United European Gastroenterology (UEG) and European Society for Neurogastroenterology and Motility (ESNM) consensus on gastroparesis

Gastroparesis is a condition characterized by epigastric symptoms and delayed gastric emptying (GE) rate in the absence of any mechanical obstruction. The condition is challenging in clinical practice by the lack of guidance concerning diagnosis and management of gastroparesis.


INTRODUC TI ON
Gastroparesis is a condition characterized by epigastric symptoms (nausea, vomiting, postprandial fullness, early satiation, and epigastric pain) and significantly delayed gastric emptying (GE) rate in the absence of any mechanical obstruction. 1,2 Gastroparesis is a complication of diabetes, especially type 1 diabetes, and may also occur following upper gastrointestinal tract surgery. Nevertheless, in the largest subgroup no underlying cause is identified and these patients are referred to as having idiopathic gastroparesis. 1,2 The epidemiology of gastroparesis is unknown, as it requires procedures such as GE tests to make a firm diagnosis. In clinical gastroenterology practice, gastroparesis is frequently encountered and considered one of the more challenging conditions, as there are uncertainties in terms of definition, symptom spectrum, diagnosis and optimal therapeutic approach, especially as there is a paucity of interventions with established efficacy. [1][2][3][4] The aim of this project was to develop a European consensus on the definition, clinical characteristics, pathophysiological concepts, diagnosis and management of gastroparesis, with a focus on idiopathic gastroparesis. The results of this consensus can offer the clinician guidance in diagnosing and managing these patients, with the aim to optimizing outcomes.

ME THODS
The European Society for Neurogastroenterology and Motility (ESNM) initiated a Delphi process, to develop consensus statements on different aspects of functional dyspepsia (FD) and gastroparesis in collaboration with other European societies. The Delphi approach, which combines the principles of evidence-based medicine, supported by systematic literature reviews and a voting process, aims to determine consensus for complex problems in medicine for which evidence from controlled trials is lacking. 5 The principal steps in the process were: (1) (Table 1).
gave feedback on clarity of the statement and made suggestions for adapting or splitting the statements into two or more questions, or for adding additional statements on a given topic. The Core Group adjusted the statement list, generating a total of 91 statements, and subdivided the Guideline Group members into 12 Working Groups with 3-4 members each. Each Working Group was allocated statements for which they needed to conduct a systematic literature search using several relevant keywords and provide narrative substantiation of the statements. The literature review and references were made available on a share-point server, accessible to all members. This was finalized by the summer of 2019, followed by a voting round in which each statement was presented with the evidence summary, and each member indicated the degree of agreement for the statement using a 6-point Likert scale (Table 1) Consensus was defined as when at least 80% of the Consensus Group agreed (A+ or A) with a statement. The strength of evidence for each statement was scored using the GRADE system (Table 2). 7 After the final voting round, the manuscript was drafted and circu-

Definitions and symptom descriptors
1.1 Gastroparesis refers to a symptom or set of symptoms that is (are) associated with delayed GE in the absence of mechanical obstruction.

Gastroparesis refers to a symptom or set of symptoms that is
(are) associated with severely disturbed gastric motor function in the absence of mechanical obstruction.
STATEMENT ENDORSED, overall agreement 85%: A+ 48%, postprandial fullness, abdominal pain/discomfort, and anorexia. 4,[9][10][11][12] A systematic review and meta-analysis, which included only studies with a high-quality emptying test showed significant associations between the presence of delayed GE and symptoms of nausea, vomiting, early satiety, postprandial fullness and epigastric pain. 9 Other studies also confirmed that gastroparesis symptoms have considerable overlap with FD defined by the Rome IV criteria. 4,10,11 A problem of this definition is that the association between delayed GE and symptoms may suggest that delayed GE is directly causing the symptoms. However, this conclusion should be considered with caution. 4 Symptoms occurring in subjects with delayed GE may also be associated with other types of gastric sensorimotor dysfunction, such as impaired gastric accommodation, hypersensitivity to gastric distention and uncoordinated intense motor activity in proximal small bowel. [12][13][14] Gastroparesis is a chronic condition and symptoms have to be present for some time before a diagnosis is triggered, and a timeframe of at least 3 months was proposed. 15 Nausea is the most common symptom in gastroparesis, affecting more than 95%, and has been associated with the severity of delay in emptying. 4,9,10,16 Vomiting is associated with nausea and with more severe delay in GE. 8,9,17  pain in a large subset of gastroparesis patients, although some of these studies included patients taking opioids, which is a major confounder. 21,22 In studies that exclude patients on opioids, pain is less prevalent, and it is also inconsistently correlated with GE delay. 4,[8][9][10][11][12][13]16,17,20 Bloating is a common symptom of functional digestive disorders including FD and IBS, but its prevalence is high in gastroparesis patients. 9 2013, with a decrease in average length of stay but an increase in the cost of each hospitalization with time. 35 Over the same time period, emergency department visits for gastroparesis more than doubled, and this was also associated with substantial healthcare costs. 36 In a population-based study in Olmsted county, overall survival of subjects diagnosed with gastroparesis was significantly lower than the age-and sex-specific expected survival computed for the reference population, and a higher mortality was seen in diabetic versus idiopathic gastroparesis. 23 In the UK CPRD database study, mortality rates were also higher in diabetic compared to idiopathic gastroparesis patients. 24 However, in these studies the excess mortality was largely driven by cardiovascular comorbidity.
