Emesis as a Model System for the Study of Functional Bowel Disease

Abstract

INTRODUCTION Wherever you go, there are children who vomit. Causes of acute vomiting episodes include infections, minor head injury, gastrointestinal obstruction, and other surgical conditions. Chronic vomiting or recurrent vomiting is less well categorised, but includes regurgitation due to gastroesophageal reflux, allergic and eosinophilic diseases of the gut, chronic gastrointestinal dysfunction, metabolic disturbance, disease of the central nervous system, and disorders such as cyclical vomiting syndrome. It has even been noted that in children it can be a “symptom of almost any disease system” (1). This article outlines the mechanics and mechanisms of emesis and related reflexes, such as belching and gastroesophageal reflux, and highlights how knowledge of the underlying physiology has implications for understanding functional upper gastrointestinal tract disease in children. In addition, we review a number of challenging areas of paediatric gastroenterology involving nausea and vomiting. The authors use the following terminology: Vomiting is a lay term, used by patients and their families, to describe previously swallowed food and gastrointestinal secretions coming up the oesophagus and out of the mouth by any mechanism. Regurgitation describes the apparently effortless movement of gastric contents from stomach to oesophagus and out of the mouth. Emesis describes the forceful expulsion (vomiting) of stomach contents out of the mouth by vigorous contraction of the anterior abdominal wall muscles and diaphragm following activation of the emetic reflex. It is vital that clinicians distinguish between regurgitation and emesis, because when they become pathological, the treatment for each is different. Inappropriate treatment risks failure, and surgical therapy for regurgitation due to gastroesophageal reflux is likely to worsen the symptoms of emesis. The emetic reflex is triggered by a wide range of peripheral or central stimuli (2). Inputs to the “vomiting centre” (the collection of nuclei in the brainstem coordinating the emetic motor outputs) include gastrointestinal vagal afferents, the area postrema (“chemoreceptor trigger zone,” subject to direct influence by blood and cerebrospinal fluid–borne factors), vestibular system (motion sickness and inner ear disease), and stimulation of the pharynx. Severe abdominal pain is also a potent stimulus, but the pathways by which it induces emesis are unclear; although the noxious stimuli are conveyed in the splanchnic afferents, stimulation of these afferents does not evoke reflex emesis, in contrast to stimulation of abdominal vagal afferents. The emetic reflex also can be activated by stimulation of more rostral regions of the brain (eg, hypothalamus, limbic system) and by unpleasant sights or smells. The reflex also is amenable to Pavlovian conditioning, but relatively little is known of the descending pathways involved and their clinical significance. However, it is likely that the threshold for activation of the reflex by a number of stimuli is capable of modulation from higher brain regions, and this is supported by observations that sensitivity to motion sickness is a predictor of emesis to anticancer chemotherapy, postoperative nausea, and vomiting and pregnancy sickness (2). The emetic reflex may be considered in 2 phases—the prodromal phase and the ejection phase. The prodromal phase is characterised by nausea, an unpleasant but not painful sensation related to the upper abdomen, associated with a desire to vomit or a feeling that vomiting is imminent. It may precede emesis or it may occur in isolation (2), as can emesis itself. Emesis may alleviate nausea. The physiological basis of nausea remains poorly understood, but there is a strong association with gastric antral dysrhythmia and a large increase in the plasma levels of arginine vasopressin (3). Nausea is often accompanied by autonomic events, including sweating, peripheral vasoconstriction (causing pallor), tachycardia, reduced gastric secretion and pupil dilatation due to sympathetic nervous activity, and increased salivation due to parasympathetic stimulation (2). The central nervous systemic pathways involved in the genesis of the sensation of nausea are not known, although the inferior frontal cortex has been implicated. The ejection phase consists of retching and vomiting. First, there is a vagally mediated relaxation of the stomach and the lower oesophageal sphincter. A retrograde giant contraction originating in the mid-small intestine sweeps to the stomach, and is also under vagal efferent control. This contraction probably accounts for the frequent presence of bile in vomitus, except when pyloric obstruction is present. Tonic longitudinal contraction of the pharyngoesophageal junction pulls up the oesophagus, helping to open up the gastroesophageal junction. Then retching begins, with the anterior abdominal wall muscles and entire diaphragm (including the crura) contracting synchronously, with displacement of the abdominal oesophagus and gastric cardia through the crural hiatus into the thorax. Although gastric contents may enter and leave the lower oesophagus, they are not ejected. During vomiting, the perioesophageal diaphragm (ie, the right crus) relaxes, and the expulsion of gastric contents