March 1, 2019
Metabolic surgery was first defined in 1978 by Buchwald and Varco, who recognized that certain diseases, such as type 2 diabetes (T2D) and obesity, were metabolic in nature and that these complex conditions benefitted more from bariatric surgery than from weight loss alone.1 The encompassing idea of metabolic surgery has led to a growing interest in new procedures, devices, and therapies based on the physiologic changes observed after bariatric surgery. Among the metabolic surgery concepts under development (many described in this series of articles), are those that concern the nervous system, which has been shown to regulate certain metabolic functions. Therefore, bariatric procedures that alter normal neurologic functions should rightfully be included in the broad definition of metabolic surgery.
Despite the nearly century-old knowledge that the vagus nerve and the brain play significant roles in appetite, energy regulation, and body weight, the last 50 years of bariatric surgery have focused on the concepts of involuntary food intake restriction and/or nutrient malabsorption. Nearly all bariatric surgery procedures were developed solely as different methods for partitioning or removing part of the stomach to create a small reservoir and bypassing various lengths of the small intestine to reduce absorptive surface area and capacity.
Research conducted in the last two decades or so has shown that the vagus nerve (and possibly other nerves), the brain, and the neuroendocrine systems have significant control over body weight and metabolism. The growing worldwide obesity epidemic, along with the unwillingness of most potential operative candidates to undergo conventional bariatric operations, has generated great interest in the development of novel therapies for obesity and its comorbidities.
This article reviews the state of development and results of these neuromodulatory procedures.
Neuromodulation refers to the application of patterned electrical impulses to target tissues, which may include the stomach, intestine, glands, nerves, or the brain. The electrical impulses can be altered to either enhance or block normal physiologic functions.
One of the primary functions of the stomach is to break down and present partially digested food to the duodenum. Gastric contractions are regulated by the myoelectrical activity of the stomach, which consist of slow waves and spike potentials.2 Slow waves originate in the proximal stomach and propagate distally at regular intervals toward the pylorus. The slow waves determine the maximum frequency, propagation velocity, and direction of gastric contractions. In humans, the normal frequency of the slow waves is generally three cycles per minute. Spike potentials are strong action potentials that also propagate from the proximal stomach to the pylorus.
When a spike potential falls on a slow wave, a strong gastric contraction will occur, which may result in rapid gastric contractions, referred to as tachygastria, or slow gastric contractions, called bradygastria.3 Bradygastria is attributed to a decrease in the frequency of the normal gastric pacemaker. In contrast, tachygastria mainly originates from an ectopic pacemaker often located in the distal stomach. Normal gastric electrical activity propagates antegrade (from the proximal stomach to the distal stomach), whereas abnormal waves propagate retrograde and may disrupt the normal antegrade contractions. Because gastric emptying plays an important role in regulating food intake, gastric distension may inhibit eating.4
Gastric electrical stimulation (GES) has been studied as a potential treatment for morbid obesity. Both antegrade and retrograde stimulation have been investigated. It is assumed that antegrade stimulation would entrain the normal slow waves. Antegrade stimulation has yet to demonstrate itself to be clinically successful for achieving weight loss, but it has been successful in the treatment of gastroparesis.5-8
Retrograde GES yielded some degree of success in treating morbid obesity. It is assumed that retrograde stimulation may retard gastric emptying, resulting in early satiety and reduced food intake. It also is possible that retrograde GES results in fundic relaxation and distention, but the actual mechanism of action has not been proven in humans.9 To achieve retrograde GES, a pacemaker-like device is attached to electrodes implanted into the distal stomach. It is theorized that the electrical impulses propagate from the distal to proximal stomach, interfering with the normal antegrade impulses.
