The esophagus is, from an endosonographer’s perspective, the window into the mediastinum, and it makes sense to consider both anatomic regions together. A good understanding of mediastinal anatomy is a prerequisite for competent image interpretation. After a basic review of cross-sectional and traditional anatomy, the endosonographer should approach mediastinal anatomy from the perspective of the ultrasound transducer in the esophagus. Numerous resources on the Internet allow familiarization of 3-dimensional anatomy of the mediastinum from every possible angle and vantage point, allowing structures and layers to be peeled away and added as desired (see Box 2.1). The number of possible cross sections that can be achieved with linear and radial endosonography probes vastly exceed those that can be reasonably expected to be represented in image collections even if they are dedicated to the topic (e.g., in an endoscopic ultrasound [EUS] anatomy atlas), although these collections are certainly helpful. Therefore, the endosonographer should make a point of identifying mediastinal structures every time an upper EUS exam is being conducted until such time when the mediastinum becomes a “second home.”

Chest radiologists have exquisite knowledge of mediastinal anatomy, and frequent consultation with them is essential for two reasons: (1) to learn and (2) to teach. Almost always it is the radiologist who, as consultant to the original requestor of the chest computed tomography (CT) or positron emission tomography (PET) scan, makes further recommendations for how a lesion could best be biopsied or further investigated. However, even chest radiologists often have an incomplete understanding of what locations can be successfully biopsied by EUS-guided fine-needle aspiration (FNA). Part of this is related to semantics, for example, mediastinal lymph nodes often are referred to by their relation to the trachea even if they are just as close to the esophagus, and this may already predicate the advice of performing an endobronchial ultrasound or mediastinoscopy instead of a transesophageal ultrasound. Furthermore, radiologists are accustomed to static images. How much the esophagus can be displaced in one direction or another by the gentle push from the tip of the echoendoscope is most often not appreciated. Lesions that appear on cross-sectional imaging as if they could not be “seen” or biopsied from a probe in the esophagus are, in fact, often within reach, and some of the cases that follow make this clear.

There are clearly situations when endobronchial ultrasound (EBUS) is the imaging modality of choice in the mediastinum and others where transesophageal EUS is the best modality. In general, EUS is most appropriate for evaluation of the posterior inferior mediastinum, whereas mediastinoscopy or EBUS are best for evaluating the anterior superior mediastinum. In many cases there will be overlap from a purely anatomic point of view. From a clinical standpoint many patients who require biopsies of lesions in the mediastinum will have coexisting pulmonary disease, and biopsies through the esophagus may be preferable to a procedure with a bronchoscope in the trachea or bronchus. The quality of biopsies (false negatives) is also correlated with how much time is available to localize the lesion and find the best biopsy trajectory and the number of biopsy passes that can be made uninterrupted by coughing and other signs of patient distress. Perspectives on the relative merits of EBUS versus EUS from the point of view of gastroenterologists and pulmonologist are listed in Box 2.1.

North American clinicians have traditionally used the Mountain-Dressler thoracic lymph node map. To reconcile differences from the Japanese Naruke system, the International Association for the Study of Lung Cancer proposed a new lymph node map in 2009 (Box 2.1). The relevant stations for the endosonographer remain the same: upper paratracheal 2L and 2R, lower paratracheal 4L and 4R, aortopulmonary window 5, subcarinal 7, and paraesophageal 8.

Mediastinal lymph node biopsies are relatively easy to perform compared with biopsies of, for example, the pancreas. The American Society for Gastrointestinal Endoscopy has an excellent video describing FNA biopsies of various anatomic locations (see Box 2.1).

The examination of small esophageal lesions and high-quality demonstrations of the esophageal wall layers is actually one of the more difficult techniques to master. An overinflated balloon that pushes hard on the esophageal wall will distort the anatomy and may lead to incorrect staging. In contrast to the stomach and rectum, it is not possible to use copious amounts of water to improve acoustic coupling. With patience and practice it is possible to combine moderate balloon inflation and small amounts of water instillation to temporarily view excellent demonstrations of small lesions. If this approach is used, regurgitation and possibly coughing and aspiration are risks. It is, therefore, preferable to elevate the head of the endoscopy table by 30 degrees.

One of the most frequent indications for EUS is the staging of esophageal carcinomas. We assume that most of our readers are familiar with the TNM (tumor, node, metastasis) system and the principles of esophageal tumor staging. There are excellent introductory textbooks that deal with this subject matter, and we will not present many esophageal tumor cases here. The problem with correct TNM staging of esophageal tumors is related to not only technical factors or operator experience but also the basic technology. Even with optimal technique there are inherent limitations, and other imaging modalities have caught up so that the contribution of EUS in the preoperative staging of esophageal cancer for surgical resection has been questioned [1]. Leading institutions discourage the restaging of esophageal cancer after combination chemotherapy and radiation [2]. A related but somewhat different problem is the evaluation of small esophageal lesions that are potentially amenable to endoscopic mucosal resection (EMR). It was previously thought that this would be an indication where EUS evaluation before the procedure would have the biggest effect. Diagnostic EMR as a first step is increasingly being promoted over EUS evaluation before EMR because the depth of submucosal invasion is difficult to accurately determine using EUS [3]. When submucosal involvement is found the patient is referred for surgical resection because of the high risk of early lymph node metastasis, which would be undetectable by imaging. We still believe that EUS is reasonable before EMR to obtain accurate measurements of the tumor, especially for those practioners who are more comfortable resecting smaller lesions and would refer larger lesions to surgeons with more experience. Furthermore, it seems that EUS will continue to contribute to the optimal management of esophageal cancers. Esophageal tumor length recently has been found to be a strong, independent predictor of long-term survival even when controlled for pTNM criteria in patients with early-stage disease who have not received neoadjuvant chemotherapy [4].

It seems that endoscopy better determines esophageal tumor length than other imaging modalities, and perhaps disease length and tumor volume defined by endoluminal ultrasound could add further value [5].