No data are available on self-cost to gastroparesis patients, but they may incur both direct and indirect costs for use of OTC medications (antiemetics, antacids, and H2 receptor antagonists), alternative therapies, medical consultations, and cofinanced treatments, as well as the cost of some dietary adjustments and nutritional support, which is often not reimbursed.
In a small US gastroparesis patient cohort, the majority of patients surveyed (54%) was not employed and considered disabled. 37 Employment rates are even lower in those on opioids. 38 In a survey of almost 500 gastroparesis patients, the symptoms reduced daily activities in 67.5% and lowered annual income in 28.5%; 11% were disabled due to gastroparesis symptoms. 39 Quality of life was significantly decreased in this cohort. Also, in a large online survey of over 1400 self-reported gastroparesis patients, quality of life as measured by the 36-item short-form health survey was decreased, mainly on the physical health component, and this was negatively correlated with increased symptom severity, especially nausea, early satiety, and abdominal pain. 40 In a study of 299 gastroparesis patients enrolled from US referral centers, anxiety and depression scores correlated with the severity of gastroparesis symptoms but did not differ between diabetic and idiopathic gastroparesis and did not correlate with the delay in GE rate. 41 In a systematic review, based on three studies in gastroparesis (n = 378) combined anxiety/depression was present in 24% of patients, severe anxiety in 12.4%, depression in 23%, and somatization in 50%. 42 Although it is often assumed that gastroparesis may lead to weight loss, this is not substantiated in the literature. In large cohorts of patients with dyspeptic symptoms, delayed GE was not associated with weight loss. 8,43 Gastroparesis may affect adolescent and young patients. When accompanied by weight loss, an eating disorder should be excluded, especially since delayed GE is a feature of anorexia nervosa. 44,45 Healthcare consulting in gastroparesis is likely driven by symptom severity and impact, but no data are available from the literature. Moreover, it is not known whether healthcare consulting is driven by psychosocial comorbidity, although this seems plausible given its correlation with symptom severity. 41 A study in the United States showed regional differences in healthcare assistance in gastroparesis patients, with admissions rates independently predicted by high overall hospitalizations within a state, suggesting that access to healthcare is a factor determining healthcare consulting. 46 4. Pathophysiology of gastroparesis STATEMENT NOT ENDORSED, overall agreement 78%: A+ 23%, A 55%, A− 15%, D− 5%, D 3%, D+ 0%. GRADE C 4.11 Altered release of peptide hormones is not a pathophysiological mechanism in gastroparesis.
STATEMENT NOT ENDORSED, overall agreement 43%: A+ high-quality emptying test showed significant associations between the presence of delayed GE and symptoms of nausea, vomiting, early satiety, postprandial fullness, and epigastric pain. 9 However, the relationship between symptom severity and emptying delay remains poor.
