Tezepelumab Reduces Alarmin Cytokine Levels in Upper Airway Epithelial Cells in Patients With Severe Asthma

Summary

• Nasal epithelial cell alarmin cytokine levels are reduced in patients with severe asthma following tezepelumab.

• This is the first report showing nasal alarmin cytokine protein expression depletion following biologics.

Epithelial cell‐derived alarmin cytokines, thymic stromal lymphopoietin (TSLP), interleukin (IL)‐25, and IL‐33, are upstream regulators of inflammatory cascades in asthma [1]. Elevated TSLP is associated with type 2 airway immune responses and severe asthma outcomes [2]. Tezepelumab, a monoclonal antibody (mAb) that binds to TSLP, reduces annualised asthma exacerbation rates, airway submucosal eosinophils, and airway hyperresponsiveness [3]. Here, we hypothesise that tezepelumab reduces TSLP‐mediated expression of alarmin cytokine levels in upper airway epithelial cells.

To test our hypothesis, we designed a prospective, observational, proof‐of‐concept study on the effects of tezepelumab on TSLP, IL‐25, and IL‐33 expression in nasal epithelial cells. We recruited 15 adult participants (> 18 years) with physician‐diagnosed asthma upon their initial visit to the clinic, who were selected for treatment with biologic therapy according to Canadian Thoracic Society guidelines [4]. Demographic and clinical data were collected at baseline (before treatment) and follow‐up (three months after the first two doses of tezepelumab), including age, sex, height, weight, and general and allergic comorbidities, particularly atopy and chronic rhinosinusitis (CRS). Patients received no oral corticosteroids within 3 months of the initial visit or between visits. This study was approved by an institutional human ethics review committee, and informed consent was obtained (Pro00104035).

Intracellular TSLP, IL‐25, and IL‐33 expression was measured in nasal epithelial cells at baseline and follow‐up [5]. In brief, we collected nasal samples by using cytology brushes (Medical Packaging Corporation, Camarillo, CA, USA) placed immediately into bronchial epithelial growth media (Lonza, Basel, Switzerland). Cells were labelled using Live/Dead Fixable Aqua Dead Cell Stain Kit (Invitrogen, ThermoFisher, Mississauga, ON, Canada), anti‐TSLP (rabbit polyclonal, Millipore Sigma, Oakville, ON, Canada), anti‐Ck8‐Alexa Fluor 647 (Abcam, Waltham, MA, USA), anti‐IL‐25 (R&D Systems, Minneapolis, MN, USA), and anti‐IL‐33‐phycoerythrin (R&D Systems). Secondary rat anti‐mouse IgG‐phycoerythrin‐cyanine‐7 (Invitrogen) and goat anti‐rabbit IgG‐Alexa Fluor 488 (Invitrogen) were used to detect IL‐25 and TSLP, respectively. Intracellular cytokine protein levels were measured as mean fluorescence intensity (MFI) (Figure 1A) [5].

FIGURE 1.

Alarmin cytokine expression in nasal epithelial cells of severe asthmatics pre‐ and post‐tezepelumab treatment. (A) Flow cytometry gating strategy to determine TSLP, IL‐25, and IL‐33 expression in cytokeratin 8⁺ epithelial cells. (B) TSLP, IL‐25, and IL‐33 expression in isolated nasal epithelial cells is shown as MFI pre‐ and post‐treatment with tezepelumab.

Descriptive variables are presented as mean (standard deviation [SD]) or frequency (%) for continuous and categorical variables, respectively. We used multivariable generalised estimating equations (GEE) with a Gaussian family and identity link functions with an exchangeable correlation matrix. We adjusted models by age, sex, BMI, smoking status, atopy (history of atopic diseases), comorbidities (allergic rhinitis, atopic dermatitis), and chronic rhinosinusitis. Model performance was evaluated using Quasi‐likelihood under the Independence model Criterion (QIC), and multicollinearity was assessed with the variance inflation factor (VIF). All analyses were performed using STATA (version 19.5). Receiver operating characteristic (ROC) curves and raincloud graphs were constructed using GraphPad Prism (v.11.0.0) and RainCloudPlots in R Statistical Software (v4.1.2), respectively.

Of participants, 9 (60%) were female, with a mean (SD) age of 48 (15) years. Ten (67%) were never smokers, 8 (53%) were atopic, 9 (60%) had CRS, and 10 (67%) reported having ≥ 1 comorbidities.

