laparoscopic al. Are Asians at greater mortality risks for being overweight adjustable gastric banding in mildly obese (BMI < 35 kg/ than Caucasians? Redefining obesity for Asians. Public Health m2) compared to severely obese. Obes Surg. 2011;21:421-5. Nutr. 2009;12:497-506. Medline:18547457 doi:10.1017/ Medline:21308421 doi:10.1007/s11695-011-0365-6 S136898000800280235 WHO. Appropriate body-mass index for Asian populations and 45 Scopinaro N, Papadia F, Marinari G, Camerini G, Adami G. Long- its implications for policy and intervention strategies. Lancet. term control of type 2 diabetes mellitus and the other major 2004;363:157-63. Medline:14726171 doi:10.1016/S0140- components of the metabolic syndrome after biliopancreatic 6736(03)15268-3 diversion in patients with BMI <35 kg/m2. Obes Surg. 2007;17:185-36 Misra A, Chowbey P, Makkar BM, Vikram NK, Wasir JS, Chadha D, 92. Medline:17476869 doi:10.1007/s11695-007-9045-y et al for Consensus Group. Consensus statement for diagnosis of obesity, abdominal obesity and the metabolic syndrome for Asian 46 Scopinaro N, Adami GF, Papadia FS, Camerini G, Carlini F, Briatore Indians and recommendations for physical activity, medical and L, et al. The effects of biliopancreatic diversion on type 2 diabetes surgical management. JAPI. 2009;57:163-70. Medline:19582986 mellitus in patients with mild obesity (BMI 30-35 kg/m2) and37 ASMBS Clinical Issues Committee. ASMBS position statement: simple overweight (BMI 25-30 kg/m2): a prospective controlled bariatric surgery in class 1 obesity (BMI 30-35 kg/m2). Surg Obes study. Obes Surg. 2011;21:880-8. Medline:21541815 doi:10.1007/ Relat Dis. 2013;9:e1-10. doi:10.1016/j.soard.2012.09.002 s11695-011-0407-038 O’Brien PE, Dixon JB, Laurie C, Skinner S, Proietto J, McNeil J, et al. Treatment of mild to moderate obesity with laparoscopic 47 Frenken M, Cho EY. Metabolic intestinal bypass surgery for type 2 adjustable gastric banding or an intensive medical program: diabetes in patients with a BMI <35 kg/m2: comparative analysis of a randomized trial. Ann Intern Med. 2006;144:625-33. 16 patients undergoing either BPD, BPD-DS, or RYGB. Obes Facts. Medline:16670131 doi:10.7326/0003-4819-144-9-200605020- 2011;4:13-7. Medline:22027284 doi:10.1159/000327038 0000539 Schauer PR, Kashyap SR, Wolski K, Brethauer SA, Kirwan JP, Pothier 48 Gianos M, Abdemur A, Fendrich I, Gari V, Szomstein S, Rosenthal CE, et al. Bariatric surgery versus intensive medical therapy in RJ. Outcomes of bariatric surgery in patients with body mass index obese patients with diabetes. N Engl J Med. 2012;366:1567-76. <35 kg/m2. Surg Obes Relat Dis. 2012;8:25-30. Medline:22019140 Medline:22449319 doi:10.1056/NEJMoa1200225 doi:10.1016/j.soard.2011.08.01240 Angrisani L, Favretti F, Furbetta F, Iuppa A, Doldi SB, Paganelli M, et al. Italian group for Lap-Band System: results of multicenter 49 Buchwald H, Buchwald JN. Evolution of surgery for morbid obesity. study on patients with BMI ≤35 kg/m2. Obes Surg. 2004;14:415-8. In: Pitombo C, Jones KB, Higa KD, Pareja JC, editors. Obesity Medline:15072665 doi:10.1381/096089204322917963 surgery: principles and practice. New York: McGraw-Hill Medical;41 Parikh M, Duncombe J, Fielding GA. Laparoscopic adjustable 2007; p. 3-14. gastric banding for patients with body mass index of ≤35 kg/ m2. Surg Obes Relat Dis. 2006;2:518-22. Medline:17015204 50 Mason EE, Tang S, Renquist KE, Barnes DT, Cullen JJ, Doherty C, et doi:10.1016/j.soard.2006.07.005 al for the National Bariatric Surgery Registry (NBSR) Contributors. A decade of change in obesity surgery. Obes Surg. 1997;7:189-97. Medline:9730547 doi:10.1381/096089297765555719 www.cmj.hr

OBES SURG (2012) 22:1298–1307DOI 10.1007/s11695-012-0684-2 OTHERLaparoscopic Greater Curvature Plication (LGCP)for Treatment of Morbid Obesity in a Series of 244 PatientsM. Fried & K. Dolezalova & J. N. Buchwald &T. W. McGlennon & P. Sramkova & G. RibaricPublished online: 1 June 2012# Springer Science+Business Media, LLC 2012Abstract Results Mean baseline BMI (±SD) was 41.4±5.5 (80.7 %Background Laparoscopic greater curvature plication women, mean age 46.1±11.0 years, 68 [27.9 %] patients(LGCP) is a new metabolic/bariatric surgical procedure that had T2DM). Mean operative time was 70.6 min; meanrequires no resection, bypass, or implantable device. We hospitalization, 36 h (24–72). Sixty-eight patients (27.9 %)report LGCP outcomes in 244 morbidly obese patients. experienced postoperative nausea and/or vomiting that wasMethods Between 2010 and 2011, patients underwent LGCP. controlled within 36 h. There was no mortality. Major com-Body mass index (BMI, kilogram per square meter) evolution, plication rate was 1.2 % (n03). Repeated-measures ANOVAexcess BMI loss (%EBMIL), excess weight loss (%EWL), indicated significant weight loss across time points (p<0.001).complications, and type 2 diabetes mellitus (T2DM) changes At 6 months (n0105), BMI, %EBMIL, and %EWL werewere recorded. Repeated-measures analysis of variance 36.1±4.7, 34.8±17.3, and 31.8±15.9. Preoperative BMI was(ANOVA) was used to assess weight change at 6, 12, and the only predictor of weight loss. Patients with BMI <40 lost18 months. Subgroup analyses were conducted to provide more weight than those ≥40, although by 9 months, differ-benchmark outcomes at 6 months. Logistic regression was used ences were no longer significant. In patients with preoperativeto identify characteristics predictive of suboptimal weight loss. BMI <40, 18-month %EWL approached 50 % and %EBMIL exceeded 50 %. At 6 months, 96.9 % of patients’ T2DM wasM. Fried significantly improved/resolved.First Faculty of Medicine, Charles University, Conclusions Over the short term, LGCP results in effectiveKaterinska 32, weight loss and significant T2DM reduction with a very low128 00 Prague 2, Czech Republic rate of complications.M. Fried (*) : K. Dolezalova : P. Sramkova Keywords Laparoscopic greater curvature plication . LGCP . Obesity . DiabetesOB Klinika a.s.,Pod Krejcarkem 975, Introduction130 00 Prague 3, Czech Republice-mail: [email protected] Of all treatments for obesity, metabolic/bariatric surgery has shown the greatest sustained effectiveness in achievingJ. N. Buchwald weight loss and comorbidity improvement [1, 2] with de-Division of Scientific Writing, creased mortality, morbidity, and health care utilizationMedWrite Medical Communications, [3–5]. The very few prior preclinical and clinical studies ofMaiden Rock, WI, USA laparoscopic greater curvature plication (LGCP)—the new- est operation in metabolic surgery—suggest the potential forT. W. McGlennon LGCP to provide safe, significant weight loss and improve-Statistical Analysis and Quality of Life Assessment, ment of metabolic parameters similar to that shown byMcGlennon MotiMetrics,Maiden Rock, WI, USAG. RibaricEuropean Surgical Institute,Ethicon Endo-Surgery (Europe) GmbH,Hamburg, Germany

OBES SURG (2012) 22:1298–1307 1299currently accepted restrictive metabolic procedures. LGCP greater curvature gastric plication techniques (enrollmentintroduces no implantable device (e.g., band, ring), requires goal of 45 patients) was approved for initiation by theno gastric or intestinal resection or intestinal bypass, is po- institutional review boards (IRBs) of three sites (Cleveland,tentially reversible, and can be augmented with more exten- OH; Columbus, OH; Prague, Czech Republic); the studysive procedures, if warranted. was funded by Ethicon Endo-Surgery. In 2010, following good early weight loss in the pilot study, no major perioper- Alternately referred to as total gastric vertical plication, ative or early postoperative complications, and patients’or gastric imbrication, LGCP consists of laparoscopically tolerability of the procedure, the Prague site, a bariatricinfolding the greater gastric curvature to reduce stomach surgery center of excellence, submitted a new protocol andvolume by placement of one or two rows of nonabsorbable received extended local IRB approval to perform the greatersutures or staples [6]. A variation, anterior gastric plication curvature plication technique in a new prospective cohort.has been studied [7]. Historical predecessors of gastric pli- The protocol provided latitude for surgeons to make slightcation include the Nissen fundoplication, a procedure modifications to the surgical technique (e.g., placing one vsintended to reduce gastroesophageal reflux disease that has two rows of sutures) as experience with the procedure wasalso been associated with some weight loss [8], and restric- gained. The IRB did not limit patient recruitment, but requiredtive metabolic surgical procedures, whose primary objective complications reports every 6 months, with the option ofis limitation of the passage of food through the alimentary termination if complications exceeded long-term hospitaltract [9]. The neuroendocrine mechanisms of the gut–brain averages for bariatric procedures.axis that affect weight loss and resolution of comorbiddisease [10, 11] are as yet unexplored for LGCP. Patient Eligibility, Informed Consent Preclinical studies have demonstrated LGCP viability. In Main study inclusion criteria were those of the internation-1969, Kirk described safe, marked weight loss in rats by ally accepted American Society for Metabolic and Bariatricinvagination of the greater curvature of the stomach [12]. Fusco Surgery [24] and National Institutes of Health [25] (i.e., ageet al., in 2006 and 2007, saw superior weight loss in Wistar rats 18 to 65; BMI ≥40, or ≥35 with comorbidities). The Praguewith greater curvature plication compared to anterior plication IRB placed no upper limit on patient BMI as an inclusiontechnique [13, 14]. Moreover, Menchaca et al., in 2011, found factor, nor were diabetic patients excluded due to anti-short-term durability of serosa-to-serosa gastric plications in diabetic treatment or medication. Patients with small hiataldogs using a variety of fastening mechanisms [15]. hernias (up to 2 cm) were also included. Ethical conduct standards of the Declaration of Helsinki were maintained; a Clinically, LGCP is still an investigational procedure asso- written, IRB-approved, informed consent form wasciated with minimal evidence [16]. Six peer-reviewed papers reviewed and signed by all patients [26]. The cost of LGCPhave been published [7, 17–21]. Four of the studies describe procedures was covered by national insurance, as are othercohorts with an N<9 at 6-month LGCP evolution; comorbid- surgical procedures in the Czech Republic.ity change is minimally reported. In these studies, a total of396 LGCP patients were described with a preoperative body Endpoints, Data Collectionmass index (BMI, kilogram per square meter) of <35 to 58.The average 6-month %EWL for LGCP ranged from 32 to Operative time and duration of hospital stay were recorded.54 %, comparable to that of other restrictive procedures (e.g., Effectiveness endpoints included 6-, 12-, and 18-monthadjustable gastric banding (AGB), 47.45 % [22]; sleeve gas- mean BMI and %EBMIL, in response to discussion in thetrectomy as a stand-alone procedure, 60.4 % [23]). literature positing this measure as more precise than % EWL for reporting weight loss [27–29]. Percentage In the current investigation, we recorded LGCP safety and EWL was also reported to facilitate comparisons witheffectiveness with follow-up to 18 months. Our focus was to prior literature. Percentage EBMIL was calculated as:establish benchmark outcomes for weight loss and changes in ½preoperative BMI À current BMIŠ=½preoperative BMI Àtype 2 diabetes mellitus (T2DM) in a substantial number of 25Š  100 [29]. Percentage EWL was calculated as:patients at the 6-month time point. We report the findings from ½preoperative weight À current weightŠ=½preoperative weighta single center, consecutive case series of 244 patients who Àideal weightŠ  100 [30] relative to the 1983 Metropolitanunderwent LGCP for treatment of morbid obesity. Life Insurance tables [31]. Six-month weight loss was ana- lyzed in subgroups by gender, age, BMI, and comorbidity.Methods Diabetes reduction/resolution was recorded. Safety endpoints analyzed were minor and major complications. Data wereProtocol, Institutional Review scheduled for collection on the day of surgery, at 7–10 days,In 2009, a prospective feasibility study (www.clinicaltrials.gov identifier NCT00721227) to test laparoscopic, two-row,

