Gastroparesis presents with symptoms of gastric retention and nongastrointestinal manifestations, with objective evidence of delayed gastric emptying in the absence of mechanical obstruction. Diabetic, idiopathic, and postsurgical gastroparesis are the most common forms, although many other conditions are associated with symptomatic delayed gastric emptying (Table 1). Gastroparesis is estimated to affect up to 4% of the US population and may produce either mild, intermittent symptoms of nausea, early satiety, and postprandial fullness with little impairment of daily function, or relentless vomiting with total disability and frequent hospitalizations. A recent report estimated that inpatient costs for patients with severe gastroparesis approach $7000/month.
Gastroparesis presents with a constellation of symptoms. In one study, nausea was reported by 93% of patients whereas early satiety and vomiting were noted by 86% and 68%, respectively. In another series, nausea, vomiting, bloating, and early satiety were reported by 92%, 84%, 75%, and 60% of patients, respectively. Many patients in both case series (89% and 46%) also reported abdominal pain (Table 2). Others experience heartburn from acid reflux into the esophagus that is facilitated by fundic distention which increases the rate of transient lower esophageal sphincter relaxations. Although some gastroparetics with frequent vomiting lose weight and develop malnutrition, most patients were overweight or obese in one series, indicating that the disorder does not necessarily restrict food intake. Phytobezoars are organized concretions of indigestible food residue that are retained within the stomach. These may increase gastroparesis symptoms or produce a palpable epigastric mass, gastric ulceration, small intestinal obstruction, or gastric perforation. Bezoars are eliminated by endoscopic disruption and lavage, enzymatic digestion (papain, cellulose, or N-acetylcysteine), and dietary exclusion of high-residue foods. Variably delayed gastric emptying may cause unpredictable food delivery in diabetics with gastroparesis, affecting glycemic control and increasing risks of both severe hypo- and hyperglycemia.
Quantification of the severity and nature of gastroparesis symptoms has been facilitated by the introduction of validated surveys. The most widely used questionnaire for this purpose is the Gastroparesis Cardinal Symptom Index (GCSI), a symptom score validated in 7 university-based clinical practices in the United States that correlates well with patient and physician ratings of gastric symptom severity. The GCSI comprises 3 subscales (postprandial fullness/early satiety, nausea/vomiting, and bloating) and represents a subset of the comprehensive Patient Assessment of Gastrointestinal Symptoms (PAGI-SYM) survey.The Patient Assessment of Upper Gastrointestinal Disorders-Quality of Life (PAGI-QOL) survey quantifies quality of life in dysmotility syndromes such as gastroparesis. Using these questionnaires, investigators have begun to stratify patients with gastroparesis into different subgroups on the basis of symptom severity and predominant symptom profiles. Although these surveys currently are most useful for research trials, in the near future they may serve a clinical role similar to the Rome criteria for functional gastrointestinal disorders in terms of helping healthcare providers to select symptom-based management approaches.
Gastroparesis has many causes. In a case series of 146 gastroparesis patients seen at a large US tertiary medical center, 29% had underlying diabetes, 13% developed symptoms after gastric surgery, and 36% were idiopathic. The mean age of onset for gastroparesis is 34 years. Eighty-two percent of cases occur in women.
The systemic disease most often complicated by development of gastroparesis is diabetes mellitus. In one population-based survey, 18% of diabetic patients reported upper gastrointestinal symptoms, a greater proportion than reported by nondiabetic controls. The prevalence of delayed emptying in those patients with long-standing type 1 diabetes ranges from 27% to 58%. Likewise, gastroparesis is present in up to 30% of patients with type 2 diabetes. Diabetes more profoundly affects gastric motor function than small bowel transit, indicating an increased sensitivity of the stomach to diabetic injury. In diabetic patients on hemodialysis, development of gastroparesis correlates with the presence of orthostatic hypotension, prior myocardial infarction or cerebrovascular accident, and gangrene of the extremities.Patients with diabetes secondary to chronic pancreatitis also may develop gastroparesis.
Delayed emptying of solids in diabetic gastroparesis traditionally is believed to result from impaired phasic antral contractions, but other factors contribute to impaired gastric evacuation. Increased postprandial antral diameter on ultrasonography suggests defects in tone. Increased liquid retention in the fundus and prolonged solid food retention in both the proximal and distal stomach occur in diabetic patients with gastroparesis, demonstrating altered intragastric distribution. Improperly timed pyloric contractions of abnormal intensity (> 10 mm Hg) and duration (> 3 minutes) leading to pylorospasm are observed in some patients with diabetes. Others exhibit abnormal postprandial jejunal burst contractions which retard gastric outflow.
Autonomic neuropathy is a likely contributor to the pathogenesis of delayed emptying in patients with long-standing diabetes. Gastric acid output in response to sham feeding is reduced by two thirds in diabetic patients, indicating vagal neuropathy. Histologic study of vagus nerves from affected patients with diabetes reveals variable degrees of myelin degeneration. Other measures of dysautonomia, including postural hypotension and loss of vagotonic cardiac reflexes, are prominent in diabetic patients with gastroparesis. In type 1 diabetes, delays in emptying correlate with the degree of cardiovascular autonomic neuropathy but not peripheral neuropathy.
