Functional dyspepsia (FD), defined as “epigastric symptoms affecting daily life, such as postprandial fullness, early satiation, epigastric pain and burning, in the absence of underlying organic abnormalities” (1), is an extremely common functional gastrointestinal disorder. In the general population, the prevalence of FD is found to be up to 21% (2, 3). Although only a minority of H. pylori infected patients remain asymptomatic after successful eradication therapy, patients reporting helicobacter pylori-associated dyspeptic symptoms are now being recognized as a separate entity referred to as H. pylori associated dyspepsia (1, 4, 5).

To facilitate the management of FD, the Rome Consensus subdivided FD into two subtypes: Postprandial Distress Syndrome (PDS) (60%) characterized by meal-related symptoms, such as postprandial fullness, early satiation, postprandial epigastric pain and other symptoms triggered by food ingestion, and Epigastric Pain Syndrome (EPS) (20%) characterized by epigastric pain and burning (4, 6). Approximately 20% of FD patients overlaps between PDS and EPS.

FD is extremely common, with estimates of 10–30% prevalence in the general population, and is associated with substantial medical care costs and a considerable health economic impact (7–9). A proportion of 20–25% of the patients with severe and refractory GI symptoms also have psychosocial co-morbidities such as anxiety, depression or somatization and severely impaired daily functioning (about 10% of these patients have work disability). Somatization, namely multiple stress-related symptoms of unknown origin resulted to be the most important risk factor for impaired QOL in patients with severe functional dyspepsia (10). This FD subgroup is often referred to advanced care, which may be associated with even higher health economic costs (11).

Finally, FD patients also show an important degree of overlap with gastro-esophageal reflux disease (GERD) (12, 13) and irritable bowel syndrome (IBS), and are, thus, often misclassified.

Gastroparesis is characterized by delayed gastric emptying and by upper gastrointestinal symptoms (nausea, vomiting, abdominal pain, early satiety, bloating) in the absence of mechanical obstruction (14). Two of the most common types of gastropareses are idiopathic gastroparesis and diabetic gastroparesis (15). Gastroparesis can also be a complication of upper gastrointestinal surgery, neurological disease, collagen vascular disorders, viral infections, or drugs use (16). It is associated with a major impact on the patients' quality of life and substantial social and health economic costs (17).

In the classical pathophysiological model, functional gastrointestinal disorders (FGIDs) are considered heterogeneous conditions, and symptoms are attributed to a combination of motility disturbances, visceral hypersensitivity, low grade mucosal immune activation, and altered processing of gut-brain signals (18). This is based on the presence of impaired gastric storage and emptying function in FD and gastroparesis, as well as findings of visceral hypersensitivity and increased levels of depression, somatization and anxiety, which are considered markers of altered gut-brain interaction (19–21).

Recent research has focused on visceral hypersensitivity as a common mechanism determining symptom severity and impact across several functional gastrointestinal disorders (19). To date, the focus of research has mainly been on hypersensitivity to mechanical stimuli, studied by balloon distention (22). However, there is increasing evidence for a role for visceral hypersensitivity to specific nutrients as well, suggested amongst other by the observation that FODMAPs induce symptoms and the observation that specific nutrients induce local immune activation in irritable bowel syndrome (IBS) patients but not in health (23, 24).

The gastrointestinal mucosa expresses a wide range of chemosensing receptors, which detect the presence and nature of nutrients in the lumen (25, 26). Nutrients are mainly sensed in the duodenum and jejunum, and initiate an avalanche-effect by releasing gut peptides from entero-endocrine cells into the blood stream. The brain receives these signals through activation of the vagus nerve or directly via the fenestrated blood brain region, the area postrema (25, 26).

There is recent evidence of nutrient-specific enhanced release of gut peptide hormones [motilin, ghrelin, peptide YY (PYY), cholecystokinin (CCK), and glucagon-like peptide 1 (GLP-1)] in FD, which was correlated to intensities of the provoked symptoms. However, most studies are somewhat artificial as they used intraduodenal tube administration of selected nutrients, rather than ingestion of a true meal (27).

The aim of this review was to describe the current evidence on the role of gastrointestinal (GI) peptides in FGID, especially in FD and gastroparesis. We will also address implications for future applications or modulations of gastrointestinal peptides for FD and idiopathic and diabetic gastroparesis treatment.

