Gastroparesis

Abstract

Gastroparesis is characterized by symptoms suggestive of and objective evidence of delayed gastric emptying in the absence of mechanical obstruction. This review addresses the normal emptying of solids and liquids from the stomach and details the myogenic and neuromuscular control mechanisms including the specialized function of the pyloric sphincter that result in normal emptying, based predominantly on animal research. A clear understanding of fundamental mechanisms is necessary to comprehend derangements leading to gastroparesis, and additional research on human gastric muscles is needed. The section on pathophysiology of gastroparesis considers neuromuscular diseases that affect non-sphincteric gastric muscle, disorders of the extrinsic neural control and pyloric dysfunction that lead to gastroparesis. The potential cellular basis for gastroparesis is attributed to the effects of oxidative stress and inflammation, with increased pro-inflammatory and decreased resident macrophages, as observed in full-thickness biopsies from patients with gastroparesis. Predominant diagnostic tests involving measurements of gastric emptying, the use of a functional luminal imaging probe and high-resolution antral duodenal manometry in characterizing the abnormal motor functions at the gastroduodenal junction are discussed. Management is based on supporting nutrition, dietary interventions including the physical reduction in particle size of solid foods, pharmacological agents including prokinetics and anti-emetics, and interventions such as gastric electrical stimulation and pyloromyotomy. These are briefly discussed, and comment is added on the potential for individualized treatments in the future, based on optimal gastric emptying measurement and objective documentation of the underlying pathophysiology causing the gastroparesis.

INTRODUCTION: NORMAL EMPTYING OF LIQUIDS AND SOLIDS FROM STOMACH

Ingestion of food results in gastric accommodation followed by development of antral contractions. After food is triturated to a small particle size, pyloric relaxation and antropyloroduodenal coordination lead to the emptying of food from the stomach. Foods of different physical nature and consistency follow distinct emptying patterns¹: exponential for liquids with low calorie content under the pressure gradient provided by fundic tone with relaxation of the pylorus; and linear for nutrient liquids and homogenized solids. Posture influences the emptying of non-nutrient liquids, being faster in the upright position.² Raised intragastric pressure increases emptying of liquids, but not solids.³

Solids are initially retained in the proximal stomach and subsequently moved to the antrum to undergo trituration. This initial period without emptying is termed the lag phase. Antral phasic pressure activity is essential and significantly correlates with the rate of emptying of solid food from the stomach, as shown in healthy stomach⁴ and by prolongation of gastric emptying in the presence of antral hypomotility.⁵ Antral hypomotility is typically characterized by an average of <1 distal antral contraction per minute in the first postprandial hour.⁶

The timing and mechanistic steps responsible for pyloric regulation of gastric emptying are still incompletely understood.⁷ From early cineradiographic studies in lab animals,⁸ it was shown that peristaltic contractions progress from corpus to pylorus, resulting in a brief period of emptying of liquid and small particles. As the peristaltic wave reaches the terminal antrum, pyloric constriction occurs, restricting emptying during the period of highest pressure in the terminal antrum. Contents are forcefully retropulsed back into the body of the stomach setting up the shearing forces that cause trituration of solids. Thus, the antropyloroduodenal junction provides a sieving function whereby food particles must be reduced in size to ≤2mm before emptying occurs.⁹ Particles retrieved from duodenums of healthy dogs were <2mm, whereas dogs with Billroth I gastrectomy with loss of the pylorus had larger solid food particles emptied,¹⁰ and this may lead to malabsorption due to inefficient digestion following this form of gastric surgery.¹¹

Failure of antral contractions or of pyloric relaxation may impede gastric emptying and constitute the predominant pathophysiological disturbances in gastroparesis, which are identified by classical clinical symptoms that include nausea, vomiting, early satiety, postprandial fullness, bloating, upper abdominal pain and documentation of delayed gastric emptying in the absence of mechanical obstruction.¹² A recent study has drawn attention to the observation that, a year after initial classification, patients with functional dyspepsia and gastroparesis assessed in tertiary referral centers are not distinguishable based on clinical and pathologic features or based on assessment of gastric emptying of a low fat, relatively low calorie, easily digestible meal.¹³ This emphasizes the importance of accurate diagnosis based on optimal measurements of gastric emptying by scintigraphy or breath test (Supplemental Table 1) and the use of robust normative data such as gastric retention of >75% at 2 hours and >25% at 4 hours, based on normative data reported in 319 healthy volunteers.¹⁴ Such strict cut-off criteria are extremely important given the mean inter- and intra-individual coefficients of variation (COV) for gastric emptying T_1/2 of 24.5% and 23.8%, respectively, in healthy individuals, and the intra-individual COV of 20% in a study of 60 patients (21 diabetic) presenting with upper gastrointestinal symptoms when evaluated twice on average 15 days apart.¹⁵ In a more recent study¹⁶ of 9 patients with proven gastroparesis (2 type 1 diabetes, 7 idiopathic) measured twice at 3-day intervals, the range of mean GE T_1/2 was 20-240 minutes and the COV_intra was 11.1%.

MYOGENIC MECHANISMS

Smooth muscle cells (SMCs) provide the forces required for trituration of food and gastric emptying. SMCs are spindle-shaped, 40-100μm long, 2-8μm in diameter and tightly packed with little connective tissue between cells. SMCs are organized into three layers (circular, longitudinal and oblique) in the stomach. As with other muscle cells, gastric contractions depend upon coupling of electrical events to contractions through an increase in cytoplasmic calcium ions (Ca²⁺). SMCs are electrically coupled by gap junctions which facilitates synchronicity of contractions by conduction of electrical impulses between the cells.

From a functional perspective, the stomach is composed of 3 major regions: the proximal stomach (fundus) generates tone that can be actively regulated to accommodate ingestion of food; the distal stomach (corpus and antrum) that processes food before emptying; and the pyloric sphincter that regulates the rate of gastric emptying.¹⁷ SMCs in these regions have both similar and different intrinsic properties that facilitate specialized functions. Common to all gastric SMCs is expression of voltage-dependent Ca²⁺ channels (dihydropyridine sensitive, L-type channels) that comprise the main Ca²⁺ entry mechanism regulating excitation-contraction coupling. However, proximal, distal and pyloric SMCs have different complements of potassium (K⁺) channels that regulate resting membrane potentials and the state of basal excitability of the SMCs. Fundus cells are intrinsically more depolarized than antral cells. Fundus SMCs sit at potentials (approximately −50 mV) that facilitate continuous leak of Ca²⁺ into cells through voltage-dependent conductances. This facilitates the development and maintenance of tone. Antral and pyloric SMCs, in contrast, have more negative resting membrane potentials at which Ca²⁺ entry is minimal, facilitating relaxation between excitable events. The resting potentials of gastric SMCs are set at negative levels by highest permeability to K⁺ ions.