is achieved by the somatic muscles compressing the relaxed stomach. In the dog, a retrograde-propagated pharyngoesophageal contraction promotes the forcible ejection of gastric contents from the mouth (4), but it is unclear whether this occurs in humans. The purpose of the emetic reflex is defensive: to remove contaminated food from the upper gastrointestinal tract (2). Nausea stops further ingestion and facilitates learned aversion. It is an aversive stimulus, much more so than pain (5). The emetic reflex is a protective gastrointestinal reflex and in the normal course of events should be activated only occasionally; however, chronic gastrointestinal disease or dysfunction may result in frequent activation. In some neurologically impaired children, the central neurological damage appears to result in loss of inhibition, or hypersensitisation of the emetic reflex, which is then activated in the course of normal everyday activity. The belch is a vago-vagal reflex that permits oral expulsion of excessive intragastric air. Accumulation of gas in the gastric fundus and distension of the region of the cardia results in sudden and complete transient relaxation of the lower oesophageal sphincter (6), accompanied by relaxation of the diaphragmatic crus (7). A common cavity phenomenon occurs, attributed to reflux of air and other gastric contents into the oesophagus, with equalisation of gastric and oesophageal pressure; 1 or more belches may then take place as air is expelled (8,9) facilitated by contraction of anterior abdominal muscles. Note that the somatic motor changes during belching are similar to those occurring during vomiting, although they are considerably less forceful, and from the person's point of view, feel effortless. Unlike activation of the emetic reflex, there is no prodrome of nausea or associated autonomic events, such as sweating or vasoconstriction. Gastroesophageal reflux is the apparently effortless leakage of gastric contents (including food and gastric secretions) up into the oesophagus. It occurs when the mechanisms of oesophagogastric competence malfunction or are overcome by exceptional factors. Episodes of gastroesophageal reflux (GER) may occur in normal, healthy individuals without significant consequences. Gastroesophageal reflux disease (GERD) is present when the reflux results in significant symptoms or harm. A major factor in the occurrence of GER is dysfunction of the mechanism of lower oesophageal sphincter complex (ie, the smooth-muscle lower oesophageal sphincter [LOS] and the encircling right crus of the diaphragm). The most common mechanism of GER, in healthy individuals and in patients of all ages, is transient lower oesophageal sphincter relaxations (TLOSR); in other words, LOS resting pressure is normal but reflux occurs during episodes of a sudden, brief drop in pressure to near zero, which is not associated with primary oesophageal peristalsis induced by swallowing (10–15). This is more frequent in the immediate postprandial period and in the presence of gastric distension, and is accompanied by selective and complete inhibition of the crural diaphragm. TLOSRs can be triggered by activation of vagal afferents supplying the gastric fundus and cardia, and the resulting reflex motor responses are presumed to be coordinated in the brainstem (13,16). The normal physiological process most closely related to TLOSR is belching (8), although again the relaxation of the crural diaphragm and the LOS is reminiscent of the mechanics of emesis. Other mechanisms of GER include very low LOS resting pressure, downward drifts in resting pressure, reflux during swallow-induced relaxations, and GER due to abdominal straining (17,18). LOS length, or lack of it, especially the abdominal oesophagus, influences the effectiveness of the LOS as a reflux barrier. Sliding hiatus hernia is common in patients with GERD. This will dissociate the diaphragmatic crus from the LOS and disrupt antireflux mechanisms (19–21). Conversely, a long intraabdominal oesophagus will have antireflux properties because it is subject to intraabdominal pressure that will tend to compress it closed. In a nonvomiting species, such as the rat, the abdominal oesophagus is disproportionately long and narrow compared with humans. There is evidence that if the distal oesophagus is replaced by a tube of sufficient intraabdominal length, then reflux is prevented even in the absence of the LOS (22,23). In adults with reflux, relatively small quantities of gastric contents reach the mouth, but in children, the amounts refluxed may be much greater and may result in a large proportion of the recently ingested feed effortlessly pouring out of the mouth. Consideration of the underlying physiological processes should make clear the difference between reflux vomiting and vomiting due to activation of the emetic reflex. Reflux vomiting is akin to belching; there is no prodrome and it appears effortless. Emetic vomiting is preceded by feeling unwell with nausea, pallor, sweating, and tachycardia, and it is accompanied by violent contractions of the anterior abdominal wall muscles and diaphragm, leaving the subject feeling drained and exhausted. CLINICAL IMPLICATIONS FOR CONFUSION BETWEEN EMESIS AND GER Chronic vomiting in children may appear to be a result of GER and regurgitation. It is reported to be particularly common in