Proton pump inhibitor–responsive esophageal eosinophilia: What is the relevance in 2018?

A meaningful discussion on the relevance of proton pump inhibitor–responsive esophageal eosinophilia (PPI-REE) in 2018 should first consider this question: How did we ever end up with this cumbersome term? Well, it arose as an artifact of early confusion regarding the need to distinguish eosinophilic esophagitis (EoE) from gastroesophageal reflux disease (GERD). This confusion led to adoption of an early, unrealistic EoE definition that excluded patients whose condition responded to proton pump inhibitors (PPIs). The term PPI-REE has persisted because physicians have been slow to abandon the traditional, erroneous notion that only an acid peptic disease like GERD responds to PPIs.

EoE is a relatively new disease, first described in 1993, and it has many features that resemble GERD. Consequently, the onus on early investigators of EoE was to prove that they indeed were studying a new disease, not just an unusual manifestation of GERD. Those investigators knew that PPIs were a highly effective treatment for GERD, presumably because they induced profound gastric acid suppression. Indeed, responsiveness to PPIs has been widely regarded as de facto evidence of acid peptic disease in any upper gastrointestinal organ. Because GERD is the only acid peptic disease of the esophagus, response to PPIs seemed like a reliable way to establish a diagnosis of GERD for patients with esophageal symptoms and eosinophilia.

Accordingly, AGA Institute guidelines published in 2007 excluded a diagnosis of EoE for such patients whose disease was responsive to PPIs. Physician recognition of EoE was not yet widespread in 2007, and the AGA guidelines were born of necessity to exclude GERD and convince skeptics of the very existence of the new disease. Nevertheless, the guidelines were unrealistic in implying that GERD and EoE were mutually exclusive disorders. This idea would make sense only if one disorder protected against the other — and that is clearly not the case.

In addition, in 2007, Dr. Bob Genta, Dr. Rhonda Souza and I published a report contending that the interaction between GERD and EoE might be complex, and that the concept of establishing a clear distinction between the two via a PPI trial was far too simplistic. During this same time, investigators began to see patients who had symptoms and findings on endoscopy and esophageal histology typical of EoE but not of GERD and whose disease responded to PPIs nevertheless. What could we call this condition? It didn’t seem to be GERD, even though it responded to PPIs, but it was not EoE because our definition excluded patients responsive to PPIs. So, for lack of a better term, we called the condition PPI-REE.

In 2011, I was a member of a working group that proposed a new, conceptual definition for EoE as an immune-/antigen-mediated esophageal disease characterized clinically by symptoms of esophageal dysfunction and histologically by eosinophil-predominant inflammation. Unlike the 2007 definition, which established primarily what EoE was not (i.e., it was not GERD), I was pleased that this new definition established what EoE was (i.e., immune-/antigen-mediated esophageal disease). However, I was disappointed when the working group still insisted that responsiveness to PPIs excluded a diagnosis of EoE.

By 2011, enough data had accumulated to suggest that PPIs might have important antiinflammatory effects entirely independent of their effects on gastric acid secretion, and that those anti-inflammatory effects might enable PPIs to heal inflammatory disorders beyond acid peptic diseases. A report published by my group in 2012 showed that PPIs blocked the secretion of an eosinophil chemoattractant (eotaxin-3) by esophageal epithelial cells that were stimulated with allergic (Th2) cytokines, and we suggested that this anti-inflammatory mechanism might underlie PPI-REE. We also suggested that some patients with PPI-REE might have subclinical GERD exacerbating antigen-mediated EoE, perhaps through GERD-induced increases in esophageal permeability that enable food antigens to penetrate the epithelium. Thus, antigen-driven EoE might benefit from both the anti-inflammatory and antisecretory effects of PPIs.


The term PPI-REE has persisted because physicians have been slow to abandon the traditional, erroneous notion that only an acid peptic disease like GERD responds to PPIs.


In 2018, the traditional notion that PPIs can only benefit patients with acid peptic disease and not an antigen-driven disease (e.g., EoE) is simply untenable. Irrespective of the mechanism, the evidence is overwhelming that many patients with the antigen-mediated, clinicopathologic syndrome we recognize as EoE also have disease that responds to PPIs. The clinical, endoscopic, histologic and gene-expression features of EoE and PPI-REE are nearly identical, and multivariate analyses have not identified any feature that distinguishes them. Studies have documented that some patients with EoE whose disease initially responds to elimination diets also responds to PPIs when those diets are stopped. Conversely, some patients with PPI-REE on unrestricted diets have disease that responds to elimination diets (with food triggers identified) when PPIs are stopped. A growing — but not yet universal — consensus is that PPI-REE is EoE in many, if not most, cases.

The term PPI-REE was proposed with all the good intentions, and confusion about PPI-REE has had consequences, specifically regarding how clinicians manage the use of PPIs during diagnostic endoscopy. For patients with symptoms of GERD whose disease incompletely responds to PPIs, it is not common clinical practice to stop PPIs before diagnostic endoscopy, even when endoscopy is specifically performed to look for alternative diagnoses such as EoE. Clinicians might be aware of the condition called PPI-REE, but they have been told that, by definition, it is not EoE. If you accept the dictum that responsiveness to PPIs excludes EoE, then why would you stop PPIs that have helped, but not eliminated, your patient’s symptoms? How could PPIs obscure a diagnosis that they already have excluded?

My group has just published a report in Gastroenterology describing patients in whom the diagnosis of EoE was considerably delayed because diagnostic endoscopies were performed while the patients were on PPI therapy and yet had eliminated all endoscopic and histologic evidence of EoE. I think it is time we started thinking of PPIs as a treatment for EoE, not as a means to exclude the diagnosis. PPI-REE is a description, not a disease.

Dr. Spechler is a consultant for Ironwood Pharmaceuticals and Takeda Pharmaceuticals.

Join the discussion

Your email address will not be published. Required fields are marked *