History

The work of many pioneers has enabled the modern physician to visualize and study the esophagus.¹ Bozzini, in 1806, was the first physician to report the ability to visualize the proximal esophagus.² After studying the performances of sword swallowers, Kussmaul performed the first rigid esophagoscopy in 1868. Advances in equipment continued to be made through the work of physicians such as Von Mikulicz (1881, first electrically lighted esophagoscope), Einhorn (1897, distal illumination in an esophagoscope), and Jackson (advances in endoscopic instruments). Interestingly, Jackson did not use anesthesia, general or local, for esophagoscopy of adults.

While advances in equipment, lighting, and optics were being made in rigid instrumentation, a major advance toward flexible esophagoscopy was made through the work of Hirschowitz and associates,³ who presented the first fiberoptic gastroscope in 1957. Since that time, the examination of the esophagus has been conducted primarily by gastroenterologists with flexible endoscopes and patient sedation. Shaker,⁴ a gastroenterologist, published the first report of unsedated transnasal esophagogastroduodenoscopy (EGD) in 1994. Although initially met with limited interest from gastroenterologists, the appeal of unsedated transnasal esophagoscopy (TNE) grew among the otolaryngology community, particularly after the reports of Herrmann and Recio⁵ and the live demonstration by Aviv and colleagues⁶ in 2001.

Equipment

One of the primary differences between a transnasal esophagoscope and a standard flexible oral esophagoscope is the instrument diameter—3.1 to 5.1 mm versus 10 to 12 mm. Most transnasal endoscopes are able to suction and insufflate air, and contain a working channel. The working channel can be used for passing biopsy forceps or brushes, laser fibers, guidewires for balloon dilators, and the topical application of anesthesia. However, ultrathin transnasal esophagoscopes may lack these capabilities and require an add-on sheath, which may increase the diameter by 1.3 mm. The length of a flexible laryngoscope is 30 to 35 cm, but that of transnasal esophagoscopes is typically at least 60 cm to allow visualization of the entire esophagus, including a retroflexed view of the gastroesophageal junction.

Two main camera systems exist for TNE. The proximal “add-on” camera is familiar to most otolaryngologists who perform endoscopy. Endoscopes utilizing this tool typically lack a working channel and require an add-on sheath. Alternatively, distal videochip technology has emerged, whereby optical impulses are converted to electric signals without the need for optical fibers.⁷ A separate video processor is required for visualization in this system (Fig. 73-1).

Figure 73-1. Video chip transnasal esophagoscope. An add-on camera is not needed with this system.

Tolerance of Unsedated Examination

Historically, examination of the esophagus has been performed with patient sedation. Analyses of the complication patterns during endoscopy reveal that most adverse events are related to sedation. Cardiopulmonary complications account for more than 50% of all reported complications, with the majority being due to aspiration, over-sedation, hypoventilation, vasovagal episodes, and airway obstruction.⁸ A 2007 national survey of endoscopists confirmed that unplanned cardiopulmonary events secondary to conscious sedation constitute the majority of endoscopy complications—67% of complications and 72% of mortalities were cardiopulmonary related.⁹

In addition to the medical risks of sedation, economic factors arise. The increased direct costs of conventional esophagoscopy include a longer procedure time, longer recovery time, and the costs associated with the needed medications, sedation, and recovery teams.¹⁰ Unsedated TNE has been found to be up to two thirds less expensive than conventional esophagoscopy.^11,12 Indirect costs accrued with conventional esophagoscopy include the loss of work time for both the patient and an adult for patient transportation. In contrast, with an unsedated examination, patients are able to resume routine daily activities shortly after completion of the study.

Physiologic studies have verified that unsedated esophagoscopy is a safe procedure. In a prospective study, the changes in systolic and diastolic blood pressures, pulse rate, and rate-pressure product, as well as the drop in arterial oxygen saturation experienced by patients who had conventional transoral EGD, were significantly greater than in those undergoing either transnasal or transoral small endoscopy. Between the two small endoscope groups, these same parameters had a significantly greater change in the transoral group than in the transnasal group at 2 minutes after esophageal intubation.¹³ Additional studies have shown that a significant increase in heart rate and a significant decrease in oxygen saturation occur in those undergoing transoral esophagoscopy but not in those undergoing transnasal esophagoscopy.¹⁴

Although initial patient anxiety is higher before unsedated TNE, numerous studies have demonstrated a very high patient satisfaction rate, usually greater than that for conventional or transoral approaches.^15–18 Crossover studies have shown that in patients who have undergone both sedated and unsedated endoscopy, the unsedated examination was better tolerated.¹⁹ Dumortier and coworkers²⁰ found that 91% of patients who had undergone unsedated TNE who had also previously had a conventional EGD stated that they preferred the TNE examination. TNE was associated with fewer gagging sensations and was rated with higher patient satisfaction than unsedated transoral examinations.^21–23