The intrinsic features of SMCs, however, cannot generate the important motor patterns of the stomach. SMCs are electrically coupled to two types of interstitial cells. Interstitial cells of Cajal (ICC) and fibroblast-like cells (PDGFRα⁺ cells) are coupled to SMCs by gap junctions,^18,19 forming an electrical syncytium known as the SIP (abbreviation for SMCs, ICC and PDGFRα⁺ cells) syncytium.²⁰ The syncytial nature of gastric muscles means that electrical responses that develop in interstitial cells can conduct to SMCs and regulate the excitability and motor activity of SMCs. There are 2 basic classes of ICC: intramuscular ICC (ICC-IM) are closely aligned with varicose projections of enteric motor neurons²¹ (Figure 1A–C²²), and ICC in the myenteric region (ICC-MY) form a complex cellular network in the region between the circular and longitudinal muscle layers²³ (Figure1 D–G^22,24). ICC-MY generate pacemaker activity.²⁵ More details about the functions of ICC-IM and PDGFRα⁺ cells in the SIP syncytium (Figure 2A²⁶ and 2E²⁷) are discussed in the section on “Neuromuscular Control Mechanisms”.

Figure 1. 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 antrum (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). These recordings are made from an impaled smooth muscle cell within a small sheet of antral muscle. In the intact stomach, both the slow waves and the contractions they induce propagate from corpus to the pyloric sphincter, constituting gastric peristalsis.

Reproduced with permissions from: Panels A-F, from ref. 22, Blair PJ et al, Cell Tissue Res 2012;350:199-213; Panel G, from ref. 24, Komuro T. Atlas of interstitial cells of Cajal in the gastrointestinal tract, Springer Science, 2012 edition; Panel H, redrawn from figures in ref. 28, Rhee PL et al, J Physiol 2011;589:6105-18.

Figure 2. A. Electron micrograph (x19,000) showing elements of the SIP (abbreviation for SMCs, ICC and PDGFRα⁺) syncytium in the stomach of guinea pig. Both PDGFRα⁺ cells and ICC (c-Kit⁺ cell) make close contacts with bundles of enteric neurons. In some places, very close contact between varicosities and ICC occur (<20 nm; see area denoted by red arrow in the inset (x44,000) which is a blowup of the varicosity noted by the black arrow in A). Both PDGFRα⁺ cells and ICC form gap junctions with smooth muscle cells (not shown in this image), forming an electrical syncytium. ICC transduce cholinergic and nitrergic inputs and PDGFRα⁺ cells transduce purinergic inputs from enteric motor neurons. B-D: enteric motor neurons (PGP9.5; green [B]) and c-Kit (red [C]) labeling in monkey stomach; D shows merged images demonstrating the close relationship between the projections of motor neurons and ICC; E: schematic of SIP syncytium with ICC in red and PDGFRα⁺ cells in green, also showing formation of gap junctions with smooth muscle cells.

Reproduced with permissions from: Panel A, from ref. 26, Mitsui & Komuro. Cell Tissue Res 2002;309:219–227; Panel B-D, from ref. 36, Sung TS, et al. J Physiol 2018;596:1549-1574;Panel E, from Fig. 2 in ref. 27, Sanders KM et al. J Physiol 2010;588:4621-4639.

ICC-MY express unique ionic conductances that generate and actively propagate electrical slow waves (Figure 1H²⁸), the electrophysiological events that generate the phasic contractions of gastric peristalsis. A dominant pacemaker region exists in the proximal corpus at or near the greater curvature. ICC-MY in this location are the most excitable and generate slow waves at the highest frequency. Slow waves are initiated by small depolarizations due to spontaneous Ca²⁺ release in ICC-MY and activation of Ca²⁺-activated Cl⁻ channels (encoded by ANO1 or anoctamin 1 gene).²⁹ Transient activation of ANO1 channels and depolarization causes activation of low-threshold voltage-dependent Ca²⁺ channels (dihydropyridine-insensitive, T-type Ca²⁺ channels), resulting in rapid depolarization, an event not unlike a Ca²⁺ action potential that can propagate regeneratively through the network of ICC-MY. This event is known as the upstroke of the slow wave. Ca²⁺ entry during the upstroke sets off additional Ca²⁺ release events, through a process known as Ca²⁺-induced Ca²⁺ release, maintaining the activation of ANO1 channels and creating the second component of the slow wave, the plateau phase. During the plateau phase, the elevated permeability of Cl⁻ ions (via sustained activation of ANO1 channels) is dominant relative to the permeability of other ions. This causes membrane potential to linger near the equilibrium potential for Cl⁻ ions, generating the depolarized conditions during the plateau potential. Termination of Ca²⁺ release causes deactivation of ANO1 channels (which are Ca²⁺ dependent), switching back from dominant Cl⁻ permeability to dominant K⁺ permeability, and this causes repolarization to the resting (i.e. inter-slow wave) potential.

Slow waves conduct to SMCs, causing depolarization, activation of L-type Ca²⁺ channels and excitation-contraction coupling. SMCs do not express the ion channels required for active propagation of slow waves, and thus an intact network of ICC-MY is necessary for normal gastric motility.³⁰

NEUROMUSCULAR CONTROL MECHANISMS

Most of what is described in the next sections is based on animal research. The enteric nervous system (ENS), which consists of about 100 million neurons throughout the entire gut,³¹ regulates tonic contraction in the proximal stomach and the amplitude and frequency of phasic contractions in the distal stomach. The ENS is organized in distinct ganglionated plexi, including the submucous plexus which is mainly involved in absorption and secretion, and the myenteric plexus which regulates motility.³¹ Innervation of the gastric muscularis results from both excitatory and inhibitory enteric motor neurons³² (Figure 3³³). Excitatory neurons release acetylcholine (ACh) and tachykinins (TK). Inhibitory neurons release nitric oxide (NO), vasoactive intestinal polypeptide (VIP), pituitary adenylate cyclase-activated peptide (PACAP), and purines. Retrograde neural tracing has shown that cell bodies of muscle motor neurons are in myenteric ganglia, with excitatory neurons projecting proximally and inhibitory neurons projecting distally.³⁴

Figure 3. Myenteric plexus showing neurons labeled with neuronal nitric oxide synthase 1 (nNOS1) (A) and vesicular acetylcholine transporter (VChaT) (B). Some of the nNOS⁺ neurons are inhibitory muscle motor neurons, and some of the VChaT neurons are excitatory muscle motor neurons. The images A and B showing respectively nNOS1 and VChaT neurons are merged in C. Panel D shows a magnified view of a single ganglion from C.

