Management of intrathoracic leakage after radical total gastrectomy
Department of Vascular and Endocrine Surgery, First Affiliated Hospital, Fourth Military Medical University, Xi’an, PR China
|
Case Report
|
|
Abstract
Background: Intrathoracic anastomotic leakage resulted from radical total gastrectomy with an end-to-side esophagojejunostomy are exclusively abdominal.
Case presentation: We report the case of a 64-year-old male who underwent radical total gastrectomy and intraabdominal end-to-side esophagojejunostomy for gastric cardiac carcinoma. Anastomotic leakage to the thoracic cavity occurred which was confirmed by contrast radiography 18 days after the operation. The symptoms included coughing and fever, with elevated white blood cells over 10×109/L. Coughing and fever disappeared after successful sealing of the fistulous orifice with an endoscopically placed covered metallic stent with the applications of antibiotics and drainage of the pleural effusion. The patient was recovered and discharged from the hospital approximately two months after the occurrence of the leakage without any symptoms except intermittent esophageal reflux which could be resolved by treatment with cisapride, or by intaking less liquid food. The patient then received 4 cycles of adjuvant chemotherapy with regimen of FOLFOX4 (fluorouracil, leucovorin and oxaliplatin). Unfortunately, he died of a disease- or treatment- unrelated accidence 5 months after the discharge.
Conclusion: The thorough drainage combined with antibiotic treatment is able to eliminate empyema without the need for a specific thoracoscopy or thoracic surgery for patients with intrathoracic leakage.
Key words
Gastric cardiac carcinoma; intrathoracic leakage; management
J Thorac Dis 2010;2:180-184. DOI: 10.3978/j.issn.2072-1439.2010.02.03.11
|
|
Intrathoracic anastomotic leakage is often taken as a complication
of thoracic surgery (1,2). On the other hand, almost all reported
leakages resulted from radical total gastrectomy with an end-toside
esophagojejunostomy are abdominal. In 2008, there was
a patient who underwent gastrectomy due to gastric cardiac
carcinoma and developed an anastomotic leakage into thoracic
cavity in our department. Here, we present this case along with
the management experience.
|
|
Case report
A 64-year-old male patient was admitted to our department,
because of epigastr ic di scomfor t for 4 years. Upper
gastroendoscopy revealed that the dentate line almost
disappeared, and an ulcer of 4 cm×3 cm was discovered under
the dentate line in the lesser gastric curvature. Histopathological examination under gastroscopy confirmed that the lesion was
poorly differentiated ulcerative adenocarcinoma (Fig 1). Distant
metastasis was excluded by computed tomography scan and
ultrasonography. The patient underwent radical total gastrectomy
with R0 lymphadenectomy by the Moynihan-type procedure
(intraabdominal end-to-side esophagojejunostomy with Braun
anastomosis). Histopathological examination of the dissected
tumor and surrounding tissues, including 4 cm of esophagus,
showed a negative margin, proved preoperative gastroscopic
diagnosis and that among 22 excised lymphatic nodes, 5 were
positive for metastasis. The tumor stage was T3N1M0.
The patient recovered 7 days after the surgery. However, he
started to cough with a small amount of white sputa, and an
elevated white blood cell count 10.96×109/L with neutrophils
being 82.4%. The body temperature was normal and there were
no any other symptoms. A chest radiograph revealed a broadened
mediastinum and bilateral reactive pleural effusion with a vague
heart image (Fig 2). The patient was treated with cefazolin 2
g, b.i.d, intravenous infusion, and the coughing was gradually
alleviated, but persisted, with the white blood cell count
increasing to 14.4×109/L and neutrophils to 91.7%. However,
on day 13, the body temperature increased to over 38°C the
first time after the surgery. The white blood cell count was even
higher (16.09×109/L with neutrophils being 91.7%) although
the repeat radiograph showed slightly narrowed mediastinum with little pleural effusion. Then, cefazolin was replaced with
ceftriaxone (Rocephin, 2 g, b.i.d, intravenous infusion) the next
day. The patient could have a rest on day 19, but slight coughing
and high temperature (about 38°C) and elevated white blood
cell count (> 10.0×109/L) remained.
