Eosinophils are seen in a number of dermatologic conditions. While the extent of their function in these diseases remains to be fully elucidated, pathogenic activity in bullous pemphigoid suggests a more significant role than previously thought. Several dermatoses have a fairly characteristic histologic morphology of eosinophil infiltration. We hypothesized that epidermal expression of eotaxins and TSLP would differ by disease, perhaps explaining the different histologic morphologies. We performed a retrospective study of eosinophil rich dermatoses to perform immunohistochemistry. We collected 49 specimens composed of bullous pemphigoid (n = 15), atopic dermatitis (n = 12), drug rash (n = 8), arthropod assault (n = 5), and non-bullous pemphigoid eosinophilic spongiosis (n = 5). We used lichen planus (n = 4) as a control for lymphocyte-mediated inflammation. TSLP was diffusely expressed in all epidermal samples, whereas eotaxins demonstrated a weaker staining. Eotaxins and TSLP demonstrated a gradient between basal and spinous keratinocytes. The correlation between overall basal keratinocyte and spinous keratinocyte staining of eotaxins and TSLP with the number of eosinophils demonstrated a significant correlation between eotaxin-1 (R = 0.404, P = 0.004), eotaxin-2 (R = 0.576, P < 0.001), and eotaxin-3 (R = 0.512, P < 0.001), but not TSLP (R = 0.164, P = 0.251). These remained significant after correcting for multiple comparisons. While we were unable to detect significant differences in epidermal expression of eotaxins and TSLP in various eosinophil rich dermatoses, we identified a significant correlation of spinous keratinocyte eotaxin staining with tissue eosinophilia. Our identification of a correlation of spinous keratinocyte eotaxin staining with tissue eosinophilia may provide insight into local eosinophil chemotaxis.
Keywords: Atopic dermatitis; Bullous pemphigoid; CCL11; CCL24; CCL26; Thymic stromal lymphopoietin.