Endoscopic Evaluation of the Esophagus and Endoscopic Ultrasonography of the Esophagus

Endoscopic Evaluation of the Gastroesophageal Junction

The gastroesophageal junction (GEJ) is the level at which the esophagus ends and the stomach begins. Unfortunately, there are no universally accepted landmarks that clearly delimit the distal esophagus and the proximal stomach, and the GEJ has been defined differently by anatomists, radiologists, physiologists, and endoscopists. Landmarks suggested by anatomists, such as the peritoneal reflection or the character of the muscle bundles in the esophageal wall, are not useful for endoscopists. Radiologists refer to the region of the GEJ as the vestibule, and they seldom attempt to localize the precise point at which the esophagus joins the stomach. Physiologists have used the distal border of the LES (determined manometrically) to define the GEJ, but it is not feasible to identify this border precisely by endoscopic techniques. Indeed, one study has shown that manometric and endoscopic localizations of the LES often differ by several centimeters. This has implications for placement of a wireless pH capsule endoscopically based on the GEJ versus based on manometry because the location and consequently the findings may differ based on technique (see later).

When considering any proposed landmark for the GEJ, it is important to appreciate that there is no clear-cut “gold standard” for the structure and, consequently, all of the suggested landmarks can be considered arbitrary. For most disorders of the esophagus and stomach that are diagnosed endoscopically, furthermore, it is not important that the GEJ be identified with great precision. For some disorders, most notably Barrett esophagus, for which the endoscopist must determine the extent of esophageal columnar lining, precise localization of the GEJ can be critical for establishing the diagnosis.

Suggested endoscopic criteria for the GEJ include: (1) the level at which the tubular esophagus flares to become the sack-like stomach, (2) the proximal margin of the gastric rugal folds when the esophagus and stomach are partially distended, and (3) the distal end of the esophageal palisade vessels. Although these landmarks may be recognized readily in still photographs of the junction region, the distal esophagus in vivo is a dynamic structure whose appearance changes from moment to moment. The location of the point of flare changes with respiratory and peristaltic activity. The proximal gastric folds can prolapse transiently up into the esophagus. The appearance of the junction region also varies with the degree of distention of the esophagus and stomach, and the palisade vessels can be difficult to identify using conventional endoscopes.

The proximal extent of the gastric folds is the landmark for the GEJ used frequently by Western endoscopists ( Figs. 7.3 to 7.5 ). This landmark was proposed by McClave et al. in 1987 based on their endoscopic observations in only four subjects who were identified as normal controls because they had “no clinical evidence of esophageal disease.” The junction between squamous and columnar epithelia (the SCJ) was located within 2 cm of the gastric folds in all of those four subjects, and so the authors concluded that the diagnosis of columnar-lined esophagus should be considered only when the SCJ is located more than 2 cm above the GEJ (i.e., the proximal level of the gastric folds). This study can be criticized both for the small number of control subjects and for the lack of documentation that the four controls were indeed normal. Esophageal pH monitoring studies were not performed, and so it is not clear that the control subjects had normal esophageal acid exposure. Biopsy specimens of the columnar-lined esophagus were not taken, and so short-segment Barrett esophagus was not excluded (see later). Furthermore, three of the four control subjects had hiatus hernias and one had reflux esophagitis. It seems surprising that a proposed landmark based on such questionable data has been so widely accepted by endoscopists.

A number of Asian investigators use the end of the esophageal palisade vessels as their landmark for the GEJ ( Fig. 7.6 ). Elegant anatomic studies of the GEJ have revealed four distinct zones of venous drainage, including a gastric zone, a palisade zone, a perforating zone, and a truncal zone. The palisade zone comprises a group of fine, longitudinal veins located largely within the lamina propria of the distal esophagus. The palisade vessels pierce the muscularis mucosae distally to join the submucosal vessels of the gastric zone and proximally to join the submucosal vessels of the perforating zone. The palisade vessels can be difficult to visualize by conventional endoscopy, especially if there is inflammation in the distal esophagus. The appearance of these vessels can be enhanced by narrow band imaging endoscopy, which uses primarily blue light that penetrates only the superficial layers of the mucosa (where the palisade vessels are found) and that is absorbed by the hemoglobin within the vessels. Furthermore, even in autopsy studies in which blood vessels of the GEJ region are injected with resins that provide exquisite detail of the venous structures, it is difficult to identify precisely the termination of the palisade vessels. Finally, it is not clear conceptually why the distal end of the palisade vessels should be considered the precise end of the esophagus.

