Electrical stimulation for gastroparesis

INTRODUCTION — Gastroparesis is defined as a delay in gastric emptying with associated nausea, vomiting, bloating, early satiety, and discomfort. In severe cases, nausea and vomiting may cause weight loss, dehydration, electrolyte disturbances, and malnutrition due to inadequate caloric and fluid intake.

Management of gastroparesis includes supportive measures (eg, hydration and nutrition), optimizing glycemic control in patients with diabetes mellitus, medications, and occasionally pyloric or surgical therapy. The limited efficacy of these options for severe gastroparesis has provided a rationale for development of novel approaches for treatment. Electrical stimuli can be delivered to the stomach as low-energy, high-frequency gastric electrical neurostimulation or high-energy, low-frequency gastric pacing. The former has emerging evidence of efficacy and is available for humanitarian treatment of refractory gastroparesis, while the latter is too bulky for implantation and is not clinically available.

This topic review will provide an overview of the methods of electrical stimulation of and their efficacy in treating gastroparesis. The pathophysiology, etiology, diagnosis, and treatment of gastroparesis are discussed separately. (See "Gastroparesis: Etiology, clinical manifestations, and diagnosis" and "Pathogenesis of delayed gastric emptying" and "Treatment of gastroparesis".)

PATHOPHYSIOLOGY — The emptying of gastric contents after a meal is controlled by motor and myoelectric activities of different gastric regions.

●The proximal stomach exhibits changes in tone after eating, which initially accommodates the ingested bolus and then, as digestion progresses, regulates delivery of food particles into the distal stomach [1].

●The distal stomach exhibits a fed motor pattern consisting of phasic contractions that propagate from the gastric body to the pylorus at a maximal frequency of three cycles per minute (cpm). These grind and mix the food into a fine suspension [2].

Impairment of normal phasic motor activity in the distal stomach produces gastroparesis. The frequency and direction of this activity is regulated by the gastric slow wave, a rhythmic electrical oscillation, which is generated by interstitial cells of Cajal in the proximal gastric body, the "pacemaker" zone of the stomach [3]. The slow wave is ubiquitously present at a frequency of 3 cpm, regardless of the contractile state of the stomach. Phasic contractions are generated when the slow wave plateau potential increases in amplitude or when action potentials are stimulated by meal-induced neurohumoral activators [4].

CANDIDATES FOR GASTRIC ELECTRICAL STIMULATION

Indications

Refractory idiopathic and diabetic gastroparesis — The gastric electrical neurostimulator (Enterra Therapy system) is not approved by the US Food and Drug Administration (FDA) for unrestricted marketing for treatment of gastroparesis but is approved as a humanitarian use device. Because of probable benefit rather than established effectiveness, the device has also received humanitarian device exemption approval for treatment of refractory diabetic and idiopathic gastroparesis, documented by objective measures of delayed gastric emptying. (See "Gastroparesis: Etiology, clinical manifestations, and diagnosis", section on 'Scintigraphic gastric emptying'.)

Other — Other potential compassionate use applications of gastric stimulation that have been evaluated in small series of patients with refractory symptoms include:

●Gastroparesis due to other causes – Patients with refractory gastroparesis from other causes including postsurgical etiologies (after fundoplication for acid reflux, esophagectomy for esophageal carcinoma, bariatric surgery) and gastroparesis secondary to malignancy, chronic intestinal pseudo-obstruction, Crohn disease, or transplantation have undergone gastric stimulation.

●Refractory cyclic vomiting syndrome – In one study of 11 patients with refractory cyclic vomiting syndrome, permanent gastric electrical stimulation was associated with a decrease in nausea and vomiting of 46 and 69 percent, respectively, compared with baseline [5].

Relative contraindications — There are no absolute contraindications to gastric neurostimulation. Due to the risk of postoperative infection, patients on immunosuppressive therapy may not be optimal candidates. Chronic opioid use adversely impacts the therapeutic response to gastric stimulation and should be considered when referring opioid-dependent patients for gastric stimulation [6]. Gastric neurostimulation should be avoided during pregnancy as the effects of stimulation on the developing fetus are unknown.