Nonneuropathic factors lead to delayed emptying in diabetes as well. In a longitudinal study of patients with type 1 and type 2 diabetes, the prevalence of autonomic neuropathy increased from 35% to 80% over 12 years, but emptying rates did not worsen, indicating that progressive neuropathy did not promote development of gastroparesis. Diabetic animals show impaired contractility at the smooth muscle level. Histopathologic examination of gastric specimens from 4 diabetic patients with gastroparesis who underwent gastric resection revealed smooth muscle degeneration and fibrosis with eosinophilic inclusion bodies. Impaired smooth muscle contractile responses in type 1 diabetes may also occur because of circulating antibodies to L-type calcium channels. Interstitial cells of Cajal (ICCs), in concert with smooth muscle and nerves, play a crucial role in regulating gastric motor function. One subset of ICCs in the circular muscle relays and amplifies information from enteric neurons to smooth muscle cells, while another population in the myenteric plexus generates rhythmic electrical depolarizations (slow waves) that control the frequency and direction of gastric contractions. Depletion of myenteric and smooth muscle ICCs was observed in one insulin-dependent diabetic patient with severe gastroenteropathy. In another study of diabetic patients undergoing gastric surgery, disruption of gastric ICC networks correlated with loss of normal slow wave cycling on preoperative testing. Tissues from patients with type 2 diabetes show ICC losses that are more prominent in circular muscle than in the myenteric plexus. One unpublished observation suggests that the histopathology of diabetic gastroparesis is highly variable, even between individuals with similar degrees of gastric stasis. Gastric tissue from one individual with an abrupt symptom onset and well-controlled diabetes showed no morphologic abnormalities, whereas tissue from another with long-standing, poorly controlled diabetes exhibited muscle fibrosis, loss of myenteric neurons, and reduced staining for many neurotransmitter and ICC markers. Some histologic deficits may be linked to single pathogenic defects. In a murine model, decreased insulin and insulin-like growth factor-I signaling reduced stem cell factor production and led to muscle atrophy that in turn promoted ICC depletion. The observation that the mitochondrial DNA mutation 3243 predisposes to gastroparesis in type 2 diabetes suggests that genetic factors also contribute to disease development.
Metabolic factors participate in the control of gastric emptying. Hyperglycemia in the absence of neuropathy or myopathy disrupts normal antral motor complexes at plasma glucose levels as low as 140 mg/dL in healthy humans. In type 1 and type 2 diabetes, gastric emptying of liquids is delayed when blood glucose levels exceed 270 mg/dL. Likewise, in type 1 diabetes, delays in solid emptying are observed during hyperglycemia which improve during euglycemia. Hyperglycemia also blunts motor responses to prokinetic drug treatment of diabetic gastroparesis.[32,33]
Idiopathic gastroparesis is at least as common as diabetic gastroparesis in most case series.Patients typically are young or middle-aged and up to 90% are women. Although many individuals report an insidious clinical course with no obvious trigger, one fourth present with an acute onset of symptoms in association with acute gastroenteritis or with viral prodromal symptoms such as diarrhea, fever, myalgias, and headache. Postinfectious idiopathic gastroparesis has a relatively good prognosis, with symptom resolution over several years in many cases. In most instances, the offending organism is not characterized. However, 8 of 11 children with acute-onset gastroparesis in one study tested positive for rotavirus infection. Transient slowing of gastric emptying also develops after infection with parvovirus-like agents and with Lyme disease. In immunosuppressed individuals, cytomegalovirus, Epstein-Barr virus, varicella zoster virus, and herpes simplex virus may be implicated.[36,37] In one case, postinfectious gastroparesis occurred as part of a larger dysautonomic syndrome also involving cardiac conduction and bladder function. Furthermore, gastroparesis has been observed after vaccination for tetanus, anthrax, and hepatitis.
The pathogenic basis of idiopathic gastroparesis is less well studied than that for diabetic gastroparesis. Most patients do not exhibit underlying neuropathy. Pre- and postprandial gastric pH levels measured by a radiotelemetry capsule are higher in diabetic patients with gastroparesis than in those with idiopathic gastroparesis, suggesting a lesser degree of vagal neuropathy in individuals with idiopathic disease. Similarly, plasma pancreatic polypeptide responses to sham feeding are normal in patients with idiopathic gastroparesis but are markedly blunted in diabetic patients with gastric stasis, indicative of greater vagal damage secondary to diabetes. As in patients with diabetes, histologic examination of gastric tissue from patients with idiopathic gastroparesis may reveal loss of ICCs and myenteric neurons.[26,42] Isolated gastric myopathy also has been described as a cause of idiopathic gastroparesis.Inflammatory mechanisms may play a role in some cases. Prominent eosinophilic infiltrates were noted in the smooth muscle layers and around nerve fibers in the gastric wall in tissues from 2 gastroparesis patients undergoing gastrectomy. Another patient with idiopathic gastroparesis exhibited increases in CD4+ and CD8+ T lymphocytes in the gastric myenteric plexus and reductions in myenteric neuronal staining for tachykinins.
Gastroparesis may complicate surgery performed on the stomach. Approximately 5% of patients who undergo vagotomy and drainage for ulcer disease or malignancy experience nausea, vomiting, and early satiety due to postoperative gastric stasis. Roux-en-Y gastrojejunostomies may be complicated by the Roux stasis syndrome, which presents with nausea, vomiting, abdominal pain, and delayed gastric emptying resulting from either spastic or retroperistaltic Roux limb contractions.Esophagectomy with gastric pull-through or with colonic interposition may cure esophageal neoplasm but can lead to gastroparesis. Half of patients undergoing pylorus-preserving Whipple procedures for pancreatic cancer or chronic pancreatitis develop delayed emptying. Gastroparesis commonly occurs after lung and heart-lung transplantation and may cause microaspiration into the transplanted lung. Gastric bypass and gastroplasty performed for morbid obesity delay gastric emptying of solids and promote fundic distention, leading to early satiety, anorexia, and weight loss. However, some individuals exhibit rapid emptying of caloric liquids which may limit the degree of weight loss. Numerous recent cases of postsurgical gastroparesis have developed after fundoplication for gastroesophageal reflux disease. In many instances, it is uncertain whether gastric motor impairments were present preoperatively or whether they occurred as a consequence of the surgery. It is estimated that 4% to 40% of patients undergoing laparoscopic fundoplication develop intraoperative vagal damage to some degree. Some centers have advocated preoperative gastric scintigraphy in those being considered for fundoplication to detect individuals at risk for gastroparesis.
Abnormal antral peristalsis and fundic tone are demonstrable in patients with postsurgical gastroparesis. In one study, 8 of 9 patients with gastroparesis had no fasting motor cycles. Increases in intragastric volumes, impaired fundic responses to meals and balloon inflation, and heightened perception of gastric distention are also observed after vagotomy, documenting combined motor and afferent defects in this condition. Gastroparesis occurring after fundoplication likely results from operative compression or severing of the vagus nerves. As evidence of this, one study reported reductions in antral postprandial phasic contractions after fundoplication.
Other Causes of Gastroparesis
Approximately 25% to 30% of cases of gastroparesis are not idiopathic and are not secondary to diabetes or postoperative gastric motor dysfunction. Etiologies in such instances include disorders with isolated gastric dysmotility, conditions with diffuse motor dysfunction involving most or all of the gastrointestinal tract, and nongastrointestinal diseases with associated emptying delays.
Other disorders with isolated gastric motor dysfunction
Some disorders present with isolated impaired gastric motor function. Delayed emptying of liquids, solids, or mixed meals has been reported in up to 57% of patients with gastroesophageal reflux, while other studies have demonstrated no abnormalities.Delays in emptying in patients with gastroesophageal reflux disease correlate poorly with symptoms, lower esophageal sphincter pressure, and esophageal acid exposure. Gastric stasis is observed in up to 60% of individuals with nonobstructing pancreatic carcinoma and in smaller numbers of patients with other malignancies.[56,57] Severe nausea, vomiting, and impaired gastric evacuation are common after abdominal irradiation. Delayed solid emptying is noted in atrophic gastritis with or without pernicious anemia. In this condition, reductions in gastric secretion increase the time needed to fragment solid foods to tiny particles. The median arcuate ligament syndrome is an entity caused by compression of the celiac axis by a fibrous band that presents with postprandial pain, nausea, vomiting, weight loss, and delayed gastric emptying and can be relieved by surgical decompression. Ischemic gastroparesis results from celiac arterial occlusion. Some patients with inactive Crohn’s disease and no radiographic evidence of obstruction may present with markedly delayed gastric emptying and severe symptoms of gastroparesis.
Disorders of diffuse gastrointestinal dysfunction with associated gastroparesis
Gastroparesis frequently occurs in patients with diffuse disorders of gut motility, such as chronic intestinal pseudo-obstruction. These individuals may also present with small intestinal bacterial overgrowth, nutritional deficiencies, bowel habit abnormalities, and pneumatosis intestinalis. The prevalence of delayed emptying in scleroderma ranges from 40% to 67%, whereas rates in polymyositis-dermatomyositis and systemic lupus erythematosus are lower. Gastric stasis is described in smooth muscle disorders such as myotonic dystrophy and progressive muscular dystrophy. Primary or secondary amyloidosis can cause neuropathic or myopathic intestinal pseudo-obstruction.[64–67] Gastric stasis is found in 19% to 64% of patients with chronic constipation, irritable bowel syndrome with constipation, and idiopathic megarectum. Other cases of intestinal pseudo-obstruction are familial, occur after a viral prodrome, or present as a paraneoplastic phenomenon, usually with small-cell lung carcinoma. Serologic markers in paraneoplastic gastroparesis or intestinal pseudo-obstruction include type 1 antineuronal nuclear antibody, anti-Purkinje cell cytoplasmic antibody, and ganglionic nicotinic acetylcholine receptor antibody. Such antineuronal antibodies are also detected in some patients with idiopathic gastrointestinal dysmotility syndromes with an autoimmune basis. In addition to producing an achalasia-like picture, Chagas disease may cause gastroparesis, megaduodenum, and chronic intestinal pseudo-obstruction. Other infectious agents that produce generalized gut dysmotility include varicella zoster, Epstein-Barr virus,Clostridium botulinum, and HIV.[37,72,73]
Nongastrointestinal disorders with associated delays in gastric emptying
Neurologic diseases may present with delayed gastric emptying. More than 7% of a large series of gastroparesis patients had underlying Parkinson’s disease. Drugs used to treat Parkinson’s disease may exacerbate gastric stasis in this condition. Disturbed gastric motility is also observed with cerebrovascular accident, migraine headaches, high cervical spine injury, and peripheral nerve disorders such as stiff-man syndrome and Charcot-Marie-Tooth syndrome. Some patients with familial dysautonomia, Shy-Drager syndrome, Guillain-Barré syndrome, multiple sclerosis, and other demyelinating disease may exhibit gastric stasis or intestinal pseudo-obstruction.
