The clinical observation that bariatric surgery seemingly “cures” type 2 diabetes has generated intensive research in humans and animal models to identify the mechanisms underlying the metabolic benefits gained from these procedures. A bariatric-surgeon-in-practice, a clinical researcher, and a basic researcher all provide unique perspectives on this phenomenon.
The prevalence of obesity and its comorbidities, including cardiovascular disease and type 2 diabetes is expanding worldwide. Therefore, there is a pressing need for the development of new and effective strategies for the prevention and treatment of obesity and metabolic disease. Bariatric surgery remains the most effective long-term treatment for obesity. It is defined as a surgical manipulation of the gut performed for the purpose of weight loss and improvement of metabolic disease. Certain bariatric procedures, such as Roux-en-Y gastric bypass (RYGB), can cause type 2 diabetes remission within days after surgery, before any significant weight loss. This clinical observation has sparked intensive research in humans and animal models to identify the mechanisms underlying the metabolic benefits gained from bariatric surgery. Ideally, such research will lead to the development of new therapies or combination of therapies that can reproduce these metabolic benefits with minimal risk and in a durable fashion. In this TriPoint article, a bariatric surgeon-in-practice discusses type 2 diabetes remission and improvement following surgery, predictors of remission, and durability of remission; a clinical researcher provides perspective on contributions that human clinical studies have made to our understanding of the metabolic benefits of bariatric surgery and highlights areas in which more clinical data is needed; and a basic researcher reviews the contributions that work in rodent models of bariatric surgery have made toward our current understanding of the mechanisms driving the metabolic improvements observed after bariatric surgery.
CLINICAL PRACTITIONER’S PERSPECTIVE
— Joshua B. Alley, MD, FACS
In 1967, a surgeon from Iowa named Edward Mason, MD reported a modifi cation of a procedure for peptic ulcer disease that could be used to induce weight loss in patients suff ering from severe obesity. He called the new procedure “gastric bypass.” Shortly after performing the fi rst several dozen procedures, it was noted that patients with type 2 diabetes very often saw a resolution of their clinical condition to a normoglycemic state. Th rough the years, this procedure has been modifi ed, refi ned, and standardized. It is performed by creating a small stomach pouch that bypasses the duodenum and is connected directly to the jejunum. It is now performed almost exclusively laparoscopically, through small abdominal incisions with the aid of a video laparoscope in the abdomen. In the past 15 years, laparoscopic adjustable gastric banding (band placed around the stomach and tightened by injecting saline) and laparoscopic vertical sleeve gastrectomy (VSG, a newer procedure involving removal of a portion of the stomach to leave behind a narrow, “sleeve”-shaped portion of the lesser curvature) are surgical procedures that have also been applied to treat morbid obesity, but laparoscopic RYGB remains the most common procedure used by bariatric surgeons in situations where the patient suffers from both severe obesity and type 2 diabetes.
Type 2 Diabetes Remission
The term “remission,” rather than “cure,” is typically used today when describing the eff ect of bariatric surgery on type 2 diabetes, to emphasize that the disease is rendered inactive rather than extirpated, and as a reminder that recurrence (or relapse) is possible. The rate and degree of remission varies by patient population and by procedure. Dixon, et al. reported type 2 diabetes remission of 73% at two years post lap band surgery (versus 13% in medically managed patients), in a very highly selected group of patients with newly diagnosed, mild type 2 diabetes. Th e more recent STAMPEDE trial was a randomized, controlled trial that enrolled 50 subjects with type 2 diabetes in three treatment arms: intensive medical therapy, RYGB, or VSG. Th e study groups each had a mean BMI of 36 (Class II obesity), with mean duration of type 2 diabetes of more than eight years and mean glycated hemoglobin A1C (HbA1C), an index of plasma glucose concentration over time of 8.9% or greater, suggesting that these research cohorts more accurately refl ect real-world conditions of poor type 2 diabetes control. Forty-four percent of each group used insulin, and more than 90% of the subjects in each group met criteria for metabolic syndrome. At one year, achievement of the primary endpoint (HbA1C ≤ 6% with no type 2 diabetes medications) was as follows: 0% in medical therapy group, 42% in RYGB group, and 27% in VSG group. At three years, the numbers were 0%, 35%, and 20%, respectively. Signifi cant improvement in glycemic control and medication use were observed in the surgery groups, even among those who did not see a complete type 2 diabetes remission.
