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

Functional dyspepsia (FD) is one of the most common conditions in clinical practice. In spite of its prevalence, FD is associated with major uncertainties in terms of its definition, underlying pathophysiology, diagnosis, treatment, and prognosis.


INTRODUC TI ON
Functional dyspepsia (FD), defined by the presence of recurrent or chronic epigastric symptoms in the absence of organic disease likely to explain them, is one of the most common conditions seen in clinical practice. 1,2 In spite of its prevalence, FD is associated with major uncertainties, as definitions and the symptom spectrum of FD have evolved over time, 3 the differential diagnosis is very broad, 1 the optimal diagnostic work-up has not been defined, 4,5 and there is a lack of available treatments with established efficacy. 6,7 The aim of this project was to develop a European consensus on the definition, pathophysiological concepts, diagnosis, management, and prognosis of FD. The results of this consensus can offer the clinician guidance in diagnosing and managing FD patients to optimize clinical outcomes.

ME THODS
The European Society for Neurogastroenterology and Motility (ESNM) initiated a Delphi process, funded by United European Gastroenterology, to develop consensus statements on different aspects of FD 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 at determining consensus for complex problems in medicine for which evidence from controlled trials is lacking. 8 The principal steps in the process were (1) selection of a working group of seven ESNM members with expertise in FD and/or Delphi consensus processes; (2) selection of a European Consensus Group consisting of experts in FD from different European countries, recruited through the ESNM board and through UEG Sister Societies; (3) drafting of statements allowing to evaluate the current knowledge on FD; (4) systematic literature reviews to identify evidence to support each statement; (5) (Table 1). Participants were blinded to the votes of other participants and also gave feedback on the 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 87 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 sev- Throughout the process, all votes were mutually anonymous and blinded.
When 80% of the Consensus Group agreed (A+ or A) with a statement, this was defined as consensus. The strength of evidence for each statement was scored using the GRADE system (Table 2). 9 After the final voting round (summarized in Table 3), the manuscript was drafted and circulated for the final approval by the participants.
The references cited in this chapter are only a selection of the articles reviewed in each area, chosen to clarify the discussion. A final meeting planned in October 2020 was canceled because of the COVID−19 pandemic. • One high-quality study

