Dyspepsia, from the Greek dys(bad) and peptein (digest) is a common presentation in general practice and a source of considerable morbidity and cost to the community. When William Brinton wrote in 1865 that dyspepsia is “not so much a single and substantive disease as a variety of ailments ... characterised not so much by the presence of certain symptoms as by the absence of structural lesions” he was not to know that in 2009 his statement would have particular relevance. In Western societies today, structural lesions previously associated with dyspepsia, such as upper gastrointestinal cancer and peptic ulcer disease, are far less prevalent than in the late 19th and 20th centuries. Today, dyspepsia is more commonly ‘dis-ease’ than a disease.
For the purposes of this review, dyspepsia is defined pragmatically as upper abdominal or retrosternal pain, discomfort, heartburn, nausea, vomiting or other symptoms considered to be referable to the proximal alimentary tract.
The overall incidence and prevalence of dyspepsia is difficult to establish because of the different definitions used but appears to be unchanged over recent years. This is probably due to the ageing population and the increasing prevalence of NSAID use counteracting the effects of falling prevalence of Helicobacter pylorirelated ulcers. Other risk factors that influence dyspepsia prevalence include:
General practice activity data estimate that about 10% of patient attendances in Australia are due to GORD, the single most common cause of upper GI symptoms. Estimates from these data suggest there are more than two million Australians with GORD. We know that only a proportion of those with dyspepsia will ever consult a GP, and estimates from community studies in a number of Western countries are that up to 40% of the population have upper GI symptoms. Despite under-presentation to general practice, the cost to the PBS of prescribed acid-suppression therapy was more than $400 million, or about 10% of our national drug budget in 2007/08. A US survey of 117,497 endoscopic reports involving 99,558 patients found that 43% of endoscopies were for dyspepsia and one third of patients were under 50 and had no alarm symptoms.1
Since the publication of the first Helicobacter pylori eradication trial in 1988 there has been great enthusiasm for attributing dyspepsia to H.pylori infection, an enthusiasm that was misplaced and is now waning. It is now clear that in most Western countries H.pylori is responsible for only a minority of the cases of dyspepsia. Infected individuals have an estimated 1:4-1:10 lifetime risk of developing ulcer disease and an increased risk of gastric cancer. As fewer people in the population acquire H.pylori in their childhood, H.pylori associated ulcer disease is becoming increasingly uncommon and gastric cancer is becoming rare. In most Western countries young people are more likely to stay uninfected than infected with H.pylori. In contrast, the elderly continue to carry much of the risk of ulcer disease and the small increased risk of gastric cancer because a higher proportion of them have acquired H. pylori as a result of the poorer domestic hygiene in their childhood. The second reason for attributing only a minority of dyspepsia cases to H.pylori is that there is little evidence to implicate this bacterium as a cause for dyspepsia in the absence of an ulcer. This is clearly demonstrated by the disappointing outcomes of H. pylori eradication therapy for GORD, gastritis and non-ulcer dyspepsia.
As the prevalence of H. pylori has fallen, so too has the proportion of patients with a lesion as a cause for their dyspepsia. Recent endoscopic studies demonstrate this and, in particular, the marked decline in the prevalence of peptic ulcer disease and gastric cancer. A recent study from 17 mainly Western countries reviewed the endoscopic findings of patients aged 18-70 presenting to primary care with dyspepsia without alarm symptoms. Only 23% were reported to have endoscopic abnormalities. The most common pathology found was reflux oesophagitis (table 1).2
About 5% of patients with dyspepsia have peptic ulcer disease. The association between H. pylori and peptic ulcer disease is well documented. More than 90% of duodenal ulcers are associated with the infection, as are a similar proportion of nonNSAID-induced gastric ulcers. As expected, with the fall in prevalence of H. pylori infection there has been a marked fall in peptic ulcer disease prevalence. Over the 10 years from 1970 some countries reported an almost 10-fold decline in hospitalisation rates for peptic ulcer disease. The second leading cause of peptic ulcer disease in our community is the use of NSAIDs, including aspirin. A recent meta-analysis showed that the risk of GI complications from NSAIDs is 1.5-2% per year in the average patient and 10% per year in the highrisk patient. The risk of ulcer complications from COX-2-selective inhibitors is about half that for naproxen.
