Dupilumab Treatment is Associated with Improvement in Small Airway Dysfunction in Children with Moderate to Severe Asthma: A Retrospective Analysis

Abstract

Background

Small airway dysfunction (SAD) is common in pediatric asthma and is associated with poor disease control and frequent exacerbations. Current therapies show limited efficacy in improving small airway function. Dupilumab suppresses type 2 inflammation and may provide benefit; however, real-world evidence in Chinese children is limited. This study evaluated the association between dupilumab and improvements in SAD in children with uncontrolled moderate-to-severe asthma.

Methods

This retrospective cohort study without a concurrent comparator group included 106 children (aged 6–14 years) with uncontrolled moderate-to-severe asthma treated with dupilumab between January 2022 and March 2025. Dupilumab was generally administered at a dose of 300 mg every 4 weeks. The primary outcome was the proportion of patients with SAD at 6 months, defined as ≥2 of maximal mid-expiratory flow (MMEF_25–75), MEF₂₅, or MEF₅₀ <65% predicted. Secondary outcomes included changes in small airway function, type 2 inflammatory biomarkers, lung function, and exacerbation rates. Longitudinal changes were assessed using paired analyses, with mixed-effects models as sensitivity analyses.

Results

Among the 106 children included in the study, the proportion of patients with SAD decreased from 63.2% at baseline to 27.0% at 6 months. Small airway indices improved significantly, including MMEF_25–75% predicted (57.2% to 75.8%), MEF₂₅% predicted (48.3% to 66.4%), and MEF₅₀% predicted (58.7% to 76.1%) (all p<0.01). Type 2 inflammatory biomarkers were reduced, including fractional exhaled nitric oxide and serum IgE. Annualized exacerbation rates declined, accompanied by improved lung function and symptom control. Improvement in SAD was associated with greater increases in FEV₁/FVC and larger reductions in FeNO.

Conclusion

Dupilumab treatment is associated with amelioration of SAD in children with moderate-to-severe asthma, accompanied by reductions in type 2 inflammatory biomarkers.

Plain Language Summary

Asthma affects both the large and small airways in the lungs. Problems in the small airways are common in children with asthma and may be linked to poor asthma control and more frequent asthma attacks. However, there is still limited information on whether dupilumab, a treatment that targets inflammation involved in asthma, can improve small airway function in children.

We reviewed the medical records of 106 children aged 6–14 years with uncontrolled moderate-to-severe asthma who received dupilumab and were followed for up to 12 months. Six months after starting treatment, the proportion of children with small airway dysfunction decreased from 63.2% to 27.0%. Measures of small airway function also improved. These changes occurred alongside reductions in markers of airway inflammation, fewer asthma attacks, and better overall lung function and asthma control.

These findings suggest that dupilumab treatment is associated with improved small airway function in children with uncontrolled asthma. Assessing small airway function may also provide useful information when evaluating treatment response. Further studies with comparison groups are needed to confirm these findings.

Introduction

Asthma is a chronic airway inflammatory disease affecting the entire respiratory tract, from the central to peripheral airway.¹ Although conventional assessment of asthma severity and treatment response has largely focused on large-airway parameters, increasing evidence indicates that small airway dysfunction (SAD) represents an important component of asthma pathophysiology. SAD is frequently observed in children with asthma, including those with mild disease, and is associated with poor asthma control, increased exacerbation risk, and impaired quality of life.^2,3 In children, small airway abnormalities may be particularly relevant because airway inflammation and remodeling can occur early in the course of asthma,⁴ and persistent inflammation may contribute to long-term impairment of lung function.⁵ Current asthma therapies, primarily based on inhaled corticosteroids (ICS) combined with long-acting β₂-agonists (LABA) or leukotriene receptor antagonists (LTRA), effectively improve disease control in many patients but may not fully address distal airway abnormalities in some patients, partly due to challenges in achieving adequate peripheral airway exposure and complete control of airway inflammation.¹

Spirometric parameters, including maximal mid-expiratory flow (MMEF_25/75)% predicted, maximal expiratory flow at 25% of forced vital capacity (MEF₂₅)% predicted, and maximal expiratory flow at 50% of forced vital capacity (MEF₅₀)% predicted, are commonly used surrogate measures for assessing small airway function. Reduced MMEF_25/75% predicted has been associated with severe asthma phenotypes and correlates with type 2 inflammatory markers, including blood eosinophilia and fractional exhaled nitric oxide (FeNO), as well as enhanced bronchial hyperresponsiveness.⁶ Increasing evidence indicates that SAD is not merely a consequence of airflow limitation but is closely linked to underlying airway inflammatory processes. Type 2 inflammatory pathways mediated by cytokines such as interleukin (IL)-4, IL-5, and IL-13 may contribute to small airway abnormalities through eosinophilic infiltration, mucus hypersecretion, and airway remodeling in the distal airways. Consistently, type 2 inflammatory biomarkers, including blood eosinophils, sputum eosinophils, serum immunoglobulin E (IgE), and FeNO, have been associated with small-airway physiological abnormalities and are widely used to characterize T2-high phenotypes and guide the selection of targeted biologic therapies.^7,8 This growing recognition of type 2 inflammatory involvement in SAD has increased interest in biologic therapies with potential effects on distal airway abnormalities.

