8. Post-Surgical Gastroparesis
KeywordsGastroparesisPostoperative gastroparesisPostsurgical gastroparesisGastric emptyingPyloroplastyGastric stimulationPer-oral pyloromyotomy
Gastroparesis is an entity most commonly discussed as a consequence of diabetes mellitus. However, diabetes is a factor in only about a third of patients who present with this clinical problem [2, 3]. Gastroparesis can be debilitating, resulting in poor quality of life, recurrent hospital admissions, and often multiple medications and procedures. Historically, postsurgical gastroparesis was diagnosed after surgical treatment of peptic ulcer disease with vagotomy to reduce gastric acid. Although acid-reduction surgery is significantly less common today, symptoms of gastroparesis following procedures like sleeve gastrectomy or fundoplication are a known cause of postoperative morbidity. This chapter addresses gastroparesis occurring in the postoperative setting. It explores the natural history of this clinical entity, reviews possible mechanisms, and discusses available treatment options.
Gastroparesis may be defined simply as a symptomatic delay in the gastric emptying of solid or liquid meals in the absence of a mechanical obstruction . Postsurgical gastroparesis is a delay in gastric emptying in the absence of an obstruction, but after a surgical procedure. Gastroparesis is an important disorder to diagnose and treat as 25% of patients with gastroparesis require at least one hospitalization and spend twice as long in care once hospitalized compared to patients with no otherwise specialized functional disorders of the stomach . Patients with gastroparesis may be prone to gastroesophageal reflux, retained gastric contents, and bezoars, which increase the risk of aspiration .
Incidence and Epidemiology
Postsurgical gastroparesis often takes a subclinical course, and treatment might not be sought until severe symptoms develop. Therefore, estimating the true incidence is a challenge. Current literature suggests that five million people within the United States suffer from gastroparesis from a variety of etiologies. This number is expected to rise as the incidence of diabetes rises . Up to 12% of patients with diabetes are diagnosed with gastroparesis . Diabetic (29%) and idiopathic (36%) gastroparesis are the most common types. The third most common type is a postsurgical state (13%). Postsurgical gastroparesis is most commonly encountered after a vagotomy or vagus nerve injury . Historically, postsurgical gastroparesis was identified in patients who underwent an acid-reducing procedure for peptic ulcer disease. Initially, their postsurgical complaints were attributed to a postvagotomy syndrome. This syndrome is characterized by nausea, vomiting, bloating, and abdominal pain, suggesting a failure of gastric emptying. The symptoms were initially managed pharmacologically; however, surgeons later began to include a drainage procedure (i.e., pyloroplasty) at the time of initial vagotomy for ulcer disease in order to facilitate postoperative emptying of the stomach. Over time, it became standard practice to add a drainage component to the initial operation. With the advent of new surgical techniques, gastroparesis has been seen in patients without intentional vagotomy. Reports of gastroparesis symptoms following procedures like sleeve gastrectomy and Nissen fundoplication have been published. Most of these cases spontaneously resolved, and literature showed prolonged postoperative gastroparesis (lasting longer than 3–4 weeks) to be a rare entity. In a series of 615 patients who underwent laparoscopic Nissen fundoplication, all had symptoms of delayed gastric emptying during the first three postoperative months. Yet by one year , symptoms suggestive of gastroparesis, e.g., bloating and flatulence, had resolved in greater than 90% of patients . Rates of postsurgical gastroparesis vary widely depending on many factors, including the site and nature of the surgical procedure. It is unclear which patients ultimately progress to chronic gastroparesis and what factors underlie or predict symptom severity.
In order to understand the treatment modalities of postsurgical gastroparesis, a review of gastric pathophysiology is useful. The stomach has multiple functions. It stores food, has mechanical and chemical processes of digestion, and works to deliver food to the duodenum in a controlled fashion. As discussed in Chap. 1, there is a complex interplay of electrical and hormonal signaling for all of these functions.
