The incidence of esophageal cancer in the United States is approximately 13,000 cases per year. Adenocarcinoma of the esophagus (EAC) is the fastest growing solid malignancy, and it is now more frequent than esophageal squamous cell carcinoma (ESCC) in the United States. Although resection offers the best chance for cure, most patients who present with symptoms already have systemic disease and are incurable. The most important prognostic factor for esophageal cancer remains the stage of disease at the time of diagnosis. Currently, the best strategy for improving survival is early diagnosis followed by resection. Neoadjuvant chemoradiation is being used increasingly in the treatment of locally advanced disease. However, despite improvements in operative technique and postoperative care, the overall 5-year survival remains less than 15 percent, highlighting the need for novel diagnostic and treatment modalities.
Esophageal Squamous Cell Carcinoma
Nutritional deficiencies, including low levels of vitamins A, C, and riboflavin as well as minerals such as selenium, molybdenum, and zinc contribute to the pathogenesis of ESCC. High levels of nitrates and nitrites, found in pickled and preserved foods, are converted to N-nitrosamines and have been associated with ESCC. Several studies have linked alcohol and tobacco use with the development of esophageal cancer, and one study found ethanol to be associated with nearly 80 percent of esophageal cancers.1 Other predisposing conditions include long-standing achalasia, caustic injury, tylosis, and Plummer–Vinson syndrome.
Barrett metaplasia, and specifically intestinal-type columnar mucosa, is the precursor lesion to EAC. Microscopically, columnar epithelium is seen with mucosal glands that contain intestinal goblet cells (Fig. 16-1A). On endoscopy, Barrett mucosa appears as red, velvety areas between smooth, pale esophageal squamous mucosa (Fig. 16-2A). After the squamous epithelium is injured chronically by reflux, there is a metaplastic change to a columnar epithelium. With further injury and multiple genetic changes, progression occurs in a metaplasia–dysplasia–adenocarcinoma sequence.
Photomicrograph. A. Barrett mucosa showing columnar epithelium with mucosal glands containing intestinal goblet cells. B. High-grade dysplasia with nuclear atypia.
Endoscopic view. A. Barrett metaplasia with tongues of red, velvety mucosa extending above the gastroesophageal junction. B. A 5-cm distal esophageal adenocarcinoma extending to the gastroesophageal junction and involving one-third of the lumen. C. Bronchoscopy showing tracheal invasion by an adjacent esophageal squamous cell carcinoma.
In patients with Barrett metaplasia, the risk of EAC is increased 30 to 125 times that of the age-matched population. Barrett esophagus results from chronic inflammation due to gastroesophageal reflux. At endoscopy, Barrett mucosa is found in 12 to 18 percent of patients with reflux.2 Although the esophageal mucosa is relatively resistant to acid, mixed reflux with bile acids, pepsin, and gastric acid appears to be more harmful, and up to 67 percent of patients with Barrett metaplasia have bile-stained duodenogastric reflux.
EAC is believed to have a 4- to 5-year preclinical phase, suggesting that surveillance may be effective. Cancers detected during surveillance tend to be less advanced, and the American College of Gastroenterology has suggested that patients with chronic reflux undergo upper endoscopy. Abnormal mucosa is biopsied to diagnose Barrett metaplasia and to detect the presence of dysplasia (Fig. 16-1B). Surveillance intervals are increased to every 3 years after two consecutive biopsies are negative for dysplasia.
Other risk factors for EAC include ectopic gastric mucosa, obesity, smoking, diets high in saturated fat and red meat, and medications that reduce lower esophageal sphincter pressure. Concerns have also been raised that the treatment of Helicobacter pylori, specifically the acid-suppressing effects of the Cag A+ strain, may actually increase the incidence of EAC.
Although the majority of ESCCs are located in the middle third of the esophagus, EACs (approximately 79 percent) arise in the distal esophagus. High-grade dysplasia (HGD) remains the best predictor of progression to adenocarcinoma. An extensive lymphatic network and the long interval during which esophageal tumors remain asymptomatic contribute to the high incidence of nodal disease with 30 to 58 percent of patients having lymph node metastases when the submucosa is involved.3 The extent of invasion and the presence or absence of lymph node metastases are important in determining patient prognosis. The most common sites of distant metastases include the liver, lungs, bone, brain, and adrenal glands (Table 16-1).