Both in FD-PDS and in gastroparesis, symptoms are triggered by ingestion of a meal (28, 29). The release of gut peptides in response to nutrient intake is expected to be triggered sequentially, driven by the location of the entero-endocrine cells that are expressing them. Thus, nutrient arrival in the stomach is thought to affect the release of gastrin, ghrelin and potentially somatostatin, while duodenal exposure to nutrients may impact on the release of CCK, motilin, PYY, and GLP-1, among others (25–27). In addition, serotonin release is expected to occur when nutrients enter the duodenum (25) (Figure 2). The association between peptide levels and symptoms in FD and gastroparesis is summarized in Table 1.

Gastrin is released by G-cells in the stomach and is a major stimulus for gastric acid secretion (25). As a group, FD patients do not seem to have altered gastrin levels according to a study of Jonsson et al. (30). However, in a study by He et al., FD patients with delayed gastric emptying had significantly higher gastrin levels (31). A recent study from Poland confirmed these findings, with elevated gastrin levels in both PDS and EPS (32). Use of acid suppressive therapy, often applied in FD as first-line therapy, may increase gastrin levels and it remains unclear to which extent the studies could rigorously exclude such confounder. In a relatively small study from Japan, gastrin serum level did not predict the response to H2 blocker therapy in FD (33).

Somatostatin is released in the stomach but also in the small bowel, and has a strong inhibitory effect on gastrointestinal motility and secretion (25). In the study by He et al., plasma somatostatin levels and mucosal expression of somatostatin in the antrum and the duodenum did not differ between health and FD, with normal or delayed emptying (31). The same was found in FD patients as a group in the study by Jonsson et al., but higher symptom burden was associated with higher fasting somatostatin levels in FD, and somatostatin levels were also correlated with heartburn severity scores (30). FD patients displayed a rapid, transient, somatostatin peak during a stress interview compared to matched controls (30). In a study by Russo et al., comparing gut peptide levels between 42 PDS and 12 EPS patients, somatostatin levels tended to be lower in PDS compared to EPS but this did not reach statistical significance (34). Itopride, a prokinetic agent with mixed dopamine-2 receptor and cholinesterase inhibitory actions, was reported to acutely increase somatostatin plasma levels (35). The somatostatin analog octreotide was reported to slow gastric emptying, enhance fasting gastric volumes and suppress meal-induced volume increments in healthy subjects (36). Clinical reports with somatostatin analogs in FD patients are lacking.

Ghrelin is produced by endocrine P/D1 cells in the stomach, with plasma levels that increase during fasting and decrease after food intake (25, 88). Ghrelin a 28 amino acid peptide which needs to have an octanoyl group attached to its third serine residue to be biologically active (25). Ghrelin levels are inversely related to body weight (89, 90) and decrease with increasing extent of gastric mucosal atrophy (91, 92). Several studies have investigated ghrelin release in FD and gastroparesis, and the findings are conflicting (38). Most studies found no difference in fasting ghrelin levels in FD compared to health (37, 39–41, 93, 94). However, in a small group of EPS and PDS patients compared to health, correlations were reported between ghrelin levels and symptom severity, in particular epigastric pain in EPS and early satiation in PDS (38). Shindo et al. found a negative correlation between plasma acylated ghrelin levels and gastric emptying rate in patients with PDS but not with EPS (39). Takamori et al. reported lower levels of des-acyl ghrelin (the inactive form after hydrolysis of the octanoyl group), in dysmotility-like dyspepsia according to the Rome II criteria (40) while Nishizawa et al. reported higher ghrelin levels in FD patients as a group (41). The ingestion of a high fat meal in FD patients did not differently affect the plasma ghrelin levels in FD compared to healthy subjects (42). Intravenous administration of ghrelin, twice daily for 2 weeks, significantly increased appetite and tended to increase daily food intake in FD patients with loss of appetite (43). Furthermore, Arai et al. observed a clear improvement in upper gastrointestinal symptoms in FD patients after administration of the Japanese Kampo medicine Rikkunshito, which increased plasma ghrelin levels (44). Suzuki et al. also showed Rikkunshito was effective among H. pylori-infected participants with increased plasma ghrelin levels (45).

Plasma ghrelin levels increased with sham feeding in healthy controls and patients with idiopathic gastroparesis but not in patients with diabetic or postsurgical gastroparesis, indicative of a role for intact vagal signaling in the control of ghrelin release (46). In pilot studies, acute intravenous administration of ghrelin enhanced gastric emptying rate in idiopathic and diabetic gastroparesis (47–49). In idiopathic gastroparesis patients, symptoms were also improved.

