Asthma Is Not One Disease and Precision Medicine Is Finally Catching Up
A review of asthma endotypes and theratypes shows how biomarkers, multi-omics, imaging and machine learning may replace one-size-fits-all asthma care.
Resumen
Asthma is no longer treated as a single disease. This review describes how patients differ in visible traits (phenotypes), underlying biology (endotypes) and treatment response (theratypes). Type 2-high asthma, driven by IL-4, IL-5 and IL-13, is well characterized and has targeted biologics. Type 2-low asthma, often neutrophilic and steroid-resistant, is poorly understood. Point-of-care biomarkers such as blood eosinophils, FeNO and IgE help stratify patients but overlap considerably and can mislead when used alone. The authors argue that multi-omics, advanced lung imaging, single-cell and spatial transcriptomics, and machine learning can define endotypes more precisely. They also introduce the new EAACI hypersensitivity nomenclature, including types V, VI and VII asthma, to support stratified management.
Resumen detallado
Asthma has long been managed as a single condition, yet patients differ widely in symptoms, triggers, inflammation and response to drugs. This narrative review explains why those differences matter and how classification is shifting toward biologically grounded, treatment-relevant categories: phenotypes (observable traits), endotypes (pathogenic mechanisms) and theratypes (response to therapy).
The authors review the current framework. Type 2 (T2)-high asthma involves IL-4, IL-5 and IL-13, IgE, mast cells and eosinophils, and these are targets of existing biologics. T2-low asthma is defined mainly by exclusion or by sputum neutrophils or paucigranulocytic inflammation. It is linked to Th1/Th17 activity, NLRP3 inflammasome activation, IL-1β and IL-17 signatures, neutrophil extracellular traps, gasdermin pathways, ORMDL3-related remodeling, and poor corticosteroid responsiveness. In one cohort 23.6% of patients were T2-low at enrollment, and the phenotype proved unstable over time. In low- and middle-income countries, 61% of cases were non-eosinophilic.
Biomarker-based classification is the practical starting point. Allergic asthma is typically defined by total IgE ≥30 kU/L plus specific IgE ≥0.35 kU/L, eosinophilic asthma by blood eosinophils ≥300 cells/µL, and T2-high by FeNO ≥35 ppb. These criteria overlap heavily: in a general adult population 78% met allergic criteria, and 59.2% of T2-high patients were also eosinophilic. FeNO and eosinophils reflect different cytokine pathways (IL-4/IL-13 versus IL-5) and are best used together. Confounders such as obesity, allergic rhinitis and corticosteroid exposure limit their reliability.
The review then describes precision tools. Transcriptomic, lipidomic and proteomic studies have revealed distinct clusters, including 9 sputum lipid phenotypes and 10 proteotypes in U-BIOPRED, 70 plasma proteins with putative causal roles, and airway microbiome signatures linked to neutrophilic disease. Single-cell and spatial methods show how IL-13 remodels the airway epithelium and how specialized lung niches drive remodeling. Imaging such as HRCT mucus scoring, SPECT, PET, MRI and OCT quantifies mucus plugging, ventilation defects and airway closure, which predict exacerbations and biologic response. Machine learning is proposed to integrate these data layers into individual molecular signatures. The paper also introduces the EAACI nomenclature, emphasizing tissue-driven (type V), metabolic/microbiome/epigenetic/neurogenic (type VI) and direct cellular activation (type VII) mechanisms.
The main implication is that stratified, mechanism-based asthma care should replace severity-only approaches. Caveats include imprecise T2-low definitions, biomarker overlap and variability, inconsistent breathomics data, and the limited clinical validation and access of many omics and imaging tools. The provided text is also truncated before the digital care section.
Hallazgos clave
- T2-high asthma (IL-4/5/13, IgE, eosinophils) is well defined and drug-targetable, while T2-low asthma remains poorly characterized and often steroid-resistant.
- Standard biomarker cut-offs overlap heavily: 59.2% of T2-high patients also met eosinophilic criteria, and 81.1% met allergic criteria.
- T2-low asthma was dynamic within patients; 23.6% were T2-low at enrollment, and 61% of cases in low- and middle-income countries were non-eosinophilic.
- Multi-omics uncovered sputum lipid phenotypes (9), proteotypes (10), causal plasma proteins (70) and microbiome patterns tied to neutrophilic asthma.
- Imaging of mucus plugs, ventilation defects and airway closure predicts exacerbation risk and biologic response, supporting machine-learning-integrated theratyping.
Metodología
This is a narrative review synthesizing clinical trial post hoc analyses, observational cohorts, multi-omics studies (e.g., U-BIOPRED), single-cell and spatial transcriptomics, and imaging research. It does not use a systematic search or formal quality appraisal. It also introduces the EAACI hypersensitivity nomenclature as a framework for classification.
Limitaciones del estudio
As a narrative review, it is subject to selection bias, and many cited findings come from post hoc analyses, small cohorts or exploratory omics studies that lack external validation. T2-low definitions are arbitrary and confounded by corticosteroid use, infection and exposures, and breathomics results are inconsistent. The supplied text is truncated, so the digital care and later EAACI type V–VII sections could not be fully assessed.
¿Te ha gustado este resumen?
Recibe la última investigación sobre longevidad en tu bandeja de entrada cada semana.
Introduce tu correo electrónico para suscribirte:
