Asthma affects more than 300 million people worldwide and remains one of the most complex chronic inflammatory disorders of the airways. Clinically, it presents a characteristic triad of bronchoconstriction, airway inflammation, and mucus hypersecretion, leading to symptoms ranging from nocturnal cough to severe dyspnea and wheezing.

While the pathogenesis involves a complex interplay of immune hyper-responsiveness and airway remodeling, the therapeutic landscape has long been dominated by targeting type 2 (T2) inflammatory axis (IL-4/5/13). However, the reality is that asthma is not a single disease. This has led to a critical shift in drug development: moving from a one-size-fits-all approach to a mechanism-based strategy that acknowledges the profound differences between T2 and non-T2 endotypes.

The classification of asthma into T2 (eosinophilic) and non-T2 (mixed granulocytic or neutrophil-dominant) is no longer just an academic exercise; it is the cornerstone of modern drug discovery. While T2 asthma is well-served by corticosteroids and biologics like Dupilumab, non-T2 asthma represents a massive unmet need (Table 1). Patients with non-T2 asthma are often insensitive to steroids and currently have few drug options other than Tezepelumab, the first biologic approved for asthma independent of T2 biomarkers. Promising clinical data for non-T2 asthma candidates like Astegolimab (anti-ST2) and Itepekimab (anti-IL-33) have shown significant efficacy in low-eosinophil populations, but this clinical progress creates an urgent preclinical challenge. Our data clearly show that therapeutic responses differ drastically between different preclinical asthma models. For instance, a standard OVA-induced mouse model is highly responsive to glucocorticoids, whereas an HDM+LPS-induced mouse model, characterized by neutrophilic infiltration, is profoundly resistant to glucocorticoids (Figure 1). Bridging the gap between bench research and these emerging non-T2 clinical outcomes requires a new generation of animal models that faithfully recapitulate the unique biology of non-eosinophilic asthma.

Table 1. The classification of asthma

Figure 1. Glucocorticoid response differences in type 2 and non-type 2 asthma models

(A) FACS analysis of BALF in OVA-induced asthma model. (B) FACS analysis of BALF in HDM+LPS-induced asthma model. Data are expressed as mean ± SEM, n=3-8. Statistics performed using One-way ANOVA. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

To empower precise, mechanism-of-action (MoA) driven drug discovery, GemPharmatech has developed a comprehensive asthma model matrix. We categorize our models based on two key clinical parameters: responsiveness to glucocorticoids and predominant immune cell infiltration. This matrix allows you to strategically select the most suitable model based on your drug’s MoA. For developers of a follow-on therapeutic targeting T2 asthma, benchmarking against Dupilumab's near-perfect efficacy in our HDM model is the recommended approach. For those targeting the unmet non-T2 space, our steroid-resistant, neutrophil-dominant asthma models provide a clinically relevant platform to demonstrate true differentiation.

Figure 2. Asthma model choice based on MoA analysis

In MoA-driven drug discovery, model-dependent efficacy is the most critical insight for asthma drug development. To illustrate, we performed a parallel study comparing the response to an anti-mouse IL-4RA antibody or Dupilumab across our T2 (HDM) and non-T2 (Alternaria alternata-induced; HDM+LPS-induced) asthma models (Figure 3, 4, 5). The results are striking and directly mirror clinical realities:

• In the classic T2 asthma model (HDM-induced), Dupilumab or anti-mouse IL-4RA treatment significantly inhibited hallmark T2 asthma pathologies. We observed a marked reduction in BALF eosinophils, and downregulation of key Th2 cytokines (IL-5, IL-33) at protein levels in the lung. Histological examination confirmed significantly reduced airway inflammation and epithelial thickening. These data establish the well-known "Dupilumab Wall”, a near-perfect efficacy ceiling that new T2 asthma targeted candidates will struggle to surpass.

• In striking contrast, in non-T2 asthma models (Alternaria alternata and HDM+LPS), anti-mouse IL-4RA treatment failed to suppress the dominant immune drivers (eosinophils and neutrophils) in BALF. More importantly, lung cytokine analysis revealed that anti-mouse IL-4RA did not reduce IL-5 or IL-33 levels—key drivers of inflammation in mixed/neutrophilic asthma models. In addition, anti-mouse IL-33 treatment, rather than anti-mouse IL-4RA treatment, significantly inhibited lung IL-5 levels in the Alternaria alternata-induced asthma model. Histology confirmed the limited efficacy of anti-mouse IL-4RA treatment in non-type 2 asthma models.

The head-to-head comparison delivers a clear message: the immune cell landscape dictates drug response. A therapeutic that is highly effective in an eosinophil-driven model can be completely ineffective in a mixed or neutrophil-driven model. This aligns perfectly with clinical observations where non-T2 asthma patients do not benefit from anti-mouse IL-4RA therapy. Therefore, selecting a model whose granulocytic profile matches the drug's target mechanism is not just important, it is essential for translational success.

Figure 3. Anti-mouse IL-4RA failed to inhibit BALF eosinophils in non-T2 asthma models

(A) eosinophils and neutrophils quantification of BALF FACS in HDM induced asthma model. (B) eosinophils and neutrophils quantification of BALF FACS in Alternaria alternata-induced asthma model. (C) eosinophils and neutrophils quantification of BALF FACS in HDM + LPS induced asthma model. Data are expressed as mean ± SEM, n=5-8. Statistics performed using One-way ANOVA. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure 4. Anti-mouse IL-4RA failed to inhibit lung IL-5 and IL-33 in non-T2 asthma models

(A) lung cytokine (IL-5, IL-33) quantification in HDM induced asthma model. (B) lung cytokine (IL-4, IL-5, IL-13, IL-33) quantification in Alternaria alternata-induced asthma model. (C) lung cytokine (IL-4, IL-5, IL-13, IL-33) quantification in HDM + LPS induced asthma model. Data are expressed as mean ± SEM, n=5-8. Statistics performed using One-way ANOVA. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure 5. Anti-mouse IL-4RA exerted limited efficacy in histology of non-T2 asthma models

(A) representative lung H&E staining images in HDM induced, Alternaria alternata-induced, and HDM + LPS induced asthma model. (B) histological examination results in HDM induced, Alternaria alternata-induced, and HDM + LPS induced asthma model. Data are expressed as mean ± SEM, n=5-8. Statistics performed using One-way ANOVA. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Summary: GemPharmatech’s asthma model portfolio

Figure 6 summarizes the key characteristics of our five core asthma models, providing a quick reference to guide your MoA-based model selection. Use this matrix to match your drug's mechanism – whether it targets T2 cytokines (IL-4/5/13), alarmins (IL-33/TSLP), or downstream effectors – to the model with the corresponding granulocytic and cytokine profile. This head-to-head comparison empowers you to de-risk your program and generates clinically translatable efficacy data from the start.

Figure 6. Comparison summary of asthma mouse models of GemPharmatech