15.7 Diseases of the oesophagus 2828

15.7 Diseases of the oesophagus 2828

ESSENTIALS Defective conduit function of the oesophagus readily induces clinical symptoms and may have serious effects on nutrition and the lungs, the latter resulting from aspiration of gastro-​oesophageal contents. Oesophageal pain and dysphagia caused by diseases of the muscular layer or epithelium are often disabling. The oesophagus is exposed to numerous hostile environments including carcinogens in food or those derived from tobacco, betel nuts, and other ingested sources and hence carcinomas can occur. Chronic reflux disease from exposure to corrosive upper gastrointestinal secre- tions is also associated with malignant disease—​adenocarcinoma—​as well as benign stricture formation related to the action of pepsin. Introduction The oesophagus is responsible for transporting food to the stomach without compromising the safety of the airway. Imprecise muscle coordination of the oropharynx and the oesophageal body can have serious consequences such as aspiration pneumonia and malnutri- tion. Diseases of the oesophagus generally present with symptoms of pain or dysphagia either alone or in combination (odynophagia). True oesophageal dysphagia should be distinguished from oropha- ryngeal dysphagia, an abnormality of transfer of food from the mouth to the oesophagus, by a careful history. Other symptoms and signs of oesophageal disease include weight loss, anaemia, cough, hoarse voice, and breathlessness secondary to aspiration pneumonia. The most common oesophageal disorder is gastro-​oesophageal reflux disease which should be taken seriously because of its profound im- pact on quality of life in some people and since it can be a precursor to oesophageal adenocarcinoma. The diagnosis of oesophageal dis- ease may generally be determined by a thorough history and a small number of focused investigations outlined in the next section. Investigation Diagnosis and treatment of suspected oesophageal disease has been enhanced by the introduction of endoscopic techniques, but in some circumstances radiological investigations and physiological studies of oesophageal function using manometry and pH-impedance provide essential guidance for diagnosis of motor disorders and nonerosive reflux disease, which can be related to systemic diseases such as systemic sclerosis. Endoscopy An oesophagogastroduodenoscopy is usually the first investigation of choice to determine whether there is a mechanical cause for the symptoms and whether there is any evidence of inflammation or mucosal damage. Direct visualization of the oesophagus, combined with biopsy as appropriate, offers immediate benefit for the diag- nosis of mucosal diseases such as (1) infections due to candida and herpes simplex virus; (2) benign and malignant strictures; (3) and bleeding lesions, including varices. Although the endoscopic ap- pearances may be suggestive of a motility disorder, this is not the investigation of choice for assessing oesophageal motor function. Endoscopic technology is advancing rapidly and enhanced magnifi- cation coupled with permutations in the parts of the light spectrum used for image processing (e.g. narrow band imaging) permit a more detailed analysis of the mucosa and superficial vasculature in real time. There is now evidence that these technologies can aid in the diagnosis of dysplasia and early oesophageal cancer. Endoscopic ultrasound is an important staging investigation for oesophageal cancer (discussed later in this chapter) and also has a role in characterizing submucosal lesions. Radiology The barium or Gastrografin radiographic swallow has largely been superseded by endoscopy for the diagnosis of structural abnormal- ities; however, it can be a useful guide prior to interventions such as endoscopic stent insertion, to evaluate conduit functioning postoperatively, and to exclude the presence of a pharyngeal pouch prior to endoscopy, especially in elderly patients. Useful information can also be obtained about the motor function of the pharynx and oesophagus by videotaping the images and analysing, in slow-​motion replay, repeated tests of standardized stimuli such as bread or barium tablets. Optimal results from radiology are achieved when there is a partnership between a clinician and a radiologist who both have a special interest in oesophageal motor disorders, so that the examin- ation technique can then be tailored accordingly. CT scanning and 15.7 Diseases of the oesophagus Rebecca C. Fitzgerald and Massimiliano di Pietro

15.7  Diseases of the oesophagus 2829 positron emission tomography, increasingly as part of a combined procedure (CT-​PET), are part of the staging investigations for oe- sophageal cancer (see ‘Staging’). A CT scanning with Gastrografin should be taken in cases where oesophageal perforation is suspected. MRI currently has a limited role for oesophageal disease. Oesophageal function testing Techniques are now available for the precise measurement of oe- sophageal motor function. These include oesophageal manometry, intraluminal pH-​impedance monitoring, video fluoroscopy, and high-​resolution oesophageal manometry with isocontour mapping. Together these techniques build a dynamic picture of the physio- logical sequence of events and the effectiveness of contraction which can be compared with resultant functional outcome. These tests may also have a clinical role for explaining symptoms to patients even in the absence of therapy. Oesophageal manometry provides the most direct indication of patterns of oesophageal motor function. It is most helpful in the diagnostic work up of patients with dysphagia, after exclusion of fixed, structural defects. Manometry may also be useful to confirm the placement of intraluminal devices such as pH probes and to as- sess pre-​ and postoperative results following surgery for achalasia or reflux. To determine the asso­ciation between reflux episodes and symptoms, pH-impedance catheters can be used to provide either a stationary or ambulatory recording. Wireless devices are also in- creasingly used. Ambulatory monitoring is particularly useful to determine the association between symptoms and episodes of acid and non-acid reflux in patients in whom the origin of troublesome symptoms is unclear. This test is generally performed off medication but may occasionally be useful to assess response to therapy. Radionuclide measurement of oesophageal transit Computerized scintigraphic analysis of the movement of swallowed radiolabelled boluses can give quantitative information about the patterns of movement of material down the oesophagus. However, its poor spatial resolution makes it an inadequate method for the display of oesophageal anatomy. If structural abnormalities have been adequately excluded, slow or interrupted transit suggests ab- normal motility, although patterns of transit are usually nonspecific. Gastro-​oesophageal reflux disease This is by far the most common oesophageal disorder, indeed the second most common outpatient gastrointestinal diagnosis according to a recent survey of nine American databases. Management should be tailored to the severity, which may vary considerably. Definition Some reflux of gastric and duodenal contents into the oesophagus oc- curs in everybody due to vagally mediated transient lower oesopha- geal sphincter relaxations. It should only be considered a disease when it gives rise to symptoms or complications sufficient to impair quality of life. Pathogenic reflux may occur without causing mucosal damage. The terms reflux or peptic oesophagitis should be reserved for circumstances when endoscopy demonstrates that the oesopha- geal mucosa is clearly breached by the action of the refluxed gastric contents. Minor changes such as erythema, oedema, or friability have been shown to be very unreliable indicators of the presence of oesophagitis. The disease can be further classified into oesophageal and extra-​oesophageal syndromes using the Montreal classification. According to this classification gastro-oesophageal reflux disease can either present with typical symptoms (regurgitation, heartburn or chest pain) or atypical ones (cough, hoarse voice and other respira- tory symptoms). While the link with typical symptoms is generally easy to establish, the correlation with atypical symptoms can be chal- lenging in clinical practice. Aetiology In most patients, reflux disease arises from the excessive exposure of the distal oesophagus to gastric contents which are primarily acid and/​or bile. This is usually because of an abnormal frequency or dur- ation of reflux episodes. In a few patients, however, symptoms arise with relatively normal levels of acid exposure, presumably because of sensitization of the oesophageal mucosa. In normal individuals, excessive reflux is prevented by the lower oesophageal sphincter and an external sphincter formed by the crural diaphragm. Pathological reflux generally occurs as a result of defective neural control of the lower oesophageal sphincter. Hiatus hernia is therefore common in patients with reflux disease and causes displacement of the sphincter from the hiatus formed by the diaphragmatic crura. Most reflux oc- curs during the day, usually after food, but lying down will impair the antireflux effect of gravity and nocturnal reflux can be a disabling symptom. Refluxate is cleared from the oesophagus by peristalsis and buffered by swallowed saliva. Slow clearance of oesophageal acidification contributes significantly to prolonged acid exposure in about 50% of patients. This can be an important contributor to se- vere gastro-​oesophageal reflux disease in systemic disorders such as scleroderma which affect gut motility. Risk factors for gastro-​oesophageal reflux disease include herit- ability (up to 30% of the risk), obesity, and age but not sex. Alcohol and smoking may play a role. Asthma increases the risk as do medi- cations which relax the lower oesophageal sphincter such as cal- cium channel antagonists, nitrates, anticholinergics, and methyl xanthines. Consequences of excessive reflux Symptoms These are an important source of disability. Evidence suggests that heartburn or regurgitations occurring on more than 2 days a week causes significant impairment of quality of life. However, presenta- tion may be with the less specific symptom of dyspepsia (or epigas- tric discomfort), chest pain, belching, nausea or regurgitation, and dysphagia due to either stricture or motor dysfunction of the oe- sophageal body. There are a number of extra-​oesophageal manifest- ations of reflux including respiratory symptoms such as hoarseness, persistent cough, globus, and asthma. Chronic sinusitis, otalgia, and glue ear may also be associated with reflux and—​according to some studies—​even idiopathic pulmonary fibrosis. Complications of reflux disease The chemical insult from excessive exposure of the mucosa to gastroduodenal contents leads to distal oesophageal erosion or ul- ceration in between 40 and 60% of patients with troublesome reflux symptoms and columnar-lined (Barrett’s) oesophagus in between 3 and 10% of individuals, as discussed in detail in the following section. The extent of ulceration varies greatly, from tiny patches of erosion to extensive circumferential ulceration in a small minority.

SECTION 15  Gastroenterological disorders 2830 Peptic stricture sufficient to cause dysphagia is typically only associ- ated with severe oesophagitis. When strictures are severe, this may lead to malnutrition. Bleeding from oesophagitis may occur but is rarely life-​threatening except when it occurs from a deep ulcer asso­ ciated with columnar metaplasia, in which case a neoplastic process is the likely cause of the ulceration. Diagnosis and assessment of severity History The history is pivotal for diagnosis because of the extremely high prevalence of reflux-​induced symptoms and the lack of a defini- tive, inexpensive diagnostic test for reflux disease. A trial of acid-​ suppression therapy can be used as an aid to diagnosis. Endoscopy When investigation is required, endoscopy is the first choice as it is the only test that can give sensitive recognition and grading of oe- sophagitis (e.g. Los Angeles classification system) and reliable diag- nosis of oesophageal columnar metaplasia (Barrett’s oesophagus). Endoscopy also allows for the effective identification of significant peptic strictures, other types of oesophagitis (discussed in later sections), and peptic ulcer disease as well as malignancies. The value of endoscopy as the initial investigation is greatly enhanced by the accurate diagnosis of endoscopic biopsy and, where indicated, cy- tology brushings. As discussed earlier, however, most patients with reflux disease do not have endoscopically visible mucosal damage, so a negative endoscopy does not exclude the diagnosis of reflux disease (so-​called nonerosive reflux disease). Oesophageal function tests The place of these is summarized in Fig. 15.7.1. Oesophageal manometry and ambulatory 24-​h pH monitoring have a limited but important role in the diagnosis of reflux disease (Fig. 15.7.2). Oesophageal pH monitoring is most useful in patients with trouble- some symptoms but without endoscopic signs of oesophagitis in whom a trial of therapy has failed, and patients with atypical symp- toms that cannot be clearly related to reflux. Patients with suspected reflux symptoms but with no endoscopic evidence of oesophagitis who are being considered for antireflux surgery should also undergo oesophageal pH monitoring to maximize the chances of successful symptom relief from this invasive intervention. Multichannel oesophageal impedance monitoring (Fig. 15.7.3) is increasingly available in clinical practice and has been shown to distin- guish reliably between swallowed or refluxed fluid and gas. Combined with pH monitoring, it may identify individuals with ‘nonacid reflux’ who fail to respond to treatment. Bile reflux can be assessed using Bilitec probes, but these are not routinely available anywhere and fur- ther studies are required to elucidate future clinical application. Barium swallow and meal This is an inappropriate primary diagnostic test; it is of no value for the detection of abnormal reflux, and is insensitive for the diagnosis of oesophagitis and cannot grade it. Other pathologies such as gas- tric ulcer and oesophageal stricture are demonstrated with reason- able sensitivity, but adequate evaluation of these findings requires endoscopic biopsy. However, barium swallow is the best method for recognizing extrinsic oesophageal compression which may be pro- ducing symptoms that could be interpreted as being due to reflux, and in the assessment of an anatomically complex hiatus hernia. The mere demonstration of hiatus hernia, however, does not necessarily indicate the presence of reflux disease. Principles of management The major aims of treatment are to provide adequate symptom- atic relief and control of oesophagitis. Reduction of oesophagitis to minor patchy erosions is probably sufficient to prevent the com- plications of oesophagitis, although adequate symptomatic relief is usually achieved only when oesophagitis is completely healed. Tailoring and titration of therapy The severity and frequency of symptoms and the endoscopic find- ings should be used to choose an appropriate level of initial therapy (Fig. 15.7.1). There is some debate about whether to treat patients with a high or low dose of acid-​suppressant drugs initially. An initial trial of low-​ to medium-​dose empirical therapy has the disadvan- tage of giving less precise diagnostic information, and often gives Symptoms mild and infrequent Reassure Nondrug measures Clear-cut history no alarm symptoms Symptoms troublesome PPI 4–8 weeks Good response Maintain PPI Titrate dose to response Endoscopy if not yet done Upper abdominal/ lower retrosternal symptoms Endoscopy ± pH studies Poor response Twice-daily PPI or surgery History indeterminate or alarm symptoms Oesophagitis No oesophagitis Symptoms typical Symptoms indeterminate Reflux disease Function tests Not reflux Manage accordingly Endoscopy Fig. 15.7.1  Principal decision paths for management of reflux disease.