Reproduced with permission from ref. 33, Cipriani G, et al. Cell Mol Gastroenterol Hepatol 2019;7:689-691.e4.

Motor neurons innervate the SMCs through ICC and PDGFRα⁺ cells^22,28,35 (Figure 2B–D³⁶). ICC-IM mediate responses to cholinergic excitatory³⁷ and nitrergic inhibitory²¹ neurotransmission. PDGFRα⁺ cells mediate responses to purinergic inhibitory neurons in GI muscles.³⁸ ICC-IM form synaptic-like connections with varicosities of enteric motor neurons that are thought to be sites of neurotransmission (Figure 2A). In the case of cholinergic neurotransmission, the tiny junctional volumes between ICC-IM and nerve varicosities facilitate rapid metabolism of ACh, thereby limiting its diffusion through the interstitium. NO may have dominant effects in ICC-IM, which would also be due to the close proximity of ICC-IM to sites of synthesis and release. PDGFRα⁺ cells, which are also closely associated with enteric motor neurons (Figure 2A), mediate purinergic response by dominant expression of key receptors and ion channels. Cholinergic nerve stimulation causes a dramatic increase in Ca²⁺ release in ICC-IM in colonic muscles;³⁹ however, this has not yet been documented in gastric ICC-IM. In contrast, NO inhibits Ca²⁺ release in ICC-IM. As noted above, release of Ca²⁺ is coupled to activation of ANO1 channels, such that enhancing Ca²⁺ release would elicit a general depolarizing trend in the SIP syncytium and increase SMC excitability. Inhibiting Ca²⁺ release in ICC-IM would reduce ANO1 activation and reduce excitability.

Excitatory nerve stimulation in the stomach has both inotropic and chronotropic consequences. The inotropic effect is to enhance the amplitude and duration of slow waves, resulting in stronger peristaltic contractions. The chronotropic effects increase the frequency of slow waves. It does not appear that the chronotropic effects are due to direct cholinergic innervation of ICC-MY, the pacemaker cells. Instead, the depolarizations, caused by cholinergic stimulation of ICC-IM, are capable of enhancing pacemaker frequency.⁴⁰ Chronotropic effects of excitatory nerve stimulation are lacking in mutant animals lacking ICC-IM,⁴¹ confirming the importance of ICC-IM in transducing neural inputs to regulate slow wave frequency.

Extrinsic innervation comes from both sympathetic and parasympathetic nerves.^30,42–44 Parasympathetic efferent neurons, with cell bodies in the dorsal motor nucleus in the brainstem, innervate myenteric neurons, with 70% or more of gastric neurons, including both excitatory and inhibitory motor neurons, receiving direct input from vagal efferent neurons.⁴⁵ Vago-vagal reflexes are responsible for many of the gastric responses to eating. For example, ingestion of food leads to relaxation of the proximal stomach, gastric accommodation, through a vago-vagal reflex that ultimately activates enteric inhibitory motor neurons to release NO.^46,47 Sympathetic innervation originates in the intermediolateral cell column of the thoracic spinal cord levels 5 to 10 and synapses with post-ganglionic neurons in the celiac and superior mesenteric ganglia. Sympathetic neurons regulating motility are focused on myenteric ganglia, but sympathetic fibers, labeled with tyrosine hydroxylase antibodies, also innervate the muscle layers directly. Sympathetic input inhibits gastric contraction, but this occurs primarily when vagal excitatory nerves are active, suggesting pre-junctional targeting of sympathetic fibers to vagal inputs to excitatory motor neurons.

SPECIALIZATIONS OF THE PYLORIC SPHINCTER

The pyloric sphincter is a narrow zone of thickened muscularis and increased luminal pressure (radiologically estimated to be 1.2cm in width) at the junction between the stomach and duodenum. Pyloric contractions and relaxations also depend upon transduction of neural signals by ICC and conduction of responses to SMCs.⁴⁸

Studies of pyloric muscles in vitro suggest that there are at least two independently-controlled, functional areas of the pyloric musculature: 1) the circular muscle close to the myenteric plexus, which is dominated by the propagation of gastric slow waves, and results in sphincteric contractions at the termination of each gastric peristaltic event, and 2) the deeper, thickened circular muscle regulated by motor neurons, as slow waves do not propagate into this region.⁴⁹ Independent control of the two regions may provide temporal and functional regulation of pyloric resistance.

At the gastroduodenal junction, there is separation of the electrical and mechanical functions of the stomach and pylorus from the duodenum, despite clear anatomical connections between the muscular tissues. ICC generate and propagate slow waves in both the stomach and small intestine,³⁰ and there is reduction in the density of ICC networks within a narrow zone of between the pylorus and duodenum that appears not to support propagation of slow waves, thereby facilitating independent electrical and motor activities in the two organs.⁴⁸

Enteric inhibitory neurotransmission, measured by inhibitory junction potentials (IJPs) in canine pyloric muscles, is mediated primarily by nitric oxide (NO) and by a purine neurotransmitter^50,51 acting through activation of small conductance Ca²⁺-activated K⁺ (SK) channels. Inhibition of IJPs unmasked excitatory junction potentials (EJPs) in the myenteric region and increased the excitability of SMCs and contractions in the submucosal region of the pylorus. Both the myenteric and submucosal layers of pyloric muscles are innervated by nitrergic neurons, as confirmed by NOS-like immunoreactivity (NOS-LI) ^48,52

As in the main areas of the stomach post-junctional nitrergic responses depend upon ICC.^48,53 Lesions in the nitrergic pathway that regulates pyloric motor activity, whether due to loss of NO synthase (NOS), loss of NOS neurons, or loss of ICC that contribute to neurotransduction of nitrergic signals, result in abnormal regulation of pyloric relaxation and, therefore, could impede gastric emptying. It should be noted that loss of ICC appears to have clinical relevance because reduction and/or morphological abnormalities in ICC have been noted in full-thickness biopsies from gastric muscles of patients with idiopathic and diabetic gastroparesis.^54,55 Recent studies have also shown reduced pyloric ICC in patients with gastroparesis.⁵⁶