1300 OBES SURG (2012) 22:1298–13074–6 weeks, and every 3 months for the first postoperative year, were incorporated into the final multivariate logistic regres-and finally, at 18 months. sion model. Statistical significance was set at p≤0.05.Surgical Technique ResultsLaparoscopic gastric volume reduction technique has been Baseline Patient Characteristicspreviously described and illustrated [6, 7]. Our techniqueconsisted of dissection beginning approximately 4 cm from Patients’ mean age was 46.1± 11.0 years; 80.7 % werethe pylorus and proceeding up toward the diaphragmatic left women (Table 1). Mean absolute weight (kilogram) wascrus, stopping approximately 2 cm below the angle of His, 118.9±21.1 and mean excess body weight, 50.8±17.6. Meanleaving the first two short gastric vessels intact. The gastric BMI was 41.4±5.5 (46.7 %<40; 46.3 % ≥40 but <50;body and fundus were mobilized. Beginning below the 7 % ≥50). Sixty-one percent of patients had at least oneangle of His and continuing distally to approximately 4 cm comorbidity. Hypertension was the most prominent comorbidof the pylorus, the greater curvature was infolded and se- illness (110, 45.1 %), followed by T2DM (68, 27.9 %), andcured with full-thickness bites of polypropylene suture in one dyslipidemia (41, 16.8 %). Mean T2DM duration was 4.75±or two plication rows. Intraoperative endoscopy was used to 3.5 years; mean preoperative HbA1C was 6.4±1.4 %; meanverify the security of the plication. preoperative glycemic level was 162.8±62.7 mg/dL.Postoperative Care Operative Time, Hospital StayFor complications, visits were scheduled as needed. For The first 50 patients received two plication rows; subsequentunusual postoperative findings, a marked weight loss pla- patients (n0194) received a single row. Mean operative timeteau, or patient complaint (e.g., postprandial or other abdom- was 70.6 min, although there was a considerable learninginal pain, vomiting), the protocol stipulated ultrasonographyas the first investigational choice, with upper gastrointestinal Table 1 Preoperative patient characteristics(GI) endoscopy, where indicated. Characteristics Value, mean ± SD (95 % CI)Statistical Analysis (N 0 244)All statistical analyses were performed with the SPSS soft- Age (years) 46.1±11.0 (44.7, 47.5)ware package (version 20, SPSS [IBM], Chicago, IL, USA). Height (m) 1.69±0.1 (1.67, 1.70)Quantitative demographic variables were reported as mean, Absolute weight (kg) 118.9±21.1 (116.2, 121.6)standard deviation (SD), and 95 % confidence intervals Ideal body weight (kg) 68.1±6.3 (67.3, 68.9)(95 % CI); qualitative variables (demographic and outcome Excess body weight (kg) 50.8±17.6 (48.6, 53.0)variables) were described as number and percentage. Com- Waist circumference (cm) 119.7±15.7 (117.7, 121.7)plications were also provided as number and percentage. Body mass index (BMI, kg/m2) 41.4±5.5 (40.7, 42.1)Generally, quantitative outcome variables were reported as BMI <40, N (%) 114 (46.7)mean and SD, range, mean change and percentage change, BMI ≥40 and <50, N (%) 113 (46.3)and 95 % CI. Within-subject analyses of BMI evolution and BMI ≥50, N (%) 17 (7.0)%EBMIL were carried out using repeated-measures analysis Gender (N, %)of variance (ANOVA). Measures of change from baseline at Male 47 (19.3)6 months were analyzed with the paired samples t test or Female 197 (80.7)Wilcoxon-signed rank test for quantitative variables and Comorbidities (N, %)McNemar’s test for qualitative variables. Between-group At least one comorbidity 148 (60.7)comparisons along quantitative measures were carried out Hypertension 110 (45.1)by means of parametric and nonparametric tests (i.e., inde- Type 2 diabetes mellitus 68 (27.9)pendent samples t test or ANOVA; Kruskal–Wallis test or Dyslipidemia 41 (16.8)Mann–Whitney U test), as determined by the Shapiro–Wilk Other 21 (8.6)test for normality. To investigate the relationships betweenqualitative variables, Fisher’s exact test was employed. Pre- Ideal body weight and excess body weight derived from the Metropol-operative patient characteristics identified as predictive of itan Weight Tables for Life Insurance [31]. Other 0 hypothyroidism,suboptimal weight loss in bivariate unadjusted analyses asthma, sleep apnea, depression, arthropathy, etc.

OBES SURG (2012) 22:1298–1307 1301curve. In the first 30 cases, operative time (skin-to-skin) was weight (kilogram) was 103.5±19.5 compared to 119.1±155 min; in the next 20 cases, 95 min; in the remaining 194, 21.1 at baseline, a mean reduction of 15.6±7.6 (95 % CIoperative time was 55 min. After the initial 50 patients, 14.1, 17.1; p<0.001). BMI was 36.1±4.7, down 13.5 % fromoperative time solely for greater curvature suturing was 41.6±5.0, a mean reduction of 5.6±2.6 (5.1, 6.1; p<0.001).reduced from 45 to 20 min. Mean postoperative time in Waist circumference (centimeter) was 111.8±13.2 relative tohospital was 36 h (24–72). 121.2±16.2 at baseline, down 9.4±10.5 (6.2, 12.6; p<0.001). Mean excess weight (kilogram) dropped 15.7±7.6 (14.2,Weight Outcomes 17.2) to 35.9±15.6 from 51.5±17.0 (p<0.001), a reduction that represents a mean %EWL of 31.8±15.9 (28.7, 34.9).Within-subject weight change over time was analyzed by Percentage EBMIL was 34.8±17.3 (31.5, 38.1).assessing mean BMI (and corresponding %EBMIL) at base-line, 6, 12, and 18 months. Two independent one-way Subgroup analyses of weight loss following LGCP byrepeated-measures ANOVAs, with Greenhouse–Geisser gender, age, BMI, and comorbidity using %EBMIL as thecorrections, indicated that mean BMI and %EBMIL meas- primary outcome measure are also presented in Table 2.ures differed significantly across time points (F[1.8, 52.3]0 There were no significant differences in weight loss between108.2, p<0.001; F[1.7, 47.8]094.1, p<0.001). Multiple males and females, age groups, general comorbidity groupspairwise comparisons were performed using Bonferroni- (comorbidity present vs not present) or specific comorbidityadjusted alpha levels. Following LGCP, significant reduc- groups (e.g., diabetic vs non-diabetic) (Table 3). A signifi-tions in BMI (and corresponding increases in %EBMIL) cant difference in weight loss was found between BMIrelative to baseline were achieved at all time points (p< subgroups (<40, ≥40). Patients with a preoperative BMI0.001); three patients (0.01 %) were lost to follow-up in the of <40 had significantly greater %EBMIL at 6 months thanseries. A significant linear effect was indicated (Fig. 1). patients with a preoperative BMI of ≥40 (42.0±20.8 vs 29.8±Based on estimated marginal means, mean BMI was re- 12.1; p<0.005).duced from 40.6±3.8 to 36.0±4.0 (p<0.001) at 6 months,with %EBMIL of 31.0±17.2. Although the downward BMI As the only significant difference demonstrated in sub-trend and corresponding upward trend in %EBMIL per- group analyses of weight loss at 6 months was found be-sisted, there were no statistically significant differences be- tween preoperative BMI groups, a retrospective analysis oftween BMI measures and %EBMIL measures taken at 6 and weight loss trends for the two groups was carried out12 months. BMI decreased further from 35.6±4.7 at (Fig. 2). Between-group differences in %EBMIL and12 months to 33.5±4.2 at 18 months (p<0.001) with a %EWL were analyzed at 1, 3, 6, 9, 12, and 18 months.corresponding increase in %EBMIL from 34.0±21.6 to Significant differences (p<0.005) were evident between46.5±20.1 (p<0.001). preoperative BMI groups in both outcome measures at 1, 3, and 6 months; however, by 9 months, statistical differ- In the follow-up of 105 LGCP patients at the 6-month ences were no longer present (BMI <40 group vs BMI ≥40;time point, paired samples t tests indicated that obesity %EBMIL, 37.6±21.6 vs 32.2±14.4, p00.32; %EWL, 35.9±indicators were significantly reduced (Table 2). Absolute 17.0 vs 30.5±13.6, p00.25). In fact, there were no significant differences in %EBMIL and %EWL between preoperativeFig. 1 Body mass index (BMI,kilogram per square meter),percentage excess BMIloss (%EBMIL), andpercentage excess weight loss(%EWL) over time afterlaparoscopic greater curvatureplication (LGCP). Means, SDs,and 95 % CIs are based onall available data at each timepoint; results from repeated-measures analyses of variance(ANOVAs) presented in textare based on estimatedmarginal means

1302 OBES SURG (2012) 22:1298–1307Table 2 Weight loss at 6 months Value (N0105)Total group Mean ± SD (range) Mean change (% change) 95 % CI P valueaAbsolute weight (kg) 103.5±19.5 (56.0 to 160.7) −15.6 (13.1) −17.1, −14.1b <0.001BMI (kg/m2) 36.1±4.7 (24.2 to 46.6) −5.6 (13.5) −6.1, −5.1b <0.001Waist circumference (cm) 111.8±13.2 (82.0 to 151.0) −9.4 (7.8) −12.6, −6.2b <0.001Excess weight (kg) 35.9±15.6 (−2.0 to 85.6) −15.7 (30.5) −17.2, −14.2b <0.001EWL (%) 31.8±15.9 (−1.67 to 107.1) – 28.7, 34.9c –EBMIL (%) 34.8±17.3 (−3.70 to 106.6) – 31.5, 38.1c –Subgroups N Mean ± SD (range) 95 % CIc P valueEBMIL (%), by gender Females 88 35.1±16.9 (−3.7 to 106.6) 31.5, 38.7 NSd Males 17 33.5±19.6 (12.5 to 66.4) 23.5, 43.6EBMIL (%), by age (years) NSe <40 31 38.9±13.9 (13.9 to 71.8) 33.8, 44.1 ≥40 to <50 29 36.4±21.2 (0.0 to 106.6) 28.4, 44.5 <0.005f ≥50 45 30.9±16.1 (−3.7 to 66.4) 26.0, 35.7EBMIL (%), by BMI NSd <40 43 42.0±20.8 (−3.7 to 106.6) 35.6, 48.4 ≥40 62 29.8±12.1 (−1.44 to 66.0) 26.7, 32.9EBMIL (%), by comorbidity Present 78 33.1±16.2 (−3.70 to 86.1) 29.4, 36.8 Not present 27 39.8±19.5 (16.0 to 106.6) 32.1, 47.5Calculations based on patients with complete preoperative and 6-month follow-up data. Due to insufficient number of patients in the ≥50 BMIcategory at 6-month follow-up to permit valid analysis, preoperative BMI categories ≥40 to <50 and ≥50 were combined to form the ≥40 subgroupcategoryBMI body mass index, EWL excess weight loss, EBMIL excess BMI loss, SD standard deviation, CI confidence intervala Paired samples t test assessing change from baselineb 95% CI of mean differencec 95% CI of the meand Independent samples Mann–Whitney U teste One-way ANOVA for independent samplesf Independent samples t testBMI groups at 12 months (%EBMIL, 46.3±25.6 vs 36.3± up to 18 months (%EBMIL, 52.4±21.8 vs 42.3±18.2, p018.9, p00.12; %EWL, 42.1±23.2 vs 34.4±17.9, p00.20) and 0.20; %EWL, 47.5±19.2 vs 40.6±17.9, p00.30). PercentageTable 3 Excess body mass in- Subgroups Number Mean ± SD (range) 95 % CIa p valuedex reduction by preoperative NSbcomorbidity at 6 months EBMIL (%), by diabetes 31.4±17.2 (−3.7 to 66.4) 25.2, 37.6 NSb 36.3±17.2 (−1.4 to 106.6) 32.3, 40.4 NScCI confidence interval, EBMIL Present 33excess body mass index loss Not present 72 33.1±16.5 (−3.7 to 86.1) 29.0, 37.2a95% CI of the mean 37.6±18.5 (10.7 to 106.6) 31.5, 43.7bIndependent samples t test EBMIL (%), by hypertensioncIndependent samples 34.4±20.1 (0.0 to 86.1) 25.0, 43.9Mann–Whitney U test Present 66 34.9±16.7 (−3.7 to 106.1) 26.2, 41.6 Not present 39 EBMIL (%), by dyslipidemia Present 20 Not present 85