In idiopathic gastroparesis, barostat studies demonstrated that 43% of patients had impaired accommodation, associated with a higher prevalence of weight loss and early satiety, 29% of patients had hypersensitivity to gastric distension, associated with higher rates of early satiety, epigastric pain, and weight loss. 12 In diabetic gastroparesis, gastric accommodation to a meal and sensory thresholds for discomfort were lower compared to healthy controls. 13 Impaired duodenal mucosal integrity, with mucosal infiltration with eosinophils and mast cells, has been reported as a putative pathophysiological mechanism in FD but studies in gastroparesis seem to be lacking. 47 Histopathological studies in an animal model of type 1 diabetes and transmural biopsy specimens from gastroparesis patients showed decreased density of interstitial cells of Cajal, possibly due to switch from anti-inflammatory M2 macrophages to pro-inflammatory M1 macrophages. 48,49 Conflicting data exist on a possible decrease of electrically coupled platelet-derived growth factor receptor α-fibroblast-like cells in gastroparesis. 50,51 In small patient cohort studies, fibrotic and inflammatory changes to gastric smooth muscle were reported in diabetic gastroparesis patients, but this was not confirmed in the larger NIDDK consortium cohort. 48,49,52 In a study which evaluated heart-rate variability responses in 41 gastroparesis patients, diabetic gastroparesis was more frequently associated with signs of vagal dysfunction than idiopathic gastroparesis. 53 In a large study of 242 patients, there were signs of both sympathetic and parasympathetic dysfunction, the latter associated with more severe symptoms and more delayed GE. 54 Few studies have evaluated the relationship between H. pylori status and GE rate, but no consistent correlation was found. [55][56][57] No studies evaluated gastric acid secretion, duodenal sensitivity to acid or lipids or duodenal microbiota composition in gastroparesis. There is also no consistent proof of altered release of gut peptides contributing to gastroparesis pathophysiology. 58 Several authors have reported higher degrees of anxiety and depression in gastroparesis patients. In those studies, anxiety and depression scores were related to gastroparesis symptom severity and hospitalization rates but not to disease etiology and GE rates. 16,17,21,31 Both in idiopathic and in diabetic gastroparesis, hypersensitivity to gastric distension is present and, in the former group, is associated with the symptom pattern and severity (higher rates of early satiety, epigastric pain, and weight loss). 12,13 As gastric compliance is not altered in these patients, the pathophysiology is likely to involve altered processing of incoming signals in the central nervous system. One study using evoked potentials showed altered processing of esophageal electrical stimulation signals in the brain in patients with diabetic neuropathy, and this was related to upper gastrointestinal symptom severity and quality of life impact. 59 An fMRI study showed altered connectivity of the insula and a tendency towards decreased insula gray matter volume in gastroparesis patients. 60 Only a few studies have evaluated genetic factors that predispose for the development of gastroparesis. A long repeat polymorphism in the heme oxygenase (HO1) has been associated with worse outcomes in several diseases, including gastroparesis, and the pres- Food retention in the stomach as seen during EGD after an overnight fast has been used as a probable sign of gastroparesis in epidemiological studies, even in the presence of a normal GE test. 23 Only one retrospective analysis study specifically addressed the relationship between GE and food retention, suggesting that gastric food retention at endoscopy could be a moderately specific sign for gastroparesis, but with poor sensitivity. 62 The definition of gastroparesis implies objective delayed GE in the absence of mechanical obstruction. To date, GE scintigraphy of a solidphase meal is considered the gold standard, and by consensus, the test is performed using a 99m Technetium-labeled standardized low-fat, egg meal with imaging at 0, 1, 2, and 4 h after meal ingestion. 63  The wireless motility capsule (WMC) is an FDA-approved device for the evaluation of GE. A systematic review from 2013 that included seven studies found that for the diagnosis of gastroparesis, as compared with gastric scintigraphy, WMC had a sensitivity of 59%-86% and specificity of 64%-81%. 66 The pitfall with WMC is that it is an indigestible solid and, therefore, it empties from the stomach in response to phase 3 migrating motor complexes rather than with the test meal. A further disadvantage is that WMC is relatively expensive and its availability across different centers is limited.
Gastric ultrasonography has been used to assess antral wall motion, patterns of transpyloric flow, and GE based on changes in the cross-sectional area or diameter of the gastric antrum. Gastric ultrasonography is noninvasive, safe, cheap, widely available, allows for bedside monitoring, does not expose the patient to ionizing radiation which is restricted particularly in children and pregnant women, and shows reasonably good interobserver agreement in the evaluation of liquid GE. 66 However, ultrasonography is unable to distinguish between the solid and liquid components of a meal and therefore is unsuitable to assess emptying of solids. Ultrasonography also requires an experienced technician, is user dependent, may be influenced by the presence of intragastric air or posture and is generally considered impractical for prolonged observations. 67 When small bowel obstruction is suspected, patients generally undergo abdominal radiography which is widely available, inexpensive and has a reported accuracy of 50%-86%. Where small bowel obstruction is strongly suspected, CT scanning is the most accurate examination. 68 6. Treatment  A diet composed of small particle size reduces gastroparesis and reflux symptoms in patients with diabetic gastroparesis. 69 In addition, gastroparesis carries an increased risk for a diet deficient in calories, which can be corrected by nutritional support and engagement with dieticians. 17 The efficacy of PPIs in gastroparesis has not been addressed in controlled trials. However, PPI intake is high in gastroparesis patients (70% or more) and more than 50% have an overlapping GERD diagnosis. 17,70 It has been suggested that the coexisting GERD rather than gastroparesis is the explanation for the PPI intake.
