Worldwide, esophageal cancer is the 6th most common cause of cancer-related death. Currently curative resection remains the cornerstone of the therapy. Despite advances in anesthesia, operative techniques and postoperative management, postoperative pulmonary complications (PPCs) occur frequently accounting for about 30% of all postoperative complications. Most importantly, PPCs have much been associated with postoperative mortality. The diaphragm is the most important respiratory muscle and its respiratory function would be inevitably damaged when esophagectomy is performed through the left posterolateral thoracotomy (Sweet procedure) because the diaphragm must be dissected for the purpose of stomach moralization. Meanwhile, Ivor-Lewis approach may effectively avoid diaphragm injury because the stomach can be managed through a laparotomy whereas an additional abdomen incision is needed. Both procedures are routinely used in practice when surgically managing esophageal cancer. The investigators hypothesize that Ivor-Lewis procedure might be superior to the left-thoracotomy route during esophagectomy in preventing PPCs.

Background:

Esophageal cancer is a common malignant tumor in China and worldwide with poor prognosis. Although the surgical resection combined with adjuvant therapy constitutes the mainstay of the treatment the esophagectomy-associated complications are still life-threatening. Postoperative pulmonary complications (PPCs), including pulmonary infection, atelectasis, edema and ARDS (acute respiratory distress syndrome) are the leading causes of the eventful postoperative episodes.

The diaphragm is the well known essential respiratory muscle whose normal function guarantees more than 30% of pulmonary function. Esophagectomy through left thoracotomy would inevitably cut off the left diaphragm for the abdominal procedure and thus greatly damage the diaphragmatic function. It has been assumed that Ivor-Lewis esophagectomy would save the pulmonary function to a great extent by retaining the intact diaphragm but the addition of a laparotomy would be unavoidable. However, this hypothesis has not been confirmed by any randomized controlled trial with adequate operation volume. At present, a large part of thoracic surgeons, for example in China, still prefer esophagectomy through Sweet approach believing that it is less traumatic. So far, there is presently no convincing evidence to guide the esophagectomy approach selection on the aspect of prevention of PPCs.

Objectives:

1. To compare the PPCs rates after Ivor-Lewis esophagectomy and Sweet esophagectomy
2. To compare the degrees of diaphragm function deficiencies between two groups.
3. To compare the pulmonary function results between two groups
4. To compare the degrees of gastric emptying between two groups

Diagnostic Criteria for PPCs: any one of the followings and not because of anastomotic leak, massive pleural effusion and non-pulmonary origin infection:

1. postoperative hypoxemia (PaO2≤ 60mmHg and/or PaCO2≥ 50mmHg) or dyspnea
2. transfer to intensive care unit due to hypoxemia or dyspnea
3. purulent pulmonary secretions requiring bronchoscopic suction
4. dependence on oxygen mask or nasal tube for more than 5 days
5. infiltrating or atelectasis in one pulmonary lobe or multiple segments identified by chest X-ray film or CT scan)
6. postoperative hiatal hernia associated pulmonary dysfunction
7. dependence on ventilator support for more than 48h

Degree of diaphragm function: The distance of diaphragmatic movement in centimeters

Pulmonary function measurement: FEV1, FVC and MMV

Gastric emptying function: Radioactivity remained at 15 minutes after ingestion of 100ml Tm99-labelled radioactive milk

Design:

Prospective randomized controlled trial.

Setting:

Thoracic Surgery Department, West China Hospital of Sichuan University, Chengdu, China

Patients and methods:

All patients with histologically confirmed esophageal carcinoma located at the middle or lower thoracic esophagus presenting to our hospital will be considered for the study

Termination criteria:

1. Intolerance of any of the treatment modality,
2. Development of PPCs
3. Development of any other severe complication, such as progressive bleeding, large pleural effusion, empyema
4. Death

Staging investigations will be standard and will include

1. Computed Tomography (CT) of Chest in all patients
2. Fibroesophagoscopy in all patients
3. Esophagogram by barium swallow in all patients
4. Endoscopic Ultrasonography (EUS) wherever possible Further investigation likes bronchoscopy, PET scan, or MRI will be applied if indicated.

Randomization:

Complete randomization will be done using a computer generated sheet. Randomization will be performed after confirming resectability of the primary tumor and enlarged lymph nodes.

All surgeries will be performed under general anesthesia with epidural analgesia. The surgery will be performed either by Dr. Longqi Chen or Dr. Yidan Lin who is the faculty and professor of the thoracic surgery department of West China Hospital. Thoracic esophageal mobilization and mediastinal lymphadenectomy will be done by open thoracotomy. Stomach dissection and gastric tube making will be done either through the left diaphragm incision when left thoracotomy is applied or through the laparotomy when Ivor-Lewis approach is applied. Operative time, blood loss, blood product replacement and all intraoperative details will be recorded in the proforma. Patients will be shifted postoperatively to the intensive care unit (ICU) for observation and subsequently to the recovery or high dependency ward once stabilized. Postoperative details including period of postoperative ventilation, hemorrhage, pulmonary and cardiac complications, arrhythmias, thoracic duct leak, anastomotic leak, wound infection and recurrent laryngeal nerve paresis or palsy will be recorded. Postoperative mortality will be defined as 30-day mortality plus death before discharge after surgery. Patients will be started on oral feeds between the 8th and 10th day in the absence of an anastomotic leak. The total duration of ICU stay and hospital stay will also be recorded.

Sample size:

The sample size considering an improvement of morbidity of sever pulmonary complications from 15% to 5% will be 280 patients (140 in each arm). The trial will aim at enrolling 320 patients in order to adjust for protocol violations. Interim analyses will be done when ½ (160 events) of the estimated events occur. We expect to complete accrual of patients in three years.

Follow up:

No follow up after discharge. The long-term pulmonary function is not the concern of this clinical trail.

Data management:

All collected data will be entered into a statistical software package for subsequent analysis

Inclusion criteria:

1. Biopsy or cytology confirmed esophageal carcinoma.
2. No-contraindication to surgery.
3. Tumor locates at middle or lower thoracic esophagus (e.g. below the level of azygos vein according to AJCC cancer staging manual, esophageal and esophagogastric junctional cancer, 7th edition, 2009), so that the anastomosis could be completed within the thorax
4. Informed consent.

Exclusion criteria:

1. Patients with low performance status (ECOG score > 1)
2. Refuse to participate
3. History of adjuvant chemo and/or radiotherapy
4. History of malignancy
5. Previous abdominal and/or thoracic surgery
6. History of pleural disease with obvious pleural adhesion on X-ray examination
7. Contraindication to any of the planned intervention procedure
8. Pregnancy
9. Age of ≥ 70 years