OBES SURG (2012) 22:1298–1307 1303Fig. 2 Weight loss trends out to 18 months following laparoscopic weight loss group.” Preoperative absolute weight, BMI,greater curvature plication (LGCP) for preoperative body mass index and waist circumference were shown to be significant pre-(BMI, kilogram per square meter) subgroups (<40, ≥40) expressed in dictors of suboptimal weight loss (p<0.05). The three con-percentage excess weight loss (%EWL) and percentage excess BMI tinuous predictors correlated at a mean r00.75 (0.69–loss (%EBMIL) 0.79) and were incorporated, along with the dichotomous variable of “presence vs absence of general comorbidity,”EBMIL and %EWL correlated at a mean r00.97 (range into the multivariate model. Logistic regression indicat-0.96–0.99, p<0.0001) across time points. Figure 2 dem- ed that only preoperative BMI remained a significantonstrates the direct relationship between the two weight predictor after adjusting for covariates and was associ-loss outcome measures. Mean %EBMIL values for both ated with an 84.0 % overall accuracy rate in classifyinggroups consistently register slightly above mean %EWL val- patients into the “successful” or “suboptimal” weight lossues at each time point. categories correctly. Figure 3 portrays the results of the logistic regression analysis in the form of a probability Table 4 provides the results of an unadjusted bivariate curve. Since BMI was shown to be the lone, significantanalysis that investigated preoperative clinical variables predictor in the multivariate model (odds ratio01.90,and their possible value in predicting suboptimal weight 95 % CI 1.50, 2.50, p < 0.001), its beta coefficientloss at 6 months. Employing the median split proce- (0.6432) and associated model constant (−26.68) weredure, patients with a residual BMI of ≤35.9 were la- used to develop an equation that projected the likeli-beled “the successful weight loss group,” and patients hood of suboptimal weight loss as a function of patient BMIwith a BMI >35.9 at 6 months were labeled “the “suboptimal preoperatively. Diabetes Outcomes Six months following LGCP, 32 of 33 preoperatively dia- betic patients (96.9 %; p<0.0001, McNemar’s test of related samples) experienced resolution or improvement: T2DM was resolved in 18 patients (54.5 %); in 14 patients (42.4 %), there was a reduction in the number of medica- tions and improvement in metabolic blood markers. Mean HbA1C was 5.1±1.3, reduced from 6.4±1.4 (p<0.001); mean glycemic level (milligram per deciliter) was reduced to 112.6±38.8 from 162.8±62.7 (p<0.001) (Table 5). At 6 months, the diabetic subgroup BMI dropped 5.2 U to 36.4Table 4 Unadjusted bivariate Variable Successfula weight Suboptimala weight p valueanalysis of preoperative clinical loss group (N054) loss group (N050)variables with respect to weight Age (years) NSbloss success at 6 months Male sex 46.3 ± 9.8 48.8 ± 10.7 NSc Absolute weight (kg) 9 (16.7 %) 8 (16.0 %) <0.001bData are expressed as mean ± Body mass index (kg/m2) 109.2 ± 17.0 129.0 ± 19.8 <0.001bSD for continuous variables and Waist circumference (cm) 38.3 ± 3.1 45.2 ± 3.9 <0.05bN (percent) for categorical Systolic BP (mmHg) 113.9 ± 14.1 123.9 ± 14.8 NSbvariables Diastolic BP (mmHg) 140.0 ± 14.9 144.9 ± 17.4 NSbaWeight loss classification deter- Glycemia (mg/dL) 91.3 ± 9.9 90.8 ± 10.5 NSdmined by BMI median split; HbA1C (%) 122.5 ± 37.2 134.9 ± 59.0 NSdsuccess group 0 BMI≤35.9, General comorbidity 5.7 ± 1.6 6.0 ± 1.4 NScsuboptimal group 0 BMI >35.9 Specific comorbidity 37 (68.0 %) 40 (80.0 %)bIndependent samples t test Type 2 diabetes mellitus NSccFisher’s exact test Hypertension 12 (22.2 %) 20 (40.0 %) NScdIndependent samples Dyslipidemia 30 (56.6 %) 35 (70.0 %) NScMann–Whitney U test 11 (20.4 %) 8 (18.0 %)

1304 OBES SURG (2012) 22:1298–1307Fig. 3 The probability curve characterizes the likelihood of a minor complications was 27.8 % (n068). Postoperativepatient with a given preoperative body mass index (BMI, kilo- (12–24-h) nausea and/or vomiting occurred in less thangram per square meter) experiencing suboptimal weight loss (i.e., one third of patients. It was possible to control the symp-residual BMI >35.9) 6 months following laparoscopic greater toms with 5-HT3 receptor antagonist medication (ondanse-curvature plication (LGCP). For example, a patient with a preop- tron) without a prolonged hospital stay; patients wereerative BMI of 40.0 is predicted to have a 28 % chance of a residual discharged in 36 h. Forty-five of these patients requiredBMI >35.9 at 6 months, whereas a patient presenting with a BMI additional parenteral fluids for 6–12 h from symptom onset.of 45 is predicted to have a 90 % chance of a residual BMI >35.9 at Six patients were later readmitted for vomiting (at 4 and6 months 7 weeks and 2, 3, 4, and 8 months), and melena occurred in two patients. All were explored endoscopically: In two, a±3.8 from 41.5±4.2 (p<0.001), corresponding to a mean mucosal lesion was found, and in two, a bleeding ulcer at%EBMIL of 31.4±17.2 (25.2, 37.6) (Table 3). the lowest portion of the plication ridge in the antral region. The ulcers were treated with clips applied to the bleedingComplications vessel and patients received high doses of proton pump inhibitor medication and were discharged after 2 days ofThere was no mortality in the series. One intraoperative hospitalization. At 6-week follow-up, no evidence of ulcera-conversion from laparoscopy to open laparotomy was per- tion or bleeding remained.formed due to bleeding from a short gastric vessel duringgreater curvature dissection at the splenic region. The rate of The rate of major complication was 1.2 % (n03). A female patient with a prior Nissen fundoplication for reflux (not taken down during the LGCP procedure) suffered a perforated stomach immediately after discharge due to noncompliance with suggested food restrictions. The patient was not able to vomit, likely due to the intact Nissen fundoplication and the substantial increase in intragastric pressure inside the greater curvature plication. On emergency reoperation, the stitches were found to be broken and gastric leak and peritonitis were observed. The greater curvature plication was taken down and the perforation oversewn. The patient made a full recovery after a 10-day hospitalization. The second major complication was a gastric band pa- tient who underwent LGCP surgery to correct weight regain. During the LGCP, an abundance of fibrous tissue adherent to the band and scarring surrounding the band area wereTable 5 Weight loss and metabolic outcomes in diabetic patients at 6 months (N033)Outcome Value Mean ± SD (range) 6 months Mean change (% change) 95 % CIa p value PreoperativeAbsolute weight (kg) 121.3±23.6 (77.0–176.0) 106.7±20.1 (73.7–160.7) −14.6 (12.1) −18.0, −11.1 <0.001bBody mass index (kg/m2) 41.5±4.2 (31.2–58.8) 36.4±3.8 (27.1–42.3) −5.2 (12.5) −6.2, −4.1 <0.001bWaist circumference (cm) 131.1±14.8 (100.0–161.0) 118.5±14.6 (97.0–151.0) −12.6 (9.6) −21.6, −3.6 <0.05bSystolic BP (mmHg) 144.3±16.0 (118.0–183.0) 135.1±20.9 (100.0–163.0) −9.2 (6.4) −18.3, −0.1 <0.05bDiastolic BP (mmHg) 92.0±8.6 (74.0–115.0) 87.5±12.5 (67.0–106.0) −4.5 (4.8) −11.9, 3.0 NSbGlycemia (mg/dL) 162.8±62.7 (84.6–313.2) 112.6±38.8 (75.6–270.0) −50.1 (30.9) −70.2, −30.1 <0.001cHbA1C (%) 6.4±1.4 (3.9–12.0) 5.1±1.3 (3.5–8.0) −1.3 (20.3) −1.7, −1.0 <0.001cCI confidence interval, HbA1C glycosylated hemoglobin, BP blood pressurea 95% CI of the mean differenceb Paired samples t test assessing change from baselinec Related samples Wilcoxon-signed rank test

OBES SURG (2012) 22:1298–1307 1305observed. The band was removed and gastric plication was criteria may have suppressed weight loss results in ourperformed below the affected region. The postoperative cohort. Looking solely at the weight trend in patients withcourse was uneventful; however, 3 days following dis- preoperative BMI <40, %EWL approached 50 % atcharge, she returned with symptoms of peritonitis and was 18 months and %EBMIL exceeded 50 %. Overall, ourreoperated. A 3×3-cm area of partial stomach wall necrosis LGCP patient group demonstrated an EWL of 43.7 %below the original band location was identified and a wedge (37.1, 50.3) at 18 months, well within the 95 % CI of thestomach resection was carried out. The patient recovered reported meta-analytic mean of 47.5 % EWL (40.7, 54.2)slowly after a complicated 12-week postoperative course. based on 1,848 AGB patients at 1–2 years [22]. The third reoperated patient experienced recurrent left When recoded as a continuous variable, logistic regres-upper quadrant pain after episodes of excessive food intake sion analysis showed that preoperative BMI was predictiveand was readmitted 4 months after LGCP. Upper GI endos- of suboptimal weight loss. The related probability curvecopy showed a swollen area at the distal end of the plication may offer preliminary guidance for LGCP patients regard-ridge inside the stomach. Exploratory surgery revealed that ing realistic weight loss expectations at 6 months. Inthe end of the continuous inverting stitch was partially addition, in time, patients who fall into the predicted “sub-buried inside the stomach wall, suggesting early migration optimal weight loss” group at 6 months may realisticallyof the stitch likely triggered by plication distention from expect to achieve results comparable to the “successfulexcessive food intake (confirmed by the patient). The suture weight loss” group by 18 months. Overall, according towas excised and the lower half of the plication loosened. subgroup analysis, LGCP appears equally effective for allSerosal gastric suturing was used to cover the affected area. ages, both sexes, and for people with a range of comorbidThe patient was discharged with no further pain after 4 days conditions.of hospitalization. With respect to lumen size calibration, in our initial Four elective reoperations were performed to address LGCP patients, we noted the location of anatomical land-premature weight loss plateaus at 4, 6, 8, and 9 months marks in relation to the suture line, and also performedfollowing LGCP. These patients were satisfied with the gastroscopy during and at the end of the procedure, ensuringinitial LGCP outcome and requested maintenance of that an adequate lumen (measured by a 36–38F bougie) wastheir plication, if possible. Pre-revisional upper GI en- obtained. In the learning curve of the first 100 cases, wedoscopy revealed plication sutures loosely hanging in- progressed to using only the calibration bougie, moving itside the stomach and a diminished plication ridge inside gradually forward into the stomach lumen while advancingthe stomach, suggesting stomach dilatation. In all four the suture line toward the pylorus. In regard to performing apatients, the plicated stomach (including the greater one-row plication in the majority of cases, after observingcurvature) was free of adhesions, and loose sutures were the initial 50 cases with two plication rows, we hypothesizedfound at the primary plication. An additional suture line that one row would make the procedure faster and mightwas applied above the previous plication row. Patients cause fewer tissue perfusion complications. We noted nowere released after 24 h with no signs of nausea or vomiting. differences between the one- and the two-row series inPatients began losing weight again and have been satisfied weight loss or in complication rates and, thus, proceededwith their outcomes. to complete the series with single-row plications, decreasing operative time by 30 min.Discussion The current study’s significant T2DM metabolic out-Current evidence regarding LGCP is scant and mostly de- comes in 33 patients at 6 months (HbA1C, −1.3 %; glycemicscribed in very small series with few patients followed level, −50.1 mg/dL) mirrored those of AGB findings bybeyond 6 months. Although the population in the current meta-analysis (HbA1C, −1.2 %; glycemic level, −56.7 mg/study included patients with higher preoperative BMIs and dL [22]). With the exception of Talebpour and Amoli, whomore extensive comorbidity than most prior LGCP inves- reported resolution of 8/13 patients’ T2DM at 6 monthstigations [7, 17–21], our 6-month %EWL (31.8; 95 % CI [17], no other study has analyzed post-LGCP diabetes out-28.7, 34.9) approached the range of reported 6-month comes. The current study reports results for the largestresults (33.0–51.7 %) with only three patients lost to diabetic population (n033) following LGCP.follow-up in the series. The current study protocol wasamong the most inclusive with respect to BMI and comorbid LGCP has not been associated with mortality. The minordisease, encompassing patients with a variety of morbidities complication of postoperative nausea and/or vomiting af-(e.g., bladder disease, gallstones, diabetes of any tenure fected 27.8 % of the 244 patients in the current series, asor medication status); thus, the highly inclusive eligibility well as nearly all patients in the six prior LGCP studies, some severely, requiring additional treatment that resolved symptoms without prolonged hospitalization. Nausea and vomiting following the majority of LGCP procedures is an

1306 OBES SURG (2012) 22:1298–1307issue of significant concern to patients that can potentially Referencesbe improved by alterations to the procedure. 1. Karlsson J, Taft C, Ryden A, et al. Ten-year trends in health-related In two of the six prior LGCP studies, Sales Puccini quality of life after surgical and conventional treatment for severe[18] and Ramos et al. [19], no major complications obesity: the SOS intervention study. Int J Obes. 2007;31:1248–61.were reported. Major complications requiring reopera-tion in the remaining four LGCP studies ranged from 2. Dixon JB, Strauss BJ, Laurie C, et al. Changes in body composition3.0 to 15.4 %. In Talebpour and Amoli’s cohort (n0 with weight loss: obese subjects randomized to surgical and medical100), four complications (prepyloric perforation, suture programs. Obesity (Silver Spring, MD). 2007;15:1187–98.line leak, liver abscess, and stomach kinking) requiredreoperation [17]; Brethauer et al. (n09) had one case of 3. Sowemimo OA, Yood SM, Courtney J, et al. Natural history ofgastric obstruction that required reoperation to reduce morbid obesity without surgical intervention. Surg Obes Relat Dis.the plication [7]; Gebelli et al. (n013) saw two major 2007;3:73–7.complications that required surgery (one total dysphagia,one herniation of the suture line) [20]; and Skrekas et 4. Christou NV, Sampalis JS, Liberman M, et al. Surgery decreasesal. (n0135) reported three acute gastric obstructions and long-term mortality, morbidity, and health care use in morbidlyone partial jejunal necrosis due to portomesenteric obese patients [with discussion]. Ann Surg. 2004;240:416–24.thrombosis which were reoperated [21]. The reoperationrate in the current study was 1.2 % (3/244), related to a 5. Sjöström L, Narbro K, Sjöström CD, et al. Effects of bariatric surgeryprior Nissen fundoplication, a prior AGB, and a case of on mortality in Swedish obese subjects. N Engl J Med. 2007;suture migration triggered by excessive food intake. Our 357:741–52.experience suggests that previous surgery that may limita patient’s ability to vomit should be considered a relative 6. Buchwald H. External gastric plication. In: Buchwald’s atlas ofcontraindication to subsequent LGCP. In our opinion, the metabolic & bariatric surgical techniques and procedures. Phila-most critical aspect of LGCP is stomach wall blood delphia: Saunders Elsevier; 2012. p. 287–92.supply. During LGCP, the highly vascular fundus must bedevascularized and infolded, providing an opportunity 7. Brethauer SA, Harris JL, Kroh M, et al. Laparoscopic gastricfor blood supply compromise and potential complication. plication for treatment of severe obesity. Surg Obes Relat Dis.This is one of the key reasons that we do not recommend 2010;7(1):15–22.LGCP after failed gastric banding. Complications are likelyto be further reduced, and the significant LGCP learning 8. Neumayer C, Ciovica R, Gadenstatter M, et al. Significant weightcurve shortened, as the procedure becomes increasingly loss after laparoscopic Nissen fundoplication. Surg Endosc.standardized. 2005;19:15–20. Limitations of this study include the lack of a con- 9. Buchwald H, Buchwald JN. Evolution of surgery for morbidtrolled comparison group and attenuated follow-up. In obesity. In: Pitombo C, Jones KB, Higa KD, Pareja JC, editors.general, LGCP is at the outset of its investigational Obesity surgery: principles and practice. New York: McGraw-Hillcourse. Further research is required to determine appro- Medical; 2007. p. 3–14.priate indications for LGCP, the value of one vs multipleplication rows, the optimum mechanism for tissue apposition 10. Cummings DE, Foster-Schubert KE, Carlson MJ, et al. Possible(e.g., suture, staple, or other clasp), the most effective plication hormonal mechanisms mediating the effects of bariatric surgery.length, and the gastric lumen size that minimizes postopera- In: Pitombo C, Jones KB, Higa KD, Pareja JC, editors. Obesitytive edema and optimizes long-term weight loss. surgery: principles and practice. New York: McGraw-Hill Medical; 2007. p. 137–47. With safety preliminarily established, larger cohorts ofpatients should be followed over the longer term. The current 11. Cummings DE, Overduin J. Gastrointestinal regulation of foodstudy found that, in the short term, laparoscopic greater cur- intake. J Clin Invest. 2007;117(1):13–23.vature plication results in effective weight loss and significantT2DM reduction with a very low rate of complications. 12. Kirk RM. An experimental trial of gastric plication as a means of weight reduction in the rat. Br J Surg. 1969;56(12):930–3.Conflict of interest M. Fried has a consulting agreement with, and isan investigator for, Johnson & Johnson (J&J). J.N. Buchwald and T.W. 13. Fusco PE, Poggetti RS, Younes RN, et al. Evaluation of gastricMcGlennon are CRO employees under contract with J&J. G. Ribaric is greater curvature invagination for weight loss in rats. Obes Surg.an employee of J&J. The study was financially supported by Ethicon 2006;16(2):172–7.Endo-Surgery Europe, GmbH, a subsidiary of J&J. 14. Fusco PE, Poggetti RS, Younes RN, et al. Comparison of anterior gastric wall and greater gastric curvature invaginations for weight loss in rats. Obes Surg. 2007;17(10):1340–5. 15. Menchaca H, Harris J, Thompson S, et al. Gastric plication: pre- clinical study of durability of serosa-to-serosa apposition. Surg Obes Relat Dis. 2011;7:8–14. 16. Clinical Issues Committee, American Society for Metabolic and Bariatric Surgery (ASMBS). Policy statement on gastric plication, March 8, 2011. Surg Obes Relat Dis. 2011;7(3):262. 17. Talebpour M, Amoli BS. Laparoscopic total gastric vertical plica- tion in morbid obesity. J Laparoendosc Adv Surg Tech A. 2007; 17:793–8. 18. Sales Puccini CE. Surset gástrico de Sales: una alternative para cirugía bariátrica restrictive. Rev Colomb Cir. 2008;23(3):131–5. 19. Ramos A, Galvao Neto M, Galvao M, et al. Laparoscopic greater curvature plication: initial results of an alternative restrictive bari- atric procedure. Obes Surg. 2010;20:913–8. 20. Gebelli JP, de Gordejuela A, Badía A, et al. Laparoscopic gastric plication: a new surgery for the treatment of morbid obesity. Cir Esp. 2011;89(6):356–61. 21. Skrekas G, Antiochos K, Stafyla VK. Laparoscopic gastric greater curvature plication: results and complications in a series of 135 patients. Obes Surg. 2011;8(9):10–5.

OBES SURG (2012) 22:1298–1307 130722. Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a Adopted by the 18th WMA General Assembly, Helsinki, Finland, systematic review and meta-analysis. JAMA. 2004;292(14):1724–37. June 1964, amended by the 59th WMA General Assembly, Seoul,23. Brethauer SA, Hammel JP, Schauer PR. Systematic review of October 2008. World Med J. 2008;54(4):122–5. sleeve gastrectomy as staging and primary bariatric procedure. Surg 27. Dixon JB, McPhail T, O’Brien PE. Minimal reporting require- Obes Relat Dis. 2009;5:469–75. ments for weight loss: current methods not ideal. Obes Surg. 2005;15:1034–9.24. Mechanick JI, Kushner RF, Sugerman HJ, et al. AACE/TOS/ 28. American Society for Bariatric Surgery Standards Committee, ASMBS Guidelines: American Association of Clinical Endocrinol- 2004–2005. Guidelines for weight calculations and follow-up in ogists, The Obesity Society, and American Society for Metabolic & bariatric surgery. Surg Obes Rel Dis. 2005;1:67–8. Bariatric Surgery Medical Guidelines for clinical practice for the 29. Deitel M, Gawdat K, Melissas J. Reporting weight loss 2007. Obes perioperative nutritional, metabolic, and nonsurgical support of the Surg. 2007;17(5):565–8. bariatric surgery patient. Surg Obes Relat Dis. 2008;4:S109–84. 30. Miller MA. A calculated method for the determination of ideal body weight. Nutr Support Serv. 1985;5(3):31–3.25. Consensus Development Conference Panel. NIH conference: gastroin- 31. Metropolitan Height and Weight Tables, 1993. Metropolitan Life testinal surgery for severe obesity. Ann Intern Med. 1991;115:956–61. Foundation. Stat Bull. 1983;64(1):2–9.26. World Medical Association (WMA). Declaration of Helsinki— ethical principles for medical research involving human subjects.

OBES SURG (2012) 22:572–581DOI 10.1007/s11695-011-0508-9 CLINICAL RESEARCHProspective, Multicenter, 3-Year Trial of LaparoscopicAdjustable Gastric Banding with the MIDBAND™Christian Gouillat & Angélique Denis & Perrine Badol-Van Straaten & Vincent Frering &Jacques Tussiot & Pierre Campan & Gérard Aulagnier & Valeria Costamagna &Jean-françois Ain & Robert Portet & Michel Roche & Celestin Esso &Jean-Marie Molasoko & Yves Claret & Jacques Desplantez & Sébastien Le Page &Marie Cécile Blanchet & Maud Robert & Isabelle Jaisson-HotPublished online: 26 August 2011# Springer Science+Business Media, LLC 2011Abstract safety and effectiveness outcomes using the MID-Background Although laparoscopic adjustable gastric BAND™ (MID, Dardilly, France).banding (LAGB) is a popular metabolic/bariatric proce- Methods Between May 2005 and September 2006, 262dure, few prospective studies have assessed its out- morbidly obese patients underwent primary gastric bandingcomes. This study aimed to prospectively assess LAGB with pars flaccida technique in 13 French medical centers.C. Gouillat : S. Le Page : M. C. Blanchet : M. Robert J.-f. Ain Polyclinique du Val de Saône,Centre de chirurgie de l’obésité, Hospices Civils de Lyon, 71000 Mâcon, FranceUniversité de Lyon,69 317 Lyon, France R. Portet Clinique de l’Occitanie,A. Denis : P. Badol-Van Straaten : I. Jaisson-Hot 31600 Muret, FranceHospices Civils de Lyon, Pôle Information Médicale Evaluation M. RocheRecherche, Clinique Mutualiste,Lyon 69424, France 01500 Amberieu, FranceA. Denis : P. Badol-Van Straaten : I. Jaisson-Hot C. Esso Polyclinique du parc,Université Lyon, 08000 Charleville-Mézières, FranceEquipe d’Accueil 4129,Lyon, France J.-M. Molasoko Clinique du Mantois,V. Frering 78200 Mantes-la-Jolie, FranceClinique de la Sauvegarde,69009 Lyon, France Y. Claret Clinique de la Dhuys,J. Tussiot 93170 Bagnolet, FranceClinique de l’Estrée,93200 Stains, France J. Desplantez Polyclinique Marzet,P. Campan 64075 Pau, FranceHôpital de La Conception, Université de Marseille,13000 Marseille, France C. Gouillat (*) Service de chirurgie digestive, Hôpital de la Croix-Rousse,G. Aulagnier 103, Grande Rue de la Croix-Rousse,Centre Hospitalier Lucien Hussel, 69317 Lyon, Cedex 04, France38000 Vienne, France e-mail: [email protected]. CostamagnaClinique Gentilly,54270 Essey les Nancy, France

OBES SURG (2012) 22:572–581 573Excess weight loss and change in body mass index (BMI, CA, USA) or Swedish Adjustable Gastric Band® (SAGB®)kilogram per square meter), percentage of patients with (known as the Realize® Band in the USA) (Ethicon Endo-comorbidities, and obesity-related complications were Surgery, Inc. [EES], Cincinnati, OH, USA). Outcomes forrecorded. Patients were followed at 6-month intervals for both of these gastric bands appeared to be comparable in a3 years. A multivariable individual growth model was used systematic review of meta-analyses [7]. Bands of ato analyze weight change over time and determine potential somewhat different design, with less rigid physical charac-predictors of weight loss. teristics, are currently available but have yet to beResults The majority of patients were female (n=233, 89%), sufficiently assessed clinically. The goal of this study waswith mean age of 36.4±9.7 years. At 3 years, LAGB with to prospectively assess the safety and efficacy of one ofMIDBAND resulted in significant decrease in mean BMI these products, the soft, low-pressure MIDBAND™ (MID,from 41.8±4.2 to 30.7±5.8 (p<0.0001). Median excess Dardilly, France, introduced in 2000), in routine practiceweight loss and excess BMI loss were 61% and 68%, across a variety of medical centers.respectively. The prevalence of obesity-related comorbiditieshad significantly decreased from 71% to 15% (p<0.0001). MethodsComplications were observed in 26 patients (10%); device-related complications occurred in 20 patients (8.2%), Study Design, Setting, and Conductrequiring band removal in 8 (3.3%), and port revision in8 (3.3%). Individual growth analysis identified significant The study was designed in 2004 as a protocol-driven, pilot,predictors of weight loss including the number of follow-up multicenter, prospective study. Consecutive patient series (n=visits. 262) underwent LAGB as a primary procedure in 13 FrenchConclusion Prospective outcomes demonstrate the safety centers (2 academic centers, 1 public general hospital, 10and efficacy of gastric banding over time using the private clinics). Three centers included between 40 and 43MIDBAND. Individual growth modeling demonstrated that patients each, the others included between 6 and 21 patientspostoperative weight loss is strongly related to the each. The study was approved by an ethics committee tofrequency and consistency of follow-up visits. ensure patients’ welfare and the ethical conduct of the study. Written informed consent was obtained from all patientsKeywords Laparoscopic adjustable gastric banding . before surgery.Obesity . Bariatric surgery . Body mass index . MIDBAND .Multivariable mixed effects Inclusion and Exclusion CriteriaBackground Patients considered for LAGB were required to meet the French guidelines for bariatric surgery (i.e., age 18–Laparoscopic adjustable gastric banding (LAGB) is a very 60 years, body mass index [BMI; kilogram per squarepopular weight loss procedure; it has become one of the meter] ≥40.0 or BMI >35.0 to <40.0 with significantbariatric procedures most frequently performed worldwide medical comorbidities, a history of obesity of more than[1]. In routine practice and in the surgical literature, LAGB 5 years, failed conservative treatment for more than 1 year,has been associated with significant excess weight loss of and a multidisciplinary preoperative preparation) [16].approximately 50% to 61% at 2–3 years and improvement Patients who did not meet all criteria for inclusion werein obesity-related comorbidities [2–6] together with a low excluded, as well as those with a life-threatening orrate of severe complications and an acceptable rate of minor psychiatric disease, an addiction to alcohol or drugs, thosecomplications [7–9]. Recently, some authors reported poor who had undergone prior bariatric surgery, and those wholong-term results regarding both LAGB weight loss and were deemed unable to meet the follow-up requirements, orcomplications [10–12]. As a probable result, a decrease has who did not speak and read French.been observed in the number of LAGB procedures inEurope [1], where the procedure was introduced in 1993 Data Collection and Primary and Secondary Endpoints[13, 14]. In the USA, where LAGB was not available until2001 [15], the procedure continues to be performed with All data were collected prospectively. Gender, age, height,increasing frequency [1]. weight, waist circumference, and obesity-related comorbid- ities were recorded preoperatively. Intraoperative complica- Despite the widespread use of LAGB, few prospective tions and postoperative data, including length of hospitaldata regarding outcomes of this procedure have been stay, adverse symptoms and events, additional hospital stay,published. In addition, nearly all LAGB studies have been reoperation, and band adjustment data were collected. Allcarried out with the LAP-BAND® (Allergan Inc., Irvine,

574 OBES SURG (2012) 22:572–581reported complications and adverse events were assessed by examining weight separately has yielded more accurate andthe scientific committee in the context of patients’ charts. better-fitting multivariable models [19]. The effect of time was centered on initial weight status before surgery. Weight The primary endpoint was change in weight, presented change across time was represented in the model in termsas change in body weight (kilogram), BMI, excess weight, of polynomials of increasing order to allow for linear,and excess BMI. Secondary endpoints were change in waist quadratic, and cubic components of developmental change.circumference, change in obesity-related comorbidities, Whenever possible, for all trend components, randomcomplications, adverse upper gastrointestinal (GI) symp- effects were allowed to account for individual differencestoms, and patient’s overall satisfaction. across time. Linear and quadratic time effects were included both as fixed and as random effects, but cubic time effectSurgical Technique and Follow-up was fitted only as fixed, to facilitate model convergence.LAGB was performed via standardized pars flaccida Potential predictors of weight loss were tested. Obesity-technique [17]. Patients were scheduled for follow-up every related comorbidities, adverse upper GI symptoms, and the6 months over the course of 3 years. There was no special band fill volume were considered time-varying predictors.requirement regarding aftercare program and diet. Band Age at entry into the study, gender, height, number ofadjustments were performed as deemed appropriate accord- follow-up visits, and location of the surgical center wereing to the discretion of surgeons at individual centers, considered as time-invariant predictors. Time-varying pre-throughout the 3 years. At the end of each clinic visit, dictors were included only as fixed effects due to apatients responded to a protocol-based physician-to-patient convergence problem in the model. Longitudinal weightquery regarding their overall satisfaction with the LAGB data on patients were nested within centers; however, weprocedure. considered center effect only as an individual variable. Although, fitting a three-level individual growth model wasStatistical Analysis considered, it was decided that it added unnecessary complexity to the model. In fact, the estimated between-Data were compiled and entered into a specifically designed center variance was not significant according to a randomdatabase using Microsoft Access 2010® (Microsoft, Inc., effects analysis of the three-level unconditional meansRedmond, WA, USA). All analyses were performed in model (p=0.3148).Statistical Analysis System (SAS®) software version 9.1(SAS Institute, Cary, NC, USA). Continuous data were The multilevel model was fit using the SAS PROCpresented as means and standard deviations and medians MIXED application [20]. Parameter estimates wereand ranges, where appropriate; categorical data were pre- obtained by the full maximum likelihood method. Nestedsented as frequencies and percentages. Analyses were models were compared using likelihood-ratio tests and non-performed on all data available at each follow-up visit. For nested models were ranked according to their Akaike18 women who were found to be pregnant during follow-up, information criterion (AIC) values. An F-statistic was usedweight data collected during their pregnancy and 6 months to test for significance of fixed effects and 95% confidenceafter their childbirth were excluded from the analysis of intervals (CIs) were calculated for each parameter estimate.weight loss. Results To detect differences between preoperative and 3-yearpostoperative patient characteristics, McNemar’s test was Preoperative patient characteristics of the 262 patientsused for paired categorical data and the Wilcoxon signed included in the study are shown in Table 1. A majority ofrank test was used for paired continuous data. A two-sided patients were female (n=233, 89%), with mean age of 36.4±p value of 0.05 was considered statistically significant. 9.7 years. The pars flaccida operative technique was used in all patients. Anterior gastric fixation sutures were placed in Individual growth modeling (i.e., a special case of 33% of patients. No intraoperative complication wasmultilevel modeling grounded in the mixed effects general observed. There was no conversion from laparoscopiclinear model) was used to assess weight loss over time after technique to an open procedure. Mean operative time wassurgery while controlling for several variables. This 41.0±35 min (median 30; range 12–300), and mean hospitalstatistical technique facilitated focus on individual changes stay was 3.0±1 days (median 3; range 2–12).in weight over time and on the manner in which individualchanges vary across patients [18]. Weight (kilogram) over The mean length of follow-up was 30.0±9 months.time was used as the dependent outcome variable rather Twenty patients (7.6%) were lost to follow-up early in thethan percentage excess weight loss (%EWL) or percentage postoperative course; at the conclusion of the study, a totalexcess body weight loss (%EBL) in order to exclude the of 69 patients (26%) were lost to follow-up.controversial concept of ideal body weight and because

OBES SURG (2012) 22:572–581 575Table 1 Baseline characteristics Characteristics Value (n=262) Median (range)of patients Age, mean ± SD, years 36.4 ± 9.7 36.0 (18.0–61.0) Women, no. (%) 233 (89) – Waist circumference, mean ± SD, cm 121.8 ± 18.9 120.0 (90.0–150.0) Body weight, mean ± SD, kg 112.7 ± 16.5 111.0 (78.0–191.0) Excess body weight, mean ± SD, kg 51.6 ± 13.8 50.0 (24.0–121.0) Body mass index, mean ± SD, kg/m2 (range) 42.2±4.5 (35.0–60.3) 41.3 (35.0–60.3) Total patients with comorbidities, no. (%) 185 (71) – Comorbidities per patient, mean ± SD 1.3 ± 1.2 1.0 (0–5) Type 2 diabetes mellitus, no. (%) 31 (12) – Change in weight loss is presented in Figs. 1, 2, and 3. Complications (Table 2) were observed in 26 patientsAt 3 years, LAGB with the MIDBAND resulted in a (10%), including device-related complications in 20 (8.2%).significant decrease in mean BMI from 41.8±4.2 (median Band-related complications occurred in 12 patients (4.9%),41.1, range 35.0–54.6) to 30.7±5.8 (median 30.6, range resulting in band removal in 8 (3.3%) after a median17.6–45.9; p<0.0001) for the 132 patients who were still postoperative period of 31 months (range 0–37). Port-followed-up. BMI was <35.0 in 77% of patients and <30.0 related complications required minor reoperation as a dayin 47% of patients. Median excess weight loss (EWL) and case in eight patients (3.3%) after a median postoperativeexcess BMI loss (EBL) were 61% and 68%, respectively. period of 10 months (range 1–38). No port infection wasAt 3 years, the mean waist circumference had decreased observed.from 123.1±21 to 96.3±14 cm (p<0.0001). Eleven patients (4.4%) had additional surgery unrelated The change in comorbidities is presented in Fig. 4. At to LAGB complications (i.e., post-bariatric abdominoplasty,3 years, the overall percentage of patients with obesity- 6 [2.4%]; cholecystectomy, 2 [0.8%]; hernia repair, 2related comorbidities and type 2 diabetes mellitus (T2DM) [0.8%]; varicose vein surgery, 1 [0.4%]). Three patientshad significantly decreased from 71% to 15% (p<0.0001) (1.2%) with insufficient weight loss were considered forand from 12% to 4% (p<0.001), respectively. conversion to an alternative bariatric surgery. One hundredFig. 1 Individual changesin weight loss

576 OBES SURG (2012) 22:572–581Fig. 2 Change in mean BMI(empty circle) and percentage ofpatients with BMI <35 (filledtriangle with solid line) andBMI <30 (filled triangle withdashed line) during the 3-yearfollow-upand thirteen patients (46%) reported at least one adverse in weight loss occurred, as indicated by the significantupper GI symptom at least one time: regurgitations (71; positive quadratic time effect (0.12; 0.09, 0.15; p<0.001),28.9%), reflux or heartburn (50; 20.4%), or dysphagia (45; followed by a further weight stabilization (negative cubic18.4%). time effect, −0.0012; −0.0018, −0.0006; p<0.001). Two hundred and thirty-four patients (97%) received at The significant predictors of weight loss included in theleast one band adjustment (median=4, range 1–14). The final model were age at entry into the study, height, adversefirst adjustment was typically performed between 4 and upper GI symptoms, obesity-related comorbidities, fill5 months. At 3 years, the mean band fill volume was 3.8± volume, number of follow-up visits, and the center where2.5 ml. In addition, 95% of patients stated that they were surgery was performed. Gender was not included in thesatisfied with the operation. model since no significant difference was found between males and females on initial weight status and on weight The results of multivariable analysis of longitudinal loss.weight loss are presented in Table 3. In the earlypostoperative period, the instantaneous rate of change in After controlling for all independent variables in the model,weight was negative (−3.87; 95% CI, −4.35, −3.40; p< results were the following: Patients older than the mean age0.001). After the period of rapid weight loss, a deceleration (36 years) weighed less at the beginning of the study (negativeFig. 3 Change in mean excessweight (dashed line) andexcess BMI loss (solid line)during the 3-yearfollow-up

