Published: August 2010
Esophageal cancer has long been considered one of the deadliest malignancies, and considerable controversy has surrounded its management. The most common histologic types are squamous cell carcinoma (SCC) and adenocarcinoma (AC), which together constitute more than 90% of esophageal malignancies. Rarely, melanoma, sarcoma, small cell carcinoma, or lymphoma may arise in the esophagus. Although SCC is more evenly distributed throughout the length of the esophagus, AC is predominantly a disease of the distal esophagus and gastroesophageal junction, and is rarely found in the cervical esophagus.
Cancers arising from the esophagus and gastroesophageal junction are relatively uncommon in the United States; there were approximately 14,520 new cases and 13,570 deaths in 2005. Worldwide, however, esophageal cancer is the eighth most common malignancy and the sixth most common cause of cancer-related death. The epidemiology of esophageal cancer changed dramatically during the latter half of the 20th century. Although 40 years ago SCC accounted for more than 90% of all esophageal tumors in the United States, diagnoses of esophageal AC have significantly increased and now represent 80% of cases. However, SCC remains the most common worldwide. The mean age at diagnosis is 67 years, and men are affected more frequently than women, particularly among patients with AC.
There are considerable geographic and racial variations in the incidence of this cancer, which is mostly explained by varying exposure to risk factors, although genetic susceptibility may play a partial role. Many of the causative and risk factors for AC and SCC have been well established (Box 1).
|Box 1 Causative and Risk Factors for Adenocarcinoma and Squamous Cell Carcinoma|
|Gastroesophageal reflux disease (GERD)|
|Obesity (by increasing the risk of GERD)|
|Squamous Cell Carcinoma|
|Dietary and environmental factors that cause chronic irritation and inflammation of the esophageal mucosa|
|Predisposing underlying conditions, such as tylosis, achalasia, esophageal diverticula and webs, Plummer-Vinson syndrome, and human papillomavirus (HPV) infection|
Smoking and heavy alcohol intake are important risk factors for the development of SCC. Smoking has a synergistic effect with heavy alcohol consumption, and heavy exposure to both increases the risk of SCC by a factor of more than 100. This is further complicated by the increased risk of other aerodigestive tract cancers in a person who smokes and drinks alcohol.
Dietary and environmental factors, and certain esophageal disorders (e.g., achalasia, diverticuli) that cause chronic irritation and inflammation of the esophageal mucosa may also increase the incidence of SCC. Plummer-Vinson syndrome—the triad of dysphagia, iron deficiency anemia, and esophageal webs—has been associated with this cancer, although it is becoming increasingly rare in the developed world as overall nutrition improves. There are few genetic factors that have been identified as being important in the development of esophageal SCC. One exception is tylosis, a rare autosomal dominant syndrome associated with hyperkeratosis of the palms and soles and a high rate of esophageal SCC. Infectious agents have also been implicated in the pathogenesis of esophageal SCC. Human papillomavirus has received the most attention. It is believed that the infection results in loss of function of the tumor suppressor genes p53 and Rb. The importance of this mechanism is not well established.
The risk factors for AC of the esophagus are different. Chronic gastroesophageal reflux is the most important, with severe, long-standing reflux symptoms increasing the risk of cancer by a factor of 40. Chronic gastroesophageal reflux disease is associated with Barrett’s metaplasia (Barrett’s esophagus), a condition in which an abnormal columnar epithelium replaces the stratified squamous epithelium that normally lines the distal esophagus. Most esophageal ACs are believed to arise from Barrett’s esophagus. Although this mucosal change appears to be a favorable adaptation to chronic reflux—columnar epithelium appears to be more resistant to reflux-induced injury than the native squamous cells—this specialized intestinal metaplasia may become dysplastic and ultimately malignant, with genetic alterations that activate proto-oncogenes, disable tumor suppressor genes, or both. Factors that increase the risk for gastroesophageal reflux, such as obesity or medications that lower the lower esophageal sphincter tone, may result in an increased risk for esophageal AC.
An infectious etiology for this disease has not been identified and, unlike AC of the gastric cardia, the role of Helicobacter pylori colonization is unknown. The genetic and molecular changes underlying the development of esophageal AC also remain poorly understood, although allelic losses at chromosomes 4q, 5q, 9p, 9q, and 18q and abnormalities of p53, Rb, cyclin D1, and c-myc have been implicated.
