Pharmacologic treatments for gastroparesis

Abstract

Gastroparesis is a neurogastrointestinal disorder of motility in which patients experience symptoms of nausea, vomiting, bloating, early satiety, postprandial fullness, upper abdominal discomfort or pain, and delayed gastric emptying of solids based on scintigraphy or stable isotope breath test when mechanical obstruction has been excluded. Symptoms of gastroparesis may result from diverse pathophysiological mechanisms, including antroduodenal hypomotility, pylorospasm, increased gastric accommodation, and visceral hypersensitivity. The most common etiologies of gastroparesis are idiopathic, diabetic, and postsurgical, and less frequent causes are neurodegenerative disorders (Parkinson’s disease), myopathies (scleroderma, amyloidosis), medication-induced (glucagon-like peptide-1 agonists and opioid agents), and paraneoplastic syndrome. This review addresses pharmacologic management of gastroparesis including prokinetic and antiemetic agents, pharmacologic agents targeting the pylorus, and effects of neuromodulators.

Significance Statement

Gastroparesis is a neurogastrointestinal motility disorder characterized by delayed gastric emptying without mechanical obstruction with numerous upper gastrointestinal symptoms, including nausea and vomiting. The management of gastroparesis involves nutritional support, medications, and procedures. The only Food and Drug Administration-approved medication for gastroparesis is metoclopramide. This article reviews the pharmacology and efficacy of all classes of antiemetics or prokinetic effects used in gastroparesis. There is still a considerable unmet need for efficacious medications specifically for the treatment of gastroparesis, especially in refractory cases.

I. Introduction to gastroparesis: Definition, etiology, and impact

Gastroparesis is a neurogastrointestinal motility disorder characterized by delayed gastric emptying where patients experience upper gastrointestinal symptoms such as nausea, vomiting, early satiety, postprandial fullness, bloating, belching, and epigastric pain or discomfort. This diagnosis cannot be made without objective evidence of delayed gastric emptying of solids and exclusion of mechanical obstruction (Camilleri et al, 2022). The key symptoms of gastroparesis are nausea and vomiting (Schol et al, 2021). There is, however, often significant overlap of symptoms with a subset of functional dyspepsia, specifically postprandial distress syndrome, which is characterized by epigastric pain or discomfort, bloating, and fullness (Pasricha et al, 2021). If the primary symptom is abdominal pain, the diagnosis of gastroparesis should be questioned, particularly if there is chronic dependence on opioids (Hasler et al, 2019; Jehangir et al, 2019a; Parkman et al, 2019). It is important to emphasize that functional dyspepsia, which presents with very similar symptoms to gastroparesis, has a much higher prevalence in the community compared to gastroparesis. For example, in Olmsted County, Minnesota, there were studies conducted around the same time to define prevalence of functional dyspepsia and gastroparesis. The prevalence of dyspepsia was 15 per 100 persons (95% confidence interval [CI], 14–17), and pain was the most prevalent symptom. Among 351 dyspeptic subjects, 51% (95% CI, 46–56) reported upper abdominal pain, 21% (95% CI, 16–25) nausea or vomiting, and 47% (95% CI, 42–52) early satiety (Choung et al, 2007). In the same community, the age-adjusted prevalence of definite gastroparesis (including delayed gastric emptying by scintigraphy) was 24.2 (95% CI, 15.7–32.6), and possible gastroparesis (based on typical symptoms >3 months) was 50.5 (95% CI, 38.1–62.8) per 100,000 persons (Jung et al, 2009).

The pivotal symptoms of gastroparesis, nausea, early satiety, postprandial fullness, upper abdominal pain, and vomiting, are documented with an instrument called the Gastroparesis Cardinal Symptom Index (GCSI) (Revicki et al, 2009) which has been extensively validated (Revicki et al, 2012, 2018, 2019). When the GCSI is used as a daily diary, it is included in the Food and Drug Administration (FDA) guidance for drug development in gastroparesis (https://www.fda.gov/regulatory-information/search-fda-guidance-documents/gastroparesis-clinical-evaluation-drugs-treatment-guidance-industry).

Idiopathic, diabetic, and postsurgical (resulting from fundoplication, bariatric procedures) gastroparesis are the most common causes of gastroparesis. Less common disorders include neuropathic (Parkinson’s disease, paraneoplastic syndrome) and myopathic (scleroderma, amyloidosis) gastroparesis (Soykan et al, 1998; Parkman et al, 2011a). There are also medications including glucagon-like peptide-1 agonists and opioid agents which can cause gastroparesis (Camilleri et al, 2017a; Sodhi et al, 2023).

The severity of symptoms in gastroparesis contributes to significant disability in 11% of patients, decreased annual income in 28.5% of patients, and reduced activities of daily living in 67.5% of patients (Lacy et al, 2018). The caregivers of patients with gastroparesis are also prone to significant caregiver fatigue (Jehangir et al, 2019b). Patients with gastroparesis have significantly decreased survival compared to age- and sex-matched reference populations (Jung et al, 2009). In patients with diabetic gastroparesis, there are significant concurrent co-morbidities (Jung et al, 2009; Ye et al, 2021). Gastroparesis has been associated with increased emergency room visits and increased costs associated with hospitalizations (Wadhwa et al, 2017; Hirsch et al, 2019).

Understanding the pharmacologic treatment of gastroparesis requires a review of the normal emptying of food from the stomach, as well as the neural, neurogenic, and myogenic mechanisms controlling gastrointestinal motor function and the specialized function of the pylorus.

A. Normal emptying of food from the stomach

The stomach is composed of 3 functional regions: the fundus, the corpus (also known as the body) and antrum (distal part of the stomach), and the pyloric sphincter. The fundus is the proximal segment of the stomach which creates tone or relaxation to accommodate ingested food. The distal stomach (corpus and antrum) aids in the mixing and processing of food before emptying. The pyloric sphincter regulates the rate of gastric emptying (Hinder and Kelly, 1977).

After food is ingested, gastric accommodation is triggered, followed by antral contractions that triturate food to a particle size of 1–2 mm. Such trituration is necessary to empty digestible solids from the stomach (Meyer et al, 1979). The tone of the fundus and resulting pressure gradient propel lower calorie containing liquid through the stomach (Moragas et al, 1993). The peristaltic waves travel through the stomach, reaching the distal terminal antrum near the pylorus. The pylorus constricts to prevent emptying of the stomach during the high-pressure period to allow for shear forces and retropulsing to help mechanically digest foods in the stomach (Carlson et al, 1966; Hinder, 1983). With pyloric relaxation and coordination of antropyloroduodenal contractions, homogenized solid food empties from the stomach. Foods of different physical nature and consistency follow distinct emptying patterns (Camilleri and Shin, 2013): exponential for low-calorie liquids; linear for nutrient liquids and homogenized solids; and for solids, an initial phase of retention in the proximal stomach corresponding to the lag time, followed by a linear emptying pattern for the adjustable solids. Solids that are not digestible empty from the stomach and pylorus long after the emptying of digestible solids (Hinder and Kelly, 1977), typically with interdigestive migrating motor complexes (Cassilly et al, 2008).

Antral phasic pressure activity correlates significantly with the rate of gastric emptying of solids in healthy stomach (Camilleri et al, 1985). Hypomotility, typically characterized by <1 distal antral contraction on average per minute in the first postprandial hour, results in prolongation of the lag duration and rate of gastric emptying (Thumshirn et al, 1997).

B. Mechanisms controlling gastric motility

1. Myogenic mechanisms

Smooth muscle cells (SMCs) generate tension and contraction forces necessary for trituration of food and gastric emptying through the pylorus. SMCs are muscle cells that are spindle-shaped, 40–100 μm long, 2–8 μm in diameter, and tightly consolidated with little connective tissue between cells. SMCs are organized into 3 layers (circular, longitudinal, and oblique) in the stomach. They are electrically coupled by gap junctions, which facilitate synchronization of the contractions. Gastric contractions depend upon an increase in cytoplasmic calcium ions (Ca²⁺).

Voltage-dependent Ca²⁺ channels (dihydropyridine sensitive, L-type channels) serve as the primary regulators for SMCs excitation-contraction coupling through Ca²⁺ entry. However, SMCs also have potassium channels that regulate resting membrane potentials and the state of basal excitability in the different regions (proximal, distal, and pyloric) of the stomach. SMCs in the fundus are intrinsically more depolarized than SMCs in the antrum, with continuous leak of Ca²⁺ into cells through voltage-dependent channels and conductances. This leads to maintenance of tone. Antral and pyloric SMCs have higher negative resting membrane potentials and less Ca²⁺ entry, and hence relaxation between excitable events. SMCs are coupled to 2 types of interstitial cells in the muscle layer and myenteric plexus: interstitial cells of Cajal (ICC) and fibroblast-like cells [platelet-derived growth factor receptor α cells (PDGFRα⁺ cells)]. They are coupled to SMCs by gap junctions (Ishikawa et al, 1997; Horiguchi and Komuro, 2000) forming an electrical syncytium known as the SIP (smooth muscle cell/ICC/PDGFRα+ cell) (Sanders et al, 2012). The intramuscular ICC (ICC-IM) are closely aligned with varicose projections of enteric motor neurons (Burns et al, 1996; Blair et al, 2012b). ICC in the myenteric region (ICC-MY) form a complex cellular network located between circular and longitudinal muscle layers (Dickens et al, 1999) and generate pacemaker activity (Ordög et al, 1999). The pacemaker activity is generated through unique ionic conductances that propagate slow electrical waves. This achieves depolarization, activation of L-type Ca²⁺ channels, and excitation-contraction coupling that generates the phasic contractions of gastric peristalsis. This intact network of ICC-MY is essential for gastric peristalsis and motility, as SMCs do not possess ion channels needed for the propagation of slow waves (Sanders et al, 2014).

C. Neuromuscular control mechanisms

1. Intrinsic innervation

The enteric nervous system is the intrinsic nervous system of the entire gut and consists of over 100 million neurons (Spencer and Hu, 2020). The enteric nervous system is organized in distinct ganglionated plexi, including the submucous plexus which is important for absorption and secretion, and the myenteric plexus which regulates motility (Spencer and Hu, 2020). The gastric muscularis receives both excitatory and inhibitory innervation from enteric motor neurons (Furness et al, 2020). Excitatory neurons release acetylcholine (ACh) and tachykinins. Inhibitory neurons release nitric oxide (NO), vasoactive intestinal polypeptide, pituitary adenylate cyclase-activated peptide, and purines. The cell bodies of muscle motor neurons reside in myenteric ganglia, with excitatory neurons projecting proximally and inhibitory neurons projecting distally (Brookes et al, 1998).