Dupilumab, a fully human monoclonal antibody targeting IL-4 receptor α, inhibits both IL-4 and IL-13 signaling,^9,10 thereby suppressing multiple components of type 2 inflammation. Although ICS-based therapies remain the cornerstone of asthma management, their effects on distal airway inflammation may be limited in some patients due to challenges in achieving adequate peripheral airway exposure. By providing targeted inhibition of type 2 inflammatory pathways, dupilumab may have the potential to improve inflammatory abnormalities throughout the airway, including the peripheral airways. Clinical trials have demonstrated the efficacy of dupilumab in reducing exacerbations and improving lung function in patients with moderate-to-severe asthma.¹¹ Emerging evidence from adult studies has also suggested potential benefits of dupilumab on small airway function.¹² In pediatric populations, clinical trial data, including findings from the VOYAGE study, have demonstrated improvements in lung function and asthma outcomes;¹³ however, the effects of dupilumab on small airway dysfunction, particularly in real-world pediatric populations, remain insufficiently characterized.

Our group previously established a retrospective cohort of children with moderate-to-severe asthma treated with dupilumab, and the overall effectiveness and safety findings from this cohort have been reported.¹⁴ With additional longitudinal follow-up data and additional eligible patients identified from the existing medical records, the current analysis uses an updated dataset and specifically evaluates small airway dysfunction and its longitudinal changes during dupilumab treatment. We hypothesized that dupilumab treatment would be associated with improvement in small airway function in children with uncontrolled moderate-to-severe asthma.

Methods

Study Population

This retrospective study analyzed electronic medical record data from the Guangzhou Institute of Respiratory Health for patients who started dupilumab treatment between January 2022 and March 2025. The study population was derived from the same ongoing retrospective real-world cohort as our previous analysis, with additional longitudinal follow-up data and additional eligible patients identified through further review of existing medical records.¹⁴ Dupilumab was generally administered at a dose of 300 mg every 4 weeks. Included patients were aged 6–14 years with uncontrolled moderate-to-severe asthma, classified according to GINA treatment-step criteria, despite medium-to-high dose ICS treatment and with type 2 inflammatory characteristics. Medium-to-high dose ICS was defined according to the age-specific daily ICS dose categories specified in the GINA strategy. Exclusion criteria included dupilumab use for non-asthma conditions, concurrent biologic therapies, and sublingual or subcutaneous desensitization treatments. Asthma severity was classified according to the Global Initiative for Asthma (GINA) criteria based on the treatment intensity required to achieve control. All eligible patients meeting the inclusion and exclusion criteria during the study period were included; therefore, no formal sample size calculation was performed. The Institutional Ethical Review Board of the First Affiliated Hospital of Guangzhou Medical University approved the study (No. ES-2024-001-01). Since it was an anonymized and retrospective research, written informed consent was waived in accordance with the Declaration of Helsinki.

Data Collection

The date of the first dupilumab injection was designated the index date. The observation period spanned from the index date to each patient’s final visit. Baseline assessments were defined as the clinical and laboratory measurements obtained before or at the initiation of dupilumab treatment. Follow-up assessments were evaluated at approximately 3, 6, 9, and 12 months after treatment initiation based on available clinical visit data. Data extracted from the routine medical care included demographics, clinical characteristics, pulmonary function test results (including small airway function indicators MMEF_25/75% predicted, MEF₂₅% predicted and MEF₅₀% predicted), serum biochemical indicators, asthma control assessments (Children-Asthma Control Test[c-ACT] and Asthma Control Questionnaire 7 Interviewer-Administered [ACQ-7-IA]), treatment patterns. All data came from the Hospital Information System/Laboratory Information System. The c-ACT and ACQ-7-IA were used to assess asthma control. Higher c-ACT scores indicate better asthma control, whereas ACQ-7-IA scores range from 0 to 6, with higher scores indicating poorer asthma control; a 0.5-point change in ACQ-7-IA is considered clinically meaningful. We got the permission to use ACQ-7-IA form from the author via http://www.qoltech.co.uk/index.htm. The c-ACT scores form was obtained from https://eprovide.mapi-trust.org/instruments/childhood-asthma-control-test.

Study Endpoints

The primary endpoint of this study was the change in the proportion of patients meeting SAD criteria from baseline to 6 months. SAD was defined according to the 2021 Chinese expert consensus on the assessment and treatment of small airway dysfunction in children with asthma, as the presence of at least two of the following three small airway parameters below 65% of predicted values: MMEF_25/75, MEF₂₅, and MEF₅₀. The 6-month time point was selected as the primary assessment because it represents a clinically relevant intermediate-term interval for evaluating changes in airway function after treatment initiation and, in this retrospective real-world cohort, also provided the most consistent availability of follow-up data.

Secondary endpoints encompassed several domains, including: 1) The change of the percentage of patients with SAD from baseline to different follow-up timepoints (3, 9 and 12 months) and changes in small-airway function indicators (MMEF_25/75% predicted, MEF₂₅% predicted and MEF₅₀% predicted) from baseline to different follow-up timepoints (3, 6, 9 and 12 months); 2) Changes in type 2 inflammation biomarkers (FeNO, serum total IgE, blood eosinophil count, and sputum eosinophil percentage) from baseline to different follow-up timepoints (3, 6, 9 and 12 months); 3) Changes in annualized asthma exacerbation rate (calculated as the total number of exacerbations divided by total person-years of follow-up), pulmonary function parameters (including pre-BD FEV₁, pre-BD FEV₁% predicted, pre-BD FEV₁/FVC% predicted, pre-BD FVC, and pre-BD FVC% predicted), asthma symptom scores (c-ACT and ACQ-7-IA) and rescue medication use from baseline to 6 months; 4) Factors associated with the SAD improvement; 5) Comparison of baseline characteristics between patients with and without SAD improvement. For analyses of small airway function improvement, patients without SAD at baseline and patients with SAD at baseline who subsequently transitioned to non-SAD status during follow-up were classified into the Improved group. Patients with SAD at baseline who remained in SAD status throughout follow-up were classified into the Not Improved group. For patients transitioning from SAD to non-SAD during follow-up, the first visit at which non-SAD status was observed was defined as the improvement visit.

