**3. Endoscopic findings**

The fact that EoE was first identified only 30 years ago is indicative of the frequently subtle and unspecific endoscopic finding present in most patients. In fact, common esophageal diseases causing dysphagia are usually characterized by evident endoscopic lesions, such us peptic erosions, ulcers, protruding masses or stenosis, all of which contrast with the relatively minor findings exhibited by the majority of patients with EoE. Successful diagnosis of the disease thus requires a high level of suspicion on the part of the clinician, who should perform a careful examination of the esophagus accompanied by mucosal biopsies, even if the mucosa appears to be normal.

Many patients diagnosed with EoE have had previous endoscopies for dysphagia or food impaction and received different diagnoses. In fact, one study reported that the average adult EoE patient underwent two endoscopic exams before being diagnosed correctly (Lucendo et al., 2007). Esophageal symptoms in these patients are frequently attributed to various causes, with some reported cases receiving referrals to mental health professionals because psychological rather than physiological disorders were suspected.

Only in past few years has EoE been extensively recognized, leaving behind its status as a broadly misdiagnosed disease (Gonsalves et al., 2005) to become the second cause of chronic esophagitis.

One important stumbling block to determining the most effective treatment for EoE is the lack of studies directly comparing different treatment strategies for the disease. Such studies

In this chapter we review the various endoscopic lesions described in EoE to date. This should help relatively inexperienced endoscopists screen for patients suspected of having EoE. We will also discuss the effects and risks of endoscopic treatment by dilation in EoE

EoE is a clinico-pathological disease characterized by symptoms related to esophageal dysfunction. Up to now, esophageal biopsies have been essential for making a diagnosis. For optimal pathological evaluation, multiple biopsies from the proximal and distal esophagus should be obtained and evaluated for a variety of pathological features, the most characteristic being an eosinophil-predominant inflammation with a minimum threshold of 15 eosinophils/high power field (hpf). However other accompanying findings reinforce the diagnosis and should also be noted by the pathologist. These include: eosinophilic microabscesses, surface layering of eosinophils, extracellular eosinophilic deposits, basal cell

The effects of EoE are isolated to the esophagus; therefore, eosinophilic inflammation should be absent from both gastric and duodenal biopsy samples (Lucendo, 2010). Furthermore, other causes of esophageal eosinophilia should be excluded, specifically gastroesophageal reflux disease (GERD). This can effectively be excluded if there is a normal pH-metry or if eosinophils persist after treatment with full doses of proton pump inhibitors (PPI). However, the prevalence of patients suffering from both EoE and GERD make the PPI trial

The fact that EoE was first identified only 30 years ago is indicative of the frequently subtle and unspecific endoscopic finding present in most patients. In fact, common esophageal diseases causing dysphagia are usually characterized by evident endoscopic lesions, such us peptic erosions, ulcers, protruding masses or stenosis, all of which contrast with the relatively minor findings exhibited by the majority of patients with EoE. Successful diagnosis of the disease thus requires a high level of suspicion on the part of the clinician, who should perform a careful examination of the esophagus accompanied by mucosal

Many patients diagnosed with EoE have had previous endoscopies for dysphagia or food impaction and received different diagnoses. In fact, one study reported that the average adult EoE patient underwent two endoscopic exams before being diagnosed correctly (Lucendo et al., 2007). Esophageal symptoms in these patients are frequently attributed to various causes, with some reported cases receiving referrals to mental health professionals

Only in past few years has EoE been extensively recognized, leaving behind its status as a broadly misdiagnosed disease (Gonsalves et al., 2005) to become the second cause of chronic

because psychological rather than physiological disorders were suspected.

hyperplasia, intercellular edema, and lamina propria fibrosis (Furuta et al., 2007).

the method of choice for diagnosing EoE in these cases (Molina-Infante et al., 2009).

will be necessary before the best therapeutic option for EoE can be established.

patients by reviewing the current literature.

**2. Diagnosis** 

**3. Endoscopic findings** 

esophagitis.

biopsies, even if the mucosa appears to be normal.

Endoscopy with esophageal biopsy remains the only reliable diagnostic test for EoE. Consequently, in order for clinicians to recognize the disease more easily, a better awareness of the distinct endoscopic features of EoE is essential. Retrospective re-evaluations of the endoscopic appearance of the esophagus in those patients eventually diagnosed with EoE have revealed that esophageal appearance had been described as normal in between one quarter to one third of the cases (Müller et al., 2007; Sgouros et al., 2006; Liacouras et al., 2005). It is important to note, however, that even though the endoscopic findings are subtle, remarkable abnormalities can still be detected in the majority of patients, as we describe below.

