The addition of pyloroplasty as a new surgical approach to enhance effectiveness of gastric electrical stimulation therapy in patients with gastroparesis


Address for Correspondence
Irene Sarosiek, Department of Internal Medicine, Texas Tech University Health Sciences Center, El Paso, TX 79905, USA.
Tel: +(915) 545-6626; fax: +(915) 545-6648;


Background  Improvement of gastroparesis (GP) symptoms has been documented in patients treated with gastric electrical stimulation (GES), but acceleration of gastric emptying (GET) is unpredictable. The aim of our study was to evaluate the advantage of adding surgical pyloroplasty (PP) to GES for improvement of GET and control of symptoms in diabetes mellitus (DM), idiopathic (ID), and postvagotomy (P-V) GP.

Methods  A total of 49 (17 – DM, 9 – ID, 23 – P-V) consecutive GP patients: 38 female; mean age 42 (21–73 years); mean weight 158 lbs (102–245), underwent GES implantation, and 26 (53%) additionally received PP. Total Symptoms Score, 4-h GET, adverse events (AEs), and days of hospitalizations were captured at baseline and at the last visit.

Key Results  The mean follow-up was 7 months. Total Symptoms Score in patients who received Enterra and PP or GES alone significantly improved compared to their baseline scores (P < 0.001). GET improved by 64% at 4 h (P < 0.001) in patients with Enterra and PP, compared to 7% observed after GES therapy alone (ns). The most impressive acceleration of GET was seen in the P-V group, who received both therapies (P = 0.004) and 8 (60%) of them normalized GET. No AEs accompanied the addition of PP to the Enterra surgery.

Conclusions & Inferences  (i) In drug-refractory GP the addition of PP to GES substantially accelerated GET; (ii) The GET response in P-V group was the most impressive; (iii) Significant symptom reductions were achieved by both procedures; and (iv) PP added to GES may sustain better long-term symptoms control particularly in the P-V setting.




gastric electrical stimulation


gastric emptying test




diabetes mellitus






Total Symptoms Score


The incidence of gastroparesis (GP) as a chronic motility disorder of the stomach with delayed gastric emptying (GET) is about 4% of the US population.1–3 The most frequent symptoms are nausea, vomiting, early satiety, bloating, and postprandial fullness with occasional abdominal pain and discomfort.4–7 The etiology of GP includes three groups: diabetes mellitus -DM (29%), idiopathic- ID (36%), and postvagotomy (P-V, 13%).1–6,8–10

The proposed mechanism of developing GP in long-standing Type I or II DM is explained by the decreased neural control of gastric motor function, impairment of the inhibitory nitric oxide-containing nerves, and the damage to the interstitial cells of Cajal.9,11–19 Idiopathic GP could follow an acute onset of viral illness caused by Cytomegalovirus, Epstein–Barr virus, and herpes (varicella-zoster) as a prelude to chronic symptoms, and delayed GET.20 Postvagotomy GP (P-V GP) occurs in up to 10% of patients who undergo vagotomy, either intentional or inadvertent, mostly related to Nissen fudoplications.1,21 It has been also reported in up to 50% of patients undergoing intentional vagotomy in setting of Billroth I and II, antral resection and Roux-en-Y surgeries.22–24

Therapeutic interventions include diet modification, pain management, and psychological counseling, use of pharmacological agents, and insertion of feeding tubes, endoscopic injection of botulinum toxin, and surgical procedures.2,3,9,25–29 Gastric electrical stimulation (GES) is an FDA approved therapy for the drug-refractory GP. Many publications have shown a 60–70% reduction in GP symptoms, which may result in sustained improvement for up to 10 years.30–34 Most studies have reported no significant changes in GET, therefore approaches to improve GET must be considered.35–40 One approach is surgical pyloroplasty (PP) which when performed alone resulted in some benefit in 30% of DM and ID-GP patients.41,42We hypothesized that by performing PP in addition to GES, the GET could be accelerated while the other therapeutic benefits and clinical outcomes of GES will also be achieved.



The criteria for patient’s selection were based on a goal to achieve an adequate number of patients4–10 from each of three etiological subgroups among all potential GES candidates who were seen at our center between 2005 and 2008. Patients were approached to undergo PP at the time when GES placement was being considered for them. None of our subjects had received a PP in the past. When adequate numbers of patients with PP were recruited in each sub-group, patients were then approached to have Enterra surgery only. In the case where patients declined to receive PP then they were offered implantation of GES alone.

