Neosquamous epithelium after radiofrequency ablation (RFA) of Barrett epithelium: Where does it come from?

Barrett esophagus (BE) is a condition in which the normal squamous lining of
the esophagus has been replaced by columnar epithelium containing specialized
intestinal metaplasia (IM), as a result of long-standing gastro-esophageal
reflux. Malignant degeneration in BE progresses through a series of morphologic
changes from IM to low-grade intraepithelial neoplasia (LGIN) and high-grade
intraepithelial neoplasia (HGIN), to early cancer (EC). To prevent malignant
progression, patients with BE can be treated with endoscopic ablation in order
to eradicate the BE.
Radiofrequency ablation (RFA) is a relatively new endoscopic technique for
ablation of BE. Data from clinical studies suggests that RFA is an efficacious
modality for eradication of all Barrett epithelium and associated neoplasia, it
has a low complication rate, preserves esophageal functional integrity, and is
relatively easy to apply. The process of regeneration and factors that
contribute to conversion of BE into neo-squamous epithelium (NSE) after RFA
are, however, poorly understood. Some patients undergoing RFA require only a
single treatment for complete reversion of their BE into NSE whereas others
show poor healing without significant regression. There are several hypotheses
concerning the origin of the NSE after ablation therapy: outgrowth from
existing pools of squamous cell progenitors, repopulation from adjacent areas
with squamous epithelium, or regeneration from multipotent progenitor cells
from the esophageal glands or bone marrow. Further insight of this process of
squamous repopulation after RFA would be valuable for two reasons. Firstly, it
might help to identify patients with an anticipated poor response to the RFA
and/or to intervene in the regeneration process. Secondly, it may answer the
question if eradication of BE by RFA results in clearance of genetic
abnormalities of the epithelium and, thus, a reduction of the risk for
developing cancer.

Aims of this study are:

1) To study the endoscopic and histological regeneration pattern of NSE after
RFA of BE.

2) To compare the histological and immunohistochemical characteristics of NSE
to normal untreated squamous epithelium.

3) To compare NSE, untreated squamous epithelium, and submucosal gland duct
epithelium by laser capture microdissection (LCM) for determination of loss of
heterozygosity and single nucleotide polymorphisms, in order to establish more
insight in the lineage of the NSE.

For this prospective study, we will include 10 patients with BE that will be
treated with primary circumferential RFA according to standard guidelines.

6-8 weeks prior to ablation, patients will undergo a work-up endoscopy to
thoroughly map the BE, as is the current standard in all patients undergoing
RFA treatment. During these work-up endoscopies, visible lesions are removed
with endoscopic resection (ER) for histological staging, and biopsies are
obtained from the remaining BE to evaluate the presence of residual neoplasia.
In the patients that will be included for this study, an additional
single-piece ER will be obtained from BE that is not suspicious for carcinoma
(based on endoscopic appearance and prior biopsies), and from squamous
epithelium.

In the majority of patients, the esophagus has healed and most of the BE has
been repopulated with squamous epithelium about 6 weeks after RFA. Patients are
then rescheduled for endoscopy to assess the treatment effect, and if
necessary, for additional RFA treatment of residual Barrett epithelium.
For this study, patients will be scheduled for two additional endoscopies
within 6 weeks after the primary RFA procedure. During these endoscopies a
single-piece ER will be performed to obtain a tissue sample of the ablated area
during the squamous regeneration phase. By scheduling patients at 1&3, 2&5,
3&6, 1.5&3.5, and 2.5&4.5 weeks after RFA, two tissue specimens will be
obtained at 10 sequential intervals during the squamous regeneration period.

With the exception of ER specimens obtained from visible lesions during work-up
endoscopy, all ER specimens from the pre-RFA BE, normal squamous epithelium,
and from the ablated area after RFA, will be divided in half. One half will be
fixated in formalin for histological evaluation of morphological changes and
IHC staining to evaluate expression of relevant proteins. The other half of the
ER specimen will be fixated in liquid nitrogen for molecular evaluation, using
LCM to selectively obtain cell structures from the submucosal glands.

Two months after complete conversion of the BE into NSE by RFA, patients are
scheduled for standard follow-up endoscopy. During follow-up endoscopy biopsies
are obtained from the NSE (usually around 16 biopsies) for histological
evaluation to confirm the eradication of IM. For this study, 8 additional
biopsies will be obtained from the untreated squamous epithelium of which half
will be fixated in formalin and half in liquid nitrogen.

During ER using the multiband mucosectomy technique, minor bleeding may occur
in 6% of cases, usually easily managed with endoscopic hemostatic techniques.