Table 16-1:Occurrence of Metastases in Patients with Esophageal Cancer |Favorite Table|Download (.pdf) Table 16-1: Occurrence of Metastases in Patients with Esophageal Cancer
|Sites ||Occurrence (%) |
|Liver ||35 |
|Lung ||20 |
|Bone ||9 |
|Adrenal glands ||2 |
|Brain ||2 |
|Pericardium ||1 |
|Pleura ||1 |
|Stomach ||1 |
|Pancreas ||1 |
|Spleen ||1 |
In the United States, the annual incidence of ESCC is 2.6 per 100,000. The incidence is four to five times higher in African Americans than in Caucasians. Regions with a high incidence are generally located in poorer areas of the world including parts of China, Central Asia, and Latin America. The overall 5-year survival rate is approximately 15 percent.
The incidence of EAC has increased progressively from the 1970s, and esophageal adenocarcinoma is the most rapidly increasing solid malignancy in the United States. Adenocarcinoma has surpassed ESCC as the most common cancer of the esophagus in America, and esophageal carcinoma is the seventh leading cause of cancer-related deaths. The median age at diagnosis is 68. Men are affected six to eight times more frequently than are women, and Caucasians are affected three to four times as often as African Americans. The incidence is higher in developed countries.
Dysphagia is the most common initial symptom but generally only occurs when the tumor involves more than 50 percent of the circumference of the esophagus. Odynophagia is the next most common symptom and may be caused by an ulcerated lesion. Constant pain in the midback or midchest suggests mediastinal invasion. Hoarseness may develop with proximal tumors and indicates involvement of the recurrent laryngeal nerve. Regurgitation may occur as the growing tumor narrows the esophageal lumen, and weight loss is frequently present at presentation. Although patients often have a history of reflux symptoms, clinical features do not distinguish patients with or without Barrett, which is frequently asymptomatic. The evaluation of a suspected esophageal adenocarcinoma is outlined in Figure 16-3.
Barium esophagram demonstrating a circumferential distal esophageal mass proximal to a hiatal hernia consistent with a Barrett adenocarcinoma associated with gastroesophageal reflux.
The management and prognosis of esophageal carcinoma depend on accurate clinical staging. Endoscopic evaluation is essential in all patients suspected of having esophageal carcinoma (Fig. 16-2B). The location of the lesion, degree of obstruction, and the length of the lesion should be determined in all patients. Any abnormal lesions should be biopsied. Other findings to be noted include the presence of Barrett mucosa or a hiatal hernia, and the fundus should be closely examined on retroflexion for any tumor involvement. Although advanced carcinoma is easily identified, early mucosal changes may be difficult to recognize. Chromoendoscopy with indigo carmine or acetic acid and narrow band imaging can highlight mucosal details and the microvasculature by using blue and green light with shorter wavelengths resulting in better penetration of superficial tissues. Lugol’s solution can be used to identify dysplastic squamous lesions. Normal epithelium, which contains glycogen, is stained by the iodine. New techniques are being developed to improve visualization including confocal endomicroscopy and fluorescent-labeled peptides designed to bind to dysplastic or cancerous lesions.4 Mapping biopsies with four-quadrant biopsies obtained every 1 to 2 cm can also help to delineate abnormal lesions.
Bronchoscopy is important to evaluate possible tracheal invasion by tumors in the upper or middle thirds of the esophagus (Fig. 16-2C). Patients with infracarinal bulky tumors or subcarinal lymphadenopathy on computed tomography (CT) should also undergo bronchoscopy. Findings may range from bulging of the bronchial wall and subtle mucosal changes to frank tumor invasion or the presence of a tracheoesophageal fistula. The carina may appear widened due to subcarinal lymphadenopathy. Bronchial biopsies are obtained to confirm that airway invasion is present.
The barium swallow is an essential diagnostic study in the evaluation of dysphagia and allows visualization of the esophageal mucosa, distensibility, and motility (Fig. 16-3). Involvement of the fundus can also be seen on the esophagram. Subtle mucosal irregularities, ulcerations, or small polypoid lesions are best seen on a double-contrast esophagram.