OBES SURG (2012) 22:572–581 577Fig. 4 Change in percentage ofpatients with obesity-relatedcomorbidities during the 3-yearfollow-up: −change in percent-age of patients having at leastone comorbidity (filled squarewith solid line); −change inpercentage of patients havingvarious main obesity-relatedcomorbidities (dotted lines)age effect, p<0.01) and lost weight more slowly over time tion, p<0.001; positive height-by-quadratic time interaction,than younger patients (positive age-by-linear time interac- p<0.01; negative height-by-cubic time interaction, p<0.05).tion, p<0.01). Patients taller than the mean height (163 cm) Patients with and without at least one adverse upper GIweighed more at the beginning of the study (positive height symptom did not significantly differ in preoperative weighteffect, p<0.001) and lost weight faster over time than shorter (p=0.161); however, those with at least one upper GIpatients (significant negative height-by-linear time interac- symptom lost significantly more weight over time (negative adverse upper GI symptoms-by-linear time interaction,Table 2 Complications p < 0.01).Complication Number (%) The coefficient associated with comorbidities was 0.94 (0.16, 1.72; p<0.05), indicating that the presence of at leastBand-related 12 (4.9)a one comorbidity was a significant predictor of weight loss Slippage 5 (2.0) over time. However, no significant interactions were noted Intolerance 5 (2.0) between comorbidities and linear, quadratic, or cubic time Erosion 1 (0.4) effects, suggesting that the trajectory of weight change over Malfunction 1 (0.4) time did not differ in patients with at least one comorbidityPort-related 8 (3.3)b compared to those without. The band fill volume was a Port twist 6 (2.5) significant predictor of weight loss over time according to Pain at port site 2 (0.8) the negative coefficient associated with this variable (−0.59;Non-device-related 6 (2.5) −0.93, −0.25; p<0.001). Significant fill volume-by-linear, Pulmonary embolism 1 (0.4) quadratic, and cubic time interactions indicated that the Deep venous thrombosis 1 (0.4) trajectory of weight change during the 3 years differed Pulmonary infection 1 (0.4) depending on fill volume. Greater weight loss was Port site hernia 2 (0.8) associated with greater fill volume, but only during the Surgical site problem 1 (0.4)c first follow-up visits, as indicated by the positive fill volume- by-linear time interaction (p<0.001) and the negative fillTotal N=242 (20 patients with incomplete data) volume-by-quadratic time interaction (p<0.001). Althougha Reoperation including band removal no difference in preoperative weight was found betweenb Port revision under local anesthetic patients who completed the full follow-up program and thosec Reoperation for suture needle left in the surgical site who did not, those who did not complete the full program

578 OBES SURG (2012) 22:572–581Table 3 Multivariable mixed model of weight loss after bariatric surgery (262 patients, 1,697 observations) 95% confidence interval Parameter estimate (SE)Fixed effects 111.4 (0.82)* 107.7, 113.0Intercept −3.87 (0.24)* −4.35, −3.40Time (linear term) 0.12 (0.016)* 0.09, 0.15Time² (quadratic term) −0.0012 (0.0003)* −0.0018, −0.0006Time3 (cubic term) 0.94 (0.40)** 0.16, 1.72Comorbidities 1.17 (0.84) −0.47, 2.81Upper GI symptoms −0.59 (0.17)* −0.93, −0.25Filling volume (ml) −0.22 (0.08)*** −0.38, −0.07Agea 1.37 (0.10)* 1.17, 1.58Heightb −0.009 (0.001)* −0.012, −0.006Number of follow-up visits × time² 0.0002 (0.00004)* 0.0001, 0.0003Number of follow-up visits × time3 −0.13 (0.04)*** −0.21, −0.05Upper GI symptoms × time 0.38 (0.04)* 0.30, 0.45Filling volume × time −0.02 (0.003)* −0.030, −0.019Filling volume × time² 0.0004 (0.00005)* 0.0003, 0.0005Filling volume × time3 0.01 (0.003)*** 0.004, 0.016Agea × time −0.04 (0.01)* −0.07, −0.02Heightb × time 0.002 (0.0007)*** 0.0007, 0.004Heightb × time² −0.00003 (0.000014)** −0.00006, −8.10−6Heightb × time3Random effects 15.8 (0.72)* 14.5, 17.3Level 1: residualLevel 2 141.6 (13.5)* 118.6, 172.2 Intercept 0.70 (0.09)* 0.55, 0.92 Linear slope (time) 0.0005 (0.00007)* 0.0004, 0.0006 Quadratic slope (time2) −0.78 (0.85) −2.44, 0.89 Covariance intercept−linear slope −0.01 (0.02) −0.06, 0.03 Covariance intercept−quadratic slopeGoodness of fit 11,633.5 –−2LL 11,647.5 –AICComorbidities and upper gastrointestinal symptoms were coded 0 = none, 1 = at least one. For space reasons, center × time, center × time², andcenter × time3 interaction significant effects (p<0.05) are not presented, but were also estimated and are available on request. The center includingthe highest number of patients was chosen as the referenceSE standard error, GI gastrointestinal, −2LL −2 log likelihood, AIC Akaike criterion information*p<0.001; **p<0.05; ***p<0.01a Age was rescaled to center around the mean (36 years)b Height was rescaled to center around the mean (163 cm)c Number of follow-up visits centered at the maximum of follow-up visits (10 visits)achieved less weight loss over time (negative interaction significantly different across the 13 participating centersbetween number of follow-up visits and quadratic time (p < 0.0001).effect, p<0.001; positive interaction between number offollow-up visits and cubic time effect, p<0.001). In addition, Regarding the random effects model, covariance betweenthe trajectories of weight change over time were significantly intercept and slope (−0.78; −2.44, 0.89) and covariancedifferent across the 13 study centers (center effects are not between intercept and curvature (−0.01; −0.06, 0.03) werepresented in Table 3). One-way analysis of variance not significant (p=0.3591 and p=0.5863, respectively),(ANOVA) indicated that attendance at clinic visits was indicating that preoperative weight was not associated with postoperative weight change over time.

OBES SURG (2012) 22:572–581 579Discussion model approach are readily comparable. Indeed, if certain significant predictor effects on weight change found usingOur prospective safety and efficacy outcomes parallel those traditional bariatric analytic techniques are also found usingLAGB outcomes reported in studies of the two most widely individual growth modeling, the data are confirmed withused gastric bands (Swedish Adjustable Gastric Band®/ greater statistical power. By contrast, if a predictor variableRealize® Band and the LAP-BAND®). In the literature, a is not significant in the multivariable model compared to50% to 60% excess weight loss is usually achieved with traditional repeated-measures techniques, additional re-LAGB [2–6]. In our series, at 3 years, the mean EWL and search using both techniques is then needed to attempt toexcess BMI loss was >60%, and nearly 80% of patients resolve this discrepancy.demonstrated a BMI <35.0, whereas nearly 50% of patientswere no longer obese. Follow-up for this study was limited In this study, individual growth analysis confirmed withto 3 years; long-term deterioration of this weight loss is greater sensitivity what traditional cross-sectional analysispossible, as has been observed by some authors [10–12]. has strongly suggested (Figs. 2 and 3). Weight loss wasHowever, these authors reported a high rate of band rapid during the first postoperative months, continued moreremoval (approximately 30% to 60%), a factor that may slowly after 6 months, and stabilized after 1 year. Theexplain the poor late results regarding weight loss. Where statistical technique also permitted an assessment of theavailable, reports of the two other primary bands’ long-term factors that had an effect on postoperative weight changeweight loss results are good in those patients who retain over time. The current study confirmed the negative effecttheir band [10]. Recent improvements in surgical technique of age on weight loss after LAGB, as previously found byand in band engineering [9] are likely to provide more others using traditional cross-sectional analysis [23].realistic and positive, long-term gastric band results. Inaddition, many of the studies of other bands that report very Patients with a high BMI are often considered to carry agood long-term results emphasize the importance of the high risk of insufficient weight loss following LAGB, asfollow-up in achieving and maintaining weight loss [21]. suggested by some observational studies [23]. However, in studies using mixed effects modeling, initial BMI has been We used individual growth modeling to assess longitu- shown to be less accurate in predicting weight loss than thedinal weight loss. This relatively modern statistical method separate variables of initial weight and height [19].is increasingly considered the most appropriate tool for Similarly, in our study, preoperative weight did notanalyzing longitudinal data, although at the current time, it demonstrate any effect on weight trajectories, and patients’is still rarely used to assess bariatric surgery outcomes [19, height was an independent positive predictor of weight loss.22]. This method overcomes some of the limitations of One must note, however, that only 18 patients (7%) had atraditional repeated-measures techniques and offers addi- BMI >50 in our series. In addition, band fill volume and thetional benefits and information. Whereas bivariate analysis presence of adverse upper GI symptoms were associatedor repeated-measures ANOVA require balanced data, with greater weight loss, possibly indicating that somewherein all individuals are measured at each time point, patients, and physicians, prioritize weight loss over comfortindividual growth modeling maintains statistical power and may tolerate significant adverse symptoms in order toeven in cases of missing data and irregularly spaced achieve weight loss.measurements across time (e.g., due to missed clinic visits).In addition, unlike with traditional statistical approaches, an The main contribution of our study pertains to theindividual growth model provides estimates of the average significant effect of follow-up on weight loss over time. Thechange trajectory as well as individual trajectories, allowing analysis demonstrated that achieving the complete plannedassessment of inter-individual differences in intra-individual follow-up resulted in better postoperative weight loss. Thechange. In bariatric surgery outcomes, multivariable mixed importance of follow-up on weight loss after LAGB haseffects models facilitate a more precise analysis of change long been suggested and was recently underscored [21, 24].in weight loss over time by statistically controlling for time- But to our knowledge, these outcomes represent the firstinvariant predictors (e.g., gender, age, height, number of time that the effect of the frequency and consistency offollow-up visits, location at which surgery was performed) follow-up on postoperative LAGB weight loss has beenas well as for time-varying predictors (e.g., comorbidities, demonstrated with the sensitivity of multivariable mixedfill volume, adverse upper GI symptoms). effects modeling. The center effect observed in our study may be explained by the large variability observed between Although the mixed model is not yet the standard centers in the frequency and consistency of their follow-upanalysis, we believe that the longitudinal aspects of process, which jibes with the conclusion of a Frenchbariatric surgery outcomes can best be assessed using this nationwide survey that demonstrates that high-volumemethodology. In addition, it should be emphasized that centers achieve the best weight loss results [23]. Additionalresults from both traditional and the more novel, mixed predictors of weight loss not collected in the current study (e.g., psychological disorders, compliance with dietary

