Esophageal Tumors

Esophageal Squamous Cell Carcinoma

ESCC is the most common form of esophageal cancer worldwide ( Fig. 48.1 ). Although it is no longer the most common form of esophageal carcinoma in Western societies, ESCC continues to be the most prevalent type of esophageal cancer in the East, representing 90% of all cancers in most Asian, African, and Eastern European countries. The areas of highest risk for ESCC are in 2 geographic belts, the Asian esophageal cancer belt across central Asia from the Caspian Sea to northern China and a belt on the eastern coast of Africa, from Ethiopia to South Africa (see Fig. 48.1 ).

Recent data on the trend of ESCC incidence in the USA between 1999 and 2008 indicate substantial regional variations. The data reviewed covered 85% of the U.S. population and found that the national age-standardized incidence rate for ESCC was 4.93 per 100,000 in men and 2.30 per 100,000 in women. The incidence was highest in patients older than age 65. Over the study period, the incidence of ESCC in the USA fell by 3.41% and 3.13% per year in men and women, respectively. The majority (70.8%) of esophageal cancers diagnosed during the study period were EAC rather than ESCC.

Although ESCC is more common in men than women, the ratio varies among low-risk areas like the USA (4:1) and high incidence areas of China and Iran where the ratio is lower, approaching or even exceeding 1:1. Environmental factors may be important in explaining the geographic variability in the incidence of ESCC. The etiologies and risk factors vary in low- and high-incidence areas. A population-attributable risk of 89% using cigarette smoking, alcoholic beverage consumption, and low consumption of fruits and vegetables was reported in a study in the USA. In contrast, a large cohort study conducted in China found that tobacco smoking played little role in ESCC risk.

In industrialized countries, the 2 most important risk factors are tobacco use and excess alcohol consumption ( Box 48.1 ). Furthermore, these 2 independent risk factors have a synergistic effect on cancer incidence. The risk of developing ESCC with active tobacco smoking increases 3- to 9-fold. Although the highest risk has been reported with smoking cigarettes, other forms of tobacco use such as pipe, cigar, hookah smoking, chewing tobacco, and the Asian betel quid have also been linked to ESCC. Exposure intensity and longer duration have been reported to be associated with the risk of ESCC. Tobacco-specific nitrosamines and polycyclic aromatic hydrocarbons are thought to be the major carcinogenic substances in tobacco. Alcohol use has been reported to have a slightly lower risk compared to tobacco, increasing risk of ESCC by 3- to 5-fold. The risk increases significantly with alcohol intake above the maximum recommended U.S. dietary guidelines of 140 g/week. Acetaldehyde, the first metabolite of ethanol metabolism, is a class I carcinogen.

In developing regions, factors such as nutritional deficiencies appear to have a stronger relation to the incidence of ESCC. Although the exact mechanism is unclear, low socioeconomic status per se is a risk factor for ESCC, even after adjusting for tobacco, alcohol, age, and many potential risk factors. Micronutrient deficiencies (such as vitamin A, C, and E) are also risk factors. These vitamins are thought to have important antioxidant properties, preventing formation of free radicals and nitrosamines. However, a 6-year randomized trial conducted in China with 20-year follow-up showed that multivitamin supplementation did not reduce risk in a sub-population of persons at high risk for ESCC. Other nutritional deficiencies, such as folic acid, zinc, and selenium, have also been reported as risk factors for ESCC. Decreased intake and impaired metabolism of folate, mostly due to gene polymorphisms, have been proposed as risk factors for several GI malignancies, including ESCC. In a case-control series, lower serum levels of folic acid were reported of patients with ESCC. The trace elements zinc and selenium are protective against ESCC. Zinc deficiency is thought to potentiate the carcinogenic effect of nitrosamines. Selenium supplementation is also thought to have chemo-preventive effects against ESCC. After a 10-year follow-up, a study showed that selenium supplements (along with β-carotene and vitamin E) reduced risk of esophageal cancer death by 17% among participants younger than 55 years old. A systematic review of predominantly case-control studies demonstrated that higher intake of fruits and vegetables probably decreases the risk of esophageal cancer with each increment of 50 g/day of raw vegetables and fruit associated with a 31% and 22% decrease in risk of esophageal cancer, respectively. Other dietary factors have also been theorized to increase the risk of ESCC. High temperature beverage drinking has been associated with increased risk of ESCC in multiple studies, likely due to the chronic thermal injury. Yerba mate, a hot herbal tea commonly consumed in South America, has been found to be a risk factor for ESCC. Other previously reported risk factors, such as ingestion of N -nitroso compounds (animal carcinogens found in cured meat products and pickles) and Fusarium verticillioides (a fungus found on maize) have mixed evidence as risk factors for developing ESCC.

Multiple studies have shown a possible association of ESCC and infection with HPV. Recently, a large international consortium performed complex serological analysis for multiple HPV serotypes in ESCC cases and control subjects and found very limited evidence for an association of HPV with ESCC. No definitive association can be proved at this time, and further investigation is necessary.