15.7  Diseases of the oesophagus 2831 only slow relief of symptoms. Patients with severe oesophagitis will usually not respond adequately to low-​dose acid suppression. Initial high-​dose proton pump inhibitor (PPI) therapy is likely to give more immediate confirmation of the diagnosis and prompt relief of symp- toms. As per recent National Institute for Health and Care Excellence (NICE) guidance, in general it is recommended to offer a full dose of PPI (e.g. omeprazole/​esomeprazole 40 mg or lansoprazole 30 mg daily) for 4 to 8 weeks for the management of symptomatic reflux. If symptoms recur after stopping treatment, then a low-​dose PPI (e.g. omeprazole 20 mg daily) should be offered as maintenance. In cases of severe oesophagitis, an 8-​week course of full-​dose PPI should be offered, or high-​dose PPI (e.g. omeprazole 80 mg daily) MEAL (C) Chest pain (H) Heartburn C H H H H pH 7 5 3 1 11:00 12:00 13:00 14:00 15:00 Time Fig. 15.7.2  Section of a 24-​h oesophageal pH monitoring study showing the association of heartburn with episodes of acid reflux after a meal. (b) (a) (c) Fig. 15.7.3  Example impedance images obtained from six impedance arranged from top (most proximal) to bottom (most distal) and two pH channels (two tracers at the bottom) which are placed in the oesophagus and stomach, respectively. (a) Impedance changes for normal liquid swallow (arrow). Usually air (higher impedance signal) is ahead of liquid bolus (low impedance signal). (b) Mixed (gas and liquid) acid reflux episode. Mixed reflux is more frequent than liquid only reflux. Gas (belch) reflux is evidenced by a short rise in impedance migrating rapidly into the proximal oesophagus. Arrowhead shows a drop in the oesophageal pH. (c) Non-acid reflux. Oesophageal pH is unchanged, while impedance channels are showing a liquid reflux episode migrating into the proximal oesophagus (arrow). Courtesy of Qasim Aziz, Professor of Physiology, Barts and the London School of Medicine and Dentistry.

SECTION 15  Gastroenterological disorders 2832 in case of no or partial response. Full-​dose PPI may be required as maintenance therapy in patients presenting with severe oesopha- gitis. For optimum cost-​effectiveness and to reduce complications, high-​dose treatment should be followed by a step-​down approach to long-​term therapy as outlined in Fig. 15.7.1. Nondrug measures and antacids The efficacy of these traditional approaches is often overrated. The most useful measures are avoidance of large meals and provocative foods, drinks, and physical activities involving bending or lifting. Stopping smoking and weight loss (if overweight) are recom- mended to help reduce symptoms and for general health improve- ment. Elevation of the bed head may be helpful, if it can be tolerated. Antacids will not usually prevent symptoms, but may be effective in aborting episodes of heartburn. These low-​cost measures are worth a trial in patients with mild intermittent symptoms and should be used as part of maintenance therapy if they prove effective. Acid suppression Inhibition of secretion of gastric acid makes gastric juice less injurious but does not stop reflux. This has deservedly become the most widely used drug therapy because of its high efficacy and adjustability. PPIs are the mainstay of treatment because of their effectiveness in reduc- tion of food-​stimulated acid secretion and their greater overall efficacy in control of acid secretion compared with histamine-​2 receptor ant- agonists. However, histamine-​2 receptor antagonists can be a useful adjunct to PPIs for controlling nocturnal reflux. Long-​term treatment with acid suppressants maintains patients free of symptoms and oesophagitis indefinitely, but withdrawal is usually associated with prompt relapse. The maintenance dose ap- pears to be the same as the lowest effective healing dose. There have been concerns about the safety of long-​term acid suppression ever since the introduction of histamine-​2 receptor antagonists. These include possible associations with susceptibility to infections such as gastroenteritis and pneumonia, malabsorption of vitamins, and increased prevalence of bone fracture. However, meta-​analyses show significant study heterogeneity and confounding effects. To date, follow-​up of many patients treated continuously for 10 years or more with acid suppression has shown no definite evidence of any effects of significance, but in the context of patients who may require treatment with these agents for decades, more extensive follow-​up is still needed. In particular, rare idiosyncratic effects, such hypomag- nesaemia and interstitial nephritis, cannot be dismissed and need to be taken into consideration. With regard to drug interactions, following extensive investi- gation it has been concluded that coadministration of PPIs with antiplatelet agents appears to be safe. However, PPIs can delay methotrexate metabolism, leading to increased toxicity, hence a switch to histamine-​2 receptor antagonists should be considered before induction with this drug. Given these theoretical safety considerations and also drug cost, long-​term treatment of reflux disease with these agents should use the lowest effective dose. In some patients, use of PPIs is limited by side effects such as diarrhoea and this may lead to consideration of surgical antireflux therapy. Motility stimulants The best researched, most efficacious motility stimulant is cisapride, which is a parasympathomimetic that acts as a serotonin 5-​HT4 receptor agonist. It mainly acts through enhancing oesophageal acid clearance as well as increasing muscle tone in the lower oesophageal sphincter. Unfortunately, cisapride can have effects on cardiac con- ductance that may rarely lead to sudden death at peak serum levels, especially when various drugs are coadministered. Cisapride has therefore been withdrawn from the market in many countries and should only be used cautiously in severe cases. Alternative prokinetics such as metoclopramide and domperidone may be useful adjuncts, again when used in combination with acid-​suppression therapy. However, long-​term therapy with these agents is also not recom- mended due to safety concerns. Recent studies with other prokinetic drugs such as acotiamide and mosapride have yielded disappointing results. Endoscopic and minimally invasive antireflux procedures Several new techniques have been evaluated to restore barrier func- tion at the level of the lower oesophageal sphincter. These include endoscopically delivered sutures (Esophyx), an electrical stimu- lator device (Endostim), radiofrequency energy delivery (Stretta), magnetic bead implantation (Lynx), and submucosal prosthetic im- plants. Effectiveness has mainly been assessed in uncontrolled trials without long-​term follow-​up data. NICE is currently monitoring the evidence for the electrical stimulator device and recommends that the use of other therapies (such as Lynx and Esophyx) is restricted to research studies until further evidence on long-​term safety and efficacy are available. Antireflux surgery Antireflux surgery is a very effective long-term therapy. Negative factors are the dependence of the results on the expertise of the surgeon and the morbidity and small (approximately 0.5%) mor- tality associated with the surgery itself. Laparoscopic antireflux surgery (usually a Nissen fundoplication) is a major advance, as it achieves good control of reflux with a major reduction in the morbidity inherent in the more traditional approach. The recent Lotus trial (a randomized controlled trial comparing antireflux surgery vs esomeprazole 20/​40 mg daily) showed similar clinical efficacy for the two arms, with significantly lower oesophageal acid exposure in the surgical group (0.7 vs 1.9%, respectively), but at the expense of increased side effects (dysphagia, flatulence, and bloating) Choice between medical and surgical therapies Selection of a medical or surgical therapy should take account of the severity of disease and the risks of antireflux surgery specific to the patient. It should also take account of the patient’s age, both from the point of view of operative risk and the time over which the patient will need treatment for reflux disease, the cost of ef- fective medical therapy, possible intolerance to PPI therapy, and, naturally, the preferences of the patient. The choice between med- ical therapies should be largely governed by the local cost of the alternatives that give the necessary level of treatment, as all of the first-​line options are effective, safe, and well tolerated. Anti-reflux surgery should not be routinely performed if any doubt persists, based on physiological studies, that gastro-oesophageal reflux be the primary cause of patients symptoms, as well as in the pres- ence of moderately to severely impaired oesophageal motility based on pre-operative manometry, as this can result in persistent dysphagia.

15.7  Diseases of the oesophagus 2833 Management of complications of reflux disease The important complication of oesophageal columnar metaplasia is discussed in a separate section (see ‘Oesophageal columnar meta- plasia (Barrett’s oesophagus)’). Abnormalities of oesophageal anatomy Non-​neoplastic abnormalities which distort oesophageal anatomy may interfere with normal function or merely pose difficulties in the interpretation of imaging findings. Sliding hiatus hernia Around 90% of hiatus hernias are of this type, in which the gastro-​ oesophageal junction is displaced upwards into the thorax, giving a simple-​shaped pouch of intrathoracic stomach. This can be mis­ taken for a columnar-lined oesophagus at endoscopy unless the ana- tomical land­marks are clearly defined. These landmarks include the proximal extent of the gastric folds and the palisading of vessels in the squamous oesophagus. The phreno-​oesophageal ligament is ef- faced in sliding hiatus hernia, but it is not clear whether this is a pri- mary defect of gastric anchorage. Many patients with hiatus hernia are asymptomatic. Despite this, physiological studies indicate that herniation of the gastro-​ oesophageal junction impairs its function as an antireflux barrier by removing the normal diaphragmatic crural compression from the lower oesophageal sphincter. Thus, hiatus hernia can be taken as a risk factor for reflux disease, but not an abnormality that makes the diagnosis. When present, symptoms of gastro-​oesophageal re- flux are the only ones of major significance. These should be treated along conventional lines (see ‘Gastro-​oesophageal reflux disease’). Prognosis is that of any associated reflux disease. Rolling or paraoesophageal hiatus hernia A variable part of the stomach herniates through the hiatus along- side a normally situated gastro-​oesophageal junction. This pattern of herniation may produce a gross disturbance of gastric anatomy, usually with a narrow exit from the herniated pouch into the main stomach cavity. Some rolling hernias are also associated with dis- placement of the gastro-​oesophageal junction above the hiatus, in which case these are known as mixed hernias. Obstruction and distension of the pouch causes upper abdom- inal discomfort and can progress to strangulation. Trapping of food bolus into the hiatus hernia can cause symptoms of post- prandial fullness, epigastric discomfort and early satiety. Gastric volvulus can occur because of the laxity of the gastric anchorage and may obstruct the gastro-​oesophageal junction. Both of these problems have a very high mortality and demand urgent surgery (Fig. 15.7.4). There are no adequate data on the degree of risk associated with rolling hiatus hernia. It is controversial whether elective surgery should be performed to reduce and anchor rolling hiatus hernias in order to remove risks of complications. Schatzki ring A Schatzki ring is a characteristic short luminal stenosis which oc- curs at the gastro-​oesophageal junction. It is made up only of mucosa and submucosa, and may narrow the lumen to a few millimetres or cause a clinically insignificant minor indentation. Aetiology is un- known: they have been postulated to be congenital, although evi- dence now suggests that they can occur in later life. Reflux and pill oesophagitis have been implicated. With mechanically significant rings, intermittent dysphagia occurs on eating solids. Meat is often the culprit, leading to the common term of ‘steakhouse syndrome’. Episodes of bolus obstruc- tion are not unusual, with associated chest pain caused by powerful oesophageal contractions. Failure to recognize a Schatzki ring fre- quently leads to the incorrect diagnosis of primary diffuse oesopha- geal spasm. Endoscopy is the investigation of choice (Fig. 15.7.5). If barium studies are performed, adequate distal oesophageal Fig. 15.7.4  Mesenteroaxial volvulus on upper gastrointestinal examination. A supine image shows the distal stomach located in a paraoesophageal hiatal hernia. There is a rotation about the short axis in the body of the stomach (arrow). A nasogastric tube is in place. Note the downward-​pointing pylorus. From Levy AD, Mortele KJ, Yeh BM (eds) (2015). Gastrointestinal imaging. By permission of Oxford University Press. Fig. 15.7.5  Endoscopic appearance of a Schatzki ring. Courtesy of Dr Ewen Cameron, Consultant Gastroenterologist, Addenbrooke’s Hospital, Cambridge.