Motor neurons innervate the SMCs through the SIP (Blair et al, 2012a; Sanders et al, 2014) (Fig. 1). ICC-IM mediate responses to cholinergic excitatory (Ward et al, 2000) and nitrergic inhibitory (Burns et al, 1996) signals of neurotransmission. PDGFRα⁺ cells facilitate responses to purinergic inhibitory neurons in gastrointestinal muscles (Kurahashi et al, 2011). Dominant expression of key receptors and ion channels enables the inhibitory purinergic response in PDGFRα⁺. Acetylcholine causes a dramatic increase in Ca²⁺ release in ICC-IM in the colon (Drumm et al, 2020). This has not yet been documented in gastric ICC-IM. In contrast, NO inhibits Ca²⁺ release in ICC-IM.

Fig. 1.

Motor neurons innervate SMCs through ICC and PDGFRα⁺ cells. SMC, smooth muscle cell; ICC, interstitial cell of Cajal; PDGFRα, platelet-derived growth factor receptor α. Reproduced with permission from Sanders et al (2014).

Excitatory nerve stimulation in the stomach results in inotropic effects (enhancing the amplitude and duration of slow waves, resulting in stronger peristaltic contractions) and chronotropic effects, increasing the frequency of slow waves caused by cholinergic stimulation of ICC-IM but not ICC-MY (Hirst et al, 2002). Figure 2 shows myogenic mechanisms controlling gastric smooth muscle (Camilleri and Sanders, 2022).

Fig. 2.

(A–C) c-Kit (A) and Ano1 (B) immunolabeling and merged (C) files of ICC-IM from monkey gastric fundus; (D–F) ICC-MY shown by same immunolabels (D = c-Kit; E = Ano1; F = merged) from monkey antrum; (G) network of ICC-MY (c-Kit labeling) shown with myenteric plexus (PGP 9.5 labeling) of guinea pig stomach; (H) gastric map used by surgeons to show where gastric muscles originate; (I–J) slow waves recorded from human gastric antral muscle (area 14). Red arrow denotes upstroke phase of slow wave, and the green arrow denotes the plateau phase; (J) also shows simultaneous recording of slow waves (above) and phasic contractions (below). In the intact stomach, both the slow waves and the contractions they induce propagate from corpus to the pyloric sphincter, constituting gastric peristalsis. ICC-MY, myenteric interstitial cells of Cajal. Reproduced with permissions from Camilleri and Sanders. Gastroparesis. Gastroenterology 2022;162:68-87.e1.

2. Extrinsic innervation

The extrinsic innervation of the stomach involves parasympathetic and sympathetic nerves (Lu et al, 2018; Camilleri, 2021). The dorsal motor nucleus of the vagus nerve houses the cell bodies of parasympathetic efferent neurons which innervate myenteric neurons. Vagal efferent neurons directly innervate more than 70% of gastric excitatory and inhibitory motor neurons (Schemann and Grundy, 1992). Vago-vagal reflexes are essential for many responses in the stomach to eating, including gastric accommodation, through activation of inhibitory motor neurons via release of NO (Desai et al, 1991). Sympathetic nerves originate in the intermediolateral cell column of levels 5–10 in the thoracic spinal cord. They synapse with postganglionic neurons in the celiac and superior mesenteric ganglia to innervate myenteric ganglia and muscle layers directly. This results in inhibition of gastric contractions, chiefly by prejunctional inhibition of vagal inputs to excitatory motor neurons.

D. Specializations of the pyloric sphincter

At the junction between the stomach and duodenum, the pyloric sphincter is a narrow zone of thickened muscularis (radiologically estimated to be 1.2 cm in width) with increased luminal pressure and tone. The transduction of neural signals by ICC and conduction of responses to SMCs are used to coordinate pyloric contractions and relaxations (Ward et al, 1998). Pyloric muscles consist of 2 independently controlled, functional areas. First, the circular muscle close to the myenteric plexus contracts with propagation of gastric slow waves at the termination of each gastric peristaltic event and participates in antropyloroduodenal coordination of contractions. The second is the deeper, thickened circular muscle regulated by motor neurons, into which antral slow waves do not propagate (Sanders and Vogalis, 1989). Temporal and functional regulation of pyloric resistance may be achieved by independent control of the circular muscle close to the myenteric plexus and the deeper, thickened circular muscle.

Enteric inhibitory neurotransmission is mediated primarily by NO and by a purine neurotransmitter (Bayguinov and Sanders, 1993b; Camilleri and Sanders, 2020) through activation of small conductance Ca²⁺-activated potassium channel. This leads to inhibitory junction potentials resulting in pyloric relaxation. Conversely, inhibition of inhibitory junction potentials leads to unopposed excitatory junction potentials in the myenteric region, increased SMC excitability, and contractions in the submucosal region of the pylorus. Both the myenteric and submucosal layers of pyloric SMCs are innervated by nitrergic neurons and are dependent on ICCs (Ward et al, 1994). Lesions in the nitrergic pathway, such as loss of NO synthase, loss of NO synthase neurons, or loss of ICCs that contribute to nitrergic signals neurotransduction, result in abnormal regulation of pyloric relaxation and may impede gastric emptying. There are reduced or abnormal ICCs in full-thickness biopsies from gastric muscles and pylorus of patients with idiopathic and diabetic gastroparesis (Grover et al, 2011; Bashashati et al, 2017). In canine pyloric muscles, enkephalinergic nerve fibers in the muscle layer mediate effects of endogenous opiates. This leads to inhibition of both cholinergic excitatory and nitrergic inhibitory junction potentials (Bayguinov and Sanders, 1993a). Exogenous opioids interfere with normal neural regulation of the pylorus and cause stimulation of pyloric tone and phasic contractility (Camilleri et al, 1986) by inhibition of nitrergic relaxation or by cholinergic stimulation (Mearin et al, 1987).

II. Pathophysiology of gastroparesis

A. Neuromuscular diseases affecting nonsphincteric gastric muscles

Figure 3 summarizes the cellular elements that could be deranged in disorders of extrinsic and enteric neural control and muscle, resulting in motility disorders (Camilleri, 2021). Idiopathic and diabetic gastroparesis are typically associated with hypomotility with reduced postprandial antral frequency (average <1/minute) and normal amplitude of contractions (Thumshirn et al, 1997). Conversely, scleroderma, and other infiltrative disorders result in low-amplitude antral (<40 mm Hg) and intestinal contractions (<10 mm Hg) (Thumshirn et al, 1997). This delays gastric emptying and small bowel transit and prevents ileocolonic bolus transfers (Greydanus et al, 1990).

Fig. 3.

Examples of diseases causing gastrointestinal motility disorders. The neurological diseases may affect extrinsic parasympathetic and sympathetic nerves as well as the enteric nervous system or smooth muscle within the gastrointestinal tract. CNS, central nervous system; ENS, enteric nervous system; MNGIE, mitochondrial neurogastrointestinal encephalopathy. Reproduced with permission from Camilleri M. Invited Review: Gastrointestinal motility disorders in neurologic disease. J Clin Invest 2021;131:e143771. (This is a free PMC article; thus, no permission is required to reuse the figure.)

Neurological diseases such as diabetic neuropathy or brainstem diseases may also cause gastroparesis as a result of abnormal parasympathetic vagal nerves and other autonomic nuclei. In addition, gastroparesis may result from neuropathies and dysautonomic disorders affecting sympathetic nerves, or from diseases affecting primarily the enteric nervous system or smooth muscle including myopathies such as amyloidosis and mitochondrial neurogastrointestinal encephalopathy (Camilleri, 2021) (Fig. 3).

B. Postsurgical or postbariatric surgery or endoscopy gastroparesis

Truncal vagotomy performed for peptic ulcer disease used to be associated with pyloroplasty or gastrojejunostomy. Partial gastrectomy used to be performed in the past for treatment of peptic ulcer disease. It is now typically performed as part of bariatric surgery. Delayed emptying from the gastric remnant is associated with either extrinsic denervation of the gastric remnant or abnormal motility in the anastomosed jejunal loop (Mathias et al, 1985; Miedema et al, 1992). Proximal gastric vagotomy inhibits gastric tone and delays the gastric emptying of liquids while maintaining antral contractility (Hould et al, 1994). Rarely, vagal denervation may result from cardiac ablation for arrhythmia (Park et al, 2017b).

Fundoplication may result in vagal injury, which is associated with impaired antral motility and gastroparesis (Stanghellini and Malagelada, 1983). Uncomplicated fundoplication (Maddern and Jamieson, 1985) and sleeve gastrectomy (Vargas et al, 2020) reduce the reservoir capacity of the stomach, accelerating the gastric emptying of solids.

Sleeve gastroplasty delays gastric emptying by creating a funnel-shaped stomach with a constricted middle and distal stomach (Abu Dayyeh et al, 2017). When the distal antrum is included in the gastroplasty, solid trituration is likely impacted, hindering the emptying of solids from the stomach into the duodenum.

C. Pyloric dysfunction

Pyloric dysfunction in diabetic gastroparesis is characterized by prolonged, but intermittent contractions characterized by increased tone at the pylorus, or “pylorospasms.” This is often associated with, antral hypomotility and evidence of extraintestinal autonomic neuropathy (Mearin et al, 1986). These effects may result from damage to the ICC-MY network (antral hypomotility) or loss of the pyloric ICC-IM. It is also known that opioid use is associated with pyloric dysfunction and gastric stasis. Recently, the pylorus has become a target for endoscopic treatment, as described below.

Pyloric studies using the EndoFLIP device (a fluid-filled cylindrical balloon housing an array of impedance electrodes spaced 4 or 8 mm apart and 1 solid state pressure transducer to determine intraballoon pressure) in 15 diabetic and 39 idiopathic patients with gastroparesis documented a wide range in pyloric diameter (5.6–22.1 mm) and distensibility (1–55 mm²/mm Hg) (Malik et al, 2015). The pyloric EndoFLIP study also found early satiety and postprandial fullness to be inversely correlated with pyloric sphincter diameter and cross-sectional area (Malik et al, 2015). Moreover, in patients with gastroparesis, there was a decrease in pyloric distensibility, and the decrease in distensibility was correlated with degree of gastric emptying and symptoms of gastroparesis (Wuestenberghs and Gourcerol, 2021).

D. Autoimmune gastritis

Another rare disorder that may present with gastroparesis-like symptoms is autoimmune gastritis (AIG), which may occur in children or adults. It is characterized by histological corpus-predominant atrophic gastritis, with or without positive antiparietal cell or anti-intrinsic factor antibodies. It is also characterized by auto-antibodies against the proton pump H⁺/K⁺ adenosine triphosphatase and other autoimmune comorbidities such as celiac disease, autoimmune thyroiditis or hepatitis, and insulin-dependent diabetes (Granot et al, 2024). The pathogenesis of AIG involves T and B lymphocytic infiltration of the gastric mucosa leading to destruction of parietal cells, hypochloridria, and vitamin B12 deficiency. Patients also experience nonspecific gastro-intestinal symptoms such as upper abdominal discomfort, bloating, and nausea (Rossi et al, 2023). Among 41 patients with AIG, 22 had evidence of delayed gastric emptying, although the data (GE T½: 241.19 ± 199 [SD]) were extrapolated based on a 2-hour test that evaluated gastric emptying of a 300-kcal radiolabeled egg-white meal (Kalkan and Soykan, 2018). Further studies are required to replicate these associations of AIG and also to assess other functions, such as gastric accommodation and sensation.

