Should patients with Barrett's oesophagus be kept under surveillance? The case for

Oesophageal adenocarcinoma is associated with high mortality rates and its incidence is increasing more rapidly than any other gastrointestinal cancer in the Western world. Several factors, including gastro-oesophageal reflux disease, smoking, alcohol and male gender, are associated with oesophageal adenocarcinoma but none can be used to identify accurately those individuals who will develop adenocarcinoma. It is generally accepted that oesophageal adenocarcinoma arises predominantly in Barrett's oesophagus and it is arguable that Barrett's oesophagus is currently the only clinically useful predictor of oesophageal adenocarcinoma.

Surveillance – periodic testing to detect adenocarcinoma or its precursor, high grade dysplasia – is widely recommended for patients with Barrett's oesophagus with the aim of reducing mortality from oesophageal adenocarcinoma. The annual incidence of oesophageal adenocarcinoma in patients with Barrett's oesophagus is 0.5%–1.0% although there is marked variation between studies, attributable variously to publication bias, concurrent acid suppression therapy and differences in patient characteristics.

There is limited evidence that surveillance reduces the incidence of oesophageal adenocarcinoma or consequent mortality and the cause of death for patients undergoing surveillance is often unrelated to oesophageal disease. There are, nonetheless, observational studies which suggest that surveillance is associated with earlier detection of malignancy and a reduction in mortality; in addition, data from modelling studies suggest that surveillance can be cost-effective. Furthermore, the advent of new, non-surgical treatments (endoscopic mucosal resection, photodynamic therapy, argon plasma coagulation) for high grade dysplasia and early cancer has reduced the risks associated with therapy for disease detected during surveillance.

Surveillance programs have high drop out rates and, for patients who continue surveillance, adherence to standard, published protocols is highly variable. The establishment of specialist Barrett's oesophagus surveillance programs, with coordinator support, has considerable potential to improve adherence to current guidelines, pending the acquisition and publication of data from ongoing studies of chemoprophylaxis and surveillance in the management of Barrett's oesophagus.

In consequence, although there is a paucity of data providing unequivocal demonstration of benefit, there is no proof that surveillance is ineffective. It is, therefore, appropriate to offer surveillance for Barrett's oesophagus in accordance with locally-applicable published guidelines after a full informed discussion of the risks and benefits of surveillance and therapy; continued participation should be reviewed regularly to accommodate changes in the patient's health and expectations.

Introduction

Oesophageal adenocarcinoma is a lethal disease with a median survival of less than 1 year and limited improvement in mortality over the last 30 years.¹ With an estimated incidence of 6500 in the United States, oesophageal adenocarcinoma is not the most prevalent of malignancies but its incidence is increasing more rapidly than that of any other gastrointestinal cancer in the Western world.2, 3, 4, 5, 6

Ideally, strategies directed at preventing deaths from oesophageal adenocarcinoma would be based on preventing the development of oesophageal adenocarcinoma; however, despite progress in our understanding of the mechanisms underlying the development of oesophageal malignancy,2, 7 initial management strategies are still based on early detection of malignancy or its precursors. Predictors of oesophageal adenocarcinoma include gastro-oesophageal reflux disease (GERD), smoking, male gender, age, ethnicity and alcohol. GERD is too common to be used as a basis for screening or surveillance and is not necessarily highly predictive of oesophageal adenocarcinoma; similarly, other risk factors are also too common to be used as a basis for screening.

It is generally accepted that oesophageal adenocarcinoma arises predominantly, if not exclusively, in Barrett's oesophagus8, 9, 10 and that there is a stepwise progression from metaplasia, through dysplasia to invasive carcinoma.⁷ Indeed, it is arguable that Barrett's oesophagus is, currently, the only clinically useful predictor of oesophageal adenocarcinoma.¹¹ The existence of this relationship is supported, in part, by the observations that Barrett's oesophagus is associated with gastro-oesophageal reflux disease and that Barrett's oesophagus is associated with more severe, more frequent and more prolonged GERD symptoms. Furthermore, there is increased reporting or diagnosis of Barrett's oesophagus12, 13, 14, 15 analogous to the reported increase in prevalence of oesophageal adenocarcinoma.16, 17

Given the presence of a relatively common, lethal disease and the availability of potential markers of disease susceptibility, there is an opportunity for population screening to identify individuals in whom the disease or a marker is present and for continued surveillance of those in whom the disease marker has been identified.