Surgical management of resectable esophageal and esophagogastric junction cancers

INTRODUCTION — Esophageal cancer presents as localized disease, defined as adenocarcinoma or squamous cell carcinoma confined to the esophagus, in approximately 22 percent of all cases [1]. Regional disease, which includes spread to regional lymph nodes, accounts for another 30 percent of patients with esophageal cancer. The goal of surgical management is curative, and a surgical resection is the traditional mainstay of multidisciplinary therapy for patients with localized disease [2-5]. The clinical spectrum of esophageal cancer has changed over the last few decades, with an increase in incidence of adenocarcinoma and a decrease of squamous cell carcinoma [6-11]. Surgical management in the United States is independent of histology.

The selection criteria and surgical management of patients with esophageal cancer are discussed in this topic. Neoadjuvant and adjuvant chemotherapy and radiation therapy, and management of patients with unresectable esophageal cancer, are reviewed separately. (See "Radiation therapy, chemoradiotherapy, neoadjuvant approaches, and postoperative adjuvant therapy for localized cancers of the esophagus" and "Management of locally advanced, unresectable and inoperable esophageal cancer".)

ANATOMY — The esophagus, which is approximately 25 to 30 cm in length, is located in the posterior mediastinum and extends from the level of the 7^th cervical vertebra to the 11^th thoracic vertebra (figure 1 and figure 2 and figure 3). It is divided into four anatomic areas including the cervical, thoracic, and lower thoracic/esophagogastric junction and the abdominal esophagus (figure 4 and figure 5 and figure 6). The esophagus is composed of the mucosa, submucosa, muscularis externa, and adventitia (image 1). There are three critical anatomic points of narrowing: the cricopharyngeus muscle, the bronchoaortic constriction, and the esophagogastric junction, which are also the most common sites of iatrogenic and mechanical perforation (figure 7 and figure 8) [12].

The arterial supply includes the inferior thyroid artery (cervical esophagus), bronchial arteries, and the aorta (thoracic esophagus) and branches of the left gastric artery and inferior phrenic artery (abdominal esophagus) (figure 9). The venous drainage is through the inferior thyroid vein (cervical esophagus); the azygous vein, the hemiazygous vein, or the bronchial veins (thoracic esophagus); and the coronary vein (abdominal esophagus) (figure 10).

The esophagus has a rich network of lymphatic channels in the submucosa that can facilitate the longitudinal spread of neoplastic cells along the esophageal wall. Lymphatic drainage is to cervical nodes, tracheobronchial and mediastinal nodes, and gastric and celiac nodes (figure 11). An important point is that the regional lymph nodes for all locations in the esophagus, including the cervical esophagus and esophagogastric junction, extend from the periesophageal cervical nodes to celiac nodes (figure 12) [13].

HISTOLOGY OF ESOPHAGEAL CANCER — The two most common histologic types of esophageal cancer are squamous cell carcinoma and adenocarcinoma. Squamous cell carcinoma arises from the cells that line the esophagus; adenocarcinoma arises from glandular cells that are most commonly located at the junction of the esophagus and stomach but can vary based on the length of Barrett's esophagus.

TNM STAGING — The tumor, node, metastasis (TNM) staging system of the American Joint Committee on Cancer (AJCC) and the Union for International Cancer Control (UICC) for esophageal cancer is used universally. The current AJCC TNM staging system (eighth edition, 2017) recognizes the biologic heterogeneity of the disease and provides separate stage groupings for adenocarcinoma and squamous cell carcinoma (table 1) [13]. (See "Clinical manifestations, diagnosis, and staging of esophageal cancer", section on 'TNM staging criteria'.)

Tumors involving the esophagogastric junction (EGJ) with the tumor epicenter no more than 2 cm into the proximal stomach are staged as esophageal cancers (table 2). In contrast, EGJ tumors with their epicenters located more than 2 cm into the proximal stomach are staged as stomach cancers, as are all cardia cancers not involving the EGJ, even if they are within 2 cm of the EGJ. Thus, regardless of histology and Siewert type, esophageal tumors arising in the cervical, thoracic, or abdominal esophagus, and those involving the EGJ that have an epicenter within 2 cm of the EGJ, share the same criteria for T stage, N stage, and M stage designation (table 1).

The TNM and Siewert (figure 13) classifications are reviewed elsewhere. (See "Clinical manifestations, diagnosis, and staging of esophageal cancer", section on 'TNM staging criteria' and "Multimodality approaches to potentially resectable esophagogastric junction and gastric cardia adenocarcinomas", section on 'Definition and classification'.)

PRETREATMENT EVALUATION — Precise pretreatment staging guides multidisciplinary management decisions, including a surgical resection [7]. Pretreatment evaluation, including endoscopic biopsy, computed tomography (CT) of the chest and abdomen, endoscopic ultrasound (EUS), positron emission tomography (PET), and diagnostic laparoscopy and thoracoscopy, is reviewed separately. (See "Clinical manifestations, diagnosis, and staging of esophageal cancer", section on 'Pretreatment staging evaluation' and "Multimodality approaches to potentially resectable esophagogastric junction and gastric cardia adenocarcinomas", section on 'Pretreatment staging evaluation'.)

SELECTION OF OPERATIVE CANDIDATES — A surgical resection is the standard treatment approach for patients with an early esophageal cancer, but its utility as a monotherapy has been challenged [14,15]. The role of neoadjuvant chemotherapy and radiation therapy as components of multimodality management is reviewed separately. (See "Radiation therapy, chemoradiotherapy, neoadjuvant approaches, and postoperative adjuvant therapy for localized cancers of the esophagus".)

Criteria for resection

Esophagectomy as first line of therapy — The indications for an esophagectomy as the initial therapeutic approach to the patient with an esophageal cancer include:

●Patients with clinical T1N0M0 lesions (table 1).