SECTION 15  Gastroenterological disorders 2834 distension during the barium swallow is essential for detection; this is best achieved by prone-​oblique views. No treatment may be necessary. Disruption of the ring by simple peroral dilatation or endoscopic diathermy or laser is very re- warding as the dysphagia and chest pain are cured, sometimes after many years of symptoms. However, there is a significant incidence of recurrence and repeated dilatations at intervals are often needed. Other rings and webs Other short oesophageal stenoses may develop because of peptic stricture, muscular rings, and cervical webs with iron-​deficiency an- aemia (Plummer–​Vinson syndrome) or without. Wide-​mouthed multiple diverticula are characteristic of sclero- derma oesophagus. In the nonsclerodermatous oesophagus, diver- ticula occur in the mid and distal oesophagus, both types probably being ‘blow-​outs’ secondary to hypercontraction motor disorders. These can become very large, but it is rare for them to cause symp- toms, although they may be associated with dysphagia and regur- gitation of retained contents. Unless symptoms are disabling, they are best left undisturbed because leakage is common following surgical removal. Multiple intramural outpouchings of barium are characteristic of intramural pseudodiverticulosis, which appears to be due to dilata- tion of the ducts of submucosal glands by an unknown process. Peptic stricture Dysphagia secondary to stricture formation needs to be distin- guished from the more common dysphagia seen in patients with reflux disease which is due to defective triggering and control of oesophageal body peristalsis (see ‘Nonspecific oesophageal motor disorders’). Peptic stricture is managed by a combination of peroral dilatation and healing of oesophagitis by either medical or surgical means. All strictures should be treated as malignant until proven otherwise by repeated biopsy. Provided oesophagitis is healed, stric- ture is usually not an ongoing problem. Respiratory complications Respiratory disease may occur either as a result of direct aspiration of refluxed gastric contents or from the effects of gastro-​oesophageal reflux on reflexes. It is difficult to prove that reflux disease which coexists with respiratory disease is actually the cause of the respira- tory problem. The best investigative approach is probably a trial of high-​level acid inhibition with at least a double dose of PPI and prokinetics for at least 2 months. Management of respiratory dis- ease by antireflux surgery is not guaranteed to be successful. The evidence on the link between reflux disease and chronic cough re- mains conflicting, although in same cases this can be demonstrated by pH-impedance. Regurgitation Voluminous regurgitation is the main symptom in a small subgroup of patients with reflux disease. They may present complaining of vomiting, but a detailed history reveals that there is no prior nausea, and no effort involved in the appearance of the gastric content in the mouth. The determinants of high-​volume reflux and regurgitation have not been well defined. Treatment with PPIs and prokinetics can have substantial benefits, but in more severe cases antireflux surgery is usually the only effective management. Chest pain: noncardiac Reflux is also an important cause of noncardiac chest pain (see ‘Noncardiac chest pain’). Oesophageal columnar metaplasia
(Barrett’s oesophagus) Definition and nomenclature Oesophageal columnar metaplasia (Barrett’s oesophagus) is defined as the conversion of the normal stratified squamous epithelium with a columnar-​lined mucosa which may have characteristics of gas- tric or intestinal epithelium. The lack of international consensus on the definition of this disease has led to confusion in the literature. However, it is widely accepted that it is the subtype characterized by intestinal metaplasia with goblet cells which has the highest risk for malignant progression and in many countries the term Barrett’s oesophagus and/​or the recommendation for surveillance has been restricted to this subtype. The extent of the metaplasia may vary from 1 cm to the entire oesophageal length and there appears to be a correlation between the length of the segment, the proximal extent of reflux exposure, and the risk of malignancy transformation. The metaplasia generally starts in the distal oesophagus although dis- crete columnar islands may occur. The accepted nomenclature for describing the endoscopic features is the Prague classification in which C indicates circumferential involvement in centimetres and M is the maximal disease extent in centimetres (e.g. C2, M5 for a 5 cm segment with 2 cm of confluent columnar epithelium and 3 cm of metaplastic tongues) (Fig. 15.7.6). Aetiopathogenesis Columnar metaplasia occurs in the context of chronic gastro-​ oesophageal reflux, and occurs more commonly in white males. Its incidence appears to have increased, even taking into account the more frequent use of endoscopy, and the reasons for this are still under intense scrutiny. Possible explanations include an increase in obesity leading to an increased susceptibility to reflux and the re- duced prevalence of Helicobacter pylori infection with a consequent increase in gastric acid secretion. Columnar metaplasia carries an increased risk for the development of oesophageal adenocarcinoma compared with the general population, which equates to an annual incidence of 0.1 to 0.4% in nondysplastic Barrett’s, depending on the length of the segment and the presence of intestinal metaplasia. The risk for cancer significantly increases with the presence of dysplasia (up to 10% per annum) and endoscopic surveillance is therefore generally advocated. Management Surveillance hinges on the subjective interpretation of dysplasia in multiple, random biopsies (usually quadrantic biopsies every 2 cm according to the Seattle protocol). The frequency of surveillance depends on the degree of dysplasia (Table 15.7.1). This arduous

15.7  Diseases of the oesophagus 2835 protocol has not been uniformly accepted since no data from ran- domized controlled trials support a reduction in population mor- tality from oesophageal adenocarcinoma using this approach. On the other hand, the outcome for individuals with symptom- atic invasive adenocarcinoma is universally poor compared with surveillance-​detected disease. Failure to discuss the risk for adeno- carcinoma and the option of endoscopic surveillance with a patient who has oesophageal columnar metaplasia could well be viewed as an indefensible lapse of practice, despite the uncertainties about cost-​effectiveness. The treatment of dysplasia in oesophageal columnar metaplasia has changed substantially in the past 10 years with the advent of ef­ficacious and safe endoscopic therapies. The key is to establish the correct diagnosis, since dysplasia is notoriously difficult to diagnose, with high levels of inter-​ and intraobserver variability. A diagnosis of dysplasia should therefore be corroborated by two independent pathologists, and the addition of p53 immunohistochemistry can be a useful diagnostic adjunct to the traditional histopathological diag- nosis (Table 15.7.1). The management of confirmed dysplasia should be discussed with the multidisciplinary team for oesophagogastric cancer, and the pa- tient should have an opportunity to discuss the options available to them. Increasingly, endoscopic mucosal resection can be used as a diagnostic adjunct for visible lesions to determine depth of invasion. This will inform management decisions and in some cases will turn out to be a therapeutic local excision. Endoscopic ultrasound may be useful but there is a danger of overestimating the depth of in- vasion. A systematic review has showed that in early oesophageal adenocarcinoma, endoscopic ultrasound staging corresponds to the pathological staging only in just over half of the cases, hence this technique is not recommended routinely in patients with high-​ grade dysplasia and early oesophageal adenocarcinoma. Management options will be determined by the extent of dys- plasia and the fitness of the individual, as well as local expertise. Oesophagectomy (which may be performed as a transhiatal or a lap- aroscopic procedure with a limited lymph node resection) is now generally reserved for invasive disease extending into the submucosa because endoscopic resection of localized lesions coupled with ab- lative therapy, generally using radiofrequency ablation, is now sup- ported by randomized control trial evidence with encouraging data on durability. For cases with confirmed low-​grade dysplasia, abla- tion therapy can also be offered following discussion with the patient about the risks and benefits of this procedure. Endoscopic ablation is generally very safe and effective, but multiple treatment sessions are usually required, especially for long segments, with a risk of stricture formation (5–​9%) requiring dilatation. Barrett’s oesophagus may also be associated with deep oesopha- geal ulceration within the columnar-​lined segment and strictures usually at the squamocolumnar junction, but these should be con- sidered malignant until proven otherwise. The diagnosis and management of invasive oesophageal adeno- carcinoma are discussed in the malignancy section. Nonreflux causes of oesophagitis Infective oesophagitis Infective oesophagitis has become more prevalent with the increasing number of people who are immunosuppressed through HIV infec- tion or chemotherapy. The more important causes of infective oe- sophagitis are summarized in Table 15.7.2. Immune status is a major determinant of the pattern of infection and simultaneous infection with two or more infective agents is not unusual (Table 15.7.2). H. pylori does not appear to be of any primary significance in the pathogenesis of oesophageal mucosal disease. Patients may present with pain or dysphagia. Viral oesophagitis can sometimes cause major haemorrhage. The dysphagia is gener- ally secondary to superficial mucosal damage and inflammation, but some disorders damage the full thickness of the oesophageal wall and so lead to stricturing. (b) (a) Fig. 15.7.6  Endoscopic appearance of Barrett’s oesophagus. (a) Columnar mucosa in the distal oesophagus on white light
imaging. (b) Corresponding image obtained with narrow band
imaging (note tongues and islands of pink mucosa amid pale white squamous mucosa).