Enkephalinergic nerve fibers are also present in the tunica muscularis of dogs,⁵⁷ cats,⁵⁸ and humans,⁵⁹ suggesting regulation of neural responses by endogenous opiates, including inhibition of both cholinergic excitatory and nitrergic inhibitory junction potentials in canine pyloric muscles.⁶⁰ Endogenous opiate peptides (e.g., met-enkephalin) participate in regulation of pyloric contraction (e.g., in response to duodenal acidification),⁶¹ and exogenous opioids interfere with normal neural regulation of the pylorus and cause stimulation of pyloric tone and phasic contractility.⁶² The effects of opioids to increase pyloric tone are mediated by inhibition of nitrergic relaxation⁵⁶ or cholinergic stimulation.⁶³

Functions of the pyloric sphincter regulating gastric emptying are not entirely understood. Instillation of acid into the duodenum, but not the antrum, increased the frequency and amplitude of phasic contractions in the pylorus, and this response was antagonized by tetrodotoxin (implying neural mediation) and intraluminal naloxone (opioid antagonist). Conversely, there was no effect noted with cholinergic, adrenergic, or serotonergic modulation of pyloric contraction in response to intraduodenal acid.⁶⁴

PATHOPHYSIOLOGY OF GASTROPARESIS

A. Neuromuscular Diseases Affecting Non-Sphincteric Gastric Muscles

Figure 4 summarizes the disorders of extrinsic and enteric neural control and muscle resulting in motility disorders.⁴⁴ Idiopathic and diabetic gastroparesis are typically associated with postprandial antral hypomotility with reduced frequency (average <1/minute postprandially) and normal amplitude contractions;⁴⁹ infiltrative disorders such as scleroderma result in low amplitude contractions antral (<40mmHg) and intestinal contractions (<2mmHg);^6,65 the latter resulting in delayed gastric emptying and small bowel transit, and absence of ileocolonic bolus transfers. In over 1280 patients with upper gastrointestinal symptoms who underwent both gastric emptying and gastric accommodation studies, it was shown that increased accommodation (postprandial to fasting ratio >3.85) was more prevalent in patients with delayed compared to accelerated gastric emptying.⁶⁶ Higher postprandial gastric volumes are also associated with delays in gastric emptying of solids measured simultaneously.⁶⁷

FIGURE 4.

A. Intrinsic excitatory cholinergic neuron (ChAT = choline acetyl transferase). B. Intrinsic inhibitory nitrergic neuron (NOS=nitric oxide synthase). C. Extrinsic neural control of gastrointestinal motor function. Parasympathetic supply is generally excitatory (indicated by + sign) to non-sphincteric muscle or excitatory to inhibitory intrinsic nerves (indicated by − sign, e.g. nitrergic neurons). Sympathetic nerves are generally inhibitory to non-sphincteric muscle and stimulatory to sphincteric muscle such as the pylorus.

B. Post-Surgical or Post-Bariatric Endoscopy Gastroparesis

Vagotomy performed for peptic ulcer disease is invariably associated with a drainage procedure, either pyloroplasty or gastrojejunostomy. Post-surgical gastroparesis⁶⁸ is associated with partial gastrectomy and is either associated with extrinsic denervation of the gastric remnant or abnormal motility in the anastomosed jejunal loop.^69,70 Proximal gastric vagotomy inhibits gastric tone and delays gastric emptying of liquids without altering antral contractility.⁷¹

Fundoplication may result in gastroparesis⁷² due to vagal injury, and this is associated with impaired antral motility.⁷³ Transient vagal injury may result from endoscopic variceal sclerotherapy⁷⁴ or from radiofrequency ablation⁷⁵ of accessory conduction pathways in the heart and may be manifest by dysphagia, nausea, vomiting, and non-specific esophageal motor disorders, accelerated gastric emptying of solids at 1 hour, or delayed gastric emptying of solids. The restoration of vagal function is associated with relief of symptoms or normalized objective tests.

Partial gastrectomy results in acceleration of gastric emptying of liquids and overall stasis if there is associated vagotomy, as was the norm in surgery performed for peptic ulceration, and was associated with either dumping syndrome (involving mainly liquids) and/or gastric stasis of solids.⁷⁶ Dogs with Billroth I gastrectomy with loss of the pylorus had larger solid food particles emptied, and this may contribute to malabsorption following this form of gastric surgery.¹⁰

Uncomplicated fundoplication⁷⁷ and sleeve gastrectomy⁷⁸ accelerate gastric emptying of solids, as the reservoir capacity of the stomach is reduced and repeated esophageal peristaltic contractions induce isobaric pressurization of the proximal stomach,⁷⁹ thus providing the drive to pressurize and empty the vertical compartment of the gastric sleeve.

Sleeve sastroplasty produces a funnel-shaped stomach with a constricted middle and distal stomach, resulting in delayed gastric emptying.⁸⁰ Inclusion of the distal antrum in the gastroplasty may impact the solid trituration and emptying of the solid phase of the meal from the stomach.

C. Pyloric Dysfunction

Pyloric dysfunction was first described in diabetic gastroparesis in 1986⁷ as unusually prolonged, but intermittent contractions characterized by marked increases in baseline tone at the pylorus, or “pylorospasms”. These findings were reported in 24 diabetic patients with symptoms of gastroparesis.⁷ Concomitant pylorospasm, antral hypomotility and evidence of extra-intestinal autonomic neuropathy in 13/24 patients suggested the pylorospasm was secondary to diabetic neuropathy ⁷ Since loss of ICC has also been reported in diabetic gastroparesis,⁸¹ it is also possible that these effects could be due to a damaged ICC-MY network (antral hypomotiity) or loss of pyloric ICC-IM (pylorospasm). In recent years, the pylorus has become a potential target for endoscopic treatment, as described below. Pyloric dysfunction and gastric stasis may also result from effects of opioid use, which is becoming increasingly prevalent in patients with gastroparesis.

In diabetic gastroparesis, excessive postprandial pyloric tonic and phasic pressure activity may accompany antral hypomotility,⁷ and both may be due to damage of ICC-MY (antral hypomotility) and ICC-IM (pylorospasm). Researchers at Temple University⁸² performed pyloric studies using the EndoFLIP™ device (composed of 17 impedance electrodes spaced 4 or 8mm apart and one solid state pressure transducer to determine intra-balloon pressure) in 39 idiopathic and 15 diabetic gastroparesis patients. They measured the pressure, diameter, cross-sectional area (CSA), and distensibility index (CSA in cm² divided by the simultaneous intra-balloon pressure in mmHg), and documented the wide range in diameter (5.6 to 22.1mm) and distensibility (1-55mm²/mmHg) of the pylorus. Symptoms of early satiety and postprandial fullness were inversely correlated with pyloric sphincter diameter and CSA.⁸² Several studies have shown that pyloric distensibility was decreased in gastroparesis and correlated with gastric emptying as well as gastroparesis symptoms.⁸³

