Medical treatments for eosinophilic esophagitis

Key messages

We found that while corticosteroids may improve patients' symptoms, they certainly reduce the amount of allergic cells (eosinophils) and they may improve what the disease looks like under visual inspection (endoscopy), when compared to placebo, for children and adults with eosinophilic esophagitis. They may be just as safe as a placebo (dummy treatment).

We found that biologics (a type of treatment that uses substances made from living organisms to treat disease) may improve patients' symptoms, that they certainly reduce the amount of allergic cells, and that they may be no different in terms of what the disease looks like under visual inspection, when compared to placebo, for children and adults with eosinophilic esophagitis. They may be just as safe as a placebo.

What is eosinophilic esophagitis?

Eosinophilic esophagitis is a long-term allergic condition in which the esophagus becomes inflamed (sore), which can lead to difficulty swallowing, vomiting, heartburn, and chest and stomach pain. Particles in foods or the air cause the immune system to have an allergic reaction and produce immune cells, which are called eosinophils. These build up in the esophagus, the tube that connects the mouth with the stomach. Eosinophilic esophagitis was first identified in the 1990s and since then it has been recognized as a major digestive illness. It is not known what causes it, but it might be related to genetics combined with environmental triggers. People with eosinophilic esophagitis tend to have other allergies as well. Currently, there is no cure for eosinophilic esophagitis, making long-term treatment necessary. Standard treatments include diets, stretching of the esophagus (dilation), and drugs such as corticosteroids, biological medications, and proton pump inhibitor medications.

What did we want to find out?

We wanted to find out if the available medical treatments for eosinophilic esophagitis work for improving patients' symptoms, reducing the amount of allergic cells when measured under a microscope, and improving what the disease looks like under visual examination. We also wanted to find out how safe they are and if they improve quality of life.

What did we do?

We searched for randomized controlled trials (studies where people are assigned to one of two or more treatment groups using a random method) comparing any medical treatment for eosinophilic esophagitis with any other medical treatment, in both adults and children.

What did we find?

We found 41 studies with 3253 participants. Eleven studies were in children only while the rest were in a mix of children and adults. We identified 19 comparisons. In this summary, we present the results of the two main comparisons: corticosteroids compared to placebo and biologics compared to placebo.

We found that corticosteroids may be better than placebo at improving patients' symptoms. We are highly certain that corticosteroids are better than placebo at reducing the amount of eosinophils (allergic cells) when measured under a microscope. Corticosteroids may be better than placebo at improving what the disease looks like under visual examination (endoscopy). We also found that people taking corticosteroids may be less likely to leave a study due to unwanted or harmful effects (side effects), and that they probably experience a similar number of both serious side effects and side effects in total, compared to placebo. There may be no difference between corticosteroids and placebo in the improvement of quality of life.

We found that biologics may be better than placebo at improving patients' symptoms. It is likely that biologics are better than placebo at reducing the amount of allergic cells when measured under a microscope. Biologics may be no different to placebo at improving what the disease looks like under visual examination. We also found that people on biologics may be equally likely to leave a study due to side effects, or have serious side effects, and may experience similar numbers of total side effects, compared to placebo. There may be no difference between biologics and placebo in the improvement of quality of life.

What are the limitations of the evidence?

The evidence in children only was quite limited and we do not know if the conclusions above can definitely apply to children specifically. Another limitation of the evidence is that the outcomes were measured in many different ways, which may have weakened our conclusions. Other treatments used by the participants were also something that varied a lot between people, and may have affected our conclusions. Finally, we were limited in the conclusions we could make about the effects of sex, age, extent of disease, dosage, and type of corticosteroid or biologic.

How up-to-date is this review?

This review is up-to-date to 3 March 2023.

Authors' conclusions: 

Corticosteroids (as compared to placebo) may lead to clinical symptom improvement when reported both as dichotomous and continuous outcomes, from the primary study definitions. Corticosteroids lead to a large increase in histological improvement (dichotomous outcome) and may increase histological improvement (continuous outcome) when compared to placebo. Corticosteroids may or may not increase endoscopic improvement (depending on whether the outcome is measured dichotomously or continuously). Withdrawals due to adverse events (dichotomous outcome) may occur less frequently when corticosteroids are compared to placebo.

Biologics (as compared to placebo) may not lead to clinical symptom improvement when reported as a dichotomous outcome and may lead to an increase in clinical symptom improvement (as a continuous outcome), from the primary study definitions. Biologics lead to a large increase in histological improvement when reported as a dichotomous outcome, but this is uncertain when reported as a continuous outcome, as compared to placebo. Biologics may not increase endoscopic improvement (dichotomous outcome), but this is uncertain when measured as a continuous outcome. Withdrawals due to adverse events as a dichotomous outcome may occur as frequently when biologics are compared to placebo.

