Article Text
Abstract
Background Children with severe therapy-resistant asthma (STRA) are refractory to treatment despite optimal management. Assessment of airway inflammation to phenotype these patients can enable targeted therapy. Samples obtained at bronchoscopy provide the most direct measure of lower airway inflammation; however, non-invasive measures (induced sputum and exhaled nitric oxide (FeNO)) are of greater clinical utility. We have previously demonstrated a poor relationship between blood and bronchoalveolar lavage (BAL) eosinophilic phenotype using clinical cut-offs for children (blood eosinophils 1.0 × 109/L).1 Recent studies of the anti-IL-5 antibody mepolizumab have used a lower cut point (0.3 × 109/L) for blood eosinophils.2 The aim of this study was to assess the concordance between BAL and non-invasive measures of inflammation.
Methods 113 children (aged 4–17 years) with STRA underwent bronchoscopy at the Royal Brompton Hospital. They had all previously been assessed and potentially modifiable factors such as poor adherence had been addressed. Inflammation was measured invasively using BAL cytology and non-invasively by blood eosinophils, induced sputum cytology, and FeNO. The eosinophilic phenotype was defined as BAL eosinophils >1.19%; blood eosinophils ≥0.3 × 109/L; sputum eosinophils ≥2.5%; and FeNO >35ppb. The relationship between measures was assessed using Spearman rank correlation and Receiver Operator Characteristic (ROC) curves were constructed to determine which cut points best determined BAL eosinophilia and positive and negative predictive values (PPV and NPV) calculated.
Results The predominant phenotype in all samples was eosinophilic. There was 75.6–77.8% concordance between the eosinophilic phenotype in BAL and each of the non-invasive measures.
Blood and BAL eosinophils had the strongest correlation (r = 0.57, p < 0.001, n = 84). Weaker correlations were found between the other measures. The most promising predictor of BAL eosinophilia was a blood eosinophil count of 0.15 × 109/L (PPV 84.1, NPV 71.4) (Table 1).
Conclusions These results suggest that blood eosinophils at a lower cut-point may be a useful measure of lower airway inflammation. However, this is still a relatively invasive test in children and there is little data available about longitudinal stability of blood eosinophils.
References 1 Ullmann N, et al. Allergy 2013;68:402
2 Pavord ID, et al. Lancet 2012;380:651