Studies investigating the potential for GES to induce weight loss were performed in 1992 and first reported by Cigaina and colleagues in 1996. They demonstrated that high-frequency GES was safe and moderated weight gain in a growing porcine model. For the first 12 weeks of the study, no differences in food intake or weight were observed between study animals and a control group. However, after 13 weeks, study animals decreased their food intake and weight relative to the control group. After 8 months, food intake was 12.8 percent lower in study animals, and their weight was 10.5 percent less than the control animals.10
In 1995, Cigaina performed the first human trial. Four women with a body mass index (BMI) of 40 kg/m2 or greater were implanted and monitored for up to 40 months. Platinum bipolar electrodes were implanted intramuscularly on the anterior gastric wall, adjacent to the lesser curve and proximal to the pes anserinus. An electrical stimulator was implanted in a subcutaneous pocket on the anterior abdominal wall. All four patients were permitted food and drink ad libitum. At 40 months after implantation, one patient had lost 32 kg, and a second had lost 62 kg. In the other two patients, malfunctions in their stimulator system were discovered.11
In 1998, a second study was performed in 10 patients, all with a BMI ≥40 kg/m2. After implantation, all subjects were permitted food and drink ad libitum during three regular meals but told not to eat between meals. Sweetened and alcoholic beverages were discouraged. No significant complications or deaths occurred during the study. After 36 months of stimulation, the mean weight loss of all 10 patients was 24 percent of excess BMI (>25 kg/m2), which was maintained until battery depletion.12 Although Cigaina’s work was instrumental in introducing gastric stimulation for weight loss to the world, his studies were small, underpowered, and lacked statistical analysis.
The first large multicenter, randomized, controlled, double-blinded trial was performed in the U.S. A total of 103 morbidly obese subjects were enrolled. One month after implantation, subjects were randomized to have their device activated or remain in the off mode. After seven months, the devices in the latter group were activated. Device settings were universal for all patients, and no dietary or behavioral counseling was provided. No deaths or complications occurred, and none of the patients experienced any adverse effects. However, 17 of the first 41 subjects were observed to have electrode lead dislodgements from the stomach wall. Interestingly, many patients admitted to having deliberately overeaten to discern whether their devices were activated. Despite the unintended consequences, 20 percent of the patients lost more than 5 percent of their total body weight, and the mean total weight loss was 11 percent after one year of stimulation.13
The U.S. Dual-Lead Implantable Gastric Electrical Stimulator Trial (DIGEST) was a smaller open-label investigation undertaken to evaluate a new device that had two leads (four electrodes) versus the previous single-lead (two electrodes) system (see Figures 1 and 2). In addition, a new lead-fixation technique was under investigation. A total of 30 morbidly obese subjects were enrolled at two clinical sites. Overall, patients experienced 15 percent excess weight loss at 38 weeks. However, results at the two institutions differed greatly. At one site, a mean excess weight loss of 30.4 percent at a mean follow-up of 9.5 months was achieved. At the other site, where a turnover of the research nurse position occurred, no significant weight loss occurred.14
Figure 1. Dual-lead implantable gastric electrical stimulation
A large prospective, randomized, placebo-controlled, double-blinded, multicenter trial was conducted, which involved 190 subjects who were randomized to either an active device or a placebo device. Again, GES was shown to be safe, but unfortunately, like the O-01 trial, the Screened Health Assessment and Pacer Evaluation (SHAPE) trial failed to demonstrate a superior weight loss for the study subjects with functioning devices versus the control group (~12 percent excess weight loss in both groups). However, in this investigation, a number of factors may have contributed to the disappointing results, including the fact that 26 percent of patients in the study group experienced exhaustion of their device batteries before the trial ended.15
In addition to the study limitations of the investigations with the continuous retrograde GES, the lower than anticipated weight loss may have been a result of the 24-hour continuous stimulation leading to therapy tolerance or even resistance. To prevent this potential phenomenon from occurring, some GES systems were designed to be meal-activated. Hence, these devices lay dormant until the patient ate. Two such systems are in development: abiliti’s Closed Loop Gastric Electrical Stimulation (CLGES) system and the Tantalus Diamond device.16,17
