Diabetes and obesity
The rapid increase in the prevalence of obesity and morbid obesity, and consequently the increase in the prevalence of type 2 diabetes mellitus, is one of the most challenging threats for health-care systems all over the world. In 2011, the International Diabetes Federation (IDF) suggested3 that 285 million people are suffering from diabetes today, and a further increase of this number to 438 million affected people by 2030 is to be expected.4 A recent World Health Organisation (WHO) analysis warns that currently 1.5 billion people are overweight, of whom 500 million have to be classified as obese.5 Apart from some variance between different nations, a steady increase can be observed globally for body mass index (BMI) over the last few decades.5 Obesity is considered to be the primary risk factor for type 2 diabetes mellitus.6 Therefore, obesity has to be assumed to be the greatest threat for public health in our century.
Besides type 2 diabetes mellitus, obesity and morbid obesity is associated with a large number of co-morbidities such as hypertension, coronary heart disease, dyslipidaemia, hyperuricaemia, obstructive sleep apnoea, gonarthrosis, coxarthrosis and even increased risk for several types of cancer7 (e.g. colon, breast and pancreas). All these factors contribute to this health-economic problem of immense dimensions, as already today one out of five dollars spent for health care in the USA is spent for the treatment of obesity 8. In times of ongoing national and global financial crises, public health care could reach its limit soon.
Classification of obesity
Obesity is generally defined as a BMI of more than 30 kg/m2, with morbid obesity starting at a BMI of 40 kg/m2 or more. The major advantage of using the BMI as a classification tool is its general acceptance and ease of calculation, based on the easily measured body weight and height. However, its application for the assessment of the medical sequela of obesity is limited. It is abdominal obesity that increases the risk for type 2 diabetes mellitus, as demonstrated by the Nurses' Health Study,9 which followed a cohort of 43,581 women between 1986 and 1994 in the USA. The risk of developing type 2 diabetes mellitus increased linearly with an increasing waist circumference, defining abdominal obesity by a waist circumference of 102 cm or more in men and 88 cm or more in women.10 Of the given measures of obesity, such as BMI, waist-to-hip ratio and waist-to-height ratio, the waist-to-height ratio (WtHR) represents the best predictor of cardiovascular risk and mortality, as presented by Schneider et al. in 2010.11
Subcutaneous fat and visceral fat contribute equally to the BMI value. Thus, visceral obesity is underrepresented within the BMI classification. As many indications and guidelines12 for surgical intervention for the morbidly obese are based strictly on the BMI and neglect the distribution of body fat, the biological impact of visceral obesity should be taken further into account for treatment recommendations in these patients.
Besides the fact that even a modest weight loss can dramatically improve glycaemic control in obese patients with type 2 diabetes mellitus,13 it is well known that a sustained weight loss is not commonly achieved in morbidly obese patients by lifestyle modifications or medical treatment. Furthermore, many diabetes drugs, such as insulin, promote further weight gain or increase the risk for hypoglycaemia. Some promising anti-obesity drugs have even been withdrawn from the market due to adverse cardiovascular effects14 or depression.15 So the possibilities of non-surgical interventions are limited in regard to body weight loss and amelioration of the metabolic syndrome. In contrast to lifestyle modifications and medical treatment, bariatric surgery with its wide range of different procedures has proven to achieve a durable and stable weight loss with an excessive weight loss of between 47–70% depending on the chosen surgical operation.16
Currently, Roux-en-Y gastric bypass, laparoscopic-adjustable gastric banding, biliopancreatic diversion, and sleeve gastrectomy are the most commonly performed bariatric procedures worldwide.17
By forming a small gastric pouch connected to an alimentary limb of 100–150 cm in length, the gastric bypass combines both mechanisms of weight loss surgery, restriction and mild malabsorption. Performed in a Roux-en-Y fashion comparable with the reconstruction after distal gastrectomy, Roux-en-Y gastric bypass was first performed laparoscopically in 1993 by Wittgrove et al.18 Nowadays, Roux-en-Y gastric bypass is expected to be the most commonly performed bariatric procedure in Europe. In 2009, Henry Buchwald et al. published a diabetes remission rate of 80.3% after gastric bypass19 within their meta-analysis. In a recent study by Mingrone et al.,1 all 20 patients who underwent gastric bypass in a randomised prospective study comparing gastric bypass, biliopancreatic diversion and medical diabetes treatment, were able to discontinue their diabetes drugs within15 days after surgery.