Some esophageal disorders increase the risk of ESCC. As discussed in Chapter 44 , it is estimated that the prevalence of ESCC among patients with achalasia is approximately 5%. Between years 1 and 24 after diagnosis of achalasia, the risk is increased 16-fold, especially in men. In fact, a large achalasia cohort from Sweden was shown to have an increased risk for both ESCC and EAC. The mechanism is likely due to stasis of food material in the esophagus, leading to chronic inflammation. Similarly, as discussed in Chapter 28 , patients with esophageal strictures from lye ingestion have an increased risk of ESCC occurring decades after the initial ingestion. Patients with esophageal webs, such as those with Plummer-Vinson syndrome (iron deficiency anemia, dysphagia, and post-cricoid webs) and with Fanconi anemia (an inherited bone marrow failure linked to several cancers) are also at increased risk for ESCC (see Chapter 43 ). ^,

Another rare condition associated with ESCC is tylosis, an autosomal-dominant disorder characterized by hyperkeratosis of the palms/soles and leukoplakia (see Chapter 25 ). The tylosis esophageal cancer gene TOC (RHBDF2) has been localized to chromosome 17q25. Recent reports have suggested that this gene may be epigenetically silenced rather than mutated. In one report, the risk of developing ESCC in those with this gene was 95% by the age of 65 years.

Certain factors are protective for ESCC. These include obesity and frequent use of aspirin and NSAIDs.

Esophageal Adenocarcinoma

EAC is the second most common form of esophageal carcinoma worldwide (see Fig. 48.1 ). The prevalence and incidence of EAC has not only changed over time, but geographically as well. ESCC used to be the most common type of esophageal cancer in both Western and Eastern countries. Based on surgical series, EAC of the esophagus was uncommon prior to 1978. In Western countries, its incidence has markedly increased such that in 1994, EAC surpassed ESCC in prevalence in the USA. Today, EAC is the predominant type of esophageal cancer in the West. In the USA, its incidence among white men rose from 0.5 to 0.9/100,000 in the 1970s to 3.2 to 4.0/100,000 over the ensuing 2 decades. This shift represents a 463% and 335% increase among white men and white women, respectively. More recent population-based studies carried out from 2003 to 2007 estimated the incidence of EAC to be 5.3/100,000. EAC was 8 times more common in men than women, and 5 times more common in Caucasians than African Americans. Furthermore, this rapid increase in the incidence of EAC is not isolated to the USA; a similar trend has been noted in other Western countries such as the United Kingdom, Iceland, and Australia. Furthermore, based on mathematical predictive models, the incidence of EAC is expected to continue to rise till 2030, at which point the predicted incidence of EAC will be 8.4 to 10.1 per 100,000 person-years. The overall 5-year observed survival rates for EAC is strongly dependent on the stage at diagnosis and is among the lowest for all cancers. A recent analysis using the USA National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) registry showed that the proportion of EAC that had localized, regional, or distant stage disease remained relatively stable (40% of EAC patients diagnosed with distant stage disease). Recent data also suggest an overall improvement in 5-year survival rates with the greatest improvement in patients with localized disease reflecting improvements in curative treatment modalities. The relationships between EAC of the distal esophagus and adenocarcinomas of the esophagogastric junction and the gastric cardia are discussed in Chapters 47 and 54

This geographic change in EAC incidence over time is thought to be related to the significant increase in GERD (see Chapter 46 ) and, more importantly, obesity in Western societies. The epidemiologic gradient across both Western and Eastern countries may be related, in part, to socioeconomic status, because high socioeconomic status is more commonly seen in EAC, and low socioeconomic status is more commonly seen in ESCC. However, other indicators of improved socioeconomic status including per capita gross domestic product and level of education have not correlated with the observed esophageal cancer gradient in different parts of the world.

Because the rise in the incidence of EAC has been so dramatic, the possibilities of misclassification and/or over-diagnosis have been raised. A population-based study using data from SEER 9, which represented approximately 10% of the U.S. population between 1973 and 2001, found that reclassification of EAC to ESCC and adjacent gastric cardia cancer was unlikely to change the incidence of EAC. The anatomic distribution of esophageal cancer has shifted from the upper third of the esophagus to the lower third. The lower third of the esophagus, the location where adenocarcinoma usually arises, was the only esophageal location with an increased incidence. Because there had been minimal change in the rates of in situ or localized disease and the fact that the mortality of EAC had also increased significantly (7-fold, from 2 to 15 deaths/million), over-diagnosis was unlikely to explain the marked increase in EAC incidence. Furthermore, the possibility of heightened surveillance with EGD as a potential reason for the observed dramatic increase in the incidence of EAC was negated by showing that the rates of EAC increased in white men and women in all stage and age groups.

