Methylation profiling for the prediction of Barrett’s esophagus progression

Article Outline

• Copyright

Patients with Barrett’s esophagus (BE) have an increased risk of developing esophageal adenocarcinoma (EAC). For years, dysplasia grade has been the sole means of risk stratification for patients with BE. Significant problems have emerged in studies of dysplasia that make it imperative for the BE field to incorporate additional detection and stratification markers including poor reproducibility of dysplasia interpretation and sampling error. We performed quantitative methylation profiling of 14 genes using real-time quantitative MSP of longitudinally derived multiple esophageal biopsies from 11 BE patients who progressed to dysplasia and/or EAC (Group P) (median follow-up 38.5 months) and compared them to an age- and sex-matched cohort of 24 BE patients who did not progress to dysplasia or EAC during a median follow-up of 56 months (Group NP). In addition, we analyzed tumor tissues from 30 patients with frank EAC and matched normal esophageal tissue, as well as from 12 patients without esophageal disease. Normal squamous esophageal epithelium, whether from noncancer non-BE patients or from patients with BE or EAC, rarely undergoes hypermethylation at the loci studied. By contrast, frank EAC are frequently methylated at these same sites, often at 5 or more loci. Five of the markers studied are novel methylation targets of EAC and BE (HPP1, CRBP1, RIZ1, RUNX3, and OST-2). Regarding BE, the methylation pattern was significantly different in BE tissues derived from P vs. NP. The number of methylated genes was significantly higher in the P group. Moreover, 3 genes were significantly more frequently methylated in BE tissues derived from P compared with NP (HPP1, TIMP3, p16). A panel of these markers was able to define groups at low, and high risk of progression to dysplasia or cancer. We identified 5 novel methylation targets in BE and EAC. In addition, methylation panel profiling of BE tissues showed significantly different methylation patterns in patients who later progressed to dysplasia and/or cancer (P) vs. patients who did not progress during long-term endoscopic follow-up (NP). These findings may have important impact regarding early detection and risk stratification and may influence current surveillance strategies if confirmed in large prospective trials.