Endoscopic and Surgical Treatments for Gastroparesis: What to do and Whom to treat?

Abstract

Gastroparesis is a complex chronic debilitating condition of gastric motility resulting in the delayed gastric emptying and multiple severe symptoms which may lead to malnutrition and dehydration. Initial management of patients with gastroparesis focuses on the diet and lifestyle modification and medical therapy. Various endoscopic and surgical interventions are reserved for refractory cases of gastroparesis not responding conservative therapy. Pyloric interventions, enteral access tubes, gastric electric stimulator and gastrectomy have been described in the care of patients with gastroparesis. In this article, the authors review current management, indications and contraindications to these procedures. Authors introduce an algorithm to aid in the decision making in the care of these patients. In short, endoscopic injection of botulinum toxin into the pylorus is regarded as a temporizing procedure and can be used as a bridge until more definitive care. Patients with severe nausea and vomiting undergo gastric electric stimulator implantation, which can be combined with pyloroplasty in cases of severe delay in gastric emptying. Without severe nausea pyloric intervention is often the procedure of choice. In patients who remain refractory to treatment, placement of a feeding jejunostomy and/or a venting gastrostomy may be considered. TPN, if used, should be limited and temporary because of the risk profile. Surgical gastrectomy should generally be discouraged and reserved for selected patients with postsurgical gastroparesis.

Introduction

Gastroparesis is a complex chronic debilitating condition of gastric motility resulting in the delayed gastric emptying and symptoms of nausea, vomiting, early satiety, postprandial fullness, and abdominal pain, and GERD often leading to malnutrition and dehydration. Initial management of patients with gastroparesis focuses on the diet and lifestyle modification and medical therapy. In those who fail conservative management, there is no uniform way to handle further therapy. Patients with refractory gastroparesis, not responding to standard antiemetic and prokinetic agents or with side effects preventing their use need more complex and advanced management options. In one study of 110 patients with “refractory” gastroparesis, 74% responded to use of another prokinetic agent and only 26% remained refractory to all prokinetic agents. These truly refractory patients underwent enteral or parenteral feedings and gastric electrical stimulation, and some underwent gastrectomy. Poor responders to prokinetic agents include postgastrectomy patients, those with myopathic connective tissue disorders, type I diabetic patients with severely delayed gastric emptying, and patients with idiopathic gastroparesis with abdominal pain (1).

An initial approach to refractory gastroparesis includes assessment of the severity of symptoms, the degree of delay of the gastric emptying, optimizing the patient’s current therapy, and changing prokinetic agents (2). Why some patients respond to one prokinetic agent and not another is uncertain. Different prokinetic agents have different mechanisms of action and the efficacy of most prokinetic agents diminishes with prolonged use. A “drug holiday,” as used for L-dopa in Parkinson disease, can be used for prokinetic agents, especially erythromycin.

In patients who remain refractory to treatment, one may consider placement of a feeding jejunostomy and/or a venting gastrostomy. Other therapies might include injection of botulinum toxin into the pylorus, pyloromyotomy and / or gastric electrical stimulation. TPN, if used, should be limited and temporary because of the risk profile. Surgical gastrectomy should generally be discouraged. Completion gastrectomy can be considered for selected patients with postsurgical gastroparesis.

Endoscopic Botox pyloric therapy.

Botulinum toxin is a potent inhibitor of neuromuscular transmission and has been used to treat spastic muscular disorders by local injection. Thus, endoscopic injection of botulinum toxin directly into the lower esophageal sphincter reduces pressure and improves symptoms in patients with achalasia (3). Extrapolating success from the achalasia experience, several small case series of botulinum toxin injection in the pylorus demonstrated mild improvement in gastric emptying and a modest improvement in symptoms (4–6). A subsequent, much larger case series of 63 patients with idiopathic and diabetic gastroparesis demonstrated clinical response in 43% of patients with a median duration of 5 months. Notably, vomiting was associated with a lack of response to the Botox therapy (7). In the large retrospective series of 179 patients with gastroparesis, clinical response was associated with a larger dose of the medication. Such, almost 77% of patient had a response to 200 Units of the toxin, whereas only 54% responded to 100 Units (8). Results from placebo controlled trials have shown that whereas botulinum toxin injection into the pylorus may mildly improve gastric emptying, there is little difference in symptom improvement with botulinum toxin compared with placebo at 1 month (9, 10). Thus, botulinum toxin injection into the pylorus does not appear a viable long-term treatment option for patients with gastroparesis. Insurance coverage for botulinum toxin has become harder to obtain for gastroparesis, although it is often covered for achalasia and lower esophageal sphincter injection. In select patients, botulinum pyloric injections may temporize a patient for several months. It may also be used to assess response prior to pyloromyotomy: a positive clinical response to pyloric Botox injection may suggest better response to pyloromyotomy (11).