Few studies have addressed specifically the problem of endoscopic localization of the GEJ and, even in those that have done so, the accuracy of the criteria used cannot be assessed meaningfully in the absence of a gold standard. It is not clear which is the best diagnostic criterion for the GEJ, and the reproducibility of the various criteria have not been established. If one cannot determine with certainty where the esophagus ends and the stomach begins, then any assessment of the extent of esophagus lined by columnar epithelium will be inherently imprecise. This unresolved problem continues to confound clinicians and investigators who deal with Barrett esophagus.

Conventional Endoscopic Diagnosis of Barrett Esophagus

Barrett esophagus is the condition in which metaplastic columnar epithelium that predisposes to cancer development replaces the stratified squamous epithelium that normally lines the distal esophagus. Endoscopic examination is required to establish a diagnosis of Barrett esophagus, and the endoscopic impression must be confirmed by histologic evaluation of biopsy specimens from the columnar-lined esophagus. Specifically, the endoscopist must ensure that the following two criteria are fulfilled : (1) columnar epithelium lines the distal esophagus and (2) biopsy specimens of the columnar-lined esophagus show specialized intestinal metaplasia. To document that columnar epithelium lines the esophagus, the endoscopist must identify both the SCJ and GEJ (see Fig. 7.3 ). Columnar epithelium has a reddish color and coarse texture on endoscopic examination, whereas squamous epithelium has a pale, glossy appearance. Narrow band imaging is invaluable for assessing this. The juxtaposition of these epithelia at the SCJ forms a visible line called the Z-line. When the SCJ and GEJ coincide ( Fig. 7.7 ), then the entire esophagus is lined by squamous epithelium. When the SCJ is located proximal to the GEJ (see Fig. 7.3 ), then there is a columnar-lined segment of esophagus. If the endoscopist takes biopsy specimens from that columnar-lined segment and histologic evaluation shows specialized intestinal metaplasia, then the patient has Barrett esophagus (in the United States definition of Barrett).

Several classification systems for Barrett esophagus have been proposed based on the extent of columnar-lined esophagus and on the appearance of the Z-line. Perhaps the most widely used system classifies patients as having either “long segment” or “short segment” Barrett esophagus. Patients have long-segment Barrett esophagus when the distance between the GEJ and the most proximal extent of the Z-line is 3 cm or more, and they have short-segment Barrett esophagus when that distance is less than 3 cm. The cutoff value of 3 cm is arbitrary, and this classification has no clear implications regarding the pathogenesis of the condition or the clinical management of affected patients. Furthermore, there can be substantial variation in the appearance of the Z-line among patients with Barrett esophagus ( Figs. 7.8 to 7.10 ), and the short-long classification provides no specific information about that appearance.

In 2000 Wallner et al. proposed the ZAP ( Z -line AP pearance) classification for evaluating the SCJ. The ZAP classification has four categories as follows : Grade 0—the Z-line is sharp and circular; grade I—the Z-line is irregular and there are tonguelike protrusions and/or islands of columnar epithelium; grade II—there is a distinct, obvious tongue of columnar epithelium less than 3 cm in length; grade III—there are distinct tongues of columnar epithelium greater than 3 cm in length, or there is a cephalad displacement of the Z-line greater than 3 cm. The likelihood of finding intestinal metaplasia (and hence having Barrett esophagus) was shown to increase significantly with increasing ZAP grades, and the classification was found to have excellent reproducibility among endoscopists. However, the clinical utility of the ZAP classification has not been established, and the system has not been used widely in clinical practice.