III. Management of gastroparesis

The management and treatment of gastroparesis require a multifaceted approach of dietary modification, restoration of hydration and nutrition, symptoms alleviation, and identifying and treating the underlying pathophysiological mechanisms.

A. Hydration and nutrition

For patients with severe fluid or metabolic derangements (ketoacidosis, renal insufficiency, altered glycemic control) due to nausea and vomiting or underlying metabolic diseases, restoration of hydration and electrolytes is essential through per-oral or intravenous routes when necessary (Camilleri, 2007). In addition, nutritional deficiencies are highly prevalent. Up to 64% of patients with gastroparesis consume diets that are deficient; that is, the patients consume less than 60% of the estimated total energy requirements, and vitamin (A, B6, C, K) and mineral (iron, potassium, zinc) deficiencies are common (Parkman et al, 2011b). The first step of dietary modification involves homogenizing solids to smaller particle size and cooking of nondigestible fibers. This simple first step significantly reduced the severity of nausea, vomiting, postprandial fullness, bloating, and regurgitation/heartburn in patients with diabetic gastroparesis (Olausson et al, 2014). High fat and nondigestible fiber retard gastric emptying and should be avoided. If solid or homogenized foods are not tolerated, stepwise nutritional interventions include use of liquid meals, oral nutrition supplements, enteral nutrition, and parenteral nutrition (Limketkai et al, 2020). Percutaneous jejunal feeding is safe and leads to weight regain (Fontana and Barnett, 1996). Parenteral nutrition is reserved for temporary use in patients with severe nutritional deficiency, as it is associated with increased risk of complications such as infections and thromboses (Bharadwaj et al, 2016).

B. Pharmacologic agents

Figure 4 shows a conceptual summary of pharmacologic approaches to treat gastroparesis (Ramprasad et al, 2018).

Fig. 4.

Conceptual summary of pharmacologic approaches to treat gastroparesis. Among these medications, metoclopramide is the only one approved by FDA. Adapted with permission from Ramprasad et al. Curr Treat Options Gastro 2018;16:489-510.

1. Prokinetics

Despite optimal supportive care with dietary strategies, most patients with gastroparesis continue to experience symptoms. The 2013 guideline from the American College of Gastroenterology recommended prokinetic agents as the first-line therapy for gastroparesis (Camilleri et al, 2013). The 2022 guideline (Camilleri et al, 2022) recommended the consideration of therapies that target gastric emptying and symptoms of gastroparesis, while weighing the benefits and risks of the agent in patients with idiopathic and diabetic gastroparesis. Although the correlation between delay in gastric emptying and severity of symptoms remains controversial, prokinetic agents in clinical trials have shown efficacy for both enhancing gastric emptying and reducing symptoms. In a systematic review of studies with optimal gastric emptying tests that assessed randomized, blinded, parallel, or crossover trials of serotonin (5-HT)₄ agonists, dopamine (D)₂ receptor antagonist, or ghrelin agonists, the meta-regression showed a positive association between improvement in gastric emptying [especially when gastric emptying half-time (T_1/2) accelerated by at least 20.4 minutes] and upper gastrointestinal symptoms (Vijayvargiya et al, 2019). This conclusion has been independently confirmed (Goelen et al, 2023).

2. Dopaminergic modulation

The only FDA-approved pharmacologic agent for gastroparesis is metoclopramide. Metoclopramide was FDA-approved for gastroparesis in 1979. It functions through antagonism of central and peripheral dopamine receptors. Metoclopramide is an antiemetic that acts centrally through the inhibition of D₂ and 5-HT₃ receptors in the area postrema located outside the blood–brain barrier and is the target of antiemetics (Sanger, 2009). Peripherally, metoclopramide exerts its prokinetic effects through 5-HT₄ receptor agonism (which leads to acetylcholine release) and through presynaptic and postsynaptic D₂ receptor antagonism (Tonini et al, 2004). Metoclopramide crosses the blood-brain barrier and can cause anxiety, agitation, somnolence, as well as usually reversible extrapyramidal symptoms including tremors, and, rarely, irreversible tardive dyskinesia. Because of the risk of neurological adverse effects, metoclopramide is only approved for a maximum of 12 weeks and carries a black box warning. The true risk of irreversible tardive dyskinesia caused by metoclopramide is actually very low, estimated to be 0.1% per 1000 patient-years (Rao and Camilleri, 2010) and more recently 0.14 per 100,000 patient years (Al-Saffar et al, 2019), rather than the 4% as estimated in FDA documents. In addition to its oral preparation, metoclopramide is also available in liquid, nasal spray (Parkman et al, 2015), and parenteral (eg, intravenous or subcutaneous) formulations (McCallum et al, 1991). Because gastroparesis is a chronic disease and there are no approved alternative agents, it is necessary to balance the FDA recommendation with the significant unmet need of patients. In particular, when there are symptoms that impair quality of life, dehydration, and malnutrition, the lowest effective dose of liquid, nasal, or tablet formulation, 5–10 mg three times a day 15 minutes before meals over the 12 weeks, should be used. If tolerated without adverse effects, experience demonstrates that with 10-day interruptions every 3 months as “drug holidays” during which patients adhere strictly to liquid or blenderized diet, the medication at a cumulative dose of 30 or 40 mg/day is tolerated and provides benefits to patients. Rescue antiemetic agents such as ondansetron, 4-mg oral dissolving tablets, or short-term (to avoid tachyphylaxis) treatment with erythromycin, 40–200 mg three times a day as tolerated, can be used as supportive agents (Camilleri, 2022). Table 1 summarizes clinical trials regarding metoclopramide for gastroparesis (Longstreth et al, 1977; Perkel et al, 1979, 1980; Snape et al, 1982; McCallum et al, 1983; Loo et al, 1984; Ricci et al, 1985; Erbas et al, 1993; Patterson et al, 1999; Parkman et al, 2014, 2015; McCallum et al, 2023).

Table 1.

Summary of clinical trials documenting efficacy of metoclopramide in gastroparesis. Adapted with permission from Camilleri et al (2022).

Design	Number, Etiology	Dose by Mouth	Duration	Results	Reference
DB, PC, PG RCT	28 pts: 5 DG, 4 vagotomy and pyloroplasty, and 19 IG	10 mg 4 times a day	3 wk	Symptom benefit vs placebo: mean TSS for metoclopramide:18.4 pre- to 7.2 post-study; for placebo, 19.1 pre- to 12.9 post-study; symptoms improved by 29%	Perkel et al, 1979
DB, PC, PG RCT	55 pts: 21 vagotomy and drainage, 5 DM, 29 IG delayed GE	10 mg 4 times a day	3 wk	Metoclopramide significantly decreased symptom scores of surgical and idiopathic patients	Perkel et al, 1980
DB, PC, XO, RCT	10 DM	10 mg 4 times a day	3 wk/arm	Improved symptoms and vomiting; ∼60% acceleration in GE liquid 150-kcal meal	Snape et al, 1982
PC, RCT	18 DG	10 mg 4 times a day	3 wk	Improved symptom score by 29%, GE by 25%	McCallum et al, 1983
DB, PC, XO, RCT	13 DM with GE accelerated by i.m. metoclopramide	10 mg 4 times a day	3 wk/arm	Improved symptoms with mean reduction of 52.6%	Ricci et al, 1985
DB, RCT	45 diabetic, domperidone- controlled, multicenter trial	10 mg 4 times a day	4 wk	Improved symptoms by 39%; similar efficacy with domperidone which had less AEs	Patterson et al, 1999
DB, XO, RCT	13 DG; erythromycin- controlled	10 mg 3 times a day	3 wk/arm	Both treatments accelerated GE compared to baseline, and improved symptoms score	Erbas et al, 1993
Open	1 diabetic	15 mg 4 times a day	6 mo	Improved symptoms, GE liquids, # of antral contraction	Longstreth et al, 1977
Open	10 GI symptomatic and 6 asymptomatic T1DM,18HC	10 mg i.v.	Single dose	Improved GE solids	Loo et al, 1984
Open, PG, RCT	89 T1DM or T2DM gastroparesis	10, 20-mg spray or 10-mg tab 4 times a day	6 wk	Nasal 10 and 20 mg had lower TSS compared to oral 10-mg group; More side effects, especially nausea with oral	Parkman et al, 2014
DB, PC, PG, RCT	285 T1DM or T2DM with delayed GE or nausea and vomiting.	10 or 14-mg nasal spray 4 times a day	4 wk	Gastroparesis symptom scores were reduced significantly in female subjects, not in males. AEs: dysgeusia, headache, and fatigue.	Parkman et al, 2015
DB, PC, PG, RCT	205 DM gastroparesis and delayed GE	10-mg nasal spray	4 wk	Overall, no significant reduction in symptoms from baseline to wk 4. However, pts with moderate-to-severe symptoms at baseline had reduction in nausea and upper abd. pain for all 4 wk vs placebo. AEs were primarily mild-to-moderate headache, abd. pain	McCallum et al, 2023

abd, abdominal; AE, adverse event; DB, double-blind; DG, diabetic gastroparesis; DM, diabetes mellitus; pts, patients; GE, gastric emptying; GI, gastrointestinal; i.m., intra-muscular; i.v., intravenous; PC, placebo-controlled; PG, parallel group; pts, patients; RCT, randomized controlled trial; T1DM, type 1 diabetes mellitus; T2DM, type 2 diabetes mellitus; tab, tablet; TSS, total symptom score.

Other marketed agents have been used in gastroparesis as off-label indications including domperidone, macrolides, and 5-HT₄ receptor agonists such as cisapride and prucalopride.

3. Domperidone

Domperidone is a peripherally-acting D₂ receptor antagonist that is available through the FDA’s Program for Expanded Access to Investigational Drugs. Its efficacy for the treatment of gastroparesis is comparable to metoclopramide (Patterson et al, 1999) and is summarized in Table 2 (Nagler and Miskovitz, 1981; Heer et al, 1983; Horowitz et al, 1985; Watts et al, 1985; Braun, 1989; Koch et al, 1989; Kozarek 1990; Soykan et al, 1997; Farup et al, 1998; Silvers et al, 1998; Patterson et al, 1999; Franzese et al, 2002; Sarosiek et al, 2022). A systematic review of 28 trials of domperidone conducted in patients with diabetic gastroparesis (11 full articles and 17 abstracts from 1981 to 2007) showed symptomatic reduction in 64%, decreased hospitalization in 67%, and accelerated gastric emptying in 60% of the studies (Sugumar et al, 2008). The risk of central nervous system side effects was much lower compared to metoclopramide, since domperidone does not cross the blood-brain barrier. However, domperidone was associated with corrected QT interval (QTc) prolongation (Dumitrascu and Weinbeck, 2000), limiting its use to small doses of <1 per week in Europe. In clinical practice, the recommended dose of domperidone is 10–20 mg three times a day and at bedtime. It should be avoided in patients with prolonged QTc (>470 milliseconds in males, >450 milliseconds in females) due to increased risk of cardiac arrhythmias (Dumitrascu and Weinbeck, 2000).

