Red Alert: The Evolution of Eosinophilic Esophagitis as a Distinct Clinicopathologic Syndrome

Introduction: EoE before 1990

Eosinophilic esophagitis (EoE) is a disease characterized by eosinophilic infiltration and characteristic macroscopic changes of the esophageal mucosa accompanied by dysphagia. Now widely recognized by gastroenterologists and allergists, EoE has rapidly become a major cause of upper gastrointestinal morbidity.¹² While still considered a rare disease,³ in practice it seems that more and more cases are diagnosed every week, research in the field is exploding, and the pipeline for treatments contains multiple agents, some of which are quite far along the development pathway.⁴ As dynamic as is the current interest in the disease, this was not always the case as substantial progress has been made over the last 30–40 years.

The first cases of EoE were described in single patient reports of Dobbins and Landres in the late 1970s.^{5, 6} The paper by Dobbins described a patient with eosinophilic gastroenteritis who had esophageal involvement with eosinophilia and symptoms of dysphagia in the setting of concomitant atopy. In contrast, the paper by Landres related a case of eosinophilic esophagitis in the setting of vigorous achalasia. This patient had solid food dysphagia as well as epigastric pain, prior gastric surgery for ulcer disease, and peripheral eosinophilia. Neither case would be considered a “classic” or straightforward EoE case today, and after these papers, there were only scattered case reports published over the next 15 years. ^7–10 At the time, EoE was vanishingly rare, the term “eosinophilic esophagitis” was not regularly used in the literature, and there was no recognition that there might be an allergic component to the disease. In fact, due in large part to a paper by Winter and colleagues, eosinophils noted in esophageal biopsies were generally attributed to gastroesophageal reflux disease (GERD).¹¹ Since the presence of these cells did not appear to impact diagnosis or management, routine esophageal biopsy was not incorporated into adult GI practice, with eosinophils often equated with reflux in children. In retrospect, reports of the corrugated esophagus or feline esophagus may have been cases of EoE that went unrecognized at the time.^12–16

In this setting came the seminal report by Stephen Attwood, Thomas Smyrk, Tom DeMeester, and James Jones, published in Digestive Diseases and Sciences in 1993 and since cited more than 800 times (Google Scholar), that described a clinicopathologic syndrome of esophageal eosinophilia with dysphagia.¹⁷ This paper, along with a report by Alex Straumman and colleagues published contemporaneously in 1994 in Schweizersche medizinische Wochenschrift (also highly cited),¹⁸ essentially defined the presentation of the syndrome that ultimately became known as EoE. In examining these papers, it is fascinating how many of the cardinal features of the disease are noted that set the stage for the future work in the field.

This review will detail the origins of this paper and compare and contrast what was observed then and what is known know about multiple aspects of EoE, including the clinical presentation, diagnosis, epidemiology, natural history, and treatments and outcomes. Moreover, it will highlight how these papers presaged a number of controversies in the field that have yet to be resolved, as well as foreshadowed the collaborative, multidisciplinary approach that has led to rapid advances.

Description of EoE as a clinicopathologic syndrome

In 1987, Dr. Attwood was working as a research fellow in the laboratory of Dr. DeMeester on a project using a model system in order to examine the relation between reflux and esophageal cancer. Dr. Attwood recounts that he wanted to work on a clinical project as well, and connected with Dr. Smyrk, a junior pathologist at the time, who had noticed that “patients with eosinophils in the esophageal biopsies had a different clinical pattern to the usual reflux patients being discussed at the weekly GI meetings”. The team was completed by James Jones, a pharmacy student at the time working as a research assistant. Dr. Attwood was then able to examine the full clinical data on these patients, including endoscopy, pH, and manometry results, and identified a set of patients who had > 20 eosinophils per high-power field of the esophageal mucosa (eos/hpf), severe dysphagia, and normal pH testing, who did not respond to antacids. Dr. Attwood recalls that he chose a cut-off of 20 eos/hpf since this statistically separated this new patient group and patients with established reflux. He also noted that “if not for Dr DeMeester’s meticulous and methodical investigations and record keeping, and the very active and respectful relationship he had with Tom Smyrk, this work would not have been possible. Even though this study was not directly related to Dr. DeMeester’s reflux work and suggested that eosinophils might not be exclusively related to reflux, Dr. Attwood goes on to say that “the man that [Dr. DeMeester] is was highlighted by the courteous attitude he always took to my academic arguments – despite me being the most junior of research fellows.” The original work from the study was presented at the British Society of Gastroenterology in 1989, and then published in this journal in 1993. Six months later, Dr. Straumann independently reported his own cohort of similar patients and used the term eosinophilic esophagitis.