A Swedish longitudinal observational study of patients undergoing a mix of various bariatric surgical procedures evaluated long-term remission rates and macrovascular and microvascular complications of type 2 diabetes after bariatric surgery and compared rates to those in a matched cohort. With remission defi ned as fasting glucose <100mg/dL and no diabetes medication, one-year remission rates for type 2 diabetes were 16% for the control group and 73% for the bariatric surgery group. Fifteen-year remission rates declined to 6.5% and 30.0%, respectively. Macrovascular and microvascular complications of type 2 diabetes were signifi cantly lower in the bariatric surgery group at a median follow-up of ~18 years.
Predictors of Remission and Relapse
A common question that I am asked in the clinic is “How likely is it that my diabetes will go away?” Several studies have examined predictors of type 2 diabetes remission and relapse, with most of them focusing on RYGB (rather than adjustable gastric banding or sleeve gastrectomy). Th e common factors that emerge from the published literature suggest that preoperative type 2 diabetes severity and duration are inversely correlated with likelihood of remission (long-term, insulin-dependent diabetic patients are less likely to see a complete remission and are more likely to relapse). One study helped reveal some reasons for this pattern. Nannipieri, et al. showed that pancreatic beta cells exhibit baseline (preoperative) glucose insensitivity in patients with type 2 diabetes, greater than in other obese patients without type 2 diabetes. Th e beta cell glucose sensitivity improved early after surgery but plateaued thereafter, and never achieved glucose sensitivity equivalent to that of normal controls. Another study bolsters this “beta cell reserve” argument, showing that preoperative C-peptide levels (a marker for insulin biosynthesis and release) correlated with achievement of type 2 diabetes remission; a preoperative C-peptide level of 6 ng/mL were associated with remission rates of 82% and 90%, respectively.
In conclusion, while bariatric surgery off ers the potential for long-term improvement or remission of type 2 diabetes and other obesity-related comorbidities, bariatric surgeons have learned to temper their enthusiasm for this important therapy with the understanding that type 2 diabetes remission is not complete or lifelong in every bariatric surgery patient. We will need to tailor combinations of medical, surgical, and lifestyle interventions, in collaborative ways, to best achieve the patient’s desired outcome as we treat the chronic diseases of type 2 diabetes and obesity.
CLINICAL RESEARCHER’S PERSPECTIVE
— Blandine Laferrère, MD
Clinical Impact of Bariatric Surgery
in Individuals with Severe Obesity
About 35% of American adults and 17% of American children are obese. By 2020, the prevalence of severe obesity, class 2 (BMI>35), class 3 (BMI> 40), and class 4 (BMI> 45), respectively, is expected to increase to 16.4%, 6.3%, and 3.1% for men and 25.3%, 12.8%, and 5.8% for women. Bariatric surgery, the treatment of choice for class 2-4 obesity, is the only treatment that results in large ~30% weight loss, often sustained over time. Data from retrospective and observational clinical studies, and from small-scale, short-term randomized control trials, support a high rate of short-term (one to two years) improvement and/or resolution of comorbidities after surgery, including remission of type 2 diabetes in 40% to 80% of cases. Individuals with diabetes of short duration, well controlled and not on insulin, are more likely to experience remission. Weight loss amount and type of surgery are also important determinants of the rate of remission, with a clear advantage for more invasive surgeries such as biliopancreatic diversion (BPD), RYGB, and VSG, over adjustable gastric banding and compared to diet and medical therapy.
Is There More Than Weight Loss
After Bypass Surgeries?
