Medwrite Portfolio_20 Surgery Sample Articles

20   BARIATRIC SURGERY ARTICLES  

OBES SURGDOI 10.1007/s11695-013-1160-3 REVIEW ARTICLEDiabetes and Weight in Comparative Studies of BariatricSurgery vs Conventional Medical Therapy: A SystematicReview and Meta-AnalysisG. Ribaric & J. N. Buchwald & T. W. McGlennon# The Author(s) 2013. This article is published with open access at Springerlink.comAbstract We performed a meta-analysis of weight loss and criteria (i.e., n<10 per arm, animal studies, reviews, caseremission of type 2 diabetes mellitus (T2DM) evaluated in reports, abstracts, and kin studies). Of 19 eligible articles,randomized controlled trials (RCTs) and observational studies two not focused on diagnosed T2DM and one with insuffi-of bariatric surgery vs conventional medical therapy. English cient T2DM data were excluded. In the final 16 includedarticles published through June 10, 2013 that compared bar- papers, 3,076 patients (mean BMI, 40.9; age, 47.0; 72.0 %iatric surgery with conventional therapy and included T2DM female) underwent bariatric surgery; 3,055 (39.4; 48.6,endpoints with ≥12-month follow-up were systematically 69.0 %) received conventional or no weight-loss therapy. Inreviewed. Body mass index (BMI, in kilogram per square bariatric surgery vs conventional therapy groups, the meanmeter), glycated hemoglobin (HbA1C, in degree), and fasting 17.3±5.7 month BMI WMD was 8.3 (7.0, 9.6; p<0.001;plasma glucose (FPG, in milligram per deciliter) were ana- I2=91.8), HbA1C was 1.1 (0.6, 1.6; p<0.001; I2=91.9), andlyzed by calculating weighted mean differences (WMDs) and FPG, 24.9 (15.9, 33.9; p<0.001; I2=84.8), with significantpooled standardized mean differences and associated 95 % differences favoring surgery. The overall T2DM remissionconfidence intervals (95 % CI). Aggregated T2DM remission rate for surgery vs conventional group was 63.5 vs 15.6 %event data were analyzed by calculating the pooled odds ratio (p<0.001). The Peto summary POR was 9.8 (6.1, 15.9);(POR) and 95 % CI. Random effects assumptions were ap- inverse variance summary POR was 15.8 (7.9, 31.4). Of theplied throughout; I2≥75.0 % was considered indicative of included studies, 94.0 % demonstrated a significant statisticalsignificant heterogeneity. Systematic review identified 512 advantage favoring surgery. In a meta-analysis of 16 studiesarticles: 47 duplicates were removed, 446 failed inclusion (5 RCTs) with 6,131 patients and mean 17.3-month follow- up, bariatric surgery was significantly more effective thanElectronic supplementary material The online version of this article conventional medical therapy in achieving weight loss,(doi:10.1007/s11695-013-1160-3) contains supplementary material, HbA1C and FPG reduction, and diabetes remission. The oddswhich is available to authorized users. of bariatric surgery patients reaching T2DM remission ranged from 9.8 to 15.8 times the odds of patients treated withG. Ribaric conventional therapy.European Surgical Institute, Ethicon Endo-Surgery (Europe) GmbH,Hamburg, Germany Keywords Bariatric . Metabolic . Comparative . Type 2 diabetes mellitus . T2DM . Systematic review .J. N. Buchwald Meta-analysis . Randomized controlled trialDivision of Scientific Research Writing, Medwrite MedicalCommunications, Maiden Rock, WI, USA IntroductionT. W. McGlennon “Diabesity,” a term coined by Dr. Ethan Sims in 1973 toStatistical Analysis & Quality of Life Assessment, denote comorbid obesity and type 2 diabetes mellitusMcGlennon MotiMetrics, Maiden Rock, WI, USA (T2DM) [1], has steadily grown into a global epidemic. The World Health Organization (WHO) estimates the number ofG. Ribaric (*)Ethicon Endo-Surgery (Europe) GmbH, MD&D EMEA (Europe,Middle East, Africa), Johnson & Johnson, HummelsbüttelerSteindamm 71, 22851 Norderstedt, Germanye-mail: [email protected]

OBES SURGoverweight adults at >1.6 billion and >400 million who are Our aim was to assess diabetes and weight outcomes inobese. By 2015, >2.3 billion adults are projected to be over- comparative studies of bariatric surgery vs conventional med-weight with >700 million obese [2]. Globally, >312 million ical therapy. In preliminary research, we found few directlypeople suffer from T2DM [3], a disease associated with a comparative, level 1, randomized controlled trial (RCT) out-markedly increased risk of heart disease and stroke, micro- comes (as defined by the Oxford Centre for Evidence-Basedand macrovascular consequences, retinopathy, and kidney Medicine [34]). As a result, we broadened our inclusionfailure [4]. While weight loss and its maintenance, by any criteria to incorporate directly comparative observational stud-means, aids in improving and managing T2DM [5, 6], long- ies (OSs). Thus, the current review systematically identifiedterm antidiabetic diet compliance is poor even when supported and screened comparative studies of bariatric surgery vsby pharmacotherapy; 50.0 to 90.0 % of patients remain unable conventional medical therapy in adults with a meanto achieve adequate diabetes control [7–9]. BMI≥25 and subjected the aggregated weight and diabetes data (BMI, HbA1C, fasting plasma glucose (FPG), and diabe- In the long-running, prospective, controlled Swedish tes remission) to meta-analysis.Obese Subjects (SOS) study, weight loss by conventionalmedical therapy was associated with T2DM remission of Methodsapproximately 21.0 % (n=248) at 2 years (compared with72.0 % remission in postbariatric surgery patients [n=342]) Inclusion Criteria and Search Strategyand 12.0 % (n=84) at 10 years (vs 37.0 % bariatric surgerygroup remission [n=118]) [10], an approximately threefold An electronic literature search and cross-referencing of articlesdifference in effective control of diabetes favoring bariatric was performed within the following databases: National Li-surgery. Typically, when even excellent weight loss has been brary of Medicine PubMed®/MEDLINE®, SpringerLink®,achieved by very low calorie diets (VLCDs) and intensive and SciVerse®. The search strategy followed the identificationlifestyle programs, neither weight loss nor diabetes resolution and screening guidelines established by the Preferredhas been maintained beyond 1–5 years [10–13] nor has dia- Reporting Items for Systematic Reviews and Meta-Analysesbetes resolved as rapidly as following most bariatric proce- (PRISMA) statement [35]. Articles were identified by Booleandures (i.e., within days to a few weeks [14]). Several series combination of keywords: “bariatric surgery,” “metabolicsuggest that although the weight-loss effect of bariatric sur- surgery,” “diabetes surgery,” “gastric band,” “sleeve gastrec-gery is attenuated in lower body mass index (BMI, in kilo- tomy,” “gastric bypass,” “duodenal switch,” “biliopancreaticgram per square meter) patients [15, 16], surgery may achieve diversion,” with “medical treatment,” “medical therapy,”a higher rate of diabetes resolution than conventional medical “conventional treatment,” “conventional therapy,” and “diet.”therapy in patients who are only overweight (BMI≥25–29.9) An additional search using keyword phrases, “bariatric sur-through those with class III obesity (BMI≥40.0) [17–19]. gery, diabetes mellitus,” and “bariatric surgery, glucose ORDiabetes and obesity are progressive, multifactorial diseases; insulin OR HbA1C or HOMA,” was run. Limits set to governit is probable that bariatric surgery and/or one of the emerging, the searches stipulated journal articles that featured compara-less-invasive, endolumenal procedures in combination with tive studies on adult human subjects written in the Englishlife-long lifestyle modification may represent an optimum language with no beginning date through June 10, 2013.management strategy [11, 20, 21]. Compiled article citations were screened by title to exclude Procedures that most effectively reduce weight, such as duplicates arising from unintentional collection of bothRoux-en-Y gastric bypass (RYGB) and biliopancreatic diver- e-publications and their follow-on print versions. The uniquesion, realize 80.0 and 95.0 % hyperglycemia remission, respec- citations were evaluated by review of abstracts. Articles withtively [22], a markedly greater treatment effect than that seen in n<10 in any study arm, reviews, animal studies, case reports,the majority of studies of conventional dietary and pharmaco- abstracts, book chapters, kin studies (i.e., reports with over-logic therapy; most patients fail to achieve the goal for glycemic lapping data, or outcomes reported for the same timeframecontrol of <7.0 % glycated hemoglobin (HbA1C) prescribed by and/or by the same author group), and Comments or Letters tothe American Diabetes Association (ADA) [23, 24]. The etiol- the Editor were excluded from eligibility. The remainingogy of diabetes remission following bariatric surgery is not articles were read in full and assessed by two researchers tofully understood. Remission may be engaged by divergent ensure that all or a subset of the overweight, mildly obese, orand/or additional mechanisms through bariatric surgery [25], morbidly obese patients in each treatment arm had beenas individual procedures reorganize the gastrointestinal (GI) diagnosed with T2DM (of any duration) prior to undergoingtract differently, activating varied neurohormonal mechanisms bariatric surgery or conventional therapy. Finally, articles[26, 27]. Preclinical [28] and clinical evidence [19, 29, 30] evaluating fewer than one of the aforementioned T2DM end-suggest that improved glycemic control is not linked exclu- points were excluded from quantitative analysis.sively with baseline weight or operative weight loss andresults from complementary processes [15, 22, 31–33].

OBES SURGDefining Diabetes Remission and Data Extraction database search. Forty-seven duplicates were removed. After screening 465 unique citations by title and abstract, 446 failedThe recommended glycemic goal for HbA1C stipulated by the to meet inclusion criteria and were excluded. The remaining2009 ADA Standards of Medical Care in Diabetes for adults is 17 articles were read and assessed for eligibility. Two that<7.0 % with a suggested normal range of 4.0–6.0 % [24]. A provided neither primary nor secondary outcomes related tovariety of definitions of T2DM remission have been used in patients with diagnosed T2DM and one that provided insuffi-bariatric surgery and conventional medical therapy studies. In cient T2DM data to assess at the 12-month analysis time pointthe current analysis, the percentage of patients that achieved were excluded, leaving a final set of 16 articles [41–56] forT2DM remission, as assessed independently in each included quantitative analysis.study, characterized T2DM remission rate. Characteristics of included studies are described in Table 1. Variable data of interest were extracted from included Comparative articles were published over a slightly less thanstudies and entered into a dedicated database. Data collection 10-year span, between December 23, 2004 and June 10, 2013,objectives centered on study characteristics (including bariat- the majority (13/16, 81.3 %) in the last 3.5 years. The respec-ric procedures, conventional treatments, study designs, and tive country of origin (based on first author’s affiliation) andanalysis time points); demographic and anthropometric mea- article distribution was: USA, six (38.0 %); Italy, foursures (age, gender, BMI [weight (in kilogram), divided by (25.0 %); Australia, two (12.5 %); Norway, two (12.5 %);height (in square meter)] [36]); markers of glycemic control Sweden, one (6.0 %); and Korea, one (6.0 %). Study designs(HbA1C and FPG); and T2DM remission rate. Assessment of included 5 (31.0 %) nonblinded RCTs, 11 OSs, 3 (19.0 %)study quality indicated that a range of diverse medical therapies nonrandomized controlled trials (nRCTs), 5 (31.0 %) prospec-characterized the non-surgical control groups. In addition, cer- tive comparison studies, and 3 (19.0 %) retrospective databasetain studies combined outcome data from multiple bariatric reviews. All studies obtained local institutional or ethics reviewsurgery procedures and presented results in the form of a board protocol approval, 12 mentioned obtaining informed con-general surgery group vs conventional therapy; no stratification sent, and 8 were nationally registered as clinical trials. Often,was used. The present authors opted to extend the concept of more than one WHO weight class was studied in a report: 11“grouping” to each study that met inclusion criteria; i.e., if (69.0 %) included morbidly obese patients, 4 (25.0 %) obese, 5studies reported on more than two arms (e.g., multiple surgery (31.3 %) mildly obese, and 1 (6.3 %) overweight. Our analysisprocedures and/or multiple forms of conventional therapy), data focused on results collected between 12 and 24 months follow-were pooled using weighted means and standard deviations to ing study commencement. Mean follow-up was 17.3±represent summary data for one “combined surgery group” vs 5.7 months (median 15.0 months). T2DM remission was de-one “combined conventional therapy group” per study. fined variably across studies; however, the target criteria was always identical for the two arms within each individual study.Statistical Analysis Bariatric surgical procedures employed were well-Data manipulation and analysis were conducted using SPSS® accepted, frequently performed operations [57]: Roux-en-Ysoftware, version 20.0 (IBM SPSS, Chicago, IL, USA) in con- gastric bypass (11; 69.0 %), laparoscopic adjustable gastricjunction with Comprehensive Meta Analysis 2.2 (Biostat, En- banding (LAGB; 5, 25.0 %), biliopancreatic diversion (BPD;glewood, NJ, USA). Body mass index, HbA1C, and FPG were 3, 18.8 %), laparoscopic sleeve gastrectomy (LSG; 3, 18.8 %),analyzed by calculating weighted mean differences (WMDs) and vertical banded gastroplasty (VBG; 1, 6.3 %), and duodenalpooled standardized mean differences (SMDs) and associated 95 % switch (DS; 1, 6.3 %). Conventional therapies ranged fromconfidence intervals (CI). Aggregated T2DM remission event data intensive weight loss programs with initial VLCDs graduatingwere analyzed by calculating the pooled odds ratio (POR) and 95 % to moderate calorie intake combined with lifestyle modifica-CI. Random effects assumptions were applied throughout; I2≥ tion training and diabetologist-managed T2DM treatment; to75.0 % was considered indicative of significant heterogeneity. those with structured, rehabilitative inpatient programs; some[See Electronic Supplementary Material, Appendix 1 includ- that included routine medical management of T2DM and self-ing references 37–40 for detailed statistical methodology.] monitored weight-loss and exercise plans; to a few programs with no educational plan or dietary supervision.Results Preliminary Analysis of RCTs vs OSsStudy Characteristics Analysis of variance using pooled summary data indicated no statistically significant baseline differences between RCT sur-The results of the systematic review are presented in Fig. 1. A gery, RCT conventional, OS surgery, and OS conventionaltotal of 512 articles were identified by extensive electronic patient groups with respect to age ((mean [SE]) (45.8 [2.0], 47.1 [1.4], 46.6 [2.0], 49.7 [1.4]), respectively; F(3, 6,127)=

OBES SURGFig. 1 Outcomes of thesystematic review of the literatureby record identification,screening, and analysis in thePreferred Reporting Items forSystematic Reviews andMeta-Analyses (PRISMA)statement flow diagram0.83, p=0.50), BMI ((37.5 [1.7], 42.5 [1.2], 37.2 [1.8], 40.3 [mean data available on 16 studies/100 %], HbA1C[1.2]); F(3, 6,127)=1.23, p=0.30, HbA1C (8.9 [0.6], 7.6 [0.4], [12/75 %], and FPG [11/69 %]) are presented in Table 2.7.8 [0.6], 7.1 [0.4]); F(3, 1,517)=1.84, p=0.14), and FPG The summary statistics indicated that, on average and relative(163.0 [8.7], 136.3 [7.9], 156.1 [11.7], 134.0 [10.43]); to the CTG, BSG patients had a higher mean baseline BMIF(3, 4,926)=0.52 ( p=0.67). In light of these findings, data (40.9 kg/m2 [38.5, 43.3] vs 39.4 kg/m2 [37.3, 41.6]; pooledwere integrated from RCTs and OSs to make direct compar- SMD=0.33 [0.16, 0.51], p<0.001; I2=84.0 %), a higherisons between bariatric surgery and conventional therapy HbA1C level (8.0 % [7.1, 9.0] vs 7.7 % [6.8, 8.5]; pooledgroups at baseline. SMD=0.39 [0.12, 0.67], p<0.01; I2=84.0 %), and a higher FPG level (150.3 mg/dL [135.7, 164.9] vs 143.1 mg/dL [129.8,Patient Characteristics and Baseline Clinical Profile 156.3]; pooled SMD=0.15 [0.02, 0.28], p<0.05; I2=84.0 %). These data suggest that the BSG and the CTG were fairly well-The total number of patients in the included studies was 6,131; matched at baseline along variables relevant to the study of3,076underwent bariatric surgery and 3,055 underwent con- T2DM remission. The mean SMD characterizing baseline dif-ventional treatment. The mean age of patients included in this ferences was 0.28 (0.15–0.39), a value considered to represent ameta-analytic research was 47.8 years, ranging from 35.8 to “small” statistical and clinical mean difference between groups62.0 years. Relative to the conventional treatment group and an approximate distribution overlap of 80.0–85.0 %.(CTG), the bariatric surgery group (BSG) was slightly youn-ger (47.0 years [95 % CI, 45.3, 48.7] vs 48.6 years [46.6, Assessment of Within-Group Change in Clinical Markers50.7]; pooled SMD =−0.23 [−0.28, −0.18], p<0.05; I2= After Treatment0.0 %), and was comprised of a somewhat greater percentageof females (72.0 vs 69.0 %, p<0.01). Overall baseline com- Table 3 presents meta-analytic data summarizing mean base-parability statistics for select clinical variables (i.e., BMI line (pretreatment), mean follow-up (post-treatment), and

Table 1 Characteristics of included comparative studies OBES SURGStudy Countrya Bariatric procedure(s) Conventional weight-loss therapy Study design + weight class, T2DM Analysis time points variables, IRB, IC, Reg. (months)Sjöström et al. [40] Sweden LAGB (19.0 %; adjustable and Nonstandardized nonsurgical treatment nRCT (“SOS study”) w/contemporaneous 24; 120O’Brien et al. [41] Australia nonadjustable) VBG (68.0 %) ranging from lifestyle intervention and subject matching 24Dixon et al. [42] Australia RYGB (13.0 %) behavior modification to no treatment 24Hofsø et al. [43] Norway Morbidly obese patients 12Adams et al. [44] USA LAGB in addition to lifestyle Intensive weight-loss program including initial 24Serrot et al. [45] USA modification instruction (i.e., VLCD (500 to 550 kcal/day, followed by T2DM variables=FPG, med. use 12Martins et al. [46] Norway increased exercise + good transition phase), pharmacotherapy, and 12 eating practices) lifestyle change tailored to patients individually IRB, IC LAGB in addition to Conventional dietary and T2DM therapy Reg. Swedish Obese Subjects (SOS) registry conventional T2DM therapy administered by diabetologist with focus on lifestyle change RCT, nonblinded RYGB after following low-calorie Mildly obese patients diet for 3–6 weeks Intensive lifestyle intervention through four 1–4 week stays at rehabilitation center T2DM variables=FPG, insulin RYGB specializing in care of morbidly obese patients; cognitive approach to motivate increased activity IRB, IC RYGB and normalized eating habits Reg. ACTRN012605000113651 RYGB No structured, monitored weight-loss intervention in either of 2 nonsurgical control groups RCT, nonblinded (1=nonintervened surgery-seekers; Mild to morbidly obese patients 2=population-based subjects not seeking surgery) T2DM variables=FPG, HbA1C, HOMA, med. use Routine medical management with nutrition, IRB, IC weight management, and exercise counseling Reg. ACTRN012605000159651 Option of residential intermittent program; nRCT (“MOBIL trial”) commercial weight-loss camp; or hospital Morbidly obese patients outpatient program T2DM variables=FPG, HbA1C, insulin, med. use IRB, IC Reg. NCT00273104 nRCT Morbidly obese patients T2DM variables=FPG, HbA1C, HOMA-IR, insulin, med. use IRB, IC Reg. NIDDK DK-55006 & NCRR M01-RR00064 Retrospective study of surgery recipients and database of matched nonsurgical controls Mildly obese patients T2DM variables=HbA1C, med. use IRB, no Reg. Prospective comparison of patients on surgery wait list given option of surgery or 1 of 3 conservative weight-loss programs Morbidly obese patients T2DM variables=FPG IRB, IC, no reg.

Table 1 (continued) Countrya Bariatric procedure(s) Conventional weight-loss therapy Study design + weight class, T2DM Analysis time pointsStudy variables, IRB, IC, Reg. (months) Individualized medical therapy with conventional 12–120Iaconelli et al. [47] Italy BPD weight, exercise, and dietary support; Prospective, matched, open case-controlled trial diabetologist available for consult Morbidly obese patients 12 every 3 months T2DM variables=FPG, HbA1C, T2DM 3; 6; 12; 18 24Scopinaro et al. [48] Italy BPD Routine medical therapy remission, HOMA, insulin IRB, IC, no Reg. study 18Leonetti et al. [49] Italy LSG Standard medical therapy with individual Obese and morbidly obese patients 12 lifestyle modification programs, access T2DM variables=FPG, HbA1C, T2DM 24 to diabetologists, dietician, nurse. remission, med. useMingrone et al. [50] Italy RYGB (50.0 %) BPD (50.0 %) Diet, exercise, lifestyle modification program, IRB, no Reg. including medication optimization and Prospective study of surgical patients matched treatment by diabetologist, dietician, nurse—1, 3, 6, 9, 12, and 24 months. with nonsurgical database controls Overweight and mildly obese patientsHeo et al. [51] Korea RYGB (28.0 %) Conventional medical therapy, including lifestyle T2DM variables=FPG, HbA1C, HOMA-IRDorman et al. [52] USA LAGB (27.6 %) LSG (44.4 %) modification, medication optimization, IRB, IC, no reg. counseling by dietician and exercise practitioner Prospective study of surgical patients matchedLeslie et al. [53] USA RYGB (60,1 %) LAGB (20.9 %) DS (18.8 %) Conventional management by endocrinologist with nonsurgical controls for medication use; lifestyle modification to Obese and morbidly obese patients promote weight loss was encouraged T2DM variables=FPG, HbA1C, med. use IRB, IC, no Reg. RYGB Conventional medical management • RCT, nonblinded • Morbidly obese patients • T2DM variables=FPG, HbA1C, med. use, OBES SURG T2DM remission IRB, IC Reg. NCT00888836 Retrospective multicenter database Retrospective case-matched database study Morbidly obese patients T2DM variables=HbA1C, med. use IRB, no Reg. Prospective study of surgical patients matched with nonsurgical database controls Morbidly obese patients T2DM variables=HbA1C, med. use IRB, no reg.

OBES SURGAnalysis time points a Country of first author’s affiliation mean change values in BMI, HbA1C, and FPG for each of the (months) RCT randomized controlled trial, nRCT nonrandomized controlled trial, HOMA homeostasis model assessment, IR insulin resistance, IRB institutional review and/or ethics board approval obtained, IC studies within the two treatment groups (i.e., BSG and CTG). 12 informed consent obtained, Reg. clinical trial registration number or governmental grant number, BPD biliopancreatic diversion, RYGB Roux-en-Y gastric bypass, VBG vertical banded gastroplasty, LAGB Pooled estimates of overall means and mean change (WMD) 12 laparoscopic adjustable gastric banding, LSG laparoscopic sleeve gastrectomy, DS duodenal switch are also provided. The weighted mean baseline BMI for the BSG was 40.9 kg/m2 (95 % CI, 38.5, 43.3), with a follow-upStudy design + weight class, T2DM weighted mean BMI of 29.4 kg/m2 (27.8, 30.9). The WMD in variables, IRB, IC, Reg. BMI for BSG patients was 11.4 kg/m2 ([95 % CI, 10.0, 12.9], RCT, nonblinded statistical significance of overall effect: p<0.001, I2=95.0 %). Obese patients In contrast, the weighted mean baseline BMI for the T2DM variables=FPG, HbA1C, HOMA CTG was 39.4 kg/m2 (37.3, 41.6), with a follow-up IRB, IC weighted mean BMI of 37.8 kg/m2 (35.6, 39.9). The Reg. NCT00432809 WMD for all patients undergoing some form of con- RCT, nonblinded, multicenter trial ventional therapy was 1.6 kg/m2 ([0.7, 2.6], p<0.01, Mildly obese and obese patients I2=86.5 %). T2DM variables=FPG, HbA1C IRB, IC Within-group mean changes in HbA1C and FPG tended to Reg. NCT00641251 follow the pattern of change observed in BMI. At baseline, the BSG had a weighted mean HbA1C of 8.0 % (7.1, 9.0), with aConventional weight-loss therapy Intensive medical therapy with lifestyle Medical management including lifestyle follow-up of 6.1 % (5.8, 6.4), a reduction that represented an counseling by diabetes educator, weight modification (diet, exercise) for maximal overall WMD of 2.0 % ([1.2, 2.8], p<0.001, I2=86.5 %) for management, encouraged to join Weight weight loss and medication optimization BSG patients. The CTG was found to have a weighted mean Watchers; clinic visits every 3 months baseline HbA1C of 7.7 % (6.8, 8.5), with a follow-up of 7.2 % (6.6, 7.7). The WMD in HbA1C levels following conventionalBariatric procedure(s) RYGB (50.0 %) LSG (50.0 %) RYGB + intensive medical therapy was 0.47 % ([0.1, 0.9], p<0.05; I2=90.1 %). Similar- management ly, the BSG had a baseline weighted mean FPG of 150.3 mg/ dL (135.7, 164.9), with a follow-up of 95.3 mg/dL (89.3,Countrya USA Ikramuddin et al. [55] USA 101.3). The WMD was 53.3 mg/dL ([40.0, 66.7], p<0.001; I2=96.8 %). Finally, baseline weighted mean FPG for theTable 1 (continued) Schauer et al. [54] CTG was 143.1 mg/dL (129.8, 156.3), with a follow-up of Study 123.2 mg/dL (113.3, 133.1). The WMD in FPG levels for conventional therapy patients was 17.4 mg/dL ([8.8, 26.0], p<0.001; I2=89.2 %). Figures 2 and 3 depict trend lines characterizing the relative changes over time in BMI and HbA1C levels for BSG and CTG groups stratified by study design. Assessment of Between-Group Differences in Clinical Markers After Treatment The WMD comparing BSG and CTG on follow-up BMI for combined OS data (k=11; n=5,257) was −8.5 kg/m2 ([−10.2, −6.9], p<0.001; I2=93.0 %); whereas, the corresponding WMD for combined RCT data (k=5; n=440) was −7.7 kg/m2 ([−10.1, −5.3], p<0.001; I2=87.5 %), with a high degree of 95 % CI overlap. No significant heterogeneity (Q p value=0.573) was found between the OSs’ summary estimate and the RCTs’ summary estimate with respect to the magnitude and direction of treatment effect on BMI, with the surgery group favored in both study designs. The WMD comparing BSG vs CTG on follow-up HBA1C, for combined OS data (k=8; n=1,131) was −0.89 % ([−1.3, −0.45], p<0.001; I2=91.7 %); whereas, the corre- sponding WMD for combined RCT data (k=4; n=370) was

