Evidence for distinct cytokine expression in allergic versus nonallergic chronic sinusitis

The purpose of this study was to characterize the relationship between tissue cytokine expression and the cellular infiltrate present in chronic hyperplastic sinusitis with nasal polyposis (CHS/NP) and to compare the immunopathology and cytokine profile of patients with allergy versus patients without allergy.

Nasal polyp tissue samples from 12 patients with CHS/NP and nasal turbinate biopsy specimens from 10 normal control patients were examined for the expression of interleukin (IL)-4, IL-2, and interferon (IFN)-gamma cytokine messenger RNA (mRNA) species by in situ hybridization. These data were analyzed in conjunction with data previously reported for the cytokine mRNA species granulocyte-macrophage colony-stimulating factor (GM-CSF), IL-3, and IL-5 and the immunocytochemical profile of the inflammatory cell infiltrate. Patients with allergy were distinguished from those without allergy on the basis of allergy skin tests.

Tissue eosinophilia was a prominent feature of both allergic and nonallergic CHS/NP and correlated in both subgroups with the density of GM-CSF and IL-3 mRNA+ cells. In comparison with normal controls, patients with allergic CHS/NP had significantly higher CHS/NP had significantly higher tissue densities of GM-CSF, IL-3, IL-4, and IL-5 (p < or = 0.025). In contrast, patients with nonallergic CHS/NP had significantly higher tissue densities of GM-CSF, IL-3, and IFN-gamma (p < or = 0.001). The allergic and nonallergic subgroups showed distinct cytokine profiles with the most distinguishing cytokines of the allergic subgroup being IL-4 (p = 0.001) and IL-5 (p = 0.017) and of the nonallergic subgroup being IFN-gamma (p = 0.004). Furthermore, patients with allergic CHS/NP showed an increased density of CD3+ T lymphocytes compared with either controls or patients with nonallergic CHS/NP (p = 0.03). The density of CD3+ T lymphocytes was the only significant difference between patients with allergic and nonallergic CHS/NP. A clinical history of aspirin sensitivity was strongly correlated with nonallergic CHS/NP, as well as the nonallergic CHS/NP profile of cytokines, including IFN-gamma.

We conclude that distinct mechanisms of eosinophilia exist in patients with allergic versus nonallergic CHS/NP. The allergic mechanism involves production of TH2-type cytokines, including GM-CSF, IL-3, IL-4, and IL-5, by infiltrating T lymphocytes. The nonallergic mechanism remains unknown but does involve production of GM-CSF, IL-3, and IFN-gamma. However, nonallergic eosinophilia is independent of IL-4 and IL-5, cytokines that contribute to tissue eosinophilia in allergic inflammation. Aspirin sensitivity is strongly correlated with nonallergic CHS/NP and production of the nonallergic CHS/NP profile of cytokines, including IFN-gamma.