15.8 Peptic ulcer disease 2849

15.8 Peptic ulcer disease 2849

ESSENTIALS Helicobacter pylori infection, use of nonsteroidal anti-​inflammatory drugs (NSAIDs) including aspirin, and smoking are the most im- portant causes of peptic ulcer disease. Peptic ulcer disease is characterized by a history of waxing and waning symptoms of localized, dull, aching pain in the upper ab- domen. Bleeding is the most common complication; free perfor- ation of the stomach or duodenum into the peritoneal cavity is uncommon but serious. The diagnosis of peptic ulcer disease is made by endoscopy, which offers an opportunity for biopsy of gastric ulcers (which may be ma- lignant) and reveals important prognostic indicators in patients with bleeding ulcers. A single daily dose of a proton pump inhibitor gives quick relief of symptoms and effective healing of peptic ulcers in 4 to 6 weeks. The management of patients with upper gastrointestinal haem- orrhage requires a multidisciplinary medical and surgical approach. Early risk stratification based on clinical and endoscopic criteria al- lows delivery of appropriate care, with endoscopic intervention now widely accepted as the first line of therapy. This should be followed by administration of a high dose of an intravenous proton pump in- hibitor to further reduce recurrent bleeding. Treatment of H. pylori is a cure for peptic ulcer disease in most patients. This usually requires at least two antimicrobial agents, with the most popular triple therapy combining a proton pump inhibitor with any two of amoxicillin, metronidazole, and clarithromycin for 7 to 14 days. Eradication of H. pylori infection, avoidance of high-​dose NSAIDs or aspirin, and the maintenance use of proton pump inhibi- tors in high-​risk individuals are the best ways to prevent recurrence of ulcer and ulcer complications. Introduction Peptic ulcer is defined as a distinct breach in the mucosa of the gastrointestinal tract as a result of caustic effects of acid and pepsin in the lumen. The term ‘peptic ulcer disease’ usually refers to ul- ceration of the stomach, duodenum, or both but it can also occur in the oesophagus in gastro-​oesophageal reflux disease and in the distal ileum as a result of a Meckel’s diverticulum lined with an acid-​ secreting gastric epithelium. Histologically, peptic ulcer is identi- fied as necrosis of the mucosa extending through the muscularis mucosae into the submucosa. In the endoscopic or radiological view, there is an appreciable depth of the lesion. When the break of epithe- lial lining is confined to the mucosa without penetrating through the muscularis mucosae, the superficial lesion is called ‘erosion’. For more than a century, peptic ulcer disease has been a major cause of morbidity and mortality. The Schwarz dictum introduced in 1910 says ‘no acid, no ulcer’, indicating that the presence of acid is essential for peptic ulceration. Indeed, peptic ulcers rarely develop in patients with achlorhydria. Thus the therapy has always been fo- cused on acid neutralization or suppression of acid secretion. The discovery of Helicobacter pylori in the highly acidic environment by Marshall and Warren in 1984 has revolutionized the concept of ulcerogenesis. In many cases, peptic ulcer disease is an infectious disease which can be cured by a single course of antimicrobial therapy. In recent decades, however, there has been a rapid change in the epidemiology of peptic ulcer disease. With the improvement in sanitary conditions in many countries, there has been a dramatic de- cline in H. pylori infection and hence the associated peptic disease in the stomach and duodenum. On the other hand, with the increasing use of aspirin and nonsteroidal anti-​inflammatory drugs (NSAIDs), a new epidemic of peptic ulcer disease and complications has arisen. Our understanding of peptic ulcer disease is not complete. With the decline of H. pylori infection, peptic ulcers emerge that are not re- lated either to H. pylori or to NSAIDs. Aetiology, pathogenesis, and pathology Gastric acid and pepsin Despite the importance of H. pylori infection and ulcerogenic drugs such as NSAIDs and aspirin as the initiating events in the devel- opment of peptic ulcer disease, gastric acid and pepsin remain the ultimate injurious factors in the development of peptic ulcers. Ulceration is a result of an imbalance between the damaging effects of acid and pepsin and the defensive effects of bicarbonate and mucin 15.8 Peptic ulcer disease Joseph Sung

SECTION 15  Gastroenterological disorders 2850 on the mucosal surface. Factors that may account for increased se- cretion of acid and pepsin include increased parietal cell mass, in- creased stimulation of acid secretion (e.g. gastrin), increased parietal cell sensitivity to stimuli, and attenuated inhibition of acid secretion (e.g. somatomedin). Patients with peptic ulcer disease often have higher parietal cell mass, leading to increased basal and nocturnal unstimulated acid output as well as peak acid output under stimu- lation by food and gastrin. Several of these abnormal physiological responses are probably related to H. pylori infection (hypersecretion of gastrin, raised basal and gastrin-​releasing peptide-​stimulated acid output, decreased inhibitory drive mediated by somatostatin and hypersecretion of pepsinogen) as they disappear after successful cure of H. pylori infection. It is known that even peptic ulcer disease induced by NSAIDs or aspirin is an acid-​dependent process. In the low secretory state, NSAID and aspirin exposure are less likely to induce peptic ulcer in the upper gastrointestinal tract. With high exposure to gastric acid, epithelium of the duodenal bulb develops gastric metaplasia. Wyatt and colleagues postulated that gastric metaplasia is essential for the colonization of H. pylori in the duodenum and subsequent development of duodenal ulcer- ation. H. pylori is found colonizing only part of the duodenum with gastric metaplasia, setting off duodenitis and eventually duodenal ulcer. However, the data for the correlation of intragastric pH with occurrence of gastric metaplasia in the duodenum are inconsistent. Duodenal bicarbonate secretion Patients with duodenal ulcer are found to have impaired bicar- bonate secretion in the proximal duodenum in the face of influx of gastric acid. This impaired response is reversed by the eradication of H. pylori. The mechanism by which H. pylori hampers duodenal bicarbonate secretion is not understood. One proposed mechanism is that nitric oxide synthase activity in the duodenum interferes with bicarbonate secretion. Helicobacter pylori Since the discovery of H. pylori in the stomach of patients with gastritis and peptic ulcer, this bacteria has been reported in approximately 90% of cases of duodenal ulcer and 60% of cases of gastric ulcer. H. pylori is a slow-​growing, microaerophilic, highly motile, Gram-​negative spiral organism aetiologically linked to gastritis, peptic ulcer disease, gastric lymphoma, and adenocarcinoma of the stomach. H. pylori infection has a long latent period before symptomatic disease appears. H. pylori is tropic for gastric epithelium (i.e. stomach and areas of gastric meta- plasia outside the stomach) and is found either attaching to the surface epithelium through a pedestal or dwelling within the mucous coating on the surface of gastric epithelium. A very small proportion of or- ganisms can be found intracellularly, but the significance of this in relation to the inflammatory response and evasion of antimicrobial therapy is still under investigation. H. pylori infection elicits robust chronic active inflammatory and immune responses that continue throughout life. H. pylori produces abundant amount of urease, which is important for its colonization and survival in the stomach. H. pylori infection is primarily acquired in childhood, such that the prevalence at the age of 20 approximates the prevalence of that birth cohort throughout life. Acquisition during adulthood is rare, with estimates ranging from 0.3 to 0.5% per year, and recurrence of infection after successful eradication is therefore uncommon. The primary mode of transmission is person to person, probably via a gastro-​oral route (through vomitus) or oro-​oral route (through contamination of saliva). There are links between the bacterial geno- type, its virulence factor, and the development of gastroduodenal disease. CagA, a 120-​ to 140-​kDa highly antigenic protein, is en- coded by the cagA gene as part of the cag pathogenicity island. In Western countries, 60 to 80% of H. pylori express CagA, compared to 90% of isolates from Asian patients. The presence of the cag pathogenicity island is associated with a more prominent inflamma- tory tissue response than is seen with strains lacking this virulence factor. This increase in inflammation is associated with an increased risk of developing of peptic ulcer disease and adenocarcinoma of the stomach. The cag pathogenicity island encodes a type IV secretory apparatus that injects CagA and possibly other bacterial proteins into mammalian cells. CagA undergoes phosphorylation in the cell and is responsible for the changes in actin polymerization seen in the infected cell, resulting in conformational change. Besides cyto- skeletal changes, CagA also enhances inflammatory response which is mediated through NF-​κB. Attachment of H. pylori to the cell is re- quired for cagA-​positive H. pylori to elicit an interleukin-​8 (IL-​8) re- sponse in the gastrointestinal epithelium triggering gastritis. Beside CagA, approximately 50% of H. pylori strains produce a protein that induces vacuole formation in eukaryotic cells. This protein, which is called VacA, has been purified and the gene vacA has been cloned. The vacA gene has two families of alleles of the middle region (m1, m2) and at least three families of alleles of the signal sequence (s1a, s1b, s2). The vacA genotype s1 is strongly, but not exclusively, associ- ated with the cagA gene. So far, studies have not found an important role for VacA in relation to histological findings, or risk of H. pylori-​ related disease. The function of VacA remains unclear. Despite the establishment of a strong association between H. pylori infection and peptic ulcer disease, it is still unclear why some pa- tients develop duodenal ulcer and others gastric ulcer. McColl and El-​Omar proposed an intriguing paradigm (Fig. 15.8.1). In patients Gastrin ↑ Gastrin ↑ No atrophy Atrophy, cancer H. pylori infection DU GU or gastric cancer Acid ↓ Acid ↑ Body predominant gastritis Antral predominant gastritis Fig. 15.8.1  Association between pattern of H. pylori gastritis and disturbance in gastric physiology. Antral-​predominant gastritis is associated with duodenal ulcer (DU), body-​predominant gastritis with gastric ulcer (GU) or cancer. Modified from McColl KEL, El-​Omar E (2000). Mechanism involved in the development of hypochlorhydria and pangastritis in Helicobacter pylori infection. In: Hunt RH, Tytgat GNJ (ed) Helicobacter pylori: basic mechanisms to clinical cure. Kluwer, Dordrecht.