The esophagus itself has several unique properties that distinguish the behavior of cancer in this organ from those of other gastrointestinal malignancies. In contrast to the rest of the gastrointestinal tract, the esophagus has no serosa, thus reducing the resistance against local spread of invasive cancer cells. Furthermore, the esophagus has an extensive network of lymphatics, allowing for early regional tumor advancement (Fig. 1). The end result is local spread and invasion into surrounding tissue, with early metastatic disease developing in most patients.
The clinical presentation of patients with esophageal cancer can be attributed to the direct effects of the local tumor, regional or distant complications of the disease, or paraneoplastic syndromes (Box 2). AC and SCC have similar clinical manifestations, which reflect the extent of local esophageal involvement. Dysphagia, the most common manifesting symptom, usually develops in response to dense solid food, and progresses gradually to interfere with the intake of softer foods and, finally, liquids. This can sometimes be accompanied by vomiting or regurgitation of saliva or food uncontaminated by gastric secretions, particularly in patients with advanced local disease. Pain is frequent and can occur in the absence of dysphagia. It can be related to swallowing itself (odynophagia) or to the local extension of the tumor into adjacent structures, such as the pleura, mediastinum, or vertebral bodies. Weight loss is common and correlates with dysphagia, dietary changes, and tumor-related anorexia. Weight loss is noted in more than 70% of patients and, if present, carries a worse prognosis. Other manifesting signs and symptoms reflect complications from disease spread, such as cough or fever from a respiratory tract fistula, upper or lower gastrointestinal bleeding, hoarseness from recurrent laryngeal nerve involvement, and hiccups from phrenic nerve involvement.
|Box 2 Manifesting Symptoms of Esophageal Cancer|
|Symptoms Caused by Local Tumor Effects|
|Cough and regurgitation|
|Upper gastrointestinal bleeding|
|Symptoms Related to Invasion of Surrounding Structures|
|Hoarseness from recurrent laryngeal nerve invasion|
|Hiccups from phrenic nerve invasion|
|Pain caused by local spread|
|Symptoms Related to Distant Disease|
|Metastatic disease to the lungs, liver, and central nervous system|
Symptoms related to distant metastasis in the lungs, bone, liver, and central nervous system, particularly in the case of AC, can also be found at the initial clinical presentation. Hypercalcemia is the most common paraneoplastic syndrome. In the absence of bone metastases, it is most common in patients with SCC and is believed to be caused by the production of a parathyroid hormone–related protein. The physical examination is often unremarkable, but should be directed toward finding evidence of metastatic disease, including supraclavicular lymphadenopathy, hepatosplenomegaly, and pleural effusion.
Dysphagia is an alarming symptom, and mandates the need for an immediate evaluation to define its exact cause and initiate appropriate therapy. Dysphagia in older adults should not be attributed to normal aging. Aging alone causes mild esophageal motility abnormalities, but these are rarely symptomatic. Evaluation of dysphagia starts with a barium swallow examination or an upper endoscopy. Endoscopy will allow the direct visualization of any tumor mass and histologic confirmation with a biopsy or brush cytology. Combining these techniques yields an overall diagnostic accuracy of 98%.
After establishing a diagnosis of esophageal cancer, adequate staging is required, because staging is the most important step in choosing appropriate therapy. More than 50% of patients have unresectable or metastatic disease at the time of presentation. For the others, survival is closely related to the stage of the disease.