Endoscopy has helped identify a great number of esophageal abnormalities in patients with EoE. These include fixed esophageal rings that sometimes reduce the esophageal lumen (a phenomenon known as trachealization) and transient esophageal rings (also called feline folds or felinization). Diffuse nodularity/granularity of the mucosa has also been described, along with widespread exudative mucosal lesions, either in the form of whitish papulae of varying sizes clustered together (white spots) or as large, white, exudative fibrinoid lesions. These whitish lesions on the mucosa resemble a mild, superficial Candida infection, but histopathology shows micro-abscesses made up of eosinophils (Lucendo et al., 2007). Furthermore, a loss of the common vascular pattern of the mucosa has been described (Lucendo, 2007; Straumann et al., 2004). Some of the most common findings are longitudinal furrows (referred to as "corrugated esophagus," which is an architectural analogy to a grooved column) (Straumann et al., 2004), diffuse esophageal narrowing, and esophageal lacerations induced by passage of the endoscope. Mucosal fragility, also called *crêpe-paper* mucosa (Straumann et al., 2003), is an important feature of this pathology as it may cause tears during upper endoscopy or even if the patient tries to dislodge impacted food by inducing vomiting (Lucendo et al., 2011). However, because all of these endoscopic features have been described in other esophageal disorders, none can be considered pathognomonic for EoE.

To shed light on the varied endoscopic appearances of EoE, we have classified them according to two independent yet complementary aspects: alterations in the caliber of the esophagus and alterations in the appearance of the mucosa (Lucendo et al., 2007).


Endoscopic Aspects of Eosinophilic Esophagitis: From Diagnosis to Therapy 67

Fig. 2. Several endoscopic aspects of eosinophilic esophagitis: a: Normal- caliber esophagus with a normal appearance mucosal surface; b: Fragile-looking mucosa, with irregular surface and whitish exudates; c: Reduced-caliber, trachealized esophagus with regular mucosal surface, which allows the passage of the endoscope; d: Longitudinal linear furrows and irregular mucosa; e: The esophageal mucosal surface may be covered in cotton-like exudates mimicking candiadiasis, but biopsy finds them to be multiple eosinophil-

containing micro-abscesses; f: Ringed esophagus with stenosis blocking the passage of the

In any case, EoE seems to be a very common cause of dysphagia, with a prevalence of up to 22% in patients with the non-obstructive version of this condition (Ricker, 2001). In addition, its incidence rates are significantly higher in men than in women and also in those of European descent than for other ethnicities. These findings underscore the importance of

Because the reliability of endoscopic findings alone for diagnosing EoE does not appear to exceed 40%, few studies deal with finding ways to improve the diagnostic efficiency of endoscopy. Indeed, only one published study has examined the ability of narrow-band imaging (NBI) endoscopy to improve reliability. While this technique proved helpful in detecting mucosal details that go unnoticed in a routine white-light examination, it only managed to identify rings and furrows with fair to good reliability; no other findings were noted. Moreover, there was also great interobserver variability. The researchers thus concluded that endoscopic findings alone were not sufficiently reliable for supporting a

As we have seen, none of the endoscopic features described above is pathognomonic for EoE; however, the presence of more than one of them in a given patient bolsters the case for a diagnosis of EoE. It is our hope that a greater awareness of these subtle characteristics will help clinicians avoid overlooking them to more accurately diagnose patients. Of course, any preliminary diagnosis must then be confirmed through biopsies. Indeed, as we emphasized above, biopsy sampling should also be performed in cases of non-obstructive dysphagia,

performing routine biopsies to screen for EoE in these patients (Ricker, 2011).

diagnosis of EoE or for making treatment decisions (Peery, 2011).

even when the esophagus appears normal.