Table 2 Characteristics of comparative patient groupsStudy N at baseline Body mass index, kg/m2 HbA1C, % Fasting plasma glucose, mg/dL Mean (SEa) Mean (SEa) Mean (SEa) Bari. Conv. Bari. Conv. p valuea Bari. Conv. p valuea Bari. Conv. p valueaSjöström et al. [40] 1,845 1,660 42.3 (0.10) 40.0 (0.11) <0.001 – – – 97.3 (0.88) 93.7 (0.84) <0.01O’Brien et al. [41] 40 40 33.7 (0.29) 33.5 (0.22) NS (0.58) – – – 95.4 (5.41) 90.1 (1.71) NS (0.35)Dixon et al. [42] 30 30 37.0 (0.49) 37.2 (0.46) NS (0.77) 7.8 (0.22) 7.6 (0.26) NS (0.55) 156.7 (7.03) 158.6 (8.89) NS (0.87)Hofsø et al. [43] 76 63 46.7 (0.65) 43.3 (0.63) <0.001 7.1 (0.15) 5.8 (0.15) <0.001 122.5 (4.76) 115.3 (3.86) NS (0.25)Adams et al. [44] 294 522 47.9 (0.47) 45.0 (0.39) <0.001 5.7 (0.05) 5.6 (0.06) NS (0.54) 96.9 (1.15) 96.3 (1.30) NS (0.76)Serrot et al. [45]c 17 17 34.6 (1.38) 34.0 (1.29) NS (0.75) 8.2 (0.32) 7.0 (0.28) <0.005 – – –Martins et al. [46] 50 129 45.2 (0.76) 45.6 (0.51) NS (0.67) – – – – – –Iaconelli et al. [47] 22 28 50.5 (1.81) 51.5 (1.17) NS (0.63) 8.0 (0.28) 8.0 (0.22) NS (0.99) 156.7 (10.7) 156.7 (7.00) NS (0.99)Scopinaro et al. [48] 30 38 30.6 (0.53) 30.2 (0.57) NS (0.62) 9.3 (0.27) 8.3 (0.13) <0.001 220.0 (12.6) 171.0 (6.16) <0.001Leonetti et al. [49] 30 30 41.3 (1.10) 39.0 (1.00) NS (0.13) 7.9 (0.38) 8.1 (0.31) NS (0.69) 166.0 (12.4) 183.0 (11.6) NS (0.32)Mingrone et al. [50] 40 20 45.0 (1.03) 45.6 (1.39) NS (0.73) 8.7 (0.25) 8.5 (0.27) NS (0.62) 173.4 (9.68) 179.0 (13.8) NS (0.74)Heo et al. [51] 261 224 39.0 (0.38) 34.3 (0.25) <0.001 – – – – – –Dorman et al. [52] 29 29 42.4 (0.56) 40.2 (0.80) <0.05 7.2 (0.20) 7.2 (0.22) NS (0.99) – – –Leslie et al. [53] 152 115 47.4 (0.54) 40.7 (0.47) <0.001 7.6 (0.10) 7.2 (0.10) <0.01 – – –Schauer et al. [54] 100 50 36.6 (0.36) 36.3 (0.42) NS (0.61) 9.4 (0.16) 8.9 (0.20) NS (0.06) 178.5 (5.27)c 155.0 (6.29)c <0.01Ikramuddin et al. [55] 60 60 34.9 (0.39) 34.3 (0.40) NS (0.28) 9.6 (0.13) 9.6 (0.16) NS (0.99) 222.0 (9.94) 207.0 (7.36) NS (0.23)Total 3,076 3,055 – – – – – – – – –IV Weighted mean – – 40.9 (1.23) 39.4 (1.10) 8.0 (0.50) 7.7 (0.45) 150.3 (7.5) 143.1 (6.77)(95 % CI) (38.5, 43.3) (37.3, 41.6) (7.1, 9.0) (6.8, 8.5) (135.7, 164.9) (129.8, 156.3)Pooled SMDb – – 0.33 (0.09) – 0.39 (0.14) 0.15 (0.07)(95 % CI) – (0.16, 0.51) (0.12, 0.67) – (0.02, 0.28)p value <0.001 <0.01 <0.05a Standard errors and p values are representative of meta-analytic data/results OBES SURGb Cohen’s d (SMD) is interpreted as: 0.2=small effect size (or small standardized mean difference), 0.5=medium effect size, 0.8=large effect sizec Median value used instead of mean for summary calculations

Table 3 Pre/postcomparative body mass index, HbA1C, and fasting plasma glucose outcomes OBES SURGStudy Bariatric procedure Conventional treatment Pre Pre Post Change (95 % CI) p valuea Post Change (95 % CI) p valueaBody mass index, kg/m2 mean (SEa) <0.001 −0.04 (−0.38,0.30) NS (0.82) <0.001 2.0 (0.36, 3.64) <0.05Sjöström et al. [40] 42.3 (0.10) 32.4 (0.12) 9.9 (9.6, 10.2) <0.001 40.0 (0.11) 40.04 (0.13) 0.6 (−1.4, 2.6) NS (0.56) 26.4 (0.58) 7.3 (6.0, 8.6) <0.001 33.5 (0.22) 31.5 (0.91) 3.7 (1.9, 5.5) <0.001O’Brien et al. [41] 33.7 (0.29) 29.5 (0.67) 7.5 (5.9, 9.1) <0.001 37.2 (0.46) 36.6 (0.92) 0.4 (−0.69, 1.5) NS (0.47) 32.7 (0.57) 14.0 (12.3, 15.7) <0.001 43.3 (0.63) 39.6 (0.68) NS (0.87)Dixon et al. [42] 37.0 (0.49) 32.2 (0.47) 15.5 (14.2,16.8) <0.001 45.0 (0.39) 44.6 (0.39) −0.3 (−3.9, 3.3) <0.001 25.8 (1.21) 8.8 (5.2, 12.4) <0.001 34.0 (1.29) 34.3 (1.32) 4.9 (3.5, 6.3) <0.001Hofsø et al. [43] 46.7 (0.65) 31.1 (0.71) 14.1 (12.1, 16.1) <0.001 45.6 (0.51) 40.7 (0.48) 7.9 (4.9, 10.9) NS (0.99) 34.6 (1.07) 15.9 (11.8, 20.0) <0.001 51.5 (1.17) 43.6 (1.02) 0.0 (−1.6, 1.6) NS (0.56)Adams et al. [44] 47.9 (0.47) 25.3 (0.42) 5.3 (4.0, 6.6) <0.001 30.2 (0.57) 30.2 (0.58) NS (0.22) 28.3 (0.99) 13.0 (10.1, 15.9) <0.001 39.0 (1.00) 39.8 (0.91) −0.8 (−3.5, 1.9) <0.05Serrot et al. [45] 34.6 (1.38) 29.3 (0.62) 15.8 (13.4, 18.1) <0.001 45.6 (1.39) 43.1 (1.51) 2.5 (−1.5, 6.5) NS (0.73) 30.2 (0.75) 8.8 (6.9, 10.7) <0.001 34.3 (0.25) 32.0 (0.89) 2.3 (0.89, 3.7) NS (0.88)Martins et al. [46] 45.2 (0.76) 27.6 (0.52) 14.8 (13.3, 16.3) <0.001 40.2 (0.80) 40.6 (0.84) <0.05 32.4 (0.53) 15.0 (13.5, 16.5) <0.001 40.7 (0.47) 40.8 (0.47) −0.4 (−2.7, 1.9) <0.001Iaconelli et al. [47] 50.5 (1.81) 27.0 (0.37) 9.6 (8.6, 10.6) 36.3 (0.42) 34.4 (0.79) −0.1 (−1.4, 1.2) 25.8 (0.46) 9.1 (7.9, 10.3) <0.001a 34.3 (0.40) 31.6 (0.49) <0.001aScopinaro et al. [48] 30.6 (0.53) 29.4 (0.81) 11.4 (0.73) WMD 39.4 (1.10) 37.8 (1.08) 1.9 (0.23, 3.57) <0.001 2.7 (1.5, 3.9) NS (0.28)Leonetti et al. [49] 41.3 (1.10) (27.8, 30.9) (10.0, 12.9) <0.05 (37.3, 41.6) (35.6, 39.9) 1.6 (0.49) WMD <0.05 NS (0.16) <0.05Mingrone et al. [50] 45.0 (1.03) 6.0 (0.15) 1.8 (1.3, 2.3) <0.001 7.6 (0.26) 7.2 (0.25) (0.7, 2.6) NS (0.80) 6.6 (0.12) 0.5 (0.1, 0.9) <0.001 5.8 (0.15) 6.3 (0.14) <0.05Heo et al. [51] 39.0 (0.38) 5.6 (0.05) 0.1 (−0.04, 0.24) <0.001 5.6 (0.06) 5.8 (0.06) 0.4 (−0.3, 1.1) <0.005 6.1 (0.25) 2.1 (1.3, 2.9) <0.001 7.0 (0.28) 7.1 (0.27) −0.5 (−0.9, −0.1) <0.05Dorman et al. [52] 42.4 (0.56) 5.2 (0.22) 2.8 (2.1, 3.5) <0.001 8.0 (0.22) 7.2 (0.21) −0.2 (−0.4, −0.1) <0.05 6.5 (0.15) 2.8 (2.2, 3.4) <0.001 8.3 (0.13) 7.7 (0.11) −0.1 (−0.9, 0.7) NS (0.77)Leslie et al. [53] 47.4 (0.54) 6.0 (0.27) 1.9 (1.0, 2.8) <0.001 8.1 (0.31) 7.1 (0.24) NS (0.99) 5.7 (0.16) 3.0 (2.4, 3.6) <0.001 8.5 (0.27) 7.7 (0.14) 0.8 (0.2, 1.4) <0.001Schauer et al. [54] 36.6 (0.36) 5.9 (0.19) 1.3 (0.8, 1.8) <0.001 7.2 (0.22) 7.3 (0.26) 0.6 (0.3, 0.9) <0.05 6.4 (0.10) 1.2 (0.9, 1.5) 7.2 (0.10) 7.2 (0.10) 1.0 (0.2, 1.8)Ikramuddin et al. [55] 34.9 (0.39) 6.5 (0.10) 2.9 (2.5, 3.3) <0.001a 8.9 (0.20) 7.5 (0.28) 0.8 (0.2, 1.4) <0.05a 6.3 (0.12) 3.3 (3.0, 3.7) 9.6 (0.16) 7.8 (0.20) −0.1 (−0.8, 0.6)Weighted mean 40.9 (1.23) 6.1 (0.15) 2.0 (.40) WMD <0.001 7.7 (0.45) 7.2 (0.27) 0.0 (−0.3, 0.3) <0.001 NS (0.35) 1.4 (0.7, 2.1) NS (0.99)(95 % CI) (38.5, 43.3) (5.8, 6.4) (1.2, 2.8) <0.001 (6.8, 8.5) (6.6, 7.7) 1.8 (1.3, 2.3) NS (0.09)HbA1C, % mean (SEa) 7.8 (0.22) 0.47 (0.19) WMDDixon et al. [42] 84.1 (0.7) 13.2 (11.0, 15.4) 93.7 (0.84) 98.5 (0.8) 88.4 (5.3) 7.0 (−7.8, 21.8) 90.1 (1.7) 89.8 (7.6) (0.1, 0.9)Hofsø et al. [43] 7.1 (0.15) 105.6 (5.5) 51.1 (33.6, 68.6) 158.6 (8.9) 139.6 (7.0) −4.8 (−7.1, −2.5)Adams et al. [44] 5.7 (0.05) 0.3 (−13.3, 13.9) 19.0 (−3.1, 41.1)Serrot et al. [45] 8.2 (0.32)Iaconelli et al. [47] 8.0 (0.28)Scopinaro et al. [48] 9.3 (0.27)Leonetti et al. [49] 7.9 (0.38)Mingrone et al. [50] 8.7 (0.25)Dorman et al. [52] 7.2 (0.20)Leslie et al. [53] 7.6 (0.10)Schauer et al. [54] 9.4 (0.16)Ikramuddin et al. [55] 9.6 (0.13)Weighted mean 8.0 (0.50)(95 % CI (7.1, 9.0)Fasting plasma glucose, mg/dL mean (SEa)Sjöström et al. [40] 97.3 (0.88)O’Brien et al. [41] 95.4 (5.4)Dixon et al. [42] 156.7 (7.0)

Table 3 (continued) Bariatric procedure Post Change (95 % CI) p valuea Conventional treatment Post Change (95 % CI) p valueaStudy Pre 88.3 (3.3) 34.2 (22.8, 45.6) <0.001 Pre 100.9 (4.3) 14.4 (3.1, 25.7) <0.05 82.0 (1.2) 14.9 (11.7, 18.0) <0.001 97.3 (1.3) −1.0 (−4.6, 2.6) NS (0.58)Hofsø et al. [43] 122.5 (4.8) 75.7 (6.2) 81.0 (56.8, 105.2) <0.001 115.3 (3.9) 126.1 (6.4) 30.6 (11.9, 49.2) <0.01Adams et al. [44] 96.9 (1.2) 149.0 (7.5) 71.0 (42.3, 99.7) <0.001 96.3 (1.3) 151.0 (4.5) 20.0 (5.0, 35.0) <0.01Iaconelli et al. [47] 156.7 (10.7) 97.0 (5.3) 69.0 (42.5, 95.5) <0.001 156.7 (7.0) 150.0 (8.8) 33.0 (4.5, 61.5) <0.05Scopinaro et al. [48] 220.0 (12.6) 86.3 (5.5) 87.1 (65.0,109.2) <0.001 171.0 (6.2) 141.1 (7.1) 37.9 (6.5, 69.3) <0.05Leonetti et al. [49] 166.0 (12.4) 98.0 (3.3) 80.5 (68.3, 92.7) <0.001 183.0 (11.6) 120.0 (6.6) 35.0 (17.0, 53.0) <0.001Mingrone et al. [50] 173.4 (9.7) 111.0 (4.5) 111.0 (89.2,132.8) <0.001 179.0 (13.8) 153.0 (7.8) 54.0 (33.0, 75.0) <0.001Schauer et al. [54]b 178.5 (5.3) 95.3 (3.1) 53.3 (6.8) WMD 155.0 (6.3) 123.2 (5.1) 17.4 (4.4) WMDIkramuddin et al. [55] 222.0 (9.9) <0.001a 207.0 (7.4) <0.001aWeighted mean 150.3 (7.5) (89.3, 101.3) (40.0, 66.7) 143.1 (6.8) (113.3, 133.1) (8.8, 26.0)(95 % CI) (135.7, 164.9) (129.8, 156.3)WMD weighted mean differencea Standard errors and p values are representative of meta-analytic resultsb Median value used instead of mean for summary calculations OBES SURG Fig. 2 Mean body mass index (BMI) reduction in bariatric surgery patients and conventional therapy patients by study design (randomized controlled trial vs observational) −1.43 % ([−2.1, −0.81], p<0.001; I 2=66.2 %), with signif- icant 95 % CI overlap. No significant heterogeneity (Q p value=0.16) was found between the OSs’ summary esti- mate and the RCTs’ summary estimate with respect to the magnitude and direction of treatment effect on HBA1C, with the surgery group favored in both study designs. The WMD comparing BSG vs CTG on follow-up FPG, for combined OS data (k=6; n=4,460) was −20.9 mg/dL ([−29.3, −12.5]; p<0.001; I2=84.3 %); whereas, the corresponding WMD for combined RCT studies (k=5; n=440) was −30.1 mg/dL ([−40.8, −19.5], p<0.001; I2=80.5 %), with a significant 95 % CI overlap. No significant heterogeneity Fig. 3 Mean glycated hemoglobin (HbA1C) reduction in bariatric surgery patients and conventional therapy patients by study design (randomizedcontrolled trial vs observational)

OBES SURG(Q p value=0.18) was found between the OSs’ summary a further breakdown of %EWL and T2DM remission byestimate and the RCTs’ summary estimate with respect to the treatment group and study design. Bariatric surgery patientsmagnitude and direction of treatment effect on FPG, with the enrolled in RCT designs reported the highest mean EWL (80.0 %); bariatric surgery patients enrolled in OSs had thesurgery group favored in both study designs. highest T2DM remission rate (65.6 %). In summary, independent treatment effect sizes for T2DM remission event data, PORs, and 95 % CIs describ- ing the effects of surgery vs conventional therapy on T2DMboth OSs and RCTs were sufficiently concordant to remission are presented in the final forest plot of Fig. 4. Independent summary estimates were calculated for RCTspermit estimation of an overall effect for each analysis and OSs using the Peto method. The POR and 95 % CI for combined OS remission event data (k=11; n=1,489) was 10.9presented in Fig. 4. The first three meta-analytic results ([6.1, 19.5], p<0.001; I2=81.6 %); whereas, the correspond- ing POR and 95 % CI for combined RCT remission event data(BMI, HbA1C, and FPG) provide individual study mean (k=5; n=404) was 7.8 ([3.3, 18.4], p<0.001; I2=70.2 %), withfollow-up differences and SEs, as well as independent a high degree of 95 % CI overlap. No significant heterogeneity (Q p value=0.52) was found between the OSs’ estimate andsummary estimates for OSs and RCTs, and the overall the RCTs’ estimate with respect to the magnitude and direc- tion of treatment effect on T2DM remission, with the surgeryWMD for combined included studies. Negative mean differ- group favored in both study designs. Evaluation by z score also indicated no statistically significant difference betweenence values indicate a treatment effect favoring surgical inter- OS and RCT summary estimates (z=−0.57; p=0.57; a z score ≤−1.96 or ≥1.96 would indicate a statistically significantvention; OR results favored surgery over conventional therapy difference at the 0.05 level). Thus, estimates were sufficiently concordant to calculate an overall effect (i.e., combiningwhere values and plotting points comprising the forest chart data from RCTs with OSs). As shown in Fig. 4, the (k=16; n=1,893) summary POR was 9.8 ([6.1, 15.9], p<0.001;are >1.0. I2=78.4 %). The overall (k=16; n=5,697) between-group WMD char- Independent PORs and 95 % CIs were also calculatedacterizing BMI outcomes was −8.3 kg/m2 ([−9.6, −6.9], using the inverse variance method as a direct comparison top<0.001; I2=91.8 %), favoring the surgery group. The corre- Peto findings. The POR and 95 % CI for combined OS datasponding pooled SMD was −1.62 ([−1.8, −1.4]; p<0.001; was 18.9 ([8.1, 43.7], p<0.001; I2=79.9 %); whereas, theI2=90.1 %); adjusted effect size, −1.95 [−2.15, −1.76]. corresponding POR and 95 % CI for combined RCT dataAnalysis of BMI using studies with complete data (k=6 was 11.0 ([3.3, 36.3], p<0.001; I2=59.9 %), with significantstudies, no imputation) yielded a WMD of −9.5 kg/m2 95 % CI overlap. No significant heterogeneity (Q p value=([−12.3, −6.6], p<0.001; I2=93.6 %) and corresponding 0.47) was found between the OS and RCT summary esti-pooled SMD of −2.1([−2.7, −1.6], p<0.001; I2=85.2 %). mates, where the surgery group was favored in both study designs. Evaluation by z score also indicated no statistically WMD calculations summarizing between-group compari- significant difference between the independent summary esti- mates (z=−0.60; p=0.55), the inverse variance methodsons along the HbA1C and FPG variable outcomes yielded yielded an overall summary POR of 15.8 ([7.9, 31.4],similar results to those found for BMI. The overall (k=12; n= p<0.001; I2=75.2 %). Thus, according to the inverse variance1,501) WMD for HBA1C outcomes was −1.1 % ([−1.6, −0.6], method, the odds of T2DM remission in patients undergoingp<0.001; I2=91.9 %), again, favoring the surgery group. The bariatric surgery were, on average, 15.8 times the odds ofcorresponding pooled SMD was −1.0 ([−1.4, −0.6], p<0.001; remission for those receiving conventional therapy. AllI2=89.2 %); adjusted effect size, −1.39 ([−1.72, −1.01]). but one of 16 studies indicated a clear statistical advantageAnalysis of HbA1C using studies with complete data (k=8 favoring surgery.studies, no imputation) yielded a WMD of −1.3 % ([−1.54,−0.98], p<0.001; I2=65.6 %) and pooled SMD of −1.13 Sensitivity and Subgroup Analyses([−1.4, −0.8], p<0.001; I2=73.2 %). Finally, the overall (k=11; n=4,900) between-groupFPG WMD was −24.9 mg/dL ([−33.9, −15.9]; p<0.001;I2=84.8 %), favoring the surgery group. The pooledSMD was −0.71 ([−0.92, −0.50], p<0.001; I2=80.5 %;adjusted effect size, −0.86 [−1.05, −0.67]). Analysis ofFPG using studies with complete data (k=5 studies, noimputation) yielded a WMD of −36.8 mg/dL ([−56.3,−17.3], p<0.001; I2=83.1 %) and pooled SMD of −0.96([−1.5, −0.5], p<0.001; I2=79.9 %).Effects of Bariatric Surgery vs Conventional Therapyon %EWL and T2DMOverall EWL means for the BSG and the CTG groups were Three sensitivity analyses were performed: (1) an analysis75.3 % (57.2–94.6) and 11.3 % (−5.7–29.8), respectively; excluding trials reporting no remission events; (2) an analysisoverall T2DM remission rates were 63.5 % (38.2–100.0) excluding studies that combined data from multiple studyand 15.6 % (0.0–46.7) (p<0.001; Table 4). Figure 5 presents arms; and (3) an analysis excluding all bariatric procedures

OBES SURGexcept RYGB (the most frequently performed procedure). 3 to read the detailed results of these subgroup analyses and an[See Electronic Supplementary Material, Appendices 2 and assessment of publication bias (includes Table 5 and Fig. 6).]

OBES SURGƒFig. 4 The first three tables and corresponding forest plots summarize loss and T2DM remission. Results indicated that bariatric surgery demonstrated greater BMI reduction, greater reduc- meta-analyses of the relative effects of bariatric surgery vs conventional tion in HbA1C and FPG, and a much greater likelihood of therapy on body mass index (BMI), glycated hemoglobin (HbA1C), and T2DM remission relative to patients receiving conventional fasting plasma glucose (FPG). Each study contributing to a particular therapy. meta-analysis is represented by a single darkened square contained on the forest plot; the size of the square being proportional to the amount of Heterogeneity weight the study was given during the calculation of the pooled summary estimate. The pooled estimate in the first three analyses is expressed as the According to the prespecified I2 value (≥75.0 %), a majority weighted mean difference (WMD) and is represented by the diamond of meta-analyses undertaken in this study were characterized shape at the base of each forest plot. Two additional diamonds in each by significant heterogeneity (mean I2, 81.4 % [0–96.8]; mean forest plot represent independent summary estimates for observational RCT, I2=72.0; mean OS, I2=86.0). Generally, when effect studies and randomized controlled trials. Negative WMD values indicate estimates from individual studies rest on opposite sides of the a treatment effect favoring surgical intervention. The fourth table (and reference line (i.e., the point of no effect), study results are, by forest plot) represents an analysis of the relative effects of surgery vs definition, heterogeneous, and conclusions questionable. conventional therapy on T2DM remission. In this case, the summary estimate However, the meta-analytic results presented in Fig. 4 showed of effect is given by the pooled odds ratio (POR). Results favor surgery over the vast majority of estimated effects were consistently an- conventional therapy when odds ratio values are greater than one chored on the same side of the reference line, with 95 % CIs overlapping to a great extent. Thus, when viewed in isolation, Discussion the mean I2 statistic may be somewhat misleading. The current meta-analysis systematically identified and inte- grated a wide range of evidence regarding the effectiveness of bariatric surgery vs conventional therapy in promoting weightTable 4 Excess weight loss and diabetes remissionStudy %EWLa T2DM remission rate Mean % (N) Bariatric Conventional Bariatric Conventional p valuebSjöström et al. [40] 57.2 −0.3 72.0 (342) 21.0 (248) <0.001O’Brien et al. [41] 83.9 23.5 93.0 (15) 46.7 (15) <0.01Dixon et al. [42] 62.5 4.9 73.0 (30) 13.0 (30) <0.001Hofsø et al. [43] 64.5 20.2 79.0 (14) 0.0 (6) <0.005Adams et al. [44] 69.6 1.9 78.7 (61) 2.6 (114) <0.001Serrot et al. [45] 91.7 3.3 64.7 (17) 0.0 (17) <0.001Martins et al. [46] 69.8 23.8 67.0 (6) 36.8 (38) NS (0.17)Iaconelli et al. [47] 62.4 29.8 100.0 (22) 45.0 (28) <0.001Scopinaro et al. [48]c 94.6 0.0 83.0 (30) 0.0 (38) <0.001Leonetti et al. [49] 79.8 −5.7 80.0 (30) 0.0 (30) <0.001Mingrone et al. [50] 79.0 12.1 85.0 (40) 0.0 (20) <0.001Heo et al. [51] 62.9 24.7 57.1 (84) 9.5 (21) <0.001Dorman et al. [52] 85.1 −2.6 65.0 (29) 3.4 (29) <0.001Leslie et al. [53] 67.0 −0.6 38.2 (152) 17.4 (115) <0.001Schauer et al. [54]d 82.8 16.8 39.4 (99) 12.0 (41) <0.005Ikramuddin et al. [55] 91.9 29.0 49.0 (57) 19.0 (57) <0.001Overall 75.3 11.3 63.5 (1,028) 15.6 (847) <0.001(Range) (57.2–94.6) (−5.7–29.8) (38.2–100.0) (0.0–46.7)Negative values in the %EWL column denote mean weight gainEWL excess weight loss, T2DM type 2 diabetes mellitusa Standardized calculation using BMI 25 as ideal weight constantb Z test for two population proportionsc Patients with controlled diabetes following treatment included in remission rate calculationd Patients recovering from metabolic syndrome following treatment included in remission rate calculation

OBES SURGFig. 5 Mean percent excess weight loss (%EWL) in bariatric surgery systematic reviews (SRs) and meta-analyses (MAs) [62, 63].patients and conventional therapy patients by study design type (random- A SR by Tsai et al., focused on weight-loss programs acrossized controlled trial vs observational) the USA, saw 15.0–25.0 % excess body weight loss over 3– 6 months, although fewer than 9.0 % of the patients main- One source of heterogeneity was expected as a result of tained their weight loss at 12 months [62]. In a MA byintegration of RCT and OS data. The majority of bariatric Dansinger et al. of 46 trials that provided dietary counseling,surgery studies are observational in design, as RCTs are less BMI was reduced by 1.9 units at 1 year (comparable to the 1.6feasible to conduct for ethical and economic reasons [58]. BMI units lost by the conventional therapy group in theMost RCTs are conducted in high-volume “centers of excel- current meta-analysis), with a per-month BMI loss of approx-lence” according to regulated protocols that conform to na- imately 0.1 unit between 3 and 12 months of the activetional guidelines [59, 60]; therefore, RCT outcomes may differ counseling programs, with a BMI regain of approximatelysignificantly from OSs. Yet, interestingly, the current review’s 0.02–0.03 unit per month during follow-on maintenancecomparative analysis of trends in outcome variables, stratified phases [63].by study design, was highly similar (Figs. 2, 3, and 5 [seeFig. 5 in online ESM 3]). In addition, meta-analytic results Interestingly, while excluded from the current analysis, aindicated no significant heterogeneity between RCT and OS 2005 comparative study, by Ritt et al., of 24 LAGB vs 16summary estimates quantifying the relative effects of bariatric conventional therapy patients reported an atypically largesurgery vs conventional therapy on BMI, HbA1C, and FPG EWL (54.5 %) [64], well outside the mean EWL (11.3 %)reduction. In addition, both study designs demonstrated the for conventional therapy patients reported in our review. Also,superiority of bariatric surgery over conventional therapy in in a 2007 OS by Anderson et al. of 1,531 morbidly obesepromoting T2DM remission. A common criticism of OSs, in conventional weight-loss patients with long-term follow-upgeneral, is that they produce exaggerated effect sizes. While it employing very intensive behavioral intervention (e.g.,is true that the POR for T2DM remission derived from OS weight-loss camps, residential nursing programs, and closelystudies was larger than that derived from RCTs (Peto, 10.9 supervised individualized outpatient programs, possibly cost[6.1, 19.5] vs 7.8 [3.3, 18.4]; inverse variance, 18.9 [8.1, 43.7] prohibitive for many patients), marked weight loss wasvs 11.0 [3.3, 36.3]), no significant heterogeneity was found achieved up to 100 lbs over the short term; however, overbetween summary estimates. Further, Shrier et al., in their 1–5-year follow-up, most patients regained 34.0–41.0 %review of the principal elements underlying this claim, found of their lost weight [65].that both study designs have strengths and weaknesses, andincluding OSs would increase precision appropriately, and Bariatric Surgerymay produce equally or more relevant and valid results [61]. The current findings for bariatric surgery in relation to BMIWeight reduction (WMD, 11.4 kg/m2 [10.0, 12.9]) and EWL (75.3 % [57.2–94.6]) were similar to those of two SR/MAs of theConventional Treatment bariatric literature by Buchwald et al. [5, 22]; aggregated weight outcomes in morbidly obese patients found 13.6 kg/m2To place the current meta-analytic findings in context, they (12.9, 14.3) and 14.2 kg/m2 (13.2, 15.1) BMI reduction, 64.7 %should be compared to publications outside of the included (32.0–93.0) and 55.9 % (54.1, 57.8) EWL, respectively [5, 22].study set. The current findings in relation to BMI reduction Also, an SR by Gill et al. of T2DM patients who underwent(WMD 1.6 kg/m2 [0.7, 2.6]) and EWL (11.3 % [−5.7–29.8]) sleeve gastrectomy (SG) with a mean 13 months of follow-upafter conventional treatment are similar to those of key noted 47.0 % EWL [66]. Conventional Treatment vs Bariatric Surgery A rigorous 2009 Cochrane review by Colquitt et al. of mildly and morbidly obese patients both with and without diabetes found that over the short term, bariatric surgery achieved substantially greater weight loss than medical therapy [67], as was concluded also by Maggard-Gibbons et al., of mildly obese diabetic patients in their 2013 SR [68]. Padwal et al.’s 2011 SR/MA of 31 RCTs, in which bariatric surgery was compared with other bariatric surgery controls or standard care (n=2,619; BMI 42.0–58.0) using network analysis, found conventional treatment significantly less effective than