The staging evaluation allows patients to be assigned a clinical stage according to the American Joint Committee on Cancer tumor-node-metastasis (TNM) classification (Box 3; Figure 2). Informed recommendations about therapy and appropriate information regarding prognosis depends on this clinical staging, an assessment that can, however, only approximate the true disease stage
|Box 3 American Joint Commission on Cancer (AJCC) Staging for Esophageal Cancer|
|Primary Tumor (T)|
|TX: Primary tumor cannot be assessed|
|T0: No evidence of primary tumor|
|Tis: Carcinoma in situ|
|T1: Tumor invades lamina propria (T1a) or submucosa (T1b)|
|T2: Tumor invades muscularis propria|
|T3: Tumor invades adventitia|
|T4: Tumor invades adjacent structures|
|Regional Lymph Nodes (N)|
|NX: Regional lymph nodes cannot be assessed|
|N0: No regional lymph node metastasis|
|N1: Regional lymph node metastasis|
|N1a: One to three nodes involved|
|N1b: Four to seven nodes involved|
|N1c: More than seven nodes involved|
|Distant Metastasis (M)|
|MX: Distant metastasis cannot be assessed|
|M0: No distant metastasis|
|M1: Distant metastasis|
|AJCC Stage Groupings|
|Tis, N0, M0|
|T1, N0, M0|
|T2, N0, M0|
|T3, N0, M0|
|T1, N1, M0|
|T2, N1, M0|
|T3, N1, M0|
|T4, any N, M0|
|Any T, any N, M1|
|Any T, any N, M1a|
|Any T, any N, M1b|
The pathologic extent of disease (pTNM) cannot truly be determined without an esophagectomy, and survival is closely linked to this pathologic stage (Fig. 3). The ability of the clinical (cTNM) stage to predict the pTNM accurately has improved with the development of more modern staging modalities.
Optimal clinical staging for this disease should include at least computed tomography (CT) scanning of the chest and abdomen, endoscopic ultrasonography and, if appropriate, a positron emission tomography (PET) scan. The importance of CT stems from its ability to assess the presence of metastatic disease (Table 1) and the extent of direct invasion of local structures, such as the aorta or major airways, any of which will preclude surgical intervention. The technique should use both oral and intravenous contrast media and should include cuts from the thoracic inlet down to the midabdomen. It must be noted, however, that CT is not very accurate in assessing the histologic depth of the tumor (T), nor is it sensitive in assigning lymph node status (N). In fact, the overall accuracy of CT in nodal detection is less than 60%.1
|Modality||Clinical Utility||Overall Accuracy (%)|
|Computed tomography (chest, abdomen)||Invasion of local structures (airways, aorta)||≥90%|
|Endoscopy||Local tumor (T) staging (operator dependent)||80%–90%|
|Ultrasonography (with or without fine-needle aspiration of lymph nodes)||Local nodal (N) staging (operator dependent)||70%–90%|
|Positron emission tomography||Metastatic disease||≥90%|
PET scanning with 18-fluorodeoxyglucose has been recently incorporated into the staging evaluation of esophageal cancer. This noninvasive test is more sensitive than CT for detecting distant metastases. Recent studies have suggested that PET scanning can detect metastatic disease in 15% of patients who were believed, on the basis of conventional diagnostic techniques, to have localized esophageal cancer.2 The superimposition of CT and PET scans is even more sensitive in identifying patients with occult metastases (Fig. 4).
Endoscopic ultrasonography has proven very useful in assessing the local depth of the tumor (T), lymph node involvement (N) and, with increased clinical experience, involvement of nonregional (M1a) lymph nodes. The overall accuracy of endoscopic ultrasonography in tumor depth assignment is about 80% and improves with more advanced stages of disease. With stringent criteria, the accuracy of detecting lymph node involvement approaches almost 75%. This accuracy can be improved further by endoscopic fine-needle aspiration of suspicious lymph nodes, which allows pathologic confirmation of involvement (Fig. 5).3
After the clinical disease stage is established, it also becomes important to assess the cardiopulmonary fitness and medical suitability of patients before surgical resection can be undertaken.
Historically, esophageal cancer has carried a dismal prognosis. This has been attributed to the late presentation of patients with this disease and the technical difficulty of an adequate surgical resection in the presence of advanced local and regional involvement. Furthermore, high-dose definitive radiation therapy, as an alternative to surgical resection, is challenging because of the anatomic location of the esophagus. Any radiation therapy portal encompassing the esophagus will also include other vital structures, such as major blood vessels, major airways, the heart, and lungs. Although modern radiation techniques have fewer adverse side effects, toxicity is still common with the radiation doses required. The frequent medical comorbidities and high incidence of second malignancies in these patients also affect the overall treatment success. Two historical reviews of the outcomes after esophageal cancer treatment, published in 1980, have demonstrated this clearly.4,5 The overall 5-year survival rate was 4% after surgical resection, with an unacceptable surgical mortality rate of almost 30%. The overall 5-year survival rate was only 6% after radiation therapy.