Gastroparesis is rarely found in eating disorders. Retarded gastric emptying and reduced antral contractions are observed in some patients with anorexia nervosa. It is possible that these motor disturbances are consequences of the disorder, as malnutrition itself impairs gastric evacuation and improved nutrition promotes accelerated solid emptying. In rare instances, patients with bulimia nervosa exhibit delayed gastric emptying. Rumination syndrome usually is not associated with gastric stasis. Characteristic simultaneous contractions on antroduodenal manometry are detected in many cases, which reflects increases in abdominal wall tone. However, one study reported small reductions in postprandial antral contractions in patients with rumination.
Other conditions affect gastric emptying and can produce gastroparesis symptoms. Studies of gastric emptying during nausea of pregnancy using nonradioactive methods have reported variable results, with some studies showing delays and others observing normal emptying. Cyclic vomiting syndrome is characterized by intermittent attacks of relentless nausea and vomiting with prolonged intervening asymptomatic periods. Most investigations observe acceleration rather than delay of emptying in adults with cyclic vomiting syndrome. However, one study reported reduced antral motor activity after eating in 5 of 8 cyclic vomiting patients. Gastroparesis or intestinal pseudo-obstruction may complicate severe hypothyroidism as well as hyper- and hypoparathyroidism. Other diseases associated with gastroparesis include chronic pancreatitis, cystic fibrosis, cirrhosis, and chronic renal insufficiency. Genetic disorders such as Turner’s syndrome may predispose to delays in gastric emptying. Many prescription medications delay emptying, as can tobacco, excess ethanol, or cannabis.[83–85] Total parenteral nutrition impairs gastric evacuation. This effect correlates with increases in serum glucose and is minimized by replacement of half of the amino acid content with branched-chain amino acids.
Symptoms of gastroparesis can mimic those of other conditions. Mucosal inflammation can exacerbate gastroparesis symptoms. In one investigation of patients with presumed gastroparesis flares, upper endoscopy revealed Candida esophagitis or acid-peptic disease in some individuals who responded to treatments other than those given specifically for gastroparesis.
Some patients with diabetes report severe nausea and vomiting yet exhibit no gastric retention. Indeed, in a recent report of patients with suspected gastroparesis, GCSI scores correlated poorly with gastric emptying rates. Only postprandial fullness showed a weak association with gastric retention. Impaired postprandial fundic accommodation has been observed in some diabetic patients. In another study involving patients with type 1 diabetes with dyspepsia, gastric distention evoked exaggerated nausea, bloating, and abdominal pain, suggesting defects in visceral afferent function. However, in studies employing evoked potential testing, afferent neuropathy was associated with fewer rather than more symptoms. The correlation of accommodation and visceral sensory defects with the degree of symptoms in diabetic patients has not been well studied.
Symptoms in some patients with idiopathic gastroparesis also stem from defects other than delayed emptying. In one recent study, 43% of patients had impaired fundic accommodation which was associated with early satiety and weight loss, and 29% exhibited heightened sensitivity to gastric distention, which was related to the degrees of epigastric pain, early satiety, and weight loss. Idiopathic gastroparesis may be difficult to distinguish from functional dyspepsia in some cases, leading some to speculate they are variants of the same disorder. As with idiopathic gastroparesis, some patients with functional dyspepsia present after an acute infection. Furthermore, 25% to 50% of patients with functional dyspepsia exhibit delayed emptying.[80,94] The postprandial distress subtype of functional dyspepsia in the current Rome III classification system is characterized by gastroparesis-like symptoms, including postprandial fullness and early satiety. Other Rome III diagnoses with symptom overlap with idiopathic gastroparesis include chronic idiopathic nausea and functional vomiting. The prevalence of delayed emptying in these syndromes has not been characterized. From a diagnostic standpoint, a presentation of predominant pain and less nausea is considered to be more typical of functional dyspepsia, whereas dominant nausea with minimal pain is more consistent with idiopathic gastroparesis.
Documentation of Impaired Gastric Function
Gastric scintigraphy is the most accepted test for diagnosing gastroparesis; it employs 99mTc-sulfur colloid bound to solid food. A major drawback of gastric scintigraphy has been a lack of standardization of criteria used to diagnose gastroparesis across all medical centers. In a Canadian hospital survey, 28% of centers defined the cutoff for gastroparesis as the degree of gastric retention greater than 2 standard deviations above the mean, whereas 26% used 1 standard deviation, 6.5% employed 1.5 standard deviations, and 40% had no objective criteria. Only 18% of centers validated their results in healthy volunteers. Recently, researchers have advocated uniform standards for gastric scintigraphy performance and interpretation. A standardized method using a meal of toast, jam, and an egg substitute with 99mTc-sulfur colloid has been validated. Gastric retention greater than 60% at 2 hours and greater than 10% at 4 hours is diagnostic of gastroparesis. Even using standardized protocols, the diagnosis of gastroparesis may not be clear-cut. In one investigation, 37% of patients with normal emptying at 2 hours exhibited delayed emptying at 4 hours, whereas 19% with delayed emptying at 2 hours normalized at 4 hours. The use of gastric scintigraphy has not been rigorously subjected to outcomes analysis. One study reported that emptying results do not influence clinical management.