It is worth noting that most NSAID-induced erosions and ulcers are asymptomatic, with GI haemorrhage often being the first indication of an NSAID-induced ulcer. NSAIDS and H. pylori are thought to induce ulcer disease by different mechanisms but with synergism between the two risk factors. H. pylori infection increases the risk of bleeding from peptic ulceration twofold, NSAIDS fivefold and together there is a sixfold increase. H. pylori-negative NSAID negative peptic ulcers do occur, but true H. pylori-negative NSAID-negative ulcers are very uncommon and are a diagnosis by exclusion (table 2). Smoking increases the risk of peptic ulcer disease in a dose-dependent manner.
The aetiology of GORD is unknown but thought to occur by three mechanisms:
Smoking, fats, caffeine, alcohol and a variety of drugs such as theophylline and calciumchannel blockers decrease lower-oesophageal sphincter pressure and may precipitate symptomatic reflux. H. pylori infection is not a cause of GORD and may even be protective, by raising gastric pH.
GORD is a risk factor for adenocarcinoma of the oesophagus and its pre-cursor Barrett’s oesophagus. A large study from Sweden found patients with oesophageal adenocarcinoma were more than seven times more likely to have symptomatic heartburn than the background population. The recent rise in the prevalence of adenocarcinoma in Australia and elsewhere is thought to be due to an increase in the prevalence of GORD (figure 1). However, oesophageal adenocarcinoma remains a rare cause of dyspepsia and may be an incidental finding at endoscopy. Figure 2 shows age-specific rates of oesophageal cancer.
In the past, one of the main arguments for investigating dyspepsia was to avoid missing gastric cancer. Gastric cancer is now known to have a rare association with dyspepsia. In Australia the age-standardised rate of gastric cancer is 9.7/100,000 in males, and lower in females, with a median age of diagnosis over 70 in both groups. As figure 3 shows, gastric cancer is rare under the age of 50 years.
Functional dyspepsia is defined as chronic dyspepsia in the absence of a structural lesion that may account for the symptoms. The pathophysiology of this condition has been poorly characterised. Recently, Rome III, a meeting of GI experts, proposed that functional dyspepsia be divided into two groups: postprandial distress syndrome, consisting of postprandial fullness and early satiety; and pigastric pain syndrome, with more consistent and less mealrelated symptoms. While this new classification may help our future understanding of the pathophysiology and ultimately lead to better treatment, its value for current management is unclear. A substantial proportion of patients with functional dyspepsia also have irritable bowel syndrome. The mechanism of both disorders may be similar and be a clue to the pathophysiology.
It is important to recognise that patients may have a combination of causes for dyspepsia, such as ulcer disease as well as GORD. A patient with peptic ulcer not improving after H. pylori eradication may have GORD requiring acidsuppression therapy.
Drug therapy is often an unrecognised cause of dyspepsia. Examples include NSAIDs, digoxin, antibiotics (including erythromycin), and iron therapy. Ten to 20 per cent of patients taking NSAIDs have associated dyspepsia, with or without an ulcer being present, and within six months of starting NSAID therapy 5-15% of patients will discontinue it because of dyspepsia. Patients taking COX-2 inhibitors such as celecoxib, rofecoxib (withdrawn worldwide) and meloxicam have a similar frequency of dyspepsia to non-selective NSAIDS but a lower risk of peptic ulcer complications.
The contribution of biliary pain and dysmotility to dyspepsia (including sphincter of Oddi dysfunction) is unclear but probably small. Systemic disorders such as diabetes and hypercalcaemia may rarely cause dyspepsia. There is debate as to whether dyspepsia is more common among patients with undiagnosed coeliac disease. The background prevalence of coeliac disease is about one in 100 and so is not infrequently found, including in patients with GI symptoms.
It is important to assess the timing of symptom onset in relation to starting of new medications and to ask about alcohol and cigarette consumption. Specific questioning about weight loss, dysphagia, symptoms resistant to acid-suppression therapy, early satiety, melaena and haematemesis are essential to evaluate the likelihood of upper GI malignancy. A dietary history of known precipitants of GORD may be very important for future management, as stopping excessive consumption of caffeine, chocolate, alcohol or fatty and spicy food may be sufficient to prevent further symptoms.