Staging CT scans of the chest and upper abdomen are useful in the initial evaluation of esophageal carcinoma as well as assessing the response to neoadjuvant therapy. CT is used to determine the local extent of the tumor, the relationship to adjacent structures, lymphadenopathy, and the presence of distant metastases. CT cannot distinguish the layers of the esophageal wall and is not as useful in determining the T stage. However, loss of fat planes can be useful in determining invasion of adjacent structures. The sensitivity for detecting lymphadenopathy (>1 cm) is 30 to 60 percent in the mediastinum and 50 to 75 percent in the abdomen, whereas CT has a 70 to 80 percent sensitivity for identifying metastatic disease greater than 2 cm.5
Endoscopic ultrasound (EUS) provides the most accurate assessment of the depth of tumor invasion, or T stage, and is able to identify five distinct layers in the esophageal wall (Fig. 16-4). EUS is most useful in the staging of early carcinoma. EUS is also more sensitive in detecting regional lymphadenopathy and can evaluate the extent of periesophageal invasion. Curvilinear arrays assist in the performance of fine-needle aspiration (FNA), which is up to 98 percent sensitive and 100 percent specific. EUS has also been evaluated for assessing the response to neoadjuvant therapy but may be prone to overstaging.
A. Endoscopic ultrasound shows a hypoechoic 6-cm distal esophageal mass with sonographic evidence of invasion into the adventitia. B. Sonographic view of an abnormal hypoechoic lymph node with well-defined margins visualized in the gastrohepatic ligament.
Positron Emission Tomography.
Positron emission tomography (PET) improves staging by identifying distant disease not seen on CT alone. PET has a higher sensitivity than CT in detecting nodal and distant metastases, although PET cannot distinguish the primary tumor from peritumoral lymph nodes. Sensitivity ranges from 78 to 95 percent, and specificity from 95 to 100 percent. The role of PET in restaging after neoadjuvant treatment needs to be further evaluated. Although studies suggest that decreased activity seen on PET corresponds with response to chemoradiation, PET is not sensitive enough to identify complete responders.
Thoracoscopy and Laparoscopy
Thoracoscopic and laparoscopic staging offer direct visualization and the ability to obtain biopsies to determine nodal status, the extent of local invasion, and the presence of metastatic disease and can be useful in specific circumstances. Laparoscopy should be considered in patients with suspected intraperitoneal metastases to assess for tumor implants whereas thoracoscopy is useful to evaluate for small pulmonary nodules or mediastinal lymph nodes not accessible by endobronchial ultrasound (EBUS) or CT-guided biopsy. Laparoscopic ultrasound is also being investigated as a staging modality and may provide improved accuracy in T and N staging. Laparoscopic ultrasound may be more accurate in staging celiac nodes than EUS, providing closer access for the ultrasound probe as well as direct inspection.
Accurate staging is essential for treatment selection, and the stage at diagnosis is the most important prognostic factor. The evaluation of patients with suspected esophageal cancer should include appropriate diagnostic studies as described above as well as biopsies of any suspected metastatic lesions (Fig. 16-5). Although patients with stage I tumors have a 65 percent 5-year survival, those with biopsy-proven distant metastatic disease have a mean survival of only 6 months and are considered unresectable. Both esophageal squamous cell and adenocarcinoma are staged according to the American Joint Committee for Cancer TNM system. Significant changes in the 7th edition take into account the histology, grade, and number of regional lymph nodes involved.
Diagnostic algorithm in the evaluation of a patient with suspected esophageal adenocarcinoma. FNA, fine-needle aspiration.
Results of a large review published by Earlam and Cunha-Melo with more than 8000 inoperable patients due to prohibitive surgical risk or unresectable esophageal carcinoma showed 1-, 2-, and 5-year survival rates of 18, 8, and 6 percent after definitive radiotherapy.6 Postoperative radiotherapy has been shown to decrease local recurrence but has not improved survival rates. Several trials of neoadjuvant radiotherapy have also failed to show increased resection rates or improved survival compared with surgery alone.
Law et al. showed that neoadjuvant chemotherapy resulted in significant downstaging of disease in nearly 50 percent of patients, but a pathologic complete response was seen in less than 10 percent.7 A marginal survival benefit was noted in this subgroup of patients. Kelsen et al. reported a large multicenter study of neoadjuvant chemotherapy versus surgery alone, which included 440 patients. At 2-year follow-up, no significant survival benefit was seen.8
The Radiation Therapy Oncology Group (RTOG) randomized trial reported by Al-Sarraf et al. provided convincing evidence of the superiority of chemoradiation over radiation alone.9 The 2- and 5-year survival rates were 36 and 27 percent in the chemoradiation group compared with 10 and 0 percent in the group treated with radiation alone. There was also a reduction in both local recurrence and distant disease. However, the toxicity of the treatments was significant. Many surgeons advocate neoadjuvant therapy for patients with esophageal carcinoma, except in patients with early disease, due to the poor survival rates following resection alone. Neoadjuvant chemoradiotherapy is routine at the University of Michigan in patients with tumors staged as T2 or greater or patients with positive nodal disease (Fig. 16-5). Neoadjuvant therapy is limited to patients younger than 75 years due to the significant toxicity and the decrease in patients able to complete trimodality therapy with increasing age.