Laparoscopic-adjustable gastric banding
Without any trans-section or stapling of the stomach or gut, gastric banding is the least invasive bariatric procedure. A band is wrapped around the upper part of the stomach, creating a small gastric pouch above the band. Advantages of gastric banding are its individual adjustability to the patients demands and its reversibility, while complications occur as band slippage, band erosion, esophageal dilatation or frequent vomiting due to food intolerance. Generally, weight loss is less after laparoscopic gastric banding compared with other bariatric procedures.16 Used in a BMI range from 30 to 40 kg/m2, Dixon et al. compared laparoscopic-adjustable banding with conventional diabetes therapy and found a significantly better diabetes remission rate in the surgical group (73% cs 13%) at two years after surgery.20 Compared with other bariatric procedures, laparoscopic-adjustable gastric banding has only minor influence on gut hormones.21
Sleeve gastrectomy was initially developed as part of the biliopancreatic diversion with duodenal switch (DS), adding restriction to the strictly malabsorptive biliopancreatic diversion. As a sole bariatric procedure it was introduced as the first part of a two-step laparoscopic gastric bypass in ‘super-super obese’ (i.e. BMI > 60kg/m2) patients.22 With a rapidly increasing number of operations performed, sleeve gastrectomy has gained enormous popularity in the bariatric world within the last five years. In 2008, sleeve gastrectomy already represented 5.4% of all bariatric procedures performed worldwide in 2008 as presented by Buchwald et al.,17 and case numbers are expected further to increase. This popularity is mainly caused by short-term results for weight loss23–25 and by a wide range of advantages compared with the surgically technically more demanding Roux-en-Y gastric bypass: no dumping syndrome is expected after sleeve gastrectomy and the biliary tract remains accessible by endoscopic retrograde cholangiopancreaticography (ERCP). Furthermore, nutrient deficiencies seem less likely after sleeve gastrectomy than after Roux-en-Y gastric bypass.26,27 In 2008, Vidal et al. 28 compared diabetes remission in gastric bypass and sleeve gastrectomy patients, finding similar remission rates (84.6% vs 84.6%). In their recent publication,2 Schauer et al. found that 37% of the patients who underwent sleeve gastrectomy reached a glycated haemoglobin (HbA1c) of < 6.0% within one year, whereas the mean HbA1c dropped from 9.5% to 6.6% in this study including 50 sleeve gastrectomy patients.
The highest diabetes remission rates are found for the biliopancreatic diversion, developed by Nicola Scopinaro in 1976.29 Combining a distal gastrectomy with a bypass, resulting in a common limb length of 50–75 cm, results in malabsorption. After division of the jejunum at 250-cm proximal to the ileocoecal valve, the distal part of the jejunum is anastomosed to the stomach and the proximal part of the jejunum is anastomosed to the ileum 50–75-cm proximal to the ileocoecal valve. Whereas malabsorption is achieved by reducing the absorptive surface of the gut, there is only minor restriction depending on the size of the stomach left in place (200–500 ml). Owing to its malabsorptive effect, the most common complications are diarrhoea, anaemia, protein deficiency, iron deficiency and flatulence.30 Besides the stable weight loss results in the long-term follow-up, Scopinaro published in 200531 a 10-year recovery rate of hyperglycaemia of 97% after biliopancreatic diversion. Patients who underwent biliopancreatic diversion were able to discontinue their diabetes drugs as fast as within 15 days after surgery, as recently presented by Mingrone et al.1
Selection criteria for surgery
Back in 1991, the US National Institutes of Health (NIH) published guidelines for the application of bariatric surgery, recommending surgery for patients with a BMI of 40 or above as well as surgery for patients with a BMI of 35 and above suffering from obesity-associated comorbidities such as hypertension and type 2 diabetes mellitus.12 Despite the fact that the BMI levels set within these guidelines for recommending surgery were never confirmed by randomized clinical trials, these 21-year-old recommendations still form the basis for all patient selection criteria in bariatric surgery. With a growing scientific evidence for lower BMI indications in the diabetic obese, ‘metabolic’ instead of ‘bariatric’ indications should be should be integrated into the treatment algorithm of type 2 diabetes mellitus.