Other techniques for measuring gastric emptying have also been promoted. Results of breath testing after consuming nonradioactive 13C-labeled nutrients such as octanoate and acetate can assess gastric emptying of solids and liquids. These methods measure liberation of 13CO2 in expired breath samples after duodenal assimilation of the ingested compound and are reliable only in persons with normal digestive and absorptive function. Emptying results from 13C-octanoate and 13C-acetate breath tests show fair-to-good correlations with scintigraphy. A second method involves swallowing a radiotelemetry capsule that continuously transmits information on luminal pH and pressure to a receiver worn by the patient. With this test, gastric emptying is detected by the abrupt pH increase as the capsule passes from the antrum to the duodenum. Correlation coefficients between capsule and scintigraphic emptying times exceed 0.8.This method also quantifies gastric motility indices as well as transit, pH, and motor patterns in the small intestine and colon. Measures of gastric emptying can also be provided by ultrasonography and magnetic resonance imaging.[101,102] A variant of gastric scintigraphy, dynamic antral scintigraphy, has been used in research to image nonocclusive antral contractions in real time, but this method has not been used clinically.
Antroduodenal manometry involves peroral placement of a catheter to monitor luminal pressure patterns. Fasting motility is recorded for 4–5 hours, during which time 1 or more fasting motor complexes are usually observed. Fed motor activity is then measured for 2 hours after a solid meal. In some centers, 24-hour ambulatory recordings or testing of motor effects of prokinetic drugs are performed. Manometry is indicated for patients who have unexplained symptoms and have not responded to treatment, or who are being considered for surgery or enteral vs parenteral nutrition. Gastroparesis is characterized by loss of normal fasting migrating motor complexes and reduced fed antral contractions and, in some cases, pylorospasm. Manometry is most useful in excluding associated small intestinal dysmotilities, including those with myopathic (contractile amplitude < 30 mmHg with normal morphology) and neuropathic (intense, uncoordinated burst contractions) patterns. Small intestinal motor dysfunction is detected in 17% to 85% of patients with gastroparesis. Clinical management is influenced in approximately 20% to 25% of patients undergoing gastrointestinal manometry.
Electrogastrography (EGG) measures gastric slow-wave activity via cutaneous electrodes overlying the stomach. In healthy persons, EGG recordings exhibit uniform waveforms of 3 cycles per minute, which increase in amplitude after ingesting water or nutrients. EGG abnormalities include rhythm disruption for more than 30% of the recording time including tachygastria (frequency of more than 4 cycles per minute) and bradygastria (fewer than 2 cycles per minute) and lack of a signal amplitude increase with eating.EGG abnormalities are prevalent in patients with gastroparesis as well as in some patients with nausea and vomiting and normal emptying. Impaired postprandial amplitude responses correlate with delayed solid emptying in gastroparesis. The importance of EGG dysrhythmias in symptom generation is supported by observations that treatment responses correlate better with slow-wave normalization than with accelerated emptying. Other electrical techniques, such as epigastric impedance and applied potential tomography, employ cutaneous electrodes to measure changes in resistance afforded by liquid meals which correlate with the rate of emptying. Tests of myoelectric function have been promoted as alternatives to gastric scintigraphy, but have not achieved widespread acceptance due to a lack of proven gastric antiarrhythmic therapies.
Other tests of gastric motor function are useful research tools but have limited clinical applicability. Single photon emission computed tomography (SPECT) employs intravenous 99mTc-pertechnetate that accumulates within the gastric wall, providing a 3-dimensional outline that can assess fundic accommodation. Satiety testing involves the ingestion of water or a liquid nutrient until the patient reports maximal fullness. Volumes consumed in functional dyspeptics with early satiety are reduced vs healthy volunteers, reflective either of impaired accommodation or visceral hypersensitivity.
Therapy of Gastroparesis
Therapies for gastroparesis include nutritional modifications, medications to stimulate gastric emptying, drugs that reduce vomiting, endoscopic and surgical approaches, and psychological interventions. To facilitate treatment selection, a gastroparesis severity classification has been proposed. Grade 1 (mild) gastroparesis is characterized by intermittent, easily controlled symptoms with maintenance of weight and nutritional status, and is treated with dietary modification and avoidance of medications that slow emptying. Grade 2 (compensated) gastroparesis is characterized by moderately severe symptoms with infrequent hospitalizations that are treated with combined prokinetic and antiemetic agents. Grade 3 (gastric failure) gastroparesis patients are medication-unresponsive, cannot maintain nutrition or hydration, and require frequent emergency department or inpatient care. Individuals with grade 3 gastroparesis may need intermittent intravenous fluids and medications, enteral or parenteral nutrition, and endoscopic or surgical therapy.
Dietary and Nonmedical Measures
Dietary recommendations, including ingesting multiple small meals, favoring liquids over solids, avoiding indigestible solids, and consuming low-fat meals, compensate for the gastric motor impairment in gastroparesis. Medications that inhibit gastrointestinal motility should be discontinued if possible. In diabetic gastroparesis, maintaining euglycemia may avoid the inhibitory effects of hyperglycemia on gastric motor function.[31,32]
Prokinetic Medication Therapy
Prokinetic medications that stimulate gastric emptying, including metoclopramide, erythromycin, and domperidone, are the mainstays of treatment for gastroparesis of moderate severity or worse (Table 3). Other drugs have motor stimulatory effects but their utility is less well defined. There have been few direct comparison trials of prokinetic agents in gastroparesis. One meta-analysis reported that erythromycin is most potent at stimulating gastric emptying while erythromycin and domperidone both are superior to metoclopramide for symptom control. This interpretation does not exclude possible publication bias favoring only positive reported clinical trials.
