Background and aim: Eosinophilic phenotype, sputum eosinophil count ≥3%, of asthma is a valuable clinical parameter. However, measurement of sputum inflammatory cells is time-consuming and requires specialized personnel. A simpler surrogate, such as peripheral eosinophil count, would be beneficial. The aim of this study was to assess the value of peripheral eosinophil count for predicting the eosinophilic phenotype of adult asthmatics.
Materials and methods: This retrospective study was conducted in 192 adult asthmatics. Sputum and venous blood were collected at the same visit and analyzed for cell differential. Receiver-operating characteristic (ROC) curve analysis was used to assess the relationship between eosinophilic phenotype and peripheral eosinophil count.
Results: Peripheral eosinophil count, derived from the ROC curve, of more than 210 cells/mm(3) yielded 67.7% sensitivity and 66.2% specificity [area under the curve (AUC)=0.698, p=0.0001] for identifying the eosinophilic phenotype in the total patients. Multiple regression analysis revealed that steroid treatment and atopy significantly influenced the peripheral eosinophil count. Within the steroid-naïve subgroup, a peripheral eosinophil count of >190 cells/mm(3) yielded 76.3% sensitivity and 67.4% specificity (AUC=0.730, p<0.0001) for identifying eosinophilic phenotype, while in the steroid treated subgroup, the AUC was 0.554 (p=0.4985). Within the steroid-naïve non-atopic subgroup, a peripheral eosinophil count of >190 cells/mm(3) yielded 73.1% sensitivity and 79.3% specificity (AUC=0.809, p=0.0001) for identifying the eosinophilic phenotype, while in the steroid-naïve atopic subgroup, the AUC was 0.588 (p=0.2723).
Conclusion: Peripheral eosinophil count is able to identify the presence of eosinophilic phenotype in steroid-naïve non-atopic adult asthmatics with moderate accuracy.