Review of Metabolic Surgery for Type 2 Diabetes in Patients with a BMI < 35 kg/m2

Bariatric/metabolic surgery is considered an accepted treatment option for type 2 diabetes mellitus (T2DM) with body mass index (BMI)  ≧  35 kg/m2. Mounting evidence also shows that metabolic surgery is effective for T2DM with BMI  <  35 kg/m2. To evaluate current status of metabolic surgery, we reviewed the available clinical studies which described surgical treatment for T2DM with mean BMI  <  35 kg/m2. 18 studies with 477 patients were identified. 30% of the patients was insulin users. The follow-up period ranged from 6 to 216 months. The weight loss effect was reasonable, not excessive. Mean BMI decreased from 30.4 to 24.8 kg/m2. Remission of T2DM was achieved in 64.7% of the patients with fasting plasma glucose and glycated hemoglobin approaching slightly above normal range. Clinical T2DM status was an important factor when selecting the eligible candidates for metabolic surgery. Postoperative complication rate of 10.3% with mortality of 0% in the studies has been acceptable. Even though it would be premature at this point to state that metabolic surgery is an accepted treatment option for T2DM with BMI < 35 kg/m2, it is clear that a high proportion of T2DM patients will derive substantial benefit from metabolic surgery.


Introduction
Diabetes mellitus represents an expanding pandemic that contributes markedly to worldwide morbidity and mortality. The world prevalence of diabetes among adults (aged 20-79 years) will be 6.4%, affecting 285 million adults, in 2010 and will increase to 7.7% and 439 million adults by 2030 [1]. There is a strong relationship between obesity and type 2 diabetes mellitus (T2DM) [2]. In a large USA population study, the prevalence of diabetes increases with increasing weight classes according to body mass index (BMI). Approximately half of those diagnosed with T2DM are obese [3]. Weight control is the key to successful T2DM management. Weight loss achieved by lifestyle interventions has been shown to be effective in preventing and treating T2DM [4][5][6][7]. However, conventional treatment, such as, lifestyle modification and pharmacotherapy has produced small improvements in weight [7][8][9][10]. By contrast, bariatric surgery has been shown to effectively provide durable weight loss [11].
Currently, bariatric surgery is now considered appropriate for T2DM patient with BMI 35 kg/m 2 . Bariatric surgery leads to remission of T2DM in the majority of patients and improvement in the rest [12]. The American Society of Bariatric and Metabolic surgery estimates that the number of bariatric procedures increased from about 16,000 in the early 1990s to more than 103,000 in 2003 and 220,000 people in the United States had bariatric surgery in 2008. Growing evidence from clinical and animal studies indicates that bariatric/metabolic surgery may be beneficial for T2DM in nonseverely obese or even nonobese patients (BMI < 35 kg/m 2 ) [13,14]. Recently, International Diabetes Federation has released its position statement [15]: surgery should be an accepted option in people who have T2DM and BMI of 35 more. Surgery should be considered as an alternative treatment option in persons with BMI 30 to 35 when diabetes cannot be adequately controlled by optimal medical regimen, especially in the presence of other major cardiovascular disease risk factors. The surgical approach is 2 Journal of Obesity now being extended to overweight and mild to moderate obese (BMI < 35 kg/m 2 ) patients with T2DM.
Herein, to evaluate current evidence of metabolic surgery for treatment of T2DM in patients with a BMI < 35 kg/m 2 , we conducted a review to date of available clinical studies.

Search Strategy and Study Inclusion
Criteria. We conducted a comprehensive review of all studies published containing data on weight loss and T2DM-related outcomes of patients treated with any form of bariatric/metabolic surgery where the mean study BMI was <35 kg/m 2 . Only reports published in English were included for review. Studies whose inclusion criteria indicated bariatric or metabolic surgery for low BMI patients were excluded if their mean BMI was 35 kg/m 2 , all participants were not T2DM patients, diabetic participants had gastric surgery with anatomical similarities to RYGB because of gastric cancer and ulcer, or they did not report diabetes-related outcomes such as fasting plasma glucose (FPG), glycated hemoglobin (HbA1c), and postoperative clinical status. Pubmed was searched from January 1, 1980, to November 1, 2011, for citations using the following keywords: "metabolic surgery", "bariatric surgery", "diabetes surgery", "T2DM", "type 2 diabetes", "obesity", "BMI < 35 kg/m 2 ", "mild to moderate obesity", and "low-BMI". To supplement the electronic search, manual reference checks were performed in the identified studies.

