Abstract
OBJECTIVES: Features of eosinophilic esophagitis (EoE) and gastroesophageal reflux disease (GERD) overlap. We aimed to determine whether staining for tissue biomarkers would differentiate EoE from GERD, suggesting utility for diagnosis of EoE.
METHODS: In this case-control study, EoE patients defined by consensus guidelines were compared to GERD patients with eosinophils on esophageal biopsy. Immunohistochemistry was performed for major basic protein (MBP), eotaxin-3, leukotriene A4 hydrolase (LTA4H), and leukotriene C4 synthase (LTC4S). After masking, the maximum staining density (cells per mm 2) was quantified for each marker and compared between groups. Receiver operator characteristic curves were constructed, and the area under the curve (AUC) calculated to assess the diagnostic utility of each of the biomarkers alone and in combination with eosinophil counts.
RESULTS: There were 51 EoE cases (mean age 24; mean 143 eosinophils per high-power field (eos per h.p.f.)) and 54 GERD controls (mean age 34; mean 20 eos per h.p.f.). The MBP density was higher in EoE than in GERD (1479 vs. 59 cells per mm 2; P < 0.001), as was the eotaxin-3 density (2219 vs. 479; P < 0.001). There were no differences for LTA4H and LTC4S. MBP density and eosinophil count correlated (R = 0.81; P < 0.001); correlation with eotaxin-3 was weaker (R = 0.25; P = 0.01). The AUC for diagnosis of EoE was 0.96 for MBP, 0.87 for eotaxin-3, 0.58 for LTA4H, 0.66 for LTC4S, and 0.99 for the combination of MBP, eotaxin-3, and eosinophil count.
CONCLUSIONS: Patients with EoE had substantially higher levels of MBP and eotaxin-3 staining than GERD patients. These markers may have utility as a diagnostic assay for EoE.