Subsequently, several ghrelin agonists have been studied, with a major focus on diabetic gastroparesis. The intravenously administered macrocylic peptidomimetic molecule ulimorelin, enhanced gastric emptying, and was subsequently mainly studied in critical care patients, with lack of differentiation from metoclopramide (50, 51). The orally administered TZP-102 showed promising results in phase 2a, but this was not confirmed in phase 2b (52, 53). Relamorelin, an injectable ghrelin receptor agonist, showed efficacy in diabetic gastroparesis patients with active vomiting symptoms in two placebo-controlled phase 2 studies and is being evaluated in phase 3 studies (54, 55).

Motilin is released from M-cells situated in the proximal duodenum during the fasted state, is a stimulus for strong antral contractions and has a hunger signaling function (25, 95). Several studies evaluated plasma motilin levels in FD and gastroparesis (34, 56, 96–98). FD patients as a group have comparable fasting plasma levels to those in health (95). Russo et al. reported higher fasting motilin plasma levels in EPS compared to PDS (34). In the same study, elevated CRF levels were also reported in PDS. The relevance of this finding is unclear. It is well-known that motilin plasma levels fluctuate with interdigestive motility and are maximal during gastric phase III (95). The study by Russo et al. did not correct for migrating motor complex (MMC) cycle, which could be a major confounder, as it is conceivable that PDS patients have less occurrence of gastric phase III (96).

In patients with FD and delayed gastric emptying, motilin plasma levels did not display the normal fluctuations during the interdigestive state and gastric phase III contractions were absent (56). In patients with diabetic gastroparesis, motilin plasma levels were elevated but still fluctuating during the interdigestive state, although antral phase III activity was absent (57, 98). In FD patients with unexplained loss of appetite, gastric phase III contractions are suppressed, suggesting low plasma levels, but these were not measured in this study (99).

Several macrolide antibiotics such as erythromycin and azithromycin have motilin receptor agonistic effects, and have a stimulatory effect on gastric emptying rate (100–102). The impact on symptoms, however, was often disappointing (101). A number of macrolides without antibiotics but with motilin receptor agonistic properties were developed for the treatment of FD and diabetic gastroparesis (103). However, invariably, they failed to provide significant symptomatic benefit in phase 2 studies and no agent progressed into phase 3 studies (58–61, 104). The main reasons that have been put forward to explain the lack of success in trials with motilin agonist drugs for gastroparesis have been the use of too high doses, which impact on gastric accommodation, and the use of long-acting agents which are prone to desensitization (62, 103).

Camicinal is a novel small molecule motilin receptor agonist with short half-life, which was shown to induce gastric phase III contractions during the fasting state and dose-dependently enhance gastric emptying rate (63, 64). In a phase 2 study, the lowest dose of camicinal significantly improved symptoms, confirming the therapeutic potential of this class of agents, whereas only the highest dose studied enhanced gastric emptying. Indicating that enhanced emptying rate does not underlie the symptom improvement (65). Camicinal was also studied in critical care patients, but the drug has not advanced to phase 3 in any indication (66).

CCK is a brain-gut peptide released from I-cells in the upper small intestine upon food intake, especially after meals containing high fat or protein amounts (25). In H. pylori associated dyspepsia patients, significantly lower CCK basal values were demonstrated in comparison to H. pylori negative patients (67). Hyper responsiveness to CCK can be one of the pathophysiological pathways for the occurrence of symptoms in FD patients (105). A recent study showed a correlation between the release of gut peptide hormones as CCK and provoked symptoms after infusion of nutrients into the duodenum (27). However, in this study, intraduodenal tube administration of selected nutrients was used, rather than ingestion of a true meal. An early increase of CCK plasma levels was found, followed later by a rise of other peptides such as GLP-1 and PYY. Previously, it has also been shown that the intraduodenal infusion of fat may trigger symptoms as fullness and discomfort and to sensitize the stomach to gastric distension (68, 105). Duodenal lipids induce higher CCK levels in patients with FD compared to health, and the CCK-A receptor antagonist dexloxiglumide, was able to reduced sensitivity to gastric distension after lipid administration (69, 70, 106). However, ingestion of a low fat meal when patients perceived intake of a high fat meal (cognitive factors) did not significantly change the CCK level but was associated with higher symptom scores (71).

In addition, a CCK antagonist accelerated the gastric emptying rate which could lead to a benefit in both functional dyspepsia as gastroparesis patients (107). The improvement in gastric emptying probably involves an effect of CCK on capsaicin-sensitive vagal pathways (107). Infusion of CCK in healthy volunteers resulted in an increase in gastric compliance, but this was not confirmed in a study with FD patients (108). Unfortunately, in spite of a number of positive mechanistic observations, CCK-receptor antagonists were not further developed for the treatment of FD.