15.8  Peptic ulcer disease 2851 with duodenal ulcer, H. pylori colonizes mainly the antrum. The antral-​predominant gastritis stimulates production of gastrin-​ releasing peptide, triggering secretion of gastrin leading to exces- sive output of gastric acid. Profuse amount of acid flooding in the duodenum leads to gastric metaplasia, which allows colonization of H. pylori in the duodenum. This sets up an intense inflammation in the duodenum, further weakening mucosal protection and eventu- ally developing into duodenal ulcer. On the other hand, in patients with gastric ulcer, H. pylori is often found throughout the entire body of the stomach, leading to diffuse gastritis. The intense inflam- mation in the body of stomach tends to reduce gastric acid secre- tion as a result of glandular atrophy. In these patients other bacterial virulence factors come into play, leading to development of either gastric ulcer or adenocarcinoma in the distal stomach. Although this schema is probably oversimplified, it provides a broad-​brush picture explaining how an infection can induce two distinctly dif- ferent diseases. The ultimate proof of a causal relationship between H.  pylori and peptic ulcer disease comes from interventional studies. If peptic ulcer disease is merely a result of altered gastric physiology in bacterial infection, eradication of H. pylori in the stomach and duodenum should rectify the physiological change and cure the dis- ease. And, if reinfection with H. pylori is rare, peptic ulcer disease should not recur. Indeed, this has been proved in clinical trials. In a study that randomized duodenal ulcer patients to receive either 1-​week bismuth triple therapy or bismuth triple therapy plus 4-​week therapy with proton pump inhibitor, ulcers healed in 90 to 95% of cases with or without acid suppressive therapy. Similarly, when non-​ NSAID-​related gastric ulcer was treated by 1-​week bismuth triple therapy or 4-​week proton pump inhibitor therapy, ulcer healing was higher with anti-​Helicobacter therapy. More importantly, ulcer re- currence was much lower after patients received anti-​Helicobacter therapy with successful eradication than with a full course of proton pump inhibitor. Studies have also shown that peptic ulcer bleeding and bowel perforation does not recur, obviating the need for acid-​ reduction surgery. Ulcerogenic drugs In the last three decades, NSAIDs and antiplatelet agents have be- come increasingly important as a cause of peptic ulcer disease. It has been estimated that NSAIDs and aspirin increase the risk of gastric ulcer fourfold and the risk of gastrointestinal bleeding threefold. The risk of drug-​induced peptic ulcer is substantially higher in older people and those with previous history of peptic ulcer disease. Patients who are taking concomitant NSAIDs, as- pirin, anticoagulants, and corticosteroids are also exposed to a higher risk of peptic ulcer disease. H. pylori infection further in- creases the risk of peptic ulcer and ulcer complication in users of NSAIDs and aspirin. Aspirin and acidic NSAIDs were initially believed to have only a topical injurious effect by direct damage to the gastric epithelium as a result of intracellular accumulation of these drugs in an ionized state. However, the fact that enteric-​coated formulations, prodrugs, and systemic administration of NSAIDs fail to reduce the frequency of gastroduodenal ulceration implies that the chief mechanism of in- jury might not be a local action. NSAIDs reduce the hydrophobicity of mucous gel on the intestinal epithelium and this may hamper the defensive mechanism of the gut. The most important mechanism of drug-​induced peptic ulcer disease is inhibition of prostaglandin synthesis by NSAIDs. Prostaglandins regulate mucosal blood flow, epithelial cell proliferation, and basal acid secretion as well as mucus and bicarbonate secretion. The rate-​limiting enzyme in prosta- glandin synthesis is cyclooxygenase (COX). Most NSAIDs are found to suppress prostaglandin synthesis via reversible inhibition of COX, but aspirin acetylates COX and inhibits its enzyme activity irrevers- ibly in a dose-​dependent manner. In the early 1990s, two struc- turally related COX isoforms, COX-​1 and COX-​2, were identified. COX-​1 is found in most of the body’s tissues, including the gastro- intestinal tract and the kidney, and COX-​2 is an inducible enzyme produced principally in inflammation. The discovery of these iso- forms has prompted the development of COX-​2 selective inhibitors as anti-​inflammatory analgesics, with the aim of protecting against gastrointestinal damage. Yet this approach is an oversimplification, as evidence indicates that both COX-​1 and COX-​2 must be inhibited for gastric ulceration to occur. Selective suppression of COX-​1 does not cause gastric damage. Clinical trials have shown that COX-​2 selective inhibitors cause less peptic ulcer and ulcer bleeding than nonselective NSAIDs. Yet, in high-​risk patients with a history of peptic ulcer disease, ulcer complication as a result of using COX-​2 selective inhibitors is still a possibility. Gastric acid exacerbates NSAID injury by disrupting the base- ment membrane to produce deep injury, impairing platelet aggre- gation and potentiating enzymatic erosion of pepsin. It is therefore logical that suppression of acid secretion by potent agents such as proton pump inhibitors can confer at least partial protection against injury induced by NSAIDs and aspirin. More recently, attention has been focused on the role of nitric oxide in maintenance of intestinal mucosal blood flow. Like pros- taglandins, nitric oxide has been shown to increase blood flow, stimulate mucin secretion, and inhibit neutrophil adherence. It may thus protect the gastroduodenal tract against injury by aspirin and NSAIDs. Nitric oxide-​releasing NSAIDs have been developed and found to produce less gastric damage than their parent drugs. Tobacco, alcohol, and stress Cigarette smoking increases the risk of peptic ulcer diseases and their complications. As with NSAID usage, tobacco decreases pros- taglandin production and inhibits acid-​stimulated bicarbonate secretion in the duodenum. Increase in gastric acidity, reduction in epithelial cell proliferation, and impairment of mucosal blood flow have also been demonstrated with consumption of tobacco. Cigarette smokers are found to have slower healing of peptic ulcers and higher relapse rate of the disease. However, when H. pylori is eradicated, the effects of tobacco appear to be mitigated. It is a misconception that alcohol as such increases the risk of peptic ulcer disease. There are no convincing data in the literature supporting this notion. Although high concentrations of alcohol can cause damage to mucosa in animal studies, normal drinks such as wine and beer do not contain a high enough concentration of al- cohol to cause ulceration in the stomach and duodenum. However, peptic ulcer disease is more common in liver cirrhosis and alcohol consumption is certainly one of the most important underlying causes of this condition. The mechanism of peptic ulcer develop- ment in cirrhosis remains to be elusive. Psychological stress has always been implicated in peptic ulcer dis- ease but there is little scientific evidence to confirm the correlation.