While they are recommended as first line symptomatic treatment, there are no formal studies establishing efficacy of traditional antiemetic agents in the treatment of nausea and vomiting associated with gastroparesis. 71 In the United States, metoclopramide, a dopamine-2 receptor antagonist, is approved for the treatment of gastroparesis. However, it carries a black box warning, as it is generally not well-tolerated and chronic use (>12 weeks) may lead to extrapyramidal side effects and potential irreversible tardive dyskinesia, which has been reported in a small percentage of cases. 71 In a phase 2, randomized controlled trial (RCT) on 285 diabetic patients with gastroparesis, 10 or 14 mg metoclopramide doses, administrated via nasal spray could not reduce the overall symptom scores significantly more than placebo. 72 However, when males and females were analyzed separately, a significant effect was observed in females. Reported adverse events were mainly headache, fatigue, and dysgeusia. Domperidone is a peripherally acting dopamine-2 antagonist that decreases nausea and increases GE rates. It does not readily cross the blood-brain barrier, making it much less likely to cause extrapyramidal side effects. However, domperidone is associated with prolongation of the cardiac QTc interval which has restricted its use. 73 The 5-HT 3 antagonists, such as ondansetron or granisetron, also have antiemetic properties, but controlled studies in gastroparesis are lacking. An open label study suggested efficacy of granisetron transdermal patches for controlling nausea and vomiting in gastroparesis. 74 Aprepitant, a neurokinin-1 receptor antagonist approved for use for the treatment of chemotherapy-induced emesis, was efficacious in the treatment of nausea in some patients with gastroparesis and related disorders. 75 Clinical trials of prokinetic agents report that, as a group, they enhance symptoms and improve GE rate. 76 However, there is disparity, as the relationship between improvement of symptoms and enhancement of GE rate is poor but becomes significant when only high-quality emptying studies are considered. 76,77 The availability of prokinetics is limited in many parts of the world, but new prokinetic agents are in the pipeline for the treatment of gastroparesis. 78 Nevertheless, GE rate enhancement per se cannot be used as a marker of clinical efficacy. 79 Based on these considerations, the panel did not support efficacy of prokinetics as a treatment group.
The panel did support the classes of dopamine-2 antagonists and 5-HT 4 agonists as having efficacy for treating gastroparesis symptoms.
Itopride, available in Asia and to some extent in Eastern Europe, is a combined dopamine-2 antagonist and cholinesterase inhibitor, which has been extensively studied in FD. A small controlled crossover study evaluating the effect of itopride 200 mg t. i.d. in 25 diabetic gastroparesis patients found no significant effect. 80 Cisapride was for a long time the preferred medication for outpatient treatment of gastroparesis but has been withdrawn from the market because of cardiologic side effects. 81 Tegaserod, another 5-HT 4 agonist, showed ability to enhance GE but was temporarily withdrawn from the market and only developed for bowel disorders. 82 Prucalopride, a 5-HT 4 agonist belonging to another chemical class, is approved in most countries for the treatment of chronic constipation. In a recent study of mainly idiopathic gastroparesis patients, prucalopride was efficacious in improving GE rate and symptoms, while a second study in mainly diabetic gastroparesis patients failed to show benefit. 83,84 Several motilin-receptor agonists have been studied as poten- A phase 3 program in diabetic gastroparesis was in progress but has recently been discontinued.
Neuromodulators are often used for managing refractory symptoms in functional and motility disorders. 90  Open-label studies all reported short term (<6 months) efficacy of intrapyloric injection of botulinum toxin both on symptoms and GE in gastroparesis. 101 However, two subsequent controlled trials failed to demonstrate an improvement in symptoms and GE. 102,103 Several open-label studies reported short and midterm (<18 months) efficacy of endoscopic pyloric myotomy on symptoms, quality of life, and GE in gastroparesis. 104  In terms of the effect of gastroparesis on life expectancy, data are conflicting. Based on 86 diabetic patients who were followed for at least 9 years, delayed GE was not related with mortality after adjustment for comorbidities, 125 but 6-year follow-up data from a tertiary care setting observed that 7% had died and 22% needed long-term parenteral or enteral feeding, suggesting gastroparesis is not a benign condition. 124 Studies conducted in referral populations demonstrate no effect of delayed GE on mortality among patients with diabetes mellitus after 12-25 years of follow-up. 126,127 In the analysis of the UK CPRD database, mortality risk was higher in diabetic compared to idiopathic gastroparesis patients (adjusted hazard ratio: 1.9, 95% CI: 1.2-3.0). 24 A population-based study compared the observed and expected mortality of patients with gastroparesis, demonstrating a significantly higher death rate in patients compared to age-and sex-matched controls, which was largely due to cardiovascular comorbidity in diabetic patients. 23

RECOMMENDATIONS
Based on the statements that achieved consensus, a number of recommendations for understanding and managing of gastroparesis can be made (Table 3), which are summarized in Table 4 and Figure 1. The Delphi process also identified several areas of uncertainty, which require additional evidence or further research.