580 OBES SURG (2012) 22:572–581instructions) would likely improve the accuracy of our longitudinal studies, including more patients and anmodel. extended follow-up period, are needed to assess long-term outcomes using the MIDBAND. In our series, LAGB resulted in a significant decrease inthe prevalence of obesity-related comorbidities and in Acknowledgments We would like to thank J. N. Buchwald,complete resolution of T2DM in more than 80% of patients Chief Bariatric/Metabolic Surgery Writer & Director of Scientific(i.e., 25/31). Insulin resistance was not assessed in our Writing and Publications, Medwrite Medical Communications,series; however, the significant decrease in waist circum- LLC, Wisconsin, USA, for research writing assistance in manu-ference to a value below the upper limit of the normal range script development; T. W. McGlennon, M3 LLC, Wisconsin, USA,suggests that LAGB with the MIDBAND is associated with for statistical consulting; and M. Perceval, Réseau EMERA, Lyon,a cure of insulin resistance, as has been observed in other France, for performing database management and researchstudies of metabolic surgery [25]. coordination of the study. Due to the mechanical restriction of the gastric band, a Conflict of Interest This work was supported by a grant from M.I.dramatic change is imposed in the way patients can eat. In D., Dardilly, France; C. Gouillat received consulting fees from MID;our series, 46% of patients reported at least once that they V. Frering received royalties and consulting fees from MID. A. Denis,experienced at least one upper GI symptom. Regurgitation P. Badol-Van Straaten, J. Tussiot, P. Campan, G. Aulagnier, V.and reflux were the most often reported symptoms and Costamagna, J.F. Ain, R. Portet, M. Roche, C. Esso, J.M. Molasoko,dysphagia was rare (18%). In a recently published cross- Y. Claret, J. Desplantez, S. Le Page, M.C. Blanchet, M. Robert, I.sectional study, the inability to consume foods of different Jaisson-Hot declare that they have no conflict of interest.textures was cited as the most troublesome problem; 67%of patients reported regurgitation, and 40% experienced Referencesreflux [26]. 1. Buchwald H, Oien DM. Metabolic/bariatric surgery worldwide We divided complications into device-related and non- 2008. Obes Surg. 2009;19:1605–11.device-related complications. Despite the prospective de-sign of our study, device-related complications were 2. Dixon JB, O’Brien PE, Playfair J, et al. Adjustable gastricobserved in only 10% of patients. Of particular interest, banding and conventional therapy for type 2 diabetes: athe band-related complication rate was just 5%; slippage, randomized controlled trial. JAMA. 2008;299:316–23.and above all, band erosion, were very rare. The resultinglow band removal rate of 3% may also be a contributing 3. Chapman AE, Kiroff G, Game P, et al. Laparoscopic adjustablefactor in the positive weight loss results. gastric banding in the treatment of obesity: a systematic literature review. Surgery. 2004;135:326–51. Some authors have reported a very high long-term rateof complications [10, 11]. Several recent papers reported 4. O’Brien PE, McPhail T, Chaston TB, et al. Systematic review oflow rates of complications when using the pars flaccida medium-term weight loss after bariatric operations. Obes Surg.approach and a new generation of gastric band [20, 27–29]. 2006;16:1032–4.The current study’s low complication rate, corroboratingtwo retrospective studies using the MIDBAND [28, 29], 5. Spivak H, Hewitt MF, Onn A, et al. Weight loss and improvementmight be related to the physical characteristics of the of obesity-related illness in 500 U.S. patients following laparo-MIDBAND, namely, its supple profile and low-pressure scopic adjustable gastric banding procedure. Am J Surg.design, in contrast with the earlier LAP-BAND design. 2005;189:29–32.Another factor that might influence complication rate is the“soft” management of adjustments employed with the 6. Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: aMIDBAND relative to the number of adjustments and band systematic review and meta-analysis. JAMA. 2004;292:1724–37.fill volume. Excessive tightening of the band may lead toincreased risk of migration and slippage. Despite the high 7. Cunneen SA. Review of meta-analytic comparisons of bariatricfrequency of adverse upper GI symptoms in our series, surgery with a focus on laparoscopic adjustable gastric banding.patient satisfaction with the MIDBAND was very high, as Surg Obes Relat Dis. 2008;4:S47–55.was previously observed in studies of other gastric bands inwhich there were significant GI symptoms [27]. 8. Garb J, Welch G, Zagarins S. Bariatric surgery for the treatment of morbid obesity: a meta-analysis of weight loss outcomes for In conclusion, our prospective, multicenter data confirm laparoscopic adjustable gastric banding and laparoscopic gastricthat LAGB using the low-pressure, silicone-based MID- bypass. Obes Surg. 2009;19(10):1447–55.BAND is a safe and effective surgical procedure for weightloss. Individual growth analysis demonstrated that postop- 9. Favretti F, Ashton D, Busetto L, et al. The gastric band: first-erative weight loss is strongly related to the frequency and choice procedure for obesity surgery. World J Surg. 2009;33consistency of follow-up visits. Additional prospective (10):2039–48. 10. Lanthaler M, Aigner F, Kinzl J, et al. Long-term results and complications following adjustable gastric banding. Obes Surg. 2010;20(8):1078–85. 11. Suter M, Calmes JM, Paroz A. A 10-year experience with laparoscopic gastric banding for morbid obesity: high long-term complication and failure rates. Obes Surg. 2006;16(7):829–35. 12. Camerini G, Adami G, Marinari GM. Thirteen years of follow-up in patients with adjustable silicone gastric banding for obesity: weight loss and constant rate of late specific complications. Obes Surg. 2004;14(10):1343–8. 13. Catona A, Gossenberg M, La Manna A, et al. Laparoscopic adjustable gastric banding: preliminary series. Obes Surg. 1993;3:207–9.

OBES SURG (2012) 22:572–581 58114. Broadbent R, Tracey M, Harrington P. Laparoscopic gastric 23. Chevallier JM, Paita M, Rodde-Dunet MH, et al. Predictive banding: a preliminary report. Obes Surg. 1993;3:63–7. factors of outcome after gastric banding: a nationwide survey on the role of center activity and patients’ behavior. Ann Surg.15. Ren CJ, Horgan S, Ponce J. US experience with the LAP-BAND 2007;246:1034–9. system. Am J Surg. 2002;184(6B):46S–50S. 24. Sivagnanam P, Rhodes M. The importance of follow-up and16. Basdevant A, Laville M, Ziegler O, et al. Recommendations for distance from centre in weight loss after laparoscopic adjustable diagnosis, prevention and treatment of obesity in France. J Nutr gastric banding. Surg Endosc. 2010;24:2432–8. and Diet. 1998;33(Suppl 1). 25. Lee WJ, Ser KH, Chen JC, et al. Improvement of insulin17. Di Lorenzo N, Furbetta F, Favretti F, et al. Laparoscopic gastric banding resistance after obesity surgery: a comparison of gastric banding via pars flaccida versus perigastric positioning: technique, complica- and bypass procedures. Obes Surg. 2008;18:1119–25. tions, and results in 2,549 patients. Surg Endosc. 2010;24:1519–23. 26. Burton PR, Brown W, Laurie C, et al. Outcomes, satiety, and18. Singer JD, Willett JB. Applied longitudinal data analysis: adverse upper gastrointestinal symptoms following laparoscopic modelling change and event occurrence. New York: Oxford adjustable gastric banding. Obes Surg. 2011;21(5):574–81. University Press; 2003. 27. Carelli AM, Youn HA, Kurian MS, et al. Safety of the19. Dallal RM, Quebbemann BB, Hunt LH, et al. Analysis of weight laparoscopic adjustable gastric band: 7-year data from a U.S. loss after bariatric surgery using mixed-effects linear modeling. center of excellence. Surg Endosc. 2010;24:1819–23. Obes Surg. 2009;19(6):732–7. 28. Swanson TW, Tang BQ, Rusnak CH, et al. A five year Canadian20. Littell R, Milliken G, Stroup W, et al. SAS system for mixed laparoscopic adjustable gastric band experience. Am J Surg. models. Cary: SAS Institute; 1996. 2010;199(5):690–4.21. Biagini J, Karam L. Ten years experience with laparoscopic 29. Blanc PM, Lagoutte JM, Picot MC, et al. Preliminary results of adjustable gastric banding. Obes Surg. 2008;18:573–7. the laparoscopic adjustable gastric banding procedure by a new generation of silicone band: MIDBAND. Obes Surg. 2008;1822. Cohen P, Chen H. Using individual growth model to analyze the (5):569–72. change in quality of life from adolescence to adulthood. Health Qual Life Outcomes. 2006;4:10.

Surgery for Obesity and Related Diseases 5 (2009) 104 –110 Video review articleSurgical implantation and adjustment technique with the Realize Band System Jaime Ponce, M.D., F.A.C.S.* Director of Bariatric Surgery, Dalton Surgical Group PC, Hamilton Medical Center, Dalton, GeorgiaKeywords: Morbid obesity; Bariatric surgery; Gastric band; Swedish Adjustable Gastric Band; SAGB; Realize Band; Lap-Band; Body mass index; Injection port Laparoscopic adjustable gastric banding (LAGB) for the Szinicz et al. [26 –28] in Austria experimentally implanted atreatment of morbid obesity (body mass index Ն40 kg/m2 silicone band with an inner balloon attached to a subcuta-or Ն35 kg/m2 with co-morbidities [1]) is the safest of the neous port that was modulated by instillation or withdrawalbariatric procedures [2– 4]. LAGB is associated with weight of saline. In 1985, at Huddinge Hospital, Sweden, Hallbergloss efficacy [5–10], resolution of weight-related co-morbid and Forsell [29] described the first open clinical implanta-disease [11–13], and improvements in quality of life [14]. In tion of a band that operated by a low-pressure, high-volumewidespread use in Europe and other countries since 1996 mechanism of action—the prototype Swedish Adjustable[15], LAGB has become the most prevalent bariatric oper- Gastric Band (SAGB, Obtech AG, Switzerland; acquired ination outside of the United States [16]. With the introduc- 2002 by Johnson & Johnson, Ethicon Endo-Surgery, Cin-tion to the United States of the Lap-Band in 2001 [17] and cinnati, OH). In 1986, Kuzmak [30] in the United Statesthe Realize Band in 2007 [18], LAGB is also growing in implanted a Silastic band with an inflatable (adjustable)popularity in America. small-diameter inner balloon that operated in a high-pres- sure, low-volume manner, the forerunner of the Lap-Band External adjustment of the stoma size [19,20] and the (Allergan, Irvine, CA).option of LAGB reversibility [21] are assets that distinguishthis operation from other metabolic procedures. LAGB ad- Laparoscopic approachjustability, however, when coupled with surgeon-dependentband placement and adjustment regimens, can contribute to Laparoscopic access to the abdomen advanced gastrica greater than optimal rate of long-term complications (e.g., banding and bariatric surgery, in general. In the early 1990s,slippage, gastric prolapse, band erosion) [22,23]. The suc- Catona et al. [31] in Italy and Broadbent et al. [32] incess with this dynamic device depends on knowledge of Australia implanted nonadjustable gastric bands laparo-technological and technical advances in the field. The scopically. Catona et al. [31] was the first to open the lesserpresent review traces the development of gastric banding omentum to achieve greater efficiency in band placement.and describes implantation of the Realize Band using the Belachew et al. [33] in Belgium described the first laparo-pars flaccida technique, an approach shown to reduce scopic placement of an adjustable band in animals in 1993.LAGB morbidity [24,25]. The technique for mechanical Subsequently, Belachew et al. [34] and Forsell et al. [35]port fixation and postoperative adjustment is also detailed. implanted the first LAGBs in humans. Belachew et al. [34] used the so-called perigastric technique, and Forsell et al.Adjustable gastric banding history [35], the “pars flaccida” dissection. Pars flaccida was con- ceived originally for use with open SAGB implantation to Between 1982 and 1989, research in animals contributed avoid some of the complications experienced in nonadjust-conceptually to the evolution of adjustable gastric banding. able banding. A third approach, the “2-step” (pars flaccida to perigastric progression) described by Weiner et al. [36] in *Reprint requests: Jaime Ponce, M.D., F.A.C.S., 1504 Broadrick Drive, Germany in 2000, combined these techniques for use withDalton, GA 30720. an extremely large fat pad. Use of the pars flaccida tech- nique was shown to reduce the incidence of gastric prolapse E-mail: [email protected]/09/$ – see front matter © 2009 American Society for Metabolic and Bariatric Surgery. All rights reserved.doi:10.1016/j.soard.2008.08.010