Then and now – clinical presentation and diagnosis

Clinical presentation

In the Attwood paper, the 12 patients identified with >20 eos/hpf had a mean age of 32, a male predominance (10:2), and dysphagia as the cardinal clinical symptom. Four required treatment for an esophageal food bolus impaction (EFBI), and 3 had associated odynophagia. Median length of symptoms before evaluation was 3 years, with a range of up to 12 years. Furthermore, though heartburn symptoms were uncommon, atopy was present in the majority (7) including 3 with asthma and 2 with chronic sinusitis. The presentations were remarkably similar in the Straumann paper, with dysphagia-predominant symptoms in 10 patients, predominantly males with an atopic predisposition.

This presentation is now what is recognized as most typical for adolescents and adults with EoE. ^19–22 Symptoms are dysphagia predominant, with EFBI requiring emergency endoscopy, but can also include chest pain or heartburn. The primacy of dysphagia in this population has led to the development of validated patient-reported outcome (PRO) measures of this symptom that have been used in clinical trials. ^{23, 24}As pediatric reports were published, the different set of nonspecific symptoms (feeding dysfunction, failure to thrive, abdominal pain, vomiting, regurgitation/reflux) was recognized^25–27 and a validated pediatric symptom score was developed.²⁸ Particularly interesting was the long duration of symptoms prior to diagnosis,²⁹ which has since be confirmed and associated with likely disease progression from an inflammatory to fibrotic phenotype.^30–34 There is also male predominance for reasons that remain unclear; the disease seems less common in non-white patients, though this may be related to surveillance and detection bias and remains an area of active investigation.^{3, 35–37} The high rate of atopy, also a cardinal feature of EoE, was quickly confirmed in pediatric populations before also being widely recognized in adults as well.^38–42

Endoscopic findings

In the Attwood paper, endoscopic findings were underwhelming. None were noted to have Barrett’s esophagus, and the authors state that the “striking feature of these patients [is] dysphagia out of proportion to the endoscopic and roentgenographic findings.” Despite this, empiric dilation was performed with symptom improvement although Dr. Attwood recalls that in 1987, these patients had been examined with fiberoptic endoscopes. Assessment of endoscopic findings for this study consisted of a retrospective review of archived hardcopy photographs, which reduced the sensitivity for detecting mucosal pathology. In contrast, the Straumann report did identify “discrete white structures” in the esophagus, as well as one patient with a web and one with an esophageal ring.

Today, particularly with the use of high resolution and high definition video endoscopes, the hallmark endoscopic features of EoE may be so striking as they define the disease for some providers.⁴³ Esophageal changes including exudates, rings, edema, furrows, strictures, narrowing, and crêpe-paper mucosa are widely known and recognized,^44–48 and new features, like the pull sign,⁴⁹ continue to be reported. The EoE Endoscopic Reference Score (EREFS) is a validated measure of endoscopic severity⁵⁰ that is responsive to treatment according to multiple EoE clinical trials,^51–56 and has been incorporated into standard endoscopic reporting software. Though these endoscopic findings were previously thought to be nonspecific, newer work shows that the combination of features and an increasing EREFS score is highly suggestive of EoE, with only a minority of patients undergoing a careful esophageal exam having an endoscopically normal-appearing esophagus.^57–59

Histologic analysis

On histologic examination, Dr. Smyrk noted that esophageal biopsies “were remarkable for their dense tissue eosinophilia and marked squamous hyperplasia”. The peak eosinophil counts were quantified with a mean of 56 (range 21–110) eos/hpf with diffuse epithelial distribution and some surface clustering. Moreover, eosinophil degranulation and basal zone hyperplasia with elongated papillary height persisted in the 5 patients with repeat biopsies. Interestingly, while the submucosa of most patients did not have pathologic findings, some patients did have eosinophilia in these deeper levels. Similar findings, with the additional note of eosinophilic microabscesses, were reported in the paper by Straumann and colleagues.