Gastric peroral endoscopic myotomy (GPOEM) / Peroral pyloromyotomy (POP).

Following the success of peroral endoscopic myotomy (POEM) for the treatment of achalasia, a similar procedure of the pyloric sphincter was introduced into clinical practice and reported in 2013 by Khashab et al. (12). In the care of a 27 years old female type one diabetic, this new experimental procedure, named GPOEM, was performed after failure of medical therapy and previous positive response to the pyloric stenting. Patient has responded favorably to this procedure and remained well at least 12 weeks after the intervention (12). In a subsequent multicenter study of 30 patients with idiopathic, diabetic and postsurgical gastroparesis authors observed clinical response in 86% of patients at the median follow up of 5.5 months. Two patients had developed adverse events – capnoperitoneum and prepyloric ulcer (13).

Due to its success (procedurally and clinically), the procedure has been widely adopted worldwide and emerging multiple reports continue to reflect positively on the outcomes. In the review article by Mekaroonkamol et al. response rate to peroral pyloromyotomy from 12 published series ranged from 73 to 100 % (14). A recent systematic review of 14 studies using G-POEM demonstrated pooled clinical symptom improvement rate of 88% with an intraoperative complication rate of 3%, thus demonstrating high efficacy with marginal risk (15, 16). In the authors’ experience, positive response to GPOEM procedure was more common in patients also responding to Botox injection previously (11).

The technique of the procedure employs principles of the submucosal endoscopy, developed in the original POEM procedure (Fig. 1). Steps of the procedure include initial mucosal entry, submucosal tunneling, myotomy and closure of the mucosotomy. GPOEM is technically more challenging procedure compared to POEM due to looping of the scope in the stomach, curving of the tunnel over the pyloric muscle, field instability due to antral contraction and difficulties in anatomic landmarking (14). There is no consensus and standardization of the procedure regarding location, length and depth of the myotomy. Initially procedure was more commonly performed on the anterior wall or greater curvature of the stomach. Location of the procedure on the lesser curvature was reported to have an advantage of the shorter scope length and potentially reduced looping in the stomach, shorter length of the tunnel, lesser probability of the tunnel deviation and not dependent position with avoidance of food residue, secretions and blood pooling (17). Initial mucosal entry is performed after submucosal dyed saline injection and elevation of the mucosa. Depending on the planned closure type, mucosotomy is either performed longitudinally for clips closure or transversely for the Overstitch device closure (18). After tunnel is created over the ridge of the pyloric sphincter, myotomy is performed to the subserosal layer. Although previously reported full thickness myotomy is not routinely advocated, accidental full thickness division is not considered a perforation and rarely requires additional attention other than a careful closure of the mucosotomy over the tunnel. Authors prefer closure with the suture for better seal of the mucosal entry.

Figures 1.

Steps of the GPOEM procedure. A. Initial mucosal entry – 5 cm proximal to the pylorus on the greater curvature. B. Dissection of the submucosal entry. C. Myotomy. D. Closure of the mucosotomy with the Overstitch device.

Surgical pyloromyotomy and pyloroplasty.