SECTION 15  Gastroenterological disorders 2836 A full history to determine the setting in which the oesophageal problem occurs is often very helpful. Cutaneous or oral disease can inform the oesophageal diagnosis. Endoscopy is the investigation of choice since the mucosal appearance and the distribution of oe- sophageal lesions can be virtually diagnostic. In addition, biopsies and brushings allow for histological diagnosis and identification of fungal elements, viral inclusions, or rarely, pathogenic bacteria. Although infective oesophagitis may be severe in immunocom- petent patients, it is characteristically self-​limiting and topical therapy is normally all that is needed. Immunocompromised pa- tients usually need aggressive, systemic therapy to resolve the infec- tion (Table 15.7.2). Some infections tend to recur, which can cause major disability. Eosinophilic oesophagitis Eosinophilic oesophagitis (in the past sometimes referred to as al- lergic oesophagitis) affects children and adults worldwide, with a male preponderance. The incidence has increased in recent years in line with other associated allergic conditions such as asthma, hay fever, allergic rhinitis, and atopic dermatitis. Although the exact aetiology is not understood, there does seem to be an aberrant immune-​mediated response and activated eosinophils may cause activation of acetyl choline via histamine. During endoscopic exam- ination, oesophageal biopsies should routinely taken in all cases referred with dysphagia and with normal appearance of the oe- sophageal mucosa. Food impaction and dysphagia are the common presenting symp- toms. Heartburn is often present, but refractory to standard treat- ments for gastro-​oesophageal reflux disease. Approximately 50% of patients will have other allergic symptoms. Endoscopy is the investigation of choice and the findings range from grossly abnormal to normal. Findings include a small-​calibre oesophagus, proximal strictures, white exudates, oesophageal rings, longitudinal furrows, and fragile mucosa. However, the sensitivity of endoscopic findings for this diagnosis is suboptimal (15–​48%); hence, in case of clinical suspicion histological corroboration is man- datory. The diagnosis is made on oesophageal biopsy in which there is a dense eosinophilic infiltration within the epithelium (>15 per high power microscopy field) in both proximal and distal biopsies; and exclusion of other possible causes including proton pump-​responsive oesophageal eosinophilia (Fig. 15.7.7). If the eosinophilic infiltration is more generalized, this suggests eosinophilic gastroenteritis. Table 15.7.1  Surveillance protocol for columnar-​lined (Barrett’s) oesophagus for patients willing and fit to undergo regular endoscopy Histopathological diagnosis Endoscopy frequency Clinical management Intestinal metaplasia No dysplasia 2–​5 years depending on the length of Barrett’s and other risk factors for cancer Treat heartburn symptoms with PPI Intestinal metaplasia Indefinite intraepithelial neoplasia/​dysplasia 6 months after index finding. Annual if changes persist. Revert 2-​yearly when two consecutive endoscopies no dysplasia Ensure adequate acid suppression to prevent inflammation confounding diagnosis Low-​grade intraepithelial neoplasia Confirm diagnosis with assessment by two independent pathologists. Rescope in 6 months if treatment not instigated. Ensure adequate acid suppression to prevent inflammation confounding diagnosis Offer ablation treatment after two endoscopies with low-​grade dysplasia. Prior to treatment consider external pathologist for confirmation of diagnosis High-grade dysplasia and intramucosal adenocarcinoma Confirm diagnosis with assessment by two independent pathologists and discuss treatment options Consider endoscopic therapy—​method depending on comorbidity and focality of lesion If endoscopic resection shows disease extending into the submucosa, surgical resection is recommended Table 15.7.2  Major causes of infective oesophagitis Pathogen Management Remarks Immunocompetent patients Candida albicans Topical/​oral antifungals By far the most common, often associated with inhaled steroid use Herpes simplex Aciclovir if severe Unusual, may denude mucosa Varicella zoster Aciclovir if severe In association with chickenpox/​herpes zoster Bacteria Rare in well individuals, and microbiome of healthy individuals becoming
better characterized Immunocompromised patients Candida albicans Systemic antifungals Most common; oral disease almost diagnostic Cytomegalovirus Prophylaxis and treatment with ganciclovir or foscarnet Serpiginous to giant ulcers in distal half Herpes simplex Prophylaxis and treatment with aciclovir or foscarnet Circumscribed ulcers, raised edges to coalescence. Oral lesions Tuberculosis Conventional From miliary and local spread Gram-​positive cocci,
Gram-​negative bacilli Intravenous antibiotics Often with systemic infection Syphilis Conventional Associated with tertiary syphilis elsewhere. Inflammatory stricture

15.7  Diseases of the oesophagus 2837 There is now robust evidence from randomized controlled trials that topical steroid (swallowed fluticasone or budesonide) is su- perior to placebo for symptom control and rate of histological re- mission, and therefore is recommended by the American College of Gastroenterology as first-​line treatment. There is also evidence from cohort studies that foods such as milk, soy, egg, nuts, fish, and wheat can trigger the eosinophilic infiltration, and their elimination im- prove symptoms and histological findings. Hence, alternative strat- egies may include dietary elimination in consultation with an allergy assessment and steroids given topically or systemically in severe cases. However, given the lack of objective tests for the assessment of food allergy, the identification of specific foods requires multiple endoscopies and biopsies, making topical therapy with steroid more practical routinely. Endoscopic food disimpaction and dilatation may be necessary. There are some data on leukotriene inhibitors and anti-​interleukin-​5 antibody therapy, but conflicting evidence on their clinical efficacy and limited data on the long-​term outcome for these patients. Given the high rate of recurrence after discontinuing the treatment, maintenance therapy with topic steroids or a restric- tion diet should be considered in all patients, but particularly those with severe symptoms and rapid relapse after discontinuation of ini- tial treatment. Medication-​induced oesophagitis This entity was only recognized in 1970. The chemical properties of medications pose hazards to the oesophageal mucosa because of its relative susceptibility to injury through pH-​dependent mechanisms. This susceptibility arises in part from the high local concentrations of medications that occur in the oesophageal lumen, since pills move surprisingly slowly through the normal oesophagus especially at the level of the aortic arch. Defective oesophageal transport, poor pill design, increased mucosal susceptibility to injury, and poor pill-​ taking technique contribute to the problem. Medications known to have an especially high risk for oesophageal damage are listed in Table 15.7.3. Symptoms are those for any form of stricture-​associated oesopha- gitis, and much pill-​induced injury probably goes unrecognized. Injury at the distal oesophagus, the other common site of hold-​up, may be commonly misdiagnosed as being due to reflux disease. Medications and formulations with a high risk of injury should be identified and avoided if possible, especially in older patients with re- flux disease or abnormal oesophageal transit. Pill transit is facilitated if medications are taken in the erect position with plenty of water. Pharmaceutical companies need to pay more attention to the use of shapes, sizes, and coatings that can assist transit of pills through the oesophagus. Stricturing may occasionally require surgery. Caustic injury See ‘Caustic ingestion’. Primary oesophageal motor disorders Idiopathic achalasia and achalasia-​like states Definition These disorders are characterized by increased lower oesophageal sphincter tone, absence of lower oesophageal sphincter relaxation with swallowing, and impairment of peristalsis of the oesophageal body. Idiopathic achalasia, which was first described over 300 years ago, accounts for more than 95% of cases and has an annual inci- dence of approximately 1 to 2 per 100 000. It affects all ages, but is diagnosed most often in early to mid-​adult life. Primary familial (b) (a) Fig. 15.7.7  Histopathological appearance of eosinophilic oesophagitis: Low power (a) and high power (b). Courtesy of Vicki Save, Lothian University Hospital, NHS Trust. Table 15.7.3  Common causes of medication-​induced oesophagitis Severe injury—​high risk Slow-​release potassium chloride Aspirin and nonsteroidal
anti-​inflammatory drugs Doxycycline/​tetracycline Quinidine Alendronate Potassium chloride Less severe injury—​high risk Many antibiotics Iron supplements Occasional injury Ascorbic acid Mexiletine Slow-​release theophylline Captopril Phenytoin Zidovudine Corticosteroids

SECTION 15  Gastroenterological disorders 2838 achalasia, which is genetically transmitted, accounts for less than 1% of cases. There are a number of causes of secondary or pseudo-​achalasia. The most common are due to malignant infiltration of the gastro-​ oesophageal junction which has been reported with carcinoma of the stomach, oesophagus, lung, pancreas, and prostate, and with lymphoma. It may also be a manifestation of paraneoplastic neural dysfunction. Chagas’ disease is an important cause worldwide and can sometimes accompany the intestinal pseudo-​obstructive syn- drome. There are a number of other secondary causes including oe- sophageal amyloidosis and sarcoidosis. Achalasia can also occur in associated with neurodegenerative diseases including Parkinson’s disease and cerebellar ataxia. Aetiology Impairment of inhibitory neural control of the distal oesophagus is the universal abnormality. The clearest evidence is degeneration of myenteric inhibitory neurons which, in the early stages, is associated with an inflammatory response. There is increasing evidence that the resulting nitric oxide deficiency may be causative. Symptoms Dysphagia with solids is almost universal, but the symptoms may extend to include liquids and regurgitation is also prominent. The regurgitated material tastes bland because it never enters the stomach. Cramping chest pain occurs in some patients during an early hypercontracting phase of the disorder. Weight loss is seen in patients with disabling dysphagia. The course of symptoms over time is variable in contrast to the progressive symptoms in patients with a malignant cause. Over a prolonged period of a hypertonic lower oesophageal sphincter, continuing oesophageal dilatation may result with increasing regurgitation. When this occurs, respira- tory problems secondary to aspiration can become a major feature. Diagnosis Idiopathic achalasia is diagnosed on average 2 years after its first presentation. Oesophageal manometry is the only sensitive method for demonstration of the characteristic motor dysfunction. The recent advent of high-​resolution manometry has significantly changed the diagnostic criteria and the classification of motor dis- orders. The most important manometric parameter for a diagnosis of achalasia is the integrated relaxation pressure, which measures both the efficacy and the duration of relaxation of the lower oe- sophageal sphincter. The new Chicago classification for motor dis- order describes three types of achalasia (I, II, and III). These are all characterized by increased integrated relaxation pressure, ac- companied by different oesophageal body motility abnormalities (failed peristalsis, pan-​oesophageal pressurization, or oesophageal spasms, respectively). It is not unusual for manometry to be diagnostic of achalasia even though barium studies have been judged to be normal. In advanced disease, a barium swallow may reveal gross oesopha- geal dilatation with a gastro-​oesophageal junction that tapers smoothly to a closed sphincter, with occasional spurts of flow into the stomach (Fig. 15.7.8). As noted earlier, idiopathic achalasia and achalasia-​like states should be distinguished from constriction of the gastro-​oesophageal junction by an infiltrating or encasing malignancy at the cardia. This can be difficult to ascertain with certainty. If there are any grounds to suspect malignancy then an endoscopy should be per- formed, including multiple biopsies from the gastro-​oesophageal junction. CT scanning may also be helpful in these circumstances. Management There are three potential approaches to treatment:  drug therapy with agents that relax the lower oesophageal sphincter, mechanical disruption of the sphincter by either pneumatic dilatation or sur- gical/​endoscopic myotomy, and pharmacological poisoning of the remaining excitatory nerves to the sphincter with botulinum toxin. The results of reduction of lower oesophageal sphincter pressure with drugs such as calcium antagonists and β-​adrenergic agonists compare poorly with mechanical techniques. Oesophagomyotomy can now be performed as an endo- scopic, laparoscopic, or thoracoscopic procedure. Randomized controlled trial evidence shows that laparoscopic myotomy has similar efficacy to pneumatic dilation, and both procedures have a risk of leading to troublesome gastro-​oesophageal reflux (10–​ 20%). In patients having surgery, this risk can be minimized by the incorporation of an antireflux procedure. Cohort studies showed that the myotomy can also be performed endoscopically via tunnelization in the submucosal space and endoscopic dissec- tion of oesophageal muscle fibres. Randomized controlled studies (a) (b) (c) Fig. 15.7.8  Chest radiograph and barium studies from a case of advanced achalasia.