Accuracy of Examination

Dean and associates²⁴ reported a high degree of correlation between the endoscopic findings (sensitivity 89%, specificity 97%) of TNE examinations and of conventional EGD. Although the biopsy sample size obtained with TNE is generally smaller than that obtained with conventional endoscopy, numerous studies have verified the accuracy of TNE-guided biopsies. In patients with known Barrett’s metaplasia, Saeian and colleagues²⁵ found that the congruence between biopsy samples obtained with conventional endoscopy and those obtained with TNE was 97%. In a crossover study of 121 patients, Jobe and colleagues²⁶ compared the endoscopic findings and the histologic detection of Barrett’s esophagus from biopsy specimens obtained from unsedated TNE and conventional esophagoscopy. A moderate to very good extent of agreement was present for endoscopic findings, such as hiatal hernias and esophagitis, between the two endoscopic modalities. In addition, although the biopsy specimen size was significantly smaller with TNE, there was a moderate extent of agreement in pathologic findings.²⁶

Technique

The most important step in performing TNE is proper nasal anesthesia and decongestion. With the patient in a sitting position, the more patent nasal cavity is sprayed with a mixture of a topical anesthetic and decongestant and then packed with a cotton pledget soaked in the same mixture. A topical anesthetic may be administered to the oropharynx. Too much anesthesia may actually make the TNE procedure more difficult owing to the accumulation of secretions in the hypopharynx that may be aspirated and induce coughing.

The endoscope is lubricated and introduced into the naris. It may be passed either along the floor of the nose—the preferred method—or between the middle and inferior nasal turbinates. As the scope approaches the post-cricoid region, the patient’s head is flexed toward the chest. The patient performs a swallow as the endoscope is gently introduced into the esophagus. Air insufflation and suction are used, as needed, to improve visualization. The scope is advanced only when the lumen is clearly seen.

The endoscope is advanced into the distal esophagus for visualization of the gastroesophageal junction and the squamocolumnar junction (SCJ), also known as the Z-line. The endoscope is advanced more deeply into the stomach and retroflexed to visualize the gastric cardia. The mucosa of the entire esophagus is examined while the instrument is withdrawn. The post-cricoid region is visualized by insufflation of a burst of air as the endoscope is withdrawn from the esophagus (Fig. 73-2).¹⁵

Figure 73-2. Examination technique. The unsedated patient is seated comfortably throughout the entire examination.

Normal Findings

The anatomy of the esophagus viewed through a thin endoscope appears different from that obtained through a rigid esophagoscope or a conventional oral esophagoscope.²⁷ By nature of the smaller size of the scope, TNE does not distort the anatomy of the esophagus, thereby allowing the natural curvatures and external compressions to be easily seen.

Three major curvatures of the esophagus are encountered before the endoscope enters the stomach. At the base of the neck, the esophagus curves to the left with a gentle return to midline at the level of T5. Another left curve is encountered at T7. Finally, an anterior curve occurs in the distal esophagus as it passes through the diaphragm.

Three external compressions are seen during a TNE examination. The first encountered is a pulsatile compression of the left anterolateral wall from the aorta. Approximately 2 cm more distal, the left main bronchus produces an anterior compression. In the distal esophagus, the diaphragm produces a circumferential constriction best visualized by having the patient sniff.

A network of terminal linear vessels can be seen in the distal esophagus that end at the gastric rugae. These vessels do not normally extend below the level of the SCJ. Folds of the gastric rugae form may extend up to 2 cm above the pinch of the diaphragm. Proximal extension of the rugae beyond 2 cm indicates a hiatal hernia. The presence of a hiatal hernia may also be suggested by visualization of a loose diaphragm crus on the retroflexed view.

Esophageal squamous epithelium normally appears as a gray-white color, whereas the gastric columnar epithelium is of a salmon-pink hue. The SCJ occurs at the meeting of the esophageal mucosa and gastric mucosa. An irregularly shaped SCJ may be found and is not always pathologic. The junction of the gastric rugae and the network of terminal linear vessels marks the level of the gastroesophageal junction. The SCJ and the junction should normally coincide. However, finding the SCJ above the gastroesophageal junction represents clinical Barrett’s esophagitis (Figs. 73-3 and 73-4).

Figure 73-3. Normal transnasal esophagoscopic appearance of the distal esophagus. The junction of the terminal vessels (short arrows) and the gastric rugae (arrowheads) corresponds to the level of the gastroesophageal junction. The squamocolumnar junction (long arrows) coincides with the gastroesophageal junction in this patient.

(From Postma GN, Belafsky PC, Aviv JE. Atlas of Transnasal Esophagoscopy. Philadelphia: Lippincott Williams & Wilkins; 2007:18).

