Abstract
Asthma is a heterogenous disease characterized by airway inflammation and variable expiratory airflow limitation resulting in variable respiratory symptoms. Characterization of airway inflammation is important to choose the optimal treatment for severe asthma patients eligible for biological treatment. However, counting cells in induced sputum samples are a time-consuming process, highly dependent on personal skills. Replacing eosinophil and neutrophil cell counting with qPCR for transcripts of selected mast cell, and basophil genes may provide more reproducible results. AIMS: The objective of this study was to compare qPCR with microscopy in asthma endotyping. METHODS: A qPCR method measuring five mast cell/basophil genes was applied on induced sputum samples from 30 severe asthma patients and compared with microscopy. Target gene Ct-values (CPA3, GATA2, HDC, MS4A2, TPSAB1/TPSB2) were referenced to household β-actin Ct values as a measure of relative mRNA abundance of the target in each sample. Target/β-actin-ratios in eosinophilic and non-eosinophilic groups determined by microscopy with an eosinophil threshold of 3% in 400 cells were compared using Mann-Whitney U Test. Spearman´s correlations were used to test for correlation between targets vs. FENO and targets vs. blood eosinophil counts. RESULTS: The study demonstrated a statistical difference in relative mRNA abundance for four mast cell/basophil specific genes. CPA3, GATA2, HDC and MS4A2 were elevated in eosinophilic asthma versus non-eosinophilic asthma patients. The study found that GATA2, CPA3, MS4A2 and TPSAB1/TPSB2 transcripts are positively correlated with FENO. Neither the five mast cell genes nor the five-gene signature correlated with blood eosinophils. The five-gene signature with a target/β-actin-ratio cut-off ≥2 generated sensitivity = 87%, specificity = 94%, NPV = 88% and PPV = 92% compared to microscopy. CONCLUSION: This study confirms the contribution of mast cells in the pathogenesis of EA and suggests that mast cell mRNA markers could be one of the biomarkers used to identify EA.