Unraveling the Molecular Heterogeneity of Asthma Through Blood Transcriptomics

Asthma is no longer regarded as a single disease, but rather as a heterogeneous syndrome driven by diverse pathophysiological mechanisms. While extensive research has sought to define phenotypes/endotypes based on clinical characteristics and laboratory parameters, type 2 (T2) inflammation remains a fundamental pathogenic mechanism exacerbating asthma symptoms and contributing to disease severity.¹,² Thus, the management of severe asthma has increasingly centered on biologic therapies targeting immunoglobulin E (IgE), interleukin (IL)-4/-13, IL-5, and thymic stromal lymphopoietin, with surrogate markers, such as high blood eosinophil counts (BECs) and fractional exhaled nitric oxide (FeNO), being used to identify patients with T2-high phenotypes.³,⁴ However, in clinical practice, some patients do not achieve adequate symptom control or clinical responses to T2-biologics despite presenting with classical T2-high phenotypes, suggesting that relying solely on BECs and/or FeNO may be insufficient given interindividual variations, particularly in response to systemic corticosteroids (SCS). Therefore, reliable biomarkers derived from non-invasive, easily accessible technologies that can reflect underlying mechanisms are required to elucidate molecular differences among patients with overlapping clinical features, ultimately enabling the selection of optimal T2-biologics for precision medicine.⁵,⁶

To overcome this discrepancy, transcriptomic analysis has emerged as a powerful tool for providing an integrative, unbiased landscape of molecular signatures that can help refine current phenotype/endotype classifications. Emerging evidence has demonstrated distinct transcriptomic profiles associated with specific clinical phenotypes/endotypes in sputum or nasal and bronchial epithelium. Notably, large-scale, multicenter cohorts, such as U-BIOPRED and SARP, have demonstrated distinct clinical and molecular characteristics associated with transcriptomic clusters defined by T2/non-T2 phenotypes or steroid responsiveness. However, transcriptomic profiles derived from local airway samples primarily reflect localized airway inflammation, highlighting the need for more comprehensive approaches.⁷,⁸

In this issue, Kim et al.⁹ addressed this challenge by characterizing blood transcriptomic profiles from peripheral blood mononuclear cells (PBMCs) in 2 adult asthma cohorts of Korean patients, the COREA (n = 456) and PRISM (n = 122) cohorts, representing a major strength of this study. PBMCs can offer a practical, reproducible, and minimally invasive approach for monitoring systemic inflammation.¹⁰ By integrating differentially expressed genes identified through meta-analyses of publicly available Gene Expression Omnibus datasets, the authors identified three distinct blood transcriptome-associated clusters (B-TACs) in the COREA cohort. The B-TAC1 represented a classic T2-high phenotype, characterized by prominent expression of IL-13/T2 and innate lymphoid cell (ILC) 2 gene sets and significantly increased BECs, with trends toward higher FeNO, total IgE, and sputum eosinophil levels, although the differences did not reach statistical significance. Similar to B-TAC1, B-TAC2 was characterized by high BECs but exhibited an intermediate/mixed molecular profile, with downregulation of the ILC3 pathway and features spanning both T2 and T1/T17 inflammation. Notably, B-TAC2 showed less improvement in asthma control test scores after 6 months of treatment with T2-biologics than B-TAC1. Lastly, B-TAC3 represented a non-T2 phenotype, characterized by lower BECs and higher reliance on SCS, with an increase in T1/T17, ILC1, neutrophil gene sets, and activation of inflammasome-related biological pathways. These findings were closely replicated in the PRISM cohort, comprising patients with severe asthma and a T2-high phenotype, further confirming the stability and reliability of the distinct signatures identified for both B-TAC1 and B-TAC2.

The most notable finding of this study⁹ is the identification and characterization of B-TAC2 as a distinct molecular phenotype. Under current treatment guidelines, patients with asthma belonging to B-TAC1 and B-TAC2 share clinical features, such as increased BECs, total IgE, and FeNO, for which T2-biologics may be considered primary treatment options.¹¹ However, transcriptomic cluster analysis revealed that B-TAC2 had a distinct mixed molecular signature and was associated with reduced response to T2-biologics and incomplete asthma control, suggesting a plausible molecular rationale for incomplete responsiveness to T2-biologics among patients with severe asthma and a T2-high phenotype. In addition, this study highlights the potential of blood transcriptomic classification for refining patient selection and predicting responsiveness to T2-biologics. Moving beyond simple biomarkers toward integrated, omics-based phenotyping could enable the development of tailored and more precise therapeutic strategies.

Although validation studies of transcriptomic signatures across independent severe asthma cohorts support the clinical relevance of this B-TAC classification, several questions warrant future investigation. First, differences in study subjects and management protocols, including pharmacologic treatments, between the training and validation cohorts should be carefully considered. Second, correlations between established transcriptomic profiles and lung function or inflammatory parameters remain inconclusive due to the lack of statistical significance in most intergroup differences. Third, the assessment of treatment response relied primarily on asthma control test scores, with no significant changes in asthma exacerbation rates or SCS requirements, representing a notable limitation. Therefore, further research is essential to determine whether these molecular clusters can predict long-term clinical outcomes in longitudinal asthma cohorts. Longitudinal assessments of therapeutic response to long-term use of T2-biologics in larger, global cohorts will be critical to establishing the clinical utility of B-TAC profiling and its applicability in real-world clinical settings.