Editorial: One small step for metaplasia, but one giant leap for biomarkers is needed.

There are 5 to 6 levels of biomarker validation. Those for Barrett's esophagus are currently at level 3, despite small prospective studies. What is ideally required is a very large prospective assessment of biopsies in large cohorts, such as the ASPirin Esomeprazole Chemoprevention Trial (AspECT) and Barrett's Oesophagus Surveillance Study (BOSS) trials, so that unbiased and random selection of cases can be subjected to rigorous pathology and biomarker assessment (level 4). Only then can the predictive power of the data be exploited in a randomized intervention trial (level 5) whereby a series of biomarkers would trigger therapy. The real trouble is that this spot is currently occupied, satisfactorily according to some researchers, by conventional histological identification of high-grade dysplasia (HGD) as used in a recent randomized study of ablation in Barrett's esophagus (BE).