OBES SURGsurgery over the short and intermediate terms. Padwal et al. Buchwald et al., diabetes was completely resolved in 76.8 %also found an average BMI WMD of 8.35 kg/m2 (2.4–11.4) in of bariatric surgery patients [5]; in the same authors’ 2009 SR/their subset comparative study of all surgery vs conventional MA that focused on T2DM outcomes in 103 treatment armstherapy [69]. The current review’s between-group BMI WMD (n=3,188), at 2-year post bariatric surgery, complete diabeteswas nearly identical (8.3 kg/m2 [7.0, 9.6]) to the results of resolution (defined therein as normal FPG and no anti-diabeticPadwal et al., which would be expected since the current medications) was attained by 78.1 %; at ≥2 years, 74.6 %authors grouped procedures for analysis. In addition, the continued resolved [22]. An MA of low-BMI bariatric surgerypooled adjusted SMD (corresponding to the BMI WMD of patients by Li et al. reported a significant mean decrease in8.3 kg/m2) was 1.95 (1.76, 2.15), indicating that the average HbA1C of 2.59 % in 80.0 % of patients after surgery,surgery patient experienced a BMI outcome superior to that of consonant with the decrease found in our meta-analysis97.1 % of the conventional therapy group. While both treat- (2.0 % [1.2, 2.8]) [72].ment groups in the current review can be said to have expe-rienced a statistically significant within-group BMI change, Conventional Treatment vs Bariatric Surgeryonly the bariatric surgery patients experienced a clinicallymeaningful weight reduction, as demonstrated by a mean shift The overall T2DM remission rates for surgery patients vsdownward by as much as two obesity classifications. conventional therapy patients presented in the current analysis were significantly different, 63.5 vs 15.6 % ( p<0.001), re-Diabetes spectively. The relative efficacy of bariatric surgery and con- ventional therapy in promoting T2DM remission was furtherConventional Treatment quantified in the meta-analytic calculation of PORs based on 16 studies, 94.0 % of which demonstrated a distinct statisticalThe long-running Look AHEAD study of weight loss by advantage in favor of bariatric surgery. When the analysis wasconventional treatment (n=5,145 overweight adults with run excluding trials reporting no remission events (which tendT2DM) found an association between intensive weight-loss to produce inflated ORs), the summary point estimates wereintervention (ILI) and a reduction in T2DM markers: Mean Peto POR of 6.9 (4.1, 11.6; inverse variance, 9.4 [5.0, 17.7],HbA1C dropped from 7.25 to 6.61 % (p<0.00). Although the p<0.001; I2=75.4 %), still indicative of a large effect sizedegree of success in ILI diabetes reduction experienced in the favoring surgery. A second sensitivity analysis indicated thatLook AHEAD study resulted in a seemingly greater outcome the odds of RYGB patients reaching T2DM remission werethan that of the current meta-analysis, the Look AHEAD ILI 8.0 times those of patients treated conventionally.group started at a lower HbA1C than the conventional group inthe current analysis, and their overall HbA1C reduction was Remission event data were supported by parallel clinicalproportionally similar. By this measure, Look AHEAD results evidence of T2DM remission in the form of HbA1C outcomesuggest that this form of therapy might be appropriate to those data. The pooled adjusted SMD quantifying the relative treat-patients hovering around the 7.0 % HbA1C mark at baseline. ment effects of surgery vs conventional therapy on HbA1CLook AHEAD FPG outcomes for the ILI group dropped from was 1.37 (1.01, 1.72), indicating that the average surgery151.9 to 130.4 mg/dL, and were essentially identical to those patient experienced an HbA1C outcome superior to that ofof the conventional group in the current analysis [70]; the FPG 90.9 % of the conventional therapy group. RCT and OS studyreduction of the current analysis compared favorably, 143.1 to surgery patients experienced both statistically and clinically123.2 mg/dL. significant reduction in HbA1C, moving from baseline levels of 8.9 and 7.6 %, respectively, to 6.1 %, well below the ADABariatric Surgery target of ≤7.0 %. Conventional therapy patients also experi- enced a statistically significantly reduction in HbA1C relativeThe current meta-analytic findings for overall T2DM remis- to baseline; however, whether participating in an RCT or OS,sion (63.5 % [38.2–100.0]) and FPG reduction of 34.5 % this group continued to experience poor glycemic control(down to 95.3 mg/dL) after bariatric surgery are similar to (≥7.0 %) at follow-up. Glycated hemoglobin is a primarythose of a 2011 cross-sectional nRCT by Reed et al., in which clinical marker and predictor of T2DM. Diabetic patients havebariatric surgery patients were observed prior to and at 1 week an 11.0 % increased risk of mortality [73] from ischemic heartand 3 months following RYGB; FPG decreased to levels disease and, those with an HbA1C >8 % are subject to asimilar to those (≤125/mg/dL) of lean controls (BMI ≤25), 150.0 % increased risk of death from heart disease [74].and diabetes was considered resolved [71]. In the aforemen- Further, each 1.0 % increase in HbA1C has been shown to betioned SR of T2DM SG patients by Gill et al., a rate of 66.0 % associated with a 20.0–30.0 % increase in cardiovascular events,remission was recorded [66], similar to our findings for the and all-cause mortality independent of diabetes status [75].bariatric surgery group (63.5 %). In the 2004 SR/MA by The future will offer additional comparative evidence from high-quality, experimental studies of diabetes remission

OBES SURGfollowing bariatric surgery vs conventional treatment. A mid- recently published studies, very similar results were found for2013 publication describes the start of a well-designed multi- patients with BMI <35 [82].center RCT of lower-BMI insulin-dependent patients who willundergo RYGB plus standard medical treatment, if needed, The current meta-analytic findings relevant to treatingcompared with controls who will receive only standard T2DM lower-BMI diabetic patients showed no significant heteroge-treatment; this RCT, the DiaSurg Trial, with a target size of neity between low- and high-BMI subgroups in summary400 participants, will provide long-term comparative data estimates quantifying the relative effects of bariatric surgery(8 years) on T2DM outcomes following bariatric surgery vs vs conventional therapy on T2DM remission. The high-BMIconventional diabetes care [76]. (BMI≥35) and lower-BMI (BMI<35) subgroups had respec- tive PORs of 15.2 (6.8, 34.1) and 17.1 (4.7, 62.9). ThisGlycemic Control suggests that lower-BMI patients may also be able to experi- ence the T2DM-reduction benefits of bariatric surgery longTwo studies in the current comparative meta-analysis provide observed in higher-BMI patients.reason to believe that the marked improvement in glycemiccontrol following bariatric surgery is not only more pro- Patients whose diabetes may be treated more effectively bynounced than with conventional therapy, but more durable bariatric surgery than conventional therapy, as found in theover the long term. Sjöström et al. reported a 96.0 % reduction current meta-analysis that sought patients with a BMI ≥25,in the risk of developing T2DM and a 36.0 % rate of T2DM may benefit from the opportunity to elect T2DM treatment byremission maintenance in bariatric surgery patients at 10-year surgery. The 2009 Asian Indian Consensus Statement, forfollow-up vs an increase in T2DM incidence in the usual care example, recommended lowering the BMI cutoff for bariatricpatient population [41]. Similarly, Iaconelli et al. reported surgery with a comorbidity to 32.5 [83], making surgical100.0 % prolongation of T2DM remission at 10-year follow- treatment for diabetes available to more patients who, becauseup vs 45.0 % in the medical therapy group [48]. In addition, a of their ethnicity, may develop severe disease at lower weightsrecently published follow-up report of the original, included, than Caucasians [84]. Also, the 2011 International DiabetesO’Brien et al. study [77] shows sustained weight loss and Federation (IDF) Task recommendation for clinical practice asmetabolic syndrome resolution after 10 years. All three of it relates to T2DM suggested that bariatric surgery should bethese long-term findings support those of Pories et al., who considered an appropriate treatment under certain circum-found >80.0 % improvement in T2DM maintained at 14 years stances, including failure of conventional weight and T2DMpost surgery [14], and of Li et al., whose MA of patients with therapy to control diabetes in those with a BMI of 30–35 [85].<35 BMI showed several patients in whom diabetes resolution Currently, however, the US National Institutes of Healthwas maintained at 18 years following surgery [72]. (NIH) guidelines for bariatric surgery for patients with one or more comorbidity stipulate a BMI cutoff of ≥35, as do theTreating Lower-BMI Diabetic Patients Centers for Medicare and Medicaid Services (CMS), who maintain their 2006 policy of bariatric surgery approval only for patients with a BMI ≥35 [59, 86].Multiple SR/MAs reveal a gradation in diabetes resolution Limitationswith specific bariatric operations, wherein the greatest effect isassociated with malabsorptive procedures, BPD and DS Only the available literature can be evaluated by a systematic(95.0–98.9 %) [5, 22]. Yet, 90.0 % of T2DM patients are not review. The analytic power of this review was limited due tomorbidly obese [78–80]. The hypothesis of metabolic sur- the diversity of T2DM diagnosis and remission standardsgery—that certain procedures, some bariatric, performed reported. Most studies determined T2DM outcomes idiosyn-without a primary weight-loss focus, but rather with the goal cratically; the majority of studies did not define T2DM remis-of achieving improved control or remission of diseases such as sion uniformly, some employing ADA or NIH measures, anddiabetes—has now been examined in several OSs, nRCTs, others, only biochemical marker of glycemic control. Al-and RCTs [15, 58, 69, 81]. In our 2010 systematic review and though the number of available RCTs was noteworthy,integrative analysis of studies of metabolic surgery between representing almost one third of included studies, there is a1979 and 2009 for treatment of T2DM in 343 patients with shortage of well-controlled observational and level 1BMI <35, we reported 85.3 % T2DM resolution based on an (experimental) evidence comparing bariatric surgery and con-FPG reduction of 93.3 mg/dL (105.2 mg/dL, −93.3) with ventional therapy outcomes; a greater number of well-patients off antidiabetic medications, and an HbA1C reduction designed experimental studies would have increased this re-of 2.7 % into the normal range (<6.0 %) in addition to view’s predictive strength. A uniform standard for reportingsignificant, not excessive, BMI loss of 5.1 (from 29.4 to T2DM remission is needed to improve the scientific evidence24.2) [15]. In a later review by Reis et al. that included our base and support clinical decision making. Another limitationSR’s original 16 papers in addition to several additional more of the analysis was that three studies, included because they

OBES SURGmet all comparative study requirements, incorporated conven- T2DM (as measured by baseline HbA1C), preoperative BMI,tional therapy arms that included no patient treatment, and age, or weight loss.simply followed patients who received no focused weight-lossintervention as a measure of “conventional treatment”. Disclosure G. Ribaric is Head of Clinical and Medical Affairs at Ethicon Endo-Surgery Europe, GmBH, a subsidiary of Johnson & Johnson. Another possible limitation of this study was the use of J.N. Buchwald is Director and Chief Scientific Research Writer at Medwritestandard deviation imputation in order to allow for inclusion LLC, Wisconsin, USA, and T.W. McGlennon is Director, Statistical Anal-of reported mean data that were not accompanied by variance ysis and Quality of Life Assessment, McGlennon MotiMetrics (M3),data. In an effort to assess imputation bias, parallel analyses of Wisconsin, USA, both CROs under contract with Ethicon. The study wasbetween-group outcomes were carried out, incorporating into financially supported by Ethicon Endo-Surgery Europe, GmBH.the meta-analyses only those studies with complete data, asoriginally abstracted. Without exception, imputation was Open Access This article is distributed under the terms of the Creativeshown to slightly constrain effect size. Also, the grouping of Commons Attribution License which permits any use, distribution, andbariatric procedures may have introduced bias; however, sen- reproduction in any medium, provided the original author(s) and thesitivity analysis demonstrated that the grouping technique, as source are credited.applied in this research, also served to constrain effect size.Thus, SMDs reported herein likely represent relatively con- Referencesservative summary estimates. Finally, our analyses affordedmainly short-term results: Comparative data for weight and 1. Sims EH et al. Endocrine and metabolic effects of experimentalT2DM remission in bariatric surgery vs conventional therapy obesity in man. Recent Prog Horm Res. 1973;29:457–96.at follow-up time points greater than 17.3 months are needed. 2. World Health Organization. Obesity and overweight—fact sheet No.Conclusions 311. Updated March 2013. Accessed June 11, 2013: http://www. who.int/mediacentre/factsheets/fs311/en/index.htmlThe current research summarizes the best available findingsfor weight loss and T2DM remission in directly comparative 3. World Health Organization. Obesity and overweight—fact sheet no.research studies of surgery vs medical therapy in mildly obese, 312. Updated March 2013. http://www.who.int/mediacentre/obese, and morbidly obese patients, a treatment effect corrob- factsheets/fs312/en/. Accessed 11 June 2013orated by the analysis of HbA1C reduction. The current reviewdemonstrates that surgery has a dramatically greater weight- 4. Andersson C, van Gaal L, Caterson ID, et al. Relationship betweenloss effect, by ≥3 orders of magnitude, and a more expeditious HbA1C levels and risk of cardiovascular adverse outcomes and all-causeand higher rate of T2DM remission than conventional therapy. mortality in overweight and obese cardiovascular high-risk women and men with type 2 diabetes. Diabetologia. 2012;55(9):2348–55. Physicians must appraise the evidence for bariatric surgeryvs conventional therapy on a patient-by-patient basis, 5. Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: aweighing complications and costs associated with surgery systematic review and meta-analysis. JAMA. 2004;292:1724–37.against those associated with uncontrolled T2DM and/or co-morbid obesity. However, only 1.0–3.0 % of individuals who 6. The Surgeon General’s call to action to prevent and decrease over-reach class I obesity ever return to normal weight through weight and obesity/ Office of Disease Prevention and Healthconventional means [87], and a significant proportion of them Promotion; Centers for Disease Control and Prevention, Nationalgo on to develop T2DM. Thus, the current meta-analytic Institutes of Health. [Rockville, MD]: U.S. Dept. of Health andfinding, that the effectiveness of treatment with bariatric sur- Human Services, Public Health Service, Office of the Surgeongery far exceeds that of conventional therapy, carries the General; Washington, D.C., 2001.additional implication that surgery should be considered as apreventive measure. 7. Resnick HE, Foster GL, Beardsley J, et al. Achievement of American Diabetes Association clinical practice recommendations among U.S. In this systematic review and meta-analysis of 16 compar- adults with diabetes, 1999–2002: the National Health and Nutritionative studies of 6,131 patients with a baseline BMI ranging Examination Survey. Diabetes Care. 2006;29:531–7.from 30.2 to 51.5 at mean 17-month follow-up, bariatricsurgery was significantly more effective in achieving weight 8. Garber A, Henry R, Ratner R, et al. Liraglutide versus glimepirideloss, HbA1C reduction, and diabetes remission than conven- monotherapy for type 2 diabetes (LEAD-3 MONO): randomized,tional medical therapy. Based on overall summary estimates, 52-week, phase III, double blind, parallel-treatment trial.the odds of T2DM remission in patients undergoing bariatric Lancet. 2009;373:473–81.surgery were calculated to be from 9.8 to 15.8 times the oddsof remission for those patients receiving conventional therapy, 9. Wong K, Glovaci D, Malik S, et al. Comparison of demographica finding that held true regardless of study design, severity of factors and cardiovascular risk factor control among US adults with type 2 diabetes by insulin treatment classification. J Diabetes Complicat. 2012;26(3):169–74. 10. Sjöström L. Review of the key results from the Swedish Obese Subjects (SOS) trial—a prospective controlled intervention study of bariatric surgery. J Intern Med. 2013;273(3):219–34. 11. Nguyen N, Champion JK, Ponce J, et al. A review of unmet needs in obesity management. Obes Surg. 2012;22(6):956–66. 12. Christiansen T, Brunn JM, Madsen EL, et al. Weight loss mainte- nance in severely obese adults after an intensive lifestyle intervention: 2- to 4-year follow-up. Obesity. 2007;15(2):413–20. 13. Goodrick GK, Foreyt JP. Why treatments for obesity don’t last. JAMA. 1991;91:1243–7.

OBES SURG14. Pories WJ, Swanson MS, MacDonald KG, et al. Who would have 36. North American Association for the Study of Obesity (NAASO) and thought it? An operation proves to be the most effective therapy for the National Heart, Lung, and Blood Institute (NHLBI). The practical adult-onset diabetes mellitus. Ann Surg. 1995;222:339–50. guide: identification, evaluation, and treatment of overweight and obesity in adults. NIH Publication #00-4084, October 2000.15. Fried M, Ribaric G, Buchwald JN. Metabolic surgery for the treat- ment of type 2 diabetes in patients <35 kg/m2: an integrative review 37. Duriak J. How to select, calculate, and interpret effect sizes. J Pediatr of early studies. Obes Surg. 2012;20(6):776–90. Psychol. 2009;34(9):917–28.16. Scopinaro N, Papadia F, Marinari G, et al. Long-term control of type 38. Furukawa TA, Barbui C, Cipriani A, et al. Imputing missing standard 2 diabetes mellitus and the other major components of the metabolic deviation in meta-analyses can provide accurate results. J Clin syndrome after biliopancreatic diversion in patients with BMI Epidemiol. 2006;59:7–10. <35 kg/m2. Obes Surg. 2007;17:185–92. 39. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency17. Ritter S, Vetter ML, Sarwer DB. Lifestyle modifications and surgical in meta-analyses. BMJ. 2003;327:557–60. options in the treatment of patients with obesity and type 2 diabetes mellitus. Postgrad Med. 2012;124(4):168–80. 40. Shikata S, Nakayama S, Yamagishi H. Comparison of effects in randomized controlled trials with observational studies in digestive18. Abbatini F, Capoccia D, Casella G, et al. Type 2 diabetes in obese surgery. Ann Surg. 2006;244(5):668–76. patients with body mass index of 30–35 kg/m2: sleeve gastrectomy versus medical treatment. Surg Obes Relat Dis. 2012;8(1):20–4. 41. Sjöström L, Lindroos AK, Peltonen M, et al. Lifestyle, diabetes, and cardiovascular risk factors 10 years after bariatric surgery. N Engl J19. Laferrere B, Teixeira J, McGinty J, et al. Effect of weight loss by Med. 2004;351(26):2683–93. gastric bypass surgery versus hypocaloric diet on glucose and incretin levels in patients with type 2 diabetes. J Clin Endocrinol Metab. 42. O’Brien PE, Dixon JB, Laurie C, et al. Treatment of mild to moderate 2008;93:2479–85. obesity with laparoscopic adjustable gastric banding or an intensive medical program. A randomized trial. Ann Intern Med. 2006;144:20. Shaw JE, Sicree RA, Zimmet PZ. Global estimates of the prevalence 625–33. of diabetes for 2010 and 2030. Diabetes Res Clin Pract. 2010;87:4– 14. 43. Dixon JB, O’Brien PE, Playfair J, et al. Adjustable gastric banding and conventional therapy for type 2 diabetes: a randomized con-21. Dixon JB, le Roux CW, Rubino F, et al. Bariatric surgery for type 2 trolled trial. JAMA. 2008;299(3):316–23. diabetes. Lancet. 2012;379:2300–11. 44. Hofsø D, Nordstrand N, Johnson LK, et al. Obesity-related cardio-22. Buchwald H, Estok R, Fahrbach K, et al. Weight and type 2 diabetes vascular risk factors after weight loss: a clinical trial comparing after bariatric surgery: systematic review and meta-analysis. Am J gastric bypass surgery and intensive lifestyle intervention. Eur J Med. 2009;122(3): 248–56-e5. Endocrinol. 2010;163(5):735–45.23. Spann SJ, Nutting PA, Galliher JM, et al. Management of type 2 45. Adams TD, Pendleton RC, Strong MB, et al. Health outcomes of diabetes in the primary care setting: a practice-based research net- gastric bypass patients compared to nonsurgical, nonintervened se- work study. Ann Fam Med. 2006;4:23–31. verely obese. Obesity (Silver Spring). 2010;18(1):121–30.24. American Diabetes Association. Standards of medical care in diabe- 46. Serrot FJ, Dorman RB, Miller CJ, et al. Comparative effectiveness of tes—2009. Diabetes Care. 2009;33 Suppl 1:S11–61. bariatric surgery and nonsurgical therapy in adults with type 2 diabetes mellitus and body mass index <35 kg/m2. Surgery. 2011;150(4):684–91.25. le Roux CW, Welbourn R, Werling M. Gut hormones as mediators of appetite and weight loss after Roux-en-Y gastric bypass. Ann Surg. 47. Martins C, Strømmen M, Stavne OA, et al. Bariatric surgery versus 2007;246:780–5. lifestyle interventions for morbid obesity—changes in body weight, risk factors and comorbidities at 1 year. Obes Surg.26. Cummings DE, Overduin J, Foster-Schubert K-E. Gastric bypass for 2011;21(7):84–9. obesity: mechanisms of weight loss and diabetes resolution. J Clin Endocrinol Metab. 2004;89:2608–15. 48. Iaconelli A, Panunzi S, De Gaetano A, et al. Effects of bilio- pancreatic diversion on diabetic complications: a 10-year follow-27. Cummings DE, Shannon MH. Ghrelin and gastric bypass: is there a up. Diabetes Care. 2011;34(3):561–7. hormonal contribution to surgical weight loss? J Clin Endocrinol Metab. 2003;88:2999–3002. 49. Scopinaro N, Adami GF, Papadia FS, et al. Effects of biliopanceratic diversion on type 2 diabetes in patients with BMI 25 to 35. Ann Surg.28. Rubino F, Forgione A, Cummings DE, et al. The mechanism of 2011;253(4):699–703. diabetes control after gastrointestinal bypass surgery reveals a role of the proximal small intestine in the pathophysiology of type 2 50. Leonetti F, Capoccia D, Coccia F, et al. Obesity, type 2 diabetes diabetes. Ann Surg. 2006;244:741–9. mellitus, and other comorbidities: a prospective cohort study of laparoscopic sleeve gastrectomy vs medical treatment. Arch Surg.29. Pories WJ, Caro JF, Flickinger EG, et al. The control of diabetes 2012;147(8):694–700. mellitus (NIDDM) in the morbidly obese with the Greenville gastric bypass. Ann Surg. 1987;206:316–23. 51. Mingrone G, Panunzi S, De Gaetano A, et al. Bariatric surgery versus conventional medical therapy for type 2 diabetes. N Engl J Med.30. Guidone C, Manco M, Valera-Mora E, et al. Mechanisms of recovery 2012;366(17):1577–85. from type 2 diabetes after malabsorptive bariatric surgery. Diabetes. 2006;55:2025–31. 52. Heo YS, Park JM, Kim YJ, et al. Bariatric surgery versus conven- tional therapy in obese Korea patients: a multicenter retrospective31. MacDonald KG, Long SD, Swanson MS. The gastric bypass opera- cohort study. J Korean Surg Soc. 2012;83(6):335–42. tion reduces the progression and mortality of noninsulin-dependent diabetes mellitus. J Gastrointest Surg. 1997;1:213–20. 53. Dorman RB, Serrot FJ, Miller CJ, et al. Case-matched outcomes in bariatric surgery for treatment of type 2 diabetes in the morbidly32. Cummings DE, Overduin J, Shannon MH. Hormonal mechanisms of obese patient. Ann Surg. 2012;255(2):287–93. weight loss and diabetes resolution after bariatric surgery. Surg Obes Relat Dis. 2005;1:358–68. 54. Leslie DB, Dorman RB, Serrot FJ, et al. Efficacy of the Roux-en-Y gastric bypass compared to medically managed controls in meeting the33. Astiarraga B, Gastaldelli A, Muscelli E, et al. Biliopancreatic diver- American Diabetes Association composite end point goals for man- sion in nonobese patients with type 2 diabetes: impact and mecha- agement of type 2 diabetes mellitus. Obes Surg. 2012;22(3):367–74. nisms. J Clin Endocrinol Metab. 2013;98(7):2765–73. 55. Schauer PR, Kashyap SR, Wolski K, et al. Bariatric surgery versus34. Oxford Centre for Evidence-Based Medicine. http://www.cebm.net/ intensive medical therapy in obese patients with diabetes. N Engl J index.aspx?o=5653. Accessed 30 March 2013. Med. 2012;366(17):1567–76.35. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for 56. Ikramuddin S, Korner J, Lee W-J, et al. Roux-en-Y gastric bypass vs systematic reviews and meta-analyses: The PRISMA Statement. intensive medical management for the control of type 2 diabetes, PLoS Med. 2009;6(6):e1000097.