Figure 73-4. Retroflexed view of the squamocolumnar junction (arrows). The diaphragm is tight around the endoscope.

Indications

The rate at which pathology is found on TNE performed in an otolaryngology practice is high, approaching 50%.^15,28 Indications for TNE can be divided into three major categories: esophageal, extraesophageal, and procedure-related. Esophageal indications include dysphagia, refractory or long-standing gastroesophageal reflux, abnormal radiographic study, and screening for Barrett’s metaplasia. Extraesophageal indications include globus pharyngeus, head and neck cancer, laryngopharyngeal reflux, and chronic cough. Reavis and colleagues²⁹ found that patients with esophageal adenocarcinoma exhibited extraesophageal reflux symptoms more frequently than typical gastroesophageal reflux complaints, with chronic cough being the most common single complaint.²⁹ Procedures for which TNE may be used are detailed in the following sections.

Transnasal Esophagoscopy in Barrett’s Esophagus

Although most cancers in the United States are showing a decline in prevalence, the rate of esophageal adenocarcinoma is on the rise. Barrett’s esophagus represents a premalignant state for this type of esophageal neoplasm. TNE is a useful screening examination for this condition. Limitations of TNE include the inability to use jumbo forceps or comfortably to obtain a full four-quadrant range of biopsy specimens spaced every 1 to 2 cm, which are standard during conventional Barrett’s esophagus surveillance. Therefore, TNE is not currently recommended for the surveillance of Barrett’s esophagus (Fig. 73-5).

Figure 73-5. Barrett’s esophagus. Tongues of gastric columnar epithelium can be seen extending into the squamous epithelium of the mid-esophagus.

Transnasal Esophagoscopy in Dysphagia

Dysphagia is a growing clinical concern, particularly among the elderly. TNE complements other diagnostic tools, including manometry, flexible endoscopic evaluation of swallow (FEES), flexible endoscopic evaluation of swallowing with sensory testing (FEESST), and radiographic studies such as the barium esophagram and the modified barium swallow. Unlike these other studies, TNE allows direct visual inspection of the mucosa. An additional advantage of this unsedated examination is the ability to subjectively assess peristaltic activity. A pill, food, or a liquid can be swallowed and observed to travel through the esophagus, in order to localize areas of decreased peristalsis or narrowing. Transit time for a liquid is usually less than 13 seconds. Conditions found during TNE for dysphagia include esophagitis, strictures, rings, large hiatal hernias, webs, primary or metastatic neoplasms, eosinophilic esophagitis, vascular anomalies, and diverticula. A tight lower esophageal stricture found on endoscopy can also suggest the diagnosis of achalasia rather than the presence of a tumor (pseudoachalasia) (Fig. 73-6).

Figure 73-6. Esophageal ring. The type B ring (Schatzki’s ring) represents the most common cause of dysphagia to solids in adults.

Procedures

Stricture Dilation

When patients complain of a primarily solid rather than liquid dysphagia, an anatomic etiology is often present, such as a stricture or web. Traditionally, the treatment of esophageal strictures involves mechanical dilation, often with bougies, in the operating room or with sedated endoscopy. Patients may need multiple treatment sessions. The need for numerous anesthetics and “blind” dilation is avoided with the use of unsedated TNE.³⁶ Controlled radial expansion (CRE) balloons can be passed over a guidewire to dilate narrowed segments.^37,38 Evidence suggests that these balloons may be less traumatic than bougies.³⁹ Unsedated TNE has been used to successfully dilate esophageal strictures in conditions such as epidermolysis bullosa as well as those induced by reflux or radiation.^40,41 The endoscope can also guide placement of controlled radial expansion balloons to treat radiation-induced stenosis of the nasopharynx. Neopharyngeal strictures that may occur after total laryngectomy have been successfully dilated with the use of office unsedated TNE as well (Fig. 73-7).⁴²

Figure 73-7. Balloon dilation of an esophageal stricture. A partially inflated controlled radial expansion balloon can be seen within an esophageal stricture.

Complications

One of the most feared complications of esophagoscopy is esophageal perforation. Among the thousands of cases performed, only one case of esophageal perforation has been reported, when unsedated TNE was performed by a gastroenterologist.¹⁶ There have been no reported cases of esophageal perforation during TNE performed by an otolaryngologist. Minor complications have been reported. In two of the largest series reported, involving 700 patients from the United States and 1100 patients from France, rates of epistaxis were reported between 0.85% and 2% and those of vasovagal events to be 0.3%.^15,20 Mild, self-resolving nasal discomfort was reported at 1.6%.

Conclusion

Transnasal esophagoscopy is a well tolerated, safe, and efficacious method to evaluate the esophagus without many of the risks associated with conventional endoscopy. Better patient care along with a significant potential cost savings can be achieved with use of TNE rather than the conventional procedure.