Abstract
Barrett’s oesophagus (BO) is a pre-malignant condition for oesophageal adenocarcinoma (OAC) [1], the incidence rate of which has been rising dramatically over the last four decades in the western world [2, 3]. Survival of an OAC affliction is poor, and one of the ways to improve this would be by focusing on identifying high-risk Barrett’s patients through surveillance programmes. Currently, histological dysplasia is the only recognized marker of progression to OAC. However, this method is hampered by many persistent obstacles.Molecular biomarkers found in tissue samples may act as a reliable tool for the stratification of patients with BO, based on risk of progression. This may allow a better utilization of resources to produce a sound targeted surveillance programme that could translate into a better survival of OAC.The literature was systematically reviewed for evidence around biomarkers of BO in order to identify the best available markers to date. Next, the data was meta-analysed to look for the most promising markers and to further assess their usefulness. We then tested some of the available markers in a pilot for patients from the Gwent Barrett’s registry, in order to examine their possible clinical utility.
Many biomarkers that could act as potential markers of progression were identified, either individually or in a panel form. The biomarker p53 was found to be the most commonly used marker but its accuracy was found to be limited by its low sensitivity. We believe that joint interpretation of biomarkers may be more useful in the detection of high risk BO. We have noted a mirroring action of p53 and cyclin D1. In addition, aberrant expression of p53 and p16, along with strong expression of cyclin D1 may indicate the presence of cancer. We found the accuracy of a model that involved the use of p53, p16 and cyclin D1 in the
diagnosis cancer/high-grade dysplasia to almost reach the 90% threshold, with an improved sensitivity of 65% (42- 79%) and a high specificity of 93% (82-97%). Long-term follow up would be needed to establish the usefulness of such approaches in the prognostication of BO. Our work suggests that there would be a need for novel markers to be tested further, in the context of longitudinal studies.
| Date of Award | 8 Jan 2019 |
|---|---|
| Original language | English |
| Supervisor | Joyce Kenkre (Supervisor) & Peter McCarthy (Supervisor) |