15.7  Diseases of the oesophagus 2839 are currently being performed to compare endoscopic and laparo- scopic techniques. Balloon dilatation is an attractive approach because of its sim- plicity and low cost, but it often needs to be repeated and may fail in up to 40% of patients, especially those who are young. A re- cent meta-​analysis compared the outcome of balloon dilatation and surgical myotomy and found that 10-​year remission rates were 48% and 80%, respectively. It also carries a risk of perforation of about 2%. Given the associated risk of perforation, dilatation should not be performed in patients who are not fit for surgical intervention. Endoscopic injection of the sphincter with botulinum toxin acts on residual excitatory nerves, thereby lowering sphincter pressure. Even though the initial response rate is as high as 75%, the effects tend to wear off. Sustained response at 12 months is seen in about 40% of patients, and the treatment needs to be repeated at 6-​ to 24-​month intervals. The toxin is relatively expensive, however, it is a simple, low-​risk procedure and most applicable to patients with significant coexisting morbidity which renders them unfit for dila- tation or myotomy. Per-oral endoscopic myotomy is an attractive minimally invasive alternative, which was introduced by the Japanese endoscopist Dr Inoue in 2008. This involves the creation of a submucosal tunnel from the mid oesophagus until beyond the oesophagogastric junc- tion, division of the inner muscle layer with an endoscopic knife, and closure of the mucosal entry with clips. Promising results have been seen in cohort studies, but definite evidence of long-​term response is currently lacking and hence this procedure is cur- rently recommended in Europe only within research trials or strict auditing programmes. When oesophageal dilatation is present, prompt treatment is indi- cated because of the morbidity and poor therapeutic outcome asso- ciated with gross oesophageal dilatation. Prognosis Results are excellent if effective treatment is applied before the de- velopment of major dilatation, despite the persistence of major physiological abnormalities. Achalasia carries a significantly in- creased risk for oesophageal malignancy (squamous cell and, more recently appreciated, adenocarcinoma) up to three decades later. The prevalence ranges from 2 to 7% (or a standardized incidence ratio of around 10) in the most comprehensive reports. There is no apparent reduction of this risk with treatment. It is not usual practice to undertake surveillance for this condition. Diffuse oesophageal spasm Definition Episodic chest pain and/​or dysphagia resulting from abnormal contractions of the distal half of the oesophageal body in the ab- sence of any precipitating structural stenosis. On high-​resolution manometry, diffuse oesophageal spasm is characterized by normal relaxation of the lower oesophageal sphincter and recurrent prema- ture high-​amplitude contractions. Aetiology The aetiology of this disorder is poorly understood. Stress is an un- likely primary precipitant but may exacerbate the problem. Good prevalence data are lacking but it affects all age groups. Symptoms Virtually all patients have episodic, crushing central retrosternal pain which can be excruciating and may be misinterpreted as car- diac ischaemia. Intermittent dysphagia occurs in about two-​thirds of patients and leads to temporary abandonment of eating until symptoms abate. Episodes of oesophageal obstruction usually last for approximately 30 min but can last for several hours. In most pa- tients, symptomatic episodes occur less than once a month but in severe cases these may occur several times a week, or each time food intake is attempted. Diagnosis Due to the intermittent nature of the problem, investigations may be normal. Most frequently the diagnosis is made on the basis of the history and the exclusion of other problems that may mimic diffuse oesophageal spasm such as myocardial ischaemia and a Schatzki ring. On oesophageal high-​resolution manometry, the typical finding is normal integrated relaxation pressure and inter- mittent vigorous premature contractions, interspersed with normal swallow-​induced peristalsis. The normal relaxation of the lower oe- sophageal sphincter allows a differential diagnosis of achalasia to be excluded. Ambulatory 24-​h manometry may improve diagnostic accuracy by increasing the likelihood of capturing symptomatic episodes. Barium swallow may show trapping of contrast beads in the distal oesophagus—​the ‘corkscrew oesophagus’—​or sustained obliter- ation of the distal oesophageal lumen. This is not the investigation of choice. Management There is no specific therapy. Smooth muscle relaxants such as ni- trites, nitrates (given as sublingual spray), and calcium antagonists may reduce symptoms but their use is often limited by side effects. In many patients, reassurance is the most important management since the intensity and nature of symptoms gives rise to great con- cern. Opiate therapy is sometimes necessary. Botox injection into the oesophageal body can offer symptomatic relief, but—​similarly to achalasia—​the response is short-​lived. In the rare case of frequent, disabling spasm, oesophagomyotomy can give good relief. Prognosis The major significance is impairment of quality of life and concern about life-​threatening cardiac disease. There is no consistent progres- sion over time. There are several reports of progression of diffuse oe- sophageal spasm to achalasia but in most of these it seems likely that achalasia was initially misdiagnosed as diffuse oesophageal spasm. Hypercontractile (jackhammer) oesophagus Definition This is defined purely by the manometric demonstration in a symp- tomatic patient of primary peristaltic pressure waves in the oesopha- geal body, with at least two swallows, that have peaks of very high amplitude involving the majority of the oesophagus (distal con- tractile integral of >8000 mmHg/​s per cm in the high-​resolution manometry) (Fig. 15.7.9). There is preservation of the normal peri- staltic pattern with a broad progression of the time of onset of the contraction wave in the oesophageal body.

SECTION 15  Gastroenterological disorders 2840 Aetiology It is not clear if this is a true motor disorder or whether it represents the upper end of a continuum of peristaltic wave amplitudes. It has been shown to vary over time within individuals. There are indica- tions that psychological factors can influence peristaltic amplitude. A  minority of patients with hypertensive peristalsis also experi- ence episodes of diffuse oesophageal spasm, suggesting that their underlying dysfunction may be related and involve neural control mechanisms. Symptoms The only clinical significance of hypertensive peristalsis is its rela- tionship to noncardiac chest pain. Hypertensive peristalsis alone does not produce dysphagia or derangement of oesophageal transit, because, by definition, peristalsis is preserved. Management and prognosis These are discussed in the section on noncardiac chest pain. Nonspecific oesophageal motor disorders Definition These are departures from normal patterns of oesophageal motor function which do not actually define specific diseases, but which are of clinical significance. These are separate from oe- sophageal dysmotility arising in association with diseases such as gastro-​oesophageal reflux disease, diabetes, and other autonomic neuropathies. Nonspecific oesophageal motor disorder is the com- monest single functional diagnosis made in most oesophageal manometric laboratories. Aetiology There are likely to be several mechanisms involved. The intermittent occurrence of dysfunctions suggests that they are due to defective neural control. Symptoms Swallow-​induced distal oesophageal body contraction waves with multiple peaks stand out from the other patterns not only functionally but also symptomatically. This pattern is loosely asso- ciated with the hypercontraction disorders of diffuse oesophageal spasm and hypertensive peristalsis, but sometimes does not appear to have any clinical significance. Hypocontraction dysfunctions, re- cently termed ‘ineffective’ peristalsis, are associated with defective triggering and progression of both primary and secondary peri- stalsis. Failure to develop a propagated pressure wave of sufficient strength to maintain closure of the oesophageal lumen leads to de- ranged oesophageal transit. This probably explains the association of these disorders with mild intermittent dysphagia which occurs characteristically with solids. The nonobstructive dysphagia and slow oesophageal acid clearance seen in gastro-​oesophageal reflux disease are due to such dysfunction. Secondary oesophageal body peristalsis has not yet been widely evaluated, but it is probably an important cause of intermittent dysphagia, since, at least in patients with nonobstructive dysphagia and reflux disease, dysfunction of secondary peristalsis is substantially more common than primary peristaltic dysfunction. Oesophageal manometry with an adequate number of recording points in the oesophageal body is the only sen- sitive means for diagnosis. Management In most cases patients seek reassurance and an explanation about the origin of their symptoms. Prokinetic agents may improve triggering and amplitude of peristaltic contractions and so, theoretically, of transit. Secondary peristaltic dysfunction may be more trouble- some, but there is no good information on the effect of prokinetic or other drugs on this. Prognosis These dysfunctions do not remit spontaneously. Patients are often helped by the measures outlined in the section on general manage- ment of oesophageal dysphagia, which minimize the demands on oesophageal transport mechanisms and provide propulsive forces that substitute for oesophageal contractions. Noncardiac chest pain Definition Implicit in this rather circuitous and negative label is the view that this pain has a cardiac-​like quality, but there is no evidence for a car- diac origin. The oesophagus is the next most likely origin, but it is unlikely that all such pain arises from the oesophagus. Aetiology Evidence for triggering of pain by reflux or oesophageal motor dysfunction has been found in between one-​fifth and one-​half of patients evaluated. Oesophageal mucosal pain due to gastro-​ oesophageal reflux is the most common and helpful diagnosis. Frank oesophageal spasm associated with achalasia and diffuse oe- sophageal spasm is an unusual but convincing cause of noncardiac chest pain. In the majority of patients, most episodes of pain occur independently of reflux and any motor abnormality, although many of these patients have nonspecific oesophageal motor dis- orders or hypertensive peristalsis (see earlier sections). Sustained contraction of the longitudinal muscle has been identified by pro- longed intraluminal ultrasonography in association with a high proportion of episodes of pain. Nevertheless, in many patients Pharynx (mmHg) 80 0 400 0 Oesophageal body 0 Time (s) 400 0 400 80 0 0 30 60 90 120 Fig. 15.7.9  Oesophageal manometric tracing in a patient with hypertensive peristalsis.

15.7  Diseases of the oesophagus 2841 noncardiac chest pain appears to be a primary oesophageal hyper- sensitivity disorder and any motor disorder may be an epiphenom- enon. Recent work to understand the neurophysiological basis for the hypersensitivity suggests that there may be distinct phenotypic subclasses of disease based on enhanced afferent transmission de- fects versus heightened secondary cortical processing. Symptoms By definition, the pain resembles cardiac pain in its sensation and distribution. It can be very intense and distressing, can disturb sleep, and may be worse during periods of emotional stress. Postprandial occurrence, in association with heartburn, suggests that it may be caused by reflux. When pain is associated with dysphagia, vigorous achalasia or oesophageal spasm are very possible. Diagnosis Myocardial ischaemia should first be excluded as the cause. Endoscopy should then be performed followed by oesophageal pH and motility studies when symptoms are disabling. Investigation can be unrewarding. Management Reassurance is essential to prevent repeated hospital admis- sions for fear of a cardiac cause. If the pain is triggered by gastro-​ oesophageal reflux, high-​level acid-​suppression therapy should be tried (see ‘Gastro-​oesophageal reflux disease’). Achalasia and dif- fuse oesophageal spasm should be treated on their own merits. In patients with no clear-cut diagnosis, treatment with anxiolytics and low dose antidepres­sants has been found to be moderately effective, given their properties in reducing nociception. Agents that reduce the strength of oesophageal contraction, such as calcium antagon- ists, appear ineffective in hypertensive peristalsis. Oesophageal motor disorders secondary to systemic disease Oesophageal motility may be affected by a number of systemic diseases (Table 15.7.4). These diseases may affect the striated or smooth muscle itself or the neural control. The division of the oesophageal musculature into striated and smooth muscle components is revealed clearly by the myopathic diseases that affect the oesophageal musculature. In patients with peripheral myopathy this would normally have already been diag- nosed. Weak or absent oesophageal contraction in the affected seg- ment has the expected adverse impact on oesophageal transit, with a pattern of symptoms similar to the hypocontraction states of nonspecific oesophageal motor disorders (see earlier discussion). The management of these dysfunctions is along general lines (see ‘General management of oesophageal dysphagia’). Table 15.7.4  Systemic diseases associated with disturbance to oesophageal symptoms Systemic diagnosis Oesophageal symptom(s) Pathophysiology Rheumatological disorders Systemic sclerosis (limited and diffuse types) Dysphagia, heartburn, regurgitation, weight loss Oesophageal dysmotility, lower oesophageal sphincter incompetence, oesophagitis, columnar-​lined oesophagus Mixed connective tissue disease Dysphagia Oropharyngeal and body dysmotility Sjögren’s syndrome Dysphagia Xerostomia, oesophageal mucosal dryness Polymyositis/​dermatomyositis Dysphagia, nasal regurgitation, aspiration Oropharyngeal and body dysmotility Systemic lupus erythematosus Dysphagia, odynophagia Hypotone of the lower oesophageal sphincter, oesophageal
body dysmotility, pill oesophagitis Seronegative spondyloarthropathies Dysphagia Oesophageal body dysmotility Rheumatoid arthritis Dysphagia, odynophagia Oesophageal body dysmotility, oesophageal vasculitis, atlantoaxial subluxation Neuromuscular disorders Dysphagia, chronic aspiration, disordered phonation Oropharyngeal and oesophageal dysmotility Myotonic dystrophy Dysphagia Oropharyngeal and oesophageal dysmotility Myasthenia gravis Dysphagia Oropharyngeal and oesophageal dysmotility Chronic intestinal pseudo-​obstruction Dysphagia Oesophageal dysmotility Other disorders Vasculitic syndromes Dysphagia, odynophagia, chest pain, haematemesis Oesophageal vasculitis, ulcers, pharyngeal stenosis Diabetes mellitus Dysphagia Oesophageal dysmotility Alcohol abuse Dysphagia, odynophagia, heartburn Oesophageal dysmotility, oesophagitis, varices and haematemesis Infiltrative disorders (amyloidosis, sarcoidosis) Dysphagia Oesophageal dysmotility