Statistical Analysis

All statistical analyses were conducted using R 4.2.3 (R Foundation for Statistical Computing, Vienna, Austria). Continuous variables were presented as mean ± standard deviation (SD) for normally distributed data and median (interquartile range, IQR) for non-normally distributed data. Categorical variables were summarized as frequencies and percentages. Normality was assessed using the Shapiro–Wilk test.

For longitudinal outcomes (small airway function and type 2 inflammation biomarkers), changes from baseline to each follow-up visit were evaluated using paired t-test for normally distributed continuous variables, and Wilcoxon signed-rank test for non-normally distributed variables. McNemar’s test was used to compare SAD status between baseline and each follow-up visit among patients with paired observations. Wilson score 95% confidence intervals were calculated for SAD prevalence at baseline and 6 months. To address potential confounding, mixed-effects models for repeated measures (MMRM) were performed as sensitivity analyses, incorporating fixed effects for time, baseline outcome value, sex, age, and body mass index (BMI), with an unstructured covariance matrix. Degrees of freedom were estimated using the Kenward-Roger method. Pairwise comparisons were adjusted for multiple testing using the Bonferroni method.

For outcomes assessed at baseline and 6 months, including annualized asthma exacerbation rate, conventional pulmonary function parameters, and asthma symptom scores, changes from baseline were evaluated using paired t-tests or Wilcoxon signed-rank tests as appropriate, with analysis of covariance (ANCOVA) as a sensitivity adjustment.

Multivariable logistic regression was used to identify factors associated with small airway function improvement. Patients without SAD at baseline and patients with SAD at baseline who transitioned to non-SAD status during follow-up were classified into the Improved group. Patients with SAD at baseline who remained in SAD status throughout follow-up were classified into the Not Improved group. For patients who transitioned from SAD to non-SAD status, the first visit at which non-SAD status was observed was used as the improvement visit. Separate multivariable logistic regression models were fitted for each candidate variable, with sex, age, and BMI included as covariates in each model. This approach was used to limit model overfitting and multicollinearity given the modest sample size and correlations among candidate lung function and inflammatory variables. Key predictors included baseline pre-BD FEV₁% predicted, pre-BD FEV₁/FVC% predicted, FeNO, blood eosinophil count, total IgE, and ACQ-7-IA score and changes from baseline to follow-up (Δ pre-BD FEV₁% predicted, Δ pre-BD FEV₁/FVC% predicted, Δ FeNO, Δ blood eosinophil count, Δ total IgE, and Δ ACQ-7-IA score). Results are presented as odds ratios (OR) with 95% confidence intervals (CI).

All statistical tests were two-sided; P < 0.05 was considered statistically significant. Missing data were not imputed, and analyses were based on available data at each time point.

Results

Baseline Characteristics

Of 106 enrolled children with moderate-to-severe asthma (mean age 8.7±2.6 years; 74.5% male), 75.5% had comorbid allergic rhinitis and 15.1% had atopic dermatitis. Mean c-ACT was 19.61±2.4, ACQ-7-IA score was 2.72±1.54, and pre-BD FEV₁ was 1.54±0.54 L at baseline. Patients averaged 2.0±1.1 asthma exacerbations in the previous year (Table 1).

Table 1.

Demographic and Clinical Characteristics of the Study Cohort at Baseline

Characteristic	All Patients N=106
General Characteristic
Male, n (%)	79 (74.53%)
Age, years, mean ± SD	8.71±2.57
BMI, kg/m2, mean ± SD	17.17±3.50
Comorbidities, n (%)
Atopic dermatitis	16 (15.09%)
Allergic rhinitis	80 (75.47%)
Asthma duration, months, median (IQR)	39.00 (24.00, 72.00)
c-ACT score, points, mean ± SD	19.61±2.40
ACQ-7-IA score, points, mean ± SD	2.72±1.54
Asthma exacerbations in past year, times, mean ± SD	2.03±1.10
≥ 1 hospitalization for exacerbation, n (%)	27 (25.47%)
Pulmonary function test
Pre-BD FEV1, L, mean ± SD	1.54±0.54
Pre-BD FEV1% predicted, %, mean ± SD	85.93±13.38
Pre-BD FEV1/FVC, %, mean ± SD	76.01±9.75
Pre-BD FEV1/FVC% predicted, %, mean ± SD	88.96±11.13
Small airway function
MMEF25/75, L/s, mean ± SD	1.30±0.60
MMEF25/75% predicted, %, mean ± SD	57.20±21.31
MEF25, L/s, mean ± SD	0.64±0.35
MEF25% predicted, %, mean ± SD	48.28±21.05
MEF50, L/s, mean ± SD	1.53±0.67
MEF50% predicted, %, mean ± SD	58.68±21.66
Small airway dysfunction, n (%)	67 (63.21%)
Type 2 inflammation biomarkers
FeNO, ppb, median (IQR)	26.00 (13.00, 46.50)
Serum total IgE, IU/mL, median (IQR)	721.34 (294.75, 1330.94)
Blood eosinophil count, ×106/L, median (IQR)	610.00 (400.00, 840.00)
Sputum eosinophil percentage, %, median (IQR)	6.00 (2.50, 21.50)

Abbreviations: SD, standard deviation; BMI, body mass index; IQR, interquartile range; c-ACT, childhood asthma control test; ACQ-7-IA, 7-item Asthma Control Questionnaire, interviewer-administered; Pre-BD, pre-bronchodilator; FEV₁, forced expiratory volume in 1 second; FEV₁/FVC, forced expiratory volume in 1 second to forced vital capacity ratio; MMEF_25/75, maximal mid-expiratory flow between 25% and 75% of forced vital capacity; MEF₂₅, maximal expiratory flow at 25% of forced vital capacity; MEF₅₀, maximal expiratory flow at 50% of forced vital capacity; FeNO, fractional exhaled nitric oxide; IgE, immunoglobulin E.