Table 2.

Summary of clinical trials documenting efficacy of domperidone in gastroparesis

Adapted with permission from Camilleri et al (2022).

Type of Study	Number, Etiology	Dose	Dura-tion	Symptom Improvement vs Baseline (OPEN) or vs Placebo (RCT)	Δ Gastric Emptying	Adverse Effects	Ref.
Open, po	3 DM	10 mg 4 times a day	1 wk	Yes, not quantified	Improved, not quantified	NA	Watts et al, 1985
Open, po	12 IG, 3 DM, 2 PS	20 mg 4 times a day	48 mo	68.3% (P < .05)	34.5% (P < .05)	↑ prolactin (100%), symptoms (17.6%)	Soykan et al, 1997
Retrospective, po	57 DM	Max. dose 80 mg/day	377 days	70% patients improved	NA	16%	Kozarek et al, 1990
Open,	6 DM	20 mg 4 times a day	6 mo	79.2% (P < .01)	26.9% (NS)	NA	Koch et al, 1989
Open	12 DM	20 mg 4 times a day	35–51 days	chronic oral administration 20 mg 3 times a day after 40 mg on day 1 reduced symptoms	↑ solid and liquid GE	NA	Horowitz et al, 1985
RCT, PG, PC, withdrawal	208 DM	20 mg 4 times a day	4 wk	53.8% lower overall score with domperidone (P = .025)	NA	2%–3% ↑ prolactin (= to placebo)	Silvers et al, 1998
RCT, PC, XO + open label 1yr	13 DM	NA	8 wk	↓ in symptom frequency and intensity (P < .03); symptomatic improvement averaging >1y	NA	NA	Braun et al, 1989
RCT, PC, XO	6 DM	10 mg i.v.	Single	NA	↑ homogen- ized solid GE	NA	Heer et al, 1983
RCT, PC, XO cisapride (C) or DOM	8 IG; 3 DM	0.8 mg/kg (C) or 0.9 mg/kg (DOM) 3 times a day	4 wk	No overall benefit over placebo; 2 of 3 DM improved	NA	Gas pains, skin rash	Franzese et al, 2002
RCT, PC, XO	11 upper GI distress: 3 DM + severe ↓GE	10 mg 4 times a day	4 wk each Rx	2 of 3 diabetics improved with DOM Rx; for 11 patients, no superiority vs. placebo	NA	Abd. gas pains, skin rash, itching, sweating, dizziness, constipation	Nagler et al, 1981
RCT, PG, DOM vs metoclo-pramide	93 DM	DOM 20 mg 4 times a day; M 10 mg 4 times a day	4 wk	41.19% improved vs baseline (NA); NS vs metoclopramide	NA	Somnolence 49% M, 29% DOM	Patterson et al, 1999
RCT, PG, PC among initial responders over 4 wk	208 DM single-blind same dose	20 mg domperidone 4 times a day	4 wk	Symptom severity increased in both groups; HRQOL (SF-36): improved physical component score, but no difference in 7/8 other subscales	NA	Not reported in study	Farup et al, 1998
Cohorts in NIH gastroparesis consortium (63% IG)	181 in DOM, 567 in non-DOM groups	Not standardized	Up to 96 wk	DOM pts: moderate but significantly more improvement in gastroparesis outcomes: GCSI, nausea, fullness, upper abdominal pain, GERD scores, and PAGI-QOL	NA	No significant CV or other DOM-related complications	Sarosiek et al, 2022

Abd, abdominal; C, cisapride; DM, diabetes mellitus; DOM, domperidone; GCSI, Gastroparesis Cardinal Symptom Index; GE, gastric emptying; GERD, gastro-esophageal reflux disease; GI, gastro-intestinal; IG, idiopathic gastroparesis; M, metoclopramide; max., maximum; NA, not applicable; NS, nonsignificant; PC, placebo-controlled; po, per-oral; PS, post-surgical; pts, patients; RCT, randomized controlled trial; XO, cross-over.

In a phase 2A trial, the D₂/D₃ antagonist, trazpiroben, had no effect on gastric emptying, although benefits in volume-to-fullness during a nutrient load test were noted, and there were nonsignificant numerical aggregate symptom scores (Kuo et al, 2021b). In a phase 2B trial, there was no clinically meaningful efficacy in treating gastroparesis (Tack et al, 2023).

4. Motilin agonists

Macrolides such as erythromycin, azithromycin, and clarithromycin are motilin receptor agonists with a prokinetic property. Among the macrolides (motilin agonists), the most widely used in gastroparesis is erythromycin, and less frequently, azithromycin. Table 3 summarizes clinical trials of erythromycin in gastroparesis clinical trials (Janssens et al, 1990; Erbas et al, 1993; Richards et al, 1993a; Arts et al, 2005; Larson et al, 2010). In a systematic review of 5 small-scaled, short-term studies, erythromycin accelerated gastric emptying and improved symptoms in 43% of patients with gastroparesis (Maganti et al, 2003). When given orally, erythromycin was associated with tolerance within days to weeks (Dhir and Richter, 2004; Thielemans et al, 2005). These medications are generally used for 1–4 weeks because of development of tachyphylaxis to motilides (Thielemans et al, 2005). Erythromycin may also prolong the QTc (possibly with greater propensity in females [Drici et al, 1998]) and reduce gastric accommodation (Liau et al, 2001). A systematic review and network meta-analysis of 33 studies involving data from 22.6 million subjects found no association between macrolide use and the risk of arrhythmia or cardiovascular mortality (Gorelik et al, 2018).

Table 3.

Summary of clinical trials documenting efficacy of erythromycin in gastroparesis

Adapted with permission from Camilleri et al (2022).

Medication/ trial design	Number, Etiology	Dose (by mouth)	Duration	Efficacy	Ref.
Erythromycin RCT, PC, XO	10 T1DM	200-mg i.v.; 250 mg by mouth 3 times a day	4 wk	Solid meal retention at 2h: 63 ± 9% with placebo; 4±1% with erythromycin; no effects on the symptoms	Janssens et al, 1990
Erythromycin open trials of i.v. and by mouth	10 IG and 4 DG; 4 patients dropped out	6 mg/kg i.v. 500 mg tid-ac and qhs	Single dose; 4 wk and open 8.4 mo	Solid meal retention at 2h: 85 ± 11% (SD) at baseline; 20 ± 29% on i.v. erythromycin (P < .001); 48 ± 21% after 4 wk of oral therapy (P < .01). Reduction in total symptom scores and a significant reduction in global assessment scores	Richards et al, 1993a
Erythromycin vs metoclopramide RCT, XO	13 DG	250 mg by mouth 3 times a day erythromycin; 10 mg by mouth 3 times a day metoclopramide	3 wk each period	Compared with baseline, improved GE parameters after both erythromycin and metoclopramide, with improved total GI symptom scores, more pronounced with erythromycin	Erbas et al, 1993
Erythromycin RCT, PC, XO	20 IG (functional dyspepsia + delayed GE)	200 mg i.v.	Single dose	Erythromycin accelerated (breath test) solid GE T½ = 146 (27) vs 72 (7) min, and liquid GE T½ = 87 (6) vs 63 (5) min; no overall symptom improvement except for bloating	Arts et al, 2005
Erythromyin vs azithromycin retrospective case-control analysis	120 patients (27 DM) underwent SGE with provocative testing	250 mg i.v. of each drug	Single dose	Both treatments accelerated gastric emptying with no difference between the 2 treatments: erythromyin GE T½ = 166 ± 68 min baseline to 11.9 ± 8.4 min; azithromycin GE T½ = 178 ± 77 min baseline to 10.4 ± 7.2 min	Larson et al, 2010

DG, diabetic gastroparesis; DM, diabetes mellitus; GE, gastric emptying; GI, gastro-intestinal; IG, idiopathic gastroparesis; i.v., intravenous; min, minutes; PC, placebo-controlled; qhs, at bedtime; RCT, randomized controlled trial; SGE, solid gastric emptying; T1DM, type 1 diabetes mellitus; tid-ac, 3 times a day, before meals; XO, cross-over.

Azithromycin and clarithromycin are other macrolides that accelerate gastric emptying (Bortolotti et al, 1999; Larson et al, 2010). There are no randomized, placebo-controlled trials to assess efficacy of symptoms in patients with gastroparesis. Azithromycin and clarithromycin have comparable efficacy and possibly better safety profile than erythromycin, based on observational studies (Larson et al, 2010; Potter and Snider, 2013), but the concern for tolerance remains. In hospitalized patients, intravenous erythromycin (infused over 45 minutes) is widely used at 1.5–3.0 mg/kg three times a day for treatment of acute gastroparesis. In addition to the concerns for tolerance and tachyphylaxis, long-term use of antibiotics should be avoided, given their association with complications such as antibiotic resistance and increased risk for infections such as Clostridioides difficile toxin induced colitis and antibiotic induced diarrhea. Their use should be balanced with the potential of tachyphylaxis, cardiac risk, infection risk, and antibiotic resistance.

5. 5-HT₄ agonists

Prucalopride is a benzofuran carboxamide that is highly selective with high affinity for 5-HT₄ receptors. It has weak affinity for human D₄ and sigma₁ and mouse 5-HT₃ receptors at concentrations exceeding the inhibitory constant (Ki) for 5-HT₄ receptors by 290-fold. Prucalopride accelerates gastric emptying of solids in healthy volunteers (Bouras et al, 1999; Kessing et al, 2014) and in patients with constipation (Bouras et al, 2001). It is approved for the treatment of chronic constipation, but not for gastroparesis. Table 4 summarizes the 2 available randomized, placebo-controlled, crossover trials of prucalopride in the treatment of gastroparesis (Carbone et al, 2019; Andrews et al, 2021). There appears to be greater benefit in patients with idiopathic gastroparesis than in gastroparesis secondary to underlying diseases such as diabetes or connective tissue.

Table 4.

Summary of randomized, controlled trials of prucalopride for gastroparesis

Reproduced with permission from Camilleri et al (2022).

Trial Design	Number, Etiology	Dose (by mouth)	Duration	Efficacy	Reference
PC, DB, XO, RCT	28 IG, 6 DG	2 mg/day	Two 4-wk treatments with 2 wk washout	Prucalopride significantly improved the total GCSI, subscales of fullness/satiety, nausea/ vomiting, and bloating/distention, overall PAC-QOL score and GE T1/2; also, all efficacies were shown only in the idiopathic group	Carbone et al, 2019
PC, DB, XO, RCT	13 DM, 2 connective tissue disease	4 mg/day	Two 4-wk treatments with 2 wk washout	GE faster on prucalopride; GCSI scores were lower than baseline but not different between treatment arms. Meal-related symptoms over time or cumulative score not significantly different between groups. GE was more rapid in the prucalopride treatment period,	Andrews et al, 2021

DB, double-blind; DM, diabetes mellitus; GCSI, Gastroparesis Cardinal Symptom Index; GE, gastric emptying; PC, placebo-controlled; RCT, randomized controlled trial; Ref, reference; XO, cross-over.

Although it had been originally approved for treatment of reflux esophagitis, the 5-HT₄ receptor agonist, cisapride, accelerated gastric emptying and improved symptoms in placebo-controlled trials conducted in patients with gastroparesis in short-term or medium-term trials (eg, 6- or 8-week duration) (Corinaldesi et al, 1987; Camilleri et al, 1989; Richards et al, 1993b), and in long-term, open-label studies (Abell et al, 1991). Despite the beneficial long-term effect in gastroparesis, there was no significant effect on glycemic control (Braden et al, 2002). As a potent inhibitor of the human ether-à-go-go-related gene (hERG) potassium channel and with reports of extremely rare cases of cardiac arrhythmias, cisapride was taken off the market by the FDA in the United States (US) due to cardiovascular concerns, although it may still be available in other countries. It is also available for compassionate use in selected cases in the US. Table 5 shows a summary of randomized, controlled trials of cisapride in treatment of gastroparesis (Corinaldesi et al, 1987; Horowitz et al, 1987; Jian et al, 1989; Richards et al, 1993b; Braden et al, 2002).

Table 5.

Summary of clinical trials of cisapride for gastroparesis

All trials were double-blinded, placebo-controlled, parallel-group or crossover, randomized, controlled trials.

Number, Etiology	Dose (by mouth)	Duration	Efficacy	Reference
12, 58% F, idiopathic gastroparesis, all abnormal GE	10 mg 3 times a day	2 wk, X-O	Significant acceleration of GE solids; total symptom score borderline significant (P = .09); relative changes in symptom scores with cisapride correlated with relative changes in GE rates (rs = 0.69; P < .025)	Corinaldesi et al, 1987
20, 60% F, diabetic gastroparesis, all abnormal GE	10 mg 4 times a day	4 wk	Significant acceleration of solid and liquid GE. Upper Gl symptoms were less after cisapride (P < .05), whereas there was no change on placebo (P > .2); no significant relationship between the changes in upper GI symptoms and changes in GE after 4 wk cisapride.	Horowitz et al, 1987
28, 64% F, idiopathic, 60% abnormal GE liquids > solids	10 mg 3 times a day	6 wk	Significant acceleration of solid and liquid GE T1/2 Decrease in global diary score was significantly higher than with placebo at wk 3 (P < .05), but not at wk 6.	Jian et al, 1989
39, 95% F, idiopathic or diabetic, all abnormal GE	20 mg 3 times a day	6-wk RCT Post 2-wk placebo run-in	16 patients in cisapride group and 12 on placebo completed the trial. Cisapride significantly increased solid GE relative to baseline (P = .005); placebo did not. Cisapride did not significantly improve symptoms of gastroparesis relative to baseline or to placebo.	Richards et al, 1993b
19, 74% F, all diabetic gastroparesis with abnormal GE based on 13C-GEBT (octanoic)	10 mg 3 times a day	52 wk	Cisapride significantly accelerated solid GE T1/2 (P = .03), dyspepsia scores (P = .002) compared to placebo, but no effect on HbA1c.	Braden et al, 2002

F, female; GE, gastric emptying; GI, gastrointestinal; RCT, randomized controlled trial.

6. Other 5-HT4 receptors in development for gastroparesis

Velusetrag, naronapride, and felcisetrag are investigational agents in the pipeline which are believed to be more selective for 5-HT₄ receptors in the gut and are promising for the treatment of gastroparesis. They are compared here with the FDA-approved, selective 5-HT₄ receptor agonist, prucalopride (Camilleri, 2012) (Table 6). Velusetrag accelerated gastric emptying based on ¹³C-octanoate breath test (Kuo et al, 2021a). In a randomized, placebo-controlled, parallel-group design, multicenter trial of 232 patients with gastroparesis (183 females; 113 idiopathic, 70 diabetic), velusetrag 5, 15, or 30 mg, or placebo were tested for 12 weeks. Symptom improvement from baseline was achieved only with the velusetrag 5-mg dose, although all doses accelerated gastric emptying, with normalization of gastric emptying (that is, <10% retention in stomach at 4 hours relative to placebo-treated subjects [43.5%, 65.0%, and 71.4% for velusetrag 5, 15, and 30 mg compared to 0% for placebo]) (Abell et al, 2023). Among “new generation” 5-HT₄ agonists, prucalopride, velusetrag, and naronapride are selective for 5-HT₄ receptors without hERG effects (Tack et al, 2012).

Table 6.

Comparison of 5-HT₄ agonists

Updated from Camilleri (2012).

	Prucalopride	Mosapride	Velusetrag	Naronapride	Felcisetrag
Chemistry	Benzofuran carboxamide	Benzamide	Quinolinone-carboxamide	Benzamide	Benzoimidazole
Selectivity and affinity for 5-HT4 receptor	Highly selective, high affinity; weak affinity for human D4 and σ1, and mouse 5-HT3 receptors at concentrations exceeding the Ki for 5-HT4 receptors by 290-fold	High selectivity and affinity for 5-HT4, But: major metabolite (M1) with 5-HT3-antagonistic activity	High affinity and selectivity for h5-HT4c over other biogenic amine receptors; >500-fold selectivity over other 5-HT receptors (including h5-HT2B, h5-HT3A)	Specific 5-HT4 full agonist activity in the GI tract, but a partial agonist activity in the heart	5-HT4 receptor agonist and substrate for P-glycoprotein (P-gp); 10- to 100-fold greater affinity for human 5-HT4 receptor splice variants (5-HT4a, 4b and 4c) compared to prucalopride
Hepatic metabolism	Limited, not CYP3A4	CYP3A4	CYP3A4	Hydrolytic esterase, not CYP3A4	CYP3A4
Pharmaco-dynamic efficacy in humans	Accelerated colonic transit in health and chronic constipation; Accelerated gastric emptying in diabetic gastroparesis	Accelerated esophageal motility, gastric emptying and small bowel transit in health	Accelerated colonic transit in health in dose-related fashion	Accelerated colonic transit in health	Accelerated gastric, small bowel and colonic transit in patients with gastroparesis; accelerated gastric emptying in critically ill patients
Clinical trial efficacy	Approved for CIC	Clinical trials in dyspepsia, GERD, IBS-C, capsule endoscopy	Phase IIB	Phase IB	Phase IIA studies in gastroparesis and critically ill, hospitalized patients
Arrhythmo-genicity	No arrhythmic activity in human atrial cells; inhibited hERG channel only at μM concentration (IC50∼4.9 10-6M); no clinically relevant cardiac AEs in clinical trials of >4000 humans	Low potency to inhibit hERG channels, no arrhythmic activity in clinical trials, no clinically relevant effects on QT-intervals	At 3μM, no effect on hERG channel current; safety ratio vs cisapride >1000-fold; no effect on QT in health or 400 patients with constipation	At 100 μM, no effect on hERG channel; affinity ratio between IKr and 5-HT4 receptors of >1000-fold	No reports to date
Cardiovas-cular safety including elderly	Healthy subjects “thorough” QTc study; safety in elderly cohort 80% on CV drugs	Healthy subjects, no effects of mosapride on heart rate, blood pressure variabilities, autonomic nervous activity parameters, QT intervals, or QT dispersions	Healthy subjects “thorough” QTc study; transient increase in heart rate not different from placebo	Healthy subjects “thorough” QTc study	No reports of EKG abnormalities in 2 phase IIA studies
Most common adverse events	Diarrhea, headache	Diarrhea, abdominal pain, headache	Diarrhea, nausea, headache	Diarrhea, headache	Diarrhea and nausea
Approval status	EMEA, Canada, Mexico, US	A variety of Asian and South American countries	NA	NA	NA
Approved dose	2 mg/day in adults; 1 mg/day in >65y	5 mg 3 times a day in adults	NA	NA	NA

5-HT, serotonin; CV, cardiovascular; EKG, electrocardiogram; EMEA, European Medicines Agency; GERD, gastro-esophageal reflux disease; NA, not applicable; US, United States.

Results of the pharmacodynamic effects of a new specific 5-HT₄ receptor agonist, felcisetrag, have been reported. Intravenously administered felcisetrag significantly accelerated gastric emptying and colonic transit versus placebo in patients with gastroparesis with previously confirmed delayed gastric emptying. Felcisetrag was well tolerated (Chedid et al, 2021). In a randomized, double-dummy, parallel-group trial in mechanically ventilated patients with enteral feeding intolerance, felcisetrag use was associated with lower gastric residual volume compared to treatment with metoclopramide while patients were fed enterally (Chapman et al, 2021). The trial demonstrated that felcisetrag resulted in a higher percentage of patients achieving normal gastric retention compared to 4 doses of 10-mg metoclopramide (Chapman et al, 2021). There is only limited cytochrome P3A4-mediated drug-drug interaction inhibition for felcisetrag (Chen et al, 2022). Felcisetrag is primarily cleared through renal excretion, and this reduces potential for clinically relevant drug-drug interaction, as cytochrome P3A metabolism plays a minor role in the overall clearance (Pusalkar et al, 2022).

Velusetrag and felcisetrag did not demonstrate significant effects on coronary artery tone in canine, porcine, and human models, human platelet aggregation, hERG potassium channel conductance, or other off-target actions (Beattie et al, 2013). Felcisetrag had high affinity (pK(i) = 9.4) for human recombinant 5-HT_4c receptors and selectivity (>2000-fold) over all other 5-HT and non-5-HT receptors, ion channels, enzymes, and transporters (n = 78 tested) (Beattie et al, 2011). Naronapride, 20-mg three times a day, accelerated gastric emptying of solids in healthy volunteers (Camilleri et al, 2007). It is currently undergoing testing for clinical efficacy in idiopathic and diabetic gastroparesis (NCT05621811).

7. Other prokinetic agents

Other medications only available in a few countries, clebopride, cinitapride, and mosapride, are not reviewed in detail, as evidence has not clearly demonstrated efficacy for gastroparesis. Clebopride (a D₂ antagonist) was effective in patients with dyspepsia with radiologically delayed gastric emptying, a criterion not currently accepted for gastroparesis (Bavestrello et al, 1985). This medication is available for prescription in several countries, but not in the US. In a systematic review and network meta-analysis by Ingrosso et al (2023), clebopride was the most effective agent for relief of global symptoms in trials that provided evidence of delayed gastric emptying in all participants (Fig. 5). However, it is important to remember that the performance was based on a single clinical trial. In addition, there is evidence that long-term use of clebopride was associated with Parkinsonism and tardive dyskinesia (Montagna et al, 1992; Sempere et al, 1994). Cinitapride (a 5-HT₁, 5-HT₄ agonist and 5-HT₂ antagonist) was superior to placebo in a parallel-design study of 19 patients with dysmotility-like dyspepsia and mild-to-moderate delayed gastric emptying (Portincasa et al, 2009). Mosapride (a 5-HT₄ agonist) enhanced gastric emptying in interferon-induced gastroparesis, but it had no significant effects on symptoms (Kawamura et al, 2012).