Pyloromyotomy and pyloroplasty are well established surgical procedures that have been applied for various indications for decades. With the minimally invasive revolution since the late 1980s, open procedures have been replaced by laparoscopic and more recently robotic interventions. Robotic technology provides better visualization of the stomach layers and better dexterity due to wrist motion and allows more precise dissection and preservation of the mucosal layer (19, 20). In the prospective survey of the 177 patients undergoing laparoscopic pyloroplasty as a sole treatment for the gastroparesis or in combination with antireflux surgery, 86% in total experienced improvement with 77% - complete normalization of the gastric emptying (21). Nineteen patients (11%) required subsequent surgical interventions: gastric stimulator implantation (7%), feeding jejunostomy and/or gastrostomy tube (3%), or subtotal gastrectomy (2%). Symptom severity scores for nausea, vomiting, bloating, abdominal pain, and early satiety decreased significantly at 3 months (21). In another report of 50 patients undergoing laparoscopic pyloroplasty for the treatment of gastroparesis, symptom improvement was observed in 82% of patients. Five patients (10%) underwent subsequent surgical interventions - gastrectomy (4%), duodenojejunostomy (4%) or gastric stimulator implantation (2%) (22). Similarly, Hibbard at al. in their analysis of 26 patients undergoing laparoscopic pyloroplasty for gastroparesis, reported normalization of gastric emptying in 71% with persistent significant improvement in symptoms nausea, vomiting, bloating, abdominal pain and GERD (23). In a systematic review of pyloric interventions, gastric stimulator implantation and gastrectomy authors demonstrated overall greater response to pyloric intervention in refractory gastroparesis patients with the best outcomes in nausea and abdominal pain (24).

Technique of laparoscopic pyloroplasty includes application of the silk stitch through the pylorus and careful division of the muscle with electrocautery hook over the stitch (Fig. 2). Once the fibers are divided over the stitch, the remaining fibers are carefully sought and severed until free bulging mucosa is exposed. Serosa is approximated in the transverse fashion with one silk stitch. If mucosal violation occurs, that we observe in about a half of all cases, procedure is converted to pyloroplasty with complete division of the mucosa and transverse Heineke-Mikulicz closure with running 3–0 absorbable V-Lock stitch. In these cases, closure is reinforced with omental flap to minimize the risk of leaks. In case of pyloroplasty, patient is often admitted for several days with nasogastric tube for stomach drainage to prevent stress on the closure and development of the leak. No drainage is performed in pyloromyotomy patients and discharge is usually planned for the next day.

Figure 2.

Surgical pyloroplasty procedure. A. Placement of the incisions across the pylorus. B. Approximation of the defect in the transverse fashion. C. Heineke Mikulicz closure of the pyloroplasty incision. D. Reinforcement of the closure with omental flap.

To facilitate dissection of the gastric wall layers, authors recently adopted saline injection technique from our submucosal endoscopy experience. Ten milliliters of saline is injected into the area of pylorus that facilitates hydro dissection of gastric wall layers and preservation of the mucosa.

Authors have strong preference for GPOEM as a pyloric intervention of choice in patients with gastroparesis. Surgical pyloric interventions are reserved for cases of concomitant surgical procedure in the patients with gastroparesis, such as antireflux surgery or gastric stimulator implantation or insurance denial of GPOEM.

Enteral access

Enteral access in the treatment of refractory patient with severe gastroparesis serves two purposes – gastric venting, to alleviate nausea and vomiting and provide nutrition and hydration. Placement of a gastrostomy tube for intermittent decompression by venting or suctioning may provide relief to the patient who has prominent abdominal distention. The gastrostomy tube can be opened for drainage or aspirated to decompress the stomach; this can be employed when symptoms are distressing to alleviate nausea, pain, and bloating (25). Venting gastrostomy tubes may be placed by the endoscopist, surgeon, or interventional radiologist. In an initial series, six of eight patients were able to return to full-time work or school (26). The percutaneous endoscopic gastrostomy (PEG) tube can be converted to a button. Gastrostomy tubes are used much less now that refractory patients can be treated with gastric electrical stimulation; the gastrostomy tube may interfere with placement of the gastric stimulating electrodes.

Use of a gastrostomy tube for liquid feedings usually aggravates gastroparesis symptoms. For this reason, postpyloric feeding via jejunostomy is usually preferred. Transgastric dual channel G-J-tubes can be utilized, allowing both feeding via jejunostomy and venting via gastrostomy (27). However, the jejunostomy limb frequently regurgitates back into the stomach. To prevent this, the tip of the jejunostomy tube can be secured in the postpyloric position by deploying an endoclip to the duodenal mucosa.