OBES SURG hypertension, and hyperlipidemia: The Diabetes Surgery Study 74. Saydah S, Tao M, Imperatore G, et al. GHb level and subsequent Randomized Clinical Trial. JAMA. 2013;309(21):2240–9. mortality among adults in the US. Diabetes Care. 2009;32:1440–6.57. Buchwald H, Oien D. Metabolic/bariatric surgery worldwide 2011. Obes Surg. 2013;23:427–36. 75. Khow KT, Warham N, Bingham S, et al. Association of haemoglobin58. Shukla AP, Moreira M, Dakin G, et al. Medical versus surgical A1c with cardiovascular disease and mortality in adults: the treatment of type 2 diabetes: the search for level 1 evidence. Surg European prospective investigation into cancer in Norfolk. J Ann Obes Relat Dis. 2012;8(4):476–82. Intern Med. 2004;141:413–20.59. No authors listed. Gastrointestinal Surgery for Severe Obesity. Consen Statement 1991; 9(1):1–20. 76. Kenngott HG, Clemens G, Gondan M, et al. DiaSurg 2 trial—60. Fried M, Hainer V, Basdevant A, et al. Inter-disciplinary European surgical vs. medical treatment of insulin-dependent type 2 diabetes guidelines on surgery of severe obesity. Int J Obes (Lond). mellitus in patients with a body mass index between 26 and 35 kg/m2: 2007;31(4):569–77. study protocol of a randomized controlled multicenter trial. Trials.61. Shrier I, Boivin J-F, Steele RJ, et al. Should meta-analyses of inter- 2013;14(1):183. ventions include observational studies in addition to randomized controlled trials? A critical examination of underlying principles. 77. O’Brien PE, Brennan L, Laurie C, et al. Intensive medical weight loss Am J Epidemiol. 2007;166:1203–9. or laparoscopic adjustable gastric banding in the treatment of mild to62. Tsai AG, Wadden TA. Systematic review: an evaluation of major moderate obesity: long-term follow-up of a prospective randomized commercial weight loss programs in the United States. Ann Intern trial. Obes Surg. 2013;23:1345–53. Med. 2005;142:56–66.63. Dansinger ML, Tabsioni A, Wong JB, et al. Meta-analysis: the effect 78. Mokdad AH, Ford ES, Bowman BA, et al. Prevalence of obesity, of dietary counseling for weight loss. Ann Intern Med. 2007;147(1): diabetes and obesity-related health risk factors, 2001. JAMA. 41–50. 2003;289:76–9.64. Ritt M, Piza H, Rhomberg M, et al. Metabolic risk factors in formerly obese women—effects of a pronounced weight loss by gastric band 79. Mokdad AH, Bowman BA, Ford ES, et al. The continuing epidemics operation compared with weight loss by diet alone. Diabetes Obes of obesity and diabetes in the United States. JAMA. 2001;286:1195– Metab. 2005;7(3):216–22. 200.65. Anderson JW, Grant L, Gotthelf L, et al. Weight loss and long-term follow-up of severely obese individuals treated with an intense be- 80. North American Association for the Study of Obesity (NAASO) and havioral program. Int J Obes. 2007;31(3):488–93. the National Heart. Clinical Guide lines on the identification, evalu-66. Gill RS, Birch DW, Shi X, et al. Sleeve gastrectomy and type 2 ation, and treatment of overweight and obesity in adults: the evidence diabetes mellitus: a systematic review. Surg Obes Relat Dis. 2010;6: report. Bethesda, MD: National Institutes of Health; 1998. NIH 707–13. publication 98–4083.67. Colquitt JL, Picot J, Loveman E, et al. Surgery for obesity. Cochrane Database Syst Rev 2009; CD003641. 81. Rubino F, Schauer PR, Kaplan LM, et al. Metabolic surgery to treat68. Maggard-Gibbons M, Maglione M, Livhits M, et al. Bariatric surgery type 2 diabetes: clinical outcomes and mechanisms of action. Annu for weight loss and glycemic control in nonmorbidly obese adults Rev Med. 2010;61:393–411. with diabetes: a systematic review. JAMA. 2013;309(21):2250–61.69. Padwal R, Klarenbach S, Wiebe N, et al. Bariatric surgery: a system- 82. Reis CEG, Alvarez-Leite JI, Bressan J, et al. Role of bariatric- atic review and network meta-analysis of randomized medical versus metabolic surgery in the treatment of obese type 2 diabetes with body surgical treatment of type 2 diabetes: the search for level 1 evidence. mass index <35 kg/m2: a literature review. Diabetes Technol Ther. Obes Rev. 2011;12:602–21. 2012;14(4):365–72.70. Gregg EW, Chen H, Wagenknecht LE, et al. for the Look AHEAD Research Group. Association of an intensive lifestyle intervention 83. Misra A, Chowbey P, Makkar BM, et al. Consensus statement for with remission of type 2 diabetes. JAMA. 2012;308(23):2489–96. diagnosis of obesity, abdominal obesity and the metabolic71. Reed MA, Pories WJ, Chapman W, et al. Roux-en-Y gastric bypass syndrome for Asian Indians and recommendations for physical corrects hyperinsulinemia implications for the remission of type 2 activity, medical and surgical management. J Assoc Phys India. diabetes. J Clin Endocrinol Metab. 2011;96(8):2525–31. 2009;57:163–70.72. Li Q, Chen L, Yang Z, et al. Metabolic effects of bariatric surgery in type 2 diabetic patients with body mass index <35 kg/m2. Diabetes, 84. Razak F, Anand SS, Shannon H, et al. Defining obesity cut points in a Obes Metab. 2012;14:262–70. multiethnic population. Circulation. 2007;115:2111–8.73. Turner RC, Mill NSH, Haw N, et al. Risk factors for coronary artery disease in non-insulin dependent diabetes mellitus: United Kingdom 85. Dixon JB, Zimmet P, Alberti KG, et al. on behalf of the International prospective diabetes study. BMJ. 1998;339:229–34. Diabetes Federation Taskforce on Epidemiology and Prevention. Bariatric surgery: an IDF statement for obese type 2 diabetes. Diabet Med. 2011;28:628–42. 86. Centers for Medicare and Medicaid Services. Decision Memo for Surgery for Diabetes (CAG-00397 N). http://www.cms.gov/ medicare-coverage-database/details/nca-decision-memo.aspx? NCAId=219&NcaName=Surgery+for+Diabetes&NCDId= 326&ncdver=1&IsPopup=y&bc=AAAAAAAACAAAAA%3D% 3D&. Accessed on 10 July 2013. 87. Finkelstein EA, Øsbye T, Malhotra R. Body mass trajectories through mid-life among adults with class-I obesity. Surg Obes Relat Dis. 2013;9:547–53.

OBES SURGDOI 10.1007/s11695-014-1311-1 REVIEW ARTICLESystematic Review and Meta-analysis of Medium-Term OutcomesAfter Banded Roux-en-Y Gastric BypassH. Buchwald & J. N. Buchwald & T. W. McGlennon# Springer Science+Business Media New York 2014Abstract This is the first systematic review and meta-analysis rates were as follows: early, 10.9 %, and late, 20.0 %. Non-band-focused exclusively on intermediate-term outcomes for the related reoperation rate was 15.2 %, and band-specific reopera-banded Roux-en-Y gastric bypass (B-RYGB). B-RYGB articles tion rate was 4.1 %. Gastric outlet stenosis, band erosion, andpublished from 1990 to 2013 were identified through band slippage were 2.8, 2.3, and 1.5 %, respectively. DiabetesMEDLINE, ScienceDirect, and SpringerLink databases aug- remitted in 80/95 (84.2 %). By systematic review and meta-mented by manual reference review. Articles were assigned an analysis, albeit with limited rates of follow-up, B-RYGB ap-evidence level (Centre for Evidence-Based Medicine [Oxford peared to result in significant, sustained excess weight loss ofUK] criteria) and Jadad quality score (randomized controlled approximately 70.0 % out to 10 years.trials). Simple and weighted means (95 % confidence interval(CI)) for excess weight loss (EWL) at follow-up (1–10+ years) Keywords Bariatric surgery . Roux-en-Y . Gastric bypass .were calculated. At 5 years, a pooled estimate for BMI (kg/m2) Banded gastric bypass . Systematic review . Meta-analysis .change (weighted mean difference and 95 % CI) for banded Diabetes . Randomized controlled trial . Weight . BMIbypass patients was computed. Rates for weighted mean com-plications, non-band- and band-related reoperations, and overall Introductioncomorbidity resolution were calculated. Three hundred twenty-one articles were identified: 286 failed inclusion criteria (i.e., Amplified by environmental factors, the obesity epidemicnon-English, B-RYGB unrelated, <10 per arm, <3-year follow- persists despite global educational efforts and treatment withup), leaving 35 articles. Manual review added 10 potentially diet, exercise, and pharmaceuticals [1–5]. Although minimal-relevant articles; 30 that failed inclusion criteria were excluded, ly invasive treatments are in development to address unmetleaving 15 for analysis. B-RYGB was performed on 8,707 needs in obesity management [7], metabolic/bariatric surgerypatients: 79.0 % female, mean age 38.7, and BMI 47.6 (41.0– remains the most effective intervention, with a record of59.4). Overall BMI weighted mean difference (reduction) at >50.0 % excess weight loss (EWL) maintained beyond5 years was 17.8 (95 % CI 12.8, 22.7; p<0.001). Five-year 10 years [7]. Since the introduction by Mason and Ito inweighted mean EWL of 72.5 % (67.5, 77.4) was sustained at 1966 [8] of the loop gastric bypass, subsequently modified10+ years (69.4 %; 58.9, 80.0). Weighted mean complication to the Roux-en-Y gastric bypass (RYGB), this procedure has proved the most consistently successful treatment for theH. Buchwald (*) greatest number of patients; it is also the most well-Department of Surgery, University of Minnesota, 420 Delaware researched bariatric operation, with >7,780 peer-reviewedStreet SE, MMC 290, Minneapolis, MN 55455, USA publications [9]. Yet, RYGB patients may contend with sig-e-mail: [email protected] nificant complications and intermediate-term weight regain is a serious concern. A growing body of studies of theJ. N. Buchwald RYGB variant, vertical banded RYGB (B-RYGB), sug-Division of Research Writing, Medwrite Medical Communications, gests that its weight loss rivals and may surpass that ofMaiden Rock, WI, USA RYGB over the longer term while presenting a compa- rable complication profile.T. W. McGlennonQuantitative Research, McGlennon MotiMetrics, Maiden Rock, WI,USA

OBES SURG Several surgical modifications to improve outcomes have reported weight and complication findings in study arms ofbeen suggested over the years. During the late 1970s and early ≥10 adults with follow-up of ≥3 years who underwent any1980s, to address eventual dilation of the gastric bypass pouch variant of the vertical B-RYGB. For multiple publications inand stoma, the technique of reinforcing the stoma with a mesh, the same year or kin studies, the most recent study thatband, or ring was introduced in the vertical banded fulfilled inclusion criteria was incorporated.gastroplasty (VBG) and banded gastric bypass variants ofMason, Eckhout et al., Torres and Oca, Laws and The National Library of Medicine PubMed®/MEDLINE®,Piantadosi, and Salmon [10–14]. In the 1980s and early ScienceDirect®, and SpringerLink® databases were searched1990s, Linner and Drew [15], Fobi et al. [16, 17], and electronically for relevant records. Our systematic reviewHoward et al. [18] improved outcomes by completely strategy observed the guidelines of the Preferred Reportingtransecting the stomach to avoid staple-line disruption. In Items for Systematic Reviews and Meta-Analyses (PRISMA)1992, Fobi and Lee reported a B-RYGB-specific technique [24] in the identification, screening, and determination ofthat incorporated both gastric transection and interposition of eligible records. In the initial search stage, articles that metthe jejunal Roux limb between the gastric pouch and bypassed the broad inclusion criteria (i.e., published between 1990 andremnant, which they believed would reduce the high morbid- November 1, 2013; duplicates eliminated) were discoveredity and reoperation rate associated with marginal ulceration through Boolean keyword assays of the databases using com-and outlet stenosis (Fig. 1) [19]. This speculation was con- binations of 19 terms that denote variants of the studiedfirmed by the consecutive B-RYGB series of Capella and procedure: “banded Roux-en-Y gastric bypass,” “bandedCapella, in which 489 initial patients sustained 64 staple-line bypass,” “banded gastric bypass,” “vertical banded by-disruptions, and the last 134, employing gastric transection pass,” “vertical banded gastroplasty–gastric bypass,” “si-and jejunal interposition, had no disruption or fistula [20]. The lastic ring vertical gastric bypass,” “silastic ring Roux-B-RYGB technique as refined by Fobi et al. and Capella and en-Y gastric bypass,” “silicone ring bypass,” “stapledCapella has been the dominant approach since the mid-1990s. silastic ring vertical gastric bypass,” “transected silastic ring vertical gastric bypass,” and relevant acronyms, In 1999, Fisher and Barber suggested that “adding the band including “BGBP,” “BRGB,” “BGB,” “B-RYGBP,”to the gastric bypass results in more weight loss for more “VBG-RGB,” “SRVGBP,” “SSRVGBP,” “SR-RYGBP,”patients that is maintained over a longer period of time” and “TSRVGBP.”[21]. In 2013, the third questionnaire-based review ofBuchwald et al. of the status of metabolic/bariatric surgery In the second search stage, titles and abstracts of identifiedworldwide estimated that 46.6 % of 340,768 operations per- records were screened. Articles with any combination of theformed annually were those of RYGB and its variants, al- following were excluded: non-English language, contentsthough B-RYGB as a unique procedure was absent from this unrelated to vertical banded bypass, non-RYGB-related,global snapshot [22]. In addition, most studies of this proce- banded bypass study arm N<10, <3-year follow-up, non-dure in the surgical literature have reported B-RYGB out- adult, comments or letters to the editor, review papers, lengthycomes as findings for RYGB, often rendering B-RYGB’s non-research commentaries, guidelines, and abstracts or caserecord of safety and effectiveness indistinguishable from that studies. The third evaluation stage entailed reading full articlesof RYGB. Yet, while new bariatric procedures come and go, and their reference sections to identify RYGB-focused studiesthe banded bypass has been performed since the mid-1980s with possible vertical banded bypass subgroups not capturedand is one of few bariatric operations for which there are >10- in the primary literature search; articles that met inclusionyear peer-reviewed safety and effectiveness data to be exam- criteria were added to the list of potentially eligible records.ined. In 2013, there is as yet no consensus on whether adding a In the final review stage, kin studies (i.e., those containingband or ring to RYGB is preferable to RYGB alone [23]. The data that overlapped with an included study), reports of non-goal of this study was to review all research reports from 1990 banded RYGB procedures, articles without usable weightto the present that pertain to the vertical B-RYGB procedure data, and those with <3-year follow-up were excluded fromand to analyze the available intermediate- to long-term weight, eligibility for quantitative analysis.complications, and comorbidity outcomes. Data Extraction and Quality AssessmentMethods Two reviewers identified and extracted data that met inclusion criteria according to systematic review of the bariatric/Inclusion Criteria and Search Strategy metabolic surgery literature on B-RYGB. Data were entered into fields comprising a dedicated Excel® (Microsoft®,Studies were included if they were full-length research articles Redmond, WA) database. Variables of interest included thein the English language published after January 1, 1990 that following: study characteristics (N [sample size], study de- sign, B-RYGB variant, follow-up time points); patient

OBES SURGFig. 1 a Fobi et al. verticalbanded Roux-en-Y gastric bypass(B-RYGB) procedure—“stapledvertical banded Roux-en-Ygastric bypass” [16]; b Capellaand Capella variant B-RYGBprocedure—“stapled verticalbanded gastroplasty–gastricbypass with window” [74]; c Fobiand Lee divided B-RYGBprocedure—“transected verticalsilastic banded Roux-en-Y gastricbypass with jejunal interposition”[19]baseline characteristics (i.e., sex, age, BMI [weight (kg) di- used to summarize variables across studies. Categorical vari-vided by height (m2)]); and weight outcomes, principally, ables were expressed as number and percentage. Quantitativepercentage excess weight loss (%EWL), and secondarily, outcome variables (i.e., EWL at time points ranging from 1 toBMI change where possible. Alternative measures of weight 10+ years postsurgery) were expressed using reported studyloss (e.g., percent total body weight lost) were not recorded means and associated sample sizes. In the analysis of EWL,with sufficient regularity to be considered for analysis. It a fixed effects model was implemented using methodsshould be noted that a recent systematic review has put for- adapted from O’Brien et al. [7]. Simple pooled EWLward that %EWL, as an outcome measure, may be some- means and ranges as well as weighted EWL pooled meanswhat biased against higher presurgical BMI patients and 95 % confidence intervals (CIs) were calculated for[25]; however, others have reported that %EWL is a each time point. Weighted means were calculated with therelatively reliable measure across studies, regardless of following formula using sample size as the weightinginitial BMI [26]. In light of this information and that factor:%EWL has been historically the most frequently report-ed weight outcome measure, the present authors believe Xn%EWL to be an appropriate measure for the current Àx ¼ Xi n¼1wixianalyses. Where studies reported banded bypass dataon multiple groups, data were pooled using weighted i ¼1wimeans and standard deviations to yield summary data.In addition to weight outcomes, data on mortality and At each time point, for the number of studies includ-complications (e.g., reoperations, gastric outlet stenosis, ed in the analysis (ranging from 1 to n, dependent onband erosion, band slippage, solid food intolerance) and studies reporting data on EWL), the weighted mean wascomorbidities (type 2 diabetes mellitus [T2DM], dyslip- calculated by summing (Σ) products of all case-idemia, hypertension, sleep apnea) were extracted and weighted study means (wixi) and dividing the result byanalyzed. The evidence level of articles was assessed the total weight (wi) (wi, combined N of included stud-according to criteria developed by the Centre for ies). Ninety-five percent CIs were calculated as approx-Evidence-Based Medicine, Oxford, UK [27], and the imations of uncertainty limits surrounding weightedmethodological quality (bias assessment) of randomized mean estimates.controlled trials (RCTs) was rated using the Jadad scor-ing method [28]. Where follow-up sample size was not specifically reported, a weighting factor estimate was derived, where possible, usingData Reporting and Statistical Analysis properties of the normal distribution or from information provided in the article’s tables and figures. Also, where BMIData analysis was performed using SPSS® software, version change but no EWL was reported at follow-up, EWL was20 (IBM SPSS, Chicago, IL) in conjunction with calculated as ([baseline BMI−follow-up BMI] / [baselineComprehensive Meta-Analysis (CMA), version 2.2 (Biostat, BMI−25])×100. For comparative purposes, a trend lineEnglewood, NJ). Quantitative baseline variables (i.e., age, depicting EWL in B-RYGB patients was incorporated withinBMI) were expressed using reported study means and stan- the 1–10-year trend lines that represent EWL findings for B-dard deviations (SD). Overall pooled means and ranges were RYGB, BPD/DS, RYGB, and laparoscopic adjustable gastric banding (LAGB) derived from the integrated 1–10 and 10+- year data of O’Brien et al.

OBES SURG BMI change at the 5-year time point was analyzed 75.0 % was considered indicative of significant hetero- geneity and was calculated using the following formulaby calculating the weighted mean difference and asso- [29]:ciated 95 % CI under a random effects model; I2 ≥I2 ¼ ð½Q‐df Š=QÞx100; where Q is the X2statistic and df ¼ degrees of freedom Where included studies provided mean BMI values at One study received a score of 2/5 and two studies, 3/5,baseline and 5-year follow-up but did not report associated although it has been suggested that these scores are 1 to 2SDs or standard errors (SEs), SDs were imputed using points lower than appropriate for surgery trials, which rarelymethods adapted from Furukawa et al. [30]. Qualitative out- employ double blinding [46].come variables (i.e., complications and comorbidity data)were expressed as number and percentage unless otherwise The number of articles that studied the following B-RYGBnoted. Weighted mean early and late complication rates, as variants were silastic or polypropylene VBG-RYGB 4/15,well as non-band and band-specific reoperation rates, were silastic ring gastric bypass (stapled [SRGBP]) 3/15, transectedalso calculated. silastic ring vertical gastric bypass (TSRVGBP) with jejunal interposition (TSRVGBP+JI) 3/15, mesh-banded RYGBResults 2/15, fascia-banded RYGB 1/15, SRGBP and transected SRGBP (TSRGBP) 1/15, and TSRVGBP 1/15. All studiesStudy Characteristics reported a minimum of 36 months of follow-up.We performed a comprehensive review of the national medi- Patient Characteristicscal databases (Fig. 2). Three hundred twenty-one records metthe broad keyword search with duplicates filtered out. After The total number of patients in the included studies thatscreening by title and abstract, 286 citations failed inclusion underwent banded bypass of some variant was 8,707. Oncriteria and were disqualified. The remaining 35 articles were average, patients were 38.7 years old (range 33.2–43.4) andscreened for RYGB-focused articles that might contain B- had a mean baseline BMI of 47.6 (41.0–59.4); 79.0 % wereRYGB subgroups; 10 such articles were identified, yielding female (Table 2).a total of 45 eligible articles. After reading the remainingarticles, 30 studies that did not meet all inclusion criteria were Weight Lossexcluded, leaving a final set of 15 articles [18, 20, 31–43] forquantitative analysis. In addition, two kin B-RYGB studies Overall, banded bypass patients tended to experience imme-that reported more complete weight and complication data diate and significant weight loss that was sustained over timethan that published in their primary studies in the same years (1–10+ years). Table 3 details individual study EWL at[44, 45] were treated as supplemental material and reported in follow-up and descriptive statistics summarizing all data col-combination with their related primary studies. lected at specific time points. For example, at 1-year postsurgery, nine study groups reported weight loss data; Table 1 describes the characteristics of included studies. however, only seven provided information on number ofArticles were published over a 21-year span, between 1992 patients that were followed to 1 year; thus, the overall pooledand 2013. The respective country of origin (based on first 1-year weighted mean EWL of 76.0 % (95 % CI 69.5, 82.5)author’s affiliation) and article distribution were as follows: was based on analysis of results from 1,288 banded bypassUSA 7/15, Brazil 4/15, Mexico 1/15, Chile 1/15, Venezuela patients from seven separate studies.1/15, and New Zealand 1/15. Study designs included 3/15RCTs (two blinded), 2/15 prospective case–control studies, At 2-year postsurgery, the weighted mean EWL for B-and 10/15 retrospective database reviews. By the Centre for RYGB patients was 78.8 % (74.8, 82.8) based on results fromEvidence-Based Medicine (CEBM) scale, there was no level I 1,587 patients in seven studies. The majority of includedevidence. Of the three RCTs, according to the 5-point Jadad studies (80.0 %) reported weight loss data, in some form, atscale for the three items directly related to the quality of bias 5-year postsurgery. As a result, the pooled weighted meanreduction measures taken (randomization, blinding, with- EWL estimate may be assumed to be the most accuratedrawals), none rated a high score for methodological quality: accounting of B-RYGB efficacy provided in this report. Based on 1,254 patients in 11 studies, 5-year weighted mean EWL was 72.5 % (67.5, 77.4).

OBES SURGFig. 2 Flow diagram ofsystematic review of the bandedRoux-en-Y gastric bypass (B-RYGB) literature based on thePreferred Reporting Items forSystematic Reviews and Meta-Analyses (PRISMA) statement White et al. [35] provided the longest follow-up of B- 53.5 % EWL) seems apparent, a statistically significant differ-RYGB weight loss data beyond 10 years, reporting 58.5 % ence could not be confirmed. A wide 95 % CI encompassed theEWL at 14 years. This value was incorporated into calculation current study’s B-RYGB EWL (69.4 %; 58.9, 80.0); the some-of the 10+-year weighted mean (69.4 %; 58.9, 80.0). B-RYGB what small number of patients (n=202) entered into the vari-weight loss data presented in Table 3 are compared in Fig. 3 to ance calculation greatly increases the likelihood of confidencemid- and longer-term weight loss trends for multiple bariatric interval overlap.procedures (RYGB, BPD/DS, LAGB) [7, 47]. At year 1, B-RYGB, RYGB, and BPD/DS achieved similarly high degrees Two studies provided weight outcomes through 5-yearof EWL—approximately 70.0±5.0 %, while LAGB achieved follow-up for revisional B-RYGB (bariatric procedure con-a significantly lower EWL of approximately 45.0 %. Between verted to a banded bypass) (i.e., Drew et al. [31]; Capella et al.years 1 and 2, each procedure resulted in increased weight loss; [32]). Although revisional B-RYGB did not achieve equiva-however, after 2 years, while B-RYGB, BPD/DS, and LAGB lent weight loss to that of primary B-RYGB (Fig. 4), signifi-weight loss continued to increase or remain relatively constant, cant differences were not observed at any time point. Also,RYGB declined significantly. The weight loss of LAGB con- revisional B-RYGB EWL was comparable to that of primarytinued to increase while that of RYGB continued to decrease RYGB at 2 years (70.0 vs 68.0 %) (Figs. 3 and 4) anduntil their trend lines intersected at approximately 55.0 % EWL substantially greater than primary RYGB at 5 years (65.0 vsat 7.5 years. At 10+ years, there is virtually no distinction in 58.0 %). Though no firm conclusions can be drawn due toweight loss efficacy between LAGB and RYGB. In contrast, limited sample size in the B-RYGB revision group (n=160),B-RYGB EWL held relatively constant at approximately these results suggest that B-RYGB, even as a revisional pro-70.0 %, comparable to BPD/DS at each time point up to 10+ cedure, does not result in the same degree of weight regain asyears. Although a marked relative difference in 10+-year primary RYGB between 3- and 5-year postsurgery.weight outcomes favoring B-RYGB over RYGB (69.4 vs Finally, as 80.0 % of included studies reported 5-year weight loss data, the effect of B-RYGB on BMI was analyzed

Table 1 Characteristics of included studiesStudy author Countrya Number Design B-RYGB variant Follow-up meanb or medianb CEBMb level Jadad scorec (range) of evidence RCTsDrew et al. [31] USA 461 Retrospective database study Fascia-banded RYGB 51 months (36–72) IV –Howard et al. [18] USA 20 RCT, double-blind, single-center Marlex mesh-banded RYGB – (12–78) II 2Capella et al. [20] USA 560 Retrospective, database study Silastic or polypropylene VBG-RYGB – (30–66) III –Capella et al. [32] USA 60 Retrospective, database study Silastic or polypropylene VBG-RYGB – (12–60) IV –Capella et al. [33] USA 652 Retrospective, database study Silastic or polypropylene VBG-RYGB – (60) IV –Fobi et al. [34] USA 576 Retrospective, database study TSRVGBP – (3–108) III –White et al. [35] New Zealand 342 Retrospective, database study SRGBP and TSRGBP – (12–168) IV –Pajecki et al. [36] Brazil 130 Retrospective, database study SRGBP 87 (24–96) IV –Bessler et al. [37] USA 46 RCT, double-blind, single-center Polypropylene VBG-RYGB – (6–36) II 3Magro et al. [38] Brazil 782 Prospective case–control series SRGBP – (18–60) IV –Salinas et al. [39] Venezuela 160 Retrospective, database study TSRVGBP + JI – (60) IV –Rasera et al. [40] Brazil 75 Retrospective, database study SRGBP – (12–48) IV –Awad et al. [41] Chile 260 Retrospective, database study TSRVGBP + JI – (12–120) III –Zarate et al. [42] Mexico 30 RCT, non-blinded, single-center Mesh-banded RYGB – (60) II 3Valezi et al. [43] Brazil 211 Prospective case–control series TSRVGBP + JI – (12–120) IV –World Health Organization obesity weight class by body mass index (BMI): I (30.0–<35), II (≥35.0–<40), and III (≥40.0)RCT randomized controlled trial, nRCT non-randomized controlled trial, B-RYGB banded Roux-en-Y gastric bypass, VBG vertical banded gastroplasty, SRGBP silastic ring gastric bypass (stapled),TSRGBP transected SRGBP, TSRVGBP transected silastic ring vertical gastric bypass, TSRVGBP+JI transected silastic ring vertical gastric bypass with jejunal interpositiona Country of first author’s affiliationb Level of evidence based on the Centre for Evidence-Based Medicine, Oxford, UK. Level I=high-quality RCT with statistically significant difference or no difference but narrow confidence intervals orsystematic review of level 1 RCT. Level 2=lesser quality RCT (<80 % follow-up, no blinding, or improper randomization), or prospective comparative study, or systematic review of level II studies or levelI studies with inconsistent results. Level III=case-controlled study, or retrospective comparative study, or systematic review of level III studies. Level IV=case series. Level V=expert opinion [27]c Jadad score for RCTs based on 5-point scale, 5=least biased/highest methodological quality [28] OBES SURG

OBES SURGTable 2 Characteristics of band- Study author Female N (%) Age mean±SD, years (range) Body mass indexed Roux-en-Y gastric (B-RYGB) mean±SD, kg/m2 (range)bypass patients Drew et al. [31] 416 (90.2) 36.0 ± 9.8 Howard et al. [18] 15 (75.0) 38.1 ± 1.9 45.0 ± 7.6 Capella et al. [20] 453 (81.0) 37.0 44.0 ± 7.8 Capella et al. [32] – 43.0 ± 8.0 52.0 ± 9.0 Capella et al. [33] 515 (79.0) 38.0 ± 10 43.0 ± 9.0 Fobi et al. [34] 501 (87.0) 40.2 50.0 ± 10 White et al. [35] 261 (76.0) 43.0 47.0 ± 7.8 Pajecki et al. [36] 99 (76.0) – 46.2 ± 9.2 Bessler et al. [37] 26 (56.5) 40.6 ± 7.4 56.7 ± 10.6 Magro et al. [38] 568 (72.6) 37.5 ± 11.5 59.4 ± 7.3 Salinas et al. [39] 121 (76.0) 33.2 ± 10.0 45.1 ± 4.4 Rasera et al. [40] 75 (100.0) 34.8 44.6 ± 9.3 Awad et al. [41] – – – Zarate et al. [42] 27 (90.0) 37.8 ± 9.6 41.0 ± 7.8 Valezi et al. [43] 101 (75.4) 43.4 ± 10.6 48.0 ± 5.0 Overall 44.2 ± 5.1 3,178 (79.0) 38.7 N (%) 33.2–43.4 47.6 Mean 41.0–59.4 Rangeby calculating a pooled estimate (weighted mean difference) ranged from 0.0 % (Pajecki et al. [36] and Awad et al. [41])of overall BMI reduction for studies reporting baseline and 5- to 79.0 % (Bessler et al. [37]). The overall weighted mean rateyear BMI data (Fig. 5). The weighted mean difference in 5- of late complications was 20.0 %. The choice of which com-year BMI for 1,254 B-RYGB patients (11 studies) was 17.8 plications to report seemed unique to each set of researchers(95 % CI 12.9, 22.7; significance of overall effect p<0.001, rather than dependent on a standardized guideline for inclu-I2=88.9 %). In addition, significant heterogeneity was found sion and assessment. The five most frequently reported com-between independent pooled BMI weighted mean differ- plications (regardless of intraoperative or postoperativeence estimates for the high BMI group (baseline BMI≥ time point) were as follows: mortality (80.0 % of stud-50) vs low BMI group (baseline BMI<50). As is typi- ies), leak (67.0 %), gastric outlet stenosis (53.0 %),cally the case with patient groups with higher baseline pulmonary embolism (47.0 %), and small bowel ob-BMI (e.g., superobese [BMI≥50], super-superobese struction (47.0 %). The non-band-related reoperation rate[BMI≥60]), B-RYGB patients (n=156) comprising the was 15.2 % (1,325/8,707) and the band-related reopera-three studies reporting a baseline BMI≥50 lost signifi- tion rate was 4.1 % (353/8,707).cantly more total body weight (weighted mean differ-ence=20.4 kg/m2 [17.9, 22.9]) at 5 years than those Comorbiditiespatients (n=1,098) comprising the eight studies reporting abaseline BMI<50 (weighted mean difference=15.4 kg/m2 A summary of the four studies reporting on comorbidity[14.0, 16.8]; Q=11.9, Q p<0.001). resolution following B-RYGB is presented in Table 5. Resolution rates for T2DM ranged from 75.0 to 92.0 %, withComplications an overall rate of 84.2 % (80/95). Dyslipidemia resolution rates ranged from 33.6 to 76.7 %, with an overall rate ofComplication rates, both early and late, differed markedly 39.8 % (129/324). Respective overall resolution rates foracross studies (Table 4). For example, of those studies obstructive sleep apnea and hypertension were 91.4 % (32/reporting on early complications, individual study rates re- 35) and 58.0 % (122/212). Due to the relatively few studiesported ranged from 0.0 % (Valezi et al. [43] and Zarate et al. included in the current comorbidity analysis, reliable estimates[42]) to 26.0 % (Bessler et al. [37]). The overall weighted of resolution were not possible; those given must be interpretedmean rate of early complications was 10.9 %. For those with caution.studies reporting late complications, individual study rates