Recently, a new system has been proposed for grading Barrett esophagus called the Prague C and M criteria. This system describes both the extent of circumferential metaplasia (C, measured from the GEJ to the most proximal extent of circumferential esophageal metaplasia) and the extent of the longest tongue of esophageal metaplasia (M, measured from the GEJ to the most proximal extent of esophageal metaplasia). For example, a patient classified as C2M5 has columnar metaplasia involving the distal 2 cm of the esophagus in a circumferential fashion with a tongue of metaplasia that extends 5 cm above the GEJ. One study has demonstrated excellent interobserver agreement among endoscopists using the Prague C and M criteria when columnar epithelium extends at least 1 cm above the GEJ, but poor agreement for shorter segments of esophageal columnar lining. The clinical utility of this system has not been established. Some have argued that the term Barrett esophagus itself is artificial and that the condition has been defined variably by investigators who have imposed arbitrary criteria that fit their personal perspectives. In 1996 Spechler and Goyal proposed a simple classification system as follows: Whenever columnar epithelium is seen in the esophagus, regardless of extent, the condition is called “columnar-lined esophagus.” In these cases, biopsy specimens can be obtained from the esophageal columnar lining to seek specialized intestinal metaplasia. The condition then can be classified as either “columnar-lined esophagus with specialized intestinal metaplasia” or “columnar-lined esophagus without specialized intestinal metaplasia.” Despite the simplicity and conceptual appeal of this system, the term Barrett esophagus has become so firmly entrenched among clinicians that it is unlikely to be abandoned.

Specialized Endoscopic Techniques for Barrett Esophagus

A variety of specialized endoscopic techniques are available for the evaluation of Barrett esophagus including chromoendoscopy, magnification endoscopy, narrow band imaging, endosonography, optical coherence tomography, and spectroscopy using reflectance, absorption, light-scattering, fluorescence, and Raman detection methods. These techniques have been used to enhance the identification of both intestinal metaplasia in the esophagus and neoplasia in Barrett esophagus. Only chromoendoscopy, magnification endoscopy, and narrow band imaging will be discussed in this chapter.

In chromoendoscopy the esophageal mucosa is painted with dyes that either stain the cells that absorb them or that accumulate in mucosal crevices to enhance the architectural features of the epithelium. When potassium iodide is absorbed by squamous epithelial cells, it binds to their glycogen and stains them brown. The application of this dye can help to delineate the SCJ. For individuals who are at high risk for squamous cell cancers of the esophagus (e.g., patients who have had cancers of the head and neck, individuals living in high-incidence areas for squamous cell carcinoma such as northern China), potassium iodide staining also has been used to identify areas of early neoplasia in the squamous epithelium. Methylene blue dye is absorbed by intestinal-type cells, and this dye can be applied to identify areas of intestinal metaplasia in a columnar-lined esophagus. In addition, areas of dysplasia and early cancer in the specialized intestinal metaplasia of Barrett esophagus can be identified by their failure to absorb methylene blue. One report has shown that methylene blue application may cause DNA damage in Barrett esophagus, and so the use of this dye conceivably could be dangerous. Indigo carmine is a chromoendoscopy dye that is not absorbed and is used to enhance architectural features. Cresyl violet dye stains the columnar cells that absorb it purple, and the dye also accumulates in crevices to enhance architectural features. Acetic acid, although not a dye, is often sprayed on the mucosa before chromoendoscopy as a mucolytic agent. Acetic acid application also causes the columnar epithelium to swell, and this effect may enhance the evaluation of architectural features.

In magnification endoscopy an optical zoom device is used to magnify the mucosa up to 150-fold. Magnification endoscopy can also be combined with chromoendoscopy. Investigators using this technique have identified a variety of “pit-patterns” that might be typical of the intestinal metaplasia of Barrett esophagus ( Figs. 7.11 and 7.12 ). Magnification endoscopy also can be combined with narrow band imaging ( Fig. 7.13 ).