Potential Cellular Basis for Gastroparesis: Oxidative Stress and Inflammation

Experimental models of gastroparesis show a reduction in or remodeling of ICC-IM, leading to secondary effects in SMCs because of the lack of trophic factors [e.g., stem cell factor (SCF)].⁸⁴ A few case reports document the reduced numbers of ICC-IM in patients with gastroparesis.⁸⁵ This loss of ICC appears to result from imbalance between processes that injure ICC networks and those that generate and maintain ICC. For example, relative insulinopenia and insulin growth factor-1 (IGF-1) deficiency in diabetes leads to reduced production of SCF by SMCs, an important ICC survival factor.⁸⁶ Moreover, diabetes is associated with high oxidative stress which may result from downregulation of macrophage heme oxygenase-1 (HO-1).^87,88 Loss of CD206⁺, anti-inflammatory M2 (resident) macrophages, and increased expression of genes associated with pro-inflammatory M1 macrophages in full-thickness gastric biopsies from patients with gastroparesis have also been reported. Depletion of M2 macrophages, which express heme oxygenase-1 (HO1), leads to oxidative damage to the pacemaker cells.

Light microscopic studies of full thickness gastric biopsies from patients with idiopathic and diabetic gastroparesis showed no significant differences in nerves or SMC cell markers, except for reduction in expression of neuronal NO synthase in diabetic compared with idiopathic gastroparesis.⁸¹ At the ultrastructural level, diabetic gastroparesis was associated with a thickened basal lamina around SMCs, and altered neuronal cell bodies and nerve endings and fibrosis around nerves were noted as more severe in idiopathic than diabetic gastroparesis.⁸⁹

From proteomics and deep sequencing of gene transcripts^90–92 of full-thickness gastric body biopsies, it was shown that granulocyte adhesion and diapedesis, as well as a macrophage-based immune dysregulation pathway are the most significantly affected pathways altered in both diabetic and idiopathic gastroparesis. M1 (pro-inflammatory) macrophages were increased in idiopathic gastroparesis samples compared to their controls. In addition, diabetic gastroparesis was associated with proteins involved in the complement and prostaglandin synthesis pathway. However, the same study revealed no enrichment of genes associated with M1 (marrow-derived) or M2 (resident) macrophages in the biopsies relative to diabetic control samples. Finally, diabetic gastroparesis biopsies had reduced expression of inflammatory markers. The significance of these findings is unclear because diabetic gastroparesis has been associated with M2 macrophage deficiency, which would be expected to result in increased inflammation.

Overall, further research on both human biopsies and animal models of diabetes is needed to understand the molecular and cellular bases for gastroparesis. The challenges include potential of sampling error impacting morphological or expression studies, the need to clarify the role of inflammatory mechanisms, the impact of vagal denervation (e.g., associated with diabetes mellitus) on inflammation given the anti-inflammatory ENS-macrophage nAChR cholinergic pathway,⁹³ and the need for treatments to effectively inhibit oxidative stress on the gut neuromuscular apparatus. A recent controlled drug trial of hemin failed to achieve the pharmacokinetic goals to test its efficacy,⁹⁴ so tests of other therapies are needed.

SYMPTOMS IN RELATION TO GASTRIC MOTOR DYSFUNCTIONS

The relationship between impairment of gastric emptying of a mixed radiolabeled meal or gastric accommodation (by SPECT) and upper gastrointestinal symptoms was examined in almost 1300 patients at Mayo Clinic.⁶⁶ Nausea, vomiting, weight loss and abdominal discomfort were more prevalent in those with delayed gastric emptying compared to those with accelerated or normal gastric emptying; bloating was associated with accelerated gastric emptying. In 108 of those tertiary referral patients with diabetes and upper gastrointestinal symptoms consistent with gastroparesis,⁹⁵ the most common presenting symptoms were nausea (80.6%), vomiting (53.7%), and abdominal pain (52.8%). The most frequent symptom associated with abnormal gastric accommodation was belching.

In the NIH Gastroparesis Consortium multicenter database,^96,97 severely delayed gastric emptying of a 2% fat, 255 kcal meal was associated with worse vomiting, more severe anorexia and overall gastroparesis symptoms, as well as stomach fullness, postprandial fullness, and early satiety.⁹⁸

On the other hand, gastroparesis symptoms showed little association with profiles of emptying of a wireless motility capsule (WMC).⁹⁹ This discrepancy may reflect the different results¹⁰⁰ obtained with an easily triturated egg-substitute meal compared to a relatively large WMC that is usually emptied from the stomach with the return of the phase 3 of the interdigestive migrating motor complex.¹⁰¹ Hinder and Kelly had previously demonstrated,¹⁷ that emptying of 7-mm diameter radiopaque spheres from canine stomach did not occur for at least 4 hours after ingestion of liver cubes and water.

In the original description of gastroparesis diabeticorum, Kassander¹⁰² identified that diabetic triopathy (retinopathy, nephropathy, and neuropathy) was highly prevalent. In more recent cohorts,⁹⁵ diabetic triopathy was uncommon at the time of presentation with upper gastrointestinal symptoms: 10% of patients with type 1 diabetes and 3% with type 2 diabetes had triopathy with gastroparesis. Indeed, 39% of patients with diabetes and gastroparesis did not have any complications of diabetes.¹⁰³

DIAGNOSTIC TESTS

Typically, diagnostic testing is primarily guided by symptom pattern and severity. Delayed gastric emptying can be documented by scintigraphy, stable isotope breath test; however, presence of retained food in the stomach at endoscopy is of limited predictive value¹⁰⁴ unless the patient has a known underlying disease predisposing to gastric retention. Barium studies and scintigraphy using labelled liquid or semi-solid meals are typically normal and of limited diagnostic value, even in the presence of moderately severe symptoms. The tests that are available to measure gastric emptying noninvasively are summarized in the Supplemental Table 1.¹⁰⁵

Scintigraphic assessment of solid emptying over 4 hours is a more sensitive test, with a well-defined normal range, and with the proportion retained at 2 and 4 hours having a sensitivity of 90% and a specificity of 70% to identify delayed emptying.¹⁰⁶ It is not accurate to extrapolate the emptying pattern from scans taken for a shorter duration (typically 90 or 120 minutes).