In general, risk factors affecting EAC incidence are commonly seen in economically developed countries. Similar to ESCC, tobacco is a risk factor for EAC (see Box 48.1 ). Pooled data from the international Barrett and Esophageal Adenocarcinoma Consortium (BEACON) confirmed association between smoking and EAC (OR 1.96, 95% CI 1.64 to 2.34). A dose-response relationship was also seen; heavy smokers had the highest risk of cancer. Furthermore, smoking cessation has been shown to be associated with reduced risk of cancer (smoking cessation ≥10 years associated with a 30% lower risk of EAC). In contrast to ESCC, alcohol does not seem to be as strong a risk factor for EAC or Barrett esophagus (BE) as it is for ESCC. Pooled analysis of 11 studies from BEACON showed no association between alcohol consumption and risk of EAC.70 GERD is the most important risk factor for the development of EAC (see Chapters 46 and 47 ). Although most of the large studies are in agreement over the increased risk, the reported risk has varied from 4- to 8-fold. A study in Sweden showed that the risk of EAC in individuals with recurrent symptoms of GERD (at least weekly heartburn and/or regurgitation symptoms) was 8-fold higher than individuals with less frequent symptoms (OR 7.7, 95% CI 5.3 to 11.4). This study also showed that the risk of EAC was especially high among individuals with more severe and longer-lasting GERD symptoms (OR 43.5, 95% CI 18.3 to 103.5). A recent pooled analysis of individual-level data from 5 large population-based case-control studies in BEACON reported a strong dose-dependent relationship between frequency of GERD symptoms and EAC (at least weekly: OR 4.81, 95% CI 3.4 to 6.8; daily symptoms: OR 7.96, 95% CI 4.51 to 14). Although GERD is a well-established risk factor for EAC, most patients with GERD never develop EAC. In fact, the majority of patients with EAC deny any substantial prior symptoms of GERD. It is believed that, in predisposed individuals, GERD leads to distal esophageal injury resulting in an aberrant healing process resulting in BE. BE is the only identifiable premalignant condition for EAC and is defined by the replacement of the normal stratified squamous epithelium with a columnar-lined distal esophagus with intestinal metaplasia (see Chapter 47 ). BE is seen in ∼7% to 15% of individuals with GERD and is estimated to be present in 1% to 2% of the general adult population. BE patients have a 10- to 55-fold higher risk of EAC; however, progression to EAC from BE has been reported in a small percentage of patients (0.12% to 0.3% per year). Similarly, obesity is another strong risk factor for EAC, and this risk increases with higher BMI. The increased risk of EAC in obesity has been estimated to be 2- to 5-fold. In addition, abdominal obesity (independent of the BMI) has been shown to be associated with an increased risk of EAC. A high intake of dietary calories and fat, by itself, is also a risk factor for EAC. Obesity may increase the risk of hiatal hernia and GERD via increased intra-abdominal pressure. However, abdominal obesity is associated with BE even after adjusting for GERD, and obesity is associated with EAC even in the absence of GERD symptoms. In addition to its mechanical effect, abdominal obesity is also associated with alterations in circulating levels of peptides that are associated with BE and EAC. Metabolic syndrome has been associated with BE and EAC. In addition, studies have also found associations between insulin-like growth factor, leptin, and adiponectin with progression of BE to EAC.

The striking male predominance in esophageal cancer has raised the question as to whether sex hormones might play a role. When male-to-female ratios of EAC and ESCC, diagnosed between 1992 and 2006, were calculated, the highest male-to-female rate ratio was seen among Hispanics (20.5), followed by Caucasians (10.8) and African Americans (7.0). In contrast, the male-to-female incidence rate ratios in ESCC were much lower. These findings supported the hypothesis that female sex hormone exposure might play a protective role in the development of EAC. The differences in the risk of EAC among men may also be related to differences in use of tobacco or types of obesity.

Several factors have been reported to have a protective role against developing EAC. The prevalence of Helicobacter pylori infection (a known risk factor for gastric cancer) has been reported to be inversely associated with EAC. This risk reduction has been estimated at 50%. The infection appears to decrease EAC risk by reducing gastric acidity. Also, as with ESCC, aspirin and NSAIDs use appears to be protective in EAC. Observational studies have also shown that the use of PPIs reduces the risk of neoplastic progression among BE patients by 71% as shown in a systematic review and meta-analysis. Similarly, statins have been shown to reduce the risk of EAC, especially in BE patients. Finally, an inverse association between fruit and vegetable consumption and EAC has been shown.

Pathogenesis

Human cancers occur in a multistep manner as a result of an accumulation of epigenetic changes and genetic alterations (see Chapter 1 ). A combination of environmental factors, hereditary factors, and acquired genetic alterations are likely to be important risk factors in the development of esophageal cancer. In EAC, there is a gradual accumulation of somatic-cell genetic abnormalities that occur during the metaplasia-dysplasia-carcinoma sequence in the esophageal epithelium (see Chapters 1 and 47 Chapter 1Chapter 47). The inflammatory microenvironment and somatic genomic alterations in stem cell populations are believed to mediate progression from BE to EAC. An understanding of the sequence of these abnormalities may lead to a more accurate stratification of patients according to their individual cancer risk. Because there is some evidence to suggest that only patients with a complete pathologic response to neoadjuvant therapy have a survival benefit (discussed later), a search is underway for prognostic and predictive genetic markers that would tailor a more efficacious multimodal therapy. ESCC, on the other hand, is thought to arise from hyperproliferative epithelium that progresses to low-, intermediate-, and high-grade dysplasia (HGD) leading, ultimately, to invasive cancer. Both concurrent genetic changes and epigenetic modifications, in the form of hypermethylation of tumor suppressor genes, occur frequently in both EAC and ESSC.