A history of predominant heartburn or reflux is useful for the diagnosis of GORD. However, up to one in five patients are unable to decide on their most bothersome symptom. A variety of respiratory and laryngeal symptoms has been attributed to GORD. These include cough, wheeze, and hoarse voice or sore throat, and may occur in the presence or absence of dyspepsia. Water brash (stimulation of salivation by the presence of gastric acid in the lower oesophagus) also rarely occurs. Dysphagia and odynophagia may result from altered oesophageal peristalsis, stricture or increased sensitivity. These symptoms should be treated with potent acid-suppression therapy, but the response is often slow and may be poor. Despite traditional teaching, the clinical diagnosis of peptic ulcer disease is difficult. Fewer than 25% of ulcers are diagnosable on history. The accuracy of clinical diagnosis can be increased by considering risk factors (table 3). Family history is a risk factor for peptic ulcer disease, with twin studies showing this to be due to both genetic and environmental factors. Smoking and a history of pain on an empty stomach have been shown to increase the likelihood of ulcer disease as the cause of dyspepsia. In the elderly, a history of aspirin and NSAID consumption is particularly important. The risk for NSAID complications increases with advanced age, H. pylori infection, smoking, previous ulcer history, concurrent anticoagulant and prednisone use.
SSRIs also slightly increase the risk of peptic ulcer disease. Comorbidities such as cardiac failure and diabetes increase the risk of peptic ulcer bleeding. The importance of a past history of peptic ulcer disease cannot be sufficiently emphasised. A recent survey in a NSW hospital found that 85% of patients admitted with a past history of peptic ulcer disease denied previous treatment with H. pylori eradication therapy and had positive H. pyloriserology indicative of ongoing infection.3
These patients should be offered H. pylori eradication therapy without further investigation. The benefits are threefold:
For most patients a trial of acid-suppression therapy is usually diagnostic of an acidrelated disorder and is safe. Rapid symptom recurrence on stopping acid-suppression therapy is virtually diagnostic of GORD. Patients should be advised to present again if symptoms do not resolve following 2-4 weeks of acid-suppression therapy.
Gastroscopy is indicated if malignancy is suspected because of the presence of alarm symptoms such as weight loss, dysphagia, early satiety, haematemesis, melaena, anaemia and bloody stools, or if dyspeptic symptoms are unresponsive to 2-4 weeks of acid-suppression therapy. Gastroscopy in the latter instance should be performed after PPIs have been ceased.
Endoscopic diagnosis is unnecessary for most patients with dyspepsia because of its low yield, particularly in those under 50 (table 4). The most frequent abnormal finding at endoscopy is erosive oesophagitis, which responds to empirical therapy, making endoscopy unnecessary. Fewer than 50% of patients with GORD have endoscopic evidence of oesophagitis. Multiple studies have shown that few treatable cancers are detected by early endoscopy. A recent large study in primary care in 17 mostly Western countries provides a useful summary of the role of endoscopy in diagnosing cancer in countries like Australia.2 In this study 2741 patients aged 18-70 with dyspepsia and without alarm symptoms had endoscopy. Upper GI cancer was found in two per 1000 patients (0.22%). The authors concluded that if endoscopy were limited to only dyspeptic patients 50 years and over, one oesophageal cancer and no gastric cancer would have been missed. If endoscopy cost $500 it would cost about $82,900 (between $35,714 and $250,000) to detect that cancer.2
Up to 40% of patients diagnosed with oesophageal adenocarcinoma have no prior history of dyspepsia. Endoscopy has had little impact on gastric cancer survival because most cancers are advanced when diagnosed and treatment options are limited. Current five-year survival for gastric cancer is 25% compared with 58% for cancer overall and 61% for colonic cancer. As 37% of oesophageal adenocarcinoma in Australia is attributable to poor physical activity and obesity, the logical management strategy for its prevention is promoting exercise and appropriate dietary intake.4
Advocates of endoscopy argue that initial empirical therapy may miss Barrett’s oesophagus. The risk of Barrett’s oesophagus increases with duration of reflux and is least likely to be found at the time of onset of symptoms. It is important to distinguish endoscopy before starting acid-suppression therapy from endoscopy after a trial of acidsuppression therapy. A study of the effect of time between withdrawing PPIs and gastroscopy showed a greater diagnostic yield the longer the time gap. A course of PPI therapy will heal more than 60% of peptic ulcers in two weeks and 94% within a month. PPIs should be stopped for at least four weeks before endoscopy is performed.