SECTION 15  Gastroenterological disorders 2842 Diseases of oesophageal smooth muscle Systemic sclerosis (scleroderma) Definition and aetiology Sixty per cent of cases are limited cutaneous scleroderma, previ- ously called CREST (calcinosis, Raynaud’s syndrome, (o)esopha- geal dysphagia, sclerodactyly, telangiectasia) syndrome. The remaining 40% of cases are now termed diffuse cutaneous sclero- derma, with widespread involvement of other organs apart from the skin. The timing of onset of symptoms from oesophageal in- volvement are very variable in relation to other manifestations but are sometimes the presenting complaint. Oesophageal muscle atrophy and fibrosis are the cardinal features, but neuropathic ab- normalities may also contribute to dysfunction. Smooth muscle peristalsis and the tone of the lower oesophageal sphincter are feeble or absent (Table 15.7.4). Symptoms Troublesome reflux symptoms are the most common consequence of loss of function. The pattern of dysphagia resembles that seen in nonspecific oesophageal motor disorder (see ‘Nonspecific oesopha- geal motor disorders’). If dysphagia is severe, peptic stricture should be excluded, as complete loss of oesophageal smooth muscle peri- stalsis rarely leads to disabling dysphagia. Treatment Reflux disease is frequently severe and should be managed by high-​level medical therapy in order to prevent complications such as stricture (discussed earlier). Antireflux surgery is relatively contraindicated because of the poor propulsive function of the oesophageal body. Other rheumatological disorders A scleroderma-​like picture of oesophageal dysfunction is some- times seen in other connective tissue disorders such as mixed connective tissue disease. The smooth muscle segment is also involved in systemic myopathies including polymyositis–​ dermatomyositis and myotonic dystrophy. It should be remem- bered that in addition to effects on motility, rheumatological disorders may also lead to oesophageal symptoms via a combin- ation of effects including mucosal dryness and associated reflux disease (Table 15.7.4). Other disorders Abnormal oesophageal motility is common in diabetes mellitus, and may be a feature of amyloidosis, chronic alcoholism, and the pseudo-​obstructive syndrome. In these disorders, the disturbance is believed to be primarily due to dysfunction of neural control mechanisms. Disorders of striated muscle Involvement of the striated muscle segment of the oesophagus is rare and patients usually present with high dysphagia, often in association with oropharyngeal dysfunction. The inflammatory myopathies (dermatomyositis, polymyositis, and inclusion body myositis), the muscular dystrophies (myotonic dystrophy and oculopharyngeal dystrophy), and myasthenia gravis are the most common causes. General management of oesophageal dysphagia Symptomatic treatment of dysphagia is frequently necessary be- cause of the limited options and efficacy of specific treatments for oesophageal disorders. Although these measures may appear ob- vious, this aspect of management is commonly neglected by both patient and physician. Optimization of bolus consistency Large particles of solid food may impact on strictures. Large boluses require greater propulsive force even in the absence of stricture, and may trigger oesophageal spasm. Boluses should therefore be small and, in some circumstances, reduced to semiliquid or liquid form. Poor dentition should be treated. In some patients, defects of oe- sophageal function may be so severe that the diet should be puréed. Consultation with a dietitian will assist patients in identifying and preparing suitable food and in maintaining nutrition. Assistance with oesophageal transit Liquids assist transit by reducing the viscosity of food and providing a pressure head in the oesophagus. Gas generated within the oesopha- geal body from effervescent drinks can act as a piston which displaces oesophageal contents into the stomach in the erect position and may be sufficient to overcome an achalasic sphincter. The value of gravity in assisting transit should never be forgotten. Patients with severely impaired oesophageal transit should be advised to swallow medica- tions in the upright position and with plenty of water so as to avoid injurious contact of the oesophagus with potentially corrosive tablets. Alternative/​supplementary approaches to feeding Rarely, the earlier mentioned measures fail to maintain nutrition. Percutaneous endoscopic gastrostomy should then be used. Oesophageal neoplasms The two most frequently occurring malignancies of the oesophagus are adenocarcinoma and squamous cell carcinoma, defined ac- cording to the histopathological characteristics. Occasionally mixed cell types are seen. Cure is only possible in the small minority (<20%) of patients whose disease presents early. Several approaches are possible for palliation, although data are somewhat conflicting about the relative merits of each. Oesophageal adenocarcinoma and tumours of the gastro-​oesophageal junction Definition Over 80% of adenocarcinomas arising in the oesophagus occur in association with oesophageal columnar metaplasia, or Barrett’s oe- sophagus (Fig. 15.7.10). In the cases with no coincident Barrett’s, the oesophageal mucous glands or the adjacent gastric cardia mucosa are presumably the source of malignant change. Aetiology As discussed previously, columnar oesophageal metaplasia occurs in the context of chronic gastro-​oesophageal reflux disease, often manifest as heartburn symptoms. The incidence of oesophageal

15.7  Diseases of the oesophagus 2843 adenocarcinoma has increased rapidly in Western countries over the past 20 to 30 years, surpassing the incidence of squamous cell car- cinoma in these populations. The disease has a marked male prepon- derance. An increase in the prevalence of reflux disease related to the rise in obesity and reduced prevalence of H. pylori infection, with con- sequent increase in gastric acid secretion, are plausible explanations but not proven. In approximately 7% of cases, Barrett’s oesophagus and oesophageal adenocarcinoma occur within familial clustering, al- though the genetic causes remain unknown and more likely related to the cumulative effect of multiple low-risk genetic polymorphisms. Given the lethality oesophageal adenocarcinoma at symptomatic presentation, endoscopic screening should be considered in the pres- ence of multiple risk factors (obesity, reflux symptoms, male sex and white race), particularly in the presence of positive family history. Clinical features Dysplasia and carcinoma in situ are asymptomatic and this has led to discussions about screening in high-​incidence areas of the world. Inexorable progression of dysphagia to solids, and eventu- ally to liquids, over several weeks is the almost universal presenta- tion. Dysphagia usually occurs only when the tumour has become circumferential or bulky. Sometimes, malignant mucosal ulceration presents with pain. Substantial weight loss has often occurred by the time of presentation. Endoscopy and biopsy is the first investigation of choice. The lesion may be stricturing or exophytic. Multiple biopsies should be taken to avoid a missed diagnosis due to sampling bias. In stricturing lesions, dilatation for diagnostic purposes should be avoided if possible due to the risk of perforation. Biopsies can usu- ally be taken from the proximal portion of the stricture even if in- tubation is limited. If performed, a barium swallow typically reveals a stricture with an irregular, lobulated mucosal outline. However, it is important to note that occasionally the barium appearance mimics a benign peptic stricture, and with modern endoscopy enabling intubation under direct vision, prior barium examination is no longer required. Occasionally, an asymptomatic oesophageal carcinoma is diag- nosed when endoscopy is done for some other reason. Early lesions arising in Barrett’s mucosa may be inconspicuous endoscopically or be associated with a nodule or ulcer. Staging Staging using the TNM system is essential for all patients fit for radical therapy in order to optimize management. The staging mo- dalities and criteria of gastro-​oesophageal junction tumours will depend on whether the cancer is thought to arise primarily in the oesophagus or proximal stomach. CT scanning is very useful to determine nodal status and the presence of any distant metastases which most commonly affect the liver, lungs, adrenal glands and peritoneum. In order to obtain the best views the stomach should be distended with water before scanning (Fig. 15.7.11). In the absence of metastatic disease on the CT scan, endoscopic ultrasound should be performed to determine the degree of invasion through the oe- sophageal wall (T stage) (Fig. 15.7.12). Mediastinal or abdominal lymph nodes can also be assessed and sampled if necessary using fine needle aspiration. Positron emission tomography can provide fur- ther information on whether lymph nodes are involved and to char- acterize distant spread. Ideally this is now increasingly performed in combination with CT (CT-​PET). Bronchoscopy or thoracoscopy can be useful to look for signs of airway infiltration in proximal dis- ease and combined with fine needle aspiration if required. For tu- mours with a significant gastric component, a laparoscopy should also be performed to document whether there is involvement of the serosa and peritoneum. Management For disease limited to the mucosa, endoscopic therapy may be suf- ficient as the risk of nodal involvement is lower than the mortality rate of an oesophagectomy (<2%). In early-​stage disease extending into the submucosa and beyond, surgery is the treatment of choice as it achieves high rates of cure. The surgical approach depends on the location of the tumour. A transthoracic approach is standard (Ivor Lewis) but increasingly the operation is performed wholly or partially laparoscopically, with improved recovery times. Other options such as a transhiatal or Merendino operation have a lower morbidity if a less radical lymph node resection is required. If lymph nodes are involved (N1) or if the cancer is invading through the muscularis and into the adventitia (T3) then multimodal therapy is usually required. By the time the disease is stage T2, it is usually associated with lymph node involvement. Multimodal therapy generally involves neoadjuvant chemotherapy, which in- creasingly is combined with radiotherapy prior to surgery if the (b) (a) Fig. 15.7.10  Barrett’s associated adenocarcinoma: (a) demonstrates the endoscopic view in which tongues of salmon pink columnar lined mucosa are seen extending beyond the exophytic tumour mass; (b) is the corresponding endoscopic ultrasound image of this case in which the tumour is seen to invade through the muscularis (T3).

SECTION 15  Gastroenterological disorders 2844 patient is fit enough, based on improved survival from randomized clinical trial data. Molecular targeted therapies have lagged behind other cancer types due to the heterogeneity of the disease and the lag-​time for clinical trial data. The precise treatment algorithm will depend on the details of the individual case and the local expertise. Combined modality therapies have significant morbidity and mor- tality and an expert multidisciplinary approach, which takes into account quality of life issues for the patient, is required to achieve the best results. When curative treatment is not possible, palliation of dysphagia poses many challenges and the field is hampered by a lack of crit- ical comparisons. Surgery is not a good palliative option because of its morbidity and mortality. Radiotherapy applied as external beam or internal brachytherapy can provide useful relief from dysphagia. There may, however, be a temporary worsening of symptoms due to mucositis and nasogastric feeding may be required over this period. Peroral dilatation of malignant strictures, except prior to stent in- sertion, should be avoided due to the risk of perforation. Peroral placement of stenting tubes (usually covered, to prevent tumour ingrowth), laser photocoagulation, argon plasma coagulation, in- jection of sclerosants, and radiotherapy (external beam or brachy- therapy) are all options for the management of dysphagia which have potential for improving the quality of life. Laser and injection therapies can be useful when there is an exophytic tumour compo- nent, but are not supported by a strong evidence base. Stent insertion has the advantage of being a once-​only treatment with immediate effect, although pain after insertion can sometimes be a problem. Alternative endoscopic therapies can provide more physiological improvements in swallowing. Oesophagopulmonary fistula is a distressing development which usually causes pneumonia and persistent cough and which can sometimes be controlled by stenting. Bleeding can be managed with radiotherapy and sometimes argon plasma coagulation, or laser. (a) (b) Fig. 15.7.11  (a) Transaxial CT staging of a type 3 tumour of the gastro-​ oesophageal junction (primarily within the proximal stomach and extending into the oesophagus). (b) Reformatted view of a gastro-​oesophageal junction tumour with extensive involvement of lower oesophagus extending into the proximal stomach with proximal oesophageal dilatation. Courtesy of Dr Nicholas Carroll, Consultant Radiologist, Addenbrooke’s Hospital Cambridge. (a) (b) Fig. 15.7.12  Lugol’s iodine spray reveals early squamous cell carcinoma as seen by white light (a). On narrow-band imaging with magnification the irregular vascular pattern can be highlighted and suggests superficial mucosal invasion (b).