neurologically impaired children and is attributed to central nervous system dysfunction (24); however, it is not always easy to get the diagnosis right. The importance of understanding the pathophysiology and the potentially devastating effects of misdiagnosis is emphasised by considering the effects of inappropriate fundoplication. Many children, particularly those with neurological impairments, do not show a full symptomatic response to antireflux medication. They are considered to have severe reflux disease and to require surgery. However, a high failure rate has been documented (25,26) and an alternative explanation must be considered. Failure to respond to antireflux therapy may not indicate severe GERD, but that some or all of the symptoms are due to another cause (eg, activation of the emetic reflex) (24). Many of the children with failure to respond to antireflux therapy have symptoms other than those of GERD. One prominent troublesome symptom that persists after fundoplication is retching (25–33), but this is a component of the emetic reflex and not a symptom of GER (24). These children may also have evidence of nausea (34). Performing fundoplication on a child whose symptoms are wholly or partly due to activation of the emetic reflex is liable to result in marked postoperative problems (35). Fundoplication does not deal with the underlying causes of emesis, and by creating a valvular mechanism at the oesophagogastric junction and obstructing the movement of gastric contents back up the oesophagus, symptoms are made worse. The child will retch repeatedly, and the accompanying nausea will persist. Children who retch preoperatively have a much higher chance of retching following fundoplication compared with nonretchers (35). In a report on anatomical wrap failure (eg, wrap herniation/wrap disruption) following laparoscopic fundoplication in adults, Soper and Dunnegan (36) found a significant association with forceful contraction of the diaphragm. The process that appears to generate the greatest pressures, and, moreover, the greatest pressure gradient from the abdomen to the thorax, is retching. Retching generates huge forces capable of causing wrap disruption, and in particular, forces that specifically drive the wrap through the crural hiatus (ie, wrap herniation into the thorax). Using radioopaque markers, Johnson and Laws (37) demonstrated elevation of the oesophagogastric junction through the crural hiatus just before retching and vomiting, with displacement of the cardia and distal oesophagus cranial to the crus, into the thorax. In a subsequent study of the mechanics of vomiting in the cat, McCarthy and Borison (38) described retches as a metronomic series of pulses, with brief negative-pressure pulses in the thorax mirrored by positive-pressure pulses in the abdomen, progressively building up a substantial transdiaphragmatic pressure gradient (200–300 mmHg) and culminating in a prolonged positive abdominal and thoracic pressure wave with sustained abdominal contraction and elevation of the diaphragm and vomit expulsion. During retching, there was a pulsing cephalad displacement of the oesophago-gastric junction, with the fundus of the stomach drawn through the diaphragmatic hiatus, into the thorax (38,39). These reports indicate that there is normally a substantial movement of the gastroesophageal junction during retching and vomiting, with significant pressure changes. Fundoplication in the retching child sets the scene for anatomical wrap failure. Repeated episodes of retching drive the oesophagogastric junction through the diaphragmatic hiatus into the thorax and pull the wrap apart. Patients classically present with increasing postoperative retching (wrap intact), which may later progress to retching and vomiting, associated with subsequent documentation of wrap disruption. Retching precedes wrap failure and is a cause rather than a symptom of wrap herniation. Not only does fundoplication fail to relieve emesis but there is also evidence that fundoplication may even sensitise the emetic reflex, reducing the threshold for activation of retching and emetic vomiting. Some children may develop new onset of retching after fundoplication, and parents of children who retched before surgery believe the postoperative retching is worse. Gastric dysrhythmias, as recorded by the surface electrogastrogram, may worsen after fundoplication, correlating with retching symptoms (24). In an animal model of fundoplication (the ferret), fundoplication was followed by an increased sensitivity to a low dose of the centrally acting emetic loperamide (40). A possible mechanism of sensitisation is the presence of peripheral nerve and muscle damage and scarring as a result of surgery. Histological examination of the fundoplication in the ferret shows clear evidence of scar tissue in the wall of the distal oesophagus and the inner gastric layer of the fundoplication wrap, in the region of the cardia (41). This effect may be even more marked in disrupted wraps. In ferrets with disrupted fundoplications, we observed an increased retching response to induction of general anaesthesia compared with intact fundoplications (in turn greater than controls), together with evidence of gastric dysmotility and increased scar tissue and nerve damage (41). This has implications for redoing fundoplications after wrap