Alternative is direct feeding jejunostomy tubes that are effective for providing nutrition, hydration, and medications, provided that the small intestine is functioning (28, 29). Carefully regulated nutrient enteral infusion may allow better glucose control in diabetic patients whose glycemic control is otherwise poor due to vomiting and dehydration (29). Jejunostomy tubes are usually inserted laparoscopically or at laparotomy; in some centers they are placed endoscopically (30, 31). Usually, a trial nasoduodenal tube for feeding is used to test the ability of patient to tolerate nutrient infusions (volume, rate of infusion and osmolality). Careful postoperative management of the specifically jejunostomy tubes is required as they are prone to septic complications due to leaks and erosion (Figure 4 A, B) (32). When starting jejunostomy feedings, one starts at a low rate (10 ml/hour) and advances slowly to goal feeding rate.

Figure 4.

Complication of jejunostomy tubes. A. Early postoperative migration of the balloon into the abdominal wall with the leak and necrotizing fasciitis of the abdominal wall. B. Late postoperative erosion of the balloon through the abdominal wall resulting in the leak and dermatitis.

Gastric electrical stimulation (GES).

There are several principles for electric stimulation of the stomach. First, in gastric electrical pacing the goal is to entrain and pace the gastric slow waves with low-frequency, high-energy, long-duration pulses. Pacing at 10% higher than the basal rate has been shown to accelerate gastric emptying and to improve dyspeptic symptoms (33). Second, stimulation with high-frequency, low-energy, short duration pulses has been shown to decrease symptoms with little effect on gastric emptying; it may affect proximal stomach function and activate sensory afferent nerves to reduce symptoms (34, 35). Third, sequential muscle stimulation using bursts of suprahigh frequency stimulation to induce direct gastric muscle contractions in a peristaltic sequence has accelerated emptying in animal studies (36). Finally, sequential gastric neurostimulation has been used to entrain gastric slow waves. Preliminary open label studies suggest that a hand held cervical vagal stimulator may provide improvements in symptoms of gastroparesis (37).

Enterra gastric electric stimulator uses high-frequency gastric electrical stimulation at 12 cycles/ min, and has received FDA humanitarian approval for treatment of chronic, refractory nausea and vomiting secondary to diabetic or idiopathic gastroparesis. Wires are placed into the gastric muscle at the greater curvature during laparoscopy or laparotomy. These leads are attached to an electric stimulator (“pacemaker”), which is placed in a subcutaneous abdominal pocket. An initial study demonstrated effectiveness in 20 of 26 patients with a decrease in nausea and vomiting and improvement in gastric emptying of liquids, but not solids (34). During long-term follow-up, 3 of 24 patients underwent total gastrectomy because of unsatisfactory results, and three had the stimulator removed because of erosion or infection. A subsequent study reported on 33 patients with chronic gastroparesis (35). After implantation, the electrical stimulator was turned on or off in a randomized, double-blind, crossover design. Patients felt better when the stimulator was on, although the decrease in vomiting that occurred was not statistically significant. Long-term follow-up over 1 year showed a decrease in the vomiting frequency from 25 to 6 times per week, with an improvement in quality-of-life and mild improvement in gastric emptying. Complications were skin erosion and pocket infection in three patients and gastric wall perforation by the electrode in one patient, requiring removal of the device. One patient required revision of the device due to its migration (35).

An open-label study in 25 patients reported an overall improvement from “severe” to “moderate” in symptoms of nausea and vomiting of 40% (38). There was an infection rate of 15%, requiring removal of the stimulator. Studies suggest that the patients with diabetic gastroparesis with main symptoms of nausea and vomiting and not taking regular narcotic medications are the most likely to respond to Enterra therapy (39). Subsequent studies in both diabetic and idiopathic gastroparesis, with blinded on and off periods of stimulation, did not show significant improvements in symptoms; however, a reduction of symptoms was seen with long-term open-label stimulation over 1 year (40, 41).