These histologic descriptions are highly consistent with the current well-known features of biopsies obtained from EoE patients.⁶⁰ In fact, the methods for histopathologic assessment in this paper are quite typical for what has become the standard for both clinical and research work in EoE. Slides were re-examined with a specific protocol in order to assess the peak eosinophil count in the most inflamed hpf. The equivalent dimensions of the hpf were provided (0. 5mm diameter; 0. 196mm²) enabling the calculation of the eosinophil density expressed as eosinophil/mm². Associated findings of degranulation, microabscesses, basal zone hyperplasia, and others abnormalities were noted. Since this time, assessment of eosinophil counts by pathologists is reproducible,^{61, 62} variability in counts by using microscopes with different hpf sizes has been noted,⁶³ the optimal number of biopsies has been established,^64–66 and the importance of basal zone hyperplasia,⁶⁷ lamina propria fibrosis,^68–70 and the potential for examination of submucosal tissue has been reported.^{71, 72} Most recently, a validated histologic scoring system (the EoEHSS) has been developed that assesses 8 histologic features of EoE and calculates grade (severity) and stage (extent) scores.⁷³ This measure is more comprehensive than the peak eosinophil count alone, can be reliably assessed, and is highly responsive to treatment in clinical trials. ^{51, 55, 56, 74}

Diagnosis

When the 1993 paper was written, not only was there no consensus on how to diagnose EoE, it would take an additional 14 years for the first guidelines to be published (Figure 1).⁷⁵ The threshold that Attwood and colleagues used, 20 eos/hpf, was just one of many histologic cut-points in the literature, with consequent substantial diagnostic variability.⁶³ The 2007 guidelines brought a measure of consistency to the field with the agreement on ≥15 eos/hpf as the key histologic threshold,⁷⁵ with this value increasingly used after this date.⁷⁶ The 2011 guidelines introduced the conceptual definition of EoE as a “chronic, immune/antigen-mediated esophageal disease characterized clinically by symptoms of related to esophageal dysfunction and histologically by eosinophil-predominant inflammation.”⁷⁷ Importantly, this idea of a clinicopathologic disorder is directly espoused in the title of the Attwood paper nearly 2 decades prior, as they termed their description “a distinct clinicopathologic syndrome”. Ever since then, the combination of symptoms and histologic abnormalities has been key not only for diagnosis but for assessing treatment outcomes.⁷⁸ Additional iterations of guidelines grappled with the contribution of proton pump inhibitor (PPI) therapy and PPI non-response, as well as potential overlap with GERD,^79–81 all issues raised in this first report (and discussed in more detail below). The current guidelines hold that EoE is diagnosed with a suggestive clinical presentation, esophageal eosinophilia ≥15 eos/hpf (approximately 60 eos/mm²), and after exclusion of disorders unrelated to EoE that could cause or potentially contribute to esophageal eosinophilia. ⁸²

Figure 1.

Timeline of the Attwood and Straumann papers, and subsequent EoE diagnosis and management guidelines.

Then and now – epidemiology and natural history

As noted above, both the Attwood and Straumann papers highlighted the key epidemiologic features of the EoE population: younger, male- and white-predominant, and atopic. Nonetheless, apart from noting that the histologic features persist, little was known at the time in terms of EoE risk factors, incidence or prevalence, or natural history.

Since then, the incidence and prevalence of EoE have substantially increased throughout the world,³ with an estimated global incidence and prevalence of 4.4/100,000 and 34.4/100,000, respectively.⁸³ Notably, the increase in incidence and prevalence have outpaced any increase related to change in endoscopy or biopsy rates or increased disease recognition.^84–86 Though potential genetic predispositions to EoE have been identified,⁸⁷ including a number of susceptibility loci,⁸⁸ these rapid changes in case numbers suggest environmental factors are at play.^89–91 There continues to be active investigation into risk factors for EoE; while no one trigger explains the rising incidence, multiple associations have been reported. These include early life factors (maternal infection, cæsarean delivery, neonatal intensive care admission, antibiotics in infancy, acid suppressants in infancy, absence of furred pets),^92–96 seasonal changes,^97–99 arid or cold climates,¹⁰⁰ low population density,¹⁰¹ certain housing components,¹⁰² and possible infections.^{103, 104} Importantly, there are likely gene-environment interactions that predispose to EoE,¹⁰⁵; an underlying epithelial barrier defect, in the setting of an appropriate exposure, may facilitate the development of EoE. ^106–109