In order to make a definite diagnosis of the cause for
coughing, fever and the elevated white blood cell count, a
contrast radiography was performed on day 18 after the surgery,
which showed a leakage through which the contrast medium was
leaked from the orifice of the anastomosis to the right thoracic
cavity, with apparent right pleural effusion (Fig 3A, Fig 3B). In
the meantime, pleural effusion cultures were carried out, and
Enterococcus faecium, which was susceptible to vancomycin, was
isolated.
On day 20, a pigtail catheter (1.5 mm in diameter) was
percutaneously placed into the right thoracic cavity guided by
ultrasonography for drainage of the pleural effusion (Fig 3C).
On day 22, ceftriaxone was discontinued when the culture
result was received. On day 23, the patient was treated with vancomycin (0.5 g, b.i.d, intravenous infusion) for 5 days, and
then with Tienam (Imipenem/cilastatin, 1 g, b.i.d, intravenous
infusion) for 3 days. However, on day 28, the body temperature
jumped to over 39°C, for the first time since the surgery, and
reached 39.2°C. Fluconazole, 0.2 g, once per day, intravenous
infusion, was administered empirically, and the temperature was
controlled under 39°C, still around 38°C. On day 30, a larger
chest tube (8 mm in diameter) was used to replace the small
one, which effectively drained the pleural effusion. Two days
later when the temperature and other conditions were stabilized,
a pre-scheduled therapeutic upper endoscopy was performed,
which showed a defect about 8mm×8mm in the right posterior
of the anastomosis. During the endoscopy, a covered metallic
stent (8 cm in length, 1.6-2.2 cm in diameter) was applied to seal
the fistulous orifice (Fig 4A), and a nutrition tube was placed
into the distal efferent loop to feed supplements. A follow up
contrast radiography performed 9 days after the therapeutic
endoscopy revealed complete sealing of the fistulae (Fig 5), and
the temperature fell to normal. The patient gradually recovered without any symptoms and the white blood cell count decreased
to 6.90×109/L. One month after the therapeutic endoscopy, the
patient resumed to a semi-liquid diet. One week later, the stent
was endoscopically removed with the fistulae being healed (Fig
4B), and the patient was discharged from the hospital without
leakage management-related complications. Initially, the patients
complained of intermittent episodes of reflux esophagitis, which
could be resolved by treatment with cisapride, or by intaking
less liquid food. The patient then received 4 cycles of adjuvant
chemotherapy with regimen of FOLFOX4 (fluorouracil,
leucovorin and oxaliplatin). Unfortunately, he died of a diseaseor
treatment- unrelated accidence 5 months after the discharge.
Fig 1 Histopathological diagnosis of a
gastric cardiac adenocarcinoma: An ulcer
of 4cm×3cm (A) under the dentate line,
poorly differentiated adenocarcinoma
(H&E staining×50) (B).
Fig 2 Comparison of the chest radiographs
before (A) and 7 days after the surgery
(B). In contrast to A, B reveals a corpulent
mediastinum, vague costophrenic
angles and heart shadow, which might be
resulted from the anastomotic leakage into
the mediastinum and subsequent bilateral
reactive pleural effusion.
Fig 3 A contrast radiograph (A) and a plain radiograph (B) on day 18 after the operation, and a plain radiograph (C) on day 20 after the
surgery. In the image A, the contrast medium (arrow A1) is traced from the orifice (arrow A2) into the right thoracic cavity. In image B ,
the right fissura interlobaris (arrow B1) is visible, and the fistulae fluid and contrast medium (arrow B2) accumulate in the right thoracic
cavity. In image C, a drainage tube (arrow C) is percutaneously placed into the right thoracic cavity directed by ultrasonography.
Fig 4 Gastroscopic procedures performed on days 32 (A) and 68 (B), respectively. Figure A shows that the efferent loop (A1) and afferent
loop (A2) are unobstructive, a covered metallic stent (A3) about 8 cm in length which has been placed into the efferent loop, seals the
fistula stoma (A4) to the thoracic cavity, together with a nutrition tube (A5) to the distal efferent loop, and figure B shows that the fistulae
has been healed (B1) with the stent being retrieved.
Fig 5 Contrast radiography performed
on day 18 after the surgery (A), and on
day 4 after insertion of the metallic stent
(B). The fistula stoma is present in the image
A (arrow A), but not seen in the same
position in image B. In image B, since the
horn shaped end of the stent (arrow B)
does not exactly match with the jejunum,
refluxed fluid enters the gap between the
stent and jejunal wall. So, there was still a
small amount of fluid observed to flow into
the thoracic cavity when contrast radiography
is dynamically monitored, although the
fistula stoma is not seen in the static image.