Another useful test for measuring solid-phase gastric emptying utilizes a standardized biscuit enriched with a ¹³C-enriched substrate. When metabolized, the proteins, carbohydrates, and lipids of the Spirulina platensis or the medium-chain triglyceride, octanoate, result in a rise in respiratory ¹³CO₂ that is measured by isotope ratio mass spectrometry, allowing for estimation of gastric emptying T_1/2.^107–109

The WMC, which senses pressure, temperature, and pH, has been approved by the U.S. Food and Drug Administration (FDA) for the evaluation of gastric emptying and colonic transit time in people with suspected slow transit constipation.¹¹⁰ Gastric emptying time is identified by a rise in pH from gastric baseline to >4.0 in the duodenum.¹¹⁰ The estimated gastric emptying time with the WMC is not as accurate as scintigraphy with a digestible solid meal. The WMC identified lower numbers of contractions and motility indices in gastroparesis.¹¹¹ Use of the capsule is contraindicated in children and those with a known history of oesophageal stricture or possible intestinal obstruction.

For people with severe upper gastrointestinal symptoms, antropyloroduodenal manometry assesses pressure profiles in the stomach and small bowel; neuropathic disorders are associated with distal antral contractile frequency of less than an average of 1 per minute during the first postprandial hour, whereas myopathic disorders are associated with amplitude less than 40mm of mercury and small intestinal amplitude less than 10mm of mercury.^6,112 Manometry may also reveal excessive tonic and phasic pressure activity at the pylorus,⁷ though the sensitivity of manometry for measurement of pyloric function is not established. Pyloric dysfunction is also documented with a functional luminal imaging probe performed during endoscopy (EndoFlip™), which assesses diameter and distensibility of the pylorus.^{106,110,111,113,114} It has been proposed that this measurement might facilitate selection of those for pyloric interventions for gastroparesis.

MANAGEMENT

Management of gastroparesis should include assessment and correction of the nutritional state, relief of symptoms, improvement of gastric emptying, reversal of iatrogenic gastroparesis (chiefly due to opioids) and, in patients with diabetes, glycemic control.

Nutrition

The first step in the management of gastroparesis is to educate the person to use a small particle diet, aided with cooking of nondigestible fiber and homogenization of solids to a small particle size.¹¹⁵ Although nutritional requirements and symptoms can be addressed to a variable extent in persons with mild and compensated gastroparesis, those with severe gastroparesis often require hospitalization for one or more of the following measures: intravenous hydration and correction of metabolic derangements (ketoacidosis, uraemia, hypoglycaemia, hyperglycaemia), nasoenteric decompression, and/or enteral nutrition to manage vomiting and nutritional requirements.¹¹⁶

For individuals with severe gastroparesis who do not respond to the measures outlined above, it may be necessary to bypass the stomach with a jejunal feeding tube.¹¹⁷ This procedure should be preceded by a trial of nasojejunal feeding for a few days, with infusion rates of at least 60 mL of iso-osmolar nutrient per hour. It is preferable to place jejunal feeding tubes directly into the jejunum either by endoscopy or, if necessary, by laparoscopy, rather than via percutaneous endoscopic gastrostomy tubes. Such tubes allow restoration of normal nutritional status, but they are not without adverse effects, and it is important to allow for a few days of habituation, escalating stepwise and slowly the infusion rate from 10mL/hour to the goal of 60ml/hour.

Parenteral nutrition may become necessary in cases of malnutrition. Ideally, this is a temporary measure for patients presenting with severe weight loss and poor oral tolerance, with conversion to long-term enteral nutrition in order to avoid risks such as central line infection and metabolic complications.

Bezoars may be mechanically disrupted during endoscopy, followed by gastric decompression to drain residual nondigestible particles.

Pharmacological Interventions

a. Stimulation of non-sphincteric muscle contractility

Erythromycin, at a dose of 3 mg/kg body weight intravenously every 8 hours, can accelerate gastric emptying.^118,119 When oral intake is resumed, treatment with oral erythromycin, 250 mg t.i.d. daily, for 1 to 2 weeks is worthwhile. The prokinetic effects of erythromycin are limited by tachyphylaxis.

In the United States, the only available medication is metoclopramide, a peripheral cholinergic and antidopaminergic agent. During acute administration, it initially enhances gastric emptying of liquids in diabetic gastroparesis, but its symptomatic efficacy (data summarised in reference¹²) is probably related to its central antiemetic effects. However, its long-term use is restricted by a decline in efficacy and by central nervous system side effects, most commonly reversible involuntary movements, and tardive dyskinesia (irreversible) in the range of 0.1% per 1000 patient years.¹²⁰ Therefore, it is preferred to prescribe a dose of 10 mg t.i.d., administered 30 minutes before meals and FDA guidance suggests only for a period of up to 3 months. Intranasal formulation of metoclopramide¹²¹ has been approved for treatment of adults with gastroparesis, though its efficacy was only proven in women in the controlled trial.

Experimental medications in development for gastroparesis are felcisetrag (5-HT4 agonist),¹²² tradipitant (NK1 antagonist),¹²³ relamorelin (ghrelin agonist),¹²⁴ and trazpiroben (dopamine D₂/D₃ receptor antagonist).¹²⁵

Supplemental Table 2 lists current^{119,122–138} and investigational¹³⁹ prokinetic drugs for gastric motility disorders, as recently summarized.¹⁴⁰

b. Two pharmacological approaches have been pursued to reverse pyloric dysfunction.

First, is reversal of the reduced nNOS expression and decreased cGMP content with the phosphodiesterase-5 inhibitor, sildenafil. Sildenafil had no significant effect on gastric emptying in gastroparesis associated with uremia.¹³⁸ Second, given that naloxone corrected gastric stasis in patients with gastric hypomotility,¹⁴¹ two studies^142,143 tested peripherally active μ- opioid receptor antagonists (PAMORA), methylnaltrexone (s.c. 0.30 mg/kg) or naloxegol (25 mg), daily on opioid-induced delay in gastric emptying and noted they did not reverse the retardation of gastric emptying in healthy volunteers treated with codeine 30mg QID.¹⁴³ However, alvimopan, administered at 1 log higher than the dose used in OIC, reversed the retardation of transit induced the same dose of codeine.¹⁴⁴

c. Anti-nausea medications

are critical for relief of nausea and vomiting; however, their efficacy is based predominantly on mechanisms of action rather than formal randomized, controlled trials. Ondansetron (4-8mg tid as tablet or oral dissolving tablet) is a 5-HT₃ antagonist that reduces nausea without affecting gastric compliance or postprandial accommodation.¹⁴⁵ The 5-HT₃ antagonist, granisetron, is available as a sustained release transdermal patch,¹⁴⁶ which was shown in an open-label study to reduce nausea and vomiting in gastroparesis. Prochlorperazine and promethazine reduce nausea through effects on dopamine (D2) and histamine (H1) receptors, respectively, and are available as oral or rectal formulations. Drowsiness, dry mouth, and constipation are common side effects. Promethazine may be habit forming and is reserved as a “rescue” agent. In a placebo-controlled trial, the NK1 receptor antagonist, aprepitant (which acts on the vomiting center and is approved for chemotherapy-induced emesis), improved nausea.¹³⁶ Many patients with gastroparesis use ^Δ9THC¹⁴⁷ for symptomatic relief; however, marijuana¹⁴⁸ and the nonselective cannabinoid receptor agonist, dronabinol,¹⁴⁹ retard gastric emptying.