Mechanisms of surgical intervention
For decades, the mechanism of weight loss after bariatric surgery was understood and simplified strictly to only two pathways. By reducing the volume of the stomach, restriction is achieved, limiting food uptake and therefore also limiting the amount of food entering the intestine. Thus, the reduced number of up taken calories leads to a negative calorie balance. On the other side, the surgical reduction of the functional absorptive intestinal length leads to malabsorption, with less calories absorbed from the intestine. Whereas some bariatric procedures were based only on restriction (e.g. the now-abandoned vertical banded gastroplasty and laparoscopic-adjustable gastric banding) or malabsorption (biliopancreatic diversion), the gastric bypass and the duodenal switch combine these two major principles for achieving weight loss.
Gut hormones and adipocytokines
Starting with leptin32 in 1994, the discovery of adipocyte-derived substances (adipocytokines) and of gut-derived hormones regulating satiety, hunger, intestinal motility and function, and interacting with glycaemic control, widened the horizon for understanding the many functional effects of bariatric surgery not explicable by restriction and malabsorption alone. The majority of adipocytokines derive from visceral fat and play essential roles in chronic pathogenic states like inflammation, atherosclerosis, thrombosis, diabetes and hypertension.33 Bariatric surgery not only reduces the visceral fat mass, but also reduces the secretion of proinflammatory adipocytokines and increases the secretion of anti-inflammatory adipocytokines.34–39
Peptide YY (PYY), the ‘ileal brake’ hormone inhibiting gastrointestinal motility and mediating satiety, is produced by the L cells in the Ileum, colon and rectum. In obese individuals, PYY is found to be decreased,40 whereas gastric bypass surgery leads to increased levels of PYY within a few days after surgery.41,42
Glucagon-like peptide 1 (GLP-1) is secreted by the ileal L cells as a response to a meal. GLP-1 reduces appetite and stimulates insulin secretion (‘incretin’) while suppressing glucagon secretion. Furthermore, it slows gastric emptying. In patients with type 2 diabetes mellitus reduced levels of GLP-1 are found,43 whereas GLP-1 levels immediately increase after gastric bypass surgery.41,44 Starting with exenatide, a growing range of GLP-1 agonists (liraglutide, lixisenatide, albiglutide and taspoglutide) have been introduced into medical diabetes therapy.45
Ghrelin is a potent 28-amino-acid peptide orexigenic hormone, discovered in 1999,46 acting as the natural ligand of the growth hormone secretagogue receptor. Approximately two-thirds are released by the stomach, mainly the fundus. Ghrelin enhances appetite and food intake by 28%. Furthermore, it leads to an increase in cortisol, adrenaline and growth hormone – as well as a decrease in insulin secretion.47 Ghrelin is secreted in a diurnal pattern and is alleviated after gastric bypass.48–49 Sleeve gastrectomy leads to permanently reduced ghrelin levels,50 whereas gastric banding and diet result in an increase in Ghrelin concentration.51 Prolonged decrease of ghrelin levels could contribute to a reduction in the feeling of hunger.52 Ghrelin and PYY 3–36 act as short-term regulators of hunger and satiety, whereas insulin and leptin are more reflective of nutritional status and are long-term regulators (‘Yin and Yang of nutrition’).53
In 2004, Buchwald et al.16 presented diabetes remission rates for bariatric surgery, analysing data from 136 studies including 22,094 patients. Diabetes completely resolved in 76.8% of the patients and resolved or improved in 86%. Thus, bariatric surgery no longer focused on weight-loss results alone, but also on metabolic improvements. With the proven efficacy of bariatric surgery for not only a stable and lasting profound weight reduction, but also metabolic changes such as glycaemic control, diabetes remission, improved lipid levels and reduced hypertension, two major surgical associations – the International Federation for the Surgery of Obesity and the American Society of Bariatric Surgeons – included the term ‘metabolic surgery’ in their name in 2007, becoming respectively the International Federation for the Surgery of Obesity and Metabolic Disorders (IFSO) and the American Society of Metabolic and Bariatric Surgery (ASMBS).