The CLGES system includes the meal-activated GES and an accelerometer to monitor subject physical activity. A total of 45 morbidly obese subjects were enrolled in a multicenter, open-label trial. At 12 months, weight loss averaged 15.7 ± 7.7 percent of the baseline body weight. With stimulation, the number of disallowed meals and between-meal snacks was reduced (p <0.05), all levels of physical activity increased (p <0.001), and activity-based energy expenditure (303 ± 53 kcal/day on average, p <0.001) improved.16
The Tantalus Diamond meal-activated GES device was developed with a focus on metabolic benefits other than weight loss. Glycemic improvements in patients with T2D were evaluated in a 48-week multicenter, randomized, blinded, crossover study. The outcomes for active devices were analyzed and compared with the results of inactive devices, which served as the controls. At six months, the subjects were crossed over. In the first 24 weeks, the two groups showed no difference in the change in HbA1C. However, when the GES was turned off in subjects with previously activated devices, the improved HbA1C values returned to baseline. In contrast, when the GES was activated in subjects who were in the control group, they maintained their improved HbA1C values.17
Duodenal stimulation (see Figure 2) was evaluated for its potential effects on T2D. In one study, Khawaled and colleagues demonstrated that electrical stimulation of the duodenum of 33 male Sprague-Dawley rats, when activated immediately after a glucose tolerance test, resulted in a significant decrease in the rising phase slope and maximal value of serum blood glucose. Insulin secretion was decreased by 21 percent. In addition, the gastric emptying rate was decreased by 80 percent, and the intestinal flow rate was increased by 40 percent. Thus, duodenal electrical stimulation (DES) may reduce postprandial blood glucose levels by changing gastrointestinal motility to reduce available intraluminal glucose for absorption.18 Liu and colleagues found similar results in humans who were subjected to intestinal electrical stimulation. Nutrient absorption was decreased and intestinal transit was accelerated.19
The vagus nerve is the 10th cranial nerve and the longest cranial nerve in the body (see Figure 3). It has numerous physiologic functions and has long been known to modulate gut function as well as body weight, appetite, and energy intake. It has been demonstrated that approximately 80 to 90 percent of vagal nerve fibers are sensory, bringing information from the gut to the brain, and the rest are motor, modulating intestinal motility, pancreatic enzyme secretion, and endogenous gluconeogenesis.20 Consequently, the vagus nerve is considered the communication link between the gastrointestinal tract and the brain.
Figure 3. The vagus nerve
For more than 100 years, truncal and partial vagotomy procedures have been performed as a treatment for peptic ulcer disease prior to the introduction of H2 receptor-blocking medications. Along with the reduction of gastric acid, other physiologic changes were observed, one of which was loss of the gastric accommodation response to eating that regulates appetite and satiety. Vagotomized individuals also experience early satiation and decreased food intake. Furthermore, subjects experience inhibition of gastric contractions with delayed gastric emptying, resulting in the feeling of fullness and bloating. Chang and colleagues reported a series of 120 patients who underwent either complete or highly selective (partial) intra-abdominal truncal vagotomy, along with duodenectomy for refractory duodenal ulcers.21 The treatment goal was to reduce gastric acid secretion to enable the patient to eat without ulcer pain, thereby increasing their body weight back to its normal level. Interestingly, less than 50 percent of patients with complete vagotomy achieved their desired body weight gain, while 94 percent of those subjects with minimal vagal nerve interruption gained weight. This finding suggested that blocking vagus nerve function may inhibit weight gain.
Intestinal nutrient absorption also may be affected by vagotomy. Patients were found to have decreased fat absorption after complete vagotomy as compared with nonvagotomized patients or patients after undergoing highly selected vagotomies.22
Gortz and colleagues discovered that truncal vagotomy had an appetite-suppressant effect. They observed in seven morbidly obese patients that truncal vagotomy resulted in reductions in body weight and caloric intake (from 2,800 kcal/day to 1,800 kcal/day in three months, and to 2,000 kcal/day at nine months after vagotomy). Interestingly, liquid calorie intake was reduced more than solid food intake. By nine months, liquid calorie intake was reduced by 50 percent, while solid calorie intake was only reduced by 27 percent.23
Based on the growing knowledge that the vagus nerve plays a significant role in weight regulation, energy intake, and appetite, surgical vagotomy was investigated as a less-invasive therapy for treating morbid obesity than conventional bariatric surgery.