Although most cases of BE and EAC are sporadic without a definite family history, there are several findings that suggest a genetic component. These findings include familial clustering, the fact that only a subset of patients with GERD develop BE, and that the amount of reflux is not an accurate predictor of BE development. In these cases, the risk of developing BE is strongly influenced by gene-gene and gene-environment interactions (see Chapter 47 ).

Based on the description of the 6 essential components in human carcinogenesis, molecular factors have been described for each of these 6 steps in esophageal cancer, as summarized below.

• 1.

  Self-sufficiency in growth signals. Cancer cells can either make their own growth factors (autocrine effect; see Chapter 4 ) or alter their growth factor receptors and signaling pathways to free themselves from exogenous growth-limiting signals. For example, expression of the gene for EGF receptor 2 ( HER2/neu ) is a prognostic factor in esophageal cancer. HER2/neu gene amplification correlates with shortened patient survival and independently predicts poor outcomes in patients with EAC. Conversely, low expression of HER2/neu is associated with an improved response to neoadjuvant chemoradiotherapy than tumors with high HER2/neu expression. Similar effects have been shown for ESCC. Furthermore, in ESCC, expression of cyclin D1, a key cell cycle regulator (see Chapter 1 ), has been associated with shorter patient survival compared to cyclin D1-negative patients.

• 2.

  Insensitivity to antigrowth signals . Inactivation of tumor suppressor genes is an important mechanism by which tumor cells become desensitized to antigrowth signals. This may happen by mutation, loss of heterozygosity, or promoter hypermethylation. Expression of the tumor-suppressor gene TP53 in ESCC is an independent prognostic factor. Tumors with low p53 staining are associated with significantly longer survival than tumors with high p53 protein expression. Also, p21 staining in ESCC has been associated with an improved survival, and ESCC patients whose tumors had high levels of p16 had longer survival.

• 3.

  Avoidance of apoptosis. Apoptosis (programmed cell death) puts a brake on expansion of the cell pool. A tumor’s capability to expand is determined not only by its rate of cell proliferation but its avoidance of apoptosis. Some important regulators of apoptosis include Bax, Bcl-2, and Bcl-X. Expression of these proteins, alone or in combination, correlates with prognosis and response to neoadjuvant chemoradiation. Survivin, another member of the inhibitor of apoptosis gene family, has been found to be a useful predictive factor in neoadjuvant therapy for esophageal cancer. Specifically, partial responders to neoadjuvant chemotherapy have lower survivin expression than nonresponders. Thus, future therapeutic strategies to reduce survivin expression or block survivin signaling pathways may increase histopathologic response rates and prognosis.

• 4.

  Uncontrolled replicative potential . Malignant cells, by an overexpression of telomerase, destabilize mechanisms that limit their proliferative capacity and they thus become resistant to cellular aging and death. Although there seems to be an increase in the expression of the human telomerase reverse transcriptase catalytic subunit in the pathogenesis of EAC, there are no studies showing any prognostic significance of the finding.

• 5.

  Sustained angiogenesis. Sustained angiogenesis is crucial for the development, progression, and eventual metastasis of cancer. Angiogenesis factors such as vascular endothelial growth factor (VEGF), COX-2, and fibroblast growth factor receptor 1 have been suggested as potential prognostic factors in esophageal cancer. Of these, VEGF seems to be the most important. High expression of VEGF is an independent, negative prognostic factor in ESCC, although a correlation in EAC has not been demonstrated. COX-2 is known to increase progressively as the tissue progresses through Barrett metaplasia, to dysplasia, and to frank EAC. In ESCC, COX-2 overexpression correlates with depth of tumor invasion, tumor stage, and reduced survival. Amplification of fibroblast growth factor receptor 1 was shown to be an independent adverse prognostic factor in ESCC. Expression of thymidine phosphorylase, another angiogenic factor, predicts an only partial response to chemoradiation in patients with ESCC.

• 6.

  Invasion and metastasis. Abnormalities in cell-to-cell adhesion molecules (i.e., E-cadherin glycoproteins, catenins, and matrix metalloproteins) and their inhibitors are associated with poor histologic differentiation and greater tumor invasion and nodal metastasis.