There were 49 consecutive, prokinetic, and anti-emetic refractory patients (38 women, 11 men; mean age of 42 years; range 21–73; mean duration of GP 3 years, range 1–11; mean weight 158 lbs, range 102–245), who underwent GES implantation between June 2005 and October 2008. Three different etiologies of GP separated these patients into the following: diabetic GP (17 patients), ID GP (9 patients), and postsurgical (P-S) GP (23 patients) groups. Nissen fundoplication was the suspected cause of GP in 16 (70%) of P-S patients. Of these 49 GP patients, 26 received PP as an adjunct to GES therapy (8 DM, 5 ID, 13 P-S) and 23 patients received GES therapy alone (9 DM, 4 ID, 10 P-S). All P-S patients had undergone a sham meal test study measuring pancreatic polypeptide to confirm vagal nerve damage. The main inclusion criteria were as follows: 1 documented diagnosis of GP for more than 1 year and refractoriness to anti-emetics and prokinetics; 2 more than seven emetic episodes per week; 3 in the setting of fundoplication, where patients could not vomit, the criterion was chronic daily nausea plus or minus retching; and 4 delayed GET (gastric retention greater than 60% at 2 h and greater than 10% at 4 h) based on a 4-h standardized radionuclide solid meal test. Patients were excluded if they had organic or pseudo-obstruction, primary eating or swallowing disorders, positive pregnancy test result, psychogenic vomiting or were utilizing peritoneal dialysis. A signed IRB approved consent form was obtained from all subjects before they entered the study.

Study protocol

This study was initiated with a baseline evaluation of medical history and upper GI symptoms, assessment of nutritional status, pregnancy testing, blood chemistries, and GET test to determine the qualification for enrolment. Surgical placement of the GES system by laparotomy with positioning of the two electrodes in the muscularis propria of the greater curvature in the intact stomach was performed.35,37 In two cases of antrectomy the electrodes were placed at 2 and 3 cm proximal to the gastric anastomosis. In addition, removal of any parenteral nutrition (TPN), and conversion of gastric decompression G-, or gastro-jejunal tubes to a surgical jejunostomy feeding approach were performed, as well as de novo placement of jejunal tubes in cases when severe malnutrition was diagnosed. All patients were followed-up at approximately 3, 6, and 12–18 months after implantation of the stimulators to repeat baseline measurements. In addition, adverse events, including hospitalization, were monitored throughout the follow-up period. The detailed descriptions of the GES system and surgical and stimulation techniques have been published previously.35,37 The Heineke-Mikulicz PP was done in the standard manner with interrupted sutures.43 Only one surgeon (JF) using the same technique at all times performed the surgeries.

Assessment of symptoms

Each patient completed a Symptoms Interview Form at baseline and at 3-, 6-, and 12–18 month follow-up visits. This form assessed the symptoms of GP occurring in the last 2 weeks before the interview for severity and frequency of vomiting, nausea, early satiety, bloating, postprandial fullness, epigastric pain, and burning. The severity of each symptom was graded by the patients as 0, absent; 1, mild (not influencing usual activities); 2, moderate (diverting from, but not requiring modifications of usual activities); 3, severe (influencing usual activities, severely enough to urge modifications); and 4, extremely severe (requiring bed rest). The frequency of each symptom was marked as 0, absent; 1, rare (1/week); 2, occasional (2–3/week); 3, frequent (4–6/week); and 4, extremely frequent (≥7/week). The sum of the severity and frequency ratings of the seven symptom sub-scores comprised the overall Total Symptom Score (TSS).

Measurement of gastric emptying

Gastric emptying scintigraphy was performed after an overnight fast and a 3-day period without prokinetics. This standardized method for GE consists of a scrambled egg substitute (120 g of Free Cholesterol & Fat Free Egg; (Sunny Fresh Foods Inc, Monticello, MN, USA) (60 Kcal) labeled with 99 m Tc sulfur-colloid (1 mCi), two slices of whole wheat bread (120 Kcal), 30 g jelly (75 Kcal), and 120 mL of water.7 The meal has a total caloric value of 255 Kcal (nutritional composition: 72% carbohydrate, 24% protein, 2% fat, and 2% fiber). Anterior and posterior images of the stomach were taken immediately after eating and then hourly for 4 h. Gastric retention of gamma counts was calculated by the Department of Nuclear Medicine by using geometric and decay correction. Again, the delayed GE was defined as the percentage of gastric retention equal to or greater than 60% at 2 h and equal to or greater than 10% at 4 h or both.