SECTION 15  Gastroenterological disorders 2852 After all, stress is difficult to measure and its effects are hard to as- sess. Historical records during natural disasters (e.g. earthquakes and tsunami) and wars report an upsurge of peptic ulcer disease. During peacetime, however, stress seldom reaches high enough levels to lead to peptic ulcers. However, hospitalized patients with multiple illnesses and critical medical conditions can develop peptic ulcer and complications such as bleeding. Stress related to serious medical conditions and multiorgan failure is likely to produce peptic ulcer bleeding and the mortality of these patients has been estimated to be 10 times higher than for those without comorbid illnesses. Other causes Zollinger–​Ellison syndrome consists of a gastrin-​secreting islet cell tumour (gastrinoma) leading to marked hypergastrinaemia, out- pouring of gastric acid, and recurrent peptic ulceration. Most cases of Zollinger–​Ellison syndrome are sporadic but some are associated with multiple endocrine neoplasia syndrome type I (MEN-​1). As well as peptic ulcer disease, these patients may complain of diar- rhoea, steatorrhoea, symptoms of gastro-​oesophageal reflux, weight loss, and other presentations of MEN-​1 (e.g. hypercalcaemia and renal stones). The diagnosis is confirmed by finding a markedly raised serum gastrin level stimulated by secretin and radiological identification of tumour in the pancreas. Beside H.  pylori, other infections such as cytomegalovirus or Helicobacter heilmannii may lead to peptic ulcer disease. H. heilman­ nii has been found to cause intense inflammation in the stomach and occasionally peptic ulcers, especially in children. Helicobacter felis, a species that usually infects dogs and cats, has also been reported to cause peptic ulcer in pet owners. Crohn’s disease affects the whole gastrointestinal tract and may be a cause of peptic ulcer disease in the stomach, duodenum, or even the oesophagus. With the rising incidence of Crohn’s disease in Asia, peptic ulcers related to it are more commonly seen. Epidemiology It is hard to follow the temporal trend and geographical variation of peptic ulcer disease as the condition may not manifest itself in clin- ical settings. From records of peptic ulcer perforation, it has been suggested that this disease was uncommon before the 19th century. Over the ensuing decades, the incidence of peptic ulcer disease es- calated. By the end of the 19th century, duodenal ulcer frequency had surpassed gastric ulcer disease in frequency. The incidence of peptic ulceration rose dramatically throughout the first half of the 20th century, and then started to decline again in the second half of the century. Thus peptic ulcer disease appears to follow the trend of urbanization. The temporal trends of frequency of peptic ulcer disease are best studied by following birth cohorts. In Western coun- tries and in Japan, the risk of developing peptic ulcer disease rose in birth cohorts born before the turn of 20th century and then declined in subsequent generations. In addition to changes in the prevalence of peptic ulcer disease over time, there is also evidence that it shows geographical variations. For example, it is more common in Scotland and northern England than in southern England. Similarly, ulcers are more common in the south of India than in the north. Environmental factors are likely to play an important role in the development of peptic ulcer disease. Human-​to-​human transmission of H. pylori in urban dwellers, im- provement of sanitation in recent decades, and increased consump- tion of tobacco and analgesics might be important factors affecting the changing epidemiology of the disease. The incidence of bleeding resulting from peptic ulcer disease is much better documented than uncomplicated peptic ulcers. Based on the American Society of Gastrointestinal Endoscopy survey and two large United Kingdom audits made available in the 1990s, the reported incidence of gastrointestinal bleeding is approximately 100 per 100 000 population. The national United Kingdom audit was a population-​based, prospective collection of data on 4185 cases in 74 acute hospitals over a 4-​month period. Acute upper gastrointestinal bleeding is a disease primarily affecting the older age groups. In this audit, 68% of patients were older than 60 years and 27% were more than 80 years of age. In comparison with historic British series, a steady rise in the incidence over the last few decades was observed. The crude mortality rate increased from 9.9% in the 1940s to 11% in the 1990s. It is often argued that advances in the care of patients with upper gastrointestinal bleeding have been offset by an ageing population. A large survey of over 10 000 peptic ulcer bleeding pa- tients shows that mortality is often related to nonbleeding causes. Cardiopulmonary decompensation, multiorgan failure, and malig- nancy account for three-​quarters of the deaths and gastrointestinal bleeding is merely a terminal event in these patients. There has also been a trend towards increasing hospital admissions among older subjects and a corresponding decline for younger patients, resulting in little change in the overall admission rate. Hospital statistics from the United Kingdom Office of National Statistics revealed that from 1989 to 1999, admission rates for peptic ulcer haemorrhage increased among older people. Over this period, admissions increased by one-​ third among older women and by almost 50% among older men. The epidemiology of peptic ulcer disease has changed in the last two decades. With the declining prevalence of H. pylori infection in the developed countries, the proportion of patients with ulcers attributed to the use of aspirin and NSAIDs as well as H. pylori-​ negative ‘idiopathic’ ulcers is on the rise. However, there is evi- dence suggesting that between 20 and 40% of peptic ulcers in North America are not associated with H. pylori infection, nor the use of aspirin or NSAIDs. Is there truly a rise in non-​H. pylori, non-​NSAID ulcer, or does this merely reflects the declining trend of the disease and therefore a proportionate rise in idiopathic ulcers? The existing evidence suggests that the prevalence of these idiopathic ulcers is probably increasing. Two prospective cohort studies in Hong Kong, each lasting 1 year, looked at idiopathic ulcers in 1997 to 1998 and 2000 to 2001. The total number of bleeding peptic ulcers was more than 1500. Comparing the two time periods, the total number of bleeding ulcers per year had decreased by 33.1% and the number of H. pylori-​associated ulcers by over 30%. On the other hand, the absolute number of idiopathic ulcers increased 4.5-​fold. Patients suffering from idiopathic ulcers are older and sicker, and the ulcers more frequently developed after patients had been admitted to hos- pital for other medical conditions. Up to one-​half of idiopathic ulcer patients have major medical conditions such as advanced cardiopul- monary or liver disease. Ulcer recurrence is higher in patients with idiopathic ulcers than for H. pylori-​associated ulcers treated with eradication therapy.

15.8  Peptic ulcer disease 2853 Clinical features Dyspepsia Peptic ulcer disease is characterized by a history of waxing and waning symptoms of localized, dull, aching pain in the upper ab- domen, which is called dyspepsia. Many patients notice that symp- toms often worsen in winter. Eating spicy food and drinking coffee and tea may aggravate the symptoms, but these dietary habits do not lead directly to ulcer formation. Pain may occur sooner after meals in gastric ulcers than in duodenal ulcers, and is not necessarily relieved by food and antacids. The relationship between symptoms and eating is an unreliable predictor of peptic ulcer disease. Gastric ulcers are more often found in older patients, especially those taking NSAIDs or aspirin. Approximately 20% of complicated ulcers pre- sents without dyspeptic symptoms. These ‘silent’ ulcers are more common in individuals consuming NSAIDs, due to their analgesic effects. The lack of warning signs in these ulcers make them more dangerous. Haemorrhage Gastrointestinal bleeding is the most common complication as- sociated with peptic ulcer disease (Fig. 15.8.2). Vomiting of fresh blood, or haematemesis, indicates that bleeding originates from a site proximal to the suspensory muscle of the duodenum (liga- ment of Treitz). A history of fresh haematemesis usually implies a significant bleed and the patients may go into haemodynamic instability due to hypovolaemia. ‘Coffee ground’ vomiting, usu- ally arising from altered black blood, often indicates that active bleeding may have ceased. Melaena is the passage of black tarry stool. It occurs when haemoglobin in the gut is converted to haem- atin by bacterial degradation. As little as 200 ml of bleeding inside the digestive tract can produce melaena. Although melaena gen- erally connotes bleeding proximal to the suspensory muscle of the duodenum, bleeding from small bowel or proximal colon may also cause it, especially when colonic transit is slow. Haematochezia, passage of pure red blood or blood admixed with stool, occurs when bleeding comes from the lower gastrointestinal tract. It can also present in a massive upper gastrointestinal bleeding. When a substantial amount of blood is lost into the gastrointestinal lumen, pulses start to rise and blood pressure drops. The haemoglobin levels at this stage may not reflect the actual amount of blood loss before haemodilution sets in. A close monitoring of vital signs and estimation of volume of vomitus offer a better prognostic indicator of the severity of the illness. Perforation Free perforation of the stomach or duodenum into the peritoneal cavity is a rare but serious complication. It is more commonly found in older patients using aspirin or NSAIDs, leading to life-​ threatening catastrophe. The use of cocaine has also been related to peptic ulcer perforation. The classic presentation is a sudden onset of intense abdominal pain at the onset with gastric juice pouring into peritoneal cavity. This is followed a period of stabilization and amelioration of symptoms. Ulcer perforation may be concealed by the omentum. However, signs of peritonitis such as guarding and rebound tenderness persist. Bowel sounds are silent and liver dull- ness to percussion diminishes. A plain abdominal radiograph may demonstrate free gas between the upper border of the liver and the diaphragm and may also outline the serosal surfaces of the bowel wall. If prompt treatment is not provided, the patient will develop frank peritonitis and severe sepsis. Body temperature will rise and breathing becomes shallow. Leucocytosis and acidosis may appear at this stage. 55 43 22 10 5 0 20 40 60 80 100 Active Bleeder NBVV Clot Dot Clean Base Risk of recurrent bleeding (%) Fig. 15.8.2  Features of bleeding ulcers and the risk of recurrent bleeding. NBVV, nonbleeding visible vessel. From Lau JY, et al. (1998). The evolution of stigmata of hemorrhage in bleeding peptic ulcers: a sequential endoscopic study. Endoscopy, 30, 513–​18. © Georg Thieme Verlag KG.