Near the beginning of a meal, the fundus relaxes to receive food and provide space for acid and enzymes to initiate the digestive process. The reflex of relaxation is mediated by nitric oxide from enteric neurons as part of a vagal reflex arc. In a postvagotomy state, there is a decrease in gastric accommodation and gastric compliance, with a resultant increase in luminal pressure. As a result, the early phase of liquid emptying is accelerated. This acceleration causes rapid emptying of hyperosmolar solutions into the proximal small intestine and can result in early dumping syndrome. When the vagus nerve is intact, the fundus contracts to facilitate movement of chyme from the fundus to the antrum. Once food particle size allows, the chyme is delivered to the pylorus. When the pylorus is denervated , the later phases of solid and liquid emptying are prolonged .
In addition to relaying signals from the central nervous system to the intrinsic motor neurons in the gastric myenteric plexus, as described in Chap. 1, the vagus nerve is predominantly a sensory nerve. It conveys physiologic information from the stomach to the brain. Some signals of sensation include satiety and nausea. Neuropathic injury of the vagus may therefore generate symptoms due to dysfunction of motor and sensory components . Truncal vagotomy can be predicted to result in several significant effects; loss of fundic relaxation (with bloating, fullness, and disorderly transit of ingested bolus to distal regions), reduced antral contractions (with loss of trituration and retention of solids), and loss of pyloric relaxation (with retention). This is the rationale for combining truncal vagotomy with a drainage procedure, leading to some unpredictability in the final outcome.
More selective vagotomy procedures may spare antral innervation of the stomach and generally are associated with milder and subtler changes in gastric function. A Roux-en-Y anastomosis is associated with gastric stasis syndrome. In addition to vagotomy and distal gastric resection, the Roux limb may contribute to gastric stasis by generating ectopic pacing activity. Nevertheless, Roux-en-Y gastric bypass (RYGB) has been used as a treatment for gastroparesis. The possible mechanism of action is possibly related to the creation of a small gastric pouch to facilitate emptying even with uncoordinated peristalsis and the bypassing of the potentially dysfunctional pylorus.
Inadvertent vagotomy has been estimated to occur in 3–5% of open surgeries on the abdomen. Postfundoplication vagal nerve injury is identified in 4–40% of patients undergoing laparoscopic fundoplication, leading to impaired pyloric relaxation and decreased pyloric compliance . Fundoplication can affect the sensorimotor function of the proximal stomach. Nissen fundoplication was the most common cause of postsurgical gastroparesis in an ongoing audit performed by the National Institute of Diabetes and Digestive and Kidney Diseases gastroparesis consortium .
The exact mechanisms responsible for postsurgical gastroparesis after gastric surgery remain unclear but are likely to be multifactorial. Vagotomy has been shown to cause accelerated emptying of liquids, weakened peristaltic contractions of the antrum, decrease in intestinal secretion of prokinetic hormones and dissociation of antral pressure waves from duodenal waves.
Presentation and Diagnosis
The most common presenting symptom of postsurgical gastroparesis is nausea (98%), followed by abdominal pain (90%), early satiety (86%), and vomiting (68%) . Patients typically complain of worsened pain after meals. Succussion splash may be auscultated on physical exam, generated by rocking the patient . There are no other consistent physical exam features associated with a diagnosis of gastroparesis. Neither are there specific laboratory studies required for diagnosis. However, patients should be screened for the presence of diabetes mellitus , thyroid dysfunction , neurological disease , prior gastric or bariatric surgery, and autoimmune disorders . An acute increase in blood glucose can result in delayed gastric emptying in both diabetic and nondiabetic adults. Hyperglycemia precipitates delayed gastric emptying, which in turn precipitates hyperglycemia. A complete review of current medications should be undertaken, and medication-induced impairment of gastric motility should be considered in patients taking narcotics, anticholinergics , glucagon-like peptide-1, and amylin analogs as these have been shown to affect emptying . Medications that affect gastric emptying should be stopped at least 48 hours prior to diagnostic testing .
Esophagogastroduodenoscopy (EGD) is commonly performed in the workup of gastroparesis. Typical findings include retained food and even bezoars. Endoscopy is important to rule out pathological postsurgical anatomic problems, including a malpositioned or slipped fundoplication, stricture, or obstruction, which may mimic symptoms of gastroparesis. An upper GI contrast exam can be performed as an alternative to EGD to detect delays in liquid phase emptying and rule out mechanical obstruction.