Natural history studies as well as placebo arms from trials have now clearly demonstrated that EoE is chronic.^{3, 110} Moreover, there is no current known cure. If any treatment is stopped, the disease nearly universally flares,^111–117 and unlike other allergic conditions, it is exceedingly rare for children to “grow out of” EoE.¹¹⁸ EoE has also been proposed as the last step in the “atopic march.”¹¹⁹ The model of the natural history of EoE also holds that for many patients EoE progresses from an inflammatory-predominant to a fibrostenotic predominant disease.^{3, 120} The longer symptoms persist prior to diagnosis (as a proxy for untreated inflammation), the more likely that a stricture or narrowing will be present.^30–34 The risk of stricture development has been estimated to double for every 10 years of symptoms prior to diagnosis. ³¹

Then and now – treatment and outcomes

Though treatment and outcomes were not the focus of the Attwood paper, interesting data are provided that all 12 EoE patients were primarily treated with bougie dilation, with resultant initial symptom response, but with the observation that “dysphagia usually recurred in three to six months and responded to a second dilation”. Furthermore, there was one patient who received a course of prednisone, also with initial improvement in dysphagia, but with subsequent recurrence of symptoms after prednisone was stopped. The Straumann paper also suggested treatment with steroids would be successful, though his group’s future work on the natural history of EoE focused on dilation therapy for symptom relief.^{110, 121} Therefore, these papers presage two of the major treatments used for EoE: steroids and esophageal dilation. Over the ensuing years, these approaches have become more refined, additional pharmacologic agents have been studied, and the effectiveness of PPIs debated (see controversy section, below). Dietary elimination has been added with the recognition that EoE is often a food antigen-driven process (Figure 2).²⁰ The treatment approach to EoE has recently been updated in evidence-based guidelines from the AGA Institute and Joint Task Force on Allergy-Immunology Practice Parameters. ^{122, 123}

Figure 2.

Timeline of key advances in treatment of eosinophilic esophagitis.

Corticosteroids

While the use of a systemic corticosteroids can be effective for EoE, as demonstrated both in case series and a randomized trial,^{124, 125} the adverse effects of short- and long-term systemic steroid use preclude routine use of prednisone and similar agents. Instead, the concept of swallowed topical steroids has come to the fore. In first reports, patients used asthma preparations, including fluticasone, which were sprayed into the mouth and swallowed, rather than inhaled.^{126, 127} This was followed by reports of aqueous budesonide (meant to be used in asthma nebulizers) being mixed into a slurry and then swallowed.^{128, 129} While both fluticasone and budesonide have been the primary topical steroids studied, other agents such as mometasone, beclomethasone, and ciclesonide,^{126, 130–132} and different delivery methods,^133–137 have been reported. Multiple clinical trials of fluticasone and budesonide have clearly demonstrated efficacy,^138–143 and with the recognition that optimizing esophageal contact or dwell time improves outcomes,¹⁴⁴ there has been substantial interest in developing esophageal-specific formulations of EoE. These now include a dissolvable budesonide tablet (the first medication to be approved for EoE, with initial approval in Europe),^{53, 115, 145} a budesonide oral solution,^{51, 146–148} and a fluticasone dissolvable tablet.⁵⁴ Swallowed/topical steroids are now considered a first-line treatment for EoE;^{122, 123} comparative effectiveness data suggest that the traditional formulations of fluticasone and budesonide are roughly equivalent.⁵²