15.7  Diseases of the oesophagus 2845 Embolization of relevant arteries should also be discussed with an interventional radiologist. Palliative chemotherapy can increase survival and help symptoms in carefully selected patients fit enough to withstand the treatment. This can be combined with HER2 inhibitors in ERBB2 amplified cases with a significant gastric component. Other critical aspects of palliative care include pain control as well as nutritional advice and support (e.g. oral supplements) which are best done within the con- text of an expert multidisciplinary team. Prognosis This remains dismal except where screening programmes identify early, asymptomatic cases. In patients presenting with symptomatic disease, only about one-​quarter of patients are deemed to be poten- tially curable by surgery, and the overall 5-​year survival rate is ap- proximately 13% for all comers, increasing to 30% in those treated with curative intent. Squamous cell carcinoma Squamous cell carcinoma has marked geographical variation. It is common in the developing world with an annual incidence of 6 in 100 000 men and 1.6 in 100 000 women. There are areas of especially high incidence (northern China, northern Iran, Kazakhstan, and the Transkei region of South Africa with >35/​100 000 cases per year). Its incidence is stable in Western countries where it is six times more common in black men than in white men. Aetiology The striking geographical variation in incidence suggests a major aetiological contribution from environmental factors. The risk fac­tors include heavy alcohol use, tobacco, and dietary factors such as high rates of consumption of very hot beverage and dietary carcinogens (nitrosamines and aflatoxins). Other factors implicated are previous treatment for head and neck cancers, Plummer–​Vinson syndrome, human papilloma virus infection, vitamin A deficiency, chronic can- dida infection, injury to the oesophageal mucosa due to ingestion of a corrosive substance years previously, and chronic irritation from oe- sophageal retention in achalasia. Invasive carcinoma is preceded by mucosal dysplasia and carcinoma in situ and there may be a lag phase of many years which affords the possibility for screening in high-​risk regions. Some patients may be genetically predisposed to this cancer. The best-​documented example is tylosis, which is an autosomal dom- inant condition with associated palmar and plantar hyperkeratosis. Linkage studies suggest that the causative gene is IRHOM2 which res- ides on chromosome 17q25. Symptoms and staging The presentation and investigation algorithms for squamous cell carcinoma are essentially as described for adenocarcinoma. Patients may present to the ear, nose, and throat department if the dysphagia is very high or if hoarseness is a key symptom. Orolaryngosocopy may reveal the cause but flexible oesophagoscopy is usually required and should not be delayed in a patient with progressive dysphagia. In very high-​risk areas, screening programmes have been set up which generally hinge on standard endoscopy coupled with Lugol’s iodine spray to highlight dysplastic mucosa for targeted bi- opsies. Nonendoscopic sampling methods have generally had low sensitivity and specificity due to reliance on a cytological diagnosis of atypia (Fig. 15.7.12). Management The treatment is essentially similar for all oesophageal malignan- cies. Squamous carcinomas are generally more radiosensitive than adenocarcinomas and chemoradiotherapy can be given as definitive treatment either alone or combined with surgery as triple modality therapy. Again, the precise algorithm will depend on the details of the individual case and the local expertise. Unfortunately, for many patients palliative therapy is all that can be offered. Critical aspects of palliative care include pain control as well as nutritional advice and support (e.g. oral supplements) which can be problematic for prox- imal cancers and are again best done within the context of an expert multidisciplinary team. Prognosis As for adenocarcinoma, this remains dismal except where screening programmes identify early, asymptomatic cases. The overall 5-​year survival rate is less than 20%. Other primary oesophageal tumours Other primary malignant tumours are rare and all have a poor prog- nosis. These include malignant melanoma, lymphoma, carcinoid, leiomyosarcoma, neuroendocrine carcinoma (small cell carcinoma), adenoid cystic carcinoma, and pseudosarcoma reflecting the cell types present within the oesophagus. These tumours show a mix- ture of polypoid and infiltrating features and are usually only clearly distinguished from the more common malignancies by histology. Although rare in the oesophagus, gastrointestinal stromal tu- mours (which were previously classified as smooth muscle tumours) can also occur and have a varying degree of malignant potential. Approximately one-​half of patients are asymptomatic and the re- mainder exhibit symptoms which may include dysphagia, retro- sternal chest pain, pyrosis, cough, odynophagia, and weight loss. Bleeding is unusual, in contrast to gastric gastrointestinal stromal tumours. They are usually intramural but can become peduncu- lated and they usually only cause symptoms if they are very large, or on a long pedicle. On endoscopy the mucosa is intact but there may be central umbilication or ulceration. Because of the submucosal nature of the lesion, biopsy is often negative and endoscopic ultra- sonography is a useful diagnostic tool which can be combined with fine needle aspiration. The majority have a gain-​of-​function muta- tion of the proto-​oncogene growth factor receptor c-​Kit. The malig- nant potential is determined by the size, mitotic index, and precise mutation. Most small lesions do not require treatment. Surgery should be considered in symptomatic patients or when the risk of malignant transformation is high. Treatment with a tyrosine kinase inhibitor can be useful for metastatic disease and is tailored to the specific mutation. Benign intramural tumours of the oesophagus include lipomas, leiomyomas and granular cell tumours. The main risk of these is that they are mistaken for malignant tumours and operated on inappropriately. Squamous cell papillomas of the mucosa can mimic a polypoid squamous carcinoma and so should be removed endoscopically for histological diagnosis.

SECTION 15  Gastroenterological disorders 2846 Extrinsic oesophageal compression This is a relatively common cause of dysphagia, and is most often a result of malignant mediastinal lymphadenopathy. Barium swallow or endoscopy usually shows a relatively long constriction of the oesophageal lumen of variable calibre, associated with a normal mucosal appearance. Dilatation of such a compression is usually unrewarding because of its elastic recoil. Mechanically significant extrinsic compression may also result from an enlarged heart, a di- lated or unfolded aorta, or an aortic aneurysm. Kyphosis may ac- centuate the mechanical impact of these abnormalities. Mechanical changes along the cervical spine can also interfere with swallowing such as atlantoaxial spurs, osteophytes associated with osteoarth- ritis, and Forestier’s disease. In patients with rheumatoid arthritis, atlantoaxial joint subluxation can lead to dysphagia and other signs of spinal cord compression (Table 15.7.4). Congenital vascular ab- normalities can also compress the oesophagus in adults, an aber- rant right subclavian artery being by far the most common. Mechanical, chemical, and radiation trauma Mallory–​Weiss tear These mucosal tears extend across the gastro-​oesophageal junction and are normally induced by vigorous straining associated with vomiting. Bleeding is the only consequence of significance. In 10% of cases, bleeding is severe enough to cause hypovolaemia. The his- tory is usually quite characteristic, but definitive diagnosis requires endoscopy. Continued bleeding usually responds to endoscopic in­jection or haemostatic clipping, electrocoagulation, vascular em- bolization, or vasopressin infusion. Very rarely, surgery is needed to under-​run a persistently bleeding artery at the base of the tear. Barogenic oesophageal rupture (Boerhaave’s syndrome) In this uncommon but serious condition, straining and vomiting cause oesophageal rupture, most often in the left lower third of the oesophagus. High-​volume spillage of the gastric contents into the pleural space causes shock and pain in the chest and upper abdomen with radiation to the back, left chest, or shoulder. The chest radio- graph becomes abnormal only some hours after rupture. Surgical repair and drainage are usually necessary, and if this is delayed be- yond 24 h the mortality is very high. Unfortunately, diagnostic delay is not unusual. Iatrogenic oesophageal perforation Physicians encounter this problem most often as a result of their involvement in dilatation of oesophageal strictures, pneumatic bag dilatation for achalasia, endoscopic resection of early oesophageal tumours, or through problems with the management of oesopha- geal varices by balloon tamponade. Even with meticulous technique and appropriate equipment, oesophageal perforation can occur. Perforation is strongly suggested by development of chest or epigas- tric pain directly after instrumentation, sometimes with dyspnoea. Pneumothorax and surgical emphysema are diagnostic. Any suspi­ cion of perforation should be acted upon by performing CT scan- ning with Gastrografin. This can demonstrate the presence of air in the mediastinum or direct leakage of water soluble contrast. Broad-​ spectrum antibiotics should be given on suspicion, as they are most effective in minimizing the risks of mediastinitis when given from the outset. Surgical consultation should occur promptly; the choice between conservative and surgical management needs to be indi- vidualized. Increasingly, instrumental perforation is being managed nonsurgically with nasogastric suction, antibiotics, and intravenous nutrition with good results, primarily because instrumental injury usually occurs when the stomach is empty. Metal clipping can be applied if the perforation is promptly recognized during the endo- scopic procedure to aid mucosal healing. Caustic ingestion Definition and aetiology Strong acids and alkalis are both very damaging to the oesophagus and are found in high concentrations in many agents commonly used in the household for cleaning and maintenance. Laryngeal and gastric injuries may overshadow oesophageal injury. Because of their relative lack of taste, alkaline solutions are more likely to be swallowed ac- cidentally in large amounts. Alkaline injury is especially deep; acid tends to form a superficial coagulant, which limits penetration. Symptoms The severity and extent of injury are immensely variable and cannot be predicted accurately from estimates of the volume ingested. Around one-​half of patients with a history of caustic ingestion have no significant injury. Oropharyngeal and laryngeal injury confirm caustic ingestion and can be a major threat to the airway, but do not predict the existence and severity of oesophageal injury which causes odynophagia, dysphagia, or haematemesis. Prompt fibreoptic pan-​endoscopy appears to be safe. This may be normal or show only patchy mucosal oedema, erythema, and small haemorrhagic ulcers, indicative of superficial damage with a good prognosis. Extensive and circumferential ulceration, and grey or brown/​black ulceration suggest transmural injury. Management Patients with severe injury must be observed closely for signs of perforation. Nasogastric suction should be used with the admin- istration of broad-​spectrum antibiotics as these appear to reduce the severity of infective complications. The use of steroids is con- troversial, the balance of evidence tending to oppose their use. Oesophageal stricture is to be expected with severe injury and ap- pears not to be prevented by routine dilatation in the first 2 weeks after injury. A barium study should be done at 2 to 3 weeks to screen for stricturing, and then subsequently at about 3-​monthly intervals thereafter for a year, so that the development of stricturing is rec- ognized at a stage when dilatation may have some impact. Prognosis The main short-​ to medium-​term risk is the development of stric- ture. Caustic strictures are difficult and hazardous to treat by peroral dilatation so that about half of patients require oesophageal resec- tion. In the long term (average onset 40  years after injury), car- cinoma of the oesophagus is a major hazard, the risk being 1000 to 3000 times the expected risk.