endoscope

density of eosinophils, and cell activation as determined with the aid of immunostaining for Major Basic Protein (MBP) (Lucendo et al., 2007), it was observed that the density of eosinophils increased with the severity of histological changes. Qualitative analysis of the patient biopsies showed a correlation between the intensity of histopathological changes and the diverse patterns of findings from endoscopic exploration of the mucosa. Consequently, four endoscopic-histopathological patters were defined: 1. Granular pattern: mucosa with relatively defined papular elevations that give it an irregular shape. Histological analysis highlighted changes in eosinophilic infiltration and derived damage, with different intensities in different areas, which implies possible mucosal effects that are not uniform in intensity. 2. Corrugated pattern: linear longitudinal ridges or striae along the folds of the esophagus also affected by mucosal edema. The histology identified edema with growth in intercellular spaces between the epithelial cells, ballooned cells, and spongiosis. 3. Undulated pattern: this may denote contraction of the muscularis mucosae (not evaluable in endoscopic biopsies). It should not be mistaken for simultaneous contraction rings in the internal or circumferential layer of the esophageal muscularis propia, which in this case reduce the size of the lumen. 4. Exudative pattern: different-sized whitish lesions (from slight spotting to squamous lesions), creating epithelial clusters or microabscesses containing eosinophils. These patients had a high density of eosinophils on the surface of the esophagus, destruction and detachment of the most superficial strata, and more intense immunostaining for MBP.

Fig. 1. Images from two patients with eosinophilic esophagitis and spontaneous esophageal tearing, withring disruption. This occurred as a result of the efforts the patients made to induce vomiting and dislodge impacted food

Several prospective studies have evaluated the utility of endoscopic findings for diagnosing EoE. In 2007, G.A. Prasad and co-workers successfully used endoscopy in conjunction with esophageal biopsies to diagnosis EoE in 15% of 222 patients who were being attended for non-obstructive dysphagia (Prasad et al., 2007). Of the 21 patients who exhibited endoscopic results characteristic of EoE, the diagnosis was confirmed in only 8 cases (38%). However, 10 of the 102 patients (9,8%) with an apparently normal endoscopic examination presented histological evidence of EoE. In 2008, S.H. Mackenzie et al. reported similar findings (Mackenzie et al., 2008). Thus, while 12% of the 261 patients suffering from dysphagia who underwent endoscopy were initially diagnosed with EoE, only 12 of 35 patients (34%) who showed esophageal rings in their endoscopic exams were confirmed to have EoE after esophageal biopsy.

Fig. 1. Images from two patients with eosinophilic esophagitis and spontaneous esophageal tearing, withring disruption. This occurred as a result of the efforts the patients made to

Several prospective studies have evaluated the utility of endoscopic findings for diagnosing EoE. In 2007, G.A. Prasad and co-workers successfully used endoscopy in conjunction with esophageal biopsies to diagnosis EoE in 15% of 222 patients who were being attended for non-obstructive dysphagia (Prasad et al., 2007). Of the 21 patients who exhibited endoscopic results characteristic of EoE, the diagnosis was confirmed in only 8 cases (38%). However, 10 of the 102 patients (9,8%) with an apparently normal endoscopic examination presented histological evidence of EoE. In 2008, S.H. Mackenzie et al. reported similar findings (Mackenzie et al., 2008). Thus, while 12% of the 261 patients suffering from dysphagia who underwent endoscopy were initially diagnosed with EoE, only 12 of 35 patients (34%) who showed esophageal rings in their endoscopic exams were confirmed to have EoE after

immunostaining for MBP.

induce vomiting and dislodge impacted food

esophageal biopsy.

density of eosinophils, and cell activation as determined with the aid of immunostaining for Major Basic Protein (MBP) (Lucendo et al., 2007), it was observed that the density of eosinophils increased with the severity of histological changes. Qualitative analysis of the patient biopsies showed a correlation between the intensity of histopathological changes and the diverse patterns of findings from endoscopic exploration of the mucosa. Consequently, four endoscopic-histopathological patters were defined: 1. Granular pattern: mucosa with relatively defined papular elevations that give it an irregular shape. Histological analysis highlighted changes in eosinophilic infiltration and derived damage, with different intensities in different areas, which implies possible mucosal effects that are not uniform in intensity. 2. Corrugated pattern: linear longitudinal ridges or striae along the folds of the esophagus also affected by mucosal edema. The histology identified edema with growth in intercellular spaces between the epithelial cells, ballooned cells, and spongiosis. 3. Undulated pattern: this may denote contraction of the muscularis mucosae (not evaluable in endoscopic biopsies). It should not be mistaken for simultaneous contraction rings in the internal or circumferential layer of the esophageal muscularis propia, which in this case reduce the size of the lumen. 4. Exudative pattern: different-sized whitish lesions (from slight spotting to squamous lesions), creating epithelial clusters or microabscesses containing eosinophils. These patients had a high density of eosinophils on the surface of the esophagus, destruction and detachment of the most superficial strata, and more intense