Elimination diets

The initial report of dietary elimination treatment for EoE, by Kelly and colleagues,¹⁴⁹ followed quickly on the heels of the Attwood and Straumann papers. This paper, which focused on a similar population of symptomatic patients who were refractory to GERD treatment and had marked esophageal eosinophilia, demonstrated that an elemental formula lacking any food antigens was an effective treatment for EoE. This also helped define EoE as a food allergy-driven disease; when food antigens were removed, disease activity was mitigated. The paper was followed by larger cohort studies demonstrating the efficacy of the elemental formula approach in both children and adults,^150–152 but also of allergy test-directed diets¹⁵³ and less restrictive diets, in particular the so-called six-food elimination diet, which removed dairy, wheat, egg, soy, nuts, and seafood.^{154, 155} Data were slower to develop in adults, but when these were published, Gonsalves and colleagues published a paper on the six food diet by,¹⁵⁶ that reported that the same approach was effective in adults. ^157–161More recent work has focused on the utility of less restrictive diets (such as the four, two, or one food diet)^162–166 and has culminated in comparative effectiveness randomized trials. ¹⁶⁷ Dietary therapy is now a mainstay of EoE treatment, considered a first-line treatment option. ^{122, 123}

Esophageal dilation

Though esophageal dilation was the “first” treatment described for EoE in the Attwood paper, it was soon to be controversial in the EoE field. Though strictures were associated with and caused by EoE, many early reports of esophageal dilation included the complications of perforations and rents, often requiring hospitalization.^168–172 This was concerning enough that strong caution was recommended in the 2007 EoE guidelines, with dilation reserved “whenever possible” for after medical or dietary therapy.⁷⁵ Since then, a large experience with esophageal dilation was accumulated in the literature.^{121, 173–177} As in the Attwood paper, it was noted to provide long-lasting symptom relief (median of > 1 year), but not impact the underlying inflammation.¹²¹ Stricture recurrence was felt to be due to persistent inflammation, with subsequent data confirming that if histologic remission is achieved, dilations can be performed less frequently.¹⁷⁸ Dilation is now considered to be safe and effective for symptoms according to the most recent guidelines,^{122, 123} with meta-analysis revealing a complication rate similar to that expected for esophageal dilation for non-EoE conditions.¹⁷⁹ The caveat with dilation is that a cautious approach should be used,¹⁸⁰ where the esophageal caliber is estimated and dilation performed with a slow increase in size and with relook demonstrating adequate dilation effect or until resistance is encountered. Several sessions are often required to achieve a goal diameter, usually 16–18mm. Notably, dilation has also been safely performed in children,^{181, 182} a population generally thought to have a lesser stricture burden than adults.⁴⁴ In sum, esophageal dilation is an important therapeutic option for EoE that addresses the fibrostenotic aspects of the disease, and can be safely performed as a complementary treatment in EoE patients with fibrostenotic changes.

Emerging treatments

While a number of other treatments for EoE have been described, including the use of montelukast (by Attwood and colleagues, in addition to others),^183–185 cromolyn sodium,^{25, 186} and immunomodulators,¹⁸⁷ with increasing knowledge about EoE disease pathogenesis the recent emphasis has been on biologic agents targeting key allergic pathways in EoE. Although this topic has been recently reviewed elsewhere,⁴ it is a measure of the increasing knowledge in the field over the past 30 years that such agents are in advanced clinical development. These include monoclonal antibodies targeting the IL-4 receptor,⁵⁶ IL-13,^{55, 188} IL-5,^189–191 the IL-5 receptor,¹⁹² siglec-8,¹⁹³ and others. Initial reports show promising efficacy and safety, including in patients refractory to other EoE treatments.

Then and now – controversies

One of the most interesting aspects of the Attwood paper is how it raises and anticipates controversies in the EoE field that remain matters of debate to this day. The most evident one is the importance of reflux, pH testing, and PPI use. One of the key features of the paper is that patients had a high density of intraepithelial eosinophils and persistent symptoms despite antireflux treatment and with normal acid exposure as assessed by 24-hour pH studies. The comparative population of 90 reflux patients included in the study had similar symptoms but positive pH monitoring, approximately half had no intraepithelial eosinophilia, and half had low grade (< 20 eos/hpf, but with a mean of 3.3). These patients, and other patients in similar future reports, became the basis for the recommendation in the 2007 guidelines that in order to diagnose EoE, patients must have either non-response to PPI treatment or negative reflux testing. ⁷⁵ Yet, it quickly became clear not only was there a complex relationship between EoE and GERD,¹⁹⁴ but that some patients who appeared to have EoE (and not reflux) had a complete response to PPI treatment.^195–202 This latter group was exemplified by the questioning title of a 2006 paper: “Eosinophils in the esophagus – peptic or allergic eosinophilic esophagitis?”.¹⁹⁵ The concept of PPI-responsive esophageal eosinophilia was developed and incorporated into the 2011 and 2013 guidelines.^{77, 79} The controversy was more directly addressed in the 2017 and 2018 guidelines where PPIs were removed from the diagnostic algorithm and repositioned as an EoE treatment.^{81, 82} Meta-analysis data show that since up to half of patients will have a histologic response to PPI treatment, and even more will have a symptomatic response, PPIs are now considered to be an important first-line option for EoE treatment and not a diagnostic hurdle.²⁰¹ In practice, however, distinguishing EoE from GERD, as well as selecting the appropriate treatment for patients with overlapping EoE and GERD, remains challenging.