In one open label study, employing the Gastroparesis Cardinal Symptom Index (GCSI) to follow symptoms of gastroparesis, 29 patients underwent GES implantation over an 18-month period, with follow-up in 28 of the patients (39). GES resulted in clinical improvement in 50% of patients with refractory gastroparesis. The overall GCSI significantly decreased with improvement in the nausea/vomiting and the post-prandial subscore, but no improvement in the bloating or abdominal pain subscore. The decrease in GCSI was greater for diabetic patients than idiopathic patients. Patients with main symptom of nausea/vomiting had a greater improvement than patients with the main symptom of abdominal pain. Patients taking narcotic analgesics at the time of implant had a poorer response compared to patients who were not. This study found three clinical parameters associated with a favorable clinical response to GES: (1) diabetic rather than idiopathic gastroparesis, (2) nausea/vomiting rather than abdominal pain as the primary symptom, and (3) independence from narcotic analgesics prior to stimulator implantation (39). Knowledge of these three factors may allow improved patient selection for GES.

A large prospective study by Heckert et al details the marked improvements and the patterns of improvement (42). Nausea, vomiting, loss of appetite, and early satiety improved significantly with stimulator use, with vomiting more improved in the diabetic cohort than idiopathic. Although GES improved symptoms in 75% of all patients, diabetic patients had a post GES Clinical Patient Grading Assessment score statistically higher than did patients with idiopathic gastroparesis. This difference is thought to be due to the neuromolecular mechanism of diabetic gastroparesis, where blunting of the enteric nervous system may contribute to symptomatology (42).

Gastric electrical stimulation in gastroparesis was studied in a pragmatic open label study of patients in the NIH Gastroparesis Registry (43). In the prospectively collected database, 92 (14.5%) of patients initiated gastric stimulation (GES). Patients who underwent GES had more delayed gastric emptying at 4 hrs (30.9 vs 21.8%) with worse GCSI scores (3.8 vs 3.0) before stimulator placement. After 48 weeks, GCSI scores in patients with GES improved by average of 0.9 compared with 0.3 in controls with 43.6% showing improvement of at least 1 point compared with only 24.7% in controls. In this observational study in multiple practice settings, patients with more severe overall symptoms were more likely to improve symptomatically, primarily for nausea (43).

A recent French study brings support for gastric electric stimulation in a double-blind study, showing gastric stimulation reduced nausea and vomiting, both in diabetic and non-diabetic patients and in both those with delayed and normal gastric emptying (44).

Surgical technique.

Implantation of GES procedure requires high level of precision for the best surgical outcomes (Fig.3). Procedure includes implantation of two covered metal wires 1 cm apart into the muscular layer of the gastric wall for one to two cm at 9 and 10 cm proximal to the pylorus on the greater curvature of the stomach. Wires are tunneled through the abdominal wall and connected to the pulse generator, placed in the subcutaneous pocket. The procedure can be performed via laparotomy, laparoscopy and lately with robotic assistance. There is no consensus on the preferred technique of implantation of the generator. In the analysis of the 36 patients undergoing GES implantation postoperative LOS was 6.4 days for the laparotomy group and only 1.1 day for laparoscopy (45). Long term outcomes between two groups was comparable, although laparotomy group had higher vomiting scores and number of previous abdominal surgeries, suggesting higher preoperative morbidity as a group in whole. In another study, mini laparotomy (n=128) had a shorter operative time (84.5 vs 137 min) and LOS (2.0 vs 3.0 days) over traditional three port laparoscopy (n=37) (46). Robotic application for the implantation of the device have been recently reported (47, 48).

Figure 3.

Gastric electric stimulator implantation procedure. A. Placement of the stimulator in the subcutaneous pocket in the left subcostal area. B. Intramural placement of the electrodes with direct visual control with intraoperative endoscopy.

Likewise, there is no consensus on the location of the subcutaneous pocket for the positioning of the device generator. Technically, with the length of the leads of 35 cm device can be located virtually in any aspect of the abdominal wall. Authors’ strong preference is to located the device in left subcostal area, below the costal margin (Fig 5 A, B). This location allows to maximally pull the wires out of the abdominal cavity, decreasing the risk of wires associated complications (49).

Figure 5.

Different location of the device in the abdominal wall. A. Right lower quadrant location with long intraabdominal course of the leads. B. Left subcostal area location with very short intraabdominal length of the wires with the majority coiled over the fascia under the device.

From Frontline Medical Communications Inc., publisher of the Journal of Clinical Outcomes Management; 2019; 26(1):31-32; with permission.