OBES SURGTable 3 Banded Roux-en-Y gastric bypass (B-RYGB) percentage excess weight loss (%EWL)Study author Follow-up (years)Baseline N 1 2 3 4 5 6 7 8 9 10+ % (n)Drew et al. [31] 74.5 76.3 72.8 70.9 73.6 – – – – –N = 461 – – – (180) –Howard et al. [18] 70.0 72.0 75.0 70.0 68.8 – – – – –N = 20 (20) – – – (6)Capella et al. [20] – – 70.0 67.0 62.0 – – – – –N = 560 (234) (152) (65)Capella et al. [32] 71.3 70.8 68.0 61.5 65.0 – – – – –N = 60 (36) (31) (15) (16) (14)Capella et al. [33] – – – – 77.0 – – – – –N = 652 (72)Fobi et al. [34] 73.5 78.2 77.7 77.0 75.7 74.7 73.7 72.5 72.2 –N = 576 (514) (475) (450) (401) (354) (327) (299) (194) (64)White et al. [35] 88.5 87.1 74.6 72.4 69.9 65.6 60.0 65.1 55.3 74.6 [58.5]*N = 342 (265) (203) (166) (115) (72) (55) (35) (37) (24) (26) [26.0]*Pajecki et al. [36] – 80.2 – – 74.0 79.0 70.0 66.0 – –N = 130 (75) (19) (19) (19) (19)Bessler et al. [37] 64.0 64.2 73.4 – – – – – – –N = 46 – – –Magro et al. [38] – 78.0 – – 69.0 – – – – –N = 782 (594) (363)Salinas et al. [39] – – – – 83.0 – – – – –N = 160 (134)Rasera et al. [40] 74.0 79.0 76.0 76.0 – – – – – –N = 75 (75) (75) (75) (49)Awad et al. [41] 74.0 – 77.0 – – – 78.2 82.2 82.3 81.7N = 260 (244) (169) (111) (78) (51) (34)Zarate et al. [42] – – – – 61.6 – – – – –N = 30 (21)Valezi et al. [43] 67.6 72.6 – – 69.7 – – 66.8 – 67.1N = 211 (134) (134) (134) (134) (116)Simple mean 73.0 75.8 73.8 70.7 70.8 73.1 70.5 70.5 70.0 70.5Range 64.0–88.5 64.2–87.1 68.0–77.7 61.5–77.0 61.6–83.0 65.6–79.0 60.0–78.2 65.1–82.2 55.3–82.3 58.5–81.7Sum of N 1,288 1,587 1,109 913 1,254 401 464 462 139 202No. studies included 7 7 6 6 11 2 4 5 3 3 73.7 73.6 73.0 69.4 in meta-analysis 76.0 78.8 75.3 73.2 72.5 67.0, 80.4 66.7, 80.5 71.6 59.5, 86.5 58.9, 80.0Weighted mean 69.5, 82.5 74.8, 82.8 71.7, 78.8 68.5, 77.8 67.5, 77.4 64.9, 78.295 % CINote: Although total number of patients undergoing bypass variants was 8,707, total baseline n for weight data=4,365Note: White et al. had 10- and 14-year data: The 14-year data have been integrated into the calculation of the 10+-year values (mean %EWL at 14 years=58.5; n=26 used in meta-analysis to give equal weight relative to 10-year data of White et al.). Also, White et al. reported mean %EWL data at 1, 2, 5, 10,and 14 years; White et al. 3-, 4-, 6-, 7-, 8-, and 9-year data were derived from O’Brien et al. [7]Discussion progenitor, RYGB. The current meta-analysis and trend line comparisons suggest that B-RYGB’s intermediate-termThe banded bypass has been performed for more than a weight loss outcomes may be superior to those of RYGB.quarter of a century, although its peer-reviewed studies have Whereas RYGB and bariatric surgery patients, on the whole,been overshadowed by the plenitude of reports about its reach their BMI nadir approximately 2 years postsurgery and

OBES SURGFig. 3 Excess weight loss (%) inbanded Roux-en-Y gastric bypass(B-RYGB) compared with Roux-en-Y gastric bypass (RYGB),biliopancreatic diversion/duodenal switch (BPD/DS), andlaparoscopic adjustable gastricbanding (LAGB). RYGB, BPD/DS, and LAGB data were fromO’Brien et al. [7]find that their subsequent weight loss decreases [7], this ≥100, they observed that B-RYGB “…appeared to retainpattern was not seen in the current analysis of B-RYGB. effectiveness better than RYGB, with the EWL being signif-From the available data, albeit derived from studies with icantly greater at 5 years for this group (74.7 vs 58.2; threelimited follow-up, the band seems to maintain the weight loss studies, n=498; p<0.05)” [7]. In addition, O’Brien et al. urgedachieved by the initial bypass well after the 2-year mark. that more data were needed to further explore the potentialWeight Loss superiority of B-RYGB. Our study has attempted to provideIn O’Brien et al.’s systematic review and meta-analysis of such data by incorporating into the analysis all studiesmedium-term weight loss after bariatric surgery in cohorts of reporting B-RYGB weight loss beyond 3 years as long as they were composed of cohorts of ≥10.Fig. 4 Excess weight loss (%) for revisional versus primary bandedRoux-en-Y gastric bypass procedures The current analysis found a 5-year EWL of 72.5 % (95 % CI 67.5, 77.4; 11 studies, n=1,254) for banded bypass patients, essentially equivalent to that reported by O’Brien et al. (74.7 %). Overall BMI reduction at 5 years in B-RYGB patients was 17.8 (12.9, 22.7), consonant with Buchwald et al.’s meta- analytic finding for BMI reduction at 2 years of 16.7 (14.9, 18.4) for all bypass patients (n=2,705, 22 studies) [48]. Also, the reviewed B-RYGB patients demonstrated a relatively con- stant EWL of approximately 70.0 % to 10+ years, rivaling weight outcomes associated with BPD/DS (Fig. 3). In Buchwald et al.’s 2004 bariatric surgery meta-analysis, with outcome time point values reflecting at least 50.0 % follow-up, EWL for BPD/DS was 70.1 % (66.3, 73.9) and for RYGB, 61.6 % (56.7, 66.5). It has been observed that typical maximal weight loss following RYGB is obtained at 1.9 years in morbidly obese patients and 2.2 years in the superobese [49]. Thereafter, approximately 10.0– 20.0 % of RYGB patients will experience weight regain or fail to lose adequate weight [50–52]. Pories et al. reported a 1-year mean EWL of 75.0–80.0 % for RYGB patients and a subsequent, significantly lower, EWL of 66.0 % at 5–14-year follow-up [53].

OBES SURGGroup by Study name Statistics for each study Difference in means and 95%CIBMI Level Difference Standard Lower Upper in means error limit limit Z-Value p-ValueHigh Capella et al (1) 18.000 1.145 15.756 20.244 15.718 0.000High Capella et al (3) 21.000 1.195 18.657 23.343 17.570 0.000High Pajecki et al 24.500 2.476 19.646 29.354 9.893 0.000High 20.409 1.274 17.913 22.906 16.023 0.000Low Howard et al 13.000 3.401 6.335 19.665 3.823 0.000Low Capella et al (2) 11.000 2.499 6.102 15.898 4.402 0.000Low White et al 15.000 1.132 12.781 17.219 13.247 0.000Low Fobi et al 16.700 0.464 15.791 17.609 35.990 0.000Low Magro et al 13.800 0.292 13.228 14.372 47.252 0.000Low Salinas et al 17.600 0.896 15.845 19.355 19.653 0.000Low Valezi et al 16.000 0.467 15.085 16.915 34.270 0.000Low Zarate et al 15.500 1.446 12.666 18.334 10.718 0.000Low 15.388 0.703 14.009 16.766 21.880 0.000Overall 17.786 2.508 12.870 22.703 7.091 0.000 -30.00 -15.00 0.00 15.00 30.00Fig. 5 The table and corresponding forest plot summarize meta-analysis mean difference [WMD] represented by the diamond shape at the base ofof body mass index (BMI) change from baseline to 5 years in banded the forest plot). Two additional diamonds represent independent summaryRoux-en-Y gastric bypass patients. Each study contributing to the meta- estimates for high BMI patients (baseline ≥50) vs low BMI patientsanalysis is represented by a single blackened square on the forest plot, its (baseline <50). WMD values >0, with 95 % CIs (i.e., box whiskers) notsize proportional to the amount of weight the study was given during intersecting the reference line (vertical line above 0 point), are indicativecalculation of the pooled summary estimate (expressed as the weighted of a significant treatment effect for banded bypass Christou et al. demonstrated that both morbidly obese and patients, the failure rate for the superobese was 18.8 vs 11.0 %superobese RYGB patients experienced significant weight for the morbidly obese group [38]. Capella et al. and Fobi et al.regain from the nadir to 5 years and again from 5 to 10 years reported a failure rate of approximately 7.0 % at 5-year[49]. In the current study, although B-RYGB studies analyzed follow-up [20, 34], and Awad et al. reported an 8.0 % failuredid not have the comparable extensive follow-up rate of rate at 10-year follow-up [41].Christou et al., the evidence available suggests that B-RYGBpatients did not appear to experience the same pattern of Although there have been few comparative trials of B-weight regain. For example, there was no significant differ- RYGB vs RYGB weight outcomes and failure rates, five suchence between the pooled mean EWL of 74.8 % (representing studies were identified in our preliminary research, four ofcombined 1–5-year follow-up data) and the pooled mean which were included in the analysis. One study comparedEWL of 72.3 % (combined 6–10-year follow-up) of the B- banded vs non-banded bypass outcomes at ≤3 years and, thus,RYGB patients analyzed. Rasera et al., Magro et al., and other did not meet our study’s inclusion criteria.researchers have observed weight regain primarily between 2and 5 years for both banded and non-banded bypass patients In 2013, Heneghan et al. reported a matched cohort analy-[38, 40, 49]. However, when weight regain occurs in B- sis of patients (n=268, 134/134) who had undergone bandedRYGB patients, it appears to do so at a lower rate (about vs non-banded LRYGB with 24-month EWL and BMI reduc-8.0 % of nadir weight) [38] relative to weight regain associat- tion as primary endpoints and secondary endpoints includinged with non-banded RYGB patients (17.0 to 22.0 % of nadir) mortality, morbidity, and band-related complications [57]. The[49]. Some researchers report that weight regain following authors found that B-RYGB patients achieved significantlyRYGB may be as high as 50.0 % of weight lost [54]. greater EWL at 2 years compared to non-banded patientsInterestingly, it may be that the most substantial amount of (58.6 vs 51.4 %, p<0.05). Further, and perhaps most interest-weight regain in B-RYGB patients occurs when the restrictive ing, subgroup analysis found the “band effect” on weight lossband must be removed due to complications such as severe localized principally in superobese patients who demonstratedfood intolerance [55]. a 10.0 % greater EWL after B-RYGB than after RYGB (57.5 vs 47.6 %, p<0.01; BMI reduction 18.0 vs 15.2, p<0.01). Weight regain also contributes to the overall failure rate of a Although the banded bypass group with baseline BMI<50particular bariatric procedure, as defined by percentage of also achieved a greater EWL than their non-banded counter-patients that do not achieve or maintain 50.0 % EWL at parts (62.9 vs 57.9 %), the difference was not statisticallyfollow-up. The long-term failure rate for RYGB has been significant (p=0.41). No significant differences in mortalitycalculated as approximately 20.0 % [49]. This rate can be as or morbidity were found; there was a 2.2 % band-relatedhigh as 40.0–60.0 % for superobese patients undergoing complication rate.RYGB [56]. In contrast, in Magro et al.’s study of banded Among the four comparative studies in the current analysis, Bessler et al. carried out a prospective randomized trial of

Table 4 Banded Roux-en-Y gastric bypass (B-RYGB) complications, by percentage of patients OBES SURGComplication Drew Howard Capella Capella Capella White Fobi [34] Salinas Pajecki Bessler Magro Rasera Awad Zarate Valezi Total [31] [18] [20] [32] [33] [35] N = 3,632 [45]a [36] [37] [38] [40] [41] [42] [43] no. (%) N = 461 N = 20 N = 623 N = 60 N = 652 N = 342 N = 130 N = 46 N = 782 N = 75 N = 260 N = 30 N = 211 E/L E/L E/L E/L E/L E/L E/L N = 1,383 E/L E/L E/L E/L E/L E/L E/L E/LAnemia –/– –/– –/– –/– –/– –/– –/– 00/17.1 –/– –/– –/– –/– –/– –/– –/– 237 (17.1)Bleeding 0.2/00 –/– 0.3/00 1.7/00 0.2/0.9 –/– –/– 1.4/00 –/– –/– –/–Deep vein thrombosis –/– –/– –/– –/– –/– –/– 2.5/00 –/– –/– –/– –/– –/– –/– –/– –/– 31 (1.0)Dumping syndrome –/– –/– –/– –/– –/– –/– –/– –/– –/– –/– –/–Fistula –/– –/– –/– –/– –/– –/– –/– 00/5.6 –/– –/– –/– –/– –/– –/– –/– 91 (2.5)Food intolerance –/– –/– –/– –/– –/– –/– 00/1.7 –/– –/– 00/79.0 –/–Gastric outlet stenosis 0.7/00 00/5.0 00/0.3 –/– 0.5/1.4 00/3.5 2.0/00 0.4/4.8 –/– –/– –/– –/– –/– –/– –/– 0 (0.0)Incisional hernia 00/6.7 –/– –/– –/– 00/4.0 00/6.7 00/5.8 00/13.7 –/– –/– –/–Leak 0.7/00 00/00 0.2/00 00/00 00/00 00/00 1.6/00 1.7/00 3.8/00 00/00 –/– –/– –/– –/– 00/0.9 80 (5.0)Pulmonary embolism 00/00 00/00 –/– –/– 0.5/00 0.3/00 0.8/00 0.4/00 –/– 00/00 –/–Ring/band erosion –/– 00/5.0 –/– –/– –/– 00/0.6 00/2.5 –/– –/– 00/00 –/– –/– –/– 00/16.7 –/– 103 (2.8)Ring/band removal –/– –/– –/– –/– 00/0.2 00/6.7 –/– –/– –/– 00/00 –/–Ring/band slippage –/– –/– 00/0.3 –/– 00/00 –/– –/– 00/2.6 –/– 00/00 –/– –/– 0.4/00 –/– –/– 176 (2.4)Small bowel obstruction 0.2/1.7 –/– –/– 1.7/00 0.5/00 00/0.6 00/2.8 0.4/1.4 –/– 2.1/00 –/–Staple line disruption –/– 00/00 00/10.3 00/5.0 –/– 00/9.6 –/– –/– –/– –/– –/– –/– –/– –/– –/– 480 (7.4)Stoma dilation 0.2/00 –/– 00/0.3 –/– –/– –/– –/– –/– –/– –/– –/–Ulceration 00/0.4 00/25.0 00/0.3 –/– 00/0.5 00/6.0 00/1.6 –/– –/– –/– –/– –/– –/– –/– –/– 90 (1.2)Wound related 8.4/00 5.0/00 –/– –/– 0.6/00 –/– 8.8/00 3.4/00 –/– 15.2/00 –/–Death 0.2/00 00/00 00/00 00/00 0.3/00 00/00 0.4/00 0.2/00 3.1/00 00/00 –/– –/– –/– –/– –/– 39 (0.6)Other 0.4/4.8 00/20.0 0.5/0.3 1.7/00 4.0/00 8.8/15.8 –/– 0.5/0.1 –/– 8.7/00 –/–Individual study rates 10.9/13.6 5.0/55.0 1.0/11.8 5.1/5.0 6.5/6.9 9.1/49.5 16.1/14.4 8.4/45.3 6.9/0.0 26.0/79.0 –/– –/– –/– –/– –/– 94 (2.3) –/– –/– 00/3.3 00/2.4 30 (2.3) –/– 1.1/00 –/– 00/2.4 46 (1.5) –/– –/– –/– –/– 143 (2.2) –/– –/– –/– –/– 100 (9.5) –/– –/– –/– –/– 3 (0.3) –/– –/– –/– –/– 92 (1.6) –/– –/– –/– –/– 418 (6.8) –/– –/– 00/00 00/00 26 (0.3) –/– –/– 00/6.7 –/– 159 (4.4) –/– 1.5/00 00/26.7 00/5.7 a,b,cNon-band-related reoperation rate=1,325 / 8,707=15.2 %; band-related reoperation rate=353 / 8,707=4.1 %E early complications, L late complications, –/– not reporteda Salinas [45] data is supplemental to Salinas [39] for complication data onlyb Weighted mean early complication rate=10.9 %c Weighted mean late complication rate=20.0 %

Table 5 Banded Roux-en-Y gastric bypass (B-RYGB) comorbidities resolved OBES SURGStudy author Type 2 diabetes, n (%) Dyslipidemia, n (%) Obstructive sleep apnea, n (%) Hypertension, n (%) 85/138 (61.6)White et al. [35] 53/62 (85.5) 89/265 (33.6) – 19/51 (37.3)Pajecki et al. [36] 13/17 (75.6) 23/30 (76.7) 29/31 (93.5) 18/23 (79.0)Bessler et al. [37] 11/12 (92.0) 7/14 (50.0) – –Awad et al. [41] 3/4 (75.0) 10/15 (66.0) 3/4 (75.0) 122/212 (58.0)Total 80/95 (84.2) 129/324 (39.8) 32/35 (91.4)superobese patients (n=90, 46/44) [37]. Differences in EWL 75 with silastic ring, 68 without) in which median EWL waswere analyzed along with complication rates and improve- calculated at 12, 24, 36, and 48 months [40]. Banded patientsments in comorbidities. Banded patients experienced consis- lost more weight at each respective time point: 74.0 vs 72.0,tently greater EWL at 6 months and 1 and 2 years (43.1 vs 79.0 vs 76.0, 76.0 vs 72.0, and 76.0 vs 71.0 %; however, it24.7, 64.0 vs 57.4, and 64.2 vs 57.2 %, respectively), although was only after 36 months that the between-group difference indifferences did not reach statistical significance. There was no EWL became significant. When surgical failure was definedsignificant difference in complication rates between banded as EWL<50.0 %, at 4 years, a significantly greater percentage(26.0 %) vs non-banded (29.5 %) patients and no difference in of non-banded patients was classified as failures relative torates of comorbidity improvement. However, at 3 years, band- banded patients (30.9 vs 8.2 %, p<0.01). Interestingly, at noed patients had achieved a significantly greater EWL than time point did the percentage of non-banded patients withnon-banded bypass patients (73.4 vs 57.7 %, p<0.05), con- 10.0–20.0 % weight regain differ significantly fromsistent with the observation in the current analysis that the their banded counterparts. For example, at 4-yearadvantages of the banded bypass become apparent at 2–3 follow-up, well beyond the point at which most re-years postsurgery, when stoma dilation is likely to occur in searchers have noted RYGB stomal dilation, 44.2 % ofnon-banded patients, potentially causing early emptying of the non-banded patients vs 30.6 % of banded patients (p=gastric pouch and increased food consumption [44, 58, 59]. 0.22) experienced >10.0 % weight regain and only 9.6 % of non-banded patients vs 10.2 % of banded The results of Bessler et al. and Heneghan et al. dovetail, patients (p=1.0) experienced >20.0 % weight regain.both showing the band being effective in promoting marked Therefore, Rasera et al. concluded that their resultsweight loss in the superobese (typically, the patients with the provided evidence corroborating the superiority of thehighest failure rates following bariatric surgery) [60]. The band with respect to overall weight loss but not withcurrent study results are in-line with and extend both respect to improved control of weight recovery follow-Heneghan et al.’s (2 years) and Bessler et al.’s (3 years) ing gastric bypass.findings for B-RYGB in superobese patients, in that the threeincluded studies with mean baseline BMI≥50 at 5-year In a final included 10-year retrospective comparative studyfollow-up yielded a weighted mean BMI reduction of by Awad et al. (n=378; 260 banded vs 118 non-banded), long-20.4 kg/m2 (95 % CI 17.9, 22.9) and a corresponding EWL term EWL, comorbidity resolution, and quality of life wereof 72.9 %; this result is significantly better than the 5-year analyzed [41]. From 2 years onward, banded patients experi-trend (approximately 58.0 %) found for all patients undergo- enced significantly greater EWL. For example, at 3 years,ing RYGB (Fig. 3). banded vs non-banded patients experienced 77.0 vs 72.4 % EWL; although the difference is significant (p<0.05), it is by The only comparative study included in the present review no means sizeable. However, at 10 years postsurgery, banded(≥3-year follow-up) that did not find superior weight loss (69.4 % follow-up rate) vs non-banded (70.8 % follow-up)results for banded bypass patients (Zarate et al. [42]) reported patients achieved 82.3 vs 61.3 % EWL, a considerable as wella 5-year randomized trial comparing banded vs non-banded as significant result (p<0.0001). Again, these results are con-LRYGB (n=60, 30/30) [42]. Although EWL was somewhat sistent with the trend lines of the current analysis (Fig. 3),greater in banded patients (61.6 vs 59.8 %), there were no wherein banded bypass patients appear to maintain theirstatistically significant differences between banded and non- weight loss to 10+ years while non-banded patients, on aver-banded bypass patients in EWL, BMI reduction, or failure rate age, do not perform as well. In Awad et al., no significant(percentage of patients not reaching a BMI<35) at 5 years. It difference between banded and non-banded groups was seenshould be noted that Zarate et al.’s 61.6 % EWL anchored the with respect to comorbidity resolution (i.e., 75.0 vslow end of the range for studies reporting 5-year mean EWL 50.0 % diabetes resolution, 66.0 vs 50 % dyslipidemiafor B-RYGB in the current research (61.6–83.0 %). resolution, 75.0 vs 60.0 % sleep apnea resolution) and overall complications. In the included comparative study of Rasera et al., a retro-spective analysis of female patients was performed (n=143;

OBES SURGComplications morbidity of 10.0 %, a late complication rate of <6.0 %, a mortality rate of 0.44 %, and a revision rate of <6.0 %. FobiAs with all surgery, bariatric surgery carries some risk. The et al. also provided complication data for laparoscopic B-complication profile associated with B-RYGB has not been RYGB alone: In a sample of 415 laparoscopic TSRVBPclearly defined; notwithstanding band-related complications, patients, there was a 9.4 % early complication rate (i.e.,complication rates for RYGB and B-RYGB should be similar. 2.2 % outlet stenosis, 4.2 % gastric leak, 0.5 % small bowelIn a case-matched JAMA study of 442 patients by Romy et al., obstruction, 1.2 % postoperative bleeding, 0.7 % marginalearly and long-term complication rates after RYGB were 17.2 ulcer, 0.3 % rhabdomyolysis, 0.3 % deep vein thrombosis,and 19.0 %, respectively, with a reoperation rate of 12.7 %. and 0.7 % perioperative mortality). The late complication rateThese numbers are comparable with those presented in the in Fobi et al.’s study was 5.3 % (i.e., outlet stenosis 3.1 %,current review of 8,707 B-RYGB procedures, in which the marginal ulcer 0.5 %, small bowel obstruction 0.7 %, chronicweighted mean early and late complication rates were 10.9 sinus 0.2 %, ring erosion 0.5 %, death 0.2 %) [44].and 20.0 %. The non-band-related reoperation rate was15.2 % (1,325/8,707) and the band-related reoperation rate Although fear of band slippage and/or erosion stopping orwas 4.1 % (353/8,707). reversing weight loss has been an issue with banded bypass [66], data presented in the current study suggest that band- Comparison of the current study’s B-RYGB complications related complications may not be frequent or severe enough towith those of RYGB in the literature yielded interesting re- warrant extreme concern: The rates of band erosion, bandsults. For example, a significant proportion of RYGB patients slippage, and band removal were 2.3, 1.5, and 2.3 %, respec-experience dumping syndrome, whereas no instances of the tively. Fobi et al. reported band erosion rates for bandedsyndrome were reported in B-RYGB patients in the included gastric bypass ranging from 1.9 to 5.9 %, depending onstudies. Postoperative leak at the anastomotic site following whether the B-RYGB was a primary or secondary operationgastric bypass is a potentially life-threatening complication or a revision. Primary B-RYGB had a band erosion rate ofreported in 0.1 to 20.0 % of patients [61–64]. The current B- 1.9 %, while B-RYGB as a secondary operation had a bandRYGB analysis found a leak rate of 1.2 %, which is in the low erosion rate of 5.9 %. The rate of revisional B-RYGB erosionend of the range reported in the literature for RYGB. In a was 4.5 %. Overall incidence of band erosion in 3,632 patientssystematic review and meta-analysis of complications after was 2.5 % [44]. Similarly, Salinas et al. reported a migrationbariatric surgery in RCTs, Chang et al. found a 3.3 % (1.2, 6.0) rate of 1.7 % in 3,800 banded bypasses and suggested that therate of small bowel obstruction [64]; whereas, the rate in the low rate of slippage is inconsequential relative to the 98.3 % ofcurrent B-RYGB study was 2.2 %. patients who typically experience better outcomes with a banded RYGB [39]. Finally, it has been shown by some With respect to mortality, bariatric surgery has developed researchers that significantly greater food intolerance is expe-to the point of being considered as safe or safer than most rienced in banded bypass patients [42] and that it might causemajor surgical procedures. A mortality-specific meta-analysis quality of life to be diminished; yet, in the longest-runningby Buchwald et al. found that among patients undergoing comparative study (10+ years of follow-up) of banded versusRYGB, the mortality rate was 0.44 % for open and 0.16 % non-banded outcomes, included in the current analysis, Awadfor laparoscopic procedures [65]. The mortality rate for B- et al. reported that even with food intolerance factored into theRYGB patients in the current review of combined open and analysis, there was no significant difference between groups inlaparoscopic outcomes was 0.3 %, a rate at the midpoint of the overall quality of life [41]. It should also be noted that in B-aforementioned rates. RYGB patients, where food intolerance (without stenosis) is severe enough to warrant some form of intervention, reoper- Although this study’s B-RYGB complication rates, early ation can be averted with aggressive endoscopic dilation inand late, are comparable to current reported complication rates approximately 97.0 % of patients and weight regain due tofor RYGB, estimated B-RYGB rates may be somewhat high band removal avoided [23].because they incorporated results across a span of 23 yearswherein improvements in the procedure’s technique and the Comorbiditiessurgical learning curve have occurred. For example, post-1990 banded bypass studies included in the current review Maintenance of weight loss following bariatric surgery is alsothat employed the evolved “Fobi technique” (complete gastric critical to maintenance of diabetes resolution and prevention.transection as well as interposition of the jejunal Roux limb) As found in the 1980–1986 Nurses Health Study of 89,941reported markedly fewer staple line disruptions than those of women, the risk of developing T2DM is >38-fold in morbidprior studies. Fobi et al. appear to have carried out the most obesity (BMI ≥35) and >20-fold in obesity (BMI 30–34.9);comprehensive B-RYGB research, reporting outcomes for further, as little as 4.2 kg of weight loss sustained for an3,632 TSRVGBP patients [34]. They found that >90.0 % of average of 3.2 years can reduce the progression to T2DM bypatients lost and, more importantly, maintained ≥50.0 % EWLat ≥5 years. In addition, they reported a perioperative