Endoscopic Diagnosis of Reflux Esophagitis

Gastroesophageal reflux disease (GERD) has been defined as the condition that develops when the reflux of stomach contents causes troublesome symptoms and/or complications. Heartburn is the most common symptom of GERD, and tissue injury results when esophageal epithelial cells succumb to the damaging effects of the refluxed acid and pepsin. When these caustic agents cause macroscopic injury to the esophageal epithelium, the endoscopist can make a diagnosis of reflux esophagitis. However, more than 50% of patients who have typical GERD symptoms have normal endoscopic examinations. Thus it appears that GERD usually does not cause visible damage to the esophageal mucosa in most patients.

Mild changes of GERD that may be visible to the endoscopist include mucosal erythema, edema, hypervascularity, friability, and blurring of the SCJ. However, identification of those changes is a subjective skill, and agreement among endoscopists regarding the presence of such minimal signs of reflux esophagitis can be very poor. More severe GERD can result in esophageal erosions and ulcerations. Histologically, erosions are defined as superficial necrotic defects that do not penetrate the muscularis mucosae, whereas ulcerations are deeper defects that extend through the muscularis mucosae into the submucosa. Endoscopically, these peptic esophageal lesions are identified on the basis of their gross features, and clinicians seldom have histologic confirmation that the lesions they call “esophageal ulcers” in fact have breached the muscularis mucosae. Thus the distinction between an esophageal ulceration and an erosion usually is based on a subjective assessment of the depth of the necrotic lesion. One modern system for grading the severity of reflux esophagitis, the Los Angeles classification, avoids the problem of distinguishing erosions from ulcerations by referring to both as “mucosal breaks.”

More than 30 systems for the classification of reflux esophagitis have been proposed over the past few decades. The endoscopic criteria for three of the most widely used systems are listed in Table 7.1 . All of the proposed systems have limitations, and no one system has been shown to be clearly superior to another for establishing the diagnosis of GERD or for predicting the response to treatment. Arguably the best validated and most widely used system now is the Los Angeles classification that was proposed at the World Congress of Gastroenterology meeting in Los Angeles in 1994. In this system a mucosal break is defined as “an area of slough or erythema with a discrete line of demarcation from the adjacent, more normal-looking mucosa” ( Fig. 7.14 ). Esophagitis is graded on a scale of A to D depending on the length and circumferential extent of the mucosal breaks ( Figs. 7.14 and 7.15 ). Los Angeles grades C and D represent severe reflux esophagitis. Originally, grade D esophagitis was defined as a mucosal break that involved the entire circumference of the esophagus, but this was modified in 1999 to the criterion shown in Table 7.1 because it can be difficult to ascertain that a mucosal break is completely circumferential.

Endoscopic Evaluation of Patients Who Have Had Antireflux Surgery

The two most commonly used fundoplication procedures (Nissen and Toupet) create characteristic folds in the proximal stomach that are best appreciated with the endoscope in the retroflexed position. The folds of the fundoplication should be located just below the diaphragm ( Fig. 7.16 ). If the folds are seen above the diaphragm, it is an indication that the fundoplication has herniated into the chest, which usually results from disruption of the crural repair. If there is a pouch of stomach proximal to the folds of the fundoplication, the condition is called a “slipped” fundoplication (e.g., a “slipped Nissen”). A slipped fundoplication can occur in two ways: (1) the fundoplication is fashioned in the correct location but a portion of the stomach later herniates (“slips”) through the fundoplication, or (2) the surgeon mistakes the proximal stomach for the distal esophagus and inadvertently fashions the fundoplication around the stomach. Although the latter situation represents an initial surgical error rather than a later slippage (herniation), the condition is called a slipped fundoplication despite the misnomer. Identification of a low or “slipped” Nissen can be challenging and is best appreciated on careful antegrade evaluation of the distal esophagus during endoscopy. Gastric rugal folds above the pinch of the fundoplication suggest a low or “slipped” Nissen. However, a Collis gastroplasty will usually have this appearance since an esophageal extension of stomach is created, so the specifics of the surgical procedure are important to know when evaluating patients with symptoms after a fundoplication. Finally, the absence of fundoplication folds suggests total disruption of the antireflux procedure (the “missin' Nissen”). Any of these abnormalities can render the antireflux surgery ineffective.