Complications of GES.

Bielefeldt reported on the 1587 related to the gastric electric stimulator adverse events from Manufacturer and User Device Experience (MAUDE) databank from 01/2001 to 10/2015 (50).

Skin erosion, wound dehiscence

This is one of the most common reported complications. Skin erosion and wound dehiscence may be related to superficial placement or inadequate securing of the device to the fascia (Fig. 5A). Abscess can develop as a seeding of the hematoma postoperatively or may be a sign of lead erosion into the lumen, tracking along the leads into subcutaneous (SQ) tissue. Direct contamination of the device can occur by insulin or glucose monitoring device needles. Infected device cannot be salvaged and requires explantation. Another implantation of a new device can be attempted once all wound issues resolve.

Device migration, flipping and twiddler syndrome.

Malposition of the device likely occurs due to inadequate device fixation to the underlying fascia or erosion of the sutures. Occasionally, this may be a result of patients constantly picking and manipulating the device (Figure 7 A, B) (51). Unless symptomatic, device migration can be observed. In case of flipping, there might be difficulties in communicating with the device, requiring its repositioning. Twiddler syndrome can lead to twisting, bradding and fracture of the wires and requires repositioning. Although approved only for cardiac pacemakers, off label use of TYRX antibiotic envelope (Fig. 8) can potentially facilitate better device fixation to prevent this occurrence.

Figure 7.

Twiddler syndrome. A. Postoperative image after implantation of the GES. B. Image of the same patient 11 month postoperatively. Patient returned with recurrence of symptoms and image reveals flipped device with braded leads.

Figure 8.

TYRX Antibiotic envelope for the cardiac stimulator.

Perforation and erosion of the leads

Very seldom leads can erode into the stomach (Fig 6C.). Usually it is associated with the loss of the device function. Endoscopy confirms the finding. In rare cases, infection of can track along the lead and present as an infection at subcutaneous pocket. This complication requires explantation of the leads and the device with planned repeat placement after all infection resolves.

Figure 6.

Complications of the gastric electric stimulator. A. Erosion of the device through the skin. B. Postoperative hematoma. C. Specimen of resected segment of the gastric wall due to erosion of the GES leads component (please note silicone retention disk protruding though the mucosa).

From Frontline Medical Communications Inc., publisher of the Journal of Clinical Outcomes Management; 2019; 26(1):31-32; with permission.

Intestinal obstruction

Positioning the device in the left upper quadrant minimizes intraabdominal length of the leads, decreasing risk of leads associated complications. In other locations long intraabdominal length of the leads can predispose to various complications. Although rare, the intestines can wrap around the leads of the device, causing different degree of obstruction (Fig. 9A, B). In case of surgical intervention for obstruction, all efforts are made to preserve the stimulator leads (Fig. 9C). In cases of bowel resection, lead contamination is a serious concern, however lead explantation is not mandatory. Close postoperative monitoring for the development of lead infection is required.

Figure 9.

Bowel obstruction due to stimulator leads. A. Preoperative scout image revealing high grade bowel obstruction and stimulator leads. B. Intraoperative image of the same patient, demonstrating loops of bowels wrapped around the leads. Leads were preserved. C. Postoperative image of another patient after laparotomy for bowel obstruction where leads were sacrificed.

From Frontline Medical Communications Inc., publisher of the Journal of Clinical Outcomes Management; 2019; 26(1):31-32; with permission.

Hematoma and Seroma

Postoperative hematomas can occur from inadequate hemostasis (Fig. 6B). Seromas frequently occur in the stimulator pocket as a reactive effusion to the trauma and a foreign body. They can be observed until full resolution if small and not complicated. In cases of large hematomas with skin compromise or dehiscence, prompt washout and drainage is required. In ideal cases, the device can be salvaged. Relocation to another site might be required if skin necrosis develops increasing risk of device contamination.

Incisional Hernia

Hernias can develop after any abdominal surgery, and as such are not unique to device. Minimally invasive technique of the GES implantation minimizes this complication.

