Devising the ideal test for early diagnosis of cancer: lessons from the oesophagus

Rebecca Fitzgerald1

1Univeristy of Cambridge, Cambridge, UK

Abstract

Oesophageal adenocarcinoma (OAC) is the most prevalent form of oesophageal cancer in the western world. OAC develops gradually from the benign metaplastic lesion Barrett's oesophagus. Due to the increased risk of developing adenocarcinoma and its exceptionally poor prognosis regular endoscopy (surveillance) is recommended for patients with Barrett's. The current endoscopic surveillance regimes are confounded by sampling bias and a subjective diagnosis of dysplasia.

One clinical approach to earlier diagnosis would be more systematic screening for Barrett's oesophagus and to couple this with molecular risk stratification. We have made progress in this area by developing a non-endoscopic cell collection tool called the Cytosponge coupled with a diagnostic biomarker specific to Barrett's oesophagus called Trefoil Factor 3 (TFF3). Additional biomarkers can then be applied to the same Cytosponge sample to stratify patients according to their likelihood for progression to adenocarcinoma and this information can be used to inform the requirement for surveillance or endoscopic therapy.

As part of the International Cancer Genome Consortium we are performing genomic analyses on unique sample cohorts of OAC and Barrett's oesophagus to precisely determine the order of mutational events in the evolution of the disease. To date WGS data has demonstrated that the OAC and Barrett's genomes are highly mutated with a unique mutational pattern which may reflect specific environmental exposures including the reflux of acid and bile. Furthermore, there are few recurrent driver gene mutations observed and the disease seems to be dominated by chromosomal rearrangements possibly as a result of catastrophes acquired during punctuated evolution. TP53 is the only truly recurrent mutation and this occurs at the stage of high grade dysplasia. As cancer develops the copy number increases markedly, including genome doubling events. These findings inform emerging molecular diagnostic strategies to distinguish between patients at low and high risk.