SECTION 15  Gastroenterological disorders 2854 Besides free perforation, peptic ulcer may also penetrate into ad- jacent organs such as the pancreas, the bile duct, or even the colon, resulting in pancreatitis or gastrobiliary or gastroenteric fistula. With prompt medical attention, these complications are rarely seen nowadays. Obstruction In patients with recurrent peptic ulceration at the prepyloric an- trum, pylorus, and duodenal bulb, oedema and/​or scarring of the tissue may lead to obstruction of the gastric outlet. Pyloric obstruc- tion usually presents with bloating, early satiety, and vomiting. The patient may recognize food ingested several days ago in their vom- itus. Weight loss may be profound and dehydration with electrolyte disturbance (metabolic alkalosis with acidic urine) is common. On examination of the abdomen, an audible splash of the gastric content can be demonstrated by shaking the patient’s abdomen (succussion splash). Aspiration of gastric content through a nasogastric tube will empty litres of fluid and undigested food from the stomach, giving quick relief for the patient. Obstruction due to acute peptic ulcer and tissue oedema usually resolve in a few weeks as the ulcer heals. On the other hand, severe scarring of the pylorus and duodenum leads to permanent gastric outflow obstruction and requires endoscopic or surgical treatment. Differential diagnosis Symptoms are neither sensitive nor specific for the diagnosis of peptic ulcer disease. A wide range of conditions ranging from func- tional dyspepsia to malignancy of the gastrointestinal tract can pro- duce symptoms mimicking peptic ulcer disease. Pancreatitis and cholecystitis often produce more severe pain than peptic ulcer, so the differentiation may not be difficult. On the other hand, it is a disaster to miss a diagnosis of malignancy of the stomach, pancreas, or hepatobiliary tract. A high index of suspicion, especially in older patients with anorexia and weight loss, is needed to minimize the possibility of missing the diagnosis. In countries where the preva- lence of gastric cancer is high, symptom of dyspepsia should be managed carefully. Clinical investigation With the advent of endoscopy, barium studies are less frequently used in the diagnosis of dyspepsia. Endoscopy serves three pur- poses in diagnosis and evaluation of the patients. First, it confirms the diagnosis of peptic ulcer disease by its morphology, location, and, in the case of gastric ulcer, offering an opportunity for biopsy. Endoscopic features of bleeding ulcers are important prognostic in- dicators. The presence of signs of recent bleeding in an ulcer con- firms the source of bleeding. The Forrest classification categorizes ulcers into those that are actively bleeding, show signs of recent bleeding, or simply have a clean base: • Forrest class I ulcers are actively bleeding, either spurting (Forrest class IA) or oozing (Forrest class IB). • Forrest class  II ulcers show signs of recent bleeding including nonbleeding visible vessel (Forrest class  IIA), adherent clots (Forrest class IIB), or flat pigmented spots (Forrest IIC). • Forrest III ulcers have a clean base. The risk of continuous or recurrent bleeding of these ulcers is related to their appearance (Fig. 15.8.3). An evolutionary scheme for the natural history of signs of haem- orrhage for peptic ulcers has been proposed. Major bleeding from a peptic ulcer is arterial in origin. A sentinel clot (a term used syn- onymously with ‘visible vessel’) plugs the bleeding point. This can initially be contiguous with a larger overlying clot, which resolves in time. The clot may be variable in colour: initially it is red, but it darkens in time and subsequently the colour disappears leaving a plug of fibrin and platelets. Eventually the plug disappears, as the healing process is complete (Fig. 15.8.3). Ulcers with an adherent clot or protuberant vessels have a 20 to 40% chance of recurrent bleeding without proper endoscopic or pharmacological therapy. Endoscopic features, including the size and the site of bleeding ul- cers, should be interpreted along with clinical factors. Ulcers at the lesser curve of the stomach or posterior duodenal bulb are a high risk because of their proximity to the left gastric artery and the gastroduodenal artery respectively.

1. Arterial bleeding
2. Large red clot contiguous with sentinel clot/vv
3. Small red sentinel clot/vv
4. Dark sentinel clot/vv
5. White sentinel clot
6. Clot disappears Fig. 15.8.3  Evolution of signs of recent haemorrhage. vv, visible vessel. From Johnston JH (1990). Endoscopic risk factors for bleeding peptic ulcer. Gastrointest Endosc, 36 Suppl, S16–​20.

15.8  Peptic ulcer disease 2855 Gastric outlet obstruction due to recurrent peptic ulcer disease may produce a pinhole pylorus and dilated stomach. The endoscope may not be able to pass through the area of obstruction, leading to difficulty in assessment. Radiological imaging such as contrast studies is useful in this situation. Management Treatment of peptic ulcer disease can be divided into two stages: (1) treatment of acute symptoms and complications such as pain and bleeding, and (2) treatment of the underlying cause to prevent ulcer recurrence. Relief of symptoms and healing of peptic ulcer Before the 1970s treatment of peptic ulcer relied on antacids, anti- cholinergics, a bland diet, and bed rest. The therapeutic efficacy was low and many patients resorted to surgery such as partial gastrec- tomy and vagotomy. H2-​receptor antagonists In 1977, the first H2-​receptor antagonist, cimetidine, was introduced. Subsequently, ranitidine, famotidine, and nizatidine became available. These are effective acid-​suppressive agents, easy to use with an excellent safety profile. H2-​receptor antagonists quickly became the treatment of choice for peptic ulcer disease, but they have several disadvantages. Cimetidine has mild antiandrogenic effects, leading to gynaecomastia and impotence in some patients. The ability of cimetidine to bind to hepatic enzyme cytochrome P450 has also led to many interactions with other drugs, altering the pharmacokinetics of medications. Cimetidine is reported to interact with theophylline, phenytoin, lido- caine, warfarin, β-​blockers, tricyclic antidepressants, benzodiazep- ines, and many others. A variety of neurological reactions have also been reported such as headache, lethargy, depression, memory im- pairment, and confusion, especially in older patients. Furthermore, the reversible competitive inhibition of histamine-​stimulated acid secretion provides only a modest suppression of acid secretion. The postprandial acidity of the stomach is still relatively high. ‘Cytoprotective’ agents In addition to the acid-​suppressive agents, drugs claimed to have ‘cytoprotective’ activities were developed in the 1980s. Sucralfate is a complex salt of sucrose in which the eight hydroxyl groups of sucrose are replaced by sulphate and aluminium hydroxide. It is insoluble in water, forming a thick, tenacious paste that covers the surface of gastro- intestinal mucosa. Sucralfate has no acid-​suppressing effects and is be- lieved to work by coating the luminal surface of the ulcer, absorbing bile salts and pepsin and protecting the injured mucosa from further insult by erosive substances in the stomach. However, its ulcer healing effect is slow. Because of its aluminium content, it is considered unsafe in patients with chronic renal insufficiency. Acute aluminium toxicity has been reported in patients with end-​stage renal failure. Sucralfate is now rarely used in the clinical management of peptic ulcers. Bismuth salts have been used for many years for the treatment of diarrhoea, dyspepsia, and abdominal pain. As in case of sucralfate, bismuth has no effect on acid secretion in the stomach. The mech- anism of action is unknown but it is found to preferentially cover the ulcer crater. In the early 1980s, bismuth was found to inhibit the growth of H. pylori, which may partially explain its ulcer healing activity. Colloidal bismuth subcitrate and bismuth subsalicylate are available for clinical use. The use of bismuth subcitrate will blacken stool which might be confusing for patients suffering from gastro- intestinal bleeding. There are isolated reports of neurotoxicity re- lated to the use of bismuth subcitrate and bismuth subsalicylate when used in large doses in older patients. Otherwise, bismuth is a fairly safe drug. Its use is now mainly confined to the treatment of H. pylori infection in combination with other antimicrobial agents. Prostaglandin analogues Prostaglandins are derivatives of unsaturated fatty acid known as eicosanoids. The human gastrointestinal tract synthesizes several prostaglandins such as PGE2 and PGF2. Prostaglandins have been found to have a modest effect in inhibiting acid secretion in the stomach as well as stimulating the production of bicarbonate and mucin. Misoprostol has been developed as a prostaglandin analogue for human use. The standard dose of the drug is 800 µg/​day taken in four divided doses. The ulcer-​healing effect is not as potent as other acid-​suppressive agent and the regimen is inconvenient, but because of the nature of its action misoprostol offers one of the best treatments for NSAID-​related ulcers and prevention of this condi- tion. Diarrhoea is the most common side effects and limits the use of misoprostol in daily practice. It is also contraindicated in pregnancy as it may induce abortion. Uterine bleeding has been reported in a substantial proportion of patients. Proton pump inhibitors The discovery of proton pump inhibitors was major advance in the treatment of acid-​related gastrointestinal disorders. The final step of hydrogen ion secretion by the parietal cells is accomplished by H+,K+-​ APTase, an acid pump that exchanges hydrogen for potassium. These pumps are located at the apical membrane of tubulovesicular appar- atus of the parietal cells. Proton pump inhibitors such as omepra­ zole, lansoprazole, pantoprazole, and rabeprazole are substituted benzimidazoles that bind to the acid pump irreversibly. These drugs turn off acid secretion stimulated by any kind of stimulants. Recovery of acid secretion requires synthesis of new enzyme in the acid pump. The most effective way of administering a proton pump inhibitor is to take the drug before meals, that is, before acid secretion is trig- gered by food. Usually, a single daily dose gives quick relief of symp- toms and effective healing of peptic ulcers in 4 to 6 weeks. Compared to misoprostol and H2-​receptor antagonists, proton pump inhibi- tors such as omeprazole, esomeprazole, lansoprazole, pantoprazole, and rabeprazole are more effective in healing the ulcer as well as preventing further peptic ulcerations and erosions. As the most po- tent inhibitors of acid secretion, proton pump inhibitors are also the treatment of choice for Zollinger–​Ellison syndrome. They have also been found to suppress the growth of H. pylori and form an integral part in combination therapy to eradicate H. pylori in the stomach (‘proton pump inhibitor triple therapy’). Newer PPIs have been de- veloped, including dexlansoprazole modified release, instant release omeprazole as well as novel molecules such as tenatoprazole have also been developed and achieve more effective control of intragastric pH especially at night. Initial suspicion that prolonged suppression of the proton pump leads to enterochromaffin-​like cell hyperplasia has not been proved in humans. There are recent reports on potential side effects of the long-​term use of proton pump inhibitors. These include osteoporosis and bone fracture, hypomagnesaemia, enteric infection (e.g. Clostridium difficile), and pneumonia. Most of these associations