The rapid and greater resolution of type 2 diabetes after BPD and RYGB has led to the hypothesis that mechanisms other than weight loss, possibly gut-derived factors, may be responsible for the improved glucose control after these surgeries. However, the greater weight loss after BPD and RYGB compared to restrictive surgeries is a confounding factor. Smaller-scale mechanistic studies have attempted to circumvent the weight loss variable by studying the eff ect of diff erent types of bariatric surgery, or diet-induced weight loss, at matched weight loss. Interestingly, these studies show little to no difference between procedures and/or between surgery and calorie-restrictive diet on insulin secretion, insulin sensitivity, and body composition. Th erefore, weight loss, rather than mode of weight loss, is the important variable for diabetes remission. One exception to that rule is the incretin effect during meals, defi ned as the largest insulin response to oral versus intravenous isoglycemic glucose challenge. The endocrine response to a meal shows a clear boost of the incretin eff ect, after RYGB. This is not surprising as post-prandial GLP-1 release is enhanced after RYGB, as a consequence of accelerated nutrient transit time. The enhanced GLP-1 release results in greater meal-related insulin secretion, and in some patients, triggers reactive hypoglycemia. In addition to the change in gut peptide release, novel areas of interest include the role of bile acid and FGF19 signaling and microbiota, all possible factors in improved metabolism and sustained weight loss after surgery. Understanding the mechanisms of the altered taste, food choices, and brain circuitry may also provide clues to understand why weight loss is sustained after surgery.
Importance of Long-Term Studies
to Measure Clinical Outcomes
The major limitation of past clinical, observational, and randomized studies is their short duration. Both obesity and type 2 diabetes are lifetime diseases. Studies looking at the effect of an intervention over one or two years are largely inadequate. Th e unique Swedish Obesity Study provides valuable long-term data. As described above, the 20-year follow-up made it possible to demonstrate a lower rate of microvascular and macrovascular complications after bariatric surgery compared to usual medical care in a subgroup of patients with diabetes. Moreover, long follow-up allowed calculation of diabetes relapse. At two years, 72.3% of patients were in diabetes remission, but only 30.4% were free of disease at 15 years. Interestingly, only minimal weight regain occurred between two years (21.2% weight loss) and 10 years (18% weight loss), and cannot alone explain diabetes relapse in two-thirds of subjects. More recently, the use of electronic medical records off ers a cheaper alternative to large multicenter clinical trials. Arterburn, et al. also showed diabetes relapse fi ve to eight years after RYGB in more than 35% of patients who initially experienced remission.
What Else Needs to be Done?
Bariatric providers, researchers, and patients have remaining questions. What are the predictors of weight loss outcomes, long-term (lifetime) complications, long-term survival, microvascular and macrovascular events, mental health outcomes, and costs? Th e importance of studies to assess durability of the eff ect of surgery and predictors of success, as well as the need for comparative studies of diff erent treatment modalities, was highlighted during an NIH symposium (Courcoulas et al. JAMA Surg 2014 Dec; 149(12):1323-9). Results from these studies will be particularly important for younger patients having bariatric surgery in their twenties, at the beginning of their reproductive life, and who may expect to spend at least 50 years of their life in the postbariatric state.
BASIC RESEARCHER’S PERSPECTIVE
— Bethany Cummings, DVM, PhD
How Does Bariatric Surgery Result
in Type 2 Diabetes Remission?
Many postoperative changes noted after bariatric surgery have been hypothesized to contribute to the metabolic benefi ts of these procedures. Th ese changes include: increases in post-prandial glucagon-like peptide-1 (GLP-1) secretion; increases in circulating bile acid concentrations; and alterations in gut microbial populations. Much of the previous work supporting these candidate mechanisms was purely correlative; however, recent work in rodent models of bariatric surgery has begun to test and quantify the actual contributions made by these proposed mechanisms.