Changes of the Percentage of SAD and Changes in Small Airway Function Indicators

Following dupilumab initiation, a decrease in the proportion of SAD patients was observed (from 63.2% at baseline [67/106, 95% CI: 53.7–71.8%] to 27.0% at 6 months [17/63, 95% CI: 17.6–39.0%]). Among the 63 patients with paired baseline and 6-month measurements, the corresponding proportions were 66.7% (42/63) and 27.0% (17/63), respectively (p < 0.001). Compared with baseline, the proportion was also lower at 3, 9, and 12 months (39.4% [39/99], 27.0% [10/37], and 16.7% [4/24], respectively), with each comparison based on patients with paired baseline and corresponding follow-up measurements (Figure 1a).

Figure 1.

Longitudinal changes of percentage of patients with small airway dysfunction and changes in small airway function indicators following dupilumab intervention over a 12-month period. (a) Percentage of patients with small airway dysfunction (b) MMEF_25/75% predicted, maximum mid-expiratory flow (between 25% and 75% of FVC) of the predicted value, (c) MEF₂₅% predicted, maximal expiratory flow at 75% of FVC of the predicted value, (d) MEF₅₀% predicted, maximal expiratory flow at 50% of FVC of the predicted value. All values represent mean percent predicted (± error bars). Percentages and summary statistics at each follow-up time point were calculated using the paired sample available at that time point (ie, patients with measurements at both baseline and the corresponding follow-up). Statistical significance is indicated as follows: ***p < 0.001. P values were derived from paired comparisons with baseline using the paired sample available at each follow-up time point.

Abbreviations: MMEF_25/75, maximal mid-expiratory flow between 25% and 75% of forced vital capacity; MEF₂₅, maximal expiratory flow at 25% of forced vital capacity; MEF₅₀, maximal expiratory flow at 50% of forced vital capacity; m, months.

All small airway indicators improved significantly compared with baseline at each follow-up time point (all p < 0.001). MMEF_25/75% predicted improved from 57.2±21.3% to 72.6±22.1%, 75.8±20.3%, 76.1±18.0%, and 77.9±19.1% at 3, 6, 9, and 12 months respectively. MEF₂₅% predicted rose from 48.3±21.1% from to 62.9±21.8%, 66.4±20.0%, 66.4±19.2%, and 69.7±25.8%. And MEF₅₀% predicted increased from 58.7±21.7% to 72.8±23.6%, 76.1±20.2%, 77.5±19.2%, and 76.8±19.6% (Figure 1b–1d).

MMRM analysis confirmed these findings, demonstrating significant least square mean (LSM) changes from baseline for all small airway parameters at each time point (all p < 0.001). The LSM change in MMEF_25/75% predicted was 16.9% at 3 months, increasing to 20.7% at 6 months, 24.5% at 9 months, and 23.9% at 12 months. For MEF₂₅% predicted, the LSM changes were 15.8%, 20.0%, 22.7%, and 22.8% at 3, 6, 9, and 12 months, respectively. MEF₅₀% predicted showed LSM changes of 15.7%, 20.3%, 22.7%, and 21.1% at the corresponding time points (Supplementary Figure 1). Notably, the greatest improvements were observed during the first 6 months of treatment, with relatively stable values maintained thereafter.

Changes in Type 2 Inflammation Biomarkers

A significant decline in all type 2 inflammation biomarkers was observed after dupilumab treatment. FeNO decreased from 26.00 (13.00, 46.50) ppb at baseline to 12.00 (8.00, 19.00) ppb at 3 months, then stabilized between 10.00–12.50 ppb through 12 months. Serum total IgE fell from 721.34 (294.75, 1330.94) IU/mL to 317.00 (115.16, 706.96) IU/mL, 162.81 (90.33, 297.70) IU/mL, 135.07 (52.59, 249.87) IU/mL and 118.09 (60.56, 201.60) IU/mL at 3, 6, 9 and 12 months. Blood eosinophil count dropped from 610.00 (400.00, 840.00) × 10⁶/L to 500.00 (270.00, 700.00) ×10⁶/L, 385.00 (260.00, 675.00) ×10⁶/L, 350.00 (170.00, 550.00) ×10⁶/L and 375.00 (285.00, 570.00) ×10⁶/L. And sputum eosinophil percentage declined from 6.00 (2.50, 21.50) % to 1.25 (0.50, 3.56) %, 1.00 (0.00, 2.00) %, 0.50 (0.00, 1.50) % and 1.00 (0.02, 3.13) %. Most improvements occurred within three months and were maintained thereafter (Figure 2).

Figure 2.

Longitudinal changes in type 2 inflammation biomarkers following dupilumab intervention. The box plots (a-d) demonstrate the temporal evolution of four key type 2 inflammation biomarkers over 12 months of treatment. (a) FeNO levels (ppb). (b) Serum total IgE concentrations (IU/mL). (c) Blood eosinophil counts (×10⁶/L). (d) Sputum eosinophil percentage (%). All four biomarkers showed significant reduction from baseline at all follow-up time points (3, 6, 9, and 12 months). Statistical significance is indicated as follows: ***p < 0.001; **p < 0.01, *p < 0.05. P values were derived from paired comparisons with baseline using the paired sample available at each follow-up time point. Boxes represent interquartile range (25th-75th percentile), horizontal lines within boxes indicate median values, and whiskers extend to 1.5× interquartile range.