Fig. 2. Several endoscopic aspects of eosinophilic esophagitis: a: Normal- caliber esophagus with a normal appearance mucosal surface; b: Fragile-looking mucosa, with irregular surface and whitish exudates; c: Reduced-caliber, trachealized esophagus with regular mucosal surface, which allows the passage of the endoscope; d: Longitudinal linear furrows and irregular mucosa; e: The esophageal mucosal surface may be covered in cotton-like exudates mimicking candiadiasis, but biopsy finds them to be multiple eosinophilcontaining micro-abscesses; f: Ringed esophagus with stenosis blocking the passage of the endoscope

In any case, EoE seems to be a very common cause of dysphagia, with a prevalence of up to 22% in patients with the non-obstructive version of this condition (Ricker, 2001). In addition, its incidence rates are significantly higher in men than in women and also in those of European descent than for other ethnicities. These findings underscore the importance of performing routine biopsies to screen for EoE in these patients (Ricker, 2011).

Because the reliability of endoscopic findings alone for diagnosing EoE does not appear to exceed 40%, few studies deal with finding ways to improve the diagnostic efficiency of endoscopy. Indeed, only one published study has examined the ability of narrow-band imaging (NBI) endoscopy to improve reliability. While this technique proved helpful in detecting mucosal details that go unnoticed in a routine white-light examination, it only managed to identify rings and furrows with fair to good reliability; no other findings were noted. Moreover, there was also great interobserver variability. The researchers thus concluded that endoscopic findings alone were not sufficiently reliable for supporting a diagnosis of EoE or for making treatment decisions (Peery, 2011).

As we have seen, none of the endoscopic features described above is pathognomonic for EoE; however, the presence of more than one of them in a given patient bolsters the case for a diagnosis of EoE. It is our hope that a greater awareness of these subtle characteristics will help clinicians avoid overlooking them to more accurately diagnose patients. Of course, any preliminary diagnosis must then be confirmed through biopsies. Indeed, as we emphasized above, biopsy sampling should also be performed in cases of non-obstructive dysphagia, even when the esophagus appears normal.

Endoscopic Aspects of Eosinophilic Esophagitis: From Diagnosis to Therapy 69

• With regard to what constitutes the best therapeutic option for EoE patients, no studies comparing different therapeutic modalities have been carried out. Moreover, several published EoE case studies involve dilation with concomitant drug therapy (either with steroids or montelukast), which makes it difficult to clearly establish the effect of the

• Additionally, esophageal symptoms are frequently intermittent in EoE patients, who can experience prolonged asymptomatic periods despite the persistence of eosinophilic inflammation. This raises doubts about the convenience of restricting therapy to

• Narrowing of the esophageal lumen can originate in two ways: by muscle contractions due to motor disturbances secondary to eosinophilic infiltration of deep esophageal wall structures, or by fibrous structures derived from fibrous remodeling and collagen deposits in the subepithelial strata. A combination of both mechanisms may also be possible. In addition, it is difficult to make routine distinctions between patients who have a definite stricture and those in whom it can be reversed through drug or diet

• A relevant difficulty in assessing the efficacy of individual therapeutic modalities in EoE patients comes from the lack of a validated, commonly accepted score for symptoms in this disease. This makes it difficult not only to extrapolate results from one study to another, but also to objectively evaluate the effect of treatment on clinical manifestations. In this scenario, and with regard to endoscopic treatment, the most

Fig. 3. Concentric esophageal short stricture, with fibrous appearance because of the absence of vascular pattern, before (a) and after (b) endoscopic dilation using a trough-the-scope

valuable criterion for response is the need for repeated dilations.

symptomatic periods only or whether to prescribe a maintenance treatment.

life (Straumann, 2008).

therapy.

individual treatment modalities.

balloon. A deep mucosal tear can be observed

epithelial inflammatory infiltrate. A group of EoE experts have recommended treating asymptomatic cases of EoE to avoid the potential consequences of fibrous remodeling of the organ (Liacouras et al., 2011), although the long-term consequences are not really known. The experience of each center and the availability of techniques and studies also limit the treatment options and the objectives established in each case. However, we should keep in mind that if left untreated, EoE is a chronic disease involving persistent histological inflammation over time, with detrimental effects on a patient's quality-of-
