Poster Presentation 37th Lorne Cancer Conference 2025

High-throughput design of hypermethylation markers for Barrett’s oesophagus with dysplasia and oesophageal adenocarcinoma (#175)

Yi Jin Liew 1 , Thejaani Udumanne 2 , Kyle Catalan 1 , Amali Ariyavidana 1 , Michelle Lee-Ng 2 , Dana Pascovici 1 , Sarah J Lord 2 3 , Reg V Lord 2 4 , Jason Ross 1
  1. CSIRO Health & Biosecurity, Westmead, NSW
  2. St. Vincent’s Centre for Applied Medical Research, Sydney, NSW
  3. NHMRC Clinical Trials Centre, University of Sydney, NSW
  4. Department of Surgery, University of Notre Dame School of Medicine, Sydney, NSW

Background: Oesophageal adenocarcinoma (OAC) has increased markedly in Australia over the last 50 years, and remains one of the most lethal cancers. There is no screening test for detecting OAC at an earlier, treatable stage; histopathological detection of dysplasia is unreliable with poor inter-observer agreement.

The main risk factor for OAC is Barrett’s oesophagus (BO), which is the metaplastic replacement of the normal squamous epithelium with intestinal-like columnar epithelium. Annually, <0.5% patients with non-dysplastic Barrett’s oesophagus (NDBO) progress to OAC via low-grade dysplasia (LGD) and high-grade dysplasia (HGD); this progression is associated with changes in DNA methylation.

Aim: Develop a PCR-based test via high-throughput screening that accurately discriminates HGD/OAC relative to NDBO by:

  1. Computational identification of differentially methylated regions (DMRs) in HGD/OAC from solid tissue biopsies
  2. Software design of PCR primers targeting regions of interest
  3. Mechanise qPCR experiments for speed and cost efficiencies
  4. Translate promising qPCR amplicons into digital PCR (dPCR) format

Methods: Whole genome enzymatic methyl-seq (EM-seq) was performed on n=86 fresh-frozen oesophageal tissue biopsies (10 normal squamous, 29 NDBO, 19 LGD, 16 HGD, 12 OAC) from 72 patients in the PROBE-NET study. DMRs were identified with metilene1.

Results: The vast majority of DMRs were hypermethylated in HGD/OAC relative to NDBO (1,478 hypermethylated vs. 67 hypomethylated, default metilene cutoffs). Whilst some DMRs overlap genes previously reported as differentially methylated e.g., CDKN2A2,3,4, MGMT2,3, RUNX33,4 and TIMP32,3, most DMRs are in genes unique to our study.

The design of primers for methylation-specific amplicons is challenging and hit-and-miss. Our automated primer design procedure also generates an order form to ease ordering; upon arrival of the primer plate, the PCR reactions are set up with an EpMotion liquid handler. Primers were initially screened against control DNA and NDBO/OAC pools; good performers were then screened against an expanded panel containing more precious clinical samples.

Conclusion: We believe our use of automation and smart screening strikes a balance between dogmatic primer design and aimless trial-and-error. Errors and labour-intensive steps are reduced, research dollars are stretched further. Our approaches can also be applied to high-throughput screening of somatic mutation assays.

  1. Juhling F, et al. metilene: fast and sensitive calling of differentially methylated regions from bisulfite sequencing data. Genome Res. 2016 Feb;26(2):256-62.
  2. Eads CA, et al. Epigenetic patterns in the progression of esophageal adenocarcinoma. Cancer Res. 2001 Apr 15;61(8):3410-8.
  3. Smith E, et al. Similarity of aberrant DNA methylation in Barrett's esophagus and esophageal adenocarcinoma. Mol Cancer. 2008 Oct 2;7:75.
  4. Jin Z, et al. A multicenter, double-blinded validation study of methylation biomarkers for progression prediction in Barrett's esophagus. Cancer Res. 2009 May 15;69(10):4112-5.