SECTION 15  Gastroenterological disorders 2856 are based on cohort studies and retrospective analysis. The evidence for serious side effects is poor and the risk is low. Potassium-competitive acid blockers (P-CABs) There are needs to overcome the shortcomings of PPI namely, the delayed release of the drug, short plasma half-lives, requirement to be given before a meal and show poor control of nocturnal acid secretion. The major advance has been the development of the po- tassium channel acid blocking drugs (P-CABs) which block the K+,H+-ATPase K+ channel. P-CABs are food independent, revers- ible, have a rapid onset of action, and maintain a prolonged and con- sistent elevation of intragastric pH. Vonoprazan, the first P-CAB, has been introduced only into a small number of Asian countries. It is in general, safe and efficacious in gastric acid secretion. Its clinical indications are still to be developed. Ulcer bleeding The management of patients with upper gastrointestinal haemor- rhage requires a multidisciplinary approach mandating cooperation among medical and surgical gastroenterologists with access to skills in endoscopic and surgical haemostasis. Endoscopic treatments Endoscopic therapy is often the first treatment in the manage- ment algorithm. Approximately 80 to 85% of upper gastrointestinal bleeding stops spontaneously. The remaining 15 to 20% continue to bleed or develop recurrent bleeding and these patients constitute the high-​risk group with substantial increased morbidity and mor- tality. Early risk stratification of patients with upper gastrointestinal bleeding based on clinical and endoscopic criteria allows delivery of an appropriate level of care to patients. Endoscopic therapy is now widely accepted as the first line of therapy for upper gastrointestinal bleeding. It should be applied to actively bleeding ulcers or ulcers covered with an adherent clot (Forrest class IIB). Many clinical trials and at least two meta-​analyses have confirmed the efficacy of endo- scopic haemostasis with dual therapy. Endoscopic therapy reduces recurrent bleeding and the need for surgical intervention. Endoscopic therapy can be broadly categorized into endoscopic injection, thermal coagulation, and mechanical haemostasis. Combining endoscopic injection with either thermal coagulation or mechanical haemostasis represents the best endoscopic therapy, with an overall success rate of around 90%. Endoscopic injection Epinephrine, polidocanol, sodium tetradecyl sulphate, absolute al- cohol, and even saline solution have been used for injection. No single agent for endoscopic injection is superior to another for achieving haemostasis. The mechanism of action is mostly related to a tamponade effect produced by the solution injected. The haemo- static effect of endoscopic injection is only transient, as the solution will be absorbed by the tissue. Injection is therefore not recom- mended as the sole therapy for peptic ulcer bleeding. Combination with other modalities is required. Thermal coagulation Thermal devices include heated probes; electrocoagulation is re- quired to secure haemostasis. Thermal devices seal blood ves- sels underneath the ulcer base by pressure and heat energy. This combined pressure–​thermal energy effect is called ‘coaptive coagu- lation’. Like endoscopic injection, no single method of endoscopic coaptive therapy is superior to the others. Mechanical haemostasis Endoscopic clipping or (in some cases) banding ligation can achieve mechanical haemostasis. The latest development is a device called over-​the-​scope clips, which takes a bigger area of ulcer base with a firmer grip on the bleeding vessel. The results appear promising. Mechanical haemostasis is found to be as effective as thermal device in controlling peptic ulcer bleeding. Haemostatic spray (Hemospray) is an inert particle which swells on exposure to water and stimulates platelet aggregation. It can be applied through an endoscope and achieve temporary control of bleeding from peptic ulcers. It is a useful stop-​gap treatment when other endoscopic methods fail. Angiographic embolization Angiographic embolization has been used as a salvage procedure when endoscopic method fails to control peptic ulcer bleeding. When compare to surgery, angiographic embolization is less ef- fective in securing haemostasis but leads to lower morbidity re- lated to procedures. Recent study has also suggested that in treating patients with large (>1.5 cm) ulcer with visible blood vessels, pre-​ emptive embolization (i.e. before recurrent bleeding occurs) after initial endoscopic haemostasis might reduce mortality of patients. Proton pump inhibitors As an acidic environment in the stomach and duodenum inhibits platelet aggregation and activates enzymatic activity of pepsinogen, suppression of acid, especially in the early phase of peptic ulcer bleeding, is useful as an adjuvant therapy. Randomized studies and a subsequent systematic review have confirmed that a high-​dose intra- venous proton pump inhibitor offers an effective inhibition of gas- tric acid secretion and reduces recurrent bleeding. Patients receiving these intravenous infusions require less blood transfusion, fewer re- peated endoscopic treatments, and fewer surgical operations. Recent data suggest that the use of an intravenous proton pump inhibitor in the early phase of bleeding, before endoscopy, may reduce the re- quirement for endoscopic therapy and shorten hospital stay. Peptic ulcer bleeding Forrest I IV PPI Endo Rx IV PPI Endo Rx Oral PPI No Endo Rx Repeat Endo Rx or angiographic emobolization Surgery Forrest II a/b Forrest IIc, III Fig. 15.8.4  Management strategy for peptic ulcer bleeding. Endo Rx, endoscopic treatment; IVPPI, intravenous proton pump inhibitor. Adapted from Sung J (2006). Current management of peptic ulcer bleeding. Nat Clin Pract Gastroenterol Hepatol, 3, 24–​32.

