PURPOSE OF THE STUDY.
The objective of this study was to assess whether exhaled biomarkers, expression of inflammation genes, and early lung function measurements can improve early asthma prediction in the wheezing preschool child when used in conjunction with the Asthma Predictive Index (API).
The study included 202 children, aged 2 to 4 years, with history of recurrent wheeze who were prospectively followed up yearly until 6 years of age. At age 6, a diagnosis of asthma or transient wheeze was determined on the basis of symptoms, use of asthma medication, and lung function features.
At initial recruitment, researchers collected exhaled breath condensate, measuring exhaled volatile organic compounds (VOCs) and cytokines, obtained blood samples for measurement of expression of inflammation genes, and measured airway resistance before and after inhalation of salbutamol. In the case of an airway infection being present, or, the use of inhaled corticosteroids, the visit was postponed for 4 weeks (with inhaled corticosteroids stopped) to limit bias. At age 6, the diagnosis of asthma versus transient wheeze was made by a pediatric pulmonologist on the basis of symptoms, use of asthma medication, and lung function features. Using a logistic regression model, researchers tested the value of separately adding these biomarkers to the API in the prediction of asthma by age 6 years.
Adding information on exhaled VOCs and gene expression to the API significantly improved the ability to predict asthma at age 6 years. For VOCs, the following were noted: area under the curve (AUC) of 89% (increase of 28%), positive predictive value (PPV) of 82%, and negative predictive value (NPV) of 83%. Sensitivity increased from 66% to 84% and specificity from 56% to 82%. Information on gene expression of toll-like receptor 4, catalase, and tumor necrosis factor-α significantly improved asthma prediction with AUC of 75% (increase of 17%), PPV of 76%, and NPV of 73%. Furthermore, in combination, the VOCs and gene expression information further improved the positive predictive value of the API. The API alone classified 60% of children accurately; the API plus VOCs and gene expression classified 89% of children accurately (AUC 95%, PPV 90%). Biomarkers in exhaled breath condensate and airway resistance measurements did not improve asthma prediction.
Adding information on pathogenic features of asthma (as represented by exhaled VOCs and expression of inflammation genes) to the API significantly improves the ability to diagnose asthma in wheezing preschool children.
Wheezing in preschool-age children is quite common (40% incidence). However, wheezing usually resolves by age 6 in the majority of these children. Thus, having a reliable tool to assist clinicians in diagnosing asthma (vs transient wheeze) earlier in life is greatly needed. The API is relatively easy to use, incorporating information on childhood eczema, childhood rhinitis, parental asthma, and eosinophilia. However, this index (among others) leaves us with much room for improvement, providing only modest sensitivity and positive predictive value, which ultimately can lead to a significant number of children without asthma receiving inappropriate treatment. In this study, information on airway resistance (pre- and postbronchodilator) did not improve an early asthma diagnosis when added to the API. Further research is needed in different populations to confirm these findings and extend their meaning into daily clinical practice.
- Copyright © 2015 by the American Academy of Pediatrics