Biomarker qualification in Barrett s oesophagus

A type of oesophageal or gullet cancer, called adenocarcinoma, is rapidly increasing in the western world. Most patients do not seek medical attention until they develop swallowing difficulties and by then the cancer is usually advanced with a 5-year survival of 20%. However, if the cancer is detected early then the survival rate can be improved to 80%. These cancers usually develop as a complication of a condition called Barrett?s oesophagus. However, most people in the population with Barrett?s oesophagus will not develop cancer. Therefore research is needed to identify the patients at greatest risk for oesophageal cancer. In this grant we are testing some promising molecular tools to see whether they can predict cancer progression using a large clinical biopsy resource from the Northern Ireland Barrett s Oeospagus Register. If we could find a molecular test that predicted cancer development this would prevent many patients with Barrett s oeosphagus having unecessary endoscopy examinations and enable us to target resources to patients at highest risk.

Introduction: The incidence of oesophageal adenocarcinoma (OAC) is increasing dramatically in the west and has a very poor prognosis. Barrett?s oesophagus (BE) is an established risk factor for OAC. Currently, BE patients are offered regular endoscopy (surveillance) to detect cancerous changes early. This hinges on the subjective histopathological assessment of multiple, random biopsies which is problematic. Biomarkers have the potential to reduce sampling bias and improve the risk-stratification of patients. Since endoscopy is invasive, unpleasant and costly, ideally biomarkers should be developed for non-endoscopic applications. Biomarkers for this study have therefore been selected on the basis of their mechanistic role, abundance and ease of detection.
Hypothesis: Molecular biomarkers (aneuploidy, cyclin A, lectins) have the potential to risk stratify BE patients. In addition because these changes occur at the epithelial surface and are abundant there is the potential to develop these markers for cytological screening (aneuploidy, cyclin A) and for novel fluorescent endoscopic diagnostic methods (lectins).
Aims: 1. Determine whether specific biomarkers can predict cancer risk in a highly characterised, population-based BE cohort. 2. Use biomarkers, appropriately qualified in phase 1, to develop: a) a quantitative assay for assessment of biomarker(s) from a cytological specimen, b) a lectin-probe suitable for fluorescence imaging detectable at endoscopy.
Study design: Phase 3 retrospective longitudinal repository study using a nested case-control approach within the N Ireland BE Register (NIBR). The NIBR includes every adult identified with BE within NI (population 1.7 million) between 1993-2005. Cases are 108 BE patients who have progressed to OAC or HGD, compared with 3 controls per patient (matched for age, sex and year of diagnosis) who have not progressed to these states; (OAC incidence 0.4% per year, HGD incidence 0.08% per year) from a total of 23,771 person years of follow-up. The expression of cyclin A (immunohistochemistry), lectins (HPA, PNA, WGA, CON-A), (fluorescence histochemistry) and aneuploidy (image cytometric analysis) will be determined in biopsies at specific time intervals. For the assay development, cytological samples (n=1000) will be taken from ongoing surveillance programmes in Cambridge, Belfast and UCLH. These patients are well characterised with longitudinal data over 5 years. This well established cohort affords our industrial partner GSK the opportunity to examine genetic aberrations in ERBB2 as a prelude to a clinical chemoprevention trial.
Clinical benefit: The use of biomarkers will benefit patients with BE by enabling accurate assessment of cancer risk and providing reliable surrogate endpoints for intervention studies.