15.8  Peptic ulcer disease 2857 There have been debates about the possible benefit of using oral instead of intravenous proton pump inhibitors in the management of bleeding peptic ulcer. While the acid-​suppressing effects of oral medi- cation are similar to intravenous administration, in the early phase of the emergency, anti-​acid effect of oral proton pump inhibitors might not be adequate. A recent randomized study, however, showed that a high-​dose oral proton pump inhibitor might be useful as an alterna- tive to intravenous bolus plus infusion of proton pump inhibitors in patients who are not candidates for high-​dose intravenous therapy. An algorithm for the management of peptic ulcer bleeding is proposed in Fig. 15.8.4. When endoscopic haemostasis fails, angiographic intervention may help to control bleeding in patients who are old and considered as a high risk for surgery. Strategy of blood transfusion Replacement of lost blood and volume with red cell transfusion has always been an important part in the resuscitation and man- agement of ulcer bleeding. In many cases of severe bleeding, transfusion of blood is considered a life-​saving measure. There is some recent data, however, indicating that restricted transfu- sion may be appropriate in some settings. In a randomized con- trolled trial in which patients were given either liberal transfusion or transfusion only when haemoglobin level dropped to less than 7 g/​dL, evidence showed that those who received restrictive trans- fusion had better survival. In this study, however, only about half of the patients had peptic ulcer bleeding. Survival benefit was more significant in those who suffered from bleeding related to portal hypertension. Prevention of ulcer recurrence Management of H. pylori infection Treatment of H. pylori is a cure for peptic ulcer disease in most patients. H. pylori is susceptible to many different antimicrobials and a variety of combinations have been used successfully. Those that have proved effective include amoxicillin, metronida- zole, tetracycline, clarithromycin, and furazolidone. Other less commonly used antimicrobials include rifabutin and several fluoroquinolones. Successful cure of infection usually requires two antimicrobial agents. Cure rates with single antimicrobial agents are poor, ranging from 0 to 35%, and monotherapy is also associated with the rapid development of antibiotic resistance, hence it is not recommended. In principle, only those regimens that give high cure rates (>80%) should be used as first-​line therapy. Generally, higher doses and longer durations provide better results. Antibiotic resistance leads to reduced efficacy of therapy, hence the antimicrobial resistance pattern of H. pylori should be moni- tored and made known in each locality. The patient’s compliance with therapy is very important for successful cure of the infec- tion, so regimens should be simple and with as few side effects as possible. The treatment of H. pylori infection has gone through substantial evolution, primarily because of the issues of antibiotic drug resistance and patient (non)compliance. The choice of therapy should depend on regional drug resistance pattern (especially to clarithromycin and metronidazole), previous treatment received by the patient, and antibiotic-​sensitivity testing of individual cases (Fig. 15.8.5). First-​line therapy Triple therapy with a proton pump inhibitor remains the first choice in areas where clarithromycin resistance is low (<15%). The regimen contains a proton pump inhibitor (omeprazole 20 mg twice daily, lansoprazole 30 mg twice daily, pantoprazole 40 mg twice daily, or rabeprazole 20 mg twice daily) plus amoxicillin (1 g twice daily) plus clarithromycin (500 mg twice daily). This regimen is simple, easy to take, and well tolerated by patients. The efficacies of 7-​day, 10-​day, or 14-​day therapy are comparable. The differences between proton pump inhibitors are minimal. H. pylori is rarely resistant to amoxicillin. Metronidazole can be used to substitute amoxicillin in penicillin-​allergic patients. Quadruple therapy that combines bismuth subsalicylate (534 mg four times daily) plus a proton pump inhibitor plus two antibiotics (usually metronidazole 250 mg four times daily and tetracycline 500 mg four times daily) given for 10 to 14 days is useful as first-​line therapy in areas where clarithromycin resistance is common. If bis- muth is not available, a proton pump inhibitor with three antibiotics (including clarithromycin), or so-​called concomitant therapy, is an option. Capsules that combine various antibiotics with bismuth or proton pump inhibitor are useful as they improve compliance. Sequential therapy using two-​phases of proton pump inhibitor–​ antibiotic combinations has recently been shown to improve effi- cacy. Lansoprazole 30 mg plus amoxicillin 1 g for the first 5 to 7 days, followed by lansoprazole 30 mg, clarithromycin 500 mg, and metro- nidazole 500 mg for another 5 to 7 days appears to be superior to lansoprazole 30 mg, amoxicillin 1 g, and clarithromycin 400 mg for 14 days. Second-​line therapy Approximately 20% of patients fail responding to first-​line therapy and eradication of H. pylori cannot be achieved. Strategies to manage treat- ment failure include (1) not repeating clarithromycin unless in vitro Peptic ulcer disease with confirmed H. Pylori infection Region with Clarithromycin Resistance ≤15% PPI-Clari-Amox/Metro or Bismuth Quadruple PPI-Levo-Amox or PPI-Rifa-Amox or Bismuth Quadruple PPI-Levo-Amox or PPI-Rifa-Amox Based on antibiotic susceptibility testing For the choice of combination therapy Bismuth Quadruple or Sequential therapy or concomitant therapy Region with Clarithromycin Resistance >15% Fig. 15.8.5  A schematic diagram for H. pylori therapy.

SECTION 15  Gastroenterological disorders 2858 sensitivity testing prove that H. pylori is still susceptible to it, (2) anti- biotics used in initial therapy should be avoided, (3) tetracycline can be used, and (4) drug compliance of patients should be checked. After failure of a proton pump inhibitor–​clarithromycin triple therapy, a bismuth-​containing quadruple therapy (bismuth subci- trate 240 mg, omeprazole 20 mg, tetracycline 500 mg, metronidazole 500 mg all given twice daily for 14 days) is recommended. In areas of high clarithromycin resistance (where proton pump inhibitor–​ clarithromycin was therefore not used but bismuth-​containing quadruple therapy failed), levofloxacin-​containing triple therapy is recommended. A 10-​day course of proton pump inhibitor plus rifab- utin (300 mg twice daily) plus amoxicillin (1 g twice daily) has also been claimed to be effective. Sequential therapy for 14 days may be considered preferable in regions where clarithromycin resistance is high and metronidazole resistance is low. Levofloxacin-​containing sequential therapy may also be considered in a sequential regimen when clarithromycin resistance is high in the region. H. pylori resistance to antibiotics Treatment of H. pylori is becoming more difficult nowadays be- cause of emerging resistance to antimicrobials. Metronidazole and clarithromycin resistance rates are alarming but vary among popu- lations. Tetracycline and amoxicillin resistance are much lower in most countries. H. pylori resistance can be detected by phenotypic or by molecular methods. Different break points may be used when performing an antimicrobial susceptibility test, so comparing resist- ance among different populations is challenging. Eradication rates are dependent on the susceptibility of the strain to metronidazole and clarithromycin, being lower in patients infected with a resistant strain. Choice of therapy should be guided by prevalence of resist- ance to antibiotics (clarithromycin in particular) in that locality. Testing for antibiotic resistance Routine culture and testing for antibiotic sensitivity for H. pylori is not recommended. However, in patients who have received at least two courses of combination therapy and fail to cure the infection, testing of drug sensitivity seems logical. Culture of H. pylori can be done by obtaining a piece of biopsy tissue and placing it on a single drop of sa- line. A culture can also be obtained on the biopsy used for CLO test if the specimen is removed from the gel within 1 h and sent immediately to the laboratory. Choice of antimicrobial should then be guided by the sensitivity testing results. However, patients should be cautioned that even in vitro culture sensitivity-​guided therapy does not guarantee 100% success as in vivo activities of these drugs might be influenced by many other factors, including acidity and food in the stomach. Prevention of NSAID-​associated ulcer NSAID-​associated ulcer and ulcer complications are more com- monly reported in high-​risk individuals, that is, older people and those with history of peptic ulcer disease or chronic medical illness. Concomitant use of NSAIDs with aspirin, anticoagulants, or cor- ticosteroid also increases the risk of bleeding from peptic ulcers. Special caution has to be exercised before prescribing NSAIDs to these patients. Various prophylactic strategies to reduce gastroduodenal injury by NSAIDs have been investigated. These include concurrent treatment with H2-​receptor antagonist, misoprostol, proton pump inhibitor, and substitution of conventional NSAIDs by COX-​2 selective inhibitors. Systematic review pooling over 30 randomized controlled clinical trials of misoprostol, H2-​receptor antagonist, or proton pump in- hibitor for the prevention of gastroduodenal ulcer showed that these drugs have different efficacy. H2-​receptor antagonists reduce the risk of duodenal ulcer, but not of gastric ulcer, except at very high dose. Misoprostol at 80 µg per day can reduce ulcer and ulcer complication but its side effects are significant. Proton pump inhibitors can reduce the risk of both duodenal and gastric ulcers associated with NSAIDs and they are much better tolerated than misoprostol. The interaction between H. pylori and NSAIDs in the develop- ment of peptic ulcer disease is complex. Clinical studies reported by different investigators have yielded conflicting results. Part of the confusion stems from the recruitment of different patient groups and use of different outcome measurement. Meta-​analysis of 16 studies showed that H. pylori infection and NSAID use increase the risk of ulcer bleeding by 1.8-​fold and 4.8-​fold respectively. The risk of ulcer bleeding increases to around sixfold when both factors are present. This implies that NSAIDs and H. pylori are independent but addi- tive risk factors for ulcer development. H. pylori-​infected individ- uals taking NSAIDs will have an increased risk of peptic ulcer and ulcer complications. If a patient known to have H. pylori infection requires a NSAIDs, eradicating H. pylori before using the NSAID may substantially reduce the risk of peptic ulcer disease. However, simply curing H. pylori infection may not be sufficient to protect the stomach and duodenum from ulcer formation in high-​risk individ- uals. In elderly patients with history of ulcer complication, concomi- tant use of a proton pump inhibitor is warranted. In these patients, even the use of COX-​2 selective inhibitors is not entirely safe. The risk of recurrent bleeding with celecoxib is comparable to the use of diclofenac combined with omeprazole, according to one study. In patients with a history of ulcer bleeding, a combination of COX-​ 2 selective inhibitors with a proton pump inhibitor offers the best safety profile for the gastrointestinal tract (Table 15.8.1). Table 15.8.1  Recommendations for reducing the risk of ulcer and ulcer complications in high-​risk patients (NSAID and aspirin users) Strategies NSAID users Aspirin users Choice of medication Choose less ulcerogenic NSAID (e.g. ibuprofen) or short-​term COX-​2 selective inhibitors Use low-​dose aspirin (80–​100 mg/​day) H. pylori infection Eradicate H. pylori infection with proton pump inhibitor triple therapy Eradicate H. pylori infection with proton pump inhibitor triple therapy Concomitant medication Avoid combining with aspirin, anticoagulants, and steroid Avoid combining with NSAID, clopidogrel, COX-​2 selective inhibitors, anticoagulant, and steroid Ulcer-​preventing drugs Proton pump inhibitor or high-​dose H2-​receptor antagonist in high-​risk individuals Proton pump inhibitors in high-​risk individuals