Abbreviations: FeNO, Fractional exhaled nitric oxide; IgE, immunoglobulin E; m, months.

MMRM analysis corroborated these findings after adjustment for baseline values, sex, age, and BMI. FeNO demonstrated significant LSM reductions of 20.4, 23.3, 23.0, and 21.9 ppb at 3, 6, 9, and 12 months, respectively (all p < 0.001; mean baseline: 36.0 ppb). Serum total IgE showed progressive LSM decreases of 585.1, 817.6, 874.3, and 936.7 IU/mL at the corresponding time points (all p < 0.001; mean baseline: 1016.1 IU/mL). Blood eosinophil count exhibited LSM reductions of 46.0, 159.9, 250.6, and 185.6 × 10⁶/L (p < 0.001 at 6 and 9 months; mean baseline: 640.0 × 10⁶/L). Sputum eosinophil percentage demonstrated LSM decreases of 13.4%, 14.3%, 13.6%, and 16.1% at 3, 6, 9, and 12 months (all p < 0.001; mean baseline: 15.1%) (Supplementary Figure 2).

Changes in Annualized Asthma Exacerbation, Pre-BD FEV₁, Asthma Symptom Score and Rescue Medication Use

Following dupilumab initiation, the annualized asthma exacerbation rate significantly decreased from 2.03 ± 1.10 at baseline to 0.37 ± 1.32 at 6 months, with a least square mean (LSM) difference of −1.18 (95% CI: −1.48 to −0.88; p < 0.001). Hospitalization rates decreased from 25.5% to ≤1%. Pre-BD FEV₁ showed significant improvement, increasing from 1.54 ± 0.54 L to 1.92 ± 0.57 L (LSM difference: 0.31 L; 95% CI: 0.26–0.36; p < 0.001), corresponding to an increase in percent predicted from 85.93 ± 13.38% to 97.64 ± 10.51% (LSM difference: 13.07%; 95% CI: 11.14–15.00; p < 0.001). Disease control also demonstrated sustained improvement: C-ACT scores increased from 19.61 ± 2.40 to 26.32 ± 1.39 (LSM difference: 6.70; 95% CI: 6.48–6.93; p < 0.001), while ACQ-7-IA scores decreased from 2.72 ± 1.54 to 1.58 ± 0.67 (LSM difference: −1.53; 95% CI: −1.69 to −1.37; p < 0.001), indicating improved symptom control(Table 2). Furthermore, rescue medication use reduced substantially: SABA from 31.1% to 4.0%, SAMA from 18.9% to 1.0%, and oral corticosteroid (OCS) from 10.4% to 0.0% at 6 months.

Table 2.

Changes in Annualized Asthma Exacerbation, Pulmonary Function Parameters, Asthma Symptom Score and Rescue Medication Use

Characteristic	Baseline, Mean (SD)	6 Months, Mean (SD)	Least Square Mean Difference (SE)	95% CIa	p-valuea
Annualized asthma exacerbation
Annualized asthma exacerbation, times	2.03±1.10	0.37±1.32	−1.18 (0.15)	(−1.48, −0.88)	<0.001
Pulmonary function parameters
Pre-BD FEV1, L	1.54±0.54	1.92±0.57	0.31 (0.02)	(0.26, 0.36)	<0.001
Pre-BD FEV1% predicted, %	85.93±13.38	97.64±10.51	13.07 (0.98)	(11.14, 15.00)	<0.001
Pre-BD FEV1/FVC, %	76.01±9.75	84.19±7.91	8.93 (0.61)	(7.72, 10.15)	<0.001
Pre-BD FVC, L	2.06±0.73	2.32±0.80	0.18 (0.02)	(0.14, 0.21)	<0.001
Pre-BD FVC % predicted, %	88.96±11.13	96.79±8.03	8.78 (0.64)	(7.50, 10.05)	<0.001
Asthma symptom score
C-ACT, points	19.61±2.40	26.32±1.39	6.70 (0.11)	(6.48, 6.93)	<0.001
ACQ-7-IA, points	2.72±1.54	1.58±0.67	−1.53 (0.08)	(−1.69, −1.37)	<0.001

Notes: ^a Adjusted least square mean differences, 95% confidence intervals, and P values were estimated using analysis of covariance (ANCOVA) models adjusted for baseline value, sex, age, and BMI. Changes from baseline to 6 months were also assessed using paired t-tests for normally distributed continuous variables or Wilcoxon signed-rank tests for non-normally distributed continuous variables, as appropriate.

Abbreviations: Pre-BD, pre-bronchodilator; FEV₁, forced expiratory volume in 1 second; FVC, forced vital capacity; FEV₁/FVC, forced expiratory volume in 1 second to forced vital capacity ratio; SD, standard deviation; SE, standard error; CI, confidence interval; c-ACT, childhood asthma control test; ACQ-7-IA, 7-item Asthma Control Questionnaire, interviewer-administered.

Factors Associated with SAD Improvement

Multivariable logistic regression analysis, adjusted for sex, age, and BMI, was performed to identify factors associated with improvement in small airway dysfunction (Table 3).

Table 3.