Fig 6 An illustration showing reflux of intestinal fluid into the
thoracic cavity. The blue color indicates that the refluxed intestinal
fluid could leak from the gap between the stent (brown color) and the jejunal wall to the thoracic cavity through the orifice.
|
|
Discussion
The diagnosis of anastomotic leakage into thoracic cavity was
not primarily suspected and thus delayed because our initial
attention was paid to the abdominal, not thoracic cavity, and
the symptoms in the chest were not typical due to antibiotic
treatment. Then, we speculated that a leakage of the anastomosis
to the abdomen might have occurred. However, to our surprise,
the contrast radiography revealed a leakage into the thoracic
cavity. Having searched in six databases from their inception to
December 2009: Medline, Embase, PubMed, Cochrane Central
Register of Controlled Trials, Cochrane Database of Systematic
Reviews, and Database of Abstracts of Review of Effectiveness,
we could not find any previous report that describes an
intraabdominally performed esophagojejunostomy with a
subsequent leakage to the thoracic cavity.
Theoretically, we believe that this rare complication occurred
because of the fact that the tumor was located at a high site of
the stomach, and the operators had to drag down the esophagus
during the surgical procedures in order to get a negative incision
margin. After the surgery, the remnant esophagus was retracted
back into the thoracic cavity, bringing the anastomosis to the thoracic cavity as well. Subsequently, the leakage occurred due
to the increased tension of the anastomosis and the possibility
of ischemia. Therefore, surgeons must keep in mind that a
transabdominal incision has a risk for thoracic leakage in cases
who require radical incision and esophagojejunostomy, and a
transthoracic incision or combined thoracoabdominal incision
may be a safer and more appropriate option (3-6). Furthermore,
a study of 285 patients with esophagojejunal anastomosis
revealed that Orr-type Roux-en-Y reconstruction method was
superior to the Moynihan-type in the case of the digestive
reconstruction, since it needs shorter operation time, and more
importantly, it effectively prevents reflux and thus improves
postoperative quality of life (7-10).
Previous studies have suggested that in case of leakage,
reoperation can be considered when the anastomotic stoma is
large, but with a high mortality (2,11). In those patients who
are unwilling or too weak to receive reoperation, the covered
metallic stent is a method to save lives (1, 12-17). In our case, we
successfully sealed the fistulous orifice with a covered metallic
stent under upper endoscopy, and all leakage-related symptoms
were resolved. However, it should be noticed that the esophageal
reflux occurred in the patient after the surgery, but the symptoms
could be resolved with cisapride or less liquid diet. Indeed, it was
also observed that there was a small amount of fluid entering into
the thoracic cavity when contrast radiography was dynamically
monitored. These findings indicate that the lower end of the
stent we used in this case might not be able to prevent the reflux
of the intestinal fluid secreted from the intestines (Fig 5, Fig
6). Thus, drugs such as cisapride or avoidance of liquid diet are
needed if the type of stent we used in our case is to be applied.
Alternatively, a more “intellective” stent that can prevent the
reflux needs to be invented to achieve a better performance.
Moynihan-type procedure is not an optimal option in patients
with gastric cardiac carcinoma because of the esophageal reflux.
A transthoracic or combined thoracoabdominal incision, but not an intraabdominal incision, i.e. Pinotti’s operation, is advisable
when the tumor is at a high site of the stomach; thorough
drainage rather than thoracoscopically therapy or thoracotomy
is enough to eliminate empyema in combination with antibiotics
and effective sealing of the fistulae, and a covered retrievable
metallic stent is an appropriate way to seal the fistulae, although
further modifications of the stent are needed to make it more
applicable.
In conclusion, the thorough drainage combined with
antibiotic treatment is able to eliminate empyema without the
need for a specific thoracoscopy or thoracic surgery for patients
with intrathoracic leakage.
|
|
References
Cite this article as: Lv YG, Yuan SF, Yun J, Yao Q, Chen JH, Yi J, Ling R, Wang L. Management of intrathoracic leakage after radical total gastrectomy. J Thorac Dis 2010;2(3):180-184. doi: 10.3978/j.issn.2072-1439.2010.02.03.11
|
