Validation of Biomarkers for predicting Barrett's esophagus that will or will not: i) progress towards cancer, or ii) recur after ablation - Abstract
This EDRN-CVC proposal is aimed at the validation of molecular biomarkers for distinguishing high versus low
risk esophageal neoplasias (Barrett’s esophagus) for the purpose of guiding selection and management of
patients for endoscopic eradication therapy (EET). Two validation studies are proposed: the first, a phase 4
prospective study to identify a patient group at low progression risk who can be spared EET; the second, a phase
3 retrospective study to distinguish individuals who following EET are at low versus high risk of disease
recurrence. Barrett’s esophagus (BE) is the precursor lesion of esophageal adenocarcinoma (EAC), a cancer
with 80% lethality whose incidence has increased more than 7-fold in the past three decades. BE progresses to
EAC in a step-wise fashion from non-dysplastic BE, to low grade dysplasia (LGD), to high grade dysplasia (HGD),
and finally cancer. EAC prevention is based on using EET to ablate HGD BE before it can progress to EAC.
However, increasingly, EET is also becoming the default therapy for LGD, a highly imprecise diagnosis about
which expert pathologists frequently disagree, and which is applied to as many as 40% of BE patients at some
point during their course. As EET has a 9% complication rate, the result is an emerging epidemic of overtreatment
of BE with LGD. In a prior EDRN-BDL award, our team developed the “BAD” technology for early detection of
BE progression. In BAD, we used a brushing device to sample a patient’s full BE esophageal segment. We then
analyzed the DNA from this sample using next-generation sequencing technology (developed for liquid biopsy
assays) to instead detect presence of BE clones that had acquired gains or losses on specific driver
chromosomes associated with EAC. Detection of driver chromosome changes (dubbed Very-BAD), typified EAC
and HGD. In contrast, 28% of LGD showed complete absence of any chromosomally aberrant clones (dubbed
Not-BAD). We will now validate Not-BAD as a biomarker that identifies LGD at such low progression risk as to
not require EET. We will do this by partnering with the SURVENT trial, that will be the first U.S. prospective study
to follow LGD patients managed by surveillance, not ablation. A second major challenge with EET is that over
25% of patients recur following ablation (with either high risk BE, HGD, or EAC). These patients face a substantial
burden of post-EET surveillance endoscopies, initially at every 3-month intervals. In our prior EDRN-BDL, our
team identified a panel of methylated DNA biomarkers for sensitive molecular early detection of BE (currently
awarded FDA breakthrough device designation). We have further identified that these markers remain retained
in a subset of patients post-EET. We accordingly now propose a retrospective Phase 3 study to further validate
these DNA markers for molecular assessment of minimal residual disease, whose post-EET elimination identifies
individuals achieving complete molecular eradication of BE, and hence at low risk of disease recurrence and not
in need of intense post-EET surveillance. We do this by partnering with the unique UNC-BEECAB biorepository
of post-EET esophageal biopsies from patients whose disease did or did not recur following ablation.