●Patients with clinical T2N0M0 lesions are candidates in many medical centers. However, the optimal approach to clinical T2N0 disease is debated, with differing guidelines from expert groups as to indications for initial resection versus upfront (neoadjuvant) therapy. (See "Radiation therapy, chemoradiotherapy, neoadjuvant approaches, and postoperative adjuvant therapy for localized cancers of the esophagus", section on 'Clinical T2N0 disease' and "Multimodality approaches to potentially resectable esophagogastric junction and gastric cardia adenocarcinomas", section on 'Patients with clinical T1/2, node-negative disease'.)

Esophagectomy following neoadjuvant chemotherapy or chemoradiotherapy — The indications for initial chemoradiotherapy rather than upfront esophagectomy include:

●Patients with thoracic esophageal or esophagogastric junction tumors and full-thickness (T3) involvement of the esophagus with/without nodal disease.

●Selected patients with T4a disease (table 1) with invasion of local structures (pericardium, pleura, and/or diaphragm only) that can be resected en bloc, and who are without evidence of metastatic disease to other organs (eg, liver, colon). (See 'Indicators of unresectability' below.)

Patients with esophagogastric junction adenocarcinomas might also be considered for upfront chemotherapy alone if they cannot tolerate trimodality therapy, or if there is a high clinical suspicion of occult metastatic disease. (See "Multimodality approaches to potentially resectable esophagogastric junction and gastric cardia adenocarcinomas", section on 'Perioperative chemotherapy'.)

For patients undergoing chemoradiotherapy for potentially resectable disease who have a clinical response, the necessity of subsequent surgery is debated. However, surgical resection is generally recommended, particularly for patients with adenocarcinoma, because of the higher local control rates and lesser need for palliative procedures when surgery is a component of therapy. If there was a complete pathologic response to chemoradiation, then it might be reasonable to avoid surgery, but since this only occurs in 20 to 25 percent of cases and it is not possible to reliably identify these patients either by endoscopic ultrasound (EUS) or repeat positron emission tomography (PET) scan, the recommendation is to proceed with resection if the patient is fit for surgery and has not progressed during chemoradiotherapy. (See "Radiation therapy, chemoradiotherapy, neoadjuvant approaches, and postoperative adjuvant therapy for localized cancers of the esophagus", section on 'Necessity for surgery'.)

Patients who remain resectable according to radiographic restaging evaluation generally undergo resection four to eight weeks following completion of chemotherapy or chemoradiotherapy [16,17]. We generally do not repeat endoscopic ultrasound for reevaluation of T and N staging prior to surgery unless specific clinical questions need to be addressed. (See "Clinical manifestations, diagnosis, and staging of esophageal cancer".)

The impact of neoadjuvant chemotherapy or chemoradiotherapy on perioperative morbidity and mortality was addressed in a meta-analysis of 23 randomized trials comparing neoadjuvant therapy versus surgery alone or neoadjuvant chemotherapy versus chemoradiotherapy [18]. Neither neoadjuvant chemotherapy nor chemoradiotherapy increased the risk of total postoperative mortality or morbidity. However, subgroup analysis suggested that patients undergoing neoadjuvant chemoradiotherapy for squamous cell cancer might be at an elevated risk for postoperative mortality relative to those treated by surgery alone (risk ratio 1.95, 95% CI 1.06-3.6).

Issues related to postchemoradiotherapy surgery in patients undergoing definitive chemoradiotherapy for locally advanced unresectable esophageal cancer are addressed elsewhere. (See "Management of locally advanced, unresectable and inoperable esophageal cancer", section on 'Role of postchemoradiotherapy surgery'.)

Relative contraindications — The relative contraindications to an esophagectomy include:

●Advanced age – Advanced age is associated with greater morbidity following esophagectomy. However, age alone should not determine operability, as selected older patients have similar outcomes to younger patients.

●Comorbid illness – Comorbid illnesses increase the risk of postoperative complications (eg, cardiorespiratory complications, anastomotic leakage, reoperation rates, wound infection) and death following esophagectomy. Obesity is not associated with any increase in overall morbidity following esophagectomy and should not be considered a contraindication for esophagectomy [19].

Indicators of unresectability — The presence of metastatic disease, such as peritoneal, lung, bone, adrenal, brain, or liver metastases, or extraregional lymph node spread (eg, paraaortic or mesenteric lymphadenopathy), precludes an attempt at resection. The regional lymph nodes for all locations in the esophagus, including the cervical esophagus and esophagogastric junction, extend from the periesophageal cervical nodes to celiac nodes. Celiac nodal metastases and mediastinal/supraclavicular nodes are scored as regional nodal disease in the tumor, node, metastasis (TNM) staging system, regardless of the primary tumor location (figure 12) [13]; it is the number of involved nodes rather than location that determines the N stage (table 1) [13]. (See "Clinical manifestations, diagnosis, and staging of esophageal cancer", section on 'Regional lymph nodes' and "Management of locally advanced, unresectable and inoperable esophageal cancer".)

PREHABILITATION — Intensive preoperative respiratory rehabilitation and nutritional support reduce postoperative pulmonary complications after an esophagectomy [20,21]. A retrospective cohort study of 100 patients undergoing an esophagectomy found that patients managed with preoperative respiratory rehabilitation (n = 63) for seven days had a lower rate of postoperative pulmonary complications compared with patients not receiving preoperative respiratory rehabilitation (6 versus 24 percent) [22]. (See "Pulmonary rehabilitation".)

OPERATIVE PROCEDURES

Cervical esophageal cancer resection — Carcinoma of the cervical esophagus presents a unique management situation, and most patients are treated primarily by chemotherapy and radiation therapy. (See "Radiation therapy, chemoradiotherapy, neoadjuvant approaches, and postoperative adjuvant therapy for localized cancers of the esophagus", section on 'Cervical esophagus tumors'.)

For patients who fail chemotherapy/radiation therapy or who opt for a surgical resection rather than neoadjuvant therapy, a surgical resection usually requires removal of portions of the pharynx, the larynx, the thyroid gland, and portions of the proximal esophagus. This one-stage, three-phase operation requires cervical, abdominal, and thoracic incisions and a permanent terminal tracheostomy. Alternatively, a transhiatal approach is reasonable. In addition, bilateral radical neck dissections are generally performed [23-33]. If there is sufficient margin on the upper esophageal sphincter (figure 6), we perform a tri-incisional esophagectomy with a hand-sewn anastomosis to preserve esophageal length. (See 'Tri-incisional esophagectomy' below.)