Data
Reporting. Study authors, country, year of publication, surgical procedure, and study design were summarized. Characteristics of the study groups, BMI, FPG, and HbA1c were recorded. Diabetes-related clinical outcomes were collected as % meds resolved (the percentage of the patients who discontinued antidiabetic medication postoperatively) and remission rate (the percentage of the patients who had remission or resolution of T2DM according to the varied definition in the studies included). These two parameters were calculated using available values if not specifically stated. Then, the studies were divided into 2 groups according to mean duration of T2DM prior to surgery. The percentage of insulin using patients prior to surgery and T2DM remission rate was compared. Regarding safety of metabolic surgery, all major and minor complications were counted because they were variably reported. Mortality was also checked.

Statistical Analysis.
Weighted means for ages and duration of T2DM were calculated. Pre-and Postoperative mean values (and corresponding 95% confidence intervals) for BMI, FPG, and HbA1c were summarized using a random effects model to account for the variability between the different studies. Only studies which provided both a pre-and post-measurement were included in the mean estimation for each of the parameters considered.
Comparisons of 2 groups with a short (<8 years) and long history (>8 years) of diabetes with respect to insulin using patients and diabetes remission were performed by the chisquare test. The statistics was performed using the software package PASW, version 18.0, for Windows (SAS Institute, Cary, NC, USA).

Patient Characteristics.
Of the 18 studies, total 477 patients underwent metabolic/bariatric surgery. 16 studies reported the patient gender, and 53% of the total study population was female. The mean age of the patients ranged from 34 to 56 and its weighted mean was 47. The followup period ranged from 6 months to 18 years, and its weighted mean was 22 months. 2 studies reported the results of a longer than 5-year followup. The duration of diabetes prior to surgery ranged from 6 months to 28 years, and its weighted mean was 8.2 years. Majority of patients were taking oral antidiabetic medication and/or insulin. The percentage of Insulin using patients was 30.1% in 16 studies. The data are summarized in Table 1.

BMI.
14 studies were included in the mean estimation. The BMI decreased from 30.4 (95% CI 28.4-32.3) to 24.8 (95% CI 24.1-25.5) kg/m 2 . There were only two studies reporting that one of the total 15 patients was in the mildly undernourished range (BMI 17-18.5 kg/m 2 ) after RYGB without any evidence of malnutrition [21] and 12 patients (17.4%) after II-DSG were underweight (BMI < 20 kg/m 2 ) without lowering serum albumin value [24]. Overall, the risk of excessive weight loss after metabolic surgery was 2.7% (13 patients). The data are summarized in Tables 2 and 3.

T2DM
Outcomes. 12 studies were included in the mean estimation of FPG. It decreased from 203.5 (95% CI 187.4-219.6) to 112.5 (95% CI 103.9-121.1) mg/dL. 10 studies were included in the mean estimation of HbA1c. It decreased from 9.0 (95% CI 8.6-9.5) to 6.3 (95% CI 6.1-6.6) %. Regarding clinical outcomes of diabetes, 86.8% of the patients stopped taking antidiabetic medication after surgery (% meds off). The definition of resolution or remission of T2DM varied. When it is defined as FPG < 126 mg/dL and/or HbA1c < 6.5% without the use of antidiabetic medication at the time of evaluation, 64.7% of the patients met the criteria. The data are summarized in Tables 2 and 3.