Metoclopramide is a prokinetic agent that acts as a combined serotonin (5-hydroxytryptamine4 [5-HT]) agonist, dopamine D2 antagonist, and direct stimulant of gut smooth muscle contraction. The drug also has antiemetic effects via brainstem D2 receptor antagonism and vagal and brainstem 5-HT3 receptor antagonism. Metoclopramide increases esophageal, antral, and small bowel contractions, accelerates gastric emptying, and facilitates gastroduodenal coordination. The drug is available in pill, liquid suspension, intravenous, and subcutaneous forms.
At least 5 controlled trials and 4 open series have studied the efficacy of metoclopramide in gastroparesis. Symptoms improved in 7 studies, while stimulation of gastric emptying was noted in 5. Patients may develop tolerance to the prokinetic action of this agent over time, but its antiemetic effects are sustained. Side effects limit metoclopramide use in 30% of patients. Fatigue, agitation, sleep disturbances, and hyperprolactinemic effects (galactorrhea, amenorrhea) are common adverse reactions. Dystonias are common, especially in patients younger than 30 years of age. Irreversible tardive dyskinesia is a catastrophic consequence which occurs in 1% to 10% of cases when metoclopramide is taken for more than 3 months. Because this condition is disabling, the risk should be discussed in detail with the patient prior to prescription and the discussion should be documented in the patient record.
Erythromycin evokes powerful, lumen-occluding antral contractions via action on neural and smooth muscle receptors for motilin, the physiologic regulator of fasting gastroduodenal motility. Unfortunately, this motor response can impair gastric sieving of solids, leading to duodenal delivery of incompletely triturated food particles. Erythromycin is available in pill, liquid suspension, and intravenous forms. When given acutely, erythromycin can facilitate passage of nasoenteric tubes for supplemental nutrition.
At least 3 controlled trials and 6 open-label trials have been published on the use of erythromycin in gastroparesis. Symptom benefits were observed in 6 of these studies. Tolerance to its prokinetic action with long-term oral use is common and likely relates to motilin receptor downregulation.[118,119] Because erythromycin has no additional antiemetic effect, the benefits of chronic therapy may be transient. Side effects with high doses include abdominal pain, nausea, and vomiting that relate to induction of intense, spastic motor activity in the stomach and upper intestine. Erythromycin also increases the risk for sudden cardiac death. In this large Medicaid cohort, the sudden death rate of current erythromycin users was 2.01 times as high as for control populations. The risk for death was further increased in those patients who also were on CYP3A (cytochrome P-450 3A) inhibitors, including selected antipsychotics, cardiac antiarrhythmics, antifungals, calcium antagonists, antidepressants, and antiemetics.
Domperidone is a peripheral dopamine D2 antagonist that does not cross the blood-brain barrier, thus central nervous system side effects are minimal. Domperidone increases lower esophageal sphincter pressure, accelerates gastric emptying, and enhances gastroduodenal coordination. Because brainstem structures regulating vomiting are outside the blood-brain barrier, domperidone is also a potent antiemetic. The oral form is approved in most countries, except for the United States. An intravenous formulation was withdrawn after fatal cardiac arrhythmias were reported.
At least 5 controlled trials and 4 open case series have assessed domperidone in patients with gastroparesis and diabetic gastropathy. Symptoms improved in all studies, but accelerated gastric emptying was not uniformly observed. The drug’s benefits on symptoms and quality of life are sustained while its prokinetic effects wane, emphasizing the importance of its antiemetic actions. Because of its action only on peripheral dopamine receptors, domperidone is also an especially useful drug for Parkinson’s disease patients who have gastrointestinal dysmotility. Adverse reactions to domperidone are uncommon, but hyperprolactinemic effects may develop due to the porous blood-brain barrier in the anterior pituitary.Domperidone has effects to prolong the QTc interval on electrocardiographic testing; an intravenous form of the drug was withdrawn from overseas markets because of fatal cardiac events. A more recent population-based study does suggest an increased risk for sudden cardiac death for patients on oral domperidone therapy.
The US Food and Drug Administration has not approved domperidone for prescription use, but the drug has been obtainable from foreign pharmacies, the Internet, and compounding pharmacies in the United States. The FDA discourages these practices but has made the drug available in the past under the auspices of a program to academic clinicians who submit an Investigational New Drug application to the FDA and who obtain Institutional Review Board approval.
Other prokinetic drugs
Although other gastrokinetic drugs are used in the treatment of gastroparesis, trials documenting their benefits have not been performed. An unpublished 8-week controlled trial of the 5-HT4partial agonist tegaserod observed accelerated gastric emptying at daily doses of 18 and 24 mg, although symptom responses were not quantified. However, tegaserod was withdrawn from the US market in 2007 due to a reported increase in the risk for cardiovascular adverse events. The FDA is allowing limited prescription of tegaserod in selected cases under an investigational new drug program.
Cisapride is a 5-HT4 agonist with weak 5-HT3 antagonist properties that once was widely used for gastroparesis. The drug was withdrawn from the US market in 2000 due to numerous sudden cardiac deaths. Although cisapride is still obtainable from overseas Web sites, a recent consensus document discouraged its use.