GLP-1, secreted by intestinal endocrine L-cells upon food intake, slows the gastric emptying in diabetes with a decrease in glycemia (108). In healthy controls, elevated GLP-1 plasma levels after administration of the Dipeptidyl peptidase-4 inhibitor vildagliptin, were associated with impaired gastric accommodation (72). In Japan, gastric emptying was measured in healthy subjects and increased significantly after ingestion of a broth with rice, which was accompanied by a significantly more rapid rise in plasma GLP-1 and glucose levels compared to rice with water (73). In an earlier study, it was shown that GLP-1 was correlated with nausea in a single meal experiment in FD patients subtype EPS as well as in healthy volunteers (74). This would be an interesting fact for the use of medication acting on the GLP-1 receptor for the treatment of gastroparesis patients with nausea as one of their main symptoms.

PYY is a gut hormone secreted from endocrine L-cells in the gut mucosa, most prominently present in the ileum and the colon, and released into the circulation after ingestion of food (25, 109). As mentioned above, the intake of lipids is often a trigger for symptoms in FD. In FD patients, ingestion of a high fat meal was associated with lower postprandial PYY levels compared to healthy volunteers (42). In addition, PYY was found to be correlated with symptoms such as a sensation of fullness in EPS patients after a single drink test and a satiety test (74). However, based on the literature, little is known about the effect of PYY in FD patients.

5-HT is also released by entero-endocrine cells in the gastrointestinal tract, in response to mechanical stimulation or the presence of nutrients or toxins (25, 110, 111). It has its effect via 14 known serotonin receptors, but we will focus on the most relevant ligands in this review. The role of 5-HT in upper gastrointestinal physiology remains unclear, due to a lack of suitable agonists and antagonists for human application (110). While 5-HT4 agonists enhance gastric accommodation and gastric emptying, 5-HT3 antagonists had no significant effect on these functions, and 5-HT_1A agonists enhance gastric accommodation and tend to slow gastric emptying (75–81). Alternative approaches to unravel a role for 5-HT in gastric sensorimotor function has been the use of tryptophan depletion (82) and the administration of selective serotonin reuptake inhibitors (SSRIs) (83, 84). Acute tryptophan depletion enhanced gastric accommodation, which was also observed with short-term SSRI use, while acute intravenous SSRI administration inhibited accommodation, suggesting that endogenous serotonin release serves to limit gastric accommodation (82–84). In terms of interdigestive gastric motility, acute intravenous SSRI administration suppresses gastric phase 3 while stimulating intestinal phase 3, and ondansetron also inhibited the occurrence of gastric phase 3 (85, 86).

Studies focusing on IBS have shown that circulating 5-HT levels rise after a meal, and that this rise is exaggerated in IBS with diarrhea and suppressed in IBS with constipation (112, 113). These studies used platelet-depleted plasma to measure circulating plasma levels of gastrointestinal origin, thereby eliminating the confounding effect of storage in thrombocytes. Similar studies in FD are lacking. One study measured plasma 5-HT in FD and found decreased basal and postprandial plasma compared to health (87). This is in agreement with a recent study reporting a decreased number of duodenal serotonin containing endocrine cells in FD patients (74).

Several 5-HT receptor agonists/antagonists, such as cisapride (5-HT4 agonist, 5-HT2, and 5-HT3 antagonist), tegaserod (5-HT4 and 5-HT1 agonist, 5-HT2a/b antagonist), mosapride (5-HT4 agonist, 5-HT3 antagonist) and revexepride (5-HT4 agonist) have been evaluated for the treatment of dyspepsia and gastroparesis, although not all studies show efficacy (114–116). A recent metanalysis showed that FD patients treated with serotonin receptor agonists have a significantly better symptom response compared to placebo (117) and the most recently published evidence indicates efficacy for prucalopride in idiopathic gastroparesis and emerging efficacy for velusetrag in gastroparesis (118, 119). Case series suggest potential benefit of the 5-HT₃ antagonists granisetron or ondansetron for symptoms of nausea and vomiting in gastroparesis, but formal studies are lacking (120, 121).

JT has given Scientific advice to AlfaWassermann, Allergan, Christian Hansen, Danone, Grünenthal, Ironwood, Janssen, Kiowa Kirin, Menarini, Mylan, Neutec, Novartis, Noventure, Nutricia, Shionogi, Shire, Takeda, Theravance, Tramedico, Truvion, Tsumura, Zealand, and Zeria pharmaceuticals and has served on the Speaker bureau for Abbott, Allergan, AstraZeneca, Janssen, Kyowa Kirin, Menarini, Mylan, Novartis, Shire, Takeda, Truvion, and Zeria. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.