Discussion
The concept of metabolic surgery was defined by Buchwald and Varco in 1978 in their book "Metabolic Surgery as the operative manipulation of a normal organ or organ system to achieve a biological result for a potential health gain" [34]. Now, metabolic surgery is defined as any modification of the gastrointestinal (GI) tract, where rerouting the food passage seems to improve T2DM, based on mechanisms that are weight loss independent. This new frontier of bariatric/metabolic surgery includes the application of conventional bariatric procedures (RYGB, BPD, SG, MGB) and the introduction of new procedures (DJB, II-SG, II-DSG, BPD-SPP) designed with the specific aim of having metabolic effects irrespective of causing massive weight loss. There is strong evidence that bariatric surgery for severely obese patients (BMI 35 kg/m 2 ) provides exceptional sustained weight loss and 50-85% remission of T2DM [35]. In view of growing enthusiasm for surgical interventions to treat T2DM, the 1st diabetes surgery summit was held in Rome in March 2007 to develop guidelines for the use of GI surgery to treat T2DM. The recommendations were made by a multidisciplinary group of 50 voting delegates [36]. Accordingly, the "Standards of Medical Care in Diabetes" published yearly by the American Diabetes Association, for the first time, mentions surgical therapy in 2009 [37]. Recently, International Diabetes Federation has released its position statement [20]. These statements     have mentioned that bariatric surgery for T2DM patients with a BMI 35 kg/m 2 is considered an accepted option as with standard medical therapy and metabolic surgery might, moreover, be considered a reasonable therapeutic alternative for low BMI (<35 kg/m 2 ) patients with T2DM who do not respond to standard medical therapy. The aim of this paper was to explore the current evidence with a view to evaluate the potential of metabolic surgery for T2DM with a BMI < 35 kg/m 2 . Metabolic surgery to treat T2DM in patients with low BMI provided desirable results regarding weight loss. The estimated mean BMI categorized as class I obesity prior to surgery reached normal weight range after surgery. Importantly, only 13 patients (2.7%) following RYGB or II-DSG in 2 studies reported excessive weight loss, and they did not show any evidence of malnutrition [21,24]. Even the procedures that typically produce the greatest reduction in BMI and excess weight in morbidly obese patients did not affect a similarly dramatic BMI reduction in the low-BMI patients [38]. Scopinaro and so forth reported that BPD does not entail risk of excessive or undue weight loss because there is a maximum energy absorption capacity after the operation, which corresponds to a weight of stabilization of low BMI patients [39]. The similar homeostatic mechanism may explain weight stabilization without causing undesirable weight loss after surgical procedures including intestinal bypass.
In this paper, diabetic status was significantly improved after metabolic surgery in the majority of studies. Discontinuation of antidiabetic medication and remission of T2DM after metabolic surgery were achieved in 86.8% and 64.7% of the patients with FPG and HbA1c approaching slightly above normal range. Moreover, metabolic surgery provided adequate glycemic control for 30.1% of the patients using insulin prior to surgery. It has been described that malabsorptive bariatric procedures have higher diabetes remission rates than restrictive ones [12,40]. T2DM typically resolves within a few days to weeks following malabsorptive procedures such as RYGB and BPD before significant weight loss is achieved. Although the exact mechanism is not yet fully understood, growing evidence shows that malabsorptive procedures involving rerouting of food might improve T2DM by enhancing insulin sensitivity and/or by improving β-cell function that is additive to weight loss and reduced caloric intake [16,41,42]. The recent studies have described that acute insulin response to intravenous glucose and early phase insulin response to oral glucose load improved significantly within a month following GI bypass surgery [16,20]. The mechanism for these changes could be due to a dramatic decrease of insulin resistance and an increase in postprandial plasma levels of glucagon-like peptide-1 (GLP-1) early after surgery. Currently, two hypotheses (hindgut and foregut theory) have been proposed to explain T2DM remission after metabolic surgery in addition to decreased calorie intake after surgery and surgical-induced weight loss which might contribute to improving insulin sensitivity. The former states that surgical rerouting of nutrients to the distal part of the small intestine results in increased secretion and concomitant glucose-lowering effects of GLP-1, and the latter emphasizes that surgical bypass of the foregut prevents the release of a hitherto unidentified nutrientinduced diabetogenic signal in susceptible individuals [43].
The novel surgical procedures such as DJB, BPD-SPP, II-SG, and II-DG were designed to apply hindgut or/and foregut hypotheses without massive weight loss and achieved 56% of T2DM remission and 84% of diabetes meds off in this paper.
The weight loss effect of metabolic surgery on T2DM in low BMI patients might be lower than that of bariatric surgery on T2DM in high BMI patients. Understanding and enhancing the abovementioned mechanism are the key to success in metabolic surgery. There is no strong evidence describing the durability of metabolic surgery in long-term followup. In this paper, 2 studies showed durable diabetes remission of T2DM during 5-18 years period after MGB and BPD [29,30]. By contrast, the recent studies of bariatric surgery for T2DM patients with Journal of Obesity 7 severe obesity showed that 24%-43% of the patients with initial remission or improvement of their T2DM subsequently developed T2DM recurrence or worsening during the midto long-term followup period [44,45]. A low preoperative BMI and severe T2DM status were associated with failure of consistent durable remission of diabetes. The common causes for failure of diabetes remission after bariatric surgery are known as inadequate weight loss or regain of weight, longstanding poorly controlled or aggressive T2DM, lower preoperative BMI, and latent autoimmune diabetes in adults (LADA) [46]. LADA comprises 10% of diabetic age 30-55 and is more prevalent in low BMI individuals [47,48].