The folds of a properly constructed fundoplication should be oriented parallel to the diaphragm. An oblique orientation of the folds suggests twisting of the fundoplication or improper construction of the wrap using the body rather than the fundus of the stomach ( Fig. 7.17 ). Either of these conditions can cause postoperative gastroesophageal reflux, dysphagia, or both. The folds should measure approximately 1 to 2 cm in span. A wider span indicates a too-generous fundoplication that can cause dysphagia. A paraesophageal hernia also can cause dysphagia by pressing on the distal esophagus ( Fig. 7.18 ). The herniated portion of the stomach in these cases often originates from the fundoplication itself and may result from attempts to construct a “floppy” wrap.

Esophageal Cancer

Esophageal cancers that are recognizable by conventional endoscopy appear as masses that protrude into the lumen of the esophagus. The masses are often nodular, irregular, and ulcerated, and the tumors may have a different color and texture than the surrounding normal mucosa. Squamous cell and adenocarcinomas of the esophagus cannot be differentiated on the basis of endoscopic appearance, but the location of the tumor and its associated features may provide important clues regarding the histology. Tumors that involve the proximal and middle esophagus and that are separated from the stomach by a segment of squamous epithelium are very likely to be squamous cell carcinomas. Tumors of the distal esophagus can be either squamous cell carcinomas or adenocarcinomas. If there is associated Barrett esophagus, the tumor is likely to be an adenocarcinoma ( Figs. 7.19 and 7.20 ). However, adenocarcinomas that cause symptoms often have grown so large that they have obliterated any evidence of the Barrett esophagus that spawned them. It can be especially difficult to determine the origin of an adenocarcinoma that straddles the GEJ ( Fig. 7.21 ). Such tumors can arise either from Barrett esophagus or from the proximal stomach. If no Barrett esophagus is apparent, investigators have relied on the location of the tumor epicenter to classify the tumor as esophageal or “cardiac.”

Eosinophilic Esophagitis

Eosinophilic esophagitis (EoE) is a modern esophageal disorder that has become recognized widely only within the past decade. EoE appears to be a manifestation of food allergy in which eosinophils infiltrate the esophageal epithelium, causing symptoms and tissue damage mediated by cytokines released from the eosinophils and surrounding tissues. The disorder commonly is diagnosed in men in the fourth and fifth decades of life who describe a long history of dysphagia for solid foods, often with hospital visits for food impactions. Heartburn is also a common complaint, and it can sometimes be difficult to distinguish EoE from GERD. Patients frequently have a personal and family history of allergic disorders such as asthma, atopic dermatitis, eczema, hay fever, and food allergies. Children with EoE may have symptoms of abdominal pain, heartburn, vomiting, feeding disorders, and failure to thrive.

Multiple esophageal rings are common endoscopic findings in patients with EoE ( Fig. 7.22 ). When pronounced, the rings may give the esophagus a trachea-like appearance. Other common esophageal endoscopic abnormalities include vertical furrows ( Fig. 7.23 ), strictures, “white specks” (which are 1- to 3-mm-diameter eosinophilic exudates), and small-caliber esophagus. In up to 25% of cases the esophagus appears normal endoscopically. Esophageal biopsy is needed to establish the diagnosis. The esophageal mucosa is unusually fragile in EoE, and esophageal dilations often are complicated by extensive mucosal tears that can be quite painful.