Restoration of gastrointestinal tract continuity is accomplished with a gastric pull-up and anastomosis to the pharynx. For tumors confined to the proximal portion of the cervical esophagus, with a sufficient distal resection margin, a free jejunal interposition graft or a deltopectoral or pectoralis major myocutaneous flap are alternative reconstructive options. Free jejunal grafts are advantageous because this approach avoids mediastinal dissection, although expertise in performing microvascular anastomosis is required. Graft necrosis, fistula formation, and later graft strictures are specific problems. When compared with gastric pull-up procedures, graft survival and leakage rates are similar. Complications of esophageal procedures are discussed separately. (See "Complications of esophageal resection", section on 'Conduit complications'.)

Thoracic cancer resection — Patients with either adenocarcinoma or squamous cell carcinoma involving the middle or lower third of the esophagus, with the exception of gastroesophageal junction cancers, generally require a total esophagectomy because of the risk of submucosal skip lesions [34-36]. In selected cases, such as a superficial or early invasive esophageal cancer arising distally in the setting of Barrett's esophagus, a more limited resection (eg, an anastomosis in the thorax or a less radical resection such as a transhiatal resection) can be performed. However, the optimal surgical approach in these clinical settings is unknown [37-39]. If a limited resection is performed, intraoperative histologic analysis of the surgical margins must show no evidence of Barrett's changes. (See "Management of superficial esophageal cancer".)

Esophagectomy is a technically difficult operation, and the complication rate is high due to the anatomic challenges of the procedure. The choice of surgical approach depends upon many factors, including:

●Tumor location, length, submucosal extension, and adherence to surrounding structures

●The type or extent of lymphadenectomy desired

●The conduit to be used to restore gastrointestinal continuity

●Postoperative bile reflux

●The preference of the surgeon

The transhiatal, Ivor-Lewis (transthoracic), and tri-incisional esophagectomy procedures are the most commonly performed esophagectomies in North America, while an esophagectomy and an extended (three-field) lymphadenectomy is commonly performed in Asia [40-46].

Although the gastric interposition is most commonly used as a conduit for reconstruction following esophagectomy, the jejunum or the colon can also be used as the conduit [47-51]. These conduits are resistant to the effects of gastric acid, and they have a shape similar to that of the native esophagus. However, their use requires two additional anastomoses, and in the case of the jejunal interposition, the fixed mesenteric length limits transposition to the proximal esophagus.

Transhiatal esophagectomy — A transhiatal esophagectomy (THE) can be performed to resect cervical, thoracic, and esophagogastric junction (EGJ) esophageal cancers; it is performed through an upper midline laparotomy incision and a left neck incision, typically without a thoracotomy [52,53]. The thoracic esophagus is bluntly dissected through the diaphragmatic hiatus superiorly and via the neck inferiorly. A cervical anastomosis is created most often with a gastric pull-up approach. Disadvantages include the inability to perform a full thoracic lymphadenectomy and lack of visualization of the midthoracic dissection (figure 11).

In the largest prospective database series of 2007 patients, the in-hospital mortality rate decreased in the 1998 to 2006 cohort (n = 944 patients) compared with the 1976 to 1998 cohort (1 versus 4 percent) [54]. In addition, the anastomotic leak rate was also lower in the 1998 to 2006 cohort (9 versus 14 percent). Other postoperative complications included atelectasis and pneumonia (2 percent) and intrathoracic hemorrhage, recurrent laryngeal nerve (RLN) paralysis, chylothorax, and tracheal laceration in <1 percent each. Similar results have been noted in other large series [52,55-59].

Ivor-Lewis transthoracic esophagectomy — The Ivor-Lewis transthoracic esophagectomy can be used to resect cancers in the lower third of the esophagus but is not the optimal approach for cancers located in the middle third because of the limited proximal margin that can be achieved. This procedure combines a laparotomy with a right thoracotomy and an intrathoracic esophagogastric anastomosis. This approach permits direct visualization of the thoracic esophagus and allows the surgeon to perform a full thoracic lymphadenectomy. We prefer a minimally invasive Ivor-Lewis approach to a thoracotomy.

Disadvantages of the transthoracic esophagectomy include a limitation to the length of proximal esophagus that can be resected to achieve a histologically negative resection margin, an intrathoracic location of the esophagogastric anastomosis, and a 3 to 20 percent risk of severe bile reflux [55,60]. A leak occurring at the intrathoracic anastomosis has been associated with morbidity and mortality rates as high as 64 percent [61-66]. With current techniques, mortality rates are substantially lower [67].

Many centers report favorable results using the conventional Ivor-Lewis esophagectomy with a right thoracic anastomosis [58,68-74]. Prospective comparisons [59,75-79], plus at least one meta-analysis [80], suggest similar long-term outcomes compared to THE [73]. In one of the largest series of 228 patients undergoing an Ivor-Lewis subtotal esophagectomy, the perioperative mortality rate was 4 percent, and major respiratory and cardiovascular and/or thromboembolic complications occurred in 17 and 7 percent, respectively [73]. Nine patients (4 percent) developed a mediastinal leak, which was anastomotic in five and due to either an ischemic gastric conduit or gastrotomy dehiscence in the remainder. Only one patient developed a chyle leak. (See "Complications of esophageal resection".)

Modified Ivor-Lewis transthoracic esophagectomy (left thoracoabdominal esophagogastrectomy) — A modification of the Ivor-Lewis transthoracic esophagectomy includes a left thoracoabdominal incision with a gastric pull-up and an esophagogastric anastomosis in the left chest [81]. This approach is most useful for tumors involving the gastroesophageal junction. Only one incision is required, but disadvantages include a high incidence of complications such as postoperative reflux and limitation of the proximal esophageal margin by the aortic arch.