Most clinical guidelines and statements have followed the BMI-based criteria established by 1991 National Institutes of Health Consensus Conference Guidelines [49]. Although BMI is convenient to classify the grade of obesity, it does not seem to be appropriate in selecting the suitable T2DM candidates for metabolic surgery. For instance, the risk of diabetes and/or metabolic syndrome is determined by ethnicity, waist circumference, fat distribution, body composition, and intrahepatic fat [50,51]. South Asian and Chinese individuals have distributions of elevated glucose and lipid levels similar to Europeans at significantly lower BMI values [52]. The natural history of type 2 diabetes is also important to consider in determining the timing of intervention. As the diabetes state progresses, there is continued beta-cell deterioration together with a decline in insulin secretion within 6-10 years of T2DM diagnosis [53,54]. Schauer et al. showed that a shorter history of diabetes and milder disease according to preoperative medication status were associated with an increased likelihood of remission after RYGB [55]. Dixon and O'Brien reported that a shorter history of diabetes and greater weight loss were positive predictive factors for remission [56]. This paper was consistent with this. A shorter history of diabetes with less number of insulin using patients prior to metabolic surgery resulted in greater remission rate of diabetes. Metabolic surgery should be considered early in the diabetic stage before irreparable beta-cell damage occurs. BMI alone is not an adequate measure to define the overall risk of morbidity and mortality in patients with established diabetes [50]. The clinical status of T2DM should be taken into account to select the suitable candidates for metabolic surgery.
The goals of treatment of T2DM are not only glycemic control but also prevention of diabetes-related complications such as macro-and microvascular diseases. The target blood pressure of <130/80 mmHg, the target cholesterol level of <200 mg/dL, and HbA1C level <7% should be achieved in diabetic patients. It has been reported that only 7.3% of adults with diabetes achieved all three recommended goals with conventional medical treatment [57]. In contrast, bariatric surgery improved hyperlipidemia in 70% or more of patients and resolved or improved hypertension in 78.5% of patients [58]. A systematic review to evaluate the effect of bariatric surgery on cardiovascular risk profile demonstrated that on average, hypertension, diabetes, and dyslipidemia resolved in 68%, 75%, and 71% and a 40% relative risk reduction for 10-year coronary heart disease risk was observed after bariatric surgery, as determined by the Framingham risk score [59]. In this paper, Shah et al. reported that RYGB for T2DM in low BMI patients reduced the predicted 10-year cardiovascular disease risk substantially for fatal and nonfatal coronary heart disease and stroke [21].
The mortality rates from bariatric operations (0.28-0.35%) [60] are compared favorably with those of other commonly performed operations, including laparoscopic cholecystectomy, whose mortality in USA ranges between 0.35 and 0.60% [61]. In this paper, no mortality was observed. Major and minor complication rate was also low (10.3%). Huang et al., and so forth reported that the operating time and duration of hospitalization of LRYGB for low BMI patients were lower than those for morbidly obese patients because of lower BMI [19]. T2DM-related additional risk should be counted, but safety of metabolic surgery for low BMI patients seems to be higher or at least similar, compared to bariatric surgery for severe obesity.
Metabolic surgery for T2DM, although not the current standard care for the disease, may be coming closer to the mainstream. The ponderable statement has suggested that metabolic surgery might be considered a reasonable therapeutic alternative for low BMI (<35 kg/m 2 ) patients with T2DM who do not respond to standard medical therapy. The data from the studies included in this paper are encouraging. Although large randomized clinical trials against best medical care and assessment of the long-term efficacy and safety should be prioritized to define the role of metabolic surgery, it is clear that a high proportion of low BMI patients with T2DM will derive substantial benefit from metabolic surgery.

Conclusions
In this paper, including 18 studies and 477 patients, T2DM patients with a BMI < 35 kg/m 2 derived benefit from metabolic surgery. The weight loss effect was reasonable without any serious excessive weight loss. The antidiabetic effect was also considered excellent. Remission of T2DM and % meds off were achieved in 64.7% and 86.8% of the patients. T2DM clinical status is important to select the eligible candidates for metabolic surgery besides current BMI criteria for bariatric surgery. Metabolic surgery can be performed safely with acceptably low complication rate and mortality. Although several concerns need to be addressed, metabolic surgery for low BMI patients is coming closer to the mainstream of diabetes treatment.

Conflict of Interests
The authors declare that they have no relevant competing interests.