Esophagectomy for esophageal cancer is a technically complex procedure which is associated with high perioperative mortality, even in high volume centers[1]. To facilitate the postoperative recovery of esophagectomies patients by reducing surgical trauma, an increasing number of surgeons have attempted minimally invasive esophagectomy (MIE) to treat patients with esophageal cancer.[2-10] However, there is no consensus regarding the optimal method for performing an esophagectomy with the minimally invasive surgical technique. In addition, the benefit of this approach has not been well confirmed based on the limited retrospective comparative studies at the present time [3, 11-12], although its potential benefit improving the immediate postoperative including the total morbidity and pulmonary complication has been demonstrated by meta-analyses[13]. Especially it is unclear whether adding laparoscopic procedures in MIE can contribute to further improvement of the perioperative outcome of the patients.[3] Previously, the investigators have found that adding of laparoscopic procedure in performing the esophageal reconstruction procedure after VATS esophagectomy can provide further benefit in reducing the postoperative major complications and fasten the postoperative recovery16. For the most cases, the patients was receiving tri-incision esophagectomy, i.e. VATS esophagectomy in the chest, laparoscopic gastric mobilization in the abdomen and left cervical esophagogastrostomy. In such circumstances, a cervical incision was required for esophagogastrostomy after esophagectomy and gastric mobilization. However, for the patients with lower-to mid third esophageal cancer, some surgeon performed Ivor Lewis esophagectomy, which performing the esophagogastrostomy in the chest after gastric mobilization without cervical incision wound. Although both of these procedures have been demonstrated to be feasible and safe, there is much debate about the advantage and disadvantage of these two approaches. For tri-incision esophagectomy, patients have the chance to have cervical lymph node dissection and the esophagus can be resected up to the neck. However, it is more time-consuming and associated with more surgical trauma by adding a cervical incisional wound and more tissue dissection around the cervical trachea as compared to that done by Ivor Lewis esophagectomy. In contrast, for the Ivor Lewis esophagectomy, the resection of esophagus was limited to the level of thoracic inlet and cervical lymph node dissection was impossible unless a neck incision was further created. However, it takes less time in performing the whole procedure by saving a neck incision.

Esophagectomy for esophageal cancer is a technically complex procedure which is associated with high perioperative mortality, even in high volume centers[1]. To facilitate the postoperative recovery of esophagectomies patients by reducing surgical trauma, an increasing number of surgeons have attempted minimally invasive esophagectomy (MIE) to treat patients with esophageal cancer.[2-10] However, there is no consensus regarding the optimal method for performing an esophagectomy with the minimally invasive surgical technique. In addition, the benefit of this approach has not been well confirmed based on the limited retrospective comparative studies at the present time [3, 11-12], although its potential benefit improving the immediate postoperative including the total morbidity and pulmonary complication has been demonstrated by meta-analyses[13]. Especially it is unclear whether adding laparoscopic procedures in MIE can contribute to further improvement of the perioperative outcome of the patients.[3] Previously, we have found that adding of laparoscopic procedure in performing the esophageal reconstruction procedure after VATS esophagectomy can provide further benefit in reducing the postoperative major complications and fasten the postoperative recovery16. For the most cases, the patients was receiving tri-incision esophagectomy, i.e. VATS esophagectomy in the chest, laparoscopic gastric mobilization in the abdomen and left cervical esophagogastrostomy. In such circumstances, a cervical incision was required for esophagogastrostomy after esophagectomy and gastric mobilization. However, for the patients with lower-to mid third esophageal cancer, some surgeon performed Ivor Lewis esophagectomy, which performing the esophagogastrostomy in the chest after gastric mobilization without cervical incision wound. Although both of these procedures have been demonstrated to be feasible and safe, there is much debate about the advantage and disadvantage of these two approaches. For tri-incision esophagectomy, patients have the chance to have cervical lymph node dissection and the esophagus can be resected up to the neck. However, it is more time-consuming and associated with more surgical trauma by adding a cervical incisional wound and more tissue dissection around the cervical trachea as compared to that done by Ivor Lewis esophagectomy. In contrast, for the Ivor Lewis esophagectomy, the resection of esophagus was limited to the level of thoracic inlet and cervical lymph node dissection was impossible unless a neck incision was further created. However, it takes less time in performing the whole procedure by saving a neck incision.