15.8  Peptic ulcer disease 2859 Prevention of ulcer associated with antiplatelet agents Aspirin and clopidogrel are increasingly used in the prevention of cardiovascular and cerebrovascular diseases. Aspirin-​induced peptic ulcer disease is dose dependent, so the lowest dose of aspirin should be prescribed. Aspirin is often used in elderly patients who require NSAID or COX-​2 selective inhibitors for musculoskeletal pain. Combinations of aspirin with NSAIDs and COX-​2 selective inhibitors have been shown to increase the risk of ulcer bleeding substantially. The gastric sparing effect of COX-​2 inhibitors is offset by the concomitant use of low-​dose aspirin, so this combination should be avoided if possible. Eradication of H. pylori infection has been shown to reduce the risk of peptic ulcer bleeding in high-​risk individuals (Table 15.8.1). Recent study shows that if aspirin is dis- continued for a prolonged period after peptic ulcer bleeding, patient survival may be jeopardized as a result of cardiovascular or cere- brovascular conditions. Clinicians are advised to exercise discretion and balance the risk and benefit of discontinuing antiplatelet agents in these patients. Clopidogrel has an improved gastrointestinal safety profile com- pare to aspirin in general use. However, in high-​risk individuals (e.g. elderly patients with a history of ulcer or ulcer complication), the risk of peptic ulcer with clopidogrel should not be underestimated. In a head-​to-​head comparison between clopidogrel and low-​dose aspirin combined with proton pump inhibitors, the risks of peptic ulcer bleeding were shown to be similar for both strategies. In recent years, there has been a trend to combine aspirin and clopidogrel in the management of patients with myocardial infarction, especially after percutaneous coronary interventions with stenting. A com- bined use of two antiplatelet agents is recommended for at least 6 months after the procedures. The combination of clopidogrel and aspirin is expected to further increase the gastrointestinal risk, so the benefit of using these antiplatelet agents must be balanced against the risk of causing gastrointestinal toxicity in these patients. This could be a difficult decision in elderly patients with life-​threatening cardiovascular disease and a history of ulcer complication in the past. In patients with bleeding peptic ulcers who have significant car- diovascular disease, discontinuing antiplatelet agents may increase the risk of cardiovascular mortality. At least one antiplatelet agent should be continued under the cover of proton pump inhibitors. Although data suggest that interactions between some proton pump inhibitors and clopidogrel at least have the potential to increase the frequency of cardiovascular events, randomized trials and a recent meta-​analysis have not revealed such an effect, and more clinical evidence will be required to reach this conclusion with certainty. New anticoagulants and risk of ulcer bleeding New anticoagulants including dabigatran, rivaroxaban, and apixaban differ from warfarin with their fixed oral dose and no requirement for routine monitoring, hence attract a lot of usage among patients who requires anticoagulation therapy. Dabigatran has been asso- ciated with an increased risk in gastrointestinal bleeding (which is probably not shared by rivaroxaban) and hence should be used with caution in patients who have a history of peptic ulcer disease. When bleeding occurs, the use of coagulation factors (fresh frozen plasma, prothrombin complex concentrates, or recombinant activated factor VII) is more successful in reversing the activity of rivaroxaban than dabigatran. Surgery and peptic ulcer complications With the improvement of ulcer treatment using proton pump in- hibitors and anti-​Helicobacter therapy, the role of ulcer surgery has diminished. Classical operative procedures such as partial gastrec- tomy (Billroth I and Billroth II gastrectomy) and vagotomy are now rarely performed except for unhealed or recurrent peptic ulcers in the stomach or duodenum. As a result, postgastrectomy complica- tions such as afferent loop syndrome, dumping syndrome, postva- gotomy diarrhoea, and bile reflux gastropathy are disappearing in clinical practice. Surgery is still the most effective method of treating peptic ulcer bleeding arising from ulcers at difficult positions or large submucosal vessels (e.g. the gastroduodenal artery). Plication of the bleeding vessels and/​or removal of part of the stomach or duodenum remain the definitive method of controlling bleeding that cannot be stopped by pharmacological and endoscopic measures. Often, this is a life-​ saving procedure. The decision to operate is best made by a team of experienced gastroenterologists and gastrointestinal surgeons with a close working relationship. Repeated, unsuccessful attempts at endoscopic haemostasis lead to undue delay in surgery, massive blood transfusion, and multiorgan failure, jeopardizing the patient’s survival. In a study comparing second-​attempt endoscopic therapy versus surgery, ulcer surgery showed a superior haemostasis result although postoperative complications were frequent. An individual-​ based decision and the exercise of clinical discretion are therefore required. Despite initial enthusiasm for endoscopic dilatation of pyloric stenosis, the long-​term result is disappointing. Gastric outlet ob- struction often recurs months or years after endoscopic balloon dilatation. Partial gastrectomy and vagotomy may solve the problem of obstruction once and for all, saving the patient repeated admis- sions to hospital. Free perforation of the ulcer into the peritoneum is another indication for ulcer surgery. Repair of perforation and vagotomy is usually adequate to control the disaster. Treatment of H. pylori infection and a maintenance dose of proton pump in- hibitor are indicated as follow-​up therapy. Areas of uncertainty and future development We have come a long way in the last two decades in the under- standing of pathogenesis of peptic ulcer disease and its management. Substantial improvements have been made in preventing recurrent disease and in the treatment of its associated complications. There are, however, areas of uncertainty and room for future improvement. Although H. pylori has been identified as the major cause of peptic ulcer disease in individuals who do not use NSAIDs or aspirin, it is still not clear why only a relatively small proportion of infected sub- jects develop peptic ulcer disease. Bacterial factors (other than the cag pathogenicity island) and host factors (other than IL-​1B poly- morphism) need further studies to elucidate the difference in out- come. With rapid emergence of antimicrobial resistance in H. pylori, cure cannot be assumed without confirmation. An effective second-​ line therapy is still much needed. The best strategy for high-​risk individuals requiring antiplatelet, NSAIDs, or COX-​2 inhibitors needs further studies. There are, at pre- sent, very few data on the effective protection of the gastrointestinal

