Clinical characteristics and predictive models of severe asthma risk in patients with late-onset asthma

Abstract

Background. Late-onset asthma (after 40 years) is characterized by more severe disease, frequent exacerbations, fixed bronchial obstruction, and high comorbidity; prognostic tools for this age group, including risk models for rehospitalization and mucus plugs, remain insufficiently developed. Aim. To develop and validate predictive models for estimating the probability of frequent severe exacerbations and the presence of mucus plugs on chest X-Ray computed tomography (CT) in hospitalized patients with late-onset asthma. Materials and methods. This prospective study included 130 hospitalized patients with late-onset asthma (71.5% female, mean age 63.5 ± 11.3 years). Clinical, laboratory (IgE, eosinophils, FeNO), and instrumental (spirometry, body plethysmography, chest CT) data were assessed. Logistic regression models with stepwise selection were built for two outcomes: ≥ 2 severe exacerbations within 12 months, and mucus plugs on CT. Evaluation included bootstrap validation (5,000 resamples), the Hosmer–Lemeshow test, and ROC analysis. Results. A T2 endotype was identified in 115 patients (88.5%); the rate of frequent severe exacerbations (≥ 2/year) was 25.4%. The most common phenotypes were allergic, eosinophilic, nonallergic, and asthma with obesity and fixed obstruction; all patients had combinations of these. Mucus plugs on CT were significantly more common with asthma-COPD overlap. Patients with four phenotypes had poorer symptom control and lower pulmonary function. Independent predictors of frequent severe exacerbations were eosinophilic phenotype (OR 3.2, 95% CI 1.3–8.1; p = 0.013) and diabetes mellitus (OR 3.3, 95% CI 1.3–8.8; p = 0.015); AUC 0.726 (95% CI 0.631–0.821), sensitivity 70%, specificity 66%. Reduced PEF% was the only independent predictor of mucus plugs (OR 0.97 per 1% increase, 95% CI 0.95–0.99; p 0.001); AUC = 0.735 (95% CI 0.649–0.821), sensitivity 68.9%, specificity 68.1%. Conclusion. In hospitalized late-onset asthma patients, eosinophilic phenotype and diabetes mellitus are independent predictors of frequent severe exacerbations (OR ~ 3.2–3.3). Reduced PEF% predicts mucus plugs on CT. The models (AUC 0.726–0.735) can be used for risk stratification in routine practice.