In multivariable analyses, we observed significant changes in intracellular cytokines following treatment with tezepelumab. Compared to baseline values (median: 36,812 MFI; interquartile range [IQR]: 28,313–89,661), TSLP expression decreased by 70%, with a post‐treatment median of 11,217 MFI (IQR: 7,333–19,291) (adjusted p‐value = 0.004) (Figure 1B). Similarly, for IL‐25, we found a decrease from the baseline median of 6586 MFI (IQR: 5,103–9,107) to a post‐treatment median of 2070 MFI (IQR: 1,619–4,744) (adjusted p‐value < 0.001). Levels of IL‐33 were reduced by 50%, with median values dropping from 1469 MFI to 703 MFI (adjusted p‐value = 0.01). In sensitivity analysis, we did not observe any collinearity among covariates, particularly among comorbidities, atopy, and CRS. Following tezepelumab treatment, TSLP and IL‐33 MFI each demonstrated a sensitivity of 80% and specificity of 90% in ROC analysis, while IL‐25 MFI exhibited 81% sensitivity and 71% specificity.

Our findings show that two doses of tezepelumab resulted in significant decreases in alarmin cytokines in nasal epithelial cells of patients with severe asthma. There are several postulated mechanisms by which blocking TSLP leads to decreased TSLP, IL‐25, and IL‐33 epithelial expression. Tezepelumab is proposed to break autocrine and paracrine feedback loops in airway epithelial cells, major sources of airway TSLP. TSLP binds through the heterodimeric receptor, TSLP receptor and IL‐7α receptor, on many cell types [6]. Receptor binding by TSLP also activates ILC2s (group 2 innate lymphoid cells) and Th2 cells to produce Th2 cytokines like IL‐4 and IL‐13, and IL‐13 further promotes epithelial production of IL‐25 and IL‐33 [7]. Clinical data show significant reductions in IL‐33 following tezepelumab treatment [8]. Finally, reduction of TSLP by tezepelumab may lead to decreased impairment of the integrity of tight junctions between epithelial cells, which then reduces “leakiness” of the epithelial cells and prevents excessive expression of IL‐25 and IL‐33 [9]. Future approaches may test this mechanism by air‐liquid interface nasal or bronchial epithelial cultures to determine if tezepelumab application to apical surfaces during cell activation (via Toll‐like receptor activation, for example) reduces expression of TSLP, IL‐25, and IL‐33. In vivo studies could also be carried out to determine TSLP‐mediated alarmin cytokine production in animal models.

One major advantage of this study is that we employed a non‐invasive sampling technique using nasal brushing to harvest upper airway epithelial cells, unlike other studies that used induced sputum, bronchoalveolar lavage (BAL), or bronchial brushing, which are technically more challenging to obtain. This also provides the added benefit of facilitating repeated sampling to study inflammation patterns in disease progression or response to treatments.

Limitations of this study are that samples were obtained from the nasal passages, which may not bear the same inflammatory signature as that seen in lower airways. Secondly, this was a proof‐of‐concept study; the sample size may be insufficient to achieve a significant effect size. Third, we considered only one follow‐up visit after treatment and were unable to determine longer trajectories of nasal alarmin cytokines. Lastly, as a single‐centre study, findings may not be generalisable to other populations due to other confounders. In summary, tezepelumab significantly reduces TSLP‐mediated elevations in alarmin cytokine levels in the upper airway epithelium of patients with severe asthma.

Author Contributions

H.M. performed laboratory investigations and wrote the manuscript. A.A. conceptualised the study, collected clinical data and revised the manuscript. S.M. performed statistical analysis, wrote the manuscript, and critically assessed the findings. P.L. conceptualised the study, acquired funding, and wrote and revised the manuscript.

Funding

This work was supported by AstraZeneca Canada.

Conflicts of Interest

H.M. and S.M. declare no conflicts of interest. A.A. reports conference travel support and/or honoraria from AstraZeneca, Covis Pharma, and GSK, and clinical trial support from AstraZeneca, outside of the submitted work. P. Lacy reports grants as well as personal fees from AstraZeneca Canada, AstraZeneca New Zealand, and GSK outside this work.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request. We are able to provide raw data from this study if requested.