Stat6 Activation and Periostin Expression in Eosinophilic Esophagitis Versus Reflux Esophagitis
WM Chandler, SK Kuwada, FC Clayton. University of Utah, Salt Lake City, UT
Background: Distinguishing eosinophilic esophagitis (EOE) from reflux esophagitis (RE) with abundant eosinophils is histologically problematic. Prior studies implicated mast cells, IL-5, IL-13, and (at the RNA level) eotaxin-3 and periostin in EOE. In other sites, stat6 phosphorylation and periostin expression are induced by IL-13.
Design: Five subject groups were studied: (pure) EOE dysphagia, >20 eosinophils maximal/hpf in an esophageal biopsy, and no clinical evidence for reflux; EOE/RE dysphagia, >20 eosinophils max./hpf, with clinical evidence of reflux (heartburn or abnormal pH probe results); RE-high clinical evidence of reflux, >20 eosinophils max./hpf and no dysphagia; RE/low clinical evidence of reflux, <20 eosinophils max./hpf and no dysphagia; and control healthy volunteers without esophageal problems. There were seven to eight subjects in each group. H&E sections of the esophageal biopsies were reviewed, as well as immunoperoxidase stains for periostin and phospho Tyr-641 Stat6. Review was blinded. Results for the different groups were compared by the nonparametric Mann-Whitney U test.
Results: Phospho-stat6 staining was nuclear. The pure EOE and EOE/RE groups had means of 59% and 45% squamous cell nuclei staining respectively, which were 3.9 and 3-fold higher than the RE/high group (15% nuclei stained). Phospho-stat6 staining was no more than 5% in each subject in the RE/low and control groups. Periostin staining was present in the connective tissue papillae and lamina propria. 7/8 EOE and 5/7 EOE/RE biopsies had abundant periostin staining in the connective tissue papillae. Only 2/7 RE/high, 1/6 RE/low, and 0/7 control biopsies had strong periostin staining. Both the EOE group and the EOE/RE group had significantly more staining for both phospho-stat6 and periostin than each of the other three groups (p<0.05, nonparametric Mann-Whitney U test). For both antigens, staining was statistically similar between the EOE and EOE/RE groups, although there were trends for more staining in the pure EOE group. There was more phospho-stat6 staining in the RE/high group (15%) than the RE/low and control groups (p<0.01).
Conclusions: As previously shown, high epithelial eosinophil counts did not distinguish EOE from some RE cases. However, periostin and phospho-stat6 staining strongly correlated with dysphagia, with or without coexisting reflux, even among cases with abundant eosinophils. This suggests a common mechanism for patients with EOE and EOE-like disease combined with reflux, involving increased expression and activation of IL-13-related proteins periostin and stat6.
Wednesday, March 11, 2009 1:00 PM
Poster Session VI # 83, Wednesday Afternoon