RE SULTS
• Several studies with some limitations C Low Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
• One or more studies with severe limitations D Very low Any estimate of effect is very uncertain.
• Expert opinion The Rome IV criteria stress that heartburn is not a dyspeptic symptom but may often coexist with FD and that the presence of heartburn should not lead to the exclusion of FD as diagnosis. 1,3 Overall, one-third of FD patients also experience typical symptoms of gastroesophageal reflux disease (GERD). 13 It seems that, in cases of overlap between FD and GERD, FD is often underestimated, favoring the diagnosis of GERD. 14  STATEMENT NOT ENDORSED, overall agreement 37%: A+ 7%, A 30%, A− 46%, D− 2%, D 15%, D+ 0%. GRADE C Approximately 10% of the adult population fulfills symptombased Rome IV criteria for (uninvestigated) FD, and its prevalence appears to disappear with increasing age. 2,19 In several studies, the peak incidence of FD seems to occur in the forties or fifties age segment. [19][20][21][22] A recent metA−analysis including 55 studies revealed a slightly higher pooled prevalence of dyspepsia in women compared with men. 19 The Rome Global Epidemiology Study, which used the most uniform criteria and approach, showed a significantly higher prevalence of (uninvestigated) Rome IV FD in women compared to men. 2 Acute gastroenteritis is associated with an increased risk of FD, with an estimated mean prevalence of 9.6% in adults. 23 Among pathogens suggested to be associated with post-infectious FD (PI-FD) are Norovirus, Giardia lamblia, Salmonella spp., Escherichia coli O157, and Campylobacter spp. [24][25][26] H. pylori does not seem to be a cause of PI-FD. 23,27 Nonsteroidal anti-inflammatory drug (NSAID) use has been identified as a risk factor for dyspepsia in two population-based studies. 28,29 It has been suggested that the development of dyspeptic symptoms during treatment with NSAIDs could be linked to alterations in gastric mechanosensory function. 30 However, NSAID intake appears to be most relevant to uninvestigated dyspepsia. 31  In addition, dyspeptic patients with high scores on anxiety and depression questionnaires report higher consultation rates than those with lower scores, indicating that psychosocial factors also influence healthcare-seeking behavior. 36,71 A longitudinal study also found that anxiety or depression precedes dyspepsia among consulters in a larger proportion than in non-consulters, confirming that basal psychosocial comorbidity is associated with healthcare seeking. 71 Consultation rates for dyspepsia vary widely between countries and regions, suggesting that access to the healthcare system may influence healthcare-seeking behavior. 70 However, differences in healthcare-consulting behavior are also modulated by socioeconomic status. Several studies have shown that low socioeconomic status is associated with higher consultation rates for dyspepsia. 68 Several studies reported delayed gastric emptying in FD for solids as well as liquids 93-10 0 without differences in the emptying times between Rome III subgroups. 91 A meta-analysis described gastric emptying in patients with FD to be 1.46 times slower than in controls. 96 Delayed gastric emptying was associated with female sex, postprandial fullness, nausea, vomiting, bloating, and early satiety. 97-10 0 In spite of the presence of delayed emptying in a subset of patients, the association with symptoms is weak. The severity of gastric emptying delay is not a good determinant of symptom severity or pattern. 94,97,99,100 Few studies also reported an acceleration of gastric emptying in FD, 101,102 but this was not confirmed in others. 99 Hypersensitivity to gastric distention has been reported in 34%-65% of FD patients by several studies, 64,100-106 without difference between Rome III subgroups, 91 and postprandial sensitivity to gastric distention was even greater than fasting sensitivity. 107 Hypersensitivity to gastric distention was associated with a higher prevalence of postprandial pain, belching, and weight loss, 64  insult in FD. 116,117 Although gastric acid secretion is reported as normal, 118 FD patients displayed increased spontaneous duodenal acid exposure during the daytime and the late postprandial phase with higher symptom severity in patients with high duodenal acid exposure. 119 However, the correlation between acid exposure and symptom severity was weak, and the increased duodenal acid exposure could be, at least in part, attributable to delayed duodenal acid clearance as The vagus nerve is also a major contributor to control upper gastrointestinal motility. 133 An early study in seven FD patients, using an insulin hypoglycemia test and plasma levels of pancreatic polypeptide, suggested a disturbed efferent vagal function. 134 The gastric response to sham feeding, a marker for vagal activity, was lower in PDS compared to controls. 135 Conversely, sham feeding was reported to improve the suppressed response to a liquid nutrient meal in FD. 136 Slow deep breathing, which is thought to activate the vagus nerve, was associated with improvement of nutrient volume tolerance and quality of life in FD. 137 Using spectral analysis of cardiac R-R intervals to evaluate vagal tone, Guo et al. showed a decreased vagal tone that was associated with delayed gastric emptying. 138 Taken together, a number of observations suggest decreased vagal activity in FD, but the studies all occurred in laboratory settings in small groups of patients.