In a group of morbidly obese patients, Kral and colleagues performed either a gastric partitioning procedure (vertical banded gastroplasty [VBG]) or the same gastric partitioning procedure with a truncal vagotomy. After one year of follow-up, the patients who had the vagotomies had a weight loss of 51 percent versus only 34 percent excess weight loss for those patients who only had the VBG (p <0.01). After 83 months of follow-up, the difference was even greater—61 percent versus 28 percent excess weight loss (p <0.001). This study and several more by Kral and others created great interest in vagotomy alone or vagotomy along with gastric partitioning as the long-awaited breakthrough in the treatment of morbid obesity.24,25
Based on these and other findings, many surgeons added truncal vagotomy to conventional bariatric surgery, such as the gastric bypass and the VBG. However, the long-term benefits were inconsistent. Angrisani and colleagues found that the addition of a truncal vagotomy to the placement of a laparoscopic adjustable gastric band did not improve weight loss compared with just gastric banding at 12 or 18 months of follow-up. However, they also found that significantly more patients who underwent banding and vagotomy did not require band adjustments at six and 12 months as compared with the cohort of patients that underwent banding without vagotomy (50 percent and 35 percent, respectively, versus 20 percent and 8 percent, respectively). The differences at both endpoints were statistically significant (p = 0.034, p = 0.024, respectively).26
In an open-label, case-controlled study, Martin and colleagues compared a cohort of patients who underwent laparoscopic truncal vagotomy and adjustable gastric banding with a matched cohort that only underwent laparoscopic adjustable gastric banding. At a mean follow-up of more than 30 months, all of the vagotomy patients reported an absence of hunger, but the difference in weight loss was statistically insignificant (38 percent versus 36 percent of excess weight, respectively).27
Okafor and colleagues investigated the effects of vagotomy on patients undergoing Roux-en-Y gastric bypass surgery. They retrospectively reviewed the outcomes of 1,278 patients who underwent gastric bypass; 40 percent of these patients also underwent concomitant vagotomies. They, too, reported that vagotomy had no effect on the percentage of excess weight loss.28 Finally, Liu and colleagues observed that truncal vagotomy did not seem to improve glucose metabolism when added to sleeve gastrectomy in a cohort of diabetic rats.29
The inconsistent results observed for truncal vagotomy as a standalone or adjunct weight-loss therapy could suggest that a truncal vagotomy, which is permanent, may result in tolerance or even the activation of compensatory mechanisms that would overpower the effects of the vagotomy. Therefore, the ability to intermittently block vagal nerve signaling for a duration of time and then allow full recovery of vagal function might achieve the desired outcomes without the loss of efficacy over time, as tolerance or compensatory mechanisms may not develop.
Because surgical vagotomy was irreversible, a device that could be implanted and reversibly block vagal nerve signaling (see Figure 4) was developed. The vBloc system, also known as the Maestro System, comprises a programmable electrical pulse generator known as a neuroregulator, with leads and electrodes. The electrodes are laparoscopically secured around the vagal nerve trunks at the gastro-esophageal junction. The electrodes are attached to the leads, which are connected to the neuroregulator, which is implanted under the skin on the lower chest wall. The procedure is relatively brief and several studies have demonstrated that it has fewer perioperative complications than nearly all other bariatric surgical procedures.30-36
Figure 4. The vBloc System
When activated, the vBloc neuroregulator generates intermittent, high-frequency electrical algorithms that are applied by the small electrodes directly onto the intra-abdominal vagus nerve trunks. These algorithms reversibly block compound action-potential transmission, resulting in reduced appetite, food intake, and body weight.30-36
A number of animal and human studies validated the vBloc therapy.37-40 Subsequently, more than 600 human patients with morbid obesity were implanted and participated in several investigations. The first two human trials were the vBloc-RF1 study, a proof of concept trial, and vBloc-RF2 study, a safety and efficacy trial. Because these studies were feasibility investigations, this review focuses on the subsequent, clinically meaningful trials.
The Eliminating Medications through Patient Ownership of End Results (EMPOWER) trial was a prospective, randomized, double-blinded, sham-controlled investigation, which included 294 patients from 13 study centers in the U.S. and two patients in Australia who were laparoscopically implanted with the Maestro System. The neuroregulator used in this study was active only when the subject wore an external power source attached to a belt. Therefore, the subjects controlled the duration of therapy by how many hours they wore the power supply belt. They were instructed to use the external electrical power source for at least nine hours daily.30
After implantation, the subjects were randomized, with 192 subjects assigned to the treatment group and 102 to the control group. The primary endpoints of the trial were the percentage of excess weight loss and serious adverse events (also known as SAEs) at 12 months.
