144
esophagus. It is difficult to investigate the developmental progres-
sion and natural history of Barrett’s esophagus–esophageal adeno-
carcinoma in humans due to the low incidence and variable risk of
progression. A large retrospective study recently revealed that the
annual risk of developing esophageal adenocarcinoma was 0.4% in
42,207 patients who had Barrett’s esophagus [ 6 ]. Due to the lim-
ited opportunity to study the disease progression in humans, ani-
mal models are great potential tools that may enhance understanding
of the underlying molecular pathophysiology.
Different animal models of GERD, Barrett’s esophagus, and
esophageal adenocarcinoma have been reported, using animals
such as rats [ 7 – 9 ], mice [ 10 , 11 ], dogs [ 12 ], and baboons [ 13 ].
Rat and mouse models have the advantages of small size, relatively
low cost maintenance, and widespread laboratory use and avail-
ability. In addition, they have a stratified squamous epithelium-
lined esophagus, similar to the human esophagus. However, they
have different histological features, namely, keratinized epithelium
with no submucosal glands. Rodents do not have spontaneous
reflux; therefore, reflux must be induced in these animals (various
surgical procedures for reflux induction have been described).
Mouse models are perhaps slightly more preferred because their
entire genome has been mapped and genetically modified lines are
easily available. Trans-genetic mouse models with specific knock-
out genes allow for improved understanding of disease pathophysi-
ology at a molecular level.
Surgical reflux models can be categorized into four subtypes:
(1) only gastric secretion reflux (GER), (2) only duodenal secre-
tion reflux (DER), (3) duodenogastric reflux with “bile predomi-
nance” (DgER), and (4) duodenogastric reflux with “acid
predominance” (dGER) [ 14 ]. The GER model is well established,
especially in rats [ 7 , 9 ]. DER and DgER models undergo proce-
dures such as esophagoduodenal anastomosis with gastrectomy
and esophagogastrojejunostomy to simulate Barrett’s esophagus
and esophageal adenocarcinoma models. This is because previous
studies have shown that contained duodenal and gastric juice
refluxate is more harmful to the esophageal mucosa than gastric
juice alone [ 15 ]. It is also more likely to induce Barrett’s esopha-
gus and esophageal adenocarcinoma [ 16 , 17 ]. Until recently, the
primary animal model used to study Barrett’s esophagus and
esophageal adenocarcinoma has been a rat model. Although mouse
reflux models have relatively lower incidence of Barrett’s esopha-
gus when compared with rat reflux models (even with the addition
of exogenous carcinogens), mouse reflux models have the major
advantage of available genetic modifications.
In 1999, Fein et al. [ 10 ] reported the first genetic model with
surgical facilitation, describing their experience performing sur-
gery (esophagojejunostomy with gastrectomy) on p53 knockout
mice with DER. Although this model showed a high rate of devel-Takahiro Masuda and Sumeet K. Mittal