H. pylori testing and eradication should be offered to patients with peptic ulcer disease or those starting NSAID therapy, as it is of proven benefit for these conditions. The yield from H. pylori testing has fallen as a result of its decreasing prevalence. The main methods for detecting H. pylori are shown in table 5. Serology remains the most readily available, inexpensive and accurate test for detection. To confirm eradication, the urea breath test is the best non-invasive option, as serology may remain positive for up to 12 months.
pH monitoring involves inserting a pH probe into the distal oesophagus and recording of the proportion of time pH falls below 4 over a 24- hour period. It has a limited role in the diagnosis and management of dyspepsia but can be useful to assess the adequacy of therapy if a patient is still symptomatic on maximal acid-suppression therapy, or, off therapy, to assess symptom correlation. Patients can have GORD symptoms despite a normal pH study.
why people present may be the key to effective management. An Australian population-based survey found that only 56% of people with dyspepsia had ever consulted about their symptoms. Consultation seems not just related to age or the severity or duration of symptoms,
but also to more complex issues such as anxiety, including fear of serious disease and stressful life events. Patients concerned about the possibility of cancer are unlikely to have symptom resolution without appropriate counselling.
It is important to see the patient’s current symptoms in the context of their overall health. It is therefore logical to address lifestyle factors known to promote dyspepsia and to impact on health in general by taking a comprehensive dietary and drug history and providing appropriate counselling (table 6).Frequent dietary components associated with reflux include coffee, orange juice, alcohol and fatty and spicy foods. Patients with suspected reflux should also be advised to avoid eating large meals just before going to bed and to avoid tight-fitting clothes after meals. A wedge pillow or elevation of the head of the bed can help to reduce nocturnal symptoms. Up to 20% of patients will respond to lifestyle intervention alone. NSAID therapy should be reviewed and stopped if possible. Patients in the high-risk group for peptic ulcer complications should be offered concurrent proton-pump inhibitor therapy and be prescribed the least-toxic NSAID (either ibuprofen or diclofenac), unless they are also high risk for cardiovascular disease, in which case NSAIDs should be avoided, or, if an NSAID is needed, naproxen with a PPI is the preferred choice.
Consensus guidelines for management with NSAIDs are now well established (figure 4).5 It is clear that patients who are H. pylori positive are at higher risk of NSAID-associated ulcer complications and that H. pylori eradication before starting NSAIDs reduces that risk. The data on eradication after an NSAID ulcer complication has occurred are less clear. Another group of patients of particular concern are those with coronary artery stents who are often receiving aggressive antithrombotic therapy with both aspirin and clopidogrel to prevent stent blockage. The GI bleeding risk increases with the number of antithrombotic agents used. Clopidogrel synergistically causes GI bleeding in patients using aspirin, non-selective NSAIDs or anticoagulants and increases the blood loss caused by NSAIDs. There has been recent concern over whether PPIs decrease effectiveness of clopidogrel if concomitantly prescribed. In November 2008 the American Heart Association, the American College of Cardiology and the American College of Gastroenterology issued a joint statement concluding that there was no definite evidence to justify changing clinical practice. A clinical trial is under way to provide more data.
A proportion of patients will respond to an initial 1-2-week trial of acid-suppression therapy without further need for medication, at least for a number of months. A single daily standard dose of a PPI results in a gastric pH >4 about 67% of the time and is sufficient for most patients requiring daily therapy, with a few requiring a twice-daily dose. Others can be well controlled with, and may prefer, ‘on demand’ therapy with acid-suppression therapy or antacids. On-demand therapy has been shown to be effective for a large proportion of patients with GORD, with similar levels of patient satisfaction to those taking regular acid-suppression therapies. For patients with evidence of oesophagitis on endoscopy, severity of disease can determine treatment needs. Patients with severe oesophagitis are more likely to develop local complications such as stricturing and bleeding. They should be maintained on PPI therapy to avoid these complications. The expected success rates for medical treatment of GORD vary widely. Recognised approaches, from the most successful to the less successful, are as follows:
There are few data showing a step-down approach to be more cost effective than a strategy of targeting therapy to symptoms. Prokinetic therapy may be useful for patients with dyspepsia unresponsive to acid-suppression therapy. Options include domperidone and metaclopramide. Treatment for functional dyspepsia is difficult. Initial therapy should comprise acid suppression, with subsequent use of prokinetic agents, antispasmodics, antidepressants and behavioural therapy or psychotherapy considered for resistant symptoms.