Predictors of response — Patients with diabetes are more likely than those with idiopathic disease to exhibit clinical improvement following stimulator surgery and show greater likelihood to remain on gastric stimulation long-term [7-10]. In a meta-analysis of 49 studies, clinical improvements with gastric stimulation were greater in patients with more severe initial symptoms [11].

Rates of gastric emptying generally do not predict responses to gastric stimulation. However, some studies noted better responses in those with more prolonged retention, while another found that more severe emptying delays predicted the need to perform stimulator implantation in diabetics but not idiopathic patients [8,12,13].

Damage to myenteric ganglia and loss of interstitial cells of Cajal in the stomach have been reported as gastric histologic predictors of response to gastric stimulation [14,15]. Furthermore, the presence of gastric smooth muscle fibrosis is predictive of a worse response to gastric stimulation [14].

HIGH-FREQUENCY GASTRIC ELECTRICAL STIMULATION

Mechanism of action — The mechanism of action of high frequency (12 cpm) gastric neurostimulation may be multifactorial, but probably does not relate to stimulating gastric emptying as the stimuli are of insufficient amplitude to generate antral contractions. Consistent acceleration of emptying has not been observed in clinical trials [16]. Gastric neurostimulation also has no effect on basal gastric electrical activity and does not reverse slow wave dysrhythmias, but it can enhance slow-wave amplitude and propagation velocity [17,18]. Intraoperative high resolution measurements of gastric electrical conduction profiles have shown no effects of gastric neurostimulation on slow wave dysrhythmias, conduction blocks, retrograde propagation, ectopic pacemakers, or colliding waveforms [19]. The device increases maximally tolerated volumes of gastric distention, reflecting blunting of luminal perception [20]. Gastric neurostimulation modifies sympathovagal activity and modulates activity in thoracic spinal neurons that are responsive to gastric distention, however, autonomic benefits only become evident months after implantation, suggesting they are not critical for symptom improvements [21-23].

In animal models, gastric neurostimulation increases ghrelin-positive cells and plasma ghrelin levels, suggesting possible mechanistic participation of this neurohumoral agent [24]. Reductions in serum tumor necrosis factor-alpha levels after surgery may indicate additional anti-inflammatory effects of gastric neurostimulation [25]. In diabetic rats, gastric stimulation enhances interstitial cell of Cajal proliferation and increases enteric nerve populations responsible for nitric oxide and acetylcholine transmission [26].

Technique

Permanent stimulator — The gastric neurostimulator can be implanted via laparoscopy or laparotomy. The device consists of a pair of leads, a pulse generator, and a programming system. The leads are placed in the muscularis propria of the greater curvature of the stomach, 10 cm proximal to the pylorus and are connected to a pulse generator. The pulse generator is typically placed subcutaneously in the right or left upper quadrants of the abdomen. An external programming device controls the stimulation parameters. The battery life is typically 5 to 10 years, but this duration can vary depending on the energy level settings.

Temporary stimulator — Endoscopically placed temporary stimulating electrodes are employed by some centers of expertise to help predict who might respond to a permanently implanted device [27].

Benefits

Clinical outcomes in uncontrolled reports — Response rates to gastric neurostimulation in uncontrolled studies range from 50 to 92 percent, which can persist for up to 15 years [28-30]. In a systematic review of 19 studies of gastric stimulation for gastroparesis, nausea and vomiting were more likely to improve than abdominal pain [6,31,32]. Symptomatic improvement has been reported as soon as three days after device implantation [5,6,33-39]. In an analysis of a large multicenter cohort using propensity score methods, nausea improved to greater degrees among 81 patients who underwent gastric stimulator surgery versus 238 patients who did not [40]. In contrast, gastric stimulation has not consistently demonstrated improvements in other gastrointestinal symptoms including fullness, bloating, or acid reflux symptoms [6,9].

Additional studies have reported associated improvements with gastric stimulation in body mass index, HbA1c, serum albumin, and reduction in the need for prokinetic medication and supplemental nutrition [34,37,41,42]. Gastric stimulation has also been associated with improvements in physical and mental quality-of-life scores and reduction in hospitalizations [34,36,41]. (See "Gastroparesis: Etiology, clinical manifestations, and diagnosis", section on 'Clinical features'.)