OBES SURG50.0 % [67, 68]. It has been known since 1994 that sustained Given outcomes of the current review, perhaps the mostweight loss of ≥50.0 % EWL results in a greater than 30-fold salient reason that B-RYGB has not been more oftenreduction in the likelihood of progression from abnormal employed is that RYGB weight loss up to 2-year follow-upglucose tolerance to diabetes [69]. Thus, the maintenance of is excellent. Yet, it is after that time point [33], when the>50.0 %EWL in B-RYGB patients in 90.0 % of patients dumping syndrome phenomenon disappears, that the bandthrough 7–10 years noted in the study of Fobi et al. [44] becomes most valuable. During the second through fourthsuggests that banded bypass may be a doubly effective oper- year postsurgery, the gastric pouch often dilates and restrictionation for those bariatric surgery candidates suffering from at the anastomosis is no longer present; this often leads todiabetes or at risk of its development. patients “out-eating” the restriction [72, 73]. The findings from the systematic review of O’Brien et al. [7] suggestedLimitations that RYGB does not maintain its EWL as consistently as was once thought. Although many questions regarding B-RYGBThe study design of a systematic review ensures assembly of a remain to be researched (e.g., most effective band material anddata set that is as comprehensive and inclusive as possible size), it may come to be viewed as the preferred operation forwith the goal of eliminating selection bias. The analytic power morbidly obese bariatric surgery patients.of the current review was limited by a shortage of experimen-tal studies, lower-quality observational studies (according to Conclusionsthe CEBM scale), no standardized B-RYGB procedure, andno specific bariatric surgery data-reporting requirements. A This is the first systematic review and limited meta-analysis ofpotential limitation of this study was the use of imputation to intermediate-term outcomes exclusively devoted to the B-allow for inclusion of reported mean data unaccompanied by RYGB. We evaluated the outcomes of 8,707 B-RYGB proce-variance data for the calculation of BMI change at 5-year dures in 15 reports to the literature published during a 23-yearfollow-up. Also, in the calculation of EWL trends associated period, from January 1, 1990 to November 1, 2013. The 5-with B-RYGB, a fixed effects model was used in an attempt to year weighted mean EWL of 72.5 % (67.5, 77.4) did notremain close to the design of the comparator O’Brien et al. In significantly vary at 10+ years (69.4 [58.9, 80.0]). The B-addition, the vast majority of included studies reporting RYGB is an effective and safe procedure, with acceptableweight outcomes at any time point did not approach the rates of early (10.9 %) and late (20.0 %) complications.recommended minimally acceptable follow-up of 61.0 % be-yond 1 year [70]; thus, the risk of positive reporting bias Banded gastric bypass appears to result in sustained weightcannot be ruled out. loss, perhaps greater than that achieved by standard RYGB. Longer prospective and RCTs with extensive follow-up rates A limitation of the available data derives from the numer- are needed to fully assess the safety and effectiveness of B-ous abstracts published on the banded bypass that have been RYGB in comparison with RYGB. The current evidencepresented in congresses but never reported formally in the suggests that the banded bypass may be an underutilizedliterature and thus are not suitable for summation by meta- procedure that might effectively serve a greater number ofanalysis. Several B-RYGB reports from South America could patients than it does currently.not be analyzed due to the English language inclusion require-ment. Finally, though we manually searched the reviewed Conflict of interest H Buchwald received no financial or other com-articles for additional citations that might have included ana- pensation for his work as primary co-author of the manuscript. Statisticallyzed banded bypass study arms, a number of articles that fit analysis and manuscript development responsibility were shared by JNour inclusion criteria could have eluded both automated and Buchwald (Medwrite Medical Communications, LLC) and TWmanual search processes because B-RYGB frequently has McGlennon (McGlennon MotiMetrics, LLC) under CRO contract withbeen reported as part of standard RYGB data. Bariatec Corporation, Palos Verdes Peninsula, CA.Assessment ReferencesBetween 10.0 and 20.0 % of RYGB procedures are considered 1. World Health Organization. Obesity and overweight—fact sheet No.failures [49] due to insufficient weight loss and/or weight 311. 2006. Accessible at: http://www.who.int/mediacentre/factsheets/regain—that is, 15,000 to 32,000 patients who suffer poor fs311/en/index.html. Accessed 27 May 2013.outcomes. Banding the bypass is already utilized as a rescue orrevision technique when weight loss has been halted or re- 2. Wang Y, Beydoun MA. The obesity epidemic in the United States—versed [71]. If up to the 32,000 failed patients had been gender, age, socioeconomic, racial/ethnic, and geographicbanded procedures, a good proportion of them might havebeen considered successful over the long term.

OBES SURG characteristics: a systematic review and meta-regression analysis. 26. Tice JA, Karliner L, Watsh J, Petersen AJ, Feldman MD. Gastric Epidemiologic Rev. 2007;29(1):6–28. banding or bypass: a systematic review comparing the two most 3. Flegal KM, Carroll M, Ogden CL, et al. Prevalence and trends in popular bariatric procedures. Am J Med. 2008;121(10):885–93. obesity among US adults, 1999–2008. JAMA. 2010;303(3):235–41. 4. US Department of Health and Human Services. National Institutes of 27. Oxford Centre for Evidence-Based Medicine. Accessed November 2, Health. Overweight and obesity statistics. WIN Weight-Control 2013: http://www.cebm.net/index.aspx?o=5653 Information Network 2010. Accessible at: http://win.niddk.nih.gov/ statistics/index.htm. Accessed 23 May 2013. 28. Jadad AR, Moore A, Carroll D, et al. Assessing the quality of reports 5. Centers for Disease Control and Prevention. National Center for of randomized clinical trials: is blinding necessary? Control Clin Health Statistics. Prevalence of overweight, obesity and extreme Trials. 1996;17:1–12. obesity among adults: United States, trends 1960–1962 through 2005–2006. Health E-Stats 2008; Dec. Accessible at: http://www. 29. Julian PTH, Simon GT, Jonathen JE, et al. Measuring inconsistency cdc.gov/nchs/data/hestat/overweight/overweight_adult.htm. in meta-analyses. BMJ. 2003;327:557–60. Accessed 02 May 2013. 6. Nguyen N, Champion JK, Ponce J, et al. A review of unmet needs in 30. Furukawa TA, Barbui C, Cipriani A, Brambilla P, Watanabe N. obesity management. Obes Surg. 2012;22(6):956–66. Imputing missing standard deviation in meta-analyses can provide 7. O’Brien PE, McPhail T, Chaston TB, et al. Systematic review of accurate results. J Clin Epidem. 2006;59:7–10. medium-term weight loss after bariatric operations. Obes Surg. 2006;16:1032–40. 31. Drew RL, Linner JH. Revisional surgery for severe obesity with 8. Mason EE, Ito C. Gastric bypass in obesity. Surg Clin North Am. fascia banded stoma Roux-en-Y gastric bypass. Obes Surg. 1992;2: 1967;47(6):1345–51. 349–54. 9. US National Library of Medicine National Institutes of Health. Accessible at: http://www.ncbi.nlm.nih.gov/pubmed. Accessed 02 May 32. Capella RF, Capella JF. Converting vertical banded gastroplasty to a 2013. lesser curvature gastric bypass: technical considerations. Obes Surg.10. Mason EE. Vertical banded gastroplasty. Arch Surg. 1982;117:701–6. 1998;8:218–24.11. Eckhout GV, Willbanks OL, Moore JT. Vertical ring gastroplasty for obesity: five year experience with 1463 patients. Am J Surg. 33. Capella JF, Capella RF. An assessment of vertical banded 1986;152:713–6. gastroplasty-Roux-en-Y gastric bypass for the treatment of morbid12. Torres J, Oca C. Gastric bypass lesser curvature with distal Roux-en- obesity. Am J Surg. 2002;183:117–23. Y. Bariatric Surg. 1987;5:10–5.13. Laws H, Piantadosi S. Superior gastric reduction procedure for mor- 34. Fobi MAL, Lee H, Felahy B, Che K, Ako P, Fobi N. Choosing an bid obesity. Am Surg. 1981;93:334. operation for weight control, and the transected banded gastric by-14. Salmon PA. Gastroplasty with distal gastric bypass: a new and more pass. Obes Surg. 2005;15:114–21.7. successful weight loss operation for the morbidly obese. Can J Surg. 1988;31:111–3. 35. White S, Brooks E, Jurikova L, Stubbs RS. Long-term outcomes after15. Linner JR, Drew RL. New modification of Roux-en-Y gastric bypass gastric bypass. Obes Surg. 2005;15:155–63. procedure. Clin Nutr. 1986;5:33–4.16. Fobi M, Lee H, Flemming A. The surgical technique of the 36. Pajecki D, Dalcanalle L, de Oliveira CP, et al. Follow-up of Roux-en- banded Roux-en-Y gastric bypass. Obesity Weight Reg. 1989;S(2): Y gastric bypass patients at 5 or more years postoperatively. Obes 99–102. Surg. 2007;17:601–7.17. Fobi MA. Why the operation I prefer is Silastic ring vertical banded gastric bypass. Obes Surg. 1991;1:423–6. 37. Bessler M, Daud A, Kim T, DiGiorgi M. Prospective randomized18. Howard L, Malone M, Michalek A, et al. Gastric bypass and vertical trial of banded vs nonbanded gastric bypass for the super obese: early banded gastroplasty—a prospective randomized comparison and 5- results. Surg Obes Relat Dis. 2007;3:480–5. year follow-up. Obes Surg. 1995;5:55–60.19. Fobi MAL, Lee H. The surgical technique of the Fobi-Pouch oper- 38. Magro DO, Geloneze B, Delfini R, Pareja BC, Callejas F, Pareja JC. ation for obesity (the transected silastic vertical gastric bypass). Obes Long-term weight regain after gastric bypass: a 5-year prospective Surg. 1998;8:283–8. study. Obes Surg. 2008;18:648–51.20. Capella JF, Capella RF. The weight reduction operation of choice: vertical banded gastroplasty or banded bypass? Am J Surg. 39. Salinas A, Salinas HM, Santiago E, Garcia W, Ferro Q, Antor M. 1996;171:74–9. Silastic ring vertical gastric bypass: cohort study with 83 % rate of 5-21. Fisher B, Barber AE. Gastric bypass procedures. Eur J Gastroenterol year follow-up. Surg Obes Related Dis. 2009;5:455–8. Hepatol. 1999;11:93–7.22. Buchwald H, Oien D. Metabolic/Bariatric surgery worldwide 2011. 40. Rasera-Junior I, Gaino NM, de Oliveira MR, et al. Ring influence on Obes Surg. 2013;23:427–36. ponderal evolution after four years of laparoscopic Roux-en-Y gastric23. Ferraz A, Campos J, Dib V, et al. Food intolerance after banded bypass. Arch Brasileiros de Cir Digest. 2012;25(4):1–5. gastric bypass without stenosis: aggressive endoscopic dilation avoids reoperation. J Metabol Surg Allied Care. 2013; March 8 41. Awad W, Garay A, Martinez C. Ten years experience of published online. banded gastric bypass: does it make a difference? Obes Surg.24. Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA 2012;22:271–8. Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA Statement. PLoS Med. 2009;6(6): 42. Zarate X, Arceo-Olaiz R, Hernandez JM, et al. Long-term e1000097. results of a randomized trial comparing banded versus stan-25. Livhits M, Mercado C, Yermilov I, et al. Preoperative predictors of dard laparoscopic Roux-en-Y gastric bypass. Surg Obes Related Dis. weight loss following bariatric surgery: systematic review. Obes 2013;9:395–7. Surg. 2012;22:70–89. 43. Valezi AC, de Almeida Menezes M, Mali Jr J. Weight loss outcome after Roux-en-Y gastric bypass: 10 years of follow-up. Obes Surg. 2013;23(8):1290–3. 44. Fobi MAL. Banded gastric bypass: combining two principles. Surg Obes Relat Dis. 2005;1:304–9. 45. Salinas A, Santiago E, Yeguez J, et al. Silastic ring vertical gastric bypass: evolution of an open surgical technique, and review of 1,588 cases. Obes Surg. 2005;15:1403–9. 46. Sinha S, Ashby E, Jayaram R, et al. Quality of reporting in random- ized trials published in high-quality surgical journals. J Am Coll Surg. 2009;209:565–71. e1. 47. O’Brien PE, MacDonald L, Anderson M, et al. Long-term outcomes after bariatric surgery: fifteen-year follow-up of adjustable gastric banding and a systematic review of the bariatric surgical literature. Ann Surg. 2013;257(1):87–94. 48. Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a systematic review and meta-analysis. JAMA. 2004;292:1724–37.

OBES SURG49. Christou NV, Look D, MacLean LD. Weight gain after short- and 62. Ballesta C et al. Management of anastomotic leaks after laparoscopic long-limb gastric bypass in patients followed for longer than 10 years. RYGB. Obes Surg. 2008;18(6):623–30. Ann Surg. 2006;244(5):734–40. 63. Alaedeen D et al. Intraoperative endoscopy and leaks after laparo-50. Henning Beck-Nielsen. The metabolic syndrome: pharmacology and scopic RYGB. Am Surg. 2009;75(6):485–8. clinical aspects. Springer, 2013. 64. Chang SH, Stoll CR, Song J. Complications after bariatric surgery: an51. Lim CS, Liew V, Talbott ML, et al. Revisional bariatric surgery. Obes updated systematic review and meta-analysis: 2003–212. JAMA Surg. 2009;19:827–32. Surg. 2013 Dec 18. [Epub ahead of print]52. Radtka JF, Puleo FJ, Wang L, et al. Revisional bariatric surgery: who, 65. Buchwald H, Estok R, Fahrbach K, Banel D, Sledge I. Trends in what, where, and when? Surg Obes Relat Dis. 2010;6:635–42. mortality in bariatric surgery: a systematic review and meta-analysis. Surgery. 2007;142(4):621–35.53. Pories WJ, Swanson MS, MacDonald KG, et al. Who would have thought it? An operation proves to be the most effective therapy for 66. Fobi MAL, Lee H, Igwe D, et al. Band erosion: incidence, etiology, adult-onset diabetes mellitus. Ann Surg. 1995;222:339–50. management, and outcome after banded vertical gastric bypass. Obes Surg. 2001;11:699–707.54. Barhouch AS, Zardo M, Padoin AV, et al. Excess weight loss varia- tion in late postoperative period of gastric bypass. Obes Surg. 67. Hu FB, Manson JE, Stampfer MJ, et al. Diet, lifestyle, and 2010;20:1479–83. the risk of type 2 diabetes mellitus in women. N Engl J Med. 2001;345:790–7.55. Elias AAG-J, Berti LV, Oliveira MR, et al. Derivacores gasticas em Y-de-Roux com anel de silicone para o tratamento da obesidade: 68. Manson JE, Willett WC, Stampfer MJ, et al. Body weight and estudo das complicacoes relacionadas com o anel. Rquivos mortality among women. N Engl J Med. 1995;333:677–85. Brasileiros de Cirurgia Digestiva. 2011;24(4):290–5. 69. Long SD, O’Brien K, MacDonald KG, et al. Weight loss in severely56. Prachand V, Da Vee R, Alverdy J. Duodenal switch provides superior obese subjects prevents the progression of impaired glucose tolerance weight loss in the super obese [BMI>50] compared to gastric bypass. to type II diabetes. A longitudinal interventional study. Diabetes Ann Surg. 2006;244:611–9. Care. 1994;17(5):372–5.57. Heneghan HM, Annaberdyev S, Eldar S, et al. Banded Roux-en-Y 70. Renquist KE, Cullen JJ, Tang S, et al. The effect of follow-up gastric bypass for the treatment of morbid obesity. Surg Obes Related on reporting success for obesity surgery. Obes Surg. 1995;5: Dis. In press, 2013. 285–92.58. Catalano MF, Rudic G, Anderson AJ, et al. Weight gain after bariatric 71. Vijgen GH, Schouten R, Bouvy ND, et al. Salvage banding surgery as a result of a large gastric stoma: endotherapy with sodium for failed Roux-en-Y gastric bypass. Surg Obes Relat Dis. morrhuate may prevent the need for surgical revision. Gastro Intest 2012;8(6):803–8. Endosc. 2007;66:240–5. 72. Fobi MAL, Lee H, Igwe D, et al. Transected silastic ring vertical59. Yimcharoen P, Heneghan HM, Singh M, et al. Endoscopic findings gastric bypass with jejunal interposition, a gastrostomy and a and outcomes of revisional procedures for patients with rate recidi- gastrostomy site marker (Fobi Pouch operation for obesity) vism after gastric bypass. Surg Endosc. 2011;25:3345–52. (Chapter 22). In: Update: Surgery for the morbidly obese patient, eds. Toronto: FD-Communications Inc, 2000, p. 203.60. Suter M, Calmes J-M, Paroz A, et al. Results of Roux-en-Y gastric bypass in morbidly obese vs superobese patients. Arch Surg. 73. Fobi MAL, Lee H. The surgical technique of the Fobi-Pouch oper- 2009;144(4):312–8. ation for obesity (the transected silastic vertical gastric bypass). Obes Surg. 1998;8:283–8.61. Agaba EA et al. Laparoscopic vs open gastric bypass in the manage- ment of morbid obesity: a 7-year retrospective study of 1,364 patients 74. Capella RF, Capella JF, Mandac H, et al. Vertical banded gastroplasty— from a single center. Obes Surg. 2008;18(11):1359–63. gastric bypass: preliminary report. Obes Surg. 1991;1:389–95.

OBES SURGDOI 10.1007/s11695-010-0113-3 REVIEWMetabolic Surgery for the Treatment of Type 2 Diabetesin Patients with BMI <35 kg/m2: An Integrative Reviewof Early StudiesM. Fried & G. Ribaric & J. N. Buchwald & S. Svacina &K. Dolezalova & N. Scopinaro# Springer Science+Business Media, LLC 2010Abstract Type 2 diabetes mellitus (T2DM) resolution in with 6–216 months follow-up. Patients lost a clinicallymorbidly obese patients following metabolic surgery meaningful, not excessive, amount of weight (from BMIsuggests the efficacy of T2DM surgery in non-morbidly 29.4 to 24.2; −5.1), moving from the overweight into theobese patients (body mass index [BMI] <35 kg/m2). This normal weight category. There were 85.3% patients wholiterature review examined research articles in English over were off T2DM medications with fasting plasma glucosethe last 30 years (1979–2009) that addressed surgical approaching normal (105.2 mg/dL, −93.3), and normalresolution of T2DM in patients with a mean BMI <35. glycated hemoglobin, 6% (−2.7). In subgroup comparison,Weighted and simple means (95% CI) were calculated to BMI reduction and T2DM resolution were greatestanalyze study outcomes. Sixteen studies met inclusion following malabsorptive/restrictive procedures, and in thecriteria; 343 patients underwent one of eight procedures preoperatively mildly obese (30.0–35.0) vs overweight (25.0–25.9) BMI ranges. Complications were few with lowM. Fried (*) : K. Dolezalova operative mortality (0.29%). Novel and/or known mecha- nisms of T2DM resolution may be engaged by surgery at aCentre for Treatment of Obesity and Metabolic Disorders, BMI threshold ≤30. The majority of low-BMI patientsOB klinika, experienced resolution of laboratory and clinical manifes-Pod Krejcarkem 975, tations of T2DM without inappropriate weight loss.130 00 Prague 3, Czech Republice-mail: [email protected] Keywords Bariatric . Metabolic surgery . Type 2 diabetes . Biliopancreatic diversion . Gastric bypass . LaparoscopicM. Fried adjustable gastric banding . Sleeve gastrectomy . IlealFirst Faculty of Medicine, Charles University, interpositionPrague, Czech Republic IntroductionG. RibaricEuropean Surgical Institute, Progress in science depends on new techniques, newEthicon Endo-Surgery (Europe) GmbH, discoveries, and new ideas, probably in that order.Hamburg, Germany —Sydney Brenner 1980J. N. Buchwald Surgical Study of DiabetesMedwrite Medical Communications, LLC,Maiden Rock, WI, USA Since the late nineteenth century, surgery has been at the axis of investigation of the mechanisms of diabetes, glucoseS. SvacinaThird Medical Department, Clinical Department of Endocrinologyand Metabolism of the First Faculty of Medicine and GeneralUniversity Hospital,Prague, Czech RepublicN. ScopinaroUniversity of Genoa School of Medicine,Genoa, Italy

OBES SURGmetabolism, and the interrelation between type 2 diabetes correlated with the extent of obesity, incidence of hyper-mellitus (T2DM) and morbid obesity (body mass index glycemia seemed less closely related. They also observed[BMI, kg/m2] ≥40, or ≥35 with comorbidities [1]). that, following all variants of the BPD operation, patients’Innovation by surgeons in the laboratory and clinic has corrected caloric intake and absorption was associated withled to development of the bariatric surgical techniques that 100% resolution of their hyperglycemia and overt T2DM,currently provide the only effective means of long-term facilitating maintained euglycemia on an unrestricted diettreatment of morbid obesity [2, 3], T2DM in morbid with no anti-diabetic medications [18].obesity [4–6], and possibly, T2DM in non-morbidly obesepatients (BMI <35). Pories et al. advanced the early bariatric surgical findings relating to T2DM in their 1987 paper on the Greenville Lithuanian surgeon Oskar Minkowski, working in 1889 gastric bypass, wherein they posited that postoperativewith the physician Josef von Mering, surgically removed the return to euglycemia might result from hormonal changespancreases of dogs and recognized that their postoperative secondary to bypass of the antrum and duodenum [19]. Insymptoms were due to the development of diabetes and that the same year, the report of patients of Jimenez et al.the pancreas regulated blood sugar through insulin production following gastroplasty was among the first to elucidate[7]. Based on this model for the study of diabetes and a 1920 factors other than adiposity (i.e., fat distribution, nutrientarticle by United States surgeon Moses Baron about the load, fat cell size, circulating fatty acids) that regulaterelation of the islets of Langerhans to diabetes [8], Canadian insulin metabolism and diabetes resolution in the morbidlysurgeon Frederick Banting isolated and extracted insulin obese [20]. The 1992 and 1995 landmark papers of Poriesfrom the dog pancreas and administered it to diabetic dogs, et al. hypothesized that bariatric procedures might be ansuccessfully lowering their blood sugars; 6 weeks later, in equally effective primary treatment for T2DM in non-1922, Banting’s research team injected a purified version of morbidly obese patients [21, 22] (whether overweight [BMIfetal calf insulin into a dying diabetic 14-year-old boy, 25.0–29.9] or mildly obese [BMI 30.0–34.9] [1]), framingclearing all signs of his disease [9]. an intriguing opportunity for future T2DM research and therapy. At mid-twentieth century, surgeons Friedman and col-leagues demonstrated the involvement of the digestive tract Late twentieth and early twenty-first century bariatricin metabolic disease in their 1955 report of T2DM surgery publications centered on reporting efficacy andresolution following gastrectomy in non-morbidly obese comorbidity reduction [3, 23–26], risk-to-benefit ratio [27–patients [10]. In the 1970s and 1980s, bariatric surgeons 30], and economic end points [31–33] associated withcharacterized key mechanisms of the dynamic relationship weight-loss procedures in high-BMI patients. Within thesebetween morbid obesity and T2DM: Bosello et al. [11] findings, clinically meaningful results regarding surgicalobserved that, in jejunoileal bypass, weight loss was the T2DM reduction were manifold. Chief among them wereprimary factor in lowering glycemia and decreasing the findings that glycemic control is restored in mostinsulinemia; Ackerman [12] and Halverson et al. [13] morbidly obese patients within weeks of surgery rather thanfound that nearly immediately after malabsorptive bariatric only secondary to extensive weight loss over time [34–37]procedures, most diabetic patients became euglycemic and and that resolution to normal levels of fasting plasmawere freed from medication use before significant weight glucose (FPG), fasting plasma insulin, glycated hemoglobinloss and that following weight loss, basal hyperinsulinemia (HbA1C), without requirement of diabetes medications,improved. Herbst et al. [14] followed morbidly obese occurs in all three traditional bariatric procedural categoriespatients with T2DM who received restrictive procedures, [38–41] (i.e., restrictive, malabsorptive/restrictive, primarilynoting their improvement in diabetes and ascribing it to malabsorptive [42]), but is most pronounced and rapid inincreased insulin receptors; Schrumpf et al. [15] corrobo- the primarily malabsorptive and malabsorptive/restrictiverated these findings in morbidly obese patients undergoing procedures [43–47]. A meta-analysis of the world bariatricgastric bypass, inferring that the primary mechanism of literature summarized diabetes outcomes for 3,188 patientsimproved glucose tolerance was decreased basal insulin in 103 treatment arms [5], finding complete T2DMsecretion coupled with enhanced hepatic insulin extraction. resolution (defined as normal FPG and cessation of diabetes medications) in 78.1% of patients at 2 years following In the late 1970s, Scopinaro et al. pioneered the bariatric surgery and by procedure, in 95.1% after BPD/biliopancreatic diversion procedure (BPD) in dogs [16] duodenal switch (DS), 80.3% with Roux-en-Y gastricand humans [17], eventually testing four versions to bypass (RYGB), 79.7% with gastroplasty, and 56.7% aftermaximize weight loss with minimized physiologic and laparoscopic adjustable gastric banding (LAGB). Finally,nutritional complications. In 1986, describing the evolution the concepts of restriction and malabsorption, mechanismsand outcomes of the BPD, Scopinaro et al. first noted that, traditionally associated with bariatric surgery and thealthough development of some comorbidities of obesity salutary response of T2DM in the morbidly obese, have(e.g., sleep apnea, arterial hypertension) were markedly