The combination of chemoradiation with surgery has resulted in significant downstaging of disease, but a survival advantage has not been consistently demonstrated in randomized trials. Most regimens are cisplatin-based and commonly combined with 5-fluorouracil. Chemoradiation is followed by resection 5 to 6 weeks after completion of therapy. Walsh et al. reported a randomized trial of 113 patients receiving neoadjuvant therapy or surgery alone and showed significantly improved survival after neoadjuvant treatment of 17 months versus 12 months.10 However, the survival in the surgery arm was lower than most published series. Urba et al. performed a randomized trial of 100 patients who underwent either transhiatal esophagectomy following neoadjuvant chemoradiation therapy or surgery alone.11 Seventy-five percent of patients had adenocarcinoma. Twenty-eight percent had a complete pathologic response. Survival at 3 years was 30 percent after neoadjuvant therapy versus 16 percent after surgery alone although this was not statistically significant. A meta-analysis of randomized trials comparing trimodality therapy with surgery alone showed a slight increase in overall survival with a complete pathologic response associated with the best survival.12 Gebski et al. performed a meta-analysis including 1200 patients undergoing neoadjuvant chemoradiation and found a hazard ratio (HR) of 0.81 (95 percent CI 0.70–0.93; p = 0.002).13
Cancer and Leukemia Group B (CALGB) 9781 was a prospective, randomized phase III trial comparing trimodality therapy with cisplatin and 5-FU to surgery alone.14 Although the study was closed early due to low accrual after 56 patients, the authors reported a 5-year survival of 39 percent compared with 16 percent in patients undergoing surgery alone supporting the use of trimodality therapy. Newer agents including taxanes and irinotecan are being evaluated in chemoradiotherapy trials and may have promising antitumor activity as well as improved tolerance.
In addition, new targeted therapies against growth factor receptors such as EGFR and HER-2/Neu and angiogenesis factors including VEGF are being investigated in the treatment of esophageal carcinoma. Inhibitors include small-molecule tyrosine kinase inhibitors and monoclonal antibodies. Several phase II trials evaluating EGFR inhibitors have been completed in patients with advanced esophageal carcinoma with modest response rates seen in 10 percent of patients.15 Larger phase III trials will be needed to determine which patients may benefit from therapy. The ToGA trial was a randomized, prospective phase III trial evaluating 594 patients with advanced HER-2-positive gastric and gastroesophageal junction (17–20 percent of patients) tumors with cisplatin and 5-FU with or without trastuzumab, an anti-HER-2 monoclonal antibody.16 There was a significant increase in survival of 13.5 versus 11.1 months (HR 0.74; 95 percent CI 0.60–0.91; p = 0.046) and progression-free survival of 6.7 versus 5.5 months. Complete response was seen in 5 percent versus 2 percent and a partial response in 42 percent versus 32 percent. Patients with high HER-2 expression had a median overall survival of 16 months (HR 0.65; 95 percent CI 0.51–0.83). For the subgroup of patients with gastroesophageal junction tumors, The HR was 0.67 (95 percent CI 0.42–1.08). Trastuzumab was recently added to the National Comprehensive Cancer Network Guidelines for combination treatment of advanced esophageal and gastric adenocarcinomas. RTOG 1010 is currently accruing patients in a phase III trial evaluating the addition of trastuzumab to trimodality therapy in HER-2 overexpressing esophageal adenocarcinomas.
At this time, surgery remains the standard treatment for resectable esophageal carcinoma. For patients with advanced disease or those unfit for surgery, chemoradiation therapy appears to be a reasonable alternative.