Obesity and life expectancy
Obesity (BMI 30–35 kg/m2) reduces not only the quality of life, but also the life expectancy for non-smoking or smoking women, by a mean of 7.1 kg/m2 or 13.3 years, respectively.54 For nonsmoking or smoking obese men, calculations reveal a tremendous lifetime reduction of 5.8 kg/m2 or 13.7 years, respectively.
Bariatric surgery and mortality
As bariatric surgery adds potential operative danger to morbidly obese patients, studies focusing on mortality are of great interest, comparing the mortality risk of patients who underwent surgery with that of non-operated morbidly obese patients. In 2007 Adams et al. published the results from an analysis including 7925 patients, focusing on long-term mortality after gastric bypass surgery.55 Disease-specific mortality could be reduced by bariatric surgery for diabetes by 92%, for coronary heart disease by 56% and for cancer by 60%. Although the Swedish Obese Subjects (SOS) study56 revealed a reduction of all-cause mortality of 22% after 11 years, Adams et al.'s analysis found a longterm all-cause mortality reduction of 40% for the surgical patients. Thus, bariatric surgery not only affects diabetes and its associated co-morbidities such as hypertension, sleep apnoea or arthritis, but also leads to a general reduction in mortality. Many studies further demonstrate that besides weight loss and co-morbidities, the quality of life is also substantially improved after bariatric and metabolic surgery.57,58
Bariatric surgery: numbers and trends
Comparing data from 2003 with 2008, a rapid increase in the number of bariatric procedures performed worldwide can be observed, with 146,301 operations in 2003 rising to 344,221 procedures performed by 1625 surgeons in 2008.17 Currently, Roux-en-Y gastric bypass, laparoscopic-adjustable gastric banding, biliopancreatic diversion and sleeve gastrectomy are the most commonly performed and accepted bariatric procedures. Whereas similar overall growth rates for bariatric procedures could be observed between 2003 and 2008 for Europe (+97%) and the USA (+113%), the types of procedures chosen show an opposing trend when comparing the USA and Europe. The number of gastric bypasses performed in the USA decreased from 85% in 2003 to 51% in 2008, with laparoscopic-adjustable gastric banding rapidly increasing from 9% to 44%. Over the same five years gastric banding lost its predominance in Europe, dropping from 64% to 43% of all operations, whereas the gastric bypass increased from 11% to 39% of all bariatric procedures performed in Europe in 2008.
Diabetes remission rates
Analysing the results from gastric bypass, gastric banding and biliopancreatic diversion, the different weight-loss mechanisms and different influences on gut hormone regulations leads to a variability in weight-loss extent and diabetes remission rates. Many factors influence the extent of diabetes remission after bariatric surgery: the choice of bariatric procedure, the duration of diabetes prior to surgery, the pre-surgical diabetes medication (insulin vs medical treatment) and the patient's age. In 2009, Buchwald et al.19 published diabetes remission rates for the most commonly performed bariatric procedures – gastric bypass, gastric banding and biliopancreatic diversion – as 80.3%, 56.7% and 95.1%, respectively, analysing data from 621 studies including 135,246 patients. Generally, the more weight loss that can be achieved by a surgical procedure, the higher the probability for a diabetes remission after surgery – making weight loss still the strongest factor in diabetes remission.
