Gastric cancer has been declining for more than half a century, whereas the incidence of oesophageal
cancer is increasing rapidly. The histopathological subtype is also changing with a predominance of oesophageal
adenocarcinoma compared with
squamous carcinoma. The reasons for these epidemiological changes are not clear, although population-based data have implicated
gastro-oesophageal reflux disease as a risk factor. In susceptible individuals reflux of duodeno-gastric contents can lead to the development of a columnar-lined oesophagus, commonly called Barrett's oesophagus. This can then progress to
adenocarcinoma via a
metaplasia-dysplasia-
carcinoma sequence. At the current time, the mortality from oesophageal
adenocarcinoma exceeds 80% at 5 years. Therefore, endoscopic surveillance programmes have been generally recommended for patients with Barrett's oesophagus in an attempt to detect early, curable lesions. Unfortunately these programmes are cumbersome and costly and have not yet been proved to reduce population mortality. In order to improve patient outcomes we need to be able to identify patients at high risk and to understand the triggers for
disease progression. There is mounting evidence that there is an underlying
genetic susceptibility to Barrett's oesophagus and oesophageal
adenocarcinoma. However, this is likely to be as a result of multiple low penetrance susceptibility genes which have yet to be identified. Once patients are identified as having Barrett's oesophagus their chance for developing
adenocarcinoma is in the order of 0.5%-1% per year. The histological assessment of dysplasia as a predictor of
cancer development is highly subjective. Therefore multiple, specific somatic mutations in the tissue have been investigated as potential
biomarkers. The most promising markers to date are the presence of
aneuploidy, loss of heterozygosity of p53 and
cyclin D1 overexpression. However, a study of evolutionary relationships suggest that mutations occur in no obligate order. Combinatorial approaches are therefore being advocated which include genomic profiling or the use of a panel of molecular markers in order to define the common molecular signatures that can then be used to predict malignant progression. An alternative approach would be to use markers for the final common pathway following genetic instability, which is the loss of proliferative control. We have demonstrated an increase in the expression of a novel proliferation marker, Mcm2, which occurs during the malignant progression of Barrett's oesophagus. These Mcm2-expressing cells are detectable on the surface, and hence a cytological approach may be applicable. In view of the role of reflux components in the pathogenesis of Barrett's oesophagus the effect of
acid and bile on the cell phenotype have been studied. These studies have demonstrated that pulsatile
acid and bile exposure induce cell proliferation. The mechanism for the hyperproliferative response appears to involve
p38 mitogen activated protein kinase (MAPK) pathways as well as
protein kinase C (PKC) and cyclo-
oxygenases. A clinical implication of the laboratory studies is that suppression of
acid and bile may need to be profound in order to suppress cell proliferation and, by inference, ultimately prevent the development of dysplasia. There is some support for this concept from short-term clinical studies, and a large randomised
chemoprevention trial is being instigated which will evaluate the effect of
proton pump inhibitors with or without
aspirin. Given the epidemic increase in oesophageal
adenocarcinoma and the dismal 5-year mortality rate, a radical approach is necessary to prevent
cancer development in individuals with pre-malignant lesions.