Serum periostin in smokers and never smokers with asthma

Background: Elevated serum periostin is associated with airway eosinophilia and may predict response to therapies targeting Th2 inflammation. Smoking in asthma is generally associated with non-eosinophilic airway inflammation and corticosteroid insensitivity. We determined the effect of smoking status on serum periostin in asthma. Methods: Serum periostin (ELISA; Aviscera Bioscience) was measured in 107 patients with stable asthma of different disease severity and 45 healthy controls. The effects on serum periostin of clinical indices including smoking status and of inflammatory biomarkers including sputum eosinophil count were analysed. Serum periostin was measured before and after two weeks of oral corticosteroids in a separate group of 33 non-smokers and smokers with stable asthma.Results: Serum periostin (median [IQR], ng/mL) was reduced in smokers with asthma compared to never smokers with asthma; 9 (9, 307) versus 233 (34, 1108) respectively, p = 0.017. Periostin was not influenced by disease severity (p = 0.786) or atopic status (p = 0.144). There was a weak correlation between serum periostin and sputum eosinophil count in smokers with asthma (r = 0.315, p = 0.020). The proportion of patients with an elevated serum periostin concentration was greater in never smokers with asthma compared to smokers with asthma [65% versus 39% respectively, p = 0.003]. Oral steroid treatment reduced serum periostin (p = 0.030) in non-smokers with asthma. Conclusion: Despite lower median serum periostin concentrations in smokers with asthma compared to never smokers with asthma, approximately forty percent of this group had a high level. The potential value of a raised serum periostin concentration in predicting a beneficial response to therapies targeting Th2 inflammation in smokers with asthma requires to be investigated.