Tri-incisional esophagectomy — The tri-incisional esophagectomy combines the transhiatal and transthoracic approaches into a transthoracic total esophagectomy with a thoracic lymphadenectomy and cervical esophagogastric anastomosis [82-86]. The three-incisional technique allows the surgeon to perform a complete two-field (mediastinal and upper abdominal) lymphadenectomy under direct vision and a cervical esophagogastric anastomosis. We prefer a thorascopic approach to the chest rather than a thoracotomy to minimize the risk of respiratory complications. (See "Complications of esophageal resection", section on 'Pulmonary'.)

The following surgical oncology principles are used to perform a tri-incisional esophagectomy:

●Thoracotomy – A right posterolateral thoracotomy or a thoracoscopy is performed first to assess resectability and exclude local invasion of contiguous structures. An en bloc resection is performed that includes the esophagus and mediastinal (figure 14) and upper abdominal lymph nodes (figure 15), including the right paratracheal, subcarinal, periesophageal, and celiac axis lymph nodes.

●Laparotomy – The abdomen is explored to exclude metastatic disease, and the stomach is mobilized in preparation for the construction of the gastric conduit.

●Neck incision – A left neck exposure is preferred for the esophagogastric anastomosis since this approach reduces the risk of injury to the RLN (figure 16) [55,68,69,82,87]. The left RLN recurs lower (around the aortic arch) than the right RLN, which recurs around the subclavian artery and is therefore more likely to be injured from a right neck approach.

The advantages of a neck anastomosis include easier management of a possible anastomotic leak, lower incidence of reflux, more extensive proximal resection margin, and location outside of radiation ports if administered preoperatively. (See "Complications of esophageal resection", section on 'Conduit complications'.)

In our reported series of 250 patients treated with the combined approach, there were no intraoperative deaths, and the perioperative mortality rate was 3.6 percent (9 of 250) [82]. The cause of death was pneumonia and progressive respiratory failure in three, aspiration and respiratory arrest in one, pulmonary embolism in two, sepsis in the setting of conduit leak and empyema in one, ischemic bowel and multisystem organ failure in one, and cirrhosis in the setting of ischemic necrosis of the conduit in one. Early postoperative complications occurred in 124 of 250 patients (50 percent); they were considered major in 83. An anastomotic or conduit leak (from either conduit necrosis or leak from the staple line) occurred in 14 and 5 patients (5.6 and 2 percent, respectively).

Esophagogastric junction cancer resection — Esophageal cancers at the EGJ or intra-abdominal esophagus have been traditionally managed surgically with either an esophagectomy with partial gastrectomy or an extended gastrectomy, with or without thoracotomy (figure 3). Regardless of the approach, complete (R0) resection, a 4 cm (distal) gastric margin, a 5 cm esophageal margin, and resection of at least 15 nodes in basins appropriate for the primary tumor location are necessary [88,89]. The extent of the esophageal resection that can be achieved solely via a transabdominal approach without thoracoabdominal incision or transhiatal esophagectomy is limited, and therefore this approach is not accepted for tumors that involve the distal esophagus, due to difficulties in achieving an adequate negative proximal margin.

The contemporary operative approach for EGJ cancer is based upon findings from two surgical phase III trials [90,91], which randomly assigned patients to the operative procedure on the basis of the Siewert classification of EGJ adenocarcinoma (figure 17) [92].

●A phase III Dutch trial randomly assigned 220 patients with Siewert type I or II EGJ adenocarcinoma to THE versus extended thoracic resection (transthoracic esophagectomy with an extended en bloc lymphadenectomy via the right thoracic approach [RTA]) [77,90]. In-hospital mortality did not differ between the groups, but pulmonary complications and postoperative chylous leakage were significantly more frequent after RTA; in addition, intensive care unit (ICU) and total hospital stays were also longer in this group [90].

Five-year overall survival rates were similar (36 versus 34 percent for RTA and THE, respectively), but there was a trend toward better survival with extended thoracic resection in the patients with a type I tumor (five-year survival 51 versus 37 percent, p = 0.33). Also, for patients with a limited number of positive lymph nodes (one to eight), the transthoracic approach had a better five-year locoregional disease-free survival. The authors concluded that given the greater hazards associated with extended transthoracic resection, it could only be recommended for patients with type I and not type II tumors.

●A Japanese trial (JCOG 9502) compared THE versus extended esophagectomy using a left thoracoabdominal approach (LTA) for patients with type II or III adenocarcinoma of the EGJ [91]. Patients in the THE group received a total gastrectomy plus a D2 lymphadenectomy (including splenectomy) and para-aortic lymph node dissection. This approach involved a laparotomy only and therefore only resection of the lower esophagus. Thoracotomy was utilized only if there was a positive proximal margin. (See "Surgical management of invasive gastric cancer", section on 'Extent of lymph node dissection'.)

The patients in the LTA group underwent thorough mediastinal nodal dissection below the left inferior pulmonary vein as well as the same lymphadenectomy procedure in the abdominal cavity as the THE group.

The trial closed prematurely when a planned interim analysis concluded that it was unlikely that LTA would be significantly better than THE. Five-year survival rate was lower in the LTA group (38 versus 52 percent), although the difference was not statistically significant. The difference was maintained with long-term follow-up (10 year survival 24 versus 37 percent), but it remained not statistically significant [93]. In addition, complications were observed more frequently after LTA, and in-hospital mortality was higher (4 versus 0 percent) [91]. The authors concluded that LTA could not be recommended for type II or III tumors.

Based upon the results of these, the surgical approach for resection is summarized as follows:

●Patients with Siewert type I tumors should preferentially undergo a transthoracic en bloc esophagectomy and partial gastrectomy with two-field lymphadenectomy. Lymphadenectomy is important to assess disease stage and evaluate response to neoadjuvant therapy with minimal added risk in morbidity or mortality rates. Other reasonable options include a transhiatal approach, which does not include a lymphadenectomy.

●Patients with Siewert type II tumors should undergo an Ivor-Lewis esophagectomy.

●Patients with Siewert type III tumors should undergo a total gastrectomy with a transhiatal resection of the distal esophagus with lymphadenectomy of the lower mediastinum and an extended lymph node dissection including nodes along the hepatic, left gastric, celiac, and splenic arteries as well as those in the splenic hilum. (See "Surgical management of invasive gastric cancer", section on 'Extent of lymph node dissection'.)