Prospective randomized studies
To assess the potential role of bariatric surgery in the treatment algorithm of type 2 diabetes mellitus in the obese, prospective randomized studies are needed. However, it is difficult to randomize patients into two interventions like bariatric surgery and medical treatment, as the extent of intervention is completely different. With two studies published this year in the New England Medical Journal, we gain more comparative information about the efficacy of these both different treatment modalities.1,2
Published in 2012 in the New England Journal of Medicine, Mingrone et al. 1 compared bariatric surgery with conventional medical therapy, including 60 patients with a BMI of > 35 kg/m2, and a history of type 2 diabetes mellitus of at least five years. Twenty patients were assigned to medical therapy, 20 underwent gastric bypass surgery and in 20 patients was performed. The primary endpoint was the rate of diabetes remission at two years, defined as a fasting glucose level of < 100 mg/dl (5.6 mmol per litre) and HbA1c levels of < 6.5% in the absence of pharmacological therapy. After a follow-up of two years, no diabetes remission was observed in the medical therapy group, whereas diabetes remission was found in 74% of the bypass patients and in 95% of the patients who underwent biliopancreatic diversion. The HbA1c level was found to be a mean of 7.69% for the medical group, whereas surgery decreased the HbA1c level to a mean of 6.35% after gastric bypass and 4.95% after biliopancreatic diversion. Excess weight loss was similar in both surgical groups, with mean 69% and 68%, respectively, whereas only 9% excess weight loss was achieved by the medically treated patients. Furthermore, surgery led to instant glycaemic control, as all gastric bypass and biliopancreatic diversion patients were able to discontinue their pharmacological diabetes treatment within only 15 days after the operation.
The second randomized prospective study focusing on this topic was entitled ‘Bariatric surgery versus intensive medical therapy in obese patients with diabetes’, published by Schauer et al.2 in the New England Journal of Medicine in 2012. The authors included 150 patients with a BMI of 27–43 kg/m2 in this study comparing medical diabetes therapy with gastric bypass or sleeve gastrectomy. The primary endpoint of this study was an HbA1c level of 6.0% or less at 12 months after treatment. Fifty patients each were randomised to receive gastric bypass surgery, sleeve gastrectomy or intensive medical treatment. The duration of diabetes was more than eight years, comparable in all groups. In each group, 44% of the patients used insulin. Focusing on the primary endpoint, 12% of the medically treated patients reached this goal, whereas 42% of the bypass patients and 37% of the sleeve patients had HbA1c levels of < 6.0%. The mean HbA1c levels decreased in all three groups: from 8.9% to 7.5% in the medical group, from 9.3% to 6.4% in the bypass group and from 9.5% to 6.6% after sleeve gastrectomy. Insulin use remained high in the medical treatment group (38%) after 12 months, although a remarkable decrease in insulin use was found in the gastric bypass group (4%) and the sleeve gastrectomy group (8%). As expected, weight loss was better in the surgical groups, with a mean body weight loss of 27% and 24% after gastric bypass and sleeve respectively, compared with mean 5% body weight loss after medical treatment alone. Furthermore, the use of drugs to lower glucose, lipid and blood pressure significantly decreased after gastric bypass and sleeve gastrectomy, whereas medication increased in the medically treated patients.
Focusing on laparoscopic gastric banding versus medical treatment for type 2 diabetes mellitus, Dixon et al. published in 2008 their study entitled ‘Adjustable gastric banding and conventional therapy for type 2 diabetes’ including 60 patients with a BMI of 30–40 kg/m2 and a history of diabetes of less than two years.20 Thirty patients were randomized into surgical and 30 patients into medical treatment. Diabetes remission was found in 73% of the gastric banding patients and in 13% patients who received best medical diabetes care, which included a dietitian, general physician and diabetes educator with frequently scheduled follow-up visits. Comparing diabetes remission results from this study with the results of Mingrone et al.1 and Schauer et al.,2 the gastric banding patients included in this study had a shorter history of diabetes and a lower mean BMI, expectedly resulting in a better diabetes remission rate. Generally, biliopancreatic diversion and gastric bypass achieve a better diabetes remission rate than laparoscopic gastric banding, as shown in the metaanalysis of Buchwald et al.19
Metabolic surgery affects not only on diabetes remission, but also the many diabetes-associated co-morbidities like hypertension, hypercholesterolaemia, hypertrigyceridaemia and hyperuricaemia, summarized by the metabolic syndrome.