Fig. 5.

Forest plots for failure to achieve an improvement in global gastroparesis symptoms showing results for individual medications (left panel) and drug class (right panel); these studies included only randomized, controlled trials that confirmed delayed gastric emptying in all patients. The 95% confidence interval not crossing 1 indicates that the individual medication or drug class has a relative risk which favors the experimental approach over placebo. Reproduced with permission from Ingrosso et al. Gastroenterology 2023;164:642-654. (This is a free PMC article; thus, no permission is required to reuse the figure.)

8. Cholinesterase inhibitors

Cholinesterase inhibitors are used as off-label prokinetics for gastroparesis. Neostigmine is a short-acting medication that induces an abnormal increase in gastroduodenal motor activity and accelerates the gastric emptying of liquids in critically ill patients with delayed gastric emptying (Bortolotti et al, 1995; Lucey et al, 2003). Neostigmine is only available in parenteral formulation and should be used in the hospital setting where cardiac electrical activity is monitored because of the potential to induce vagotonia and bradycardia.

Pyridostigmine has a longer duration of action, and oral formulation is used off-label as liquid or tablet at a dose of 60 mg three times a day. There are, however, no clinical trials to support its use in gastroparesis. In an open-label case series in children with different gastrointestinal motility problems (chronic intestinal pseudo-obstruction, gastroparesis with delayed small bowel transit, chronic constipation with failure to thrive, and prolonged ileus after pelvic surgery with chronic opioid use), pyridostigmine was beneficial in relief of symptoms; the effective dosing ranged between 0.25 and 2.0 mg/kg per day (Manini et al, 2017). Pyridostigmine may also have ameliorated small intestinal bacterial overgrowth in 15 patients with HIV-related autonomic neuropathy, although there was no significant effect on gastric emptying of solids, possibly because only 3 of 15 had delayed emptying at baseline (Robinson-Papp et al, 2019).

9. Ghrelin receptor agonist

Ghrelin is a 28-amino acid orexigenic hormone found primarily in the stomach. Administration of a pharmacologic dose of ghrelin increased proximal gastric tone through central and peripheral effects (Peeters, 2003; Tack et al, 2006) and accelerated gastric emptying in some studies of patients with gastroparesis (reviewed by Camilleri et al, 2009).

Relamorelin, a pentapeptide ghrelin receptor agonist, exhibited powerful prokinetic effects estimated to be 15- to 130-fold more potent than natural ghrelin (Van der Ploeg et al, 2014). In patients with diabetic gastroparesis with documented delay in gastric emptying, relamorelin, 100 mg subcutaneous, accelerated gastric emptying of solids (Shin et al., 2013a, Shin et al., 2013b). In healthy controls, relamorelin increased the frequency of distal antral contractions without inhibiting gastric accommodation or inducing satiation (Nelson et al, 2016). Relamorelin had proven clinical efficacy and safety in phase 2A and 2B, randomized, controlled trials in patients with diabetic gastroparesis (Lembo et al, 2016; Camilleri et al., 2017a, Camilleri et al., 2017b, 2020). However, subsequent phase III trials did not document efficacy of relamorelin for gastroparesis symptoms (NCT03285308; NCT 03426345; NCT 03383146; NCT 03420781). Table 7 shows a summary of the clinical trials of relamorelin in gastroparesis. As reported in clinicaltrials.gov, “the relamorelin program is being terminated solely based on a business decision.”

Table 7.

Clinical trials of relamorelin in gastroparesis.

A portion of this table is reproduced with permission from Camilleri et al (2022).

Trial design	Number, etiology	Dose, SQ	Duration	Efficacy	Reference
DB, PC, PG multi-center, 2-dose RCT	204, 67% F, Diabetic gastroparesis	10 μg once daily or twice a day	4 wk; after 7 d placebo run-in	10 μg twice a day relamorelin significantly accelerated GE and reduced vomiting vs placebo. Among 119 patients with baseline vomiting,10 μg b.i.d. relamorelin significantly accelerated GE, reduced vomiting and nausea, abdominal pain, bloating, and early satiety	Lembo et al, 2016
DB, PC, PG multicenter, 3-dose RCT	393, 63% F, diabetic gastroparesis	10, 30 or 100 μg twice a day	12 wk	Reduced nausea, abdominal pain, postprandial fullness, and bloating (but not vomiting) and accelerated GE by 13C-spirulina GEBT with all 3 doses, compared to placebo. Worse glycemic control due to accelerated GE	Camilleri et al, 2017b; 2020
DB, PC, PG, multi-center RCT	336, 66% F, diabetic gastro-paresis, abnormal GEBT in all	10 μg twice a day	12 wk; after 2 wk baseline symptom score collection	No significant treatment difference on change in severity (relative to baseline) during 12 wk using the DGSSD, recorded in electronic diary; No differences in proportion achieving responder criterion for individual symptoms of gastroparesis during wk 6–12 or in (TEAEs	NCT03285308
Safety 40 wk active, plus 6 wk PC randomized withdrawal trial	57, 70% F, diabetic gastro-paresis	10 μg twice a day	6 wk; after 40 wk safety open trial	No difference in TEAEs, mortality or serious AEs; Participants previously enrolled in NCT03420781 or NCT03383146	NCT03285308
DB, PC, PG, multicenter RCT	311, 73% F, Diabetic gastro-paresis, abnormal GEBT in all	10 μg twice a day after 2 wk placebo run-in	12 wk	No significant treatment difference on change in severity (relative to baseline) during 12 weeks using the DGSSD, recorded in electronic diary; No differences in proportion achieving responder criterion for individual symptoms of gastroparesis during week 6–12 or in TEAEs	NCT 03426345
DB, PC, PG, multicenter RCT; randomized 2:1 for relamorelin: placebo	450, 72.4% F, diabetic gastro-paresis, abnormal GEBT in all	10 μg twice a day (after 2 wk placebo run-in)	52 wk	No significant treatment difference on change in severity (relative to baseline) or weekly average score during 52 weeks using the DGSSD, recorded in electronic diary; Participants previously enrolled in NCT03285308; NCT03426345	NCT 03383146
Safety; 40 wk active plus 6-wk PC, randomized withdrawal trial	91 (sex % not specified) Diabetic gastro-paresis, abnormal GEBT in all	10 μg twice a day	6 wk after 40 wk PC RCT	No significant treatment difference on change from baseline to week 46 for DGSSD or individual symptoms Participants previously enrolled in NCT03285308; NCT03426345	NCT 03420781

DGSSD, Diabetic Gastroparesis Symptom Severity Diary; F, female; GE, gastric emptying; RCT, randomized controlled trial.

10. Network meta-analyses of prokinetic agents

Forest plots from the network analysis for failure to achieve an improvement in global gastroparesis symptoms are presented for individual medications or for drug class in Figure 5. This documents greatest benefit over placebo for the oral dopamine antagonist class and the tachykinin-1-antagonist class, whereas other classes were not statistically superior to placebo.

C. Antiemetics

Pharmacologic agents aimed at improving nausea, or antiemetics including mechanistically diverse agents are used for symptom relief in gastroparesis.

1. 5-HT₃ antagonists

Dosed at 4–8 mg every 8 hours as needed, ondansetron is a 5-HT₃ antagonist that targets stomach distention, alleviating nausea without affecting gastric compliance, volume, or accommodation (Janssen et al, 2011). Ondansetron is available in several formulations: oral tablet, oral dissolution, and intravenous. Ondansetron can cause QTc prolongation and, in rare cases, can even lead to Torsades de pointes, a life-threatening cardiac arrhythmia. Baseline and continual cardiac monitoring with electrocardiogram is recommended. Granisetron is also a 5-HT₃ antagonist available orally and intravenously. The sustained release transdermal patch of granisetron was shown in an open-label study to significantly improve nausea and vomiting in 51 patients with gastroparesis (of which 11 were diabetic) in a prescription registry study (Midani and Parkman, 2016). Side effects reported included redness at the site of the patch in 7 patients, pruritus in 5, and constipation in 5 patients. The latter is a known adverse effect of this class of medications.

2. Agents targeting multiple receptors

Prochlorperazine exerts its primary action though blocking D₂ receptors, although it can also block histaminergic, cholinergic, and noradrenergic receptors. Promethazine primarily acts as an antagonist for H₁ histamine receptors, although it also has antagonistic properties against dopamine, adrenergic, N-methyl-D-aspartate receptor (also known as the NMDA receptor; it is a glutamate receptor and predominantly Ca⁺⁺ ion channel found in neurons), and muscarinic cholinergic receptors. Scopolamine competes for binding at muscarinic (M₁) receptors, inhibiting cholinergic nerve stimulation. These antiemetics are available in orally disintegrating tablets, dermal, or rectal formulations for patients with gastroparesis. Cholinergic side effects like sedation, dry mouth, and constipation are frequently encountered. Additionally, promethazine may be habit forming and is reserved as a “rescue” agent.

Mirtazapine acts on several receptors: presynaptic alpha-2 adrenergic receptors, several subtypes of the 5-HT receptors, and the H₁ histamine receptor. Of particular interest in relation to gastroparesis, mirtazapine is an agonist of central and peripheral 5-HT (serotonin)_1A receptors which influence gastric receptive fundic relaxation. Among 30 adult patients with gastroparesis, 24 (80%) completed 4 weeks of therapy. There were statistically significant improvements in nausea, vomiting, retching, perceived loss of appetite, and the clinical patient grading assessment scale score at 2 and 4 weeks (all P values <.05) compared with pretreatment. Of the total patients, 14 (46.7%) experienced adverse effects from mirtazapine, and due to this, 6 patients stopped therapy (Malamood et al, 2017).

3. Neurokinin-1 antagonists

The neurokinin-1 (NK₁) receptor antagonist, aprepitant (approved for chemotherapy-induced emesis), affects the vomiting center in the brainstem and enhances gastric accommodation without slowing of gastric emptying (Jacob et al, 2017). Similarly, tradipitant, 85 mg twice a day, did not significantly affect gastric motor functions (gastric volumes or gastric emptying) in healthy participants (Khanna et al, 2022). In a randomized, double-blind, placebo-controlled trial of 126 patients with chronic nausea and vomiting of presumed gastric origin, aprepitant, 125 mg daily, significantly reduced severity of nausea, vomiting, and overall symptoms (Pasricha et al, 2018). Tradipitant is a similar investigational agent which was demonstrated to decrease nausea score, increase nausea-free days, and improve the GCSI score in patients with gastroparesis compared to placebo (Carlin et al, 2021). Table 8 summarizes the randomized, controlled trials of NK₁ antagonists for gastroparesis (Pasricha et al, 2018; Carlin et al, 2021, 2024).