Bethanechol is an approved smooth muscle muscarinic agonist that increases lower esophageal sphincter pressure and evokes fundoantral contractions but does not induce propulsive contractions or accelerate gastric emptying. Prominent adverse effects include abdominal cramps, nausea, vomiting, diaphoresis, bronchoconstriction, urinary urgency, hypotension, and atrial fibrillation. Likewise, cholinesterase inhibitors have little effect on gastric emptying. However, pyridostigmine has recently been noted to reduce symptoms in a patient with gastroparesis secondary to underlying autoimmune disease.
Other macrolides, including clarithromycin and azithromycin, have similar prokinetic effects as erythromycin, but their use in gastroparesis is uninvestigated. Mitemcinal, a newer oral motilin agonist without antimicrobial action, produced symptom benefits in patients with diabetic gastropathy which were restricted to those with body mass indexes < 35 kg/m2 and with hemoglobin A1c values < 10%.Recent studies demonstrate potent stimulatory effects for parenteral ghrelin, an endogenous neurohumoral mediator involved in food intake.
Antiemetic Medication Therapy
Antiemetic agents without motor stimulatory activity are used alone or in combination with prokinetic drugs for gastroparesis; however, no trials have been performed to support this practice. Antiemetic medications act on a broad range of distinct receptor subtypes (Table 4). The most common antiemetic drugs are the phenothiazines (eg, prochlorperazine, thiethylperazine), which are brainstem dopamine and muscarinic antagonists. Thiethylperazine reportedly reduced symptoms in a patient with diabetic gastroparesis. Tricyclic antidepressants reduce symptoms in patients with functional vomiting.In a recent publication, 88% of diabetic patients with nausea and vomiting reported benefits with tricyclic antidepressants. One third of patients had delayed gastric emptying, suggesting that these agents may have utility in gastroparesis. The antidepressant mirtazapine also reduced gastroparesis symptoms in a patient with type 1 diabetes. However, formal prospective trials of these antidepressants for the treatment of gastroparesis have not been performed, thus their use is considered off-label. The benefits of other antiemetic drug classes in gastroparesis, including serotonin 5-HT3 antagonists, muscarinic M1antagonists, histamine H1 antagonists, neurokinin NK1 antagonists, cannabinoids, and benzodiazepines, are unproved. Corticosteroids are employed as antiemetics in the postoperative setting or in the prevention of chemotherapy-induced emesis. One individual with idiopathic myenteric ganglionitis exhibited improvement with corticosteroid therapy, confirming the inflammatory basis of some cases of upper gut dysmotility.
Complementary and alternative therapies are increasingly used to treat nausea and vomiting of diverse etiologies. Ginger, a Chinese remedy with weak 5-HT3 antagonist properties, has antiemetic actions in some settings but its benefit in gastroparesis is unexplored. Acupressure and acustimulation on the P6 point reduce nausea postoperatively, after chemotherapy, and during first-trimester pregnancy. One group observed benefits with acupuncture in diabetic gastroparesis.
Control of Symptoms Other Than Nausea and Vomiting
Symptoms other than nausea and vomiting may predominate in gastroparesis. Early satiety relates to impaired fundic accommodation in functional dyspepsia. Nitrates, buspirone, sumatriptan, and selective serotonin reuptake inhibitors promote fundic relaxation in this condition.[139,140] Epigastric pain is disabling in some gastroparesis patients. As in functional dyspepsia, pain in gastroparesis may stem from sensory rather than motor dysfunction. The benefits of agents that relax the fundus or reduce visceral sensation have not been studied in gastroparesis. Finally, the ability of prokinetics to stabilize glycemic control has been studied in gastroparetics with poorly controlled diabetes. Some studies report reductions in glucose levels with erythromycin or cisapride but others observe no effect on either short-term or long-term glycemic control, and thus no consensus opinion has formed on this issue.
Therapeutic endoscopy may provide benefit in some patients with gastroparesis. Injection of botulinum toxin into the pylorus reduces phasic contractions and tone by preventing unopposed cholinergic contractile activity.[1,143] Several uncontrolled case series have reported reduced symptoms and acceleration of gastric emptying after botulinum toxin treatment.[144,145] In the largest series, 43% of 63 patients experienced benefits for a mean of 5 months. An unpublished study involving 78 patients observed similar response rates in diabetic (55%), idiopathic (51%), and postsurgical (44%) gastroparesis. In this study, higher doses of botulinum toxin (150–200 units) had greater symptom benefits than lower doses (75–100 units). Recent unpublished, blinded controlled trials have not demonstrated efficacy of botulinum toxin vs placebo, but these studies are small and may not be adequately powered to detect a therapeutic benefit.[148,149] Use of botulinum toxin for gastroparesis is considered off-label. Other endoscopic procedures provide relief in some cases of gastroparesis. Venting gastrostomies intermittently release retained gas and liquid, reducing fullness, discomfort, and distention-related nausea.
Surgical intervention is increasingly employed to treat refractory gastroparesis. The most common operation, gastric electrical stimulator implantation, has been performed in more than 1500 patients over the past decade. Gastric resection and pancreatic transplantation for diabetic patients are performed less often and their benefits are not well documented.