In 2004, Sjöstrom et al. published ‘Lifestyle, diabetes and cardiovascular risk factors 10 years after bariatric surgery’ in the New England Journal of Medicine, presenting the 10-year data from the SOS study.56 In this analysis, the 2-year and 10-year recovery rates from diabetes, as well from hypertriglyceridaemia, low levels of high-density lipoprotein-cholesterol, hypertension and hyperuricaemia, were more favourable in the surgically treated group. Interestingly, this study revealed better two-year rates for diabetes remission (72%) compared with 10-year data36. The patients in the SOS study not only underwent gastric bypass (5%) or gastric banding (24%), but in the majority of the patients a vertical banded gastroplasty (71%) was also performed. In all the patients, a weight regain was observed explaining the 10-year outcome. Nowadays vertical-banded gastroplasty has been abandoned owing to frequent staple-line disruption in the longer term, subsequently leading to significant weight regain.59
In their 2004 meta-analysis16 including 136 studies with a total of 22,094 patients, Buchwald et al. found a diabetes remission rate of 76% for all different surgical procedures, whereas hyperlipidaemia improved in 76% of the patients. Hypertension was found to resolve in 62% of the patients and obstructive sleep apnoea was resolved in 85%. Thus, bariatric surgery leads to an dramatic improvement in obesity-associated co-morbidities, aside from the effects on type 2 diabetes mellitus itself.
Lower body mass index indications
With the proven success in morbidly obese patients, bariatric procedures were performed even in patients with a lower BMI, focusing more on the metabolic effects of surgery than weight reduction itself. In 2006, O'Brien et al.60 published the results of a study including patients with a BMI of 30–35 randomized for laparoscopic gastric banding or a non-surgical program with a very-low-calorie (Optifast) diet, pharmacotherapy (orlistat) and lifestyle modifications. Metabolic syndrome was reduced from 38% of the patients prior to surgery to 3% postoperatively. Another study published by Dixon et al. focused on laparoscopic-adjustable gastric banding in low-BMI (30–40 kg/m2) obese patients.20 In the 30 patients randomized for surgery, a diabetes remission rate of 73% was found. Cohen et al. 61 performed gastric bypass surgery in low-BMI patients with type 2 diabetes mellitus. The BMI ranged from 32 to 35 kg/m2. All 37 patients enrolled into this study were able to discontinue their oral anti-diabetic medication postoperatively, whereas 36 of the 37 patients achieved a total remission of all their co-morbidities. Nicola Scopinaro and coworkers were able to demonstrate a reduction in HbA1c levels from 9.3% to 6.3% in 30 patients after biliopancreatic diversion with BMI from 25–35 kg/m2.62 In recent years, several new operations (omentectomy, duodeno–jejunal bypass63 and ileal transposition with or without sleeve gastrectomy64) have been designed to optimize effects on gut hormones and adipocytokines. Although they show promising results, it is too early to recommend their application outside of study protocols.
Future aspects of metabolic surgery
Metabolic surgery has proved to be an effective therapy for morbidly obese patients with type 2 diabetes mellitus. With a long duration of diabetes being the strongest predictor of postoperative non-remission of diabetes,65,66 these patients should be considered to be too late for surgery. Furthermore, BMI should not be considered the best measure of obesity for the indication for surgery. Within the near future, metabolic surgery could be considered an alternative treatment option in patients with a BMI between 30 and 35 kg/m2 when diabetes cannot be controlled adequately by an optimal medical regimen, especially in the presence of other major cardiovascular risk factors, according to the 2011 statement of the IDF.3