A gastric emptying study is the gold standard. A patient is fed a bolus of solid food (usually a radio-labeled egg). This meal should be substantial enough to generate gastric contractions. Imaging is obtained immediately and at two and four hours postingestion. At two hours, a delay in movement of the food bolus is 90% sensitive and specific for delayed gastric emptying (normal to have 30–60% empty). At four hours, a delay of movement of the food bolus is 100% sensitive and 70% specific, with normal residual of 0–10% of gastric contents . The most reliable parameter to report gastric emptying is the gastric retention at four hours . Blood glucose concentration should be measured during emptying and considered in the interpretation of the result .
Solid gastric-emptying studies appear to be more sensitive for the detection of gastroparesis compared to liquid phase studies. Liquid emptying studies were performed in those who could not tolerate solids or in patients with dumping syndrome. Ziessman et al. tested this hypothesis by studying the emptying of liquids and solids in healthy volunteers and compared them to those with symptoms of gastroparesis. They found that of all patients who had solid normal emptying, 32% had delayed liquid emptying . Liquid emptying is also positional. They concluded that for the detection of gastroparesis, a 30-minute study of liquid gastric emptying has limited diagnostic value over solid emptying alone but can be considered if the diagnosis is still in question.
Nonoperative management includes a combination of lifestyle and dietary modifications in addition to medical therapy. Patients should eat several small liquid meals throughout the day. Diet should consist of low fat, low fiber foods. Poor tolerance of a liquid diet is predictive of a poor outcome with oral nutrition . Carbonated beverages can exacerbate symptoms. Patients should also avoid alcohol and tobacco as these can affect emptying. Tight glucose control is imperative as hyperglycemia has been associated with symptom exacerbation [2, 16]. It is important to perform a careful review of patient medications that may be contributing to delayed emptying and discuss modifying these with the patient and prescribing provider.
Pharmacologic therapy consists of two main options that have been approved in the U.S. As discussed in Chap. 4, metoclopramide is a dopamine and serotonin receptor antagonist and is an antiemetic and prokinetic agent. Its promotility effect results in esophageal and antral contractions, decreased pyloric, and duodenal tone, thus facilitating gastric emptying. Side effects of metoclopramide include dystonic reactions, extrapyramidal movements, diarrhea, fatigue, depression. It is FDA approved for a maximal period of 12 weeks .
Erythromycin is another agent used for the treatment of gastroparesis. The chemical compound is an analog of motilin. Motilin , and thus erythromycin, binds to the migrating motor complex (MMC) and generates strong contractions. It has shown to be effective in patients following vagotomy, gastrectomy, and esophagectomy. Unfortunately, with continued use, tachyphylaxis can occur, limiting its utility. The side effects of erythromycin include nausea, hearing loss, muscle weakness, prolonged QT, fatigue.
However, approximately 10% of all patients with gastroparesis of any etiology do not respond to conventional medical therapy . Shafi et al. suggest patience and time when treating postsurgical gastroparesis. Symptoms often resolve spontaneously over time, especially in patients who are symptomatic postfundoplication. An improvement in gastric emptying may occur, with an improvement in symptoms, suggesting that either the central nervous system may be able to adapt to loss of vagal input or that vagal reinnervation may occur. Therefore, watchful waiting and conservative symptom management is a reasonable strategy. The likelihood of spontaneous improvement declines after one year, however.
Surgical and Endoscopic Therapy
As discussed elsewhere, operative management for patients with gastroparesis may run the spectrum from minimally invasive techniques, such as gastric pacing, to surgical resection or bypass. Decompressive gastrostomy can be offered to patients who do not respond to changes in diet or medical therapy. A gastrostomy tube can be placed endoscopically, laparoscopically or in an open fashion to facilitate drainage from the stomach and reduce the patient’s abdominal pain, nausea, and vomiting. Some patients may also undergo a feeding jejunostomy tube during or after their decompressive gastrostomy tube is placed to facilitate long-term nutrition. As discussed in Chap. 5, it has been recommended that a trial of nasojejunal feeding be performed prior to placement of a jejunostomy tube as small bowel dysfunction has also been observed in patients with gastroparesis . Indications for enteral feeding include unintentional weight loss of 10% or more in a period of three to six months, as well as repeated hospitalizations for refractory symptoms.