Factors Influencing SAD Improvement

Variable	OR	95% CI	p-valuea
Type 2 inflammation biomarkers
Δ FeNO, ppb	1.02	1.00, 1.06	0.049
Δ Serum total IgE, IU/mL	1.00	0.99, 1.00	0.157
Δ Blood eosinophil count, ×106/L	1.00	0.98, 1.01	0.576
Δ Sputum eosinophil percentage, %	1.01	0.97, 1.07	0.640
Annualized asthma exacerbation
Baseline annualized asthma exacerbation, times	1.02	0.67, 1.62	0.913
Δ Annualized asthma exacerbation, times	0.81	0.55, 1.18	0.281
Pulmonary function test
Δ Pre-BD FEV1, L	0.99	0.97, 1.01	0.297
Δ Pre-BD FEV1% predicted, %	0.98	0.94, 1.02	0.292
Δ Pre-BD FEV1/FVC, %	0.96	0.90, 1.03	0.204
Δ Pre-BD FEV1/FVC% predicted, %	1.27	1.09, 1.51	0.003
Asthma symptom score
Baseline c-ACT score, points	1.52	1.19, 2.05	< 0.001
Δ c-ACT score, points	1.02	0.66, 1.47	0.905
Baseline ACQ-7-IA score, points	0.69	0.50, 0.91	0.010
Δ ACQ-7-IA score, points	1.76	0.44, 7.49	0.416

Notes: ^a P values were derived using Wald tests from separate multivariable logistic regression models. Each candidate variable was evaluated in a separate model adjusted for sex, age, and BMI. Δ indicates change from baseline to follow-up (positive values indicate increase). OR > 1 indicates higher odds of SAD improvement. Bold values indicate statistical significance (p < 0.05).

Abbreviations: OR, odds ratio; CI, confidence interval; FeNO, fractional exhaled nitric oxide; IgE, immunoglobulin E; Pre-BD, pre-bronchodilator; FEV₁, forced expiratory volume in 1 second; FEV₁/FVC, forced expiratory volume in 1 second to forced vital capacity ratio; c-ACT, childhood asthma control test; ACQ-7-IA, 7-item Asthma Control Questionnaire, interviewer-administered.

Among pulmonary function changes during follow-up, improvement in Δ Pre-BD FEV₁/FVC % predicted was significantly associated with SAD improvement (OR = 1.27, 95% CI: 1.09–1.51, p = 0.003), indicating that each 1% increase in FEV₁/FVC % predicted was associated with a 27% higher odds of SAD improvement. Additionally, reduction in FeNO (Δ FeNO) was marginally associated with SAD improvement (OR = 1.02, 95% CI: 1.00–1.06, p = 0.049).

Baseline asthma control status was also associated with SAD improvement. Higher baseline C-ACT scores (indicating better asthma control) were associated with greater likelihood of SAD improvement (OR = 1.52, 95% CI: 1.19–2.05, p < 0.001), with each 1-point increase in baseline C-ACT score associated with 52% higher odds of improvement. Conversely, higher baseline ACQ-7-IA scores (indicating poorer asthma control) were associated with lower likelihood of SAD improvement (OR = 0.69, 95% CI: 0.50–0.91, p = 0.010).

Characteristics of Patients with Improvement in Small Airway Function and Those Without Improvement in Small Airway Function

In this study, we further compared baseline characteristics between patients who demonstrated SAD improvement and those who did not (Table 4). Overall, patients in the “improved” group exhibited a more favorable baseline clinical profile. They had better asthma control, reflected by higher c-ACT scores (p < 0.001) and lower ACQ-7-IA scores (p = 0.004). Baseline pulmonary function was also less impaired in this group, with higher values for pre-BD FEV₁% predicted (p < 0.001) and FEV₁/FVC (p < 0.001), as well as consistently better small-airway indicators including MMEF_25/75% predicted, MEF₂₅% predicted, and MEF₅₀% predicted (all p < 0.001).

Table 4.

Baseline Clinical Characteristics of Patients with Improvement in Small Airway Function and Those Without Improvement in Small Airway Function

Characteristic	Improved N = 82	Not Improved N = 24	p-valuea
General Characteristic
Male, n (%)	61 (74.39%)	18 (75.00%)	0.952
Age, years, mean ± SD	8.55±2.65	9.25±2.21	0.208
BMI, kg/m2, mean ± SD	16.85±2.87	18.26±5.04	0.637
Comorbidities, n (%)
Atopic dermatitis	11 (13.41%)	5 (20.83%)	0.352
Allergic rhinitis	66 (80.49%)	14 (58.33%)	0.027
Asthma duration, months, median (IQR)	36.00 (12.00, 72.00)	48.00 (31.5, 79.00)	0.095
c-ACT score, points, mean ± SD	20.09±1.86	18.00±3.27	<0.001
ACQ-7-IA score, points, mean ± SD	2.50±1.45	3.46±1.64	0.004
Asthma exacerbations in past year, times, mean ± SD	2.05±1.15	1.96±0.91	0.981
≥1 hospitalization for exacerbation, n (%)	23 (28.05%)	4 (16.67%)	0.260
Pulmonary function test
Pre-BD FEV1, L, mean ± SD	1.56±0.55	1.49±0.49	0.853
Pre-BD FEV1% predicted, %, mean ± SD	88.70±12.55	76.47±11.90	<0.001
Pre-BD FEV1/FVC, %, mean ± SD	78.38±8.73	67.90±8.78	<0.001
Pre-BD FEV1/FVC% predicted, %, mean ± SD	91.63±10.03	79.83±9.97	<0.001
Small airway function
MMEF25/75, L/s, mean ± SD	1.39±0.62	0.96±0.36	0.001
MMEF25/75% predicted, %, mean ± SD	62.34±20.81	39.65±11.44	<0.001
MEF25, L/s, mean ± SD	0.70±0.37	0.45±0.17	<0.001
MEF25% predicted, %, mean ± SD	53.09±20.96	31.84±10.52	<0.001
MEF50, L/s, mean ± SD	1.64±0.69	1.15±0.44	0.002
MEF50% predicted, %, mean ± SD	63.75±21.37	41.35±11.36	<0.001
Type 2 inflammation biomarkers
FeNO, ppb, median (IQR)	26.00 (13.00, 49.00)	26.00 (11.50, 43.00)	0.706
Serum total IgE, IU/mL, median (IQR)	743.89 (292.67, 1303.00)	506.29 (285.00, 1525.13)	0.766
Blood eosinophil count, ×106/L, median (IQR)	625.00 (400.00, 850.00)	535.00 (405.00, 770.00)	0.394
Sputum eosinophil percentage, %, median (IQR)	7.25 (3.50, 25.75)	4.50 (1.50, 13.00)	0.121