Gastric electrical stimulation
Recent research suggests that some cases of gastroparesis may respond to stimulation with electrical depolarizing stimuli. In a trial of high-energy gastric pacing at a rate slightly above the normal slow-wave frequency in 9 patients with medication-resistant gastroparesis (5 diabetic, 3 idiopathic, 1 postsurgical), slow-wave entrainment was observed, underlying dysrhythmias were abolished, gastric emptying accelerated, and symptoms improved to the point that 8 patients no longer required supplemental jejunal nutrition. However, this method was impractical for long-term use because the current source needed to pace the stomach was too bulky for implantation.
Because of these drawbacks of pacing, researchers investigated the benefits of other lower-energy stimulation parameters. Initial unpublished series employing an implantable gastric neurostimulator that delivered low-energy impulses at 12 cycles per minute reported reduced symptoms in patients with medication-unresponsive gastroparesis. Due to this apparent benefit, the stimulator received FDA approval as a humanitarian use device and was granted a humanitarian device exemption for application in patients with diabetic and idiopathic gastroparesis. For this judgment, the FDA deems the device to have probable benefit with the expectation that controlled clinical trials will ultimately be performed, and restricts its use to centers in which Institutional Review Board approval has been granted.
The gastric stimulator has since shown efficacy in uncontrolled studies in diabetic, idiopathic, and postsurgical gastroparesis. In one uncontrolled multicenter trial, 35 of 38 patients experienced greater than 80% reductions in nausea and vomiting which persisted for 2.9–15.6 months, with an associated 5.5% increase in weight and reduced requirements for supplemental nutrition. Other studies reported similar long-term symptom benefits which may persist for at least 10 years, with associated improvements in body mass index, serum albumin, and glycemic control.[153,154] In the only controlled trial of gastric electrical stimulation, 33 patients with idiopathic or diabetic gastroparesis completed a 2-month double-blind, crossover, sham stimulation-controlled phase followed by a 12-month uncontrolled observation period with the device activated. During the blinded phase, frequency of vomiting was 14% lower when the device was on as opposed to off. Symptom reductions were more impressive during the unblinded phase of the study. Complications including infection, lead dislodgement, and bowel obstruction necessitated device removal in over 10% of patients; up to one fourth of patients have undergone gastrectomy in some series because of a lack of response. The mechanism of action of gastric stimulation does not involve acceleration of gastric emptying or slow-wave stabilization.[152,155] Rather, in animal and human models, gastric stimulation acts via vagal pathways to promote fundic relaxation and reduce afferent hypersensitivity.[156–158]
Other operative interventions
Other operations rarely are considered for patients with gastroparesis unresponsive to drug therapy. Surgical pyloroplasty produced benefit in 1 unpublished series of diabetic patients with gastroparesis, but a more recent report observed symptom improvements in only one third of patients.[159,160] Gastroenteric anastomosis reconstruction (ie, conversion of a Billroth I to a Billroth II or vice versa) is rarely effective. Performance of completion gastrectomy with preservation of only a small cuff of gastric tissue provides long-term symptom relief in 43% to 67% of patients with postsurgical gastroparesis.[161,162] The utility of subtotal gastrectomy for diabetic gastroparesis was promoted in a small series of 7 patients. However, extraintestinal complications developed in 3 patients, 2 of whom died; thus, the benefits of surgery were difficult to evaluate. Pancreas transplants prevent progression of preexisting diabetic complications such as neuropathy or retinopathy, although studies report conflicting data on the benefits of this operation in patients with diabetic gastroparesis.
Enteral and Parenteral Nutrition
Some patients with refractory gastroparesis benefit from enteral or parenteral nutrition intermittently for symptom flares or for permanent support. The benefits of parenteral nutrition are clinical observations and have not been reported in the literature. In diabetic gastroparesis, jejunostomy placement for enteral feeding improves overall health status with trends towards reduced symptoms and hospitalization rates and enhanced nutrition. Indications for enteral nutrition include significant malnutrition (eg, greater than 10% weight loss over 6 months) unresponsive to dietary modification, essential mineral deficiencies or electrolyte disturbances, and frequent hospitalizations producing profound disability. Short-term total parenteral nutrition (TPN) can reverse rapid weight decline and ensure adequate fluid delivery, but permanent TPN usually is needed only for individuals with superimposed severe intestinal dysmotility who cannot tolerate enteral feeding.
Patients with gastroparesis commonly experience psychological sequelae of their gastrointestinal disease, including anxiety, depression, and somatization. The degree of psychological impairment correlates strongly with symptom severity in patients with diabetic gastroparesis. Despite these observations, the role of the mental health specialist in managing gastroparesis has not been defined. Small studies report benefits with biofeedback or hypnosis.
Investigation into the natural history, pathogenesis, evaluation, and management of gastroparesis will be active in the coming years. Use of quantitative symptom questionnaires will facilitate stratification of patients into subtypes on the basis of symptom predominance and severity. Research on novel prokinetic therapies, as well as on treatments that relieve symptoms resulting from pathophysiologic defects other than delayed gastric emptying, will be emphasized.
This work was supported by a grant 1 U01 DK073985-01 from the National Institutes of Health.
Reader Comments on: Gastroparesis – Current Concepts and Considerations See reader comments on this article and provide your own.
Readers are encouraged to respond to the author at ude.hcimu@relsahw or to George Lundberg, MD, Editor in Chief ofThe Medscape Journal of Medicine, for the editor’s eyes only or for possible publication as an actual Letter in the Medscape Journal via email: ten.epacsdem@grebdnulg