Clinical Features

Patients with EAC and ESCC have a similar clinical presentation despite the differences in demographics and risk factors. In the early stages, most patients are asymptomatic. However, as the disease progresses, progressive dysphagia and weight loss are the most common symptoms. The diagnosis is often delayed because patients experiencing dysphagia tend to avoid the foods causing the symptom and adjust their dietary intake. Dysphagia is initially with solids but progresses to liquids in the later stages of the disease. Solid food dysphagia typically occurs with an esophageal luminal diameter of 13 mm or less. The severity of dysphagia and concomitant weight loss from decreased oral intake is proportional to the degree of luminal obstruction. The point of difficulty with swallowing as localized by the patient is a poor predictor of the actual location of the mass. Odynophagia is a less common symptom and usually indicates the presence of an ulcerated lesion.

Other less common clinical presentations include iron deficiency anemia, palpable cervical lymphadenopathy, and/or chest pain. Chest pain, often radiating to the back, suggests the possibility of invasion into peri-esophageal structures. Tumor erosion can lead to an esophageal-respiratory fistula, which can present as refractory cough, recurrent pneumonia, or pleural effusions. A rare complication is esophageal-aortic fistula, which can cause massive upper GI hemorrhage and exsanguination. Hoarseness is another rare presentation due to recurrent laryngeal nerve injury from the tumor per se or associated lymphadenopathy. Metastatic lesions can be found not only in lymph nodes, but also in lungs, liver, brain, and bone.

Diagnosis

Laboratory tests are nonspecific and can reveal an anemia (iron deficiency or chronic disease type), hypoalbuminemia, and/or hypercalcemia (usually associated with osteolytic metastasis). Although paraneoplastic syndromes are rare with esophageal cancer, ESCC rarely can cause hypercalcemia due to tumoral production of a circulating parathyroid hormone-related protein. There are no specific serologic markers for esophageal cancer.

The diagnosis of esophageal cancer is primarily made by endoscopic biopsies in a patient presenting with progressive dysphagia to solids ( Fig. 48.2 ). The endoscopic appearance is similar between advanced ESCC and EAC; however, approximately three quarters of all EAC lesions are found in the distal esophagus whereas ESCC is more frequent in the proximal to middle esophagus. Esophageal cancer can appear as a mass, raised nodule, ulceration, depression, stricture, or subtle irregularity in the mucosa. It is critical for endoscopists to spend adequate time inspecting the esophagus and document landmarks such as the gastroesophageal junction, extent of BE using the Prague C (circumferential) and M (maximal extent) criteria, the presence or absence of extension into the stomach, and the exact location of the tumor. These data points determine the surgical technique used in the management of patients with esophageal cancer. Clear description of morphologic features of tumors may also help in determining candidacy for endoscopic eradication therapies, therapies that have become standard of care in the management of early esophageal cancer (ESCC and EAC).

Various other imaging modalities can aid in diagnosis. Routine chest radiography can reveal nonspecific findings such as aspiration pneumonia, a dilated esophagus with an air-fluid level (pseudo-achalasia), metastatic lesions in the lung parenchyma, pleural effusions, or signs of fistulas. Barium esophagogram can be helpful in diagnosis of esophageal cancer. A sign of early cancer with this modality is an abnormal esophageal mucosal lining, which can represent a plaque, polypoid lesion, ulceration, or nonspecific focal irregularity. Advanced tumors may be seen as overt masses, strictures with distinct shoulders, or luminal narrowing. Although it generally has fallen out of favor owing to widespread availability of endoscopy, barium esophagogram can be very useful prior to endoscopy in suspected esophageal-respiratory fistulas. When used for this concern, the endoscopist can have a “roadmap” of the anatomy prior to endoscopic stenting. Specific care should be taken to use barium as a contrast agent as opposed to hyperosmolar agents (diatrizoate meglumine and diatrizoate sodium), which carry a risk of severe pulmonary edema and pneumonitis with aspiration.

CT may demonstrate esophageal wall thickening/irregularity, focal esophageal stricture with proximal dilation, or an intraluminal mass may be seen. Signs of aspiration pneumonia, metastatic lesions, lymphadenopathy, and esophageal-respiratory fistula may be seen.

As stated, endoscopy with biopsy has the highest yield for diagnosis of esophageal cancer and is the standard for diagnosis ( Figs. 48.3 and 48.4 ). Several imaging modalities have been used in the context of 3 major clinical applications: (a) improved detection and identification of patients with early cancer during screening and surveillance endoscopy; (b) prediction of histology and real-time diagnosis during endoscopy; and (c) guiding endoscopic eradication therapies. Technologic advancements have allowed the creation of smaller charge-couple device chips that are capable of producing images with high resolution (over 850,000 to 2.1 million pixels) displayed on monitors with a 16:9 aspect ratio resulting in superior imaging quality compared to standard definition white light endoscopy (WLE). High-resolution magnification endoscopes can optically magnify images up to 150 times (although not routinely available in the USA). High-resolution endoscopy, with or without magnification endoscopes, has been shown in several studies to increase the yield for detection of dysplasia and early cancer. White-light endoscopy overall has a low sensitivity rate of 55% to 63% in the detection of early ESCC. The use of high-resolution WLE, where available, should be the minimum standard for evaluation of BE patients undergoing surveillance or being considered for endoscopic eradication therapy (EET) and for other subtle lesions in the esophagus (esophageal squamous dysplasia and early ESCC), a practice endorsed by GI society guidelines. Endoscopic findings can vary from relatively normal-appearing mucosa (submucosal infiltrative pattern), to ulcers, nodules, and overt masses. Several other endoscopic techniques can also help identify areas of dysplasia and malignancy, including narrow band imaging, chromoendoscopy (conventional or electronic), autofluorescence imaging (AFI), and confocal laser endomicroscopy (CLE). The chapter on Barrett esophagus discusses these imaging modalities in further detail (see Chapter 47 ).

