Endoscopic Treatment of Early Esophageal Neoplasia

Introduction

Esophageal cancer remains the eighth most common cancer and the sixth leading cause of cancer-related mortality worldwide. Two main subtypes exist: esophageal adenocarcinoma (EAC) and esophageal squamous cell carcinoma (ESCC). EAC predominates in North America and Europe, and there has been a steep rise in its incidence over the last several decades. ESCC is most prevalent in Asia and Africa, and accounts for over 90% of esophageal cancers worldwide. The prognosis for esophageal cancer is poor, with an overall 5-year survival of less than 20%. This stems in large part from challenges related to early detection, as the majority of cancers present at advanced stages (T3 and T4) for which effective treatments do not yet exist. However, when esophageal cancer is encountered early (i.e., stage Tis or T1a), survival can be dramatically improved with therapy. Until recently, the only curative treatment option was esophagectomy, which carries substantial risk and may not be advised in patients with significant comorbidities. Improvements in endoscopic resection techniques, namely endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD), devised to procure larger amounts of tissue for accurate staging, have also enabled endoscopists to curatively treat early esophageal cancers. Moreover, residual mucosa that could become neoplastic is treated with ablative therapies that can be applied with less morbidity than previously. Endoscopic methods to treat cancer have become an important tool for the gastroenterologist, and their mechanisms, techniques, and evidence are reviewed in this chapter. Screening and surveillance for Barrett's esophagus is covered in Chapter 25 .

ESCC is believed to follow a stepwise process of neoplastic progression. Established risk factors include alcohol, tobacco, exposure to environmental toxins such as polycyclic aromatic hydrocarbons, and potentially nutritional deficiencies. Recent studies (2015) suggest a cost benefit for screening endoscopy or balloon brush cytology in areas of high incidence such as northern and rural areas of China. At this time, however, there are no widely accepted guidelines for ESCC screening.

Endoscopic Detection and Diagnosis

At an early stage, esophageal cancer is usually asymptomatic (without dysphagia, weight loss, or anemia, which tend to reflect advanced disease), and is typically detected during screening or surveillance endoscopy or during the endoscopic evaluation of unrelated abdominal symptoms. Moreover, both histologic subtypes may appear subtle and be easily missed during endoscopy, underscoring the importance of a careful, systematic endoscopic evaluation. EAC is predominantly detected in the distal third of the esophagus, arising from a background of Barrett's epithelium, possibly in the form of an ulceration, an altered mucosal pattern, or a nodularity ( Fig. 27.1 ). EAC may also be detected during random biopsies in the absence of any apparent endoscopic abnormality. Enhanced techniques can be used to complement white light endoscopy (WLE) for a more detailed inspection of Barrett's mucosa and abnormalities. Dye-based chromoendoscopy involves the topical application of methylene blue or acetic acid using a catheter spray. These techniques are detailed in Chapter 25 .

ESCC is more commonly found in the middle and proximal third of the esophagus. Early signs of ESCC may be as subtle as a mucosal irregularity, sometimes with a thin white coating or a reddish hue. Dye chromoendoscopy, specifically using Lugol solution, can aid in the detection of early ESCC. Lugol solution is sprayed over the esophageal mucosa during endoscopy, and within minutes preferentially stains normal squamous epithelium brown/black. Demarcated, unstained tissue may represent dysplasia, carcinoma, or inflammation ( Fig. 27.2 ). In a study investigating the detection of high-grade dysplasia (HGD) or cancer in a high-risk population from China, the application of Lugol solution improved the sensitivity of WLE from 62% to 96%, while retaining a specificity of 63% (79% prior to staining). In another study, the additional finding of reddish or rose-pink appearance (so-called pink sign) was helpful in differentiating ESCC and HGD from low-grade dysplasia (LGD) or inflammation, with a sensitivity and specificity of 92% and 94%, respectively. Thus, Lugol chromoendoscopy is generally recommended during screening for ESCC. The utility of virtual chromoendoscopy in improving ESCC detection has also been investigated. Using narrow band imaging (NBI), ESCC or dysplasia may appear as a well-demarcated brownish area with distorted patterns of intrapapillary capillary loops ( Fig. 27.3 ). In a prospective, randomized controlled study, virtual chromoendoscopy significantly improved the detection of head, neck, and esophageal squamous cell cancer compared to WLE alone (97% vs. 55%). A 2015 randomized noninferiority trial comparing Lugol and virtual chromoendoscopy concluded there was no significant difference in diagnostic yield.