Surgical results have improved significantly over recent years, however. Multiple surgical series from major medical centers now report that patients undergoing surgery alone have median survival rates between 13 and 19 months, 2-year survival rates between 35% and 42%, and 5-year survival rates of 15% to 24%.6 Although these numbers are certainly more promising, they can hardly be characterized as a medical success story, especially when we keep in mind that much of this improvement is the result of better clinical staging and better patient selection.
The goal of oncologic surgery for this disease is the resection of the primary tumor and draining lymph nodes. Given the propensity for submucosal skip lesions in the esophagus, this usually requires a subtotal esophagectomy, using a transthoracic or transhiatal approach. The transhiatal approach requires a laparotomy, blunt dissection of the thoracic esophagus, and esophagogastric anastomosis in the neck. This approach saves the patient the cardiopulmonary complications of a thoracotomy, but no prospective clinical trials have demonstrated superiority of this procedure over a thoracoabdominal approach.7 In patients with tumors of the gastroesophageal junction and significant gastric involvement, a total gastrectomy and Roux-en-Y esophagojejunostomy may be required.
Radiation therapy has been used in the past as a single-modality approach with curative intent. However, except for those with very early-stage disease, radiation has had little impact on long-term survival.8 For more-advanced disease, single-modality radiation therapy should, in general, be considered a palliative intervention in patients whose underlying medical comorbidities preclude surgical resection or aggressive multimodality treatment.
Multimodality treatment approaches have evolved over recent years in response to the frequent locoregional and distant recurrences identified after surgery or radiation therapy alone. Several different combinations and sequences of treatment modalities have been tried, with mixed results.
Chemotherapy has been given preoperatively, postoperatively, or both. A cisplatin-based regimen is often used. Preoperative chemotherapy has been studied in several randomized clinical trials that compared surgery alone with chemotherapy followed by surgery. These studies have demonstrated that induction chemotherapy can produce up to a 50% clinical response rate but less than a 10% pathologic complete response rate and that the 2-year survival rate after subsequent surgery is approximately 35%. However, the results of these studies have been mixed, and no clear survival advantage has been identified with the induction regimens.9
More intensive multimodality approaches have attempted to exploit the radiosensitizing properties of chemotherapy by using concurrent cisplatin-based chemotherapy and radiation as definitive treatment or as a preoperative adjuvant. In 1992, Herskovic and colleagues8 reported a phase III prospective randomized trial that compared chemotherapy given concurrently with radiation with radiation therapy alone. Surgery was not a planned part of disease management in this trial. A clear survival benefit for the combined approach was identified, with a 5-year survival rate of 25% compared with radiation therapy alone, which produced no long-term survivors.8 How this approach can be integrated with surgery has remained unclear, however.
Three small phase 3 trials have randomized patients to surgery alone or to preoperative concurrent chemoradiation followed by surgery. Unfortunately, the results have been conflicting, with no consistent survival advantage identified for one approach or another. It is of note, however, that 25% to 30% of patients treated in this fashion achieved a pathologically complete response after concurrent chemoradiation therapy. For obvious reasons, this trimodality approach produces considerable toxicity, with some series reporting treatment-related mortality in excess of 12%.10 Currently, in many major referral centers in the United States, trimodality therapy is used for suitable patients who have at least T3 lesions, any nodal involvement with esophageal cancer, or both.
Patients with metastatic disease are treated with palliative intent. Palliation should first be directed toward relief of dysphagia and esophageal obstruction. This can be achieved in several ways, including palliative radiation, endoscopic dilation, endoscopic stenting, endoscopic laser therapy, or other light-based therapy (e.g., photodynamic therapy).11 Palliative chemotherapy has only a limited role in this setting and only a marginal impact on survival. Current efforts are directed toward developing targeted therapies that may prove more active in esophageal cancer.
More recently, there has been an effort to discover early, asymptomatic, esophageal adenocarcinoma by screening patients with Barrett’s esophagus. The identification of high-grade dysplasia (i.e., carcinoma in situ) is considered to be an indication for esophagectomy, because occult invasive cancer is frequently identified at the time of resection, and because invasive cancer will develop in almost 50% of patients with high-grade dysplasia who do not undergo esophageal resection.10