Postmarketing surveillance data for use of H2 -receptor antagonists compared with community controls show them to be relatively safe. Long-term postmarketing surveillance data for PPIs are not yet available despite the first PPI being released in 1988. The most common side effects of PPIs are headache and diarrhoea and they are uncommon. Other rarer but serious reported side effects are interstitial nephritis, which is not always reversible and, hypomagnesaemia. PPIs act by reducing gastric acid production. Reduced gastric acid promotes the growth of enteric and swallowed flora in the proximal gut and leads to a small increased risk of community-acquired pneumonia and of enteric infection. An acid environment is also needed for calcium absorption, and longterm PPI use may predispose to osteoporosis-associated hip fracture. The medications appear to be safe in pregnancy.
Two H pylori-post-testing treatment strategies have been recommended:
• H. pylori-negative patients: treatment with a PPI.
• H. pylori positive patients: either— referral for endoscopy, or — H. pylori eradication treatment, with subsequent referral only of those who have persistent symptoms after a trial of acid-suppression therapy (see table 4).
The strategy of referral for endoscopy is based on the association between H. pylori infection and peptic ulcer and gastric cancer, and increases the diagnostic yield of selective endoscopy. The H. pylori eradication strategy is based on the principle that H. pylori eradication will cure non-NSAID ulcer disease, and acid-suppression therapy most of the remainder. Antibiotic intolerance, side effects and the impact of inappropriate use of antibiotics on community antibiotic resistance are some of the disadvantages of H. pylori-eradicationbased strategies. A meta-analysis of H. pylori-eradication studies in patients with nonulcer dyspepsia concluded that H. pylori eradication therapy has a small but statistically significant effect in H. pylori-positive non-ulcer dyspepsia.
The number needed to treat to cure one case of dyspepsia was 14. The number needed to harm was not calculated.7 The benefits of secondary prevention of gastric cancer through H. pylori eradication are theoretical, although biologically plausible, but there are no data to indicate optimal time for treatment to be effective. It is reasonable to offer H. pylorieradication therapy to patients with dyspepsia and a family history of gastric cancer, as a history of a first degree relative with gastric cancer increases the risk threefold.
For H. pylori infection the main issues surrounding treatment are which regimen to use and whether eradication of infection needs to be confirmed. Controversy remains as to the preferred combination of antibiotics to use for eradication therapy. PPI-based therapies have been shown to have a higher efficacy and to be better tolerated than bismuth-based therapies. The main PPI-based combination packs are Nexium Hp7 and Klacid Hp7, seven-day regimens of twicedaily clarithromycin 500mg, amoxycillin 1g and esomeprazole 20mg or omeprazole 20mg. In the event of allergy or resistance to either of these antibiotics, a second prescription for metronidazole 400mg twice daily can be written and the offending antibiotic substituted. All regimens involve antibiotics to which H. pylori has been demonstrated to develop resistance. Australian studies show resistance for clarithromycin is about 6%.
In vitro metronidazole resistance has been shown to be up to 40%. The need to confirm the success of eradication therapy has been questioned because of the high efficacy of eradication therapies and the cost involved detecting few failures. A pragmatic approach is to confirm eradication for patients who have had complicated ulcer disease, and who, because of comorbidity, would poorly tolerate ulcer complications. Options to detect failed eradication are limited to endoscopic methods (which are invasive), and urea breath testing. There is a prolonged antibody response to infection, making serology unhelpful. Patients failing H. pylori eradication can be referred to a gastroenterologist to determine the most appropriate antibiotic combination or for consideration of H. pylori culture and sensitivities, if the initial indication for H. pylori eradication was justified. After eradication, the risk of re-infection is small (of the order of 0.5-1% per annum). In the absence of aspirin or NSAID use, eradication therapy virtually cures H. pylori-associated peptic ulcer disease without any further treatment.