Some investigators have had success with mucosal ablation of Barrett mucosa with squamous re-epithelialization. Effective antireflux therapy is critical. Techniques include laser, photodynamic therapy, cryotherapy, and radiofrequency ablation. Salo et al. reported successful Nd-YAG laser ablation after antireflux surgery in 11 patients who had complete squamous regeneration after a mean of 26 months.17 However, Wang and Sampliner have found persistent areas of intestinal metaplasia underlying the squamous epithelium.18
Photodynamic therapy (PDT) involves the systemic injection of a photosensitizer, porfimer sodium, targeting dysplastic cells. After 24 to 48 hours, light is used to produce oxygen radicals. All devitalized tissue is irrigated away 48 hours later. Using this technique, Overholt et al. reported 90 percent of lesions cleared or downstaged to low-grade dysplasia although 83 percent required repeat laser treatment 3 to 6 months later for residual Barrett mucosa.19 Esophageal mucosal re-epithelialization occurred over 2 to 3 months and was complicated by strictures in 34 percent. Patients also need to avoid sun exposure due to photosensitivity for up to 3 months.
In the prospective, multicenter AIM-II trial endoscopic radiofrequency ablation was evaluated in patients with nondysplastic Barrett mucosa. Complete eradication was achieved in 98.4 percent of patients at 2.5 years and 92 percent at 5 years (46/50 patients). The four patients with focal Barrett were re-treated with RFA. The authors reported no buried glands, dysplasia, or strictures. Although this technique may be useful for Barrett mucosa, as an ablative procedure, pathology specimens are not obtained, and ablation should not be used for HGD or esophageal carcinoma where pathologic examination to determine the histology, depth of invasion, lymphovascular involvement, and margins is critical.
Endoscopic Mucosal Resection.
Endoscopic mucosal resection (EMR) was first reported in Japan and involves injecting saline and epinephrine into the submucosal space to elevate the target area, which is then lifted with a suction cap or banded and removed with a polypectomy snare. Failure of the lesion to lift after saline injection suggests submucosal invasion. High-frequency EUS is useful in defining the depth of invasion prior to resection. Risks include bleeding, perforation, and stricture. The overall complication rate is 13 to 17 percent.20 Strictures are reported in 6 to 23 percent of cases and are more likely after circumferential EMR or in combination with ablation such as PDT. The risk of bleeding is 0.6 to 11 percent. Bleeding requiring transfusions occurs in 3.8 percent, and perforation occurs in 0 to 2.5 percent.
Recurrence of dysplasia or carcinoma has been reported in 12 to 21 percent of patients with a mean follow-up of 43 to 63 months although these lesions were often treated with repeat EMR.20 The risk of recurrence is increased with long-segment Barrett (relative risk 1.9; 95 percent CI 1.06–3.3), larger lesions requiring piecemeal resection (RR 2.44; 95 percent CI 1.13–4.89), and multifocal carcinoma (RR 2.1; 95 percent CI 1.16–3.99). Residual metaplastic mucosa can also persist beneath the regenerating squamous mucosa. Close endoscopic follow-up is mandatory and persistent lesions, involvement of the submucosa, or positive margins warrants an esophagectomy.
HGD is being treated increasingly with EMR although esophagectomy remains the gold standard. Another alternative is intensive surveillance with four-quadrant biopsies taken every centimeter and repeated every 3 months. However, there are inherent false negatives as biopsies may miss dysplasia and carcinomas separated by large areas of Barrett epithelium. Several retrospective studies have shown difficulty in differentiating HGD from adenocarcinoma using endoscopic surveillance. In addition, the majority of patients with HGD will have invasive adenocarcinoma develop within 10 years with 25 percent at 1.5 years, 50 percent at 3 years, and 80 percent at 8 years on surveillance biopsies. Zhu et al. evaluated 127 esophagectomy specimens performed for high-grade dysplasia.21 Patients with biopsies showing only HGD were found to have adenocarcinoma in 5 percent of esophagectomy specimens while more than 80 percent of patients with HGD suspicious for carcinoma were found at resection to have adenocarcinoma suggesting that these patients may be more appropriately treated as patients with carcinoma.
Early intramucosal carcinomas (T1a) have also been treated with endoscopic resection especially in debilitated patients who are not surgical candidates. EACs confined to the mucosa have 1 to 5 percent risk of lymph node metastases while invasion of the submucosa increases the risk to 30 to 58 percent.3,20 Long-term results are not yet available and surgery remains the gold standard with curative intent in these patients.
Surgical therapy currently offers the best chance for cure with a complete R0 resection and provides effective palliation with relief of dysphagia. Although mortality rates have decreased significantly over the past few decades, relatively high postoperative morbidity remains. Appropriate pulmonary and cardiac evaluation as well as aggressive preoperative preparation including preoperative ambulation, nutrition, incentive spirometry, patient education, and smoking cessation are critical. Although the goal is to proceed with resection 5 to 6 weeks after completion of neoadjuvant therapy, it is important to wait until patients have returned to their baseline function prior to proceeding with esophagectomy.