As mentioned above, there is an association of anxiety with FD, and anxiety may precede FD. Moreover, FD is associated with altered brain processing of gastrointestinal (GI) stimuli, altered central nervous system connectivity, and structure and altered expression of neurotransmitter pathways. [139][140][141][142][143] However, a causal relation between anxiety and FD has not been established. As mentioned above, depression is also associated with FD, but also in this case, there is a lack of evidence for a causal relation.
Numerous studies have shown that FD patients report more symptoms or earlier symptoms following gastroduodenal stimulation with either balloon, liquid volume, or food compared to healthy controls. A systematic review evaluated studies on central processing of signals from the gastroduodenal region in FD patients and controls, mostly by balloon distention of the stomach, using PET or fMRI technology. 144 The results show that FD is associated with functional abnormalities in sensory and pain modulation, emotion, saliency, and homeostatic processing regions, suggesting that disordered central processing of incoming signals from the gastroduodenal region is indeed a relevant pathophysiological mechanism at least in a subgroup of FD patients. However, this does not exclude an involvement of peripheral mechanisms as well.
Family studies support a genetic component in FD susceptibility. 145 A mea-analysis of eight studies indicated that the GNβ3 C825T polymorphism is significantly associated with FD, and susceptible to racial variation. 146 However, a meta-analysis based on 12 studies has failed to confirm a significant association. 147  In population-based symptom analyses, dyspeptic symptoms were shown to group around clusters, representing EPS and PDS. 3,156 The literature is divided on the usefulness of distinguishing PDS and EPS for patient management. When the Rome III subdivision is used, a major overlap is found between both, which is largely corrected with the Rome IV subdivision as a good separation between both subtypes is now found both in epidemiological studies and in clinical practice. 3,11,40,[156][157][158] While some studies report different treatment responses, 159,160 others do not. 161,162 To date, no fully published study has evaluated differential pathophysiological mechanisms or treatment outcomes according to the Rome IV subdivision. 163,164 Kraag et al. performed a meta-analysis of 21 controlled studies on the association between gallstones and dyspeptic symptoms and found no reasonable association between gallstones and "classical" dyspeptic symptoms other than upper abdominal pain. 165 In a systematic review of 24 publications, biliary colic was the only single symptom associated with gallstones. 166 Hence, guidelines do not recommend upper abdominal ultrasound for exclusion of biliary pathology in the diagnosis of FD. 1,150 The prevalence of delayed gastric emptying in FD ranges between 20% and 50%, but its association with symptoms and response to therapy has shown inconsistent results. In a systematic review and meta-analysis of 37 studies, both mast cell and eosinophil counts in the duodenum were increased in FD compared to controls, both in PDS and EPS. 172 Nevertheless, as there is overlap with health and as other conditions may also be associated with increased eosinophil numbers, this cannot be used as a diagnostic marker.
Several studies have reported decreased volume tolerance in FD compared to health, using liquid nutrients but also water intake. 63  cisapride, which has been withdrawn from the market because of cardiac adverse events. 188 When cisapride was removed from the meta-analysis, the overall effect was still significant, but the NNT increased to 12. 187 The rationale for prokinetic therapy in FD is the presence of motor abnormalities such as delayed gastric emptying, especially in PDS, but in a large study, similar prevalence of gastric motor abnormalities was found in PDS, EPS, and the overlap group. 91 In the 2019 meta-analysis, prokinetics demonstrated similar efficacy in PDS and EPS subgroups, although only two studies included a total of 124 EPS patients, 187 one of which is a study on cisapride from 1989. 189 The evidence supporting the use of prokinetics in FD is rather poor and no target subgroup can be defined based on the available evidence.