Roux-en-Y gastric bypass is another surgical option. In a small study of seven patients who underwent laparoscopic Roux-en-Y gastric bypass between December 2010 and February 2013 for chronic gastroparesis, the average percent of gastric retention preoperatively was 72%. A total symptom severity score was measured for each patient, before and after their bypass surgery. The severity and frequency of symptoms significantly improved from preop (M severity = 13.8, range = 12–16; M freq = 15.4, range = 12–22) to postop (M severity = 7.6, range = 5–10, P = 0.002; M freq = 8.4, range = 6–11, P = 0.04) . Three out of seven patients stopped antiemetic therapy postoperatively. There was a mean follow-up of 315 days, mean BMI change of 9.1 units, and mean weight loss of 56 pounds. One patient had remission of diabetes and one with decreased medication use. Five out of six patients with preoperative reflux resolved.
Subtotal gastrectomy with Roux-en-Y reconstruction is advocated by some for gastric atony and persistent nausea secondary to postsurgical gastroparesis . In a study of 62 patients with refractory postsurgical gastroparesis, subtotal or completion gastrectomy provided symptomatic improvement in 67% of patients . Completion gastrectomy should be considered in patients with postsurgical gastroparesis who remain markedly symptomatic and fail medical and other surgical therapies .
Historically, pyloroplasty has been a component of surgical treatment for peptic ulcer disease . A patient would undergo an acid-reducing procedure concurrently with a pyloroplasty. Several groups have investigated the utility of pyloroplasty as treatment for postsurgical gastroparesis [5, 6]. In a study by Gilsdorf et al., outcomes after pyloroplasty for the relief of postfundoplication gastroparesis were evaluated. Patients included in this study had at least one prior fundoplication, and several had multiple additional foregut surgeries. They selected ten patients who had botox injections prior to pyloroplasty as a positive predictor of response to pyloroplasty. Botox injection was used as a surrogate for pyloric defunctionalization. This study demonstrated that in a carefully selected group of patients, there was durable improvement in gastroparesis symptoms, as evidenced by objective measurements in gastric emptying function, in subjective reporting of symptoms, and in retrospective completion of a symptom scoring questionnaire. Symptomatic improvement occurred in 90% of patients and was durable in 89% of patients at long-term follow-up .
Per-oral pyloromyotomy (POP) has also been performed to treat medically refractory gastroparesis. In a study by Landreneau et al., patients who underwent pyloroplasty (either laparoscopic or POP) between 2014 and 2017 for gastroparesis were assessed for outcomes and symptoms. There were an equal number of postsurgical gastroparetic patients (20%) in each group. POP had statistically significant less mean operative time, blood loss, and mean length of stay compared to laparoscopic pyloromyotomy. Patients had no statistically significant difference in gastric emptying studies, postoperatively. Both groups had significant improvement in emptying, though several patients did not undergo repeat testing. Both groups had significant improvement in the quality of life scoring, again not statistically different. The additional benefit of POP or pyloroplasty is that these are organ-sparing interventions that do not affect future surgical options in the event that they are not successful .
For patients who have failed medical management and are neither a surgical candidate nor amenable to surgery, transpyloric stenting may be of benefit. A small retrospective study of 30 patients was performed to assess this question. Patients who were diagnosed with gastroparesis (either idiopathic (53%), diabetic (26%), or postsurgical in origin (20%)) and did not respond to medical therapy underwent endoscopic transpyloric stenting. Seventy-five percent of patients had weight gain postprocedure. All inpatients were discharged with symptomatic improvement or resolution. One patient was able to stop parenteral nutrition as their oral intake dramatically improved. There was a 59% stent migration rate but no complications from stent placement or migration. The authors of this study suggested that transpyloric stenting may be utilized in an algorithm for triaging patients for definitive therapy at the level of the pylorus or to facilitate discharge during a severe episode .