The operation should be extended into a transhiatal or transthoracic esophagectomy in the setting of a positive proximal resection margin at frozen section examination.

PRINCIPLES OF SURGICAL RESECTION

Minimally invasive approaches — An Ivor-Lewis esophagectomy can be performed through an open, totally minimally invasive (thoracoscopic/laparoscopic), or hybrid approach. The hybrid approach can either combine thoracoscopic resection of the thoracic esophagus and mediastinal nodes with open laparotomy or combine laparoscopic resection of the intra-abdominal esophagus, stomach, and lymph nodes with an open right thoracotomy. The transhiatal esophagectomy can be performed open or laparoscopically.

Early reports indicated that minimally invasive esophagectomy was safe [94-100], but benefit in terms of long-term outcomes compared with open esophagectomy were uncertain [101]. More studies and two trials associated totally minimally invasive and hybrid Ivor-Lewis esophagectomy with fewer postoperative pulmonary complications and similar oncologic outcomes compared with open surgery, at least in the intermediate term (up to three-year follow-up). A 2019 systematic review and meta-analysis of 55 studies, including over 14,000 patients, concluded that minimally invasive esophagectomy was associated with an 18 percent lower five-year all-cause mortality compared with open esophagectomy (hazard ratio 0.82, 95% CI 0.76-0.88) [102]. Additional data are available from an analysis of over 5500 patients from the National Cancer Database who underwent esophagectomy between 2010 and 2015, which found comparable survival between minimally invasive, robotic-assisted, and open approaches [103]. Both minimally invasive and robotic approaches produced a higher median lymph node yield than the open approach.

Totally minimally invasive Ivor-Lewis esophagectomy — A totally minimally invasive approach to Ivor-Lewis esophagectomy is technically feasible, but data for oncologic outcomes are limited due to short follow-up and the small number of patients evaluated [97,98,104-110].

The best evidence comes from the TIME trial, which found that patients undergoing a totally minimally invasive esophagectomy have a better perioperative hospital course. In this multicenter, randomized trial of 115 patients with esophageal cancer, patients undergoing a minimally invasive esophagectomy had a lower rate of pulmonary infections in-hospital compared with patients undergoing an open esophagectomy (12 versus 34 percent, relative risk [RR] 0.35, 95% CI 0.16-0.78) [111]. Patients undergoing minimally invasive esophagectomy also had a lower perioperative (first two weeks after surgery) rate of pulmonary infections compared with patients undergoing an open esophagectomy (9 versus 29 percent, RR 0.30, 95% CI 0.12-0.76). A follow-up study showed similar three-year disease-free (36 versus 40 percent) and overall (40 versus 51 percent) survival for patients who underwent open surgery versus minimally invasive esophagectomy [112].  

The robot-assisted minimally invasive thoracolaparoscopic esophagectomy (RAMIE) technique was evaluated in another trial of 112 patients with esophageal cancer [113]. Compared with open transthoracic esophagectomy, RAMIE resulted in fewer complications, faster functional recovery, and better short-term quality of life. Oncologic outcomes were comparable at a median follow-up of 40 months.

The oncologic outcomes of totally minimally invasive esophagectomy were also addressed in three other nonrandomized series:

●A propensity score-matched analysis of 184 patients also found comparable early outcomes between robot-assisted versus thoracoscopic-assisted Ivor-Lewis esophagectomy for the treatment of middle and distal esophageal cancer [114].

●In a British series of 65 patients who underwent totally minimally invasive esophagectomy for invasive esophageal cancer, the two-year overall and disease-free survival rates were 81 and 74 percent, respectively. Recurrence was documented in 14 patients, 11 of which were distant recurrences [97].

●A retrospective Australian series compared outcomes among 114 patients who had open esophagectomy, 309 patients who had thoracoscopic-assisted surgery, and 23 who had a totally minimally invasive esophagectomy [115]. There were no differences in the rate of margin positivity or the number of lymph nodes retrieved in any of the groups and no differences in the time to recurrence or median or three-year overall survival when patients were compared stage for stage.

Although data suggest potential benefit for a totally minimally invasive approach to Ivor-Lewis esophagectomy, they are not sufficient to conclude that this is a standard approach. Additional data, particularly regarding long-term complications and oncologic outcomes, are needed before it can be concluded that totally minimally invasive Ivor-Lewis esophagectomy should replace open esophagectomy. For now, the National Comprehensive Cancer Network (NCCN) guidelines regard minimally invasive esophagectomy as an acceptable approach for esophageal cancer. However, minimally invasive esophagectomy should only be attempted by qualified surgeons; these procedures come with a steep learning curve and high learning-associated morbidity [116,117]. There is evidence that the outcomes improve with experience [118].

The technique of totally minimally invasive Ivor-Lewis esophagectomy with intrathoracic anastomosis is discussed in another topic. (See "Minimally invasive esophagectomy".)

Hybrid Ivor-Lewis esophagectomy — A thoracoscopic approach to mobilization of the intrathoracic portion of the esophagus and node dissection is an accepted alternative to open thoracotomy in centers with surgeon expertise in these techniques and is combined with open laparotomy for completion of the mobilization and upper abdominal node dissection. This represents the most popular minimally invasive esophagectomy technique with the most extensive published experience [95,96,119-121], despite none of the studies being prospective or randomized. Relative contraindications to thoracoscopic surgery include inadequate pulmonary function, extensive pleural adhesions, prior pneumonectomy, bulky tumors, and locally infiltrative tumors, particularly those with airway involvement [94].

An alternative hybrid approach combines laparoscopic resection of the intra-abdominal esophagus, stomach, and lymph nodes with an open right thoracotomy. The purported advantage of this approach is reduced risk of intraoperative pulmonary complications. The benefit of this approach was shown in a randomized trial of 207 patients with cancer in the middle or lower third of the esophagus, in which a hybrid approach of laparoscopic gastric mobilization combined with open right thoracotomy resulted in fewer major intraoperative and postoperative complications (36 versus 64 percent), especially pulmonary complications (18 versus 30 percent), compared with open esophagectomy [122]. The overall (67 percent hybrid versus 55 percent open) and disease-free survival (57 versus 48 percent) at three years was not affected by the operative approach.

