How EGPA CT Phenotypes Unmask Hidden Vasculitis Relapses

Eosinophilic granulomatosis with polyangiitis presents marked pulmonary heterogeneity that complicates clinical management. Chest high-resolution computed tomography frequently reveals diverse lung manifestations, yet standardized radiological classifications remain scarce. Recently, researchers demonstrated that characterizing distinct EGPA CT phenotypes through latent class analysis provides vital prognostic insight. Consequently, this novel imaging stratification helps clinicians identify patients who face heightened risk for clinical relapse.

Classifying EGPA CT Phenotypes Using Latent Class Analysis

Investigators retrospectively evaluated 140 patients with pulmonary eosinophilic granulomatosis with polyangiitis treated between 2014 and 2025. In this cohort, two thoracic radiologists analyzed chest CT scans using latent class analysis to establish reproducible imaging clusters. Moreover, the statistical clustering revealed four unambiguous disease subcategories. First, the bronchial phenotype occurred in 32.1 percent of patients and showed airway-predominant lesions. Second, the infiltrative phenotype accounted for 20.7 percent and presented parenchymal opacities. Third, the mixed phenotype affected 14.3 percent and combined both bronchial and parenchymal features. Finally, the nonspecific phenotype appeared in 32.9 percent of cases without a dominant single pattern.

Clinical Outcomes and Relapse Risk by Phenotype

The derived radiological clusters directly correlated with treatment response and one-year prognosis. Specifically, patients displaying bronchial and infiltrative phenotypes achieved favorable short-term outcomes after standard therapy. Similarly, individuals with nonspecific imaging features maintained clinical stability over the follow-up period. In contrast, patients with the mixed phenotype experienced significantly higher rates of disease reactivation. Furthermore, multivariable modified Poisson regression confirmed this increased relapse risk after adjusting for baseline treatments, age, and sex. Radiologists also demonstrated outstanding diagnostic reliability, achieving inter-observer and intra-observer agreement coefficients of 0.85 and 0.87.

Implications for Indian Pulmonologists and Physicians

In clinical practice across India, diagnosing and managing eosinophilic vasculitis poses substantial challenges. For instance, clinicians often mistake early EGPA presentations for severe asthma or refractory allergic bronchopulmonary aspergillosis. Therefore, routine high-resolution CT assessment remains indispensable during the initial evaluation. Because mixed phenotype manifestations signal an elevated relapse risk, these patients require proactive surveillance. In addition, attending physicians might consider earlier escalation or maintenance with targeted biologic therapies, such as anti-IL-5 agents, for this vulnerable subset. Ultimately, integrating radiological phenotyping into daily workflows enhances risk stratification and personalized therapy.

Frequently Asked Questions

Q1: What are the four primary EGPA CT phenotypes identified in the study?

The four phenotypes include the bronchial phenotype (airway-predominant), infiltrative phenotype (parenchyma-predominant), mixed phenotype (combined airway and parenchymal disease), and nonspecific phenotype.

Q2: Which imaging phenotype carries the highest risk of disease relapse?

Patients exhibiting the mixed phenotype demonstrate a significantly higher risk of one-year clinical relapse compared to other radiological categories.

Q3: How reliable is the visual assignment of these CT phenotypes by radiologists?

Thoracic radiologists achieved excellent diagnostic reproducibility, with inter-observer and intra-observer agreement coefficients reaching 0.85 and 0.87, respectively.

References

1. Yang J et al. Latent class analysis of CT-based phenotypes in eosinophilic granulomatosis with polyangiitis and pulmonary involvement. Eur Radiol. 2026 Oct 23. doi: 10.1007/s00330-026-12859-z. PMID: 42909531.
2. Cottin V, Bel E, Bottero P, et al. Respiratory manifestations of eosinophilic granulomatosis with polyangiitis (Churg–Strauss). Eur Respir J. 2016;48(5):1429-1441.
3. Gu Y, Liu Y, Zhang T, Han Y, Peng M, Shi J. Latent Class Analysis to Explore Subtypes of EGPA: Focusing on Respiratory Involvement and Inflammation Markers. J Clin Med. 2024;13(4):1042.