Notes: ^a P values were derived using the chi-square test of independence, Wilcoxon rank-sum test, or Fisher’s exact test, as appropriate.

Abbreviations: SD, standard deviation; BMI, body mass index; IQR, interquartile range; c-ACT, childhood asthma control test; ACQ-7-IA, 7-item Asthma Control Questionnaire, interviewer-administered; Pre-BD, pre-bronchodilator; FEV₁, forced expiratory volume in 1 second; FEV₁/FVC, forced expiratory volume in 1 second to forced vital capacity ratio; MMEF_25/75, maximal mid-expiratory flow between 25% and 75% of forced vital capacity; MEF₂₅, maximal expiratory flow at 25% of forced vital capacity; MEF₅₀, maximal expiratory flow at 50% of forced vital capacity; FeNO, fractional exhaled nitric oxide; IgE, immunoglobulin E.

Discussion

In this study, dupilumab treatment was associated with a significant reduction in the proportion of pediatric asthma patients with SAD and with improvements in small airway function indicators. Persistent SAD despite conventional inhaled therapy remains an important clinical challenge in children with uncontrolled moderate-to-severe asthma. This result aligns with previous studies. In the VESTIGE study, dupilumab was associated with improvements in small airway function in adults with moderate-to-severe asthma,¹² while another study involving 25 pediatric patients reported similar trends in lung function parameters, including small airway indices.¹⁵ The present study extends these findings by providing longitudinal evidence from a larger real-world pediatric cohort, with systematic evaluation of SAD status and multiple small airway function parameters during dupilumab treatment.

After dupilumab was added to background ICS–LABA therapy, improvements in small airway function were observed early during follow-up, with the greatest changes occurring within the first 3–6 months and generally maintained through the available 12-month follow-up. Beyond the reduction in SAD prevalence, all small-airway function indicators—including MMEF_25/75% predicted, MEF₂₅% predicted, and MEF₅₀% predicted—showed substantial and sustained improvement. These findings corroborate and extend the limited pediatric evidence on dupilumab and small-airway function. In the phase-III VOYAGE trial, MMEF_25/75 significantly improved by 0.60–0.65 L/s in children (6–11 years) with uncontrolled moderate-to-severe asthma after 52 weeks of dupilumab treatment versus placebo.¹¹ The dosing regimen in the present real-world cohort, which generally administered at 300 mg every 4 weeks, differed from the weight-based dosing schedules used in previous pediatric trials. Although improvements in lung function were observed across these different treatment settings, the present study was not designed to compare dosing regimens, and their potential differential effects on efficacy or safety cannot be determined from our data. Improvement in small-airway function may be clinically relevant, as reduced MMEF_25/75 has been associated with future exacerbation risk independent of FEV_1.¹⁶ Asthma exacerbations remain an important cause of morbidity and healthcare utilization in children, and substantial variability persists in their clinical management across existing guidelines.¹⁷ Consistent with the improvements in small-airway function observed in this cohort, exacerbation rates and rescue medication use also decreased during follow-up. However, given the retrospective design and absence of a comparator group, these concurrent improvements should not be interpreted as evidence that restoration of small-airway function directly mediated the observed clinical benefits.

Type 2 inflammation biomarkers also declined significantly following treatment. The rapid reduction in FeNO, together with decreases in serum IgE, blood eosinophils, and sputum eosinophils, aligns with the known pharmacologic mechanism of dupilumab. Blockade of IL-4Rα inhibits both IL-4- and IL-13-mediated processes, reducing eosinophilic inflammation, mucus hypersecretion, and airway epithelial nitric oxide production—processes closely related to small-airway narrowing and remodeling.^18,19 Notably, sputum eosinophil percentage, an airway inflammatory marker that has been less frequently reported in pediatric asthma studies, was systematically reported in this study. Sputum eosinophil percentage decreased from 6.00 (2.50, 21.50) % at baseline to 1.00 (0.02, 3.13) % at 12 months, and remained at a low level after the first 3 months of treatment. Compared with peripheral blood eosinophil, sputum eosinophils may more directly reflect the local airway type 2 inflammatory state.²⁰ Blood and sputum eosinophils, serum IgE, and FeNO are also recognized as clinically relevant non-invasive biomarkers of type 2 inflammation in children with severe asthma.²¹ Although both sputum and blood eosinophils showed downward trends during follow-up, the sputum eosinophil percentage decreased more rapidly and reached near normal levels (< 2.5%) earlier than blood eosinophil count. This difference may reflect distinct temporal patterns of inflammatory changes in the airway and peripheral blood following dupilumab treatment. However, this interpretation should be considered exploratory because sputum assessments were available in fewer patients and induced sputum measurements may show greater variability in children.