Controversy continues to surround the optimal surgical approach to esophagectomy as well as the extent and necessity of regional lymph node dissection. Proponents of the combined thoracotomy and laparotomy approach described by Ivor Lewis emphasize the direct visualization during mediastinal dissection, an en bloc resection, and a wider lymphadenectomy resulting in improved staging. Two meta-analyses comparing the two approaches showed higher perioperative morbidity and mortality after the transthoracic approach with no difference in long-term survival.22,23 The choice of approach depends on the location of the tumor, planned extent of lymphadenectomy, and the preference of the surgeon. The role of palliative resection is limited with the availability of esophageal stents, laser ablation, and radiation therapy.
The transhiatal approach avoids the morbidity of a thoracotomy in debilitated patients as well as the potential occurrence of a disastrous intrathoracic anastomotic disruption. Although accessible subcarinal, paraesophageal, and celiac lymph nodes are sampled, no attempt is made to perform an en bloc resection. Although critics have questioned the oncologic principles arguing for the need to perform an aggressive mediastinal lymph node resection, survival has been reported in several series to be equivalent to that seen after a transthoracic resection. Transhiatal esophagectomy is technically possible in the majority of patients even with a history of prior operations, esophageal perforation, or radiation treatment and provides a maximal esophageal margin. In addition, the stapled side-to-side cervical anastomosis has decreased the anastomotic leak rate to less than nine percent at the University of Michigan, and most cervical leaks are treated conservatively by opening the wound at the bedside and dressing changes.24,25 Less frequent complications include intrathoracic hemorrhage, recurrent nerve paralysis, and tracheal laceration.
The abdomen is entered through a midline supraumbilical incision, and the liver and peritoneal cavity are examined for metastatic disease. The triangular ligament is divided, and the stomach is carefully examined for tumor involvement. The right gastroepiploic artery is identified and protected. This may be difficult if there is a history of previous abdominal surgery. The left gastroepiploic and short gastric vessels are ligated, and the esophagogastric junction is mobilized. After examining the gastrohepatic ligament for a replaced left hepatic artery, the left gastric artery is ligated near its origin taking the left gastric lymph nodes with the specimen. The right gastric artery is preserved as the dissection is continued along the lesser curvature. A Kocher maneuver is then performed to allow mobilization of the pylorus to the level of the hiatus. A pyloromyotomy is performed to decrease the incidence of postvagotomy delayed gastric emptying. Clips are placed at the level of the pyloromyotomy for future radiographic studies. A jejunostomy tube is placed and is used as needed to supplement postoperative nutrition.
The distal 5 to 10 cm of esophagus is mobilized through the hiatus. The mobility of the esophageal tumor is then assessed to ensure that there is no invasion of the prevertebral fascia, aorta, or surrounding mediastinal structures. Deaver retractors inserted into the hiatus allow ligation of the periesophageal tissues to the level of the carina under direct vision using long right-angle clamps.
An oblique incision is made along the anterior border of the sternocleidomastoid muscle. Care is taken to avoid retraction on the left recurrent laryngeal nerve in the tracheoesophageal groove. The middle thyroid vein and inferior thyroid artery are ligated as needed. After dissecting down to the prevertebral fascia, blunt finger dissection is continued into the superior mediastinum. Sharp dissection is used along the anterolateral surface of the esophagus staying posterior to the recurrent laryngeal nerve. The upper thoracic esophagus is mobilized almost to the level of the carina using blunt dissection keeping the fingers directly on the esophagus.
One hand is inserted through the diaphragmatic hiatus posterior to the esophagus while a sponge stick is placed through the cervical incision dissecting the esophagus off the prevertebral fascia (Fig. 16-6A). The blood pressure is carefully monitored to prevent hypotension, and a sump catheter is inserted through the cervical incision to evacuate blood from the posterior mediastinum. The anterior mobilization is performed from both the abdominal and the cervical incisions with the fingers directly against the anterior esophagus to avoid injury to the membranous trachea. The esophagus is held in the superior mediastinum between the index and middle fingers of the hand inserted through the hiatus, and the remaining attachments are lysed with a downward motion (Fig. 16-6B). A partial upper sternotomy can be performed for improved exposure in patients with a short, broad neck or poor neck extension. The upper esophagus is then divided obliquely, and the thoracic esophagus is delivered through the diaphragmatic hiatus. Using Deaver retractors, the posterior mediastinum is inspected for hemostasis through the hiatus. If the pleural cavities have been entered, which occurs in three-quarters of cases, chest tubes are placed. The posterior mediastinum is temporarily packed with gauze pads to tamponade minor bleeding.