Endoscopic Esophageal Ultrasonography

The advent of endoscopic ultrasonography (EUS) has extended endoscopic examination of the esophagus beyond the mucosa into the esophageal wall and paraesophageal tissues. The diagnostic capabilities of cutaneous ultrasound have been expanded by endoscopic placement of ultrasound transducers adjacent to the gastrointestinal mucosa. These transducers, operating at relatively high frequencies, provide detailed examination of the esophageal wall and surrounding tissues. EUS is the most significant advance in the diagnosis of esophageal disease since the introduction of flexible fiberoptic endoscopy. These intracorporeal examinations have proved beneficial in the diagnosis and treatment of both benign and malignant diseases of the esophagus and adjacent structures.

Fundamentals of Ultrasonography

Sound is produced by vibration of a source within a medium. Vibration produces waves, cyclic compression, and rarefaction (expansion) of molecules in the medium, thus transmitting the sound wave through the medium. The number of cycles (compression and rarefaction) of a sound wave occurring in 1 second is the frequency and is measured in hertz (Hz). The frequency of sound waves audible to the human ear is between 20 and 20,000 Hz. Sound waves with frequencies higher than 20,000 Hz are ultrasound waves. Frequencies used in medical ultrasound imaging range from 1 to 20 million Hz (1 to 20 MHz).

Ultrasound waves may be produced by electrical excitation of a piezoelectric crystal. The application of voltage across a crystal causes it to deform. Alternating electrical energy vibrates the crystal and produces sound waves. Conversely, if a sound wave deforms a crystal, electrical energy is produced. It is this ability to convert electrical energy into sound energy and, conversely, to convert sound energy into electrical energy that allows these crystals to function as both transmitters and receivers (i.e., as transducers ). These transducers are responsive to a limited range of frequencies; hence more than one transducer may be required for an ultrasound examination.

The speed of a sound wave within a medium (tissue) is defined by the following relationship: V = ( K / p ) ^1/2 , where V is the velocity of the sound wave, K is the bulk modulus of the tissue (a measure of stiffness), and p is the density of the tissue.

The resistance to passage of a sound wave through tissue is called the acoustic impedance ( Z ), which is defined by the following relationship: Z = pV = ( pK ) ^1/2 .

Sound waves travel best through dense or elastic tissue. Absorption of some of the energy of an ultrasound wave occurs as the wave passes through tissue. The amount of absorption is determined by tissue characteristics and the frequency of the sound wave. Higher-frequency waves have greater absorption.

Interactions occurring as a sound wave encounters different tissues are critical to the diagnostic capabilities of ultrasound. As a sound wave passes from one tissue to the next, a portion of the wave is transmitted and a portion is reflected. The reflected wave is received by the transducer, thereby providing the diagnostic information of ultrasound. The difference in acoustic impedance between the two tissues and the angle at which the sound wave enters the new medium (angle of incidence) determine the portion of the wave that is reflected and the portion that is transmitted. In tissue with similar acoustic impedance, most of the wave is transmitted. Soft tissue has excellent transmission qualities; the density and velocity vary only by 12% to 14% among different soft tissues. Because acoustic impedance is the product of velocity and density, the product of these small changes results in a 22% difference in acoustic impedance between fat and muscle. Useless, bright echo images are obtained when an ultrasound wave encounters air or bone. Air is very compressible and of low density, whereas bone, although dense, has low compressibility and high reflectivity. These properties account for the poor transmission of ultrasound waves from tissue to air or tissue to bone. The amount of reflected sound is also related to the angle of incidence: as the angle of incidence increases, less sound is reflected. In addition, sound waves are bent as they travel from one tissue to the next. This process is termed refraction .

Absorption, reflection, and refraction are major sources of energy loss. Some ultrasound wave energy is also lost by scattering (diffusion), which occurs when a sound wave encounters heterogeneous tissue. Tiny particles within tissue (such as fat in muscle), smaller than the ultrasound wavelength, scatter the ultrasound wave. As a sound wave passes through tissue, a portion of its energy is lost; this is called attenuation . Attenuation increases as more tissues are encountered and as the wave travels farther from the source. If the returning ultrasound wave is not processed, the same tissue would be imaged differently, depending on its distance from the transducer. The intensity of the returning waves must be amplified (gain) to ensure that distant waves are correctly represented. Attenuation increases as ultrasound frequency increases.