Moreover, many prokinetics such as domperidone and acotiamide, are not widely available. Adverse effects, such as extrapyramidal syndrome for many dopamine-2 (D2) receptor antagonists and QTc prolongation with domperidone, are limiting chronic use in FD. This has led the ACG/CAG guideline to recommend treatment with a TCA in patients refractory to PPI treatment before prokinetics. 150 A systematic analysis of 34 studies failed to demonstrate a correlation between the acceleration of gastric emptying and symptom improvement. 167 One possible explanation is the heterogeneous pathophysiology of FD, involving not only delayed gastric emptying but also impaired gastric accommodation and hypersensitivity to gastric distention which are often not taken into account. However, in a recent meta-analysis by Vijayvargiya et al., the authors found that when optimal test methods were used, a selection of promotility agents significantly accelerated gastric emptying and produced significant symptom improvement in gastroparesis patients. 168 Nevertheless, it has not been established that the effect of prokinetics in FD depends on baseline emptying rate or improvement of emptying rate with therapy.
Itopride is a combined D2 antagonist and acetylcholinesterase inhibitor and is available in Asia and several countries in Eastern Europe.
A phase IIb placebo-controlled trial found significantly more responders to itopride, based on a global efficacy measure. 190 However, no significant improvement over placebo in reduction of FD symptoms was observed in two subsequent Phase III trials. 191 These trials suffered from issues with patients and endpoint selection, 192  Psychotropic drugs appear to be an effective treatment for FD, as demonstrated by a systematic review and meta-analysis, with an NNT of 6 when data from all studies were pooled (1241 patients, 673 assigned to psychoactive drugs, and 568 to placebo). However, this beneficial effect appeared to be limited to TCAs and antipsychotics. 194 In a randomized placebo-controlled trial including 292 FD patients assigned to either placebo, 50 mg amitriptyline or 10 mg escitalopram for 12 weeks, subjects with "ulcer-like" FD (likely equivalent to EPS) receiving amitriptyline reported more adequate relief of symptoms than those receiving placebo or escitalopram (p = 0.06). 195 There were adverse events in 30% (n = 29) individuals in the amitriptyline arm, leading to discontinuation of treatment in two of them. Those with delayed gastric emptying were less likely to report adequate relief on amitriptyline compared with FD patients with normal emptying, but this was not related to an amitriptylineinduced delay in gastric emptying. 194 Amitriptyline appeared to derive its benefit predominantly through improving abdominal pain, since no change in psychological distress measures nor gastric emptying rates was found. 195,196 In the subset of patients with PDS, little evidence exists so far supporting the use of TCA's. A double-blind, randomized controlled trial including 107 patients with refractory FD, treated with either imipramine or placebo for 12 weeks, showed efficacy in symptom relief (p = 0.0051), but 18% of the patients on imipramine discontinued the study due to adverse effects. In this study, no conclusion was made regarding efficacy in FD subtypes. 197 The ACG and CAG clinical guidelines on dyspepsia considered that ing almost 400 patients, which were negative. 194,195,198 Thus, it seems reasonable to assume that these drugs are of no benefit in FD.
A double-blind clinical trial randomly assigned 160 FD patients to 8 weeks of treatment with venlafaxine or placebo. 199 At none of the measurement times there was a statistically significant difference in symptom severity, quality of life or anxiety, and depression scores between venlafaxine and placebo. 199 The dropout rate among venlafaxine-treated patients was high due to side effects.
While this single study with venlafaxine in FD was negative, it remains to be elucidated whether certain groups of patients might benefit from treatment with serotonin/noradrenaline reuptake inhibitors with a more potent analgesic effect at lower doses, for example, duloxetine.
In FD, weight loss is normally considered an alarming symptom, but may be present in up to 40% of tertiary care FD patients. 10,65,66 A controlled trial to assess mirtazapine's efficacy in FD and weight loss randomly assigned 34 patients to placebo or mirtazapine 15 mg daily for 8 weeks. 200 Mirtazapine significantly improved early satiation scores and nutrient tolerance compared to placebo. A trend was found for overall dyspepsia symptom score at week 4 in the mir- A single double-blind, placebo-controlled randomized study examined the efficacy of rifaximin in subjects with Rome III criteria defined FD who were H. pylori negative. 211 The authors found that rifaximin was superior to placebo for the relief of global dyspeptic symptoms, postprandial fullness/bloating, and belching. Additional future trials are needed to examine the efficacy of rifaximin in FD and to elucidate the underlying mechanism of action.
In a 2017 review, a total of 12 controlled trials of psychological therapies involving 1563 FD patients were identified. 150 All trials reported a statistically significant benefit of psychological therapies over control, which was most commonly usual management.