A. Esophageal mobilization from the prevertebral fascia is performed using a half-sponge stick from the cervical wound and a hand inserted through the diaphragmatic hiatus. (From Orringer MB, Sloan H. Esophagectomy without thoracotomy. J Thorac Cardiovasc Surg 1978;76(5):643–654. Copyright Elsevier.) B. A downward raking motion with the esophagus held between the index and the middle fingers is used to take down the lateral periesophageal attachments. (From Orringer MB. Transhiatal blunt esophagectomy without thoracotomy. In: Cohn LH (ed.). Modern Techniques in Surgery, Vol. 62. Cardiovascular Surgery. New York: Futura Publishing, 1983:1.)
A partial proximal gastrectomy 4 to 6 cm distal to the esophagogastric junction is performed using a GIA stapler (Fig. 16-7A). For tumors of the middle esophagus or benign disease, the amount of stomach resected is minimized to preserve collateral circulation to the fundus. The staple line is oversewn. The mobilized stomach is passed through the hiatus and delivered 4 to 5 cm above the clavicles by pushing the stomach up through the mediastinum and guiding it with a Babcock clamp from the cervical incision (Fig. 16-7B). Traction sutures and suction devices to pull the stomach through the mediastinum are avoided to minimize trauma to the gastric conduit. Care is taken to ensure that there is no torsion of the conduit.
A. Partial proximal gastrectomy is performed by sequentially applying the GIA stapler. (Reproduced with permission from Orringer MB, Sloan H. Esophageal replacement after transhiatal esophagectomy without thoracotomy. In Nyhus LM, Baker RJ (eds). Mastery of Surgery, 2nd ed. Boston: Little, Brown, 1992:569.) B. The mobilized stomach is gently passed into the posterior mediastinum. A Babcock clamp is used to carefully grasp the stomach in the superior mediastinum, but the stomach is delivered primarily by pushing from below. (Reproduced with permission from Orringer MB, Marshall B, Iannettoni MD. Eliminating the cervical esophagogastric anastomotic leak with a side-to-side stapled anastomosis. J Thorac Cardiovasc Surg 2000;119(2):277–288. Copyright Elsevier.)
The abdominal phase is then completed to avoid contamination with bacteria from the cervical esophagus. The diaphragm is reapproximated using silk sutures so that the hiatus allows three fingers to pass alongside the stomach. The anterior gastric wall is approximated to the edge of the hiatus, and the triangular ligament is re-attached to the diaphragm to help prevent a hiatal hernia. The pyloromyotomy is covered with omentum.
Cervical Esophagogastric Anastomosis.
A traction suture is placed in the anterior gastric wall to elevate the stomach into the wound. A 1.5-cm vertical gastrotomy is made in the anterior gastric wall low enough down to allow the Endo GIA to be fully inserted into the stomach (Fig. 16-8A). The cervical esophageal staple line is then amputated obliquely with enough redundancy for a tension-free anastomosis and sent to pathology as the proximal esophageal margin. After placing stay sutures to align the stomach and esophagus, the Endo GIA is inserted and closed (Fig. 16-8B). Before opening the stapler, two suspension sutures between the esophagus and adjacent stomach are placed on either side of the anastomosis. After removing the stapler, a nasogastric tube is guided across the anastomosis, and the anterior wall of the anastomosis is reapproximated in two layers (Fig. 16-8C). Clips are placed beside the anastomosis as radiographic markers, and the wound is closed after placing a penrose drain.
Cervical stapled side-to-side anastomosis after transhiatal esophagectomy. A. The esophagogastrostomy is aligned using stay sutures. B. An Endo-GIA stapler is inserted. C. The semimechanical anastomosis is completed. D. The completed cervical side-to-side anastomosis. E. Postoperative barium swallow. (Reproduced with permission from Orringer MB, Marshall B, Iannettoni MD. Eliminating the cervical esophagogastric anastomotic leak with a side-to-side stapled anastomosis. J Thorac Cardiovasc Surg 2000;119(2):277–288. Copyright Elsevier.)