Resolution is the ability to discriminate among different tissues with ultrasound waves. Depth or axial resolution is the ability to differentiate between two tissues along the path of the ultrasound wave. Lateral resolution is the ability to distinguish between adjacent tissues. Transducer characteristics and focus determine resolution. Higher frequencies allow better resolution but decreased tissue penetration.

Pulse-echo technique is used in EUS. Ultrasound waves are emitted for a brief period, followed by a subsequent listening period during which the reflected waves are received. The returning ultrasound waves are displayed such that the brightness is proportional to the amplitude of the returning ultrasound waves. This is known as B-mode ultrasonography . Because the amplitude is presented in a range from white to gray to black, the display is also termed grayscale ultrasound . Individual scans are shown at a rate at which the eye cannot detect single images (12/second). This fast-frame display is called real-time ultrasound and allows the ultrasonographer to study tissue temporally as well as spatially.

Instruments and Techniques

Because EUS does not provide adequate endoscopic inspection of the upper gastrointestinal tract, every ultrasound study should be preceded by a standard flexible endoscopic upper gastrointestinal examination. This provides precise location and mucosal definition (including biopsy) of the esophageal lesion and guides the ultrasound examiner. The ultrasound endoscope is generally passed blindly through the oropharynx and hypopharynx. Care must be taken because the distal tip containing the transducer is rigid. For complete examination, the endoscope must be passed beyond the esophagus into the stomach.

In the past, the radial mechanical ultrasound endoscope ( Fig. 7.24 ) was the principal instrument used for EUS. The ultrasound transducer is housed in the tip of the endoscope. It produces up to a 360-degree sector scan perpendicular to the transducer tip. Because the transducer is adjacent to tissues to be examined, higher frequencies than those used in extracorporeal ultrasound can be used. In the newest models, a range of transducer frequencies, from 5 to 20 MHz, are available. These transducers allow adequate visualization of anatomic structures to a depth of 3 to 12 cm. An acceptable acoustic interface between the transducer and the tissue being examined must be obtained to ensure good-quality ultrasound images. This is most commonly accomplished by covering the tip of the endoscope with a latex balloon, which can be filled with water to provide an excellent acoustic interface (see Fig. 7.24 ). A less commonly used technique is rapid insufflation of the esophageal lumen with water. This provides an excellent, but transient acoustic interface without the tissue compression that may occur with the latex balloon. Current echoendoscopes also provide a video endoscopic image, albeit with a somewhat limited view in a forward oblique direction. The control section contains the deflection controls and air/water and suction valves, similar to those on a standard endoscope (see Fig. 7.24 ). A water inflation/deflation system for the balloon is incorporated into the air/water and suction valve mechanisms. A direct-current motor and drive mechanism that rotates the ultrasound transducer are housed in the control section. Current ultrasound endoscopes are totally immersible in liquids.

A radial mechanical blind probe ( Fig. 7.25 ) is available for the evaluation of esophageal strictures. This echoendoscope provides images similar to those of larger-diameter radial mechanical echoendoscopes, but it has no endoscopic optical capabilities and is less than 8 mm in diameter. More commonly used in current practice are higher-frequency miniprobes passed through the operating channel of standard endoscopes ( Fig. 7.26 ); these miniprobes provide radial images from 12 to 30 MHz.

These three instruments are used in conjunction with an image processor ( Fig. 7.27 ). The image processor allows for adjustment of gain, contrast, and sensitivity time control to regulate the strength of the returning echo at different depths. Onscreen calibration and labeling can be done with the image processor. The image may be displayed on a video monitor or stored digitally or on videotape. The image processor has been refined and miniaturized with successive generations of endoscopic ultrasound equipment.