OBES SURGled to elucidation of neurogenic and hormonal mechanisms (DJB), essentially a stomach-preserving RYGB, that theythat are sharpening our understanding of surgery’s effect on employed in GK rats to investigate the upper intestinaldiabetes [48–53]. hypothesis of diabetes resolution. When compared to sham- operated controls, DJB rats achieved the same caloric intake In 1978, Buchwald and Varco published a book entitled, and weight with substantial, durable glucose homeostasis, anMetabolic Surgery [54], heralding a new field whose effect directly incurred by exclusion of the proximal intestinemission encompassed, but was not limited to, weight loss. rather than weight loss [59]. In contrast, the 2005 studyPories, in his 2003 article, “So you think we are bariatric demonstrated normalization of hyperglycemia after DJB insurgeons? think again,” presaged the field’s evolving Zucker fatty rats, both fasting and fed, but with diminishedresearch opportunities and echoed the aptness of the caloric consumption and weight gain, suggesting theoriginal comprehensive term, “metabolic surgery,” framing viability of the lower intestinal hypothesis of diabetesthe developments of the preceding 25 years [55]. In resolution [60]. In 2006, Rubino et al. performed a DJBparticular, outcomes for diabetes resolution in the morbidly variant in GK rats that directed nutrients from the stomach toobese suggest the viability of surgically treating T2DM in both the proximal jejunum and duodenum with no anti-low-BMI patients and have provided fresh impetus toward diabetic effect. Subsequently, rats with only the gastro-preclinical investigations. jejunostomy subjected to exclusion of the proximal intestine showed markedly improved glucose tolerance, and when ratsPreclinical Studies of T2DM with resolved diabetes were surgically restored to gastro- jejunostomy alone, their diabetes returned; this suggestedT2DM surgical experiments published between 2004 and that an upper intestinal segment bypass could resolve2009 (and surgical techniques now in initial clinical use for diabetes regardless of caloric intake, nutrients diverted toT2DM reduction) were extensions of surgical progenitors the distal intestine, weight, or weight loss [61].dating to the mid-twentieth century, including the ilealbypass procedure, employed originally to investigate Patriti et al. showed similar improvement of glucosemechanisms of cholesterol absorption and metabolism and tolerance with no alteration of caloric intake or weight inprovide therapeutic reduction of cholesterol and heart GK rats following II [62], confirming and expanding thesedisease [56]; biliopancreatic diversion devised as a means findings in a 2007 study [63]. Pacheco et al. achievedof inducing selective malabsorption with weight loss [16]; significantly improved glucose tolerance in GK rats withand ileal transposition, or interposition (II), introduced as a gastrojejunal bypass with decreased levels of leptin and nostrategy for weight reduction through increased release of change in insulin and enteroinsular hormone levels [64].intestinal glucagons [57]. These, and other metabolic Wang et al. compared modified DJB and II in GK rats withsurgical strategies, have been adapted for use in current sham-operated controls wherein the two operations werelean-animal experiments to elucidate mechanisms of glu- found to result in equivalent glucose control and weightcose homeostasis and T2DM resolution. [65]. Recently, Strader et al. [66], in a controlled compar- ison of II in streptozotocin-treated Long-Evans rats (i.e., Mistry et al. published a very informative review sham-diabetic rats) and sham-operated normal euglycemicstratifying rodent models according to their appropriateness rats, found that II improved glucose tolerance in bothfor use in bariatric vs diabetic surgery research, citing the groups with no change in weight.Goto Kakazaki (GK), Cohen (C), and Torri (T) strains oflean rodent as best suited to investigation of the effect of The intriguing preclinical findings showing glucose con-glucose metabolism on T2DM following diabetes surgery trol with little to no weight change after metabolic surgery[58]. They noted the value of the GK rat particularly as a suggest the possibility of definitive surgical resolution ofmodel to study the relationship between beta cell mass and T2DM in humans. These historic antecedents led us tocomplications associated with T2DM; the C rodent, as conduct a review and integration of the outcomes to date ofuseful for observing expression of the T2DM phenotype in the published clinical studies of metabolic surgery in thegenetically predisposed subjects due to nutritional and treatment of T2DM in patients with BMI <35.environmental factors; and the T strain as appropriate forstudying long-term complications of T2DM, although no Methodsmetabolic surgical study has yet been published using the Cor T model. Search Strategy and Study Inclusion Criteria Eight recent preclinical studies describe T2DM-specific An extensive literature search was undertaken. Inclusionsurgical research in rodents. Between 2004 and 2006, criteria specified English language studies of all designsRubino and colleagues published a series of experiments onthe anti-diabetic properties of metabolic procedures. Rubinoand Marescaux [59] developed a duodenal-jejunal bypass

OBES SURGcontaining data on treatment of T2DM by any form of subgroup analyses, in the text. In studies where T2DMbariatric/metabolic surgery in human subjects where the was present in a small subset of patients, and/or was notmean study BMI was <35. An electronic search was reported for all end points collected for this review, theperformed in MEDLINE, Current Contents, Science Cita- number of patients in that subset for each variable as welltion Index, and the Cochrane Library for peer-reviewed as in the total study was reported. Though new metabolicjournal articles published during the last 30 years (January surgery nomenclature that incorporates actions of hormonal1, 1979 to October 1, 2009). The search terms used and and neurogenic agents is evolving, procedure subgroupcross-referenced were: “non-insulin dependent diabetes,” results were organized and tabulated by traditional bariatric“type 2 diabetes,” “T2DM,” “bariatric surgery,” “BMI surgery categories [42].<35,” “low-BMI,” “obesity,” “diabetes surgery,” “metabolicsurgery,” “hyperinsulinemia,” “normoglycemia,” “euglyce- Studies reported data at a varying range of time points,mia,” “insulin,” and “metabolism.” To supplement the from 6 to 216 months (mean 23 months); therefore, theelectronic results, focused reading and review of the follow-up time point for data integrated in this review isreferences in the identified studies was employed. expressed as “≥6 months.” Values for total cholesterol, plasma insulin, and homeostasis model assessment (anData Reporting estimate of steady-state beta cell function and insulin sensitivity as percentages of a normal reference populationT2DM is defined according to the American Diabetes [69]) were reported with insufficient frequency to beAssociation (ADA) 2009 Standards of Medical Care in assessed.Diabetes as FPG (8-h) ≥126 mg/dL [61]. The NationalInstitutes of Health (NIH) guidelines define normal FPG as Statistical Analysis≤99 mg/dL [67]. The ADA glycemic goal for HbA1C fornonpregnant adults is <7%; the ADA document references Weighted means for pre- and postoperative BMI, FPG, andthis range to “a nondiabetic range of 4.0–6.0% using a HbA1C (and their respective change values) were calculatedDCCT-based assay” [68]. to allow for some correction in the integration of data derived from studies of varying series sizes and data In this review, study authors, country of study origin availability. A general equation for calculation of weighted(i.e., first author), surgical procedure, and study design are means (also presented in expanded form) was used. For thepresented. Mean BMI (and BMI range) and primary T2DM number of studies included in a particular outcome analysisoutcomes, including means and ranges for FPG (mg/dL), (ranging from 1 to n, dependent on studies reporting dataHbA1C (%), and the change (drop or rise) in these measures on the variable of interest), the weighted mean ( ) wasfrom baseline, were compiled and reported in as many calculated by summing (Σ) products of all case-weightedinstances as data were available. Consensus mean estimates study means (wixi) and dividing the result by the totaland associated 95% confidence intervals [CI] for total weight (wi) (wi, combined N of included studies):review group analyses are reported in tables and for A second estimate of the consensus mean for in- Resultsdividual variables was derived using the mean of meansmodel (i.e., the simple mean of each of the individual Study Characteristicsstudy means). Standard 95% CIs are presented asapproximations of uncertainty limits surrounding consen- Sixteen studies [71–86] met inclusion criteria assus mean estimates. FPG reported in SI units (mmol/L) identified by literature search and focused reading. Tenwas divided by the conversion factor 0.0555 [70] to of the 16 studies identified (63%) were published in theprovide the conventional units (mg/dL) reported by most past 2 years (2008–2009); one of the remaining sixreviewed studies. Calculations were performed using reports dates to 1998. The majority of studies (14,SPSS® (version 16.0, SPSS Inc, Chicago, IL, USA) 87%) were performed outside of the United States, instatistical software. Brazil (6, 38%), Italy (5, 31%), Australia (1, 6%), the

OBES SURGTable 1 Characteristics of studies of T2DM patients with BMI <35 kg/m2Study Countrya Procedure Design1. Noya et al. [79] Italy BPD-SPP Prospective: observational study2. Angrisani et al. [71] Italy LAGB Retrospective: multi-center prospectively collected database experience of 27 Italian sites3. Parikh et al. [72] USA with Australian database LAGB Retrospective: prospectively collected database study4. Cohen et al. [74] Brazil RYGB Prospective: IRB-approved observational study5. Scopinaro et al. [81] Italy BPD Retrospective: prospectively collected database study6. Cohen et al. [80] Brazil DJB Prospective: preliminary results in two of 50 IRB-approved patients in an observational study7. Lee et al. [75] Republic of China MGB Retrospective: prospectively collected database study8. Sultan et al. [73] USA LAGB Retrospective: prospectively collected database study; surgery in <35 vs >35 BMI groups9. DePaula et al. (a) [76] Brazil II-DSG Prospective: IRB-approved observational study10. DePaula et al. (b) [77] Brazil II-SG and II-DSG Prospective: IRB-approved observational study11. Ramos et al. [82] Brazil DJB Prospective: IRB-approved observational study12. Geloneze et al. [83] Brazil DJB Prospective: IRB-approved study, surgery vs matched control group on standard medical care13. Ferzli et al. [84] USA DJB Prospective: IRB-approved observational study14. Chiellini et al. [85] Italy BPD Prospective: IRB-approved pilot study, surgery vs matched group on energy-restricted diet15. Shah et al. [78] India RYGB Prospective: IRB-approved observational study16. Scopinaro et al. [86] Italy BPD Prospective: IRB-approved pilot study, surgery in 25.0–35.0 BMI (with comparison of 25.0–29.9 and 30–35)IRB institutional review board, BPD biliopancreatic diversion, BPD-SPP stomach- and pylorus-preserving BPD, RYGB Roux-en-Y gastric bypass,LAGB laparoscopic adjustable gastric banding, DJB duodenal-jejunal bypass, MGB mini-gastric bypass, II-SG ileal interposition with sleevegastrectomy, II-DSG ileal interposition with diverted sleeve gastrectomya Country of first author’s affiliationPeople’s Republic of China (1, 6%), and India (1, 6%; (66%), and patients’ mean age at baseline was 46.2 years.Table 1). Twenty-seven patients underwent a form of restrictive procedure; 223, a malabsorptive/restrictive procedure; and Studies featured the following surgical procedures: BPD 93, a primarily malabsorptive procedure (Table 2). At(three studies); stomach- and pylorus-preserving BPD baseline, the total review group weighted mean BMI was(BPD-SPP, one); RYGB (two); LAGB (three); DJB (four); 29.4 kg/m2; FPG, 198.5 mg/dL; and HbA1C, 8.7% (Table 3).mini-gastric bypass (MGB, one); II with sleeve gastrectomy Most patients were taking a form of daily oral or injectable(II-SG, one); and II with diverted sleeve gastrectomy (II- diabetes medication and/or insulin or a combination thereof.DSG, two). Three of the total 16 studies (19%) employed Where specified in the studies reviewed, the majority ofrestrictive procedures, five (31%) were malabsorptive/ patients had been diagnosed with T2DM within 1 to 10 yearsrestrictive, and eight (50%) were primarily malabsorptive. prior to surgery; the complete range of studies that metStudy designs ranged from a preliminary report of two inclusion criteria encompassed patients diagnosed withpatients in a larger prospective study approved for 50 T2DM as many as 18 years prior to surgery.patients, to retrospective case series, and matched con-trolled trials. Eleven (69%) of the studies were prospec- Efficacytively performed, nine of which specified that they soughtinstitutional review board (IRB) approval to conduct their BMI, FPG, HbA1C, and Medication Status in the Integratedstudies (Table 1). Studies defined resolution of T2DM Review Groupfollowing surgery variably, not consistently applying theNIH definition of normal FPG (≤99 mg/dL). The integrated study postoperative weighted mean BMI was 24.2 kg/m2 (with a weighted mean change of −5.1).Patient Characteristics Postoperative FPG was 105.2 mg/dL (−93.3). HbA1C was 6.0% (−2.7) (weighted and simple means and CIs inIn a total of 343 patients with T2DM who underwent Table 3). The majority of low-BMI patients experiencedmetabolic surgery, most study participants were female

Table 2 Characteristics of patients with T2DM and BMI <35 kg/m2 by procedure category OBES SURG Study N Female/Male (%) Mean age (range; years)Restrictive Angrisani et al. [71] 4 w/ T2DM (210 study total) 176 (84)/34 (16) study total 38.2 (17–66) study meanMalabsorptive/Restrictive Parikh et al. [72] 8 w/ T2DM (93 study total) 76 (82)/17 (18) study total 44.6 (16–76) study mean Sultan et al. [73] 15 w/ T2DM (53 study total) 44 (83)/9 (17) study total 46.9 (16–68) study meanPrimarily Malabsorptive Cohen et al. [74] 37 30 (81)/7 (19) 34.0 (28–45) Lee et al. [75] 44 w/ T2DM 38 (86)/6 (14) 39.0 DePaula et al. (a) [76] 69 22 (32)/47 (68) 51.4 (41–63) DePaula et al. (b) [77] 58 18 (31)/40 (69) 51.4 (40–66) S Shah et al. [78] 15 7 (47)/8 (53) 45.6 (33.6–57.6) Noya et al. [79] 10 5 (50)/5 (50) 52.1 (40–62) Cohen et al. [80] 2 studied of 50 approved 0 (0)/2 (100) 47.0 (43–51) Scopinaro et al. [81] 7 2 (29)/5 (71) 49.0 (39–60) Ramos et al. [82] 20 9 (45)/11 (55) 43.0 (29–60) Geloneze et al. [83] 12 3 (25)/9 (75) 50.0 Ferzli et al. [84] 7 – 43.3 (33–52) Chiellini et al. [85] 5 2 (40)/3 (60) 48.3 Scopinaro et al. [86] 30 11 (37)/19 (63) 56.0 (43–69) Total Mean 343 66%/34% 46.2 (34–56)BMI body mass index, HbA1C glycated hemoglobinT2DM resolution within days or weeks of surgery that was HbA1C were not reported and could not be summed for thedurable to individual study end points (mean follow-up category.23 months, range 6–216 months). T2DM resolution(euglycemia) was defined and reported in a variety of ways Respective pre- and postoperative variables in fivein the 16 reviewed studies. When defined as an FPG≤99 mg/dL, the patient group approached resolution. studies (223 patients) with malabsorptive/restrictive proce-Defined as an HbA1C ≤6% and off all diabetes medications, dures [74–78] were, for BMI, 28.9 (28.5–29.2; 29.4; 25.6–85.3% resolved. When defined as an HbA1C <7% with 32.8) and 22.4 (22.3–22.5; 22.7; 20.8–24.5), change ofmarked reduction of FPG, the patient group trended toward −5.7 (−6.1 to −5.3; −6.1; −11.8 to −0.4). FPG was 196.7resolution. (192.7–200.6; 196.2; 150.3–242.2) and 97.1 (96.1–98.1; 94.6; 84.2–105), change of −99.6 (−102.9 to −96.3; Specific preoperative diabetes medications, number of −101.6; −143.0 to −60.2). HbA1C was 8.5 (8.4–8.7; 8.8;medications, and percentage of patients taking medications 6.9–10.6) and 5.7 (5.7–5.8; 5.8; 5.4–6.2), change of −2.9were not consistently reported, precluding calculation of the (−3.0 to −2.7; −3.0; −4.6 to −1.4).change in medication status for the total reviewed group.Yet, the number of patients not taking diabetes medications Eight studies with primarily malabsorptive procedurespostoperatively was provided in most studies, facilitating (93 patients) [79–86] had a pre- and postoperative BMI ofcalculation of a weighted mean percentage. When defined 29.5 (29.0–30.0; 29.8; 27.5–32.1) and 25.7 (25.5–25.9;as not taking any anti-diabetic medication, 85.3% of low- 26.3; 25.2–27.5), change of −3.8 (−4.3 to −3.4; −3.5; −5.7BMI patients experienced diabetes resolution. to −1.3). FPG was 203.7 (197.8–209.6; 207.6; 168.4– 247.0) and 128.4 (123.2–133.6; 126.4; 96.5–156.2), changeBMI, FPG, HbA1C, and Medication Status by Procedure of −74.4 (−80.6 to −68.2; −74.6; −122.8 to −26.4). HbA1CSubgroups was 9.1 (8.9–9.2; 9.0; 8.5–9.4) and 6.8 (6.6–7.0; 6.7; 5.5– 8.0), change of −2.2 (−2.4 to −2.0; −2.0; −3.1 to −0.8).Results for BMI, FPG, and HbA1C by procedure sub-groups were compiled. In the three studies describing In the restrictive and malabsorptive/restrictive categories,outcomes for restrictive procedures (27 patients) [71–73],pre- and postoperative weighted mean BMI results were BMI was reduced into a lower weight category; the33.1 (95% CI 32.9–33.3; simple mean 33.2; 95% CI 31.7–34.8) and 26.9 (26.3–27.5; 27.7; 22.9–32.5), change of primarily malabsorptive procedure BMI remained in its−6.2 (−6.7 to −5.7; −5.5; −9.5 to −1.6); FPG and mean preoperative (overweight) category. Preoperative FPG was only slightly higher in the primarily malabsorptive proce- dures than in the malabsorptive/restrictive, but unlike the malabsorptive/restrictive, achieved prediabetic rather than normal FPG values. With respect to HbA1C, preoperative values for the primarily malabsorptive procedures were

Table 3 Weighted mean outcomes: body mass index, fasting plasma glucose, glycated hemoglobin, and diabetes medication status in T2DM patients with BMI <35 kg/m2 OBES SURG Study N BMI (kg/m2) Fasting Plasma Glucose (mg/dL) HbA1C (%) Off T2DM Medications at ≥6 months Mean Preop Mean Post Mean P Mean Preop Mean Post Mean P Mean Preop Mean Post Mean P N% Change Range Range Change Range Range Range Range ChangeRestrictive Angrisani 4 33.9 29.7 study −4.2 –– –– – – – – – 4 100 et al. [71] (25.1–35) mean −5.1 study mean −7.3 Parikh 8 27.6 study – –– –– – – – – – 8 100 32.7 mean et al. [72] (30.0–34.0) study mean 25.8 study Sultan 15 mean –– –– – – – – – 8 50 33.1 et al. [73] (28.2–35.0) not stated study mean (∼78% EWL)Malabsorptive/ Cohen 37 – 146.0 88.0 –58.0 <0.01 – ≤6.0 – – 37 100 Restrictive et al. 2006 32.5 (126.0–242.0) (60.0–94.0) [74] 44 (32.0–34.9)Primarily 69 Malabsorptive Lee et al. 31.7 23.2 –8.5 – 168.7 88.6 –80.0 – 7.3 5.6 −1.7 – –– 2008 [75] 58 <0.001 66 96 25.7 21.8 −3.9 <0.001 218.1 102.0 −116.0 <0.001 8.7 5.9 (4.8–8.5) −2.8 DePaula 15 <0.05 – – et al. (a) 10 (21.8–29.2) (17.7–25.8) (90.0–334.0) (73.0–161.0) (7.5–13.7) [76] 2 <0.001 15 100 7 28.2 – – – 215.3 105.4 −110.0 <0.001 8.9 5.4 −3.5 – 10 100 DePaula 20 (20.0–34.8) – 2 100 et al. 12 (78.0–333.0) (66.0–167.0) (7.5–12.8) – 7 100 (b) [77] 7 <0.001 18 90 5 28.9 23.0 −5.9 <0.001 233.0 89.2 −144.0 <0.001 10.1 6.1 −4.0 <0.01 0 0 S Shah (22.0–35.0) −5.7 NS 1 14 et al. [78] 30 27.5 −1.4 –– –– –– – – <0.042 5 100 33.3 (20.46–32.4) −6.4 Noya 343 (24.0–38.9) −2.7 – – 91.8 – –– 5.35 (5.0–5.7) – et al. [79] 28.3 −0.5 – (89.0–94.0) −132.0 29.6 (27.0–29.5) <0.001 252.7 −75.0 –– – – Cohen (29.0–30.3) NS (131.0–400.0) 121.0 −27.0 et al. [80] 27.0 (68.0–146.0) <0.001 8.8 6.8 (5.8–7.9) −2.0 33.4 (22.0–31.2) 171.3 (7.5–10.2) 7.8 −1.1 Scopinaro (32.0–34.6) (127.0–242.0) 96.3 et al. [81] 24.4 (78.0–118.0) <0.01 8.9 27.1 (20.2–28.3) 185.4 Ramos (25.0–30.0) 158.2 et al. [82] 25.6 26.1 Geloneze et al. [83] 27.5 27.3 −0.2 – 208.8 154.8 −55.0 NS 9.4 8.5 (6.3–12) −0.9 (21.7–33.0) (23–33) −5.8 Ferzli (112.0–286.0) (63.0–315.0) (6.6–11.8) et al. [84] 30.9 25.1 <0.043 – –– – 8.48 5.7 −2.8 Chiellini et al. 2009 30.6 25.3 −5.3 – 220.0 136.0 −84.0 – 9.3 (7.1–12.9) 6.3 (4.8–8.3) −3.0 – 25 83 [85] (25.3–34.9) (123.0–362.0) (83.0–250.0) Scopinaro et al. 2009 206 [86] 85.3 Total Weighted 29.4 24.2 −5.1 198.5 105.2 −93.3 8.7 6.0 –2.7 Mean

Off T2DM OBES SURG Medications N% higher than those for the malabsorptive/restrictive, but both categories achieved postoperative values within the ADA 82.0–88.7 goal range. 80.9 61.2–100 By procedure category, restrictive procedures achieved a 72.2% (62.2–82.2%; 83.3%; 11.6–100%) resolution rate Mean P −2.8–−2.6 relative to medication use. Malabsorptive/restrictive and Change −2.4 primarily malabsorptive procedures achieved 97.7% (97.3– −3.2–−1.6 98.1%; 98.7%; 93–100%) and 72.9% (65.7–80.3%; 73.4%; 38.6–100%) resolution off medications, respectively. Mean Preop Mean Post Range 5.9–6.1 6.3 BMI, FPG, HbA1C, and Medication Status by Overweight 5.6–7.0 and Obese SubgroupsHbA1C (%) Range 8.6–8.8 A comparison of change in BMI, FPG, and HbA1C in 2 8.9 BMI subgroups within the total <35 BMI T2DM population 8.3–9.5 at ≥6 months is shown in Fig. 1; that is, preoperatively overweight (BMI 25.0–29.9; seven studies; 183 patients) vs P preoperatively mildly obese (BMI 30.0–35.0; nine studies; 160 patients). The respective mean pre- and postoperative Mean Change −96.3–−90.0 BMI for the subset of overweight vs the subset of mildly −88.1 obese patients was 27.0 (26.9–27.2; 27.6; 26.3–28.9) and −114.5–−61.6 23.1 (22.8–23.5; 25.1; 22.4–27.7), change of −3.4 (−3.7 to −3.1; −2.4; −4.7 to −0.1) vs 32.1 (31.9–32.2; 32.5; 31.6–Fasting Plasma Glucose (mg/dL) Mean Post Range 103.0–107.4 33.3) and 25.2 (24.8–25.5; 26.4; 24.7–28.1), change of 111.9 −6.8 (−7.0 to −6.5; −6.0; −7.2 to −4.9). The mean BMI of 94.1–129.8 the preoperatively overweight group dropped into the normal weight category and the obese group dropped into Mean Preop the overweight range. This constitutes a notable divergence Range between the BMI <30 and >30 groups in trajectory, with the 195.1–201.8 preoperatively heavier subset of patients losing almost 201.9 twice as much weight. 178.3–225.5 The respective mean pre- and postoperative FPG for Mean P overweight and mildly obese patients was 210.7 (208.3– Change 213.2; 205.3; 181.4–229.2) and 107.0 (104.4–109.6; 114.0; −5.3–−4.8 86.6–141.3), change of −103.6 (−107.7 to −99.5; −87.8; −4.5 −133.3 to −42.4) vs 179.6 (173.5–185.8; 196.9; 119.8–273.9) −5.9–−3.0 and 102.3 (98.5–106.2; 108.4; 70.2–146.5), change of −77.2 (−80.4 to −73.9; −88.5; −138.1 to −38.9). HbA1C was 8.9 at ≥6 months Mean Preop Mean Post Range 23.9–24.4 (8.8–9.0; 9.1; 8.6–9.7) and 6.0 (5.9–6.2; 6.6; 5.4–7.7), 25.8 change of −2.9 (−3.0 to −2.7; −2.4; −3.7 to −1.1) vs 8.1 24.5–27.1 (7.9–8.3; 8.4; 5.9–10.9) and 5.9 (5.8–6.0; 5.9; 5.4–6.4), change of −2.3 (−2.4 to −2.1; −2.5; −4.2 to −0.8). TheBMI (kg/m2) Range 29.1–29.7 BMI body mass index, HbA1C glycated hemoglobin preoperatively mildly obese group experienced reduction of 30.3 FPG closely approaching the normal range, and both groups 28.8–31.8 achieved the HbA1C target range. These results are based on 44% of reviewed studies reporting FPG and HbA1C in whichTable 3 (continued) 95% CI statistically significant differences were present. Simple Mean Study N 95% CI In addition, the mean percentage of preoperatively overweight vs mildly obese T2DM patients not taking any diabetes medications at ≥6 months was 81.8% (75.9– 87.6%; 67%; 18.0–100%) vs 89.1% (86.0–92.2%; 91.6%; 76.7–100%; Fig. 2). Nearly 10% more patients in the preoperatively mildly obese group were taking no diabetes

OBES SURG 33 BMI* Overweight Group: Preop BMI 25.0–29.9 32 Obese Group: Preop BMI 30.0–35.0Fig. 1 Weighted mean change 31in body mass index (BMI), 30fasting plasma glucose (FPG), 29and glycated hemoglobin 28(HbA1C) in preoperatively 27overweight (BMI 25.0–29.9 26kg/m2) and mildly obese 25(BMI 30.0–35.0 kg/m2) T2DM 24patients 23 kg/m2 25.2 –6.8 Overweight† (BMI 25.0–29.9) Normal weight† (BMI 20.0–24.9 23.1 220 –3.4 210 200 _> 6 months 190 180 FPG* 170 mg/dL 160 150 140 107.0 –103.6 NIH normal FPG = >_99 mg/dL 130 102.3 –77.2 ADA target HbA1C = <7.0% 120 110 100 90 80 70 9 _> 6 months 8 HbA1C* 7 6.0 –2.9 5.9 –2.3 6 % 5 4 >_ 6 months * In the preoperatively 25.0–29.9 BMI group [7 studies, 183 patients] and the preoperatively 30–35 BMI group [9 studies, 160 patients], not all studies reported BMI, FPG, or HbA1C; therefore, weighted means were calculated. † National Institutes of Health (NIH) weight categories for bariatric surgery.