Minimally Invasive Approaches.
Luketich et al. reported a series of 222 patients undergoing a minimally invasive laparoscopic and thoracoscopic approach with a cervical anastomosis.26 A pyloroplasty is performed and the cervical anastomosis is completed using an EEA stapler. Neoadjuvant chemotherapy was used in 35.1 percent and radiation in 16.2 percent. Conversion to an open esophagectomy was performed in 7.2 percent. The average hospital stay was 7 days, and operative mortality was 1.4 percent. The anastomotic leak rate was 11.7 percent. The leak rate was affected by the width of the gastric conduit with a leak rate of 6.1 percent with a conduit greater than 6 cm in width versus 25.9 percent with a narrower conduit. Recurrent laryngeal nerve injury occurred in 3.6 percent and pneumonia in 7.7 percent.
More recently, this group has described performing a minimally invasive Ivor Lewis esophagectomy in 50 patients.27 Fifteen of the patients were approached by thoracoscopy while the remainder had a planned mini thoracotomy for the anastomosis that was performed with an EEA stapler. The operative mortality was 6 percent and the leak rate 6 percent. Although there were no recurrent laryngeal nerve injuries with avoidance of the cervical dissection, a return to an intrathoracic anastomosis could increase the morbidity of an intrathoracic leak compared with the relatively benign consequences of a cervical anastomotic leak.
Minimally invasive esophagectomy is technically demanding and should be performed in centers with extensive laparoscopic and thoracoscopic experience. Care must be taken to balance the benefits of smaller incisions and less pain from retraction with not compromising key components of the esophagectomy including avoiding trauma to the gastric conduit, adequately unrolling the stomach so that it reaches to the neck without tension, making an adequate mediastinal tunnel to avoid venous congestion of the conduit, and performing an adequate gastric drainage procedure. Further prospective studies are needed to determine if there are significant differences in costs, complication rates, and postoperative pain and recovery.
Five-year survival for patients with esophageal carcinoma remains poor at less than 15 percent. Tumor stage and lymph node status are the strongest predictor of survival. Other factors associated with a poor prognosis include increased age, African American race, lower esophageal tumors, and increasing length and depth of the tumor. Five-year survival greater than 80 percent can be seen after esophagectomy for early lesions limited to the mucosa, and early diagnosis continues to be the best strategy for improving survival. Overall, esophagectomy is potentially curative in 20 percent of patients.
Orringer et al. reported the largest experience with transhiatal esophagectomy in 2007 patients.24 When comparing group I (1976–1998) and group II (1998–2006), there was a significant decreases in in-hospital mortality (4 vs. 1 percent). A gastric conduit was used in 97 percent of patients and only 4 percent of patients in group II required intensive care postoperatively. Complications included recurrent nerve palsy (<1 percent), atrial fibrillation (6.6 percent), and pneumonia (2 percent). The incidence of anastomotic leak decreased from 14 to 9 percent in group II (p <0.001) and correlated with the introduction of the side-to-side cervical esophagogastric anastomosis in 1997.25 Ninety-four percent of those with leaks were treated with wound packing alone. Chylothorax occurred in less than 1 percent and should be treated aggressively with thoracic duct ligation. Fifty percent of patients in group II were discharged at 1 week with a median length of stay of 8 days. In cancer patients, functional results were good in 80 and 75 percent reporting no regurgitation or dumping. Fifty-two percent of patients in group II underwent neoadjuvant chemoradiation. A complete pathologic response was found in 21 percent of patients with a significant improvement in 5-year survival of 58 percent in complete responders compared with 22 percent for those with residual disease.
Several studies have evaluated the relationship of hospital volume to outcomes after a variety of complex procedures including coronary artery bypass grafting, aortic aneurysm repair, heart transplantation, and pancreatic resections. Esophagectomy is technically demanding, and outcomes have been shown to be related to hospital volume. Dimick et al. reviewed 1136 patients in a Maryland statewide database and found that high-volume centers, defined as hospitals performing greater than 15 esophagectomies per year, had a fivefold reduction in hospital mortality.28 Kuo et al. found similar results in 1193 patients in Massachusetts with a 2-day decrease in length of stay and a 3.7-fold decrease in hospital mortality in centers performing greater than six esophagectomies per year.29 Outcomes have also been associated with the experience of the individual surgeon. Miller et al. found a significant decrease in hospital mortality rates in patients treated by surgeons who perform greater than six esophagectomies per year.30