Conventional chromoendoscopy involves the use of special stains to highlight subtle architectural changes to help direct biopsies and predict histology. Lugol’s iodine, methylene blue, acetic acid, crystal violet, and indigo carmine are the most commonly used stains. Lugol’s iodine solution consists of a 0.5% to 3.0% aqueous solution of potassium iodide. Iodine stains glycogen-containing cells of the normal esophageal epithelium and is not taken up by dysplastic or malignant cells that are glycogen depleted (“pink color sign”). Chromoendoscopy with Lugol’s iodine staining has become the standard of care for screening of ESCC in high-risk populations and has been shown to have a high sensitivity rate of 89% to 100% with highly variable specificity rates due to false positive lesions. In contrast, methylene blue, acetic acid, and indigo carmine staining are more useful in the detection of glandular abnormalities, as seen in EAC. These stains are sprayed in the esophagus with the intent of improving characterization of the mucosa resulting in selective uptake (vital staining—methylene blue) or enhancement of mucosal surface pattern (contrast staining—indigo carmine, acetic acid). Of importance, there are many limitations of conventional chromoendoscopy including: (a) dysplasia and inflammation (esophagitis) are not distinguishable from each other; (b) these techniques are generally cumbersome, time-consuming, and require dye spraying equipment; (c) difficulty in achieving complete and uniform coating of the mucosal surface with the dye; (d) inability to detect vascular patterns; (e) conflicting published data; (f) the need for magnification endoscopy; and (g) lack of standardized classification. Conventional chromoendoscopy has largely been replaced by optical chromoendoscopy (see below) while evaluating patients with Barrett esophagus and EAC.

Optical chromoendoscopy is another modality to detect signs of dysplasia and cancer by using selective light filters to highlight subtle architectural and vascular changes in the mucosa. This method avoids some of the concerns associated with conventional chromoendoscopy highlighted above. Several variations of electronic chromoendoscopy are available by different manufacturers, but most of the published studies have described the use of narrow band imaging (NBI; Olympus, Tokyo, Japan) in the setting of BE. NBI is an imaging technique that is based on the optical phenomenon that the depth of light penetration into tissue depends on the wavelength; the shorter the wavelength, the more superficial the penetration. Use of blue light with narrow band filters enables detailed imaging of the mucosal and vascular surface patterns with a high level of resolution and contrast without the need for dye chromoendoscopy. NBI is the most widely studied electronic chromoendoscopy technique to predict histology during surveillance, improve detection of dysplasia, and guide endoscopic eradication therapies. An international working group of experts recently established a classification system for BE using NBI. A systematic review and meta-analysis showed that advanced imaging techniques (conventional and optical chromoendoscopy) increased dysplasia or cancer detection by 34% (95% CI 20% to 56%).

NBI without magnification has been evaluated in the detection of ESCC and squamous dysplasia. Endoscopically suspicious lesions for early ESCC appear as well-demarcated brown areas at NBI without magnification ( Fig. 48.5 ). A prospective comparative study showed that NBI examination was reliable for the detection of early ESCC with sensitivity of 88% and specificity of 75%.

AFI involves a technique that uses short-wavelength blue or ultraviolet light to stimulate biological fluorophores (e.g., collagen, porphyrins, flavins, aromatic amino acids) in the esophagus. This technology has been mostly used in clinical trials for detecting BE-associated neoplasia. Pooled analysis from 5 prospective trials showed that AFI had little effect on the diagnosis of neoplasia with additional diagnostic value of 2% compared to WLE and with a high false positive rate (78%). Given this limited additional diagnostic and therapeutic value, AFI is not widely used clinically.

CLE allows real-time, in vivo microscopic imaging of the esophageal mucosa. It involves IV administration of a fluorescent dye (most commonly fluorescein sodium) that is taken up by normal mucosal cells. Fluorescent dye uptake is not seen in dysplastic cells, and thus they appear dark. This method creates magnification up to 1000 times. The only commercially available method currently for CLE is a probe-based instrument that can be passed through an accessory channel on the standard upper endoscope. Several prospective randomized controlled studies have reported that adding CLE to high-definition WLE significantly improves the diagnostic yield for BE-associated neoplasia.

Optical coherence tomography emits near-infrared light to provide cross-sectional images of tissue, which also has the potential advantage of identifying submucosal lesions. Initial studies showed encouraging results for detecting dysplasia, but limitation for its use included small field of view and slow imaging processing.