In line with existing definitions and guidelines, gastroparesis is based on the presence of upper gastrointestinal symptoms and delayed GE in the absence of an obstructive lesion. [1][2][3][4] The panel also agrees that gastroparesis is associated with other manifestations of severely disturbed gastric motor function. [12][13][14] The European consensus identifies nausea and vomiting as the cardinal symptoms of gastroparesis. While the panel also agrees that there are often coexisting FD/PDS symptoms, the presence of predominant nausea and vomiting offers an approach to differentiate (idiopathic) gastroparesis from PDS, where early satiation or postprandial fullness are dominant symptoms. 2,4,18,19 The overlap with EPS is considered less prevalent. 4,17,20 There is consensus that the true epidemiology of gastroparesis is not known, but that diabetes, gastric surgery, certain neurological and connective tissue diseases, as well as the use of certain drugs are associated with an increased risk of gastroparesis. [23][24][25][29][30][31][32][33][34] The panel agreed that gastroparesis has an important impact on quality of life and healthcare costs, and also concurs that gastroparesis is often associated with psychological comorbidities such as anxiety and depression. 35,[40][41][42] There is no consensus that gastroparesis may lead to unintended weight loss, but when present, eating disorders must be excluded. 9,[43][44][45] In terms of pathophysiological mechanisms that are relevant to gastroparesis, while the panel agreed that severely impaired gastric motor function is present in these patients, there was no consensus for a direct role for delayed emptying, impaired gastric accommodation or gastric hypersensitivity in determining symptom pattern and severity. [1][2][3][4][8][9][10][11][12][13]16,17,20,21,32 There was also no qualifying majority support for pivotal pathophysiological roles for loss of interstitial cells of Cajal, intrinsic or extrinsic (vagus) nerves, changes in smooth muscle, peptide hormone release, gastric acid secretion, duodenal mucosal changes, psychosocial comorbidity, or altered central processing in the generation of gastroparesis symptoms.  There is consensus that an abnormal GE test as well as an EGD, to rule out mechanical obstruction, are mandatory for establishing a diagnosis of gastroparesis, but that the presence of food at endoscopy is not a reliable diagnostic marker. 1,2,19,23,62 Radiological examinations, preferably using a CT scan, can be added in case of uncertainty regarding the absence of a mechanical obstructive factor. Scintigraphy and breath tests are agreed to be reliable diagnostic tests, but there is no support for the WMC or gastric ultrasound to detect delayed GE. [63][64][65][66][67] Besides the statements on pathophysiology, the section on treatment approaches is a second one to display a major lack of consensus. In spite of very few literature data, 69 the panel supported dietary intervention in the treatment of gastroparesis. There is no consensus on the efficacy of PPIs, nor for different types of antiemetics. [70][71][72][73][74][75] There is borderline support (78% agreement) for the use of prokinetics as a group, but the panel agreed on the use of 5-HT 4 receptor agonists as a class. [76][77][78][79][80][81][82][83][84][85][86][87][88][89] There is also no consensus on the use of other neuromodulators, herbal therapies, acupuncture or psychological therapies in gastroparesis. [90][91][92][93][116][117][118] The same is true for invasive therapies such as botulinum toxin injection, GES,  Gastroparesis is a major source of healthcare costs 3.1,

E S N M G A S TRO PA R E S I S CO N S EN S US G RO U P
Gastroparesis is associated with a significant decrease in quality of life and with psychosocial co-morbidities 3.5,3.6 In case of weight loss, an eating disorder must be ruled out 3.8 Upper GI endoscopy to rule out mechanical obstruction and an abnormal gastric emptying test are mandatory for establishing a diagnosis of gastroparesis. Gastric emptying can reliably be assessed by scintigraphy or breath test. Additional radiological imaging can be used to exclude obstruction in case of doubt