If emptying was actually accelerated after the surgery we utilized the criteria that have been published for the dumping syndrome, where <30% of the study meal is remaining at 1 h and <20% at 2 h after the scintigraphic GET test.

Statistical analysis

The TSS scores in three groups based on GES alone or GES + PP were reported as means at baseline and follow-up visits and were compared by paired t-test analysis. The Wilcoxon’s signed rank test was used for comparison of not normally distributed GE data; therefore results of GE are reported as median. Statistical significance was assigned for P-value <0.05.


The mean follow-up for all patients was 7 months (range 3–18). All groups were similar in regard to mean age, weight, BMI (range 19–40), duration of GP symptoms, baseline TSS, and GET results. All three etiological groups of gastroparetics maintained their weight with an average 158 lbs at baseline versus 161 lbs at the time at their last visit. The admissions to hospital for Enterra & PP patients were significantly reduced from a mean of 28 days in the year preceding the surgery to a mean of 1 day at the last follow-up visit; versus 24 days before and 6 days after GES in the Enterra alone group.

Overall, in regard to mean total symptoms scores, the clinical efficacy reached statistical significance and was similar in both groups. Patients who received only GES therapy had decreased their TSS severity score by 35%; (P ≤ 0.001) and TSS frequency from by 40%; (P < 0.001). Those patients receiving PP and GES had improvement in severity of gastroparetic symptoms by 45% (P < 0.001), and TSS frequency score decreased by 43% (P < 0.001) (Table 1).

Table 1.   Changes of gastroparesis symptoms and gastric retention in all patients who received one or two surgical procedures
Total symptoms score (mean)
  1. GP, gastroparesis; TSS, total symptoms score; GES, gastric electrical stimulation; PP, pyloroplasty; GET, gastric emptying test; F/U, follow up. Overall TSS improvements and changes in GET results for both therapeutic groups: GES only and GES/PP are presented in this table. P-value compares follow-up to baseline data.

All GPs with GES only201335%; P < 0.001201246%; P < 0.001
All GPs with GES & PP201145%; P < 0.001211243%; P < 0.001
Gastric emptying test (mean)
 2 h4 h
All GPs with GES only776713%; ns47447%; ns
All GPs with GES & PP734445%; P < 0.001471764%; P < 0.001

Also 2 h and 4 h GET results at follow-up visits were significantly slower in the GES group compared to patients who received GES & PP therapy, and reached statistical significance at these two points of measurements (P < 0.001). The GES group had no significant change in GET at 4 h, whereas the GES & PP patients significantly accelerated GET by 64% at 4 h. (P < 0.001; Table 1)

On the basis of our results we conclude that combining all patients with drug-refractory GP the addition of PP to the currently available GES therapy significantly accelerates GET at 4 h, while also significantly improving GP symptoms without any adverse event.

The data were also analyzed based on the underlying three causes of GP (Tables 2 and 3). Diabetic groups of GP patients improved their symptoms regardless of the surgical approach. The 4-h gastric retention was reduced by 48% in the group of GES and PP compared with 17% in GES alone. Comparable results regarding symptoms responses were observed in both ID-GP groups. Idiopathic GPs who underwent implantation of GES and PP improved GET at 4 h by 65% compared to baseline whereas GES alone resulted in 44% reduction.

Table 2.   Total Symptom Score (TSS) with severity and frequency of gastroparesis (GP) symptoms in each etiological subgroup of patients
 TSS severityTSS frequency
  1. TSS, total symptoms score; GES, gastric electrical stimulation; PP, pyloroplasty; ID, idiopathic; DM, diabetes mellitus; P-V, postvagotomy; F/U, follow up. P-value compares baseline and follow-up data.

