Biomarkers diagnosed EoE, could not distinguish EoE from PPI-REE

Immunohistochemical analysis of major basic protein, eotaxin-3 and tryptase levels in esophageal tissues effectively diagnosed eosinophilic esophagitis, but could not differentiate the disease from proton pump inhibitor-responsive esophageal eosinophilia, according to new research data.

“It can be difficult based on standard clinical metrics to distinguish eosinophilic esophagitis (EoE), gastroesophageal reflux disease (GERD), and proton pump inhibitor-responsive esophageal eosinophilia (PPI-REE),” Evan S. Dellon, MD, from the University of North Carolina School of Medicine, told Healio Gastroenterology. “For example, the level of esophageal eosinophilia in esophageal biopsies is not specific for any of these conditions.”

Evan S. Dellon

Aiming to determine whether major basic protein (MBP), eotaxin-3 and tryptase staining in the esophageal epithelium can be used to diagnose EoE and discriminate between EoE and PPI-REE, Dellon and colleagues performed a prospective study of adults undergoing outpatient esophagogastroduodenoscopy at the University of North Carolina from 2009 to 2012.

They performed immunohistochemistry on 123 non-EoE controls with GERD or dysphagia (mean age, 52 years; 41% men, 76% white, 58% with dysphagia), 50 patients with EoE (mean age, 48 years; 87% men, 87% white, 96% with dysphagia) and 23 patients with PPI-REE (mean age, 52 years; 41% men, 76% white, 58% with dysphagia).

Esophageal tissues from the EoE group had a median of 951 (interquartile range [IQR], 322-1,849) MBP-positive cells/mm², 155 (IQR, 63-783) eotaxin-3-positive cells/mm² and 249 (IQR, 163-378) tryptase-positive cells/mm² vs. 2 (IQR, 0.4-6) MBP-positive cells/mm², 18 (10-31) eotaxin-3-positive cells/mm² and 11 (7-19) tryptase-positive cells/mm² in the control group (all P<.001). MBP, eotaxin-3 and tryptase individually and in combination demonstrated excellent utility for identifying EoE (area under the receiver operating characteristic curve [AUC], 0.99, 0.94, 0.99 and 1, respectively). Comparable results were found from analysis of samples with eosinophil counts of 10 to 100 eosinophils per high-power field. None of the markers distinguished EoE from PPI-REE; compared with EoE, PPI-REE samples had 987 (IQR, 655-1,917) MBP-positive cell/mm² (P=.18), 160 (IQR, 53-334) eotaxin-3-positive-cells/mm² (P=.33) and 243 (IQR, 163-280) tryptase-positive cells/mm² (P=.28).

“The first main finding was that this set of esophageal biopsy stain clearly distinguished patients with EoE from those with GERD with a high degree of accuracy, essentially 100%,” Dellon said. “Because of this, we have created an online predictive calculator that can be used to interpret the level of staining and predict whether a patient might have EoE based on biopsy staining alone.

“The second main finding was that this set of stains did not distinguish EoE from PPI-REE,” he added. “However, for the first time we were able to show that patients with PPI-REE have some of the same immune/allergic inflammatory factors increased in esophageal biopsies, and that after PPI treatment, these factors return to normal levels.”

David A. Katzka

“One also might ask from the compelling data in this study whether performing routine immunohistochemical staining for cytokines, mast cells, and eosinophil-derived proteins associated with EoE would help clinically to distinguish EoE from [GERD],” David A. Katzka, MD, from the Mayo Clinic, wrote in an accompanying editorial. “Although future studies of greater numbers of patients may corroborate this, particularly those that prospectively measure response to PPI or steroid therapy on the basis of these stains, I am not sure it is ready to become a routine method of analysis for our pathology colleagues.

“With the strength of data, such as that provided in the article by Dellon et al, it may be time to diagnostically commit, reclassify, and treat these patients with esophageal eosinophilia and response to PPIs as EoE, GERD, or an overlap of the two,” he concluded.

For more information:

Dellon ES. Clin Gastroenterol Hepatol. 2014;12:2015-2022.

Katzka DA. Clin Gastroenterol Hepatol. 2014;12:2023-2025.

Disclosure: The researchers report no relevant financial disclosures.