SECTION 15  Gastroenterological disorders 2860 tract when patients are prescribed double antiplatelet agents. Scepticism still persists about the safety of eradicating H. pylori as the only prophylaxis for aspirin users. In view of the complicated interaction between NSAIDs, COX-​2 inhibitors, and antiplatelet agents in vascular and gastrointestinal safety, a matrix for choice of therapy under different circumstances is much desired. Guidelines need to be developed for primary care physicians, cardiologists, and gastroenterologists looking after these patients. The role of nitric oxide is receiving more attention in under- standing the gastrointestinal toxicity of NSAIDs and analgesics. The development of NSAIDs and aspirin coupled with a molecule of nitric oxide is an exciting area that opens up a new horizon in the management of peptic ulcer disease in those who requires anti-​ inflammatory medication and antiplatelet therapy. The efficacy and safety of these drugs can only be confirmed by carefully designed clinical studies. Despite the advances in pharmacological and endoscopic therapy, the mortality of ulcer bleeding remains at 7 to 10%. Especially in the elderly patients, death is often related to nonbleeding causes such as cardiopulmonary conditions, multiorgan failure, and terminal malignancy. Clinicians are reminded to manage the comorbid ill- ness instead of focusing on bleeding lesions alone. In difficult cases of ulcer bleeding, endoscopy and surgery are the two common ap- proaches available at this stage. The role of radiological intervention, that is, embolization of the feeding artery at the ulcer base, deserves more careful investigation. FURTHER READING Abraham NS, et al. (2013). Novel anti-​coagulants: bleeding risk and management strategies. Curr Opin Gastroenterol, 29, 676–​83. Bhatt DL, et al. (2010). Clopidogrel with or without omeprazole in cor- onary heart disease. N Engl J Med, 363, 1909–​17. Blaser MJ (1996). Role of vac A and the cag A locus of Helicobacter pylori in human disease. Aliment Pharmacol Ther, 10, 73–​77. Calvet X, et al. (2004). Addition of a second endoscopic treatment following epinephrine injection improve outcome in high-​risk bleeding ulcers. Gastroenterology, 126, 441–​50. Chan FKL, et al. (2001). A randomised comparison of Helicobacter pylori eradication and omeprazole for the prevention of recurrent upper gastrointestinal bleeding in chronic users of low-​dose aspirin and non-​aspirin non-​steroidal anti-​inflammatory drugs. N Engl J Med, 344, 967–​73. Chan FKL, et al. (2002). Celecoxib versus diclofenac and omeprazole in preventing recurrent ulcer bleeding in patients with arthritis. N Engl J Med, 347, 2104–​10. Chan FKL, et al. (2002). Screen-​and-​treat Helicobacter pylori to reduce the risk of peptic ulcers for patients starting long-​term non-​steroidal anti-​inflammatory drug treatment:  a double blind randomised placebo-​controlled trial. Lancet, 359, 9–​13. Chan FKL, et al. (2005). Clopidogrel versus aspirin and esomeprazole to prevent recurrent ulcer bleeding. N Engl J Med, 352, 238–​44. Chan HLY, et al. (2001). Is non-​Helicobacter non-​NSAID peptic ulcer a common cause of upper gastrointestinal bleeding? A prospective study of 977 patients. Gastroint Endosc, 53, 438–​42. Cook DJ, et al. (1992). Endoscopic therapy for acute non-​variceal upper gastrointestinal haemorrhage:  a meta-​analysis. Gastroenterology, 102, 139–​48. Dorward S, et al. (2006). Proton pump inhibitor treatment initiated prior to endoscopic diagnosis in upper gastrointestinal bleeding. Cochrane Database Syst Rev, CD005415. El-​Omar EM, et al. (1995). Helicobacter pylori infection and abnor- malities of acid secretion in patients with duodenal ulcer disease. Gastroenterology, 109, 681–​91. El-​Omar EM, et al. (1997). Helicobacter pylori infection and chronic gastric acid hypo-​secretion. Gastroenterology, 113, 15–​24. European Helicobacter pylori Study Group. (1997). Current European concepts in the management of Helicobacter pylori infection. The Maastricht Consensus Report. Gut, 41, 8–​13. Forrest JA, Finlayson ND, Shearman DJ. (1974). Endoscopy in gastro- intestinal bleeding. Lancet, 2, 392–​7. Graham DY, Yamaoka Y (2000). Disease-​specific Helicobacter pylori virulence factors—​the unfulfilled promise. Helicobacter, 5, 3–​9. Hawkey CJ (1990). Non-​steroidal anti-​inflammatory drugs and peptic ulcers. BMJ, 300, 764. Hawkey CJ, et  al. (1998). Omeprazole compared with misoprostol for ulcer associated with non-​steroidal anti-​inflammatory drugs. Omeprazole versus misoprostol for NSAID-​induced ulcer manage- ment (OMNIUM) study group. N Engl J Med, 338, 727–​34. Hosking SW, et al. (1994). Duodenal ulcer healing by eradication of Helicobacter pylori without anti-​acid treatment: randomized con- trolled trial. Lancet, 343, 508–​10. Huang JQ, Sridhar S, Hunt RH (2002). Role of Helicobacter pylori in- fection and non-​steroidal anti-​inflammatory drugs in peptic ulcer disease: a meta-​analysis. Lancet, 359, 12–​22. Hung LCT, et al. (2005). Long-​term outcome of H. pylori-​negative bleeding ulcers: a prospective cohort study. Gastroenterology, 128, 1845–​50. Kwok CS, et  al. (2013). No consistent evidence of differential car- diovascular risk amongst proton pump inhibitors when used with clopidogrel: meta-​analysis. Int J Cardiol, 167, 965–​74. Lanas A, et al. (2006). Risk of upper gastrointestinal ulcer bleeding associated with selective cyclo-​oxygenase 2 inhibitors, traditional non-​aspirin non-​steroidal anti-​inflammatory drugs, aspirin and combinations. Gut, 55, 1731–​8. Lau JYW, et  al. (1999). Endoscopic re-​treatment versus surgery in patients rebleeding after initial endoscopic ulcer hemostasis:
a prospective randomized controlled trial. N Engl J Med, 340, 751–​6. Lau JYW, et al. (2000). A comparison of high-​dose omeprazole infu- sion to placebo after endoscopic hemostasis to bleeding peptic ulcer. N Engl J Med, 343, 310–​316. Lau JYW, et al. (2007). Omeprazole before endoscopy in patients with gastrointestinal bleeding. N Engl J Med, 356, 1631–​40. Leontiadis G, Sharma VK, Howden CW (2006). Proton pump in- hibitor treatment for acute peptic ulcer bleeding. Cochrane Database Syst Rev, CD002094. Liou JM, et al. (2013). Sequential versus triple therapy for the first-​ line treatment of Helicobacter pylori: a multicentre, open-​label, ran- domised trial. Lancet, 381, 205. Malfertheiner P, et al. (2011). Helicobacter pylori eradication with a capsule containing bismuth subcitrate potassium, metronidazole, and tetracycline given with omeprazole versus clarithromycin-​ based triple therapy:  a randomised, open-​label, non-​inferiority, phase 3 trial. Lancet, 377, 905–​13. Malfertheiner P, et al. (2012). Management of Helicobacter pylori in- fection –​ the Maastricht IV/​Florence Consensus Report. Gut, 61, 646–​64.

15.8  Peptic ulcer disease 2861 Marmo R, et al. (2007). Dual therapy versus monotherapy in the endo- scopic treatment of high-​risk bleeding ulcers:  a meta-​analysis of controlled trials. Am J Gastroenterol, 102, 270–​89. Marshall BJ, Warren JR (1984). Unidentified curved bacilli in the stomach of patients with gastritis and peptic ulceration. Lancet, 1, 1311–​15. Masferrer JL, et  al. (1994). Selective inhibition of inducible cyclooxygenase-​2 in vivo is anti-​inflammatory and non-​ulcerogenic. Proc Natl Acad Sci U S A, 91, 3228–​32. McColl KE, El-​Omar E (1996). Helicobacter pylori and disturbance of gastric function associated with duodenal ulcer disease and gastric cancer. Scand J Gastroenterol Suppl, 215, 32–​7. Patrono C, et  al. (2005). Low dose aspirin for the prevention of atherothrombosis. N Engl J Med, 353, 2373–​83. Rockall TA, et al. (1995). Incidence of and mortality from acute upper gastrointestinal haemorrhage in the United Kingdom. Steering Committee and members of the National Audit of Acute Upper Gastrointestinal Haemorrhage, BMJ, 311, 222–​6. Rostom A, et al. (2002). Prevention of NSAID-​induced gastroduodenal ulcers. Cochrane Database Syst Rev, 4, CD002296. Sacks HS, et al. (1990). Endoscopic haemostasis: an effective therapy for bleeding peptic ulcers. JAMA, 264, 494–​9. Silverstein FE, et al. (1981). National ASGE survey on upper gastro- intestinal bleeding:  study design and baseline data. Gastrointest Endosc, 27, 73–​9. Silverstein FE, et  al. (1995). Misoprostol reduces serious gastro- intestinal complications in patients with rheumatoid arthritis receiving non-​steroidal anti-​inflammatory drugs. A  random- ized double-​blind placebo-​controlled trial. Ann Intern Med, 123, 241–​9. Sonnenberg A. (1995). Temporal trends and geographical variations of peptic ulcer disease. Aliment Pharmacol Ther, 9 Suppl 2, 3. Sung JJY, et al. (1995). Antibacterial treatment of gastric ulcers associ- ated with Helicobacter pylori. N Engl J Med, 332, 139–​42. Sung JJY, et  al. (2007). Endoscopic clipping versus injection and thermo-​coagulation in the treatment of non-​variceal upper gastro- intestinal bleeding: a meta-​analysis. Gut, 56, 1364–​73. Sung JJY, et al. (2009). Intravenous esomeprazole for prevention of re- current peptic ulcer bleeding: a randomized trial. Ann Intern Med, 150(7), 455–​64. Sung JJY, et  al. (2010). Causes of mortality in patients with peptic ulcer bleeding:  a prospective cohort study of 10,428 cases. Am J Gastroenterol, 105(1), 84–​9. Sung JJY, et al. (2010). Continuation of low-​dose aspirin therapy in peptic ulcer bleeding: a randomized trial. Ann Intern Med, 152, 1–​9. Sung JJY et al. (2014). Effects of intravenous and oral esomeprazole in the prevention of recurrent bleeding from peptic ulcers after endo- scopic therapy. Am J Gastroenterol, 109(7), 1005–​10. Taha AS, et al. (1996). Famotidine for the prevention of gastric and duodenal ulcers caused by non-​steroidal anti-​inflammatory drugs. N Engl J Med, 334, 1435–​9. Vane JR (1971). Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-​like drugs. Nat New Biol, 231, 232–​5. Villanueva C, et  al (2014). Transfusion strategies for acute upper gastrointestinal bleeding. N Engl J Med, 368, 11–​21. Walan A, et al. (1989). Effect of omeprazole and ranitidine on ulcer healing and relapse rates in patients with benign gastric ulcer. N Engl J Med, 320, 69–​75. Wallace JL, et al. (2000). NSAID-​induced gastric damage in rats: require- ment for inhibition of both cyclooxygenase 1 and 2. Gastroenterology, 119, 706. Wyatt JI, et al. (1987). Campylobacter pyloridis and acid induced gastric metaplasia in the pathogenesis of duodenitis. J Clin Pathol, 40, 841–​8. Yeomans ND, et  al. (1998). A comparison of omeprazole with ra- nitidine for ulcers associated with nonsteroidal anti-​inflammatory drugs. Acid suppression trial:  Ranitidine versus omeprazole for NSAID-​associated ulcer treatment (ASTRONAUT) study group. N Engl J Med, 338, 719–​26.