Dyspepsia is a common reason for primary care consultation and secondary care referral. Yet, despite a national guideline¹ to help identify patients who merit referral for endoscopy (table 1, attached), concerns remain that underlying upper gastrointestinal (GI) malignancy may be overlooked in young patients with non-specific symptoms.

Dyspepsia or gastrooesophageal reflux?

There is a variety of definitions for dyspepsia; the most commonly used is that dyspepsia covers all symptoms suggestive of an origin in the proximal GI tract. However, this definition does not differentiate between dyspepsia and gastrooesophageal reflux, yet the management of one differs markedly from the other; table 2, attached, shows the symptoms of dyspepsia and of reflux.

Of all patients with dyspepsia, on review 27% will be found to have erosive oesophagitis or symptoms more in keeping with acid reflux.

NICE Recommendations

The NICE guideline on dyspepsia, issued in 2004, advocates a 'test and treat' approach to the initial management of dyspepsia, either by a trial of a proton pump inhibitor (PPI) or by testing for Helicobacter pylori followed by eradication therapy.

The exceptions to this are patients with clear symptoms of gastro-oesophageal reflux disease and those with symptoms or signs suggesting the need for upper gastrointestinal endoscopy (table 3, attached).

Symptoms suggesting malignancy

Estimates vary as to the percentage of patients with dyspepsia who, on investigation, are found to have a malignancy. Of all patients with gastric malignancy, less than 8% are younger than 55 years. Within this group, close questioning reveals that 96% have 'red flag' symptoms.^2,3 In one study of 2,867 patients younger than 45 without alarm symptoms, four (0.14%) had oesophagogastric neoplasia.⁴ In older populations, however, some studies suggest that up to 25% of all patients with upper GI cancers do not have any associated red flag characteristics.

One of the problems in making a clinical diagnosis of malignancy in the upper GI tract is the poor correlation between symptoms and the presence of a cancer. Whereas dysphagia and weight loss are indicators of oesophageal malignancy, their positive predictive values (PPV) are low: 6.6% for dysphagia and 7.8% for weight loss. The strongest indicators for gastric malignancy are early satiety, with a PPV of 9.9%, and weight loss, with a PPV of 8%.

Overall, however, even the pooled predictive value of 'any red flag symptom' is low; only 5.9% in a recent meta-analysis.⁵ These findings emphasise that clinical symptoms are poor indicators of cancer.

Physical findings, for example a palpable epigastric or abdominal mass or lymphadenopathy in the supra-clavicular fossa, can be useful, but tenderness alone is more likely to result from erosive oesophagitis than from a cancer.

Laboratory investigations are of limited use; iron-deficiency anaemia (a hypochromic, microcytic picture) should be of concern, notably in men, or women after the menopause. However, tumour markers, with their high false-positive rate, and faecal occult blood testing, which has a high false-negative rate, have no role in the investigation of symptomatic individuals.

H. pylori

and upper GI pathology

In the 23 years since Warren and Marshall described organisms colonising the stomach and their role in gastric inflammation and ulceration,⁶ the role of H. pylori in both dyspepsia and gastric cancer has been clearly defined.

Carriage rate in UK populations runs at 20-30%, depending on location and socioeconomic status.⁷ The presence of infection confers a relative risk of 3 of developing a gastric cancer;⁸ this fact, and the improvement in dyspeptic symptoms seen in 30-50% of individuals with positive serology who undergo eradication therapy, has resulted in the current 'test and treat' policy.

In patients who respond initially then relapse after therapy, the infection may have failed to clear; some 8-10% of all patients receiving standard therapy will remain H. pylori positive. Alternatively, their disease may be responsive to a proton pump inhibitor (PPI) and show a clinical improvement on triple therapy. A further trial of a PPI alone is therefore appropriate for these patients.

In patients with reflux-predominant symptoms, H. pylori eradication may exacerbate symptoms, because acid is suppressed by the infection but synthesis increases after eradication therapy.

Some data suggest that by the same mechanism H. pylori eradication may increase the risk of oesophageal adenocarcinoma. In view of this, the role of H. pylori eradication is less clear and there is no consensus on it.

In general, with a clear increased risk of gastric carcinoma, H. pylori eradication followed by a PPI for reflux symptoms seems reasonable. However, some authorities advocate, rather than testing for the infection in acid reflux, managing with a trial of a PPI.

H. pylori eradication following therapy should always be confirmed by carbon urea breath testing (prescribable on an FP10). Serology remains positive for up to 12 months after successful therapy.

Who should undergo endoscopy?

The importance of upper GI endoscopy has lessened over the past four years, and the number of procedures carried out is falling in most units. This reflects the fact that the NICE guideline does not recommend endoscopy as a primary diagnostic tool. Who then merits referral for investigation?

Among patients with typical dyspeptic symptoms, those with red flag or alarm features should be referred urgently for endoscopy (see table 3, attached). Those without these features represent the majority. The guideline's recommendation, updated after first publication, is that:

Routine endoscopic investigation of patients of any age, presenting with dyspepsia and without alarm symptoms, is not necessary. However in patients aged 55 years and older with unexplained and persistent recent-onset dyspepsia alone, an urgent referral for endoscopy should be made.

NICE defines 'unexplained' as: 'a symptom(s) and/or sign(s) that has not led to a diagnosis being made by the primary care professional after initial assessment of the history, examination and primary care investigations'.

Persistent symptoms are generally determined to be those lasting longer than 4-6 weeks. In essence, this suggests that all patients older than 55 years with new-onset symptoms of more than four weeks' duration should be referred for urgent endoscopy.

In patients younger than 55 years, endoscopy should only be carried out for atypical symptoms or if symptoms have failed to respond to H. pylori and PPI therapies. Endoscopy is rarely needed in this group of patients, and it may be more appropriate to consider other diseases, for example coeliac disease or irritable bowel syndrome, or management approaches for functional dyspepsia, such as prokinetics.

Conclusion

The NICE guideline has provided a clear protocol for the management of dyspepsia, but doubts remain about the safety of a policy that excludes endoscopy in most patients. However, on close questioning, most patients with malignancy younger than 55 years are found to have red flag symptoms.