Based on statement(s)
Dyspepsia refers to a symptom or set of symptoms that are considered to originate from the gastroduodenal region. Early satiation, postprandial fullness, epigastric pain and epigastric burning are the cardinal dyspeptic symptoms.
FD is a condition characterized by chronic dyspeptic symptoms in the absence of organic, systemic or metabolic condition(s) that is (are) likely to explain symptoms. The vast majority of patients with dyspeptic symptoms and no alarm symptoms in the general population would be identified as FD after investigation (if performed).

1.3, 1.4
Two main subtypes of FD are distinguished which may overlap: postprandial distress syndrome (PDS) characterized by mealinduced symptoms (early satiation, postprandial fullness) and epigastric pain syndrome (EPS), with epigastric pain and/or epigastric burning.

1.5
Dyspeptic symptoms often co-exist with other symptoms such as bloating in the upper abdomen, nausea and belching. Typical reflux symptoms and irritable bowel syndrome often coexists with FD. In case of severe weight loss in FD, nutritional support may be needed.

6.27
The long-term prognosis is favorable in the majority of patients with FD, whose life expectancy is similar to that of the general population.

7.1, 7.2
Information on individual types of psychological therapies is variable. For hypnotherapy in FD, only one small randomized controlled study reported benefit. 212 For CBT, two studies showed positive short-term effects on FD symptoms. 213,214 Meta-analyses of numerous low-quality randomized, controlled studies suggest manual and electric acupuncture being effective in the treatment of FD, as shown by improved symptom scores and health-related quality-of-life scores. [215][216][217] Effects are most pronounced in sham-controlled trials and less pronounced in trials comparing to prokinetic medication or traditional Chinese medicine. A recent sham-controlled trial adds that the effect of acupuncture, following 20 treatment sessions in a 4-week episode, is sustained for 24 weeks. 218 Besides the overall small number of patients included and the different acupuncture protocols followed, selection bias, performing bias, reporting bias, attrition bias, and blinding difficulties remain the major concerns when interpreting findings in the meta-analyses.
In a meta-analysis, the quality and effectiveness of mindfulnessbased therapy in FGIDs was evaluated. 219

RECOMMENDATIONS
Based on the statements that achieved consensus (Table 3), a number of recommendations for understanding and managing FD can be made, which are summarized in Table 4. The Delphi process also identified several areas of uncertainty, which require additional evidence or further research. In terms of pathophysiological mechanisms that are relevant to FD, consensus supports a role for impaired gastric accommodation, delayed gastric emptying, hypersensitivity to gastric distention, H.
pylori infection, and altered central processing of incoming signals from the gastroduodenal region. There is no consensus on duodenal mucosal alterations, sensitivity to luminal contents, peptide release, or microbiota. Anxiety is a risk factor for the development of functional dyspepsia; however, anxiety, depression, or stress is not considered a pathophysiological mechanism that underlies FD symptom generation.
There is consensus that endoscopy is mandatory for establishing a firm diagnosis of FD, but that patients in primary care with dyspeptic symptoms and no alarm symptoms or risk factors can be managed without endoscopy. There is consensus that endoscopy is mandatory in case of alarm symptoms or risk factors, and that H. pylori status should be determined at endoscopy or non-invasively in every patient.
There is no consensus on the benefit of additional examinations including laboratory testing, abdominal ultrasound, gastric emptying testing, or esophageal pH monitoring.
The biggest area of lack of consensus is the section on treatment approaches for FD. There is an agreement to use to subdivision in PDS and EPS to guide management, but the vast majority of treatment options are not supported for a specific subgroup. There is consensus to eradicate every H. pylori positive FD patient, and PPI therapy is considered an effective therapy for FD, although there is no consensus that it is the preferred initial therapy. There is no consensus on the indication and efficacy of prokinetics or antidepressants, but an almost-agreement (78%) on the use of TCA in EPS. There is also no consensus on the use of other neuromodulators, herbal therapies, acupuncture, or psychological therapies in FD. There is agreement on the use of nutritional support in case of severe weight loss. Finally, there is consensus that the long-term prognosis of FD is favorable and that life expectancy is not shortened in FD.
The areas of uncertainty revealed by this consensus are multiple.
Further unraveling of the FD symptom pattern is useful, and especially the concept of using pictograms deserves additional, preferably multilingual and multicultural studies. These may aid further diagnostic refinement, where the currently only supported tools are endoscopy and H. pylori status assessments, with their limited sensitivity and impact on management and outcome. While there is acceptance of a role for gastric sensorimotor dysfunction in triggering FD symptoms, the role of duodenal luminal or mucosal alterations is not established, but the 76% amount of agreement is still considerable. The putative entity of H. pylori-associated dyspepsia is poorly described and needs further characterizing studies, especially in the West. The voting on the therapy statements clearly establishes the need for more therapeutic trials, preferably in a multi-center setting using validated endpoints. In the absence of a well-established path to regulatory approval, scientific or professional organizations such as the ESNM or the Rome Foundation may consider taking the lead here.

CON CLUS ION
FD is a highly prevalent and impactful clinical condition. This Delphi process used a multinational and multidisciplinary group of European experts to summarize the current state of consensus on definition, symptom characteristics, pathophysiology, diagnosis, treatment, and prognosis of this condition. The Consensus Group voted on several statements that may guide clinicians in recognizing, diagnosing, and treating FD in clinical practice, whereas the statements without consensus identify areas in need of future research.