In this study, we would conduct a prospective randomized study to compare the surgical results between the Tri-incision and Ivor Lewis approaches for esophagectomy in treating esophageal cancer. Both of the procedures will be performed by video-assisted thoracic surgical (VATS) esophagectomy and laparoscopic gastric mobilization and esophageal reconstruction. The primary end-point will be the overall survival and secondary end-point would be perioperative complication, postoperative recovery and quality of life.

Patients and Methods This study will include patients with a diagnosis of esophageal cancer who will undergo curative surgical resection in the surgical department of the National Taiwan University Hospital. All the patients will receive staging study for the tumor including computed tomography (CT) of the brain, neck, chest and abdomen, panendoscopy with endoscopic ultrasound (EUS), position emission tomography with computed tomography (optionally) and bronchoscopy for the tumor locating at the mid-to-low third of thoracic esophagus. The patient was cared according to the perioperative routine protocols of the thoracic surgical department of the National Taiwan University Hospital, including preoperative respiratory and exercise training, nutritional support and postoperative bronchoscopic toileting, and chest physical therapy. Jejunostomy feeding was started in the postoperative day 2 or 3, the oral intake was started 10 days to 2 weeks after surgery once no anastomotic leakage is demonstrated by the contrast swallowing image studies.

Exclusion criteria:

Poor lung function with FEV1 less than 70% of prediction. Major systemic co-morbidity : e.g. CVA, end-stage renal disease, coronary artery disease, congestive heart failure.

Presence of tracheal invasion or distant metastasis of the tumor

Primary end point:

Overall survival duration

Secondary end pint:

Disease-free survival duration Surgical complication Postoperative ICU stay and hospital stay Quality of Life Change of lung function after surgery

Anesthesia and perioperative care Epidural analgesia once agreed by the patients was administered before surgery. The patients were intubated and ventilated with double-lumen endotracheal tube during surgery. After surgery, extubation was given once satisfactory general condition including oxygenation, spontaneous breathing and vital signs are observed. However, temporary ventilator support will be given to the patients with high surgical risk in the intensive care unit (ICU) and was weaned off based on weaning parameters and the general condition of the patients. Jejunostomy feeding was begun after stool or flatus passage was detected. Oral intake began after an esophagogram examination revealed no anastomotic leakage, and was usually 10 to 14 days after surgery.

Postoperative clinical follow-up The patients will be followed up in the out-patient clinics after discharge at least tri monthly. Pan-endoscopy, computer tomography of brain, neck, chest and abdomen will be done every three months immediately 2 years after surgery and every six months thereafter. Life quality and lung function will be evaluated one, three and six months after surgery.

Power calculation:

With the difference of 10% in postoperative complication between the two groups of study, 50 patients will be required to recruited in each study group. The status of surgical complications, disease progression or recurrence and survival will be evaluated each year. Any significant difference once detected under analysis will call to early termination of the study to protect the patients from injury by inadequate treatment.

Inclusion Criteria:

• Patients with a diagnosis of esophageal cancer
• Age: below 75 years old.
• Tumor location: 2 cm above GEJ and 5 cm below thoracic inlet.
• Tumor stage: less than TNM stage III

Exclusion Criteria:

• Poor lung function with FEV1 less than 70% of prediction.
• Major systemic co-morbidity : e.g. CVA, end-stage renal disease, coronary artery disease, congestive heart failure.
• Presence of tracheal invasion or distant metastasis of the tumor