Gastric Electrical Stimulation (GES)

A systematic review and meta-analysis¹⁵⁰ documented that 5 studies that randomly allocated patients to periods with or without GES showed no significant differences in the total symptom severity (TSS) scores between these periods in contrast to 16 open-label studies of GES which showed a significant TSS decrease. A 48-week observational study from the NIH Gastroparesis Consortium¹⁵¹ demonstrated significant benefit for one symptom (nausea) in 81 patients with GES as compared to 238 patients without GES. In a randomized, double-blind, crossover trial¹⁵² of 172 patients with refractory vomiting (133 with confirmed gastroparesis), GES improved vomiting score, but not gastric emptying or quality of life.

RELATIONSHIP OF SYMPTOMS, GASTRIC EMPTYING, AND RESPONSE TO PROKINETIC AGENTS

Efficacy of pharmacological agents in the treatment of gastroparesis has been documented in systematic reviews and meta-analyses in critically ill patients with feeding intolerance requiring enteral nutrition,¹⁵³ as well as in patients with gastroparesis.^154,155 However, questions have arisen regarding the association of gastric emptying delay and symptoms suggestive of gastroparesis.¹⁵⁶ Recent studies based on meta-regression of symptoms and gastric emptying measurements obtained with scintigraphy and stable isotope breath tests demonstrated that, when emptying was measured using a solid meal and there was assessment of emptying for >2 hours, that is 3 or 4 hours, after ingestion of the test meal, there was a significant association between gastric emptying and symptoms.¹⁵⁷ In addition, when the same criteria to identify studies with optimal measurement of gastric emptying were used to select the studies for analysis, meta-regression demonstrated an association between acceleration of gastric emptying and reduction in symptoms;¹⁵⁵ specifically, a 20.4 minute reduction in gastric emptying T_1/2 was associated with a clinically relevant improvement in symptoms.

INTERVENTIONS TO INCREASE PYLORIC DIAMETER/DISTENSIBILITY AND IMPACT ON SYMPTOMS AND GASTRIC EMPTYING

The advent of procedures directed at the pylorus (botulinum toxin injection, surgical or endoscopic pyloromyotomy) has renewed interest in the role of the region in the pathophysiology of gastroparesis.

Multiple open-label trials of intrapyloric botulinum toxin injections for gastroparesis suggested efficacy,^158–160 but the two randomized, controlled trials^161,162 failed to show any improvement in symptoms. It is unclear whether participants had pyloric dysfunction at baseline, and it is unclear if the neurotoxin had the expected effects on the pylorus, although one trial¹⁶¹ confirmed improved gastric emptying compared to placebo. However, fasting pyloric compliance (based on measured pressure and pyloric diameter) was decreased in patients with gastroparesis and was associated with delayed gastric emptying, symptoms and quality of life.¹⁶³ These data suggest that it may be useful to target pyloric compliance by pyloric dilation or botulinum toxin injection in patients with gastroparesis;¹⁶³ however, rigorous controlled trials are required that also include measurements of pyloric diameter and distensibility and impact on patient response outcomes in addition to motor functions.

Surgical pyloroplasty has been associated with short-term improvements in symptom severity scores three months post-procedure and accelerated gastric emptying in several case series.^164–166 The standard Heineke-Mikulicz pyloroplasty involves transverse closure of a longitudinal incision across the pylorus which involves both longitudinal and circular muscle layers. These studies set the stage for gastric peroral endoscopic myotomy (G-POEM)¹⁶⁷ which divides the pylorus from the mucosal surface and presumably cuts predominantly the circular muscle layer while maintaining the longitudinal muscle to avoid perforation. Improved gastric emptying with G-POEM has also been documented.¹⁶⁸ Current evidence on the efficacy of G-POEM for gastroparesis is summarized in Table 1,^169–179 which shows results of exclusively uncontrolled trials. There is a critical need for a controlled study of the effect of G-POEM in patients with well-characterized gastroparesis, including baseline and post-treatment measurements of antropyloroduodenal motor function and the diameter, distensibility, and compliance of the pylorus itself, in order to ascertain whether there are predictors of responsiveness to the G-POEM procedure.

Table 1. Published studies on gastric peroral endoscopic myotomy (G-POEM).

COPD=chronic obstructive pulmonary disease; EDS=Ehlers Danlos syndrome; GE=gastric emptying; GES=gastric electrical stimulation; QOL=quality of life. The Gastroparesis Cardinal Symptom Index (GCSI) scores are mean values pre- or post-G-POEM.

# Pts	Types of gastroparesis pts	Changes in GE	Changes in symptoms	Duration follow up	Adverse events	Ref.#
29	diabetic=7 idiopathic= 15 post-surgical=5 scleroderma=2	70% Normalized	79% at 3 months; 69% at 6 months. GCSI improved from 3.5 to 0.9 at 3 months	3 and 6 months	17% (2/12) Pneumoperitoneum requiring decompression	169
16	diabetic = 9 idiopathic = 5 post-surgical = 1 post-infectious= 1	75% normalized, 25% improved	81% improvement. GCSI improved from baseline of 3.4 to 1.46 12 months later	12 months	None	170
47	diabetic =12 idiopathic =27 post-surgical=8	4h retention improved: from 37.2 to 20.4%	GCSI improved from 4.6 to 3.3	3 months (follow-up in 31/47 pts)	1 death (unrelated)	171
30	diabetic =11 idiopathic =7 post-surgical =12	47% Normalized	No validated outcome measure available	6 months	2/30 (6%): 1 prepyloric ulcer and 1 capnoperitoneum	172
13	diabetic =1 idiopathic = 4 post-surgical = 8	4/6 improved; % retention at 4h improved from 49 to 33%	In 11:4 considerably, 4 somewhat better, 1 no Δ, 2 worse	3 months	3 accidental mucosotomy closed with clips; 1 pulmonary embolism	173
16	Diabetes =3 post-surgical =13	Mean % retention (radiolabeled bread) at 2h from 69.3% to 33.4%	Mean total symptom score from 24.25 to 6.37; 13/16 substantial improvement	3 months	1 pyloric stenosis at day 45	174
20	diabetic =10 non-diabetic=10	% retention at 4h improved from 57.5 to 15%; and 30% normalized	GCSI improved from 3.5 to 1.3; QOL improved	3 months	3 mild hemorrhage, 3 gastric perforation, 1 moderate dyspepsia	175
40	diabetic =15 nondiabetic =25 (of which 18 idiopathic)	% retention at 4h reduced by 41.7%	Improved GCSI, nausea/ vomiting, not bloating	median 15 months	1 tension capnoperitoneum, 1 exacerbation of COPD; 1 (Ehlers-Danlos syndrome) disrupted mucosotomy + ulcer	176
22	diabetic =8, idiopathic=14, all with GES and most with diverse other procedures	7/11 with post-G-POEM GE were normal	GCSI improved (reduction 1.63 points); improved all sub-scores	1 and 3 months	1 laparoscopy for pain due to capnoperitoneum and adhesions	177
38	Post-surgical gastroparesis (76% for fundoplication or hiatal hernia repair)	% retention at 4h improved from 46.4 to 17.9%; 50% normalized	GCSI improved (mean reduction 1.29 points); improved all sub-scores	1 month	2 readmissions: 1 melena; 1 dehydration	178
80	Idiopathic (41.3%), postsurgical (35%) and diabetes (23.8%).	GE improvement in 64.2% and normalized in 47.2% (of 53 cases with test) at 3 months	Decrease in total GCSI >1 + >25% decrease in at least two of the subscales in 66.6% at 12 months	3 months GES, 12 months clinical	3 symptomatic capnoperitoneum, 1 mucosotomy; 1 thermal mucosal injury	179