Multivariable logistic regression model found that improvement in pre-BD FEV₁/FVC% predicted was significantly associated with small airway function improvement, while reduction in FeNO showed a marginal association with improvement. The relationship between FeNO and small-airway function is consistent with previous pediatric studies. One study of 248 children with asthma reported negative correlations between FeNO and MMEF_25/75% predicted, MEF₂₅% predicted, and MEF₅₀% predicted (r = –0.46 to –0.49, p < 0.001).²² Another study similarly reported a positive correlation between elevated FeNO and SAD (r = 0.52, p < 0.0001).²³ These findings support a potential relationship between type 2 airway inflammation and small-airway dysfunction. Dupilumab also attenuates IL-13–related processes involved in airway remodeling.²⁴ The observed marginal association between FeNO reduction and MMEF_25/75 improvement (OR 1.02 per 1-ppb decrease, 95% CI 1.00–1.06) is consistent with the hypothesis that suppression of IL-4/IL-13-mediated airway inflammation may contribute to improved peripheral airway function.²⁵ However, given the modest effect size, borderline statistical significance, and exploratory nature of the analysis, this finding should not be interpreted as establishing FeNO as a predictor of SAD improvement and requires confirmation in larger prospective studies.

In addition, comparison of baseline characteristics between patients with and without small airway function improvement showed that patients in the Improved group had better asthma control, reflected by higher c-ACT scores and lower ACQ-7-IA scores, as well as less impaired baseline lung function, including higher pre-BD FEV₁% predicted and pre-BD FEV₁/FVC. These findings suggest that baseline disease status may influence the likelihood of subsequent small airway function improvement during dupilumab treatment. One possible explanation is that patients with relatively preserved baseline function may have a greater component of reversible inflammatory dysfunction and therefore greater potential for functional recovery after suppression of type 2 inflammation. Conversely, patients with more severe baseline impairment may have a greater contribution from established airway remodeling, which could limit functional recovery even when inflammation is effectively controlled. However, these interpretations remain speculative because airway remodeling was not directly assessed in this study. Therefore, although the findings raise the possibility that earlier intervention before substantial structural airway changes may be beneficial, the present study cannot define the optimal timing of biologic treatment or establish specific lung-function or biomarker thresholds for treatment initiation.

This study has some limitations. As a single-center retrospective cohort study without a concurrent control group, the observed improvements in SAD and type 2 inflammation cannot be definitively attributed to dupilumab alone. Natural disease evolution, regression to the mean, changes in medication adherence, differential follow-up intensity, and other unmeasured confounding factors may have influenced the findings. Second, the sample size was modest, and the number of patients with available data decreased at later follow-up time points. Missing data were not imputed and analyses were based on available data; therefore, the findings at 9 and 12 months should be interpreted cautiously because of the potential for selection bias. Third, MMEF_25/75, MEF₂₅, and MEF₅₀ are spirometry-derived surrogate measures of small airway function and are subject to intra-individual variability and volume dependency. Changes in FVC during follow-up may therefore have partly influenced the observed changes in these parameters, and more direct assessments of peripheral airway function, such as impulse oscillometry or multiple-breath washout, were not available in this retrospective dataset. Fourth, the logistic regression analyses were exploratory and based on a relatively limited sample size. Candidate predictors were evaluated in separate models adjusted for sex, age, and BMI to reduce model overfitting and multicollinearity; therefore, the independent predictive value of these factors relative to one another cannot be established. In addition, although Bonferroni correction was applied to MMRM pairwise comparisons, the number of secondary and exploratory comparisons may still increase the risk of type I error. Finally, seasonal variation and environmental allergen exposure, including pollen, were not captured in the retrospective data and may have influenced longitudinal changes in asthma-related outcomes. Prospective multicenter controlled studies with larger pediatric populations and longer follow-up are needed to confirm these findings.

Conclusion

Dupilumab treatment was substantially associated with the improvement of SAD in children with moderate-to-severe asthma, accompanied by reduction in type 2 inflammation biomarkers. A decrease in FeNO may serve as a potential indicator associated with improvement in small airway function, although this finding requires further validation.

Funding Statement

This study was supported by State Key Laboratory of Respiratory Disease (Special Project for Clinical and Epidemiological Research, Project No. SKLRD-L-202603).

Use of Artificial Intelligence

ChatGPT (GPT-5.6 Sol; OpenAI) was used during manuscript preparation solely for language editing and refinement, including improving grammar, clarity, and readability of the authors’ original text. It was not used to generate study data, perform statistical analyses, or independently formulate scientific conclusions. The authors reviewed and verified all AI-assisted revisions and confirm the originality and accuracy of all manuscript content. The authors take full responsibility for the integrity of the entire manuscript, including the accuracy and appropriateness of all references. The authors also confirm that the terms of use of ChatGPT were reviewed and that its use was considered suitable for preparation of content intended for publication.

Data Sharing Statement

The datasets analyzed for the current study are available from the corresponding authors (Jinping Zheng and Nanshan Zhong) on reasonable request.

Ethics Approval and Consent to Participate

The Institutional Ethical Review Board of the First Affiliated Hospital of Guangzhou Medical University approved the study (No. ES-2024-001-01).

Consent for Publication

Consent for publication was not applicable.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

This paper has been uploaded to Research Square as a preprint: https://www.researchsquare.com/article/rs-8288806/v1. Dr Jinping Zheng reports honoraria from Sanofi; meeting/travel support from Sanofi; data safety/advisory board participation from Sanofi, outside the submitted work. The authors report no other conflicts of interest.