Efficacy compared with other treatments — A systematic review of 38 studies reported smaller improvements in nausea and abdominal pain with gastric neurostimulation compared with pyloromyotomy or pyloroplasty and lesser antiemetic effects than observed after pyloric surgery or gastrectomy [31]. Similarly, responses to gastric stimulation were less than with gastric peroral endoscopic myotomy of the pylorus in a study [43]. However, another paper observed comparable benefits of gastric stimulation and pyloric therapies in 82 patients with refractory gastroparesis [44]. Superior long-term outcomes were seen after total gastrectomy versus gastric stimulation in one study, but this outcome may have been influenced by a higher prevalence of postsurgical gastroparesis in the gastrectomy group [45].

Sham-controlled trials of gastric stimulation — In three initial randomized crossover trials, there was only marginal benefit of gastric stimulation compared to sham stimulation treatment arms where no electrical current was delivered [46-48]. However, a sham-controlled study observed significant symptom decreases with active stimulation, which were associated with reduced health care utilization [49,50].

●The first suggested that there was benefit only in diabetic patients with active gastric stimulation versus sham treatment, while patients with idiopathic benefits showed no symptom reductions [46].

●The second study in diabetic gastroparesis showed no difference in symptoms during the blinded treatment phase whether the device was turned on or off [47].

●A third controlled study in idiopathic gastroparesis patients had a similar trial design as the investigation in diabetics and results paralleled the diabetic study, showing no significant differences in symptoms in the blinded crossover phase regardless of whether the device was on or off [48].

●In a multicenter randomized trial in which 149 patients with gastroparesis were assigned to on (active) versus off (sham) stimulation for four months with crossover to the other arm for an additional four months, vomiting scores were significantly improved with the device on versus off in both diabetic and nondiabetic patients [49]. However, no other symptom benefits were reported in this study. In additional analyses from this trial, active stimulation was associated with reductions in hospitalizations, lower health care costs, and less work absenteeism [50].

Risks — Risks of high-frequency gastric neurostimulation include infection, lead migration or erosion or dislodgement, electrode penetration into the gastric mucosa, seroma, and bowel obstruction.

Reasons for device explantation in one series included nonresponse to therapy (4 percent), mechanical issues (3 percent), and infection (2 percent) [51]. Battery replacement is often required within 10 years of initial device implantation [28]. Forty-three percent of patients in one longitudinal study required an average of 2.15 additional surgeries over eight years of follow-up, mostly for battery exchange and device relocation [30].

Management of treatment failures — In patients who fail to respond to gastric stimulation or who develop recurrent symptoms after surgery, additional treatment options include pyloric myotomy or replacement of the gastric stimulator.  

●Pyloric myotomy – In patients with a poor response to gastric neurostimulation, addition of pyloroplasty may improve symptoms. In one study, addition of pyloroplasty to gastric stimulation promoted greater acceleration of gastric emptying compared with stimulator implantation alone with greater symptom improvements [52].

●Replacement of gastric electrical stimulator – In patients initially unresponsive to gastric electrical stimulation, implanting a replacement stimulator with positioning of the new electrodes to alternate locations on the gastric serosa has been associated with reduced symptoms [53].

OTHER INVESTIGATIONAL MODES OF ELECTRICAL STIMULATION

Newer experimental methods of gastric neurostimulation — New approaches to gastric neurostimulation are in development. Progress in nonoperative methods of electrode insertion, device miniaturization, and battery technology may expand the options available for these patients. A percutaneous electrode system was developed to deliver neurostimulating pulses for up to eight weeks in patients with gastroparesis [54]. With this technique, a cannula with an internal needle is introduced percutaneously and advanced to the gastric submucosa. A self-anchoring electrode is then placed through the needle. Electrodes that can be placed during percutaneous gastrostomy placement have also been devised [55].