Staging of Early Esophageal Cancer

The depth of esophageal cancer invasion (i.e., T staging) is critical in determining the risk of lymph node involvement and whether a patient is a candidate for curative endoscopic treatment or will need esophagectomy and/or (neo)adjuvant chemoradiotherapy. To understand why, it is helpful to remember that the esophageal mucosal layer is comprised of epithelium, lamina propria, and muscularis mucosa layers. The mucosa is separated from the underlying submucosa by the basement membrane. The esophageal cancer staging system most commonly used is outlined by the American Joint Committee on Cancer. The staging system was updated in 2010, at which time separate staging systems were developed for EAC and ESCC. However, they remain identical with respect to staging of early esophageal cancers. Early esophageal cancers (i.e., stage Tis and T1a) are characterized by confinement to the esophageal mucosal layer. Specifically, stage Tis is characterized by HGD limited to the epithelium (formerly known as carcinoma in situ), and stage T1a by lamina propria or muscularis mucosa involvement (intramucosal cancers). Later-stage tumors, beginning with T1b, invade the submucosa ( Fig. 27.4 ).

Submucosal tumor involvement marks an important distinction because it confers a higher risk of lymph node metastasis and usually precludes endoscopic treatment with curative intent. In a large systematic review of 70 studies including 1874 patients undergoing esophagectomy with lymph node dissection for HGD or intramucosal adenocarcinoma, lymph node metastasis was found in 0% of 524 patients with HGD and 1.9% of patients with intramucosal cancer. In another study reviewing the outcomes of the National Cancer Database, the risk of nodal metastasis for T1a disease in patients undergoing esophagectomy for EAC and ESCC was 6.4% and 6.9%, whereas for T1b disease it was 19.6 and 20%, respectively. Hence, the risk of nodal metastasis is much greater for submucosally invasive tumors, and patients with T1b disease should be referred for consideration of esophagectomy. A subset of patients with T1b disease limited to the superficial third of submucosa (i.e., T1b sm1) and well-differentiated histology without lymphovascular invasion who are not good candidates for surgery may still be treated endoscopically with encouraging outcomes, but this decision should involve a multidisciplinary surgical and oncology discussion.

Staging of potential early esophageal cancers is best achieved by endoscopic methods of tissue acquisition (discussed later). If confinement to mucosa is confirmed, imaging modalities such as endoscopic ultrasound (EUS) and computed tomography (CT) with or without positron-emission tomography (PET) do not appear to impact the management of early esophageal cancer.

Endoscopic Inspection

Although tissue acquisition is the gold standard for early esophageal cancer staging, inspection for macroscopic clues may be helpful in predicting tumor depth, particularly for superficial ESCC. Various classification schemes have been described. The Paris classification categorizes superficial-appearing esophageal carcinoma accordingly: type 0-I (protruding), 0-II (without definite protrusion or depression), and 0-III (excavated). Type 0-I and 0-III lesions carry a high likelihood of submucosal invasion (94.7% and 100%, respectively) consistent with stage T1b disease. Type 0-IIa (slightly elevated) and 0-IIb (completely flat) tumors suggest stage Tis or T1a disease. Type 0-IIc (slightly depressed) may be either stage Tis, T1a, or T1b ( Fig. 27.5 ). The utility of NBI magnification in detecting microvascular irregularities such as intrapapillary capillary loops (to differentiate T1a from T1b tumors) has also been reported to have high sensitivity (78%) and specificity (95%). It is worth remembering that documentation of the shape, size, and location (with respect to the gastroesophageal junction) of all potential tumors can aid in planning, if surgery is recommended.

Endoscopic Resection

The presence of an esophageal nodule or mucosal irregularity in patients with risk factors, especially Barrett's esophagus (BE), should immediately raise concern, due to the high probability of harboring cancer. Biopsies alone have been found to underestimate the degree of dysplasia because of difficulty in judging the extent of invasion, and therefore have been largely supplanted by EMR and ESD, which enable acquisition of larger amounts of tissue for histology. Studies have demonstrated that EMR alters the final histological stage in 30% to 50% of cases when compared to biopsies alone. Given the low rate of lymph node metastasis in early esophageal cancer, endoscopic resection has the potential to be simultaneously curative in patients. Resection also has the potential to spare the cost, patient anxiety, and delays in care stemming from other staging modalities in this low-yield setting. Histology from EMR or ESD also provides information on prognostic factors for the survival of patients with early stage esophageal cancer, such as lymphovascular invasion and deep margin involvement ( Fig. 27.6 ). For these reasons, endoscopic resection should be the initial step in the evaluation of visible esophageal nodules or lesions suspicious for early esophageal cancer.