OBES SURGOverweight Group: Preop BMI 25–29.9 (mean 27.0) this review. A 2008 RCT by Dixon et al. [89] comparedObese Group: Preop BMI 30–35 (mean 32.1) T2DM patients with a BMI of 30 to 40 undergoing surgery or medical therapy, although their final recruitment resulted81.8% 89.1% in a mean BMI of 37.1. The goal of the current review and analysis was strictly to capture results of surgery in a subsetPostop Resolution Off All Diabetes Medications (_>6 months) of T2DM patients at, or below, a mean BMI of 35 so as to provide a more distinct, albeit smaller, study populationFig. 2 Weighted mean percentage of preoperatively overweight vs with which to gauge the state of the research.preoperatively obese T2DM patients off all diabetes medications at≥6 months The principal outcomes of the current review of metabolic surgery studies in low-BMI patients are asmedications at study end point relative to the preoperatively follows: The great majority of studies of low-BMI T2DMoverweight group. patients undergoing metabolic surgery corroborated the results of lean rodent research previously reported thatSafety demonstrated postoperative euglycemia without undesirable weight loss. BMI <35 T2DM patients had a low rate ofOverall mortality in the entire analysis was one of 343 total complications, and mortality virtually identical to the lowpatients (0.29%). In a study in the restrictive procedure mortality associated with bariatric surgery, by meta-analysiscategory, one patient died 20 months postoperatively from (0.28%) [5, 90]. Mean BMI was markedly reduced, with asepsis subsequent to perforation of a dilated upper gastric desirable level of change from the overweight category intopouch. The rate of total complications was low overall, under the normal weight category. Mean postoperative FPG4%; one study in the malabsorptive/restrictive procedure closely approached normal glucose values established bycategory reported a 10.3% of major complications. the NIH and mean HbA1C dropped to well within the ADA target range. Clinical glycemic control was evident in a With respect to patient satisfaction with the overall high proportion (85.3%) of low-BMI patients who were notoutcome of their T2DM reduction, there was no reporting taking diabetes medication after metabolic surgery. Resultsof administration of quality of life instruments. were maintained to a follow-up of 6–216 months.Discussion Findings for two subsets of patients are of interest. In the first subset, procedure categories, restrictive, malabsorptive/Integrated Review Findings restrictive, and primarily malabsorptive procedures were all effective, by varying definitions of resolution, in reducing,Clinical reports focused on low-BMI T2DM patients and, in most cases, resolving, T2DM in low-BMI patients.effectively begin with the 1998 publication of Noya et al. T2DM resolution, by all laboratory and clinical measureson BPD employed to address hypercholesterolemia and available for compilation, was greatest for patients follow-T2DM end points [79]. Subsequent studies whose inclusion ing any form of malabsorptive/restrictive procedure fol-criteria indicated a probable BMI <35, but whose final lowed by primarily malabsorptive procedures and then,study group mean BMI was >35, were excluded from this restrictive procedures. Interestingly, the primarily malab-report: In a noteworthy 2002 study by Fobi et al. [87], sorptive category of procedures, the class of operation thatpatients with T2DM and a BMI of 32 to 40 were recruited, typically produces the greatest reduction in BMI and excessyet, had a mean preoperative study BMI of 37.8. In 2006, weight in morbidly obese patients, did not effect a similarlyO’Brien et al. [88] published an important RCT of patients dramatic BMI reduction in the low-BMI patients reviewed.with a mean BMI of 33.6 undergoing LAGB or an intensive As Scopinaro et al. have posited in their low-BMI studiesmedical program that observed significant reductions in [81, 86], this suggests a homeostatic mechanism involvedmeasures of the metabolic syndrome, but did not record in BPD, and perhaps in all primarily malabsorptivemedication and HbA1C status and so were not included in procedures, that permits only a loss of weight relative to the amount of calories for which the procedure facilitates absorption. In the second subset of interest, the total low-BMI category broken into overweight (n=183, preoperative BMI 27.0) and mildly obese (n=160; BMI 32.1) groups, both groups experienced mean BMI reductions that placed them in lower weight categories, normal and overweight, respectively; yet, the preoperatively mildly obese group lost nearly twice the weight postoperatively. HbA1C change

OBES SURGand medication status were more robust as well in the BMI LADA patients would be predicated on underlyingpreoperatively heavier group, while their mean FPG mechanisms that are distinct, to some degree, from low-BMIapproached normal values. In the reviewed studies, the BMI T2DM patients, and options may not include metabolicgroup with <35, but >30 experienced clinical and laboratory surgery or specific metabolic procedural categories; RYGBresults similar to those attained in morbidly obese patients. At would likely be contraindicated, while BPD/DS or othera BMI threshold of around ≤30, an unclear mechanism, or primarily malabsorptive procedures might be appropriatecascade of interactions, appears to blunt weight loss while options. Because of such potential stumbling blocks and thefacilitating T2DM resolution in both groups, but more complex arena of neurohormonal mechanisms to be inves-profoundly in the preoperatively lower weight group. This tigated, we must proceed with care toward metabolic surgeryobservation corroborates the prior original findings of in low-BMI patients. In 1936, in “The Lancet,” HP Hims-Scopinaro et al., who, in their study (included in this review worth differentiated diabetes into 2 primary modes of action,[86]) followed 30 patients with BMI <35, separating them insulin sensitive and insensitive (T1 and T2 DM) [100].into two subgroups of overweight (n=15) and mildly obese Today, a T3 and T4 DM, or T2DM subtype A and B, may(n=5) patients undergoing BPD. The overweight group, like await discovery, and may help to define new diagnostics andthe reviewed overweight patients, lost less weight than the therapy. Going forward, patient safety must be balancedmildly obese patients, markedly reduced their FPG to a low against such promising opportunities.prediabetic level (slightly above the normal range), andachieved the target range for HbA1C. Limitations of this review include the relatively small number of low-BMI/T2DM-inclusive studies and their Scopinaro et al. propose that T2DM may be more mostly small numbers of patients. Rather than a systematicadvanced, severe, and less responsive to surgical rearrange- literature review, an extensive standard literature review,ment of the gastrointestinal tract at a BMI ≤30 [89]. Others complemented by an integration and weighted meanshave found that in the non-morbidly obese, glucose analysis of available data, was conducted, an approachtransport is negatively correlated with the degree of obesity, deemed appropriate to characterize the very few, primarilybut only in low-BMI individuals up to a BMI of observational studies published to date. This methodapproximately 30 [91, 92]. The somewhat suppressed represents our bias that the highly formulaic design ofT2DM response to surgery in the preoperatively overweight systematic review is more suitable to sorting a large body ofpatients may also be influenced by genetic or developmen- research in a well-developed field (e.g., bariatric surgery intal defects of beta-cell function or insulin action, diseases of morbidly obese patients). The published studies on surgerythe endocrine or exocrine pancreas, infection, chemicals, for T2DM at this time afford only short- to medium-termimmune disorders, or gestational diabetes. As has been data without the robust evidence provided by well-shown in morbidly obese patients undergoing bariatric conducted RCTs. Thus, the current “footprint” articlesurgery, presurgery duration of T2DM may also critically presents a snapshot of a discipline on the verge ofinfluence whether a low-BMI patient can achieve complete expansion. Numerous studies, providing high-quality studyremission [93, 94]. designs, including RCTs, will be forthcoming; a systematic review and meta-analysis eventually will be warranted. Herron and Tong [95], Dixon [96], and Cummings et al.[97] have carefully summarized the prevailing theories of Another study limitation—varied and incomplete datamechanism involved in surgical resolution of T2DM in reporting among the identified reports with respect tomorbidly obese patients. These incorporate, primarily, the measures of T2DM—is, in fact, a finding of this review:action of weight loss and decreased calories, the role of The treatment effect of metabolic surgery cannot be gaugedhormones, intestinal malabsorption, and the role of gastro- without “best evidence” containing clear T2DM resolutionintestinal rearrangement (i.e., the upper and lower intestinal end points consistently reported across studies in thishypotheses [97]). While detailed discussion of mechanism emerging discipline.is beyond the scope of this review, it can be noted that themechanisms of surgical T2DM resolution in morbidly Future Studyobese and low-BMI patients may, or may not, prove tofunction on a continuum; all of the theories require long- Metabolic surgery research is producing new evidence andterm testing. In addition, what appears to be T2DM in some procedures that will require adjustment of the guidelines forlow-BMI patients under consideration for metabolic surgery patient selection. Most studies in this review were notmay be late-onset type 1 diabetes (latent autoimmune diabetes American in origin, likely due to the restriction of bariatricin the adult, LADA), which comprises 10% of diabetics age surgery insurance coverage in the US to patients with ≥3530–55 and is more prevalent in low-BMI individuals [98, 99]. BMI. Insufficient data on surgery for T2DM in categoriesIdentification of LADA should be part of the standard of BMI <35 preclude consideration of US reimbursement attreatment evaluation of low-BMI patients. Therapy for low- this time; 1991 National Institutes of Health Consensus

OBES SURGDevelopment Panel guidelines govern [101]. Similarly, in amount of their preoperative weight, resulting in a BMIEuropean and other member countries of IFSO, published within the upper end of the normal weight category.guidelines have been in place since 1997 that restrict bariatric Approximately 85% were resolved off T2DM medicationssurgery to individuals with BMI ≥35 with comorbidities with slightly above normal FPG at ≥6 months and HbA1C[102]. Nine of the 11 prospective studies of surgery for markedly diminished to 6%. Complications were relativelyT2DM patients with BMI <35 summarized in the current few with a very low operative mortality. The trend toward areview specified that they sought IRB approval, and in some lesser percentage reduction in clinical and laboratorycases, governmental health body and insurance approval, to measures of T2DM in the lowest subset of patients studiedpursue this novel indication. Recently, Misra and the Asian (BMI 25.0–29.9) relative to those with BMI 30–35 is aIndian Consensus Group published a statement [103] recom- finding of interest requiring future study. This representsmending that the BMI guidelines for bariatric surgery be the best available evidence to date of the efficacy and safetylowered to >32.5 with comorbidity, >37.5 without comorbid- of metabolic surgery for resolution of T2DM in the non-ity, with a greater emphasis in patient selection on waist-to-hip morbidly obese, at an early stage in the development of theratio due to the genetic differences that cause Asian Indians to new therapy. The literature provides preliminary evidencedevelop comorbidities of obesity at a lower BMI than white to support continued investigation of surgery as a possibleCaucasians [104, 105]. Selection criteria for surgery in non- means of achieving lasting resolution of T2DM in a vastmorbidly obese T2DM patients may need to discriminate for non-morbidly obese population.body habitus, metabolic and neurogenic variables, andduration, severity, and possibly subtype of T2DM. Acknowledgments The authors would like to thank T. McGlennon of M3, LLC, WI, USA for statistical consultation during manuscript Future trials should be designed to provide the highest development.classes of evidence feasible; refinement and consistent useof standardized end points is needed to provide studies of Disclosures This work was supported by Ethicon Endo-Surgeryvalue. Trials of novel surgical procedures should be (Europe) GmbH.approached with caution; the mode of action of establishedprocedures that have proven to be safe and effective must Referencescontinue to be investigated, and our understanding broad-ened, prior to, or concomitant with, attempting new 1. North American Association for the Study of Obesity (NAASO) andconstructions in humans. Collaborative medical–surgical the National Heart, Lung, and Blood Institute (NHLBI). The Practicalinvestigation of T2DM mechanism, particularly in incre- Guide: Identification, Evaluation, and Treatment of Overweight andmentally distinct BMI ranges, is likely to provide insights Obesity in Adults. NIH Publication #00-4084, October 2000.into the management and resolution of T2DM. 2. Adams TD, Gress RE, Smith SC, et al. Long-term mortality after Differences in outcome of T2DM in low-BMI patients gastric bypass surgery. N Engl J Med. 2007;357:753–61.treated with surgery or standard medical therapy should beexplored. Two trials of this design are under way: The 3. Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a“Surgical Therapy And Medications Potentially Eradicate systematic review and meta-analysis. JAMA. 2004;292:1724–37.Diabetes Efficiently (STAMPEDE),” study in progress atthe Cleveland Clinic Bariatric and Metabolic Institute and 4. Christou NV, Sampalis JS, Liberman M, et al. Surgery decreasesthe Department of Endocrinology, Diabetes, and Metabo- long-term mortality, morbidity, and health care use in morbidlylism, is a randomized trial that will compare the effective- obese patients. Ann Surg. 2004;240:416–23.ness of advanced medical therapy alone vs metabolicsurgery and therapy combined in mildly obese patients 5. Buchwald H, Estok R, Fahrbach K, et al. Weight and type 2over a 5-year study period. The second is a collaboration diabetes after bariatric surgery: systematic review and meta-between Columbia University, the University of Minnesota, analysis. Am J Med. 2009;122(3):248–56.e5. Review.and National Taiwan University, in which 150 patients withT2DM and a BMI of 30–35 will be randomized to either 6. Pontiroli AE, Folli F, Paganelli M, et al. Laparoscopic gastricRYGB or standard medical care. banding prevents type 2 diabetes and arterial hypertension and induces their remission in morbid obesity: a 4-year case-Conclusions controlled study. Diabetes Care. 2005;28(11):2703–9.In this review, the 343 low-BMI T2DM surgery patients 7. Von Mering JV, Minkowski O. Diabetes nach pankreas extirpa-integrated for summary lost a meaningful, not excessive, tion. Arch Exp Path Pharmakol. 1889;26:371. 8. Barron M. The relation of the islets of langerhans to diabetes with special reference to cases of pancreatic lithiasis. Surg, Gynecol and Obstetrics. 1920;31(5):437–48. 9. Banting FG, Best CH, Collip JB, et al. Pancreatic extracts in the treatment of diabetes mellitus: preliminary report. CMAJ. 1922;12(3):141–6. 10. Friedman MN, Sancetta AJ, Magovern GJ. The amelioration of diabetes mellitus following subtotal gastrectomy. Surg Gynecol Obstet. 1955;100(2):201–4. 11. Bosello O, Armellini F, Pelloso M, et al. Glucose tolerance in jejunoileal bypass for morbid obesity: a fifteen month follow-up. Diabetes Metab. 1978;4(3):159–62.

OBES SURG 35. Hickey MS, Pories WJ, MacDonald Jr KG, et al. A new paradigm for type 2 diabetes mellitus: could it be a disease of the 12. Ackerman NB. Observations on the improvements in carbohydrate foregut? Ann Surg. 1998;227:637–43. metabolism in diabetic and other morbidly obese patients after jejunoileal bypass. Surg Gynecol Obstet. 1981;152(5):581–6. 36. Gumbs A, Modlin IM, Ballantyne GH. Changes in insulin resistance following bariatric surgery: role of caloric restriction 13. Halverson JD, Kramer J, Cave A, et al. Altered glucose toler- and weight loss. Obes Surg. 2005;15:462–73. ance, insulin response, and insulin sensitivity after massive weight reduction subsequent to gastric bypass. Surgery. 1982;92 37. Brolin RE. Update: NIH consensus conference. Gastrointestinal (2):235–40. surgery for severe obesity. Nutrition. 1996;12:403–4. 14. Herbst CA, Hughes TA, Gwynne JT, et al. Gastric bariatric 38. Rubino F, Gagner M. Potential of surgery for curing type 2 operation in insulin-treated adults. Surgery. 1984;95:209–13. diabetes mellitus. Ann Surg. 2002;236:554–9. 15. Schrumpf E, Bergan A, Djoseland O, et al. The effect of gastric 39. Levy P, Fried M, Santini F, et al. The comparative effects of bypass operation on glucose tolerance in obesity. Scand J bariatric surgery on weight and type 2 diabetes. Obes Surg. Gastroenterol. 1985;20 Suppl 107:24–31. 2007;17(9):1248–56. Review. 16. Scopinaro N, Gianetta E, Civalleri D, et al. Biliopancreatic 40. Schauer PR, Ikramuddin S, Gourash W, et al. Outcomes after bypass for obesity: I. An experimental study in dogs. Br J Surg. laparoscopic roux-en-Y gastric bypass for morbid obesity. Ann 1979;66:613–7. Surg. 2000;232:515–29. 17. Scopinaro N, Gianetta E, Civalleri D, et al. Biliopancreatic 41. Scopinaro N, Gianetta E, Adami GF, et al. Biliopancreatic diversion bypass for obesity: II. Initial experience in man. Br J Surg. for obesity at eighteen years. Surgery. 1996;119(3):261–8. 1979;66:618–20. 42. Buchwald H, Buchwald JN. Evolution of surgery for morbid 18. Scopinaro N, Gianetta E, Friedman D, et al. Evolution of obesity. In: Pitombo C, Jones KB, Higa KD, Pareja JC, editors. biliopancreatic bypass. Clin Nutr. 1986;5(Suppl):137–46. Obesity Surgery: Principles and Practice. New York: McGraw- Hill Medical; 2007. p. 3–14. 19. Pories WJ, Caro JF, Flickinger EG, et al. The control of diabetes mellitus (NIDDM) in the morbidly obese with the Greenville 43. Pories WJ. Why does the gastric bypass control type 2 diabetes Gastric Bypass. Ann Surg. 1987;206(3):316–23. mellitus? Obes Surg. 1992;2:303–13. 20. Jimenez J, Zuniga-Guajardo S, Zinman B, et al. Effects of weight 44. Cowan GS, Buffington CK. Significant changes in blood pres- loss in massive obesity on insulin and C-peptide dynamics: sure, glucose, and lipids with gastric bypass surgery. World J sequential changes in insulin production, clearance, and sensi- Surg. 1998;22:987–92. tivity. J Clin Endocrinol Metab. 1987;64(4):661–8. 45. Segal JB, Clark JM, Shore AD. Prompt reduction in use of 21. Pories WJ, MacDonald Jr KG, Flickinger EG, et al. Is type II medications for comorbid conditions after bariatric surgery. Obes diabetes mellitus (NIDDM) a surgical disease? Ann Surg. 1992;215 Surg. 2009;19:1646–56. (6):633–42. discussion 643. 46. Cummings DE, Overduin J, Foster-Schubert KE. Gastric bypass 22. Pories WJ, Swanson MS, MacDonald KG, et al. Who would have for obesity: mechanisms of weight loss and diabetes resolution. J thought it? An operation proves to be the most effective therapy for Clin Endocrinol Metab. 2004;89:2608–15. adult-onset diabetes mellitus. Ann Surg. 1995;222:339–52. 47. Ferrannini E, Mingrone G. Impact of different bariatric surgical 23. Maggard M, Shugarman LR, Suttorp M, et al. Meta-analysis: sur- procedures on insulin action and beta-cell function in type 2 gical treatment of obesity. Ann Intern Med. 2005;142(7):547–59. diabetes. Diabetes Care. 2009;32:514–20. 24. Sjöström CD, Lissner L, Wedel H, et al. Reduction in incidence 48. Vencio S, De Paula A, Macedo A, et al. Effect of laparoscopic of diabetes, hypertension and lipid disturbances after intentional ileal interposition on beta cell function and insulin sensitivity in weight loss induced by bariatric surgery: the SOS Intervention nonobese patients with type 2 diabetes mellitus. Diabetologia. Study. Obes Res. 1999;7(5):477–84. 2009;52 Suppl 1:S44. 25. Sugerman HJ, Wolfe LG, Sica DA, et al. Diabetes and hypertension 49. Mingrone G. Role of the incretin system in the remission of type in severe obesity and effects of gastric bypass-induced weight loss. 2 diabetes following bariatric surgery. Nutr Metab Cardiovasc Ann Surg. 2003;237(6):751–6. discussion 757–8. Dis. 2008;18:574–9. 26. Dixon JB, Dixon ME, O’Brien PE. Quality of life after Lap- 50. Polyzogopoulou EV, Kalfarentzos F, Vagenakis AG, et al. Band placement: influence of time, weight loss, and comorbid- Restoration of euglycemia and normal acute insulin response to ities. Obes Res. 2001;9(11):713–21. glucose in obese subjects with type 2 diabetes following bariatric surgery. Diabetes. 2003;52:1098–103. 27. Sjöström L, Narbro K, Sjöström CD. Effects of bariatric surgery on mortality in Swedish obese subjects. NEJM. 2007;357(8):741–52. 51. Laferrère B, Teixeira J, McGinty J, et al. Effect of weight loss by gastric bypass surgery versus hypocaloric diet on glucose and 28. Flum DR, Dellinger EP. Impact of gastric bypass operation incretin levels in patients with type 2 diabetes. J Clin Endocrinol on survival: a population-based analysis. J Am Coll Surg. Metab. 2008;93:2479–85. 2004;199:543–51. 52. Salinari S, Bertuzzi A, Iaconelli A, et al. Twenty-four hour 29. Lara MD, Kothari SN, Sugerman HJ. Surgical management of insulin secretion and beta cell NEFA oxidation in type 2 diabetic, obesity: a review of the evidence relating to the health benefits morbidly obese patients before and after bariatric surgery. and risks. Treat Endocrinol. 2005;4(1):55–64. Review. Diabetologia. 2008;51:1276–84. 30. Buchwald H, Estok R, Fahrbach K, et al. Trends in mortality in 53. le Roux CW, Aylwin SJ, Batterham RL, et al. Gut hormone bariatric surgery: a systematic review and meta-analysis. Sur- profiles following bariatric surgery favor an anorectic state, gery. 2007;142:621–35. facilitate weight loss, and improve metabolic parameters. Ann Surg. 2006;243(1):108–14. 31. Sampalis JS, Liberman M, Auger S, et al. The impact of weight reduction surgery on health-care costs in morbidly obese 54. Buchwald H, Varco RL. Metabolic surgery. New York, NY: patients. Obesity Surg. 2004;14(7):939–47. Grune & Stratton; 1978. 32. Finkelstein EA, Brown DS. Return on investment for bariatric 55. Pories W. So you think we are bariatric surgeons? Think again. surgery. Am J Managed Care. 2008;14(9):561–2. Obes Surg. 2003;13(5):673–5. 33. Cremieux PY, Buchwald H, Shikora SA, et al. A study on the 56. Buchwald H. Lowering of cholesterol absorption and blood levels economic impact of bariatric surgery. Am J Manag Care. by ileal exclusion: experimental basis and preliminary clinical re- 2008;14(9):589–96. port. Circulation. 1964;29:713–20. 34. Scopinaro N, Adami GF, Marinari GM, et al. Biliopancreatic diversion. World J Surg. 1998;22:936–46.

57. Koopmans HS, Sclafani A. Control of body weight by lower gut OBES SURG signals. Int J Obes. 1981;5:491–5. 78. Shah S, Todkar JS, Shah PS, et al. Diabetes remission and reduced58. Mistry SB, Omana JJ, Kini S. Rat models for bariatric surgery and cardiovascular risk after gastric bypass in Asian Indians with body surgery for type 2 diabetes mellitus. Obes Surg. 2009;19:655–60. mass index <35 kg/m2. Surg Obes Rel Dis. 2009. doi:10.1016/j. Review. soard.2009.08.009.59. Rubino F, Marescaux J. Effect of duodenal-jejunal exclusion in a 79. Noya G, Cossu ML, Coppola M, et al. Biliopancreatic diversion non-obese animal model of type 2 diabetes: a new perspective preserving the stomach and pylorus in the treatment of for an old disease. Ann Surg. 2004;239(1):1–11. hypercholesterolemia and diabetes type II: results in the first 10 cases. Obes Surg. 1998;8:67–72.60. Rubino F, Zizzari P, Tomasetto C, et al. The role of the small bowel in the regulation of circulating ghrelin levels and food intake in the 80. Cohen RV, Schiavon CA, Pinheiro JS, et al. Duodenal-jejunal obese Zucker rat. Endocrinology. 2005;146(4):1745–51. bypass for the treatment of type 2 diabetes in patients with BMI 22– 34: a report of two cases. Surg Obes Relat Dis. 2007;3(2):195–7.61. Rubino F, Forgione A, Cummings DE, et al. The mechanism of diabetes control after gastrointestinal bypass surgery reveals a 81. Scopinaro N, Papadia F, Marinari G, et al. Long-term control of role of the proximal small intestine in the pathophysiology of type 2 diabetes mellitus and the other major components of the type 2 diabetes. Ann Surg. 2006;244(5):741–9. metabolic syndrome after biliopancreatic diversion in patients with BMI <35 kg/m2. Obes Surg. 2007;17:185–92.62. Patriti A, Facchiano E, Annetti C, et al. Early improvement of glucose tolerance after ileal transpositition in a non-obese type 2 82. Ramos AC, Galvao Neto MP, de Souza YM, et al. Laparoscopic diabetes rat model. Obes Surg. 2005;15:1258–64. duodenaljejunal exclusion in the treatment of type 2 diabetes mellitus in patients with BMI <30 kg/m2. Obes Surg. 2009;19:307–12.63. Patriti A, Aisa MC, Annetti C, et al. How the hindgut can cure type 2 diabetes. Ileal transposition improves glucose metabolism 83. Geloneze B, Geloneze S, Fiori C, et al. Surgery for nonobese and beta-cell function in Goto-Kakazaki rats through an type 2 diabetic patients: an interventional study with duodenal– enhanced proglucagon gene expression and L-cell number. jejunal exclusion. Obes Surg. 2009;19:1077–83. Surgery. 2007;142:74–85. 84. Ferzli GS, Dominique E, Ciaglia M, et al. Clinical improvement64. Pacheco D, de Luis DA, Romero A, et al. The effects of duodenal- after duodenojejunal bypass for nonobese type 2 diabetes despite jejunal exclusion on hormonal regulation of glucose metabolism in minimal improvement in glycemic homeostasis. World J Surg. Goto-Kakizaki rats. Am J Surg. 2007;194(2):221–4. 2009;33:972–9.65. Wang TT, Hu SY, Gao HD, et al. Ileal transposition controls 85. Chiellini C, Rubino F, Castagneto M, et al. The effect of bilio- diabetes as well as modified duodenal jejunal bypass with better pancreatic diversion on type 2 diabetes in patients with BMI lipid lowering in a nonobese rat model of type II diabetes by <35 kg/m2. Diabetologia. 2000;52:1027–30. increasing GLP-1. Ann Surg. 2008;247(6):968–75. 86. Scopinaro N, Adami GF, Papadia FS, et al. Effects of bilio-66. Strader A, Clausen TR, Goodin SZ, et al. Ileal interposition pancreatic diversion on type 2 diabetes in patients with BMI 25 improves glucose tolerance in low dose streptozotocin-treated to 35 kg/m2. J Clin Endocrinol Metab 2009; submitted. diabetic and euglycemic rats. Obes Surg. 2009;19:96–104. 87. Fobi M, Lee H, Igwe D, et al. Gastric bypass in patients with67. National Diabetes Information Clearinghouse. Diagnosis of Diabe- BMI <40 but >32 without life-threatening co-morbidities: tes. USDHHS. NIH Publication No. 09–4642; October 2008. preliminary report. Obes Surg. 2002;12:52–6.68. American Diabetes Association. Diagnosis and classification of 88. O’Brien PE, Dixon JB, Laurie C, et al. Treatment of mild to mod- diabetes mellitus. Diabetes Care. 2009;32(1):S62–6. erate obesity with laparoscopic adjustable gastric banding or an intensive medical program. Ann Intern Med. 2006;144(9):625–33.69. HOMA Calculator, v2.2.2, released 12 Dec 2007. Diabetes Trials Unit. The Oxford Centre for Diabetes, Endocrinology and Metab- 89. Dixon JB, O'Brien PE, Playfair J, et al. Adjustable gastric olism. Available at: http://www.dtu.ox.ac.uk/index.php?maindoc=/ banding and conventional therapy for type 2 diabetes: a homa/index.php. randomized controlled trial. JAMA. 2008;299(3):316–23.70. Rowlett R and the University of North Carolina at Chapel Hill. 90. Dimick JB, Welch HG, Birkmeyer JD. Surgical mortality as an Dictionary of Units of Measure. SI units for clinical data. Available indicator of hospital quality. JAMA. 2004;292:847–51. at: http://www.unc.edu/∼rowlett/units/scales/clinical_data.html. 91. Kolterman OG, Gray RS, Griffin J, et al. Receptor and post-71. Angrisani L, Favretti F, Furbetta F, et al. Italian Group for Lap- receptor defect tribute to the insulin resistance in non-insulin Band System®: results of multicenter study on patients with BMI dependent diabetes mellitus. J Clin Invest. 1981;68:957–69. ≤35 kg/m2. Obes Surg. 2004;14:415–8. 92. Elton CW, Tapscott EB, Pories WJ, et al. Effect of moderate72. Parikh M, Duncombe J, Fielding GA. Laparoscopic adjustable obesity on glucose transport in human muscle. Horm Metab Res. gastric banding for patients with body mass index of 35 kg/m2. 1994;26:181–3. Surg Obes Rel Dis. 2006;2:518–22. 93. Schauer PR, Burguera B, Ikramuddin S, et al. Effect of73. Sultan S, Parikh M, Youn H, et al. Early U.S. outcomes after laparoscopic Roux-en-Y gastric bypass on type 2 diabetes laparoscopic adjustable gastric banding in patients with a body mellitus. Ann Surg. 2003;238(4):467–84. mass index less than 35 kg/m2. Surg Endosc. 2009;23:1569–73. 94. Ponce J, Haynes B, Paynter S, et al. Effect of Lap-Band-induced74. Cohen R, Pinheiro JS, Correa JL, et al. Laparoscopic Roux-en-Y weight loss on type 2 diabetes mellitus and hypertension. Obes gastric bypass for BMI <35 kg/m2: a tailored approach. Surg Surg. 2004;14(10):1335–42. Obes Relat Dis. 2006;2:401–4. 95. Herron D, Tong W. Role of surgery in management of type II75. Lee WJ, Wang W, Lee YC, et al. Effect of laparoscopic mini- diabetes. Mount Sinai Med J. 2009;76:281–93. gastric bypass for type 2 diabetes mellitus: comparison of BMI >35 and <35 kg/m2. J Gastrointest Surg. 2008;12:945–52. 96. Dixon J. Obesity and diabetes: the impact of bariatric surgery on type-2 diabetes. World J Surg. 2009;33(10):2014–21.76. DePaula AL, Macedo ALV, Mota BR, et al. Laparoscopic ileal interposition associated to a diverted sleeve gastrectomy is an 97. Cummings D, Overduin J, Foster-Schubert KE, et al. Role of the effective operation for the treatment of type 2 diabetes mellitus bypassed proximal intestine in the anti-diabetic effects of patients with BMI 21–29. Surg Endosc. 2009;23:1313–20. bariatric surgery. Surg Obes Relat Dis. 2007;3(2):109–15.77. DePaula AL, Macedo ALV, Schraibman V, et al. Hormonal 98. Zimmet P, Turner R, McCarty D, et al. Crucial points at evaluation following laparoscopic treatment of type 2 diabetes diagnosis. Type 2 diabetes or slow type 1 diabetes. Diabetes mellitus patients with BMI 20–34. Surg Endosc. 2009;23:1724–32. Care. 1999;22 Suppl 2:B59–64. 99. Deitel M. Surgery for diabetes at lower BMI: some caution. Obes Surg. 2008;18:1211–4.