Volumetric laser endomicroscopy (VLE) is a second-generation optical coherence tomography technique that produces real-time cross sectional images of large areas in the esophagus. NvisionVLE (nVLE; NinePoint Medical Inc., Bedford, MA, USA) is approved by the Food and Drug Administration and is currently commercially available. This device has the ability of laser marking of suspicious areas for subsequent treatment. Using a diagnostic algorithm in a prospective cohort, VLE was shown to have a sensitivity of 86% (95% CI 69 to 96), specificity of 88% (95% CI 60 to 99), and diagnostic accuracy of 87% (95% CI 86 to 88) to detect BE-associated dysplasia. This new technology is still not widely available.

Screening and Surveillance

EAC screening and surveillance are covered under Barrett esophagus (see Chapter 47 ). The histologic precursors of ESCC include dysplasia and carcinoma in situ ( Fig. 48.6 ). Given the time needed for dysplasia to develop into malignancy, screening and surveillance programs are designed to detect these early histologic precursors in an effort to improve survival. Early-stage carcinoma is associated with a substantially improved survival (up to 86% at 5 years) when treated surgically.

The primary screening strategy in areas of high ESCC incidence involve Lugol’s chromoendoscopy, with endoscopic biopsies obtained from the unstained areas. Vital staining with Lugol’s is based on the reversible chemical reaction between the iodine and glycogen present in normal epithelial cells. Dysplastic, cancerous, and inflammatory cells are usually devoid of stain (because of less abundant glycogen) and targeted biopsies can be obtained to confirm dysplasia/malignancy. As mentioned above, Lugol’s chromoendoscopy has a sensitivity and specificity of >90%. Some studies showed low specificity of Lugol’s solution to detect early ESCC that is likely secondary to difficulty differentiating between dysplasia/neoplasia and inflammation, but targeted biopsies should be helpful in this situ ation. Vital staining with Lugol’s solution, although simple and inexpensive, is associated with some limitations, including possible iodine allergy, risk of aspiration, or chest pain due to esophageal spasm.

NBI allows for clearer depiction of capillary networks and superficial mucosal patterns. Detailed classifications of these abnormalities have been proposed and validated in the identification and staging of squamous dysplasia and early ESCC (see Fig. 48.5 ). A recent meta-analysis of 18 studies that included >1900 patients showed that NBI had a significantly improved specificity compared with chromoendoscopy with no difference in sensitivity.

Cost-benefit studies have been carried out to assess esophageal cancer screening strategies. One study found that a strategy of 1-time screening at age 50 would be the best approach in underdeveloped high-risk areas, whereas a 3-time screening strategy starting at age 40 (10-year intervals) would be preferable in areas with better health care resources.

A community assignment study in China with 10 years of follow-up showed a one third reduction in the cumulative ESCC-related mortality in communities where 40- to 69-year-old adults were screened once by Lugol’s chromoendoscopy compared with communities without screening.

Routine screening for ESCC is also recommended in patients with head and neck cancer. A retrospective study from Taiwan found a higher prevalence of secondary ESCC in head-and-neck cancer patients receiving routine endoscopies compared with the non-screening group (4.5% vs. 3%, P = 0.04), with earlier stage at diagnosis ( P = 0.03).

The current screening guidelines include :

• 1.

  One-time Lugol’s chromoendoscopy for high-risk Asian and African populations beginning at the age of 40.

• 2.

  Endoscopy with Lugol’s or NBI every 6 months to 1 year after completion of therapy for head and neck squamous cell cancer, for 10 years.

• 3.

  Screening could also be considered for patients at high risk (tylosis, achalasia, and caustic injury).

Other endoscopic techniques covered earlier, such as NBI and AFI, have also been evaluated as screening tools. NBI has been shown in multiple studies to detect early esophageal lesions based on superficial vasculature and surface pattern with excellent sensitivity and specificity but is still used in conjunction with Lugol’s chromoendoscopy. AFI is not sufficiently sensitive for lesions smaller than 10 mm, rendering it a less useful test for screening.

Unsedated transnasal endoscopy (TNE) using an ultrathin endoscope has been shown to be safe and well tolerated. A study of high-risk patients showed that TNE (with or without optical chromoendoscopy) has excellent specificity and sensitivity for early ESCC. TNE is still not widely available but could be an important tool for ESCC screening.

Additional endoscopic techniques include endocytoscopy and high-resolution microendoscopy which allow for real-time pathological examination of the tissue. Limited initial data are promising, but these techniques are not widely available.

Non-endoscopic methods for screening such as the use of per-oral balloons and sponges have the advantage of being low cost and less invasive compared to endoscopy. However, these techniques have been shown to be inadequate owing to the poor sensitivities for dysplasia and even invasive carcinoma. Combining non-endoscopic cytologic samples with biomarker (such as P53) may improve the sensitivity of the test. Other non-invasive screening methods under investigation include breath markers (volatile organic compounds), antibodies to tumor-associated antigens, circulating microRNAs, and methylated DNA markers.