As discussed in detail in Chap. 6, gastric stimulation has also been offered for several years for the treatment of gastroparesis. The Enterra device is a type of electrical stimulator available in the United States. It has been commercially available since 2000 for patients with severe gastroparesis unresponsive to medical management. It provides high-frequency, low-energy gastric stimulation via electrodes that are implanted into the serosa of the antrum. The generator is within the anterior abdominal wall. This device is placed surgically, either laparoscopically or open. A group from the UK recently completed a study of gastric electrical stimulation using the Enterra device . They placed temporary stimulators in patients endoscopically (transnasal, implanted at the junction of the body and the antrum), and pulses were generated in a standard fashion for 7–14 days. Patients who had improvement in symptoms were candidates for a permanent device. Permanent stimulators were placed laparoscopically, 9 and 10 cm proximal to the pylorus on the lesser and greater curves, respectively. The placement was confirmed by the absence of visible electrodes on endoscopy. Patients were evaluated at six weeks and three months. A total of 71 patients were included in this study. Two thirds of their patient population had idiopathic gastroparesis and 18% were postsurgical. Of the postsurgical gastroparesis patients, four had a prior Roux-en-Y gastric bypass, six with prior antireflux surgery, one prior cardiomyotomy, one prior splenectomy, and one likely vagotomy for peptic ulcer disease . They concluded that 71% of their patients (18% postsurgical and 61% idiopathic) with intractable gastroparesis had a good response to permanent stimulation. The crux of their study relied upon patient selection. By implanting permanent stimulators only in patients who showed improvement during the trial period, they were able to select out the potential responders and minimize cost and risk to likely nonresponders. A criticism of this study is that no objective measure of symptoms was recorded pre- or postintervention. When reviewing consensus guidelines from Camilleri et al. about gastric stimulation, it is stated that improvement in gastric emptying is variable. Complications such as infection or lead migration occur in up to 10% of patients implanted, and in general, symptomatic improvement appears to be greater for diabetic patients than for those with idiopathic gastroparesis .
Of note, the postsurgical gastroparesis population exhibits a different response to GES therapy than the diabetic or idiopathic group. Gastric stimulation has been advocated for patients who have failed the surgical management of their symptoms. In the same UK study discussed above, two patients with a prior pyloroplasty, one with prior antrectomy and Roux-en-Y jejunostomy and one with a loop gastrojejunostomy, all sought stimulation to treat their unrelenting symptoms. In this study, 71% of patients with postsurgical gastroparesis saw greater than 50% improvement in symptoms after implantation of a gastric stimulator . Though the sample size of this study was small, it does raise the question as to whether gastric stimulation is a reasonable option even when more aggressive surgical intervention has failed. Further studies in this area are needed.
There are a variety of surgical options to treat gastroparesis, as outlined above. In several studies, patient preference was a major driver of the surgical option ultimately adopted. As more data become available, surgical strategy will likely change with published outcomes. In 2017, a group from the Oschner Clinic attempted to determine which surgical approach was the most efficacious. The study of 58 patients included five postsurgical gastroparesis patients. Patients underwent pyloroplasty only, insertion of a gastric stimulator only, both pyloroplasty and gastric stimulator, or sleeve gastrectomy. Patient preference was the main determinant of the operative approach. Comparing preoperative and postoperative symptom scores, they investigated the long-term efficacy of gastric stimulation, pyloroplasty, and combined gastric stimulation with pyloroplasty for the treatment of refractory gastroparesis. The gastric stimulator only group demonstrated the greatest symptomatic relief. The pyloroplasty alone group demonstrated the least amount of improvement. Combining both gastric stimulator and pyloroplasty demonstrated an additional symptomatic improvement in those patients who did not have adequate relief from the initial procedure. There was a significant improvement in all symptoms except for severity of vomiting, early satiety, postprandial fullness, and epigastric pain. They concluded that gastric stimulation should be the initial treatment unless there is a contraindication. While this work offers potentially promising therapy, given the low power of these studies, more investigation and perhaps multicenter trials need to be undertaken to determine when patients should undergo additional procedures to better control their symptoms .
Postsurgical gastroparesis can be a debilitating disease for patients, resulting in poor quality of life, recurrent hospital admissions, and often multiple medications and procedures. The postsurgical population is a unique patient group, sometimes requiring a different treatment algorithm. As evidence accumulates and technology improves, the approach to patient management is likely to change. Watchful waiting in the early postoperative period with medical management remains the first-line therapy at the present time. Surgical intervention is typically offered if medical therapy fails, and the type of procedure pursued is often driven by a patient’s prior surgical history and patient preference.