Electric Shock Sensations

Shocks may occur from breakage of the plastic lining of the device wires or from fluid buildup around the insertion of the wires into the stimulator. This can also happen from shortening of the leads to the muscles of the abdominal wall. Patients describe periodic muscle cramps with the frequency of the device impulses (every 5 seconds). Authors successfully employed omental flap coverage of the freshly implanted leads to isolate them from the abdominal wall (Fig. 10 A, B) (49). In patients who continue to feel shocks despite all efforts, the possibility of visceral hypersensitivity should be considered. A trial of nortriptyline for symptoms modulation and lowering of the output current can be undertaken. If nothing works, the device will have to be turned off for a time period, occasionally requiring explantation.

Figure 10.

Placement of gastric stimulator leads. A. Fixation of the leads in the muscular layer of the gastric wall. B. Omental flap coverage of the implanted leads.

Lack of effect / persistent symptoms

When a patient has persistent symptoms after device implantation (no response) or recurrence of symptoms after initial favorable response, a thorough workup is undertaken aimed to investigate any problems. In case of abnormal impedance values, KUB can be performed to rule out leads migration or fracture (Fig. 11 A, B). If no abnormalities are detected, the output of the device can be turned up. After adjusting device settings, at least 1 to 3-month period is undertaken to assure any improvement. One report suggests repositioning the stimulator leads on the gastric wall in patients not responding to GES (52).

Figure 11.

Abdominal imaging in the investigation of patients with recurrence of symptoms and abnormal Impedance values. A. Fracture of the lead along the medial border of the device. B. Migration of the lead. Please note wide spaced distance between the leads and lack of association of the medial lead with retention clips.

From Frontline Medical Communications Inc., publisher of the Journal of Clinical Outcomes Management; 2019; 26(1):31-32; with permission.

Combination GES and Pyloromyotomy

Recently, the combination of GES with pyloric intervention procedure has been introduced in the clinical practice. Initial concerns of the device contamination have dissipated after several favorable reports. In the group of 49 patients with 26 receiving combined GES with pyloroplasty procedure, significant improvement in total symptom score was observed. However, improved gastric emptying time was seen by 64% at 4 hours in combined procedure whereas in GES alone therapy only modest 7% improvement was observed (53). In the analysis of long-term efficacy of combined GES and pyloroplasty procedure in 24 patients, 71% improvement in total symptom score at follow up between 3-38 months (mean 17 months) with significant improvement in gastric emptying times and normalization in 60% of patients (48). Patients with refractory symptoms of gastroparesis undergoing either stimulator placement, pyloromyotomy or combined stimulator with pyloromyotomy each have improvement of their gastroparesis symptoms in open label studies (54). Gastric stimulation and combined stimulator with pyloromyotomy improved nausea/vomiting, whereas pyloromyotomy alone tended to improve early satiety and postprandial fullness (24).

Gastrectomy.

Gastrectomy is a morbid and invasive procedure regardless of the approach (open or minimally invasive) and usually regarded as a last resort intervention in the care of gastroparesis patients. The outcomes of surgical gastric resection for the treatment of idiopathic and diabetic gastroparesis has generally been disappointing (55).

For selected patients with refractory postsurgical gastroparesis, usually due to vagotomy, in whom medical therapy has failed, subtotal gastrectomy may occasionally be considered. In the report of 40 patients (32 – postvagotomy, 6 – idiopathic and 2 diabetic), a subtotal or a near-total gastrectomy with Roux-Y reconstruction was performed. 22 patients (56%) had moderate response to resection with improvements in their symptoms (56). In a series of 62 patients who underwent completion gastrectomy for severe postvagotomy gastric stasis, good symptomatic improvement in nausea, vomiting, and postprandial abdominal pain was obtained in only 43%. There was no improvement in chronic pain, diarrhea or dumping syndrome. The combination of nausea, need for TPN, and retained food at endoscopy predicted a poor outcome (57). In another series of 81 patients with severe postsurgical gastroparesis, near-complete gastrectomy with a 55-cm Roux-en-Y reconstruction was performed. Follow-up averaged 56 months for 52 patients; 78% reported improvement of their GI symptoms, 7% believed that there was no change, and 15% stated that their condition had worsened. A subtotal (70%) gastrectomy with resection of the antrum and pylorus, closure of the duodenum, and restoration of GI continuity with a 60-cm Roux-en-Y jejunal loop was reported in four patients with insulin-dependent diabetes with intractable vomiting from gastroparesis (58). In total, three of the four patients did well, eliminating frequent hospital admissions.