Laparoscopic transhiatal esophagectomy — The transhiatal esophagectomy can be performed open or laparoscopically. A 2016 Cochrane systemic review and meta-analysis of six observational studies (five retrospective) concluded that laparoscopic transhiatal esophagectomy was associated with fewer overall (risk ratio 0.64, 95% CI 0.48-0.86) and serious complications (risk ratio 0.49, 95% CI 0.24-0.99), as well as shorter hospital stays (by three days) compared with open transhiatal esophagectomy [123]. However, randomized prospective trials are needed to determine the optimal approach to transhiatal esophagectomy.  

Circumferential resection margin — The prognostic role of a positive circumferential resection margin (CRM) for patients with esophageal cancer was unclear, in part due to a lack of clear definition of a positive margin [124]. The College of American Pathologists (CAP) defines a positive CRM as the presence of esophageal cancer at the resection margin [125]. However, the United Kingdom Royal College of Pathologists (RCP) defines a positive CRM as the presence of esophageal cancer within 1 mm of the resection margin [126,127].

CAP criteria differentiate a higher-risk group of patients with resectable esophageal cancer than the RCP criteria. A meta-analysis of 14 cohort studies including 3566 patients with resectable esophageal cancer found the overall five-year mortality rates were higher for patients with a positive CRM compared with a negative CRM [124]:

●CAP criteria – 15.3 percent positive margins (173/1133), odds ratio (OR) 4.02, 95% CI 2.25-7.20, p <0.001

●RCP criteria – 36.5 percent positive margins (889/2433), OR 2.52, 95% CI 1.96-3.25, p <0.001

In addition, patients with CRM involvement between 0.1 and 1 mm of the resection margin had a significantly higher five-year mortality rate compared with patients with tumor more than 1 mm of the resection margin (OR 2.05, 95% CI 1.41-2.99, p <0.001).

Extent of lymphadenectomy — The appropriate extent of lymphadenectomy during esophageal cancer surgery is debated. The minimum number of lymph nodes that should be removed during potentially curative esophagectomy has not been established. However, as many lymph nodes should be removed as is feasible since more extensive lymphadenectomy has been associated with better survival [46,128-132]. For example, in a retrospective review of 972 patients with node-negative esophageal cancer, the five-year disease-specific survival was 55 percent when fewer than 11 nodes were resected, 66 percent for 11 to 17 nodes resected, and 75 percent for 18 or more nodes resected [128]. A greater number of retrieved lymph nodes generally reflects more accurate staging, which generally comes with more extensive resections.

Many high-volume surgical centers routinely perform en bloc esophagectomy with a two-field (mediastinal, upper abdomen) lymph node dissection in the belief that this contributes to better locoregional control because of removal of metastatic lymph nodes. An even more extensive lymphadenectomy, three-field lymphadenectomy of the mediastinal, abdominal, and cervical nodes, is commonly practiced in Asian countries for upper thoracic esophageal cancers [40-46,133]. In a retrospective review of 1361 patients with squamous cell carcinoma of the thoracic esophagus, the frequency of nodal metastasis was neck (9.8 percent), upper mediastinum (18.0 percent), middle mediastinum (18.9 percent), lower mediastinum (11.8 percent), and upper abdomen (28.4 percent) [133]. However, while this approach increases the accuracy of staging, whether local control or survival is improved compared with two-field lymphadenectomy is uncertain.

Proponents of extended lymphadenectomy emphasize the relationship between total lymph node count and prognosis [46,129-131] and quote long-term survival rates as evidence of its therapeutic benefit. As an example, in one American series of 80 patients undergoing this technique in conjunction with en bloc esophagectomy, the overall five-year survival rate was 51 percent (88 percent for node-negative and 33 percent for node-positive patients) [134]. However, since the extent of lymph node dissection can affect the assignment of the final stage of disease, this resulting stage migration phenomenon hampers a stage-by-stage comparison between different forms of surgical resection [135]. Furthermore, although unsuspected metastases in the recurrent laryngeal or cervical nodes were detected in 36 percent of patients in the above-noted series [134], others report a low incidence of cervical nodal recurrence following a two-field lymphadenectomy [136]. The location of the tumor (upper versus middle to lower third) may have an influence on the frequency of finding cervical nodal metastases [137].

At least two randomized trials have compared different extents of lymphadenectomy during esophageal cancer surgery, although neither provides a conclusive result as to the benefit of extended lymphadenectomy [90,138]. In the United States, en bloc resection of the mediastinal and upper abdominal lymph nodes is considered a standard component of transthoracic esophagectomy, and a three-field lymphadenectomy is not considered a standard treatment for patients with esophageal cancer. However, if a lymph node dissection is not done, then lymph node sampling should be carried out to accurately stage the patient and to gauge the response to induction treatment in patients enrolled in trials using neoadjuvant therapy.

Hand-sewn versus stapled anastomosis — Anastomotic closure techniques include hand-sewn (single versus double layer), stapled (circular versus side-to-side linear), and hybrid linear stapled technique, with surgeon experience likely being the most important determinant at present [139-148]. For example, in a meta-analysis of 12 randomized controlled trials that included 1407 patients, the rate of anastomotic leak was similar for a circular stapled anastomosis compared with a hand-sewn anastomosis (OR 1.02, 95% CI 0.66-1.59) [148]. However, a stapled anastomosis was associated with a significantly higher rate of anastomotic stricture (OR 1.67, 95% CI 1.16-2.42).

A hybrid linear stapled technique was developed in the 1990s that demonstrated a 65 percent increase in the anastomotic cross-sectional area [145] and reduced morbidity for patients undergoing a cervical esophagogastric anastomosis [147]. In a review of 274 patients, those undergoing a hybrid anastomosis (modified Collard technique, n = 86 patients) had fewer cervical wound infections compared with patients undergoing a hand-sewn anastomosis (8 versus 29 percent) [147]. Patients undergoing a hybrid anastomosis had a statistically similar leak rate but required fewer anastomotic dilatations (4 versus 11 percent, mean 2.4 versus 4.1 per patient, respectively).