The vBloc therapy demonstrated itself to be very safe in this study, with no deaths or serious complications. After the 12 months of follow-up, comparable excess weight loss was achieved in both groups. Treatment subjects achieved a 17 ± 2 percent excess weight loss in contrast to 16 ± 2 percent in the control group. Post hoc analysis suggested that weight loss was correlated with hours of use, and almost half of the treatment group subjects used the therapy for less than nine hours daily. The post hoc analysis also revealed that the safety checks performed by the neuroregulators in the control group subjects delivered a minimal amount of electricity to the vagal trunks. This amount of electrical energy, although less than 1/1,000th of the energy delivered to the treatment group, still may have provided therapeutic effects.
Subsequent research with rat sciatic nerve demonstrated that the safety algorithm used for the control patients resulted in a 31 percent reduction in the amplitude of the action potential of the sciatic nerves. Hence, it is likely that the subjects in the control group may have, in fact, received some level of vagal nerve impulse blockade that may have led to the unexpected weight loss in the control subjects.
Although the weight loss achieved in the EMPOWER trial was moderate, it was accompanied by medically significant metabolic benefits. Hypertensive subjects demonstrated statistically significant reductions in systolic and diastolic pressures. The EMPOWER trial created as many questions as it answered, and a true analysis of reversible vagal blocking was yet to be achieved.
Ikramuddin and colleagues reported the results of a subsequent study, the ReCharge trial, which was designed to correct the earlier trials’ unanticipated design flaws. A next-generation neuroregulator with a rechargeable internal battery was used to reduce patient noncompliance. In addition, all control patients were treated as if they had electrodes implanted laparoscopically, even though no leads were implanted to eliminate any possibility of partial blockade.
The ReCharge trial was a randomized, double-blinded, multicenter, sham-control investigation, with 162 subjects in the vBloc group and 77 in the control group. The one-year results of the ReCharge trial were mixed. At 12 months, vBloc therapy again proved to be safe. Serious adverse events occurred in only 3.7 percent (95 percent CI, 1.4 to 7.9 percent, p <0.001). No deaths occurred. Although the difference in mean excess weight loss between the vBloc and sham-control subjects was statistically significant (24.4 percent versus 15.9 percent, p = 0.002), the mean difference did not achieve the prespecified superiority of ≥10 percent that was a primary endpoint in the protocol. The sham control group outperformed the prestudy estimation of 5 percent excess weight loss. However, at 18 and 24 months of follow-up, the vBloc group’s weight loss was essentially stable at 21 percent of excess weight, whereas the control group patients’ mean excess weight loss decreased by 40 percent, which was notably only 3.8 percent of excess weight by 18 months.33,41
As in the EMPOWER trial, beneficial metabolic effects were observed in the ReCharge trial. Subjects with abnormal preoperative cardiovascular risk factors demonstrated significant improvements in mean LDL cholesterol (−16 mg/dL), HDL (+4 mg/dL), triglycerides (−46 mg/dL), HbA1C (−0.3 %), and systolic (−11 mm Hg) and diastolic blood pressures (−10 mm Hg). Approximately 50 percent of patients who preoperatively met the criteria for metabolic syndrome or who were classified as pre-diabetic had normal values at both 12 and 24 months. Improved quality-of-life measures seen at 12 months also were sustained at 24 months.41
The vBloc therapy also was evaluated in a small group of morbidly obese patients with T2D mellitus (T2DM). A total of 28 subjects with T2DM and BMI 30–40 kg/m2 had the vBloc device implanted and were prospectively followed. At one year, mean excess weight loss was 24 percent and remained stable at both two and three years, 22 percent and 21 percent of excess, respectively. The mean baseline HbA1C was 7.8 percent. At one year, it decreased 1 percent (p<0.0001), and at two and three years, the HbA1C was still reduced by 0.6 percent (p = 0.0026). Fasting plasma glucose dropped significantly from a mean of 151 mg/dL at baseline to 123 mg/dL at 12 months (p = 0.0003), 136 mg/dL at 24 months (p = 0.0564), and 133 mg/dL at 36 months.31,42,43