In the analysis of 35 patients (43% postoperative, 34% idiopathic and 23% diabetic) undergoing near total gastrectomy different symptoms improved in 70-89% of patients. Six patient suffered postoperative leak, requiring reintervention. There were no postoperative mortalities (59). Landreneau et al reported on the 53 patients undergoing Roux-Y reconstruction for the treatment of gastroparesis with either gastrectomy (27 patients) or stomach left in situ (essentially a gastric bypass) (26 patients) (60). Patients had similar symptom response to intervention. Gastrectomy patients required longer operative times (223 vs 155 min), longer hospital stay (7 vs 4 days) and experienced higher rate of postoperative complications (44% vs 8%). However, patients with retained stomach were more likely to require subsequent surgical interventions (23% vs 4%), suggesting that gastrectomy may be a more definitive procedure for the treatment of gastroparesis (60). Recent report on sleeve gastrectomy in 19 gastroparesis patients suggests good response with improvement in Barium Emptying Radiography Index and Gastrointestinal Quality of Life Index (61).

Decision making for surgical treatments

With the array of the endoscopic and surgical options for the management of patients with refractory gastroparesis, it is frequently difficult to decide on the best course of action and sequence of intervention (Fig. 12). At our institution, we are able to offer patients each of the treatments discussed in this chapter. As described above, we reserve gastrectomy for the “end of the road situations” where no other intervention has proved helpful.

Figure 12.

Decision algorithm for the choice of procedure in patients with gastroparesis.

Botox pyloric injection is considered a temporizing intervention and we utilize it sparingly to bridge patient until more definitive intervention is performed, which is sometimes is delayed by the insurance approval process. Occasionally, we use Botox injection to see if the patient is more likely to respond to surgical or endoscopic pyloromyotomy.

Enteral access we also reserve to refractory patients in whom pyloric intervention and or GES implantation has not produced the desired effect and patient continues to lose weight, remains TPN dependent for nutritional support and hydration or requires gastric venting for uncontrolled nausea and vomiting.

In the majority of the cases, the decision comes to either gastric stimulator or pyloromyotomy, and or combination of both. Patients with contraindications to GES implantation (anticipated need for MRI, lack of insurance coverage, patient preference or abdominal wall infection, precluding safe implant) are offered pyloric intervention – preferably GPOEM or surgical pyloromyotomy or pyloroplasty as a backup. In patients with severe and predominant symptoms of nausea and vomiting we proceed with gastric electrical stimulation. In patients with significantly If delayed gastric emptying in the absence of severe nausea and vomiting our preferred intervention is endoscopic pyloromyotomy. If patients have significant nausea and vomiting with markedly delayed gastric emptying, patients often undergo a combined stimulator placement with surgical pyloromyotomy or pyloroplasty. Those patents who had undergone one of the procedures – either GES implantation or pyloric interventions and remain symptomatic later can crossover and undergo the other procedure, essentially ending up with the combined intervention in the sequential fashion.

In patients who remain refractory to treatment, placement of a feeding jejunostomy or a venting gastrostomy may be considered. TPN, if used, should be limited and temporary because of the risk profile. Surgical gastrectomy should generally be discouraged and reserved for selected patients with postsurgical gastroparesis.

Studies are currently being performed to evaluate this type of patient-oriented approach – selecting between pyloric intervention and gastric stimulation.

Key points.

1. Failure of medical therapy of gastroparesis requires surgical or endoscopic intervention. Between multitude treatment options available decision usually comes to gastric electric stimulator vs pyloric intervention.

2. Patients with predominant symptoms of nausea and vomiting benefit more from GES placement.

3. Patient with severely delayed gastric emptying without severe nausea or vomiting proceed with pyloric intervention. Those with severely delayed gastric emptying and severe nausea and vomiting proceed with combined procedure – GES and pyloroplasty or pyloromyotomy.

4. Endoscopic pylori Botox injection is a temporizing intervention and enteral access tubes and gastrectomy are reserved for recalcitrant cases with failure of all previous interventions.