Background  Blood eosinophil count has become an indication for the application of inhaled corticosteroids in patients with chronic obstructive pulmonary disease (COPD), but the value of fractional exhaled nitric oxide (FeNO) in the diagnosis and treatment of patients with stable COPD is still unclear.

  Objective  To observe the correlation between blood eosinophil count and FeNO in patients with stable COPD with different characteristics, and explore the value and limitations of FeNO in stable COPD.

  Methods  Patients with stable COPD or asthma who had FeNO and peripheral blood eosinophil count results in our hospital from March 2017 to March 2020 were retrospectively included. The correlation between FeNO and blood eosinophil count was analyzed for both groups. A binary logistic regression analysis was performed to identify risk factors for an elevation in FeNO. According to the degree of airway obstruction (FEV1pred%) and the degree of hyperinflation (RVpred%), the patients with COPD were divided into FEV1pred%≤50% and FEV1pred%>50% subgroups, RVpred%>150 and RVpred%≤150 subgroups, and RVpred%>200 and RVpred%≤200 subgroups to perform the correlation analysis of peripheral blood eosinophil count and FeNO. A receiver operating characteristic (ROC) curve analysis was conducted to determine the efficacy of blood eosinophil count in predicting an elevated FeNO level in the asthma and COPD groups.

  Results  A total of 113 patients with COPD (90 males and 23 females with a mean age of 69.04±9.59 years) and 43 patients with asthma (31 males and 12 females with a mean age of 68.37±9.45 years) were included. FeNO was moderately correlated with blood eosinophil count in the asthma group (r=0.439, P<0.001), but without significant correlation in the COPD group (r=0.104, P=0.272). The binary logistic regression analysis found that smoking history was a risk factor and RVpred% was a protective factor for an elevated FeNO level in the COPD group. Further subgroup analysis of the patients with COPD showed that blood eosinophil count was positively correlated with FeNO in the FEV1pred%≤50% subgroup (r=0.370, P=0.020), but without significant correlation in the other subgroups (P >0.05). The ROC analysis found that blood eosinophil count had almost no differential value for FeNO in the COPD group, but with certain differential efficacy in the FEV1pred%<50% subgroup (the area under the curve AUC=0.688, 95% CI: 0.444-0.931, sensitivity=0.56, and specificity=0.90). Blood eosinophil count showed good differential efficacy for FeNO in the asthma group (AUC=0.869, 95% CI: 0.721-1.018, sensitivity=0.83, and specificity=0.90).

  Conclusion  There is no correlation between FeNO and blood eosinophil count in patients with stable COPD, which is different from that of patients with asthma, but the two indicators are positively correlated in patients with stable COPD with FEV1pred%≤50%. FeNO can not replace blood eosinophil count to determine whether to use inhaled corticosteroids in patients with stable COPD, but it can be an alternative choice to a certain extent in the FEV1pred%≤50% subgroup. However, FeNO can replace blood eosinophil count to evaluate eosinophilic airway inflammation in patients with asthma.