After successful eradication, about 5% of ulcers recur over a 6-12-month period, compared with 25% of those given H2 –receptor antagonists, and just under 75% of those untreated. It needs to be remembered that a proportion of patients will have concurrent GORD and may therefore need treatment for this.
Effective acid-suppression therapy has meant that surgery is rarely indicated for ulcer disease except in the presence of complications. Indications for anti-reflux surgery (fundoplication) include:
Up to 80-90% of patients have good to excellent outcomes at 10 or more years. Laparoscopic fundoplication has less short-term morbidity than open surgery.
Despite reports of benefits for both psychotherapy and hypnotherapy for functional dyspepsia, this is not confirmed in formal studies.
Some patients are dissatisfied out of proportion to their residual symptoms after treatment. A major cause of dissatisfaction is unmet expectations. A US survey of patients visiting primary care physicians for various reasons found 98% had at least one pre-visit expectation.8 The most frequent expectations were:
Resurvey immediately after their visit found the most common unmet expectations were prognostic information (51%) followed by diagnostic information (33%). Of note, patients with no unmet expectations had less worry about serious illness (54% vs 27%), significantly greater satisfaction (59% vs 19%) and were more likely to have symptom alleviation. A study of 91 patients starting PPI therapy for gastro-oesophageal reflux symptoms found that patients expected
Only four patients had no real expectations of the medication. There are clearly common themes in regard to what patients want to know and most patients want to be involved in their care.
All patients with alarm symptoms and those who, on clinical grounds, are suspected of having peptic ulcer disease should be referred for endoscopy before acid-suppression therapy. It is also not unreasonable to refer the elderly patient with new-onset dyspepsia in the absence of a medication change. Whether to refer the patient with longstanding symptoms of GORD for screening for Barrett’s oesophagus remains controversial. The higher-risk group are male smokers with longstanding and severe symptoms. If referred for surgery, these patients should be fit enough for the procedure.
The largest referral group will be those who have not responded to high-dose acid-suppression therapy, and these can be the most difficult group to treat. A major role for the gastroenterologist is to exclude other pathology and to document oesophageal acid exposure by pH testing. For some patients, concern about the significance of their symptoms may only be allayed by referral to a specialist.
The natural history of GORD is not well documented. Most data come from patients with symptoms severe enough to warrant referral to secondary care. Some community-based longitudinal studies report a steady prevalence, with the proportion of affected individuals having symptom resolution being matched by a similar proportion becoming symptomatic. These studies also report “poorer quality of life”. A systematic review of patient outcomes for GORD found 40% of those taking prescription medications reported residual symptoms.10
H. pylori eradication virtually cures duodenal ulcer disease, with relapse rates at one year of about 5%. Without treatment, about 40% of duodenal ulcers and about 30% of gastric ulcers heal at four weeks, but 50-80% recur in the subsequent 6-12 months. It is controversial whether recurrences continue indefinitely.
Functional dyspepsia has a high placebo response of up to 50%, consistent with a generally good prognosis for this disorder. Longitudinal studies of functional dyspepsia that have followed patients for more than a decade show a large amount of overlap with IBS and a high rate (up to 40%) of change of predominant symptomatology from the upper to lower gut, and vice versa.
Several endoscopic anti-reflux techniques are under evaluation and may obviate the need for surgery in people with severe GORD. This is unlikely to have a major impact at a population level. In contrast, the rise in obesity — a risk factor for reflux — may result in a substantial rise in the number of consultations for GORD in the future and in turn the incidence of oesophageal adenocarcinoma. Whether the fall in H. pylori prevalence can be implicated in these events remains to be determined.
Gastroenterological Society of Australia:
www.gesa.org.au/pdf/Healthy_Gut_2nd_03.pdf, www.gesa.org.au/pdf/Heartburn_3Ed_07.pdf, www.gesa.org.au/pdf/ H_Pylori_2nd_03.pdf
Dr Anne Duggan
gastroenterologist and associate director, clinical governance, Hunter New England Area Health, and conjoint professor, school of medical practice and population health, The University of Newcastle, NSW.