From June 2015 to June 2017, a total of 37 patients with lesions in the EGJ underwent
RYAO in our center. Clinicopathological features, surgical data, perioperative and
long-term complications and postoperative follow-up data of the patients who underwent
RYAO were collected and compared with those of 36 patients who underwent EGPP and
31 patients who underwent DT during the same period. Preoperative assessments included
gastroscopy and biopsy and, if necessary, endoscopic ultrasonography, contrast-enhanced
computed tomography (CT), and other routine examinations.
The inclusion criteria for patients were as follows: (1) the lesion was located at
the EGJ; (2) surgery was indicated for these lesions, which were not suitable for
endoscopic therapy after discussion with the endoscopist; (3) the lesions were precancerous
or borderline lesions confirmed by pathology or lesions of an unknown nature preoperatively;
(4) the resection range of proximal gastrectomy was ≤ 1/2 of the stomach volume, and
the surgical margin met the criteria of the guidelines; (5) there were no surgical
contraindications, and (6) the patients had no history of previous surgery.
The Ethics Committee of the Second Military Medical University had approved this study.
Patients and their family members were informed about the details of each specific
surgical procedure (RYAO or EGPP or DT) and were educated about the advantages and
disadvantages of each surgical approach before they made any decisions. The patients
and their family members provided consent and signed the informed consent form before
Surgical procedure for proximal gastrectomy with RYAO
If no ascites, adhesions and other abnormal lesions were noted in the abdominal cavity
and the regional lesion could be surgically resected, proximal gastrectomy was carried
out first, as described below.
The perigastric vessels were divided with preservation of the distal gastric artery
and its branches to the gastric wall, including the right gastroepiploic artery and
right gastric artery (Fig. 1A, B). The left gastroepiploic artery and left gastric
artery as well as other gastric vessels could be excised. The esophagus was transected
2-3 cm above the gastric cardia, and an anvil was inserted into the esophageal lumen.
A 75-mm linear cutter (NTLC75, Ethicon Endo-Surgery, OH, USA) was used to incise the gastric corpus and
close the gastric stump in the greater curvature (Fig. 1C, D). Another 75-mm linear
cutter was used to incise and close the gastric body in the lesser curvature; then,
the residual gastric tube was formed (Fig. 1E, F).
Next, digestive tract reconstruction was performed in the following order: esophagogastric
anastomosis, obstruction of the antrum and RNY anastomosis.
At the lowest point in the greater curvature of the stomach, a 3-cm incision was made
in the anterior wall of the gastric body. A circular stapler was inserted into the
residual stomach. The trocar of the circular stapler was advanced through the posterior
wall (approximately 3 cm distal to the stump) and attached to the anvil for side-end
anastomosis of the esophagus and stomach (Fig. 2A, B) to complete the esophagogastric
anastomosis. The pylorus was located, and the omentum from the vascular arch to the
greater and lesser curvatures of the gastric antrum were separated. At a site approximately
2-3 cm from both the pylorus and the previous gastric incision, a 60-mm stapler (XF-60,
XINNENGYUAN, Changzhou City, China) was used to close but not cut off the antrum (Fig.
2C, D). If rupture of the stomach wall did not occur and no bleeding from the wound
was noted, then the obstruction of the antrum was considered complete (Fig. 2E, F).
The jejunum was divided at a site 15 cm distal to the ligament of Treitz. Gastrojejunal
side-side anastomosis was completed with hand-sewn sutures between the efferent loop
incision and the previous incision in the residual stomach wall (Fig. 2G, H). Then,
the stump of the afferent loop was anastomosed to the efferent loop 40 cm distal to
the gastrojejunal anastomosis. Finally, the RNY anastomosis was completed.
After completion of RYAO, the novel reconstructed digestive tract included an esophageal-gastric-jejunal
tract and a partially effective gastric antrum-duodenal tract with an obstructed proximal
antrum (Fig. 3A, B).
Routine esophagogastric anastomosis with pyloroplasty
The pylorus was located, and a 2-cm incision was made in the anterior wall of the
gastric pylorus along the longitudinal axis of the stomach. A circular stapler was
inserted into the residual stomach. The trocar of the circular stapler was advanced
through the posterior wall and attached to the anvil in the esophagus to complete
the esophagogastric anastomosis. Then, the primary incision was hand-sutured vertical
to the longitudinal axis of the stomach.
Double tract reconstrucion with jejunal interposition
The anvil of the circular stapler was inserted into the esophageal stump. The jejunum
was divided 15 cm distal to the ligament of Treitz. An end-to-side jejunojejunostomy
was created by an anastomosis between the 30 cm of anal jejunum from the oral jejunal
stump and divided oral jejunum. An entry hole for the circular stapler was made halfway
(15 cm) along the anal jejunal stump, and the circular stapler was used to achieve
an end-to-side esophagojejunostomy. After removing the circular stapler, the anastomosis
between the entry hole and the oral edge of the remnant stomach was made by hand sewing.
The length of the jejunogastrostomy was 6 cm.
After surgery, routine gastrointestinal decompression was used. If no complications,
including digestive tract leakage, hemorrhage or obstruction, were evident, then the
gastric decompression tube was removed after gastrointestinal motility was restored
and passage of flatus was reported. The patients were given an appropriate amount
of water. One day later, a liquid diet was initiated if no discomfort was reported.
Two days later, a semiliquid diet was initiated. If no discomfort was reported on
the second day after starting the semiliquid diet and no abnormalities were noted
in the drainage volume and color, then the abdominal drainage tube was removed, and
the patient was discharged from the hospital.
Between June 2015 and June 2017, 37 patients underwent RYAO, 36 patients underwent
EGPP, and 31 patients underwent DT. The clinicopathological features, the surgical
procedure notes, perioperative complications, long-term complications, weight changes,
health status and quality of life were documented and compared between the three reconstruction
All 104 patients who underwent proximal gastrectomy were reevaluated at a 6-month
postoperative follow-up examination, which included symptom assessment, physical examination,
routine blood and liver and kidney function tests, endoscopy, iodine radiography of
upper GI tract, abdominal ultrasonography and, gastrointestinal symptoms (EORTC QLQ-STO22
scores) [20, 21] and quality of life (EORTC QLQ-Core 30 scores) [22, 23].
Bile reflux gastritis was diagnosed by upper GI endoscopy and bilirubin level assessment
of gastric aspirate. The diagnostic criteria was as follows: intragastric bile and
residual gastritis was found under endoscopy, bilirubin level was exceeded normal
serum range (> 1.3 mg/dl). Reflux esophagitis were diagnosed by gastroscopy and iodine
radiography of upper GI tract. The diagnostic criteria was as follows: endoscopically
visible breaks in the distal esophageal mucosa, retrograde flow of gastrointestinal
contents into the esophagus was detected by radiography.
All statistical analyses were performed with IBM SPSS Statistics for Windows, version
22.0 (Armonk, NY, USA). The Oneway ANOVA test was used to compare differences in normal
distribution continuous variables between the groups, the Kruskal-Wallis test was
used to compare differences in ordinal categorical and non-normal distribution continuous
variables between the groups, and the Pearson Chi-Square test or Fisher’s exact test
was used to compare differences in unordered categorical variables between the groups.
LSD-t test or Bonferroni test was used for pairwise comparison between groups, P < 0.05 was considered statistically significant.