Serum periostin: Population reference range and daily variation

<p><b>Background: Periostin, previously termed Osteoblast-specific factor 2, is an extracellular matrix protein involved in development, stress or injury response, epithelial mesenchymal transition, restructuring of the extracellular matrix and remodelling of tissues. It has been studied across a wide range of tissues and pathologic processes including heart valve and periodontal ligament formation, asthma, myocardial infarction, bone fractures, and atopic dermatitis. It is currently being evaluated as a potential biomarker for directing and monitoring asthma treatment including monoclonal antibody treatment against IL13 and IgE. The clinical utility of periostin as a biomarker is severely limited by the lack of understanding of its serum levels and the factors affecting these levels. The absence of a normal reference range and daily variation may affect its clinical applicability as a TH2 biomarker.</b></p> <p>Objectives: To derive age- and sex-related reference intervals from an adult population sample without asthma or COPD and to determine the effect of sampling time on serum periostin levels in adult participants with and without asthma.</p> <p>Methods: - Reference RangeSerum periostin levels were measured in 480 individuals, comprising 60 female and 60 male adults in each of the 18-30, 31-45, 46-60 and 61-75 year age groups. Key exclusion criteria included a doctor’s diagnosis of asthma, chronic bronchitis or COPD, and a history of wheezing or use of respiratory inhalers in the last 12 months. The distribution of periostin and logarithm-transformed periostin levels were derived, and 90% confidence intervals for an individual prediction calculated.</p> <p> - Daily VariationSerum periostin was measured at 2 hourly intervals from 0800 to 1800 hours in 16 participants with stable asthma prescribed inhaled corticosteroid and long-acting beta-agonist therapy, and in 16 otherwise healthy participants without asthma. Mixed linear models were used to compare time zero (0800) with subsequent measurement time for serum periostin for both asthma and non-asthma groups.</p> <p>Results: The distribution of serum periostin was right skewed with a mean (SD) periostin of 51.2 (11.9) ng/ml, median (IQR) 50.1 (43.1 to 56 .9) ng/ml and range 28.1 to 136.4 ng/ml. There was no association between logarithm periostin and age or sex. The 90% confidence limits for periostin were 35.0 and 71.1 ng/ml. There were weak positive associations between logarithm periostin and blood eosinophil count and log IgE, but no association with log FeNO, and weak negative associations with FEV₁, BMI and creatinine clearance.</p> <p>In both asthma and non-asthma, the mean (SD) serum periostin continuously reduced during the day from 53.5 (13.6) ng/ml at 0800 hours to 50.9 (13.4) ng/ml at 1800 hours (difference log periostin -0.05, P=<0.001) and 50.5 (13.0) ng/ml at 0800 hours to 46.2 (11.5) ng/ml at 1800 hours (difference log periostin -0.08, P=<0.001) respectively. In a post hoc analysis, 1/16 asthma participants changed classification from ‘high’ (≥50ng/ml) to ‘low’ (<50ng/ml) periostin, based on 0800 and 1800 hour measurements.</p> <p>Conclusions: These findings provide reference values for serum periostin levels in adults without asthma or COPD. The weak and inconsistent association between periostin levels and other Th2 biomarkers suggests that they may measure related but different components of the Th2 inflammatory processes. A daily variation in periostin was also established with higher periostin values in the morning compared with the afternoon in both asthmatic and non-asthmatic adults. The small magnitude of the variation in serum periostin levels suggests that the time of day in which the serum periostin measurements are made is unlikely to influence treatment decisions if a specific serum periostin level is used to predict treatment responsiveness.</p>