J. Ponce / Surgery for Obesity and Related Diseases 5 (2009) 104 –110 105and other late complications to .6 –3.5% relative to the Fig. 1. Ponce-modified lithotomy trocar placement.perigastric technique [37–39]. By 2005, the pars flaccidatechnique had largely supplanted perigastric band place- approximately 1-cm incision lateral on the left side of thement [39]. gastroesophageal junction within the gastrophrenic liga- ment; this small “window” initiates the plane of dissectionPort implantation innovation for the retrocardial tunnel. Care must be taken to preserve the gastrophrenic attachments to help secure the band and The most recent innovation in the LAGB technique is maintain stability of the fundus [44]. Fat pads surroundingbased on mechanical port-fixation technology, an alterna- the gastroesophageal junction can be excised, if needed, totive to suture fixation of the port (known as the Velocity expose the stomach wall or identify and repair a cruralPort outside the United States, created for use with the defect or hiatal hernia [45].SAGB [Ethicon Endo-Surgery]; in the United States, theVelocity Port is named the Realize Port and has been in- In the lesser omentum (“pars flaccida”) dissection at thecorporated into the Realize Band system). A 1-year random- lesser curvature, the pars flaccida is stretched to make aized study of this technology versus standard suture fixation 5-cm incision cranial to the left gastric artery but caudal todemonstrated its more rapid speed-to-implantation and re- the gastroesophageal junction. It is important to avoid mak-duced complications (i.e., port site pain, use of pain medi- ing an opening in the omental bursa to avert future pouchcations) [40]. A second randomized study showed an eight- dilation, and care should be taken to avoid injury to thefold greater speed-to-implantation (i.e., from Ͼ8 min to Ͻ1 nerves and vessels in the area of the pars flaccida. If a largemin, P Ͻ.0001) with the Velocity technology [41]. In a hepatic artery emanates from the left gastric artery, thepreliminary report, a mechanical port fixation technology “2-step” dissection can be used [38]. The right crus is thenfor use with the Lap-Band, the RapidPort system (Allergan), incised, and blunt dissection of the tissue along the posteriorwas found to decrease the port implant time from 15–20 medial border of the right crus is performed, initiating aminutes to 3.02 minutes, with a reduction of complications 0.5-cm incision for the retrocardial tunnel. An articulatingin the early postoperative period (P Ͻ.001) [42]. This sys- blunt dissector (Realize Endoscopic Dissector, Ethicontem is not available in the United States. Endo-Surgery) is introduced, inserted at the incision pre- pared at the right crus, and gently advanced to emerge at the The Lap-Band and the SAGB are both approved for use angle of His incision (Fig. 2a).in the United States. The SAGB is marketed as the RealizeBand in America, although no technical differences exist Introduction and placement of bandbetween the Realize and SAGB bands. After being prepared and checked for leaks, air is aspi-Implantation technique with Realize Band rated from the Realize balloon and a knot is tied in the distal tubing to prevent air from re-entering. The locking end flap Implantation of the Realize Band can be seen in Video 1. of the Realize Band is introduced with forceps, without force, through the 15-mm trocar into the abdominal cavity.Patient position and trocar placement With the balloon side of the band facing the stomach wall, the band’s suture loop is transferred by forceps to the notch Patient position and trocar placement are determined by of the articulating dissector tip (Fig. 2b), which guides thethe surgeon’s experience and personal preference. In ourapproach, a 5-port technique is used, with four 5-mm portsand one 15-mm port. In the modified lithotomy position,with the surgeon standing between the patient’s legs, theoptical trocar (trocar 1) is inserted subxiphoid in a mid-epigastric position. Trocar 2, used for liver retraction, isinserted at the patient’s right subcostal margin in the mid-clavicular line; trocars 3 and 4, in an arc extending fromtrocar 2 toward the patient’s left; and trocar 5, laterally onthe left subcostal margin (Fig. 1). Other approaches includeplacing the patient in the supine position with the surgeonstanding on the patient’s right. Satisfactory access to thesuperior portion of the peritoneal cavity should be ensuredto introduce the Realize Band with minimal risk of banddamage or infection [43].Angle of His and pars flaccida dissections The omentum covering the angle of His is gently sweptto the side to facilitate exposure during the creation of an

106 J. Ponce / Surgery for Obesity and Related Diseases 5 (2009) 104 –110Fig. 2. (a) “Pars flaccida” dissection with the articulating blunt dissector; (b) suture loop placed into dissector’s notched tip; (c) band pulled gently throughretrogastric tunnel with balloon side oriented toward stomach wall; (d) 2-grasper technique used to close Realize Band; (e) 2 gastrogastric plication suturesplaced to surround band anteriorly, leaving band without tension within tunnel; and (f) applier used to attach Realize port in fascia.band through the retrocardial tunnel from the angle of His to band is introduced through the 15-mm trocar, and the tubingthe lesser curvature (Fig. 2c). is left outside. With the grasper still holding the distal tab, the band is stable enough to mobilize and place the suture The author does not use the calibration balloon routinely, loop around the dissector’s tip notch. The distal tab is thenbut only if there is a question about the anatomy or hiatal released, and the band is pulled through the retrocardialhernia dissection and repair. Band closure is achieved by tunnel while the grasper is kept in the buckle hole. Thisinserting the dissector through the band’s buckle opening to maneuver retains the band’s orientation and facilitatesgrasp the end flap, which is pulled through the buckle until grasping of the distal tab again to close the band after it hasthe diamond-shaped locking indicator is visible to the right passed the posterior tunnel (Fig. 3).of the buckle (Fig. 2d). The band buckle is drawn downwardtoward the lesser curvature in preparation for placement of Port fixationan average of 2 nonabsorbable gastrogastric plication su-tures. A curved needle is used to ensure secure “bites” of At T4, the incision is expanded to 3– 4 cm to contain thetissue between the seromuscular layer of the fundus and the length of the port applier [46]. Fat is dissected away frompouch. These stitches protectively tunnelate the anterior the fascia on the anterior rectus sheath. The locking con-band, except at the location of the buckle, to minimize the necter of the port is placed around the band tubing withrisk of erosion. The band should lie without tension within the tubing snug against the port housing. Before slidingthe tunnel to allow for filling during the adjustments (Fig. the connector into the housing, the locking tabs should be2e). The free end of the band tubing is pulled out through T4 aligned with the connector notch in the port housing; thein preparation for fixation of the port system. connector is turned clockwise until it stops rotating. A Huber needle is used to aspirate air from the port.Alternative “1-step technique” The actuator ring is checked to ensure that it is rotated The alternative “1-step technique” is shown in Video 2. into the unlocked position and that the port’s fastening After passing the “learning curve,” we have also used the hooks are retracted. (The actuator ring may need to be“1-step” technique to introduce, place, and close the Realize rotated again to achieve the unlocked state if the port’sBand in a single step. This method was conceptualized by hooks are not completely retracted.) The port snaps into theBrengman (personal communication, 2008, Richmond, applier with its red safety cap on the outside and the con-VA). In this technique, the grasper used to introduce the nection housing fitted into 1 of the 2 possible guide notches.band is placed through the buckle to grab the distal tab. The To fix the port to the fascia, the red safety cap is removed

J. Ponce / Surgery for Obesity and Related Diseases 5 (2009) 104 –110 107Fig. 3. “One-step” technique. Realize band introduced, placed, and closed in 1 step.from the port’s underside. The applier is grasped in the palm periodic adjustments [48]. Approximately 5– 6 adjustmentsand inserted into the incision at an angle so that the port are performed in the first year, tapering to 2 or 3 annually inbase lies uniformly flat on the fascia. The applier’s safety the next 2 years.trigger is released, the palm compresses the applier firinglever, and the port is fixed to the fascia. To release the port Indicationsfrom the applier, the firing lever must remain compressed asthe applier is pulled away (Fig. 2f). Aligned with the port to Adjustments to instill fluid in the band are performedprevent kinking, the tubing is carefully guided into the when there is an insufficiency of satiety between meals,abdominal cavity with a strain-release loop to facilitate the small food portions at meal times (i.e., hunger beyond 1 Cpatient’s free movement without band or port dislodgement. of solid food), or appropriate weight loss (i.e., ideally,The band balloon is left unfilled at the conclusion of sur- 0.5–1.0 kg/wk in the first year). Adjustments to remove fluidgery. (If proper mechanical fixation has not been attained, from the band are performed in response to symptoms ofthe port should be secured with sutures by way of the holes band overtightness [49].in the actuator ring.) Locating and accessing the port Port placement for ease of access at adjustments is ofparamount importance. Whether by mechanical or suture The patient is placed in a supine position with the armsfixation, the port must be placed securely to avoid rotation behind the head, and the port location is identified by pal-or flipping, and it should be situated in a consistent location pation. The skin above the port is disinfected, and a syringefrom patient to patient, with the tubing in a predictable with 3– 4 mL of sterile 0.9% saline solution is prepared.orientation with respect to the port to facilitate safe, long- While stabilizing the port with the thumb and first finger, aterm access. long Huber needle is inserted vertically through the skin and port septum into the port chamber. Great care should beAdjustment technique taken to avoid puncturing the port tubing. If unable to identify the exact orientation of the port by palpation, port The band is left unfilled for 4 weeks after surgery to access should be accomplished under fluoroscopic control.promote healing and to avert tightness of the stoma, peri- Ultrasonography is another option for port identification. Ifoperative edema, and emesis. Although a consensus has rotated, the port can be righted manually, and its locationnote been reached regarding guidelines for follow-up [47], marked on the skin with a pen. With the needle in the portgood outcomes are associated with patient compliance and chamber, the fluid is instilled into the band [49].

108 J. Ponce / Surgery for Obesity and Related Diseases 5 (2009) 104 –110 The Realize Band is a low-pressure system throughout removal from the band using fluoroscopy. Removal of 0.5the whole range of adjustments; thus, when the needle is mL typically resolves such symptoms. Patients who are notintroduced, no back pressure will be noted, not even at the losing weight, or who are gaining weight, might have a bandmaximal fill volume. If not sure of the needle location, the leak or might not be complying with the LAGB eating re-injection of 0.5 mL and aspiration can be done to verify an quirements. This can be evaluated by removal of the fluid fromadequate position within the port chamber. the band and instillation of Յ9 mL of non-ionic iodinated contrast under fluoroscopy. Patients struggling with weightCalibrating the band adjustment loss, who have appropriate band restriction, might require more frequent follow-up visits, counseling, and/or nutritionist At the first filling, all air should be removed from the support.band system. Approximately 3– 4 mL of fluid are instilledinto the band. Subsequent band adjustments should be made Discussionat 1-month intervals, with an additional .5–1.0 mL added ateach visit. Generally, the addition of .2–.5 mL of fluid at an The contributions of the SAGB/Realize System to gastricadjustment is sufficient to maintain appropriate weight loss. banding include 360° soft-surfaced stoma coverage and aIf a patient is losing Ͼ1.0 kg/wk and has good restriction, high-volume (9-mL) inner balloon designed to operate in ano fluid should be added to the band. To avoid overtightness low-pressure manner [51]. The pars flaccida technique, de-after an adjustment, the patient should be able to swallow veloped for SAGB/Realize implantation, facilitates higherwater comfortably before leaving the office. Annual barium placement of the band on the stomach, out of the lesser sac,swallow studies are advised to evaluate esophagus, pouch, markedly reducing the slippage rates relative to the perigas-and band orientation. tric approach [38,39]; the pars flaccida approach also aids in the creation of a smaller pouch and preserves the retrogas-Fluoroscopic-guided band calibration tric attachments that help secure the band [51]. Finally, maintenance of the band in an unfilled state for 4 weeks The Realize port’s fastening hooks and connector tube after surgery, always a requirement of SAGB/Realize im-are radiopaque for recognition under fluoroscopy should plantation, has been shown to foster healing, reduce edema,this adjustment approach be necessary. With the patient in a and avert vomiting [35].standing position, the level of band restriction can be de-termined by observing the stoma during the barium swallow Bariatric surgeons outside of the United States, somestudy. Gastric emptying should proceed at a rate such that with Ͼ10 years of experience implanting the SAGB/Real-flow is restricted, but not obstructed (Fig. 4). An added ize, have found it effective and safe during the intermediateadvantage of fluoroscopy is its capacity for diagnosis of term, with good long-term data emerging [5– 8,10,16,25]. Inoutlet stenosis, reflux, esophageal or pouch dilation, gastric a meta-analytic comparison of worldwide SAGB/Realizeprolapse, and band malposition (i.e., from an overtight and the Lap-Band data, the mean percentage of excessband, hiatal hernia, or prolapse) [50]. weight loss was similar at 3 years after surgery, ranging from 56.3% for the SAGB/Realize (n ϭ 4273) to 50.2% forGauging adjustment success the Lap-Band (n ϭ 24,707). The 2 bands also appeared to have similarly low rates of complications and equivalent If a patient is losing Ͼ1 kg/wk and has symptoms of early mortality (Ͻ.1%) [52]. Because the Realize Band wasobstruction (e.g., protracted vomiting, regurgitation, or in-ability to eat), the patient should be evaluated for fluidFig. 4. Realize port view under x-ray. (a) Radiopaque parts are hooks, locking connector, and septum. (b) Appropriately adjusted band observed to restrict,but not obstruct, flow of barium through stoma.