disruption, with a higher risk of failure, increasing damage, physiological dysfunction, and ever-increasing symptoms. CHALLENGES IN NAUSEA AND VOMITING IN PAEDIATRIC GASTROENTEROLOGY How Do We Identify Nausea in Young Children and Neurologically Impaired Children? Nausea is a self-reported subjective sensation, so how can the presence of this significant clinical symptom be established in patients who are unable to self-report, such as young children and those with severe neurological impairments? The situation is analogous to that in animals, in which behavioural changes and physiological markers have been used as surrogate markers for the presence of the sensation (2). Traditional teaching states that pain from reflux oesophagitis causes food refusal in children. In our study of children before and after Nissen fundoplication, we observed a striking association between food refusal and retching, strongly suggesting that in the context of recurrent vomiting, food refusal is a manifestation of the nausea that accompanies the emetic reflex (34). Refusal of specific foods is particularly suggestive of nausea-induced taste aversion. An infant with activation of the emetic reflex secondary to intolerance of cow's milk protein may refuse whole-protein cow's-milk formula, but will readily drink a hydrolysed formula or water. Parents of a child with emesis may say that they appear unsettled or “pull a face” immediately before emesis, again suggesting distress and nausea. Other indicators of nausea may include the surface electrogastrogram; development of a gastric dysrhythmia in response to food may indicate nausea in the same way that the appearance of tachygastria correlates with the onset of motion sickness (42). Measurements of plasma vasopressin have not been made in children with suspected nausea, but such measurements may be helpful in ensuring that this distressing symptom does not go unrecognised and hence untreated. Cyclical Vomiting Cyclical vomiting is a that is most reported in children 2 to but which can also occur in It is characterised by a of or more of acute nausea and vomiting to with to with of and central nervous system disease of this a to into the physiology and of the emetic reflex. The vomiting is particularly and prolonged and is associated with abdominal pain in of is the that the nausea is not by vomiting The in is for at 2 It is an of a in which there is and prolonged activation of the emetic reflex without an emetic and as such may a in which the emetic reflex has become under include of abdominal afferents, such that normally stimuli evoke emesis in these such a mechanism be with the high of abdominal pain in patients with A in the brainstem nuclei that to or the sensitivity of the emetic reflex as an or vomiting a number of with properties a range of emetic stimuli (eg, is considered to the a in these then this an to to mechanisms in the emetic such as and metabolic a number of of the and are known to be potent emetic in their as well as lower the threshold to other emetic stimuli such as which a in the of in the brainstem nuclei involved in emesis, such as the to induction of emesis without an stimulus, although often is as a There is evidence for some of in with in the also have been to the increased of in the of children with These may into the that the sensitivity of the emetic reflex. There is evidence that nausea responses to are higher in compared with and The of the effects of in the and their subsequent into the for the treatment of and emesis and postoperative nausea and vomiting to a significant in of this distressing effect in children and These by the activation of on the peripheral and central of abdominal vagal afferents, with the peripheral of the afferents activated by from in the of the The are particularly in the acute phase of but are less or in the and the for to this phase was 1 of the that to an effect of in emesis, in (eg, and in patients anticancer The is the for the a The with this of that they a effect with stimuli acting abdominal vagal afferents, the area and the vestibular system, for a central of the nuclei the of the emetic reflex This is the of a selective such a of and a on that it may be possible to a or for use in humans. The of of the only is under but a study has evidence that it emesis in patients with severe to other we are to the of children with chronic vomiting we to distinguish between activation of the emetic reflex and GERD. nausea other prodromal such as sweating, or and retching point effortless A and is Parents of children with GER distinguish the regurgitation of GER from the emesis of Parents who report that their child vomiting the stomach is are the forceful and of activation of the emetic reflex. In the has increasing evidence that that the emetic reflex are the cause of symptoms previously attributed to GER (eg, eosinophilic oesophagitis and allergic oesophagitis in suspected should not be for evidence of they should be other specific causes of and for evidence of emesis or general to be with shows it does not make a diagnosis of reflux An of retching during a contrast is liable to be as reflux and hiatus at are are subjective and can only be this of more and the reports of as a common cause of vomiting in children with severe neurological The and of these children with recurrent emesis due to dysmotility our greatest