172
FC Zhi et al.
connective tissue, the density and strength of the
regenerated esophagus also increased.
The main cause of death within the perioperative
period of esophagus reconstruction was anastomotic
fistula, the incidence of which has been reported as
10–60% in previous reports.3,8,10 It was 6.7% in the
present study, which was much lower than in the
other reports and we think the low incidence is
related to the material used for the reconstruction
and the method of anastomosis. The artificial
biological esophagus has good tissue compatibility
and is pyknotic, elastic, and soft. Needle punctures
close well after suturing so leakage is less likely.
Number 0 or 1 nontraumatic suture and 5 × 17
radian intact round needles were used to perform the
sectional, whole layer and end to end anastomosis
with two-thirds introversion of the back wall and
one-third extroversion of the front wall. Other
factors, such as the blood supply of the esophagus,
the dog’s hygiene and manner of eating, and poor
nutritional support after operation might lead to
anastomotic fistula. In addition, the dog naturally
eats its food rapidly and frequently regurgitates,
which makes the lower anastomotic ostium vulner-
able to the formation of fistula. This conditions are
less likely in humans.
The dogs began to take food 7 days after operation
and the continuous stimulation from food intake as
well as bacterial action broke the sutures at both ends
of the anastomosis. Gastroscopic examination at
2 weeks after operation revealed that the sutures had
detached from the original esophagus and the artifi-
cial esophagus had shrunk because of the effects of
saliva, food and bacteria. Dysphagia occurred to a
certain degree. Separation proceeded from the ends
toward the center and finally the artificial esophagus
was completely separated from the ‘regenerated
esophagus’ and was dislodged into the stomach,
digested, vomited out of the body or manually
removed during gastroscopy. After detachment of the
artificial esophagus, the esophageal lumen became
patent and eating became easy. Therefore the trans-
planted artificial esophagus was only a short-term
replacement and support.
After detachment of the artificial esophagus, there was
an acute inflammatory reaction (edema, proliferation)
in the internal wall of the ‘regenerated esophagus’.
The squamous epithelial layer at both ends gradually
moved towards the center and during this period,
some dogs developed dysphagia to some degree
because of the inflammation. After approximately
3 months, squamous epithelium covered the whole
wall of the esophagus and there was stratified
squamous epithelium after 6 months. During this
period, the ‘regenerated esophagus’ was stable and
cicatricial proliferation ceased. The circular and longi-
tudinal contraction lessened and the ‘regenerated
esophagus’ gradually became soft and formalized.
The ‘regenerated esophagus’ showed an average of
30–40% longitudinal shrinkage, which was similar
to the report by Tian et al.6 Two experimental dogs
underwent X-ray examination, which showed
smooth passage of the barium, but no peristalsis. The
‘regenerated esophagus’ was composed of pyknotic
fibrous and connective tissue with squamous epithe-
lium covering the surface and many blood vessels
could be seen in the fibrous and connective tissues,
which indicated the squamous epithelium was
stable. Regeneration of muscles and glands has been
noted in previous reports,10,11 but we did not find
any evidence in the dogs that survived for more than
6 months. The fibrous and connective tissues were
well ordered in those dogs, so the remodeling
phenomenon might be related to the stress of food
intake.
Two dogs dislodged the artificial esophagus into the
digestive tract after the perioperative period and
another 2 vomited them out. In the remaining 23
dogs (81.5%), the artificial esophagi were removed by
a snare during gastroscopy. In most cases detachment
of the transplanted esophagi was not smooth, which
will be a major issue to overcome in the future.
Another important and difficult problem is the occur-
rence of stenosis after reconstruction, which directly
affects survival. Our experiment showed that stenosis
occurred after operation under two conditions: (i)
inflammation and granulation tissue hypertrophy of
the ‘regenerated esophagus’ before and after detach-
ment of the artificial esophagus;12 and (ii) during the
period of remodeling of the ‘regenerated esophagus’.
There are four ways to solve these problems: (i)
increase the time the artificial esophagus remains
within the body to reduce the reactive inflammation
before and after detachment; (ii) feed soft or semi-
liquid food to reduce the irritation caused by food
intake and prevent the overproliferation of granula-
tion tissue; (iii) dilatation treatment, if required for
severe stenosis (e.g. 2 months after operation) should
be performed gently, starting with the smallest caliber
probe and gradually increasing in size; and (iv) an
esophageal stent should be inserted if none of the
other measures are successful.
Chinese Journal of Digestive Diseases 4, 168–173