 Therapy groupBaselineF/UImprovementBaselineF/UImprovement
 GES (n = 9)20860%; P < 0.00117853%; P = 0.01
GES & PP (n = 8)181045%; P = 0.01201145%; P = 0.019
 GES (n = 4)221150%; (P < 0.001)211529%; ns
 GES & PP (n = 5)211624%; ns211529%; ns
 GES (n = 10)19764%; P < 0.00121766%; P < 0.001
 GES & PP (n = 13)201145%; P = 0.001211148%; P < 0.001
Table 3.   Gastric emptying test (GET) results with median retention of the study meal at 2 and 4 h in subgroups of gastroparetic (GP) patients before and after surgical treatment
 GET(2 h)GET(4 h)
  1. GES, gastric electrical stimulation; PP, pyloroplasty; ID, idiopathic; DM, diabetes mellitus; P-V, postvagotomy; F/U. P-value compares baseline and follow-up data.

 Therapy groupBaselineF/UImprovementBaselineF/UImprovement
 GES (n = 9)78737%; ns594917%; ns
 GES & PP (n = 8)705620%; ns 462448%; ns
 GES (n = 4)924852%; ns 522944%; ns
 GES & PP (n = 5)664345%; ns 401465%; ns
 GES (n = 10)73679%; ns 332040%; ns
 GES & PP (n = 13)793655%; P = 0.002501472%; P = 0.004

The most impressive control of symptoms was achieved in P-V GP patients where statistical significance reached P ≤ 0.001 in both surgical groups. In addition, patients with P-V GP who received both GES and PP improved GET retention at 2 h by 55%, (P = 0.002) and by 72% at 4 h (P = 0.004). Eight of 13 (67%) of these P-V patients actually normalized their GET and two (15%) reached criteria for rapid GET (retention <30% at 1 h and <20% at 2 h). The GES alone patients improved GET by 40%, which was not significant (see Table 3 for details).

Adverse or serious adverse events in both groups of patients were similar. The wound infection rate was similar in GES and GES plus PP population, 3%versus 2% (respectively). No technical and mechanical problems related to the utilization of GES system were observed in any patient.


In the clinical world of treatment of severe symptoms of drug-refractory GP, there are not many pharmacological or surgical options.44,45 The list of available prokinetic and anti-emetic agents has not changed in last 20 years. The same medications with their sometimes severe side effects as well as tachyphylaxis problems provide limited options for the growing number of gastroparetic patients around the world.

Since its FDA approval in March 2000, GES with Enterra Therapy has demonstrated in the three major etiologically different groups of gastroparetic patients that better symptoms control can be achieved as evidenced by the supporting literature.30–40,46 These studies include double blind clinical trials (30, 31, and 47) as well as supportive publications from unblinded studies providing evidence for successful Enterra therapy in those gastroparetics not responding to medical therapies and where nausea and vomiting are the dominant symptoms.32–38 Overall diabetic and P-S patients are most consistently improved by Enterra, and the ID sub-group shows the least improvement.36–39 At the same time, this control of GP symptoms is being achieved by GES without a significant acceleration of the delayed GET.47 This disconnection set the stage for our protocol, where patients received either GES alone or the addition of a surgical PP at the time when GES was implanted.

The sub-group of diabetics undergoing GES and PP showed an improvement in GP symptoms similar to the cohort with GES alone, but had a faster GET that did not reach statistical significance. Less impressive data were generated in the ID sub-group. However, both these groups had small numbers for statistical comparison. Finally, the last and generally least studied etiology of GP is the P-S subgroup, where the pathogenesis of their condition was related to inadvertent or accidental vagal nerve injury during a fundoplication procedure in the majority (80%) of these patients. In this group, after GES and PP, not only were the cardinal symptoms of GP controlled as well as with GES alone but also GET significantly improved by 72% at 4 h, and 67% of the P-S patients actually normalized their GET. In addition, during long-term follow-up hospitalization days were decreased to only 1 day for post-PP compared to 6 days for Enterra alone group. The main reason for this better long-term follow-up effect after GES and PP is based on the P-V group, which comprised 41% of all patients studied.

Pyloroplasty used as an adjunct to GES therapy for patients with GP secondary to DM and ID etiologies shows promise, although larger studies should be done to strengthen further statistical analysis. However, the combination of the two procedures can be proposed as the best surgical recommendation for the P-V category of gastroparetic patients, because it provides enhanced benefits in GET without changing morbidity.

It has to be acknowledged that one limitation of our study was that it was not a randomized clinical trial (RCT). The major reasons for not going immediately to a RCT were that we regarded our work as a preliminary pilot study to investigate if: (i) PP is effective and safe in gastroparetic patients and (ii) which GP etiological group benefits the most. It is important to mention that we compared two procedures, thus the group of solo GES patients served as the control population of GP patients.48 Because GP has multiple etiological factors we felt that a double blind study should only be done after we had more understanding of the responses of each group of gastroparetics. This would also allow us to plan the size of such a RCT based on statistical principals in the future.