High-energy, low-frequency gastric pacing — Gastric pacing differs from gastric neurostimulation in that it resets the regular slow wave rhythm by delivering low duration electrical pulses (30 to 500 ms) and promotes increased gastric contractions with a consequence of accelerating gastric emptying. In an open-label trial of gastric pacing, nine gastroparesis patients received high-energy, long-duration electrical pulses at a rate slightly higher than the normal slow wave frequency through surgically implanted electrodes [56]. When delivered before and after meals, the pacing stimuli entrained the slow wave in all individuals and corrected any underlying slow wave rhythm disturbances. After one month, gastroparetic symptoms decreased, eight patients no longer required jejunal tube feedings, and gastric emptying was enhanced. Multichannel stimulation protocols have been devised to deliver high-energy pulses in sequential fashion that evoke propagating contractions and enhance gastric emptying with lower current requirements [57]. Two-channel pacing with reduced current requirements has shown efficacy in stimulating gastric emptying and reducing symptoms in patients with severe diabetic gastroparesis [58]. A feedback-controlled mechanism that can turn the pacer on or off depending on the intrinsic contractile state of the stomach has been developed [59]. Such devices that are activated only when triggered may reduce current requirements and prolong battery life. Despite these advances, gastric pacing remains impractical because the external current source needed to generate the energy to entrain the slow wave is too large for subcutaneous implantation.

Research into newer gastric pacing methods is focusing on portable implantable devices that do not require permanent externally-wired connections. Endoscopically-delivered miniaturized stimulators that can be affixed to or implanted in the gastric mucosa have been developed that enhance slow wave regularity and amplitudes [60,61]. An investigational battery-powered gastric pacemaker combined with high-resolution electrical mapping, which is activated in wireless fashion by a mobile phone or computer was shown to entrain slow waves in a porcine model [62]. Other methods that have been shown to entrain and enhance slow wave activity include a leadless inductively powered pacemaker implanted in the gastric serosa in a pig model, which is activated by an external transmitter, and a similar implantable magnet system in the gastric wall with an external drive coil [63,64].

Another area of investigation relates to devices that deliver dual stimuli, which elicit beneficial effects both on symptoms and gastric emptying by combining effective features of both neurostimulation and pacing. In canine studies, stimulation with low-energy pulses of short duration alternating with high-energy pulses of long duration exerted antiemetic effects and reversed slow wave dysrhythmias caused by the emetic stimulus vasopressin [65].

Extragastric electrical stimulation — Studies suggest that electrical stimulation at sites distant from the stomach may provide symptom benefits in gastroparesis. Transcutaneous vagal nerve stimulators, like those used for migraine treatment, have improved symptoms in some uncontrolled gastroparesis studies [66]. A noninvasive thoracic spinal nerve magnetic neuromodulator treatment reduced symptoms in seven patients with diabetic gastroparesis in a proof-of-concept report [67].  

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Gastroparesis".)

SUMMARY AND RECOMMENDATIONS

●Electrical stimuli can be delivered to the stomach as low-energy, high-frequency gastric neurostimulation or high-energy, low-frequency gastric pacing. (See 'Introduction' above.)

●Gastric neurostimulation has no effect on basal gastric emptying or electrical activity but modifies sympathovagal activity and perception of gastric distention. Uncontrolled studies suggest clinically important benefits of gastric neurostimulation, including improvements in nausea and vomiting associated with gastroparesis. Emerging evidence from a multicenter crossover trial suggests there may be benefit from active versus sham stimulation. (See 'Mechanism of action' above and 'Benefits' above.)

●Patients with severe nausea and vomiting refractory to antiemetic and prokinetic drug therapy for at least one year may be candidates for gastric stimulation. In the United States, the gastric neurostimulator (Enterra Therapy system) is approved as a humanitarian exemption device only for diabetic and idiopathic gastroparesis. (See 'Indications' above.)

●The benefits of gastric neurostimulation for symptoms other than nausea and vomiting are uncertain. Patients with nausea and vomiting and those without narcotic dependence have more favorable clinical responses than those with predominant abdominal pain, bloating, or fullness. Patients with diabetic gastroparesis have greater symptom reductions versus individuals with idiopathic or postsurgical disease. (See 'Predictors of response' above.)

●Risks of gastric neurostimulation include infection, lead migration or erosion, lead dislodgement, electrode penetration into the gastric mucosa requiring re-operation, seroma, and bowel obstruction. (See 'Risks' above.)

●Progress in device miniaturization, battery technology, devices that deliver variable stimuli, and remote leadless delivery may offer greater options for electrical neurostimulation and pacing of the stomach for refractory gastroparesis in the future. Studies describing symptom reductions with vagal or spinal stimulation methods suggest potential benefits of extragastric electrical protocols as well. (See 'Other investigational modes of electrical stimulation' above.)