Newer electronic endoscopes are now more commonly used. The electronic radial echoendoscope provides an enhanced image as a result of use of tissue harmonic echo and can provide color and power Doppler ( Fig. 7.28 ). The curvilinear electronic echoendoscope ( Fig. 7.29 ) also has video endoscopic capability and can produce up to a 180-degree oblique forward field. It allows a range of scanning frequencies from 5 to 10 MHz with a depth of penetration of 4 cm or greater. This echoendoscope provides color and power Doppler examination and direct visualization of cytology needles passed into and beyond the esophageal wall.

Radial and curvilinear echoendoscopes have increased the accuracy of EUS. For diagnostic purposes, the radial scanner is preferable because it allows a 360-degree view and is known as the “workhorse” of EUS. Because the radial scanner does not allow safe directed passage of a needle into the esophageal wall or adjacent tissue if a tissue sample is required for cytologic evaluation, the electronic curvilinear echoendoscope is used. It is possible to perform both diagnosis and fine-needle aspiration (FNA) with the electronic linear echoendoscope alone, but the limitation in viewing field requires significant torque on the insertion tube to image the esophageal wall and adjacent tissues for a 360-degree view. However, comparable results for staging examinations have been reported with the electronic curvilinear echoendoscope. Both systems must be available for adequate EUS evaluation. Electronic radial and linear echoendoscopic examinations can be accomplished using one image processor ( Fig. 7.30 ).

Esophageal Wall and Ultrasound Anatomy

The esophageal wall is composed of three distinct layers: mucosa, submucosa, and muscularis propria ( Fig. 7.31 ). The mucosa has three elements: epithelium, lamina propria, and muscularis mucosae. The innermost layer is stratified, nonkeratinizing squamous epithelium. It is separated and isolated from the remainder of the esophageal wall by a basement membrane. Immediately beneath is the lamina propria. This loose matrix of collagen and elastic fibers forms a superficial undulating layer; invaginations into the epithelium produce epithelial papillae. Lymphatic channels in the lamina propria are an anatomic feature unique to the esophagus. The muscularis mucosae surrounds the lamina propria. This smooth muscle layer pleats the two inner layers of the mucosa into folds that disappear with distention of the lumen.

The submucosa is composed of connective tissues that contain a rich network of blood vessels and lymphatics. The dense submucosal lymphatic plexus facilitates early dissemination of esophageal malignancies. Submucosal glands of mixed type are characteristic of the esophagus.

The muscularis propria is the muscular sleeve that provides the propulsive force necessary for swallowing. There are two layers of muscle: an inner circular layer and an outer longitudinal layer. The upper cervical esophagus is composed entirely of striated muscle. There is a gradual transition from striated to smooth muscle within muscle bundles until the esophagus is entirely smooth muscle at the upper and midthird junction. Lymphatic channels pierce the muscularis propria and drain into regional lymphatics or directly into the thoracic duct.

The esophagus has no investing adventitia. The paraesophageal tissue is composed of fibrofatty tissue that lies directly against the outer fibers of the muscularis propria.

The normal esophagus is usually viewed as five discrete layers by EUS ( Fig. 7.32 ). These layers are seen as alternating hyperechoic (white) and hypoechoic (black) rings. Studies demonstrate that the five layers seen by EUS correspond to the balloon-mucosa interface, the mucosa deep to this interface, the submucosa and the acoustic interface between the submucosa and muscularis propria, the muscularis propria minus the acoustic interface between the submucosa and the muscularis propria, and the periesophageal tissue. For clinical purposes, these layers represent the superficial mucosa, deep mucosa, submucosa, muscularis propria, and periesophageal tissue. In the upper part of the esophagus, with overdistention of the examining balloon or if the transducer is too close to the esophageal wall, only three layers of the esophageal wall may be apparent because the superficial mucosa, deep mucosa, and submucosa compose one hyperechoic layer. The thickness of each ultrasound layer is about equal and does not represent the thickness of the tissue layer but, instead, the time that it takes the ultrasound wave to traverse this layer.