On the other hand, the question could be asked whether PP alone without the Enterra Therapy could generate similar results. In the most recent publication on this question 49 the authors present a very short-term follow-up (up to 3 months), to assess laparoscopic Heineke-Mikulicz PP. Their study was a retrospective review of prospectively collected data of 28 gastroparetics with 21 ID and 7 DM patients. They excluded all other gastroparetics due to concurrent or prior gastric or esophageal surgery. This minimally invasive procedure resulted not only in a highly significant (P < 0.0001) improvement in GP symptoms severity scores but also it significantly decreased mean T1/2 of GET time. Interestingly, 10% of patients who were observed for at least 3 months did require implantation of gastric neurostimulators to control severe persistent nausea. In their investigation, the GES system was not introduced to patients during PP surgery because of concern for potential risks and their confidence that a positive outcome of GP symptoms control would be achieved by one surgery alone. In our study, we have not observed any increase in post-op complications related to the combination of the two procedures at this same time.

In our opinion, there is not enough existing data addressing this issue to support efficacy of PP performed as a solo procedure in patients with delayed emptying of the stomach. However, pyloric injection of botulinum toxin has been studied as a therapeutic option to overcome the tonic motor activity of the pylorus termed ‘pylorospasm’. This approach of pyloric Botox injections could be viewed as having a similar, but obviously more transient effect than surgical PP.29 Gumaste 9 compared botulinum injection and dilation of the pylorus and he concluded that both techniques may produce the same benefit, but more studies were required to reach a verdict.11 Overall when diabetics and IDs were studied in a controlled, randomized trial setting the data were negative.9 In the P-V subset, this approach seems to be more effective. Our group 50 noted that in patients with GP secondary to vagal injury who received endoscopic Botox, GP symptoms at 1 and 3 months after treatment were reduced, but returned by 6 months. Other authors noted similar transient efficacy of intrapyloric Botox injections to control GP symptoms and GET time in recipients of lung transplantation 51 where vagal nerve damage occurs.

An explanation or hypothesis as to why Botox and hence PP may be effective in P-S states is as follows. In postvagal nerve damage GP, the onset of symptoms occurs in a stomach that had been entirely normal before this injury. Hence, with no organ damage or systemic diseases and the problem being recognized soon after the injury one could assume this process could be easily reversed and the previously normal gastric function may be restorable. In diabetic and ID cases, the diseases are often longstanding ones and chronic injury is still continuing and unlikely to be reversible. Hence, simply performing a PP does not overcome severe antral hypomotility and stagnation of gastric contents. In P-V states the vagal nerve damage is more specific for inducing pyloric ‘spasm’ and impaired pyloric relaxation both of which are addressed by a PP.

With our current data we can now propose that the addition of a Heineke-Mikulicz PP to the standard GES procedure markedly improves and often normalizes delayed GET is most pronounced in P-V gastroparetic patients. This combined GES-PP enhances long-term symptom control and augments nausea and vomiting control as well as maintains the increased vagal nerve function that has been shown to accompany GES.52


This part of our research and all data obtained from gastroparetic patients were not supported by any sponsors. Compassionate use of GES therapy HDE protocol generated presented clinical results. The authors would like to acknowledge the following individuals for their contributions: Kathy Roeser, Pernilla Foran LPN, and Teri Lavenbarg RN as well as faculty, fellows, and nursing staff in the Division of Gastroenterology, Surgery residents and OR personnel at KUMC Kansas City, KS. In addition, we wish to recognize Dr. R. Dusing, James Trayler, and Chris McMillin from the Nuclear Medicine Department of KUMC.


No funding declared.


Data for this manuscript were obtained from patients who have not been participating in any sponsored clinical trials.

Author contribution

IS offered consultation on study design, administrative issues, insurance pre-approval, patient recruitment, patient retention; OR procured assistants, acquired data, and wrote the manuscript; JF offered surgical support; SC was administrative assistant; ZL gave technical support; JS looked after the data analysis, reviewed the manuscript; RWMcC designed the study concept, looked after clinical care and patient recruitment, writing, and reviewing of manuscript.