• Barrett's oesophagus
• minichromosome maintenance proteins
• oesophageal cancer
• oesophageal squamous epithelium

Oesophageal cancer contributes about 3% of the cancer burden in the UK, 5% of cancer mortality, and the five year survival is a dismal 6% (www.crc.org.uk). There are two major types of oesophageal carcinoma, squamous cell carcinoma (SCC) and adenocarcinoma, each with different risk factors and epidemiologies. SCC arises from squamous cells lining the oesophagus and the geographical distribution of the disease shows wide variations, being virtually unknown in North Africa but common, for example, in eastern Turkey, Iraq, Iran, and northern China; high risk areas are generally associated with local food preservation practices that favour the generation of nitroso compounds from mould growing on pickled vegetables. In Western populations, heavy alcohol and cigarette consumption are well known risk factors for oesophageal SCC. Most adenocarcinomas on the other hand appear to arise from within areas of metaplasia known as Barrett's oesophagus, the metaplasia probably being caused by prolonged reflux of gastric acid and digestive enzymes (reflux oesophagitis). With the passage of time the epithelial lining becomes progressively more abnormal as it passes through a series of sequential steps that eventually result in the development of invasive adenocarcinoma. These steps include the development of glandular dysplasia, signalled by an increased nuclear:cytoplasmic ratio and loss of nuclear polarity within the cells lining the metaplastic glands: presumably morphological correlates of the underlying genetic alterations (commonly seen in neoplastic progression in other tissues) that are found in such glands.¹

In the colon, the adenoma (by definition dysplastic)-carcinoma sequence is reasonably well understood, with large severely dysplastic villous adenomas having the most sinister reputation. Moreover, prophylactic removal of adenomatous polyps is the “norm”, such is the seemingly inevitable progression of such lesions. The natural history of oesophageal dysplasia to invasive adenocarcinoma is not as clearly defined, and herein lies the problem. For example, two recent articles report somewhat different outcomes for severe dysplasia: Buttar et al found a cumulative cancer incidence at three years of 56% among patients initially presenting with diffuse (affecting >5 crypts or in multiple biopsies) high grade dysplasia,² whereas Schnell et al found a five year cumulative cancer incidence of only 9% among a group of 79 high grade dysplasia patients.³ These apparent discrepancies may be partly attributable to the ways in which the pathology is “read”⁴; nevertheless, high grade dysplasia is an ominous finding. In this issue of Gut, Going et al report the aberrant overexpression of proteins called minichromosome maintenance (Mcm) proteins, along with Ki-67, at the mucosal surface of both dysplastic oesophageal squamous epithelium and dysplastic Barrett's mucosa [see 373].⁵ As expression of Mcms was observed in almost all surface cells in high grade dysplasias, anti-Mcm antibodies may thus have a significant role in the recognition of this important precancerous state.

Initiation of DNA synthesis in eukaryotes is a complex multistep process involving the sequential loading of initiation factors into prereplicative complexes (pre-RCs) at replication origins, resulting in particular regions of chromatin being “licensed” for replication in the ensuing S phase.^6–,⁹ The process begins with the binding of the origin recognition complex (ORC), and recruitment of Cdc6 and Mcm2–7 (minichromosome maintenance). The Mcms are highly conserved and were originally discovered and named as factors that supported minichromosome maintenance in yeast; the assembled Mcms are presumed to act as an enzymatically active (DNA unwinding) helicase.¹⁰ As cells enter S phase, Cdc6 is released and other factors are added to the replication origin to initiate DNA replication; critically, Mcm proteins gradually dissociate from chromatin as the DNA is replicated. This negative regulation ensures that each region of DNA is replicated only once during a single cell cycle because replicated DNA lacks functional pre-RCs.

In non-neoplastic tissues, expression of Mcms is generally confined to the proliferative compartment in a pattern similar to conventional proliferation markers, such as Ki-67.^{11, 12} As such, conventional proliferation markers have not made a huge impact on tumour pathology diagnosis or cancer management, although they have been useful in areas such as the differential diagnosis of smooth muscle tumours, grading of soft tissue sarcomas, and prediction of metastases in thin melanomas. However, Mcms are not your average proliferation marker; even in some normal tissues such as premenopausal breast, Mcm expression far exceeds that of Ki-67 and may identify mammary gland progenitor cells.¹² Thus Mcms may be biomarkers of cells with replication potential (licensed to cycle!). In this respect, Mcms came to the attention of histopathologists/cytologists from work originating from the same stable as the current study of Going et al. In CIN of the ectocervix, where Ki-67 labelled approximately 10% of surface epithelial cells, by contrast Mcm5 and Cdc6 expression was observed in almost all surface cells. Such a huge discordance suggested that Mcm expression had utility as a marker of dysplasia and may identify that rare (“litigation”) cell that might be missed by more conventional staining.¹³ Likewise, in a variety of other dysplastic states (actinic keratosis, Bowen's disease, colonic tubulovillous adenomas) Mcms are more highly expressed than Ki-67, for example.¹⁴ The present study by Going et al also suggests that Mcm expression may be a useful adjunct in identifying dysplasia in the oesophagus although curiously they did not observe the large discrepancy between Ki-67 and Mcm labelling seen in the cervix, epidermis, and colon. However, in common with previous studies,¹⁴ Mcm expression was inversely correlated with tumour differentiation (keratinisation), supporting the notion that Mcms are indicative of proliferation potential.

The incidence of Barrett's oesophagus is increasing rapidly in the Western world and identification of high grade dysplasia within an area of Barrett's oesophagus has profound implications for the patient and gastroenterologist alike. If there is a method of surveillance for high grade dysplasia within patients who have Barrett's oesophagus that is practical, sensitive, and specific, and there is a defined treatment option which results in increased survival, then it would seem reasonable to pursue this surveillance programme. Although surveillance guidelines have been produced by the Practice Parameters Committee of the American College of Gastroenterology,¹⁵ other workers have found little enthusiasm for pursuing unselected Barrett's surveillance programmes.¹⁶ This is where detection of Mcm expression may prove useful; Mcm expression along with other molecular markers may help in stratifying patients into low and high risk groups.¹⁷ Furthermore, within an area of Barrett's oesophagus the endoscopist cannot biopsy all tissue and hence Mcm expression may allow detection of deregulated cell proliferation in a wider area than can be recognised on morphological criteria alone or perhaps allow brush cytology sampling of a wide area.

In summary, evaluation of Mcm expression probably represents an incremental step in the armamentarium for the detection of oesophageal dysplasia (particularly of high grade type) in Barrett's oesophagus, and thus may help facilitate the identification of a cohort of patients most at risk of disease progression.