15.7  Diseases of the oesophagus 2847 Oesophageal problems caused by oncological therapies Chemotherapy causes oesophageal problems in several ways. Therapy may impair mucosal defences by affecting cell turnover leading to ‘mucositis’. This in turn may reduce resistance of the mucosa to damage from other agents, and increase susceptibility to infective oesophagitis from immune suppression. Oesophageal transit and acid clearance may be impaired through the neuro- toxic effects of some agents. Fistulation or perforation may occur through cytotoxic effects on a malignancy in the oesophageal wall. It has been reported that combination chemotherapy is associ- ated with the development of oesophageal columnar metaplasia in women being treated for breast cancer. Radiotherapy as treat- ment of oesophageal malignancy or for lung, breast or other me- diastinal tumours can induce chronic oesophageal inflammation (radiation oesophagitis). This can cause dysphagia and chest pain. Endoscopic findings corroborated by the oncological history are sufficient for the diagnosis. However, radiation oesophagitis can occasionally cause stricture and mimic the endoscopic appearance of oesophageal cancer. In these cases endoscopic biopsies help in the differential diagnosis. Other non-​neoplastic mucosal diseases Skin and systemic diseases associated with lesions of the oropharynx may also involve the oesophagus. These include epidermolysis bullosa, Behçet’s disease, lichen planus, pemphigus vulgaris, bullous pemphigoid, benign mucous membrane (cicatrial) pemphigoid, and drug-​induced disease (Stevens–​Johnson syndrome and toxic epi- dermal necrolysis). Chronic, and less frequently acute, graft-​versus-​host disease may cause severe oesophageal problems through mucosal desquamation or mural damage. Resultant stricturing shows considerable variation in appearance. Rarely, Crohn’s disease can cause indolent, craggy ulceration and/​or stricturing. Oesophageal sarcoidosis can mimic Crohn’s disease. FURTHER READING Bhat S, et  al. (2011). Risk of malignant progression in Barrett’s esophagus patients:  results from a large population-​based study.
J Natl Cancer Inst, 103, 1049–​57. Blaydon DC, et  al. (2012). RHBDF2 mutations are associated with tylosis, a familial esophageal cancer syndrome. Am J Hum Genet, 90, 340–​6. Boeckxstaens GE, et  al. (2011). Pneumatic dilation versus laparo- scopic Heller’s myotomy for idiopathic achalasia. N Engl J Med, 364, 1807–​16. Boerwinkel DF, et al. (2014). The clinical consequences of advanced imaging techniques in Barrett’s esophagus. Gastroenterology, 146, 622–​9. Clouse R, Diamant N (2006). Motor function of the esophagus. In: Johnson L (ed.) Physiology of the gastrointestinal tract, 4th edi- tion, pp. 913–​26. Elsevier, Boston, MA. Cowgill SM, et al. (2007). Ten-​year follow up after laparoscopic Nissen fundoplication for gastroesophageal reflux disease. Am Surg, 73, 748–​52. Decalmer S, et  al. (2012). Chronic cough:  relationship between microaspiration, gastroesophageal reflux, and cough frequency. Chest, 142, 958–​64. Dellon ES, et al. (2013). ACG clinical guideline: evidenced based approach to the diagnosis and management of esophageal eosino- philia and eosinophilic esophagitis (EoE). Am J Gastroenterol, 108, 679–​92. Desai TK, et al. (2011). The incidence of oesophageal adenocarcinoma in non-​dysplastic Barrett’s oesophagus:  a meta-​analysis. Gut, 61, 970–​6. Desai TK, et al. (2012). The incidence of oesophageal adenocarcinoma in non-dysplastic Barrett’s oesophagus: a meta-analysis. Gut, 61(7), 970–6. Di Pietro M, et  al. (2015). Screening for Barrett’s Esophagus. Gastroenterology, 148(5), 912–23. Ellis FH Jr. (1998). Long esophagomyotomy for diffuse esophageal spasm and related disorders: an historical overview. Dis Esophagus, 11, 210–​14. Enzinger PC, Mayer RJ (2003). Esophageal cancer. N Engl J Med, 349, 2241–​52. Falk GWM, Fennerty B, Rothstein RI (2006). AGA Institute technical review on the use of endoscopic therapy for gastroesophageal reflux disease. Gastroenterology, 131, 1315–​36. Fang Y, et al. (2013). Cellular origins and molecular mechanisms of Barrett’s esophagus and esophageal adenocarcinoma. Ann N Y Acad Sci, 1300, 187–​99. Fitzgerald RC, Triadafilopoulos G (1997). Esophageal manifestations of rheumatic disorders. Semin Arthritis Rheum, 26, 641–​66. Fitzgerald RC, et al. (2014). British Society of Gastroenterology guide- lines on the diagnosis and management of Barrett’s oesophagus. Gut, 63, 7–​42. Frankell AM, et al. (2019). The landscape of selection in 551 esopha- geal adenocarcinomas defines genomic biomarkers for the clinic. Nat Gen, 51(3), 506–16. Galmiche JP, et  al. (2006). Functional esophageal disorders. Gastroenterology, 130, 1459–​65. Galmiche JP, et  al. (2011). Laparoscopic antireflux surgery vs esomeprazole treatment for chronic GERD: the LOTUS randomized clinical trial. JAMA, 305, 1969–​77. Giuliano C et al. (2012). Are proton pump inhibitors associated with the development of community-​acquired pneumonia? A  meta-​ analysis. Expert Rev Clin Pharmacol, 5, 337–​44. Harvey PR, et al. (2019). Outcomes of pneumatic dilatation and Heller’s myotomy for achalasia in England between 2005 and 2016, Gut, Jan 3. Hvid-​Jensen F, et al. (2011). Incidence of adenocarcinoma among pa- tients with Barrett’s esophagus. N Engl J Med, 365, 1375–​83. Hobson AR, et al. (2006). Neurophysiologic assessment of esopha- geal sensory processing in noncardiac chest pain. Gastroenterology, 130, 80–​8. Jankowski JAZ, et al. (2018). Esomeprazole and aspirin in Barrett’s oesophagus (AspECT): a randomised factorial trial. Lancet, 392(10145), 400–8. Kahrilas PJ, et al. (2015). The Chicago Classification of esophageal mo- tility disorders, v3.0. Neurogastroenterol Motil, 27, 160–​74. Kahrilas PJ (2017). Expert consensus document: Advances in the management of oesophageal motility disorders in the era of high- resolution manometry: a focus on achalasia syndromes. Nat Rev Gastroenterol Hepatol, 14(11), 677–88. Katz PO, et al. (2013). Guidelines for the diagnosis and management of gastroesophageal reflux disease. Am J Gastroenterol, 108, 308–​28.

SECTION 15  Gastroenterological disorders 2848 Lagergren J, et al. (1999). Symptomatic gastroesophageal reflux as a risk factor for esophageal adenocarcinoma. N Engl J Med, 340, 825–​31. Liacouras CA, et al. (2011). Eosinophilic esophagitis: updated con- sensus recommendations for children and adults. J Allergy Clin Immunol, 128, 3–​20.e6. Mainie I, et al. (2006). Acid and non-​acid reflux in patients with per- sistent symptoms despite acid suppressive therapy:  a multicentre study using combined ambulatory impedance-​pH monitoring. Gut, 55, 1398–​402. Malfertheiner P, et al. (2005). Prognostic influence of Barrett’s oe- sophagus and Helicobacter pylori infection on healing of erosive gastro-​oesophageal reflux disease (GORD) and symptom reso- lution in non-​erosive GORD: report from the ProGORD study. Gut, 54, 746–​51. Matthews PJ, Aziz Q (2005). How useful are proton-​pump inhibitors for diagnosis and therapy of patients with noncardiac chest pain? Nat Clin Pract Gastroenterol Hepatol, 2, 506–​7. Mulhall BP, Wong RK (2003). Infectious esophagitis. Curr Treat Options Gastroenterol, 6, 55–​70. National Institute for Health and Care Excellence (2014). Dyspepsia and gastro-​oesophageal reflux disease: Investigation and management of dyspepsia, symptoms suggestive of gastro-​oesophageal reflux di­ sease, or both. http://​www.nice.org.uk/​guidance/​cg184 Ngamruengphong S, et al. (2011). Proton pump inhibitors and risk of fracture: a systematic review and meta-​analysis of observational studies. Am J Gastroenterol, 106, 1209–​18. Pandolfino JE, et al. (2006). Obesity: a challenge to esophagogastric junction integrity. Gastroenterology, 130, 639–​49. Pandolfino JE, et al. (2006). Transient lower esophageal sphincter re- laxations and reflux: mechanistic analysis using concurrent fluor- oscopy and high-​resolution manometry. Gastroenterology, 131, 1725–​33. Pech O, et al. (2014). Long term efficacy and safety of endoscopic re- section for patients with mucosal adenocarcinoma of the esophagus. Gastroenterology, 146, 652–​60. Peery AF, et al. (2015). Burden of Gastrointestinal, Liver, and Pancreatic Diseases in the United States. Gastroenterology, 149, 1731–​41. Phoa KN, et al. (2014). Radiofrequency ablation vs endoscopic surveil- lance for patients with Barrett esophagus and low-​grade dysplasia: a randomized clinical trial. JAMA, 311, 1209–​17. Rastogi A, et al. (2007). Incidence of esophageal adenocarcinoma in patients with Barrett’s esophagus and high-​grade dysplasia: a meta-​ analysis. Gastrointest Endosc, 67, 394–​8. Ross WA, et al. (2007). Evolving role of self-​expanding metal stents in the treatment of malignant dysphagia and fistulas. Gastrointest Endosc, 65, 70–​6. Ross-​Innes CS, et  al. (2015). Whole-​genome sequencing provides new insights into the clonal architecture of Barrett’s esophagus and esophageal adenocarcinoma. Nat Genet, 47, 1038–​46. Ruigómez A, et al. (2007). Endoscopic findings in a cohort of newly diagnosed gastroesophageal reflux disease patients registered in a UK primary care database. Dis Esophagus, 20, 504–​9. Scottish Intercollegiate Guidelines Network (SIGN) (2006) Manage­ ment of oesophageal and gastric cancer. http://​www.sign.ac.uk Shaheen NJ, et  al. (2009). Radiofrequency ablation in Barrett’s esophagus with dysplasia. N Engl J Med, 360, 2277–​88. Shapiro J, et al. (2015). Neoadjuvant chemoradiotherapy plus surgery versus surgery alone for oesophageal or junctional cancer (CROSS): long-term results of a randomised controlled trial. Lancet Oncol, 16(9), 1090–8. Sharma P, et al. (2006). The development and validation of an endo- scopic grading system for Barrett’s esophagus: the Prague C & M criteria. Gastroenterology, 131, 1392–​9. Sheikh M, et al. (2019). Individual and Combined Effects of Environ­ mental Risk Factors for Esophageal Cancer Based on Results From the Golestan Cohort Study. Gastroenterology, Jan 3. Spechler SJ (2007). Screening and surveillance for Barrett’s esophagus—​an unresolved dilemma. Nat Clin Pract Gastroenterol Hepatol, 4, 470–​1. Vakil N, et al. (2006). The Montreal definition and classification of gastroesophageal reflux disease: a global evidence-​based consensus. Am J Gastroenterol, 101, 1900–​20. Weber CE, et al. (2012). Medium and long-​term outcomes after pneu- matic dilation or laparoscopic Heller myotomy for achalasia: a meta-​ analysis. Surg Laparosc Endosc Percutan Tech, 22, 289–​96. Young PE, et  al. (2010). Endoscopic ultrasound does not accur- ately stage early adenocarcinoma or high-​grade dysplasia of the esophagus. Clin Gastroenterol Hepatol, 8, 1037–​41. Zendehdel K, et  al. (2007). Risk of esophageal adenocarcinoma in achalasia patients, a retrospective cohort study in Sweden. Am J Gastroenterol, 102, 1–​5.