Read the full abstract...
Background: 

Eosinophilic esophagitis (EoE) is a chronic antigen-mediated eosinophilic inflammatory disease isolated to the esophagus. As a clinicopathologic disorder, a diagnosis of EoE requires a constellation of clinical symptoms of esophageal dysfunction and histologic findings (at least 15 eosinophils/high-powered microscope field (eos/hpf)). Current guidelines no longer require the failure of response to proton pump inhibitor medications to establish a diagnosis of EoE, but continue to suggest the exclusion of other etiologies of esophageal eosinophilia.

The treatment goals for EoE are improvement in clinical symptoms, resolution of esophageal eosinophilia and other histologic abnormalities, endoscopic improvement, improved quality of life, improved esophageal function, minimized adverse effects of treatment, and prevention of disease progression and subsequent complications.

Currently, there is no cure for EoE, making long-term treatment necessary. Standard treatment modalities include dietary modifications, esophageal dilation, and pharmacologic therapy. Effective pharmacologic therapies include corticosteroids, rapidly emerging biological therapies, and proton pump inhibitor medications.

Objectives: 

To evaluate the efficacy and safety of medical interventions for people with eosinophilic esophagitis.

Search strategy: 

We searched CENTRAL, MEDLINE, Embase, ClinicalTrials.gov, and WHO ICTRP to 3 March 2023.

Selection criteria: 

Randomized controlled trials (RCTs) comparing any medical intervention or food elimination diet for the treatment of eosinophilic esophagitis, either alone or in combination, to any other intervention (including placebo).

Data collection and analysis: 

Pairs of review authors independently selected studies and conducted data extraction and risk of bias assessment. We expressed outcomes as a risk ratio (RR) and as the mean or standardized mean difference (MD/SMD) with 95% confidence interval (CI). We assessed the certainty of the evidence using GRADE.

Our primary outcomes were: clinical, histological, and endoscopic improvement, and withdrawals due to adverse events. Secondary outcomes were: serious and total adverse events, and quality of life.

Main results: 

We included 41 RCTs with 3253 participants. Eleven studies included pediatric patients while the rest recruited both children and adults. Four studies were in patients with inactive disease while the rest were in patients with active disease. We identified 19 intervention comparisons. In this abstract we present the results of the primary outcomes for the two main comparisons: corticosteroids versus placebo and biologics versus placebo, based on the prespecified outcomes defined of the primary studies.

Fourteen studies compared corticosteroids to placebo for induction of remission and the risk of bias for these studies was mostly low.

Corticosteroids may lead to slightly better clinical improvement (20% higher), measured dichotomously (risk ratio (RR) 1.74, 95% CI 1.08 to 2.80; 6 studies, 583 participants; number needed to treat for an additional beneficial outcome (NNTB) = 4; low certainty), and may lead to slightly better clinical improvement, measured continuously (standard mean difference (SMD) 0.51, 95% CI 0.17 to 0.85; 5 studies, 475 participants; low certainty).

Corticosteroids lead to a large histological improvement (63% higher), measured dichotomously (RR 11.94, 95% CI 6.56 to 21.75; 12 studies, 978 participants; NNTB = 3; high certainty), and may lead to histological improvement, measured continuously (SMD 1.42, 95% CI 1.02 to 1.82; 5 studies, 449 participants; low certainty).

Corticosteroids may lead to little to no endoscopic improvement, measured dichotomously (RR 2.60, 95% CI 0.82 to 8.19; 5 studies, 596 participants; low certainty), and may lead to endoscopic improvement, measured continuously (SMD 1.33, 95% CI 0.59 to 2.08; 5 studies, 596 participants; low certainty).

Corticosteroids may lead to slightly fewer withdrawals due to adverse events (RR 0.64, 95% CI 0.43 to 0.96; 14 studies, 1032 participants; low certainty).

Nine studies compared biologics to placebo for induction of remission.

Biologics may result in little to no difference in clinical improvement, measured dichotomously (RR 1.14, 95% CI 0.85 to 1.52; 5 studies, 410 participants; low certainty), and may result in better clinical improvement, measured continuously (SMD 0.50, 95% CI 0.22 to 0.78; 7 studies, 387 participants; moderate certainty).

Biologics result in better histological improvement (55% higher), measured dichotomously (RR 6.73, 95% CI 2.58 to 17.52; 8 studies, 925 participants; NNTB = 2; moderate certainty). We could not draw conclusions for this outcome when measured continuously (SMD 1.01, 95% CI 0.36 to 1.66; 6 studies, 370 participants; very low certainty).

Biologics may result in little to no difference in endoscopic improvement, measured dichotomously (effect not estimable, low certainty). We cannot draw conclusions for this outcome when measured continuously (SMD 2.79, 95% CI 0.36 to 5.22; 1 study, 11 participants; very low certainty).

There may be no difference in withdrawals due to adverse events (RR 1.55, 95% CI 0.88 to 2.74; 8 studies, 792 participants; low certainty).