Several individual studies and published systematic reviews and meta-analyses based on 10 studies involving 292 patients¹⁸⁰ have documented early and medium-term efficacy of gastric per oral endoscopic myotomy (G-POEM) in the treatment of gastroparesis and improvement of gastric emptying, as well as the superiority of this approach to gastric electrical stimulation for gastroparesis and equivalence of results to surgical pyloroplasty based on 332 patients with G-POEM (11 studies) and 375 patients who underwent surgical pyloroplasty (seven studies).¹⁸¹

It has been suggested that the change in diameter and distensibility index of the pylorus after G-POEM is associated with improved therapeutic outcome.^83,114,182 Two studies have shown that the average change in the diameter of the pylorus may be about 2mm¹¹⁴ or 1.2mm,¹⁸² and this has led to the reiteration that, if antral hypomotility is a component of the pathophysiology in a patient undergoing G-POEM, then merely widening the pyloric diameter may not be sufficient therapy.¹⁸³ In fact, the average post-G-POEM changes in the diameter of the pylorus in those with clinical success and clinical failure were estimated to be 7.46 and 1.92mm respectively¹⁸² with the EndoFLIP™ device distended to 50mL. Watts et al.¹¹⁴ also showed that, following G-POEM, the number of phasic contractions of the pylorus remained unchanged compared to recordings prior to G-POEM, and this is consistent with the known antropyloric coordination since the circular muscle close to the myenteric plexus propagates gastric slow waves to the pyloric region resulting in sphincter contraction.⁴⁹ It is unknown whether the phasic contractions at the pylorus actually provide resistance to flow of content out of the stomach; the option of extending the pyloromyotomy to the distal antrum in order to reduce pyloric phasic contractions might further aggravate gastric emptying by impacting antral trituration of solids. A recent study reported that gastroparesis symptoms were unrelated to the inner diameter or length of the pylorus or thickness of the pyloric wall, before or after G POEM.¹⁸⁴ Further details on how to optimize functional benefit from the endoscopic pyloroplasty are therefore required. In addition, it is important to note that such surgical or endoscopic procedures may be complicated by development of dumping syndrome.

FUTURE TREATMENT OF GASTROPARESIS: INDIVIDUALIZE TREATMENT BASED ON THE DISORDER OF FUNCTION

Based on the analyses^155,157 demonstrating the importance of optimal measurement of gastric emptying of solids over >2 hours (that is, 3 or 4 hours) to show correlation with symptoms and beneficial symptom responses, patients should be selected for prokinetic agents based on delayed gastric emptying of solids, based on meta-regression¹⁵⁵ that included 9 studies with low risk of bias and optimal gastric emptying test methodology which found a significant association between change in gastric emptying and upper gastrointestinal symptoms in response to prokinetic agents (specifically, cisapride, revexepride, domperidone, ghrelin, relamorelin, and the motilin receptor agonists, TZP-101 and TZP-102).

There is increased recognition of the overlap of functional dyspepsia and gastroparesis with common symptom profiles and variation in gastric emptying in the patients over time.¹⁸⁵ Further studies are required to explore functions of ICC-IM and ICC-MY in functional dyspepsia. Since a link between loss of ICC and gastroparesis has been made, future investigation into how loss of these cells might be avoided or how the ICC phenotype might be restored should be pursued. Given that the origin of symptoms suggestive of gastroparesis may reflect other mechanisms such as gastric hypersensitivity or alteration in gastric accommodation, other pathophysiologic mechanisms may also be targets for pharmacological therapy. Pharmacological agents have been used to target sensory mechanisms such as mirtazapine,¹⁸⁶ an antagonist of the histamine receptor HI, the α2 adrenergic receptor, and the serotonin receptors 5-HT_2C and 5-HT₃, and the NK1 antagonist, aprepitant,¹³⁶ which in addition to effects on nausea, also increases gastric accommodation.¹³⁷

Finally, it is anticipated that selection of patients for G-POEM should be based on objective measurements such as diameter and distensibility index by EndoFLIP™ ^{105, 182} or postprandial pyloric hypercontractile responses on antropyloroduodenal manometry, EndoFLIP™ can measure antropyloric propagated contractions or isolated pyloric contractions,^114,182 and the impact of such measurements on outcomes of G-POEM are awaited. Nevertheless, a recent study did not identify predictors of benefit with G-POEM based on EndoFLIP™ or antroduodenal manometry.¹⁸⁷

FIGURE 5. Examples of diseases causing gastrointestinal motility disorders. These disorders, including gastroparesis, may result from abnormalities in parasympathetic nerves as occurs in diabetic neuropathy or brainstem diseases affecting vagal and other autonomic nuclei, disorders affecting sympathetic nerves including neuropathies and dysautonomias, and finally diseases affecting primarily the enteric nerves (ENS) or smooth muscle including myopathies such as amyloidosis and mitochondrial neurogastrointestinal encephalopathy (MNGIE).

Reproduced with permission from ref. 44, Camilleri M. Invited Review: Gastrointestinal motility disorders in neurologic disease. J Clin Invest 2021;131:e143771.