Current dogma dating back to the 1970s suggests that the development of cancer occurs through clonal selection such that the primary tumour contains all the genes necessary to invade and metastasise.1 In this model heterogeneity within the tumour cell population arises as a result of genetic instability and as a result a variety of non-causative or hitchhiker mutations may be observed.2 Barrett’s oesophagus is an ideal system in which to study the genesis of cancer in the human since the oesophagus is readily accessible to biopsy and the metaplasia–dysplasia sequence can be observed longitudinally.

To date, detailed molecular studies of the Barrett’s epithelium have tended to depend on the analysis of an entire snap-frozen endoscopic biopsy or alternatively paraffin-embedded sections are probed for gene expression status using immunohistochemistry or in situ hybridisation methods. Although these in situ methods enable the construction of spatial maps this is at the expense of detailed information regarding the mechanisms underlying loss or gain of gene function. Advances in molecular biology mean that it is now possible to extract high-quality DNA from archival formalin-fixed paraffin-embedded material. Combined with microdissection and powerful DNA amplification techniques this enables analysis of the molecular status of specific genetic loci on a crypt-by-crypt basis with preservation of spatial information so that clonality can be determined. Using this innovative approach …