Modified triangulating stapling technique for esophagogastrostomy after esophagectomy for esophageal cancer

We typically performed thoracic esophagectomy and mediastinal lymph node dissection in the left decubitus position via a thoracoscopic approach. After the thoracic procedure was completed, the position of the patient was changed to supine. A gastric tube was made using linear stapler via a laparoscopic approach. If the tumor invaded the upper thoracic esophagus, a bilateral cervical lymph node dissection was performed after the abdominal procedure. We used a narrow gastric tube that was sufficiently long to transverse the posterior mediastinal route to the cervix to serve as an esophageal substitute. Esophagogastric anastomosis was performed on the left side of cervix. A modified triangular stapled anastomosis was made using three linear staplers. The first instrumental anastomosis was applied to the anterior wall of the remnant esophagus and the superior end of the gastric tube in an everted fashion. The first linear stapler was applied after three or four stay sutures through the whole layer were added to secure the first anastomosis, and then these stay sutures were pulled up and all stay sutures were then completely removed with the linear stapler (Fig. 1A, B). We applied the stay sutures at the both ends of the first anastomotic staple line (Fig. 2A). The anastomotic region was half-turned using one thread of these sutures (Fig. 2B). Furthermore, a stay suture was applied to the center of the posterior wall of the remnant esophagus and gastric tube in the whole layer, and one or two more additional stay sutures were applied between the end of the first staple line at the anterior wall and the center of posterior wall (Fig. 3A). Half of the posterior wall was anastomosed evertly with a linear stapler while pulling up the stay sutures. The most important feature of this surgical technique was that the suture lines had to be securely intersected. After the stay sutures were inserted to the end of the second staple line, we pulled up the stay sutures and performed anastomosis with a third linear stapler (Fig. 3B). Then an anastomosis was completely formed evertly (Fig. 4). After confirming hemostasis, the gastric tube was pulled down from an abdominal cavity, and torsion was spontaneously relieved when the gastric tube returned to the intrathoracic cavity.

The postoperative endoscopic views of the anastomotic site are shown in Fig. 5. The lumen of the triangular anastomosis (Fig. 5A) was wider than that of the circular stapled anastomosis (Fig. 5B). All sides of the triangle were everted, and no mucosal defect was evident.

Medical records of 48 patients with thoracic esophageal cancer treated at our hospital from January 2006 to December 2009 were reviewed retrospectively. All patients were pathologically diagnosed with esophageal squamous cell carcinoma, and radical esophagectomies were performed. All of the operations were performed by a single surgeon (M.T.).

The clinical characteristics of the patients are shown in Table 1. Patients comprised 36 men and 12 women with a mean (range) age of 59.4 (42–77) years at the time of esophagectomies. The tumors were located in the upper thoracic esophagus in 3 patients, middle thoracic esophagus in 30, and lower thoracic esophagus in 15. The pathological stage classification was according to the Japanese Classification of Esophageal Cancer of the Japanese Society for Esophageal Diseases [9] and 4 patients had stage 0, 5 had stage I, 7 had stage II, 25 had stage III, and 7 had stage IV disease.

The mean (range) operating time and estimated blood loss were 440 (262–784) minutes and 486 (120–1,310) milliliters, respectively. Postoperative barium swallowing was performed on the eighth postoperative day to evaluate the swallowing function and anastomotic leakage. If anastomotic leakage was diagnosed, endoscopic examination and chest computed tomography were immediately performed.

Frequencies of postoperative complications are shown in Table 2. Anastomotic leakage occurred in 4 patients (8.3 %). All patients with anastomotic leakage recovered quickly after draining from the cervical wound. Anastomotic stenosis was observed in 6 patients (12.5 %). The average number of endoscopic pneumatic dilatations in patients with anastomotic stenosis was 2.4. There was no hospital death in these patients. The median (range) length of hospital stay was 40.8 (19–154) days.