Table 8.

Summary of randomized, controlled trials of NK1 antagonists for gastroparesis

A portion of this table is reproduced with permission from Camilleri et al (2022).

Medication/trial design	Number, Etiology	Dose (by mouth)	Duration	Efficacy	Reference
Aprepitant PC, DB, multicenter RCT	126, 80% F, IG or DG symptoms, 57% delayed GE	125 mg per day	4 wk	Aprepitant did not reduce symptoms of nausea (primary endpoint), but there was reduction in symptom severity for nausea (P = .005), vomiting (P = .001), and overall symptoms (P = .001).	Pasrichaet al, 2018
Tradipitant PC, DB, multicenter RCT	152 DM, 90% F, all with documented delayed GE	85 mg twice a day	4 wk	At wk 4, significant decrease in nausea score (P = .0099) and increase in nausea-free days at wk 4 (28.8% increase on tradipitant vs 15.0% on placebo; P = .016); greater effect in 101 patients with nausea and vomiting at baseline; improvement in GCS ≥1 point in 46.6% on tradipitant vs 23.5% on placebo (P = .0053).	Carlin et al, 2021
Tradipitant PC, DB, multicenter RCT	201 patients: 103 IGs and 98 DG, 80.1% F, all with documented abnormal GE	85 mg twice a day	12 wk	In ITT population, tradipitant did not meet the prespecified primary endpoint (difference in nausea severity change drug vs placebo) at week 12; post hoc sensitivity analyses, tradipitant treatment demonstrated statistically significant improvements in nausea at wk 12 based on baseline severity inflation, PK, and removing confounding by rescue medication	Carlin et al, 2024

DB, double-blind; DG, diabetic gastroparesis; DM, diabetes mellitus; F, female, GE, gastric emptying; IG, idiopathic gastroparesis; ITT, intention to treat; PC, placebo-controlled; PK, pharmacokinetics; RCT, randomized controlled trial.

4. Cannabinoid agents

Among 506 patients evaluated in a National Institutes of Health Gastroparesis Consortium study, 12% of the patients with gastroparesis used medical or recreational marijuana for symptomatic relief (Parkman et al, 2020). The primary ingredient in marijuana is tetrahydrocannabinol (THC), a nonselective cannabinoid receptor agonist. THC has been found to delay gastric emptying of solids (McCallum et al, 1999) and, with chronic use, may lead to cannabinoid hyperemesis syndrome (Rubio-Tapia et al, 2024). Dronabinol is an FDA-approved synthetic THC for human immunodeficiency virus-acquired immunodeficiency syndrome-induced anorexia and chemotherapy-induced nausea and vomiting. Consensus guidance for its use has been published (Alderman et al, 2022). For gastroparesis, dronabinol serves as a second-line therapy for patients with refractory nausea, and it is a potent appetite simulant helpful for those with weight loss. It is limited to short-term use due to adverse effects such as cannabis-like highs and binge eating.

Cannabidiol (CBD), a low-THC extract from Cannabis sativa approved for seizure disorders, was efficacious in a randomized, double-blinded, placebo-controlled study of CBD twice daily (Epidiolex escalated to 20 mg/kg per day in 44 patients (32 idiopathic, 6 diabetes mellitus type 1, and 6 diabetes mellitus type 2) (Zheng et al, 2023). Compared with placebo, CBD reduced the total GCSI score (P = .008), the inability to finish a normal-sized meal (P = .029), the number of vomiting episodes/24 hours (P = .006), and the overall symptom severity (P = .034). Patients treated with CBD had a higher volume to comfortable fullness and maximum tolerance, and slower gastric emptying of solids. The most common adverse events reported were diarrhea (14 patients), fatigue (8 patients), headache (8 patients), and nausea (7 patients).

CBD’s effectiveness seems tied to its diverse actions: blocking cannabinoid receptor 1 and cannabinoid receptor 2, inhibiting endocannabinoid breakdown via the fatty acid amide hydrolase enzyme to elevate endocannabinoids like anandamide which activate cannabinoid receptors, as well as activating 5-HT_1A receptors and transient receptor potential vanilloid 1 channels. These actions contribute to its anxiolytic and analgesic properties, even though it may slow gastric emptying due to heightened endocannabinoid activation of cannabinoid receptors (de Almeida and Devi, 2020).

5. Neuromodulators for pain relief

Patients frequently experience abdominal pain with gastroparesis; however, those primarily presenting with abdominal pain should be evaluated for alternative diagnoses. Neuromodulators, including tricyclic antidepressants such as amitriptyline and nortriptyline, are typically considered first-line treatment for functional abdominal pain. Yet, in a randomized trial involving 130 patients diagnosed with idiopathic gastroparesis, nortriptyline did not demonstrate superiority over placebo in alleviating symptoms, as measured by the GCSI score (Parkman et al, 2013).

D. Targeting the fundus

Simultaneously measured gastric accommodation and emptying of a solid egg meal identified a direct relationship between the degree of gastric accommodation and the gastric emptying lag duration and T_1/2. This relationship suggests that, in some patients with gastroparesis, excessive gastric accommodation with delayed movement of solid food from the fundus to the antrum of the stomach results in impaired gastric emptying (Wang et al, 2021). Thus, stimulation of the proximal stomach with reduced gastric accommodation may actually improve gastric emptying in patients with gastroparesis.

Therefore, it is logical that erythromycin, with its motilin receptor agonism and stimulation of cholinergic mechanisms, is associated with marked acceleration or dumping of food from the stomach. Erythromycin exerts a dual effect, improving both fundic contraction and antral motor function, increasing longitudinal axial forces in the antrum in healthy participants and patients with dysmotility and acceleration of gastric emptying (Surrenti et al, 1996; Coulie et al, 1998; Liau et al, 2001).

Another medication, buspirone, a 5-HT_1A agonist with anxiolytic properties, improved aggregate symptoms and nausea in response to a nutrient challenge meal without significant alteration in gastric accommodation in healthy controls (Chial et al, 2003). However, patients with moderate-to-severe early satiety or postprandial fullness and other symptoms of gastroparesis did not benefit from treatment with buspirone (Parkman et al, 2023).

As discussed above, in an open-label, 4-week treatment study of 30 patients with gastroparesis, 15 mg mirtazapine resulted in statistically significant improvements in nausea, vomiting, retching, and perceived loss of appetite at 2 and 4 weeks (all P values <.05) compared with pretreatment (Malamood et al, 2017). However, the mechanism appears unrelated to alteration in gastric accommodation (Carbone et al, 2017).

Mechanistic studies showed that acotiamide (inhibitor of acetylcholinesterase [AChE] and antagonist at the M1 and M2 receptor sites, decreasing degradation and increasing the release of acetylcholine from cholinergic enteric neurons) enhanced gastric accommodation and gastric emptying of a liquid meal (Kusunoki et al, 2012) and improved symptoms in patients with functional dyspepsia (Matsueda et al, 2012). However, another study suggested acotiamide reduced antral motility without affecting fundic intragastric pressure or gastric emptying (Masuy et al, 2019). Acotiamide is not available in the US.

Additional research into the effectiveness of medications that stimulate the fundus and antrum and alleviate gastric symptoms would be of significant interest.

Table 9 provides a simple summary of the efficacy and main adverse effects or safety signals of medications for the treatment of gastroparesis.

Table 9.

Efficacy and main adverse effects of medications (approved drugs in bold text) with formal trials in gastroparesis

Drug Name	Disease	Effect on Gastric Motor Function	GP Symptoms	Main Adverse Effects/ Comments
5-HT4 Receptor Agonist
Prucalopride	IG and DG	↑ GE	Improved, mostly IG	Diarrhea, abdominal pain
Velusetrag	IG and DG	↑ GE	Improved	Diarrhea, nausea, and abdominal or upper abdominal pain
Naronapride	IG and DG	↑ GE	NCT05621811
Felcisetrag	IG and DG	↑ GE	Not studied	Administered i.v.; addo. distension, pain, nausea
D2/3Receptor Antagonist
Trazpiroben	IG and DG	↑ volume to fullness, No Δ in GE	Not improved in phase 2A and 2B trials	Constipation, headache, nausea
Ghrelin Receptor Agonist
Relamorelin	DG	↑ GE, ↑ antral contractions	Improved in phase 2, not in phase 3 trials	Hyperglycemia
Motilin Receptor Agonist
Erythromycin	IG and DG	↑ GE, ↑ fundic and antral contractions, ↓ pyloric contractions	Improved	Abdominal pain, postprandial fullness, diarrhea
Azithromycin	Gastro-paresis	↑ GE	Not studied
Clarithromycin	FD	↑ GE	Not studied
NK1Receptor Agonist
Aprepitant	IG and DG	↑ GA, No Δ in GE	Improved	Gastrointestinal (no additional details)
Tradipitant	IG and DG	No Δ in GE	Improved	Diarrhea, UTI, sinusitis, dizziness, headache

DG, diabetic gastroparesis; FD, functional dyspepsia; GA, gastric accommodation; GE, gastric emptying; IG, idiopathic gastroparesis; UTI, urinary tract infection.

E. Targeting the pylorus

Rationale for targeting the pylorus: in a subset of patients with gastroparesis, pyloric dysfunction, characterized by abnormally prolonged and intense tonic contractions of the pylorus, was noted (Mearin et al, 1986). This observation of “pylorospasm” prompted the development of several procedural interventions aimed at the pylorus, including botulinum toxin injection, pyloric dilation and/or stenting, as well as surgical or endoscopic pyloromyotomy.

Iatrogenic pyloric dysfunction: there is growing awareness that individuals using opioid medications for extended periods may develop gastroparesis (Hasler et al, 2019). The mechanisms underlying the effects of opioids are reviewed elsewhere (Camilleri and Sanders, 2020). Briefly, the gastric effects of opioids include pyloric dysfunction and inhibition of antral motor function, resulting in delayed gastric emptying. Given the effects of opioids on the pylorus, a novel therapeutic approach may include pharmacologic manipulation of the pylorus or inhibition of opioid effects at the pylorus.

Botulinum toxin injection: in patients with refractory gastroparesis despite pharmacotherapies, multiple open-label studies showed intrapyloric injection of botulinum toxin had short-term (<6 months) efficacy in accelerating gastric emptying and improving symptoms (Thomas et al, 2018). However, 2 randomized, placebo-controlled trials did not confirm efficacy in achieving symptom improvement, although 1 trial did confirm acceleration of gastric emptying with botulinum toxin compared to saline (Arts et al, 2007; Friedenberg et al, 2008). Given the discordant results, it is important to note that analysis of a large, open-label, retrospective study of 179 gastroparesis patients demonstrated dose-dependent, short-term (1–4 months) symptom improvement in 51.4% patients after botulinum toxin injection (Coleski et al, 2009). The factors associated with improved response included female sex, age <50 years, and etiology not related to diabetes or surgery (Coleski et al, 2009). There is, however, concern that repeated injection with high-dose botulinum toxin may induce pyloric fibrosis over time, thus reducing the feasibility of pyloromyotomy or gastric peroral endoscopic myotomy.