Deep brain stimulation (DBS) as a treatment for neurodegenerative conditions such as Parkinson’s disease, multiple sclerosis, essential tremor, and amyotrophic lateral sclerosis has been found to achieve weight loss. However, few animal and human trials have been conducted. Melega and colleagues investigated the effects of DBS of the ventromedial hypothalamus in overweight mini pigs. They discovered that the mini pigs subjected to DBS gained significantly less weight compared with a control group, 6.1 ± 0.4 kg vs 9.4 ± 1.3 kg (p <0.05). However, no significant differences in fasting morning blood glucose levels were detected.44
DBS also may influence blood sugar regulation. In a group of rats rendered diabetic by the administration of alloxan, transcranial DBS enabled those animals to maintain normal serum blood glucose levels compared with a control group that was rendered hyperglycemic by alloxan but did not receive DBS. Furthermore, the study results suggested that DBS may have normalized the functional state of damaged β-cells and their proliferation.45
DBS targeting different regions of the brain will likely elicit different results.46 In contrast to Lebedev and colleagues, who reported improved blood glucose homeostasis in diabetic rats with transcranial DBS, Diepenbroek and colleagues reported that DBS of the nucleus accumbens actually increased blood glucose levels. Furthermore, they discovered that a 10 uA output had no effect, but an increased output of 200 uA did increase blood glucose levels.47
Patients with Parkinson’s disease progressively lose weight over time. In a case control study, Strowd and colleagues reported that patients with Parkinson’s undergoing DBS of the sub-thalamic nucleus experienced weight gain, while control patients demonstrated weight loss. At 21.3 months follow-up, the DBS patients gained a mean of 2.92 ± 9.4 kg while control patients lost 1.82 ± 6.9 kg. The mean weight difference of 1.1 ± 2.5 kg was statistically significant (p <0.02).48
In 2002, Burneo and colleagues unexpectedly observed that vagal nerve stimulation in patients who suffered from intractable epilepsy resulted in weight loss; 17 of the 27 patients lost weight. Of the patients who lost weight, eight lost more than 5 percent of their body weight, and five patients lost more than 10 percent. Furthermore, the weight loss seemed to correlate with increased stimulator output current.49
Although neuromodulatory therapies have been under investigation for more than 20 years, they are far from reaching their full potential. Many questions remain that prevent these technologies from becoming more commonly applied in humans. A significant knowledge gap persists in understanding their mechanisms of action, optimal therapy targets, patient selection, and markers of efficacy (such as neuropeptides, gastric distention, electrocardiogram changes, and symptoms). These devices can be programmed in an infinite number of ways. For example, a device can be set to continuous output or triggered output. The voltage, current, pulse frequency, pulse width, pulse train length, and total energy delivered can be adjusted to effect. Although the most effective way to maximize weight loss remains to be determined, these interventions have been associated with an extraordinarily good safety profile. Given the high prevalence of obesity throughout the world, neuromodulation as a treatment modality likely will continue to develop, improve in efficacy, and become more commonly used in clinical practice.
Awareness of the concept of metabolic surgery and its benefits in the treatment of metabolic disorders, such as T2DM, morbid obesity, and hyperlipidemia, is growing. The nervous system has significant control over many metabolic pathways including body weight, energy intake, and serum glucose control. It is readily perceptible that neurologic metabolic therapies as highlighted in this overview have potential to benefit patients with metabolic diseases.
This work was supported by the American College of Surgeons (ACS). The authors declare that they have no relevant conflict of interest.
We are grateful to the ACS for their generous sponsorship of the Metabolic Surgery Symposium and associated journal publication development. We thank Jane N. Buchwald, Chief Scientific Research Writer, Medwrite Medical Communications, Maiden Rock, WI, for manuscript editing and publication coordination. And we thank Patrick Beebe and Donna Coulombe, ACS Executive Services, for their expert organization of the Metabolic Surgery Symposium.