The influence of dysmotility in EoE is also controversial. EoE patients may have dysmotility not only from fibrosis of the esophagus but because inflammatory factors (including those produced by mast cells) can directly alter smooth muscle function.^{203, 204} The subjects in the Attwood paper had esophageal motility carefully characterized by manometric testing. Two patients had diffuse spasm and two had nutcracker esophagus, whereas three had low contraction amplitudes and four had short duration contractions; notably, lower esophageal sphincter relaxation was normal in all. In the 90 reflux comparator patients, about half had hypotensive lower sphincters but only 2 had nutcracker and 1 had achalasia. These data were suggestive that perhaps EoE was associated with motor abnormalities. In fact, as noted above, one of the first cases of EoE was in a patient with vigorous achalasia.⁶ Nevertheless, there remains controversy as to whether eosinophilic infiltration, perhaps in the muscle layers of the esophagus (now accessible by EUS-guided biopsy or during per-oral endoscopic myotomy) may predispose to major motility disorders or even cause them, or whether they are an epiphenomenon.²⁰⁵ To support this, there are reports of motility disorders resolving and manometry findings improving after successful EoE treatment.²⁰⁶ Series continue to be published examining motility abnormalities in EoE, now using high resolution manometry, which suggest that there still may be an increased prevalence of dysmotility in EoE, at least when evaluated at some centers.^207–209

Besides these, there are many ongoing controversies in the field. Although it is beyond the scope of this article to review them in detail, some include: the comparative effectiveness of first-line treatments for EoE (PPI, steroids, diet) and which should be used first; how EoE phenotypes and endotypes impact presentation, treatment response, and natural history; what are the best outcome metrics for EoE treatment and how much is knowledge of the deeper layers in the esophagus important; what is the importance of non-eosinophil inflammatory cells for the diagnosis and treatment of EoE; identification of predictors of disease progression; where will biologics (when available) fail in the EoE treatment algorithm; and what is the optimal approach to maintenance therapy and long-term disease monitoring.

Conclusions: Enduring impact of the original descriptions of EoE

After the initial case reports of EoE in the late 1970s and 1980s, the report of EoE as a distinct clinical syndrome by Attwood and colleagues, followed closely by similarly impactful reports from Straumann and colleagues as well as Kelly and colleagues, set the stage for the explosion in knowledge about EoE (Figure 3). It is striking that these articles, none of which report more than 12 patients, are able to succinctly encapsulate what is now recognized to be the classic presentation of EoE (with essential symptoms, demographics, and allergic comorbidities), suggest treatments that have become mainstays, and hint at controversial issues that persist to this day. The paper by Attwood and colleagues further highlights the multidisciplinary nature of EoE. Their team consisted of providers (in this case surgeons) with a focus on gastrointestinal and esophageal conditions, a pathologist, and a pharmacist, most at a junior level of training, with an experienced and encouraging senior mentor. It is not an understatement to say that many of the rapid advances in the field are due to multidisciplinary collaborations in adult and pediatric specialties which now encompass gastroenterology, allergy, pathology, nutrition, pharmacology, surgery, esophageal physiology, in the context of mentoring. Finally, it shows the importance of clinical observation. While there is a tendency in medicine to view large randomized trials as the most impactful clinical research possible, many advances start with clinicians having the creativity and courage to be alert to and explore something new or out of the ordinary rather than ignoring it. In this case, it was prominent esophageal eosinophilia in patients with dysphagia who did not respond to antireflux treatments. The stage was then set for the EoE field to develop and continue to advance in the future.

Figure 3.

Explosion of citations for eosinophilic esophagitis according to PubMed.