Role of GLP-1
Post-prandial increases in GLP-1 secretion have been noted in many human clinical studies, and after several types of bariatric surgery, such as RYGB and VSG. GLP-1 is a hormone produced by gut enteroendocrine L-cells that has numerous antidiabetic actions. Th erefore, the initial fi ndings that GLP-1 secretion is elevated after various bariatric procedures was met with enthusiasm and the assumption that GLP-1 is a major driver for type 2 diabetes remission after bariatric surgery. However, studies using genetic and pharmaceutical ablation of GLP-1 receptor signaling in rodent models of VSG and RYGB have revealed no signifi cant detriment in bariatric-associated improvements in body weight and glucose homeostasis. Th ese studies suggest that increases in GLP-1 receptor signaling may not be playing a prominent role in the glucoregulatory benefi ts of bariatric surgery.
Role of Bile Acids
Bile acids are amphipathic steroid molecules that have a well-known role in lipid digestion and absorption. However, bile acids also play an important role in the regulation of glucose homeostasis. Two bile acid receptors, Farnesoid X Receptor (FXR) and G-protein coupled bile acid receptor (TGR5), have been identified as mediators of the glucoregulatory effects of bile acids. The antidiabetic effects of bile acid signaling through FXR include: decreased liver triglyceride content, decreased hepatic gluconeogenesis, and increased insulin secretion. The antidiabetic effects of TGR5 signaling include increased GLP-1 secretion, increased energy expenditure, and decreased inflammatory cytokine release from immune cells.
Circulating bile acid concentrations are elevated after RYGB and VSG in human bariatric patients and in rodent models. Increased bile acid signaling has been suggested to contribute to metabolic improvements after bariatric surgery. This was assessed by studying the effects of VSG in high-fat fed wild type and wholebody FXR knockout mice. In the absence of FXR signaling, VSG-induced improvements in body weight and glucose tolerance were blunted compared with wild type mice. Furthermore, my laboratory has unpublished data demonstrating a relative impairment in the improvement of glucose tolerance after VSG surgery in whole-body TGR5 knockout mice compared to wild type controls. Overall, the emerging mouse data suggest that increases in bile acid signaling may be contributing to improvements in glucose homeostasis after bariatric surgery.
Role of the Gut Microbiome
In recent years, specific gut microbial populations have been implicated as regulators of metabolic homeostasis. Pioneering studies demonstrated that obesity is associated with an increased Firmecutes to Bacteroides ratio. Furthermore, obese and lean phenotypes are transmissible via inoculation of germ-free mice with obese or lean microbiomes, strongly suggesting that gut microbial populations can influence body weight and may represent a new therapeutic target in the treatment of obesity. Rodent and human studies report increases in Gammaproteobacteria and decreases in Firmicutes after RYGB. Furthermore, inoculation of gnotobiotic mice with gut microbiome derived from mice after RYGB produces greater weight loss and metabolic improvement than inoculation of gnotobiotic mice with gut microbiome derived from sham-operated mice.
Notably, bile acids have an interdependent relationship with the gut microbiota. Gut microbes can deconjugate and convert primary to secondary bile acids during enterohepatic recirculation. Conversely, bile acids regulate gut microbial growth and composition. In addition to the reported reliance of VSGinduced improvements in body weight and glucose tolerance on FXR signaling, Ryan, et al. also reported that some of the postoperative changes in gut microbial composition rely on FXR signaling. In particular, the prevalence of Roseburia is elevated after VSG surgery in wild type mice, but this is not the case in VSGoperated FXR knockout mice. These data suggest that there is important cross-talk between postoperative changes in bile acid signaling and alterations in the gut microbiome. While alterations in gut microbial populations appear to play an important role in producing the metabolic benefits of bariatric surgery, the specific role of these bacteria remain to be elucidated.
In summary, genetic and pharmaceutical studies in rodent models of bariatric surgery are improving our understanding of the relevance of certain postoperative changes in mediating the metabolic benefits of bariatric surgery. Such studies suggest that changes in bile acid signaling and the gut microbiome are key contributors to postoperative improvements in glucose homeostasis. It is likely that further studies in rodents will identify other key contributors. However, whether these systems can be safely and effectively targeted in human patients to recapitulate the metabolic benefits of bariatric surgery remains to be determined.