Oral pharmacotherapy: two pharmacologic mechanisms have been pursued to improve pyloric dysfunction in patients with gastroparesis. One approach is the use of sildenafil (Dishy et al, 2004), a phosphodiesterase-5 inhibitor, which increases intracellular cGMP and mimics the effect of NO. In diabetic mice, insulin and sildenafil reversed the reduced expression of neuronal NO synthase in the pylorus (Watkins et al, 2000). However, sildenafil had no significant effect on gastric emptying in uremic gastroparesis (Dishy et al, 2004). A second approach that has been explored is the use of opioid antagonists. In an older study, naloxone did not stimulate gastric emptying in healthy subjects, nor did it correct gastric stasis in patients with gastric hypomotility (functional dyspepsia/idiopathic gastroparesis) (Narducci, 1986). Two studies tested opioid antagonists on opioid-induced delay in gastric emptying in opioid-naïve healthy participants. As peripherally active μ-opioid receptor antagonists have emerged as a treatment option for chronic opioid-induced constipation, 2 proof of concept studies were conducted with short-term administration of methylnaltrexone (subcutaneous 0.30 mg/kg) or naloxegol (25 mg) daily in healthy, opioid-naïve volunteers (Wong et al, 2010; Halawi et al, 2018). However, in both studies there were no differences in gastric emptying between the peripherally active μ-opioid receptor antagonists and placebo, nor between 30-mg 4 times a day codeine-treated healthy participants concomitantly receiving naloxegol or methylnaltrexone and placebo.

F. Anti-inflammatory approaches

A potential mechanism to enhance neuromuscular function in the stomach is through modulation of anti-inflammatory pathways that induce vagal stimulation, and the 5-HT₄ agonist, prucalopride, which modified T helper 2 cell responses that was associated with shortened postoperative ileus (Matteoli et al, 2013; Bosmans et al, 2017; Stakenborg et al, 2019). Gut-resident macrophages can impact enteric nervous system function (De Schepper et al, 2018). This anti-inflammatory mechanism could be pertinent since certain animal models exhibit damage in the enteric nervous system and pacemaker cells caused by inflammation and oxidative stress.

1. Targeting M₂ macrophages and oxidative stress

There is a different lineage of resident macrophages termed M2 or mannose receptor C-type 1 gene (CD206)-positive in gastric tissues which are distinct from the M1 macrophages that are involved in immune responses. The M2 macrophages are involved in phagocytosis of bacteria and tissue repair. Enteric mechanisms play a role in the development of gastroparesis, at least in part, by macrophage-based immune dysregulation, which was reported to be associated with delayed gastric emptying in diabetic mice (Cipriani et al, 2018). Abnormalities in patients with idiopathic or diabetic gastroparesis include, in some studies but not all, reduced pacemaker cells (ICCs), PDGFRα fibroblast-like cells, and numbers of nitrergic neurons and CD206-positive macrophages (Grover et al, 2011, 2012; Bernard et al, 2014; Grover et al, 2017; Herring et al, 2018). Depletion of anti-inflammatory resident M₂ macrophages expressing heme oxygenase-1 is believed to cause oxidative stress, leading to loss of pacemaker cells in animal models of gastroparesis (Choi et al, 2008; Cipriani et al, 2018).

3. Potential pharmacologic promotion of neuronal cell differentiation

In the enteric nervous system, there is a dynamic balance between cell loss by apoptosis and macrophage phagocytosis to remove dead neurons. Neurogenesis in precursor cells that behave like stem cells are prominent in the submucosal zone and in the muscular layers (Kulkarni et al, 2017) and may transform to neurons. A selective estrogen receptor β agonist, LY3201, stimulated glial-to-neuron cell differentiation in vitro, promoted neurogenesis, and increased the recovery of neurons in the damaged myenteric plexus in two in vivo murine models of enteric neuronal damage (D'Errico et al, 2018).

In rats, exogenous brain-derived neurotrophic factor increased myoelectric activity and peristalsis in the gastrointestinal tract and colon (Chai, 2003; Grider et al, 2006), and exogenous recombinant human brain-derived neurotrophic factor and neurotrophin-3 accelerated gastrointestinal and colonic transit in healthy human volunteers and in patients with constipation (Coulie et al, 2000). These studies suggest that neurogenesis has the potential to improve enteric nervous system function in patients with gastroparesis.

IV. Other treatments not addressed

Gastroparesis is a motility disorder with multiple underlying pathophysiological mechanisms. Multidisciplinary management is indicated, particularly dietary modification (essentially use of low-fat diet with cooking and homogenizing to particle size ≤2 mm of any foods such as fruits and vegetables with high cellulose content) and enteral or parenteral nutrition support, which are outside the scope of this article. Intubations are sometimes required for decompression and feeding in patients with gastroparesis. A summary of the different approaches as well as their usefulness and disadvantages are summarized in Table 10, based on an authoritative review (Maple et al, 2005).

Table 10.

Intubations for decompression and feeding in patients with gastroparesis (adapted from Maple et al [2005] and experience of the authors)

Type of Access	Usefulness/ Advantages	Disadvantages
Nasogastric tube	Gastric decompression in managing acute exacerbations of gastroparesis	Not intended for long-term use Large tube size often causes discomfort Poor choice for feeding due to delayed GE Significant GER can occur
Nasoduodenal/ nasojejunal tube	Used when trial of enteral feedings are indicated to determine if jejunal feedings are tolerated at an infusion rate that will suffice for patient’s nutritional needs	Not intended for long-term use Vomiting may expel the tube into the stomach
Gastrostomy tube	May be used for venting of secretions to decrease vomiting and fullness	Poor choice for feeding with delayed GE May prevent other interventions, eg, G-POEM or (rarely performed) sleeve gastrectomy for gastroparesis
PEG-J or direct PEJ	Allows the patient to vent gastric secretions to decrease persistent emesis Provides jejunal feedings Distal feeding ports reduce duodenogastric reflux	Migration of the J-tube extension into stomach following vomiting, and requires replacement Potential pyloric obstruction from J-tube
Jejunostomy (surgical, endoscopic, radiographic)	Stable access for reliable jejunal nutrient delivery Avoids gastric penetration, which would potentially interfere with G-POEM	Cannot vent/decompress stomach
Dual gastro-stomy and jejunostomy	Two sites—1 for venting and 1 for enteral nutrition	Increased risk of leakage, infection at entry sites; Cosmetic issues

GE, gastric emptying; GER, gastroesophageal reflux; G-POEM, gastric peroral endoscopic myotomy; PEG-J, percutaneous endoscopic gastrostomy jejunostomy; PEJ, percutaneous endoscopic jejunostomy.

Because electrical approaches and endoscopic or surgical approaches do not involve pharmacotherapy, they have not been reviewed in detail. The reader is referred to a recent guideline that documented clinical trial-based efficacy as well as systematic reviews and meta-analyses of gastric electrical stimulation, electro-acupuncture, acupuncture, gastric peroral endoscopic myotomy, and sleeve gastrectomy in patients with gastroparesis or chronic nausea and vomiting (Camilleri et al, 2022).

V. Challenges associated with drug development for gastroparesis

Among the challenges associated with drug development in gastroparesis, three stand out as dominant factors. First, the gastric emptying test protocol for eligibility may be suboptimal, such as use of a liquid or easily digestible egg protein substitute with relatively low calorie (250 kcal) and low fat (2%) content, documentation of emptying for less than 3 hours, and using cut-off for diagnosis of gastroparesis based on ≥10% retention at 4 hours. A second major concern pertains to patient confounders such as the use of opioids, anticholinergics, tetra-hydrocannabinol, and glucagon-like peptide-1 receptor agonists, as well as fasting hyperglycemia >250 mg/dL, that can retard gastric emptying. Third, although the FDA draft guidance endorses a patient reported outcome instrument, specifically the American Neurogastroenterology and Motility Society Gastroparesis Cardinal Symptom Index-Daily Diary (ANMS GCSI-DD), recent phase 3 trials have used variations of this patient response outcome such as Diabetic Gastroparesis Symptom Severity Diary in relamorelin trials and Gastroparesis Core Symptom Daily Diary in tradipitant trials. Fourth, current FDA guidance on gastroparesis trials suggests a treatment duration of at least 12 weeks before measuring the change in symptoms from baseline. This presents significant challenges associated with use of rescue medications, given the long duration of the trial in patients who frequently have to resort to symptomatic relief, visits to emergency departments, or even hospitalization.

Thus, a recent review (Yang and Camilleri, 2024) detailed the goals for successful development of treatment in gastroparesis, the evidence-based criteria for treatment success based on current scientific understanding of gastroparesis as well as patient response outcomes, and proposed evidence-based principles for the successful development of treatments for gastroparesis. The recommendations included considerations regarding eligibility based on robust documentation of delayed gastric emptying of solids appraised over 4 hours, design of randomized, placebo-controlled trials including a baseline period to be considered as a covariate in assessment of efficacy, and endorsement of the ANMS GCSI-DD as a validated patient response outcome. In contrast to recommendations from regulatory agencies such as the FDA, it is justifiable based on clinical practice to shorten clinical trials to 4 weeks of treatment, given the necessity to achieve efficacy with any treatment proposed for gastroparesis.

Given that the pathophysiology underpinning the development of symptoms may vary across individuals, an additional challenge is that there is limited access to validated tests to measure gastric accommodation (eg, single-photon emission computed tomography imaging) and hypersensitivity (intragastric balloon distention or nutrient drink test) that may be the pathophysiological mechanisms leading to the symptoms. Indeed, in almost 1300 patients with upper gastrointestinal symptoms evaluated with a 320 kcal, 30% fat gastric emptying test and single-photon emission computed tomography-based gastric accommodation, it was shown that about a quarter of patients had only delayed gastric emptying, a quarter had only impaired gastric accommodation, a quarter had both abnormalities, and a quarter had neither (Park et al, 2017a). This illustrates the point that selection of pharmacologic treatment would be enhanced if the precise mechanism underpinning the development of symptoms could be targeted such as with the use of prokinetic for those with delayed gastric emptying, or medications that enhance gastric accommodation, or central neuromodulators for those with hypersensitivity.

VI. Conclusion

Gastroparesis is diagnosed with a well validated gastric emptying study such as scintigraphy or stable isotope breath test with robust cut-off values to differentiate normal versus delayed emptying. Management of gastroparesis takes on a multidisciplinary approach involving nutritional support, prokinetic agents, and antiemetics. There is still a considerable unmet need for efficacious prokinetic and antiemetic medications specifically developed for the treatment of gastroparesis.

Conflicts of interest

The authors have no conflicts of interest.

Associate Editor: Ali Eid