Staging

Staging of esophageal cancer is carried out using the American Joint Committee on Cancer staging system ( Fig. 48.7 ), which was last updated in 2017 (8th edition). This system not only separates staging for EAC and ESSC but also presents 3 classifications systems separately for EAC and ESCC: the classic reference pathologic (pTNM) stage groups ( Fig. 48.8 ), the newly introduced post-neoadjuvant pathologic (ypTNM) stage groups ( Fig. 48.9 ), and clinical (cTNM) stage groups ( Fig. 48.10 ). This staging system recently revised the definition of the esophagogastric junction such that cancers involving it with epicenters no more than 2 cm into the gastric cardia are staged as EAC and those with >2 cm involvement of the gastric cardia are staged as gastric cancers.

The depth of tumor invasion (T stage) is an important factor because the rich lymphatic supply of the esophagus can provide a route of metastasis. Superficial cancers are defined as either carcinoma in situ (Tis) or T1 tumors. T1 is divided into T1a and T1b depending on whether the submucosa is spared or involved, respectively. Further classification for T1a includes M1 (intraepithelial cancer), M2 (invasion into the lamina propria), and M3 (invasion to the muscularis mucosa). T1b lesions also can be subdivided into SM1 (invasion into the upper third of the submucosa), SM2 (invasion into the middle third), and SM3 (invasion into the lower third). Tis and T1a lesions have a predicted lymph node metastasis rate up to 8% compared to T1b lesions, which have up to a 56% lymph node metastasis rate. In addition to the mediastinal nodes, lymph node metastases can occur in the neck and upper abdomen. The risk of lymph node involvement is related to several factors (in decreasing order of frequency): grade III histology, SM3 invasion, lymphatic invasion, vascular invasion, SM2 invasion, and SM1 invasion. In ESCC, the best predictors of lymph node invasion are SM3 invasion and vascular invasion, whereas in EAC, the most important predictor is lymphatic invasion. Figure 48.11 shows correlation of survival and the T stage.

Staging can be performed by several methods: endoscopy with mucosal biopsies, endoscopic resection, multidetector CT (MDCT) with 18F-fluoro-2-deoxy- d -glucose PET (18F-FDG-PET), and EUS with FNA for cytology.

Multidetector CT and 18F-FDG-PET

MDCT and 18F-FDG-PET scans are commonly used in the staging of esophageal cancer ( Fig. 48.12 ). Both ESCC and EAC have high affinity for 18F-FDG which makes PET scan very helpful in esophageal cancer evaluation. The sensitivity and specificity for 18F-FDG-PET in detecting distant metastases in esophageal cancer are 71% (95% CI 62 to 79) and 93% (95% CI 89 to 97), respectively, whereas for CT they are 52% (95% CI 33 to 72) and 91% (95% CI 86 to 96), respectively. Studies have shown that CT-PET can identify 5% to 28% with metastatic disease especially at sites not detected on CT alone. In addition, restaging after neoadjuvant therapy seems to be better accomplished with FDG-PET/CT than EUS-FNA or MDCT alone.

Treatment

A multidisciplinary approach to the treatment of esophageal cancer is essential and requires input from experts in surgical oncology, radiation oncology, medical oncology, gastroenterology, radiology, pathology, and often palliative care. Tumor location, staging, histologic type, medical comorbidities, and patient preference are factors that must be considered for selecting the proper treatment. Some general principles can be summarized as follows :

• Surgery is the standard treatment for a medically optimized surgical candidate with a localized, non-superficial tumor.

• For a patient with a localized tumor who is not a surgical candidate, definitive chemoradiation with curative intent may be considered.

• For all others (metastatic disease), palliation is recommended.

With a wide spectrum of treatment options for esophageal cancer, accurate staging is essential to selecting the appropriate treatment modality. The primary objective is to identify those patients who may benefit from neoadjuvant therapy and those with widespread metastatic disease who are better candidates for palliation. Patient selection is another very important component of the management of esophageal cancer. The evaluation should include medical comorbidities and the patient’s current performance status, nutrition, and cardiopulmonary function. Pulmonary complications, pneumonia in particular, are important determinants of early postoperative outcome and are associated with more than 4-fold increase in mortality. Interestingly, preoperative chemoradiation has been found to be a risk factor for postoperative pneumonia. A scoring system involving spirometry results, age, and performance status is available to predict cardiopulmonary complications. Selection based on mathematical scores may not be superior to a physician’s clinical assessment; the 2 methods are complementary. Importantly, the scores obtained may not be applicable to all institutions or populations. For example, patients with ESCC from Asia have a different risk profile (e.g., pulmonary and hepatic comorbidities from smoking and alcohol use) compared with patients in Western societies with EAC (e.g., cardiac risk factors or comorbidities).

Patients with cervical or cervicothoracic esophageal tumors (<5 cm from the cricopharyngeus) are usually poor candidates for surgery because of limitations in surgical techniques. These patients are typically treated with definitive chemoradiation, with an encouraging response rate in a recent study.