Short-term longitudinal variation in exhaled nitric oxide fraction predicts future moderate-to-severe exacerbations...

Fractional exhaled nitric oxide (F _ENO) reflects type 2 airway inflammation, yet the prognostic utility of longitudinal F _ENO changes in COPD remains unclear. We aim to explore whether 6-month F _ENO variation patterns predict subsequent moderate-to-severe exacerbations in COPD patients receiving triple therapy.

We prospectively evaluated COPD patients receiving triple inhaled therapy from 2023 to 2024. Four F _ENO-related variables were used to derive F _ENO variation phenotypes via K-means unsupervised clustering.

Three distinct F _ENO-variation phenotypes were identified: a High F _ENO-Persistently Elevated subtype (Cluster 1, 11.2%), a High F _ENO-Declining subtype (Cluster 2, 20.3%) and a Low F _ENO-Stable subtype (Cluster 3, 68.5%). During 6-12 months of follow-up, 75 patients (33.8%) experienced a moderate-to-severe exacerbation. Compared with Cluster 3, both Cluster 1 (OR 6.97, 95% CI 2.38-20.47) and Cluster 2 (OR 20.08, 95% CI 7.01-57.58) were independently associated with future exacerbations (p<0.001). The final prediction model incorporating F _ENO variation phenotypes demonstrated excellent discrimination (AUC 0.889) and provided greater net clinical benefit than models using traditional clinical predictors alone (p=0.004). Subgroup analyses revealed consistent effect sizes across categories of age, prior-year exacerbations, baseline FEV₁ % predicted, symptom burden, smoking status and proportion of days covered, supporting the robustness and internal consistency of the F _ENO-variation phenotyping approach.

Short-term longitudinal F _ENO variation over 6 months was independently associated with future moderate-to-severe exacerbations in COPD patients receiving triple therapy, supporting risk stratification and hypothesis generation for future interventional studies rather than direct treatment guidance.