Abstract
BACKGROUND: Initial asthma management often relies on a one-size-fits-all approach; however, the heterogeneity of treatment-naive, uncontrolled asthma remains insufficiently characterized. Specifically, the complex relationships among symptom burden, type 2 (T2) inflammation, and pulmonary function in this population are poorly understood. OBJECTIVE: We sought to identify distinct clinical phenotypes in treatment-naive adults via multidimensional analysis and evaluate the association between symptom scores and T2 inflammation. METHODS: This single-center, retrospective study analyzed 171 treatment-naive adults with uncontrolled asthma (5-item Asthma Control Questionnaire [ACQ-5] score ≥ 1.5). We performed k-means clustering using 20 variables, including demographics, symptoms, T2 biomarkers (blood eosinophils and fractional exhaled nitric oxide), and lung function (FEV(1) and forced expiratory flow at 25% to 75% of forced vital capacity). Logistic regression assessed the ability of ACQ-5 scores to predict T2-high status (blood eosinophils ≥ 300/μL and fractional exhaled nitric oxide > 50 ppb). RESULTS: Cluster analysis identified 3 phenotypes: relative T2-low/preserved function (32.7%); severe T2/obstructive (29.2%); and (3) T2-moderate/small airway dysfunction predominant (38.0%). The prevalence of T2-high status varied significantly (30.4%, 64.0%, and 50.8%, respectively; P = .002). ACQ-5 scores showed limited ability to discriminate T2-high status (area under the curve 0.615). Multivariable modeling provided only modest improvement (area under the curve 0.644). CONCLUSIONS: In this study population, treatment-naive uncontrolled asthma comprises 3 clinically distinct phenotypes with divergent inflammatory and functional profiles. Symptom burden alone appeared to be a limited predictor of T2-high inflammation. These exploratory findings suggest that incorporating objective biomarkers and functional assessments into initial care may help tailor therapeutic strategies for patients with small airway dysfunction or discordant symptom-inflammation profiles. Further validation in larger cohorts is required.