Cervical versus thoracic anastomosis — When performed using a standardized technique, cervical and thoracic esophagogastric anastomoses are equally safe [141,149]. In a randomized trial of 83 patients undergoing an esophagogastric anastomosis, the additional 5 cm esophageal resection to permit a cervical esophagogastric anastomosis did not increase tumor removal and survival or adversely influence morbidity, anastomotic diameter, or body weight development compared with a thoracic esophagogastric anastomosis [149].

At present, the choice of anastomotic location remains clinician dependent. A cervical anastomosis has a higher leak rate and risk of injury to the recurrent laryngeal nerve [59,141,150]. However, the anatomic confines of the neck and thoracic inlet limit surrounding tissue contamination and thus limit morbidity. A systematic review of four clinical trials that included 267 patients found that patients undergoing a cervical esophagogastric anastomosis (n = 132) had a higher rate of anastomotic leak compared with those undergoing a thoracic esophagogastric anastomosis (18 versus 4 percent, OR 3.43, 95% CI 1.09-10.78) [141]. A cervical esophagogastric anastomosis has a significantly higher rate of recurrent laryngeal nerve injury (OR 7.14, 95% CI 1.75-29.14), but there was no difference in the rate of pulmonary complications, perioperative mortality, benign stricture formation, or tumor recurrence at the anastomotic site. (See "Complications of esophageal resection", section on 'Anastomotic leak'.)

Additionally, utilizing the Surveillance, Epidemiology, and End Results (SEER) database, a retrospective cohort review was performed comparing 225 transhiatal resections to 643 transthoracic resections [59]. The unadjusted anastomotic complication rate, identified by need for postoperative endoscopic dilation, was higher (43.1 versus 34.5 percent; p = 0.02) in the transhiatal group; however, there was no significant difference in risk-adjusted five-year survival between groups.

Orthotopic placement — Orthotopic placement of the neoesophagus in the posterior mediastinum is generally preferred by most thoracic surgeons. A meta-analysis of trials comparing the posterior mediastinal route and the retrosternal route was performed and was unable to demonstrate any difference in postoperative morbidity [151]. Other series, however, have revealed a higher anastomotic leak rate in the retrosternal route, likely due to increased length requirements for the conduit as well as compression [54,152,153]. Conduit choice may also affect anastomotic integrity, although the majority of surgeons presently prefer the stomach due to its ease of preparation, robust blood supply, and sufficient length, with alternative conduits, including colon and jejunum, only being chosen out of necessity.

Role of pyloroplasty or pyloromyotomy — The role of a pyloroplasty or pyloromyotomy to reduce the risk of gastric outlet obstruction following a gastric pull-up procedure has been challenged by prospective studies and randomized trials, including:

●A prospective study of 242 patients undergoing an esophagectomy with gastric conduit found that patients with a pyloromyotomy (n = 159) did not have significantly lower rates of gastric outlet obstruction compared with those without a pyloromyotomy (9.6 versus 18.2 percent, respectively) [154]. In addition, there was no significant difference for rates of pneumonia or mortality (27.7 versus 19.5 percent, and 2.4 versus 2.5 percent, respectively). Management with pyloric dilation was effective in relieving symptoms in approximately 97 percent of symptomatic patients.

●A meta-analysis of nine trials and 553 esophagectomy patients randomized to pyloromyotomy versus none found a lower risk of gastric outlet obstruction for patients treated with a pyloromyotomy (OR 0.18, 95% CI 0.03-0.97, p <0.046) [155]. There was no difference for operative mortality, esophagogastric anastomotic leaks, pulmonary morbidity, or fatal pulmonary aspiration.

Functional conduit disorders and their management are reviewed separately. (See "Complications of esophageal resection", section on 'Functional disorders'.)

Recurrent laryngeal nerve identification — Recurrent laryngeal nerve (RLN) injury can occur during dissection of the cervical and thoracic esophagus [55,68,69,87]. Rates of injury range from 2 to 17 percent and occur more commonly when a cervical approach is utilized [54,82,156,157]. Principles for the avoidance of RLN injury include precise knowledge of cervical esophageal anatomy (figure 18), maintaining the plane of dissection as close as possible to the esophagus, and avoidance of metal or rigid retractors along the tracheoesophageal groove.

Jejunal feeding tube placement — A feeding jejunostomy tube is inserted at the time of the surgical resection for all patients undergoing an esophagectomy and for select patients who require nutritional support during induction chemotherapy and/or radiation therapy. The jejunostomy tube is inserted 40 cm distal to the ligament of Treitz, using either a laparoscopic approach if technically feasible or through a small laparotomy incision. For patients who will likely resume oral intake early, it is reasonable to omit a jejunal feeding tube placement.

POSTOPERATIVE MANAGEMENT — Enteral feedings are started on postoperative day 2 and slowly advanced until feeding goals are attained at approximately postoperative day 5. A barium swallow is performed on postoperative day 7 to evaluate for leak and emptying of the conduit. The nasogastric tube generally remains in place until the barium swallow is performed and demonstrates no anastomotic leak. Patients are maintained on a minimal liquid diet for approximately two weeks to allow the conduit to remain decompressed and straight in the mediastinum. Early oral feeding may have some advantage without compromising outcome [158]. (See "Complications of esophageal resection", section on 'Conduit complications'.)

Measures to reduce the risk of postoperative cardiovascular death, nonfatal myocardial infarction, and nonfatal cardiac arrest have included perioperative beta blockade, although the results of the Perioperative Ischemic Evaluation (POISE) trial [159] have generated some measure of caution for this practice [160]. In a randomized trial that included 8351 patients with or at high risk of atherosclerotic disease and scheduled for a noncardiac operation, patients receiving extended-release metoprolol succinate had a higher mortality rate compared with patients receiving a placebo (3.1 versus 2.3 percent, hazard ratio [HR] 1.33, 95% CI 1.03-1.74) [159]. In addition, the risk of a cardiovascular accident (stroke) was also higher for the patients receiving metoprolol (1.0 versus 0.5 percent, HR 2.17, 95% CI 1.26-3.74). We prefer not to routinely use beta blockers for patients who have a minimally invasive esophagectomy, since their risk of atrial fibrillation is low.

The guidelines for the perioperative management of patients with cardiac disease, including those of the American College of Cardiology/American Heart Association (ACC/AHA), are reviewed elsewhere. (See "Management of cardiac risk for noncardiac surgery".)

The American College of Chest Physicians Guidelines on the Prevention of Venous Thromboembolism [161,162] classify esophagectomy as a high-risk procedure and thus recommend postoperative thromboprophylaxis with low-molecular-weight heparin, unfractionated subcutaneous heparin three times daily, or fondaparinux. Conversely, some would classify esophagectomy as high risk for bleeding, especially in the setting of blunt mediastinal dissection, and thus argue for less aggressive prophylaxis. Further complicating the matter is the frequent use of neuraxial anesthesia, which further limits the use of perioperative anticoagulants for thromboprophylaxis [163]. Unfortunately, a paucity of data exists to help clarify these issues, and, therefore, clinical practice varies. We use subcutaneous heparin and pneumatic boots beginning in the operating room and continuing until postoperative day 7 or until patients are ambulatory. For those with high risk of having a thromboembolic event, thromboprophylaxis may be extended to 30 days. (See "Prevention of venous thromboembolic disease in adult nonorthopedic surgical patients".)

MORBIDITY AND MORTALITY — Morbidity and mortality rates vary, depending upon hospital and/or surgeon volume, patient comorbidities, and operative approach (eg, open, minimally invasive, tri-incisional) [164-167]. Complications, mortality, and management of complications are discussed in a separate topic. (See "Complications of esophageal resection".)

QUALITY OF LIFE — An esophagectomy can exert temporary [168-175] and long-term detrimental impacts on health-related quality of life (HR-QOL) [172,174,176-178]. In most patients, recovery seems to occur within 12 to 24 months, but even after three or more years, a substantial number of long-term survivors still report residual problems with eating, breathlessness, diarrhea, reflux, fatigue, and odynophagia [172,174,176,177]. In a self-reported quality-of-life study among 117 long-term-year survivors of esophageal cancer surgery, physical function was stable or improved at five years in 86 percent [179].

Recovery of HR-QOL may be related, at least in part, to the occurrence of postoperative complications. A prospective study that included 141 five-year survivors following an esophagectomy for cancer found that patients who sustained a major postoperative complication (eg, pneumonia, anastomotic leak; n = 46) had clinically and statistically significant more dyspnea compared with patients who did not have a major complication (mean score difference 15, 95% CI 6-23) [178]. Patients with major postoperative complications also had significantly more fatigue and eating restrictions (mean score difference 13, 95% CI 5-20, and mean score difference 10, 95% CI 2-17, respectively).

SPECIAL CONSIDERATIONS DURING THE COVID-19 PANDEMIC — The COVID-19 pandemic has increased the complexity of cancer care. Important issues include balancing the risk from delaying cancer treatment versus harm from COVID-19, minimizing the number of clinic and hospital visits to reduce exposure whenever possible, mitigating the negative impacts of social distancing on delivery of care, and appropriately and fairly allocating limited healthcare resources. Specific guidance for decision-making for cancer surgery on a disease-by-disease basis is available from the American College of Surgeons, from the Society for Surgical Oncology, and from others. These and other recommendations for cancer care during active phases of the COVID-19 pandemic are discussed separately. (See "COVID-19: Considerations in patients with cancer".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Esophageal cancer" and "Society guideline links: Esophagectomy".)

SUMMARY AND RECOMMENDATIONS

●The clinical spectrum of esophageal cancer has changed over the last few decades, with an increase in incidence of adenocarcinoma in the distal esophagus and a decrease in squamous cell carcinoma. However, patients with squamous cell carcinoma and adenocarcinoma are managed similarly from an operative point of view. (See 'Introduction' above.)

●The selection criteria for eligibility for a surgical resection include localized T1 to T3 lesions and select T4a lesions that involve the pericardium, pleura, or diaphragm (table 1). (See 'Criteria for resection' above.)

●Because most patients with a cervical esophageal cancer present with advanced disease, a surgical resection usually requires removal of portions of the pharynx, the larynx, the thyroid gland, and portions of the proximal esophagus. If there is sufficient margin on the upper esophageal sphincter, we perform a tri-incisional esophagectomy with a hand-sewn anastomosis to preserve esophageal length. (See 'Cervical esophageal cancer resection' above.)

●For patients with a thoracic esophageal cancer, we perform a total thoracic esophagectomy with cervical esophagogastrostomy, radical two-field lymph node dissection, and jejunostomy feeding tube placement. (See 'Thoracic cancer resection' above.)

●We generally prefer a tri-incisional approach that consists of initial right posterolateral thoracotomy or thoracoscopic approach, followed by a laparotomy or via laparoscopy to obtain complete esophageal dissection and mobilization of the gastric conduit, en bloc resection of both mediastinal and upper abdominal lymph nodes, and a left neck incision and cervical anastomosis. (See 'Tri-incisional esophagectomy' above.)

●For patients with an esophageal gastric junction cancer, we generally perform a total esophagectomy with cervical esophagogastrostomy with a partial gastrectomy or an extended gastrectomy, depending upon the extensiveness of the gastric involvement. An alternative to the total esophagectomy is the minimally invasive Ivor-Lewis procedure, with a thoracic esophagogastrostomy. However, there are insufficient data from randomized trials to determine the superiority of one approach over the other. (See 'Esophagogastric junction cancer resection' above and "Complications of esophageal resection", section on 'Conduit complications'.)

●The integration of surgery into multimodality treatment of localized cancer of the esophagus is discussed elsewhere. (See "Radiation therapy, chemoradiotherapy, neoadjuvant approaches, and postoperative adjuvant therapy for localized cancers of the esophagus".)